Firstly, I very much appreciate the grant made by the LTF Fund! On the discussion of the paper by Stephen Cave & Seán Ó hÉigeartaigh in the addenda, I just wanted to briefly say that I’d be happy to talk further about both: (a) the specific ideas/approaches in the paper mentioned, and also (b) broader questions about CFI and CSER’s work. While there are probably some fundamental differences in approach here, I also think a lot may come down to misunderstanding/lack of communication. I recognise that both CFI and CSER could probably do more to explain their goals and priorities to the EA community, and I think several others beyond myself would also be happy to engage in discussion.
I don’t think this is the right place to get into that discussion (since this is a writeup of many grants beyond my own), but I do think it could be productive to discuss elsewhere. I may well end up posting something separate on the question of how useful it is to try and “bridge” near-term and long-term AI policy issues, responding to some of Oli’s critique—I think engaging with more sceptical perspectives on this could help clarify my thinking. Anyone who would like to talk/ask questions about the goals and priorities of CFI/CSER more broadly is welcome to reach out to me about that. I think those conversations may be better had offline, but if there’s enough interest maybe we could do an AMA or something.
Neat! I’d be very interested in talking about/debating this, perhaps in the comments of another post. In particular, the sections above that feel most cruxy to me are the ones on the centrality of conceptual progress to AI strategy/policy work: what that looks like, how to figure out what new concepts are needed, or whether this is even an important part of AI policy, are all things I’d be interested to discuss.
I would definitely also be interested in talking about this, either somewhere on the forum, or in private, maybe with a transcript or summarized takeaways from the conversation posted back to the forum.
Firstly, I very much appreciate the grant made by the LTF Fund! On the discussion of the paper by Stephen Cave & Seán Ó hÉigeartaigh in the addenda, I just wanted to briefly say that I’d be happy to talk further about both: (a) the specific ideas/approaches in the paper mentioned, and also (b) broader questions about CFI and CSER’s work. While there are probably some fundamental differences in approach here, I also think a lot may come down to misunderstanding/lack of communication. I recognise that both CFI and CSER could probably do more to explain their goals and priorities to the EA community, and I think several others beyond myself would also be happy to engage in discussion.
I don’t think this is the right place to get into that discussion (since this is a writeup of many grants beyond my own), but I do think it could be productive to discuss elsewhere. I may well end up posting something separate on the question of how useful it is to try and “bridge” near-term and long-term AI policy issues, responding to some of Oli’s critique—I think engaging with more sceptical perspectives on this could help clarify my thinking. Anyone who would like to talk/ask questions about the goals and priorities of CFI/CSER more broadly is welcome to reach out to me about that. I think those conversations may be better had offline, but if there’s enough interest maybe we could do an AMA or something.
Neat! I’d be very interested in talking about/debating this, perhaps in the comments of another post. In particular, the sections above that feel most cruxy to me are the ones on the centrality of conceptual progress to AI strategy/policy work: what that looks like, how to figure out what new concepts are needed, or whether this is even an important part of AI policy, are all things I’d be interested to discuss.
I would definitely also be interested in talking about this, either somewhere on the forum, or in private, maybe with a transcript or summarized takeaways from the conversation posted back to the forum.