Thanks, I'm looking forward to this! Some questions that seem worth considering to me are:

1. Is AGI likely to lock in values? (if so it's probably bad for animals)
2. Is the answer to this question even knowable? (a lot of what I've heard on the topic has been like "AI could mean X but also not X")
3. If AGI is good/bad, how steerable is it? (e.g. maybe making sure that AGI goes well for humans is actually much easier)

I think these are fair points, but the tone seems deconstructive and a bit condescending. I think it's possible to disagree and to caution loudly while still respecting that the post was made in good faith.

For what it's worth I'm also surprised by the reaction. Within government departments in NZ (where I worked before) this is not allowed. Of course it still happens but it seems good to me for the organization to discourage it. 

*Edit for spelling

 Want to add this here: https://www.reddit.com/r/dataisbeautiful/comments/1rhv521/oc_dietary_v_nondietary_veganism_interest_over_16/

Reddit might not be the best source of data. But it confirms what I've heard elsewhere that 2018-2020 was the height of veganism as a health craze, and at least indicates that ethical vegans (if they are reflected by those who buy vegan clothes) are still rising. 

I think this is a super important question and want way more conversation about it - but could we re-frame your conclusion as being not that we shouldn't use AI, but should be mindful about how we're using AI?

The scenario you described appears to be a pretty bad use. But I think much of the harm you're seeing could be mitigated. Here are some ideas, just off the top of my head, addressing the issues you listed (in order):

Use of AI in research should -

1. Consider the appropriateness of AI in that context (e.g. is this an area where we need the most up-to-date answers? Is this an area where we want to consider non-western perspectives?)
2. Approach AI-generated answers critically, treating them as vibes-based answers rather than having any authority (and in group-work contexts, leaders should encourage this)
3. Have AI write up its answers in bullet points rather than full text, so that a human is always contributing to the style
4. Be a second or later-resort option (try to think creatively/critically first, rather than relying on AI - again, leaders can encourage this)
5. In group settings: encourage new or unusual ideas (addresses the last two points). 

I know these are far from perfect solutions. Point 4 is admittedly quite hard to keep up (I feel myself struggling with this). But to me it feels similar to how a calculator makes people lazy (I'm sure I can't do mental arithmetic now as well as I could when I was 12), but is still a net win. It seems likely that if we create good habits/culture about using AI, its benefits can significantly outweigh the downsides, even in research.^[1] But I do think that requires a lot of conversations, and maybe some research, into how to use it well and avoid those pitfalls. So I would love to see more posts discussing this.

1. ^{^}

   I think these benefits are pretty significant. For instance, (and as a counter-point to 5), I find AI can actually help to reign in crazy ideas, by acting as a sanity-check tool; I also find it's helpful to quickly spot holes in an argument when otherwise I would have only gotten feedback from a colleague some days later; and it can quickly structure disorganized ideas. But surely there are many more.

By power I mean: ability to change the world, according to one's preferences. Humans clearly dominate today in terms of this kind of power. Our power is limited, but it is not the case that other organisms have power over us, because while we might rely on them, they are not able to leverage that dependency. Rather, we use them as much as we can.

No human is currently so powerful as to have power over all other humans, and I think that's definitely a good thing. But it doesn't seem like it would take much more advantage to let one intelligent being dominate all others.

The argument I'm referring to is the AI doom argument. Y&S are its most prominent proponents, but are widely known to be eccentric and not everyone agrees with their presentation of it. I'm not that deep in the AI safety space myself, but I think that's pretty clear.

The authors of this post seemed to respond to the AI doom argument more generally, and took the book to be the best representative of the argument. So that already seems like a questionable move, and I wish they'd gone further.

I don't think the point about alien preferences is a crux of the AI doom argument generally. I think it it's presented in Bostrom's Superintelligence and Rob Miles videos (and surely countless other places) as: "an ASI optimising for anything that doesn't fully capture collective human preferences would be disastrous. Since we can't define collective human preferences, this spells disaster." In that sense it doesn't have to be 'alien', just different from the collective sum of human preferences. I guess Y&S took the opportunity to say "LLMs seem MUCH more different" in an attempt to strengthen their argument, but they didn't have to.

So, as I said, I'm not really that deep into AI safety, so I'm not the person to go to for the best version of these arguments. But I read the book, sat down with some friends to discuss it... and we each identified flaws, as the authors of this post did, and then found ways to make the argument better, using other ideas we'd been exposed to and some critical reflection. It would have been really nice if the authors of the post had made that second step and steelmanned it a bit. 

Thanks Yarrow, I can see that that was confusing.

I don't think that Yudkowsky & Soares's argument as a whole is obviously wrong and uninteresting. On the contrary, I'm rather convinced by it, and I also want more critics to engage with it.

But I think the argument presented in the book was not particularly strong, and others seem to agree: the reviews on this forum are pretty mixed (e.g.). So I'd prefer critics to argue against the best version of this argument, not just the one presented in the book. If these critics had only set out to write a book review, then I'd say fine. But that's not what they were doing here. They write "there is no standard argument to respond to, no single text that unifies the AI safety community" - true, but you can engage with multiple texts in order to respond to the best form of the argument. In fact that's pretty standard, in academia and outside of it. 

It doesn't seem like a straw man to me when 1) the effectiveness of these interventions is evaluated against their short term impact (as far as I'm aware ACE doesn't consider this kind of long term impact much at all), and 2) the orgs don't publish any long term theory of change to help donors or critics decide if they agree with it. This strongly implies that their long term theory of change is far less important than the short term wins, at least at the organization level. 

Just out of interest, do you believe that animal welfare wins are moving us AWAY from abolition? I agree with you that it's possible but I haven't ever seen any evidence that there is this effect. It also seems very possible to have incremental improvements and then eventually abolition, as people become more empathetic and aware.