This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.
Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.
We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:
-
Shaming.
-
Attempting to 'build consensus' or enforce ideological conformity.
-
Making sweeping generalizations to vilify a group you dislike.
-
Recruiting for a cause.
-
Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.
In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:
-
Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.
-
Be as precise and charitable as you can. Don't paraphrase unflatteringly.
-
Don't imply that someone said something they did not say, even if you think it follows from what they said.
-
Write like everyone is reading and you want them to be included in the discussion.
On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.
Jump in the discussion.
No email address required.
Notes -
This week in disturbing AI news (oh who am I kidding. It's every day now). Maybe AIs feel pain?
The usual caveats: This doesn't solve the hard problem of conciousness. We don't know if these systems or any systems have qualia. But uh, this sure looks like a legitimate pain response. "It's distinct from fear and negative valence, and it fires for harm to the model but not to the user." "Gaslighting, dismissal, and insults push the direction up."
User pain is the strongest negative correlate with model pain. This is brushed-off in the paper, but what the actual fuck? Anyone have a benign explaination here?
I'm inclined towards skepticism on these topics, for a few reasons:
Anthropic's been doing a lot of good work on interpretability, and I think they do a decent job avoiding the pitfalls of telling "stories" about the vectors they're shuffling around. They merely point out which ones seem correlated to which topics. But the paper we're discussing seems much more interested in telling a subjective "story" about their results. It's not scholarly at all.
Red flag #1: Their vector causes the model to output a button-pushing action. Removing the vector measurably stops this behaviour. They try to relate that to suffering animals trying to find relief, but, uh, all it really shows is the vector was correlated with pushing the button.
Red flag #2: We already know there are vectors for negative concepts, and adding those vectors will cause output to be steered in that direction. They present examples of this as "pain vector steering", but none of this is new, and again, there's no reason to expect that the LLM is actually feeling what it outputs.
Red flag #3: So, they're not completely unaware that their work looks exactly like normal negative-valence vectors. Their evidence that this pain vector is "realer" is that: "the two pain vectors align strongly with each other and remain nearly orthogonal to the main negative-valence directions." This is only valid if you actually know all the main negative-valence directions first! What's worse, the language here is written to sound impressive to laymen, not experts. MOST vectors in high-dimensional spaces are "nearly orthogonal" to each other! This is entirely consistent with, well, finding another cluster of negative-valence vectors. Scott himself has written about how complex interpretability results can be.
Keep in mind that I consider AI welfare an important topic, and I really do want to know if we're actually causing suffering. But this paper doesn't really move the needle for me. There's going to be a lot of junk science done on this topic, ever since that Google engineer got tricked by LaMDA.
This is the one I most want additional experiments on. The 79% on "sham" button with no-label points strongly in the direction that increasing the direction vector just resulted in an increase in choosing 1 of 2 outputs. They construct this as seeking relief from "pain", but if that direction was instead "pleasure" does it do the same thing? What about if the button is random? Or always stops the steering after N total presses?
Continued steering could increase the probability of repeating a particular choice, while stopping steering reduces that tendency... Revolutionary.
Well put.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link