804 Comments
Comment deleted
Expand full comment

Do you think I’ll get a tenure track job with my PhD thesis - Hitler Wasn’t Wrong About Anything - The True History of Europe 1920-1945?

Expand full comment
Comment deleted
Expand full comment

To be fair, his idiocy is slightly overrated. I recently watched a French documentary series about WW2, Apocalypse, and they portrayed him as reasonably competent, for a warmongering megalomaniac. His allies, in comparison, were actual idiots. Mussolini couldn't get anything done in relatively irrelevant Africa, and Japanese stupidly dragged the reluctant US into the war by strategically and tactically abysmal Pearl Harbor.

Expand full comment
Comment deleted
Expand full comment

How many warmongering megalomaniacs ever accomplished a tenth of what hitler did (for better or for worse)?

Expand full comment

Hitler was in many ways more competent than Stalin, who managed to squander much of the massive demographic and industrial advantage the USSR had relative to Germany in the early 1930s, gut his military leadership and bring his country very close to collapse in late 1941. People forget that the Tsarist military performed much better encounter for encounter against a far better trained and led German army in WWI. Soviet victory in WWII was mostly due to German exhaustion, and basically inevitable once the initial German gamble failed.

Expand full comment

Lend-lease also played an important part. Even considering all that, without the US invasion some sort of stalemate might've been reached, and Heisenberg could've eventually stumbled upon the bomb first. You really don't need too many counterfactuals for history taking a much different course.

Expand full comment

It's horribly tempting to speculate - suppose Hitler hadn't gone all-in on race theory? Suppose instead Nazi Germany treated Jews under an equivalent of the Jim Crow laws? Could the regime have lasted? I think it probably would have eventually collapsed, given the economic constraints, and if Hitler had still decided to go to war to restore German greatness.

But without the Holocaust, there would be a lot less easy way of saying "Fascism is Evil" and a lot more excuse for keeping out of the war, if it went ahead. Even with that, suppose he hadn't invaded Poland but had declared war on the Soviet Union at the very start? Is it totally impossible he would have not been on at least tenuous basis of alliance or "we're not going to interfere" with the Western nations for that?

Expand full comment

(Very weird, that double-posted).

Ahem. As I was saying:

He managed to turn his silly little boys' club into a party that became the government of Germany and a world power (even if for a very brief time). That argues some level of competence and ability, even given the chaotic nature of German politics at the time.

Expand full comment

> But I'm also not defending the Jews, who are trying to destroy my rights and liberties

wait what? Are you somehow convinced that they operate in some union against you?

Expand full comment
Comment deleted
Expand full comment
Comment deleted
Expand full comment
Comment deleted
Expand full comment
Comment deleted
Expand full comment

What counts as sinister and hidden? Not wanting people to believe that Covid is the fault of virologists seems like an entirely natural motive for virologists to have — but implies that claims by virologists that Covid couldn't possibly be a lab leak should be viewed with great suspicion. Attributing to the other two the motive of wanting people to believe that what they do is very important also seems natural, and a ground for at least some skepticism.

Expand full comment

You don't need sinister hidden motivations to be influenced by your unconscious biases. That said, saying "this work is probably biased, I just can't tell how, so I won't believe it" is seeing the speck in your neighbor's eye while not noticing the log in your own.

Expand full comment
Comment deleted
Expand full comment

> and lacks historical perspective

This seems subject to reference class tennis. What if the relevant history is all countries over all time? Modern countries in the past 50 years? Any entity that can commit violence upon it's subjects?

I don't know if you've chosen the correct reference class or not.

Expand full comment
Comment deleted
Expand full comment

CFS seems pretty solid to me. Their reactor design is very similar to conventional research tokamaks. It's broadly accepted in the field that energy-profitable fusion with a conventional tokamak design is a matter of scale, and that it can be scaled up in terms of size or in terms of magnet power. The big multinational government-backed project (ITER) is taking the "size" approach, since at the time it was designed, the best available superconducting magnets were near their limit in terms of the field density they could generate. CFS is taking the other approach, using newer and better superconductor materials to increase the magnetic field power. A lot of the senior people are experienced fusion researchers, many of whom had previously been involved with ITER. The magnet assemblies have been demonstrated at component scale, and the company has published peer-reviewed papers on their reactor design.

The other fusion startups are trying to develop varying degrees of novel reactor designs. Some are modelled off other research reactor designs that are somewhat well studied and show some promise at scaling up better than conventional tokamaks but which haven't been studied as thoroughly and thus have some degree of additional physics risk and engineering challenge. Tokamak Energy uses a spherical tokamak design (a variant of a conventional tokamak where the "donut hole" in the center is as small as possible, which is thought to scale up more efficiently but is challenging because there's a bunch of magnets and stuff that need to go in the donut hole).

General Fusion and Helion are using different hybrids approaches between magnetic confinement and inertial confinement. Both of their approaches seem to be theoretically sound as far as we know, but they're a lot further from the beaten trail and there's considerably more risk of unknown unknowns rearing their ugly heads on either the engineering or the physics fronts.

Expand full comment

"For a long time, fusion energy has been 10-20 years away. Now it is 3 years away."

Given that the first part of this is true - they have indeed been promising fusion for a long time - I am very darn interested in the second part.

It's a big, big claim and I wonder if by 2024 they will be coming out with "sorry guys, unforeseen circumstances, but in another 10 years we promise!"

Expand full comment
Comment deleted
Expand full comment

So they've already hit the "yeah, three years was too optimistic" obstacle?

Expand full comment

No, about 7 years ago Helion was claiming that they would reach breakeven within 5 years. They walked that back about 4 years ago citing a combination of failure to secure sufficient funding and unforeseen technical hurdles, they then went into stealth mode and haven't made a peep until this new announcement.

Expand full comment

I'm not convinced at all by 3, but this is the first time in a while I've been willing to consider that it *might* actually be, say, 8-10 years away. Priors still say probably not, but they've passed an "I'm not even going to bother paying attention" bar for me at least.

Expand full comment

Keep in mind that the original claim, back in the 70s, was "if you fund our research, fusion power will be available within 20 years". Over the last 50 years, there has been less total funding for fusion than they asked to get over those 20 years, so it's hardly surprising they didn't manage to succeed without funding

Expand full comment

If they were heading to any kind of success funding, private or public, would have materialised.

Expand full comment

None of the alternatives to the regular tokamak have had full scale facilities before so it's hard to say what their chances are. I'm no insider, though I've been soaking it up for a few months

I believe if you draw a straight line from Helion, TAE or General Fusion's experimental results to a full scale experiment you have a commercial scale output, probably with better economics than tokamaks. Experience with tokamaks and laser inertial however is that straight lines are not to be expected

CFS is notable in this regard in that they are a tokamak design with much more validated physics and with a new superconductor magnet that makes commercial levels of output possible. The physics will be tested by SPARC in about 2025 along with General Fusion's 70% scale facility and Helion in 2024. From what I can tell the economics aren't a home run but are possible. If you compare to coal with air pollution costs it appears the CFS tokamak concept is immediately competitive. Their ARC concept proposes a first of its kind reactor producing electricity at twice the cost of fission and similar to the first of its kind fission pilot plant

One of the CFS guys put it this way - the non tokamak concepts have bigger physics unknowns and the tokamaks have bigger engineering challenges

It's worth noting Helion isn't claiming a commercial level plant and I suspect not a sustained net electricity output either, but that a single pulse will put out net electricity. Their pulses are quite short. Their concept is really interesting because it proposes that the expansion of the magnetically crushed plasma will induce a current in the same electromagnets which can be captured as electricity at 85% efficiency. The other concepts go by the usual thermal conversion of heat to steam turbines at 40% efficiency

Expand full comment

It's hard to accurately judge where Helion is right now, since they've been very secretive. CFS comes to conferences and publishes papers (with simulations and not yet any plasma experiments, but their concept is backed by decades of knowledge from tokamak research.). General Fusion comes to conferences and has fewer papers, and they haven't made outwardly -super-visible progress in overall plasma confinement in the past few years. It's worth nothing that their 'Fusion Demonstration Plant' which they plan to build in the UK is not in line with what other companies or governments would call a 'Fusion Demonstration Plant': as I understand it will just be doing deuterium plasmas (no tritium) and will not be even trying to produce electricity; it's just a first integrated test of the plasma and driver system; this is kind of on par with where tokamaks were in the 1980s.

Helion energy is planning to do D-3He fusion, which is much harder from a plasma perspective (iirc about 5x higher temperatures are required). They claim on their website to have 9keV ion temperatures, but they don't (often?) come to conferences or (recently) publish papers and there's not (to my knowledge) independent validation of their claims. I do think that if their plasmas and machine engineering could work out, it would be a very exciting and potentially low-cost overall system. However especially on the plasma side this is a big 'if'.

Expand full comment
Comment deleted
Expand full comment

They might just be better financed now. If getting to fusion is a matter of scale, then having the resources to actually build a working plant to scale would make a big difference in progress.

Expand full comment

Yeah people bring up that the private companies have consistently said something like "working reactor in 10 years" and say this shows it's just hype but the statement fully fleshed out is more like "working reactor in 10 years with full funding if things go smoothly"

Private funding is now getting to the point where funding will be sufficient, this should especially become the case by around 2025 if one of the demonstrators gets the net energy Kitty Hawk moment. (This appears to be a high likelihood for CFS) Then we'll start to see what the roll of the dice is on things going smoothly

Expand full comment

The magnets are the big breakthrough. Older niobium-based superconductors can't handle much more than about 10 T worth of magnet fields without breaking down and losing their superconductivity and get really finicky as you approach that. You can (we think) make a working fusion reactor within that limit by making it bigger, but bigger means a lot more expensive, on the order of tens of billions of dollars. Expensive enough that no one government wants to put up all the money, so you get all the headaches of a government megaproject multiplied by trying to juggle cooperation and coordination between the US, Russia, the EU, Japan, Korea, India, and China. That megaproject (ITER) has been semi- languishing in government megaproject hell since the 1980s, finally broke ground in 2007, and is scheduled to start experiments in 2025.

REBCO-based superconductors have been around since the 1980s, but they're tricky materials to work with and it's only in the last 5-10 years that the engineering challenges of making a practical big giant magnet out of it have come close to getting worked out. REBCO stays superconducting at mich higher temps that Niobium alloys, liquid nitrogen rather than liquid helium, and it can handle much higher magnetic field powers (20-40 T).

This lowers the scale of the project by quite a bit, to the point where a project with ITER-level goals can reasonably be done with commercial funding.

A secondary factor is computer modelling. A ton of work has been done in parallel with ITER on smaller research reactors to flesh out our understanding of near-breakeven fusion plasma, which gives us a great dataset for physics simulations, and there's enormously more computing power available now than even a few years ago to do the sims with. It's the computer simulation capacity that's made all the novel reactor design projects feasible to attempt.

Expand full comment

Materials science has always been the most underappreciated science.

So many of the wonderful inventions in the world seem to come down to "Well, this idea has been around for ages, but we haven't had materials with the right properties until now".

Expand full comment

Cf. fully artificial hearts. Almost entirely a materials science problem.

Expand full comment

cf beanstalk. If we have a material that is suitable, then we could build a beanstalk for no more than what Apollo cost - possibly for about what ISS cost. Obviously, those are big amounts of money, but they are also big amounts that humanity is already demonstrably prepared to commit to space.

And a beanstalk would have a huge impact on access to space - it would make SpaceX Starship look expensive.

Expand full comment

People were talking in another thread about physics people getting government funding, but I think that funding physics and materials science is ultimately how we get all our advancements.

Expand full comment

Improvements in materials science and better computer modelling of high energy plasmas based on a lot more experimental data.

Expand full comment

I recently finished a Ph.D. in plasma physics.

I would give Helion <1% chance of success (meaning fusion (Q_plasma > 5) within 10 years), General Fusion <10% chance of success, and Commonwealth Fusion Systems / SPARC > 50% chance of success.

One simple thing to check is how engaged the company is with the broader fusion community. At the American Physical Society - Division of Plasma Physics (APS-DPP) conference two weeks ago, SPARC was discussed in two 30 min talks, nine 12 min talks, and eight posters. General Fusion was discussed in seven posters. Helion didn't have a presence there at all.

Three other startups that I think are worth watching are: Type One Energy (stellarator in Wisconsin), Renaissance Fusion (stellarator in France), and Tokamak Energy (spherical tokamak in the UK). Other startups might make the news, but I think that they can be largely ignored.

Expand full comment

"One simple thing to check is how engaged the company is with the broader fusion community."

Why is this a meaningful thing to check?

Expand full comment

Plasma physics and associated sciences are not simple and it's easy to trick yourself. As an example, there was an early experiment ( https://en.wikipedia.org/wiki/ZETA_(fusion_reactor) ) where scientists thought that their neutron measurements (evidence of fusion reactions) were from a true 'hot' plasma (indicator of success) but in reality they were due to effects from specific instabilities, which was evidence that their idea would *not* scale to a reactor.

Expand full comment

Fusion is complicated. There are lots of important sub-problems which the broader fusion community has had to deal with. Even if your design is very different, you're likely to encounter at least some of these. So it is valuable to tap into the decades of expertise of the broader community.

I should also note that my estimates are based on more than just whether they showed up at the conference. I'll write some more details about why I think Helion is implausible from a physical (instead of community) standpoint.

Expand full comment
founding

One obvious counterexample to this is the Wright Brothers. Aviation, ca. 1900, was a Very Hard Problem with a large community trying to simultaneously tackle hard problems in propulsion, aerodynamics, controls, and materials. And basically all following each other down the same wrong track (airplanes with two-axis flight controls and inherently stable in roll). The Wrights *followed* that community, and to some extent engaged with select members, but mostly hid in their bicycle shop and did their own thing.

It wasn't until four years after they invented the airplane, and two years after Santos-Dumont was credited for inventing something very charitably described as an "airplane", that the Wrights went public with an airplane that actually worked.

W/re fusion, there are a lot of people who have been doing much the same thing for fifty years with not much to show for it. And some people who have been trying novel approaches while deeply engaged with the mainstream fusion community and ditto. I'm not going to rule someone out because they aren't part of the Official Fusion Community.

Expand full comment

Not being engaged in the broader community is not sufficient reason to rule out Helion, although it's not a good sign. See my other comment on the physical reasons why I'm skeptical that Helion will get fusion anytime soon.

In the last 50 years, the mainstream fusion community has improved fusion performance (as measured by the triple product) by five orders of magnitude. We have less than an order of magnitude to go. The start-ups that use standard techniques only have to close this final gap, instead of achieving all of this progress on their own.

Expand full comment

Helion has several important problems that make their success implausible.

They are using D-He3 fusion. This requires 5 times hotter temperatures as the usual goal of D-T fusion. The reason they want to use the D-He3 reaction is because it doesn't produce neutrons. But their plasma will also have D-D fusion reactions, which either produces tritium and a proton or He3 and a neutron, with equal probability. Their plasma will be producing neutrons, just not as many as a D-T plasma. He3 does not exist in significant quantities on the surface of the earth. Other fusioneers who want to use He3 suggest that the best way to get it is to mine the surface of the moon. Helion claims that they're going to produce the He3 themselves using D-D reactions. So they will need at least twice as many D-D fusion reactions as D-He3 reactions.

Helion is planning on compressing their fuel. This sounds nice because it means that you can get higher density (causing more collisions) without having to maintain that high density in steady state. The problem is that compressing fuel tends to cause instabilities. When you try to squeeze something, it tends to squish out between your fingers instead of being compressed. This has been the main challenge for the inertial confinement fusion program at NIF - and it has already cost them more than 3 years. They have made progress recently by manufacturing their capsule with extremely highly precision. Helion is starting with something fluid, so they can't use NIF's strategy.

Helion's plan is for a pulsed power source. You create two plasmas, compress them and collide them, and then extract energy. Each time you do this is called a "shot". Current experiments like NIF and Z-pinch can do about 1 shot per day. To make a feasible power plant, you would need at least 1 shot per second. I don't know how they're going to speed things up by a factor of 100,000 compared to existing experiments.

While I don't think that their approach is particularly plausible, I applaud them for their efforts and for their ability to get people excited about fusion. $500,000,000 is a lot of money - possibly more than what Commonwealth has raised.

My predictions: They will get an experiment built within 2-3 years. The experiment is small enough that they can just buy the He3. They will have problems with compressing the plasma, so their Q will be about an order of magnitude lower than expected. Maybe after another 2-5 years of experiments, they'll have figured out how to make the compression work, so they can announce that they GOT FUSION. The shot rate will still be only a few times per day, but they can use the hype to get more money. I would guess that increasing the shot rate to once per second is at least as hard as dealing with the compression, so they will be not be selling electricity to the grid anytime soon.

Expand full comment

+5 Informative

Expand full comment

I agree D-He3 is perhaps pie in the sky. However I think it's less of a moonshot than it appears for a few reasons. The neutron motivation is solid - D-T neutrons are 14 MeV vs D-D neutrons at 2.5 MeV and less of them. This creates two advantages in that the reactor can be far smaller without the need for serious shielding and at the same time may not have its lifetime limited by the neutron flux

Helion proposes that their full scale reactor could be 50MW vs a CFS style D-T tokamak needing to be 200MW at the smallest size to contain enough shielding. This would allow Helion to serve a larger market potentially including powering container ships, and their larger plants would benefit from modularity where they should not only need less downtime but would have a much easier time with one or a few modules out of service at a time. Combining this with much less worry about capital intensive plants having their lifetimes cut short by D-T neutron flux makes up some substantial part of the ground lost in having a much more difficult reaction to create

If Helion's energy capture system really works there is another large gain. I may be incorrect here but it appears in theory Helion would need a power gain + initial power of 1.18 to reach breakeven. They only need to gain 20% energy for net breakeven after the input energy plus gained energy are recovered at 85% efficiency. By the same numbers a thermal plant would need a gain + input power of 2.5 for breakeven electricity conversion - a gain of 150%

So Helion's basic reaction is less likely but if they make it to the starting line the finish line is relatively close. Their sourcing of He3 seems like a pretty notable question mark but the pulse rate is actually not as big of a concern. The current Trenta machine can pulse every 10 mins while the 2024 Polaris is proposed to be able to pulse once per second temporarily. The ten minutes is nice and it will certainly be interesting to see if they can close in on second pulses. Supposedly the limiting factor is removing the fusion product 'exhaust' between pulses and they have done it fast enough on some small scale and are hoping they can scale that up

If all else fails I don't know how their concept compares as a D-T reactor but they could try it as I believe TAE have indicated they would license their design for D-T use while they continued work on trying to get it to aneutronic levels

Expand full comment

I did not know that Trenta can do a shot every 10 minutes. They only need to improve this by 2.5 orders of magnitude instead of 5 orders of magnitude.

The best way to measure how close you are to fusion is with the triple product: temperature * density * confinement time. Helion has not published triple products for Trenta, so I don't know how close they are. Their last experiment, Venti, had a triple product 300 times too small for fusion. (Source: random comments.) Trenta has five times higher temperature than Venti, but we don't know its density or confinement time. <Do you think that this means it's better than expected?> I expect that they're already having trouble with compression. It's common in fusion to try to scale your experiment up by an order of magnitude or two, but then run into new instabilities or turbulence that take years to work around.

Helion's energy capture system could also be a significant improvement, although I don't know what the best efficiency they have achieved so far is. They also don't have the extra ~25% energy from the Li6 + n -> He4 + T reaction in the breeding blanket for tokamaks.

The size of fusion power plants is mostly determined by the energy confinement time. The farther a particle has to go (across magnetic field lines) to get to the wall, the longer it stays in the plasma. When you make a fusion device this big, there ends up being enough room for shielding - except for the thin central column of spherical tokamaks.

D-He3 seems like a great reaction for second generation fusion devices. Once we get good at confining plasmas, then we can push to higher temperatures and smaller sizes with He3. For a first generation device, you're just making your life more difficult.

I'm reminded of Type One Energy, which says that they hope to use a D-D reaction, but will check whether D-D or D-T is more feasible in their experiment. I am confident that they will "find out" that D-T is easier. Helion might "end up" using D-T for their first generation reactors as well.

Expand full comment

Helion's claiming they'll be able to run the same or similar equipment with D-D fuel to generate the He3. I'm skeptical of this on a couple fronts:

D-D fusion is even harder than D-He3 and releases considerably less energy, so your fuel generation cycle is likely to be significantly energy negative. Not necessarily a deal-breaker, but it does need to be factored into the calculation. It also multiplies your capital costs, since every pulse you run with D-D fuel to generate He3 consumes equipment time that is now no longer available for your power generation cycle.

The other problem is that D-D fusion isn't aneutronic: it yields 50% He3+n and 50% T+p. So in addition to the tritium (which isn't really a waste product, since you can sell it), you've got a high-energy neutron to deal with for every He3, and the whole point of using D+He3 in the first place was so you'd have a nice, well-behaved high-energy proton to handle instead of a neutron. So you'll still have the same number of neutrons to deal with (or somewhat more, since you're likely to lose some He3 in the process somewhere), and you're still going to need an answer to the same engineering challenges of neutron shielding that the D-T fueled reactors have.

There are a couple residual benefits, thought, if the can get it all working. Neutron shielding and power generation are now separable, simplifying the engineering challenges somewhat. Using the neutrons to generate tritium is demoted from a critical requirement to a stretch goal for the neutron shielding system. The neutrons from D-D reaction are considerably less energetic, so there's correspondingly less waste heat to vent. You can potentially capture useful energy from the protons a lot more efficiently than from neutrons. And you can use different facilities for you D-D reactions and your D-He3 reactions, so you could have a big centralized facility for He3 generation with all the neutron shielding, leaving your power generation reactors small and relatively free of the burden of neutron shielding.

The D-D reactions also offer a potential offramp short of successful power generation. If Helion gets the fuel generation cycle working but not the power generation, and CFS or one of the other D-T fueled projects does succeed, then there's going to be a huge market for Tritium that Helion could fill. A D-T reactor should be able to make a lot of its own tritium (the neutron is captured by a lithium in the shielding system, which them splits into a tritium and a helium 4), but would need to be 100% efficient at this to be fully self sufficient: there are proposed solutions to this involving neutron multipliers (materials like beryllium or lead that can emit multiple neutrons after absorbing one), but it'd be cleaner and simpler to just buy it if it's available in the required quantities at a vaguely reasonable price.

Expand full comment

> D-D fusion is even harder than D-He3

I thought so, too. It makes concerns about stray D-D reactions in the D-He3 mix seem overblown.

Expand full comment
founding

D-D fusion is not so much harder than D-He3 that you can avoid having an unpleasant amount of D-D fusion going on anyplace you have enough D to make D-He3 happen.

Well, maybe in something like a two-beam system where the D is kept "cold" and fuses with He3 only by directed kinetic energy, but I don't think any of the current contenders are proposing that (and it probably wouldn't work if they did).

Expand full comment

"Unpleasant" is relative. D-T fusion is already producing 2 or 3 orders less radioactive waste than fission. Depending on the conditions in the D-He3 reactor, we'd drop another few orders of magnitude.

I'm still skeptical we'll somehow leapfrog to success with D-He3 before D-T.

Expand full comment

Well described! re: How much harder is D+D or D+3He fusion than D+T, here's the classic "triple product" plot from https://en.wikipedia.org/wiki/Lawson_criterion: https://upload.wikimedia.org/wikipedia/commons/thumb/a/ad/Fusion_tripleprod.svg/1920px-Fusion_tripleprod.svg.png. You can see that D3He fusion requires about an order of magnitude higher temperature at the same density and confinement time -- that is, an order of magnitude higher pressure (P = n * T). DD fusion requires two orders of magnitude higher density or confinement at the same temperature, so still an order of magnitude higher pressure. (Note this is a pretty rough mathematical guideline: given all the challenges of getting plasmas to do what you want, e.g. instabilities, this is a lower-limit than a threshold for fusion ignition.) So a reactor that can just not reach the pressure conditions necessary for either of those will probably ignite D-T almost trivially. This is why if they're not talking about at least demonstrating first burn with DT, it's hard to take them seriously.

re: How much better are the CFS magnets? It's been a long time since I studied tokamaks, but if I recall correctly, magnetic confinement fusion power is expected to scale with peak magnetic field to the fourth power (B^4). This means that by doubling the peak field strength, CFS can build an igniting tokamak (in theory) that is 1/16 the volume of ITER. This takes it from the international megaproject scale (>$10bn, 40 yrs) to something achieved more locally in space and time (< $1bn, "5 years"). The reason to be bullish on CFS is that tokamak physics is about as well understood as anything in fusionland, and their breakthrough directly pushes a highly critical factor.

I think it's also generally underappreciated that igniting a plasma will result in novel conditions that, if history is any indicator, will not behave quite how we expect. The recent (near) ignition event on the National Ignition Facility is exciting primarily because it will let us test ignited plasma behavior, assuming we can repeat it. CFS's SPARC will also be a research machine to understand how to control and engineer magnetically-confined burning plasmas.

Expand full comment

Hot off the presses, CFS has raised 2 billion to build SPARC and begin work on ARC. In a week private fusion funding has more than doubled

I was reading on the B^4th power relationship recently and it doesn't result in 16x more density due to one of the other factors being radius squared - so I think you would have 16x more density at the same size as ITER but the ARC wiki page describes it as allowing the same power as ITER at a quarter of the size

Expand full comment
founding

It is one of those great cosmic coincidences that America has produced the two greatest chess players of all time (as defined by dominance over their contemporaries), both of whom quit at the height of their dominance to never play chess again.

The best way of describing Bobby Fischer to a non-chess player is to imagine an Eskimo tennis player who cleared away the snow to create his own tennis court, trained exclusively by himself, then showed up at Wimbledon, where every single player colluded and conspired against him (sharing training techniques, intentionally throwing matches to ensure better rest for his opponents), but nevertheless won every match without losing a set or a game, and then immediately disappeared, only to show up decades later to blame the Eskimos for 9/11.

Expand full comment

Fischer did have to famous successful parents, so really not Eskimo like.

In fact someone whose mother defected to Stalin's Soviet Union to study medicine is exactly the kind of person you expect to be a 9/11 truther and a chess prodigy.

Expand full comment
Comment deleted
Expand full comment

What sort of guns you like to shoot in the nude, Paula?

Expand full comment
Comment deleted
Expand full comment

Paula, you have more pressing problems.

Your Mccaffrey antivirus subscription has lapsed!

Expand full comment

Or she likes to kill or maim people who aren't wearing clothes? That's just awful!

Expand full comment

I feel like Substack could use a report button... and maybe Akismet or something

Expand full comment

Yeah, it's not like his family history predicted stability for him.

Expand full comment

"It is one of those great cosmic coincidences that America has produced the two greatest chess players of all time (as defined by dominance over their contemporaries)..."

Both Morphy and Fischer had amazing peaks, but also *short* peaks.

Capablanca went undefeated for eight years, which included games in tournaments and defeating Lasker to become world champion...

Expand full comment

Capablanca is not really considered a serious contender for the greatest player of all time, while Morphy and (especially) Fischer are.

Expand full comment

I have never seen a realistic claim that Morphy was the "greatest of all time." Greatest at the time he was playing, sure.

But if we define "greatest of all time" to be "dominance over their contemporaries" as Jackie'sDad did, then how does going eight years without losing (while still playing in tournaments and a world chess championship match) not even get into the conversation?

Expand full comment

The problem is that, when you really consider the situation involved, the "8 years without losing" streak becomes significantly less impressive than it may appear at first because so he played so few games. It's not his fault, of course, but limited capability for international travel + fewer top players + political considerations etc meant that there were far fewer chess tournaments in which he could face other top players. More specifically, he played 63 consecutive games without losing, while a super grandmaster like Ding Liren, for instance, went 100 straight with no losses 3 years ago (mostly against 2700+ players).

This isn't meant to disqualify Capablanca as one of the greatest players ever, mind you; he was a world champion, and he was certainly dominant in his time. But he's not in contention for greatest player ever, since, in the eyes of most observers, he can't really hold a candle to Kasparov, or Fischer (or even Carlsen, honestly).

With regards to Morphy's argument for being the greatest ever, I would recommend you watch some of the videos in which Ben Finegold explains why that's a reasonable position to hold.

Expand full comment

And here's a pop song that references Fischer; Prefab Sprout, 1984, "Cue Fanfare":

https://www.youtube.com/watch?v=6DBE3H06uUg

Expand full comment

Small typo: “ets qualified" should be "Lets qualified"

Expand full comment

1. In addition to Bezos, Steve Jobs and Larry Ellison were also abandoned by their biological parents. At least some of their motivation must have been to make their biological parents deeply regret having ever abandoned them.

Expand full comment

Perhaps success in big tech is enabled by having the kind of genes that would allow you to abandon a child, plus being raised in a environment by the kind of parents who would adopt an abandoned child.

Expand full comment

I hadn’t thought about the callousness angle…but it does makes sense.

Expand full comment

I don't know about Bezos or Ellison, but what I know of Jobs suggest he was a pretty cold father.

Expand full comment

Cold to all his daughters but not to his son, Reed.

Expand full comment

This just blew my mind.

Expand full comment

I don't know why their parents abandoned them. Poverty/desperation strikes me as less callous than just not wanting to raise one's child.

Expand full comment

Steve Jobs was not an uncritical person, but he had nothing but appreciation for how well his adoptive parents raised him. For example, when Steve was getting picked on in school, he told his parents he had to go to a better school so they sold their house and moved to a better school district.

The person I felt sorry for was Steve's adoptive little sister. Did anybody ask her if she wanted to move and leave her school friends behind?

In general, can you imagine having for an older brother the World's Greatest Salesman? Especially if you don't share any of his Reality Distortion Field genes?

Expand full comment
author

So you're saying if I want my kids to be rich and successful, I need to abandon them at birth? Worked for Moses, Romulus, and King Arthur...

Expand full comment

Perhaps a surprising number of presidents too--Ford and Clinton.

Expand full comment

And Obama.

Expand full comment

Three of the last ten Presidents have had a different surname at times in their lives, including Gerald Ford who was in general extremely non-exotic. The lesson I take from that is that lots of people have had family drama.

Expand full comment

"So you're saying if I want my kids to be rich and successful, I need to abandon them at birth? Worked for Moses, Romulus, and King Arthur..."

Yes! Also, if you want to become rich you should bet everything on 17 five times running...

Expand full comment

Seems something of a high-rise strategy. After all, Arthur and Romulus's siblings didn't turn out quite so well...

I'd suggest 20% chance of criminal career, 79% chance of fairly normal life and 1% chance of being a legendary archetype for your people. Odds of the last may lower if said archetype already exists...

Expand full comment

Arthur's half-sisters and adopted brother were all raised by their birth parents. He had no other siblings to the best of my knowledge.

Expand full comment

I'm not sure there's a canonical childhood history for Morgana (because there's no real canonical Arthurian tradition), but I'd missed that possibility... As Moses had at least one brother (Aaron) raised by his parents maybe the abandonment strategy works best if you just abandon one child: it probably creates quite a driven (if vengeful) personality, and avoids the risks caused by association with siblings seen in the case of Remus.

We may need some more data points to be sure though.

Expand full comment

I say Geoffrey of Monmouth is the canonical version, since it actually got accepted as actual history before people concluded he had an "inordinate love of lying".

Expand full comment

But he doesn't have Lancelot or Guenivere! He certainly doesn't underlie the French romance tradition, which has its own core stories.

Expand full comment

I say Malory is the canonical version. He's got all of the major characters, and to the best of my knowledge everything in English written after Malory references Malory, directly or indirectly.

Expand full comment

Probably improves their odds. On the other hand, if you want them to be happy, it's a different calculation entirely.

Expand full comment

Moses wasn't really abandoned. Mom was running a number to save him from the Gestapo. She left Miriam to watch him and Miriam talked the Princess into letting Mom be his wet nurse.

Expand full comment

You need to (have the genes to) be the kind of person who would abandon them, and yet raise them in a nurturing environment. Kavka's toxin IRL.

Expand full comment

It wouldn't help, since you don't have the genes for the personality which *spontaneously* abandons them, which is what gives the child the necessary characteristic for success. If you abandon your children because you want to improve their life outcomes, that just proves you have unusually strong conscientiousness and willingness to sacrifice for the future well-being of your children, that is, the act selects for exactly the opposite of what abandonment selects for in nature.

This is actually a parable about all attempts at intentional design of society.

Expand full comment

And if you want a boy to be able to look out for himself after you've abandoned him, name him Sue.

Expand full comment

The Orphan of Destiny ...

Expand full comment

This raises an interesting question: Why is abandonment as children so popular a precursor in our hero stories and myths? (They often end up with terrible step-mothers, which just makes things that much worse.) Is it that we love the intrigue of the come-from-way-behind mythos to conquer all? Or is there something about that kind of adversity that really does foster psychological hero potential in people and our myths have captured it?

Expand full comment

My son calls it the Orphan of Destiny motif seen in Moses, Oedipus, Romulus and

Expand full comment

And? And what? The tension of this orphaned comment!

Expand full comment

We need to look at base rates.

What % of people are abandoned by their fathers and what % of billionaires are.

This Daily Mail article is the best I can find, it says one in five with second families don't see the kids from their first family, not the relevant statistic but doesn't indicate an order of magnitude.

https://www.google.com/amp/s/www.dailymail.co.uk/news/article-2510371/amp/Britains-130-000-absent-dads-One-fathers-lose-contact-children-earlier-relationships.html

Expand full comment

The generic response to this is that an unusually large % of highly successful people were orphaned, which wouldn't fit your hypothesis, or at least be a stretch.

The thing is, the same logic applies to inmates. I suspect that the burden of huge responsibility early in life makes some great and some break. I think Jonathon Haidt said something like that.

Expand full comment

What percentage, though?

2.5% of children are adopted and another 4% don't live with either parent. That would be 6.5% of children.

How many billionaires are adopted?

Another possibility, though: people who adopt children are disproportionately likely to be wealthy, so the odds of being adopted and ending up rich are probably distorted. Especially if you are adopted as a baby/small child rather than later on due to behavioral problems.

Expand full comment

Steve Jobs was the biological child of two grad students and was placed for adoption through an agency. He was placed with two working class parents of fairly similar ethnicity to his more upscale biological parents (one German-American parent in each pair, while one biological parent was Syrian, the nephew of the Syrian foreign minister IIRC, and one adoptive parent was Armenian). His biological parents had one more child together, the prominent novelist Mona Simpson. (These days, I presume, it's much harder to adopt a child from higher class biological parents than yourselves.)

Jeff Bezos's mother was from Cowboy-American gentry and her first husband was a Danish-American. The marriage broke up after a few years and his mother remarried a kindly Cuban immigrant manager at Exxon.

Larry Ellison was born during WWII when his biological father was an Italian-American pilot in the military and his mother was a Jewish-American teen. At 9 months she gave him up for adoption to her uncle and aunt and didn't see him for 48 years.

Other zillionaires: Bill Gates' father was a distinguished figure in the Puget Sound legal profession and his paternal grandfather owned a furniture store. Mark Zuckerberg is a classic Ellis Island immigrant progression across four generations of peddler to Post Office employee to very prosperous dentist to tycoon.

Sam Walton's father was something of a ne'er do well, but his uncle owned a big regional chain of stores. Michael Milken's father was a middle class accountant but his uncle was a tycoon.

Expand full comment

Not true of Bill Gates, though. And he is a much more psychologically stable individual.

Expand full comment

There's a whole post on gwern.net about how very successful people were disproportionately likely to have grown up in a bad family environment OR a perfect family environment, but no inbetween.

Expand full comment

Or, you know, that's the story that is more interesting.

Expand full comment

> I can’t read the full text,

Pubmed notes on the right hand side that it's "free", because there's an Open Access link you can read: https://onlinelibrary.wiley.com/doi/10.1111/j.1939-1676.2009.0407.x

Expand full comment

Came here to say this, you definitely can read the full text. Also to say that clinically as well as in studies like this one, we are often relying on pet owners' reports of symptoms, so you're just back in the territory of human expectations influencing perceived outcome. Seizures are less subjective for people to assess in their pets than for instance "severity of arthritis pain" but still more subjective than you might think. My typical seizure monitoring visit might consist of me asking "About how many seizures has Fluffy had in the last three months?" and if I'm really lucky, Fluffy's owner saying "I saw one that lasted about 2 minutes, and there was a puddle of urine on the kitchen floor once so I think there was another since he doesn't usually break housetraining." The pet owners who are in these studies are generally more diligent than usual and are supposed to keep diaries, but the fundamental problem remains.

Expand full comment

In the discussion the authors write that they cannot distinguish placebo effects in their study from regression to the mean:

Regression to the mean is a statistical term used to describe the fluctuations of biological variables that occur over time and take the form of a sine wave around the mean.11 Epilepsy is a waxing and waning disorder, and fluctuations in seizure frequency are common over the course of the disease. Owners are most likely to seek a change in therapy for their pet when seizures are under poor control. Over the short term, improvement in the seizure frequency is probable, regardless of the treatment administered. However, this improvement is often erroneously attributed to a recently instituted change in therapy, whereas in fact it is because of an effect of time.

Expand full comment

"A broad survey of academic philosophers found that the only group who majority one boxes on newcomb's problem are those who study ancient Greek and Roman philosophy." - Interestingly, my (weakly held) belief is that rationalists tend to be way more into classical aesthetics / classical history in general than the average nerd. I don't care much for classical stuff; also I'd two-box.

Expand full comment
author

I think this is probably just a coincidence or that classical philosophers don't study the weird math-y stuff that would make two-boxing seem justified. But if I had to come up with a cuter theory, it would be that mostly upper-class people get into classics, and upper-class people are less likely to make the choice that makes them seem like grubby money-maximizers desperate for an extra $1K.

Expand full comment

I wonder if the focus on virtue as the ground of ethics is also important.

Expand full comment

Classical literature/philosophy actually includes a problem roughly isomorphic to Newcomb's, called the "lazy argument" (ἀργὸς λόγος), that makes two-boxing seem undesirable:

You are suffering from a disease, and Fate has determined whether you will or will not recover from it. You reason that if you are fated to recover, then it is useless to see a doctor (because you will recover with or without their help) and that if you are fated *not* to recover, then it is *also* useless to see a doctor (because you will die with or without their help); you thus conclude that, live or die, you shouldn't bother with the expense of seeing the doctor.

(For Newcomb mapping, let the $1000 box correspond to what you save by not going to the doctor, and let the million-or-nothing box correspond to the value of your life; the lazy argument's conclusion is that you should two-box.)

The ancients were not happy with this—'if anything about the future is true, you may as well not bother doing anything' seemed like an unhealthy outcome—and their usual answer, attributed to Chrysippus, was that obviously Fate would have also determined whether you are going to see a doctor and that it's reasonable your predetermined prognosis might depend on that, so there could still be good in giving up the value of the doctor's fee.

Expand full comment

Academic Philosophy in the last century has relied a lot on thought experiments. As opposed to ancient philosophers who tended to make broader appeals to people's intuitions about day to day life. So would make sense classical philosophers are less inclined to seperate out their normal intuitions

Expand full comment

When I dug around in these demographics, it generally looked like the closer you are to decision theory, the more likely you were to two-box. Just as some of the questions about non-classical logic got a lot of continental philosophers, it looks like one-boxing is primarily philosophers who aren't familiar with the problem.

What I really want to know, and I don't think I can validate from their interface yet, is whether my anecdotal sense is right that decision theorists older than me are mostly two-boxers while decision theorists younger than me are often one-boxers or hold some weird mixed view like me.

Expand full comment

Your theory about a possible classical philosophy/one-box connection spurred me to think. If it isn't just a coincidence, then it might be that a character trait (disinclination to optimize) influences choice of specialization (choosing classical philosophy over other fields of philosophy).

What do I mean by 'disinclination to optimize'? Most living philosophers seem to evaluate classical philosophy as an unsatisfactory foundation for Western philosophy. (Go check the reference list for Plato and Aristotle.)

A philosopher seeking to optimize/maximize* their potential contribution to philosophy would be more likely to avoid specializations which are considered unsatisfactory. Therefore, this philosopher would be more likely to avoid classical philosophy.

By contrast, a philosopher who is disinclined to optimize lacks that reason. Thus this philosopher would be less likely to avoid classical philosophy. They would be more likely to find it good enough.

Similarly, taking one box instead of two boxes would be good enough for them.

*I leave open what optimizing/maximizing one's potential contribution to philosophy actually means for philosophers, if it means anything at all.

Expand full comment

Latin is the gateway to studying classics, and in the US lots of people study Latin in part because they are Catholic. Does being raised Catholic make you more likely to one box?

Expand full comment

I was raised Catholic and I'd definitely one-box.

I would imagine that Christians in general are more likely to one-box than atheists. Christianity is all about making deals with all-powerful beings, and if an all-powerful being tells you they're doing something nice for you then you'd better darn well believe it. Atheists hear about the existence of a supposedly all-powerful being and immediately start thinking about how they can outsmart it.

As for Catholicism versus Protestantism, I get the impression that Catholicism encourages a somewhat more philosophical approach to religion than many Protestant denominations (e.g. meditating on the mysteries of the rosary) but possibly not quite so much as Judaism.

Expand full comment

"Atheists hear about the existence of a supposedly all-powerful being and immediately start thinking about how they can outsmart it."

Weird take on atheists, who hear about the existence of a supposedly all-powerful being and simply don't believe in such a thing.

Expand full comment

I like the classical side, but find newcomb's problem to be like other similar thought experiments (trolleys, Chinese boxes) that to me are just bogus. If you assume X, where X happens to be impossible/false, everything/anything can be proved. My aspirations to become a philosopher get foiled by what I read about philosophy.

Expand full comment

There's nothing logically impossible about Newcomb's thought experiment.

Expand full comment

Are you saying that it's not impossible, or that it's some other kind of impossibility that doesn't qualify as a _logical_ impossibility?

Expand full comment

> There's nothing logically impossible about Newcomb's thought experiment.

Yes, a 100% accurate predictor of human behaviour is logically impossible. See Rice's theorem.

Even a very accurate predictor is implausible due to complexity theory.

Expand full comment

With payoffs of 1 million and 1,000, you don't even need a very accurate predictor:

Payoff from 1-boxing: 1m * P(correct)

Payoff from 2-boxing: 1m * P(incorrect) + 1000

The two payoffs are equal at p(correct) = 0.5005. Above that, 1-boxing will have a better payoff.

Do you think a person could guess someone's decision on the Newcomb problem more than 50.05% of the time?

Expand full comment

I wonder how it would change people's answers if it was phrased as being "a pretty good predictor, gets things right at least 3/4 of the time" - putting it in the territory where people picture a savvy human rather than a god.

Expand full comment

With a "pretty good predictor" I'd two box, with a perfect predictor I'd one-box.

With a mere mortal as the predictor it becomes a bit of a "he thinks I think he thinks I think" game. I'm betting I probably _could_ predict what people would do 75% of the time, because I think that 75% of the time people will two-box it when facing a mere mortal.If I ever get rich enough to be eccentric instead of just crazy, maybe I will try this.

Expand full comment

Rice's theorem has no relevance here, the predictor just needs to simulate your brain for a finite time.

Expand full comment

Sure, "just" needs to simulate. You're hiding a lot of non-trivial semantic properties behind that "just", and per Rice's theorem, all non-trivial semantic properties are undecidable.

Expand full comment

Perhaps you are being confused by the popular informal statement of Rice's theorem? The confusing thing is that "semantic property of the program" is supposed to mean a property of the partial function computed by the program. A program in the sense of the theorem (a Turing machine) can use unbounded time and memory before it outputs anything, or it can use unbounded time and memory and not output but just run indefinitely. So informally, when it runs for a long time there is no point at which we could say "ok, it definitely won't output anything" because it could always run for a little longer and output. But the predictor needs to test the behavior of a finite system (presumably also in finite time, but it doesn't matter). The computer science equivalent is running a program for n steps and looking what it did.

Expand full comment

> If you assume X, where X happens to be impossible/false, everything/anything can be proved.

Newcombs paradox generalizes pretty gracefully to versions of omega that are merely very good at predicting what you are going to do, so I don't know if this objection is as valid here as it is for eg. the Trolley Problem.

Expand full comment

If omega is merely very good, then it necessarily admits some two box solutions depending on specific circumstances, in which case it's no longer a paradox and no longer as a single solution.

Expand full comment

No longer has a single solution.

Expand full comment

I think it don't : either you have a non super good predictor with comparabke pay off, in which case twi boxer thinking they can beat the predictor makes sense.... Or you have super good predictor and order of magnitude appart pay off. In which sense, it still makes sense to be a two boxer, like it does to play at the lottery : even if you loose on average, there is on non zero chance of life changing event while the average gain (or loss) does not change anything in your life. This is a existing thing, and we can observe a huge amount of people bet in this case, even if not playing have a better average payoff...

Expand full comment

A lot of philosophers are bad at writing thought experiments, and try to stipulate things that our imagination resists. Newcomb is an example like that.

The standard move is to re-describe the Newcomb problem in a medical way. Suppose we discover that masks do nothing to block viruses, but that people who regularly wear masks have lower covid infection rates (presumably because of some common cause). In that case, do you wear a mask? (Wearing a mask is like one boxing and not wearing one is like two boxing.)

I believe the standard internet-rationalist one-boxers claim that this version is importantly different from the standard presentation, and go for the two-boxing equivalent here.

Expand full comment

Thanx for this. When I think about it this way, my prior is to wear a mask. So I guess that makes me a one-boxer and confirms the classics stereotype.

Expand full comment

I don't think they're the same at all, in the above situation I wouldn't wear a mask, but I'm definitely a one-boxer.

Expand full comment

(And I expect that I'll end up with a million dollars and no covid. One should try to win.)

Expand full comment

[Maybe add EY's talk with Paul Christiano on takeoff speeds](https://www.lesswrong.com/posts/vwLxd6hhFvPbvKmBH/yudkowsky-and-christiano-discuss-takeoff-speeds)

I think this was, in some sense, the most awaited one of the discussions.

Expand full comment

Re nature vs. nurture for criminality, I'm reminded of this study from Ohio.

https://www.gwern.net/docs/sociology/2021-norris.pdf

Key points:

- You have a parent who committed a crime in Ohio

- Your parent is randomly assigned a judge. Some judges are much harsher than others

- If your parent got a harsher judge, you are slightly less likely to be a criminal yourself.

The study generated interest because it came to the unpleasant conclusion that having your criminous parent thrown in prison (this being more likely under harsh judge) was better than having him around in your life. Or maybe it scares you into following the law. But any way you cut it, this is a nurture effect.

Outside my field, but assuming it was conducted competently it has beautiful randomization and relevance to the question.

Expand full comment

Isn't there an obvious third option there involving revoking custory without a long prison sentence?

Expand full comment

Locking people up has a lot of beneficial effects. Turns out most criminals won't change, so locking them up for a long time keeps them from committing more crimes and pulling more people into crime.

Expand full comment

What's the benefit of eventually letting them out then?

Expand full comment

There's a reason why we used to hang so many criminals...

Some people do "age out" of criminal behavior. Others don't, and need to be incarcerated indefinitely.

Expand full comment

Sure, but there's also a pretty common type of a career criminal where nobody really expects them to become straight again, and yet this alone isn't enough for a life sentence, if they're caught over a trifle off they go after a short sentence.

It seems like justice system wasn't really designed with these considerations in mind, judging the "character" comes distant second to the severity of the crime at hand. Maybe at some point this led to a beneficial balance, but by now everything is so far gone that maybe it's time to redesign the system from the ground up. Sadly society doesn't appear to be up to the task...

Expand full comment

This is what "three strikes" laws and similar things are designed to do - someone keeps ending up in court time and again for committing various crimes, they clearly aren't going to change their ways.

Expand full comment

The fraction that change does also depend on the nature of the carceral system - the American one seems to actively make people *more* criminal, whereas other countries do successfully reform at least some of the criminals incarcerated.

Expand full comment

False, actually. Criminals have very high recidivism rates globally. Countries which solve a lower percentage of crimes have "lower" recidivism rates because they simply fail to arrest as high of a percentage of criminals.

This is similar to thinking that you are lowering COVID rates by not testing it.

People just lie about it for propaganda purposes.

Expand full comment

Cutting off the right hand off thieves helps them to reform not to steal so much, and try out a career in begging instead.

Expand full comment

There's no good evidence for this.

People always point to Scandinavian countries as evidence but can never separate out the confounders. There are more Americans per capita who commit crimes for the first time compared with these countries, which obviously cannot be a direct result of recidivism ​(maybe you could argue its indirect because their criminal parents are reformed and therefore better parents than American criminals, but this ought to be observable simply through a higher heritability of criminality in the US).

You also have to assume that none of these differences between the US and Scandinavia are demographic based. Certain ethnic groups in the US commit the vast majority of crime per capita, and these ethnic groups are a much smaller part of the population of scandanavian countries. Maybe, you know, people of Northern European ancestry are more easily "rehabilitated" than certain other demographic? Scoff at that all you want, but if your analysis does not make a bona fide consideration of this possibility, then your analysis is junk.

Expand full comment

The state, for very good reasons, wants to keep kids and parents together, even if the parents are not stellar or are even actively bad. I hate the idea of kids staying with their terrible parents, but hate the idea of the state routinely deciding who is a bad parent even more.

When parents are in jail, there are fewer, possibly no, opportunities for the parents to be around the kids, so the state has to go with other options. That also has the benefit of a fair (hopefully) judiciary to review a case and determine if the parent really should be punished and put in jail. There are significant checks on the system, and the state has to make a compelling case. It's not perfect, of course, but I have a lot more confidence in that system for taking kids away than a bureaucratic option.

Expand full comment

This isn't terribly surprising and is almost certainly one of the reasons why mass incarceration lowers crime - throw criminals in jail for a long time, they aren't around to mess up their kids/suck them into their own criminal stuff. Likewise, the kid isn't exposed to that stuff and doesn't see it as normalized, and also looks at it as bad - they got locked up forever, they should avoid such things to have the same effect to them.

Expand full comment

this requires defining aggression against other prisoners as not-crime, I think.

Expand full comment

They commit fewer crimes in prison than they do outside of it due to the higher degree of supervision.

Also, to some extent, we care less about the well-being of prisoners than the general public.

Expand full comment

As a veterinarian, I frequently think about placebo effects, but for humans, not for animals. I rarely get to observe the seizures or other clinical signs ("symptoms" are for humans) myself. Instead, I have to rely on the owner/guardian's assessment and report which can be highly biased. The placebo effect in question should be for the humans reporting the number of seizures their dogs are experiencing after giving them a new medication.

Expand full comment

I think the dog placebo study linked in #5 is also terrible because it's comparing placebo to baseline instead of comparing placebo to no placebo. The effect could be all regression to the mean. If they want to run this study properly, they need to randomize dogs into placebo vs no placebo and blind the owners about which group their dog is in.

Expand full comment

You could carefully construct different placebos to test against.

Perhaps you can make a placebo that only the humans notice (perhaps praying for the dog?), vs placebos that the pets notice more. Eg neutral tasting Vs bitter pills. Presumably, the owners wouldn't sample the pills for taste.

Expand full comment

On 10 (painted classical sculptures), a friend of mine said something similar: “I don't buy it. Why would they spend weeks or months carefully carving the marble and then spend an afternoon slapping some flashy colors on it? … I think archaeologists are only finding scant molecular evidence of the *base coat* on the statue. I bet the statues were also covered with highlights, layers, shades, etc.”

He suggests that the statues might have looked like life-size versions of today's painted miniatures, which is intriguing: https://www.facebook.com/steeleky/posts/10120183205650904

Expand full comment

My favorite theory is that the paint we have found is (in at least some cases) an amateurish repaint done some years after the work was painted (presumably much better) the first time. No idea how viable the theory is, but it's amusing to think about.

Expand full comment

Given how often this seems to have happened to great paintings, I wouldn't at all be surprised. E.g., the history of attempts to restore da Vinci's Last Supper: https://en.wikipedia.org/wiki/Conservation-restoration_of_Leonardo_da_Vinci%27s_The_Last_Supper

Expand full comment

Or maybe an amateurish first paint! Perhaps the original sculptors preferred the bare marble aesthetic, which then later became unpopular.

Expand full comment

We need to let a warhammer nerd loose on the Elgin Marbles. Or at the very least, SLA print them a 1:72 reproduction of that Augustus statue to paint.

Expand full comment

Service Level Agreement?

Symbionese Liberation Army?

Student Leadership Academy?

Expand full comment

Probably StereoLithographic Apparatus (for 3d printing: https://en.wikipedia.org/wiki/Stereolithography )

Expand full comment

I think that might be the British Museum's plan B just in case anyone forces them to give them back.

Expand full comment

Why bother with 3D printing when we could just have a robot carve a replica out of marble? https://www.nytimes.com/2021/07/11/world/europe/carrara-italy-robot-sculptures.html

Expand full comment

I think there's a lot to be said for using practical domain knowledge to inform history. This reminds me of Janet Stephens, the real-life hairdresser whose research into Roman hairstyles has revealed just how much those hairstyles were sewn; or the historical re-enacters who put their clay pipes down by the fire to dry overnight, and when the next morning they threw a new log on the fire, realised why archaeology finds so many broken pipes.

Expand full comment

If a culture made detailed naturalistic sculptures, then painted them in flat colors, it would be unique in the history of the world. Flat colors go with abstraction; shading goes with naturalism.

Expand full comment

The full-room light approach they tested was notably much less bright than even a normal light box would be (more light generated, but the bulbs were much less close to the eyes, so effectively less bright). See EA Forum here https://forum.effectivealtruism.org/posts/Ei2uYbn2zrzmBjEsp/sleep-effective-ways-to-improve-it?commentId=smDmE2vgCtpC4Ftnw

Still a nice idea, but arguably not really a test of the lumenator idea.

Expand full comment

So how is this different from just having some bright lightbulbs in your room?

Expand full comment
founding

That's the direction they're going for: might we get away with just some bright lightbulbs? They managed to confirm that having a bunch of them strung out in the room works.

Honestly, I doubt a single source of light on the ceiling will be enough - it's too weak if it's not right in front of your face. And if you make it bright enough, it'll probably be painful when accidentally looked at.

I'm currently experimenting with a setup of ceiling light plus around a 2m bar at the top of a wall. It's ok, but it'd probably be easier to tolerate if it were spread on several walls, or even better all over the ceiling.

Expand full comment

I recently added some pretty bright plant growth light to my living room, and I've noticed it's livened it up a lot.

Expand full comment

On the U of A topic - I *really* struggle to see what they're going to do to promote diversity of ideology that isn't either a.) just being a right-leaning culture overall, which, fine, or b.) allowing some pretty heinous stuff because they don't want to draw a line on, say, anti-LGBT speech and actions.

Their thesis - that universities are too woke to function - is not nearly as true in my experience (at a Midwestern public school, so admittedly not the MOST woke school in America) as they seem to think it is.

Expand full comment

I believe (for anecdotal reasons) that the wokeness level varies enormously from college to the next. U of A might not even turn out to the most anti-woke school out there.

Expand full comment

Given various of the founders were involved in hounding out people for not supporting the Iraq war, or making anti Israel statements, seems unfortunately likely that"free speech" will mean "my in group gets to decide what is allowed not yours".

Expand full comment
Comment deleted
Expand full comment

Unfair. There's a lot of universities that protect all speech. You don't hear much about them simply because if they have that culture then there tends not be calls for cancellations and the like, or the university shoots them down.

Expand full comment

Name three. The University of Chicago was the bastion of free speech, but searching for 'woke and University of Chicago' I get lots of articles...

And yes free speech means that Nazis and racists and the KKK and... get to voice their opinions. (No one has to listen...)

Expand full comment

Unfortunately I *worked* at their number 2 college and it was terrible.

For example, I was part of their cancer moonshot deal while Trump was running the first time, and one of the leads would spam hipchat shitting on Trump supporters (Not Trump; sometimes, yes, but usually specifically the people who like him). I asked him to stop because my family is full of them and I HAVE to work and would prefer not to hear my boss's boss shitting on my family all day while I do it. He refused.

I tried to report it up the ladder and the U was awfully unresponsive. In fact they used their commitment to free speech *to defend someone with institutional power pushing his views on his subordinates all day*.

Okay, then. If *that's* what they mean by a commitment to free speech, so be it. So I started refuting what he said in slack every. single. time. he said it. I tried to do it as calmly and rationally - but persistently and consistently as possible.

LO AND BEHOLD by contract wasn't renewed despite good performance. And This isn't the only thing I saw while I was there by far.

In summary, UofC's speech code seems nice on paper but *in practice* seems to be deployed to protect a progressive majority's right to push their view on their subordinates (including students) and to use their power to lock people they disagree with out of the conversation (or bully them into silence). It simply uses the type of language FIRE (god bless 'em unironically) likes to hear so they get the good ranking.

All that being said, given my personal experience, as far as I'm concerned you have to eliminate everything below UofC. So you've shown one University.

Expand full comment

> anti-LGBT speech and actions

Why should that speech be banned? And not pro-LGBT speech instead?

I think we are all in opposition to assault and battery. What actions did you have in mind?

Expand full comment

Openly discriminating against LGBT people by for example refusing to enroll them in courses, allow them to join clubs, that sort of thing. Definitely free association! But...

Expand full comment

...that's not speech.

If you followed the people that want to do this I doubt you'd have any concern about this happening.

Expand full comment

The impression I get is that wokeness flows from the top down for colleges with a bias to the liberal arts, racial minorities, and left politics in the area broadly. Small Elite liberals arts schools > Ivies > elite public universities > mid-tier colleges > trade schools

Expand full comment

> 20: Variations on the fable of The Frog And The Scorpion.

THANK YOU. I saw this once back when it was first published, forgot the reference, and had been looking for it since!

Expand full comment

My dad used to recite this fable as a cautionary tale (which was its original usefulness.) But all these variations on a theme illustrate its power as a larger commentary on trust, and how much trust comes into play in situations of love and war...

Expand full comment

If you watch a lot of commentary by GM Ben Finegold, you come to love Paul Morphy. Apparently there's a lot of disagreement about Morphy's true strength; they'll say "Everybody he played against sucked. How can you say he was objectively strong?" But Ben argues that that's exactly the point. How exactly do you become so dominant over everybody in the world without there being anybody to teach you?

Expand full comment

Came here to note exactly this. It's like if Steph Curry showed up 5 years after James Naismith invented basketball; what the heck is this guy doing here absent the evolutionary process that should have produced him?

And, as Finegold points out, although he obviously made his name flaying dudes alive in knife-edge tactical positions, on the rare occasions that he got into a slow positional game he played it in a way that, again, shows a deep understanding of those types of positions that's just totally out of place in his context.

Expand full comment

In his match against Andersen (the best player besides him at the time), he crushed Andersen quickly in a couple of game. Then Andersen won a 70 move slug fest and said "Morphy needed only 20 moves to beat me, while I needed 70".

Expand full comment

Ben Finegold is hilarious! You can see from Morphy's aggression and sacrifices how much he was just toying with contemporaries the same way you can disrespect people way below your skill level in other domains by trying things that shouldn't work.

Expand full comment

I think your summary of Caplan is fair. Granted, it's admirable that he's open about it, and he was clear he'd say these things upfront, and he adhered to the terms of the bet. But even so it seems a little weak.

Then again, in the same vein -- his half-hearted rebuttal to me serves as a ringing endorsement of the book and I will probably read it now!

Expand full comment

Heart-eyes! So gratifying to read this. Thank you!

Expand full comment

The left wing bias was only one point in the review.

The bigger points were eg not taking nuclear or geoengineering serious.

Expand full comment

No no no, those points were strictly in service to his "look how leftist they are" point!

With geoengineering, Bryan failed to engage at all with the climatological arguments in the book against geoengineering. I think he basically bulverized the authors, saying that they only dislike geoengineering because it would allow a billionaire to unilaterally solve warming without government involvement, which Bryan Caplan wrongly assumed the authors inherently would hate, based on his sterotypes of leftists. So circular!

With nuclear, Bryan has a fair point: the authors ought to have gone out of their way to praise it, instead of failing to mention it at all. That would've given the book more credibility, particularly in Bryan's eyes. My defense of the authors is to presume they just weren't looking for ways to signal their lack of left-wing bias, or just thought the topic riles people up and would distract from their core points. Or maybe they just think that nuclear power is good exactly to the extent that the market says it's good, assuming externalities are priced correctly (that being primary message of the book). I.e., no need to praise it just for the sake of signaling.

Expand full comment
founding

It's not what I got from his rebuttal. It's not a situation where the book makes a good point, but Caplan dismisses it due to an unreliable source. It's that the book is basically asking the reader to take a lot on faith, and Caplan just doesn't have the necessary level of faith in the author. That's a fair conclusion, and the reason why he didn't expect to be convinced in the first place.

If I'm considering a complicated topic and I'm about to read an untrusty source on that topic, I can tell in advance that I'll only be convinced if the author manages to give me a gears-level understanding of his arguments. If I can't "grok" his points, I am not going to take them on faith.

And since climate change is a rather complicated topic, Caplan could have been pretty sure in advance that it's unlikely he'll change his mind. The chance of it happening is proportional to the chance the authors can clearly explain their reasoning on that complicated topic. It's not proportional to how convinced you are, but with how complicated the topic is.

Expand full comment

Wait, did you read the book? I thought it left as little to faith as was possible in a single volume book! But that could be my own bias, since I already thought climate change was a big problem before reading the book.

I'd love to find someone who started skeptical and whose mind was changed by _Climate Shock_.

Expand full comment
founding

I'm just explaining Caplan's likely thought process, since I think Scott's and others' in this thread are using a rather uncharitable interpretation.

Personally, I'm very much not interested in climate science :) It's the perfect mix of being an imprecise and complex science, culture war topic and in any case, there's nothing I can do about it so it can only, at most, increase my anxiety level. In this respect I could make the same bet with you and be reasonably sure to win, because my conviction is not based on the content of the book.

And even on a purely intellectual level it's unlikely to change much for me. I already think humans are changing the climate - that much is pretty obvious. I'm not concerned with small to moderate change, because I think we'll grow economically and technologically much faster than the climate. I'm somewhat afraid of a Venus-like scenario, but the consensus seems to be it is unlikely. I'm already pro-technology. I'm rabidly anti-pollution, and this overlaps a lot with climate. So see, yet again a case of not considering the book relevant, even before reading it. Having read it and saying "yeah, it's interesting, but I'm still 90% what I said above" is by far the most likely outcome.

Expand full comment

> I’d previously cited a claim in Joseph Henrich’s Secret Of Our Success that people liked spicy foods because they were antibacterial, but an article in Nature says there is “little evidence” to support that claim.

Does this generalize to all spices, or just capsaicin and similar 'hot' things? Spices being antimicrobial is still the best explanation I've heard for why we've evolved to like their taste, in small amounts.

Expand full comment

I always thought hot spice was to hide the taste of food that is spoiling. It’s not that we evolved the spicy taste, it’s that we can tolerate it better than the evolutionary learning that spoiled food should not be eaten.

Expand full comment

That's one of those weird ideas people made up about the medieval Europe.

In actual medieval Europe, rich people ate spice because it tasted good, and poor people didn't because they couldn't afford it. Then global trade opened up and the poor started eating spice, so the rich started eating bland food to set themselves apart. Having convinced themselves that spice doesn't actually taste good, they made up weird reasons for their ancestors to have eaten spice.

If spice actually did cover up the taste of things you shouldn't eat, wouldn't having a taste for spice be dysgenic?

Expand full comment

One possible source of the claim that medieval food was overspiced is the introduction to _Two Fifteenth Century Cookerybooks_, a 19th c. publication of the Early English Text Society. It comments on the strong stomachs of our ancestors as shown by the cinnamon soup on page X. Turning to that page one finds not a recipe but a menu. The fact that the editor thought putting cinnamon in soup was evidence of a strong stomach tells us more about 19th c. English cooking than about 15th c.

Most medieval recipes don't give quantities, so although you can tell what spices they used you can't tell how much. I've tried to estimate spice quantities from a couple of sources, one a recipe that did have quantities (for a spiced wine), one the shopping list for an enormous several day feast, which had quantities. My conclusion was that the spicing was no greater than in many modern cuisines.

Expand full comment

Could you provide citations for your contention that rich people decided spice didn’t taste good when it became more available, and that they made up the claim that spices were used to cover up spilled food?

Expand full comment

Fashions come and go in food like everything else. I don't think people did really believe spices tasted 'bad', but it was like fashions in clothing.

Purple was a very expensive dye-stuff because it was hard to produce, so it was a luxury good which became a signifier of royalty.

Then in 1856 an eighteen year old chemist accidentally synthesised an artificial purple dye, which became known as "mauve", and now everyone could afford it from royalty down to housemaids. There was a craze for it, and like all crazes, it ran its course by becoming so ubiquitous that people tired of it.

When even the ploughman or parlour maid can afford pepper on their meals, the cachet of the spice is lost. Trendsetters turn to new styles and means of preparing food and now perhaps instead of vivid spicy flavours, you go for minimalist 'freshness' or rich, creamy, buttery recipes.

Expand full comment

People in the past weren't stupid, they knew that eating spoilt food makes you sick (and quite possibly die, given that this was before modern medicine).

Expand full comment

This is my opinion - not that hot spice allows you to eat food you shouldn't but that it makes it a lot nicer to eat food that you have no better option for. Evolution isn't the only explanation for all things, I say

Expand full comment

What do you mean by food that you have no better option for?

Expand full comment

Like cheese that's gone off but it's February and food's short. Or somewhat less in terms of necessity, if 90% of what you eat is rice or potatoes and you can make it spicy that is what you will happily do

Expand full comment

I don't know about that, if you're eating 90% rice and potatoes to survive, how would you afford spice?

Expand full comment

Yeah I definitely speak from grounds of ignorance on this but I assume hot spices have traditionally been less expensive than something like being able to eat meat weekly

Expand full comment

I know spice was expensive in Europe, because it had to be imported, but I don't think it was expensive in the places it grew.

Expand full comment

Marginal utility in the other 10%. It takes very little habanero. But spicy cheesy mashed potatoes? Yum.

Expand full comment