by Trevor Chow, Basil Halperin, and J. Zachary Mazlish In this post, we point out that short AI timelines would cause real interest rates to be high, and would do so under expectations of either unaligned or aligned AI. However, 30- to 50-year real interest rates are low. We argue...
by Trevor Chow, Basil Halperin, and J. Zachary Mazlish [Note: This is an appendix to "AGI and the EMH: markets are not expecting aligned or unaligned AI in the next 30 years"] One naive objection would be the claim that real interest rates sound like an odd, arbitrary asset price...
(Cross-posted from my blog.) Confidence level: medium I want to argue that Newcomb’s problem does not reveal any deep flaw in standard decision theory. There is no need to develop new decision theories to understand the problem. I’ll explain Newcomb’s problem and expand on these points below, but here’s the...
You might like this blog post on 'the inefficiency of perfect price discrimination'. (Everyone, including me, hates links to LLM shared queries ...but here is chatgpt fleshing out the math.)
Forgive if a naive question, but what about distributed training runs — any view on whether progress on that front will result in training runs larger than what you describe here?
PS thank you for the extremely useful analysis as always — I would 100% subscribe if you had a newsletter or something (with a strictly positive willingness-to-pay even!).
Very nice. I recently did a similar exercise, and -- because as you note the Epoch data (understandably) doesn't have estimates of training compute for reasoning models -- I had o3 guesstimate "effective training compute" by OpenAI model (caveat: this doesn't really make sense!). You can see the FLOP by model in the link. And:
By this metric, it's ~3.5 more OOMs from o3 to 1-month-AGI. If -- as was often said to be the case pre-reasoning models -- effective compute can still be said to be growing at ~10x a year, then 1-month-AGI arrives around early 2029
Extremely cool.
As you say, looking at treatment effect heterogeneity is hard given power. But -- if I understand correctly, you did not take any melatonin between nights in which you randomized -- have you looked "treatment effect vs. length of time since last experimental night"? This would be a very crude way of getting at tolerance effects.
I have a (completed) 5-year melatonin self-experiment that I will hopefully write up later this year (although... I have been saying that for 12+ months at this point), will be fun to compare notes.
~3x slower after funding stops growing at current rates in 2027-2029
Have you (or others) written about where this estimate comes from?
Very cool!
To deal with the imperfect compliance of the randomization, you could use the "instrumental variables" approach. In this case, since it is (one-sided) noncompliance in an experiment, this amounts to:
I emphasize that this is a very simple econometric technique and does not rely on unreasonable assumptions ("Wald estimator" is another search term here).
Isn't the fact that Manifold is not really a real-money prediction market very important here? If there was real money on the table, for example, it's less likely that the 1/1/26 market would have been "forgotten" -- the original traders would have had money on the line to discipline their attention.
Every time someone calls Manifold (or Metaculus) a "prediction market", god kills an arbitrageur [even though both platforms are still great!].
Since the multiple upvotes seem indicate multiple people looking for an explanation: a link
by Trevor Chow, Basil Halperin, and J. Zachary Mazlish
In this post, we point out that short AI timelines would cause real interest rates to be high, and would do so under expectations of either unaligned or aligned AI. However, 30- to 50-year real interest rates are low. We argue that this suggests one of two possibilities:
by Trevor Chow, Basil Halperin, and J. Zachary Mazlish
[Note: This is an appendix to "AGI and the EMH: markets are not expecting aligned or unaligned AI in the next 30 years"]
One naive objection would be the claim that real interest rates sound like an odd, arbitrary asset price to consider. Certainly, real interest rates are not frequently featured in newspaper headlines – if any interest rates are quoted, it is typically nominal interest rates – and stock prices receive by far the most popular attention.
The importance of real rates. However, even if real interest rates are not often discussed, real interest rates affect every asset price. This is because asset prices always reflect some discounted value of future cash... (read 548 more words →)
(Cross-posted from my blog.)
Confidence level: medium
I want to argue that Newcomb’s problem does not reveal any deep flaw in standard decision theory. There is no need to develop new decision theories to understand the problem.
I’ll explain Newcomb’s problem and expand on these points below, but here’s the punchline up front.
TLDR:
This is interesting. Caveating the below as a "research note comment" in a similar spirit to the caveats around the above being a "research note":
1. Generalizing the model to include bottlenecks
- I generalized the model to have R be CES over L and C, rather than Cobb-Douglas --> to allow for bottlenecks
- Based on conversation with Tom Houlden, though he hasn't reviewed anything I've written here
- [a la Ben Jones (2025) or Jones and Tonetti (2026), if references are useful]
- In the most extreme version, this would be R = min(L, C)
- (This notion of substitutability must be different from that discussed in the post / in the original AIFM model, but I don't follow what's going
... (read more)