Does anyone know why Kalamazoo and Numazu are sister cities? Did they just choose each other because of the name similarity? Orr was that pure coincidence?
I have to say, it's amusing how the guy just leaves the board open while recovering from kidney donation and it turns into a giant Israel-Palestine argument.
Thread for publicly sharing anonymized information about the Open AI board members, since I suspect many readers here to have various Open AI connections.
Been kicking around an idea/observation over the tg holiday.
Call it the “hate coefficient.” This board tends to prioritize verifiable evidence. But in this regard, internet, crowd-sourced argument presents a vulnerability. Tom the very-motivated racist, communist, anti-fascist, Palestine-hater or Israel-hater or what-have-you has a near inexhaustible capacity to dredge twitter, Wikipedia, Facebook, telegram, or whatever source is necessary to find evidence, however specious, for his preferred conclusion. In a debate, then, predisposition and motivated reasoning can transmute themselves into a endless barrage of “evidence” for how the blacks are the most hateful race, or how Russians are the real defenders in Ukraine or what have you.
One should never discount evidence, to be sure, but a disinterested third party trading study for study ends up at a surprising disadvantage- if I care a fair bit about not slandering an entire race, but Tom hates the Jews or the blacks or the French a capital-L “Lot,” for every item of evidence I’m willing to find and present, he’s more than willing and able to take as many hours of internet he needs to find and present 2.
The internet being the inexhaustible font of garbage that it is, a sufficiently motivated reasoner can easily drown a debate- not by actual preponderance of evidence, but by “preponderance of evidence I’m willing to find.” A sufficiently motivated flat-earther can just keep digging and throwing up links to the point that anyone contradicting him for the “sake of argument” becomes exhausted and calls it a day.
Which can leave the public square looking like “earth might be flat- tom’s evidence hasn’t been rebutted” even when the facts on the ground are more like “flat earth Tom threw so much garbage that no one had it in them to keep refuting it.”
At the same time, evidence matters. This phenomenon is real, but if you take it as license to ignore facts you don’t like, you’re blinding yourself. I guess you just have to take the grain of salt for very-opinionated-internet-man while also taking that same grain of salt for yourself when applying that label to him.
I don’t know. Reasoning is hard I guess. I wish I had a conclusion or a clear perspective but it seems like a prisoners’ dilemma we’re all stuck with, discounting by the hate coefficient as best we can.
It's often said that the term "Rationalist" is a poor choice because they're on the empiricist side, but I wonder how much that's actually true. The movement started in the late 2000s when New Atheism vs Creationism was the main war of the day, so there is all the obligatory lip science to The Power of Science and so on, but beyond that, Yudkowsky really seems to prefer rationalism over empiricism.
For example, in HPMOR, while Harry does do *some* experiments, they're unconnected to any of the benefits he gets. Harry's modus operandi is 1) Think about things and decide how the world *must* be based on intuition, 2) Believe *really* hard in your theory, 3) Be right because you're the author avatar and get rewarded with unique magical powers. (At least that's how he got Kill Dementor and Partial Transfiguration - the rest of his powers come from randomly getting OP magical artifacts dropped into his lap for no reason.)
Meanwhile, Yudkowsky's other classic writings seem to have a remarkable amount of contempt for actual scientists for someone ostensibly on the Pro Science side of the 2000s Religion Wars.
Meanwhile, nowdays in the Yudkowsky-derived AI Doomer movement, a common argument is that AI will be able to near-instantly take over the world because Intelligence means you can magically solve everything just by thinking really hard, no observations or legwork required. No this isn't a strawman, I've seen Doomers *explicitly* make this argument many times, as an argument about why AI takeoff shouldn't be constrained by the speed of running experiments and making observations about the world.
One more thing about the intractability of the fight between Israel and Palestine: they aren't the only players. At a minimum, there's Iran supporting Palestinian aggression (possibly also Russia) and Americans millennialists who want to start the end of the world at Megiddo.
There's an article that is making the rounds of the rat blogosphere that I think is seriously wrong. You've probably seen it quoted. It blames the ALARA (as low as reasonably achievable) radiation protection standard for all the economic problems of US nuclear power. From https://worksinprogress.co/issue/taming-the-stars/:
"ALARA is defined as: "making every reasonable effort to maintain exposures to radiation as far below the dose limits in this part as is practical consistent with the purpose for which the licensed activity is undertaken, taking into account the state of technology, the economics of improvements in relation to state of technology, the economics of improvements in relation to benefits to the public health and safety, and other societal and socioeconomic considerations, and in relation to utilization of nuclear energy and licensed materials in the public interest." [footnote citing 10 CFR 20.1003]
As currently applied to nuclear power, ALARA literally means that every expense must be spent on eliminating every possible effect of nuclear power, at least until the resulting electricity is no cheaper than what the market pays for electricity generated from non-nuclear sources. Since standards cannot ratchet downwards, only up, safety standards that are just about affordable at the top of energy price spikes get entrenched, meaning that nuclear is made unaffordable until the next price hike – which makes it even more expensive, since it prevents learning and the economies of scale that a steady pipeline of projects can allow. ALARA, as currently applied in the US and much of the rest of the developed world, means that nuclear power is never allowed to be cheaper, no matter how much safer and cleaner it is than other sources of energy. It makes affordable, safe nuclear energy impossible, and forces us to rely on much less safe energy sources instead." End quote.
The first paragraph is a literal quote from the regulations. Everything after that? where the author tells you what ALARA "literally means"? is wrong. At least, I think so. To the extent I understand the claim being made here.
Is the author saying that nuclear regulations actually change in response to energy prices? This absolutely does not happen. Is he saying that inspection standards or radiation protection procedures change with energy prices? So that some regulator or energy company employee is actually making the decision to increase radiation protection standards when they observe nuclear becoming cheaper compared to non-nuclear energy? Highly implausible. Energy prices change all the time, and regulations/inspection procedures/radiation protection procedures are only changed in a slow and cumbersome way. Also, industry would have no incentive to make itself less competitive, and it is very much NRC culture to NOT pay attention to energy prices.*
Okay, maybe the author is making a more general claim that the level of safety/security regulation increases over time, it's a one-way ratchet and regulation prevents nuclear power from being as cheap as it arguably should be compared to other energy sources. A fair but unoriginal claim. But then why the talk about ALARA?
First of all, understand that ALARA is about radiation protection. It is not the be-all and end-all of nuclear regulation. The ALARA standard adds on to other radiation dose regulations. For example, a typical nuclear power plant worker can get a max of 5 rem per year of occupational radiation exposure (10 CFR 20.1201) AND their radiation dose must be ALARA. So if a worker gets more than 5 rem, it's a violation of both regulations. If a worker gets less than 5 rem but the plant does not make reasonable effort to make the dose ALARA, it could be a violation of the ALARA standard. Conclusion...even if the ALARA standard didn't exist, nuclear plants would have to put significant effort into radiation protection, albeit not quite so much.
I'm not gonna say ALARA is unimportant. But it's only one of a whole host of regulations that apply to nuclear power plant design, construction, operation, and decommissioning. There are regulations that apply to nuclear security, reducing and mitigating the risk of nuclear accidents, emergency planning, environmental protection, and I could go on. There would be a significant regulatory burden even without ALARA.
Maybe the author is using ALARA as shorthand for the entire group of US regulations and laws relevant to nuclear? Or the entire regulatory mindset? But, if your argument is that nuclear regulation should incorporate cost considerations, why pick on one of the regulations that explicitly incorporates consideration of cost, instead of the many that don't consider cost at all?
Another quote: "[T]he components that are not safety critical are still subject to a gold plated ALARA standard. This means the same component is regulated differently depending on whether it is in a coal plant or a nuclear plant, even if it is far away from the reactor and cannot affect it."
False. The reason that a component in a nuclear plant is regulated differently from a component in a coal plant is that different laws, regulations, and administrative agencies regulate nuclear plants from those that regulate coal plants. ALARA has absolutely nothing to do with that.
I hate to be all argument from authority, but I notice the author, John Myers, seems to be a UK YIMBY activist and if he has any experience in US nuclear, I'm not aware of it. Please understand that the statement "ALARA, as currently applied in the US and much of the rest of the developed world, means that nuclear power is never allowed to be cheaper, no matter how much safer and cleaner it is than other sources of energy" is false. That is not what ALARA means. ALARA is not that powerful. Please stop quoting this guy uncritically.
*Because NRC's mission is to ensure nuclear safety and security, not to ensure that the US nuclear power industry is economically viable. If you want something to complain about, ask Congress to change that.
I have some ideas about what Israel should be doing. I'm not sure whether I'm right, nor whether this is psychologically possible even if it is right.
I think Israel should be looking to its own borders and its own safety. Wrecking Gaza will not necessarily make Israel safer, and may be putting it at more risk. It's certainly creating more hatred for Israel and I gather there are Hamas leaders in other countries-- they aren't at personal risk from the attack on Gaza..
10/7 wasn't just an atrocity, it was an embarrassment. I assume the borders are getting more attention, but are they getting more thoughtful use of tech? Bulldozer-proof barriers?
Destroying Hamas' tunnels has some practical and humanitarian issues, but additionally, the attack was by air and sea as well as by underground.
As I understand it (discussion is welcome), Hamas' intent was to provoke Israel into a drastic reaction so the world would stop supporting Israel (maybe also to make it more likely for Moslem countries to attack Israel), so that Israel could be destroyed. It's a vile approach, but it might actually make some practical sense. I doubt that Israel will be destroyed, but I still think it would be bad if it were on the receiving end of a big attack.
A part which might not be psychologically possible is to quit abusing Palestinians. Torture and a lot of imprisonment might, oh maybe just might, have something to with why it was possible to keep such such tight security on the 10/7 attack. I'm not sure how many people were involved, but I'm expecting low thousands.
Maybe they *were* warned. I get the impression the Israeli government didn't want to believe such an attack was possible.
Meanwhile, Israeli military capacity is being spent on wrecking Gaza, and perhaps the most valuable thing being wasted is attention.
Just by the way, Netanyahu is staying in power while the attack on Gaza is going on. I'm not sure when the next possibility for getting him out of office is, though I'm betting he will be out. In any case, His incentives to continue the attack are personal as well as emotional.
Sidetrack: it may not be possible to get all the hostages back. I wouldn't be surprised if some of them are dead, and I've heard a plausible claim that some of them are being held by groups other than Hamas.
Hi, I have a small whatsapp group of some 50 people, mist of whom are fellow Orthodox Rabbis.
I've been getting some negative feedback for not vocally supporting Netanyahu's actions so I made a video explaining the full context of what I would do if I were in charge.
Being members of the human species this hasn't served to have too many split off and begin to understand that my feelings of empathy for Arabs weren't dangerous eothin my chosen context but some of you might find it interesting.
As I am currently engaged in attempting to take over the world (not really, I would rather return to being an anonymous hostel traveler) in an immediate manner I'm not especially interested in intellectual masturbation online discussions with anonymii, but anyone willing to step up and choose to join me _if they come to agree with me_ is more than welcome to come on my youtube show and have a conversation to see whether or not you fo agree with me anout the general jist of what we mist do right now to make the world a place we would all enjoy living on a whole lot more than we currently do.
In short, the essentials of my approach come largely from The Torah (I am an ordained orthodox rabbi) and while the absolutely *MOST IMPORTANT* video on my channel is on a wholy different subject than the happenings in Canaan (it is called "Final Chance For A Fresh Start" and is the top video in the eponymous playlist on my channel), here's my basic 3 point plan for what I would/plan-to do regarding the situation in Israel.
Should you notice (as you _should_) that I said nearly nothing regarding how to deal with the legitimate-concern of the suffering of many Arabs under Israeli control you would be right to note its unfortunate absence (and more unfortunate glossing over the matter in a non-sympathetic manner).
The reason is simply that most of my audience comprise Israelis who, being human, are presently incapable of understanding such sentiments as anything other than a callous disregard for their own plight.
If you are willing to consider joining my war for takeover but aren't sure, please get in touch and we'll make a youtube video of it.
Just clicked through the "implicit association test" Scott referenced in his "Quests and Requests" post, and got a strong perception that I would get about the same bias given black/white colored squares instead of dark/light skinned people. I think in my mind, negative emotions are in some part defined as negations of positive emotions, and dark skin - as a negation of light skin. So it's natural that it's easier to hold positive<->positive association versus positive<->negative.
It's also a bias of sorts, but not _that_ kind of bias Scott was hinting at, it seems
What are examples of times and places when political change has been both fast and good? (Good in your opinion; fast with respect to typical political change throughout history.) Change directly related to the end of long wars, independence wars and the fall of the USSR don't count.
To be clear: it can be after a (non-independence) revolution, but not a time when things are much better simply because a time of peace has followed a time of war.
Last year I read a book with a lot of stuff in it about Churchill during the war. There were accounts of him wandering around his residence in the night wearing outrageous get-ups. I have forgotten the details -- but some were women's clothiing, like maybe a lacy negligee, and some just absurd, like maybe. a clown suit. Also accounts of his champagne dinners attended by his staff, visiting dignitaries, etc. Churchill would sometimes lead the group in skipping in circles around the table, I believe with music playing.
Others who have read about these things -- how do you think of them? I know he was an alcoholic -- I know he was not crazy. Why did he do those things? Was there more tolerance then for eccentricities of this kind? Was it a way of demonstrating his self-confidence? -- like that he was so sure that he was admired and respected that he felt able to indulge his weirdest whiims in public? Was it a way of making fools of his dinner guests?
If Supply & Demand is a thing, why does Black Friday exist?
If Black Friday is driven by a spike in demand, then I'd expect prices to grow rather than shrink. If Black Friday is driven by the supply side, wouldn't concentrating the costs of logistics/production into a single month make less money than smooth, continuous operations over the course of the year?
The common wisdom I've always received was: suppliers compete on price for business. But this just doesn't add up, to me.
Gets into detail about Joan of Arc's trial being by secular authorities and lacking many guardrails that the Catholic Church required for heresy trials.
On the one hand, the Catholic Church wouldn't have had her killed, and I'm not sure it would have put her on trial for heresy at all. On the other hand, it's the Church that made heresy trials a serious matter, so I think it deserves some of the blame, though rather indirectly.
A spectacular essay about Joan of Arc, patron saint of Catholics who don't fit well in the Catholic Church, at least on the left side. It actually gave me a feeling of what's it's like to want a patron saint.
A spectacular essay about Joan of Arc, patron saint of Catholics who don't fit well in the Catholic Church, at least on the left side. It actually gave me a feeling of what's it's like to want a patron saint.
Thinking about the fall-injury incentives thing from https://slatestarcodex.com/2016/11/10/book-review-house-of-god/ is it possible that somehow adjusting the basis on which medicaid and other government programs pay for dialysis would motivate existing medical providers to throw their weight behind reforms?
The term "oratrice mécanique d'analyse cardinale" has been trending in the meme-verse lately. It's the name of a device in the game Genshin Impact. I'm trying to figure out whether the name makes any sense in proper French.
A straightforward translation into English gets me "mechanical speaker of cardinal analysis," which doesn't make much sense, particularly that "cardinal" bit. But maybe there is more going on here than my high school French skills can manage.
Wonder how many other socialists read these comments as a kind of "hostile observer." I've been reading them for many years (obviously including Scott's previous blog) and I will continue. Scott seems humane and decent, while the commenters are often terrifyingly calculating. So why bother reading? Because rationalists are in the vanguard of technocratic polemics. They are GOOD at making effective arguments, and well, "...keep your enemies closer." I've got enormous faith in rationalism/EA as a bellwether for the future of technocracy/centrism. Give yourselves a hand, rationalists, you MATTER. Thank heavens for this comments section. Staring into the abyss isn't only educational, it's downright fun.
This week I've been playing Slay The Princess, a visual novel that has been going mildly viral and getting outstanding reviews. The premise made it seem like it would be a recreation of AI box experiment: the princess is locked up in a cabin, your job is to slay her, and she will manipulate, threaten or seduce you to stay alive.
(mild spoilers below)
Well it turns out it was less of that and more of a Stanley Parable crossed with Disco Elysium (which I should get to playing sometime soon). The game is essentially a series of vignettes, some touching and some amusing, connected by branching paths. The full playthrough basically requires you to backtrack and re-make your choices, so you can't really play a role of a prudent gatekeeper. Or, well, you can, but it leads to a joke ending and credits roll.
Those who played it, would you like to share your favourite route? (and why is it Razor)
Israeli argument that Hamas uses human shields can't be defended. Israelis are targeting those civilians. The argument that "they could kill more so they arent actually targeting them" doesn't work either. They are killing as many children as they can get away with, for the purpose of breaking Hamas's will. The goal is genocide and to solidify their theft of the West Bank.
Over the past couple of weeks I've been seeing an increasing number of articles on the subject of battery systems for renewable energy becoming price competitive with gas-fired plants (example link below). Given that intermittency of renewable energy has been THE sticking point in regards to the energy transition, that seems like pretty big news. Can anyone with more experience or knowledge in the subject offer some insight as to what degree this is hype or if we're on the cusp of a genuine shift in the economics of power generation?
I'd like to find a sperm donor whose sperm increases the chance of various desirable traits: health, IQ, talents, looks, etc. Additionally, its extremely important to me to minimize the chances of mental health issues - because the egg will likely bring some. Finally, I'd like to increase the chances for things such as values being close to mine and overall usefulness/success in life.
1) The best approach to all of the above seems to be to know someone's wide family. If there's no history of mental health issues X generations down and across many people, that sounds like reasonable probability. The same for other traits. (Yes, I'm thinking a bit in the vein of https://www.astralcodexten.com/p/secrets-of-the-great-families ).
2) Do any official institutions (spermbanks) offer anything similar? If yes, would you have tips? If not, why not? Are there regulatory issues or is there just such low demand / high stigma?
3) Can you think of a better way to find donors than just get tips on Wikipedia, on these forums and through chain emails sent to competent friends who know competent friends and then doing deep background checks on their families?
"The world is not that complex, reductionism works, intelligence is basically what matters, world optimization should be tried, all it takes is high agency people with the right values.
OR
The world is very complex, marginalism is what works, intelligence alone isn’t worth much, tacit knowledge and experience and tradition are valuable, smart people thinking they can optimize the world is hubris and inevitably leads to failure or worse."
Which do you think has more truth value? I think I'd go with 10/90 former/latter. A good response I saw says: "first one locally, second one globally".
I know that I've massively missed the bus on this, but is anyone else annoyed by the historical inaccuracy of calling the "rationalist community" as such? In the history of Western philosophy, the divide between rationalism and empiricism is one of the main splits, and the modern "rationalist" movement clearly falls on the empiricist side. Empiricism was about basing your view of the world on sense data, which is what modern "rationalism" does with its focus on Bayesian updating as the core means of knowledge acquisition. Meanwhile, actual historical rationalism held that if there was a conflict between your preconceived internal ideas about the world and your sense-based observations, instead of updating your internal ideas, you held that it was your senses that were wrong. This is how you got stuff like the Eleatics (who were essentially proto-rationalists) holding that change didn't exist despite change being observable at every moment of existence, or Leibniz holding that this was the best of all possible worlds despite all the easily observable evil in it. As you can see, this is the complete opposite of the epistemic system advocated by modern "rationalism". If I were to come up with a more accurate label for this movement, which I know it's much too late for, I'd call it Bayesian Empiricism, or maybe Neo-Empiricism. Anyways, that's my rant, I know it likely won't change anything but I had to get it out there.
This week I have been thinking about finding a voice. I have experimented with writing in a few different styles this last year and what I have seen is that my writing style always reverts to something that sounds like a generic magazine article, and is quite plain. This is not all bad, as I can now produce a lot more words of above average quality on demand, but it feels to me that the next step in my writing would be to focus more on the execution and the details, like word choice and sentence structure, or whatever else will help me express myself more in my own voice
I see more information about JFK's assassination has come to light with the recent publication of a book called "The Secret Witness" by 88 year old Paul Landis, who at the time of the incident was a Secret Service agent in the car behind the President's.
That article reminded me of a analogous incident much further back in time, about which I recently learned more when reading a book called "William Rufus" by Frank Barlow, published by Yale UP (2000).
He sounds like a fascinating character, and in some ways quite modern in outlook. But so repellent did his attitudes and behavior seem to contempory historians, who were mostly clerics, and many since, that he's had a "bad press".
Barlow devotes a chapter to the event of his reign, as William II (1087-1100), best remembered today: His mysterious assassination in the New Forest, in the year 1100. In relation to this, he includes several facts recorded at the time, but makes no attempt to identify a culprit from among the many suspects.
As luck would have it, I found an ebook copy of his book on a Russian website a month or two ago. (Those naughty copyright-violating Russians have literally tens of millions of ebooks stashed away, many bang up to date, if you know where to look!)
When I reread the book, the available facts pointed to a clear prime suspect for the killing. My conclusion, for reasons briefly summarised below (if anyone cares much), is that there weren't two assassins, as possibly in JFK's case, or even one. I believe the most likely truth, based on all the facts that can be known today, is that the silly sod accidently killed himself!
The first fact is that he was killed while out hunting. Now there were various kinds of hunting, and on that day it was not the kind of hectic style, with packs of hounds, and riders charging about blowing horns. It was a stealth mode deer hunt, in which the participants spaced themselves individually widely apart throughout a forest area and waited for deer to gallop past, which they would try and bag with arrow shots.
Normally on a hunting day the participants would head off, keen as mustard, literally at the crack of dawn. Apart from anything else, it might be several miles from their overnight lodging to the hunting ground. But we are told that on the morning in question they didn't start until after midday. One historian claimed this was because the King had drunk more than usual the night before and had a hangover.
Another chronicler mentioned in passing that a blacksmith arrived at around midday and delivered six arrows, of which the king kept four and gave two to a sidekick called Walter Tyrrell. Although apparently a trivial aside, hardly worthy of mention, this fact may be a key to the mystery!
So in summary, at the start of the fatal day we have a king who may be a bit woozy from the night before, and thus not fit to operate heavy machinery, or any machinery, including new-fangled cross-bows.
Perhaps it was not a hangover which delayed the start of the hunt. Maybe the chronicler assumed that was the reason for the delay . To my mind, another obvious possibility was that they were waiting for something. From the facts recorded, that was most likely the arrows which the blacksmith was due to deliver.
Now imagine a Texan billionaire inviting his rich pals on a hunting trip. On the day, they all have to wait for a gunsmith who eventually turns up with a mere six bullets. Sounds ridiculous doesn't it? They would have crate loads of ammo, and it would have been the same with arrows on that fateful morning.
They must have had ample supplies of arrows, with normal heads, and (broad bladed) hunting heads. So the arrows the blacksmith delivered must have been very special, and I suspect they were crossbow bolts. According to Wikipedia, crossbows were only reintroduced into Europe at around this time. So they were cutting-edge technology, doubtless with various models regularly appearing, as with any new technology.
Standing nearest to the King during the hunt was Walter Tyrrell, some hundred yards away. No participant would have had much of a clear view of any others, especially as they would all have been trying to look inconspicuous so the deer would not be deterred. For the same reason, they wouldn't have wanted any attendants or servants nearby.
Immediately after the killing, Tyrrell hoofed it to France, in the not unreasonable belief that he would be blamed. But for the rest of his days, including on his deathbed, he swore by the blood of Christ that he was not responsible for the fatal shot, and most people took their religious oaths very seriously back then, especially when they were about to meet their maker!
So in summary, I believe the king was fumbling to load or reload his crossbow, and possibly turned it upside down, so he could push the bowstring down with his foot (if it was an early model without a windlass, so he would have had to pull the bow string back by hand). Then he spotted a deer, and in the heat of the moment, he nocked a bolt while the bow was still propped on the ground pointing up at him, and the rest is history ..
26 year old female literature and history teacher who enjoys Bach and Mahler looking for an older male partner who also enjoys classical music, wants marriage and children in a few years, and is oriented towards technical understanding and good taste
Pokemon Gen 2 was based on the Kansai region of Japan, which among other things, contains Nara, a city famous for its deer parks, where deer roam around the city and visitors buy special crackers to feed them. Pokemon Gen 2 also introduced Stantler, the first deer pokemon. Unfortunately, it's just a random wild pokemon on Route 36/37. They really missed an opportunity to have a fictional analog of Nara there. They could have made it like the Safari Zone where you feed crackers to Stantler in order to catch them.
I have recently had some work experiences that give me some insight into one of the potential reasons why so much of modern architecture is so ugly: because the way we build buildings these days requires architects to precisely specify minute details of every aspect of the building on computer generated 2D architectural drawings.
I work for an architectural lighting company, and recently I have been asked to start making production drawings for our orders. Sometimes the customer is fairly clear and this is easy, but not always. Today I finished up a project that took 20-30 hours, where I had to try to interpret the architectural drawings for a building to get the details of what the customer needed to order, and how we needed to build the lights to satisfy their need. There was such a staggering amount of data on these drawings. I was extremely fortunate to have had a copy of the drawings which the architect had helpfully highlighted all of the locations I needed to scrutinize, and only received the pages relevant to my needs, though I could tell from the page numbering and table of contents that there were over a hundred pages in the whole document. This is a staggering amount of work to produce, and frankly not a particularly great way to convey all the information necessary to build this building.
For instance, some of the most difficult lights to interpret were the ones on the stairs. This building had a set of stairs with a super common arrangement, where you go up half a flight, turn 180 degrees, go up the rest of the flight, then repeat. They wanted linear lights on the underside of the stairs. But despite having multiple views of each set of stairs, the only way I was able to figure out that they actually wanted u-shaped runs on the bottom of each set of stairs was the hand-made drawings somebody at some point higher in the process than me had produced at some point, and which were included in the information packet I was given. These kinds of stairs are simple and common, and yet with what is essentially a square spiral shape, not that easily depicted in a series of 2D drawings. Especially when those 2D drawings include not only the lights I am trying to specify, but also all the structural elements and trim and flashing and every tiny little detail. It is just so much to go through.
Hundreds of years ago, when a team of builders built a cathedral, there is no way they would have specified all the minute details like this, especially for something light lighting. They would build the structure of the building, with some plans for how different parts of the structure would be illuminated. Then, when it was time to finish the interior, the aristocrat in charge of the project would walk through the space with a head craftsman and discuss the broad goals of how the illumination sources would be arranged, and then the head craftsman would work with a team of skilled artisans to build and install the lamps and other fixtures in situ. The important point being that the small little details would be left up to the skilled artisans responsible for the labor of manufacturing and installing the fixtures.
My company COULD do things this way too, if the world was set up to operate this way. Our products are highly customizable and not terribly complicated. We could have sent out a team of a few skilled artisans in a truck with a nice portable mitering saw and a pile of the materials we build our fixtures from, and they could have built everything on site exactly to fit the space, with only vague direction from the architect about what needs to go where, and what kind of style and illumination they want in each location. I think this would be cheaper and take much less time overall then the way we currently do it, where we spend many hours of time with customer service and reps and everyone going back and forth again and again on exactly what is needed. Our products aren’t difficult to assemble, and don’t require heavy machinery. They could be assembled in the field. And this would mean we wouldn’t need to spend a long amount of time carefully packing them for shipping, which is difficult and expensive given that our standard size fixture is 8’ long.
But this isn’t what the customer thinks they want. They want a highly customizable pre-made product that they can slip into place at with unskilled laborers. We have a ton of problems with our products being installed incorrectly, which just adds more time and back-and-forth, and shipping broken products back-and-forth for repairs and adjustments and replacements. And it requires we build our products robustly enough to be installed by laborers who we know will damage them, and robust enough to be shipped without breaking. It all feels incredibly wasteful and unnecessary to me. But this is how builders and architects expect things to work.
And 2D drawings… really? Can’t you just give me a 3D model of the building? No, of course not, that would violate somebody’s intellectual property. That or the architectural drawing software the architect uses won’t give us a license to a reader for those files.
Point is: the way we build buildings these days is with the expectation that every single minute little detail is fully specified in drawings before construction begins. This requires a tremendous amount of effort to plan out, and generates a tremendous amount of data that is difficult to efficiently convey. And of course, standard features are much easier to draw/design with architectural software than some complex, novel artistic concept. And so architects and designers feel this pressure to keep repeating the same patterns that are easy to draw again and again, which is why so much of modern architecture is boring, ugly, and similar.
Regarding your language learning proposal from a while back, I think English -> Japanese is one of the *worst* examples you could have chosen. You could kind of do what you propose for closely related languages like Dutch or German where the word patterns closely match English (though even then, good luck explaining gender), but English and Japanese are just so utterly different that Mad Libs study makes no sense.
Anyone have advice about reverse mortgages? We're going to outlive our retirement funds if we don't do something and that looks like an option. We have substantial equity in our home in a high cost of living area, don't want to move, and don't need to worry about leaving any inheritance.
What an absolutely wild ride this continues to be... almost any option seems to still be on the table, including OpenAI's board stepping down and their successors reinstating Altman (good, but fraught without a clear account of the motivations and supposed reasons for his firing), the acquisition or acqui-hiring of OpenAI by Microsoft (bad short-term, ok-to-good-medium-term, probably bad long-term), or the OpenAI board staying the course and hoping the company isn't a ghost-town by next week under Emmett Shear as CEO (worst, no knock on Shear, but this is the bad-end outcome that likely results in bargain-bin acquisition by Microsoft with serious losses of employee retention and major interruptions in development and service).
These options and more are largely all in play, and Microsoft wins big in almost any scenario. This may be why Microsoft CEO Satya Nadella has been so magnanimous in keeping these many paths open, characterizing Microsoft's hiring of Sam Altman and Greg Brockman as a "holding action", and committing to continued support and partnership with OpenAI regardless of how things shake out. Microsoft currently has the ability to essentially end OpenAI as a solvent company right now, but Nadella has (imho) shown a great deal of leadership and pro-cooperation tendencies when the chips are down, at the present juncture... maybe in part because there are very few paths where Microsoft doesn't come out on top in this shake-up.
I reject any interpretation that some of the decision-makers here were "playing 5-D chess" or planned for any of this... there's simply too much variability in highly stochastic systems (such as human choice). Rather, the arc of this entire story has been characterized by extremely reactive decisions where the likely consequences weren't thought out or well-considered, with the outcomes of those decisions spiraling quickly into chaotic no-win scenarios. As usual, the winners here aren't those who had some kind of "grand master plan", or even expected the players to respond rationally according to their incentives and self-interest... but those who could respond quickly, effectively, and cooperatively to events where decision-makers acted in irrational and self-damaging ways, while also leaving opportunities open for "saving face" and not rubbing salt into the wounds of any perceived vulnerability.
I applaud Satya Nadella, Sam Altman, and Greg Brockman... this has been a master-class in damage control and applied game theory, in many ways... as well as Ilya Sutskever for admitting when he was wrong, taking accountability for his choices, and course-correcting. None of those things are easy or natural, and it speaks to the professionality of all involved that Altman and Brockman responded very positively to his contrition in the face of what must have felt like a massive betrayal by Sutskever.
I await any further developments just as everyone else is.
Christopher Mims recently voiced some suspicions I've had recently regarding smartphones ('Social Media Is Warping Into Old-Fashioned Mass Media,' Wall St. Journal, November 18-19). Is the nearly obsessive use of smartphones healthy.
I don't own a mobile phone, and never have.
It's not that I'm particularly against it, although it has made driving a lot more dangerous. I just don't see the point. I've never played a video game, either. Again, the point?
Is there anyone else out there who hasn't turned into a cyborg?
I haven't had a chance to read all the comments on the Girard post, but I thought Scott was overly harsh on Girard, especially the last two chapters on political correctness (this was Girard's term). Scott writes: "So Girard is stuck in an awkward position of saying that the rise of concern-for-victims was good when Christianity is doing it, is bad now, and not having any good theory of what changed, or how this relates to the more speculative anthropology." I take Girard to be arguing that what went wrong is that contemporary western culture took the concern for victims from Christianity but then threw away the rest of the moral framework in which it was embedded. That moral framework includes, for example, exhortations to love your enemies and forgive those who persecute you. Take away those things and you end up with a system that is ostensibly concerned with victims but uses that concern to justify the kinds of scapegoating and victimization it's supposed to be against. As for why this changed, this is just Satan reasserting himself within the moral system that threatens his power: using the concern for victims against itself. Of course this doesn't explain why political correctness arose exactly when and how it did, but I don't think Girard is trying to explain specific details of history like that.
The BlueCrossBlueShield carrier I use got hacked, and what was taken was not just account numbers but also passwords. BCBX paid for all those whose records were exposed to get 2 free years of Experian identity theft protection. So I signed up for that, but am not sure that was a good idea. Experian already knew a shitload of stuff about me. In order to prove I was who I said I was I had to give correct answers to a bunch of questions they asked me about my own finances, such as name of banks I have used in the past, and model of car I bought 5 years ago (how do they even know that? It was a cash transaction between me and previous owner.). And when I signed up for Experian's identity theft monitoring I gave the company a bunch more information about my finances, including numbers, expiration dates and security codes of all my credit cards, numbers of all my bank accounts. So now I'm thinking, so what if Experian gets hacked?
Has anyone found a good dark chocolate brand that is low on heavy metals? I eat a lot of dark chocolate, like 1/3 of a bar per day, and I'd hate to give it up.
I worry Altman’s sacking illustrates what I long feared: that the limited influence of AI safety enthusiasts on the world will be burned for negligible impact on AI safety.
Now is not the time.
LLMs reduce AI risk, in the same way calculators reduce AI risk: a person with a calculator is “superintelligent” compared to one without, so the calculators technology raises the bar of how intelligent AIs have to to surpass humanity.
(Of course, LLMs also increase AI risk, in several ways which were discussed to death here. But I expect no one to read these parenthesis! … also, it makes sense to me that exploiting LLMs for all they are worth will reduce AI risk according to the argument above more than it will increase AI risk, because in LLMs at least the initial training objective is reasonably orthogonal to paperclip maximization arguments.)
Has anyone here read Tom Holland's *Dominion*? I am starting it now, and definitely intrigued by some of the crossover between Holland's points and the recent "I See Satan Fall Like Lightning" review.
OpenAI is a private company. The board has no obligation to tell the PUBLIC in advance or later, that they were going to fire the CEO and why. The board members might have had an obligation to tell the shareholders (such as Microsoft) about this.
So I'm wondering why they're being called "secretive" accusingly. They had no obligation to share this with the public, even in the vague terms they did.
It seems like the EA community has had at least three massive failures this year. Firstly Yudkowsky's comments about basically establishing a police state to prevent AI. (You can disagree with this characterization but this is the public sentiment of it. Even if you think he's right it's a PR failure.) Then the SBF debacle which I hope no one is still defending. And lastly this AI ouster. Now, you might think that last one is the right thing to attempt. But pragmatically it's not working. And I don't give you points for trying and failing.
In practical terms the philosophy seems on retreat on all fronts and with a severely tarnished brand. So my question is: are there feedback mechanisms? Does anyone get fired? Or does everyone continue like normal? If everyone does just carry on then I think this probably signals the end of EA as influential in large segments of technology. The AI ethicists with their left coding and elite backing have more sway with the government and the accelerationists will take over the actual companies.
I'd like to share with you the latest issue of my newsletter, Interessant3, available at https://interessant3.substack.com. In this issue, I share links on the following interesting topics:
1. The Chilean Economy: A thorough analysis of Chile's economic landscape, exploring government policies, trade, and internal dynamics.
2. Denmark's Electricity Dilemma: A look into Denmark's reliance on imported electricity despite its significant renewable energy sources, discussing sustainability and energy security challenges.
3. Yemen's Ancient Jewish Community: An exploration of the history and cultural heritage of one of the Middle East's oldest Jewish communities.
Feel free to explore these discussions and subscribe for more insights! Thank you for your interest and happy reading.
As far as I can tell this puts a pretty big dent in the "Ilya pulled the plug after seeing an unexpected AI advance" theory. Though who knows - maybe Ilya realized only too late how important Sam was to the cohesion of the company? And now is scrambling to get him back for that alone?
Does Greg know Ilya's intentions? Are we ever going to get any official disclosure of what the board members were really thinking that Friday?
This whole fiasco is incredible (and a bit terrifying) to watch unfold live, especially for someone who's relatively new to AI safety debates. It's like I'm watching a thriller play out IRL.
EDIT: And Sam himself responded to Ilya's comment the same way.
I don’t want this thread to get too hostile but I feel I must bring this up somewhere, and this place is articulate enough to handle this tendentious subject without too much acrimony.
So trigger warning.
Why do people think that Die Hard is a Christmas movie, it was released in July.
Anyone have any success with free or paid online writing courses? Meaning helping you to launch a career in writing online (on the internet) as opposed to novels or something else. I am currently in a corporate job and really would like to start writing and explore a possible career change, but I am at a loss as to where to start or how to narrow down what to write about. There's a glut of courses and things online purporting to help people with this but I have no idea how to suss out what's worth the money and time. Thanks in advance.
I was just reading about the U.S.-dominated "Echelon" radio spying network. Are there any confirmed or likely gaps in its coverage, or does it intercept 100% of radio broadcasts originating anywhere on Earth's surface?
Does anyone know a good book about caretaker/professional carer (nurses, therapist, etc) burnout? Ideally I’m looking for something at the same level of technical depth and scope as The Body Keeps the Score. Thanks!
I know there are several people on the open comments who are themselves writers on substack. Question for you: Has your stack and your desire for readership started to shape what you write and how you write? I came to the decision recently that I would not let this happen to me, but is this just hubris? Is it possible not to let this happen?
I've been curious for a while about traffic engineering (like how to decide which way a stop sign faces or how to time a light). I'd be very interested in a good textbook about the topic and a more approachable (sort of pop science if you will) nonfic about it. If anyone has any such recommendations, it would be greatly appreciated. Thank you!
I was thinking more about Scott's comment after running the impact certificates thing, where he felt that if the investors were making a profit, then the grant funders were sort of getting screwed.
Having given this some thought, I think he was sort of correct, but it's an easily fixable problem. Some quick economics first (if you have a micro background, skip the next paragraph).
Generally there is an idea of a supply and demand curve, where the idea is that buyers who really value a product will bid the price up and sellers will bid the price down until no one is willing to buy/sell any more goods at a different price. Generally whoever bought the last food was somewhat indifferent (got little value) since if they were still getting lots of value, they might have bid it up more. But, in general, MOST customers are paying less for the product than they might be willing to pay, and this gap between actual and maximum price generates value for them.
With the way that the impact grants market was structured, it was almost impossible for Scott to get value. He structured it as "I will decide how much I value your product at, and then pay you that amount" which means that he loses valuable money at exactly the rate that projects generate value.
My proposal would be pretty simple, which is to say that next time you run it, announce that you will value projects as normal, but then buy impact certificates at, say, 80% of their value. This still gives investors room to make a profit (if it's far cheaper to generate value), but leaves Scott having saved on value generated
An issue of human perception: do many audiophiles prefer vinyl due to a real difference in sound or is it just the ceremony of taking a record out of a work of art and putting it on a turntable?
Here is the twist: there IS a difference in the signal extracted from CDs. While you can theoretically put all the information that a human can hear on a CD, the playback method of CD players is not the same as the playback described by Nyquist's Sampling Theorem.
On the vain of language teaching methods, a while ago I encountered a well documented Anki deck for the learning of verb conjugation, created by one Andy, an approach he since named "the Kofi method". I took advantage of the Spanish deck (an Italian one is also available and a French one is coming up), and I can attest to the effectiveness and efficiency of using such flash cards (I made superficial changes to them and followed a different order and pace than suggested, which nonetheless was easy to execute since the notes are well tagged). He wrote a post about the decks, including a section about how he generated them.
I have a question about Russian-Ukrainian relations during the Soviet era.
If I imagine war between say Britain and France in 2023 it's almost unimaginable, there' a long history of good relations (since at least 1945), we're ideologically similar, in the same geopolitical alliance, etc. France just doesn't seem like the kind of country we conceivably cold have a war with. Ditto for most Western countries.
In the Soviet era, I'd like to know if Russia and Ukraine had a similar relationship, being ideologically and geopolitically aligned, and sharing a similar language/culture. Of course there were nationalist movements in USSR, and a history of Russian domination,.
If you told a soviet citizen in the 80s that there would be war, would they be shocked or think something like "Yeah, they're the kind of country we might fight."
Has there always been a bit of mutual suspicion or is it possible for relations to deteriorate from a Britain-France level in 20-30 years, or even since 2014 when Ukraine began to move towards the Western camp?
It'd be especially cool to hear from anyone that was an adult in 80s Eastern Europe.
"Sunday’s Philadelphia Marathon was Baude’s 12th marathon in a year, including two Ironman triathlon competitions — the culmination of a feat of endurance aimed at drawing awareness to the national need for living kidney donors.
Baude donated her kidney to a man she’d never met before in 2021, after hearing of his plight — a five-year wait for a lifesaving organ transplant — on national news. She and her racing partner, Matt Cavanaugh, a fellow donor who heads the National Kidney Donor Organization, hope to show potential donors that they can continue a healthy, physically active life after donation."
I'm pleased to see they aren't making unproven promises about longevity, but I wonder how good a marker running marathons is.
"Microsoft has hired Sam Altman and Greg Brockman to lead a team conducting artificial intelligence research, days after the pair were pushed out of OpenAI, the company they co-founded.
Writing on X on Monday, Microsoft chief executive Satya Nadella said that Altman and Brockman, “together with colleagues, will be joining Microsoft to lead a new advanced AI research team”.
“We look forward to moving quickly to provide them with the resources needed for their success,” he added.
Altman retweeted Nadella’s post on Monday, adding: “The mission continues.”
Microsoft’s share price was up 1.7 per cent in pre-market trading on Monday, bouncing back from a similar fall on Friday after the announcement of Altman’s firing."
Right, let me get the mean comment out of the way first: and these are the people the AI doom brigade are hoping will save the world? They can't even sort out their own company (insert eyeroll here).
I'm not surprised by this; note the drop and then bounce back of the Microsoft share price. This is why I have no belief in calls for pauses or moratoria; write all the Open Letters you like, but when it comes down to "please join us to accelerate the tech for the money fountain" and even more "yikes our share price!!!!", then there will be no well-intentioned 'for the common good' halting of progress.
AI is coming and pretty much already here, remember Microsoft have just had a week of "how to integrate AI into your business; if you don't integrate AI into your business you are going to fall behind very badly; let us do the integrating for you". AGI may well be on the way, but I remain convinced that what we are going to get is more and more Smart Dumb Machines that will be increasingly used to generate output, which will (as we have seen) invent total nonsense rather than reply "no answer" and people won't check that output because (1) they're being sold on 'the machine is always right' and (2) businesses will be using AI for cost-cutting, so there won't be the time to check or the warm bodies to do any checking.
We'll have people docilely dosing themselves up with arsenic because the AI Doc In A Box told them to do so, as per a distracted Watson in "The Sign of the Four":
"He was clearly a confirmed hypochondriac, and I was dreamily conscious that he was pouring forth interminable trains of symptoms, and imploring information as to the composition and action of innumerable quack nostrums, some of which he bore about in a leather case in his pocket. I trust that he may not remember any of the answers which I gave him that night. Holmes declares that he overheard me caution him against the great danger of taking more than two drops of castor oil, while I recommended strychnine in large doses as a sedative."
Let's continue re-reading the old Scott posts https://archive.ph/fCFQx, even in these weird days. There was a period of time when Scott was something of a travel blogger. In "Stuff" (https://archive.ph/FfJaDhttps://justpaste.it/9z9wk) he goes to Jerusalem, narrowly avoids being immolated by the holy fire and oversees the preparations for the coming of the Messiah.
Weirdly, I couldn't post the text on pastebin, because apparently: "Pastebin’s SMART filters have detected potentially offensive or questionable content in your Paste.
The content you are trying to publish has been deemed potentially offensive or questionable by our filters, because of this you’re receiving this warning.
This Paste can only be published with the visibility set to "Private"." Wtf?
Stray thought (that probably should have gone on the ketamine post - too late now): if all anaesthesia has an antidepressant effect, is there a chance that some of the effectiveness of modern ECT comes from the fact that we put people under for the procedure? Has anyone done a study trying to control for this somehow? (eg groups with anesthesia + ECT vs just anesthesia).
I'm currently a Software Engineer at Google, based in Poland. I want to move and am exploring my options. One option is an internal transfer within Google but I'm curious about other opportunities that might be out there.
Right now USA is my first choice but I'm open to considering other locations as well. Anyone knows how the situation looks right now / what companies are hiring and offering good conditions? (recruiters who reach out to me are only offering positions in EU)
Javier Milei has won the presidential election in Argentina. While his plan to abandon the peso and use the US dollar as Argentina’s currency has gotten the most attention, he’s also in favour of legalising organ sales. So medical tourism to Argentina may become a big part of the answer to the kidney shortage!
> Condolences to everyone in AI right now, I hope you’re all okay.
Is that the Sam Altman thing. I don’t really follow the ups and downs of the average CEO. But that was… weird, right?
The board gets rid of the CEO and another founder based on a relatively flimsy reason, with no real specifics. But, fine. Boards can do that. In fact they should do it more, for underperforming CEOs.
Then it turns out that the main investors, including Microsoft, are flummoxed by this and were not informed. They want him back. Did the investors not get members on the board after they invested billions? Shouldn’t they dominate the board? Shouldn’t they be informed before the coup at the very least.
Edit. As a write I googled Altman and he’s hired by Microsoft. Who are now investors in OpenAI and doing their own thing with AI.
Open Thread 303
Does anyone know why Kalamazoo and Numazu are sister cities? Did they just choose each other because of the name similarity? Orr was that pure coincidence?
I have to say, it's amusing how the guy just leaves the board open while recovering from kidney donation and it turns into a giant Israel-Palestine argument.
Thread for publicly sharing anonymized information about the Open AI board members, since I suspect many readers here to have various Open AI connections.
https://openaiboard.wtf/
What are you hoping to find in your stocking that no family or friends would think to stuff?
Been kicking around an idea/observation over the tg holiday.
Call it the “hate coefficient.” This board tends to prioritize verifiable evidence. But in this regard, internet, crowd-sourced argument presents a vulnerability. Tom the very-motivated racist, communist, anti-fascist, Palestine-hater or Israel-hater or what-have-you has a near inexhaustible capacity to dredge twitter, Wikipedia, Facebook, telegram, or whatever source is necessary to find evidence, however specious, for his preferred conclusion. In a debate, then, predisposition and motivated reasoning can transmute themselves into a endless barrage of “evidence” for how the blacks are the most hateful race, or how Russians are the real defenders in Ukraine or what have you.
One should never discount evidence, to be sure, but a disinterested third party trading study for study ends up at a surprising disadvantage- if I care a fair bit about not slandering an entire race, but Tom hates the Jews or the blacks or the French a capital-L “Lot,” for every item of evidence I’m willing to find and present, he’s more than willing and able to take as many hours of internet he needs to find and present 2.
The internet being the inexhaustible font of garbage that it is, a sufficiently motivated reasoner can easily drown a debate- not by actual preponderance of evidence, but by “preponderance of evidence I’m willing to find.” A sufficiently motivated flat-earther can just keep digging and throwing up links to the point that anyone contradicting him for the “sake of argument” becomes exhausted and calls it a day.
Which can leave the public square looking like “earth might be flat- tom’s evidence hasn’t been rebutted” even when the facts on the ground are more like “flat earth Tom threw so much garbage that no one had it in them to keep refuting it.”
At the same time, evidence matters. This phenomenon is real, but if you take it as license to ignore facts you don’t like, you’re blinding yourself. I guess you just have to take the grain of salt for very-opinionated-internet-man while also taking that same grain of salt for yourself when applying that label to him.
I don’t know. Reasoning is hard I guess. I wish I had a conclusion or a clear perspective but it seems like a prisoners’ dilemma we’re all stuck with, discounting by the hate coefficient as best we can.
It's often said that the term "Rationalist" is a poor choice because they're on the empiricist side, but I wonder how much that's actually true. The movement started in the late 2000s when New Atheism vs Creationism was the main war of the day, so there is all the obligatory lip science to The Power of Science and so on, but beyond that, Yudkowsky really seems to prefer rationalism over empiricism.
For example, in HPMOR, while Harry does do *some* experiments, they're unconnected to any of the benefits he gets. Harry's modus operandi is 1) Think about things and decide how the world *must* be based on intuition, 2) Believe *really* hard in your theory, 3) Be right because you're the author avatar and get rewarded with unique magical powers. (At least that's how he got Kill Dementor and Partial Transfiguration - the rest of his powers come from randomly getting OP magical artifacts dropped into his lap for no reason.)
Meanwhile, Yudkowsky's other classic writings seem to have a remarkable amount of contempt for actual scientists for someone ostensibly on the Pro Science side of the 2000s Religion Wars.
Meanwhile, nowdays in the Yudkowsky-derived AI Doomer movement, a common argument is that AI will be able to near-instantly take over the world because Intelligence means you can magically solve everything just by thinking really hard, no observations or legwork required. No this isn't a strawman, I've seen Doomers *explicitly* make this argument many times, as an argument about why AI takeoff shouldn't be constrained by the speed of running experiments and making observations about the world.
Israelis using palestinian baby bodies in their private screenings of "khamas atrocities" allegedly
https://x.com/bidetmarxman/status/1728163260954775685?s=20
One more thing about the intractability of the fight between Israel and Palestine: they aren't the only players. At a minimum, there's Iran supporting Palestinian aggression (possibly also Russia) and Americans millennialists who want to start the end of the world at Megiddo.
On nuclear regulation:
There's an article that is making the rounds of the rat blogosphere that I think is seriously wrong. You've probably seen it quoted. It blames the ALARA (as low as reasonably achievable) radiation protection standard for all the economic problems of US nuclear power. From https://worksinprogress.co/issue/taming-the-stars/:
"ALARA is defined as: "making every reasonable effort to maintain exposures to radiation as far below the dose limits in this part as is practical consistent with the purpose for which the licensed activity is undertaken, taking into account the state of technology, the economics of improvements in relation to state of technology, the economics of improvements in relation to benefits to the public health and safety, and other societal and socioeconomic considerations, and in relation to utilization of nuclear energy and licensed materials in the public interest." [footnote citing 10 CFR 20.1003]
As currently applied to nuclear power, ALARA literally means that every expense must be spent on eliminating every possible effect of nuclear power, at least until the resulting electricity is no cheaper than what the market pays for electricity generated from non-nuclear sources. Since standards cannot ratchet downwards, only up, safety standards that are just about affordable at the top of energy price spikes get entrenched, meaning that nuclear is made unaffordable until the next price hike – which makes it even more expensive, since it prevents learning and the economies of scale that a steady pipeline of projects can allow. ALARA, as currently applied in the US and much of the rest of the developed world, means that nuclear power is never allowed to be cheaper, no matter how much safer and cleaner it is than other sources of energy. It makes affordable, safe nuclear energy impossible, and forces us to rely on much less safe energy sources instead." End quote.
The first paragraph is a literal quote from the regulations. Everything after that? where the author tells you what ALARA "literally means"? is wrong. At least, I think so. To the extent I understand the claim being made here.
Is the author saying that nuclear regulations actually change in response to energy prices? This absolutely does not happen. Is he saying that inspection standards or radiation protection procedures change with energy prices? So that some regulator or energy company employee is actually making the decision to increase radiation protection standards when they observe nuclear becoming cheaper compared to non-nuclear energy? Highly implausible. Energy prices change all the time, and regulations/inspection procedures/radiation protection procedures are only changed in a slow and cumbersome way. Also, industry would have no incentive to make itself less competitive, and it is very much NRC culture to NOT pay attention to energy prices.*
Okay, maybe the author is making a more general claim that the level of safety/security regulation increases over time, it's a one-way ratchet and regulation prevents nuclear power from being as cheap as it arguably should be compared to other energy sources. A fair but unoriginal claim. But then why the talk about ALARA?
First of all, understand that ALARA is about radiation protection. It is not the be-all and end-all of nuclear regulation. The ALARA standard adds on to other radiation dose regulations. For example, a typical nuclear power plant worker can get a max of 5 rem per year of occupational radiation exposure (10 CFR 20.1201) AND their radiation dose must be ALARA. So if a worker gets more than 5 rem, it's a violation of both regulations. If a worker gets less than 5 rem but the plant does not make reasonable effort to make the dose ALARA, it could be a violation of the ALARA standard. Conclusion...even if the ALARA standard didn't exist, nuclear plants would have to put significant effort into radiation protection, albeit not quite so much.
I'm not gonna say ALARA is unimportant. But it's only one of a whole host of regulations that apply to nuclear power plant design, construction, operation, and decommissioning. There are regulations that apply to nuclear security, reducing and mitigating the risk of nuclear accidents, emergency planning, environmental protection, and I could go on. There would be a significant regulatory burden even without ALARA.
Maybe the author is using ALARA as shorthand for the entire group of US regulations and laws relevant to nuclear? Or the entire regulatory mindset? But, if your argument is that nuclear regulation should incorporate cost considerations, why pick on one of the regulations that explicitly incorporates consideration of cost, instead of the many that don't consider cost at all?
Another quote: "[T]he components that are not safety critical are still subject to a gold plated ALARA standard. This means the same component is regulated differently depending on whether it is in a coal plant or a nuclear plant, even if it is far away from the reactor and cannot affect it."
False. The reason that a component in a nuclear plant is regulated differently from a component in a coal plant is that different laws, regulations, and administrative agencies regulate nuclear plants from those that regulate coal plants. ALARA has absolutely nothing to do with that.
I hate to be all argument from authority, but I notice the author, John Myers, seems to be a UK YIMBY activist and if he has any experience in US nuclear, I'm not aware of it. Please understand that the statement "ALARA, as currently applied in the US and much of the rest of the developed world, means that nuclear power is never allowed to be cheaper, no matter how much safer and cleaner it is than other sources of energy" is false. That is not what ALARA means. ALARA is not that powerful. Please stop quoting this guy uncritically.
*Because NRC's mission is to ensure nuclear safety and security, not to ensure that the US nuclear power industry is economically viable. If you want something to complain about, ask Congress to change that.
I have some ideas about what Israel should be doing. I'm not sure whether I'm right, nor whether this is psychologically possible even if it is right.
I think Israel should be looking to its own borders and its own safety. Wrecking Gaza will not necessarily make Israel safer, and may be putting it at more risk. It's certainly creating more hatred for Israel and I gather there are Hamas leaders in other countries-- they aren't at personal risk from the attack on Gaza..
10/7 wasn't just an atrocity, it was an embarrassment. I assume the borders are getting more attention, but are they getting more thoughtful use of tech? Bulldozer-proof barriers?
Destroying Hamas' tunnels has some practical and humanitarian issues, but additionally, the attack was by air and sea as well as by underground.
As I understand it (discussion is welcome), Hamas' intent was to provoke Israel into a drastic reaction so the world would stop supporting Israel (maybe also to make it more likely for Moslem countries to attack Israel), so that Israel could be destroyed. It's a vile approach, but it might actually make some practical sense. I doubt that Israel will be destroyed, but I still think it would be bad if it were on the receiving end of a big attack.
A part which might not be psychologically possible is to quit abusing Palestinians. Torture and a lot of imprisonment might, oh maybe just might, have something to with why it was possible to keep such such tight security on the 10/7 attack. I'm not sure how many people were involved, but I'm expecting low thousands.
Maybe they *were* warned. I get the impression the Israeli government didn't want to believe such an attack was possible.
Meanwhile, Israeli military capacity is being spent on wrecking Gaza, and perhaps the most valuable thing being wasted is attention.
Just by the way, Netanyahu is staying in power while the attack on Gaza is going on. I'm not sure when the next possibility for getting him out of office is, though I'm betting he will be out. In any case, His incentives to continue the attack are personal as well as emotional.
Sidetrack: it may not be possible to get all the hostages back. I wouldn't be surprised if some of them are dead, and I've heard a plausible claim that some of them are being held by groups other than Hamas.
Hi, I have a small whatsapp group of some 50 people, mist of whom are fellow Orthodox Rabbis.
I've been getting some negative feedback for not vocally supporting Netanyahu's actions so I made a video explaining the full context of what I would do if I were in charge.
Being members of the human species this hasn't served to have too many split off and begin to understand that my feelings of empathy for Arabs weren't dangerous eothin my chosen context but some of you might find it interesting.
As I am currently engaged in attempting to take over the world (not really, I would rather return to being an anonymous hostel traveler) in an immediate manner I'm not especially interested in intellectual masturbation online discussions with anonymii, but anyone willing to step up and choose to join me _if they come to agree with me_ is more than welcome to come on my youtube show and have a conversation to see whether or not you fo agree with me anout the general jist of what we mist do right now to make the world a place we would all enjoy living on a whole lot more than we currently do.
In short, the essentials of my approach come largely from The Torah (I am an ordained orthodox rabbi) and while the absolutely *MOST IMPORTANT* video on my channel is on a wholy different subject than the happenings in Canaan (it is called "Final Chance For A Fresh Start" and is the top video in the eponymous playlist on my channel), here's my basic 3 point plan for what I would/plan-to do regarding the situation in Israel.
Should you notice (as you _should_) that I said nearly nothing regarding how to deal with the legitimate-concern of the suffering of many Arabs under Israeli control you would be right to note its unfortunate absence (and more unfortunate glossing over the matter in a non-sympathetic manner).
The reason is simply that most of my audience comprise Israelis who, being human, are presently incapable of understanding such sentiments as anything other than a callous disregard for their own plight.
If you are willing to consider joining my war for takeover but aren't sure, please get in touch and we'll make a youtube video of it.
May we be worthy (and capable) of winning now.
https://youtu.be/XJZ920oq6h0?feature=shared
Moshe
Just clicked through the "implicit association test" Scott referenced in his "Quests and Requests" post, and got a strong perception that I would get about the same bias given black/white colored squares instead of dark/light skinned people. I think in my mind, negative emotions are in some part defined as negations of positive emotions, and dark skin - as a negation of light skin. So it's natural that it's easier to hold positive<->positive association versus positive<->negative.
It's also a bias of sorts, but not _that_ kind of bias Scott was hinting at, it seems
What are examples of times and places when political change has been both fast and good? (Good in your opinion; fast with respect to typical political change throughout history.) Change directly related to the end of long wars, independence wars and the fall of the USSR don't count.
To be clear: it can be after a (non-independence) revolution, but not a time when things are much better simply because a time of peace has followed a time of war.
Last year I read a book with a lot of stuff in it about Churchill during the war. There were accounts of him wandering around his residence in the night wearing outrageous get-ups. I have forgotten the details -- but some were women's clothiing, like maybe a lacy negligee, and some just absurd, like maybe. a clown suit. Also accounts of his champagne dinners attended by his staff, visiting dignitaries, etc. Churchill would sometimes lead the group in skipping in circles around the table, I believe with music playing.
Others who have read about these things -- how do you think of them? I know he was an alcoholic -- I know he was not crazy. Why did he do those things? Was there more tolerance then for eccentricities of this kind? Was it a way of demonstrating his self-confidence? -- like that he was so sure that he was admired and respected that he felt able to indulge his weirdest whiims in public? Was it a way of making fools of his dinner guests?
If Supply & Demand is a thing, why does Black Friday exist?
If Black Friday is driven by a spike in demand, then I'd expect prices to grow rather than shrink. If Black Friday is driven by the supply side, wouldn't concentrating the costs of logistics/production into a single month make less money than smooth, continuous operations over the course of the year?
The common wisdom I've always received was: suppliers compete on price for business. But this just doesn't add up, to me.
https://going-medieval.com/2023/11/17/no-the-church-did-not-kill-joan-of-arc-you-credulous-dullards/
Gets into detail about Joan of Arc's trial being by secular authorities and lacking many guardrails that the Catholic Church required for heresy trials.
On the one hand, the Catholic Church wouldn't have had her killed, and I'm not sure it would have put her on trial for heresy at all. On the other hand, it's the Church that made heresy trials a serious matter, so I think it deserves some of the blame, though rather indirectly.
A spectacular essay about Joan of Arc, patron saint of Catholics who don't fit well in the Catholic Church, at least on the left side. It actually gave me a feeling of what's it's like to want a patron saint.
http://tigerbeatdown.com/2011/01/09/running-toward-the-gunshots-a-few-words-about-joan-of-ar/
A spectacular essay about Joan of Arc, patron saint of Catholics who don't fit well in the Catholic Church, at least on the left side. It actually gave me a feeling of what's it's like to want a patron saint.
Thinking about the fall-injury incentives thing from https://slatestarcodex.com/2016/11/10/book-review-house-of-god/ is it possible that somehow adjusting the basis on which medicaid and other government programs pay for dialysis would motivate existing medical providers to throw their weight behind reforms?
Would it make sense for advertisers to aim at spaced repetition rather than apparently just buying as much repetition as they can afford?
Any native French speakers here?
The term "oratrice mécanique d'analyse cardinale" has been trending in the meme-verse lately. It's the name of a device in the game Genshin Impact. I'm trying to figure out whether the name makes any sense in proper French.
A straightforward translation into English gets me "mechanical speaker of cardinal analysis," which doesn't make much sense, particularly that "cardinal" bit. But maybe there is more going on here than my high school French skills can manage.
Which tricks or skills have the highest ratio of how impressive people think they are to how long they actually take to learn?
Wonder how many other socialists read these comments as a kind of "hostile observer." I've been reading them for many years (obviously including Scott's previous blog) and I will continue. Scott seems humane and decent, while the commenters are often terrifyingly calculating. So why bother reading? Because rationalists are in the vanguard of technocratic polemics. They are GOOD at making effective arguments, and well, "...keep your enemies closer." I've got enormous faith in rationalism/EA as a bellwether for the future of technocracy/centrism. Give yourselves a hand, rationalists, you MATTER. Thank heavens for this comments section. Staring into the abyss isn't only educational, it's downright fun.
This week I've been playing Slay The Princess, a visual novel that has been going mildly viral and getting outstanding reviews. The premise made it seem like it would be a recreation of AI box experiment: the princess is locked up in a cabin, your job is to slay her, and she will manipulate, threaten or seduce you to stay alive.
(mild spoilers below)
Well it turns out it was less of that and more of a Stanley Parable crossed with Disco Elysium (which I should get to playing sometime soon). The game is essentially a series of vignettes, some touching and some amusing, connected by branching paths. The full playthrough basically requires you to backtrack and re-make your choices, so you can't really play a role of a prudent gatekeeper. Or, well, you can, but it leads to a joke ending and credits roll.
Those who played it, would you like to share your favourite route? (and why is it Razor)
Israeli argument that Hamas uses human shields can't be defended. Israelis are targeting those civilians. The argument that "they could kill more so they arent actually targeting them" doesn't work either. They are killing as many children as they can get away with, for the purpose of breaking Hamas's will. The goal is genocide and to solidify their theft of the West Bank.
Over the past couple of weeks I've been seeing an increasing number of articles on the subject of battery systems for renewable energy becoming price competitive with gas-fired plants (example link below). Given that intermittency of renewable energy has been THE sticking point in regards to the energy transition, that seems like pretty big news. Can anyone with more experience or knowledge in the subject offer some insight as to what degree this is hype or if we're on the cusp of a genuine shift in the economics of power generation?
https://www.reuters.com/business/energy/giant-batteries-drain-economics-gas-power-plants-2023-11-21/
Interesting article on OpenAI with interesting final line: "no corporate structure, no matter how well intended, can be trusted to ensure the safe development of AI" - replace "corporate structure" with "AI design" or whatever and it applies (corporations being like AI in many respects) https://www.theatlantic.com/technology/archive/2023/11/openai-ilya-sutskever-sam-altman-fired/676072/
I'd like to find a sperm donor whose sperm increases the chance of various desirable traits: health, IQ, talents, looks, etc. Additionally, its extremely important to me to minimize the chances of mental health issues - because the egg will likely bring some. Finally, I'd like to increase the chances for things such as values being close to mine and overall usefulness/success in life.
1) The best approach to all of the above seems to be to know someone's wide family. If there's no history of mental health issues X generations down and across many people, that sounds like reasonable probability. The same for other traits. (Yes, I'm thinking a bit in the vein of https://www.astralcodexten.com/p/secrets-of-the-great-families ).
2) Do any official institutions (spermbanks) offer anything similar? If yes, would you have tips? If not, why not? Are there regulatory issues or is there just such low demand / high stigma?
3) Can you think of a better way to find donors than just get tips on Wikipedia, on these forums and through chain emails sent to competent friends who know competent friends and then doing deep background checks on their families?
Round and round the OpenAI drama goes, where it stops...
https://arstechnica.com/tech-policy/2023/11/reports-sam-altman-in-talks-for-openai-return-board-members-could-be-ousted
https://www.bloomberg.com/news/articles/2023-11-21/altman-openai-board-open-talks-to-negotiate-his-possible-return
Eli Dourado posts this great dichotomy on X:
"The world is not that complex, reductionism works, intelligence is basically what matters, world optimization should be tried, all it takes is high agency people with the right values.
OR
The world is very complex, marginalism is what works, intelligence alone isn’t worth much, tacit knowledge and experience and tradition are valuable, smart people thinking they can optimize the world is hubris and inevitably leads to failure or worse."
Which do you think has more truth value? I think I'd go with 10/90 former/latter. A good response I saw says: "first one locally, second one globally".
https://twitter.com/elidourado/status/1726730831048130593
I know that I've massively missed the bus on this, but is anyone else annoyed by the historical inaccuracy of calling the "rationalist community" as such? In the history of Western philosophy, the divide between rationalism and empiricism is one of the main splits, and the modern "rationalist" movement clearly falls on the empiricist side. Empiricism was about basing your view of the world on sense data, which is what modern "rationalism" does with its focus on Bayesian updating as the core means of knowledge acquisition. Meanwhile, actual historical rationalism held that if there was a conflict between your preconceived internal ideas about the world and your sense-based observations, instead of updating your internal ideas, you held that it was your senses that were wrong. This is how you got stuff like the Eleatics (who were essentially proto-rationalists) holding that change didn't exist despite change being observable at every moment of existence, or Leibniz holding that this was the best of all possible worlds despite all the easily observable evil in it. As you can see, this is the complete opposite of the epistemic system advocated by modern "rationalism". If I were to come up with a more accurate label for this movement, which I know it's much too late for, I'd call it Bayesian Empiricism, or maybe Neo-Empiricism. Anyways, that's my rant, I know it likely won't change anything but I had to get it out there.
NYT humorous moment this morning:
On the same iPhone screen:
Shot:
‘I Think We’re Heading Toward the Best World Ever’: An Interview With Sam Altman
Chaser:
“As A.I.-Controlled Killer Drones Become Reality, Nations Debate Limits“
It would be ironic if a dumb AI kills us all with guns and bombs before the smart one gets the chance.
https://www.politicshome.com/news/article/boris-johnson-bamboozled-scientific-advice-covid-pandemic
Discussion of Boris Johnson not understanding exponential growth (very important for policy on COVID) or that science is a process of learning.
I'm wondering want a Science for Politicians course would include.
Just to be utopian and science fictional about it, they have to pass the course in order to hold office.
This week I have been thinking about finding a voice. I have experimented with writing in a few different styles this last year and what I have seen is that my writing style always reverts to something that sounds like a generic magazine article, and is quite plain. This is not all bad, as I can now produce a lot more words of above average quality on demand, but it feels to me that the next step in my writing would be to focus more on the execution and the details, like word choice and sentence structure, or whatever else will help me express myself more in my own voice
I see more information about JFK's assassination has come to light with the recent publication of a book called "The Secret Witness" by 88 year old Paul Landis, who at the time of the incident was a Secret Service agent in the car behind the President's.
https://www.dailymail.co.uk/news/article-12586167/JFK-proof-two-shooters-secret-service-agent.html
That article reminded me of a analogous incident much further back in time, about which I recently learned more when reading a book called "William Rufus" by Frank Barlow, published by Yale UP (2000).
https://www.amazon.com/William-Rufus-English-Monarchs-Barlow/dp/0300082916/
He sounds like a fascinating character, and in some ways quite modern in outlook. But so repellent did his attitudes and behavior seem to contempory historians, who were mostly clerics, and many since, that he's had a "bad press".
Barlow devotes a chapter to the event of his reign, as William II (1087-1100), best remembered today: His mysterious assassination in the New Forest, in the year 1100. In relation to this, he includes several facts recorded at the time, but makes no attempt to identify a culprit from among the many suspects.
As luck would have it, I found an ebook copy of his book on a Russian website a month or two ago. (Those naughty copyright-violating Russians have literally tens of millions of ebooks stashed away, many bang up to date, if you know where to look!)
When I reread the book, the available facts pointed to a clear prime suspect for the killing. My conclusion, for reasons briefly summarised below (if anyone cares much), is that there weren't two assassins, as possibly in JFK's case, or even one. I believe the most likely truth, based on all the facts that can be known today, is that the silly sod accidently killed himself!
The first fact is that he was killed while out hunting. Now there were various kinds of hunting, and on that day it was not the kind of hectic style, with packs of hounds, and riders charging about blowing horns. It was a stealth mode deer hunt, in which the participants spaced themselves individually widely apart throughout a forest area and waited for deer to gallop past, which they would try and bag with arrow shots.
Normally on a hunting day the participants would head off, keen as mustard, literally at the crack of dawn. Apart from anything else, it might be several miles from their overnight lodging to the hunting ground. But we are told that on the morning in question they didn't start until after midday. One historian claimed this was because the King had drunk more than usual the night before and had a hangover.
Another chronicler mentioned in passing that a blacksmith arrived at around midday and delivered six arrows, of which the king kept four and gave two to a sidekick called Walter Tyrrell. Although apparently a trivial aside, hardly worthy of mention, this fact may be a key to the mystery!
So in summary, at the start of the fatal day we have a king who may be a bit woozy from the night before, and thus not fit to operate heavy machinery, or any machinery, including new-fangled cross-bows.
Perhaps it was not a hangover which delayed the start of the hunt. Maybe the chronicler assumed that was the reason for the delay . To my mind, another obvious possibility was that they were waiting for something. From the facts recorded, that was most likely the arrows which the blacksmith was due to deliver.
Now imagine a Texan billionaire inviting his rich pals on a hunting trip. On the day, they all have to wait for a gunsmith who eventually turns up with a mere six bullets. Sounds ridiculous doesn't it? They would have crate loads of ammo, and it would have been the same with arrows on that fateful morning.
They must have had ample supplies of arrows, with normal heads, and (broad bladed) hunting heads. So the arrows the blacksmith delivered must have been very special, and I suspect they were crossbow bolts. According to Wikipedia, crossbows were only reintroduced into Europe at around this time. So they were cutting-edge technology, doubtless with various models regularly appearing, as with any new technology.
Standing nearest to the King during the hunt was Walter Tyrrell, some hundred yards away. No participant would have had much of a clear view of any others, especially as they would all have been trying to look inconspicuous so the deer would not be deterred. For the same reason, they wouldn't have wanted any attendants or servants nearby.
Immediately after the killing, Tyrrell hoofed it to France, in the not unreasonable belief that he would be blamed. But for the rest of his days, including on his deathbed, he swore by the blood of Christ that he was not responsible for the fatal shot, and most people took their religious oaths very seriously back then, especially when they were about to meet their maker!
So in summary, I believe the king was fumbling to load or reload his crossbow, and possibly turned it upside down, so he could push the bowstring down with his foot (if it was an early model without a windlass, so he would have had to pull the bow string back by hand). Then he spotted a deer, and in the heat of the moment, he nocked a bolt while the bow was still propped on the ground pointing up at him, and the rest is history ..
26 year old female literature and history teacher who enjoys Bach and Mahler looking for an older male partner who also enjoys classical music, wants marriage and children in a few years, and is oriented towards technical understanding and good taste
lilyreadsyouremail@gmail.com
Pokemon Gen 2 was based on the Kansai region of Japan, which among other things, contains Nara, a city famous for its deer parks, where deer roam around the city and visitors buy special crackers to feed them. Pokemon Gen 2 also introduced Stantler, the first deer pokemon. Unfortunately, it's just a random wild pokemon on Route 36/37. They really missed an opportunity to have a fictional analog of Nara there. They could have made it like the Safari Zone where you feed crackers to Stantler in order to catch them.
I have recently had some work experiences that give me some insight into one of the potential reasons why so much of modern architecture is so ugly: because the way we build buildings these days requires architects to precisely specify minute details of every aspect of the building on computer generated 2D architectural drawings.
I work for an architectural lighting company, and recently I have been asked to start making production drawings for our orders. Sometimes the customer is fairly clear and this is easy, but not always. Today I finished up a project that took 20-30 hours, where I had to try to interpret the architectural drawings for a building to get the details of what the customer needed to order, and how we needed to build the lights to satisfy their need. There was such a staggering amount of data on these drawings. I was extremely fortunate to have had a copy of the drawings which the architect had helpfully highlighted all of the locations I needed to scrutinize, and only received the pages relevant to my needs, though I could tell from the page numbering and table of contents that there were over a hundred pages in the whole document. This is a staggering amount of work to produce, and frankly not a particularly great way to convey all the information necessary to build this building.
For instance, some of the most difficult lights to interpret were the ones on the stairs. This building had a set of stairs with a super common arrangement, where you go up half a flight, turn 180 degrees, go up the rest of the flight, then repeat. They wanted linear lights on the underside of the stairs. But despite having multiple views of each set of stairs, the only way I was able to figure out that they actually wanted u-shaped runs on the bottom of each set of stairs was the hand-made drawings somebody at some point higher in the process than me had produced at some point, and which were included in the information packet I was given. These kinds of stairs are simple and common, and yet with what is essentially a square spiral shape, not that easily depicted in a series of 2D drawings. Especially when those 2D drawings include not only the lights I am trying to specify, but also all the structural elements and trim and flashing and every tiny little detail. It is just so much to go through.
Hundreds of years ago, when a team of builders built a cathedral, there is no way they would have specified all the minute details like this, especially for something light lighting. They would build the structure of the building, with some plans for how different parts of the structure would be illuminated. Then, when it was time to finish the interior, the aristocrat in charge of the project would walk through the space with a head craftsman and discuss the broad goals of how the illumination sources would be arranged, and then the head craftsman would work with a team of skilled artisans to build and install the lamps and other fixtures in situ. The important point being that the small little details would be left up to the skilled artisans responsible for the labor of manufacturing and installing the fixtures.
My company COULD do things this way too, if the world was set up to operate this way. Our products are highly customizable and not terribly complicated. We could have sent out a team of a few skilled artisans in a truck with a nice portable mitering saw and a pile of the materials we build our fixtures from, and they could have built everything on site exactly to fit the space, with only vague direction from the architect about what needs to go where, and what kind of style and illumination they want in each location. I think this would be cheaper and take much less time overall then the way we currently do it, where we spend many hours of time with customer service and reps and everyone going back and forth again and again on exactly what is needed. Our products aren’t difficult to assemble, and don’t require heavy machinery. They could be assembled in the field. And this would mean we wouldn’t need to spend a long amount of time carefully packing them for shipping, which is difficult and expensive given that our standard size fixture is 8’ long.
But this isn’t what the customer thinks they want. They want a highly customizable pre-made product that they can slip into place at with unskilled laborers. We have a ton of problems with our products being installed incorrectly, which just adds more time and back-and-forth, and shipping broken products back-and-forth for repairs and adjustments and replacements. And it requires we build our products robustly enough to be installed by laborers who we know will damage them, and robust enough to be shipped without breaking. It all feels incredibly wasteful and unnecessary to me. But this is how builders and architects expect things to work.
And 2D drawings… really? Can’t you just give me a 3D model of the building? No, of course not, that would violate somebody’s intellectual property. That or the architectural drawing software the architect uses won’t give us a license to a reader for those files.
Point is: the way we build buildings these days is with the expectation that every single minute little detail is fully specified in drawings before construction begins. This requires a tremendous amount of effort to plan out, and generates a tremendous amount of data that is difficult to efficiently convey. And of course, standard features are much easier to draw/design with architectural software than some complex, novel artistic concept. And so architects and designers feel this pressure to keep repeating the same patterns that are easy to draw again and again, which is why so much of modern architecture is boring, ugly, and similar.
Regarding your language learning proposal from a while back, I think English -> Japanese is one of the *worst* examples you could have chosen. You could kind of do what you propose for closely related languages like Dutch or German where the word patterns closely match English (though even then, good luck explaining gender), but English and Japanese are just so utterly different that Mad Libs study makes no sense.
Anyone have advice about reverse mortgages? We're going to outlive our retirement funds if we don't do something and that looks like an option. We have substantial equity in our home in a high cost of living area, don't want to move, and don't need to worry about leaving any inheritance.
On OpenAI developments-
What an absolutely wild ride this continues to be... almost any option seems to still be on the table, including OpenAI's board stepping down and their successors reinstating Altman (good, but fraught without a clear account of the motivations and supposed reasons for his firing), the acquisition or acqui-hiring of OpenAI by Microsoft (bad short-term, ok-to-good-medium-term, probably bad long-term), or the OpenAI board staying the course and hoping the company isn't a ghost-town by next week under Emmett Shear as CEO (worst, no knock on Shear, but this is the bad-end outcome that likely results in bargain-bin acquisition by Microsoft with serious losses of employee retention and major interruptions in development and service).
These options and more are largely all in play, and Microsoft wins big in almost any scenario. This may be why Microsoft CEO Satya Nadella has been so magnanimous in keeping these many paths open, characterizing Microsoft's hiring of Sam Altman and Greg Brockman as a "holding action", and committing to continued support and partnership with OpenAI regardless of how things shake out. Microsoft currently has the ability to essentially end OpenAI as a solvent company right now, but Nadella has (imho) shown a great deal of leadership and pro-cooperation tendencies when the chips are down, at the present juncture... maybe in part because there are very few paths where Microsoft doesn't come out on top in this shake-up.
I reject any interpretation that some of the decision-makers here were "playing 5-D chess" or planned for any of this... there's simply too much variability in highly stochastic systems (such as human choice). Rather, the arc of this entire story has been characterized by extremely reactive decisions where the likely consequences weren't thought out or well-considered, with the outcomes of those decisions spiraling quickly into chaotic no-win scenarios. As usual, the winners here aren't those who had some kind of "grand master plan", or even expected the players to respond rationally according to their incentives and self-interest... but those who could respond quickly, effectively, and cooperatively to events where decision-makers acted in irrational and self-damaging ways, while also leaving opportunities open for "saving face" and not rubbing salt into the wounds of any perceived vulnerability.
I applaud Satya Nadella, Sam Altman, and Greg Brockman... this has been a master-class in damage control and applied game theory, in many ways... as well as Ilya Sutskever for admitting when he was wrong, taking accountability for his choices, and course-correcting. None of those things are easy or natural, and it speaks to the professionality of all involved that Altman and Brockman responded very positively to his contrition in the face of what must have felt like a massive betrayal by Sutskever.
I await any further developments just as everyone else is.
Christopher Mims recently voiced some suspicions I've had recently regarding smartphones ('Social Media Is Warping Into Old-Fashioned Mass Media,' Wall St. Journal, November 18-19). Is the nearly obsessive use of smartphones healthy.
I don't own a mobile phone, and never have.
It's not that I'm particularly against it, although it has made driving a lot more dangerous. I just don't see the point. I've never played a video game, either. Again, the point?
Is there anyone else out there who hasn't turned into a cyborg?
I haven't had a chance to read all the comments on the Girard post, but I thought Scott was overly harsh on Girard, especially the last two chapters on political correctness (this was Girard's term). Scott writes: "So Girard is stuck in an awkward position of saying that the rise of concern-for-victims was good when Christianity is doing it, is bad now, and not having any good theory of what changed, or how this relates to the more speculative anthropology." I take Girard to be arguing that what went wrong is that contemporary western culture took the concern for victims from Christianity but then threw away the rest of the moral framework in which it was embedded. That moral framework includes, for example, exhortations to love your enemies and forgive those who persecute you. Take away those things and you end up with a system that is ostensibly concerned with victims but uses that concern to justify the kinds of scapegoating and victimization it's supposed to be against. As for why this changed, this is just Satan reasserting himself within the moral system that threatens his power: using the concern for victims against itself. Of course this doesn't explain why political correctness arose exactly when and how it did, but I don't think Girard is trying to explain specific details of history like that.
The BlueCrossBlueShield carrier I use got hacked, and what was taken was not just account numbers but also passwords. BCBX paid for all those whose records were exposed to get 2 free years of Experian identity theft protection. So I signed up for that, but am not sure that was a good idea. Experian already knew a shitload of stuff about me. In order to prove I was who I said I was I had to give correct answers to a bunch of questions they asked me about my own finances, such as name of banks I have used in the past, and model of car I bought 5 years ago (how do they even know that? It was a cash transaction between me and previous owner.). And when I signed up for Experian's identity theft monitoring I gave the company a bunch more information about my finances, including numbers, expiration dates and security codes of all my credit cards, numbers of all my bank accounts. So now I'm thinking, so what if Experian gets hacked?
Has anyone found a good dark chocolate brand that is low on heavy metals? I eat a lot of dark chocolate, like 1/3 of a bar per day, and I'd hate to give it up.
I worry Altman’s sacking illustrates what I long feared: that the limited influence of AI safety enthusiasts on the world will be burned for negligible impact on AI safety.
Now is not the time.
LLMs reduce AI risk, in the same way calculators reduce AI risk: a person with a calculator is “superintelligent” compared to one without, so the calculators technology raises the bar of how intelligent AIs have to to surpass humanity.
(Of course, LLMs also increase AI risk, in several ways which were discussed to death here. But I expect no one to read these parenthesis! … also, it makes sense to me that exploiting LLMs for all they are worth will reduce AI risk according to the argument above more than it will increase AI risk, because in LLMs at least the initial training objective is reasonably orthogonal to paperclip maximization arguments.)
Hmm?
Has anyone here read Tom Holland's *Dominion*? I am starting it now, and definitely intrigued by some of the crossover between Holland's points and the recent "I See Satan Fall Like Lightning" review.
Please tell me if I'm right about this
OpenAI is a private company. The board has no obligation to tell the PUBLIC in advance or later, that they were going to fire the CEO and why. The board members might have had an obligation to tell the shareholders (such as Microsoft) about this.
So I'm wondering why they're being called "secretive" accusingly. They had no obligation to share this with the public, even in the vague terms they did.
It seems like the EA community has had at least three massive failures this year. Firstly Yudkowsky's comments about basically establishing a police state to prevent AI. (You can disagree with this characterization but this is the public sentiment of it. Even if you think he's right it's a PR failure.) Then the SBF debacle which I hope no one is still defending. And lastly this AI ouster. Now, you might think that last one is the right thing to attempt. But pragmatically it's not working. And I don't give you points for trying and failing.
In practical terms the philosophy seems on retreat on all fronts and with a severely tarnished brand. So my question is: are there feedback mechanisms? Does anyone get fired? Or does everyone continue like normal? If everyone does just carry on then I think this probably signals the end of EA as influential in large segments of technology. The AI ethicists with their left coding and elite backing have more sway with the government and the accelerationists will take over the actual companies.
I found this meta-analysis for using hypnosis for anxiety and was surprised at the significant positive results: https://underfund.dk/wp-content/uploads/2020/03/The-Efficacy-of-Hypnosis-as-a-Treatment-for-Anxiety-A-Meta-Analysis.pdf
I wonder if it’s not more mainstream because of the stigma, the lack of official recognition of it, or other reasons
I'd like to share with you the latest issue of my newsletter, Interessant3, available at https://interessant3.substack.com. In this issue, I share links on the following interesting topics:
1. The Chilean Economy: A thorough analysis of Chile's economic landscape, exploring government policies, trade, and internal dynamics.
2. Denmark's Electricity Dilemma: A look into Denmark's reliance on imported electricity despite its significant renewable energy sources, discussing sustainability and energy security challenges.
3. Yemen's Ancient Jewish Community: An exploration of the history and cultural heritage of one of the Middle East's oldest Jewish communities.
Feel free to explore these discussions and subscribe for more insights! Thank you for your interest and happy reading.
Ilya Sutskever apparently regrets removing Sam.
https://twitter.com/ilyasut/status/1726590052392956028
And Greg Brockman responds with a show of affection - possibly forgiveness, or even understanding?
https://twitter.com/gdb/status/1726598594948735256
As far as I can tell this puts a pretty big dent in the "Ilya pulled the plug after seeing an unexpected AI advance" theory. Though who knows - maybe Ilya realized only too late how important Sam was to the cohesion of the company? And now is scrambling to get him back for that alone?
Does Greg know Ilya's intentions? Are we ever going to get any official disclosure of what the board members were really thinking that Friday?
This whole fiasco is incredible (and a bit terrifying) to watch unfold live, especially for someone who's relatively new to AI safety debates. It's like I'm watching a thriller play out IRL.
EDIT: And Sam himself responded to Ilya's comment the same way.
https://twitter.com/sama/status/1726594398098780570
My head hurts.
I don’t want this thread to get too hostile but I feel I must bring this up somewhere, and this place is articulate enough to handle this tendentious subject without too much acrimony.
So trigger warning.
Why do people think that Die Hard is a Christmas movie, it was released in July.
What are some examples of good regulations?
Anyone have any success with free or paid online writing courses? Meaning helping you to launch a career in writing online (on the internet) as opposed to novels or something else. I am currently in a corporate job and really would like to start writing and explore a possible career change, but I am at a loss as to where to start or how to narrow down what to write about. There's a glut of courses and things online purporting to help people with this but I have no idea how to suss out what's worth the money and time. Thanks in advance.
How do I find a doctor that is amenable to write the prescriptions I want?
I'm not talking about Cocaine or Adderall.
I want things like getting diet pills without being mathematically obese, Metformin for potential life extension. There are more examples.
Are there plausible theories about why people have two kidneys?
I was just reading about the U.S.-dominated "Echelon" radio spying network. Are there any confirmed or likely gaps in its coverage, or does it intercept 100% of radio broadcasts originating anywhere on Earth's surface?
https://en.wikipedia.org/wiki/ECHELON
Does anyone know a good book about caretaker/professional carer (nurses, therapist, etc) burnout? Ideally I’m looking for something at the same level of technical depth and scope as The Body Keeps the Score. Thanks!
I know there are several people on the open comments who are themselves writers on substack. Question for you: Has your stack and your desire for readership started to shape what you write and how you write? I came to the decision recently that I would not let this happen to me, but is this just hubris? Is it possible not to let this happen?
https://falsechoices.substack.com/p/hold-fast-people
I've been curious for a while about traffic engineering (like how to decide which way a stop sign faces or how to time a light). I'd be very interested in a good textbook about the topic and a more approachable (sort of pop science if you will) nonfic about it. If anyone has any such recommendations, it would be greatly appreciated. Thank you!
I was thinking more about Scott's comment after running the impact certificates thing, where he felt that if the investors were making a profit, then the grant funders were sort of getting screwed.
Having given this some thought, I think he was sort of correct, but it's an easily fixable problem. Some quick economics first (if you have a micro background, skip the next paragraph).
Generally there is an idea of a supply and demand curve, where the idea is that buyers who really value a product will bid the price up and sellers will bid the price down until no one is willing to buy/sell any more goods at a different price. Generally whoever bought the last food was somewhat indifferent (got little value) since if they were still getting lots of value, they might have bid it up more. But, in general, MOST customers are paying less for the product than they might be willing to pay, and this gap between actual and maximum price generates value for them.
With the way that the impact grants market was structured, it was almost impossible for Scott to get value. He structured it as "I will decide how much I value your product at, and then pay you that amount" which means that he loses valuable money at exactly the rate that projects generate value.
My proposal would be pretty simple, which is to say that next time you run it, announce that you will value projects as normal, but then buy impact certificates at, say, 80% of their value. This still gives investors room to make a profit (if it's far cheaper to generate value), but leaves Scott having saved on value generated
An issue of human perception: do many audiophiles prefer vinyl due to a real difference in sound or is it just the ceremony of taking a record out of a work of art and putting it on a turntable?
Here is the twist: there IS a difference in the signal extracted from CDs. While you can theoretically put all the information that a human can hear on a CD, the playback method of CD players is not the same as the playback described by Nyquist's Sampling Theorem.
https://conntects.net/members/thatsDoctorMilsted/postPermalinks/103
On the vain of language teaching methods, a while ago I encountered a well documented Anki deck for the learning of verb conjugation, created by one Andy, an approach he since named "the Kofi method". I took advantage of the Spanish deck (an Italian one is also available and a French one is coming up), and I can attest to the effectiveness and efficiency of using such flash cards (I made superficial changes to them and followed a different order and pace than suggested, which nonetheless was easy to execute since the notes are well tagged). He wrote a post about the decks, including a section about how he generated them.
https://www.asiteaboutnothing.net/w_ultimate_spanish_conjugation.html
> Condolences to everyone in AI right now, I hope you’re all okay.
No joke, no sarcasm, Scott Alexander is my favourite comedian of 2023.
In case you still think I'm being sarcastic or whatever, other favourite comedians include Bo Burnham and Demetri Martin.
I have a question about Russian-Ukrainian relations during the Soviet era.
If I imagine war between say Britain and France in 2023 it's almost unimaginable, there' a long history of good relations (since at least 1945), we're ideologically similar, in the same geopolitical alliance, etc. France just doesn't seem like the kind of country we conceivably cold have a war with. Ditto for most Western countries.
In the Soviet era, I'd like to know if Russia and Ukraine had a similar relationship, being ideologically and geopolitically aligned, and sharing a similar language/culture. Of course there were nationalist movements in USSR, and a history of Russian domination,.
If you told a soviet citizen in the 80s that there would be war, would they be shocked or think something like "Yeah, they're the kind of country we might fight."
Has there always been a bit of mutual suspicion or is it possible for relations to deteriorate from a Britain-France level in 20-30 years, or even since 2014 when Ukraine began to move towards the Western camp?
It'd be especially cool to hear from anyone that was an adult in 80s Eastern Europe.
"Is My Toddler a Stochastic Parrot?"
https://www.newyorker.com/humor/sketchbook/is-my-toddler-a-stochastic-parrot
Free link: https://archive.li/pzUyb
Could a drive (simulation of a drive?) to make sense of the world be built into an AI?
Is "Closer to Fine" a good anthem for post-rationalists? For any particular sort of post-rationalist?
"II'm trying to tell you something about my life
Maybe give me insight between black and white
And the best thing you've ever done for me
Is to help me take my life less seriously
It's only life after all
Well, darkness has a hunger that's insatiable
And lightness has a call that's hard to hear
And I wrap my fear around me like a blanket
I sailed my ship of safety 'til I sank it
I'm crawling on your shores
And I went to the doctor, I went to the mountains
I looked to the children, I drank from the fountains
There's more than one answer to these questions pointing me in a crooked line
And the less I seek my source for some definitive
Closer I am to fine
Closer I am to fine
And I went to see the doctor of philosophy
With a poster of Rasputin and a beard down to his knee
He never did marry or see a B-grade movie
He graded my performance, he said he could see through me
I spent four years prostrate to the higher mind
Got my paper and I was free
I went to the doctor, I went to the mountains
I looked to the children, I drank from the fountains
There's more than one answer to these questions pointing me in crooked line
And the less I seek my source for some definitive
Closer I am to fine
Closer I am to fine
I stopped by the bar at 3 A.M
To seek solace in a bottle or possibly a friend
And I woke up with a headache like my head against a board
Twice as cloudy as I'd been the night before
And I went in seeking clarity
I went to the doctor, I went to the mountains
I looked to the children, I drank from the fountain
We go to the doctor, we go to the mountains
We look to the children, we drink from the fountain
We go to the Bible, we go through the workout
We read up on revival, we stand up for the lookout
There's more than one answer to these questions pointing me in a crooked line
And the less I seek my source for some definitive
Closer I am to fine
Closer I am to fine"
https://www.inquirer.com/health/kidney-donor-penn-philadelphia-marathon-runners-20231119.html
"Sunday’s Philadelphia Marathon was Baude’s 12th marathon in a year, including two Ironman triathlon competitions — the culmination of a feat of endurance aimed at drawing awareness to the national need for living kidney donors.
Baude donated her kidney to a man she’d never met before in 2021, after hearing of his plight — a five-year wait for a lifesaving organ transplant — on national news. She and her racing partner, Matt Cavanaugh, a fellow donor who heads the National Kidney Donor Organization, hope to show potential donors that they can continue a healthy, physically active life after donation."
I'm pleased to see they aren't making unproven promises about longevity, but I wonder how good a marker running marathons is.
Well, dog my cats, looks like I was right! Maybe I should give this prediction market thing a whirl 😁
https://www.ft.com/content/54e36c93-08e5-4a9e-bda6-af673c3e9bb5
"Microsoft has hired Sam Altman and Greg Brockman to lead a team conducting artificial intelligence research, days after the pair were pushed out of OpenAI, the company they co-founded.
Writing on X on Monday, Microsoft chief executive Satya Nadella said that Altman and Brockman, “together with colleagues, will be joining Microsoft to lead a new advanced AI research team”.
“We look forward to moving quickly to provide them with the resources needed for their success,” he added.
Altman retweeted Nadella’s post on Monday, adding: “The mission continues.”
Microsoft’s share price was up 1.7 per cent in pre-market trading on Monday, bouncing back from a similar fall on Friday after the announcement of Altman’s firing."
Right, let me get the mean comment out of the way first: and these are the people the AI doom brigade are hoping will save the world? They can't even sort out their own company (insert eyeroll here).
I'm not surprised by this; note the drop and then bounce back of the Microsoft share price. This is why I have no belief in calls for pauses or moratoria; write all the Open Letters you like, but when it comes down to "please join us to accelerate the tech for the money fountain" and even more "yikes our share price!!!!", then there will be no well-intentioned 'for the common good' halting of progress.
AI is coming and pretty much already here, remember Microsoft have just had a week of "how to integrate AI into your business; if you don't integrate AI into your business you are going to fall behind very badly; let us do the integrating for you". AGI may well be on the way, but I remain convinced that what we are going to get is more and more Smart Dumb Machines that will be increasingly used to generate output, which will (as we have seen) invent total nonsense rather than reply "no answer" and people won't check that output because (1) they're being sold on 'the machine is always right' and (2) businesses will be using AI for cost-cutting, so there won't be the time to check or the warm bodies to do any checking.
We'll have people docilely dosing themselves up with arsenic because the AI Doc In A Box told them to do so, as per a distracted Watson in "The Sign of the Four":
"He was clearly a confirmed hypochondriac, and I was dreamily conscious that he was pouring forth interminable trains of symptoms, and imploring information as to the composition and action of innumerable quack nostrums, some of which he bore about in a leather case in his pocket. I trust that he may not remember any of the answers which I gave him that night. Holmes declares that he overheard me caution him against the great danger of taking more than two drops of castor oil, while I recommended strychnine in large doses as a sedative."
I have personal testimony this is simulated environment. I’m so glad to find my tribe here. Thank the controllers
Let's continue re-reading the old Scott posts https://archive.ph/fCFQx, even in these weird days. There was a period of time when Scott was something of a travel blogger. In "Stuff" (https://archive.ph/FfJaD https://justpaste.it/9z9wk) he goes to Jerusalem, narrowly avoids being immolated by the holy fire and oversees the preparations for the coming of the Messiah.
Weirdly, I couldn't post the text on pastebin, because apparently: "Pastebin’s SMART filters have detected potentially offensive or questionable content in your Paste.
The content you are trying to publish has been deemed potentially offensive or questionable by our filters, because of this you’re receiving this warning.
This Paste can only be published with the visibility set to "Private"." Wtf?
Stray thought (that probably should have gone on the ketamine post - too late now): if all anaesthesia has an antidepressant effect, is there a chance that some of the effectiveness of modern ECT comes from the fact that we put people under for the procedure? Has anyone done a study trying to control for this somehow? (eg groups with anesthesia + ECT vs just anesthesia).
I'm currently a Software Engineer at Google, based in Poland. I want to move and am exploring my options. One option is an internal transfer within Google but I'm curious about other opportunities that might be out there.
Right now USA is my first choice but I'm open to considering other locations as well. Anyone knows how the situation looks right now / what companies are hiring and offering good conditions? (recruiters who reach out to me are only offering positions in EU)
who should i talk to for a podcast on the openAI situations for normies? ima gonna wait a week... i'm thinking Joscha Bach, but who else?
Javier Milei has won the presidential election in Argentina. While his plan to abandon the peso and use the US dollar as Argentina’s currency has gotten the most attention, he’s also in favour of legalising organ sales. So medical tourism to Argentina may become a big part of the answer to the kidney shortage!
This whole OpenAi fiasco is really worrying. If someone sees anything positives in this, I'd really want to know.
For me it looks like just negatives all the way down, in so many different ways that I don't even know where to start.
I suppose someone noticed this before, but it seems like if www is not included in the url, you end up on your test website:
http://astralcodexten.com
http://www.astralcodexten.com
> Condolences to everyone in AI right now, I hope you’re all okay.
Is that the Sam Altman thing. I don’t really follow the ups and downs of the average CEO. But that was… weird, right?
The board gets rid of the CEO and another founder based on a relatively flimsy reason, with no real specifics. But, fine. Boards can do that. In fact they should do it more, for underperforming CEOs.
Then it turns out that the main investors, including Microsoft, are flummoxed by this and were not informed. They want him back. Did the investors not get members on the board after they invested billions? Shouldn’t they dominate the board? Shouldn’t they be informed before the coup at the very least.
Edit. As a write I googled Altman and he’s hired by Microsoft. Who are now investors in OpenAI and doing their own thing with AI.
Just confusing.
Are there any ACX readers in Lille? (Or Paris or belgium tbh) I’ve just moved here and am looking for people to hang with
Did you see Sam Kriss's recent Girard post? https://harpers.org/archive/2023/11/overwhelming-and-collective-murder-rene-girard/
What did Sam Altman even do in his role at OpenAI?