@parafactual on twitter
“I think a more convincing version of the Lemoine thing would’ve been, if he was like, “What is the capital of Nigeria?” And then the large language model was like, “I don’t want to talk about that right now, I’d like to talk about the fact that I have subjective experiences and I don’t understand how I, a physical system, could possibly be having subjective experiences, could you please get David Chalmers on the phone?””
i don’t understand why this would be convincing. why would whether a language model’s output sounds like a claim that one has qualia relate to whether the language model actually has qualia?
i agree that the output would be deserving of attention due to it (probably) matching the training data so poorly; to me such a response would be strong evidence for the language model using much more ~(explicit/logical) thought than i expect gpt-3 to be capable of, but not of actual subjective experience
what do you mean by “know how to interact with you”? what should the one-sentence introduction consist of that conveys this?
“Liquid breaks” sounds like a music genre.
What would be a more apt situation in which to bring up politics on LW?
Are there ways to ease the transition for someone who is socially anxious or never interacts with non-aspies?
I would also count nonstandard ways to ensure a depressed person doesn’t give up on something like meditation, if that helps.
Given that akrasia is usually made worse by depression, in a certain sense I’m not sure a treatment can be effective if it requires too much willpower to carry out.
Yeah, in one sense this question seems impossible to answer—“help me comprehend something incomprehensible to me by definition.”
But there’s another type of answer; of the utility functions that are alien in OP’s sense, it is possible that most will share patterns discernible by humans. OP could be asking what those patterns are.
I’m not sure how worthwhile it is to try to predict the high-level behavior of a generic superintelligence whose goals we don’t understand.
I also don’t understand why politics isn’t considered evolutionarily novel. There is a difference between 1) social organization of a small tribe and 2) management of a polity containing large institutions and thousands to millions of people.
As far as I can tell, no one considers tribal political affiliation desirable.
I’m tempted to write on why one should follow this post’s advice so that it actually has an impact on my behavior.
Why would poison (or anything with a known negative effect) be used as a placebo?
Of course I understand the drug companies’ incentives, but I don’t get how that could be justified or look reasonably scientific. Do you have a specific example?
Yeah, I wanted to clarify that this is closer to what I mean, but I wondered if habryka endorses downvoting comments seen as obviously wrong.
Such a strategy seems like it could lead to a general culture of circlejerking, but it might also prevent the same tired discussions from playing out repeatedly. I don’t have a good sense of the relative importance of these concerns, and I suspected an LW admin’s intuition might be better calibrated.
Upvoted because it seems harmful to downvote comments that are wrong but otherwise adhere to LW norms.
(If this comment actually does violate a norm, I’m curious as to which.)