Thanks for engaging. I have karma upvoted but disagreed voted.
You wrote a lot, so I’ll focus on your two bounded points.
About Goodfire
I think you are wrong. If I had to guess, it’s in the same way that people are misguided about Anthropic.
For comparison, this is the copy at the top of Anthropic’s Research page:
Our research teams investigate the safety, inner workings, and societal impacts of AI models – so that artificial intelligence has a positive impact as it becomes increasingly capable.
This is their copy at the top of Goodfire’s About page:
Goodfire is a research company using interpretability to understand, learn from, and design AI systems. Our mission is to build the next generation of safe and powerful AI—not by scaling alone, but by understanding the intelligence we’re building.
Fundamental interpretability research to understand and intentionally design advanced AI systems
They specifically state their wish to design advanced and the next generation of AI systems. To the extent I may have mis-characterised them in that people may misunderstand that they are not just doing capabilities on top of existing systems. What I have described could lead to people thinking they are just doing RAG or something.
About Revolving Doors
I think it is better to criticize specific decisions made by evaluations companies, or decisions they haven’t taken, rather than pointing at associations between people and inferring harmful behavior without evidence (I think your example of people at evaluations orgs being concerned about losing API access if they speak out is a much better criticism here). In general I think this sort of guilt-by-association is not very fruitful.
I think you are fractally wrong in this paragraph. I’ll try to show it to you.
1) “It is better to [A] than [B]” is a false dilemma. In the article, as you point out, I just do both.
2) “Pointing associations between people and inferring harmful behavior without evidence” does not describe what I am doing, again as you point out.
3) It is crucial to point association between people and infer harmful behaviour from it even without confirmation! One should not reject evidence just because it is not “very fruitful” when used by itself. It is obviously relevant that the head of safety of US-AISI, the CEO of Open Phil, and the CEO of Anthropic were all roommates, and it is trivial to infer things from that.
4) I literally wrote “This is not only about having a couple of senior staff from the industry. That in itself can be good! It’s the whole picture that looks bad.”, and then proceeded to list the whole picture.
So probably this cashes out as more fundamental generator disagreements that aren’t worth hashing out here. Broadly I think it’s okay for a company to say “We are designing and advancing the next generation of AI systems” and I think to analyze whether they should be bucketed with capabilities labs like OpenAI or Anthropic (which I think are also meaningfully different places), one should look at and critically assess their research output.
Like, if someone believes that interpretability with be both helpful to build better systems and helpful to build safer systems, I think it’s justifiable for them to do the thing that builds better systems in the hopes that those systems are also safer than the next-best-thing that would’ve been built (and that’s probably reliant on a bunch of other beliefs where we may differ as I said before).
About revolving doors
Sorry I may have been unclear. What I meant is that [A] is good and [B] is bad. I am criticizing [B] here which is playing the associations-game. I think that is generally bad and you should not do it.
I think it describes some of what you are doing though not all of it, yes.
I am not saying one should reject it as evidence. It’s fine to say it publicly. What I disagree with is the inference that is implied. I do not think it is trivial to infer things from that. I have had roommates with whom I disagree and I would say so publicly. It is, in my opinion, better to apply criticism to actual actions you disagree with, or strategic choices, or whatever, instead of who people were roommates with or not. Otherwise you get to games like what happened with your point about Goodfire/Apollo where we have: Apollo is suspect because it was co-founded by two people who then went to Goodfire and anyone who works at Goodfire is suspect because in some way they think what they’re working on will be useful for AI development (even though I’m pretty sure those people continue to see their work as being heavily motivated by safety, and differentially useful for safety). I think this chain of inferences is bad, and so we should cut it at the root. Most professional criticism should be about what people have actually done, not about who they’ve been associated with.
It looks bad in the same way as other revolving door cultures can look bad to outsiders, but the question should be is it bad? Is it an actual structural problem? Or is it a PR problem? Or just a feature of the space? And I think for answering these questions, it is better to look at behaviour (which again, you did do as well as I mentioned originally, I just think the association-game stuff isn’t great).
Genuine thanks for your response. I have again karma-upvoted and disagreed-voted.
About Goodfire
Broadly, I don’t care much about the moral intents of OpenAI, Anthropic and Goodfire. I think everyone is the hero of their own story.
To the extent that people state their intents, I basically remove all the moral colours and emotional valence. For instance, “we build capabilities for The Good and for Good Reasons” becomes “we build capabilities”.
Another example may be “we care a lot about rationality”. Rationality is the practice of “reason”, which is “thinking correctly”. “Care” is also “we have good feelings for”. So, through this process, it becomes “we have a lot of feelings around the practice of thinking”.
A last example could be “effective altruism”. Here, it’s hard to see what’s left after you go through this process. Something like “we do things connected to others (altruism) that perform highly on measures of our own choice (effective)”.
I strongly recommend adopting this frame of analysis to any of my readers, not only “GenericModel”. If you are this deep in this thread, you most likely need it.
About revolving doors
On 3, I think you are forgetting the basics of LessWrong.
I am not saying one should reject it as evidence. It’s fine to say it publicly. What I disagree with is the inference that is implied. I do not think it is trivial to infer things from that. I have had roommates with whom I disagree and I would say so publicly.
“If [X], then necessarily [Y] is true” is logical entailment. Evidence is “If [X], then [Y] is more likely.” A single counter-example does not disprove it.
If you mean that housemates (by choice, not necessity!) who have worked together are not more likely to be synchronised on their view, then I think you are clearly wrong.
Just so you know, Holden has moved from CEO of Open Philanthropy to Anthropic a year ago. This type of revolving door was entirely predictable through this type of evidence.
I have been polite in the article. Had I included public links of marriage, things would already look worse. Had a journalist included private affairs, the picture would look much worse. This would be entirely valid probabilistically.
On 4, I think you are forgetting that everything is adversarial.
it is better to look at behaviour
Yes, but people lie, they deceive, they “strategically withhold” information, and so on. Even when they don’t, it also just takes time to document everything, and vet it for publication.
This lack of transparency is why we use other types of evidence. They tend to be less accurate, but also more plentiful, less filtered, and harder to fake.
It looks bad in the same way as other revolving door cultures can look bad to outsiders, but the question should be is it bad? Is it an actual structural problem?
Yes, and I have listed the adversarial considerations in the article. It is the part that literally starts with “The considerations around independence are structural. Namely, [...]”
Thanks for engaging. I have karma upvoted but disagreed voted.
You wrote a lot, so I’ll focus on your two bounded points.
About Goodfire
I think you are wrong. If I had to guess, it’s in the same way that people are misguided about Anthropic.
For comparison, this is the copy at the top of Anthropic’s Research page:
This is their copy at the top of Goodfire’s About page:
And at the top of their Research page:
They specifically state their wish to design advanced and the next generation of AI systems. To the extent I may have mis-characterised them in that people may misunderstand that they are not just doing capabilities on top of existing systems. What I have described could lead to people thinking they are just doing RAG or something.
About Revolving Doors
I think you are fractally wrong in this paragraph. I’ll try to show it to you.
1) “It is better to [A] than [B]” is a false dilemma. In the article, as you point out, I just do both.
2) “Pointing associations between people and inferring harmful behavior without evidence” does not describe what I am doing, again as you point out.
3) It is crucial to point association between people and infer harmful behaviour from it even without confirmation! One should not reject evidence just because it is not “very fruitful” when used by itself. It is obviously relevant that the head of safety of US-AISI, the CEO of Open Phil, and the CEO of Anthropic were all roommates, and it is trivial to infer things from that.
4) I literally wrote “This is not only about having a couple of senior staff from the industry. That in itself can be good! It’s the whole picture that looks bad.”, and then proceeded to list the whole picture.
About Goodfire
So probably this cashes out as more fundamental generator disagreements that aren’t worth hashing out here. Broadly I think it’s okay for a company to say “We are designing and advancing the next generation of AI systems” and I think to analyze whether they should be bucketed with capabilities labs like OpenAI or Anthropic (which I think are also meaningfully different places), one should look at and critically assess their research output.
Like, if someone believes that interpretability with be both helpful to build better systems and helpful to build safer systems, I think it’s justifiable for them to do the thing that builds better systems in the hopes that those systems are also safer than the next-best-thing that would’ve been built (and that’s probably reliant on a bunch of other beliefs where we may differ as I said before).
About revolving doors
Sorry I may have been unclear. What I meant is that [A] is good and [B] is bad. I am criticizing [B] here which is playing the associations-game. I think that is generally bad and you should not do it.
I think it describes some of what you are doing though not all of it, yes.
I am not saying one should reject it as evidence. It’s fine to say it publicly. What I disagree with is the inference that is implied. I do not think it is trivial to infer things from that. I have had roommates with whom I disagree and I would say so publicly. It is, in my opinion, better to apply criticism to actual actions you disagree with, or strategic choices, or whatever, instead of who people were roommates with or not. Otherwise you get to games like what happened with your point about Goodfire/Apollo where we have: Apollo is suspect because it was co-founded by two people who then went to Goodfire and anyone who works at Goodfire is suspect because in some way they think what they’re working on will be useful for AI development (even though I’m pretty sure those people continue to see their work as being heavily motivated by safety, and differentially useful for safety). I think this chain of inferences is bad, and so we should cut it at the root. Most professional criticism should be about what people have actually done, not about who they’ve been associated with.
It looks bad in the same way as other revolving door cultures can look bad to outsiders, but the question should be is it bad? Is it an actual structural problem? Or is it a PR problem? Or just a feature of the space? And I think for answering these questions, it is better to look at behaviour (which again, you did do as well as I mentioned originally, I just think the association-game stuff isn’t great).
Genuine thanks for your response. I have again karma-upvoted and disagreed-voted.
About Goodfire
Broadly, I don’t care much about the moral intents of OpenAI, Anthropic and Goodfire. I think everyone is the hero of their own story.
To the extent that people state their intents, I basically remove all the moral colours and emotional valence. For instance, “we build capabilities for The Good and for Good Reasons” becomes “we build capabilities”.
Another example may be “we care a lot about rationality”. Rationality is the practice of “reason”, which is “thinking correctly”. “Care” is also “we have good feelings for”. So, through this process, it becomes “we have a lot of feelings around the practice of thinking”.
A last example could be “effective altruism”. Here, it’s hard to see what’s left after you go through this process. Something like “we do things connected to others (altruism) that perform highly on measures of our own choice (effective)”.
I strongly recommend adopting this frame of analysis to any of my readers, not only “GenericModel”.
If you are this deep in this thread, you most likely need it.
About revolving doors
On 3, I think you are forgetting the basics of LessWrong.
“If [X], then necessarily [Y] is true” is logical entailment.
Evidence is “If [X], then [Y] is more likely.” A single counter-example does not disprove it.
If you mean that housemates (by choice, not necessity!) who have worked together are not more likely to be synchronised on their view, then I think you are clearly wrong.
Just so you know, Holden has moved from CEO of Open Philanthropy to Anthropic a year ago. This type of revolving door was entirely predictable through this type of evidence.
I have been polite in the article. Had I included public links of marriage, things would already look worse. Had a journalist included private affairs, the picture would look much worse. This would be entirely valid probabilistically.
On 4, I think you are forgetting that everything is adversarial.
Yes, but people lie, they deceive, they “strategically withhold” information, and so on. Even when they don’t, it also just takes time to document everything, and vet it for publication.
This lack of transparency is why we use other types of evidence. They tend to be less accurate, but also more plentiful, less filtered, and harder to fake.
Yes, and I have listed the adversarial considerations in the article. It is the part that literally starts with “The considerations around independence are structural. Namely, [...]”