It’s worth highlighting that the two expectations do not condition on the same event. This explains why we can have E[A | all even] < E[B | even] even though A ≥ B almost surely: the two “all even”s actually refer to different events.
Arjun Pitchanathan
Karma: 43
He says that he only cares about the learning aspect, and that AI cannot help, because he isn’t bottlenecked by typing speed, i.e., it would take as much time for him to write the code as to read it. But it’s easier to learn from a textbook than figure things out yourself? Perhaps he meant that he only cares about the “figuring out” aspect.
Just to be sure, are you on the “Latest” feed? (as opposed to “Enriched” or “Recommended”)
Would you say that you knew they wouldn’t like the new thing, but you didn’t care because it wasn’t against the rules?
I would like to practice your form of relaxation, do you have some channel suggestions?
You can try to make a prediction about what future you will think. For example, “in 2 years, I will think that working on project X was a good idea”. If other people don’t want to bet on those terms (since you can technically say whatever you want at the end), you can just write down predictions and then see whether your predictions in the past were correct.
You might object that now you don’t have skin in the game, but I think you do, if you care about trying to win the game of writing down good predictions.
You need to have beliefs to test how well-calibrated your beliefs are. Studying is one way to form new beliefs. You could avoid that effort by just testing your existing beliefs.
Suleyman acknowledges that consciousness as a concept is ill defined and tautological.
Few concepts are as elusive and seemingly circular as the idea of a subjective experience.
I don’t think the quote from the article supports your claim here. “Elusive” does not mean “ill-defined”, and “seemingly circular” does not mean “tautological”.
Thank you for the post, I found it very informative on Suleyman’s views.
Serious decisions which have consequences that will effect billions of lives, and potentially billions more minds, should not be made on the basis of “Invisible” concepts which cannot be observed, measured, tested, falsified, or even defined with any serious level of rigor.
I don’t think that the difficulty of ascertaining whether something results in qualia is a valid basis to reject its importance. I would say that one has to extrapolate from oneself to the entity in question and evaluate whether its features suggest associated qualia.
(Some reject that substrate could be a relevant feature and others claim that substrate is the only relevant feature… I have not yet undersood why people hold these beliefs. Prima facie, behavior, algorithm and substrate all seem like they could be relevant.)
Yes, and this might be a crux between “successionists” and “doomers” with highly cosmopolitan values.
Yeah, and to nitpick, “minimum” seems redundant with “necessary” anyway
Active noise cancellation is a huge quality of life improvement when flying.
I think it’s a feature of the local dialect. I’ve seen it multiple times around here and never outside.
Interesting turn of events for the US to now disincentivize trade with Japan, wonder what Commodore Perry would think
My impression is that prior discussion focused on discretizing . is already boolean here, so if the hypothesis is true then it’s for a different reason.
On this particular example you can achieve deterministic error with latent , but it seems easy to find other examples with ratio > 5 (including over latents ) in the space of distributions over with a random-restart hill-climb. Anyway, my takeaway is that if you think you can derandomize latents in general you should probably try to derandomize the latent for variables , for distributions over boolean variables .
(edited to fix typo in definition of )
To check my understanding: for random variables , the stochastic error of a latent is the maximum among . The deterministic error is the maximum among . If so, the claim in my original comment holds—I also wrote code (manually) to verify. Here’s the fixed claim:
Let . With probability , set , and otherwise draw . Let . Let and . We will investigate latents for . Let be the stochastic error of latent . Now compute the deterministic errors of each of the latents , , , , , , . Then for , all of these latents have deterministic error greater than .
It should be easy to modify the code to consider other latents. I haven’t thought much about proving that there aren’t any other latents better than these, though.
Yeah, my comment went through a few different versions and that statement doesn’t apply to the final setting. I should’ve checked it better before hitting submit, sorry. I only used LLMs for writing code for numerical calculations, so the error is mine. [1]
I think that I didn’t actually use this claim in the numerical calculations, so I’d hope that the rest of the comment continues to hold. I had hoped to verify that before replying, but given that it’s been two weeks already, I don’t know when I’ll manage to get to it.
- ^
I did try to see if it could write a message explaining the claim, but didn’t use that
- ^
Thanks! (I knew enough about Avatar to know what you wrote in your last paragraph, but the rest is new to me)
In the “all positions” page, why is the second sentence of most summaries referring to a “detail” or “full description”? I see no way to access anything like that