Engineer at CoinList.co. Donor to LW 2.0.
ESRogs
Karma: 5,129
I wouldn’t call this “AI lab watch.” “Lab” has the connotation that these are small projects instead of multibillion dollar corporate behemoths.
Disagree on “lab”. I think it’s the standard and most natural term now. As evidence, see your own usage a few sentences later:
They’ve all committed to this in the WH voluntary commitments and I think the labs are doing things on this front.
Yeah I figured Scott Sumner must have been involved.
Nitpick: Larry Summers not Larry Sumners
If “—quine” was passed, read the script’s own source code using the
__file__variable and print it out.
Interesting that it included this in the plan, but not in the actual implementation.
(Would have been kind of cheating to do it that way anyway.)
Worth noting 11 months later that @Bernhard was more right than I expected. Tesla did in fact cut prices a bunch (eating into gross margins), and yet didn’t manage to hit 50% growth this year. (The year isn’t over yet, but I think we can go ahead and call it.)
Good summary in this tweet from Gary Black:
$TSLA bulls should reduce their expectations that $TSLA volumes can grow at +50% per year. I am at +37% vol growth in 2023 and +37% growth in 2024. WS is at +37% in 2023 and +22% in 2024.
And apparently @MartinViecha head of $TSLA IR recently advised investors that TSLA “is now in an intermediate low-growth period,” at a recent Deutsche Bank auto conference with institutional investors. 35-40% volume growth still translates to 35-40% EPS growth, which justifies a 60x-70x 2024 P/E ($240-$280 PT) at a normal megacap growth 2024 PEG of 1.7x.And this reply from Martin Viecha:
What I said specifically is that we’re between two major growth waves: the first driven by 3/Y platform since 2017 and the next one that will be driven by the next gen vehicle.
let’s build larger language models to tackle problems, test methods, and understand phenomenon that will emerge as we get closer to AGI
Nitpick: you want “phenomena” (plural) here rather than “phenomenon” (singular).
I’m not necessarily putting a lot of stock in my specific explanations but it would be a pretty big surprise to learn that it turns out they’re really the same.
Does it seem to you that the kinds of people who are good at science vs good at philosophy (or the kinds of reasoning processes they use) are especially different?
In your own case, it seems to me like you’re someone who’s good at philosophy, but you’re also good at more “mundane” technical tasks like programming and cryptography. Do you think this is a coincidence?
I would guess that there’s a common factor of intelligence + being a careful thinker. Would you guess that we can mechanize the intelligence part but not the careful thinking part?
Happiness has been shown to increase with income up to a certain threshold ($ 200K per year now, roughly speaking), beyond which the effect tends to plateau.
Do you have a citation for this? My understanding is that it’s a logarithmic relationship — there’s no threshold. (See the Income & Happiness section here.)
Why antisocial? I think it’s great!
I would imagine one of the major factors explaining Tesla’s absence is that people are most worried about LLMs at the moment, and Tesla is not a leader in LLMs.
(I agree that people often seem to overlook Tesla as a leader in AI in general.)
I don’t know anything about the ‘evaluation platform developed by Scale AI—at the AI Village at DEFCON 31’.
Looks like it’s this.
Here are some predictions—mostly just based on my intuitions, but informed by the framework above. I predict with >50% credence that by the end of 2025 neural nets will:
To clarify, I think you mean that you predict each of these individually with >50% credence, not that you predict all of them jointly with >50% credence. Is that correct?
I’d like to see open-sourced evaluation and safety tools. Seems like a good thing to push on.
My model here is something like “even small differences in the rate at which systems are compounding power and/or intelligence lead to gigantic differences in absolute power and/or intelligence, given that the world is moving so fast.”
Or maybe another way to say it: the speed at which a given system can compound it’s abilities is very fast, relative to the rate at which innovations diffuse through the economy, for other groups and other AIs to take advantage of.
I’m a bit skeptical of this. While I agree that small differences in growth rates can be very meaningful, I think it’s quite difficult to maintain a growth rate faster than the rest of the world for an extended period of time.
Growth and Trade
The reason is that: growth is way easier if you engage in trade. And assuming that gains from trade are shared evenly, the rest of the world profits just as much (in absolute terms) as you do from any trade. So you can only grow significantly faster than the rest of the world while you’re small relative to the size of the whole world.
To give a couple of illustrative examples:
The “Asian Tigers” saw their economies grow faster than GWP during the second half of the 20th century because they were engaged in “catch-up” growth. Once their GDP per capita got into the same ballpark as other developed countries, they slowed down to a similar growth rate to those countries.
Tesla has grown revenue at an average of 50% per year for 10 years. That’s been possible because they started out as a super small fraction of all car sales, and there were many orders of magnitude of growth available. I expect them to continue growing at something close to that rate for another 5-10 years, but then they’ll slow down because the global car market is only so big.
Growth without Trade
Now imagine that you’re a developing nation, or a nascent car company, and you want to try to grow your economy, or the number of cars you make, but you’re not allowed to trade with anyone else.
For a nation it sounds possible, but you’re playing on super hard mode. For a car company it sounds impossible.
Hypotheses
This suggests to me the following hypotheses:
Any entity that tries to grow without engaging in trade is going to be outcompeted by those that do trade, but
Entities that grow via trade will have their absolute growth capped at the size of the absolute growth of the rest of the world, and thus their growth rate will max out at the same rate as the rest of the world, once they’re an appreciable fraction of the global economy.
I don’t think these hypotheses are necessarily true in every case, but it seems like they would tend to be true. So to me that makes a scenario where explosive growth enables an entity to pull away from the rest of the world seem a bit less likely.
Ah, good point!
I suppose a possible mistake in this analysis is that I’m treating Moore’s law as the limit on compute growth rates, and this may not hold once we have stronger AIs helping to design and fabricate chips.
Even so, I think there’s something to be said for trying to slowly close the compute overhang gap over time.
0.2 OOMs/year was the pre-AlexNet growth rate in ML systems.
I think you’d want to set the limit to something slightly faster than Moore’s law. Otherwise you have a constant large compute overhang.
Ultimately, we’re going to be limited by Moore’s law (or its successor) growth rates eventually anyway. We’re on a kind of z-curve right now, where we’re transitioning from ML compute being some small constant fraction of all compute to some much larger constant fraction of all compute. Before the transition it grows at the same speed as compute in general. After the transition it also grows at the same speed as compute in general. In the middle it grows faster as we rush to spend a much larger share of GWP on it.
From that perspective, Moore’s law growth is the minimum growth rate you might have (unless annual spend on ML shrinks). And the question is just whether you transition from the small constant fraction of all compute to the large constant fraction of all compute slowly or quickly.
Trying to not do the transition at all (i.e. trying to growing at exactly the same rate as compute in general) seems potentially risky, because the resulting constant compute overhang means it’s relatively easy for someone somewhere to rush ahead locally and build something much better than SOTA.
If on the other hand, you say full steam ahead and don’t try to slow the transition at all, then on the plus side the compute overhang goes away, but on the minus side, you might rush into dangerous and destabilizing capabilities.
Perhaps a middle path makes sense, where you slow the growth rate down from current levels, but also slowly close the compute overhang gap over time.
See my edit to my comment above. Sounds like GPT-3 was actually 250x more compute than GPT-2. And Claude / GPT-4 are about 50x more compute than that? (Though unclear to me how much insight the Anthropic folks had into GPT-4′s training before the announcement. So possible the 50x number is accurate for Claude and not for GPT-4.)
Doesn’t this part of the comment answer your question?
We can very easily “grab probability mass” in relatively optimistic worlds. From our perspective of assigning non-trivial probability mass to the optimistic worlds, there’s enormous opportunity to do work that, say, one might think moves us from a 20% chance of things going well to a 30% chance of things going well. This makes it the most efficient option on the present margin.
It sounds like they think it’s easier to make progress on research that will help in scenarios where alignment ends up being not that hard. And so they’re focusing there because it seems to be highest EV.
Seems reasonable to me. (Though noting that the full EV analysis would have to take into account how neglected different kinds of research are, and many other factors as well.)
Have you thought much about the feasibility or desirability of training an ML model to do deductive estimation?
You wouldn’t get perfect conformity to your three criteria of linearity, respect for proofs, and 0-1 boundedness (which, as you say, is apparently impossible anyway), but you could use those to inform your computation of the loss in training. In which case, it seems like you could probably approximately satisfy those properties most of the time.
Then of course you’d have to worry about whether your deductive estimation model itself is deceiving you, but it seems like at least you’ve reduced the problem a bit.