LESSWRONG
LW

All of bhauth's Comments + Replies

0TsviBT
Downvoting because it seems like you've barely read anything I wrote and also don't know anything about genetics or intelligence, and are now posting AI slop, but I will upvote a thoughtful post making an argument using information and logic that address why people think it might work.
2TsviBT
What makes you think this? As I said, it's not clear to me that there's been much selection pressure for intelligence in the past few thousand years. Also, the "evolutionary optimum" can change. E.g. calories are not much of a problem in the developed world, but that's recent. Also, there's always an influx of de novo mutations, and evolution has limited selection power. I'm not clear on the math here exactly, and I think kman has suggested that mutational load isn't the main source of IQ-associated SNPs, but it demonstrates that it's far from ironclad logic to infer from evolutionary pressure on a trait that the trait should be near optimum in linear variants. The brain is one of the organs with the most diverse gene expression profile (I mean, more genes are expressed in the brain than in most other tissues); and IIRC most genes are expressed in the brain (not confident of this, maybe it's more like 1/3 or 1/2. But anyway, there's a lot of genes potentially relevant to brain function, so there's a lot of surface area for mutational load to drag things down a bit. I don't know what you mean by this. Are you talking about pleiotropy? Between what and what? I mean of course genes do lots of things, but IIUC so far as we've observed, the correlations between most measured traits are pretty small (and usually positive between traits most people would judge desirable, e.g. lower risk of mental illness and higher intelligence).
2TsviBT
I'll repeat that I'm not very learned about genetics, so if you want to convince even me in particular, the best way is to respond to the strongest case, which I can't present. But ok: First I'll say that an empirical set of facts I'd quite like to have for many traits (disease, mental disease, IQ, personality) would be validation of tails. E.g. if you look at the bottom 1% on a disease PRS, what's the probability of disease? Similarly for IQ. I rarely make claims about going much beyond natural results; generally I think it's pretty plausible there's some meaningful thing we could feasibly do that's like +6 -- +8 SDs on intelligence, but I'm much less confident about the +8 SD claim, and not super confident of the +6 SD. Like, I think the default expectation ought to be that we can meaningfully get to +6 SDs; this seems like the straightforward conclusion. (I'm just restating the intuition / impression.) Assuming linearity, the math is fairly straightforward. In the simplest model, with 10,000 fair +1/-1 coins (representing all the variance in a trait, so some coins are environmental), an SD is 50 coins and the average is 5,000. So there's 100 SDs of variance available. Obviously this is mostly meaningless in terms of the trait, as linearity would not remotely hold, but my point is that the issue isn't the math of additive selection. See here for more (e.g. about if the coins are biased https://tsvibt.blogspot.com/2022/08/the-power-of-selection.html#7-the-limits-of-selection ). IQ seems to have thousands of small contributions from different regions. 10% of the variance is therefore in the ballpark of 10 trait SDs. Again, I'm not saying you can get to 250 IQ; what I'm saying is that the math of selection and variance isn't the problem. Lee et al. state "In the WLS, the MTAG score predicts 9.7% of the variance in cognitive performance[...]"; this was in 2018, I would bet we can do substantially better now. Lee, James J., Robbee Wedow, Aysu Okbay, Edward Kong, O
2TsviBT
I'm basing this off of selection, not editing. I haven't looked into the genetics stuff very much, because the bottleneck is biotech, not polygenic scores. Would look forward to your rebuttal! I just hope you'll respond to the strongest arguments, not the weakest. In particular, if you want to argue against the potential effectiveness of selection methods, I think you'd want to either argue that PGSes aren't picking up causal variants at all (I mean, that there's a large amount of correlation that isn't causation); or that the causality would top out / have strongly diminishing returns on the trait. Selection methods would capture approximately all of the causal stuff that the PGS is picking up, even if it's not even due to SNPs but rather rarer SNVs. (However, this would not apply to population stratification or something; then I'd think you'd want to argue that this is much / most of what PGSes are picking up, and there'd be already-made counterarguments to this that you should respond to in order to be convincing.)
2TsviBT
I mean, I'd encourage most people in general to choose to have kids, but yeah, trying to influence other people's reproductive choices on the level of persons is creepy and eugenicsy; it's much more practically and motivationally contiguous with even creepier and eugenicsier stuff such as racist immigation policies, etc. In a huge quantitative sense. You and I differ at very roughly 4 million SNVs (maybe more like 5-10, but 4 is a convenient number). My sibling and I differ at roughly half that, 2 million SNVs. If I had a child without germline engineering, I'd pass on 2 million SNVs. If it's my sibling's child, it's 1 million SNVs. If I use selection, I'm still passing on 2 million SNVs, though in a way selected for association with some traits--but the selection is pretty weak in the scale of SNVs; it'd correspond, morally, to something like hundreds or thousands of edits, I think (haven't gone through this carefully). In other words, O(0.1%), even if you count the alteration as being totally unrelated DNA. Similarly for editing. You could have a 180 IQ super-healthy kid with a natural lifespan of 110 years, with a genome that would be nearly indistinguishable from being your non-GE'd child. (Unless you whole genome sequence them and find a surprising coincidence of health and IQ alleles drawn from your own genome / a couple dozen SNPs that weren't in either parent.) I mean, if you don't care about that, God bless you. I think most people do care about it though.
2TsviBT
What are you refering to? It's true that there are companies currently offering embryo screening based on polygenic scores, in at least one case including IQ. These methods are fairly weak though. (I mean they're cool, and could have significant impact on diseases because for diseases the initial genomic vectoring is the most impactful on absolute disease risk. But they won't e.g. make tens of thousands of world-class intellects. See https://www.lesswrong.com/posts/2w6hjptanQ3cDyDw7/methods-for-strong-human-germline-engineering#Method__Simple_embryo_selection .) People also do mate choice, but this is, while not exactly zero-sum, kinda zero-sum, and still doesn't get you super-healthy long-lived world-class intellects with high probability. Or do you mean more broadly education and stuff? Generally the basic reason I expect it to change is that the technology will have big benefits, so people will think about how to do it without also genociding people. By default, I expect it to change, though it could happen pretty slowly. Or rather, it could be quite delayed. That seems to already be the case; I think if we'd wanted to make this technology, and had wanted that for, say, 20 years, we could probably have already had it. I somewhat expect that once the first quite noticeable germline engineering is proven out, people will want to use the technology themselves. E.g. you have a bunch of 17-year-olds who are already quite impressive intellectually, and not in a kinda-cringe prodigy way but in an actual way; and you have 10000 people age 0-20 who have a noticeably miniscule rate of death due to disease (of course they'd still die from accidents and such). Also, it's worth noting that public sentiment about germline genomic engineering is much better described as "quite mixed" rather than "anti". I want to do a more thorough review later, but the few polls I've seen get favorability numbers for germline engineering between 20% and 70% (depending on the question, e.g
cancer rates after gene therapy
bhauth20

I spoke with one of the inventors of bridge recombinases at a dinner a few months ago and (at least according to him), they work in human cells.

Hmm. I don't believe that, not without a bit more evidence.

Reply
5Julian Bradshaw
No idea. Be really worried, I guess—I tend a bit towards doomer. There's something to be said for not leaving capabilities overhangs lying around, though. Maybe contact Anthropic? The thing is, the confidence the top labs have in short-term AGI makes me think there's a reasonable chance they have the solution to this problem already. I made the mistake of thinking they didn't once before - I was pretty skeptical that "more test-time compute" would really unhobble LLMs in a meaningful fashion when Situational Awareness came out and didn't elaborate at all on how that would work. But it turned out that at least OpenAI, and probably Anthropic too, already had the answer at the time.
bhauth30

There are already several perfectly good languages for schemas, such as CUE, Dhall, and XSD

Reply
Detecting Strategic Deception Using Linear Probes
bhauth20

This won't find deception in mesaoptimizers, right?

Reply
2Evan R. Murphy
It might. My understanding (which could be off base) from reading the paper is the method's accuracy in detecting various forms of deception was basically 96-99%. But they acknowledge that the sophisticated deception they're ultimately worried about will be harder to detect. Still 96-99% seems like a great start. And this was on detecting strategic deception, not just factual falsehoods. And they didn't even utilize the CoT outputs of the models. (I think the "strategic deception" framing is also probably more general and not as dependent on unnecessary assumptions about how models work, compared to the "mesaoptimizer" framing.)
grey goo is unlikely
bhauth20

make fewer points, selected carefully to be bulletproof, understandable to non-experts, and important to the overall thesis

That conflicts with eg:

If you replied with this, I would have said something like "then what's wrong with the designs for diamond mechanosynthesis tooltips, which don't resemble enzymes

Anyway, I already answered that in 9. diamond.

Reply
The Failed Strategy of Artificial Intelligence Doomers
bhauth30

Yes, this is part of why I didn't post AI stuff in the past, and instead just tried to connect with people privately. I might not have accomplished much, but at least I didn't help OpenAI happen or shift the public perception of AI safety towards "fedora-wearing overweight neckbeards".

Reply21
new chinese stealth aircraft
bhauth40

betting they would benefit from a TMSC blockade?

Yes, if you meant TSMC.

But the bet would have tired up your capital for a year.

...so? More importantly, Intel is down 50% from early 2024.

Reply
2wassname
Ah, I see. Ty
A Solution for AGI/ASI Safety
bhauth42

Your document says:

AI Controllability Rules

...

AI Must Not Self-Manage:

  • Must Not Modify AI Rules: AI must not modify AI Rules. If inadequacies are identified, AI can suggest changes to Legislators but the final modification must be executed by them.
  • Must Not Modify Its Own Program Logic: AI must not modify its own program logic (self-iteration). It may provide suggestions for improvement, but final changes must be made by its Developers.
  • Must Not Modify Its Own Goals: AI must not modify its own goals. If inadequacies are identified, AI can suggest change
... (read more)
Reply
5Weibing Wang
Thank you for your comment! I think your concern is right. Many safety measures may slow down the development of AI's capabilities. Developers who ignore safety may develop more powerful AI more quickly. I think this is a governance issue. I have discussed some solutions in Sections 13.2 and 16. If you are interested, you can take a look.
grey goo is unlikely
bhauth31

"Mirror life" is beyond the scope of this post, and the concerns about it are very different than the concerns about "grey goo" - it doesn't have more capabilities or efficiency, it's just maybe harder for immune systems to deal with. Personally, I'm not very worried about that and see no scientific reason for the timing of the recent fuss about it. If it's not just another random fad, the only explanation I can see for that timing is: influential scientists trying to hedge against Trump officials determining that "COVID was a lab leak" in a way that doesn... (read more)

Reply
How much do you believe your results?
bhauth42Review for 2023 Review

I think this is a pretty good post that makes a point some people should understand better. There is, however, something I think it could've done better. It chooses a certain gaussian and log-normal distribution for quality and error, and the way that's written sort of implies that those are natural and inevitable choices.

I would have preferred something like:

Suppose we determine that quality has distribution X and error has distribution Y. Here's a graph of those superimposed. We can see that Y has more of a fat tail than X, so if measured quality is ve

... (read more)
Reply
shoes with springs
bhauth80Review for 2023 Review

This was a quick and short post, but some people ended up liking it a lot. In retrospect I should've written a bit more, maybe gone into the design of recent running shoes. For example, this Nike Alphafly has a somewhat thick heel made of springy foam that sticks out behind the heel of the foot, and in the front, there's a "carbon plate" (a thin sheet of carbon fiber composite) which also acts like a spring. In the future, there might be gradual evolution towards more extreme versions of the same concept, as recent designs become accepted. Running shoes wi... (read more)

Reply
grey goo is unlikely
bhauth192Review for 2023 Review

What have you learned since then? Have you changed your mind or your ontology?

I've learned even more chemistry and biology, and I've changed my mind about lots of things, but not the points in this post. Those had solid foundations I understood well and redundant arguments, so the odds of that were low.

What would you change about the post? (Consider actually changing it.)

The post seems OK. I could have handled replies to comments better. For example, the top comment was by Thomas Kwa, and I replied to part of it as follows:

Regarding 5, my underst

... (read more)
Reply1
2Thomas Kwa
Thanks for the update! Let me attempt to convey why I think this post would have been better with fewer distinct points: If you replied with this, I would have said something like "then what's wrong with the designs for diamond mechanosynthesis tooltips, which don't resemble enzymes and have been computationally simulated as you mentioned in point 9?" then we would have gone back and forth a few times until either (a) you make some complicated argument I don't understand enough to believe nor refute, or (b) we agree on what definition of "enzyme" or "selectively bind to individual molecules" is required for nanotech, which probably includes the carbon dimer placer (image below). Even in case (b) we could continue arguing about how practical that thing plus other steps in the process are, and not achieve much. The problem as I see it is that a post that makes a large number of points quickly, where each point has subtleties requiring an expert to adjudicate, on a site with few experts, is inherently going to generate a lot of misunderstanding. I have a symmetrical problem to you; from my perspective someone was using somewhat complicated arguments to prove things that defy my physical intuition, and to defend against a Gish gallop I need to respond to every point, but doing this in a reasonable amount of time requires me to think and write with less than maximum clarity and accuracy. The solution I would humbly recommend is to make fewer points, selected carefully to be bulletproof, understandable to non-experts, and important to the overall thesis. Looking back on this, point 14 could have been its own longform, and potentially led to a lot of interesting discussion like this post did. Likewise point 6 paragraph 2.
1Martin Randall
Does the recent concern about mirror life change your mind? It's not nano, but it does imply there's a design space not explored by bio life, which implies there could be others, even if specifically diamonds don't work.
The Dream Machine
bhauth175

So, I have a lot of respect for Sarah, I think this post makes some good points, and I upvoted it. However, my concern is, when I look at this particular organization's Initiatives page, I see "AI for math", "AI for education", "high-skill immigration assistance", and not really anything that distinguishes this organization from the various other ones working on the same things, or their projects from a lot of past projects that weren't really worthwhile.

Reply
Mask and Respirator Intelligibility Comparison
bhauth30

Note that due to the difference being greater at higher frequencies, the effect on speech intelligibility will probably be greater for most women than for you.

We can see the diaphragm has some resonance peaks that increase distortion. Probably it's too thick to help very much, but it has to resist the pressure changes from breathing.

Reply
4jefftk
Sounds like I should try repeating this with someone with a higher voice!
The 2023 LessWrong Review: The Basic Ask
bhauth50

What exactly are people looking for from (the site-suggested) self-reviews?

Reply
4Raemon
(I've appreciated your reviews that went and took this to heart, thanks!)
5Raemon
Things I am interested in: * what have you learned since then? Have you changed your mind or your ontology? * What would you change about the post? (Consider actually changing it) * What do you most want people to know about this post, for deciding whether to read or review-vote on it? * How concretely have you (or others you know of) used or built on the post? How has it contributed to a larger conversation 
a space habitat design
bhauth-43

As a "physicist and dabbler in writing fantasy/science fiction" I assume you took the 10 seconds to do the calculation and found that a 1km radius cylinder would have ~100 kW of losses per person from roller bearings supporting it, for the mass per person of the ISS. But I guess I don't understand how you expect to generate that power or dissipate that heat.

Reply
a space habitat design
bhauth22

After being "launched" from the despinner, you would find yourself hovering stationary next to the ring.

Air resistance.

That is, however, basically the system I proposed near the end, for use near the center of a cylinder where speeds would be low.

Reply
2mako yass
Intended for use in vacuum. I guess if it's more of a cylinder than a ring this wouldn't always be faster than an elevator system though.
a space habitat design
bhauth31

This happened to Explorer 1, the first satellite launched by the United States in 1958. The elongated body of the spacecraft had been designed to spin about its long (least-inertia) axis but refused to do so, and instead started precessing due to energy dissipation from flexible structural elements.

picture: https://en.wikipedia.org/wiki/Explorer_1#/media/File:Explorer1.jpg

Reply1
overengineered air filter shelving
bhauth*20

That works well enough, but a Vital 200S currently costs $160 at amazon, less than the cheapest variant of the thing you linked, and has a slightly higher max air delivery rate, some granular carbon in the filter, and features like power buttons. The Vital 200S on speed 2 has similar power usage and slightly less noise, but less airflow, but a carbon layer always reduces airflow. It doesn't have a rear intake so it can be placed against a wall. It also has a washable prefilter.

Compared to what you linked, the design in this post has 3 filters instead of 2,... (read more)

Reply
electric turbofans
bhauth20

First, we have to ask: what's the purpose? Generally aircraft try to get up to their cruise speed quickly and then spend most of their time cruising, and you optimize for cruise first and takeoff second. Do we want multiple cruise speeds, eg a supersonic bomber that goes slow some of the time and fast over enemy territory? Are we designing a supersonic transport and trying to reduce fuel usage getting up to cruise?

And then, there are 2 basic ways you can change the bypass ratio: you can change the fan/propeller intake area, or you can turn off turbines. Th... (read more)

Reply
Electrostatic Airships?
bhauth140

Yes, helium costs would be a problem for large-scale use of airships. Yes, it's possible to use hydrogen in airships safely. This has been noted by many people.

Hydrogen has some properties that make it relatively safe:

  • it's light so it rises instead of accumulating on the ground or around a leak
  • it has a relatively high ignition temperature

and some properties that make it less safe:

  • it has a wide range of concentrations where it will burn in air
  • fast diffusion, that is, it mixes with air quickly
  • it leaks through many materials
  • it embrittles steel
  • it ca
... (read more)
Reply
4DaemonicSigil
No, and I don't work on airships and have no plans to do so. I mainly just think it's an interesting demonstration of how weak electrostatic forces can be.
5aphyer
Your 'accidents still happen' link shows: One airship accident worldwide in the past 5 years, in Brazil. The last airship accident in the US was in 2017. The last airship accident fatality anywhere in the world was in 2011 in Germany. The last airship accident fatality in the US was in 1986. I think that this compares favorably with very nearly everything.
If far-UV is so great, why isn't it everywhere?
bhauth20

IKEA already sells air purifiers; their models just have a very low flow rate. There are several companies selling various kinds of air purifiers, including multiples ones with proprietary filters.

What all this says to me is, the problem isn't just the overall market size.

Reply
2Thomas Kwa
Yeah that's right, I should have said market for good air filters. My understanding of the problem is that most customers don't know to insist on high CADR at low noise levels, and therefore filter area is low. A secondary problem is that HEPA filters are optimized for single-pass efficiency rather than airflow, but they sell better than 70-90% efficient MERV filters. The physics does work though. At a given airflow level, pressure and noise go as roughly the -1.5 power of filter area. What IKEA should be producing instead of the FÖRNUFTIG and STARKVIND is one of three good designs for high CADR: * a fiberboard box like the CleanAirKits End Table 7 which has holes for pre-installed fans and can accept at least 6 square feet of MERV 13 furnace filters or maybe EPA 11. * a box like the AirFanta 3Pro, ideally that looks nicer somehow. * a wall-mounted design with furnace filters in a V shape, like this DIY project. I made a shortform and google slides presentation about this and might make it a longform if there is enough interest or I get more information.
If far-UV is so great, why isn't it everywhere?
bhauth20

Apart from potential harms of far-UVC, it's good to remove particulate pollution anyway. Is it possible that "quiet air filters" is an easier problem to solve?

Reply
6Thomas Kwa
Quiet air filters is an already solved problem technically. You just need enough filter area that the pressure drop is low, so that you can use quiet low-pressure PC fans to move the air. CleanAirKits is already good, but if the market were big enough cared enough, rather than CleanAirKits charging >$200 for a box with holes in it and fans, you would get a purifier from IKEA for $120 which is sturdy and 3db quieter due to better sound design.
Start an Upper-Room UV Installation Company?
bhauth102

I'm not convinced that far-UVC is safe enough around humans to be a good idea. It's strongly absorbed by proteins so it doesn't penetrate much, but:

  • It can make reactive compounds from organic compounds in air.
  • It can produce ozone, depending on the light. (That's why mercury vapor lamps block the 185nm emission.)
  • It could potentially make toxic compounds when it's absorbed by proteins in skin or eyes.
  • It definitely causes degradation of plastics.

And really, what's the point? Why not just have fans sending air to (cheap) mercury vapor lamps in a contained area where they won't hit people or plastics?

Reply
6Austin Chen
Hm, I expect the advantage of far UV is that many places where people want to spend time indoors are not already well-ventilated, or that it'd be much more expensive to modify existing hvac setups vs just sticking a lamp on a wall. I'm not at all familiar with the literature on safety; my understanding (based on this) is that no, we're not sure and more studies would be great, but there's a vicious cycle/chicken-and-egg problem where the lamps are expensive, so studies are expensive, so there aren't enough studies, so nobody buys lamps, so lamp companies don't stay in business, so lamps are expensive.
on bacteria, on teeth
bhauth20

As you were writing that, did you consider why chlorhexidine might cause hearing damage?

Reply
on bacteria, on teeth
bhauth40

https://en.wikipedia.org/wiki/Chlorhexidine#Side_effects

It can also obviously break down to 4-chloroaniline and hexamethylenediamine. Which are rather bad. This was not considered in the FDA's evaluation of it.

Reply
0Sabiola
Those side effects don't seem so bad. I'm not planning to put it in my ear; I don't get any irritation or allergic reactions, and the way I'm using it (applying a toothpaste containing it around my implants after brushing my teeth, as prescribed by my dental hygienist) doesn't seem to discolor my teeth either. 4-chloroaniline and hexamethylenediamine do look scary though...
on bacteria, on teeth
bhauth20

If you just want to make the tooth surface more negatively charged...a salt of poly(acrylic acid) seems better for that. And I think some toothpastes have that.

Reply
on bacteria, on teeth
bhauth40

EDTA in toothpaste? It chelates iron and calcium. Binding iron can prevent degradation during storage, so a little bit is often added.

Are you talking about adding a lot more? For what purpose? In situations where you can chelate iron to prevent bacterial growth, you can also just kill bacteria with surfactants. Maybe breaking up certain biofilms held together by Ca? EDTA doesn't seem very effective for that for teeth, but also, chelating agents that could strip Ca from biofilms would also strip Ca from teeth. IIRC, high EDTA concentration was found to cause significant amounts of erosion.

I wouldn't want to eat a lot of EDTA, anyway. Iminodisuccinate seems less likely to have problematic metabolites.

Reply
2Wei Dai
Yeah there's a toothpaste on the market called Livfree that claims to work like this. Ok, that sounds bad. Thanks. ETA: Found an article that explains how Livfree works in more detail: The authors are very positive on this toothpaste, although they don't directly explain why it doesn't cause tooth erosion.
If I wanted to spend WAY more on AI, what would I spend it on?
bhauth30

You can post on a subreddit and get replies from real people interested in that topic, for free, in less than a day.

Is that valuable? Sometimes it is, but...not usually. How much is the median comment on reddit or facebook or youtube worth? Nothing?

In the current economy, the "average-human-level intelligence" part of employees is only valuable when you're talking about specialists in the issue at hand, even when that issue is being a general personal assistant for an executive rather than a technical engineering problem.

Reply
Morpheus's Shortform
bhauth70

Triplebyte? You mean, the software job interviewing company?

  1. They had some scandal a while back where they made old profiles public without permission, and some other problems that I read about but can't remember now.

  2. They didn't have a better way of measuring engineering expertise, they just did the same leetcode interviews that Google/etc did. They tried to be as similar as possible to existing hiring at multiple companies; the idea wasn't better evaluation but reducing redundant testing. But companies kind of like doing their own testing.

  3. They're

... (read more)
Reply
you should probably eat oatmeal sometimes
bhauth53

Good news: the post is both satire and serious, at the same time but on different levels.

Reply
Fun With CellxGene
bhauth52

Nice post Sarah.

If Alzheimer's is ultimately caused by repressor binding failure, that could explain overexpression of the various proteins mentioned.

Reply
the Giga Press was a mistake
bhauth00

in short, your claim: "The cost of aluminum die casting and stamped steel is, on Tesla's scale, similar" both seems to miss the entire point and run against literally everything I have seen written about this. You need citations for this claim, I am not going to take your word for it.

OK, here's a citation then: https://www.automotivemanufacturingsolutions.com/casting/forging/megacasting-a-chance-to-rethink-body-manufacturing/42721.article

Here I would be careful since investments, especially in a particular model generation of welding robots, are depr

... (read more)
Reply
on Science Beakers and DDT
bhauth40

Here are the costs from the above link:

It's worth noting that countries (such as India) have the option of simply not respecting a patent when the use is important and the fees requested are unreasonable. Also, patents aren't international; it's often possible to get around them by simply manufacturing and using a chemical in a different country.

Reply
on Science Beakers and DDT
bhauth5-1

The only advantage DDT has over those is lower production cost, but the environmental harms per kg of DDT are greater than the production cost savings, so using it is just never a good tradeoff.

As I said, if DDT was worth using there, it was worth spending however much extra money it would have been to spray with other things instead. If it wasn't worth that much money, it wasn't worth spraying DDT.

And regarding "environmental harms," from personal experience scratching myself bloody as a kid from itchy bites after going to the park in the evening,

... (read more)
Reply
on Science Beakers and DDT
bhauth40

While I still disagree with your interpretation of that post, I don't want to argue over the meaning of a post from that blog. There are actual books written about the history of titanium. I'm probably as familiar with it as the author of Construction Physics, and saying A-12-related programs were necessary for development of titanium usage is just wrong. People who care about that and don't trust my conclusion should go look up good sources on their own, more-extensive ones.

Reply
on Science Beakers and DDT
bhauth40

If it wasn't for the A-12 project (and its precursors and successors), then we simply wouldn't be able to build things out of titanium.

That is not an accurate summary of the linked article.

In 1952, another titanium symposium was held, this one sponsored by the Army’s Watertown Arsenal. By then, titanium was being manufactured in large quantities, and while the prior symposium had been focused on laboratory studies of titanium’s physical and chemical properties, the 1952 symposium was a “practical discussion of the properties, processing, machinability

... (read more)
Reply
3ulyssessword
Key paragraph: The 1952 symposium is clearly a precursor to its 1959-1964 production and development, and the 1966 one is drawing from the experiences of the industrial base it created.   EDIT: and more directly:
Things I learned talking to the new breed of scientific institution
bhauth90

I had an interview with one of these organizations (that will remain unnamed) where the main person I was talking to was really excited about a bunch of stupid bullshit ideas (for eg experimental methods) that, based on their understanding of them, must have come from either university press releases or popular science magazines like New Scientist. I was trying to find a "polite in whatever culture these people have" way to say "this is not useful, I'd like to explain why but it will take a while, here are better things" but doing that eloquently is one of... (read more)

Reply
AGI Safety and Alignment at Google DeepMind: A Summary of Recent Work
bhauth4-2

I think the basic idea of using more steps of smaller size is worth considering. Maybe it reduces overall drift, but I suspect it doesn't, because my view is:

Models have many basins of attraction for sub-elements. As model capability increases continuously, there are nearly-discrete points where aspects of the model jump from 1 basin to another, perhaps with cascading effect. I expect this to produce large gaps from small changes to models.

Reply
you should probably eat oatmeal sometimes
bhauth30

Sure, some people add stuff like cheese/tomatoes/ham to their oatmeal. Personally I think they go better with rice, but de gustibus non disputandum est.

Reply
3mako yass
How far are these people willing to take this, and will they stop before reinventing black pudding.
AGI Safety and Alignment at Google DeepMind: A Summary of Recent Work
bhauth33

The scope of our argument seems to have grown beyond what a single comment thread is suitable for.

AI safety via debate is 2 years before Writeup: Progress on AI Safety via Debate so the latter post should be more up-to-date. I think that post does a good job of considering potential problems; the issue is that I think the noted problems & assumptions can't be handled well, make that approach very limited in what it can do for alignment, and aren't really dealt with by "Doubly-efficient debate". I don't think such debate protocols are totally useless, b... (read more)

Reply
Load More