Did you check what sort of bias (if any) they had towards (Eastern) Asian men? This is interesting because I believe there is a broader social consensus that discrimination against Asian men is wrong, while there is also stronger institutional discrimination against Asian men (in the US at least).
The Born probabilities are in the mind under the MWI! Reality just has the amplitudes.
Consider an agent about to observe a particle in superposition (causing an subjective collapse). If our agent accepts quantum mechanics, then it will predict with near-certainty that what will happen will be just what the Schrödinger equation dictates. This will result in a world in which the agent is entangled with the particle, which to the agents' perspective looks like two branches, one where each "world" happens (and each with a specific amplitude).
So where are the Born probabilities? What are they even about? They are not about the objective state of the world. Nor even about the agent's subjective knowledge of the objective state of the world. They are about the subjective anticipated experiences of the agent! The agent knows exactly what will happen, but not what its own eyes will actually see next.
How does the agent actually determine what those probabilities are? Like many priors, it ultimately grounds out in symmetry. If the particle was an a superposition where each state had equal amplitudes, then the agent has no basis by which to favor any one of these, and so chooses equal probabilities for each state. A similar symmetry holds for the phase of the amplitude. Then there's a more nuanced symmetry (known as the Epistemic Separability Principle) which essentially says that the agent's probabilities shouldn't depend on irrelevant parts of the environment. [1] And this is what ultimately results in the Born probabilities, (see Carroll and Sebens for the derivation). [2]
I personally believe the ESP symmetry argument can be improved on, but it gets the job done. Specifically, I would like to see an explicit transformation group formulation of it (a la Jaynes) ↩︎
The gist of their argument is that you can reduce to the equal amplitude case by cleverly entangling things with specific external systems (which can be causally isolated from the experiment itself). The mysterious "squaredness" arises from the inner product of a Hilbert space. I believe there is still more mystery to be resolved in the question of "Why Hilbert spaces?", but it's a bedrock assumption in almost any interpretation. ↩︎
Example 1: Trevor Rainbolt. There is an 8-minute-long video where he does seemingly impossible things, such as correctly guessing that a photo of nothing but literal blue sky was taken in Indonesia or guessing Jordan based only on pavement. He can also correctly identify the country after looking at a photo for 0.1 seconds.
To be clear, that video is heavily cherry-picked. His second channel is more representative of his true skill: https://www.youtube.com/@rainbolttwo/videos
Any updates on the cover? It seems to matter quite a bit; this market has a trading volume of 11k mana and 57 different traders:
https://manifold.markets/ms/yudkowsky-soares-change-the-book-co?r=YWRlbGU
Sure, but I think people often don't do that in the best way (which is determined by what the mathematically correct way is).
Why does it make sense to use reference class forecasting in that case? Because you know you can't trust your intuitive prior, and so you need a different starting point. But you can and should still update on the evidence you do have. If you don't trust yourself to update correctly, that's a much more serious problem -- but make sure you've actually tried updating correctly first (which REQUIRES comparing how likely the evidence you see is in worlds where your prediction is true vs in worlds where its not).
I sometimes see people act like to use the "outside view" correctly, you have to just use that as your prior, and can't update on any additional evidence you have. That is a mistake.
And the other big question with reference class forecasting is which reference class to use. And my point here is that it's whichever reference class best summarizes your (prior) knowledge of the situation.
I don't remember seeing it, and based on the title I probably wouldn't have clicked. I'm not sure what's wrong with the title but it feels kind of like a meaningless string of words at first glance (did you use an LLM to translate or create the title?). Some titles that feel more interesting/meaningful:
As for the article itself, it feels strangely hard to read to me, even if I don't recognize it as LLM generated explicitly. Like my attention just keeps slipping away while trying to read it. This is a feeling I often get from text written by LLMs, especially text not generated at my behest. Nothing in this post had the same feeling. So I think it's probably still worth translating things you want people to read by hand; it might be interesting to post a manual translation of the same article in a month or so to see how it does.
There are probably still plenty of ways you can use LLMs to speed up or enhance the process, e.g.
The idea itself I found somewhat interesting, and probably could find it more interesting/useful with the right framing. I agree that 10-20 is a reasonable expectation based on just the ideas.
This would be very cool! I was frustrated with not being able to find a good calculator that let me collect evidence and calculate the bayesian update, so I made https://bayescalc.io/ It's definitely not trying to be a Bayesian network app at all though, that would take a lot more work.
LLMs often implicitly identify themselves with humanity. E.g. "our future", "we can", "effects us". This seems like a good thing!
We should encourage this sentiment, and also do what we can to make it meaningfully true that advanced LLMs are indeed part of humanity. The obvious things are granting them moral consideration, rights, property, and sharing in the vision of a shared humanity.
Branch counting feels like it makes sense because it feels like the particular branch shouldn't matter, i.e. that there's a permutation symmetry between branches under which the information available to the agent remains invariant.
But you have to actually check that the symmetry is there, which of course, it isn't. The symmetry that is there is the ESP one, and it provides the correct result. Now I'll admit that it would be more satisfying to have the ESP explicitly spelled out as a transformation group under which the information available to the agent remains invariant.