When you can predict that beliefs won’t update towards convergence, you’re predicting a mutual lack of respect and a mutual lack of effort to figure out whose lack of respect is misplaced.
Are you saying that the interlocutors should instead change to attempting to resolve their lack of mutual respect?
As a relatively new person to lesswrong, I agree.
The number of conversations which I’ve read which end in either party noticeably updating one way or the other have been relatively rare. The one point I’m not sure if I agree with is being able to predict a particular disagreement is a problem?
I suppose being able to predict the exact way in which your interlocutors will disagree is the problem? If you can foresee someone disagreeing in a particular way, and then accounting for it in your argument, and then they disagree anyway, in the exact way you tried to address, that’s generally just bad faith.
(though sometimes I do skim posts, by god)
I try to ask myself whether the tenor of what I’m saying overshadows definitional specificity, and how I can provide a better mood or angle. If my argument is not atonal—if my points line up coherently, such that a willing ear will hear, definitionalist debates should slide on by.
As a descriptivist, rather than a prescriptivist, it really sucks to have to fall back on Socratic methods of pre-establishing definitions, except in highly-technical locations.
Thus, I prefer to avoid arguments which hinge on definitions altogether. This doesn’t preclude examples-based arguments, where for example, various interlocutors are operating off different definitions of the same terms but have different examples.
For example, take the term tai.
For some, tai means not when ai is agentic, but when ai can transform the economy in some large or measurable way. For others, it is when the first agentic ai deployed at scale occurs. Yet still, others have differing definitions! Definitions which wildly transform predictions and change alignment discussions. Despite using the term with each other in different ways- with separate definitions- interlocutors often do not notice (or perhaps are subconsciously able to resolve the discrepancy?)!
I don’t think there’s any place quite like lesswrong on the entire internet. It’s a lot of fun to read, but it tends to be pretty one-note, and even if there is discord in lesswrong’s song, it’s far more controlled, Eru Illuvitar’s hand can yet be felt, if not seen. (edit: that is to say, it’s all the same song)
For the most part, people are generally-tolerant of Christians. There is even a Catholic who teaches (taught?) at the Center For Applied Rationality, and there’s a few other rationalist-atheists who hopped to christianity, though I can’t remember them by name.
Whether or not it’s the place for you, I think you’ll find that there’s more pop!science, and if you are a real physicist, there’s more and more posts where people who do not know physics will act like they do, and correcting them will be difficult, and it depends on if you can tolerate that.
I wasn’t aware that Eliezer was an experienced authority on SOTA LLMs.
I don’t agree, but for a separate reason from trevor.
Highly-upvoted posts are a signal of what the community agrees with or disagrees with, and I think being able to more easily track down karma would cause reddit-style internet-points seeking. How many people are hooked on Twitter likes/view counts?
Or “ratio’d”.
Making it easier to track these stats would be counterproductive, imo.
It seems pretty standard knowledge among pollsters that even the ordering of questions can change a response. It seems pretty blatantly obvious that if we know who a commenter is, that we will extend them more or less charity.
Even if the people maintaining the site don’t want to hide votes + author name on comments and posts, it would be nice if user name + votes were moved to the bottom. I would like to at least be presented with the option to vote after I have read a comment, not before.
Re: Papers- I’m aware of papers like you’re alluding to, though I haven’t been that impressed.
The reason why I don’t want a scratch-space, is because I view scratch space and context equivalent to giving the ai a notecard that it can peek at. I’m not against having extra categories or asterisks for the different kinds of ai for the small test.
Thinking aloud and giving it scratch space would mean it’s likely to be a lot more tractable for interpretability and alignment research, I’ll grant you that.
I appreciate the feedback, and I will think about your points more, though I’m not sure if I will agree.
Given my priors + understanding of startups/silicon valley culture, it sounds more like Openai started to leave the startup phase and is entering into the “profit-seeking” phase of running the company. After they had the entire rug pulled under them by stable diffusion, I would expect them to get strategic whiplash, and then decide to button things up.
The addition of a non-compete clause within their terms of service and the deal with Microsoft seems to hint towards that. They’ll announce GPT-3.5, and “next-gen language models”, but it doesn’t match my priors that they would hold back GPT-4 if they had it.
Time may tell, however!
What I’m asking with this particular test is, can an ai play blindfold chess, without using a context in order to recant every move in the game?
I’m confused. What I’m referring to here is https://en.wikipedia.org/wiki/Blindfold_chess
I’m not sure why we shouldn’t expect an ai to be able to do well at it?
My proposed experiment / test is trying to avoid analogizing humans, but rather scope out places where the ai can’t do very well. I’d like to avoid accidentally overly-narrow-scoping the vision of the tests. It won’t work with an ai network where the weights are reset every time.
An alternative, albeit massively-larger-scale experiment might be:
Will a self-driving car ever be able to navigate from one end of a city to another, using street signs and just learning the streets by exploring it?
A test of this might be like the following:
Randomly generate a simulated city/town, complete with street signs and traffic
Allow the self-driving car to peruse the city on its own accord
(or feed the ai network the map of the city a few times before the target destinations are given, if that is infeasible)
Give the self-driving car target destinations. Can the self-driving car navigate from one end of the city to the other, using only street signs, no GPS?
I think this kind of measuring would tell us how well our ai can handle open-endedness and help us understand where the void of progress is, and I think a small-scale chess experiment like this would help us shed light on bigger questions.
I just wanted to break up the mood a bit. Reading everything here is like listening to a band stuck in the same key.
Unfortunately, what I am proposing is not possible with current language models, as they don’t work like that.
Thanks. fyi, i tried making the post i alluded to:
I was rereading and noticed places where I was getting a bit too close to insulting the reader. I’ve edited a couple places and will try to iron out the worst spots.
If it’s bad enough I’ll retract the article and move it back to drafts or something. Idk.
So long as the “buffer” is a set of parameters/weights/neurons, that would fit my test.
Do you believe encouraging the site maintainers to implement degamification techniques on the site would help with your criticisms?