I think he's talking about (1) extracting the plutonium produced as a by-product of U-235 fission, and (2) extracting and re-concentrating the remaining U-235. I don't think he's talking about using any of the actual high-level wastes, e.g. the Cs-137 or Sr-90 et cetera that are usually considered the problem children in spent nuclear fuel, isotopes produced in abundance that have half-lives ~10-100 years (so they're very radioactive but don't decay very fast). So, yes, those latter isotopes are just always going to be a pain, so you just have to put them in a (dry) hole somewhere and let them decay over a few centuries.
Those two isotopes in particular are also annoying because they have the chemical properties of K and Na (in the case of Cs) and Ca (in the case of Sr), and so they can slot into the places where those elements are used in the body and just hang out, irradiating you internally for years. Cs also forms very soluble salts, so it disperses easily in any kind of watery environment.
Many of the fission products you isolate during reprocessing are useful for other things, such as medical isotopes, radioisotope thermal generators (including Sr-90), etc. I'm not sure what they do with the cesium. But France after reprocessing is left with a pretty negligible amount of spent fuel waste and they just keep it all in a building somewhere.
Yes, I think that's true to an extent, and I suspect Cs-137 is used in radiotherapy applications for just this reason (it's conveniently obtained). But my vague impression is that the requirements of medical-grade isotopes are such that you need to use special reactors constructed for (or amenable to modification to) that purpose, I don't get the impression that reprocessed fuel from power reactors is usually considered a great option. Indeed, my impression these days is that the bigger nuclear medicine departments are moving towards installing their own synchotron to make their isotopes to order, and on the spot.
But maybe they do it in France, and maybe I'm generally wrong, this is not my area. I am little suspicious that any of these ancillary applications can absorb the few hundred kg of high-level waste you'd collect from a refueling. I'm still thinking you'd end up with at least a tonne or two of stuff you just needed to stick in a hole in the ground somewhere. (Not that I think this is a big problem, it's a big planet, there's plenty of places to put such a hole.)
People tend to worry about waste that has a half-life of the order of 10,000 years, but IIUC almost all of that long-lived stuff is plutonium, which is nuclear fuel that we can and should separate out and burn in reactors (which will split the atoms into smaller radioactive atoms).
The smaller radioactive "fission products" (which can't be burned) mostly have half-lives below 100 years, and after 500 years their radioactivity level has fallen below the level of uranium ore (not even uranium metal).
The goal of being able to store waste without any human maintenance for 20,000 years always seemed to be of dubious value — I mean, how about we focus on preventing an apocalypse, rather than focusing on guaranteeing that post-apocalyptic settlers who somehow decide to settle in the Nevada desert 8,000 years from now won't suffer from a 0.1% increased risk of cancer because they don't know what the ☢ symbol means?
But since society seems devoted to storing the waste For As Long As It Takes, it's worth noting that it's much easier to design a structure to last 500 years than 20,000. But the whole issue is kinda moot because plutonium emits alpha radiation, which is easily shielded (unlike the gamma radiation from many shorter-lived isotopes). So, you can safely sleep atop a drumful of plutonium, just don't ingest it, or grind it into dust and breathe it, or grind it into dust and dump it in your drinking well.
As for isotopes with very long half-lives (over 100,000 years or so), the crucial thing to understand is that radioactivity is something that happens with radioactive decay; therefore radioactivity is inversely proportional to half-life. As the half-life gets longer, the danger level decreases: all else being equal, if something has a half-life of one year, it is 10,000 times more hazardous than something with a half-life of 10,000 years. Therefore, things with extremely long half-lives aren't scary; it's the intermediate half-lives (10 to 10,000 years) that are problematic, as they are both long-lasting and radioactive enough to potentially cause cancer.
P.S. A one-ton drum of plutonium contains as much usable energy as 3,000,000 tons of coal, so I'd just like to add that nuclear waste is awesome compared to other kinds of waste. Because it's tiny.
Coal ash might be stored in open ponds or giant buildings that occasionally split open and dump their entire contents into rivers: https://appvoices.org/coalash/disasters/ — but because nuclear waste is so small, it's practical to encase it in concrete so that it never harms humans, wildlife or even plants. Why didn't those activists worried that containment isn't 100% perfect spend the last 60 years fighting coal instead of nuclear?
A geologist friend of mine worked on Yucca Mountain and he called it a sham. They were being asked to guarantee there would be no major earthquakes or volcanoes for 10,000 years. According to him that was impossible to state with high confidence.
But even the reason your geologist friend was asked to ensure that for 10,000 years was a sham. It's totally not necessary. Even high level waste becomes less radioactive than a granite countertop within a few hundred years.
And there are plenty of rocks that haven't been touched in millions of years! Pick any one rock that hasn't moved for 10 million years, drill a hole into it, and drop the waste in. Honestly the whole storage "problem" is mostly regulatory theater.
I've never understood that requirement. Like, why are we *planning* for a dystopian future in which humanity has degenerated so much it can't exercise competent stewardship of a waste respository?
We also have 5,000 nuclear weapons lying around . Do we need to weld instructions for their care and responsible storage to them, written in pictograms at a 4th grade level, just in case the future turns out to be stupid?
Yeah honestly even the fact that we're planning to ensure that this waste repository is safe for the post-apocalyptic humans who might stumble upon it is...simply not a standard we hold literally anything else to.
What the heck is an "accidental near detonation?" Is that like "not a detonation at all" after someone dropped one by accident? If an airman bumps into an W80 in a warehouse and clonks it with his wrench, and nothing happens, is that an "accidental near detonation?" I'm deeply unimpressed.
Anyway, the fact that a maximum of 20,000 weapons has been supervised by a passel of 19- to 25-year-olds with at most a HS diploma for 75 years without a single notable slipup is proof enough to me that it can be done. At this point I think the burden of proof is on the people who think future American government will screw up something as obvious as digging an irrigation channel right next to the big yellow warning sign that says YOU WILL DIE IF YOU DIG HERE .
Even with the school board's actions, it appears that most likely there were no negative health effects to the people who lived near Love Canal. Source: Wikipedia.
I remember Love Canal. At the time I thought the reaction was hysterical, and I still think that. Perhaps I need to learn something, but my experience was that in the 70s public attitudes towards random assorted chemical in the environment shifted radically -- in the 1960s after we got done cleaning a carburetor with a bowl of gasoline, we just threw it onto the ground. Heck, people changed the oil on cars and threw it on the ground. Turpentine after painting? Down the sink. Same with business. If a gas station leaked 1000 gallons of gasoline into the soil, big deal, nobody much cared.
All of that changed in the 70s. All of a sudden we got really upset about those things, started suspecting them of causing cancer and allergies and whatnot. (And some of that may have been true, after all, cancer has to come form somewhere.) Love Canal was just a symptom of the changing times.
So that would be a better argument if we were really careless about nuclear waste *now* and the future might get a lot more uptight and be pissed at us. But we're talking the other way around -- we're uptight *now* and, strangely, feeling like we need to assume the future will be slobs. But that has never happened before, and it seems strange to assume.
It’s massive recency bias to think the next 75 years will be just as stable as the last 75.
Look, I support nuclear power in general, but I also know humans are very bad at assessing risk, especially over very long timelines.
I’m old enough to remember the Berlin Wall falling and a tremendous amount of entrepreneurial dynamism and overall optimism in the 1990s. America was the dominant hyper power and most of the world *liked* it that way.
But then a few dozen followers of Osama Bin Laden changed everything.
It’s foolish to assume something like that won’t happen again.
What do you mean by "changed everything"? I don't see that anything changed at all except for the greater clampdown on the tiny amount of terrorism. Or are you talking about the war in Iraq & Afghanistan, which I'm pretty sure was small compared to wars of previous generations?
In any case, the ~3,000 people that died on 9/11 is to me a very small number when you're talking about the entire U.S. over a long time span. I mean, there have been about 200 times as many deaths due to COVID.
Well, no, it's *logical* to assume the next 75 years will be rather like the previous 75. Unless you have evidence to the contrary, or at least a good working hypothesis. I don't mind a *certain* amount of caution, e.g. it's definitely prudent to put your nuclear waste out in the desert, far from anywhere even remotely habitable, and mark it plainly, big warning signs and such. But assuming that within the next century we're going to be in some Thunderdome situation where nobody reads English any more and the wild kids dig up the shiny metal bits to make into necklaces -- this goes well beyond prudence.
Look, the whole point of the future is that they're almost certainly going to know more than us (technologically speaking), the same way we have known more than the people in the past, for the past 1000 years or so. So why try to do their job? Do ours. Leave them a growing economy, healthy happy new generation of people, some new inventions -- and yes, a few inherited problems with which they're expected to cope. After all, our parents handed us nuclear weapons and the Cold War, and we've done OK with that.
People seem to manage extensive hours of training, constant practice, and a focus on situational awareness and response while driving. Are you saying flying is intrinsically much harder? Or that you wouldn't expect flying to become a primary mode of transportation, so it would be harder to maintain the practice?
Flying is intrinsically much harder than driving. It's tricky to get an apples-to-apples comparison because the units of measure are different (fatalities per passenger mile vs fatalities per hour flight time), but general aviation is at least 10x more deadly than driving. And that's in our current regime, where people who are doing the flying are self-selected to be more focused/situationally aware/skilled than your average random car driver. There's obvious reasons for this - mechanical issues in a car generally mean you pull over and wait for a tow truck. Mechanical issues in a Cessna mean you do everything exactly right or you die.
I have a PL, IFR, and High Performance rating. Flying an airplane is intrinsically much harder because of the procedures involved, but I might seem like a heretic to other aviators for saying that I don't believe these are things that couldn't be overcome by the public at large by more and better automation and inherent safety protocols built into the systems as available to the public at large.
I absolutely could see a future where the general public had access to and used flying as their primary *individual* form of transportation, but in that form a huge amount of the system routine and implementation would have to be prohibited to the owner/user.
99% of aircraft incidents are still (again) pilot error, and that is with operators that are trained and certified above the level of the general public. For the general public to have it as their primary form of transport, even more freedom would need to be limited.
Think "Minority Reports" self-driving rail-pods, but airborne.
+1 Yes, flying today's airplanes is too hard, but a market for flying cars would put pressure on the control system design to make it simpler. Using "cockpits are too confusing for normal people" as an excuse is 1st order thinking.
Carl Pham below points out that "we shouldn't do it because it's too dangerous" means that we never begin the process of learning how to make it better.
Air taxis would make a ton of sense, though, right? People who fly all day can afford extra training; that we never got air taxis supports the thesis of excessive regulations.
...though it occurs to me that many people would want to hire a taxi to fly downtown, where airstrips might be scarce, even assuming you can build them directly on top of the normal roadways. So another explanation is chicken-and-egg: you need a lot of taxi traffic to justify the airstrips, but lots of airstrips including downtown ones to justify the taxis. Another popular destination should be airports, but air traffic control wouldn't want small aircraft encroaching on its airspace. I doubt this is insurmountable tho.
A third explanation is risk to third parties. Air taxis may be safer than cars (certainly per mile traveled), but a fatal crash will occasionally kill a random person on the ground. This is mostly true of car crashes too! However, while a car crash may kill a random person on a road or sidewalk, an air-taxi could crash into a house, which probably makes them feel psychologically more dangerous even it they're not.
Maybe to get past choked highways and reduce infrastructure costs. But the thought of then weaving between skyscrapers gives me goosebumps. I'd keep them out of downtown or any residential neighborhood. That would make them more like an airBus anyway, i think.
I am a pilot, and flying is much harder than driving. Maneuvering in three dimensions is part of it, but bigger parts are cross-coupling of controls (where every input does at least two different things), a higher level of critical multitasking, and the part where you can't pull over to the side of the road or even slow down if things get temporarily overwhelming or the hardware glitches. Also, all-weather flying (necessary for routine travel) becomes a complex exercise in real-time systems management and requires training *out* some deeply instinctive behavior. In my particular case, that makes me safer in the sky than on the ground because flying effectively commands my attention but driving leaves me too easily distracted, but I'm not exactly typical.
The average person can learn to fly a light plane safely if you insist on it. And we sort of do, for people who want to fly. But the difficulty, expense, and tedium of the necessary training and the non-trivial fraction of the population that will wash out and be aeronautically disenfranchised, will likely make this politically infeasible at flying-car levels of utilization. The remaining hope for "flying cars" is increased automation, but safe autopilots are a classic 80/20 problem where, for this application, we'd really need that last 20%.
I'm reading the book because of this review, and one point Hall makes is that the autogyro from the 1920's has a reverse risk profile from an airplane. It's unstable on takeoff, but will glide to a stop on landing, even if engines fail. So the "can't slow down and can't pull over" might be less of a problem with another technical model.
Nice!! I want one now. One thing Hall doesn't answer well is why US regulations blocked progress across the world. The other review raises that issue as well - financial and health regulation has slowed things down, but not stopped us in our tracks. Are flying cars just on the wrong side of the cost disease? Or is there some other issue (like, say, physics)?
"People seem to manage extensive hours of training, constant practice, and a focus on situational awareness and response while driving."
And yet *look at how dangerous driving is.* Even if flying drivers were no more prone to accidents than terrestrial drivers, there would still be too many accidents, and they'd probably be more deadly.
Self-flying cars are an easier problem now, but that would quickly devolve if even 5% of the US population had them and used them as their primary form of transport. Granted, working an integrated Terrain/Collision & Avoidance System (TCAS) for objects moving in 3 dimensions instead of 2 is slightly easier, but if you see my comment above it would be a catastrophe if the car/system gave the choice to the pilot/operator on how to avoid a collision as opposed to the linked computers making that choice.
If you want to see a horror story of just how confusing human/computer/system interfaces can be vis-a-vis collision avoidance, watch this simulation of the Flight 2937 collision of Switzerland.
As a matter of handling traffic, I imagine flying cars would be restricted to long-distance travel, between or just outside cities, and restricted to tired travel within them. Also freight trucks would probably remain wheeled, given the weight, and given that shippers already have the option of freight airplanes, and aren't using them.
That's pretty much what Heinlein was figuring 70 years ago for domestic transport: the wealthy and important government agents would have city cars that can quickly covert into aircraft in the countryside, then land in the hinterlands of the destination and roll into town.
Why does the *average* person need to be able to do it well for *anyone* to be allowed to do it at all? This seems indicative of the weird current hyperfocus on equity over opportunity which I mentioned elsewhere. None of us can have flying cars until all of us can? Why?
Is flying a flying car so hard that we can't, say, take the 10% best drivers in the nation (about 20 million people) and let *them* fly around? What would that do to boost their productivity? Maybe there are a ton of really smart people from whom we are not getting X new inventions a year because they are freaking stuck in traffic.
I think because on one hand, the gains are kind of speculative: "spend less time in traffic -> more inventions" seems like a stretch, especially considering how many other ways there are to solve that problem: remote presence being the obvious 2020 solution, public transit or a private driver to make 'commuting' time more productive, etc.
And the potential downsides are obvious and specific - the potential for loss of life and destruction is pretty clear (and not just for the people behind the wheel!) in case of accidents, to say nothing about the potential deliberate acts of destruction.
Right. And I get this. As I said elsewhere, we have just become across-the-board more timid about things. "But is it safe? What could go wrong?" tends to trump "But what marvels might we unlock if it goes right?" pretty much all the time.
There are argument both ways, of course, and either "safe at any speed" or "damn the torpedoes" extremes are unwise. But whatever choice we make comes with costs. If we are much more concerned about what will go wrong, then we will be generally less entrepreneurial, take fewer risks, and our social progress will increasingly resemble NASA's progress on human space flight -- very, very slow, but also very very safe.
"we have just become across-the-board more timid about things."
Can't this be seen as a reflection of the increased opportunity cost associated with various forms of risk taking? As measures of expected life satisfaction/life expectancy substantially increased over the 20th century, the costs of risk taking in the form of foregoing expected life increased.
I think it's a good argument in general, but I'm not sure the data support it in this case. Life expectancy in the US has only gone from 70 to 79 since 1960. That's nothing to sneeze at, but it's hard to see it having that big an effect on risk-taking. I don't think life satisfaction has increased at all, and indeed the rising rates of middle-aged suicide, and middle-aged drug use and overdose, would kind of suggest the opposite. Heck, even the teenagers are having less sex, apparently. Four hours of World o' Warcraft doesn't compare to getting it on with the fox from Algebra, so I can't see how *they're* happier.
'Life satisfaction' was a poor choice of words on my part. Maybe it would be more accurate to say video games/the internet has largely solved the "problem" of boredom. What equivalent way of wasting time by yourself did teenagers in the 50s and 60s have? Comics? Playing records? Seems like there's no real equivalent to the video game slob stereotype we have now. I wonder if a lot of early-life entrepreneurship (broadly speaking) is just a way to avoid boredom.
We already do let a few highly trained people fly around. Furthermore, we allow them to take passengers, so the rest of us can have the benefit of their skill. They're called airline pilots.
I think a lot of the problem is that there isn't a clear Schelling point for regulation in between "The average person should be able to get a license without it being unduly burdensome" and the current very cautious regime where it takes hundreds of hours of expensive training to get a pilot's license.
Regulations for cars can't get that strict because the average person expects to be able to drive and the population as a whole wouldn't tolerate that changing, but once you accept that only a fraction of unusually skilled people can do it safely the standards might tend to err on the side of caution because of the incentives on regulators.
Well it's a good thing the automobile wasn't invented last year then, isn't it? "You're going to let 16-year-olds sit at the controls of a two-ton metal machine capable of accelerating to 100 MPH? Insane!"
If ten percent of the population is flying freely above the crowded freeways at 100+ mph on their daily commute, the remaining ninety percent are going to really notice this. And be jealous. And vote.
Right, well, you could say the same thing about absolutely any manifestation of wealth. It's not like people don't already notice private planes, yachts, gated communities, private schools, expensive condos with beautiful views in ski resorts.
The way we traditionally deal with this is social mobility. People have the *opportunity* to get rich themselves, through hard work and talent, and then they join the ranks of the 10%. People are smart enough to understand that if everyone has to have the same stuff, it's going to be a low level of stuff, and so they're generally willing to instead enter the lottery of who ends up in the 10% -- as long as they believe it isn't pure chance that produces the winning ticket.
They also usually tend to believe that what the 10% have usually moves down after a while anyway, as long as economic growth continues -- at one point you had to be decently well off to afford a car at all, then everyone had one car but being a two-car family was pretty tony, and now these days it's buying a Tesla for the 3rd car that marks you out as in the 1%.
I agree such attitudes are less common today, though, and there are many more like what you're saying: "Only 10% of us will ever fly above the Blade Runner dystopian squalid streets below, and it's going to be some connected/powerful/aristocratic 10% that will never include me or my kids, so screw that." But the existence of that curious change, and wild speculation on its origin, is kind of the point of the reviewed book.
I think a lot of the military research successes have come because they have goals of making a practical technological application as their first objective. Academic scientists are frequently driven by raw curiosity and/or obsession with the function of some very specific part of the natural world. There's nothing morally wrong with that, but it's not optimal for finding practical applications. They stretch the words in their grant proposals to try to convince funding sources that their work will have some practical benefit--and every so often one of the many projects does--but fundamentally their interests are usually anchored to the thing they're studying rather than a solution to a practical problem.
In contrast the military's objectives are to increase real world combat effectiveness. That's a practical goal, and therefore they're much more willing to sideline or abandon research avenues that are less likely to result in new solutions to practical problems and to do so sooner.
> Academic scientists are frequently driven by raw curiosity and/or obsession with the function of some very specific part of the natural world
Worse than this, I think a lot of academic scientists aren't even all that interested in the very specific thing they study, they're just kinda stuck with it.
Here's the way it works: your PhD supervisor was the world expert on the subject of tantalum oxide. When you came along he racked his brain for five minutes and assigned you a project looking at the ever-so-slightly-different-and-more-obscure subject of tantalum sulphide. You work hard for five years and sure enough, you're one of the world's top experts on the subject of tantalum sulphide. You don't really care about tantalum sulphide. It turns out that tantalum sulphide is completely boring and unimportant. But you _can_ think of a dozen possible tantalum sulphide related projects that you could potentially get some grant money to work on, and you don't know enough about anything else. So you start churning out grant proposals on the subject of tantalum sulphide, starting with "Tantalum sulphide has potential applications in X, Y and Z".
I feel like this is kinda the case for at least half the academic scientists I've met; they're stuck in some kind of dead end doing research that they're not especially interested in but they suspect they can get funding for.
Yeah, the 'just stuck with it cause it's what your advisor had to bequeath upon you' is why I left the academic research field. Naively I originally went in with some very specific research goals and did so specifically because i felt that basically no one was working on them. I eventually found that the fact no one is working on them already, means it's nearly impossible to start doing so within the framework of academia and grant funding. You need to either be rich enough to self-fund, or you need to be fortunate enough to discover an exciting breakthrough in your target direction while still a student--and even then it probably needs to be monetizable to work out. Even if you do one of those it's incredibly hard to start down the new direction in academia because there may only be a handful of people in the world who'd be a good fit as your advisor and there's no guarantee you'll be accepted to (or want to attend) their particular institution.
> As for government science not being worth a damn, that may be true now, as I'm told the funding process has become completely corrupted. However, there are notable successes like GPS and the internet
I'm old enough to remember the exact same criticisms of government science being made decades ago, while those things were being researched. I don't think there's been any major change, just that as always the kind of research that governments fund is unsexy and far from the point at which it would be implemented, so it will only seem valuable in retrospect
My pat answer to the flying car question has always been "We do have flying cars, they're called helicopters".
Now of course there's a good reason why most of us don't own helicopters; they burn a lot of fuel, they need a lot of expensive maintenance, they require a lot of specialised training in order to fly them, they're relatively dangerous, and they're so loud that you're not allowed to land them in most places. But all of those problems are intrinsic to hovering transport, so they preclude your "flying car".
An electric automatically-piloted quad-copter might be able to solve most of these problems to some extent, though.
hard disagree. couple hours in a sport plane and in a robinson helicopter plus regular motorcycle rider. of the three, plane is the easiest and motorcycle the most difficult/dangerous. yes, crashing the plane would be bad, but after takeoff and setting it on course and trimming it out, i could practically take my hands off the controls (extremely low-inertia 600 pound aircraft with no autopilot or stabilization aids). it’s much less stressful as you’re making fewer inputs/decisions per minute. as far as take off/landing, the amount of computing power in the average tesla is probably more than enough to automate those anyway esp if vtol as the book proposed.
On the other hand the first time a flying car breaks down and crashes on a kindergarten there will be no end of additional restrictions made about who can fly what over built up areas.
I’ve never heard anyone before say government-funded science was bad for science!" Ayn Rand has as character in Atlas Shrugged who was a brilliant physicist before he got government funding. Then he became useless. Nassim Taleb (I forget which of his books) also argues that many of our significant inventions came from outside academia.
Funny that the author uses a machine learning analogy, as ML is definitely an invention of academics.
I don't think it would be hard to imagine that there is an *appropriate* level of government funding for science and technology, and that at levels that are too high it sucks all the oxygen out of the room -- and then all your science becomes done in the way you do government science, which is not the only way to do it, and in some cases not the best way to do it.
It is certainly the case that 50 years ago there was a far more vigorous realm of basic R&D outside of government and academia. You think of Bell Labs, or IBM Yorktown, Xerox PARC, Exxon Annandale -- these place attracted absolutely first-rank talent, and in their day invented amazing technologies. But corporate R&D has been eviscerated, and it isn't *completely* out of the question that part of that is the entry of the 800lb gorilla -- government funded academic research. From the point of view of the new researcher, there is much to like (initially) about the government/academic model: you don't have nearly as strong a deadline/results pressure, and you can often work on more abstract problems. And for some areas of research that is an excellent shift. But...there are areas of research where a certain amount of bottom-line and practical focus *is* good for results, and can even be quite beneficial to the individual, in the era when it was possible to become very handsomely rewarded by commerce.
I think there's a very active world of corporate research in things programming, and for this we can thank many of the brilliant innovations in computational stuff. And to some extent we still see that in biology, although perhaps not as much as we'd hoped 30 years ago -- biotech is still very tightly tied to academia, and the bio giants (e.g. Big Pharma) don't seem to be *expanding* their basic R&D programs. If Utopia could be created just by brilliant programming, we'd be in good shape, but alas it needs progress in things made of actual stuff also.
I'm pretty sure a lot of private research money is gone because companies realized that groundbreaking research usually doesn't actually pay itself off to the company funding the research.
Leo Szilard wrote a short story in *Voice of the Dolphins* where a character predicted pathologies of government-funded science. IIRC it was written in the late 40s or early 50s, satirical in the form of proposing NSF-style funding as a way of sabotaging an enemy society.
(Szilard was a physicist who proposed the fission chain reaction in the 30s and clashed with Manhattan Project management.)
I've heard it before. A number of years ago I met a very smart person whose name escapes me (associate of current MIRI researcher Eliezer Yudkowsky), at a SENS research conference on aging who brought it up during a Q&A. Most of the attendees get their paychecks via grant money so his question was laughed off, but at the time I recognized his name and went to talk to him about it afterwards. We had an interesting discussion about historical funding of research, but most of it was about pre-industrial revolution research where discoveries in the sciences were mostly made by rich or patronized-by-rich people and no one had even considered government funding go through a bureaucracy to be distributed to researchers.
It's not that nothing is ever discovered when research is funded via government grants, but it shouldn't be a surprise to anyone most of the funding goes to things which enhance the prestige of the researchers of yesteryear who are now in charge, rather than to people who intend to make progress in completely new directions. I'm not sure I'd attribute ML primarily to academics either--the math of it is a very old idea. Rather computers have recently reached a point in all their capacity types where the technique is useful. Academics working in ML are primarily "just" fiddling with the variables of layer counts and compounding systems to get new results. It's important work, but not fundamentally a new paradigm.
I don't have a particularly strong opinion for/against public research funding, but I think dismissing the idea without a lot more discussion is exactly the kind of failure that "Where's My Flying Car?" argues we already suffered from. Similarly labeling the idea "libertarian", and especially associating it with Ayn Rand, makes it more political, and therefore more controversial, than it needs to be and wrongly biases a lot of people against the idea. We'd need much better evidence that the bureaucratic funding model benefits ML research more than a private funding model would before I'd find its use in the review ironic.
"I'm not sure I'd attribute ML primarily to academics either--the math of it is a very old idea. Rather computers have recently reached a point in all their capacity types where the technique is useful."
Those facts are correct, but the old math was originally invented by academics as well. So it is certainly attributable primarily to academics.
As a side note on this idea of government funding in historic periods...The idea of trying to separate 'the wealthy' such as the random Lord, Duke, Earl, and sundry types of lesser nobleman from the idea of government is invalid in my view when talking about that era.
The rich and powerful people funding or lending their power to various researchers, artists, and merchants in the enlightenment era were either directly members of the government through their aristocratic positions or were themselves strongly beholden to support from such nobles.
The idea of trying to say this funding wasn't run through large government bureaucracies in the 1700s doesn't make sense as such structures by and large did not exist at the time with the major government actions being tax collection, the military, and running a fairly simple court system to arbitrate disputes and process criminal charges. I don't think there was an NIH or NAS type body at the time and the main equivalent would have been the nascent University sector and the not-so-private actions of nobles with governmental authority to individually fund whomever caught their interest.
It sounds like a coloured view of history seen through a modern lens which ignores how people lived at that time. I'm obviously reading into that position from only a few words, but it sounds fairly anachronistic to me.
You're understating the case for ML. The fact is that there have been major advancements in ML. GANs, Transformers, reinforcement learning, just really large networks, RNNs, etc... are all serious innovations in ML. And most of them have come out of private labs (in fact most of them have come from FB and Google...).
I was about to say the same thing. Arguments against government funded/controlled science go back to the 1950s.
And the thing is - they kinda have a point. Instead of coming up with new and innovative methods, we end up taking years to publish papers that are mostly forgotten and ignored.
It also freezes in the horrible bachelor-master-phd requirement. Let's get real: I can train anyone to be a good experimental biologist in about a 100 hours or less. 12 years' effective apprenticeship? Get. Real.
I've yet to get to the point in my PhD at which they sit me down and tell me how to think independently, I personally just think you just have to learn from experience.
> And the thing is - they kinda have a point. Instead of coming up with new and innovative methods, we end up taking years to publish papers that are mostly forgotten and ignored.
Would we suddenly get a lot better at coming up with new and innovative methods if government funding were removed, though?
> Let's get real: I can train anyone to be a good experimental biologist in about a 100 hours or less
You can train someone to do biology experiments, yes; the part that requires an expert is understanding which experiments are worth doing.
(Or, less pithily, the job of the expert is planning out a full research program that will advance a particular sub-sub-field of biology, assembling a team of 100-hour doofuses aka PhD students to actually do it, and then properly communicating these results to the other experts in the field.)
This depends a lot on how government funding works. In the US and UK, yes. In most remaining Europe, no. It is part of the job, but by far not the main part.
In most universities in Germany or France, you can get by without raising any external money at all. You will have less PhD students, and you won't be a superstar, but even superstars do not spend the main chunk of their time allocating money.
This does have downsides. Whether it is overall a good or a bad thing, this is really complicated.
"In most universities in Germany or France, you can get by without raising any external money at all. You will have less PhD students, and you won't be a superstar, but even superstars do not spend the main chunk of their time allocating money."
In France, the current annual funding of a researcher is about 2000 euros. Its is almost impossible to do anything without grants, at least from experimental people.
Well, in principle you can do it in the US once you have tenure, but...you'll never be promoted, and you'll get assigned to all kinds of painful committees, get shit teaching assignments, that kind of thing. The university really loves its overhead :(
In fact, that's the corruption I'd most like to see rooted out of the system. When research overhead can make up twice as much of the university budget as tuition, the incentives are pretty screwy.
If it weren't for the 12 years of apprenticeship AND that said process effectively locks you into researching some specific side project of your advisor's specialty, biology research would be my current occupation. The current paradigm is actively turning away people who want to go in directions that the fewest people were interested in historically, and that's inherently going to bias results against real breakthroughs.
"A survey and analysis performed by the OECD in 2005 found, to their surprise, that while private R&D had a positive 0.26 correlation with economic growth, government funded R&D had a negative 0.37 correlation!”
It seems widely unlikely to me that this strong negative correlation is causal. Has the - 0.37 correlation by any chance been calculated by mixing developped countries (High public RD, low ec grotwh, and developping countries (low public RD, high economic growth) by any chance?
I find causality believable. When I left my job as a bioinformatician, I had in mind to do a bioinformatics start-up. I eventually gave that idea up, because nearly all of the bioinformatics software used in the US is developed using government grants, and made available for free. It's difficult to find anything in the bioinformatics space where you don't have to compete with people who are getting paid by government grants and giving away their software. But that free software isn't easy to use, and usually isn't supported or maintained.
It seems to me that your observation suggests that goverment money makes producing well maitained product more difficult, but it really doesn't seem obvious to me that well maintained software is a key to innovation.
Firstly, if you use economic growth correlations and/or practical applications as measures for how well science is being done, something is wrong - especially in what concerns fundamental research.
What should be analysed is how public funding criteria have often created an environment where risky projects are disincentivised, especially as more and more people got PhDs and exhausted what little slack existed for trying wacky stuff without fear of being outcompeted by conventional incremental stuff.
Anyway, since private companies will do R&D in search of profit regardless, it seems logical that public funding should go primarily to that research with no obvious profit routes. The question is how that funding can be more efficiently allocated in today's highly competitive academic world.
If you haven't ever wondered whether government funding of research was bad for science, you haven't done research with government funds.
I have so many horrible stories that I don't know where to begin. But let's start with Congress. Congress understands that we need basic research, but we also need that research to lead to profitable businesses, and to solve urgent technological problems. So, in a lot of basic research funding, they've decided to kill 3 birds with one stone by requiring that basic research must also be applied research, and must be the basis for a profitable start-up business. In the SBIR program, which I'm most-familiar with, your grant proposal must explain how it is basic research, how it is applied research, and what clients you have interested in the product that will come out of this "basic research". If you get to Phase 2, which you must get to at least 1/4 of the time in order for SBIR grants to be profitable, your Phase 2 grant proposal must have a client lined up and committed to provide part of your funding. To get to Phase 2, you must focus on your business plan, and spend about 90% of your budget on producing a cool-looking Phase 1 demo.
The result is that few SBIR grants do much basic research. In some years, I've looked at most of the unclassified "research" grants offered by the US government, and concluded that the only actual basic research being done by any government agency was by DARPA. DARPA is the very best agency we have for doing basic research.
When I submitted, won, and ran a DARPA contract, the COTR (Contracting Officer's Technical Representative) in charge of my $100,000 project was also managing a project which, if I recall, had a budget of $100 million, so that it was literally not worth his time to read any of the reports I read, or to answer any of my emails or phone calls. He had some underlings inform me of this. My team was one of 5 or 6 other teams awarded a Phase 1 contract. Near the end of the contract, after I'd worked 2 months of 12-hour days and weekends preparing the final report and demo, I got an email telling me not to bother completing it, because the COTR had already decided who he wanted to award the Phase 2 grant to. It was given to the only team which appeared, from its slide presentations, not to have produced anything other than slide presentations.
That was one of the most-successful government research projects I ever worked on--the software I developed did eventually make the company a lot of money--for the simple reason that, although no one was interested in the results, at least no one on the project wanted it to fail.
Contrast this with the NASA/FAA grants I worked on. Back around 1970, Congress ordered the FAA to use NASA engineers for airspace research projects, in order to avoid suddenly firing all those engineers after the moon landing. So air transportation research projects were managed by NASA, and carried out by government contractors (so those NASA engineers got fired anyway). NASA was monitored by the FAA, which saw NASA as a bunch of pointy-headed nerds with no practical experience who should stop telling them what to do. Plus the FAA didn't really want to automate air traffic control, because while it would save lives and a lot of fuel, it would put FAA employees out of work.
Or contrast it with the government-funded bioinformatics work I did for a genome research institute, where I was supposed to automate the work of the genome annotators, who were supposed to help me with the program and approve it when it was ready, after which they would be fired. That worked about as well as you'd expect it to.
The most-successful project I ever ran was a NASA project. As often happens, the original COTR who wrote up the project solicitation had been rotated out before the project entered Phase 2, and no one else in NASA or the FAA had the slightest interest in the project. Even the new COTR, whose job is to ensure that I carry out the work approved in the contract rather than repurposing it to some other objective, encouraged me to repurpose it to some other objective that someone actually cared about. So I did; and that project made the company a lot of money, and saved NASA $40 million, though it will probably never be used for its intended purpose (to automate air traffic control).
But we haven't even begun to talk about the main reasons government research grants waste money. One is that government funding centralizes funding, so for every agency there's somebody in a room in Washington DC who's responsible for $10 billion of grant funding every year. They'd much rather manage ten $1 billion projects than ten-thousand $1 million projects, even though the 10,000 $1 million projects would be much, much, MUCH more efficient.
(I once read an NIH blog post describing a study of the relative efficiency of NIH research grants. They compared the output of grants of between $1 million and $20 million by counting the number of research papers produced per project. They concluded, IIRC, that projects costing more than $10 million were almost twice as productive as projects costing less than $5 million. But they forgot to divide project output by project cost. The blog post summarizing the report was written by the director of an Institute, managing hundreds of millions of dollars worth of grants yearly, who DID NOT KNOW YOU NEED TO DIVIDE OUTPUT BY COST to compute productivity, because Institutes aren't incentivized to check on overall monetary efficiency.)
Most of the money in government contracting goes to pay for reputation. Big agencies prefer to award contracts of $50 million and up, because otherwise they have too many projects to manage. So some COTR has to award a $100 million contract. She's gonna award it to some big company, like Raytheon or IBM, because most big contracts fail, and she won't get fired if IBM fails, but she will get fired if Name-You-Don't-Recognize fails. BigCorp will try to line up a bunch of subcontractors to do the work. The competing subcontractors each line up famous experts who claim they'll support the work.
So with a big contract, you end up with a hierarchy, with BigCorp at the top, subcontracting corporations below them, and PIs at the bottom who manage the productive work. BigCorp's job is to vouch for the reliability of the subcontractors, for which they take about a 50% cut. Each subcontractor's job is mostly to vouch for the reliability of the PI, for which they take a more than 50% cut. The famous experts might get a 10-20% cut of the remaining money, usually do approximately nothing, and are being paid for the use of their names, like the famous people on a company's Board of Directors.
None of this is irrational, given the premise that projects must be big. Big projects fail so often that taking a 50% cut at each reputation level in order to put some credible reputation on the line is worth it.
Do the math, and you'll see that little of the money on big projects is left to do work. Whereas with small projects, there's more money left to do work, but most of it is put into making a cool demo (think the MIT Media Lab), and (rough guess) 90% of small government projects are killed or thrown away without anybody ever using them, either because they threatened someone's job, or because nobody really wanted them in the first place.
This was excellent. And a bit of a shame it's buried in the substack comments. I know at some point in future I'll fruitlessly look for "That ACX comment on science funding stuff".
This needs one or another of Scott's epistemic status notes, but anyway, here's a story for you: One of the "A.I. Winter" events happened after a period in which the Department of Defense had decided that Artificial Intelligence was a priority, but then professors at America's most venerable universities had perfected the art of getting DoD grant money by laying out mumbo-jumbo that promised the moon and delivered nothing. Somebody at one of the big private research labs (probably Bell Labs or IBM) eventually collected some convincing enough evidence that this shitshow was what was going on. DoD responded by just turning off the money spigot. Then for a decade or so everything in AI research got very quiet - though some smart people may have been laying the groundwork for the more productive research directions that eventually emerged later.
As one of those 1980s AI researchers, I think all of the AI researchers I knew were sincere, though a few had crackpot theories. Mostly, symbolic AI never worked out as well as people had hoped. Expert systems were hard to deploy because they were by definition designed to replace experts, and most fields with "experts" have regulations about who is and is not an expert. That turns out not to include computer programs.
The DoD had its own problem with AI. I won't claim this was endemic, but it did happen more than once: The military wanted automated "red teams" for training exercises, and would periodically award grants to people who had state-of-the-art (symbolic) AI systems to control the red team. Then the product would be deployed, and the red team had to be programmed by some Spec 3 technician who /might/ know how to program computers, and that went poorly. Eventually the military would give up and look for something simpler. When they got the simpler thing, it was too simple to be a good red team, and someone would say "We should use more AI!", and the cycle would begin again.
The utter lunacy of this book quite aside, flying cars are around the corner and we have development of greentech to thank: electric cars beget better batteries that can do the job.
Agreed, as Tom explained (much better) below me "Energy density. We could have built flying cars off gasoline a long time ago". We seem to have read Matej's question differently. I took their question to mean "why do we need better batteries before they can be used for flying cars", not "why are batteries are better than gasoline".
Energy density. We could have built flying cars off gasoline a long time ago, but they would have been extremely pollutive (helicopters get around two miles per gallon). Now, what exactly is the distinction between flying car, helicopter, and airplane is mostly just, like, the branding, but basically nobody is going to release something called a "flying car" and not have it be emission-less today, so that means batteries.
Energy density matters because a flying car starts to run into a rocket equation kind of problem, where the more range you want, the more batteries you need to carry, which means a greater mass, which decreases your range. This means that more batteries doesn't really solve your problem. So to get a longer range flying car, you need higher energy density batteries.
Oh absolutely, the theoretical limits could be insanely better than we have now. But this is one of those "it's actually just really hard" problems.
For lithium ion, we're near the limit. Tesla's 4680 batteries are probably going to be within 30% of the theoretical limit for lithium ion chemistries. I don't know that number for sure but it's what I recall talking with battery experts I work with.
Solid state chemistries can theoretically do a lot better but there's a reason they're taking a very long time. The history of the battery industry is kind of like fusion power, always 10 years away or something. You should ignore all news that you hear about a "battery tech breakthrough" until it's in a consumer product that you could theoretically buy.
Side point - if we can work from home and in virtual reality and do not need to commute anywhere (for even 20-30% of the population)...do we need flying cars anymore? Might the technological innovation which allows us to work virtually reduce traffic and the need to live near our workplaces or have any kind of commute...might that innovation outpace the battery technology needed for flight? Flying with an electric car requires clearing quite a high bar of innovation and production which could still be 15+ years away from a cheap consumer model. While the battery technology for ground transport is essentially already or very soon to be a consumer grade product and getting better and cheaper over time.
It is sort of like the strange move for 'smart thermostats' which are a short term stopgap until we can build better houses such as passivehaus designs which passively self regulate with a small up front investment to reduce running costs of a building dramatically over its lifespan.
A flying car is cool and fun, but I don't see it being more than a toy like an ATV for Jetski for a very long time in the 30+ year range until a person at home might 'call for a flying car automated taxi' to take them somewhere unless that house is a large mansion - in which case they can just afford to hire a human helicopter pilot right now.
I would say that flying cars are finally becoming available today not just because of dramatically improved battery technology, but also because of dramatically improved control software (and the hardware to run it, of course).
I think they are both quite difficult. Sure, it's possible to design the optimal control system on paper; but we have only recently gained the power to actually implement one that is small enough to fit into a flying car, and fast/smart enough to be usable by the average person (as opposed to a trained helicopter pilot). Doing so required massive advances in computer hardware as well as machine learning.
Not really. The reason that the computer science problem appears to be being solved isn't that it was easy to do, but that computers are getting so fast and so parallel that we don't have to shrink away from brute forcing it so much. Really this goes for a lot of things. Most changes in software stacks today aren't to enable new capability but are instead to make the work of software developers slightly less tedious, and to allow less-smart people to successfully accomplish software development tasks, at the cost of program performance (since we have increasing performance to spare).
It's so aggravating. Like a level 1 engineering analysis would say "oh, renewables are intermittent? Well a large capacitor (battery) can solve that problem."
A level 2 analysis is like "well, do we have enough batteries?" The automotive industry is making incredible strides here. The problem is that all of those batteries are going to be used by the automotive industry. Vehicle-2-grid is not actually a good idea, people tend to want their car to be charged when they go to use it. Battery production MIGHT just scale extremely well now that demand is of no consequence, but it's just weird to see so many people (who are acting out of a very respectable abundance of caution over climate change) betting the future on the assumption that battery production will scale like semiconductors did. Do I hope that's going to happen? Absolutely, in fact my career is trying to make that happen.
What you or I "back" is irrelevant. Fission nuclear is too expensive and too slow and a huge expansion simply isn't going to happen. Internet debates are one thing, but the brute economics and politics are another. Nobody wants to deal with the many tricky problems associated with radioactive waste and you're not going to force them to. It's time to move on.
I'm strangely optimistic about fusion, which I think is a reasonable possibility within the decade. But fission is EOL.
D-T fusion still makes radioactive waste, although you have better control over what it is (there's a rather long list of elements you can't use in a wasteless fusion reactor because of (n,y) or (n,p) creating something with an annoying half-life - "annoying", here, being "too long to just put it in a pond for a couple of years, short enough to make significant radiation").
"Fission nuclear is too expensive and too slow and a huge expansion simply isn't going to happen."
Did you read the review at all? It had giant sections about just why fission is expensive and it had nothing to do with the underlying technology.
Also I was interested in fusion (especially after hearing about MIT's SPARC) but then I read a bit more about energy densities and... they're dismal. Back to fusion.
It has to do with regulation forcing companies to pay the costs that they would otherwise externalize, which makes the projects uneconomic. The risk of the underlying technology is what makes it so expensive. The only thing the book has to add is radiation denialism, which is not super useful.
Batteries are well and good, but how are these things going to get around?
My limited experience with VTOL drones of appropriate size to carry humans makes me highly skeptical that we have technology to automate key system functions (primarily take-off and landing as well as collision avoidance) sufficiently to where people would be able to drive them with an equivalent level of training that we give drivers.
Landing on a ship (which is always moving) is much more difficult than landing on the ground (which doesn't move). This is like saying that we don't allow cars to dock with trucks in motion so we shouldn't allow drivers to be licensed at all.
hmm. the book’s framing would say that contemporary battery/energy storage tech is finally on the level of 40s-era petrol tech, thus limited flying cars with likely similar range/speed/cost as a pitcairn autogyro.
We shouldn't have flying cars until people are willing to require that only computers may fly them. There isn't enough airspace above a city for humans to avoid crashing into each other.
"The public is wrongly terrified of nuclear energy, but they shouldn’t be. Radiation killed 0 people at Fukishima"
You really lost me with the Fukishima minimization. As if deaths at the time of the incident are the only relevant concern. How much land exactly is contaminated forever?
Well, the worse places in the Fukushima area are giving out 90 mS a year right now. That's about twice what we allow a radiation worker to be exposed to every year and given that these limts are conservative it's not clear that there's anything wrong with living there now. That's slightly smaller than the smallest yearly dose linked to cancer but given that we let people smoke, have wood stoves in cities, etc we should probably let people live in Fukushima right now. I wouldn't move there until levels are down to the 10mS/year where Ramsar and we know the people there aren't getting more cancer than usual. But that's most of the Fukushima exclusion zone.
That is no small thing. And who is to say the next nuclear disaster doesn't create a far worse contamination problem. For a looong long time. To not even mention that aspect of the issue in a blithe dismissal of concerns about nuclear energy is just crazy to me. This is a real pattern in the arguments of nuclear apologists I've noticed, and doesn't exactly inspire confidence.
I don't mean to come across as blithe. I wouldn't endorse someone smoking even 1 cigarette each day, eating bacon for breakfast every day, or other risks in roughly the same range. I wouldn't live in Fukushima right now myself. But it's still not a huge risk and if other people have more tolerance for risk than I do I think that they should be allowed to smoke or eat bacon or live in Fukushima. And it isn't forever. Different isotopes decay at different rates. One with a half-life of 1 year is roughly 10 times more radioactive than one with a half-life of 10 years so the most radioactive isotopes tend to decay fastest, though the least radioactive ones will be with us for a long time. Still, I think that in a few decades when the worst parts of Fukushima are merely as radioactive as Denver it will be unreasonable to worry about the radiation levels and at that point I will actually move my attitude to blithe dismissal.
Yes, I meant the author was being blithe (I assume that is not you). Your clarification was quite helpful, thanks. My use of "forever" was of course not literal, but along the lines of "ruined for human use for decades/centuries"
Yes, that's certainly a serious cost to the accident. But it's also the case that many other kinds of power generation have even higher environmental costs, on average. For my part I'd say that once we finish getting rid of every fossil fuel power plant we should start getting rid of fission but I'd hold off on it until then and I wish we'd built more in the past before solar became cheap.
I'm certainly willing to entertain a cost/benefit comparison of nuclear with other options. That is why I find it frustrating when I am presented with what seems to be a deliberately incomplete or even misleading one such as a we find in this review.
The fact that you are calling people apologists is not a sign of good faith intent to weigh the risks here. All electricity generation comes with risks. Solar panel manufacturing involves toxic chemicals which have a half-life of infinity years (non-radioactive chemicals don't decay) The silicon sand refinement process releases tiny silica particulate in the air which leads to silicosis, responsible for tens of thousands of deaths a year. More people will die this year falling off a rooftop installing solar panels than will die building nuclear power plants. Are you an outspoken advocate against solar panel manufacturing?
You shouldn't be. Energy, on net, saves lives and improves global welfare. Why would anybody just be a nuclear apologist? Have you considered that so-called nuclear apologists genuinely believe that it's the safest form of energy generation?
Pretending that day-of-meltdown causalities are the-only relevant metric of the nuclear power risk is ridiculous to the point that anyone who employs such a ridiculous metric can fairly be called an apologist in good faith.
I think a lot hinges on you saying the phrase "only relevant metric." Those are words that you have attributed to the author, and yet the author did not write, yet you are using to cast aspersions over his intentions and all else who may just genuinely think this is a good policy decision. I urge you to have greater patience and a more open mind.
That is the only metric he or she uses, to the exclusion of all the other relevant metrics I mentioned. It was their decision to make that strange choice, not mine just because I pointed it out.
To be uncharitably pedantic, exactly zero land is contaminated "forever" - the radioactive contaminants will, eventually, decay to undetectable levels.
But if you're asking the more reasonable question of "how much land is contaminated to the point of uselessness on a timescale longer than a couple decades"...well, it honestly doesn't seem like there's very much at all outside of the plant itself. The exclusion zone, such as it is, has pretty much shrunk to a couple towns in the immediate vicinity of the plant, and that's *with* the Japanese government's conservative-bordering-on-paranoid safety regulations.
It's also important to note the discharge of contaminated wastewater into the ocean adjacent to Fukushima, which can concentrate in sea animals, including food animals.
The question is Trutium which is constantly created by cosmic rays in the upper atmosphere, has a half-life of 12 years, and is NOT concentrated by sea animals.
Umm, no: A study published in the journal Science in August 2020 found traces of several other radioactive isotopes in the Fukushima wastewater, many of which take much longer to decay than tritium.
Some of that radioactive material may have already made its way into local wildlife; In February, Japanese media reported that shipments of rockfish were halted after a sample caught near Fukushima was found to contain unsafe levels of radioactive cesium.
Assuming you're talking about this study (https://science.sciencemag.org/content/sci/369/6504/621.full.pdf), the amount of other radioisotopes is utterly negligible. There's 500,000 Bq/liter of tritium radioactivity and around 10 Bq/liter for the other isotopes. For comparison, your body contains 8000 Bq of radioactivity, or about 100 Bq/liter. That is, if you took the tritium out of the wastewater and drank the rest, you'd probably lower the average radioactivity of your body.
As an aside, the 500,000 Bq/liter of tritium becomes 100 Bq/liter if you dilute it by a factor of 5000. Dumping it into the ocean is a good way to do that.
I don't want to dismiss the concern of waste leakage into the ocean...but it's not established that the waste leakage into the ocean is going to cause significant ecological damage. I know that may sound like a crazy claim to make, but radiation is extremely not intuitive.
Most scientists expected Pripyat (outside Chernobyl) to be some kind of toxic, uninhabitable wasteland for a hundred years. Instead, wildlife there is thriving beyond anybody's expectation. Vegetation growth even exceeds what we would expect if there was no meltdown. It's quite possible that Chernobyl-like radiation levels are harmful to large mammals like humans but good for smaller life forms like bacteria and plants. Nature is antifragile.
Now that is absolutely not to say that we should just go dumping nuclear waste into the ocean because it might be good for it, but it's not obviously a catastrophe.
I don't think so. Even with a standard "linear no threshold" model of radiation damage, the chance that wildlife could notice the difference is ~zero; I expect a flourishing of plants and wildlife purely as a result of the missing humans.
Specifically, it looks like a typical radiation exposure in Pripyat is 0.7 µSv/hr or 6 mSv/year as of 2009 (without any attempt at environmental cleanup AFAIK). For comparison, "In Europe, average natural background exposure by country ranges from under 2mSv annually in the United Kingdom to more than 7mSv annually in Finland." http://www.chernobylgallery.com/chernobyl-disaster/radiation-levels/
I'm sure human scientists have methods sensitive enough to notice a tiny increase in cancer rates due to exposures substantially below 100 mSv — but wildlife is more concerned with how to find its next meal. (Btw, this figure of 7 mSv is the the highest I've ever heard and, if true, ought to make a good place to do a study on radiation risks.)
Regarding your link: beware the man of one study, but especially beware the study by one man! I don't know how best to read the radiation-risk debate, but it's clear there *is* a debate, i.e. evidence for the effects of low-dose radiation seems inconclusive (apart from universal agreement that the effect, whatever it is, is small), and making this especially hard is the near-impossibility of finding healthy human subjects exposed to intermediate levels of radiation (10 to 100 mSv per year). An illustrative paragraph from a study shows this:
"Most of the previous ecological studies investigating associations between childhood leukaemia and naturally occurring sources of ionising radiation have found positive associations for radon (13-15) while for gamma radiation and cosmic rays results have been inconsistent (16-22). Early case-control studies of the association between natural sources of radiation and childhood leukaemia were underpowered and have reported mixed results (23-26). The largest of these, the UK Childhood Cancer Study, included over 2000 cases of childhood cancer and reported weak evidence of a negative association between childhood leukaemia and measured radon concentrations (25) but no evidence of an association with measured gamma dose rates (26). However, the proportion of eligible subjects participating in the measurements was low and varied by socio-economic status. Because exposure to these sources is ubiquitous and variation in cumulative doses received by children of similar age are small, large sample sizes are needed to detect the small predicted risk. Given the rarity of childhood cancer, the only way to achieve such sample sizes is by combining data over long periods of systematic cancer registration." https://boris.unibe.ch/135621/15/Mazzei_JRadiolProt_2019_AAM.pdf
In the same paragraph we see childhood leukaemia associated positively and (in the largest study) negatively with radiation from radon. Also mentioned is the difficulty of measuring the effect (it's especially difficult because cancer is normally caused by something other than radiation, so the effect we're looking for could easily be swamped by other things that affect cancer rates.)
In general this is the problem with the "precautionary principle": if something is banned for being potentially harmful, it becomes extremely difficult to clearly show that it IS harmful.
My understanding on the effect of the relatively low dose of radiation outside Chernobyl on plants and animal is that they are thriving despite clear negative effects of radiation (increase of the ferquency of abnormality, decreased growth ring in trees, etc), because the absence of human pressure more than compensate the effect of radiation. But given our current knowledge, it seems to me extremely unlikely that radiation is good for any known living organism.
"But given our current knowledge, it seems to me extremely unlikely that radiation is good for any known living organism." -- A study here suggests that some amount of radiation isn't just good, it's required: https://www.pbs.org/wgbh/nova/article/life-without-radiation/
At some point, you have to say what form of radiation you're talking about. You know that sunlight is radiation, right? I assume you agree that sunlight is good for life.
Radio waves are radiation. If you're using wireless internet, you're using radiation. Sound is radiation.
As with anything else in science, numbers matter, and what amount of radiation you're talking about is crucial. But it's clear that the idea that all radiation in any amount is bad for life is just ignorant.
"At some point, you have to say what form of radiation you're talking about. "
I thought it was obvious that when I said "the relatively low dose of radiation outside Chernobyl" I meant α and β radiation due to radioactive contaminanation, not sunlight!
This certainly does not demonstrate that (hard!) radiation are necessary. The only well estblished result is that if you hit bacteria with a high dose of radiation they survive better if they had previous exposure to a low dose. That is a far cry from some (hard!) radiation is necessary for life. α particles really do a lot of damage to complex molecules, it seems extremely unlikely that this damage can become necessary.
" But it's clear that the idea that all radiation in any amount is bad for life is just ignorant." Obviously true if you include lower energy radiation. But I maintain that "given our current knowledge, it seems to me extremely unlikely that (hard!) radiation is good for any known living organism".
I didn't know if you meant just alpha and beta, or also gamma radiation.
I think the Nova article says (or implies) that the scientists restricted cosmic rays and maybe natural radiation from heavy isotopes. Cosmic rays are protons and other nuclei (like an alpha particle, I guess); I don't know what the other consists of, but probably neutrons, and alpha and beta particles.
I was wrong to say this radiation is necessary for life, but the experiment suggests that hard radiation is beneficial to life. That contradicts your statement that radiation is not good for any known living organism. Am I missing something here?
Aside: I believe that radiation in some form, probably electromagnetic, is required for all life, because we need it to get mutations in DNA, and we need mutations in DNA to get evolution. It could be that DNA polymerase could've evolved to be more error-prone to make up for less radiation.
Whoa, so the ecological impact of the worst civilian nuclear disaster in history is exceeded by the prior ecological impact of humans simply living there? That definitely lowers my estimate of the ecological consequences of civilian nuclear power.
I am not sure. The impact of humans "simply living there" on plants and animals is huge! Imagine an alien civilization landing on earth and colonizing it for their own purpose, whithout noticing or caring about us, the impact on us would probably be extremely high.
The findings are gradually decreasing (it's still only 45 years ago) and was probably just bad luck that it is only 1,400 km away (800 miles might be close for US citizens but for Europeans there are 4 countries in between)
Well, none of it is contaminated "forever," but the current restrictions on residence apparently apply to about 371 km^2. You may want to bear in mind that industrial pollution routeinly affects much larger swathes of living area, and routinely cases far more deaths. For example, China's Huai River policy encouraging the use of coal north of the river has been argued to cut 5y of life expectancy off the 500 million people who live there:
Industry pollutes, and if we choose to live technological lives we always run health risks (particularly when industrial plant is combined with things like earthquakes and tsunamis). It's not possible to run zero risks without living in caves and hunting antelopes with spears. So perhaps the focus should be on *relative* risks, and trade-offs. Being blinding Polyanna enthusiastic about nuclear power is dumb. But so is being "terrified."
I thought it was clear that I wasn't using "forever" literally. As I said elsewhere, I'm certainly willing to entertain a cost/benefit comparison of nuclear with other options. That is why I find it frustrating when I am presented with what seems to be a deliberately incomplete or even misleading one such as a we find in this review.
My critique was not oversimplified. I just used a term loosely in such a way that the meaning was still obvious. Neither did the author oversimplify. They lied by omission. So maybe don't misrepresent the whole exchange?
I don't agree that your oversimplification was obviously not literal, while the author choosing deaths as the comparison criterion is a lie. If you're not willing to be charitable to the author, you shouldn't expect others to be charitable to you.
Really? You thought that by "forever" I meant that when the sun turns into a black hole Fukushima will still be irradiated? Come on. That could not be more different than choosing a deeply misleading metric.
I had the same reaction to the Chernobyl minimization. If only 43 people had died, that'd be one thing, but glossing over the health impacts of the disaster on hundreds, if not thousands, of other people seems dishonest to me. I'm pro-nuclear power, and while I think when it fails it fails due to human error and not scientific error, the costs of failure really are very, very high.
You really lost me with the "forever" bit. Yes, yes, you didn't mean it literally. But either this is irrational blind fear, or it's a fundamentally quantitative problem. If it's irrational blind fear, then there's nothing for it but to try and point the stupid people somewhere else. If it's a fundamentally quantitative problem, then you don't address it by saying "how much?" and then introducing a spurious infinity term for rhetorical effect.
About 300 square kilometers near Fukushima have been turned into a de facto nature preserve for the next few decades. Turns out nuclear-power-plant levels of radioactive contamination are a pretty good deal for most animals, because it poses relatively low risk over their natural lifespan but is enough to drive off their chief natural predator and habitat-paver-over.
An analysis I did on SSC or DSL a year or so ago, suggests that if the human race generated 100% of its electric power using nuclear power plants built to 20th-century standards (but no new atom bomb factories), the rotating nuclear nature preserve would at any point be equivalent to IIRC Macedonia or Haiti in size, with individual zones rotating in and out every fifty years or so. Or, if you prefer human habitats to the wildlife sort, you can build your power plants to more modern designs and cut that down.
Compare and contrast to the amount of land we'd have to turn into e.g. solar farms, where the sun never shines and nothing grows. No, it's not enough to just put solar panels on the roofs of existing buildings, and we're not even going to talk about solar roads.
I'm open to the possibility that radiation from meltdown concerns can be addressed. But the author didn't even acknowledge them, let alone make a case as you have here
If somebody proposes to do something useful, like providing abundant cheap clean energy, then the burden of making the case ought to fall on the person saying "no you're not allowed to do that because it's too dangerous". You stepped up to the plate and completely failed to make that case, offering only a rhetorical question including an objectively false assumption.
And that's pretty much par for the course in this business. Lots of people believe that they know that everybody knows that nuclear power is "unacceptably dangerous" and that simply alluding to that "fact" is a slam-dunk win. Almost nobody actually makes the case.
I think the real reason for the turn against nuclear was that the public was used to thinking of fallout in terms of being downwind of thermonuclear groundbursts, where it meant dying puking your guts out in hours or days rather than a theoretical increase in your risks of cancer. Order of magnitude comparisons are made hard by the fact that radiation is invisible. I wrote about the matter [on my blog](http://hopefullyintersting.blogspot.com/2019/06/sometimes-you-need-new-word.html) at more length. I already see a change of attitude between the generation who grew up in the shadow of the mushroom cloud die off and as those of us who grew up with reactor meltdows as our image of fallout so I'm optimistic about that aspect of the future.
At the core of The Green Religion is something I call the environmentalist's habitat paradox: if you really like the environment, a natural first-order desire would be to live in a cottage deeply secluded in nature, far from civilization. But this is either unscalable (and therefore antisocial) as you cannot allow too many others to indulge in the same lifestyle, or you DO proselytize this lifestyle and it becomes environmentally catastrophic. The paradox is that if you love the environment (as in truly want to protect it), you must live in a city.
Eco-pragmatism needs better branding. We need extremely lush, literally-covered-in-plants cities powered by cheap nuclear.
Part of the branding problem is lack of good definitions that leads many to group serious thought on the cost/benefit of various levels of protection for the long standing state of the environment, in with the dumbest environmental protesters--chained to trees while using their iPhone to tweet about how a plant has the same moral value as a human. Worse any attempt at serious problem solving or compromise in public debate very quickly devolves into extremes shouting at each other because government's control via regulation makes things winner-take-all at the entrepreneurial investor level and a status signal of tribal politics and virtue at the level of regular voters.
Well I think it's more the same problem which exists with everything else (and why I'm a rationalist) which is that people choose the right answer with their feelings. If you're someone that really likes the aesthetics of environmentalism, well the last place you want to live is a concrete jungle.
Enter Derek Jensen, an anti-civilization advocate who lives in the remote wilderness with bears. He thinks civilization, cities, and industrialization were horrible ideas, and everyone should give them up and go live "more natural lifestyles" where they hang out with bears in the woods. Nevermind that there are probably less than 300,000 bears alive in the world and so, at best, each bear would have to befriend tens of thousands of people. Not only would this certainly not be good for the bears (not sure what the Dunbar number is for bears but I think it's safe to assume it's less than ten thousand), but Jensen himself would almost certainly not want to share his bear friends with ten thousand other people. I know countless environmentalists/activists who would say that Jensen's lifestyle is idyllic.
I'm a person that just naturally has to think through what the natural consequences of things are. Apparently most people don't do that, they just think "I'd like to live in the woods." and that's the end of any kind of consequential analysis. And so to me, these ideologies, when people say "we should all abandon civilization and live in the woods" somewhere in there, either nature is just absolutely destroyed beyond the likes of which we have ever seen (which doesn't sound like their goal), or colossal numbers of humans vanish somehow. And so I am extremely skeptical of these ideologies.
Looking at some very rough numbers, there is enough forest in the world for everyone to live on a little more than one acres of forest. Maybe we could live in small communities of 50-100 people on 50-100 acres, and maybe some people would be happier living on the savannah or other open grasslands, freeing up space for the rest of us to have 1.5-2 acres or whatever.
Some Googling did not get me an answer to how much forested land a person needs to live in a sustainable way, but I know it's a lot more than an acre. That's especially true for the many many people who would have to live in Siberia and other inhospitable forests, where heating fuel for the winters would be a big issue.
On the bright side, the billions of deaths in that first year would certainly help with the "colossal numbers of humans vanish somehow" problem!
I'm quite sympathetic to Jensen's goal, but I think there may be much more sustainable options available. For instance, I choose to live in an area where there are so many forests around that I can literally see at least one from any vantage point within 50 miles. There's several hundred thousand people in that range, mostly clustered in a series of small towns and surrounded by farmland.
I don't think it's intellectually responsible to be sympathetic to Jensen's goals. Climate change and sustainability have always been human problems. We have no direct business case for reducing emissions, but we know we have to, can we make ourselves do it? We can hardly get people to stop feeding wild deer, raccoons, or ducks. Somehow we're supposed to expand the average person's sphere of altruistic concern to contain the ecosystems of all of the "forest[s] of the world" and give them a scientifically rigorous understanding of how to achieve that?
This is a fantasy, perhaps worth thinking about, debating, and taking seriously back in the 1960s when it emerged in force, but given humanity's track record since then we have no reason to believe such a psychological/sociological/educational stunt can be performed, and it's frankly dangerous to continue entertaining it. Green hippies are in their bubble, waiting for everyone to come join the drum circle, while mainland America is still rolling coal and ICEing EV charging stations.
To be sympathetic is not to actually support. I'm "sympathetic" in the sense that most humans (and incidentally also environmentalists) have been misled by their ignoring the numbers in favor of policy ideas based on feelings, and who can blame them for just doing the usual human thing?
Now, when I point out the numbers to someone and they're like "you red tribe bastard!" and I'm like "I'm not red tribe" and they're like "whatever I'm outta here", that's when my sympathy dries up.
Yes. We don't have to live in cities of a million persons though. We can live in 100 dense pedestrian pockets connected by silent inconspicuous hyperloop.
I don't think that it's obvious the ideal size is 10,000 people and afterwards you see diminishing returns to scale for agglomeration. Even if it were, a 10 billion human planet would require one million such small cities. A quick google search tells me that there are currently 10,000 "cities" worldwide. We should absolutely not want to find 100x as many locations around the world for more cities. That would necessarily mean fewer nature reservations.
It seems to me that we've moved on to trickier problems to solve that are mostly based on coordination rather than simply maximizing consumption. For instance, imagine a 40-story office building with about 2,000 workers where each one commuted by flying car. How many landing strips do you need? Remember that unlike parking spots, you can't stack them -- each one needs to be open to the sky -- and you probably need *minimum* five minutes' clearance between cars. If everyone arrived between 8am and 9am, that means that each landing strip can serve a dozen employees, so you'd need 166 total just for this one building. From a perspective of land use and of time spent getting from your parking spot to your destination, this just sounds terrible. So we should be happy we don't have flying cars, because the societal equilibrium they'd put us in would be terrible.
One could argue that this particular problem is specific to transportation technologies, but social media has amply demonstrated that it's possible for many technologies to lead to bad equilibrium outcomes.
All this is not to say that techological stagnation isn't a problem, but when people talk *specifically* about flying cars, I discount their arguments specifically for the above reasons.
I'm not really sympathetic to this kind of argument because you could probably make similar sounding arguments about a lot of technologies we take for granted today.
Imagine if you had to propose making, say, the national power grid in todays climate if it didn't already exist. You are going to make a giant country spanning grid of copper wires that make a circuit into every single household? The wires themselves are ugly and you want to have them on every street? And these wires are dangerous if anyone fiddles with them or damges them. And if they are burried then and someone were to hit some burried ones with a spade? Wouldn't that hurt them? And then you have the wiring in the walls of your wooden houses? Wouldn't that be a fire risk? And the sockets themselves have a potentially deadly voltage just sitting there on the wall where any kid could stick a fork and kill themselves.
How could you ever possibly scale this?
If something is useful on an individual level, then it will be slowly adopted and then we will find a way to scale it later.
I don't think the argument I'm making is about "how do you scale this" -- spatially inefficient transportation technologies have negative returns to scale. If you have the first car? Sure there aren't any gas stations, but it's still awesome (especially compared to the first telephone) because you can get anywhere 5x faster than anyone else. But if everyone has a car, it clearly makes your car go slower.
So in fact, I would make exactly this argument about technology (the car) that we do have today. How come scientists haven't solved traffic? It's because road space is a special kind of resource where the richer you get, the scarcer it gets.
If you are saying that your argument about Flying Cars applies equally to regular cars, then how do you explain the fact that Cars are widely in use today?
Doesn't that imply that Flying Cars would still be in use widely even with the challenges that you forsee?
You mentioned a bad equalibrium before and compared it to social media, but that implies that regular cars are in a bad equalibrium too and that we would be better off without them? I don't think I can agree with that.
I do think that regular cars in the way currently used in the US are a bad equilibrium! If you look at Europe or Asia, the number of car trips per capita is about half of what's in the US, and it's not because they're behind in car technology. Similarly, I don't think that making cars better technologically solves the problems we have with them. Teslas can accelerate much faster than internal-combustion cars, but do they actually get you where you're going any faster?
Counterargument: The most appealing argument for flying cars was reduced reliance on public infrastructure, highways. Also flying around buildings is dangerous. Restrict flighted travel to outside cities. Continue using urban parking garages designed for tired vehicles.
Honestly, this is what we should have said with ground-based cars. Driving around people is dangers. Restrict >20mph travel to outside cities. Continue using urban transport using small and/or dense vehicles like bikes, feet, and streetcars.
Stand back a few paces from this discussion, and ask yourself honestly: Do you really believe there will ever be ten million flying cars buzzing around the US?
For me is just ludicrous to imagine that.
For comparison, can you imagine millions of persons moving around the US at 300 mph in silent hyperloops?
> so you'd need 166 total just for this one building. From a perspective of land use and of time spent getting from your parking spot to your destination, this just sounds terrible
I mean, if you had a 2000-person building where everyone commutes by non-flying car then you'd need a pretty big parking structure too. In order to just maintain the same 166-space footprint that you already think is too big, you'd need a twelve-storey parking structure, which is very large.
From a land use point of view, flying cars (idealised flying cars anyway) would allow us to reclaim all the space currently devoted to streets and roads.
Also if we had (idealised) flying cars we probably wouldn't bother having 40-storey office buildings anyway, our cities would be less dense because you could travel greater distances with ease.
One nit to pick: general (private) aviation was not done to death by regulation as much it was by product liability torts. Lawyers somehow got extremely good at convincing juries that the aviation accident equivalent of "16 year old who just got their driver's license buys a Ferrari, drives it at 120mph on a twisting mountain road at night in a rainstorm, and predictably winds up dead after careening off a cliff" was somehow Ferrari's fault, and awarding the idiot's family millions of dollars in damages. Given that Ferraris are already a low volume market, it doesn't take too many such lawsuits to drive the cost to buy a new one through the stratosphere.
(The actual scenario would be that a rich retired athlete or businessman would buy an expensive, complex high performance airplane, do the minimum amount of training required, then fly off into bad weather in unfamiliar areas - which they should have known not to do if they had been paying attention in flight school - and run into a mountain, or building, or just plain crash. And their widow would then sue the airplane manufacturer, and usually win.)
In 1994, Congress passed https://en.wikipedia.org/wiki/General_Aviation_Revitalization_Act, which was supposed to fix this. Lawyers just switched targets from the manufacturers to the mechanics who work on planes, with the predictable result that airframe & powerplant mechanics refuse to sign off on an airplane's annual inspection unless everything is perfect, increasing cost of ownership for private airplanes.
All that being said, as a private pilot, the idea of having to share the skies with several orders of magnitude more aircraft, being flown by the equivalent of your average automobile driver who can't be bothered to use their turn signal or put down their phone while driving, is terrifying.
I've read quite a few reports of amateur pilots crashing their planes here in Ireland. These are guys who have their own little plane, are experienced, often are going out for a routine short flight maybe taking friend/family along as passengers, and then something goes wrong and there's a fatal crash. So it's not even at the level of "dumb 16 year old".
We've come to accept the level of road accidents and we still try to bring them down. Cars in the sky will have a much, much tougher barrier to acceptance.
To make matters worse, I don't know how it is in Ireland, but in the US the FAA certification process for a commercially sold aircraft is very expensive, even for a small two-seater. It's often more than the plane would be otherwise. So a lot of small aircraft manufacturers don't bother selling assembled aircraft, they instead sell the parts as kits and hobbyists build them themselves.
My wife's uncle was a professional commercial airline pilot who in his retirement flew a Piper Cub. Despite all his experience, he crashed the thing not once but *twice* and nearly died both times.
Also all the celebrities who have some flying experience (certainly more formal training than the average motorist has) who crash and die.
If there's one unnecessary risk I've learned, it's to never set foot in a private aircraft.
> All that being said, as a private pilot, the idea of having to share the skies with several orders of magnitude more aircraft, being flown by the equivalent of your average automobile driver who can't be bothered to use their turn signal or put down their phone while driving, is terrifying.
Right. I've witnessed two minor accidents in the last two weeks, both completely avoidable and the result of obviously reckless driving by the person who caused the accident. I'm very skeptical of an environment where the same morons are flying, meaning every one of those stupid reckless accidents becomes very likely fatal and, if it occurs over populated land, potentially a mass casualty incident.
Then there's maintenance... I heard a boom several years ago while working from home - it was a small plane crashing into an occupied house in my neighborhood due to a mechanical failure. The occupants happened to be on the other side of the house, or it would have been a 4-fatality incident due a maintenance oopsie.
I'm sympathetic to the idea that regulation has hampered technological progress, but the negatives of this are (and I understand, yes, can only be) argued from the perspective of our current world.
What I mean is, flying cars have all sorts of logistical challenges associated with them related to infrastructure and training and higher possibilities of disaster if an accident were to occur in a densely populated place, where most of the cars would be. I imagine if the government had been hands off and let them develop, they still wouldn't be common. More of a novelty if anything. It just doesn't seem feasible in an urban area, and probably not desirable from a livability standpoint.
Nuclear, I agree, was a massive mistake to oppose on the basis of climate change, and probably would have gone differently if it were invented today. But we might not feel that way if we had built hundreds or thousands of additional reactors, and one horrible disaster had taken place that killed hundreds of thousands or millions of people. I understand that isn't likely, but we know it to be possible. Current energy sources just don't have that doomsday potential, and I think humans are willing to accept lots of incremental damage but have a strong aversion to potential apocalyptic accidents.
I sympathize with the thesis. We could be living in a very different world. But it's easy to speculate about the positives and give your alternate history a rosier outcome that it might have achieved in reality.
No, we don't. The idea of a nuclear plant melt down killing thousand of persons is purely speculative.
You could perhaps argue that there is a one in a million chance that operating 1000 nuclear power plants for 100 years would result in an accident that killed 1000 persons. So, yeah, "it's possible". But this kind of logic proves nothing.
I stopped fearing nuclear power plant accidents when I learned that the Chernobyl Exclusion Zone is now a thriving nature preserve.
If you know how a nuclear plant works you know it is certainly possible. Highly unlikely, as I stated, but from a purely physics-based standpoint it certainly could happen.
I know how a US nuclear reactor works: a controlled fission reaction produce heat which is transferred to pressurized normal water which boils water to make steam which drives a turbine which produces electricity.
Now explain how an accident would kill 1000 persons.
For US reactors, consequence analyses show that even in the case of a meltdown and large release of radiation to the environment (like Fukushima), the average number of expected prompt fatalities is zero. The high end of the uncertainty range is tens of prompt fatalities. At the high end, latent cancer fatalities can run into the low thousands, but this is based on the Linear No Threshold model (which few experts find plausible) and most of those fatalities are from people moving back into the area after an evacuation ends.
Honestly, Chernobyl was about as bad as a power plant accident can physically get. It was a big reactor, it had no containment structure, and the whole thing pretty much vaporized and blew itself up all over the place. Pretty much whatever radioactive stuff was in the fuel got released into the environment. For modern reactor accidents, we look at the percentage of the reactor's inventory of each radionuclide that gets released. Anything more than 1 percent of the iodine or cesium is considered pretty high, and 100% would be unthinkable.
I take that back, Chernobyl could have been worse if several units melted down instead of just one. Like 3-5 times worse, max. The response also could have been worse (though I wouldn't say it was good). Still, no way we'd be talking about hundreds of thousands of deaths.
On the other hand, Chernobyl should *not* be considered as a typical fission reactor accident.
On the third hand, the fact that the author of the book seemingly dismisses/ignores/minimizes what happened there (also the times that we got *really* close to starting a global thermonuclear war ?! And remember Putin's doomsday device?) makes me want to dismiss the rest of the book as as sloppy...
How would a nuclear plant disaster kill "hundreds of thousands or millions of people"? Even Chernobyl, based on a power plant design that would be have been illegal in the free world (even as far back as the 1960s), killed at most ~9000 people (if you accept the linear no-threshold model).
Even scraping the bottom of the barrel for the most unsafe plausible design, how do you arrange for it to kill over 100,000?
But even accepting this argument, the Banquio dam disaster killed roughly 171,000 people and this doesn't usually scare people away from hydro-electric power. I think it's more of an argument against shoddy communist engineering (ditto for the USSR's reactors in Chernobyl).
This book, and Peter Thiels’s famous quote, seem quite odd to me — I am glad we don’t have average humans flying cars, because they would be a massive hazard.
Driving is one of the most dangerous things that people habitually do. Flying a plane is way harder, and should be left to the pros and dedicated amateur hobbyists. Especially given the rates of alcohol, cannabis, opiate, benzo, etc. usage, it would be reckless to encourage casual flight by human amateurs.
Contrary to the idealism of this book and the libertarian movement, it’s government investment in computing, software, the internet, and AI that enables the development of automated flight systems, manufacturing, and maintenance — airplanes need to be in tip-top-shape.
We still don’t have level 5 autonomy for cars. Once that nut is cracked, then flying cars are a viable product.
Arguably, level 5 autonomy for cars is a much harder problem than something like level 5 for aircraft. The extra vertical dimension means collisions are an order of magnitude easier to avoid. Biggest problem would be large birds but wind turbines seem to be making progress on that one.
The lack of pedestrians and cyclists in the sky helps a lot too.
Plus we can start regulations from scratch, meaning that all flying cars can be required to communicate with all other flying cars in the area, and to all follow some specific protocol to avoid collisions.
I wonder what the potential energy of a flying car is vs. a 'road car'. I imagine the flying car's in substantially higher both because of the energy requirements to fly them and because they're, you know, in the air.
If they have more energy, they can do more damage if misused and therefore for similar levels of risk acceptance require more substantial regulation and safety requirements.
Why would they be going faster? VTOL + fast = super expensive and complex. Are we talking Apaches (max airspeed: 150 kts)? V-22s (270 kts)? This seems insane to me.
Seems insane to me too, I just took this as part of the premise of the book. Flying cars are less transformative if they don't let us get places faster. Also, it sounds like you know a lot more about engineering constraints on flying machines than I do and I appreciate that perspective.
Even if they only go 100mph, they could pretty much go that speed continuously in a straight line to the destination. For almost any journey by car, that will cut your time in half or better...
Is it? My impression is that very few airplane accidents are just falling out of the sky like a stone. I thought they were generally shallow-angle flights into terrain, in which the horizontal component of the plane's velocity was indeed the largest.
Yeah now that I say it I think I was definitely wrong. Unless like the wings fall off. Otherwise almost anybody is gonna try to pull off some kind of glide.
This was a Reddit TIL today, a helicopter in autorotation has a glide angle ratio of 4:1 (4 horizontal to 1 vertical) so the horizontal velocity is still the large component (the TIL was actually about the space shuttle, which is scarcely better)
There is no such thing as a self-driving car. We have cars that are driven by the human being right behind the wheel right now, and we can imagine cars that are being driven by a human being in a cubicle in front of a computer some time ago, writing down all the future decisions in an enormous if-then-else tree.
Now, if you believe the *hard* part of driving a car is Newtonian mechanics -- deciding exactly when to press down the brake pedal to achieve 0mph in X feet -- or sheer reaction time -- noticing the stop lights on the car in front of you have gone on in 0.05ms instead of 100ms -- then it makes sense that turning the whole thing over to a pre-written computer program will succeed.
On the other hand, if you believe the hard part of driving a car is subtle situational judgments -- *should* I be driving the speed limit in this weather, this lighting? How probable is it that the truck in the driveway ahead is going to foolishly pull out in front of me? When that bus ahead of me comes to the intersection where I can see some construction at right, what are the chances it will move into my lane? -- then you are much more skeptical.
I've no doubt a computer-programmed car could beat the pants off any human driver in a video-game driving game, and if the real world were as perfect as a video game (or could be made that way), then it would be an obvious win for the programs.
But personally I'm very doubtful the real world can ever be made as cleanly predictable as a video game at anything approaching a realistic cost. And fortunately one of the things at which the human eye-brain-hand system has been superbly optimized by evolution is the perception of, and appropriate reaction to, weird unexpected threats from the environment.
So I can see driver-assistance technology getting very very good, and much more pervasive, and it preventing a lot of obvious dumb things humans sometimes do, through inattention, distraction, exhaustion, and I can see it taking over certain routine tasks -- super-duper cruise control, roughly. But I cannot easily see a day when you can just dispense with the driver entirely[1], except for narrow-purpose restricted situations, the equivalent of driving a closed-loop track.
------------
[1] Which doesn't mean he has to be literally in the car: another area I can see occupying a highly productive niche is remote driving, the equivalent of "driving" a drone. This would solve the much-talked-about mobility problem of people who cannot drive, e.g. grandma who can't be trusted behind the wheel can be "driven" places in her own car by her son in a city 200 miles away, or she can hire some whippersnapper in her own city who sits in a building somewhere at a console and "drives" a bunch of old people where they need to go in their own cars all day for $30/hour.
> There is no such thing as a self-driving car. [... Just] cars that are being driven by a human being in a cubicle in front of a computer some time ago
I hear people make this point a lot, but I've never understood why this is an interesting or useful distinction. Would you be able to explain that?
(And sorry to pick on a minor aspect of what you said - I liked your point about subtle situational judgments.)
It's a useful distinction because it emphasizes how very much rests on the crystal ball sitting next to the programmer's desk. He needs to foresee *everything* that could happen, in the future, to the car-brain he's programming, and design an appropriate response, because there is going to be no intelligence on the spot to take over if the program can't handle whatever weird situation has arisen.
I mean, discussions often center around the quotidian challenges of driving, and how easy these are to automate -- and I agree, a suitably trained chimp or 12-year-old human could do *most* of my driving. But as the discussion of nuclear power elsewhere in this thread demonstrates, we cannot ignore Black Swan events.
And I understand the programmer has tons of real-world testing data to help him. But one hopes the continued existence of zero-day malware suggests that this is less of a guarantee than one might hope.
Thanks, this helped. I misunderstood what you were getting at the same way as Andrew below. And now I'm wondering whether all the other times I rolled my eyes at people making what I thought were dumb semantic points about whether a capable AI was 'really' intelligent because it had been built by humans, they were perhaps trying to get at a point about robustness or flexibility or originality or something in a way I wasn't picking up on.
(This made it sound like I thought you were making a dumb semantic point, but actually due to priors and you seeming thoughtful in general, I didn't. I figured I was probably missing something. That's just usually been my reaction to similar strings of words in the past.)
I am perfectly capable of making a dumb semantic point, so I am not in the least offended by your suspicion that I might have been. I appreciate the opportunity to clarify which you created by asking the question rather than running with the suspicion.
I'll add that the quote is like saying there's no such thing as a one-person-driven car, just cars that are being driven by two people in a bed some time ago.
How so? I think you misunderstood what I said. I mean that the *actual driving decisions* ("press down on the accelerator when the light turns green") are being made by a programmer considerably distanced in time and space from the actual event, because he wrote down if (light == green) { press_accelerator; } and then that got executed later.
> we can imagine cars that are being driven by a human being in a cubicle in front of a computer some time ago, writing down all the future decisions in an enormous if-then-else tree.
I'm not sure if I'm reading you correctly, but this is not how the things that are today referred to as 'self-driving cars' work. They use deep neural networks, which only resembles what you describe if you use very fuzzy meanings of "writing down", "all future decisions", and "if-else tree".
Well, I'm not in the business myself, but I know a little of neural networks, and I wouldn't agree with that at all. During the training of a neural network, one might indeed experiment different metrics for success and failure, and look for novel new patterns, but my understanding is that none of that happens when it comes time to actually execute the learning. The net looks for patterns, based on weights established during training, and then makes decisions based on those patterns which have been precomputed and predecided. It's certainly possible that those decisions are a set of weighted options, among which the best at the moment will be chosen -- but none of this corresponds to the completely de novo decisions which a human being can make. It may be pre-programmed in a complicated way, but's still pre-programmed -- still deterministic.
It's certainly possible I'm wrong, because as I said this is not my field, but I would need to dig into it in much more detail to figure that out.
Well, there's a prominent (maybe even majority these days?) stance that humans can also be described as pre-programmed and determenistic, at least in regards to any single de novo decision they are faced with. In any case, the relevant question here, are they on average better or worse at making it than the state of the art algorithm? By how much does the algorith have to beat them before we approve it? I don't think that society is rational enough to handle these questions, sadly.
I agree that there's going to be a lot of issues between automated cars and manual cars. One possible solution is to have all cars automated, rather than allowing manual cars around the automated ones. A group of automated cars can talk to each other and coordinate movements.
As far as the transition period, that's going to be a really hard problem to solve. We can segregate types of traffic fairly easily, as with HOV lanes now, or even automated-only roads. There are still going to be a lot of problems with segregation once we reach a certain critical number of automated vehicles, but there are still too many manual cars to remove them.
I'm not sure society has the gumption to make these kinds of changes anymore.
On one hand, there's good reason to be skeptical that a computer will (maybe ever, or maybe any time soon) be able to match what a human does.
On the other hand, so many people are incredibly bad drivers. The next time you are stopped at a light, count how many drivers are looking down at the phone in their laps.
Anyone who would trade computers and smartphones for flying cars is out of their goddamned mind.
And we have flying cars - they're called "helicopters". Turns out they're not very practical and I wouldn't trust any regular person with them, but possibly self-flying ones might change that?
And to add to the same point. There are many states in the world. They differ quite a lot with regards to legislation and other societal aspects impacting experimentation and potential uptake of flying cars.
Take the australian outback or alaskan wilderness as examples. Airplanes are used as complicated and expensive flying cars in these locations. Solutions to the complicated and expensive part of that would be very welcomed, but the solutions still fail to manifest.
Sort of agree, but we have terrible flying cars and really good smartphones. Someone from a world with amazing flying cars and terrible pocket abacuses probably wouldn't want to trade either.
I'm extremely pro-nuclear, but I'm skeptical of arguments that act like the United States is the only country on earth. France, Japan (until Fukishima), South Korea, China . . . lots of countries to pick up the technology and run with it. It's certainly true that regulation made nuclear plants impossible to build it the US after the 1970s, but why don't the Chinese have flying cars?
Also it's perfectly normal for technologies to get cheaper and cheaper until they don't, you can't just assume things would have stayed on the trajectory they were on in the early days of nuclear power. DNA sequencing is a great example, which has been mentioned on this blog before.
Yeah, "why did this happen everywhere" is the best rebuttal to US-specific explanations, like the stories of US regulators shutting down specific projects over the decades. I actually find the arguments extremely compelling except for this one point. Why doesn't any other country on Earth pick up the slack on energy technology and outcompete the US?
The answer is relatively simple- nuclear power just isn't that great. It has some advantages over other sources of power (low carbon emissions in comparison to fossil fuels, more consistent in comparison to some renewables), but those advantages aren't enough to overcome the fact that it's extremely expensive. And while high regulatory costs account for a big portion of the cost of nuclear power (about a third, according to Eash-Gates et al, 2020), even if those regulatory costs dropped to zero, nuclear power would still be substantially more expensive than other options- the LCOE of nuclear energy is more than twice that of both onshore wind and solar. Just like with regulation, there are anecdotes of green activists blocking construction of specific plants, but neither environmentalism nor regulation is why nuclear power plants largely aren't built- it just comes down to nuclear power not being a cost-effective energy source.
I still think nuclear beats wind and solar when you factor in the cost of the batteries you'd need for a mostly wind/solar grid. Cost to generate electricity in the Mojave at noon is not really a fair measure.
You think wrong, then. LCOE does not measure the absolute cheapest production, it's a measure of average cost over the lifetime of an energy source. Nor does adding the cost of batteries make nuclear power cheaper. It's not as if these numbers are some secret or something that needs to be deduced from first principles- you can go look them up. I'm using as a source this EIA report: https://www.eia.gov/outlooks/aeo/assumptions/pdf/table_8.2.pdf. Even taking the most generous possible assumptions for nuclear power, that a full third of the costs could be shaved by dropping all regulation (which implies that there's no reason to spend any money on safety), nuclear underperforms solar and wind with batteries.
Well, whatever the theoretical arguments, I live in an area where a big nuke was shut down a few years ago *and* people have been installing solar like crazy. Before the shutdown, electricity rates were cheap. Since, they have done nothing but climb and are among the most expensive in the state now. So as a practical matter, this doesn't hold water for me, and I don't care how many reports say otherwise.
"You may have objective data compiled from detailed studies, but I have a single anecdote where one event happened after a different event, and therefore I've concluded that the first must have caused the second and no amount of facts will ever convince me to change my mind!"
You realize this is supposed to be a rationalist blog, right?
According to Lazard rooftop solar is $150–227/MWh, which is a tad more expensive than their estimate for new nuclear ($129–198), marginal (running) cost of existing nuclear ($25–32) or new utility scale solar or wind ($29–38 and $28–34). However, it is true that renewables were more expensive a few years ago (e.g. solar $49–61 in 2016).
If solar is really that cheap, why hasn't much of the grid been replaced by solar yet? It's only 2% in the US. Even with enough batteries to cover the night, you can still get screwed by weather, so you need a whole backup plant that's idle most of the time, which may or may not be included in that LCOE.
Also, it's easy to imagine scenarios where regulators increase costs by 10x or more. I'm not sure a third is sufficiently pessimistic. Apparently nuclear is competitive enough to still be worth building in China, even though the average cost of energy there is only 8 cents per kwh.
The avoided cost (money saved) isn't the same as the old levelized cost minus the new. It's closer to the old *marginal* cost minus the new levelized cost, since when you build a new power plant early you don't get back the money you spent on the old one. EIA's modeling is, as I understand, a little more detailed than that (though still simplified for obvious tractability reasons).
TL;DR: ignoring whether there are any other factors effecting the cost difference for the sake of the argument, a third is probably high, a tenth is probably low, I'd put my money the average being between the two. The wholesale LCOE in general estimates isn't actually that much higher than $8/kWh though, Lazard puts it at $129 to 198/MWh. It's just that (utility scale) solar and wind are much cheaper than that.
I read the wiki article on LCOE and it seems to imply none of the costs of backup capacity are included in LCOE, so LCOE is not a fair metric by which to compare intermittent solar/wind with more consistent sources.
A big part of the answer should be that solar only became cheap about 5 years ago, and the buildout will take time. I'm not sure how long it takes to secure the necessary land and transmission lines, but the U.S. isn't known for speed in such matters.
I find it very hard to believe that regulations only account for a third of costs, because the realized cost of nuclear reactors more than doubled around 1980, and in turn, the regulations in the 1970s were stricter than the ones that governed the USSR and their Chernobyl reactor, or even the U.K. Magnox reactors. Something quite suddenly made costs ridiculous. If not regulations, then what?
If it took ten years to build a car, how much would it cost? There is no inherent reason why a small nuclear power plant should cost more than a comparably-sized car.
The problem with this argument is that it ignores the way electricity is bought and sold on the open market. Electricity is a unique commodity which must be consumed the moment it is produced; it cannot sit on a shelf.
So the raw per kWh costs are extremely misleading. What everybody obviously cares about is total system costs and total system reliability. And as an electrical engineer that studied power networks and have my name published on a conference preceding for solar (thank you very much) (these credentials aren't to say that I'm an expert, just evidence that I'm not biased against solar), you can't run a whole grid off renewables. Running a grid is all about matching the electricity supply to the electricity demand. You need to forecast electricity demand and then try to crank up and down your electricity generation to meet it. Unfortunately, standard fuels like coal and gas are extremely well suited to this task. Solar energy has been cheaper per kWh than natural gas for literally two decades now, why are we still debating this issue? Why hasn't it simply been deployed everywhere? Because each kWh of electricity is worth a lot less when you can't control when it's going to be generated. Gas may seem more expensive, but you can combust it literally whenever you want. There are no simple equations to express this, you need to consult elaborate models, but rest assured the people building the power grids of the world are doing this and that's why they keep building gas.
I'm certainly not arguing that solar/wind are unequivocally better than fossil fuels in terms of cost or effectiveness for supplying a grid- there's a reason we've used fossil fuels for a long time. The reason people are advocating switching away from standard fuels is their high externalities. Once you've made the policy choice to convert some production from cheap, easy fossil fuels to power sources that come with more complications but less carbon emissions, you compare costs among those alternative energy sources, and nuclear power rarely is the best option. Of course solar isn't as cheap and easy as coal, but it's cheaper than building nuclear power plants.
And in any case, as you know, it's not so easy to scale nuclear power generation up and down quickly either- both nuclear and solar/wind baseloads need to be supplemented with power sources whose generation can be varied more quickly, to compensate for variations in demand and supply.
I don't think the biggest problem with solar is handling quick fluctuations, it's "the sun sets and you have zero electricity". As far as I know, no place stores a meaningful fraction of electricity generated during the day for use at night. Batteries are just way too expensive, and this isn't reflected in the LCOE number.
FCAS is actually fairly easy to supply via batteries, to the point that the market is rapidly becoming too saturated. Energy arbitrage and resource adequacy on the other hand, could really benefit from batteries that are much lower $/kWh, even if they have to sacrifice kWh/kg for it. IMO the high density types like Li-ion aren't well suited in that role.
There are reactors we know how to build now that can modulate their output to a sufficient degree. We have a clear technical path to nuclear power that can modulate it's output to a high degree, at least as clear of a technical path as our path to better and cheaper batteries. Nuclear is, by default, a baseload. Solar and wind are not.
Even combined with batteries, whether solar and wind can adequately supply a base or variable load is highly dependent upon geography. In many places, we can expect fairly consistent sun and wind (although the whole idea of climate change is that the future of the climate and therefore weather is highly uncertain, which means that even these are not necessarily great assumptions). But in other locations, some days you get 10% the solar irradiance as you might get other days. This would suggest that you need to overbuild on the order of 10x in these locations if you want to reliably attain a minimum output, something that all other energy sources are capable of. None of these can be logically factored into a single number, which is why LCOE numbers are extremely misleading. You can't talk about per-unit rate of renewables agnostic of where you deploy them, it doesn't make any sense. Policy makers should not be using such simplified metrics. The utility companies themselves are not using them when they decide what to build. The fossil fuel lobby (specifically natural gas companies) are lobbying FOR policy makers to use these numbers because they know they're misleading and they will lead to a policy future where natural gas cannot be eliminated because it's the only way to deliver reliable service. Gas turbines can spin up and down very quickly and honestly are the most versatile type of power generator, easily functioning as a base or variable load at the same unit costs. This is why in Japan, New York, and everywhere else that nuke has been taken offline, it's been replaced by gas, not renewables.
> There are reactors we know how to build now that can modulate their output to a sufficient degree.
Yeah, and we can modulate the output of renewables as well, its called "turning the damn thing off". We've been able to do that with nuclear for ages now, France has *done* that for ages with their nuclear fleet. That doesn't change the economics of "if you have an expensive reactor that costs next to nothing to run, you better run it as much as you can or you won't be able to pay your loans off".
EIA's LACE calculations are regional, so unless they're as misinformed as the guy who thinks solar had a lower LCOE than gas for two decades, I'd say their doing some forcasts is better than that same guy going "trust me bro, thats what the oil lobby wants you to think".
Yes, gas is cheap, and it's one of the only sources that has a better LACE than breakeven.
It's an OK rebuttal, but not super duper, because human beings are rather monkey-see-monkey-do creatures, and are perfectly capable of following a leader (e.g. the US) even where it makes no sense. One would like to *think* each country thinks independently, but then again, we'd like to think each US state thinks independently, or even that each individual human thinks independently, and we are often sadly disappointed in this expectation.
Seeing as Fukishima got whammed with a tsunami, I think it's a little unfair of him to sniff that the reactor would have been perfectly fine had it not been for those panicking officials. Yeah, they were panicking - because a tsunami was coming to wham them! And I think we can cut Japan some slack for being Very Nervous about nuclear radiation and civilian populations.
I think you're missing the point of my comment here. I'm saying that Japan was pretty okay with nuclear technology from 1970 to 2010, had an economy roughly half the size of the US's and an advanced tech industry, and yet they didn't produce insanely super cheap nuclear power. That can't be blamed on American regulators.
Right, but the effect of their being a power plant in the earthquake/tsunami zone is that a natural disaster that killed some 19,000 people and devastated several thousand square kilometers, killed maybe one extra person and a few hundred square kilometers of the affected area will take decades rather than years to recover.
The rational lesson of Sendai is "massive earthquakes are really bad things", not "nuclear power plants are really bad things".
Well, France does make (much) cheaper reactors. So does South Korea for example. Here is a link (super-quick googling) from Vox. I'm not a fan of Vox, but on cursory reading this seems alright:
There's a few issues there- while it does show that France built reactors for cheaper in the past, it hasn't built any new reactors since the early '90s. And while Korea does show that reactors can be built for cheaper than they are in the US, the lowered costs on Korean reactors (2.5 million krw(2010)/kW, which converts to $2,755/kW) are still not competitive with renewables like solar and wind. And an important difference between those countries and the US is that they have largely centrally controlled utilities, whereas the US has a bunch of different companies that face harsher competition to provide low-cost power and are less centrally controlled.
Just a note: nuclear does not compete with solar/wind. It competes with gas/coal/hydro etc. They provide different services and there isn’t an apples-to-apples way of comparing their costs.
That's not really accurate either. The difference in role of gas peakers and your regular coal, nuclear, or even CCGT plant is way larger than the difference between that CCGT plant and wind. Ignoring the jargon of marginal cost or merit order or whatever, if you build loads of renewables the coal and nuclear plants are what's going to be driven out of business because they're they type of plants that cost loads to build and next to nothing to run, and therefore are the ones that you want to have trundling along at 100% of whatever it can produce for it to make economic sense.
A nuclear design like Moltex's seems, roughly speaking, like baseload combined with a peaker, truly complimentary with renewables: https://www.youtube.com/watch?v=cQCm-kmUWA8
One important part here is that France built a whole set of fairly standardized reactors. In a world that actually cared about the environment, there would be an EU or a U.S. "standard nuclear plant" that could then be produced on a large scale and more easily maintained, but clearly we're never getting _that_.
That's Robin Hanson's take on why the book applies globally. I don't know why it would also apply to China. It's much more plausible that India would inappropriately adopt regulation popular in the first world
Other countries have even more stilted regimes of regulation than the US though. There's really no example of a first world country that maintained a lower level of regulation than in the US and would therefore make for a good comparison on what would have happened. Additionally the shear total size of the scientifically educated in population in the US absolutely dwarfs that of these other countries.
In China for example, having well over a billion people doesn't get you to flying cars faster when most of them are living at levels 1, 2, or 3 and your regime stifles new thought far more dramatically than in the US among the few level 4s who might try to make flying cars a thing.
Exactly. Soviet Union might be an even better example. They had much less restrictive rules on nuclear energy, great engineers... but this did not make energy ridiculously cheap, compared to the US.
This was my thought reading the review - I'm surprised the reviewer didn't mention it. The book's author appears to be saying not just "nuclear is the best of our available power sources" but "widespread nuclear energy would represent a paradigm shift on the order of wood -> coal or coal -> oil." 1970s USA might be more innovative than 1970s USSR, but if less-restrictive nuclear was this level of a gamechanger you would still expect the Soviets to have been dominating the world economy with all the cheap stuff they made, not desperately struggling to maintain military parity with the US while running out of grain and toilet paper. Obviously communist inefficiencies do count for a lot, but I would expect a communist Industrial Age country to easily outcompete a capitalist Baroque-era one.
Well, the U.S. has been very good at exporting its culture. You know all those cultures in National Geographic where nudity is normal? Mostly they've been shamed into wearing clothes by now. So, maybe the U.S. culture of fear of nuclear power spread really well around the world. The cold war, the bombs on Japan, and the activists working so hard to equate power generation with bombs enhanced the effect.
It's surprising that this would affect the whole world, but it's not literally the whole world we're talking about, only countries that already had reactor-building programs. In any country that didn't have regulations for nuclear power, it would be de-facto illegal to build a reactor — or so I infer, because you could save a lot of money if you didn't have to meet any regulations, and yet no one has ever taken advantage of this. AFAIK the closest thing to unregulated nuclear power is the USSR's Chernobyl design, the RBMK. Any country that is considering opening up to nuclear power is likely to copy regulations from another country, so the regulations are likely to be very strict post-1979 regs rather than "medium strictness" pre-1979 regs.
So, assuming a strict regulatory regime, I have high hopes for Molten Salt Reactors, which, for example, should be able to avoid the need for a super-expensive "nuclear-grade" concrete containment structure, because plausible accident scenarios with MSRs don't need to contain large amounts of high-pressure steam, and for the most part, plausible accident scenarios don't produce large amounts of airborne radioactive material.
P.S. Robin Hanson points out that far beyond the nuclear industry, regulations weirdly end up almost the same everywhere. And he's not even talking about international treaties that force regulations to be the same (e.g. copyright, WTO TRIPS, and isn't there a treaty about nuclear fuels?) https://www.overcomingbias.com/2020/09/the-world-forager-elite.html
The ML analogy is interesting. One of the most pernicious problems in designing a good ML tool is getting the objective function right. In the 'economy-as-ML', I'm not sure what the objective function would look like at the 'economy' level.
The objective function at each node is price discovery--profit maximization for sellers and the tradeoff of purchase for buyers. The analogy only goes so far but basically as buyers have money to flow one direction and sellers have products to flow the other, prices (the values at each node adjust to balance the two.
The interesting thing about the analogy is that you effectively have a recursive two-layer neural network, with a separate 'dimension' for each product in which you use a different subset of the available nodes for buyers and for sellers, and a different set of values for the prices, with the max of the values being limited by each buyer node's participation as both seller and buyer in all other dimensions (products), and the minimum of the values being limited by a seller's production capacity and net gain vs the net value of participation in a different dimension. Trying to actually simulate something like in a computer would immediately trip over several exponents with only a couple products and a handful of people, but I think it has a lot of merit as a thought experiment.
Alternatively, the objective function is value creation. People do not buy your product or service unless you deliver value for them. Price discovery is to ascertain a quantitative measure of the value. But we are not discovering prices for the sake of it.
> Then there should have been nuclear fission and nanotech, letting you fit a lifetime's worth of energy in your pocket.
No, there shouldn't have been, because pocket fission and nanotech (that is, the non-trivial molecular assembler kind) are both physically impossible. Science is not a magic box that takes in money and outputs anything you want. No matter how earnest and enthusiastic you are, and how efficient you are at spending money, you will never get FTL travel or pocket fission or cold fusion or magic transmutation nanotech or anything else of the sort. You *would*, however, get high-energy-density electric batteries and intelligent control software, both of which are finally enabling flying cars today.
Ribosomes are not particularly useful to human operators. Sure we know how it takes the code and puts out amino acids, but there's a lot of processing afterwards and people have devoted their lives to figuring out how proteins fold into their fully functional forms.
My understanding is there have been great advances in understanding protein folding lately. You can't think of a use for being able to make arbitrary proteins?
What exactly are you trying to accomplish with ribosomes ? If you're trying to nudge existing living organisms to be slightly more useful, then ribosomes are the way to go, for sure. If you're trying to nano-assemble computer circuits and space elevators out of carbon, silicon, and other raw materials, then ribosomes won't help you in the slightest.
It certainly won't. But I'm trying to broaden the definition a little here. After all, in 1969 "technology" included missiles and spacecraft and nuclear stuff -- but computer programming for commercial applications was boring stuff people with HS diploma only did, punching cards to transfer a lot of payroll data into a mainframe.
I don't think Feynman would be disappointed by what we've accomplished in the biomolecular realm.
And I wouldn't be so quick to count the ribosome out yet: what if we could engineer an alga that turns CO2 and water into diesel economically? Shazam, no more climate change worries. What if we could engineer another that would turn 98% of Venus's atmosphere into O2 and water? Or...what if by learning how Nature does it, we get some big key insights into how we can do it? All very speculative stuff, but not as daffy as imagining the Alcubierre drive will work out.
> what if we could engineer an alga that turns CO2 and water into diesel economically? Shazam, no more climate change worries. What if we could engineer another that would turn 98% of Venus's atmosphere into O2 and water?
It would be nice, sure, but it's not enough to just say "what if"; you have to have at least some indication that your plan is possible. Right now, it doesn't look like it is; for example, while extremophile bacteria do exist, AFAIK none of them would last very long on Venus (and those that would, wouldn't convert CO2 to O2 and water anyway). Sure, you could posit hitherto unknown algae that work in some completely undiscovered way in order to do all the things you want them to do -- but then, how are they different from magical pixies ?
It seems that you classify the set of "every thing any living organism can do" as "trivial," which sounds like we're working with very different definitions of the term.
Yes, I do classify it as "trivial", for the purposes of building flying cars and space elevators. Yes yes, I know that we are technically living organisms, and we could build flying cars and such, but I was referring to using organisms as tools on the molecular level.
Why doesn't rational drug design and the like count as "nanotech?" You may not be building the molecules with molecule-sized tweezers, but they certainly operate on the atomic scale, and you certainly need an understanding of the atomic-scale processes to design them.
Rational drug design is what stopped the AIDS epidemic. Rational mRNA design is what has just stopped the COVID epidemic. Every drug with an "ib" at the end of its name is a monoclonal antibody raised against some particular atomic-scale target, based on atomic-level understanding, and there are a lot of them these days. There was a kid in France who was *cured* of sickle-cell disease by laboratory editing of the genes in his stem cells.
I think the issue is that organic 'nanotech' is for the most part limited to a much smaller subset of the periodic table than what we'd want in order make durable products that have a desirable set of functions outside the body. And what happens at the biological nano-level mostly stays at the biological nano-level. The promise of nanotech is that we can have designable nano-scale features in a macro-scale product arraigned in ways that create desirable function. Today we can mostly only create nano-scale components as a soup of many copies of components, and worse we can't really control how they are assembled into larger scale objects without adjusting the components themselves. That's makes for a nasty engineering problem that is currently pretty much impossible.
I don't think rational drug design is nearly as developed as you make it sound. We can't really predict or simulate the function and interaction of molecules very well yet at all outside of knowing the function and interaction of very similar molecules. Drug design is mostly a lot of testing of molecules that are analogues of ones with known interactions, and using strong artificial evolutionary pressures to both find a random solution and eliminate non-solutions. We're not designing large enzymes from scratch, or coming up with molecules that can chemically attract a disease protein without the aid of some animal's immune system.
Sure, I agree "nanotech" as it is written in sf magazines has not come to pass, and I never thought it would. Materials science has not stopped advancing, and very nice things are being done with it, but it's not a revolution and only credulous people thought it would be.
But I don't agree that there hasn't been *remarkable* progress in biology with applications to medicine. For some reason -- perhaps because it is has a lot more women in it than physics or materials -- these things don't seem to become part of that future-loving speculative discussion. I'm suggesting it is actually some pretty revolutionary stuff happening at the atomic scale that is about as exciting as anything "nanotech" was going to delivre.
I hope by "rational drug design" you don't think I'm limiting this to in silico studies. Nothing of the sort. I'm including the insights and collective wisdom of all the med chemists, too, and all the immunologists and studiers of genomes. All that is part of the way people go about attacking disease with molecules (or biomolecules) these days, and it's *way* more "rational" (meaning deliberate, based on genuine understanding of what's going on) than it was even 20-30 years ago.
We've literally done cold fusion with muon catalysis. The problem is getting to breakeven with it, because muons have a half-life of 2 microseconds and there's no obvious way to efficiently make them.
I can see why pocket fission is impossible, because fission relies on criticality and there's such a thing as a critical mass. But why is a molecular assembler impossible? Life clearly does molecular assembly all the time.
Critical mass isn't that big a problem. I could hold a bare-sphere critical mass of plutonium in my hand (although I'd need my other arm to hold it up). The issue's more with control and shielding.
For the critical mass to be "small" (small enough for a car), wouldn't you need nuclear material almost pure enough to make a bomb? Can't imagine that being legal.
But yeah, neutron and gamma shielding is hard. Neutrons, being neutral, can fly some distance through solid objects until finally hitting and ionizing something. Gammas are also hard to stop for some reason. (Betas & alphas are charged, and so easier.)
A less pressing challenge is heat dissipation — it should be easy to manage for a small reactor, but maybe not so easy that any random person would be allowed to own a reactor.
You can get around the "lol homemade nukes" issue by including a bunch of SF-happy stuff that would predetonate and fizzle a bomb - ideally, SF-happy isotopes of the same element the main core's made out of (that way you'd need an isotopic separator to fix it, and if you have one of those you can build nukes anyway).
On reflection, there are ways to get control (and you need control, both to turn the reactor off/on and to allow high burnup before the pile fails) without needing neutron poison and the associated large increase in critical mass - removable moderators and reflectors can do it, and can even potentially do it in a failsafe way. Gamma and neutron shielding is still a bitch, though.
This book sounds as if it would tie in very nicely with the previous review, "The Collapse of Complex Societies"; to the question "What went wrong in the 1970s?", my initial response was "It was the 70s".
Tainter would probably say "complexity"; the people alive at the time would have said "nothing much wrong (well, crime in some cities, sure, and the whole domestic terrorism thing, and oil shortages and things, but otherwise pretty okay)"; but I never heard anyone, until now, say "a lack of nuclear power plants".
If this review is faithful to the book, the guy sounds like a crank. Let's take a few points:
"The average American moved from Level 2 in 1800, to level 3 in 1900, to Level 4 in 2000. We can state the Great Stagnation story nearly as simply: There is no level 5."
Oh yes there is; aeroplanes. Private citizens can own their own, from the basic models to the billionaire's private jet, and the affordability of commercial flight has never been so within the range of the ordinary person (to the point of griping about airlines packing passengers in like sardines and crowding more and more is smaller and smaller seating, which shows the success of 'I want to take a flight and I can afford it' for ordinary people).
In fact, the progress that Hall describes from Shank's Mare to Your Very Own Automobile was forecast by many writers to take that Level 5 step: Your Very Own Aeroplane. Indeed, Kipling for one went so far as to put the entire world, now that air traffic was cheap and commonplace, under the control of the Aerial Board of Control: https://en.wikipedia.org/wiki/Aerial_Board_of_Control
As for the wicked interference with developing flying cars, Hall seems to want highways in the skies. And think about that - think about Highways. In. The. Skies. Think of the worst snarled-up rush hour traffic, cubed (because you can stack flying cars), in the skies overhead for major cities. Think of your commute, only now you're up in the air.
Hall has the notion of primitive days before everyone could afford their own car, only transferred to flying cars. Hop into your jalopy, rev 'er up, and flit from home to town and back again in the clear blue yonder. A charming notion, but in reality, it'd be you and six thousand others up in those skies.
The terror of nuclear energy comes from it being used first in war, not in peacetime. As the most destructive and terrible of weapons, not as a source of energy generation.
"If you put a completely legal luminous watch in a barrel containing half a tonne of dirt, that dirt would technically be intermediate-level nuclear waste according to the regulations."
Ahhh - when did radium watches become legal again? I think he's doing some three-card-trick shuffling of terms here; radium is radioactive (read this Wikipedia article on what happened to women working with it for watch dials https://en.wikipedia.org/wiki/Radium_Girls) but now luminescent dials are using " phosphorescent- or occasionally tritium-based light sources" which are safer and hence legal.
He is definitely pining for the 70s, when you could wear radium watches on your wrist and fluoroscope your feet! https://en.wikipedia.org/wiki/Shoe-fitting_fluoroscope A little dose of radiation will do growing kids no harm, in fact it'll encourage them to grow in new and interesting ways!
I do very strongly get the impression that Hall isn't interested in flying cars qua cars, he's interested in them to get that interfering government off his lawn (quick, where's his shotgun, here come those dadgum Revenuers again!)
As for the flying cities, James Blish had an entire series about them https://en.wikipedia.org/wiki/Cities_in_Flight . He had them powered by spindizzies, something I think perhaps superior to the idea of "one single line of nuclear power plants (every 250 feet) along the wing would suffice to keep Aero City flying indefinitely".
My main objection here is that he envisages everything working perfectly all the time (once tech has been so perfected by being liberated) so that nuclear plants will never fail, flying cars will never crash into each other or drop out of the sky, and as for his flying city... well, let's not even imagine how catastrophic that could be.
1. Tritium is radioactive, and it's likely that's what he meant. The problem with radium on watches is in its manufacture, not its use. In terms of wearing it, you'd rather have the alpha emitter (radium) than the beta emitter (tritium) because the alpha is stopped dead by even the thinnest watch back, while the beta might well burrow through and into your skin.
2. It's much harder to get traffic jams in the sky, because you can move in three dimensions. Each evening in Carlsbad Caverns, about 1 million bats exit the cave in an hour or two. If it were a million equal-sized mice trying to crawl out of the same size hole, it would take days.
The traffic-james-in-the-sky line of thought is interesting. No, you wouldn't have traffic jams, but the ground storage for aircraft would surely be finite and you could have a much, much worse problem of too many aircraft trying to land in the same area. Cars have an advantage here where they don't require much energy to sustain themselves in totally deadlocked traffic, and worst case scenario they need a tow. Aircraft have a landing problem. Flying cars would only move the problem of congestion from the via to the destination, but "too many people want to go to the same place at the same time" doesn't have an easy solution (except dense living and public transit).
Sure, and the congestion near existing commercial airports is already pretty hairy at times. One assumes that flying cars would be something you only really considered for a daily driver if you lived pretty far out and went to work not in a dense urban center. But that's still a lot of people. One could argue the automobile presented a similar challenge to the cities of 1900, and what happend is that a lot of people and jobs moved to the suburb.
If I think about it personally, as in, I just won a nifty flying car by falling through a wormhole and winning an AD 2200 version of "The Price Is Right," my main hesitations are (1) cost -- I feel like this is going to cost even more than my R8 -- and (2) safety, as in, am I going to feel like I *cannot* get in this thing unless I am well-rested, calm, know exactly where I am going, have planned everything out meticulously? I don't *drive* like that, all the time, and if I had to be in mental tip-top shape to go to work Monday morning or jump out for a burger and fries...well, meh.
But on the other hand...I like to go skiing about 400 miles away, and if I could jump in the Jetcar and be there in 3 hours, door to door...the pulse quickens. An afternoon on the slopes whenever I want to make a bit of an effort? I would need less scotch after certain days at the office.
"congestion near existing commercial airports is already pretty hairy at times"
Some of that is because we have taken the problem of motorcycles and tractor-trailer trucks co-existing on the same roads and multiplied it by several times. I think the difference in vehicle weight between a jumbo jet and a Cessna is probably bigger than that.
Radium daughter isotopes can emit substantial amounts of gamma radiation, and radium is also taken into the bones, unlike tritium which, being a hydrogen isotope, doesn't stick around for too long in the body. Also, tritium is an extremely low energy beta emitter, stopped dead in its tracks by a sheet of paper. You're wrong about radium being safer than tritium.
Sure, that's possible. You're right I should have considered the whole decay chain in the case of Ra -- that was silly. But I will also note unless you eat your watch dial, it's not much of a worry that it's in Group 2A.
As others have pointed out, plenty of other countries have embraced nuclear energy. That doesn't automatically enable flying cars. Unless we're going to put reactors in individual vehicles, which we won't do, you'd need better battery technology to take advantage of widespread nuclear energy. Besides that, there are so many other reasons we don't have and probably don't even want flying cars. Is every apartment building going to have its own air traffic controller? Automated systems on roads work fine because they're roads. Who keeps arbitrary flying machines on pre-defined flight paths? We can do that now because of how controlled and limited airplanes are. Are you going to file a flight plan to go to the grocery store? What happens when a flying car breaks down? It's fine to pull over to the side of the road. Falling out of the sky in the middle of a crowded city is not so fine.
What we actually want and are pretty close to having is widespread fast delivery, ideally via self-driving cars, so people don't need to go to stores at all, and better telecommuting and VR technology to limit the amount of interstate travel and daily commuting that needs to happen in the first place. Flooding the sky with amateur aircraft is not the answer.
Ok, so this book review isn't bad, but the book sounds too error-filled to be worth anything. DDT didn't get banned for causing cancer; but it is associated with increased Hodgkin's lymphoma, testicular cancer, and liver cancer. https://en.wikipedia.org/wiki/DDT#Carcinogenicity It got banned for it's impact on wildlife, IIRC. The author sounds like some kind of fundamentalist. Hard pass.
This definitely stuck out to me on first reading. I don't know if Where's My Flying Car goes into any more depth, but... the impact on wildlife is why Silent Spring is called Silent Spring. Because places where one would previously hear sounds of birds had gone silent. Characterizing it as weighing against the use of DDT on the basis of it being carcinogenic is not a mistake one could reasonably make given cursory reading of the book, or even brief reading *about* the book.
After reading the Wikipedia article on DDT, it appears that that the DDT ban was due to fear of it causing cancer *and* its impact on wildlife, particularly birds.
At least the book doesn't continue the lie that DDT would have solved malaria (DDT is still allowed for malaria control purposes, but many places have moved on anyway because the mosquitoes developed DDT resistance).
>Imagine if you had to bike to a train station to get anywhere (not such a leap of imagination for me in New York City! But it wouldn’t work in the suburbs).
Wouldn't work in the suburbs we have in the US, perhaps, but my understanding is it would and does work in many European suburbs. (Plus you can get around more effectively just by bike there too.)
Here in the US we have this sharp divide between cities and suburbs (where it's impossible to get around effectively without a car), but it doesn't have to be like that in the first place, and in many parts of the world it's not.
Anecdote from the last time I was in Brescia - the streets were jammed with parked cars during a weekday. Not sure if they were from people coming into the city and taking the train, or from people taking cars instead - I was there on vacation and had taken the train into the city.
There's an idea in economics called the Environmental Kuznets Curve which posits that as societies become wealthier, people become less willing to tolerate environmental degradation. Or, alternatively, are willing to spend more money to remediate or avoid environmental degradation. A few years back, I read somewhere about a similar effect regarding certain types of personal risk. E.G., if you wish to understand why children 50 years ago were allowed so much more unsupervised play time than kids today, it's because parents were less risk averse. If you wonder about why the list of mandated safety features on automobiles keeps getting longer (crumple zones, airbags, electronic stability controls, etc.), it's because drivers are more risk averse. And so on and so forth. As someone who leans libertarian, I'm quite sympathetic to concepts like "bureaucratic incentives are to calcify," but I think it has to be pointed out that our bureaucratic paralysis wasn't exactly imposed on an unwilling or skeptical public.
Exactly, and relevant to the comment by Crouchback just above yours.
The reason nuclear reactors cost 3 times in the US what they cost in France is because the voters in the US don't want them while nobody asked French voters how they felt. (I exaggerate for effect.)
Pinker makes the point somewhere that the death penalty still exists in the US (and not Western Europe) is, counter-intuitively, because the US is more democratic. I think this is an insight.
I too am a libertarian. I watch with disgust a government that keeps being a bigger slice of the economic pie. But then I talk to my neighbors and realize we have the government we can afford. As we create more wealth we get a bigger government. Likewise, as we get richer we spend more on education and healthcare. Etc.
The late Mark Kleiman theorized that the difference in capital punishment between the U.S & elsewhere is because civil servants are more education and that results in them opposing it. But when I looked at the General Social Survey (which is US-specific), a majority supported the death penalty at every level of education:
You can explain it in terms of impact on quality of life rather than gdp. E.g. The marginal benefits of cheaper energy become smaller compared to the marginal harms of the air being less breathable. Even if lower cost energy more directly contributes to gdp on paper
So how does he explain France? It's famous for dirigisme and arguably is the most statist economy among Western democracies. It is also famous for being #1 at exploiting nuclear power. If anything you could argue that nuclear power is the most socialist form of energy. Likewise does he assume that every single nation as developed as the US made the exact same mistakes so we're all equally handicapped in technological development? I can't help remembering that libertarians were big fans of the Dale.
Personally I found Robert Gordon most convincing - we solved the easiest problems by the 1970s and further gains in productivity were slower because they were more difficult.
It's impossible to answer your question without knowing what you mean by "tons." But I see this on Forbes: "During their 13th Five-Year Plan period from 2016 to 2020, China built 20 new nuclear power plants with a total capacity of 23.4 GW, doubling their total capacity to 47 GW. And that is expected to happen again during their next 5-year plan, which has a new target of 70GW of nuclear generation before 2025."
I would count 20 new nuclear plants in 5 years as "tons."
Ok, so then this looks like evidence that the theory has something to it. I didn't realize they were building at this rate, but i guess i didn't do much research. Thank you for sharing!
Fukushima put the brakes on the Chinese nuclear power program exactly as it did the Japanese. The CCP is very risk averse; it was terrified that a Chinese Fukushima would lead to a large scale revolt and societal instability.
You underestimate how much the world follows the American lead on regulations and decision making.
I think you raise a good point, but I don't think I would put it that baldly. I was living near Three Mile Island at the time, and while people were concerned, it wasn't a freakout, and it wasn't nearly as uniform a response as your sentence suggests -- and I grant you may mean a more nuanced thing, and be just summarizing here.
I would say (and I have not researched this, just going on having lived through the period in question) that TMI was more a symbol and possibly an accelerant of a transformation of public opinion that was well underway already.
But I certainly agree the establishment of the DoE had nothing to do with it. If anything, that was just a symbol of transforming attitudes. As I recall, the main argument for it was the "energy crisis" and Carter's (as it turned out futile) wish not to be seen as doing nothing about it, come the election of 1980. So kind of like the Bush Administration's creation of the TSA in response to 9/11.
Costs actually started rising well before either of those events (see https://imgur.com/a/fEG1WKx, from https://gordianknotbook.com/). TMI might have contributed some, but I think it's pretty clear that there was a lot more going on.
My impression is that the opposition to all things nuclear started in the 1960s or even since the bombs on Japan; it gradually gained steam until Three Mile Island supercharged it in 1979. I think this is related to the cost increases. See also: http://www.phyast.pitt.edu/~blc/book/chapter9.html
I wouldn't really call this a review. It's just a naive summary of the main points of the book that doesn't critically engage with any of the ideas. There's not a single word written on whether the author's claims hold water or whether his argument is logically coherent, not a single outside source referenced, nothing to suggest that the reviewer has any understanding of the subjects where Hall is making controversial and unsubstantiated claims. Bluntly, it's the level of writing I'd expect from a 5th grade book report, not a finalist in a review contest.
Flying cars are an inherently bad idea for many reasons. This is the best analysis I've found that lays out all the reasons why. It concludes that it would be much better to normalize travel on small, autonomous planes based at dense networks of small airports than to try putting a flying car "in every driveway."
I was going to object that many (though not all) of those problems go away given the orders-of-magnitude improvements in energy technology Hall thinks we should have... But then I remembered that Hall thinks an actual '70s design would have been successful if not for regulators, so really it all applies.
With the benefit of being able to remember the last 40-odd years, I would say the general description of the social change is accurate: there's no question to my mind that modern society hugely more timid than what it was 30-60 years ago, and far more concerned with equity than opportunity.
But there's still some serious chicken-and-egg questions about that. Did we become more timid because of this change or that? Or was it our timidity that drove those other changes? There's a strong interconnect between social attitudes and technological change, but one can make an argument for the causality running either way, and probably in different areas it runs in different directions. Plus there are probably feedback loops, in the sense that once a change of attitudes causes a change in technological progress, the delta in progress informs further change in attitudes, et cetera.
It's definitely the case, though, that the difference between American culture and Soviet culture has narrowed in completely surprising (if you remember the difference in the 60s) ways. We really are much closer that I would ever have thought possible to the culture that, were it told the chocolate ration had been "increased" to 20g, would pretty meekly go along, and viciously choke off any troublemakers who attempted to point out it was 40g last week.
I certainly wish I knew why. I'm not convinced by this it can be easily traced to a failure to go full bore on nuclear power. That feels more like another symptom to me.
I wonder whether this has something to do with the wars. In a society where a world war was just 20-30 years ago, and the one before that 50 years, then you are not so confident that your society still exists in 20 years from now. If you have these looming risks anyway, then you are less concerned about an exploding power plant, or about destroying your environment.
In most European minds, facing a society-threatening risk like a war in your own country is just of of scope nowadays (except for the far right). And if you expect firmly that your 10-year old child is going to live until 80 in the same stable society, then you worry a lot more that it might have a bad experience tomorrow, or that it is long-term poisoned by some chemicals. And the same for yourself.
I wonder whether this explains part of the more progressive philosophy of the rationalist community. A lot of people here are concerned about X-risks like a badly aligned AI. If you believe that there are real society-threatening risks, then you are less concerned of adding more minor risks. On the other hand, if you firmly believe that society will do well for the next 80 years, you are more conservative.
Ironically, this expectation of doom is something that (parts of) the rationalist community share with right-wing parties like the Republicans. At least Scott Alexander made a convincing argument that this is something that sets apart Republicans and Democrats in the US. But where the rationalist community is concerned about society destroyed by an AI, Republicans are more concerned about (their) society being destroyed by too much immigration.
Are you familiar with the "hygiene hypothesis" for the higher rates of autoimmune and allergic disease in First World countries[1]? The idea is very roughly that in the absence of enough real threats (infections to fight off) to occupy its attention the immune system starts to find imaginary threats and attacks your own cells.
My pet theory is a sort of social hygiene hypothesis, in that if there aren't enough real threats to a society people just start making them up, or at least magnifying small threats until they look like big ones. Id est, if in your lifetime Hitler might get The Bomb before Patton takes him out, or Stalin might kick off World War III at any moment if he feels sufficiently pissed about Berlin, that might...focus your existential concerns. Contrariwise, if you live in an era devoid of much in the way of plausible immediate existential threats, and where there hasn't even been a serious recession in 12 years...maybe you start finding all kinds of rather modest problems and blowing them up into problems that feel around the size of the Cuban Missile Crisis.
I'm not convinced humans are well-designed for (general) happiness. If we were let back into Eden, I'm pretty sure we'd find some reason why it was awful, or we had to fight each other to the death over whether to pick pears or apples first from the Tree of Life.
It's not that the problems are just imagined. (Ok, some are.) It's more that the level of agitation is not adapted to how serious the problem is.
Also, when there is a real existential threat, then society is aligned. All people fear the same thing. But without an existential threat, people focus on different fears and problems, and get unsatisfied when others are not concerned about the same issues. (There are some nice satire pieces by Kishon about how absurdly nice and polite people in Israel became towards each other during the six day war, and how it immediately degraded with each success story from the their military.)
"Before the War, America was an individualistic nation. Then came the Depression, the New Deal, and most of all the War. America won the war with a “completely centralized bureaucratic government structure” - and it was a huge success. And for a while, that worked: the generation forged in the war had a “cooperative “same boat” spirit” that “[made] the centralized corporate structures work.” But then it didn’t. Hall blames the hippies"
I never thought of it this way before -- and the above paragraph doesn't note it explicitly -- but one tends to think of the hippies' individualist ethos as left-wing and the older, Emersonian self-reliance ethos as right-wing, yet perhaps it's the same strain of American culture that gave us both.
One could argue the hippies had the creative, individualistic drive of a child, whereas Emerson was talking about the responsible-minded independence of an adult.
One thing I find disturbing about this 'fear of nuclear is just hysteria' argument is how its proponents always casually brush off the cost and the danger associated with cleaning up the few disasters that have occurred. Even allowing for the argument that the cleanups themselves have inflated costs because of excessive fears of radiation, we're talking about thousands of people who HAVE to be displaced, land that remains uninhabitable for decades, and cleanup costs in the tens of billions. And this is just the accidents that occurred at the heavily regulated plants that do exist! Saying Chernobyl killed 37 people is such a bizarre understatement of a) how serious it nearly was and b) just how extreme the actions taken were to prevent more deaths. As an energy source it also generates waste that remains dangerous to people for thousands of years, which has to be safely transported and stored and remains a huge security risk for the duration of its lifespan.
I'm prepared to accept arguments that the impact of climate change could be worse (probably true) or that fear of nuclear is somewhat overblown, but I always find it shocking when people are as dismissive of anti-nuclear viewpoints as they are of e.g. anti-vaxx.
What are the dangers of modern nuclear plants vs. the dangers of modern coal and gas plants? isn't that the question?
We also need to factor in the dangers related to getting the fuel to the plants: mining, drilling, fracking, transportation, be it by tanker, pipeline, truck or rail.
Most experts who have done the cost-benefit analysis have concluded the dangers of not having nuclear are much higher than the costs of having it, given current technologies in all areas. I believe that is true even_if you don't think climate change is a danger.
I agree with the cost-benefit analyses, and I am mildly pro-nuclear. But at the same time, I think they miss a really important point. There are expectations, and there are tail risks. The cost-benefit analysis is usually about expectations or typical outcomes. Here, nuclear energy is not so bad.
But in tail risks, they are much worse than coal and gas plants. What is the risk that a country like Czechia becomes uninhabitable because things go really, really, really catastrophic in in a coal mine or a gas plant? Pretty much zero.
What is the risk that the same happens because things go really, really, really catastrophic in a power plant? Small, but not zero.
You could argue that "uninhabitable" is a too strong word. But then replace it by some slightly less catastrophic scenario, and the argument still works.
To say that the plants that failed were "highly regulated" is odd. Chernobyl is a case study on the Soviet Union's spectacularly bad political incentive structure. The rest of the world at the time knew that those reactors were unsafe. Several soviet scientists themselves attempted to publish research saying as much and they were dismissed and discredited, who knows how many MORE scientists agreed with them (or would have) but who was willing to go against one of the darlings of Soviet scientific achievement, the RBMK reactor, so many of which had already been built? One of the frustrating things about the Chernobyl HBO documentary, which is extremely good and entertaining, is that a lot of people took away from it the lesson "nuclear reactors are unsafe!" when imo the proper lesson to take away was "holy shit look how dysfunctional the USSR was on the inside."
I would never claim that "Chernobyl killed only 37 people", but the more one learns about nuclear power, the less one is worried about it. I look back on history and see there was an anti-nuclear movement for 60 years or so, during most of which, most of their victories were victories for fossil fuels (as solar & wind only became cheap over the last 10 years). So I take all the anti-nuclear arguments of the last 60 years and compare them with the alternative they effectively supported: fossil fuels. Okay, so Chernobyl killed a few thousand? Fossil fuels killed a few million! Yet nuclear is the one they demonized?
The opposition vaguely made sense if you think "nuclear power makes it easier to build nuclear bombs, and bombs are horrible!" The bombs are horrible, yes, but why not just focus on opposing bombbuilding? Some nuclear plant designs aren't useful for bombmaking, so why not promote those? But no: 60ish years of opposition, and we end up with tons of coal plants and also huge stockpiles of nuclear weapons. Great job fellas.
My other comments above speak to many of your concerns. Let me know if you have questions.
1. I'm not talking about enriching uranium, I'm talking about the risk of dirty bombs being created with plain old radioactive waste. The consequences of a dirty bomb being used in a Boston Marathon-style attack are unthinkable. It's also an environmental hazard that needs to be stored somewhere it's safe from every kind of natural disaster imaginable, then kept safe from theft or corrupt/ideological sale. There is NO option of failure on either of these points. That waste is a security liability for as long as it's sufficiently radioactive.
2. Every time I hear this comparison, I see the existing state of fossil fuels being compared against the existing state of nuclear power. This isn't what we're arguing. It needs to be compared instead to a hypothetical much, much wider and less cautious rollout of nuclear power that would have happened in the absence of the cautious approach we took, over decades with industrial safety standards that saw the Bhopal disaster, the Sellafield disaster, the 2020 Beirut explosion, the Tianjin blasts, etc. etc. Also, we have to hope that there were no nuclear plants in Kuwait or Syria, to name two places where there was a complete security breakdown in the last few decades. Disasters happen and when they are nuclear, the cost is incredible.
Some people might *say* dirty bombs are "unthinkably bad", but I've never seen evidence that dirty bombs could realistically be "better" from a terrorists perspective than a conventional attack or a chemical weapon attack — especially if the terrorist wants to *actually* hurt people instead of mainly causing a fear-based response. A quick Google finds me an info sheet from the DHS — an organization that benefits not at all from minimizing any threat to America: https://www.dhs.gov/xlibrary/assets/prep_radiological_fact_sheet.pdf
> It is very difficult to design an RDD that would deliver radiation doses high enough to cause immediate health effects or fatalities in a large number of people. Therefore, experts generally agree that an RDD would most likely be used to:
• Contaminate facilities or places where people live and work, disrupting lives and livelihoods.
• Cause anxiety in those who think they are being, or have been, exposed.
"Cause anxiety" indeed: Bernard Cohen once sent out a survey to the Health Physics Society and Radiation Research Society. The survey asked experts to compare "the general public's fear of radiation with actual dangers of radiation". 91% of respondents agreed that the public's fear is greater than realistic, and 58% thought that the public's fear is "substantially" or "grossly" greater than realistic. Further, 82% of respondents thought that television coverage of the dangers of radiation "substantially exaggerated" or "grossly exaggerated" the danger.
If indeed fear is the main benefit from the terrorists' perspective, I don't see how hyping up the dangers of radiation is beneficial to the rest of us. Any time you hype something out of proportion to the actual danger, you open the door to other dangers you didn't hype (such as fly ash, which is empirically about 1000 times more dangerous.)
If we're talking about "existing state of fossil fuels" vs "a hypothetical much, much wider and less cautious rollout of nuclear power", does "less cautious" mean "a 1978 level of caution (back when nuclear plants were affordable), plus lessons learned from all nuclear accidents since then"? If so, I'm comfortable with that. The TMI and Fukushima accidents, for example — keeping in mind that these were both built in the 1970s — are avoidable without significant cost increases, and all western reactors already had safeguards against the kind of accident that happened at Chernobyl.
I should add, though, that assuming we stick with modern super-safety standards, it should still be practical to affordably replace fossil fuels with nuclear plants (especially if there were a carbon tax or clean energy subsidies). The most promising way to do this is via 4th-generation reactors, which mostly aren't "children" of traditional reactors, but more like back-to-the-drawing-board designs such as MSRs, which reduce costs by (1) eliminating hazards rather than managing them, and (2) avoiding the use of water for coolant, which is challenging from a containment perspective because it is kept at 150 atmospheres of pressure and over 300°C, and that's under normal rather than accident conditions. (the concept of Small Modular Reactors or SMRs, by the way, is a separate concept with separate cost benefits, though obviously SMRs can also take steps like eliminating hazards rather than managing them.)
The case about nuclear energy is one that I always think to myself, but how does it explain non USA (or non Western for that matter) nuclear prices? Since they don't have the same regulatory hell shouldn't Russian or Indian or Chinese nuclear electricity should be so cheap they'll just be selling the rest of the world electricity? I really want more research made on nuclear so we make most of our electricity out of it (and some architectures like MSRs using Thorium etc theoretically can both decrease the prices and increase safety), but also somebody needs to write an article why it's still expensive in other regulatory environments. One more caveat about Chernobyl, there's no way in hell there's so little casualties. Just the extra cancer deaths from the Black Sea coast of Turkey is way more than that. From the top of my head I'd expect in total worldwide somewhere around a few tens of thousands to maybe a hundred thousand dead?
Re: Flying cars well safety regulations are one thing when the vehicle can just stop in case there's something wrong and stay where it is, and another thing when it just falls off the sky. I want a flying car more than anybody but (maybe it's my lack of imagination) I cannot see a world when they're not falling off the skies like flies.
Is the public R&D spending somehow stop the private investors from investing in nanotech? How can the public be blamed for this?
Overall I think this book sounds like one where the ideological fervor of the author clouded his judgement so heavily it doesn't make much sense.
>but how does it explain non USA (or non Western for that matter) nuclear prices?
Well it is a complicated beast to wrangle, the nuclear power plant. I would imagine high prices outside the western world are high because it's a highly technologically complex thing to make. But China is building nuclear plants anyway, we'll see over the next few years how long it takes them to make and how expensive they are to run but even those datapoints I don't think will be that generalizable because China is kind of a special case in the world economy.
But within the western world you have France, and they have among the cheapest electricity in Europe. The difference between France and US nukes is that France picked basically one reactor design and built a bunch of them. By contrast, basically every plant in the US is a special, one-off snowflake. The US used its nuclear energy program as an input to its nuclear weapons program, they were essentially treated as science experiments to produce useful weapons-grade material, the more the better. They were not optimized for cost, they had a healthy chunk of the defense budget.
In general, I don't think prices in Europe suggest that nuclear energy is cheaper. Belgium has the forth-highest fraction of electricity produced by nuclear, and the second-highest electricity prices. (Numbers slightly outdated.)
You're right I overstated it; France's power is cheap for nuclear, average overall. But I still think the general argument holds: 70% of France's power is nuclear, the highest anywhere, offering an existence-proof that a mostly-nuclear grid can provide clean electricity at reasonable prices.
One just has to figure out what France did to achieve that and replicate it, and certainly good debate could be had on what the causal factors were, but I think there's a good argument that it was the use of a common and consistent design that led to plants being fast to build and cheap to construct and operate.
Yes, I agree. It proves that nuclear energy is competitive. And nuclear energy might have been yet cheaper in the counterfactual world where the US and the EU went even stronger for nuclear power.
It just does not really support the extreme position of the book that this would have blown us into a new age of cheap energy.
France is often touted as a success story for nuclear, and yet it is planning to gradually phase out nuclear plants over the next few decades (even while attempting to decarbonize). It certainly isn't building new standardized nuclear plants hand over fist. Nuclear plants in the US were cheap at one point, too. I think France has experienced the same trend of increasing nuclear construction costs (though not so extreme), and then more recently public opinion there has turned somewhat against nuclear because of Fukushima.
Fukushima is part of it. France is subject to EU regulation, and for a while there it was looking like the EU wasn't going to consider nuclear a "green" energy source, qualified for public funds, after what looked like lobbying from German LNG. Within the last year the EU commission decided that nuclear was a valid thing to invest in in order to meet climate goals, and so I expect France to switch tracks on their nuclear divestment plans.
> Since they don't have the same regulatory hell shouldn't Russian or Indian or Chinese nuclear electricity should be so cheap they'll just be selling the rest of the world electricity?
Good question. I've been wondering if there are international treaties (regarding fuel, reprocessing, nonproliferation and safety) that all-but-guarantee that nuclear power as traditionally built will be expensive for any country that does it. Another thing is that maybe the overblown fear of nuclear power is a global phenomenon (I noticed it a little in the Philippines when I was there - maybe not the best example, as the influence of America is pretty obvious there).
The runner-up review says of why we don't have flying cars,
"Is it because of air traffic control being difficult? No, not really. Because the skies are vast and there's plenty of room up there, what with three dimensions to play with and no "road" restrictions."
I'd really like the explication of this, because I find it hard to believe. For one thing, the flying cars wouldn't be scattered evenly throughout the air. They'd all be congregating towards the places where people want to go. 2-dimensional traffic jams are bad enough as it is: 3-dimensional ones with no road markers would be far, far worse, and more dangerous, because there effectively wouldn't be such a thing as a minor collision.
Ok, but imagine we all did have flying cars. Would we live so close together? Perhaps there would be urban areas without flying cars, but everyone in the "suburbs" would have flying cars and those suburbs would be a long ways away. For instance, imagine all the Southern California suburbs existing far into the desert. You'd have all new sorts of new cities pop up all over the country (maybe even Pop-Up Cities), because someone could start a new entertainment/business district wherever, as long as there was plenty of airplane parking.
Couple the idea of flying cars with remote working. People would still want to get together, but the geographic considerations for everything completely change.
Two-dimensional ground autos enabled us to live not so close together. Our homes spread out enormously, with Consequences. And it didn't prevent traffic jams.
Those consequences were a net positive, right? Cities like Houston designed entirely with autos in mind don't have bad traffic jams. The bad traffic jams are in cities not designed for autos or places with geographical constraints.
Anyway, imagine the entire American West, currently almost unpopulated, as the suburbs of America. In the West, everyone has a private plane and an enormous plot of private property. Meanwhile all these entertainment/meeting hubs also pop up all over the desert. There's little law and what law there is made mostly by ray-gun.
Those who want to live in cities can live their separate lives in cities. They don't have to own planes or cars. They can bicycle everywhere and take trains. They don't even have to carry ray-guns if they don't want.
Wouldn't that be great? Everyone could live something closer to their ideal than now.
"For instance, imagine all the Southern California suburbs existing far into the desert"
And now I'm imagining the first American Water Wars. It's bad enough (seemingly) trying to manage the demand for water in California, given that agriculture guzzles a lot and agriculture is important for feeding the rest of the nation, as it stands. Houses sprawling in the desert all wanting the same kind of water usage they do today, competing with the existing demands?
Only about 10% of California water usage is for houses[1]. And there are quite a lot of Californians (like me) who think the problem is less one of exploding demand than of government that hasn't proposed or constructed a single significant water infrastructure project -- or even properly maintained those that already exist[2] -- since the 1960s, over a time period when the population of California doubled. The state loves the tax revenue from growing population and economy, but they want to spend it on things that are much more fun than water projects, with their inevitable painful political struggles, long time horizons, payoff in the far distant future.
There's tons of water in California, but it's all in the Sierras, of which only a small amount is current diverted to human needs (e.g. the Hetch Hetchy project provides 80% of the water for the entire Bay Area from a watershed that spans at most about 1200 km^2, something like 2% of the Sierra Nevada). But that water needs to be collected and transported to where it's needed, since the natural rivers don't do it, and that's where water projects come in. There were many in the early decades of the 20th century, and they were big and bold and economical, and there's a good argument they did a lot to enable the explosion of the California economy and population post-WW2. But since then...zip.
One can make the argument that you should not build cities as big as LA or San Diego in the desert, where there just isn't enough local water to sustain them, but if you *do* accept the proposition that it's OK to do that and ship the water in from the mountains, then you kind of have to keep up with the times, as the population and economy grow. This the California state government has not done.
4. At $64 per day, the one billion people at Level 4 own a car.
Wow, $64 per day. That is within single digits of my income (US SS). Except I can't.afford a car. Or perhaps I should put it this way: If I had a car I would have to stop helping many of my Haitian friends. Since I exist quite well without a car, that is a no-brainer.
There were only 43 deaths at Chernobyl (assuming that estimate is right) because they undertook an expensive and difficult mitigation project, and more or less permanently evacuated an area 2600 square kilometers in size. That's from a single reactor disaster, with a plant design that was being run with cost first and foremost in mind (the main virtue of the RBMK reactor was that it was cheap to build and operate - also why there was no containment structure at the time of the accident).
That's why nuclear power is heavily regulated - the potential downside is enormous, and very much not something to blow off casually as exaggeration. The Soviets actually got lucky in that the disaster happened at Chernobyl rather than, say, St Petersburg (which also had the same type of reactor).
The stuff about the private sector and nuclear power rings false to me as well. The only time nuclear power ever got cheaper either in the US or abroad was when it was being built on a huge scale, by large, quasi-monopolistic utilities with implicit state backing and support (or explicit). Even in the US, nuclear power got a huge boost from having its liability limited in the Price-Anderson Nuclear Industries Indemnity Act in the 1950s - if they'd had to actually accept the liability for the plants in full, they likely would not have been built at all.
Incidentally, the French are running into cost overruns on new nuclear plants too.
I'm surprised nobody has mentioned the noise issue with flying cars. Choppers and aircraft are not quiet.
The first part is not even close to being true. The only people who died of radiation sickness at Chernobyl were extremely close to the damaged core itself, i.e. operators of the plant in the immediate aftermath of the explosion, and the fire crew that came to put out the fire on the roof of No. 4 reactor (which was covered with core debris).
The evacuation of Pripyat and vicinity 36 hours later was precautionary, in case of further explosions, and because the amount of radiation released was then unknown, but I know of no evidence that radiation levels that far away from the plant were ever high enough to produce radiation sickness. Had it been, there would surely have been hundreds of deaths from radiation acquired before the evacuation. Mind you, something like 3,000 people refused to evacuate at all, and there was no massive radiation sickness among them.
Chernobyl was bad enough without exaggeration, which only serves to deepen cynicism about nuclear criticism.
An explanation based on the establishment of the department of energy only really works if you ignore the world outside the USA. Arguably similar pressures exist, but there are many places, e.g. France, where nuclear makes up a much larger share of their energy supply. And while they've gained in efficiency due to the benefits of mass production and expertise, there hasn't been the major shifts that would be necessary for this theory to pan out.
A more parsimonious explanation would be that, as in many things, each additional marginal unit of energy is more difficult to acquire. And the returns on additional energy usage in terms of increased quality of life are diminishing. (A flying car would be cool but didn't make a difference to your life in the same scale as having shoes).
I would guess the returns on marginal units of energy don't diminish quickly. Energy is an input to the cost of almost everything. Every good you buy has energy costs in its production and its transportation to take it to market. Services have energy costs as well. Housing has energy costs. Computers have energy costs.
A reduction in energy costs is an increase in wealth for almost everyone.
Putting aside the specific merits of flying cars or nuclear power, I think the author is onto something about the change in our society.
Specifically, risk tolerance. Somewhere along the way, we changed from a society that favored growth and change, to one that favored safety and stability. All sorts of things in the 50s and 60s would be considered horribly unsafe nowadays. Nuclear power plants and the Saturn V rocket are big, obvious examples. But there's also smaller examples like car safety measures, drinking, smoking, and letting kids roam freely outside. Or banning lawn darts, for a really trivial example.
I'm not completely against safety measures- they really do save a lot of lives, when you multiply them over a huge population. But there needs to be a countermeasure too- some things are worth risking a few lives over, and I think transformational changes in technology are one of them.
I wonder if the difference is the generational memory of war? The decision makers in the 50s and 60s were people who remembered ww2, and many had fought in it. Compared to that, a fatality rate of 1/100 might seem acceptable, or even trivial. But everyone since then is either a civilian, or used to extremely limited wars where relatively few people die. Now we would think of a 1/100,000 death rate as unacceptable. That's my crank theory, anyway.
And the lower risk tolerance has potentially massive (and unaccounted-for) hidden costs. As life gets really comfy for all the new homeowners, it's really easy to lower their appetite for risk. After all, they don't need a ton of new growth, they are quite content with a much higher living standard than they had growing up. This seems like an almost inevitable trap for a wealthy society to run into.
Yes, I like your point about risk tolerance. Not just memories of war, but memories of much harder times. If you grew up getting kicked in the head by mules, or just having them step on your feet, and your dad brought gas from the pharmacy in a bucket for his flivver while smoking his pipe- flying cars might seem no greater risk. All the men in Norfolk jackets who died to make heavier-than-air flight possible would say flying cars sounded safe from their end too.
Radiation risk was taken a lot less seriously before the brilliant martinet Rickover was in a place of power over nuclear power plants. In the 1930's my granddad's brother used to scare my mom with his giant grey Frankenstein hands- he was a doctor experimenting with radium. But how many people think of the risk of nuclear power or flying cars without being influenced by moral panics whipped up by lawyers ?
Yeah, it's probably not just the war, but the general lifestyle. *Everything* back then was more dangerous, especially jobs. That would fit the general pattern that we see, not just in the US but in every country. As their economy develops, everything gets safer, and then people become unwilling to take the risks that are necessary to push things futher.
Indeed. One thing in the review made me smile : the idea that small planes were not more dangerous that motorbikes would somehow show that regulation of planes is not normal. I think it's the other way around: motorbikes in current society are an anachronism, living on borrowed times: they have not yet been regulated to death, because they were already well established before modern safety-focused world, they trigger nostalgia and romanticism which protect them a little bit from being anathema to women, and dealing with angry bikers is not worth the trouble. But they are killed, one segment at a time: for example dirtbikes in europe, the easier target, have moved from a not so small hobby to specialized extreme sports (the Instagram-hosted refuge of something that was much more common at one time)
> All sorts of things in the...60s...horribly unsafe...nuclear power plants...are big, obvious examples.
1960s reactors weren't all that great. But successfully stopping nuclear power was a boon to the coal industry. If all the coal plants were replaced with nuclear plants, and if every one of those nuclear plants had Fukushima-style meltdowns, it would probably cause far fewer deaths than coal does, even assuming that people around the reactors do not relocate: https://www.reddit.com/r/nuclear/comments/jtm6hm/how_bad_is_meltdown_world/
It mentions in passing "stagnant wages" and I can't resist swatting that one down every time I hear it.
The stagnation of wages in the US is a frequently repeated myth with a kernel of truth. Real median *personal* income has risen a lot since 1981. *household* income growth has been smaller because the number of earners per household declined due to huge increases in fatherlessness. Which is a big problem but not the one most people have in mind when they're complaining about "stagnant wages".
The average commute in the USA is 27.6 minutes. And you can do that while listening to a podcast, or great music in a climate-controlled environment. I don't see much economic benefit to shaving time off a moderate commute, as it is unlikely to increase productivity for the average Joe. Small planes are noisy, and it would be obnoxious if there were runways everywhere and constant air traffic. And of course, any kind of "weather" would ground all the small planes. And small aircraft are dangerous and we are increasingly risk-averse. So perhaps physics and cost-benefit calculations are the reason we don't have flying cars yet.
As a proponent of nuclear energy, I really want to tell this guy to get off my side.
Also: TSCM is working on 2 nm process semiconductor manufacturing. If that isn't "development of nanotech", I'm not sure what is.
Finally: An average car (according to Google) is 1500 kg. If I assume very modest flight altitudes of 100 m (which to my understanding is unrealistically low?), that's 1.5 sticks of dynamite worth of gravitational potential energy *before considering the car's speed*. If flying cars become widely available where I live, I'm moving. For the same reason I'd move if they started selling dynamite to use on the 4th of July.
On nuclear I agree, he's a little too gung ho, makes the rest of us look bad.
For nanotech, the originally idea of nanotech as an industry was "atomically precise manufacturing of arbitrary things." Semiconductors are an extremely specific thing. The idea of nanotech was to treat it like a Manhattan Project, or at least like fusion: put a lot of money and effort into this long moonshot, even if it takes many decades. The idea was that, if we had good nanotech, making 2nm semiconductors is just a (highly nontrivial) design problem of figuring out where you want the atoms to go, but figuring out how to get those atoms in that arrangement would be a paved path.
I'm not sure if it was ever a coherent goal, it certainly would be more complicated than "use one factory to build a smaller factory, and use that one to build a smaller factory..."
Isn't it physically impossible to have a small nuclear reactor? To sustain a nuclear reaction, you need an amount of fissile material that's a good fraction of the critical mass. The critical mass of uranium-235 is about 52 kg; for plutonium-239, about 10 kg. The Kilopower project, intended to produce nuclear reactors for space travel, is probably as small as you can get a nuclear reactor, and their 10 KW reactor still weighs 1500 kg.
Of course, no government in their right mind is going to let private citizens own enough U-235 or Pu-239 for a nuclear bomb, and it takes the resources of a country to produce that a critical mass of those isotopes anyway. A heavy water reactor that uses natural uranium has to use much more uranium. I'm not sure how small you can get a heavy water reactor, but the KANUPP nuclear power plant is one of the smallest in the world, and it still has 30 tons of uranium. Of course, in addition to the uranium, you also need the moderator, control rods, some way to turn heat into usable energy, and shielding (if any humans are going to be around). I'm not convinced it's physically possible to create a nuclear powered car, or even a nuclear powered airliner, with natural uranium.
You need to be at or above criticality, although reactors can stay in the zone between criticality and prompt criticality that weapons must be above. So yeah with the shielding, cooling, and power generation accoutrement I think it's pretty impossible to build a reactor for something as small as a car. Maybe a big airplane would be good, though: an A380 weighs 360 tonnes empty, so adding 5-10 tonnes of reactor and subtracting 200 tonnes of fuel would probably be a win.
Nuclear-powered airplanes were studied during the Cold War as a way to allow bombers to remain airborne indefinitely. They didn't get beyond some very early tests before the development of ICBMs made them obsolete, but they're probably workable. However, I can't find much information on the reactor design they used, so I don't know how much harder it would be to do it with non-enriched uranium.
How do you do it at all with natural uranium? Did you mean not using highly-enriched U the way the Navy does? That does seem dubious. So it has to go into a pretty stout containment vessel, to survive the inevitable controlled flights into terrain, and we need some special licensing requirements around the fuel handling so none of it gets diverted to North Korea ha ha. Maybe we should have nuclear-powered dirigibles instead.
"Early tests" isn't quite a fair description. They built two working nuclear jet engines, which are sitting in a parking lot near Idaho Falls (if you're ever nearby they're worth a visit!). They certainly didn't use natural uranium, but I don't think it was weapons-grade either. The main problem wasn't the weight of the reactor, it was the neutron shielding you would need to have for the crew. Also, the risk of fuel degradation that could cause it to shed fission products in the exhaust.
Another partial explanation for the less adventurous, risk adverse climate since the 60-70: the aging of the population, together with a feminization of political and economical power. Even if not so strong at decision making level (which was always old - ish and remained largely masculine), political support and typical consumers changed for older and more feminine, so technology directly aimed at consumers changed targets, and regulations followed a much more risk averse population. Even for IT, this can be seen a little, with transition from pc and gaming to smartphones, e-commerce and social media, with supporting server infrastructure hidden from the public.
The extreme irony of someone waving their flag about how superlatively anti-racist they are, then using - for triple irony points during the month dedicated to Pride - a homophobic insult? I suppose I could put something in here about "pot" and "kettle" or do a dissection on why licking dicks is a shameful thing (if we're all liberated now and are no longer homophobic).
But instead I think I would much prefer a higher standard of invective, or at least some imagination in the Billingsgate - "you dick licker" is something I'd expect a ten year old today to come out with in the playground, and promptly be hauled in to sit through a Special Class on not using naughty words, Tommy.
"Greetings, sports fans, and welcome to the qualifying rounds of the Insult, Invective and Rhodomontade Championship! Tonight will see only four contestants moving forward to the semis, so stiff competition here!
Hi, I'm your host, Rob Ennybodie, and this is my co-host for tonight, Bob S. Ure-Unkell. Bob, what do you think of our first competitor?"
"Well Rob, Jack Wilson doesn't have much of a reputation built up here yet, so there's all to play for. We'll have a better idea once we see who he's drawn against - and it's Sailer! Steve Sailer".
"A controversial pick there, Bob. Soft target or water off a duck's back?"
"It's anyone's call, Rob. It all depends on Wilson's opening move".
"And he's started with a third-party appeal about racism".
"A solid opener, if unimaginative. The conventional choice, you might say."
"Maybe Wilson is holding back for the semis, Bob?"
"He has to get there first, Rob".
"Let's see what he brings for his follow-up, Bob."
"Oh dear, and it's racism again. Mere repetition is not going to cut it at this level, Rob".
"But a strong close can still see him pull through! And he ends with - "dick licker". Well, a disappointing finish there, Bob".
"Agreed, Rob. Juvenile abuse may work on the playground, but it's a different game at this level. Even the classic "gobshite", as in the friendly between Carragher and Neville, https://www.football365.com/news/neville-gobshte-carragher-car-park-debate-liverpool-manchester-united, would have brought this up a badly-needed notch. To have any hope of winning, then turning your back on with a sniff of disdain, the grand trophy for the Championship of Insult, Invective and Rhodomontade, you need imagination, you need flow. Inventiveness and novelty are the hallmarks of a true champion, but even then a solid, workmanlike, steady, conventional performance might have been enough - had it not been for the weak ending.
"A pity, Bob; we were all hoping for an entertaining bout. Well that's it for tonight, fans, I've been Rob Ennybodie and this has been Bob S. Ure-Unkell for the I, I & R League! See you next week!"
If one wants a "high standard" of invective against Steve Sailer you can simply point out that Steve Sailer made false pedophilia claims (he never marshalled a single shred of evidence) against one of his ideological opponents.
My dad's first job as an aeronautical engineer in the late 1930s was designing a small part for a flying car company. It was going to a tricycle road vehicle/aircraft.
He said the project was coming along pretty well. But then the FAA decided that having only one wheel in the front was too risky: while driving, you'd run your front wheel into a curb and put a hairline crack in the front wheel/landing gear pylon and then the next time you are landing it, the lone front landing gear snaps off and you go nose first into the runway. But having two wheels/landing gears in the front would be too heavy to get off the ground. So the company sold their work in progress to the Japanese, who maybe were planning to use flying cars to invade Los Angeles, land them on the Pasadena Freeway. So my dad then got a job at Lockheed from the 1930s into the 1980s
Why do you allow such racist comments on your blog? My theory is that you aren't as big a racist as your commenters but you have made yourself beholden to racists so you can't afford to reject them.
I'm pretty sympathetic to the central thesis of this book. However, I must point out its glaring scientific weaknesses. First, it treats cold fusion as respectable science and insinuates that the scientific establishment covered it up. Second, it says that "material science [...] trashed the reputation of actual nanotech." In reality, nanoscale materials behave very differently due to quantum mechanics. So the dream of geometrically decreasing "factory" sizes based on everyday technology is wrong.
Re science funding, a very well-known microbiologist once told me his theory that the (rough) doubling of the NIH budget in the 90s had had a strong negative effect on the pace and quality of biological research.
His reason was simple: a ton of mediocre scientists were now able to get their careers funded, and to make matters worse many of them ended up in high-up administrative positions deciding where to direct future funding, and their decisions were often poor.
The reviewer summarizes the author’s basic claim as “The Great Stagnation was caused by energy usage flatlining, which was caused by our failure to switch to nuclear energy, which was caused by excessive regulation, which was caused by ‘green fundamentalism.’”
I find plausible the claim that nuclear hasn’t taken off because of regulatory burden, but contra some other commenters, my intuition is that the burden of regulation in nuclear energy should be comparable across countries, given that nuclear energy is already highly regulated at the international level.
On the other hand, I’m skeptical that the reason energy use has “flatlined” in some countries is because of the lack of expansion in nuclear power generation. Here is a chart for comparison of energy consumption per year for various countries:
If you try China, India, France, Russia, and the US, each of which has a home-grown nuclear power industry, you’ll find that energy consumption in the first two (especially the first) is increasing rapidly, while in the last three, it has flatlined. So I’m guessing that the relevant variables are rate of energy use and rate of economic growth, and while regulation must affect both, regulation of nuclear energy in particular seems unlikely to be so relevant.
While we are on the topic of energy, I recently read Energy and Human Ambition, which does some easy and fun Fermi calculations such as how long, at current growth rate in energy use, would it take us 1) to eat the Sun, the Milky Way, the visible Universe, and 2) to boil off the oceans:
In short, not long on a civilizational time scale, from which we can conclude that energy use will probably either flat line or collapse in the next few hundred years (a sub-exponential growth rate would buy a lot of time, but isn’t observed).
I read the article yesterday (though skipping most of the exercises) and was quite impressed how easily our perception of the facts of live can be unmasked as self deception by a couple simple calculations. I always thought that I had a pretty realistic view of the world (having an engineering background) and was very optimistic that humanity could overcome future and present challenges with science and rational thinking. Since yesterday I have serious doubts, exacerbated by some comments here…
Flying cars? Sounds to me like dreams of junkies in an exstatic high, while the average American is already consuming resources more than 6 times than sustainable
"I’m more convinced than ever that not embracing nuclear power was one of humanity’s worst mistakes (partially because I’m more afraid of climate change than Hall is)"
It occurs to me the failure mode of an overregulated nuclear energy is a reduction in the rate of technological progress. The failure mode underregulated nuclear energy is the backstory of On the Beach. If a country can build a lot of nuclear reactors, it can build a lot of nuclear bombs. If every country can do that, then international politics will get very interesting. At least for a moment.
1) It will be harder to stop people building nuclear bombs when plutonium, the essential raw material for making them, can be produced in any country on earth. Making it easier to make things will always make it harder to stop them from being made.
2) The failure mode of anti-nuclear-war regulation is always going to be nuclear war
I've heard multiple experts say you can't build plutonium bombs contaminated substantially with Pu-240, so if you want bombs the fuel must be removed and processed within 30 days. This is not normal in civilian plants and reactors can be designed to make such rapid cycling difficult or impractical.
Forcing people to build their nuclear power plants that way would be a case of regulating the nuclear power industry. And if you get the regulations wrong then maybe they will be able to produce weapons grade material after all. As I said, the failure of anti-nuclear-war regulation is likely to be dramatic.
Like most people, I'm not opposed to regulations in general, just regulations that unnecessarily increase costs. For the most part, affordable 1970s power plants were already unsuitable for bomb making, whether due to reasonable regulations or simply because they weren't interested in making bombs - see http://www.phyast.pitt.edu/~blc/book/chapter13.html
Generally a power reactor benefits from high burnup, opposite the low burnup desirable for bombmaking. In addition to his natural reason for reactors not to produce weapons-grade Pu-239, designers are aware that nuclear proliferation is bad, and therefore may make design decisions that make their reactors as unattractive as possible for bombmaking. Thorcon, for instance, talks about this explicitly as a reason to mix thorium and uranium together in molten salt. Unfortunately they don't really explain *why* plutonium would be hard to separate from thorium - I don't get it, any chemists want to take a guess? - but the point is that designers are actually disincentivized to facilitate proliferation, not just for some regulatory reason but because public opposition has historically been a serious risk to the nuclear business, and if they find something they can use to affordably to reduce that risk, they will do it.
Why would people designing nuclear reactors care about the opinion of members of the public? Members of the public are not their launch customers. And historically, the people prepared to pay top dollar for nuclear reactors have wanted them precisely because they need them to make nuclear weapons.
Regulations exist, from a certain point of view, precisely to stop people from buying what they want.
I'm a big fan of nuclear power and a big fan of flying cars. To the point where I basically designed my current (well, pre-COVID) lifestyle around the latter. But these are two great tastes that don't taste great together, and I'm wondering how the author makes the connection.
Energy is relevant to flying cars, but it's the wrong kind of energy. The driving requirement there is having to fit on the order of a megawatt of peak power into something the size of a mid-sized automobile (for VTOL performance out of a practical urban "parking space") with no direct connection to the power grid. There is no plausible nuclear option for that. Even assuming too-cheap-to-meter electric power at the wall outlet, your flying car doesn't get off the ground without some non-electric, non-nuclear power supply, or batteries better than anything that were plausibly available in the era the author seems to be talking about.
And there are other critical problems that aren't at all energy-related. Pilot training and/or automation, as has been discussed here already. Plus, disc loading and associated scaling issues mean that if you shrink a helicopter into something that can take off from a driveway, it will be dangerously (literally dangerously) loud, and also dangerous in that the downdraft will e.g. kick up stray beer bottles and accelerate them to potentially lethal velocity.
I like "flying cars", but it's not realistically possible to have them deliver people directly from their homes to their workplaces and favorite commercial districts. Nor is it practical for them to double as groundcars. This is fundamentally a last-mile problem. I can and do fly 95% of the distance from my home to my office, and we can *maybe* make it so most people can do that. Now find an answer for the last 5%, and if you insist that it be "we have to fly!", you're going to fail. Likewise if you think "because nuclear!" or "because nanotech!" is going to solve the problem.
For me, a folding bicycle in the back of the airplane works nicely. And quite possibly the reason we don't have "flying cars" is that the golden age of general aviation, when airplanes were affordable and airports were everywhere, corresponded with a period when bicycles and public transportation were both considered horribly, unacceptably lower-class. Now we have Uber, which might provide an answer but we've almost priced airplanes out of reach of the working class (don't even ask about helicopters) and we've closed most of the airports.
Nuclear power is a completely different discussion.
Can you say a bit more about "I designed my lifestyle around the latter" and "I can and do fly 95% of the distance from my home to my office"? Do you mean that you actually have a flying car?
I have an airplane (Grumman AA-5B Tiger) with a folding bicycle (Montague 20" frame) that fits in the baggage area. In Southern California of the 1930s through 1950s, that would very nearly have been a flying car - there were enough small airports, even in urban areas, that I could have flown to within easy bicycling distance of most anyplace I wanted to go. Most of those airports are now closed, but there's one less than three miles from my office. Enough that I can live far enough out that houses are affordable, and still have a reasonable commute.
Many people make the argument that government centralization of science is a mistake. The problem isnt that money is bad, it's the way the money was allocated. Scientific Freedom by Donald W Braben makes the same case but he has empirics to prove it. He got a company to agree funding scientific research in a way that basically gave academics free reign over the money (no grant process etc), complete freedom. Of course the academics were hand picked by Donald. They managed to produce 1 noble prize winning research and 1 company (>100m net worth). This was out of around 20 I think, which is pretty better than grant funded research. The problem with academic grant system is basically : 1) makes professors spend massive amount of time writing grants 2) very conservative with regards to projects it will fund (unless the prof is very imminent)
3) comes with strings attached that slows down the entire process
Does the author explain why flying cars didn't happen outside the US? My guess at a steel man would be that everywhere else is even more regulated (or, if not poor).
A very quick and rough estimate: up to 2020 there are about 287 million road vehicles in the US (this includes trucks and everything). There are 212,335 aircraft (this includes commercial airlines and Joe with his cropduster).
You tell me how the US scales up from 200,000 to 200 million craft flying around.
"Then there should have been nuclear fission and nanotech, letting you fit a lifetime's worth of energy in your pocket."
and also
"Nanotech, Hall says, is to nuclear energy as the steam engine was to coal - the technology that will unlock the potential of a new energy source."
This statement causes me to update massively in the direction "The book author does not know what they are talking about".
While I am not a nuclear engineer, a central concept in nuclear reactor design seems to be criticality, which is the point where one neutron will, on average, cause exactly as many fission reactions as needed for another neutron to be released. For different fissile isotopes, you would need different amounts to sustain a chain reaction, generally upwards of 5kg. You can use neutron deflectors and the like to lower it a bit, but in the end, that is a minimal scale for a criticality-based nuclear reactor.
You might or might not be able to use nanotech to scale down a cars internal combustion engine to fit into a hair, but you will definitely not be able to do the same for a submarines nuclear reactor, or even fit the reactor into your pocket.
(On the bright side, this is also why scaling down the mass of Little Boy by six orders of magnitude will not yield explosive small arms ammo with a blast of 15kg of TNT.)
I have tried -- and failed -- to parse the first quote in a more charitable way, e.g. that nanotech would provide the batteries to store the nuclear energy, but that would just mean that nanotech would provide energy densities similar to nuclear, which is hardly an easier sell.
For me no amount of Gell-Man Amnesia can overcome the scepticism about a book with its core argument based on "nuclear power + nanotech = reactor in your pocket".
OK, now I'm sure about how I feel. There's just too many entries for me to judge and it seems almost unfair to the entrants at this point. If you do this again next year (and I hope you do!) I would urge you not to go above 5-7 finalists.
I agree, I feel like I've already forgotten many and will be unsure how to vote due to how long this has gone and how many there were. As a contest to vote in, not great.
In terms of content, though, this is great! I'm loving these book reviews and there are tons of them.
On the other hand, even if it makes the voting less reliable, I bet for at least some authors getting the review posted on ACX and the readership that comes along with that is a prize in itself. Personally that would be worth at least $1k to me. And as a reader, I'm glad for the larger number of reviews getting posted - while in theory I could go to the google doc and read all the reviews, or a curated 'top 20' list from Scott, in practice I'm much more likely to read a review if it's posted to the blog.
Fair point about the exposure for the authors. I wouldn't mind lots of "special mentions" being published outside of the contest but I feel this is unmanageable for me. Maybe just do around 5 or so posts for the main competition and publish a select number of worthy entries that didn't quite make it once a week for a while after the voting is done?
That's only a problem if you have to deliver all your ratings at the same time. They should instead be rated independently on a 4-to-10 scale, so you don't have to remember how you voted on the last one to vote on this one. (why 4 to 10? because come on, they're all finalists, they can't plausibly be below 4, now, can they? now, your preferences/feelings might not be constant in time, so that you might well feel that the same review is a 5/10 in May but 7/10 in June, but given enough voters, such inconsistencies ought to average out to zero)
The phrase "green fundamentalism" really raised an eyebrow for me here - it's very hard to seriously claim that the US regulatory environment is even adequately protecting the environment, let alone overdoing it. Americans by and large live in a very toxic environment, and the lack of progress on climate change is an obvious red flag against this argument.
"Climate change will only cost a few percent of GDP" is also a HUGE handwave of trillions of dollars per year in additional *permanent costs* added to the US budget alone. See also: human suffering, likely abandonment of some affected populations, and extinction risk. That sort of claim makes me doubt his intellectual honesty overall.
The estimated annual cost of unmitigated climate change, accounting for hurricane damage, real estate losses, energy sector costs, and water costs. https://www.nrdc.org/sites/default/files/cost.pdf
Can you choose one of those things, perhaps real estate losses, and prove that there will be a *permanent* trillions of dollars per year cost?
I'm not going to read a 62-page report right now, but from skimming, I didn't see anything about the benefits that climate change would bring the U.S. I think these could actually outweigh the costs, especially if people stop being stupid and move away from the areas of highest hurricane threat.
"Tropical cyclone intensities globally will likely increase on average (by 1 to 10% according to model projections for a 2 degree Celsius global warming)"
"The global proportion of tropical cyclones that reach very intense (Category 4 and 5) levels will likely increase due to anthropogenic warming over the 21st century. There is less confidence in future projections of the global number of Category 4 and 5 storms, since most modeling studies project a decrease (or little change) in the global frequency of all tropical cyclones combined."
To me this sounds a lot like "We will get more [disastrous] hurricanes".
One of the most interesting points I think is why no country has done better at keeping alive a good regulatory environment and encouraging innovation in atoms. Was the US an outlier before, and does the trajectory of the US therefore have an outsized global role? I also recall some Robin Hanson post about elite consensus across countries being surprisingly robust (regardless of whether it is correct), and maybe this effect has grown with globalisation and the internet, and means that policy between countries is much more correlated than we might think.
> we can and have built them ever since the 1930s. They got interrupted by the Great Depression (people were too poor to buy private airplanes), then WWII (airplanes were directed towards the war effort, not the market), then regulation mostly killed the private aviation industry. But technical feasibility was never the problem.
It's because they kept making them again and again, and it was always a crappy solution to a non-existing problem.
Road vehicles and flying machines have different tradeoffs (weight is one). A flying car sounds good abstractly, but is the worst of both worlds.
The 20th century was the Pre-Cambrian explosion of technology, during which a single generation went from seeing people ride horses, to landing on the moon. Who wouldn’t be optimistic after that?
I can’t find the authors exact age, but he’s an older gentleman and I can assumed he grew up at a time when everything seemed possible. “Technology Progress” can become an ideology, and I think nothing typifies this better than the Jetsons.
I also find that sometime STEM types are not great at inspecting their own ideological biases, and so the author might just not be willing to give up on the future he imagined.
Could somebody wake me up on wether nanotech (as defined in the book) is pseudoscientific or has real potential. The so-called Drexler-Smalley debate has made me think the former.
Nanotech seems to be mostly propagated by computer scientists and this is something Smalley reproaches Drexler for. I'd be especially interested in the opinions of trained chemists or physicists among us.
My hot take is that we SHOULD have regulated cars out of existence.
Driving is an extraordinarily dangerous activity, but we've just come to accept that tens of thousands dying on the road every year is fine.
Driving has also completely ruined the urban fabric of most American cities. People often say they strongly prefer to live in walkable neighborhoods, but the necessities of dealing with urban traffic flows makes walkability very difficult to achieve. We've gone from a highly social society where people live in tight-knit communities to a fully atomized society where people routinely never meet their neighbors, and though cars aren't the whole story, they're definitely part of it.
Cars exist as they do today BECAUSE governments stepped in to change the rules in favor of cars. Jaywalking was invented as a crime because car manufacturers wanted to crank up speed limits and not have to worry about hitting people all the time. So all of a sudden, the streets themselves were taken away from pedestrians and given to cars, with tiny strips along intersections allowed for occasional pedestrian use. Miles and miles of highway were built to make cars a convenient long-distance transportation option.
So yeah, flying cars are probably a bad idea, and I think on the whole we are better off for not having them.
Knowing your neighbors is generally a good thing that I wish I had more of, but I've also had neighbors that I really wish I had never known, because all their problems became my problems.
How would a city with no urban motor vehicle traffic even work? You'd have to carve out tons of exceptions, for commercial deliveries, construction, moving (imagine trying to get your couch from one apartment to another across town, if you can't throw it on a truck!). Or else you'd need some completely different last-mile transportation infrastructure suitable for cargo. Before cars, I think this stuff was basically done with wagons, which are just cars that suck.
I don't really mean that we should have had a total ban on all motorized vehicles, in the same way that today we don't have a total ban onaircraft. I think various sorts of trucks and vehicles would still be useful and necessary. But we should have regulated them to fit into the existing urban fabric instead of rebuilding the urban fabric around them. That would mean stuff like: all streets would be mixed use and vehicle drivers would be responsible for not hitting anyone, surface parking would not be heavily subsidized (in other words, parking would be expensive most places), and on a general cultural level, motor vehicles would be for industrial and commercial use, not something most people own and drive around in every day (unless their job was to make deliveries or something).
I could see that being nice. It's striking that there aren't really any examples of this in the US, when it's something that could absolutely be done at the local level. I think there are some neighborhoods in Europe that are sort of like this, mostly in downtown areas of old cities where the roads weren't designed for cars.
Probably the closest in the US that I'm aware of would be Mackinac Island, where they really have just banned all motor vehicles. But being a small island in the middle of the Great Lakes, it's hard to get a motor vehicle there in the first place, and it's not really scalable on a full-society level.
As a person living living in Europe I can only shake my head increduously:
You in the USA already have the most suicidal mass killings world wide, and you dream of putting another instrument/weapon in each garage allowing a frustrated pupil or employee to steer the vehicle right into the classroom or office?
The gratuitous jingoistic nationalism is not appreciated. Yes, it's possible to use motor vehicles to kill crowds of innocent people. That may be worth considering. But doesn't seem to be a thing Americans are any more inclined than Europeans to do, so why bring that part up if you're not trying to stir up trouble?
As far as I can tell nobody is talking about the plasma fusion reactors that are planned (e.g. ITER) which would have no waste. I understand that fission reactors are already developed, but if the idea is to not repeat a past mistake, why not rebrand on a type of reactor that doesn't carry all that baggage.
"Nuclear Fusion Is Always 30 Years Away" as the joke goes. ITER was started in 1988 and they hope to have deuterium-tritium fusion by 2035. As ITER's info page says, "ITER will not capture the power it produces as electricity, but as the first of all fusion experiments in history to produce net energy... it will prepare the way for the machine that can."
In contrast, the companies designing new (fission-based) Molten Salt Reactors over the last 5-10 years hope to start their first reactors by 2030, and they are mostly privately funded, operating on small budgets, and intending to build actual power plants, not proofs of concept.
There's a bunch of discussion here comparing the results of crashes between flying cars versus non-flying cars. Is it possible to build craft that can fly around, crash into another such craft, get dinged up a bit, but still limp along in the air until it can reach a safe landing point? Or will physical contact with another craft always imply hurtling into the ground with more-or-less gravitational constant based acceleration?
Electricity was government funded. Radio technology was government funded. Metallurgy was government funded. The railroad was government funded. The automobile was government funded. Aviation was government funded. Interchangeable parts was government funded. Sure, the technologies were rolled out to consumers by private businesses, but the innovations were due to government research initiatives. The government has always provided the initial push and provided the necessary subsidies. Businesses can do just fine with stagnant technologies.
The problem with nuclear power plants was the spent fuel disposal problem. Even the French, who went all in on nuclear power and made it work, have been unable to solve this problem. Every nuclear power plant in the US has a big storage pool full of spent fuel rods and no place to put them. Sure, we could eliminate regulations and accept a higher background radiation the way we accept warmer, less settled weather in exchange for burning hydrocarbons. We'd also have to accept a landscape full of damaged nuclear plants. Finally, run the numbers. Solar is still cheaper and could become even cheaper.
As for flying cars, the roadblocks aren't energy and regulation. There are technological barriers. Unless we totally redesign our cities and suburbs, anything that requires a runway is out. Right now, vertical take off and landing means lots of noise which is a regulatory problem unless you live nearby. Then there's the airspace control problem. The airlines were the ones who got the government into airspace control. Technology can help; TCAS came out in the 1990s, but the general problem is unsolved, and the everyone-has-a-flying-car problem wasn't going to be solved with 1960s technology.
This review makes the book sound like an anti-government screed based on technological ignorance. I hope it's better than that.
Overall I find the review to be strongly lacking in critical analysis, indeed it is a summary more than a review with limited commentary beyond agreement. Ideas are summarised and well presented by the reviewer, but I don't think the original ideas hold much merit.
The author of the book asserts many things and I can only find agreement in this book review. This is not a totally wrong approach and perhaps the reviewer fully agrees with the book author on the major points they make, but I don't think this is a strong or suitable entry for the ACX community where Scott typically gives a balanced case for and against various aspects and arguments made in the book.
Indeed many of the arguments in the book lean into thoughtless mind killing partisan areas and are founded in ideology more than in sound logic. The reviewer overlooked the style, tone, and approach typical of book reviews in Scott's blog and in the fairly pro-logic and pro-analysis less wrong type community which is the audience.
Indeed the non-categorical reference to flying cars as in...actual flying cars rather than to technological innovation in general severely limits the scope and meaning of the book author's analysis. Besides some, misguided in my view, boomer style 1950's childhood dreams....do regular people need, want, or have a use for flying cars which likely have travel distances not hugely different from cars to take passengers no more than 300-400 miles without refuelling or recharging? I don't know of anyone living in any urban or suburban area whose life would be improved by such expensive and noisy machines.
We are having enough problems when considering the usage and approval of small unmanned drones flying over residential areas to deliver packages and takeaway, along with the problems they pose from noise pollution and hazards of drones falling out of the sky causing damage to people and property. Several pilot studies of drones flying around at over communities to do this was strongly strongly disliked and hated by the communities subject to the endless noise of drones flying overhead. Meanwhile we could have networks of autonomous trucks with boston dynamics style robo-dogs going around to delivery small packages instead with existing technology or very near term technologies....or you know...just humans delivery drivers in trucks.
I can certainly say I'd be 100% against my neighbour having a flying car or a helicopter pad in their yard and low and behold the allowance of helipads is very limited and nearly non-existent in residential areas. Such flying cars would be taking off and landing with huge amounts of noise everyday. Perhaps such vehicles could be forced to drive to specific take off and landing locations, but wouldn't' that just create huge congestion there at the mini aircarports if tens or hundreds of thousands of commuters wanted to use them? Along with the high cost of devoting land and airspace to such purposes. Would it be safe for such craft to fly over residential areas...and at what altitudes? Do I even care about the answer to such a question when the vehicles have nearly zero utility compared to their huge social costs?
A flying car is soo incredible unattractive and useless that it is hard to imagine why anyone would bother talking about them beyond as a toy for the wealthy or as a dream from their childhood watching the Jetsons.
I'm sympathetic to the libertarian perspective here, and It definitely seems reasonable that nuclear was smothered in the crib with no valid justification. But:
1. Government research -- Seems like a lot of the innovations of the 20th century came as a result of research funded due to world war 2 and the early cold war. Obviously though new government institutions ossify overtime and become less beneficial. Am i missing something here?
2. Flying Cars -- Are we talking about anti-gravity or glorified helicopters? What kind of noise pollution would we be dealing with? What would American levels of driving skill mean in terms of accidents and fatalities of these flying cars?
The book doesn’t sound as if it recognizes there are countries outside America. Japan and France went all-in on nuclear power, continuing that way for several decades after the 70s. Japan especially has obviously changed direction since, but France hit about 70% nuclear generation I think (much of the rest being hydro; a good combination as nuclear is base load and hydro can mostly be turned on and off).
Even if electricity is a bit cheaper in France, they don’t have flying cars.
Sure, the nuclear industrial base would be a lot bigger from the US adopting nuclear power than just France and Japan, but that doesn’t seem like something that should make quite the ‘all or nothing’ difference the book posits. Oh, and the Soviet Union built a lot of nuclear plant too, notably unconstrained by safety regulations. They didn’t develop flying cars either.
I am actually very pro-nuclear by the way. I just thought the book sounded a little blinkered: stuff does happen outside the US, after all.
> At $64 per day, the one billion people at Level 4 own a car. ... The average American moved from Level 2 in 1800, to level 3 in 1900, to Level 4 in 2000.
This rang weird to me, because 50% of American households owned cars in like 1925. (It went back below 50% thanks to the depression and WWII, I think, but above it again long before 2000.) If 2000 is when the average American hit $64/day, then "owning a car" might not be a good handle for level 4. That said, this might be "the average American", not "the average American household"; do 50% of Americans period (or at least 50% of adults) own cars?
I feel as though something is missed about radiation here. Radiation exposure as energy is almost certainly not a big deal at low levels. But the author seems to miss the fact that radiatioactive materials ingested into the body are a big deal. Walking past a light bulb is not going to harm you unless we are talking about some exceptionally powerful bubl pumping out a huge amount of light. Swallowing a lightbulb that stays lite inside yourbody for the rest of your life, though is likely to cause damage over time.
Netflix's show "Dark Tourist" has an episode where the host tours Fukishima...where "no one died". The tourists go about in the bus looking at the empty town and the host even sneaks off to go look inside some buildings where things are covered with dust. Over time, though, their geiger counters start increasing, going beyond what they were told is 'safe'...as they continue to rise, even when they confine themselves to remaining inside the bus, nerves start rising and the group leaves the area. This is years after Fukishima and the place is still covered in radioactive dust. Do I believe no one would die if somehow you hypnotised the entire population to have simply remained there and ignored everything? Not for a moment.
The fear of radioactive materials is more rational than the author realizes. They are cumulative over time. They are not able to be seen, smelled or tasted when they are in the environment and in a catastrophic event they are dispersed over a large area and essentially never go away in human history.
This is missed if you are concentrating on doses. In normal environments, you only get doses of radioactive energy (say by getting an x-ray or flying in a plane). Here you normally should not have much to worry about unless you have a major malfunction of the machine or you are exposed to a nuclear explosion for some reason or your finger slips while messing around with a 'demon core'. However if the reactor explodes, it isn't just the dose from the immediate radiation levels you have to worry about but asorbing the material itself.
Reasonably interesting and well written, but I don't feel like I learned that much from it-- which may be a critique of the book as much as the review. Won't be voting for this one.
The core problem isn't a lack of a storage facility, it's pretending that perfectly viable nuclear fuel is "waste".
I think he's talking about (1) extracting the plutonium produced as a by-product of U-235 fission, and (2) extracting and re-concentrating the remaining U-235. I don't think he's talking about using any of the actual high-level wastes, e.g. the Cs-137 or Sr-90 et cetera that are usually considered the problem children in spent nuclear fuel, isotopes produced in abundance that have half-lives ~10-100 years (so they're very radioactive but don't decay very fast). So, yes, those latter isotopes are just always going to be a pain, so you just have to put them in a (dry) hole somewhere and let them decay over a few centuries.
Those two isotopes in particular are also annoying because they have the chemical properties of K and Na (in the case of Cs) and Ca (in the case of Sr), and so they can slot into the places where those elements are used in the body and just hang out, irradiating you internally for years. Cs also forms very soluble salts, so it disperses easily in any kind of watery environment.
Many of the fission products you isolate during reprocessing are useful for other things, such as medical isotopes, radioisotope thermal generators (including Sr-90), etc. I'm not sure what they do with the cesium. But France after reprocessing is left with a pretty negligible amount of spent fuel waste and they just keep it all in a building somewhere.
Yes, I think that's true to an extent, and I suspect Cs-137 is used in radiotherapy applications for just this reason (it's conveniently obtained). But my vague impression is that the requirements of medical-grade isotopes are such that you need to use special reactors constructed for (or amenable to modification to) that purpose, I don't get the impression that reprocessed fuel from power reactors is usually considered a great option. Indeed, my impression these days is that the bigger nuclear medicine departments are moving towards installing their own synchotron to make their isotopes to order, and on the spot.
But maybe they do it in France, and maybe I'm generally wrong, this is not my area. I am little suspicious that any of these ancillary applications can absorb the few hundred kg of high-level waste you'd collect from a refueling. I'm still thinking you'd end up with at least a tonne or two of stuff you just needed to stick in a hole in the ground somewhere. (Not that I think this is a big problem, it's a big planet, there's plenty of places to put such a hole.)
People tend to worry about waste that has a half-life of the order of 10,000 years, but IIUC almost all of that long-lived stuff is plutonium, which is nuclear fuel that we can and should separate out and burn in reactors (which will split the atoms into smaller radioactive atoms).
The smaller radioactive "fission products" (which can't be burned) mostly have half-lives below 100 years, and after 500 years their radioactivity level has fallen below the level of uranium ore (not even uranium metal).
The goal of being able to store waste without any human maintenance for 20,000 years always seemed to be of dubious value — I mean, how about we focus on preventing an apocalypse, rather than focusing on guaranteeing that post-apocalyptic settlers who somehow decide to settle in the Nevada desert 8,000 years from now won't suffer from a 0.1% increased risk of cancer because they don't know what the ☢ symbol means?
But since society seems devoted to storing the waste For As Long As It Takes, it's worth noting that it's much easier to design a structure to last 500 years than 20,000. But the whole issue is kinda moot because plutonium emits alpha radiation, which is easily shielded (unlike the gamma radiation from many shorter-lived isotopes). So, you can safely sleep atop a drumful of plutonium, just don't ingest it, or grind it into dust and breathe it, or grind it into dust and dump it in your drinking well.
As for isotopes with very long half-lives (over 100,000 years or so), the crucial thing to understand is that radioactivity is something that happens with radioactive decay; therefore radioactivity is inversely proportional to half-life. As the half-life gets longer, the danger level decreases: all else being equal, if something has a half-life of one year, it is 10,000 times more hazardous than something with a half-life of 10,000 years. Therefore, things with extremely long half-lives aren't scary; it's the intermediate half-lives (10 to 10,000 years) that are problematic, as they are both long-lasting and radioactive enough to potentially cause cancer.
P.S. A one-ton drum of plutonium contains as much usable energy as 3,000,000 tons of coal, so I'd just like to add that nuclear waste is awesome compared to other kinds of waste. Because it's tiny.
Coal ash might be stored in open ponds or giant buildings that occasionally split open and dump their entire contents into rivers: https://appvoices.org/coalash/disasters/ — but because nuclear waste is so small, it's practical to encase it in concrete so that it never harms humans, wildlife or even plants. Why didn't those activists worried that containment isn't 100% perfect spend the last 60 years fighting coal instead of nuclear?
Good point !
A geologist friend of mine worked on Yucca Mountain and he called it a sham. They were being asked to guarantee there would be no major earthquakes or volcanoes for 10,000 years. According to him that was impossible to state with high confidence.
But even the reason your geologist friend was asked to ensure that for 10,000 years was a sham. It's totally not necessary. Even high level waste becomes less radioactive than a granite countertop within a few hundred years.
And there are plenty of rocks that haven't been touched in millions of years! Pick any one rock that hasn't moved for 10 million years, drill a hole into it, and drop the waste in. Honestly the whole storage "problem" is mostly regulatory theater.
I've never understood that requirement. Like, why are we *planning* for a dystopian future in which humanity has degenerated so much it can't exercise competent stewardship of a waste respository?
We also have 5,000 nuclear weapons lying around . Do we need to weld instructions for their care and responsible storage to them, written in pictograms at a 4th grade level, just in case the future turns out to be stupid?
Yeah honestly even the fact that we're planning to ensure that this waste repository is safe for the post-apocalyptic humans who might stumble upon it is...simply not a standard we hold literally anything else to.
Yes? Just since 1950 there have been dozens of accidental near detonations of nuclear weapons.
https://www.atomicarchive.com/almanac/broken-arrows/index.html#:~:text=Since%201950%2C%20there%20have%20been,been%20lost%20and%20never%20recovered.
And on your second question: yes? That was part of the planning for Yucca Mountain
https://www.bbc.com/future/article/20200731-how-to-build-a-nuclear-warning-for-10000-years-time
What the heck is an "accidental near detonation?" Is that like "not a detonation at all" after someone dropped one by accident? If an airman bumps into an W80 in a warehouse and clonks it with his wrench, and nothing happens, is that an "accidental near detonation?" I'm deeply unimpressed.
Anyway, the fact that a maximum of 20,000 weapons has been supervised by a passel of 19- to 25-year-olds with at most a HS diploma for 75 years without a single notable slipup is proof enough to me that it can be done. At this point I think the burden of proof is on the people who think future American government will screw up something as obvious as digging an irrigation channel right next to the big yellow warning sign that says YOU WILL DIE IF YOU DIG HERE .
Even with the school board's actions, it appears that most likely there were no negative health effects to the people who lived near Love Canal. Source: Wikipedia.
I remember Love Canal. At the time I thought the reaction was hysterical, and I still think that. Perhaps I need to learn something, but my experience was that in the 70s public attitudes towards random assorted chemical in the environment shifted radically -- in the 1960s after we got done cleaning a carburetor with a bowl of gasoline, we just threw it onto the ground. Heck, people changed the oil on cars and threw it on the ground. Turpentine after painting? Down the sink. Same with business. If a gas station leaked 1000 gallons of gasoline into the soil, big deal, nobody much cared.
All of that changed in the 70s. All of a sudden we got really upset about those things, started suspecting them of causing cancer and allergies and whatnot. (And some of that may have been true, after all, cancer has to come form somewhere.) Love Canal was just a symptom of the changing times.
So that would be a better argument if we were really careless about nuclear waste *now* and the future might get a lot more uptight and be pissed at us. But we're talking the other way around -- we're uptight *now* and, strangely, feeling like we need to assume the future will be slobs. But that has never happened before, and it seems strange to assume.
75 years < 150 < 500 < 1,000
It’s massive recency bias to think the next 75 years will be just as stable as the last 75.
Look, I support nuclear power in general, but I also know humans are very bad at assessing risk, especially over very long timelines.
I’m old enough to remember the Berlin Wall falling and a tremendous amount of entrepreneurial dynamism and overall optimism in the 1990s. America was the dominant hyper power and most of the world *liked* it that way.
But then a few dozen followers of Osama Bin Laden changed everything.
It’s foolish to assume something like that won’t happen again.
What do you mean by "changed everything"? I don't see that anything changed at all except for the greater clampdown on the tiny amount of terrorism. Or are you talking about the war in Iraq & Afghanistan, which I'm pretty sure was small compared to wars of previous generations?
In any case, the ~3,000 people that died on 9/11 is to me a very small number when you're talking about the entire U.S. over a long time span. I mean, there have been about 200 times as many deaths due to COVID.
Well, no, it's *logical* to assume the next 75 years will be rather like the previous 75. Unless you have evidence to the contrary, or at least a good working hypothesis. I don't mind a *certain* amount of caution, e.g. it's definitely prudent to put your nuclear waste out in the desert, far from anywhere even remotely habitable, and mark it plainly, big warning signs and such. But assuming that within the next century we're going to be in some Thunderdome situation where nobody reads English any more and the wild kids dig up the shiny metal bits to make into necklaces -- this goes well beyond prudence.
Look, the whole point of the future is that they're almost certainly going to know more than us (technologically speaking), the same way we have known more than the people in the past, for the past 1000 years or so. So why try to do their job? Do ours. Leave them a growing economy, healthy happy new generation of people, some new inventions -- and yes, a few inherited problems with which they're expected to cope. After all, our parents handed us nuclear weapons and the Cold War, and we've done OK with that.
People seem to manage extensive hours of training, constant practice, and a focus on situational awareness and response while driving. Are you saying flying is intrinsically much harder? Or that you wouldn't expect flying to become a primary mode of transportation, so it would be harder to maintain the practice?
Flying is intrinsically much harder than driving. It's tricky to get an apples-to-apples comparison because the units of measure are different (fatalities per passenger mile vs fatalities per hour flight time), but general aviation is at least 10x more deadly than driving. And that's in our current regime, where people who are doing the flying are self-selected to be more focused/situationally aware/skilled than your average random car driver. There's obvious reasons for this - mechanical issues in a car generally mean you pull over and wait for a tow truck. Mechanical issues in a Cessna mean you do everything exactly right or you die.
I have a PL, IFR, and High Performance rating. Flying an airplane is intrinsically much harder because of the procedures involved, but I might seem like a heretic to other aviators for saying that I don't believe these are things that couldn't be overcome by the public at large by more and better automation and inherent safety protocols built into the systems as available to the public at large.
I absolutely could see a future where the general public had access to and used flying as their primary *individual* form of transportation, but in that form a huge amount of the system routine and implementation would have to be prohibited to the owner/user.
99% of aircraft incidents are still (again) pilot error, and that is with operators that are trained and certified above the level of the general public. For the general public to have it as their primary form of transport, even more freedom would need to be limited.
Think "Minority Reports" self-driving rail-pods, but airborne.
yeah, but how much automation was available up until, say, 10 years ago? Suppressing flying cars in the 70s might not have been such a bad idea.
+1 Yes, flying today's airplanes is too hard, but a market for flying cars would put pressure on the control system design to make it simpler. Using "cockpits are too confusing for normal people" as an excuse is 1st order thinking.
Carl Pham below points out that "we shouldn't do it because it's too dangerous" means that we never begin the process of learning how to make it better.
Air taxis would make a ton of sense, though, right? People who fly all day can afford extra training; that we never got air taxis supports the thesis of excessive regulations.
...though it occurs to me that many people would want to hire a taxi to fly downtown, where airstrips might be scarce, even assuming you can build them directly on top of the normal roadways. So another explanation is chicken-and-egg: you need a lot of taxi traffic to justify the airstrips, but lots of airstrips including downtown ones to justify the taxis. Another popular destination should be airports, but air traffic control wouldn't want small aircraft encroaching on its airspace. I doubt this is insurmountable tho.
A third explanation is risk to third parties. Air taxis may be safer than cars (certainly per mile traveled), but a fatal crash will occasionally kill a random person on the ground. This is mostly true of car crashes too! However, while a car crash may kill a random person on a road or sidewalk, an air-taxi could crash into a house, which probably makes them feel psychologically more dangerous even it they're not.
A fourth problem: noise pollution. Not only are aircraft louder than cars, but the walls put up on highways to reduce noise wouldn't work on them.
Maybe to get past choked highways and reduce infrastructure costs. But the thought of then weaving between skyscrapers gives me goosebumps. I'd keep them out of downtown or any residential neighborhood. That would make them more like an airBus anyway, i think.
Both, I'd imagine.
I am a pilot, and flying is much harder than driving. Maneuvering in three dimensions is part of it, but bigger parts are cross-coupling of controls (where every input does at least two different things), a higher level of critical multitasking, and the part where you can't pull over to the side of the road or even slow down if things get temporarily overwhelming or the hardware glitches. Also, all-weather flying (necessary for routine travel) becomes a complex exercise in real-time systems management and requires training *out* some deeply instinctive behavior. In my particular case, that makes me safer in the sky than on the ground because flying effectively commands my attention but driving leaves me too easily distracted, but I'm not exactly typical.
The average person can learn to fly a light plane safely if you insist on it. And we sort of do, for people who want to fly. But the difficulty, expense, and tedium of the necessary training and the non-trivial fraction of the population that will wash out and be aeronautically disenfranchised, will likely make this politically infeasible at flying-car levels of utilization. The remaining hope for "flying cars" is increased automation, but safe autopilots are a classic 80/20 problem where, for this application, we'd really need that last 20%.
I'm reading the book because of this review, and one point Hall makes is that the autogyro from the 1920's has a reverse risk profile from an airplane. It's unstable on takeoff, but will glide to a stop on landing, even if engines fail. So the "can't slow down and can't pull over" might be less of a problem with another technical model.
Nice!! I want one now. One thing Hall doesn't answer well is why US regulations blocked progress across the world. The other review raises that issue as well - financial and health regulation has slowed things down, but not stopped us in our tracks. Are flying cars just on the wrong side of the cost disease? Or is there some other issue (like, say, physics)?
"People seem to manage extensive hours of training, constant practice, and a focus on situational awareness and response while driving."
And yet *look at how dangerous driving is.* Even if flying drivers were no more prone to accidents than terrestrial drivers, there would still be too many accidents, and they'd probably be more deadly.
Self flying cars seems to be an easier problem than self driving cars so maybe there's hope there.
Self-flying cars are an easier problem now, but that would quickly devolve if even 5% of the US population had them and used them as their primary form of transport. Granted, working an integrated Terrain/Collision & Avoidance System (TCAS) for objects moving in 3 dimensions instead of 2 is slightly easier, but if you see my comment above it would be a catastrophe if the car/system gave the choice to the pilot/operator on how to avoid a collision as opposed to the linked computers making that choice.
If you want to see a horror story of just how confusing human/computer/system interfaces can be vis-a-vis collision avoidance, watch this simulation of the Flight 2937 collision of Switzerland.
https://www.youtube.com/watch?v=iYJWWngRxus
As a matter of handling traffic, I imagine flying cars would be restricted to long-distance travel, between or just outside cities, and restricted to tired travel within them. Also freight trucks would probably remain wheeled, given the weight, and given that shippers already have the option of freight airplanes, and aren't using them.
That's pretty much what Heinlein was figuring 70 years ago for domestic transport: the wealthy and important government agents would have city cars that can quickly covert into aircraft in the countryside, then land in the hinterlands of the destination and roll into town.
Why does the *average* person need to be able to do it well for *anyone* to be allowed to do it at all? This seems indicative of the weird current hyperfocus on equity over opportunity which I mentioned elsewhere. None of us can have flying cars until all of us can? Why?
Is flying a flying car so hard that we can't, say, take the 10% best drivers in the nation (about 20 million people) and let *them* fly around? What would that do to boost their productivity? Maybe there are a ton of really smart people from whom we are not getting X new inventions a year because they are freaking stuck in traffic.
I think because on one hand, the gains are kind of speculative: "spend less time in traffic -> more inventions" seems like a stretch, especially considering how many other ways there are to solve that problem: remote presence being the obvious 2020 solution, public transit or a private driver to make 'commuting' time more productive, etc.
And the potential downsides are obvious and specific - the potential for loss of life and destruction is pretty clear (and not just for the people behind the wheel!) in case of accidents, to say nothing about the potential deliberate acts of destruction.
Right. And I get this. As I said elsewhere, we have just become across-the-board more timid about things. "But is it safe? What could go wrong?" tends to trump "But what marvels might we unlock if it goes right?" pretty much all the time.
There are argument both ways, of course, and either "safe at any speed" or "damn the torpedoes" extremes are unwise. But whatever choice we make comes with costs. If we are much more concerned about what will go wrong, then we will be generally less entrepreneurial, take fewer risks, and our social progress will increasingly resemble NASA's progress on human space flight -- very, very slow, but also very very safe.
"we have just become across-the-board more timid about things."
Can't this be seen as a reflection of the increased opportunity cost associated with various forms of risk taking? As measures of expected life satisfaction/life expectancy substantially increased over the 20th century, the costs of risk taking in the form of foregoing expected life increased.
I think it's a good argument in general, but I'm not sure the data support it in this case. Life expectancy in the US has only gone from 70 to 79 since 1960. That's nothing to sneeze at, but it's hard to see it having that big an effect on risk-taking. I don't think life satisfaction has increased at all, and indeed the rising rates of middle-aged suicide, and middle-aged drug use and overdose, would kind of suggest the opposite. Heck, even the teenagers are having less sex, apparently. Four hours of World o' Warcraft doesn't compare to getting it on with the fox from Algebra, so I can't see how *they're* happier.
Good points.
'Life satisfaction' was a poor choice of words on my part. Maybe it would be more accurate to say video games/the internet has largely solved the "problem" of boredom. What equivalent way of wasting time by yourself did teenagers in the 50s and 60s have? Comics? Playing records? Seems like there's no real equivalent to the video game slob stereotype we have now. I wonder if a lot of early-life entrepreneurship (broadly speaking) is just a way to avoid boredom.
We already do let a few highly trained people fly around. Furthermore, we allow them to take passengers, so the rest of us can have the benefit of their skill. They're called airline pilots.
I said 10% not 0.1%.
We don't need 10%. A little tweaking and we'd have enough capacity on our planes.
Not sure I follow. I proposed 10% of drivers could drive flying cars. We do not have 10% of people flying airplanes.
We don't need 10% of people flying airplanes. A little tweaking and we'd have enough capacity already.
Also, I expect that 10% is far in excess of the amount of the population capable of being trained as adequate pilots.
I think a lot of the problem is that there isn't a clear Schelling point for regulation in between "The average person should be able to get a license without it being unduly burdensome" and the current very cautious regime where it takes hundreds of hours of expensive training to get a pilot's license.
Regulations for cars can't get that strict because the average person expects to be able to drive and the population as a whole wouldn't tolerate that changing, but once you accept that only a fraction of unusually skilled people can do it safely the standards might tend to err on the side of caution because of the incentives on regulators.
Well it's a good thing the automobile wasn't invented last year then, isn't it? "You're going to let 16-year-olds sit at the controls of a two-ton metal machine capable of accelerating to 100 MPH? Insane!"
If ten percent of the population is flying freely above the crowded freeways at 100+ mph on their daily commute, the remaining ninety percent are going to really notice this. And be jealous. And vote.
Right, well, you could say the same thing about absolutely any manifestation of wealth. It's not like people don't already notice private planes, yachts, gated communities, private schools, expensive condos with beautiful views in ski resorts.
The way we traditionally deal with this is social mobility. People have the *opportunity* to get rich themselves, through hard work and talent, and then they join the ranks of the 10%. People are smart enough to understand that if everyone has to have the same stuff, it's going to be a low level of stuff, and so they're generally willing to instead enter the lottery of who ends up in the 10% -- as long as they believe it isn't pure chance that produces the winning ticket.
They also usually tend to believe that what the 10% have usually moves down after a while anyway, as long as economic growth continues -- at one point you had to be decently well off to afford a car at all, then everyone had one car but being a two-car family was pretty tony, and now these days it's buying a Tesla for the 3rd car that marks you out as in the 1%.
I agree such attitudes are less common today, though, and there are many more like what you're saying: "Only 10% of us will ever fly above the Blade Runner dystopian squalid streets below, and it's going to be some connected/powerful/aristocratic 10% that will never include me or my kids, so screw that." But the existence of that curious change, and wild speculation on its origin, is kind of the point of the reviewed book.
I think a lot of the military research successes have come because they have goals of making a practical technological application as their first objective. Academic scientists are frequently driven by raw curiosity and/or obsession with the function of some very specific part of the natural world. There's nothing morally wrong with that, but it's not optimal for finding practical applications. They stretch the words in their grant proposals to try to convince funding sources that their work will have some practical benefit--and every so often one of the many projects does--but fundamentally their interests are usually anchored to the thing they're studying rather than a solution to a practical problem.
In contrast the military's objectives are to increase real world combat effectiveness. That's a practical goal, and therefore they're much more willing to sideline or abandon research avenues that are less likely to result in new solutions to practical problems and to do so sooner.
> Academic scientists are frequently driven by raw curiosity and/or obsession with the function of some very specific part of the natural world
Worse than this, I think a lot of academic scientists aren't even all that interested in the very specific thing they study, they're just kinda stuck with it.
Here's the way it works: your PhD supervisor was the world expert on the subject of tantalum oxide. When you came along he racked his brain for five minutes and assigned you a project looking at the ever-so-slightly-different-and-more-obscure subject of tantalum sulphide. You work hard for five years and sure enough, you're one of the world's top experts on the subject of tantalum sulphide. You don't really care about tantalum sulphide. It turns out that tantalum sulphide is completely boring and unimportant. But you _can_ think of a dozen possible tantalum sulphide related projects that you could potentially get some grant money to work on, and you don't know enough about anything else. So you start churning out grant proposals on the subject of tantalum sulphide, starting with "Tantalum sulphide has potential applications in X, Y and Z".
I feel like this is kinda the case for at least half the academic scientists I've met; they're stuck in some kind of dead end doing research that they're not especially interested in but they suspect they can get funding for.
Yeah, the 'just stuck with it cause it's what your advisor had to bequeath upon you' is why I left the academic research field. Naively I originally went in with some very specific research goals and did so specifically because i felt that basically no one was working on them. I eventually found that the fact no one is working on them already, means it's nearly impossible to start doing so within the framework of academia and grant funding. You need to either be rich enough to self-fund, or you need to be fortunate enough to discover an exciting breakthrough in your target direction while still a student--and even then it probably needs to be monetizable to work out. Even if you do one of those it's incredibly hard to start down the new direction in academia because there may only be a handful of people in the world who'd be a good fit as your advisor and there's no guarantee you'll be accepted to (or want to attend) their particular institution.
> As for government science not being worth a damn, that may be true now, as I'm told the funding process has become completely corrupted. However, there are notable successes like GPS and the internet
I'm old enough to remember the exact same criticisms of government science being made decades ago, while those things were being researched. I don't think there's been any major change, just that as always the kind of research that governments fund is unsexy and far from the point at which it would be implemented, so it will only seem valuable in retrospect
My pat answer to the flying car question has always been "We do have flying cars, they're called helicopters".
Now of course there's a good reason why most of us don't own helicopters; they burn a lot of fuel, they need a lot of expensive maintenance, they require a lot of specialised training in order to fly them, they're relatively dangerous, and they're so loud that you're not allowed to land them in most places. But all of those problems are intrinsic to hovering transport, so they preclude your "flying car".
An electric automatically-piloted quad-copter might be able to solve most of these problems to some extent, though.
We also have jumped which have the same problems ,but even worse.
*jumpjets
Does a quadcopter help with the noise part?
Or an autogyro? I wonder if they could be modified to act more like a VTOL. https://www.youtube.com/watch?v=Etcq3lfqIbs
hard disagree. couple hours in a sport plane and in a robinson helicopter plus regular motorcycle rider. of the three, plane is the easiest and motorcycle the most difficult/dangerous. yes, crashing the plane would be bad, but after takeoff and setting it on course and trimming it out, i could practically take my hands off the controls (extremely low-inertia 600 pound aircraft with no autopilot or stabilization aids). it’s much less stressful as you’re making fewer inputs/decisions per minute. as far as take off/landing, the amount of computing power in the average tesla is probably more than enough to automate those anyway esp if vtol as the book proposed.
On the other hand the first time a flying car breaks down and crashes on a kindergarten there will be no end of additional restrictions made about who can fly what over built up areas.
I’ve never heard anyone before say government-funded science was bad for science!" Ayn Rand has as character in Atlas Shrugged who was a brilliant physicist before he got government funding. Then he became useless. Nassim Taleb (I forget which of his books) also argues that many of our significant inventions came from outside academia.
Funny that the author uses a machine learning analogy, as ML is definitely an invention of academics.
I don't think it would be hard to imagine that there is an *appropriate* level of government funding for science and technology, and that at levels that are too high it sucks all the oxygen out of the room -- and then all your science becomes done in the way you do government science, which is not the only way to do it, and in some cases not the best way to do it.
It is certainly the case that 50 years ago there was a far more vigorous realm of basic R&D outside of government and academia. You think of Bell Labs, or IBM Yorktown, Xerox PARC, Exxon Annandale -- these place attracted absolutely first-rank talent, and in their day invented amazing technologies. But corporate R&D has been eviscerated, and it isn't *completely* out of the question that part of that is the entry of the 800lb gorilla -- government funded academic research. From the point of view of the new researcher, there is much to like (initially) about the government/academic model: you don't have nearly as strong a deadline/results pressure, and you can often work on more abstract problems. And for some areas of research that is an excellent shift. But...there are areas of research where a certain amount of bottom-line and practical focus *is* good for results, and can even be quite beneficial to the individual, in the era when it was possible to become very handsomely rewarded by commerce.
I think there's a very active world of corporate research in things programming, and for this we can thank many of the brilliant innovations in computational stuff. And to some extent we still see that in biology, although perhaps not as much as we'd hoped 30 years ago -- biotech is still very tightly tied to academia, and the bio giants (e.g. Big Pharma) don't seem to be *expanding* their basic R&D programs. If Utopia could be created just by brilliant programming, we'd be in good shape, but alas it needs progress in things made of actual stuff also.
I'm pretty sure a lot of private research money is gone because companies realized that groundbreaking research usually doesn't actually pay itself off to the company funding the research.
Leo Szilard wrote a short story in *Voice of the Dolphins* where a character predicted pathologies of government-funded science. IIRC it was written in the late 40s or early 50s, satirical in the form of proposing NSF-style funding as a way of sabotaging an enemy society.
(Szilard was a physicist who proposed the fission chain reaction in the 30s and clashed with Manhattan Project management.)
Thanks for that. I read the Szilard biography, "Genius in the Shadows" years ago and loved it.
I've heard it before. A number of years ago I met a very smart person whose name escapes me (associate of current MIRI researcher Eliezer Yudkowsky), at a SENS research conference on aging who brought it up during a Q&A. Most of the attendees get their paychecks via grant money so his question was laughed off, but at the time I recognized his name and went to talk to him about it afterwards. We had an interesting discussion about historical funding of research, but most of it was about pre-industrial revolution research where discoveries in the sciences were mostly made by rich or patronized-by-rich people and no one had even considered government funding go through a bureaucracy to be distributed to researchers.
It's not that nothing is ever discovered when research is funded via government grants, but it shouldn't be a surprise to anyone most of the funding goes to things which enhance the prestige of the researchers of yesteryear who are now in charge, rather than to people who intend to make progress in completely new directions. I'm not sure I'd attribute ML primarily to academics either--the math of it is a very old idea. Rather computers have recently reached a point in all their capacity types where the technique is useful. Academics working in ML are primarily "just" fiddling with the variables of layer counts and compounding systems to get new results. It's important work, but not fundamentally a new paradigm.
I don't have a particularly strong opinion for/against public research funding, but I think dismissing the idea without a lot more discussion is exactly the kind of failure that "Where's My Flying Car?" argues we already suffered from. Similarly labeling the idea "libertarian", and especially associating it with Ayn Rand, makes it more political, and therefore more controversial, than it needs to be and wrongly biases a lot of people against the idea. We'd need much better evidence that the bureaucratic funding model benefits ML research more than a private funding model would before I'd find its use in the review ironic.
"I'm not sure I'd attribute ML primarily to academics either--the math of it is a very old idea. Rather computers have recently reached a point in all their capacity types where the technique is useful."
Those facts are correct, but the old math was originally invented by academics as well. So it is certainly attributable primarily to academics.
As a side note on this idea of government funding in historic periods...The idea of trying to separate 'the wealthy' such as the random Lord, Duke, Earl, and sundry types of lesser nobleman from the idea of government is invalid in my view when talking about that era.
The rich and powerful people funding or lending their power to various researchers, artists, and merchants in the enlightenment era were either directly members of the government through their aristocratic positions or were themselves strongly beholden to support from such nobles.
The idea of trying to say this funding wasn't run through large government bureaucracies in the 1700s doesn't make sense as such structures by and large did not exist at the time with the major government actions being tax collection, the military, and running a fairly simple court system to arbitrate disputes and process criminal charges. I don't think there was an NIH or NAS type body at the time and the main equivalent would have been the nascent University sector and the not-so-private actions of nobles with governmental authority to individually fund whomever caught their interest.
It sounds like a coloured view of history seen through a modern lens which ignores how people lived at that time. I'm obviously reading into that position from only a few words, but it sounds fairly anachronistic to me.
You're understating the case for ML. The fact is that there have been major advancements in ML. GANs, Transformers, reinforcement learning, just really large networks, RNNs, etc... are all serious innovations in ML. And most of them have come out of private labs (in fact most of them have come from FB and Google...).
I was about to say the same thing. Arguments against government funded/controlled science go back to the 1950s.
And the thing is - they kinda have a point. Instead of coming up with new and innovative methods, we end up taking years to publish papers that are mostly forgotten and ignored.
It also freezes in the horrible bachelor-master-phd requirement. Let's get real: I can train anyone to be a good experimental biologist in about a 100 hours or less. 12 years' effective apprenticeship? Get. Real.
Nah. You can train someone to be a half-way decent technician in that period.
I've yet to get to the point in my PhD at which they sit me down and tell me how to think independently, I personally just think you just have to learn from experience.
> And the thing is - they kinda have a point. Instead of coming up with new and innovative methods, we end up taking years to publish papers that are mostly forgotten and ignored.
Would we suddenly get a lot better at coming up with new and innovative methods if government funding were removed, though?
> Let's get real: I can train anyone to be a good experimental biologist in about a 100 hours or less
You can train someone to do biology experiments, yes; the part that requires an expert is understanding which experiments are worth doing.
(Or, less pithily, the job of the expert is planning out a full research program that will advance a particular sub-sub-field of biology, assembling a team of 100-hour doofuses aka PhD students to actually do it, and then properly communicating these results to the other experts in the field.)
Yes, you do need that extreme knowledge - to _plan a full research program_. Which is something you only start planning late into your postdocs.
The cycle at the moment is:
- Go to 4 year undergraduate course and learn a bunch of theoretical stuff
- Forget it
- Use your grades to get into a PhD program
- Start learning practical skills
- Start seeing the use of theoretical stuff from your practical skills
- Start learning a bunch of theoretical stuff...
Wouldn't it make a lot more sense to start with the practical skills, and then build the theoretical knowledge on top of that?
I wish. The main job of the PI is to raise money. It's like running a start-up in the VC stage forever.
This depends a lot on how government funding works. In the US and UK, yes. In most remaining Europe, no. It is part of the job, but by far not the main part.
In most universities in Germany or France, you can get by without raising any external money at all. You will have less PhD students, and you won't be a superstar, but even superstars do not spend the main chunk of their time allocating money.
This does have downsides. Whether it is overall a good or a bad thing, this is really complicated.
"In most universities in Germany or France, you can get by without raising any external money at all. You will have less PhD students, and you won't be a superstar, but even superstars do not spend the main chunk of their time allocating money."
In France, the current annual funding of a researcher is about 2000 euros. Its is almost impossible to do anything without grants, at least from experimental people.
Well, in principle you can do it in the US once you have tenure, but...you'll never be promoted, and you'll get assigned to all kinds of painful committees, get shit teaching assignments, that kind of thing. The university really loves its overhead :(
In fact, that's the corruption I'd most like to see rooted out of the system. When research overhead can make up twice as much of the university budget as tuition, the incentives are pretty screwy.
If it weren't for the 12 years of apprenticeship AND that said process effectively locks you into researching some specific side project of your advisor's specialty, biology research would be my current occupation. The current paradigm is actively turning away people who want to go in directions that the fewest people were interested in historically, and that's inherently going to bias results against real breakthroughs.
"A survey and analysis performed by the OECD in 2005 found, to their surprise, that while private R&D had a positive 0.26 correlation with economic growth, government funded R&D had a negative 0.37 correlation!”
It seems widely unlikely to me that this strong negative correlation is causal. Has the - 0.37 correlation by any chance been calculated by mixing developped countries (High public RD, low ec grotwh, and developping countries (low public RD, high economic growth) by any chance?
I find causality believable. When I left my job as a bioinformatician, I had in mind to do a bioinformatics start-up. I eventually gave that idea up, because nearly all of the bioinformatics software used in the US is developed using government grants, and made available for free. It's difficult to find anything in the bioinformatics space where you don't have to compete with people who are getting paid by government grants and giving away their software. But that free software isn't easy to use, and usually isn't supported or maintained.
It seems to me that your observation suggests that goverment money makes producing well maitained product more difficult, but it really doesn't seem obvious to me that well maintained software is a key to innovation.
Firstly, if you use economic growth correlations and/or practical applications as measures for how well science is being done, something is wrong - especially in what concerns fundamental research.
What should be analysed is how public funding criteria have often created an environment where risky projects are disincentivised, especially as more and more people got PhDs and exhausted what little slack existed for trying wacky stuff without fear of being outcompeted by conventional incremental stuff.
Anyway, since private companies will do R&D in search of profit regardless, it seems logical that public funding should go primarily to that research with no obvious profit routes. The question is how that funding can be more efficiently allocated in today's highly competitive academic world.
If you haven't ever wondered whether government funding of research was bad for science, you haven't done research with government funds.
I have so many horrible stories that I don't know where to begin. But let's start with Congress. Congress understands that we need basic research, but we also need that research to lead to profitable businesses, and to solve urgent technological problems. So, in a lot of basic research funding, they've decided to kill 3 birds with one stone by requiring that basic research must also be applied research, and must be the basis for a profitable start-up business. In the SBIR program, which I'm most-familiar with, your grant proposal must explain how it is basic research, how it is applied research, and what clients you have interested in the product that will come out of this "basic research". If you get to Phase 2, which you must get to at least 1/4 of the time in order for SBIR grants to be profitable, your Phase 2 grant proposal must have a client lined up and committed to provide part of your funding. To get to Phase 2, you must focus on your business plan, and spend about 90% of your budget on producing a cool-looking Phase 1 demo.
The result is that few SBIR grants do much basic research. In some years, I've looked at most of the unclassified "research" grants offered by the US government, and concluded that the only actual basic research being done by any government agency was by DARPA. DARPA is the very best agency we have for doing basic research.
When I submitted, won, and ran a DARPA contract, the COTR (Contracting Officer's Technical Representative) in charge of my $100,000 project was also managing a project which, if I recall, had a budget of $100 million, so that it was literally not worth his time to read any of the reports I read, or to answer any of my emails or phone calls. He had some underlings inform me of this. My team was one of 5 or 6 other teams awarded a Phase 1 contract. Near the end of the contract, after I'd worked 2 months of 12-hour days and weekends preparing the final report and demo, I got an email telling me not to bother completing it, because the COTR had already decided who he wanted to award the Phase 2 grant to. It was given to the only team which appeared, from its slide presentations, not to have produced anything other than slide presentations.
That was one of the most-successful government research projects I ever worked on--the software I developed did eventually make the company a lot of money--for the simple reason that, although no one was interested in the results, at least no one on the project wanted it to fail.
Contrast this with the NASA/FAA grants I worked on. Back around 1970, Congress ordered the FAA to use NASA engineers for airspace research projects, in order to avoid suddenly firing all those engineers after the moon landing. So air transportation research projects were managed by NASA, and carried out by government contractors (so those NASA engineers got fired anyway). NASA was monitored by the FAA, which saw NASA as a bunch of pointy-headed nerds with no practical experience who should stop telling them what to do. Plus the FAA didn't really want to automate air traffic control, because while it would save lives and a lot of fuel, it would put FAA employees out of work.
Or contrast it with the government-funded bioinformatics work I did for a genome research institute, where I was supposed to automate the work of the genome annotators, who were supposed to help me with the program and approve it when it was ready, after which they would be fired. That worked about as well as you'd expect it to.
The most-successful project I ever ran was a NASA project. As often happens, the original COTR who wrote up the project solicitation had been rotated out before the project entered Phase 2, and no one else in NASA or the FAA had the slightest interest in the project. Even the new COTR, whose job is to ensure that I carry out the work approved in the contract rather than repurposing it to some other objective, encouraged me to repurpose it to some other objective that someone actually cared about. So I did; and that project made the company a lot of money, and saved NASA $40 million, though it will probably never be used for its intended purpose (to automate air traffic control).
But we haven't even begun to talk about the main reasons government research grants waste money. One is that government funding centralizes funding, so for every agency there's somebody in a room in Washington DC who's responsible for $10 billion of grant funding every year. They'd much rather manage ten $1 billion projects than ten-thousand $1 million projects, even though the 10,000 $1 million projects would be much, much, MUCH more efficient.
(I once read an NIH blog post describing a study of the relative efficiency of NIH research grants. They compared the output of grants of between $1 million and $20 million by counting the number of research papers produced per project. They concluded, IIRC, that projects costing more than $10 million were almost twice as productive as projects costing less than $5 million. But they forgot to divide project output by project cost. The blog post summarizing the report was written by the director of an Institute, managing hundreds of millions of dollars worth of grants yearly, who DID NOT KNOW YOU NEED TO DIVIDE OUTPUT BY COST to compute productivity, because Institutes aren't incentivized to check on overall monetary efficiency.)
Most of the money in government contracting goes to pay for reputation. Big agencies prefer to award contracts of $50 million and up, because otherwise they have too many projects to manage. So some COTR has to award a $100 million contract. She's gonna award it to some big company, like Raytheon or IBM, because most big contracts fail, and she won't get fired if IBM fails, but she will get fired if Name-You-Don't-Recognize fails. BigCorp will try to line up a bunch of subcontractors to do the work. The competing subcontractors each line up famous experts who claim they'll support the work.
So with a big contract, you end up with a hierarchy, with BigCorp at the top, subcontracting corporations below them, and PIs at the bottom who manage the productive work. BigCorp's job is to vouch for the reliability of the subcontractors, for which they take about a 50% cut. Each subcontractor's job is mostly to vouch for the reliability of the PI, for which they take a more than 50% cut. The famous experts might get a 10-20% cut of the remaining money, usually do approximately nothing, and are being paid for the use of their names, like the famous people on a company's Board of Directors.
None of this is irrational, given the premise that projects must be big. Big projects fail so often that taking a 50% cut at each reputation level in order to put some credible reputation on the line is worth it.
Do the math, and you'll see that little of the money on big projects is left to do work. Whereas with small projects, there's more money left to do work, but most of it is put into making a cool demo (think the MIT Media Lab), and (rough guess) 90% of small government projects are killed or thrown away without anybody ever using them, either because they threatened someone's job, or because nobody really wanted them in the first place.
This was a very enlightening write-up, thank you.
Thank you. From the not selected book reviews you might like, "Scientific Freedom: Elixir of Civilization".
Where can I find the not selected book reviews?
Item 3. here,
https://astralcodexten.substack.com/p/open-thread-169
This was excellent. And a bit of a shame it's buried in the substack comments. I know at some point in future I'll fruitlessly look for "That ACX comment on science funding stuff".
This needs one or another of Scott's epistemic status notes, but anyway, here's a story for you: One of the "A.I. Winter" events happened after a period in which the Department of Defense had decided that Artificial Intelligence was a priority, but then professors at America's most venerable universities had perfected the art of getting DoD grant money by laying out mumbo-jumbo that promised the moon and delivered nothing. Somebody at one of the big private research labs (probably Bell Labs or IBM) eventually collected some convincing enough evidence that this shitshow was what was going on. DoD responded by just turning off the money spigot. Then for a decade or so everything in AI research got very quiet - though some smart people may have been laying the groundwork for the more productive research directions that eventually emerged later.
As one of those 1980s AI researchers, I think all of the AI researchers I knew were sincere, though a few had crackpot theories. Mostly, symbolic AI never worked out as well as people had hoped. Expert systems were hard to deploy because they were by definition designed to replace experts, and most fields with "experts" have regulations about who is and is not an expert. That turns out not to include computer programs.
The DoD had its own problem with AI. I won't claim this was endemic, but it did happen more than once: The military wanted automated "red teams" for training exercises, and would periodically award grants to people who had state-of-the-art (symbolic) AI systems to control the red team. Then the product would be deployed, and the red team had to be programmed by some Spec 3 technician who /might/ know how to program computers, and that went poorly. Eventually the military would give up and look for something simpler. When they got the simpler thing, it was too simple to be a good red team, and someone would say "We should use more AI!", and the cycle would begin again.
I'll even hop in and say it!
The utter lunacy of this book quite aside, flying cars are around the corner and we have development of greentech to thank: electric cars beget better batteries that can do the job.
can you expand on why better batteries are pivotal?
Not the person you responded to but the simple explanation is increasing energy density.
That's simply nonsense, best batteries have about 2 orders of magnitude lower energy density than gasoline.
I'm comparing newer, better batteries to older, worse batteries, not to gasoline.
How is that enabling flying cars? You can run them off gas (or other similar fuel) much easier.
Agreed, as Tom explained (much better) below me "Energy density. We could have built flying cars off gasoline a long time ago". We seem to have read Matej's question differently. I took their question to mean "why do we need better batteries before they can be used for flying cars", not "why are batteries are better than gasoline".
Check my substack, i have a podcast interview with an EVTOL battery expert that goes into detail. Cell Siders episode 8.
Energy density. We could have built flying cars off gasoline a long time ago, but they would have been extremely pollutive (helicopters get around two miles per gallon). Now, what exactly is the distinction between flying car, helicopter, and airplane is mostly just, like, the branding, but basically nobody is going to release something called a "flying car" and not have it be emission-less today, so that means batteries.
Energy density matters because a flying car starts to run into a rocket equation kind of problem, where the more range you want, the more batteries you need to carry, which means a greater mass, which decreases your range. This means that more batteries doesn't really solve your problem. So to get a longer range flying car, you need higher energy density batteries.
Is there a plausible pathway towards batteries with vastly better energy density?
Oh absolutely, the theoretical limits could be insanely better than we have now. But this is one of those "it's actually just really hard" problems.
For lithium ion, we're near the limit. Tesla's 4680 batteries are probably going to be within 30% of the theoretical limit for lithium ion chemistries. I don't know that number for sure but it's what I recall talking with battery experts I work with.
Solid state chemistries can theoretically do a lot better but there's a reason they're taking a very long time. The history of the battery industry is kind of like fusion power, always 10 years away or something. You should ignore all news that you hear about a "battery tech breakthrough" until it's in a consumer product that you could theoretically buy.
Side point - if we can work from home and in virtual reality and do not need to commute anywhere (for even 20-30% of the population)...do we need flying cars anymore? Might the technological innovation which allows us to work virtually reduce traffic and the need to live near our workplaces or have any kind of commute...might that innovation outpace the battery technology needed for flight? Flying with an electric car requires clearing quite a high bar of innovation and production which could still be 15+ years away from a cheap consumer model. While the battery technology for ground transport is essentially already or very soon to be a consumer grade product and getting better and cheaper over time.
It is sort of like the strange move for 'smart thermostats' which are a short term stopgap until we can build better houses such as passivehaus designs which passively self regulate with a small up front investment to reduce running costs of a building dramatically over its lifespan.
A flying car is cool and fun, but I don't see it being more than a toy like an ATV for Jetski for a very long time in the 30+ year range until a person at home might 'call for a flying car automated taxi' to take them somewhere unless that house is a large mansion - in which case they can just afford to hire a human helicopter pilot right now.
I would say that flying cars are finally becoming available today not just because of dramatically improved battery technology, but also because of dramatically improved control software (and the hardware to run it, of course).
The computer science problem is trivial compared with the chemical engineering problem.
I think they are both quite difficult. Sure, it's possible to design the optimal control system on paper; but we have only recently gained the power to actually implement one that is small enough to fit into a flying car, and fast/smart enough to be usable by the average person (as opposed to a trained helicopter pilot). Doing so required massive advances in computer hardware as well as machine learning.
Not really. The reason that the computer science problem appears to be being solved isn't that it was easy to do, but that computers are getting so fast and so parallel that we don't have to shrink away from brute forcing it so much. Really this goes for a lot of things. Most changes in software stacks today aren't to enable new capability but are instead to make the work of software developers slightly less tedious, and to allow less-smart people to successfully accomplish software development tasks, at the cost of program performance (since we have increasing performance to spare).
Problem: we cannot get enough batteries to run 'green' - namely, wind, solar - energy. It can't be done.
If you're serious about stopping carbon emissions, back nuclear.
It's so aggravating. Like a level 1 engineering analysis would say "oh, renewables are intermittent? Well a large capacitor (battery) can solve that problem."
A level 2 analysis is like "well, do we have enough batteries?" The automotive industry is making incredible strides here. The problem is that all of those batteries are going to be used by the automotive industry. Vehicle-2-grid is not actually a good idea, people tend to want their car to be charged when they go to use it. Battery production MIGHT just scale extremely well now that demand is of no consequence, but it's just weird to see so many people (who are acting out of a very respectable abundance of caution over climate change) betting the future on the assumption that battery production will scale like semiconductors did. Do I hope that's going to happen? Absolutely, in fact my career is trying to make that happen.
But we should build nukes.
What you or I "back" is irrelevant. Fission nuclear is too expensive and too slow and a huge expansion simply isn't going to happen. Internet debates are one thing, but the brute economics and politics are another. Nobody wants to deal with the many tricky problems associated with radioactive waste and you're not going to force them to. It's time to move on.
I'm strangely optimistic about fusion, which I think is a reasonable possibility within the decade. But fission is EOL.
D-T fusion still makes radioactive waste, although you have better control over what it is (there's a rather long list of elements you can't use in a wasteless fusion reactor because of (n,y) or (n,p) creating something with an annoying half-life - "annoying", here, being "too long to just put it in a pond for a couple of years, short enough to make significant radiation").
"Fission nuclear is too expensive and too slow and a huge expansion simply isn't going to happen."
Did you read the review at all? It had giant sections about just why fission is expensive and it had nothing to do with the underlying technology.
Also I was interested in fusion (especially after hearing about MIT's SPARC) but then I read a bit more about energy densities and... they're dismal. Back to fusion.
It has to do with regulation forcing companies to pay the costs that they would otherwise externalize, which makes the projects uneconomic. The risk of the underlying technology is what makes it so expensive. The only thing the book has to add is radiation denialism, which is not super useful.
Batteries are well and good, but how are these things going to get around?
My limited experience with VTOL drones of appropriate size to carry humans makes me highly skeptical that we have technology to automate key system functions (primarily take-off and landing as well as collision avoidance) sufficiently to where people would be able to drive them with an equivalent level of training that we give drivers.
For example: https://news.usni.org/2021/04/27/mq-8b-fire-scout-crashes-into-littoral-combat-ship-uss-charleston-on-deployment
Landing on a ship (which is always moving) is much more difficult than landing on the ground (which doesn't move). This is like saying that we don't allow cars to dock with trucks in motion so we shouldn't allow drivers to be licensed at all.
While landing on ships can be harder than landing on a fixed runway, that isn't the problem here:
https://www.navytimes.com/news/your-navy/2020/11/18/fire-scout-drone-crashes-at-california-base/
hmm. the book’s framing would say that contemporary battery/energy storage tech is finally on the level of 40s-era petrol tech, thus limited flying cars with likely similar range/speed/cost as a pitcairn autogyro.
We shouldn't have flying cars until people are willing to require that only computers may fly them. There isn't enough airspace above a city for humans to avoid crashing into each other.
"The public is wrongly terrified of nuclear energy, but they shouldn’t be. Radiation killed 0 people at Fukishima"
You really lost me with the Fukishima minimization. As if deaths at the time of the incident are the only relevant concern. How much land exactly is contaminated forever?
Well, the worse places in the Fukushima area are giving out 90 mS a year right now. That's about twice what we allow a radiation worker to be exposed to every year and given that these limts are conservative it's not clear that there's anything wrong with living there now. That's slightly smaller than the smallest yearly dose linked to cancer but given that we let people smoke, have wood stoves in cities, etc we should probably let people live in Fukushima right now. I wouldn't move there until levels are down to the 10mS/year where Ramsar and we know the people there aren't getting more cancer than usual. But that's most of the Fukushima exclusion zone.
That is no small thing. And who is to say the next nuclear disaster doesn't create a far worse contamination problem. For a looong long time. To not even mention that aspect of the issue in a blithe dismissal of concerns about nuclear energy is just crazy to me. This is a real pattern in the arguments of nuclear apologists I've noticed, and doesn't exactly inspire confidence.
I don't mean to come across as blithe. I wouldn't endorse someone smoking even 1 cigarette each day, eating bacon for breakfast every day, or other risks in roughly the same range. I wouldn't live in Fukushima right now myself. But it's still not a huge risk and if other people have more tolerance for risk than I do I think that they should be allowed to smoke or eat bacon or live in Fukushima. And it isn't forever. Different isotopes decay at different rates. One with a half-life of 1 year is roughly 10 times more radioactive than one with a half-life of 10 years so the most radioactive isotopes tend to decay fastest, though the least radioactive ones will be with us for a long time. Still, I think that in a few decades when the worst parts of Fukushima are merely as radioactive as Denver it will be unreasonable to worry about the radiation levels and at that point I will actually move my attitude to blithe dismissal.
Yes, I meant the author was being blithe (I assume that is not you). Your clarification was quite helpful, thanks. My use of "forever" was of course not literal, but along the lines of "ruined for human use for decades/centuries"
Yes, that's certainly a serious cost to the accident. But it's also the case that many other kinds of power generation have even higher environmental costs, on average. For my part I'd say that once we finish getting rid of every fossil fuel power plant we should start getting rid of fission but I'd hold off on it until then and I wish we'd built more in the past before solar became cheap.
I'm certainly willing to entertain a cost/benefit comparison of nuclear with other options. That is why I find it frustrating when I am presented with what seems to be a deliberately incomplete or even misleading one such as a we find in this review.
Who can say there are no aliens walking around the Earth? Can you prove there are none? Should we formulate policy based on that premise?
The fact that you are calling people apologists is not a sign of good faith intent to weigh the risks here. All electricity generation comes with risks. Solar panel manufacturing involves toxic chemicals which have a half-life of infinity years (non-radioactive chemicals don't decay) The silicon sand refinement process releases tiny silica particulate in the air which leads to silicosis, responsible for tens of thousands of deaths a year. More people will die this year falling off a rooftop installing solar panels than will die building nuclear power plants. Are you an outspoken advocate against solar panel manufacturing?
You shouldn't be. Energy, on net, saves lives and improves global welfare. Why would anybody just be a nuclear apologist? Have you considered that so-called nuclear apologists genuinely believe that it's the safest form of energy generation?
Pretending that day-of-meltdown causalities are the-only relevant metric of the nuclear power risk is ridiculous to the point that anyone who employs such a ridiculous metric can fairly be called an apologist in good faith.
I think a lot hinges on you saying the phrase "only relevant metric." Those are words that you have attributed to the author, and yet the author did not write, yet you are using to cast aspersions over his intentions and all else who may just genuinely think this is a good policy decision. I urge you to have greater patience and a more open mind.
That is the only metric he or she uses, to the exclusion of all the other relevant metrics I mentioned. It was their decision to make that strange choice, not mine just because I pointed it out.
You keep mentioning "day-of-meltdown" casualties. How big a deal do you think the other casualties were?
The point is casualties aren't the only relevant metric - there is also the issue of land contamination
To be uncharitably pedantic, exactly zero land is contaminated "forever" - the radioactive contaminants will, eventually, decay to undetectable levels.
But if you're asking the more reasonable question of "how much land is contaminated to the point of uselessness on a timescale longer than a couple decades"...well, it honestly doesn't seem like there's very much at all outside of the plant itself. The exclusion zone, such as it is, has pretty much shrunk to a couple towns in the immediate vicinity of the plant, and that's *with* the Japanese government's conservative-bordering-on-paranoid safety regulations.
It's also important to note the discharge of contaminated wastewater into the ocean adjacent to Fukushima, which can concentrate in sea animals, including food animals.
The question is Trutium which is constantly created by cosmic rays in the upper atmosphere, has a half-life of 12 years, and is NOT concentrated by sea animals.
Umm, no: A study published in the journal Science in August 2020 found traces of several other radioactive isotopes in the Fukushima wastewater, many of which take much longer to decay than tritium.
Some of that radioactive material may have already made its way into local wildlife; In February, Japanese media reported that shipments of rockfish were halted after a sample caught near Fukushima was found to contain unsafe levels of radioactive cesium.
Assuming you're talking about this study (https://science.sciencemag.org/content/sci/369/6504/621.full.pdf), the amount of other radioisotopes is utterly negligible. There's 500,000 Bq/liter of tritium radioactivity and around 10 Bq/liter for the other isotopes. For comparison, your body contains 8000 Bq of radioactivity, or about 100 Bq/liter. That is, if you took the tritium out of the wastewater and drank the rest, you'd probably lower the average radioactivity of your body.
As an aside, the 500,000 Bq/liter of tritium becomes 100 Bq/liter if you dilute it by a factor of 5000. Dumping it into the ocean is a good way to do that.
seafood: https://pubmed.ncbi.nlm.nih.gov/22863967/
Tritium
I don't want to dismiss the concern of waste leakage into the ocean...but it's not established that the waste leakage into the ocean is going to cause significant ecological damage. I know that may sound like a crazy claim to make, but radiation is extremely not intuitive.
Most scientists expected Pripyat (outside Chernobyl) to be some kind of toxic, uninhabitable wasteland for a hundred years. Instead, wildlife there is thriving beyond anybody's expectation. Vegetation growth even exceeds what we would expect if there was no meltdown. It's quite possible that Chernobyl-like radiation levels are harmful to large mammals like humans but good for smaller life forms like bacteria and plants. Nature is antifragile.
Now that is absolutely not to say that we should just go dumping nuclear waste into the ocean because it might be good for it, but it's not obviously a catastrophe.
You're talking about radiation hormesis, which has a substantial literature:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2477686/
I don't think so. Even with a standard "linear no threshold" model of radiation damage, the chance that wildlife could notice the difference is ~zero; I expect a flourishing of plants and wildlife purely as a result of the missing humans.
Specifically, it looks like a typical radiation exposure in Pripyat is 0.7 µSv/hr or 6 mSv/year as of 2009 (without any attempt at environmental cleanup AFAIK). For comparison, "In Europe, average natural background exposure by country ranges from under 2mSv annually in the United Kingdom to more than 7mSv annually in Finland." http://www.chernobylgallery.com/chernobyl-disaster/radiation-levels/
I'm sure human scientists have methods sensitive enough to notice a tiny increase in cancer rates due to exposures substantially below 100 mSv — but wildlife is more concerned with how to find its next meal. (Btw, this figure of 7 mSv is the the highest I've ever heard and, if true, ought to make a good place to do a study on radiation risks.)
Regarding your link: beware the man of one study, but especially beware the study by one man! I don't know how best to read the radiation-risk debate, but it's clear there *is* a debate, i.e. evidence for the effects of low-dose radiation seems inconclusive (apart from universal agreement that the effect, whatever it is, is small), and making this especially hard is the near-impossibility of finding healthy human subjects exposed to intermediate levels of radiation (10 to 100 mSv per year). An illustrative paragraph from a study shows this:
"Most of the previous ecological studies investigating associations between childhood leukaemia and naturally occurring sources of ionising radiation have found positive associations for radon (13-15) while for gamma radiation and cosmic rays results have been inconsistent (16-22). Early case-control studies of the association between natural sources of radiation and childhood leukaemia were underpowered and have reported mixed results (23-26). The largest of these, the UK Childhood Cancer Study, included over 2000 cases of childhood cancer and reported weak evidence of a negative association between childhood leukaemia and measured radon concentrations (25) but no evidence of an association with measured gamma dose rates (26). However, the proportion of eligible subjects participating in the measurements was low and varied by socio-economic status. Because exposure to these sources is ubiquitous and variation in cumulative doses received by children of similar age are small, large sample sizes are needed to detect the small predicted risk. Given the rarity of childhood cancer, the only way to achieve such sample sizes is by combining data over long periods of systematic cancer registration." https://boris.unibe.ch/135621/15/Mazzei_JRadiolProt_2019_AAM.pdf
In the same paragraph we see childhood leukaemia associated positively and (in the largest study) negatively with radiation from radon. Also mentioned is the difficulty of measuring the effect (it's especially difficult because cancer is normally caused by something other than radiation, so the effect we're looking for could easily be swamped by other things that affect cancer rates.)
In general this is the problem with the "precautionary principle": if something is banned for being potentially harmful, it becomes extremely difficult to clearly show that it IS harmful.
My understanding on the effect of the relatively low dose of radiation outside Chernobyl on plants and animal is that they are thriving despite clear negative effects of radiation (increase of the ferquency of abnormality, decreased growth ring in trees, etc), because the absence of human pressure more than compensate the effect of radiation. But given our current knowledge, it seems to me extremely unlikely that radiation is good for any known living organism.
"But given our current knowledge, it seems to me extremely unlikely that radiation is good for any known living organism." -- A study here suggests that some amount of radiation isn't just good, it's required: https://www.pbs.org/wgbh/nova/article/life-without-radiation/
At some point, you have to say what form of radiation you're talking about. You know that sunlight is radiation, right? I assume you agree that sunlight is good for life.
Radio waves are radiation. If you're using wireless internet, you're using radiation. Sound is radiation.
As with anything else in science, numbers matter, and what amount of radiation you're talking about is crucial. But it's clear that the idea that all radiation in any amount is bad for life is just ignorant.
It was perfectly clear what kinds of radiation Emma was talking about.
It's not clear to me, then: what kinds?
Thank you, I also thought it was obvious from the context.
"At some point, you have to say what form of radiation you're talking about. "
I thought it was obvious that when I said "the relatively low dose of radiation outside Chernobyl" I meant α and β radiation due to radioactive contaminanation, not sunlight!
"A study here suggests that some amount of radiation isn't just good, it's required: https://www.pbs.org/wgbh/nova/article/life-without-radiation"
This certainly does not demonstrate that (hard!) radiation are necessary. The only well estblished result is that if you hit bacteria with a high dose of radiation they survive better if they had previous exposure to a low dose. That is a far cry from some (hard!) radiation is necessary for life. α particles really do a lot of damage to complex molecules, it seems extremely unlikely that this damage can become necessary.
" But it's clear that the idea that all radiation in any amount is bad for life is just ignorant." Obviously true if you include lower energy radiation. But I maintain that "given our current knowledge, it seems to me extremely unlikely that (hard!) radiation is good for any known living organism".
I didn't know if you meant just alpha and beta, or also gamma radiation.
I think the Nova article says (or implies) that the scientists restricted cosmic rays and maybe natural radiation from heavy isotopes. Cosmic rays are protons and other nuclei (like an alpha particle, I guess); I don't know what the other consists of, but probably neutrons, and alpha and beta particles.
I was wrong to say this radiation is necessary for life, but the experiment suggests that hard radiation is beneficial to life. That contradicts your statement that radiation is not good for any known living organism. Am I missing something here?
Aside: I believe that radiation in some form, probably electromagnetic, is required for all life, because we need it to get mutations in DNA, and we need mutations in DNA to get evolution. It could be that DNA polymerase could've evolved to be more error-prone to make up for less radiation.
Whoa, so the ecological impact of the worst civilian nuclear disaster in history is exceeded by the prior ecological impact of humans simply living there? That definitely lowers my estimate of the ecological consequences of civilian nuclear power.
I am not sure. The impact of humans "simply living there" on plants and animals is huge! Imagine an alien civilization landing on earth and colonizing it for their own purpose, whithout noticing or caring about us, the impact on us would probably be extremely high.
Sure, it's a real paradise there in Pripyat/Tschernobyl ...
Couple questions still remain:
We are the European nations paying Ukraine for building a new "Sarcophagus" on top of the old one?
Why are European countries still routinely measuring radiation levels in food (mushrooms, boar, deer, vegetables, ...) and destroying those exceeding radiation thresholds? (https://www.lgl.bayern.de/lebensmittel/chemie/kontaminanten/radioaktivitaet/ue_2018_radioaktivitaet.htm)
The findings are gradually decreasing (it's still only 45 years ago) and was probably just bad luck that it is only 1,400 km away (800 miles might be close for US citizens but for Europeans there are 4 countries in between)
For the same reason they are banning/overregulating GMO - they are pretty paranoid, European nations.
Well, none of it is contaminated "forever," but the current restrictions on residence apparently apply to about 371 km^2. You may want to bear in mind that industrial pollution routeinly affects much larger swathes of living area, and routinely cases far more deaths. For example, China's Huai River policy encouraging the use of coal north of the river has been argued to cut 5y of life expectancy off the 500 million people who live there:
https://www.pnas.org/content/110/32/12936
Industry pollutes, and if we choose to live technological lives we always run health risks (particularly when industrial plant is combined with things like earthquakes and tsunamis). It's not possible to run zero risks without living in caves and hunting antelopes with spears. So perhaps the focus should be on *relative* risks, and trade-offs. Being blinding Polyanna enthusiastic about nuclear power is dumb. But so is being "terrified."
I thought it was clear that I wasn't using "forever" literally. As I said elsewhere, I'm certainly willing to entertain a cost/benefit comparison of nuclear with other options. That is why I find it frustrating when I am presented with what seems to be a deliberately incomplete or even misleading one such as a we find in this review.
Well, maybe try not to respond to oversimplification with an oversimplified critique?
My critique was not oversimplified. I just used a term loosely in such a way that the meaning was still obvious. Neither did the author oversimplify. They lied by omission. So maybe don't misrepresent the whole exchange?
I don't agree that your oversimplification was obviously not literal, while the author choosing deaths as the comparison criterion is a lie. If you're not willing to be charitable to the author, you shouldn't expect others to be charitable to you.
Really? You thought that by "forever" I meant that when the sun turns into a black hole Fukushima will still be irradiated? Come on. That could not be more different than choosing a deeply misleading metric.
I think someone should enforce a dictum:
EVERY DISCUSSION OF RADIATION RISKS MUST INCLUDE A DIRECT COMPARISON TO COAL!
To pick just one factor: land made inhabitable. And mountain top removal?
And don't forget that some of the pollutants associated with burning coal are themselves radioactive.
Speaking of which, the "meltdown world" in which every single nuclear reactor suffers a Fukushima-style meltdown seems clearly safer than the coal-heavy world we actually live in: https://www.reddit.com/r/nuclear/comments/jtm6hm/how_bad_is_meltdown_world/
I had the same reaction to the Chernobyl minimization. If only 43 people had died, that'd be one thing, but glossing over the health impacts of the disaster on hundreds, if not thousands, of other people seems dishonest to me. I'm pro-nuclear power, and while I think when it fails it fails due to human error and not scientific error, the costs of failure really are very, very high.
You really lost me with the "forever" bit. Yes, yes, you didn't mean it literally. But either this is irrational blind fear, or it's a fundamentally quantitative problem. If it's irrational blind fear, then there's nothing for it but to try and point the stupid people somewhere else. If it's a fundamentally quantitative problem, then you don't address it by saying "how much?" and then introducing a spurious infinity term for rhetorical effect.
About 300 square kilometers near Fukushima have been turned into a de facto nature preserve for the next few decades. Turns out nuclear-power-plant levels of radioactive contamination are a pretty good deal for most animals, because it poses relatively low risk over their natural lifespan but is enough to drive off their chief natural predator and habitat-paver-over.
An analysis I did on SSC or DSL a year or so ago, suggests that if the human race generated 100% of its electric power using nuclear power plants built to 20th-century standards (but no new atom bomb factories), the rotating nuclear nature preserve would at any point be equivalent to IIRC Macedonia or Haiti in size, with individual zones rotating in and out every fifty years or so. Or, if you prefer human habitats to the wildlife sort, you can build your power plants to more modern designs and cut that down.
Compare and contrast to the amount of land we'd have to turn into e.g. solar farms, where the sun never shines and nothing grows. No, it's not enough to just put solar panels on the roofs of existing buildings, and we're not even going to talk about solar roads.
I'm open to the possibility that radiation from meltdown concerns can be addressed. But the author didn't even acknowledge them, let alone make a case as you have here
If somebody proposes to do something useful, like providing abundant cheap clean energy, then the burden of making the case ought to fall on the person saying "no you're not allowed to do that because it's too dangerous". You stepped up to the plate and completely failed to make that case, offering only a rhetorical question including an objectively false assumption.
And that's pretty much par for the course in this business. Lots of people believe that they know that everybody knows that nuclear power is "unacceptably dangerous" and that simply alluding to that "fact" is a slam-dunk win. Almost nobody actually makes the case.
That is not at all what I said, you are dancing with a strawman.
I'd wager it's a lot less land than will be rendered arid by global warming as a result of not having enough nuclear reactors.
I think the real reason for the turn against nuclear was that the public was used to thinking of fallout in terms of being downwind of thermonuclear groundbursts, where it meant dying puking your guts out in hours or days rather than a theoretical increase in your risks of cancer. Order of magnitude comparisons are made hard by the fact that radiation is invisible. I wrote about the matter [on my blog](http://hopefullyintersting.blogspot.com/2019/06/sometimes-you-need-new-word.html) at more length. I already see a change of attitude between the generation who grew up in the shadow of the mushroom cloud die off and as those of us who grew up with reactor meltdows as our image of fallout so I'm optimistic about that aspect of the future.
At the core of The Green Religion is something I call the environmentalist's habitat paradox: if you really like the environment, a natural first-order desire would be to live in a cottage deeply secluded in nature, far from civilization. But this is either unscalable (and therefore antisocial) as you cannot allow too many others to indulge in the same lifestyle, or you DO proselytize this lifestyle and it becomes environmentally catastrophic. The paradox is that if you love the environment (as in truly want to protect it), you must live in a city.
Eco-pragmatism needs better branding. We need extremely lush, literally-covered-in-plants cities powered by cheap nuclear.
Part of the branding problem is lack of good definitions that leads many to group serious thought on the cost/benefit of various levels of protection for the long standing state of the environment, in with the dumbest environmental protesters--chained to trees while using their iPhone to tweet about how a plant has the same moral value as a human. Worse any attempt at serious problem solving or compromise in public debate very quickly devolves into extremes shouting at each other because government's control via regulation makes things winner-take-all at the entrepreneurial investor level and a status signal of tribal politics and virtue at the level of regular voters.
Well I think it's more the same problem which exists with everything else (and why I'm a rationalist) which is that people choose the right answer with their feelings. If you're someone that really likes the aesthetics of environmentalism, well the last place you want to live is a concrete jungle.
Enter Derek Jensen, an anti-civilization advocate who lives in the remote wilderness with bears. He thinks civilization, cities, and industrialization were horrible ideas, and everyone should give them up and go live "more natural lifestyles" where they hang out with bears in the woods. Nevermind that there are probably less than 300,000 bears alive in the world and so, at best, each bear would have to befriend tens of thousands of people. Not only would this certainly not be good for the bears (not sure what the Dunbar number is for bears but I think it's safe to assume it's less than ten thousand), but Jensen himself would almost certainly not want to share his bear friends with ten thousand other people. I know countless environmentalists/activists who would say that Jensen's lifestyle is idyllic.
I'm a person that just naturally has to think through what the natural consequences of things are. Apparently most people don't do that, they just think "I'd like to live in the woods." and that's the end of any kind of consequential analysis. And so to me, these ideologies, when people say "we should all abandon civilization and live in the woods" somewhere in there, either nature is just absolutely destroyed beyond the likes of which we have ever seen (which doesn't sound like their goal), or colossal numbers of humans vanish somehow. And so I am extremely skeptical of these ideologies.
Looking at some very rough numbers, there is enough forest in the world for everyone to live on a little more than one acres of forest. Maybe we could live in small communities of 50-100 people on 50-100 acres, and maybe some people would be happier living on the savannah or other open grasslands, freeing up space for the rest of us to have 1.5-2 acres or whatever.
Some Googling did not get me an answer to how much forested land a person needs to live in a sustainable way, but I know it's a lot more than an acre. That's especially true for the many many people who would have to live in Siberia and other inhospitable forests, where heating fuel for the winters would be a big issue.
On the bright side, the billions of deaths in that first year would certainly help with the "colossal numbers of humans vanish somehow" problem!
I'm quite sympathetic to Jensen's goal, but I think there may be much more sustainable options available. For instance, I choose to live in an area where there are so many forests around that I can literally see at least one from any vantage point within 50 miles. There's several hundred thousand people in that range, mostly clustered in a series of small towns and surrounded by farmland.
I don't think it's intellectually responsible to be sympathetic to Jensen's goals. Climate change and sustainability have always been human problems. We have no direct business case for reducing emissions, but we know we have to, can we make ourselves do it? We can hardly get people to stop feeding wild deer, raccoons, or ducks. Somehow we're supposed to expand the average person's sphere of altruistic concern to contain the ecosystems of all of the "forest[s] of the world" and give them a scientifically rigorous understanding of how to achieve that?
This is a fantasy, perhaps worth thinking about, debating, and taking seriously back in the 1960s when it emerged in force, but given humanity's track record since then we have no reason to believe such a psychological/sociological/educational stunt can be performed, and it's frankly dangerous to continue entertaining it. Green hippies are in their bubble, waiting for everyone to come join the drum circle, while mainland America is still rolling coal and ICEing EV charging stations.
To be sympathetic is not to actually support. I'm "sympathetic" in the sense that most humans (and incidentally also environmentalists) have been misled by their ignoring the numbers in favor of policy ideas based on feelings, and who can blame them for just doing the usual human thing?
Now, when I point out the numbers to someone and they're like "you red tribe bastard!" and I'm like "I'm not red tribe" and they're like "whatever I'm outta here", that's when my sympathy dries up.
Yes. We don't have to live in cities of a million persons though. We can live in 100 dense pedestrian pockets connected by silent inconspicuous hyperloop.
I don't think that it's obvious the ideal size is 10,000 people and afterwards you see diminishing returns to scale for agglomeration. Even if it were, a 10 billion human planet would require one million such small cities. A quick google search tells me that there are currently 10,000 "cities" worldwide. We should absolutely not want to find 100x as many locations around the world for more cities. That would necessarily mean fewer nature reservations.
Call it what you want, but not green.
Pockets of 10,000 persons.
It seems to me that we've moved on to trickier problems to solve that are mostly based on coordination rather than simply maximizing consumption. For instance, imagine a 40-story office building with about 2,000 workers where each one commuted by flying car. How many landing strips do you need? Remember that unlike parking spots, you can't stack them -- each one needs to be open to the sky -- and you probably need *minimum* five minutes' clearance between cars. If everyone arrived between 8am and 9am, that means that each landing strip can serve a dozen employees, so you'd need 166 total just for this one building. From a perspective of land use and of time spent getting from your parking spot to your destination, this just sounds terrible. So we should be happy we don't have flying cars, because the societal equilibrium they'd put us in would be terrible.
One could argue that this particular problem is specific to transportation technologies, but social media has amply demonstrated that it's possible for many technologies to lead to bad equilibrium outcomes.
All this is not to say that techological stagnation isn't a problem, but when people talk *specifically* about flying cars, I discount their arguments specifically for the above reasons.
I'm not really sympathetic to this kind of argument because you could probably make similar sounding arguments about a lot of technologies we take for granted today.
Imagine if you had to propose making, say, the national power grid in todays climate if it didn't already exist. You are going to make a giant country spanning grid of copper wires that make a circuit into every single household? The wires themselves are ugly and you want to have them on every street? And these wires are dangerous if anyone fiddles with them or damges them. And if they are burried then and someone were to hit some burried ones with a spade? Wouldn't that hurt them? And then you have the wiring in the walls of your wooden houses? Wouldn't that be a fire risk? And the sockets themselves have a potentially deadly voltage just sitting there on the wall where any kid could stick a fork and kill themselves.
How could you ever possibly scale this?
If something is useful on an individual level, then it will be slowly adopted and then we will find a way to scale it later.
I don't think the argument I'm making is about "how do you scale this" -- spatially inefficient transportation technologies have negative returns to scale. If you have the first car? Sure there aren't any gas stations, but it's still awesome (especially compared to the first telephone) because you can get anywhere 5x faster than anyone else. But if everyone has a car, it clearly makes your car go slower.
So in fact, I would make exactly this argument about technology (the car) that we do have today. How come scientists haven't solved traffic? It's because road space is a special kind of resource where the richer you get, the scarcer it gets.
If you are saying that your argument about Flying Cars applies equally to regular cars, then how do you explain the fact that Cars are widely in use today?
Doesn't that imply that Flying Cars would still be in use widely even with the challenges that you forsee?
You mentioned a bad equalibrium before and compared it to social media, but that implies that regular cars are in a bad equalibrium too and that we would be better off without them? I don't think I can agree with that.
I do think that regular cars in the way currently used in the US are a bad equilibrium! If you look at Europe or Asia, the number of car trips per capita is about half of what's in the US, and it's not because they're behind in car technology. Similarly, I don't think that making cars better technologically solves the problems we have with them. Teslas can accelerate much faster than internal-combustion cars, but do they actually get you where you're going any faster?
Yeah, the way cars are used in the US seems like a nightmare for many French people...
Counterargument: The most appealing argument for flying cars was reduced reliance on public infrastructure, highways. Also flying around buildings is dangerous. Restrict flighted travel to outside cities. Continue using urban parking garages designed for tired vehicles.
Heinlein's "The Puppet Masters" has a lot on how flying cars would be a boon in the rural Midwest, but his seem dangerous in NYC or even low rise DC.
Honestly, this is what we should have said with ground-based cars. Driving around people is dangers. Restrict >20mph travel to outside cities. Continue using urban transport using small and/or dense vehicles like bikes, feet, and streetcars.
And the noise!
Stand back a few paces from this discussion, and ask yourself honestly: Do you really believe there will ever be ten million flying cars buzzing around the US?
For me is just ludicrous to imagine that.
For comparison, can you imagine millions of persons moving around the US at 300 mph in silent hyperloops?
I sure can.
I don't think the argument from personal incredulity is especially strong here; I find both scenarios equally easy/hard to imagine.
> so you'd need 166 total just for this one building. From a perspective of land use and of time spent getting from your parking spot to your destination, this just sounds terrible
I mean, if you had a 2000-person building where everyone commutes by non-flying car then you'd need a pretty big parking structure too. In order to just maintain the same 166-space footprint that you already think is too big, you'd need a twelve-storey parking structure, which is very large.
From a land use point of view, flying cars (idealised flying cars anyway) would allow us to reclaim all the space currently devoted to streets and roads.
Also if we had (idealised) flying cars we probably wouldn't bother having 40-storey office buildings anyway, our cities would be less dense because you could travel greater distances with ease.
One nit to pick: general (private) aviation was not done to death by regulation as much it was by product liability torts. Lawyers somehow got extremely good at convincing juries that the aviation accident equivalent of "16 year old who just got their driver's license buys a Ferrari, drives it at 120mph on a twisting mountain road at night in a rainstorm, and predictably winds up dead after careening off a cliff" was somehow Ferrari's fault, and awarding the idiot's family millions of dollars in damages. Given that Ferraris are already a low volume market, it doesn't take too many such lawsuits to drive the cost to buy a new one through the stratosphere.
(The actual scenario would be that a rich retired athlete or businessman would buy an expensive, complex high performance airplane, do the minimum amount of training required, then fly off into bad weather in unfamiliar areas - which they should have known not to do if they had been paying attention in flight school - and run into a mountain, or building, or just plain crash. And their widow would then sue the airplane manufacturer, and usually win.)
In 1994, Congress passed https://en.wikipedia.org/wiki/General_Aviation_Revitalization_Act, which was supposed to fix this. Lawyers just switched targets from the manufacturers to the mechanics who work on planes, with the predictable result that airframe & powerplant mechanics refuse to sign off on an airplane's annual inspection unless everything is perfect, increasing cost of ownership for private airplanes.
All that being said, as a private pilot, the idea of having to share the skies with several orders of magnitude more aircraft, being flown by the equivalent of your average automobile driver who can't be bothered to use their turn signal or put down their phone while driving, is terrifying.
I've read quite a few reports of amateur pilots crashing their planes here in Ireland. These are guys who have their own little plane, are experienced, often are going out for a routine short flight maybe taking friend/family along as passengers, and then something goes wrong and there's a fatal crash. So it's not even at the level of "dumb 16 year old".
We've come to accept the level of road accidents and we still try to bring them down. Cars in the sky will have a much, much tougher barrier to acceptance.
To make matters worse, I don't know how it is in Ireland, but in the US the FAA certification process for a commercially sold aircraft is very expensive, even for a small two-seater. It's often more than the plane would be otherwise. So a lot of small aircraft manufacturers don't bother selling assembled aircraft, they instead sell the parts as kits and hobbyists build them themselves.
My wife's uncle was a professional commercial airline pilot who in his retirement flew a Piper Cub. Despite all his experience, he crashed the thing not once but *twice* and nearly died both times.
Also all the celebrities who have some flying experience (certainly more formal training than the average motorist has) who crash and die.
If there's one unnecessary risk I've learned, it's to never set foot in a private aircraft.
> All that being said, as a private pilot, the idea of having to share the skies with several orders of magnitude more aircraft, being flown by the equivalent of your average automobile driver who can't be bothered to use their turn signal or put down their phone while driving, is terrifying.
Right. I've witnessed two minor accidents in the last two weeks, both completely avoidable and the result of obviously reckless driving by the person who caused the accident. I'm very skeptical of an environment where the same morons are flying, meaning every one of those stupid reckless accidents becomes very likely fatal and, if it occurs over populated land, potentially a mass casualty incident.
Then there's maintenance... I heard a boom several years ago while working from home - it was a small plane crashing into an occupied house in my neighborhood due to a mechanical failure. The occupants happened to be on the other side of the house, or it would have been a 4-fatality incident due a maintenance oopsie.
I'm sympathetic to the idea that regulation has hampered technological progress, but the negatives of this are (and I understand, yes, can only be) argued from the perspective of our current world.
What I mean is, flying cars have all sorts of logistical challenges associated with them related to infrastructure and training and higher possibilities of disaster if an accident were to occur in a densely populated place, where most of the cars would be. I imagine if the government had been hands off and let them develop, they still wouldn't be common. More of a novelty if anything. It just doesn't seem feasible in an urban area, and probably not desirable from a livability standpoint.
Nuclear, I agree, was a massive mistake to oppose on the basis of climate change, and probably would have gone differently if it were invented today. But we might not feel that way if we had built hundreds or thousands of additional reactors, and one horrible disaster had taken place that killed hundreds of thousands or millions of people. I understand that isn't likely, but we know it to be possible. Current energy sources just don't have that doomsday potential, and I think humans are willing to accept lots of incremental damage but have a strong aversion to potential apocalyptic accidents.
I sympathize with the thesis. We could be living in a very different world. But it's easy to speculate about the positives and give your alternate history a rosier outcome that it might have achieved in reality.
> but we know it to be possible
No, we don't. The idea of a nuclear plant melt down killing thousand of persons is purely speculative.
You could perhaps argue that there is a one in a million chance that operating 1000 nuclear power plants for 100 years would result in an accident that killed 1000 persons. So, yeah, "it's possible". But this kind of logic proves nothing.
I stopped fearing nuclear power plant accidents when I learned that the Chernobyl Exclusion Zone is now a thriving nature preserve.
If you know how a nuclear plant works you know it is certainly possible. Highly unlikely, as I stated, but from a purely physics-based standpoint it certainly could happen.
Could you explain how a nuclear power plant accident could kill millions of people?
I know how a US nuclear reactor works: a controlled fission reaction produce heat which is transferred to pressurized normal water which boils water to make steam which drives a turbine which produces electricity.
Now explain how an accident would kill 1000 persons.
For US reactors, consequence analyses show that even in the case of a meltdown and large release of radiation to the environment (like Fukushima), the average number of expected prompt fatalities is zero. The high end of the uncertainty range is tens of prompt fatalities. At the high end, latent cancer fatalities can run into the low thousands, but this is based on the Linear No Threshold model (which few experts find plausible) and most of those fatalities are from people moving back into the area after an evacuation ends.
Honestly, Chernobyl was about as bad as a power plant accident can physically get. It was a big reactor, it had no containment structure, and the whole thing pretty much vaporized and blew itself up all over the place. Pretty much whatever radioactive stuff was in the fuel got released into the environment. For modern reactor accidents, we look at the percentage of the reactor's inventory of each radionuclide that gets released. Anything more than 1 percent of the iodine or cesium is considered pretty high, and 100% would be unthinkable.
I take that back, Chernobyl could have been worse if several units melted down instead of just one. Like 3-5 times worse, max. The response also could have been worse (though I wouldn't say it was good). Still, no way we'd be talking about hundreds of thousands of deaths.
Well, it could have been much much worse, though we're kind of unsure how much :
https://www.physicsforums.com/threads/did-chernobyl-divers-prevent-a-multi-megaton-explosion.971999/post-6194382
On the other hand, Chernobyl should *not* be considered as a typical fission reactor accident.
On the third hand, the fact that the author of the book seemingly dismisses/ignores/minimizes what happened there (also the times that we got *really* close to starting a global thermonuclear war ?! And remember Putin's doomsday device?) makes me want to dismiss the rest of the book as as sloppy...
Anything's possible. What is the probability of a nuclear power plant killing hundreds of thousands or millions of people?
How would a nuclear plant disaster kill "hundreds of thousands or millions of people"? Even Chernobyl, based on a power plant design that would be have been illegal in the free world (even as far back as the 1960s), killed at most ~9000 people (if you accept the linear no-threshold model).
Even scraping the bottom of the barrel for the most unsafe plausible design, how do you arrange for it to kill over 100,000?
But even accepting this argument, the Banquio dam disaster killed roughly 171,000 people and this doesn't usually scare people away from hydro-electric power. I think it's more of an argument against shoddy communist engineering (ditto for the USSR's reactors in Chernobyl).
This book, and Peter Thiels’s famous quote, seem quite odd to me — I am glad we don’t have average humans flying cars, because they would be a massive hazard.
Driving is one of the most dangerous things that people habitually do. Flying a plane is way harder, and should be left to the pros and dedicated amateur hobbyists. Especially given the rates of alcohol, cannabis, opiate, benzo, etc. usage, it would be reckless to encourage casual flight by human amateurs.
Contrary to the idealism of this book and the libertarian movement, it’s government investment in computing, software, the internet, and AI that enables the development of automated flight systems, manufacturing, and maintenance — airplanes need to be in tip-top-shape.
We still don’t have level 5 autonomy for cars. Once that nut is cracked, then flying cars are a viable product.
Arguably, level 5 autonomy for cars is a much harder problem than something like level 5 for aircraft. The extra vertical dimension means collisions are an order of magnitude easier to avoid. Biggest problem would be large birds but wind turbines seem to be making progress on that one.
The lack of pedestrians and cyclists in the sky helps a lot too.
Plus we can start regulations from scratch, meaning that all flying cars can be required to communicate with all other flying cars in the area, and to all follow some specific protocol to avoid collisions.
I wonder what the potential energy of a flying car is vs. a 'road car'. I imagine the flying car's in substantially higher both because of the energy requirements to fly them and because they're, you know, in the air.
If they have more energy, they can do more damage if misused and therefore for similar levels of risk acceptance require more substantial regulation and safety requirements.
The idea is also that they'd be going much faster, and KE = 1/2mv^2. 10x the speed means 100x the energy.
Why would they be going faster? VTOL + fast = super expensive and complex. Are we talking Apaches (max airspeed: 150 kts)? V-22s (270 kts)? This seems insane to me.
Seems insane to me too, I just took this as part of the premise of the book. Flying cars are less transformative if they don't let us get places faster. Also, it sounds like you know a lot more about engineering constraints on flying machines than I do and I appreciate that perspective.
Even if they only go 100mph, they could pretty much go that speed continuously in a straight line to the destination. For almost any journey by car, that will cut your time in half or better...
The horizontal component is very likely to be dwarfed by the vertical component.
Is it? My impression is that very few airplane accidents are just falling out of the sky like a stone. I thought they were generally shallow-angle flights into terrain, in which the horizontal component of the plane's velocity was indeed the largest.
Yeah now that I say it I think I was definitely wrong. Unless like the wings fall off. Otherwise almost anybody is gonna try to pull off some kind of glide.
Well, depends on if they're more like helicopters or planes I think.
This was a Reddit TIL today, a helicopter in autorotation has a glide angle ratio of 4:1 (4 horizontal to 1 vertical) so the horizontal velocity is still the large component (the TIL was actually about the space shuttle, which is scarcely better)
Rough calculation:
Two-ton car moving at 100 km/h: 1.4 megajoules
Two-ton car hovering stationary at 2000 ft: 5.4 megajoules
There is no such thing as a self-driving car. We have cars that are driven by the human being right behind the wheel right now, and we can imagine cars that are being driven by a human being in a cubicle in front of a computer some time ago, writing down all the future decisions in an enormous if-then-else tree.
Now, if you believe the *hard* part of driving a car is Newtonian mechanics -- deciding exactly when to press down the brake pedal to achieve 0mph in X feet -- or sheer reaction time -- noticing the stop lights on the car in front of you have gone on in 0.05ms instead of 100ms -- then it makes sense that turning the whole thing over to a pre-written computer program will succeed.
On the other hand, if you believe the hard part of driving a car is subtle situational judgments -- *should* I be driving the speed limit in this weather, this lighting? How probable is it that the truck in the driveway ahead is going to foolishly pull out in front of me? When that bus ahead of me comes to the intersection where I can see some construction at right, what are the chances it will move into my lane? -- then you are much more skeptical.
I've no doubt a computer-programmed car could beat the pants off any human driver in a video-game driving game, and if the real world were as perfect as a video game (or could be made that way), then it would be an obvious win for the programs.
But personally I'm very doubtful the real world can ever be made as cleanly predictable as a video game at anything approaching a realistic cost. And fortunately one of the things at which the human eye-brain-hand system has been superbly optimized by evolution is the perception of, and appropriate reaction to, weird unexpected threats from the environment.
So I can see driver-assistance technology getting very very good, and much more pervasive, and it preventing a lot of obvious dumb things humans sometimes do, through inattention, distraction, exhaustion, and I can see it taking over certain routine tasks -- super-duper cruise control, roughly. But I cannot easily see a day when you can just dispense with the driver entirely[1], except for narrow-purpose restricted situations, the equivalent of driving a closed-loop track.
------------
[1] Which doesn't mean he has to be literally in the car: another area I can see occupying a highly productive niche is remote driving, the equivalent of "driving" a drone. This would solve the much-talked-about mobility problem of people who cannot drive, e.g. grandma who can't be trusted behind the wheel can be "driven" places in her own car by her son in a city 200 miles away, or she can hire some whippersnapper in her own city who sits in a building somewhere at a console and "drives" a bunch of old people where they need to go in their own cars all day for $30/hour.
> There is no such thing as a self-driving car. [... Just] cars that are being driven by a human being in a cubicle in front of a computer some time ago
I hear people make this point a lot, but I've never understood why this is an interesting or useful distinction. Would you be able to explain that?
(And sorry to pick on a minor aspect of what you said - I liked your point about subtle situational judgments.)
It's a useful distinction because it emphasizes how very much rests on the crystal ball sitting next to the programmer's desk. He needs to foresee *everything* that could happen, in the future, to the car-brain he's programming, and design an appropriate response, because there is going to be no intelligence on the spot to take over if the program can't handle whatever weird situation has arisen.
I mean, discussions often center around the quotidian challenges of driving, and how easy these are to automate -- and I agree, a suitably trained chimp or 12-year-old human could do *most* of my driving. But as the discussion of nuclear power elsewhere in this thread demonstrates, we cannot ignore Black Swan events.
And I understand the programmer has tons of real-world testing data to help him. But one hopes the continued existence of zero-day malware suggests that this is less of a guarantee than one might hope.
Thanks, this helped. I misunderstood what you were getting at the same way as Andrew below. And now I'm wondering whether all the other times I rolled my eyes at people making what I thought were dumb semantic points about whether a capable AI was 'really' intelligent because it had been built by humans, they were perhaps trying to get at a point about robustness or flexibility or originality or something in a way I wasn't picking up on.
(This made it sound like I thought you were making a dumb semantic point, but actually due to priors and you seeming thoughtful in general, I didn't. I figured I was probably missing something. That's just usually been my reaction to similar strings of words in the past.)
I am perfectly capable of making a dumb semantic point, so I am not in the least offended by your suspicion that I might have been. I appreciate the opportunity to clarify which you created by asking the question rather than running with the suspicion.
Ah so, well thank you for the feedback, as it seems therefore I did not write as clearly as I thought, and this is good to know for the future.
I'll add that the quote is like saying there's no such thing as a one-person-driven car, just cars that are being driven by two people in a bed some time ago.
How so? I think you misunderstood what I said. I mean that the *actual driving decisions* ("press down on the accelerator when the light turns green") are being made by a programmer considerably distanced in time and space from the actual event, because he wrote down if (light == green) { press_accelerator; } and then that got executed later.
> we can imagine cars that are being driven by a human being in a cubicle in front of a computer some time ago, writing down all the future decisions in an enormous if-then-else tree.
I'm not sure if I'm reading you correctly, but this is not how the things that are today referred to as 'self-driving cars' work. They use deep neural networks, which only resembles what you describe if you use very fuzzy meanings of "writing down", "all future decisions", and "if-else tree".
Well, I'm not in the business myself, but I know a little of neural networks, and I wouldn't agree with that at all. During the training of a neural network, one might indeed experiment different metrics for success and failure, and look for novel new patterns, but my understanding is that none of that happens when it comes time to actually execute the learning. The net looks for patterns, based on weights established during training, and then makes decisions based on those patterns which have been precomputed and predecided. It's certainly possible that those decisions are a set of weighted options, among which the best at the moment will be chosen -- but none of this corresponds to the completely de novo decisions which a human being can make. It may be pre-programmed in a complicated way, but's still pre-programmed -- still deterministic.
It's certainly possible I'm wrong, because as I said this is not my field, but I would need to dig into it in much more detail to figure that out.
Well, there's a prominent (maybe even majority these days?) stance that humans can also be described as pre-programmed and determenistic, at least in regards to any single de novo decision they are faced with. In any case, the relevant question here, are they on average better or worse at making it than the state of the art algorithm? By how much does the algorith have to beat them before we approve it? I don't think that society is rational enough to handle these questions, sadly.
I agree that there's going to be a lot of issues between automated cars and manual cars. One possible solution is to have all cars automated, rather than allowing manual cars around the automated ones. A group of automated cars can talk to each other and coordinate movements.
As far as the transition period, that's going to be a really hard problem to solve. We can segregate types of traffic fairly easily, as with HOV lanes now, or even automated-only roads. There are still going to be a lot of problems with segregation once we reach a certain critical number of automated vehicles, but there are still too many manual cars to remove them.
I'm not sure society has the gumption to make these kinds of changes anymore.
On one hand, there's good reason to be skeptical that a computer will (maybe ever, or maybe any time soon) be able to match what a human does.
On the other hand, so many people are incredibly bad drivers. The next time you are stopped at a light, count how many drivers are looking down at the phone in their laps.
Anyone who would trade computers and smartphones for flying cars is out of their goddamned mind.
And we have flying cars - they're called "helicopters". Turns out they're not very practical and I wouldn't trust any regular person with them, but possibly self-flying ones might change that?
Yes.
And to add to the same point. There are many states in the world. They differ quite a lot with regards to legislation and other societal aspects impacting experimentation and potential uptake of flying cars.
Take the australian outback or alaskan wilderness as examples. Airplanes are used as complicated and expensive flying cars in these locations. Solutions to the complicated and expensive part of that would be very welcomed, but the solutions still fail to manifest.
Sort of agree, but we have terrible flying cars and really good smartphones. Someone from a world with amazing flying cars and terrible pocket abacuses probably wouldn't want to trade either.
I'm extremely pro-nuclear, but I'm skeptical of arguments that act like the United States is the only country on earth. France, Japan (until Fukishima), South Korea, China . . . lots of countries to pick up the technology and run with it. It's certainly true that regulation made nuclear plants impossible to build it the US after the 1970s, but why don't the Chinese have flying cars?
Also it's perfectly normal for technologies to get cheaper and cheaper until they don't, you can't just assume things would have stayed on the trajectory they were on in the early days of nuclear power. DNA sequencing is a great example, which has been mentioned on this blog before.
Yeah, "why did this happen everywhere" is the best rebuttal to US-specific explanations, like the stories of US regulators shutting down specific projects over the decades. I actually find the arguments extremely compelling except for this one point. Why doesn't any other country on Earth pick up the slack on energy technology and outcompete the US?
The answer is relatively simple- nuclear power just isn't that great. It has some advantages over other sources of power (low carbon emissions in comparison to fossil fuels, more consistent in comparison to some renewables), but those advantages aren't enough to overcome the fact that it's extremely expensive. And while high regulatory costs account for a big portion of the cost of nuclear power (about a third, according to Eash-Gates et al, 2020), even if those regulatory costs dropped to zero, nuclear power would still be substantially more expensive than other options- the LCOE of nuclear energy is more than twice that of both onshore wind and solar. Just like with regulation, there are anecdotes of green activists blocking construction of specific plants, but neither environmentalism nor regulation is why nuclear power plants largely aren't built- it just comes down to nuclear power not being a cost-effective energy source.
I still think nuclear beats wind and solar when you factor in the cost of the batteries you'd need for a mostly wind/solar grid. Cost to generate electricity in the Mojave at noon is not really a fair measure.
You think wrong, then. LCOE does not measure the absolute cheapest production, it's a measure of average cost over the lifetime of an energy source. Nor does adding the cost of batteries make nuclear power cheaper. It's not as if these numbers are some secret or something that needs to be deduced from first principles- you can go look them up. I'm using as a source this EIA report: https://www.eia.gov/outlooks/aeo/assumptions/pdf/table_8.2.pdf. Even taking the most generous possible assumptions for nuclear power, that a full third of the costs could be shaved by dropping all regulation (which implies that there's no reason to spend any money on safety), nuclear underperforms solar and wind with batteries.
Well, whatever the theoretical arguments, I live in an area where a big nuke was shut down a few years ago *and* people have been installing solar like crazy. Before the shutdown, electricity rates were cheap. Since, they have done nothing but climb and are among the most expensive in the state now. So as a practical matter, this doesn't hold water for me, and I don't care how many reports say otherwise.
"You may have objective data compiled from detailed studies, but I have a single anecdote where one event happened after a different event, and therefore I've concluded that the first must have caused the second and no amount of facts will ever convince me to change my mind!"
You realize this is supposed to be a rationalist blog, right?
According to Lazard rooftop solar is $150–227/MWh, which is a tad more expensive than their estimate for new nuclear ($129–198), marginal (running) cost of existing nuclear ($25–32) or new utility scale solar or wind ($29–38 and $28–34). However, it is true that renewables were more expensive a few years ago (e.g. solar $49–61 in 2016).
If solar is really that cheap, why hasn't much of the grid been replaced by solar yet? It's only 2% in the US. Even with enough batteries to cover the night, you can still get screwed by weather, so you need a whole backup plant that's idle most of the time, which may or may not be included in that LCOE.
Also, it's easy to imagine scenarios where regulators increase costs by 10x or more. I'm not sure a third is sufficiently pessimistic. Apparently nuclear is competitive enough to still be worth building in China, even though the average cost of energy there is only 8 cents per kwh.
The avoided cost (money saved) isn't the same as the old levelized cost minus the new. It's closer to the old *marginal* cost minus the new levelized cost, since when you build a new power plant early you don't get back the money you spent on the old one. EIA's modeling is, as I understand, a little more detailed than that (though still simplified for obvious tractability reasons).
As for how much China's nukes cost, I'll defer to the WNA: https://www.world-nuclear.org/information-library/country-profiles/countries-a-f/china-nuclear-power.aspx
TL;DR: ignoring whether there are any other factors effecting the cost difference for the sake of the argument, a third is probably high, a tenth is probably low, I'd put my money the average being between the two. The wholesale LCOE in general estimates isn't actually that much higher than $8/kWh though, Lazard puts it at $129 to 198/MWh. It's just that (utility scale) solar and wind are much cheaper than that.
I read the wiki article on LCOE and it seems to imply none of the costs of backup capacity are included in LCOE, so LCOE is not a fair metric by which to compare intermittent solar/wind with more consistent sources.
A big part of the answer should be that solar only became cheap about 5 years ago, and the buildout will take time. I'm not sure how long it takes to secure the necessary land and transmission lines, but the U.S. isn't known for speed in such matters.
I find it very hard to believe that regulations only account for a third of costs, because the realized cost of nuclear reactors more than doubled around 1980, and in turn, the regulations in the 1970s were stricter than the ones that governed the USSR and their Chernobyl reactor, or even the U.K. Magnox reactors. Something quite suddenly made costs ridiculous. If not regulations, then what?
If it took ten years to build a car, how much would it cost? There is no inherent reason why a small nuclear power plant should cost more than a comparably-sized car.
You mean other than the fact that they're completely different things?
The problem with this argument is that it ignores the way electricity is bought and sold on the open market. Electricity is a unique commodity which must be consumed the moment it is produced; it cannot sit on a shelf.
So the raw per kWh costs are extremely misleading. What everybody obviously cares about is total system costs and total system reliability. And as an electrical engineer that studied power networks and have my name published on a conference preceding for solar (thank you very much) (these credentials aren't to say that I'm an expert, just evidence that I'm not biased against solar), you can't run a whole grid off renewables. Running a grid is all about matching the electricity supply to the electricity demand. You need to forecast electricity demand and then try to crank up and down your electricity generation to meet it. Unfortunately, standard fuels like coal and gas are extremely well suited to this task. Solar energy has been cheaper per kWh than natural gas for literally two decades now, why are we still debating this issue? Why hasn't it simply been deployed everywhere? Because each kWh of electricity is worth a lot less when you can't control when it's going to be generated. Gas may seem more expensive, but you can combust it literally whenever you want. There are no simple equations to express this, you need to consult elaborate models, but rest assured the people building the power grids of the world are doing this and that's why they keep building gas.
I'm certainly not arguing that solar/wind are unequivocally better than fossil fuels in terms of cost or effectiveness for supplying a grid- there's a reason we've used fossil fuels for a long time. The reason people are advocating switching away from standard fuels is their high externalities. Once you've made the policy choice to convert some production from cheap, easy fossil fuels to power sources that come with more complications but less carbon emissions, you compare costs among those alternative energy sources, and nuclear power rarely is the best option. Of course solar isn't as cheap and easy as coal, but it's cheaper than building nuclear power plants.
And in any case, as you know, it's not so easy to scale nuclear power generation up and down quickly either- both nuclear and solar/wind baseloads need to be supplemented with power sources whose generation can be varied more quickly, to compensate for variations in demand and supply.
I don't think the biggest problem with solar is handling quick fluctuations, it's "the sun sets and you have zero electricity". As far as I know, no place stores a meaningful fraction of electricity generated during the day for use at night. Batteries are just way too expensive, and this isn't reflected in the LCOE number.
FCAS is actually fairly easy to supply via batteries, to the point that the market is rapidly becoming too saturated. Energy arbitrage and resource adequacy on the other hand, could really benefit from batteries that are much lower $/kWh, even if they have to sacrifice kWh/kg for it. IMO the high density types like Li-ion aren't well suited in that role.
There are reactors we know how to build now that can modulate their output to a sufficient degree. We have a clear technical path to nuclear power that can modulate it's output to a high degree, at least as clear of a technical path as our path to better and cheaper batteries. Nuclear is, by default, a baseload. Solar and wind are not.
Even combined with batteries, whether solar and wind can adequately supply a base or variable load is highly dependent upon geography. In many places, we can expect fairly consistent sun and wind (although the whole idea of climate change is that the future of the climate and therefore weather is highly uncertain, which means that even these are not necessarily great assumptions). But in other locations, some days you get 10% the solar irradiance as you might get other days. This would suggest that you need to overbuild on the order of 10x in these locations if you want to reliably attain a minimum output, something that all other energy sources are capable of. None of these can be logically factored into a single number, which is why LCOE numbers are extremely misleading. You can't talk about per-unit rate of renewables agnostic of where you deploy them, it doesn't make any sense. Policy makers should not be using such simplified metrics. The utility companies themselves are not using them when they decide what to build. The fossil fuel lobby (specifically natural gas companies) are lobbying FOR policy makers to use these numbers because they know they're misleading and they will lead to a policy future where natural gas cannot be eliminated because it's the only way to deliver reliable service. Gas turbines can spin up and down very quickly and honestly are the most versatile type of power generator, easily functioning as a base or variable load at the same unit costs. This is why in Japan, New York, and everywhere else that nuke has been taken offline, it's been replaced by gas, not renewables.
> There are reactors we know how to build now that can modulate their output to a sufficient degree.
Yeah, and we can modulate the output of renewables as well, its called "turning the damn thing off". We've been able to do that with nuclear for ages now, France has *done* that for ages with their nuclear fleet. That doesn't change the economics of "if you have an expensive reactor that costs next to nothing to run, you better run it as much as you can or you won't be able to pay your loans off".
EIA's LACE calculations are regional, so unless they're as misinformed as the guy who thinks solar had a lower LCOE than gas for two decades, I'd say their doing some forcasts is better than that same guy going "trust me bro, thats what the oil lobby wants you to think".
Yes, gas is cheap, and it's one of the only sources that has a better LACE than breakeven.
Just note that nuclear power doesn't have to be expensive: it wasn't in the 1970s.
http://www.phyast.pitt.edu/~blc/book/chapter9.html
We could build reactors inexpensively; we choose not to. Moltex thinks it has the answer though: https://www.youtube.com/watch?v=cQCm-kmUWA8
It's an OK rebuttal, but not super duper, because human beings are rather monkey-see-monkey-do creatures, and are perfectly capable of following a leader (e.g. the US) even where it makes no sense. One would like to *think* each country thinks independently, but then again, we'd like to think each US state thinks independently, or even that each individual human thinks independently, and we are often sadly disappointed in this expectation.
Yes.
Seeing as Fukishima got whammed with a tsunami, I think it's a little unfair of him to sniff that the reactor would have been perfectly fine had it not been for those panicking officials. Yeah, they were panicking - because a tsunami was coming to wham them! And I think we can cut Japan some slack for being Very Nervous about nuclear radiation and civilian populations.
I think you're missing the point of my comment here. I'm saying that Japan was pretty okay with nuclear technology from 1970 to 2010, had an economy roughly half the size of the US's and an advanced tech industry, and yet they didn't produce insanely super cheap nuclear power. That can't be blamed on American regulators.
Right, but the effect of their being a power plant in the earthquake/tsunami zone is that a natural disaster that killed some 19,000 people and devastated several thousand square kilometers, killed maybe one extra person and a few hundred square kilometers of the affected area will take decades rather than years to recover.
The rational lesson of Sendai is "massive earthquakes are really bad things", not "nuclear power plants are really bad things".
Well, France does make (much) cheaper reactors. So does South Korea for example. Here is a link (super-quick googling) from Vox. I'm not a fan of Vox, but on cursory reading this seems alright:
https://www.vox.com/2016/2/29/11132930/nuclear-power-costs-us-france-korea
There's a few issues there- while it does show that France built reactors for cheaper in the past, it hasn't built any new reactors since the early '90s. And while Korea does show that reactors can be built for cheaper than they are in the US, the lowered costs on Korean reactors (2.5 million krw(2010)/kW, which converts to $2,755/kW) are still not competitive with renewables like solar and wind. And an important difference between those countries and the US is that they have largely centrally controlled utilities, whereas the US has a bunch of different companies that face harsher competition to provide low-cost power and are less centrally controlled.
Just a note: nuclear does not compete with solar/wind. It competes with gas/coal/hydro etc. They provide different services and there isn’t an apples-to-apples way of comparing their costs.
That's not really accurate either. The difference in role of gas peakers and your regular coal, nuclear, or even CCGT plant is way larger than the difference between that CCGT plant and wind. Ignoring the jargon of marginal cost or merit order or whatever, if you build loads of renewables the coal and nuclear plants are what's going to be driven out of business because they're they type of plants that cost loads to build and next to nothing to run, and therefore are the ones that you want to have trundling along at 100% of whatever it can produce for it to make economic sense.
A nuclear design like Moltex's seems, roughly speaking, like baseload combined with a peaker, truly complimentary with renewables: https://www.youtube.com/watch?v=cQCm-kmUWA8
One important part here is that France built a whole set of fairly standardized reactors. In a world that actually cared about the environment, there would be an EU or a U.S. "standard nuclear plant" that could then be produced on a large scale and more easily maintained, but clearly we're never getting _that_.
https://www.overcomingbias.com/2020/09/the-world-forager-elite.html
That's Robin Hanson's take on why the book applies globally. I don't know why it would also apply to China. It's much more plausible that India would inappropriately adopt regulation popular in the first world
https://marginalrevolution.com/marginalrevolution/2019/09/premature-imitation-and-indias-flailing-state.html
Other countries have even more stilted regimes of regulation than the US though. There's really no example of a first world country that maintained a lower level of regulation than in the US and would therefore make for a good comparison on what would have happened. Additionally the shear total size of the scientifically educated in population in the US absolutely dwarfs that of these other countries.
In China for example, having well over a billion people doesn't get you to flying cars faster when most of them are living at levels 1, 2, or 3 and your regime stifles new thought far more dramatically than in the US among the few level 4s who might try to make flying cars a thing.
I take it you're American then
Exactly. Soviet Union might be an even better example. They had much less restrictive rules on nuclear energy, great engineers... but this did not make energy ridiculously cheap, compared to the US.
This was my thought reading the review - I'm surprised the reviewer didn't mention it. The book's author appears to be saying not just "nuclear is the best of our available power sources" but "widespread nuclear energy would represent a paradigm shift on the order of wood -> coal or coal -> oil." 1970s USA might be more innovative than 1970s USSR, but if less-restrictive nuclear was this level of a gamechanger you would still expect the Soviets to have been dominating the world economy with all the cheap stuff they made, not desperately struggling to maintain military parity with the US while running out of grain and toilet paper. Obviously communist inefficiencies do count for a lot, but I would expect a communist Industrial Age country to easily outcompete a capitalist Baroque-era one.
Well, the U.S. has been very good at exporting its culture. You know all those cultures in National Geographic where nudity is normal? Mostly they've been shamed into wearing clothes by now. So, maybe the U.S. culture of fear of nuclear power spread really well around the world. The cold war, the bombs on Japan, and the activists working so hard to equate power generation with bombs enhanced the effect.
It's surprising that this would affect the whole world, but it's not literally the whole world we're talking about, only countries that already had reactor-building programs. In any country that didn't have regulations for nuclear power, it would be de-facto illegal to build a reactor — or so I infer, because you could save a lot of money if you didn't have to meet any regulations, and yet no one has ever taken advantage of this. AFAIK the closest thing to unregulated nuclear power is the USSR's Chernobyl design, the RBMK. Any country that is considering opening up to nuclear power is likely to copy regulations from another country, so the regulations are likely to be very strict post-1979 regs rather than "medium strictness" pre-1979 regs.
So, assuming a strict regulatory regime, I have high hopes for Molten Salt Reactors, which, for example, should be able to avoid the need for a super-expensive "nuclear-grade" concrete containment structure, because plausible accident scenarios with MSRs don't need to contain large amounts of high-pressure steam, and for the most part, plausible accident scenarios don't produce large amounts of airborne radioactive material.
P.S. Robin Hanson points out that far beyond the nuclear industry, regulations weirdly end up almost the same everywhere. And he's not even talking about international treaties that force regulations to be the same (e.g. copyright, WTO TRIPS, and isn't there a treaty about nuclear fuels?) https://www.overcomingbias.com/2020/09/the-world-forager-elite.html
The ML analogy is interesting. One of the most pernicious problems in designing a good ML tool is getting the objective function right. In the 'economy-as-ML', I'm not sure what the objective function would look like at the 'economy' level.
The objective function at each node is price discovery--profit maximization for sellers and the tradeoff of purchase for buyers. The analogy only goes so far but basically as buyers have money to flow one direction and sellers have products to flow the other, prices (the values at each node adjust to balance the two.
The interesting thing about the analogy is that you effectively have a recursive two-layer neural network, with a separate 'dimension' for each product in which you use a different subset of the available nodes for buyers and for sellers, and a different set of values for the prices, with the max of the values being limited by each buyer node's participation as both seller and buyer in all other dimensions (products), and the minimum of the values being limited by a seller's production capacity and net gain vs the net value of participation in a different dimension. Trying to actually simulate something like in a computer would immediately trip over several exponents with only a couple products and a handful of people, but I think it has a lot of merit as a thought experiment.
Alternatively, the objective function is value creation. People do not buy your product or service unless you deliver value for them. Price discovery is to ascertain a quantitative measure of the value. But we are not discovering prices for the sake of it.
> Then there should have been nuclear fission and nanotech, letting you fit a lifetime's worth of energy in your pocket.
No, there shouldn't have been, because pocket fission and nanotech (that is, the non-trivial molecular assembler kind) are both physically impossible. Science is not a magic box that takes in money and outputs anything you want. No matter how earnest and enthusiastic you are, and how efficient you are at spending money, you will never get FTL travel or pocket fission or cold fusion or magic transmutation nanotech or anything else of the sort. You *would*, however, get high-energy-density electric batteries and intelligent control software, both of which are finally enabling flying cars today.
Do ribosomes not count as 'non-trivial molecular assemblers?'
Ribosomes are not particularly useful to human operators. Sure we know how it takes the code and puts out amino acids, but there's a lot of processing afterwards and people have devoted their lives to figuring out how proteins fold into their fully functional forms.
My understanding is there have been great advances in understanding protein folding lately. You can't think of a use for being able to make arbitrary proteins?
Of what use would extremey pure silicon crystals be to the entrepreneur of the 1800s?
What exactly are you trying to accomplish with ribosomes ? If you're trying to nudge existing living organisms to be slightly more useful, then ribosomes are the way to go, for sure. If you're trying to nano-assemble computer circuits and space elevators out of carbon, silicon, and other raw materials, then ribosomes won't help you in the slightest.
Well, if you feed them some artificial mRNA, they stop a person from dying of COVID, which has proven rather useful over the past year.
Yeah, don't get me wrong, biology is great -- but it won't get you self-assembling flying cars or space elevators anytime soon.
It certainly won't. But I'm trying to broaden the definition a little here. After all, in 1969 "technology" included missiles and spacecraft and nuclear stuff -- but computer programming for commercial applications was boring stuff people with HS diploma only did, punching cards to transfer a lot of payroll data into a mainframe.
I don't think Feynman would be disappointed by what we've accomplished in the biomolecular realm.
And I wouldn't be so quick to count the ribosome out yet: what if we could engineer an alga that turns CO2 and water into diesel economically? Shazam, no more climate change worries. What if we could engineer another that would turn 98% of Venus's atmosphere into O2 and water? Or...what if by learning how Nature does it, we get some big key insights into how we can do it? All very speculative stuff, but not as daffy as imagining the Alcubierre drive will work out.
> what if we could engineer an alga that turns CO2 and water into diesel economically? Shazam, no more climate change worries. What if we could engineer another that would turn 98% of Venus's atmosphere into O2 and water?
It would be nice, sure, but it's not enough to just say "what if"; you have to have at least some indication that your plan is possible. Right now, it doesn't look like it is; for example, while extremophile bacteria do exist, AFAIK none of them would last very long on Venus (and those that would, wouldn't convert CO2 to O2 and water anyway). Sure, you could posit hitherto unknown algae that work in some completely undiscovered way in order to do all the things you want them to do -- but then, how are they different from magical pixies ?
It seems that you classify the set of "every thing any living organism can do" as "trivial," which sounds like we're working with very different definitions of the term.
Yes, I do classify it as "trivial", for the purposes of building flying cars and space elevators. Yes yes, I know that we are technically living organisms, and we could build flying cars and such, but I was referring to using organisms as tools on the molecular level.
Okay, but how do you know that there cannot be a technology for building almost arbitrary nanostructures with almost arbitrary materials?
Why doesn't rational drug design and the like count as "nanotech?" You may not be building the molecules with molecule-sized tweezers, but they certainly operate on the atomic scale, and you certainly need an understanding of the atomic-scale processes to design them.
Rational drug design is what stopped the AIDS epidemic. Rational mRNA design is what has just stopped the COVID epidemic. Every drug with an "ib" at the end of its name is a monoclonal antibody raised against some particular atomic-scale target, based on atomic-level understanding, and there are a lot of them these days. There was a kid in France who was *cured* of sickle-cell disease by laboratory editing of the genes in his stem cells.
I think the issue is that organic 'nanotech' is for the most part limited to a much smaller subset of the periodic table than what we'd want in order make durable products that have a desirable set of functions outside the body. And what happens at the biological nano-level mostly stays at the biological nano-level. The promise of nanotech is that we can have designable nano-scale features in a macro-scale product arraigned in ways that create desirable function. Today we can mostly only create nano-scale components as a soup of many copies of components, and worse we can't really control how they are assembled into larger scale objects without adjusting the components themselves. That's makes for a nasty engineering problem that is currently pretty much impossible.
I don't think rational drug design is nearly as developed as you make it sound. We can't really predict or simulate the function and interaction of molecules very well yet at all outside of knowing the function and interaction of very similar molecules. Drug design is mostly a lot of testing of molecules that are analogues of ones with known interactions, and using strong artificial evolutionary pressures to both find a random solution and eliminate non-solutions. We're not designing large enzymes from scratch, or coming up with molecules that can chemically attract a disease protein without the aid of some animal's immune system.
Sure, I agree "nanotech" as it is written in sf magazines has not come to pass, and I never thought it would. Materials science has not stopped advancing, and very nice things are being done with it, but it's not a revolution and only credulous people thought it would be.
But I don't agree that there hasn't been *remarkable* progress in biology with applications to medicine. For some reason -- perhaps because it is has a lot more women in it than physics or materials -- these things don't seem to become part of that future-loving speculative discussion. I'm suggesting it is actually some pretty revolutionary stuff happening at the atomic scale that is about as exciting as anything "nanotech" was going to delivre.
I hope by "rational drug design" you don't think I'm limiting this to in silico studies. Nothing of the sort. I'm including the insights and collective wisdom of all the med chemists, too, and all the immunologists and studiers of genomes. All that is part of the way people go about attacking disease with molecules (or biomolecules) these days, and it's *way* more "rational" (meaning deliberate, based on genuine understanding of what's going on) than it was even 20-30 years ago.
> FTL travel or pocket fission or cold fusion or magic transmutation nanotech
Agree with your overall point but the items on this list vary pretty widely in how plausible they are.
We've literally done cold fusion with muon catalysis. The problem is getting to breakeven with it, because muons have a half-life of 2 microseconds and there's no obvious way to efficiently make them.
I can see why pocket fission is impossible, because fission relies on criticality and there's such a thing as a critical mass. But why is a molecular assembler impossible? Life clearly does molecular assembly all the time.
Critical mass isn't that big a problem. I could hold a bare-sphere critical mass of plutonium in my hand (although I'd need my other arm to hold it up). The issue's more with control and shielding.
For the critical mass to be "small" (small enough for a car), wouldn't you need nuclear material almost pure enough to make a bomb? Can't imagine that being legal.
But yeah, neutron and gamma shielding is hard. Neutrons, being neutral, can fly some distance through solid objects until finally hitting and ionizing something. Gammas are also hard to stop for some reason. (Betas & alphas are charged, and so easier.)
A less pressing challenge is heat dissipation — it should be easy to manage for a small reactor, but maybe not so easy that any random person would be allowed to own a reactor.
You can get around the "lol homemade nukes" issue by including a bunch of SF-happy stuff that would predetonate and fizzle a bomb - ideally, SF-happy isotopes of the same element the main core's made out of (that way you'd need an isotopic separator to fix it, and if you have one of those you can build nukes anyway).
On reflection, there are ways to get control (and you need control, both to turn the reactor off/on and to allow high burnup before the pile fails) without needing neutron poison and the associated large increase in critical mass - removable moderators and reflectors can do it, and can even potentially do it in a failsafe way. Gamma and neutron shielding is still a bitch, though.
This book sounds as if it would tie in very nicely with the previous review, "The Collapse of Complex Societies"; to the question "What went wrong in the 1970s?", my initial response was "It was the 70s".
Tainter would probably say "complexity"; the people alive at the time would have said "nothing much wrong (well, crime in some cities, sure, and the whole domestic terrorism thing, and oil shortages and things, but otherwise pretty okay)"; but I never heard anyone, until now, say "a lack of nuclear power plants".
If this review is faithful to the book, the guy sounds like a crank. Let's take a few points:
"The average American moved from Level 2 in 1800, to level 3 in 1900, to Level 4 in 2000. We can state the Great Stagnation story nearly as simply: There is no level 5."
Oh yes there is; aeroplanes. Private citizens can own their own, from the basic models to the billionaire's private jet, and the affordability of commercial flight has never been so within the range of the ordinary person (to the point of griping about airlines packing passengers in like sardines and crowding more and more is smaller and smaller seating, which shows the success of 'I want to take a flight and I can afford it' for ordinary people).
In fact, the progress that Hall describes from Shank's Mare to Your Very Own Automobile was forecast by many writers to take that Level 5 step: Your Very Own Aeroplane. Indeed, Kipling for one went so far as to put the entire world, now that air traffic was cheap and commonplace, under the control of the Aerial Board of Control: https://en.wikipedia.org/wiki/Aerial_Board_of_Control
As for the wicked interference with developing flying cars, Hall seems to want highways in the skies. And think about that - think about Highways. In. The. Skies. Think of the worst snarled-up rush hour traffic, cubed (because you can stack flying cars), in the skies overhead for major cities. Think of your commute, only now you're up in the air.
Hall has the notion of primitive days before everyone could afford their own car, only transferred to flying cars. Hop into your jalopy, rev 'er up, and flit from home to town and back again in the clear blue yonder. A charming notion, but in reality, it'd be you and six thousand others up in those skies.
The terror of nuclear energy comes from it being used first in war, not in peacetime. As the most destructive and terrible of weapons, not as a source of energy generation.
"If you put a completely legal luminous watch in a barrel containing half a tonne of dirt, that dirt would technically be intermediate-level nuclear waste according to the regulations."
Ahhh - when did radium watches become legal again? I think he's doing some three-card-trick shuffling of terms here; radium is radioactive (read this Wikipedia article on what happened to women working with it for watch dials https://en.wikipedia.org/wiki/Radium_Girls) but now luminescent dials are using " phosphorescent- or occasionally tritium-based light sources" which are safer and hence legal.
He is definitely pining for the 70s, when you could wear radium watches on your wrist and fluoroscope your feet! https://en.wikipedia.org/wiki/Shoe-fitting_fluoroscope A little dose of radiation will do growing kids no harm, in fact it'll encourage them to grow in new and interesting ways!
I do very strongly get the impression that Hall isn't interested in flying cars qua cars, he's interested in them to get that interfering government off his lawn (quick, where's his shotgun, here come those dadgum Revenuers again!)
As for the flying cities, James Blish had an entire series about them https://en.wikipedia.org/wiki/Cities_in_Flight . He had them powered by spindizzies, something I think perhaps superior to the idea of "one single line of nuclear power plants (every 250 feet) along the wing would suffice to keep Aero City flying indefinitely".
My main objection here is that he envisages everything working perfectly all the time (once tech has been so perfected by being liberated) so that nuclear plants will never fail, flying cars will never crash into each other or drop out of the sky, and as for his flying city... well, let's not even imagine how catastrophic that could be.
Two modest comments:
1. Tritium is radioactive, and it's likely that's what he meant. The problem with radium on watches is in its manufacture, not its use. In terms of wearing it, you'd rather have the alpha emitter (radium) than the beta emitter (tritium) because the alpha is stopped dead by even the thinnest watch back, while the beta might well burrow through and into your skin.
2. It's much harder to get traffic jams in the sky, because you can move in three dimensions. Each evening in Carlsbad Caverns, about 1 million bats exit the cave in an hour or two. If it were a million equal-sized mice trying to crawl out of the same size hole, it would take days.
The traffic-james-in-the-sky line of thought is interesting. No, you wouldn't have traffic jams, but the ground storage for aircraft would surely be finite and you could have a much, much worse problem of too many aircraft trying to land in the same area. Cars have an advantage here where they don't require much energy to sustain themselves in totally deadlocked traffic, and worst case scenario they need a tow. Aircraft have a landing problem. Flying cars would only move the problem of congestion from the via to the destination, but "too many people want to go to the same place at the same time" doesn't have an easy solution (except dense living and public transit).
Sure, and the congestion near existing commercial airports is already pretty hairy at times. One assumes that flying cars would be something you only really considered for a daily driver if you lived pretty far out and went to work not in a dense urban center. But that's still a lot of people. One could argue the automobile presented a similar challenge to the cities of 1900, and what happend is that a lot of people and jobs moved to the suburb.
If I think about it personally, as in, I just won a nifty flying car by falling through a wormhole and winning an AD 2200 version of "The Price Is Right," my main hesitations are (1) cost -- I feel like this is going to cost even more than my R8 -- and (2) safety, as in, am I going to feel like I *cannot* get in this thing unless I am well-rested, calm, know exactly where I am going, have planned everything out meticulously? I don't *drive* like that, all the time, and if I had to be in mental tip-top shape to go to work Monday morning or jump out for a burger and fries...well, meh.
But on the other hand...I like to go skiing about 400 miles away, and if I could jump in the Jetcar and be there in 3 hours, door to door...the pulse quickens. An afternoon on the slopes whenever I want to make a bit of an effort? I would need less scotch after certain days at the office.
"congestion near existing commercial airports is already pretty hairy at times"
Some of that is because we have taken the problem of motorcycles and tractor-trailer trucks co-existing on the same roads and multiplied it by several times. I think the difference in vehicle weight between a jumbo jet and a Cessna is probably bigger than that.
Radium daughter isotopes can emit substantial amounts of gamma radiation, and radium is also taken into the bones, unlike tritium which, being a hydrogen isotope, doesn't stick around for too long in the body. Also, tritium is an extremely low energy beta emitter, stopped dead in its tracks by a sheet of paper. You're wrong about radium being safer than tritium.
Sure, that's possible. You're right I should have considered the whole decay chain in the case of Ra -- that was silly. But I will also note unless you eat your watch dial, it's not much of a worry that it's in Group 2A.
As others have pointed out, plenty of other countries have embraced nuclear energy. That doesn't automatically enable flying cars. Unless we're going to put reactors in individual vehicles, which we won't do, you'd need better battery technology to take advantage of widespread nuclear energy. Besides that, there are so many other reasons we don't have and probably don't even want flying cars. Is every apartment building going to have its own air traffic controller? Automated systems on roads work fine because they're roads. Who keeps arbitrary flying machines on pre-defined flight paths? We can do that now because of how controlled and limited airplanes are. Are you going to file a flight plan to go to the grocery store? What happens when a flying car breaks down? It's fine to pull over to the side of the road. Falling out of the sky in the middle of a crowded city is not so fine.
What we actually want and are pretty close to having is widespread fast delivery, ideally via self-driving cars, so people don't need to go to stores at all, and better telecommuting and VR technology to limit the amount of interstate travel and daily commuting that needs to happen in the first place. Flooding the sky with amateur aircraft is not the answer.
Ok, so this book review isn't bad, but the book sounds too error-filled to be worth anything. DDT didn't get banned for causing cancer; but it is associated with increased Hodgkin's lymphoma, testicular cancer, and liver cancer. https://en.wikipedia.org/wiki/DDT#Carcinogenicity It got banned for it's impact on wildlife, IIRC. The author sounds like some kind of fundamentalist. Hard pass.
This definitely stuck out to me on first reading. I don't know if Where's My Flying Car goes into any more depth, but... the impact on wildlife is why Silent Spring is called Silent Spring. Because places where one would previously hear sounds of birds had gone silent. Characterizing it as weighing against the use of DDT on the basis of it being carcinogenic is not a mistake one could reasonably make given cursory reading of the book, or even brief reading *about* the book.
After reading the Wikipedia article on DDT, it appears that that the DDT ban was due to fear of it causing cancer *and* its impact on wildlife, particularly birds.
At least the book doesn't continue the lie that DDT would have solved malaria (DDT is still allowed for malaria control purposes, but many places have moved on anyway because the mosquitoes developed DDT resistance).
>Imagine if you had to bike to a train station to get anywhere (not such a leap of imagination for me in New York City! But it wouldn’t work in the suburbs).
Wouldn't work in the suburbs we have in the US, perhaps, but my understanding is it would and does work in many European suburbs. (Plus you can get around more effectively just by bike there too.)
Here in the US we have this sharp divide between cities and suburbs (where it's impossible to get around effectively without a car), but it doesn't have to be like that in the first place, and in many parts of the world it's not.
Anecdote from the last time I was in Brescia - the streets were jammed with parked cars during a weekday. Not sure if they were from people coming into the city and taking the train, or from people taking cars instead - I was there on vacation and had taken the train into the city.
There's an idea in economics called the Environmental Kuznets Curve which posits that as societies become wealthier, people become less willing to tolerate environmental degradation. Or, alternatively, are willing to spend more money to remediate or avoid environmental degradation. A few years back, I read somewhere about a similar effect regarding certain types of personal risk. E.G., if you wish to understand why children 50 years ago were allowed so much more unsupervised play time than kids today, it's because parents were less risk averse. If you wonder about why the list of mandated safety features on automobiles keeps getting longer (crumple zones, airbags, electronic stability controls, etc.), it's because drivers are more risk averse. And so on and so forth. As someone who leans libertarian, I'm quite sympathetic to concepts like "bureaucratic incentives are to calcify," but I think it has to be pointed out that our bureaucratic paralysis wasn't exactly imposed on an unwilling or skeptical public.
Exactly, and relevant to the comment by Crouchback just above yours.
The reason nuclear reactors cost 3 times in the US what they cost in France is because the voters in the US don't want them while nobody asked French voters how they felt. (I exaggerate for effect.)
Pinker makes the point somewhere that the death penalty still exists in the US (and not Western Europe) is, counter-intuitively, because the US is more democratic. I think this is an insight.
I too am a libertarian. I watch with disgust a government that keeps being a bigger slice of the economic pie. But then I talk to my neighbors and realize we have the government we can afford. As we create more wealth we get a bigger government. Likewise, as we get richer we spend more on education and healthcare. Etc.
The late Mark Kleiman theorized that the difference in capital punishment between the U.S & elsewhere is because civil servants are more education and that results in them opposing it. But when I looked at the General Social Survey (which is US-specific), a majority supported the death penalty at every level of education:
https://entitledtoanopinion.wordpress.com/2010/09/12/support-for-the-death-penalty-by-education/
You can explain it in terms of impact on quality of life rather than gdp. E.g. The marginal benefits of cheaper energy become smaller compared to the marginal harms of the air being less breathable. Even if lower cost energy more directly contributes to gdp on paper
So how does he explain France? It's famous for dirigisme and arguably is the most statist economy among Western democracies. It is also famous for being #1 at exploiting nuclear power. If anything you could argue that nuclear power is the most socialist form of energy. Likewise does he assume that every single nation as developed as the US made the exact same mistakes so we're all equally handicapped in technological development? I can't help remembering that libertarians were big fans of the Dale.
Personally I found Robert Gordon most convincing - we solved the easiest problems by the 1970s and further gains in productivity were slower because they were more difficult.
The best counter argument I can think of, to “nuclear power would enable so much more progress, but greens killed it with regulation”, is China.
If what he’s saying is true, why isn’t China building tons of nuclear plants? Are they also constrained by the same regulations?
It's impossible to answer your question without knowing what you mean by "tons." But I see this on Forbes: "During their 13th Five-Year Plan period from 2016 to 2020, China built 20 new nuclear power plants with a total capacity of 23.4 GW, doubling their total capacity to 47 GW. And that is expected to happen again during their next 5-year plan, which has a new target of 70GW of nuclear generation before 2025."
I would count 20 new nuclear plants in 5 years as "tons."
Source: https://www.forbes.com/sites/jamesconca/2021/04/23/china-will-lead-the-world-in-nuclear-energy-along-with-all-other-energy-sources-sooner-than-you-think/?sh=385ceeb0778c
Ok, so then this looks like evidence that the theory has something to it. I didn't realize they were building at this rate, but i guess i didn't do much research. Thank you for sharing!
That's not tons. China builds 70GW of wind a year, that evens out to 18GW a year with capacity factor 0.2.
China is not building 18 reactors a year.
Fukushima put the brakes on the Chinese nuclear power program exactly as it did the Japanese. The CCP is very risk averse; it was terrified that a Chinese Fukushima would lead to a large scale revolt and societal instability.
You underestimate how much the world follows the American lead on regulations and decision making.
China *is* building tons of nuclear power plants.
“ Then, in 1977, Jimmy Carter established the Department of Energy. Costs immediately skyrocketed, and never came back down.”
So nothing to do with Three Mile Island?
I almost 100% agree with the guy. But why include such bull $hit?
Why did he include such bill $hit - I mean.
The Three Mile Island accident happened in 1979, so it's unlikely that it was that much of a factor on the creation of the DoE.
What are you calling "bull $hit"?
We aren’t taking about the creation of the DoE. We’re talking about the sky rocking expense.
Since we can’t edit: The creation of the DoE wasn’t the problem. It was the public freak out about Three Mile Island that caused the cost explosion.
I think you raise a good point, but I don't think I would put it that baldly. I was living near Three Mile Island at the time, and while people were concerned, it wasn't a freakout, and it wasn't nearly as uniform a response as your sentence suggests -- and I grant you may mean a more nuanced thing, and be just summarizing here.
I would say (and I have not researched this, just going on having lived through the period in question) that TMI was more a symbol and possibly an accelerant of a transformation of public opinion that was well underway already.
But I certainly agree the establishment of the DoE had nothing to do with it. If anything, that was just a symbol of transforming attitudes. As I recall, the main argument for it was the "energy crisis" and Carter's (as it turned out futile) wish not to be seen as doing nothing about it, come the election of 1980. So kind of like the Bush Administration's creation of the TSA in response to 9/11.
Costs actually started rising well before either of those events (see https://imgur.com/a/fEG1WKx, from https://gordianknotbook.com/). TMI might have contributed some, but I think it's pretty clear that there was a lot more going on.
My impression is that the opposition to all things nuclear started in the 1960s or even since the bombs on Japan; it gradually gained steam until Three Mile Island supercharged it in 1979. I think this is related to the cost increases. See also: http://www.phyast.pitt.edu/~blc/book/chapter9.html
I wouldn't really call this a review. It's just a naive summary of the main points of the book that doesn't critically engage with any of the ideas. There's not a single word written on whether the author's claims hold water or whether his argument is logically coherent, not a single outside source referenced, nothing to suggest that the reviewer has any understanding of the subjects where Hall is making controversial and unsubstantiated claims. Bluntly, it's the level of writing I'd expect from a 5th grade book report, not a finalist in a review contest.
Flying cars are an inherently bad idea for many reasons. This is the best analysis I've found that lays out all the reasons why. It concludes that it would be much better to normalize travel on small, autonomous planes based at dense networks of small airports than to try putting a flying car "in every driveway."
https://www.militantfuturist.com/why-flying-cars-never-took-off-and-probably-never-will/
I was going to object that many (though not all) of those problems go away given the orders-of-magnitude improvements in energy technology Hall thinks we should have... But then I remembered that Hall thinks an actual '70s design would have been successful if not for regulators, so really it all applies.
With the benefit of being able to remember the last 40-odd years, I would say the general description of the social change is accurate: there's no question to my mind that modern society hugely more timid than what it was 30-60 years ago, and far more concerned with equity than opportunity.
But there's still some serious chicken-and-egg questions about that. Did we become more timid because of this change or that? Or was it our timidity that drove those other changes? There's a strong interconnect between social attitudes and technological change, but one can make an argument for the causality running either way, and probably in different areas it runs in different directions. Plus there are probably feedback loops, in the sense that once a change of attitudes causes a change in technological progress, the delta in progress informs further change in attitudes, et cetera.
It's definitely the case, though, that the difference between American culture and Soviet culture has narrowed in completely surprising (if you remember the difference in the 60s) ways. We really are much closer that I would ever have thought possible to the culture that, were it told the chocolate ration had been "increased" to 20g, would pretty meekly go along, and viciously choke off any troublemakers who attempted to point out it was 40g last week.
I certainly wish I knew why. I'm not convinced by this it can be easily traced to a failure to go full bore on nuclear power. That feels more like another symptom to me.
I wonder whether this has something to do with the wars. In a society where a world war was just 20-30 years ago, and the one before that 50 years, then you are not so confident that your society still exists in 20 years from now. If you have these looming risks anyway, then you are less concerned about an exploding power plant, or about destroying your environment.
In most European minds, facing a society-threatening risk like a war in your own country is just of of scope nowadays (except for the far right). And if you expect firmly that your 10-year old child is going to live until 80 in the same stable society, then you worry a lot more that it might have a bad experience tomorrow, or that it is long-term poisoned by some chemicals. And the same for yourself.
I wonder whether this explains part of the more progressive philosophy of the rationalist community. A lot of people here are concerned about X-risks like a badly aligned AI. If you believe that there are real society-threatening risks, then you are less concerned of adding more minor risks. On the other hand, if you firmly believe that society will do well for the next 80 years, you are more conservative.
Ironically, this expectation of doom is something that (parts of) the rationalist community share with right-wing parties like the Republicans. At least Scott Alexander made a convincing argument that this is something that sets apart Republicans and Democrats in the US. But where the rationalist community is concerned about society destroyed by an AI, Republicans are more concerned about (their) society being destroyed by too much immigration.
Are you familiar with the "hygiene hypothesis" for the higher rates of autoimmune and allergic disease in First World countries[1]? The idea is very roughly that in the absence of enough real threats (infections to fight off) to occupy its attention the immune system starts to find imaginary threats and attacks your own cells.
My pet theory is a sort of social hygiene hypothesis, in that if there aren't enough real threats to a society people just start making them up, or at least magnifying small threats until they look like big ones. Id est, if in your lifetime Hitler might get The Bomb before Patton takes him out, or Stalin might kick off World War III at any moment if he feels sufficiently pissed about Berlin, that might...focus your existential concerns. Contrariwise, if you live in an era devoid of much in the way of plausible immediate existential threats, and where there hasn't even been a serious recession in 12 years...maybe you start finding all kinds of rather modest problems and blowing them up into problems that feel around the size of the Cuban Missile Crisis.
I'm not convinced humans are well-designed for (general) happiness. If we were let back into Eden, I'm pretty sure we'd find some reason why it was awful, or we had to fight each other to the death over whether to pick pears or apples first from the Tree of Life.
------------
[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2841828/
I like the theory! It rings true to me.
It's not that the problems are just imagined. (Ok, some are.) It's more that the level of agitation is not adapted to how serious the problem is.
Also, when there is a real existential threat, then society is aligned. All people fear the same thing. But without an existential threat, people focus on different fears and problems, and get unsatisfied when others are not concerned about the same issues. (There are some nice satire pieces by Kishon about how absurdly nice and polite people in Israel became towards each other during the six day war, and how it immediately degraded with each success story from the their military.)
Thanks, I'll have to look one of those up, sounds amusing.
The big difference seems to me to be that we changed from two (or more) children to (roughly) one child per family.
Since all of your eggs are in the one basket (as it were) you protect the basket much more.
"Before the War, America was an individualistic nation. Then came the Depression, the New Deal, and most of all the War. America won the war with a “completely centralized bureaucratic government structure” - and it was a huge success. And for a while, that worked: the generation forged in the war had a “cooperative “same boat” spirit” that “[made] the centralized corporate structures work.” But then it didn’t. Hall blames the hippies"
I never thought of it this way before -- and the above paragraph doesn't note it explicitly -- but one tends to think of the hippies' individualist ethos as left-wing and the older, Emersonian self-reliance ethos as right-wing, yet perhaps it's the same strain of American culture that gave us both.
One could argue the hippies had the creative, individualistic drive of a child, whereas Emerson was talking about the responsible-minded independence of an adult.
Insightful comment. Kudos.
One thing I find disturbing about this 'fear of nuclear is just hysteria' argument is how its proponents always casually brush off the cost and the danger associated with cleaning up the few disasters that have occurred. Even allowing for the argument that the cleanups themselves have inflated costs because of excessive fears of radiation, we're talking about thousands of people who HAVE to be displaced, land that remains uninhabitable for decades, and cleanup costs in the tens of billions. And this is just the accidents that occurred at the heavily regulated plants that do exist! Saying Chernobyl killed 37 people is such a bizarre understatement of a) how serious it nearly was and b) just how extreme the actions taken were to prevent more deaths. As an energy source it also generates waste that remains dangerous to people for thousands of years, which has to be safely transported and stored and remains a huge security risk for the duration of its lifespan.
I'm prepared to accept arguments that the impact of climate change could be worse (probably true) or that fear of nuclear is somewhat overblown, but I always find it shocking when people are as dismissive of anti-nuclear viewpoints as they are of e.g. anti-vaxx.
What are the dangers of modern nuclear plants vs. the dangers of modern coal and gas plants? isn't that the question?
We also need to factor in the dangers related to getting the fuel to the plants: mining, drilling, fracking, transportation, be it by tanker, pipeline, truck or rail.
Most experts who have done the cost-benefit analysis have concluded the dangers of not having nuclear are much higher than the costs of having it, given current technologies in all areas. I believe that is true even_if you don't think climate change is a danger.
"costs" = "dangers" in the above
I agree with the cost-benefit analyses, and I am mildly pro-nuclear. But at the same time, I think they miss a really important point. There are expectations, and there are tail risks. The cost-benefit analysis is usually about expectations or typical outcomes. Here, nuclear energy is not so bad.
But in tail risks, they are much worse than coal and gas plants. What is the risk that a country like Czechia becomes uninhabitable because things go really, really, really catastrophic in in a coal mine or a gas plant? Pretty much zero.
What is the risk that the same happens because things go really, really, really catastrophic in a power plant? Small, but not zero.
You could argue that "uninhabitable" is a too strong word. But then replace it by some slightly less catastrophic scenario, and the argument still works.
I'm not sure that any of your statements are true. In any case, they need to be supported.
To say that the plants that failed were "highly regulated" is odd. Chernobyl is a case study on the Soviet Union's spectacularly bad political incentive structure. The rest of the world at the time knew that those reactors were unsafe. Several soviet scientists themselves attempted to publish research saying as much and they were dismissed and discredited, who knows how many MORE scientists agreed with them (or would have) but who was willing to go against one of the darlings of Soviet scientific achievement, the RBMK reactor, so many of which had already been built? One of the frustrating things about the Chernobyl HBO documentary, which is extremely good and entertaining, is that a lot of people took away from it the lesson "nuclear reactors are unsafe!" when imo the proper lesson to take away was "holy shit look how dysfunctional the USSR was on the inside."
I would never claim that "Chernobyl killed only 37 people", but the more one learns about nuclear power, the less one is worried about it. I look back on history and see there was an anti-nuclear movement for 60 years or so, during most of which, most of their victories were victories for fossil fuels (as solar & wind only became cheap over the last 10 years). So I take all the anti-nuclear arguments of the last 60 years and compare them with the alternative they effectively supported: fossil fuels. Okay, so Chernobyl killed a few thousand? Fossil fuels killed a few million! Yet nuclear is the one they demonized?
The opposition vaguely made sense if you think "nuclear power makes it easier to build nuclear bombs, and bombs are horrible!" The bombs are horrible, yes, but why not just focus on opposing bombbuilding? Some nuclear plant designs aren't useful for bombmaking, so why not promote those? But no: 60ish years of opposition, and we end up with tons of coal plants and also huge stockpiles of nuclear weapons. Great job fellas.
My other comments above speak to many of your concerns. Let me know if you have questions.
A couple of things to reply to here:
1. I'm not talking about enriching uranium, I'm talking about the risk of dirty bombs being created with plain old radioactive waste. The consequences of a dirty bomb being used in a Boston Marathon-style attack are unthinkable. It's also an environmental hazard that needs to be stored somewhere it's safe from every kind of natural disaster imaginable, then kept safe from theft or corrupt/ideological sale. There is NO option of failure on either of these points. That waste is a security liability for as long as it's sufficiently radioactive.
2. Every time I hear this comparison, I see the existing state of fossil fuels being compared against the existing state of nuclear power. This isn't what we're arguing. It needs to be compared instead to a hypothetical much, much wider and less cautious rollout of nuclear power that would have happened in the absence of the cautious approach we took, over decades with industrial safety standards that saw the Bhopal disaster, the Sellafield disaster, the 2020 Beirut explosion, the Tianjin blasts, etc. etc. Also, we have to hope that there were no nuclear plants in Kuwait or Syria, to name two places where there was a complete security breakdown in the last few decades. Disasters happen and when they are nuclear, the cost is incredible.
Some people might *say* dirty bombs are "unthinkably bad", but I've never seen evidence that dirty bombs could realistically be "better" from a terrorists perspective than a conventional attack or a chemical weapon attack — especially if the terrorist wants to *actually* hurt people instead of mainly causing a fear-based response. A quick Google finds me an info sheet from the DHS — an organization that benefits not at all from minimizing any threat to America: https://www.dhs.gov/xlibrary/assets/prep_radiological_fact_sheet.pdf
> It is very difficult to design an RDD that would deliver radiation doses high enough to cause immediate health effects or fatalities in a large number of people. Therefore, experts generally agree that an RDD would most likely be used to:
• Contaminate facilities or places where people live and work, disrupting lives and livelihoods.
• Cause anxiety in those who think they are being, or have been, exposed.
"Cause anxiety" indeed: Bernard Cohen once sent out a survey to the Health Physics Society and Radiation Research Society. The survey asked experts to compare "the general public's fear of radiation with actual dangers of radiation". 91% of respondents agreed that the public's fear is greater than realistic, and 58% thought that the public's fear is "substantially" or "grossly" greater than realistic. Further, 82% of respondents thought that television coverage of the dangers of radiation "substantially exaggerated" or "grossly exaggerated" the danger.
http://www.phyast.pitt.edu/~blc/book/chapter5.html
If indeed fear is the main benefit from the terrorists' perspective, I don't see how hyping up the dangers of radiation is beneficial to the rest of us. Any time you hype something out of proportion to the actual danger, you open the door to other dangers you didn't hype (such as fly ash, which is empirically about 1000 times more dangerous.)
If we're talking about "existing state of fossil fuels" vs "a hypothetical much, much wider and less cautious rollout of nuclear power", does "less cautious" mean "a 1978 level of caution (back when nuclear plants were affordable), plus lessons learned from all nuclear accidents since then"? If so, I'm comfortable with that. The TMI and Fukushima accidents, for example — keeping in mind that these were both built in the 1970s — are avoidable without significant cost increases, and all western reactors already had safeguards against the kind of accident that happened at Chernobyl.
Also, the badness of coal vs nuclear is so high (millions of deaths) that, AFAICT, all the nuclear reactors melting down en masse isn't bad enough to come close to coal: https://www.reddit.com/r/nuclear/comments/jtm6hm/how_bad_is_meltdown_world/
I should add, though, that assuming we stick with modern super-safety standards, it should still be practical to affordably replace fossil fuels with nuclear plants (especially if there were a carbon tax or clean energy subsidies). The most promising way to do this is via 4th-generation reactors, which mostly aren't "children" of traditional reactors, but more like back-to-the-drawing-board designs such as MSRs, which reduce costs by (1) eliminating hazards rather than managing them, and (2) avoiding the use of water for coolant, which is challenging from a containment perspective because it is kept at 150 atmospheres of pressure and over 300°C, and that's under normal rather than accident conditions. (the concept of Small Modular Reactors or SMRs, by the way, is a separate concept with separate cost benefits, though obviously SMRs can also take steps like eliminating hazards rather than managing them.)
The case about nuclear energy is one that I always think to myself, but how does it explain non USA (or non Western for that matter) nuclear prices? Since they don't have the same regulatory hell shouldn't Russian or Indian or Chinese nuclear electricity should be so cheap they'll just be selling the rest of the world electricity? I really want more research made on nuclear so we make most of our electricity out of it (and some architectures like MSRs using Thorium etc theoretically can both decrease the prices and increase safety), but also somebody needs to write an article why it's still expensive in other regulatory environments. One more caveat about Chernobyl, there's no way in hell there's so little casualties. Just the extra cancer deaths from the Black Sea coast of Turkey is way more than that. From the top of my head I'd expect in total worldwide somewhere around a few tens of thousands to maybe a hundred thousand dead?
Re: Flying cars well safety regulations are one thing when the vehicle can just stop in case there's something wrong and stay where it is, and another thing when it just falls off the sky. I want a flying car more than anybody but (maybe it's my lack of imagination) I cannot see a world when they're not falling off the skies like flies.
Is the public R&D spending somehow stop the private investors from investing in nanotech? How can the public be blamed for this?
Overall I think this book sounds like one where the ideological fervor of the author clouded his judgement so heavily it doesn't make much sense.
>but how does it explain non USA (or non Western for that matter) nuclear prices?
Well it is a complicated beast to wrangle, the nuclear power plant. I would imagine high prices outside the western world are high because it's a highly technologically complex thing to make. But China is building nuclear plants anyway, we'll see over the next few years how long it takes them to make and how expensive they are to run but even those datapoints I don't think will be that generalizable because China is kind of a special case in the world economy.
But within the western world you have France, and they have among the cheapest electricity in Europe. The difference between France and US nukes is that France picked basically one reactor design and built a bunch of them. By contrast, basically every plant in the US is a special, one-off snowflake. The US used its nuclear energy program as an input to its nuclear weapons program, they were essentially treated as science experiments to produce useful weapons-grade material, the more the better. They were not optimized for cost, they had a healthy chunk of the defense budget.
The electricity prices of France are in the normal range within Europe.
https://www.cleanenergywire.org/sites/default/files/styles/paragraph_text_image/public/paragraphs/images/eurostat-average-electricity-price-households-2ndhalf-2018.jpg?itok=vbrZx8N1
In general, I don't think prices in Europe suggest that nuclear energy is cheaper. Belgium has the forth-highest fraction of electricity produced by nuclear, and the second-highest electricity prices. (Numbers slightly outdated.)
https://energypost.eu/wp-content/uploads/2016/04/Svr-nuclear-energy-in-the-mix-EU.png
You're right I overstated it; France's power is cheap for nuclear, average overall. But I still think the general argument holds: 70% of France's power is nuclear, the highest anywhere, offering an existence-proof that a mostly-nuclear grid can provide clean electricity at reasonable prices.
One just has to figure out what France did to achieve that and replicate it, and certainly good debate could be had on what the causal factors were, but I think there's a good argument that it was the use of a common and consistent design that led to plants being fast to build and cheap to construct and operate.
Yes, I agree. It proves that nuclear energy is competitive. And nuclear energy might have been yet cheaper in the counterfactual world where the US and the EU went even stronger for nuclear power.
It just does not really support the extreme position of the book that this would have blown us into a new age of cheap energy.
France is often touted as a success story for nuclear, and yet it is planning to gradually phase out nuclear plants over the next few decades (even while attempting to decarbonize). It certainly isn't building new standardized nuclear plants hand over fist. Nuclear plants in the US were cheap at one point, too. I think France has experienced the same trend of increasing nuclear construction costs (though not so extreme), and then more recently public opinion there has turned somewhat against nuclear because of Fukushima.
Fukushima is part of it. France is subject to EU regulation, and for a while there it was looking like the EU wasn't going to consider nuclear a "green" energy source, qualified for public funds, after what looked like lobbying from German LNG. Within the last year the EU commission decided that nuclear was a valid thing to invest in in order to meet climate goals, and so I expect France to switch tracks on their nuclear divestment plans.
> Since they don't have the same regulatory hell shouldn't Russian or Indian or Chinese nuclear electricity should be so cheap they'll just be selling the rest of the world electricity?
Good question. I've been wondering if there are international treaties (regarding fuel, reprocessing, nonproliferation and safety) that all-but-guarantee that nuclear power as traditionally built will be expensive for any country that does it. Another thing is that maybe the overblown fear of nuclear power is a global phenomenon (I noticed it a little in the Philippines when I was there - maybe not the best example, as the influence of America is pretty obvious there).
The runner-up review says of why we don't have flying cars,
"Is it because of air traffic control being difficult? No, not really. Because the skies are vast and there's plenty of room up there, what with three dimensions to play with and no "road" restrictions."
I'd really like the explication of this, because I find it hard to believe. For one thing, the flying cars wouldn't be scattered evenly throughout the air. They'd all be congregating towards the places where people want to go. 2-dimensional traffic jams are bad enough as it is: 3-dimensional ones with no road markers would be far, far worse, and more dangerous, because there effectively wouldn't be such a thing as a minor collision.
Ok, but imagine we all did have flying cars. Would we live so close together? Perhaps there would be urban areas without flying cars, but everyone in the "suburbs" would have flying cars and those suburbs would be a long ways away. For instance, imagine all the Southern California suburbs existing far into the desert. You'd have all new sorts of new cities pop up all over the country (maybe even Pop-Up Cities), because someone could start a new entertainment/business district wherever, as long as there was plenty of airplane parking.
Couple the idea of flying cars with remote working. People would still want to get together, but the geographic considerations for everything completely change.
Two-dimensional ground autos enabled us to live not so close together. Our homes spread out enormously, with Consequences. And it didn't prevent traffic jams.
Those consequences were a net positive, right? Cities like Houston designed entirely with autos in mind don't have bad traffic jams. The bad traffic jams are in cities not designed for autos or places with geographical constraints.
No.
Also, that's a very limited definition of "bad" in regard to traffic jams.
Anyway, imagine the entire American West, currently almost unpopulated, as the suburbs of America. In the West, everyone has a private plane and an enormous plot of private property. Meanwhile all these entertainment/meeting hubs also pop up all over the desert. There's little law and what law there is made mostly by ray-gun.
Those who want to live in cities can live their separate lives in cities. They don't have to own planes or cars. They can bicycle everywhere and take trains. They don't even have to carry ray-guns if they don't want.
Wouldn't that be great? Everyone could live something closer to their ideal than now.
"For instance, imagine all the Southern California suburbs existing far into the desert"
And now I'm imagining the first American Water Wars. It's bad enough (seemingly) trying to manage the demand for water in California, given that agriculture guzzles a lot and agriculture is important for feeding the rest of the nation, as it stands. Houses sprawling in the desert all wanting the same kind of water usage they do today, competing with the existing demands?
Only about 10% of California water usage is for houses[1]. And there are quite a lot of Californians (like me) who think the problem is less one of exploding demand than of government that hasn't proposed or constructed a single significant water infrastructure project -- or even properly maintained those that already exist[2] -- since the 1960s, over a time period when the population of California doubled. The state loves the tax revenue from growing population and economy, but they want to spend it on things that are much more fun than water projects, with their inevitable painful political struggles, long time horizons, payoff in the far distant future.
There's tons of water in California, but it's all in the Sierras, of which only a small amount is current diverted to human needs (e.g. the Hetch Hetchy project provides 80% of the water for the entire Bay Area from a watershed that spans at most about 1200 km^2, something like 2% of the Sierra Nevada). But that water needs to be collected and transported to where it's needed, since the natural rivers don't do it, and that's where water projects come in. There were many in the early decades of the 20th century, and they were big and bold and economical, and there's a good argument they did a lot to enable the explosion of the California economy and population post-WW2. But since then...zip.
One can make the argument that you should not build cities as big as LA or San Diego in the desert, where there just isn't enough local water to sustain them, but if you *do* accept the proposition that it's OK to do that and ship the water in from the mountains, then you kind of have to keep up with the times, as the population and economy grow. This the California state government has not done.
-----------
[1] https://www.ppic.org/publication/water-use-in-california/
[2] https://en.wikipedia.org/wiki/Oroville_Dam_crisis
4. At $64 per day, the one billion people at Level 4 own a car.
Wow, $64 per day. That is within single digits of my income (US SS). Except I can't.afford a car. Or perhaps I should put it this way: If I had a car I would have to stop helping many of my Haitian friends. Since I exist quite well without a car, that is a no-brainer.
There were only 43 deaths at Chernobyl (assuming that estimate is right) because they undertook an expensive and difficult mitigation project, and more or less permanently evacuated an area 2600 square kilometers in size. That's from a single reactor disaster, with a plant design that was being run with cost first and foremost in mind (the main virtue of the RBMK reactor was that it was cheap to build and operate - also why there was no containment structure at the time of the accident).
That's why nuclear power is heavily regulated - the potential downside is enormous, and very much not something to blow off casually as exaggeration. The Soviets actually got lucky in that the disaster happened at Chernobyl rather than, say, St Petersburg (which also had the same type of reactor).
The stuff about the private sector and nuclear power rings false to me as well. The only time nuclear power ever got cheaper either in the US or abroad was when it was being built on a huge scale, by large, quasi-monopolistic utilities with implicit state backing and support (or explicit). Even in the US, nuclear power got a huge boost from having its liability limited in the Price-Anderson Nuclear Industries Indemnity Act in the 1950s - if they'd had to actually accept the liability for the plants in full, they likely would not have been built at all.
Incidentally, the French are running into cost overruns on new nuclear plants too.
I'm surprised nobody has mentioned the noise issue with flying cars. Choppers and aircraft are not quiet.
The first part is not even close to being true. The only people who died of radiation sickness at Chernobyl were extremely close to the damaged core itself, i.e. operators of the plant in the immediate aftermath of the explosion, and the fire crew that came to put out the fire on the roof of No. 4 reactor (which was covered with core debris).
The evacuation of Pripyat and vicinity 36 hours later was precautionary, in case of further explosions, and because the amount of radiation released was then unknown, but I know of no evidence that radiation levels that far away from the plant were ever high enough to produce radiation sickness. Had it been, there would surely have been hundreds of deaths from radiation acquired before the evacuation. Mind you, something like 3,000 people refused to evacuate at all, and there was no massive radiation sickness among them.
Chernobyl was bad enough without exaggeration, which only serves to deepen cynicism about nuclear criticism.
An explanation based on the establishment of the department of energy only really works if you ignore the world outside the USA. Arguably similar pressures exist, but there are many places, e.g. France, where nuclear makes up a much larger share of their energy supply. And while they've gained in efficiency due to the benefits of mass production and expertise, there hasn't been the major shifts that would be necessary for this theory to pan out.
A more parsimonious explanation would be that, as in many things, each additional marginal unit of energy is more difficult to acquire. And the returns on additional energy usage in terms of increased quality of life are diminishing. (A flying car would be cool but didn't make a difference to your life in the same scale as having shoes).
I would guess the returns on marginal units of energy don't diminish quickly. Energy is an input to the cost of almost everything. Every good you buy has energy costs in its production and its transportation to take it to market. Services have energy costs as well. Housing has energy costs. Computers have energy costs.
A reduction in energy costs is an increase in wealth for almost everyone.
Putting aside the specific merits of flying cars or nuclear power, I think the author is onto something about the change in our society.
Specifically, risk tolerance. Somewhere along the way, we changed from a society that favored growth and change, to one that favored safety and stability. All sorts of things in the 50s and 60s would be considered horribly unsafe nowadays. Nuclear power plants and the Saturn V rocket are big, obvious examples. But there's also smaller examples like car safety measures, drinking, smoking, and letting kids roam freely outside. Or banning lawn darts, for a really trivial example.
I'm not completely against safety measures- they really do save a lot of lives, when you multiply them over a huge population. But there needs to be a countermeasure too- some things are worth risking a few lives over, and I think transformational changes in technology are one of them.
I wonder if the difference is the generational memory of war? The decision makers in the 50s and 60s were people who remembered ww2, and many had fought in it. Compared to that, a fatality rate of 1/100 might seem acceptable, or even trivial. But everyone since then is either a civilian, or used to extremely limited wars where relatively few people die. Now we would think of a 1/100,000 death rate as unacceptable. That's my crank theory, anyway.
And the lower risk tolerance has potentially massive (and unaccounted-for) hidden costs. As life gets really comfy for all the new homeowners, it's really easy to lower their appetite for risk. After all, they don't need a ton of new growth, they are quite content with a much higher living standard than they had growing up. This seems like an almost inevitable trap for a wealthy society to run into.
Yes, I like your point about risk tolerance. Not just memories of war, but memories of much harder times. If you grew up getting kicked in the head by mules, or just having them step on your feet, and your dad brought gas from the pharmacy in a bucket for his flivver while smoking his pipe- flying cars might seem no greater risk. All the men in Norfolk jackets who died to make heavier-than-air flight possible would say flying cars sounded safe from their end too.
Radiation risk was taken a lot less seriously before the brilliant martinet Rickover was in a place of power over nuclear power plants. In the 1930's my granddad's brother used to scare my mom with his giant grey Frankenstein hands- he was a doctor experimenting with radium. But how many people think of the risk of nuclear power or flying cars without being influenced by moral panics whipped up by lawyers ?
Yeah, it's probably not just the war, but the general lifestyle. *Everything* back then was more dangerous, especially jobs. That would fit the general pattern that we see, not just in the US but in every country. As their economy develops, everything gets safer, and then people become unwilling to take the risks that are necessary to push things futher.
Indeed. One thing in the review made me smile : the idea that small planes were not more dangerous that motorbikes would somehow show that regulation of planes is not normal. I think it's the other way around: motorbikes in current society are an anachronism, living on borrowed times: they have not yet been regulated to death, because they were already well established before modern safety-focused world, they trigger nostalgia and romanticism which protect them a little bit from being anathema to women, and dealing with angry bikers is not worth the trouble. But they are killed, one segment at a time: for example dirtbikes in europe, the easier target, have moved from a not so small hobby to specialized extreme sports (the Instagram-hosted refuge of something that was much more common at one time)
> All sorts of things in the...60s...horribly unsafe...nuclear power plants...are big, obvious examples.
1960s reactors weren't all that great. But successfully stopping nuclear power was a boon to the coal industry. If all the coal plants were replaced with nuclear plants, and if every one of those nuclear plants had Fukushima-style meltdowns, it would probably cause far fewer deaths than coal does, even assuming that people around the reactors do not relocate: https://www.reddit.com/r/nuclear/comments/jtm6hm/how_bad_is_meltdown_world/
It mentions in passing "stagnant wages" and I can't resist swatting that one down every time I hear it.
The stagnation of wages in the US is a frequently repeated myth with a kernel of truth. Real median *personal* income has risen a lot since 1981. *household* income growth has been smaller because the number of earners per household declined due to huge increases in fatherlessness. Which is a big problem but not the one most people have in mind when they're complaining about "stagnant wages".
https://fred.stlouisfed.org/series/MEPAINUSA672N
https://fred.stlouisfed.org/series/MEHOINUSA672N
From those graphs, household income appears to have risen a lot since 1984, too. About 30% increase for household, 50% increase for personal.
The average commute in the USA is 27.6 minutes. And you can do that while listening to a podcast, or great music in a climate-controlled environment. I don't see much economic benefit to shaving time off a moderate commute, as it is unlikely to increase productivity for the average Joe. Small planes are noisy, and it would be obnoxious if there were runways everywhere and constant air traffic. And of course, any kind of "weather" would ground all the small planes. And small aircraft are dangerous and we are increasingly risk-averse. So perhaps physics and cost-benefit calculations are the reason we don't have flying cars yet.
As a proponent of nuclear energy, I really want to tell this guy to get off my side.
Also: TSCM is working on 2 nm process semiconductor manufacturing. If that isn't "development of nanotech", I'm not sure what is.
Finally: An average car (according to Google) is 1500 kg. If I assume very modest flight altitudes of 100 m (which to my understanding is unrealistically low?), that's 1.5 sticks of dynamite worth of gravitational potential energy *before considering the car's speed*. If flying cars become widely available where I live, I'm moving. For the same reason I'd move if they started selling dynamite to use on the 4th of July.
On nuclear I agree, he's a little too gung ho, makes the rest of us look bad.
For nanotech, the originally idea of nanotech as an industry was "atomically precise manufacturing of arbitrary things." Semiconductors are an extremely specific thing. The idea of nanotech was to treat it like a Manhattan Project, or at least like fusion: put a lot of money and effort into this long moonshot, even if it takes many decades. The idea was that, if we had good nanotech, making 2nm semiconductors is just a (highly nontrivial) design problem of figuring out where you want the atoms to go, but figuring out how to get those atoms in that arrangement would be a paved path.
I'm not sure if it was ever a coherent goal, it certainly would be more complicated than "use one factory to build a smaller factory, and use that one to build a smaller factory..."
Isn't it physically impossible to have a small nuclear reactor? To sustain a nuclear reaction, you need an amount of fissile material that's a good fraction of the critical mass. The critical mass of uranium-235 is about 52 kg; for plutonium-239, about 10 kg. The Kilopower project, intended to produce nuclear reactors for space travel, is probably as small as you can get a nuclear reactor, and their 10 KW reactor still weighs 1500 kg.
Of course, no government in their right mind is going to let private citizens own enough U-235 or Pu-239 for a nuclear bomb, and it takes the resources of a country to produce that a critical mass of those isotopes anyway. A heavy water reactor that uses natural uranium has to use much more uranium. I'm not sure how small you can get a heavy water reactor, but the KANUPP nuclear power plant is one of the smallest in the world, and it still has 30 tons of uranium. Of course, in addition to the uranium, you also need the moderator, control rods, some way to turn heat into usable energy, and shielding (if any humans are going to be around). I'm not convinced it's physically possible to create a nuclear powered car, or even a nuclear powered airliner, with natural uranium.
You need to be at or above criticality, although reactors can stay in the zone between criticality and prompt criticality that weapons must be above. So yeah with the shielding, cooling, and power generation accoutrement I think it's pretty impossible to build a reactor for something as small as a car. Maybe a big airplane would be good, though: an A380 weighs 360 tonnes empty, so adding 5-10 tonnes of reactor and subtracting 200 tonnes of fuel would probably be a win.
Nuclear-powered airplanes were studied during the Cold War as a way to allow bombers to remain airborne indefinitely. They didn't get beyond some very early tests before the development of ICBMs made them obsolete, but they're probably workable. However, I can't find much information on the reactor design they used, so I don't know how much harder it would be to do it with non-enriched uranium.
How do you do it at all with natural uranium? Did you mean not using highly-enriched U the way the Navy does? That does seem dubious. So it has to go into a pretty stout containment vessel, to survive the inevitable controlled flights into terrain, and we need some special licensing requirements around the fuel handling so none of it gets diverted to North Korea ha ha. Maybe we should have nuclear-powered dirigibles instead.
"Early tests" isn't quite a fair description. They built two working nuclear jet engines, which are sitting in a parking lot near Idaho Falls (if you're ever nearby they're worth a visit!). They certainly didn't use natural uranium, but I don't think it was weapons-grade either. The main problem wasn't the weight of the reactor, it was the neutron shielding you would need to have for the crew. Also, the risk of fuel degradation that could cause it to shed fission products in the exhaust.
Another partial explanation for the less adventurous, risk adverse climate since the 60-70: the aging of the population, together with a feminization of political and economical power. Even if not so strong at decision making level (which was always old - ish and remained largely masculine), political support and typical consumers changed for older and more feminine, so technology directly aimed at consumers changed targets, and regulations followed a much more risk averse population. Even for IT, this can be seen a little, with transition from pc and gaming to smartphones, e-commerce and social media, with supporting server infrastructure hidden from the public.
Isn't the phrase: "Dude, Where's My Flying Car?"
http://www.2blowhards.com/archives/000830.html
Maybe stick to your own racist blog, you dick licker.
Enchante ...
I'm not a fan of Sailer either, but there's nothing wrong with licking dicks.
The extreme irony of someone waving their flag about how superlatively anti-racist they are, then using - for triple irony points during the month dedicated to Pride - a homophobic insult? I suppose I could put something in here about "pot" and "kettle" or do a dissection on why licking dicks is a shameful thing (if we're all liberated now and are no longer homophobic).
But instead I think I would much prefer a higher standard of invective, or at least some imagination in the Billingsgate - "you dick licker" is something I'd expect a ten year old today to come out with in the playground, and promptly be hauled in to sit through a Special Class on not using naughty words, Tommy.
Even in 1760, this was thought not high-quality wordplay: https://media.sciencephoto.com/image/c0173597/800wm
Sure, triple irony points. But did he stick the landing?
"Greetings, sports fans, and welcome to the qualifying rounds of the Insult, Invective and Rhodomontade Championship! Tonight will see only four contestants moving forward to the semis, so stiff competition here!
Hi, I'm your host, Rob Ennybodie, and this is my co-host for tonight, Bob S. Ure-Unkell. Bob, what do you think of our first competitor?"
"Well Rob, Jack Wilson doesn't have much of a reputation built up here yet, so there's all to play for. We'll have a better idea once we see who he's drawn against - and it's Sailer! Steve Sailer".
"A controversial pick there, Bob. Soft target or water off a duck's back?"
"It's anyone's call, Rob. It all depends on Wilson's opening move".
"And he's started with a third-party appeal about racism".
"A solid opener, if unimaginative. The conventional choice, you might say."
"Maybe Wilson is holding back for the semis, Bob?"
"He has to get there first, Rob".
"Let's see what he brings for his follow-up, Bob."
"Oh dear, and it's racism again. Mere repetition is not going to cut it at this level, Rob".
"But a strong close can still see him pull through! And he ends with - "dick licker". Well, a disappointing finish there, Bob".
"Agreed, Rob. Juvenile abuse may work on the playground, but it's a different game at this level. Even the classic "gobshite", as in the friendly between Carragher and Neville, https://www.football365.com/news/neville-gobshte-carragher-car-park-debate-liverpool-manchester-united, would have brought this up a badly-needed notch. To have any hope of winning, then turning your back on with a sniff of disdain, the grand trophy for the Championship of Insult, Invective and Rhodomontade, you need imagination, you need flow. Inventiveness and novelty are the hallmarks of a true champion, but even then a solid, workmanlike, steady, conventional performance might have been enough - had it not been for the weak ending.
"A pity, Bob; we were all hoping for an entertaining bout. Well that's it for tonight, fans, I've been Rob Ennybodie and this has been Bob S. Ure-Unkell for the I, I & R League! See you next week!"
Grin... +1
"Tune in next week, sports fans!" Glad it entertained you 😀
If one wants a "high standard" of invective against Steve Sailer you can simply point out that Steve Sailer made false pedophilia claims (he never marshalled a single shred of evidence) against one of his ideological opponents.
To be fair, there's nothing in Jack Wilson's post that implies that licking dicks is a bad thing.
My dad's first job as an aeronautical engineer in the late 1930s was designing a small part for a flying car company. It was going to a tricycle road vehicle/aircraft.
He said the project was coming along pretty well. But then the FAA decided that having only one wheel in the front was too risky: while driving, you'd run your front wheel into a curb and put a hairline crack in the front wheel/landing gear pylon and then the next time you are landing it, the lone front landing gear snaps off and you go nose first into the runway. But having two wheels/landing gears in the front would be too heavy to get off the ground. So the company sold their work in progress to the Japanese, who maybe were planning to use flying cars to invade Los Angeles, land them on the Pasadena Freeway. So my dad then got a job at Lockheed from the 1930s into the 1980s
Why do you allow such racist comments on your blog? My theory is that you aren't as big a racist as your commenters but you have made yourself beholden to racists so you can't afford to reject them.
I'm pretty sympathetic to the central thesis of this book. However, I must point out its glaring scientific weaknesses. First, it treats cold fusion as respectable science and insinuates that the scientific establishment covered it up. Second, it says that "material science [...] trashed the reputation of actual nanotech." In reality, nanoscale materials behave very differently due to quantum mechanics. So the dream of geometrically decreasing "factory" sizes based on everyday technology is wrong.
Re science funding, a very well-known microbiologist once told me his theory that the (rough) doubling of the NIH budget in the 90s had had a strong negative effect on the pace and quality of biological research.
His reason was simple: a ton of mediocre scientists were now able to get their careers funded, and to make matters worse many of them ended up in high-up administrative positions deciding where to direct future funding, and their decisions were often poor.
The reviewer summarizes the author’s basic claim as “The Great Stagnation was caused by energy usage flatlining, which was caused by our failure to switch to nuclear energy, which was caused by excessive regulation, which was caused by ‘green fundamentalism.’”
I find plausible the claim that nuclear hasn’t taken off because of regulatory burden, but contra some other commenters, my intuition is that the burden of regulation in nuclear energy should be comparable across countries, given that nuclear energy is already highly regulated at the international level.
On the other hand, I’m skeptical that the reason energy use has “flatlined” in some countries is because of the lack of expansion in nuclear power generation. Here is a chart for comparison of energy consumption per year for various countries:
https://ourworldindata.org/energy/country/united-states?country=~USA#how-much-energy-does-the-country-consume-each-year
If you try China, India, France, Russia, and the US, each of which has a home-grown nuclear power industry, you’ll find that energy consumption in the first two (especially the first) is increasing rapidly, while in the last three, it has flatlined. So I’m guessing that the relevant variables are rate of energy use and rate of economic growth, and while regulation must affect both, regulation of nuclear energy in particular seems unlikely to be so relevant.
While we are on the topic of energy, I recently read Energy and Human Ambition, which does some easy and fun Fermi calculations such as how long, at current growth rate in energy use, would it take us 1) to eat the Sun, the Milky Way, the visible Universe, and 2) to boil off the oceans:
https://escholarship.org/uc/item/9js5291m
In short, not long on a civilizational time scale, from which we can conclude that energy use will probably either flat line or collapse in the next few hundred years (a sub-exponential growth rate would buy a lot of time, but isn’t observed).
Thanks for the links!
I read the article yesterday (though skipping most of the exercises) and was quite impressed how easily our perception of the facts of live can be unmasked as self deception by a couple simple calculations. I always thought that I had a pretty realistic view of the world (having an engineering background) and was very optimistic that humanity could overcome future and present challenges with science and rational thinking. Since yesterday I have serious doubts, exacerbated by some comments here…
Flying cars? Sounds to me like dreams of junkies in an exstatic high, while the average American is already consuming resources more than 6 times than sustainable
Oh wow, I had no idea that Dr Murphy not only started blogging again, but even wrote a whole book, thank you for this !
https://dothemath.ucsd.edu/2021/03/textbook-debut/#more-1586
"I’m more convinced than ever that not embracing nuclear power was one of humanity’s worst mistakes (partially because I’m more afraid of climate change than Hall is)"
It occurs to me the failure mode of an overregulated nuclear energy is a reduction in the rate of technological progress. The failure mode underregulated nuclear energy is the backstory of On the Beach. If a country can build a lot of nuclear reactors, it can build a lot of nuclear bombs. If every country can do that, then international politics will get very interesting. At least for a moment.
Unless, of course, nuclear bombs are more regulated than nuclear plants.
1) It will be harder to stop people building nuclear bombs when plutonium, the essential raw material for making them, can be produced in any country on earth. Making it easier to make things will always make it harder to stop them from being made.
2) The failure mode of anti-nuclear-war regulation is always going to be nuclear war
I've heard multiple experts say you can't build plutonium bombs contaminated substantially with Pu-240, so if you want bombs the fuel must be removed and processed within 30 days. This is not normal in civilian plants and reactors can be designed to make such rapid cycling difficult or impractical.
Forcing people to build their nuclear power plants that way would be a case of regulating the nuclear power industry. And if you get the regulations wrong then maybe they will be able to produce weapons grade material after all. As I said, the failure of anti-nuclear-war regulation is likely to be dramatic.
Like most people, I'm not opposed to regulations in general, just regulations that unnecessarily increase costs. For the most part, affordable 1970s power plants were already unsuitable for bomb making, whether due to reasonable regulations or simply because they weren't interested in making bombs - see http://www.phyast.pitt.edu/~blc/book/chapter13.html
Generally a power reactor benefits from high burnup, opposite the low burnup desirable for bombmaking. In addition to his natural reason for reactors not to produce weapons-grade Pu-239, designers are aware that nuclear proliferation is bad, and therefore may make design decisions that make their reactors as unattractive as possible for bombmaking. Thorcon, for instance, talks about this explicitly as a reason to mix thorium and uranium together in molten salt. Unfortunately they don't really explain *why* plutonium would be hard to separate from thorium - I don't get it, any chemists want to take a guess? - but the point is that designers are actually disincentivized to facilitate proliferation, not just for some regulatory reason but because public opposition has historically been a serious risk to the nuclear business, and if they find something they can use to affordably to reduce that risk, they will do it.
Why would people designing nuclear reactors care about the opinion of members of the public? Members of the public are not their launch customers. And historically, the people prepared to pay top dollar for nuclear reactors have wanted them precisely because they need them to make nuclear weapons.
Regulations exist, from a certain point of view, precisely to stop people from buying what they want.
I'm a big fan of nuclear power and a big fan of flying cars. To the point where I basically designed my current (well, pre-COVID) lifestyle around the latter. But these are two great tastes that don't taste great together, and I'm wondering how the author makes the connection.
Energy is relevant to flying cars, but it's the wrong kind of energy. The driving requirement there is having to fit on the order of a megawatt of peak power into something the size of a mid-sized automobile (for VTOL performance out of a practical urban "parking space") with no direct connection to the power grid. There is no plausible nuclear option for that. Even assuming too-cheap-to-meter electric power at the wall outlet, your flying car doesn't get off the ground without some non-electric, non-nuclear power supply, or batteries better than anything that were plausibly available in the era the author seems to be talking about.
And there are other critical problems that aren't at all energy-related. Pilot training and/or automation, as has been discussed here already. Plus, disc loading and associated scaling issues mean that if you shrink a helicopter into something that can take off from a driveway, it will be dangerously (literally dangerously) loud, and also dangerous in that the downdraft will e.g. kick up stray beer bottles and accelerate them to potentially lethal velocity.
I like "flying cars", but it's not realistically possible to have them deliver people directly from their homes to their workplaces and favorite commercial districts. Nor is it practical for them to double as groundcars. This is fundamentally a last-mile problem. I can and do fly 95% of the distance from my home to my office, and we can *maybe* make it so most people can do that. Now find an answer for the last 5%, and if you insist that it be "we have to fly!", you're going to fail. Likewise if you think "because nuclear!" or "because nanotech!" is going to solve the problem.
For me, a folding bicycle in the back of the airplane works nicely. And quite possibly the reason we don't have "flying cars" is that the golden age of general aviation, when airplanes were affordable and airports were everywhere, corresponded with a period when bicycles and public transportation were both considered horribly, unacceptably lower-class. Now we have Uber, which might provide an answer but we've almost priced airplanes out of reach of the working class (don't even ask about helicopters) and we've closed most of the airports.
Nuclear power is a completely different discussion.
Can you say a bit more about "I designed my lifestyle around the latter" and "I can and do fly 95% of the distance from my home to my office"? Do you mean that you actually have a flying car?
I have an airplane (Grumman AA-5B Tiger) with a folding bicycle (Montague 20" frame) that fits in the baggage area. In Southern California of the 1930s through 1950s, that would very nearly have been a flying car - there were enough small airports, even in urban areas, that I could have flown to within easy bicycling distance of most anyplace I wanted to go. Most of those airports are now closed, but there's one less than three miles from my office. Enough that I can live far enough out that houses are affordable, and still have a reasonable commute.
Many people make the argument that government centralization of science is a mistake. The problem isnt that money is bad, it's the way the money was allocated. Scientific Freedom by Donald W Braben makes the same case but he has empirics to prove it. He got a company to agree funding scientific research in a way that basically gave academics free reign over the money (no grant process etc), complete freedom. Of course the academics were hand picked by Donald. They managed to produce 1 noble prize winning research and 1 company (>100m net worth). This was out of around 20 I think, which is pretty better than grant funded research. The problem with academic grant system is basically : 1) makes professors spend massive amount of time writing grants 2) very conservative with regards to projects it will fund (unless the prof is very imminent)
3) comes with strings attached that slows down the entire process
Does the author explain why flying cars didn't happen outside the US? My guess at a steel man would be that everywhere else is even more regulated (or, if not poor).
I imagine it's because flying cars are, functionally, planes (or helicopters) and those are regulated already quite heavily.
Being regulated didnt stop planes happening at all.
A very quick and rough estimate: up to 2020 there are about 287 million road vehicles in the US (this includes trucks and everything). There are 212,335 aircraft (this includes commercial airlines and Joe with his cropduster).
You tell me how the US scales up from 200,000 to 200 million craft flying around.
"Then there should have been nuclear fission and nanotech, letting you fit a lifetime's worth of energy in your pocket."
and also
"Nanotech, Hall says, is to nuclear energy as the steam engine was to coal - the technology that will unlock the potential of a new energy source."
This statement causes me to update massively in the direction "The book author does not know what they are talking about".
While I am not a nuclear engineer, a central concept in nuclear reactor design seems to be criticality, which is the point where one neutron will, on average, cause exactly as many fission reactions as needed for another neutron to be released. For different fissile isotopes, you would need different amounts to sustain a chain reaction, generally upwards of 5kg. You can use neutron deflectors and the like to lower it a bit, but in the end, that is a minimal scale for a criticality-based nuclear reactor.
You might or might not be able to use nanotech to scale down a cars internal combustion engine to fit into a hair, but you will definitely not be able to do the same for a submarines nuclear reactor, or even fit the reactor into your pocket.
(On the bright side, this is also why scaling down the mass of Little Boy by six orders of magnitude will not yield explosive small arms ammo with a blast of 15kg of TNT.)
I have tried -- and failed -- to parse the first quote in a more charitable way, e.g. that nanotech would provide the batteries to store the nuclear energy, but that would just mean that nanotech would provide energy densities similar to nuclear, which is hardly an easier sell.
For me no amount of Gell-Man Amnesia can overcome the scepticism about a book with its core argument based on "nuclear power + nanotech = reactor in your pocket".
I thought it was "the fuel for your lifetime would fit in your pocket," yet you would still need an at-scale reactor to process it.
OK, now I'm sure about how I feel. There's just too many entries for me to judge and it seems almost unfair to the entrants at this point. If you do this again next year (and I hope you do!) I would urge you not to go above 5-7 finalists.
Yeah, I’m running up against the seven plus or minus two issue myself.
https://en.m.wikipedia.org/wiki/The_Magical_Number_Seven,_Plus_or_Minus_Two
I agree, I feel like I've already forgotten many and will be unsure how to vote due to how long this has gone and how many there were. As a contest to vote in, not great.
In terms of content, though, this is great! I'm loving these book reviews and there are tons of them.
Yes, good point about the content, I love it too. But as you say it doesn't work too well as a contest
For me the contest is a side show. The main event is finding good books to read. Great stuff! Thanks again to all the authors.
Yup, lots of good suggested reading here and even more in the Google doc, I'm sure.
I liked the review of "Where's my flying car" on the google docs better than the one above. A bit more critical of the author.
On the other hand, even if it makes the voting less reliable, I bet for at least some authors getting the review posted on ACX and the readership that comes along with that is a prize in itself. Personally that would be worth at least $1k to me. And as a reader, I'm glad for the larger number of reviews getting posted - while in theory I could go to the google doc and read all the reviews, or a curated 'top 20' list from Scott, in practice I'm much more likely to read a review if it's posted to the blog.
Fair point about the exposure for the authors. I wouldn't mind lots of "special mentions" being published outside of the contest but I feel this is unmanageable for me. Maybe just do around 5 or so posts for the main competition and publish a select number of worthy entries that didn't quite make it once a week for a while after the voting is done?
That's only a problem if you have to deliver all your ratings at the same time. They should instead be rated independently on a 4-to-10 scale, so you don't have to remember how you voted on the last one to vote on this one. (why 4 to 10? because come on, they're all finalists, they can't plausibly be below 4, now, can they? now, your preferences/feelings might not be constant in time, so that you might well feel that the same review is a 5/10 in May but 7/10 in June, but given enough voters, such inconsistencies ought to average out to zero)
The phrase "green fundamentalism" really raised an eyebrow for me here - it's very hard to seriously claim that the US regulatory environment is even adequately protecting the environment, let alone overdoing it. Americans by and large live in a very toxic environment, and the lack of progress on climate change is an obvious red flag against this argument.
"Climate change will only cost a few percent of GDP" is also a HUGE handwave of trillions of dollars per year in additional *permanent costs* added to the US budget alone. See also: human suffering, likely abandonment of some affected populations, and extinction risk. That sort of claim makes me doubt his intellectual honesty overall.
What are the trillions of dollars per year in additional permanent costs?
The estimated annual cost of unmitigated climate change, accounting for hurricane damage, real estate losses, energy sector costs, and water costs. https://www.nrdc.org/sites/default/files/cost.pdf
Can you choose one of those things, perhaps real estate losses, and prove that there will be a *permanent* trillions of dollars per year cost?
I'm not going to read a 62-page report right now, but from skimming, I didn't see anything about the benefits that climate change would bring the U.S. I think these could actually outweigh the costs, especially if people stop being stupid and move away from the areas of highest hurricane threat.
The areas of highest hurricane threat are also the most populated parts of the Eastern half of the country.
They are not. You think New York City and Boston and Atlanta and Chicago and D.C. etc. are the areas of highest risk of a hurricane?
"We will get more hurricanes" is the kind of thing that gets repeated by people who listen to what journalists say the scientists say.
It is not what the scientists say.
https://www.gfdl.noaa.gov/global-warming-and-hurricanes/
From your source:
"Tropical cyclone intensities globally will likely increase on average (by 1 to 10% according to model projections for a 2 degree Celsius global warming)"
"The global proportion of tropical cyclones that reach very intense (Category 4 and 5) levels will likely increase due to anthropogenic warming over the 21st century. There is less confidence in future projections of the global number of Category 4 and 5 storms, since most modeling studies project a decrease (or little change) in the global frequency of all tropical cyclones combined."
To me this sounds a lot like "We will get more [disastrous] hurricanes".
Yes, but not more hurricanes. Climate science is complicated, so accidental misrepresentation is very common.
This book has attracted many reviews; there's a (long) review here: https://strataoftheworld.blogspot.com/2021/03/review-where-is-my-flying-car.html
And also this one: https://rootsofprogress.org/where-is-my-flying-car
One of the most interesting points I think is why no country has done better at keeping alive a good regulatory environment and encouraging innovation in atoms. Was the US an outlier before, and does the trajectory of the US therefore have an outsized global role? I also recall some Robin Hanson post about elite consensus across countries being surprisingly robust (regardless of whether it is correct), and maybe this effect has grown with globalisation and the internet, and means that policy between countries is much more correlated than we might think.
Or are some countries succeeding? It looks like Iceland and Singapore have bucked the plateauing energy use trend: https://ourworldindata.org/grapher/per-capita-energy-use?tab=chart&country=DEU~JPN~SGP~SWE~TWN~GBR~USA~ISL~CHN
> we can and have built them ever since the 1930s. They got interrupted by the Great Depression (people were too poor to buy private airplanes), then WWII (airplanes were directed towards the war effort, not the market), then regulation mostly killed the private aviation industry. But technical feasibility was never the problem.
It's because they kept making them again and again, and it was always a crappy solution to a non-existing problem.
Road vehicles and flying machines have different tradeoffs (weight is one). A flying car sounds good abstractly, but is the worst of both worlds.
The 20th century was the Pre-Cambrian explosion of technology, during which a single generation went from seeing people ride horses, to landing on the moon. Who wouldn’t be optimistic after that?
I can’t find the authors exact age, but he’s an older gentleman and I can assumed he grew up at a time when everything seemed possible. “Technology Progress” can become an ideology, and I think nothing typifies this better than the Jetsons.
I also find that sometime STEM types are not great at inspecting their own ideological biases, and so the author might just not be willing to give up on the future he imagined.
> I’ve never heard anyone before say government-funded science was bad for science!
You don't read much economics. This isn't even an uncommon claim.
Could somebody wake me up on wether nanotech (as defined in the book) is pseudoscientific or has real potential. The so-called Drexler-Smalley debate has made me think the former.
Nanotech seems to be mostly propagated by computer scientists and this is something Smalley reproaches Drexler for. I'd be especially interested in the opinions of trained chemists or physicists among us.
My hot take is that we SHOULD have regulated cars out of existence.
Driving is an extraordinarily dangerous activity, but we've just come to accept that tens of thousands dying on the road every year is fine.
Driving has also completely ruined the urban fabric of most American cities. People often say they strongly prefer to live in walkable neighborhoods, but the necessities of dealing with urban traffic flows makes walkability very difficult to achieve. We've gone from a highly social society where people live in tight-knit communities to a fully atomized society where people routinely never meet their neighbors, and though cars aren't the whole story, they're definitely part of it.
Cars exist as they do today BECAUSE governments stepped in to change the rules in favor of cars. Jaywalking was invented as a crime because car manufacturers wanted to crank up speed limits and not have to worry about hitting people all the time. So all of a sudden, the streets themselves were taken away from pedestrians and given to cars, with tiny strips along intersections allowed for occasional pedestrian use. Miles and miles of highway were built to make cars a convenient long-distance transportation option.
So yeah, flying cars are probably a bad idea, and I think on the whole we are better off for not having them.
Knowing your neighbors is generally a good thing that I wish I had more of, but I've also had neighbors that I really wish I had never known, because all their problems became my problems.
How would a city with no urban motor vehicle traffic even work? You'd have to carve out tons of exceptions, for commercial deliveries, construction, moving (imagine trying to get your couch from one apartment to another across town, if you can't throw it on a truck!). Or else you'd need some completely different last-mile transportation infrastructure suitable for cargo. Before cars, I think this stuff was basically done with wagons, which are just cars that suck.
I don't really mean that we should have had a total ban on all motorized vehicles, in the same way that today we don't have a total ban onaircraft. I think various sorts of trucks and vehicles would still be useful and necessary. But we should have regulated them to fit into the existing urban fabric instead of rebuilding the urban fabric around them. That would mean stuff like: all streets would be mixed use and vehicle drivers would be responsible for not hitting anyone, surface parking would not be heavily subsidized (in other words, parking would be expensive most places), and on a general cultural level, motor vehicles would be for industrial and commercial use, not something most people own and drive around in every day (unless their job was to make deliveries or something).
I could see that being nice. It's striking that there aren't really any examples of this in the US, when it's something that could absolutely be done at the local level. I think there are some neighborhoods in Europe that are sort of like this, mostly in downtown areas of old cities where the roads weren't designed for cars.
Probably the closest in the US that I'm aware of would be Mackinac Island, where they really have just banned all motor vehicles. But being a small island in the middle of the Great Lakes, it's hard to get a motor vehicle there in the first place, and it's not really scalable on a full-society level.
As a person living living in Europe I can only shake my head increduously:
You in the USA already have the most suicidal mass killings world wide, and you dream of putting another instrument/weapon in each garage allowing a frustrated pupil or employee to steer the vehicle right into the classroom or office?
One living too many in first sentence
Let me rephrase my previous comment:
You might not get another Nine-Eleven but quite a number of Four-and-a-half/Five-and-a-halves…
Cowering in fear of terrorists instead of living life to the fullest is letting the terrorists win.
I guess you‘re from the Land of the Brave, where people dare to leave their homes only with a gun in the Pocket …
But I wasn‘t talking about terrorists, I was talking about your fellow citizens
I don't know who told you that no one dares leave their houses armed but a majority of Americans own zero guns and get through their day just fine.
The gratuitous jingoistic nationalism is not appreciated. Yes, it's possible to use motor vehicles to kill crowds of innocent people. That may be worth considering. But doesn't seem to be a thing Americans are any more inclined than Europeans to do, so why bring that part up if you're not trying to stir up trouble?
Working biotech does exist -- it's called molecular biochemistry, and it's something we've made a lot of advances in since the 1970s.
As far as I can tell nobody is talking about the plasma fusion reactors that are planned (e.g. ITER) which would have no waste. I understand that fission reactors are already developed, but if the idea is to not repeat a past mistake, why not rebrand on a type of reactor that doesn't carry all that baggage.
"Nuclear Fusion Is Always 30 Years Away" as the joke goes. ITER was started in 1988 and they hope to have deuterium-tritium fusion by 2035. As ITER's info page says, "ITER will not capture the power it produces as electricity, but as the first of all fusion experiments in history to produce net energy... it will prepare the way for the machine that can."
In contrast, the companies designing new (fission-based) Molten Salt Reactors over the last 5-10 years hope to start their first reactors by 2030, and they are mostly privately funded, operating on small budgets, and intending to build actual power plants, not proofs of concept.
There's a bunch of discussion here comparing the results of crashes between flying cars versus non-flying cars. Is it possible to build craft that can fly around, crash into another such craft, get dinged up a bit, but still limp along in the air until it can reach a safe landing point? Or will physical contact with another craft always imply hurtling into the ground with more-or-less gravitational constant based acceleration?
Electricity was government funded. Radio technology was government funded. Metallurgy was government funded. The railroad was government funded. The automobile was government funded. Aviation was government funded. Interchangeable parts was government funded. Sure, the technologies were rolled out to consumers by private businesses, but the innovations were due to government research initiatives. The government has always provided the initial push and provided the necessary subsidies. Businesses can do just fine with stagnant technologies.
The problem with nuclear power plants was the spent fuel disposal problem. Even the French, who went all in on nuclear power and made it work, have been unable to solve this problem. Every nuclear power plant in the US has a big storage pool full of spent fuel rods and no place to put them. Sure, we could eliminate regulations and accept a higher background radiation the way we accept warmer, less settled weather in exchange for burning hydrocarbons. We'd also have to accept a landscape full of damaged nuclear plants. Finally, run the numbers. Solar is still cheaper and could become even cheaper.
As for flying cars, the roadblocks aren't energy and regulation. There are technological barriers. Unless we totally redesign our cities and suburbs, anything that requires a runway is out. Right now, vertical take off and landing means lots of noise which is a regulatory problem unless you live nearby. Then there's the airspace control problem. The airlines were the ones who got the government into airspace control. Technology can help; TCAS came out in the 1990s, but the general problem is unsolved, and the everyone-has-a-flying-car problem wasn't going to be solved with 1960s technology.
This review makes the book sound like an anti-government screed based on technological ignorance. I hope it's better than that.
Overall I find the review to be strongly lacking in critical analysis, indeed it is a summary more than a review with limited commentary beyond agreement. Ideas are summarised and well presented by the reviewer, but I don't think the original ideas hold much merit.
The author of the book asserts many things and I can only find agreement in this book review. This is not a totally wrong approach and perhaps the reviewer fully agrees with the book author on the major points they make, but I don't think this is a strong or suitable entry for the ACX community where Scott typically gives a balanced case for and against various aspects and arguments made in the book.
Indeed many of the arguments in the book lean into thoughtless mind killing partisan areas and are founded in ideology more than in sound logic. The reviewer overlooked the style, tone, and approach typical of book reviews in Scott's blog and in the fairly pro-logic and pro-analysis less wrong type community which is the audience.
Indeed the non-categorical reference to flying cars as in...actual flying cars rather than to technological innovation in general severely limits the scope and meaning of the book author's analysis. Besides some, misguided in my view, boomer style 1950's childhood dreams....do regular people need, want, or have a use for flying cars which likely have travel distances not hugely different from cars to take passengers no more than 300-400 miles without refuelling or recharging? I don't know of anyone living in any urban or suburban area whose life would be improved by such expensive and noisy machines.
We are having enough problems when considering the usage and approval of small unmanned drones flying over residential areas to deliver packages and takeaway, along with the problems they pose from noise pollution and hazards of drones falling out of the sky causing damage to people and property. Several pilot studies of drones flying around at over communities to do this was strongly strongly disliked and hated by the communities subject to the endless noise of drones flying overhead. Meanwhile we could have networks of autonomous trucks with boston dynamics style robo-dogs going around to delivery small packages instead with existing technology or very near term technologies....or you know...just humans delivery drivers in trucks.
I can certainly say I'd be 100% against my neighbour having a flying car or a helicopter pad in their yard and low and behold the allowance of helipads is very limited and nearly non-existent in residential areas. Such flying cars would be taking off and landing with huge amounts of noise everyday. Perhaps such vehicles could be forced to drive to specific take off and landing locations, but wouldn't' that just create huge congestion there at the mini aircarports if tens or hundreds of thousands of commuters wanted to use them? Along with the high cost of devoting land and airspace to such purposes. Would it be safe for such craft to fly over residential areas...and at what altitudes? Do I even care about the answer to such a question when the vehicles have nearly zero utility compared to their huge social costs?
A flying car is soo incredible unattractive and useless that it is hard to imagine why anyone would bother talking about them beyond as a toy for the wealthy or as a dream from their childhood watching the Jetsons.
There are a bunch of companies aiming for flying taxi networks within a few years.
https://www.google.com/amp/s/amp.cnn.com/cnn/2019/05/16/tech/lilium-flying-taxi-germany/index.html
I'm sympathetic to the libertarian perspective here, and It definitely seems reasonable that nuclear was smothered in the crib with no valid justification. But:
1. Government research -- Seems like a lot of the innovations of the 20th century came as a result of research funded due to world war 2 and the early cold war. Obviously though new government institutions ossify overtime and become less beneficial. Am i missing something here?
2. Flying Cars -- Are we talking about anti-gravity or glorified helicopters? What kind of noise pollution would we be dealing with? What would American levels of driving skill mean in terms of accidents and fatalities of these flying cars?
The book doesn’t sound as if it recognizes there are countries outside America. Japan and France went all-in on nuclear power, continuing that way for several decades after the 70s. Japan especially has obviously changed direction since, but France hit about 70% nuclear generation I think (much of the rest being hydro; a good combination as nuclear is base load and hydro can mostly be turned on and off).
Even if electricity is a bit cheaper in France, they don’t have flying cars.
Sure, the nuclear industrial base would be a lot bigger from the US adopting nuclear power than just France and Japan, but that doesn’t seem like something that should make quite the ‘all or nothing’ difference the book posits. Oh, and the Soviet Union built a lot of nuclear plant too, notably unconstrained by safety regulations. They didn’t develop flying cars either.
I am actually very pro-nuclear by the way. I just thought the book sounded a little blinkered: stuff does happen outside the US, after all.
> At $64 per day, the one billion people at Level 4 own a car. ... The average American moved from Level 2 in 1800, to level 3 in 1900, to Level 4 in 2000.
This rang weird to me, because 50% of American households owned cars in like 1925. (It went back below 50% thanks to the depression and WWII, I think, but above it again long before 2000.) If 2000 is when the average American hit $64/day, then "owning a car" might not be a good handle for level 4. That said, this might be "the average American", not "the average American household"; do 50% of Americans period (or at least 50% of adults) own cars?
I feel as though something is missed about radiation here. Radiation exposure as energy is almost certainly not a big deal at low levels. But the author seems to miss the fact that radiatioactive materials ingested into the body are a big deal. Walking past a light bulb is not going to harm you unless we are talking about some exceptionally powerful bubl pumping out a huge amount of light. Swallowing a lightbulb that stays lite inside yourbody for the rest of your life, though is likely to cause damage over time.
Netflix's show "Dark Tourist" has an episode where the host tours Fukishima...where "no one died". The tourists go about in the bus looking at the empty town and the host even sneaks off to go look inside some buildings where things are covered with dust. Over time, though, their geiger counters start increasing, going beyond what they were told is 'safe'...as they continue to rise, even when they confine themselves to remaining inside the bus, nerves start rising and the group leaves the area. This is years after Fukishima and the place is still covered in radioactive dust. Do I believe no one would die if somehow you hypnotised the entire population to have simply remained there and ignored everything? Not for a moment.
The fear of radioactive materials is more rational than the author realizes. They are cumulative over time. They are not able to be seen, smelled or tasted when they are in the environment and in a catastrophic event they are dispersed over a large area and essentially never go away in human history.
This is missed if you are concentrating on doses. In normal environments, you only get doses of radioactive energy (say by getting an x-ray or flying in a plane). Here you normally should not have much to worry about unless you have a major malfunction of the machine or you are exposed to a nuclear explosion for some reason or your finger slips while messing around with a 'demon core'. However if the reactor explodes, it isn't just the dose from the immediate radiation levels you have to worry about but asorbing the material itself.
Brief review-of-the-review:
Reasonably interesting and well written, but I don't feel like I learned that much from it-- which may be a critique of the book as much as the review. Won't be voting for this one.