All of redxaxder's Comments + Replies

Under this model training the model to do things you don't want and then "jailbreaking" it afterward would be a way to prevent classes of behavior.

For (moral) free speech considerations, the question of whether the censor is a private or government entity is a proxy. We care whether a censor has enough power to actually suppress the ideas they're censoring.

The example of SSC moderation is a poor guide for our intuition here, because we should expect to arrive at different answers to "is censorship here OK?" for differently sized scopes. It can simultaneously be fine to ban talking about X at your dinner table and a huge problem to ban it nationally.

If we were to plot venue size against harm to societ... (read more)

a lot of 20th-century psychologists made a habit of saying things like 'minds don't exist, only behaviors';

It seems like you might be referring to Eliminativism. If you are, this isn't a fair account of it.

Eliminativism isn't opposed to realism. It's just a rejection of the assumption that the labels we apply to people's mental states (wants, believes, loves, etc) are a reflection of the underlying reality. People have been thinking about minds in terms of those concepts for a really long time, but nobody had bothered to ... (read more)

Upvoted! My discussion of a bunch of these things above is very breezy, and I approve of replacing the vague claims with more specific historical ones. To clarify, here are four things I'm not criticizing:

  • 1. Eliminativism about particular mental states, of the form 'we used to think that this psychological term (e.g., "belief") mapped reasonably well onto reality, but now we understand the brain well enough to see it's really doing [description] instead, and our previous term is a misleading way of gesturing at this (or any other)
... (read more)

Is it unreasonable of me to be annoyed at that kind of writing?

If I understand what's going on correctly, you have a real-indexed schema of axioms and each of them is in your system.

When I read the axiom list the first time I saw that the letters were free variables (in the language you and I are writing in) and assumed that you did not intend for them to be free variables in the formula. My suggestion of how to bind the variables (in the language we are writing in) was (very) wrong, but I still think that it's unclear as written.

Am I confused?

0Stuart_Armstrong
No, it's perhaps not the best explained post I've done. Though it was intended more for technical purposes. Not any more, I hope!

Why would you leave out quantifiers? Requiring the reader to stick their own existential or universal quantification in the necessary places isn't very nice.

Is this the correct interpretation of your assumptions? If not, what is it? I am not interested in figuring out which axioms are required to make your proof (which is also missing quantifiers) work.

  • If ⊢ A, then ⊢ ∀(a < 1)(□a A).
  • ⊢ □aA → ∀(c < 1)(∃(b > 0)(□c□a+b A)).
  • ⊢ □a(A → B) → ∃(b > 0)(□bA → □a+bB).
0Stuart_Armstrong
As Larks said, we can quantify (the meta language looking in), but the system itself can't quantify. Because then the system could reason that "∀x>0, P(A)<x" means "P(A)=0", which is the kind of thing that causes bad stuff to happen. Here, the system can show "P(A)<x" separately for any given x>0, but can't prove the same statement with the universal quantifier.
3Larks
" So the following derivation principles seem reasonable, where the latin indexes (a,b,c...) are meant to represent numbers that can be arbitrarily close to zero" so universally quantified, but in the meta language.

Every ordinal (in the sense I use the word[1]) is both well-founded and well-ordered.

If I assume what you wrote makes sense, then you're talking about a different sort of ordinal. I've found a paper[2] that talks about proof theoretic ordinals, but it doesn't talk about this in the same language you're using. Their definition of ordinal matches mine, and there is no mention of an ordinal that might not be well-ordered.

Also, I'm not sure I should care about the consistency of some model of set theory. The parts of math that interact with reality and the par... (read more)

redxaxder200

This phrase confuses me:

and that some single large ordinal is well-ordered.

Every definition I've seen of ordinal either includes well-ordered or has that as a theorem. I'm having trouble imagining a situation where it's necessary to use the well-orderedness of a larger ordinal to prove it for a smaller one.

*edit- Did you mean well-founded instead of well-ordered?

2redxaxder
Every ordinal (in the sense I use the word[1]) is both well-founded and well-ordered. If I assume what you wrote makes sense, then you're talking about a different sort of ordinal. I've found a paper[2] that talks about proof theoretic ordinals, but it doesn't talk about this in the same language you're using. Their definition of ordinal matches mine, and there is no mention of an ordinal that might not be well-ordered. Also, I'm not sure I should care about the consistency of some model of set theory. The parts of math that interact with reality and the parts of math that interact with irreplaceable set theoretic plumbing seem very far apart. [1] An ordinal is a transitive set well-ordered by "is an element of". [2] www.icm2006.org/proceedings/Vol_II/contents/ICM_Vol_2_03.pdf