I have a slight concern that this would contribute to the vague celebrity/elitist culture that EA has. Sure, I'd find it fun/enjoyable to see people who are EA-famous on stage, but that plays into my vague perception of the haves and the have nots, elitism, and other poorly defined/not well thought-out ideas.
I'm not sure how to avoid that. Maybe design it so that none of the debaters are well-known people?
Could you share how the debate format may be more likely to play into celebrity/elitist culture in EA than a standard conference talk format?
Hmmm. I don't have anything well thought out, but my rough and off-the-top-of-my-head impressions are this:
None of these are particularly high-confidence, and I want to emphasize that these are vague impressions. I'll also add that how the debate is structured/designed has a huge affect, and could minimize or amplify various characteristics. Merely saying "a debate" really leaves a lot of room open for interpretation regarding what it will actually be. (4-10 people around a conference table in a small room? 2-4 people on stage with podiums? Time limits? Interruptions? Prepared notes? Audience votes?)
otoh I've found watching "famous" people debate helpful for taking them off the pedestal. It's demystifying to watch them argue things and think about things in public rather than just reading their more polished thoughts. The former almost always makes impressive-looking people seem less impressive.
I heard that the EAG London debate wasn’t really a debate as such, it was two talks back to back presenting different views on something. That seems not as valuable as a back and forth.
Mike Berkowitz and Ezra Klein (both have been on 80k) arguing for why US democracy is at risk and why it could be one of the most impactful things for EA to address. This is a position not popular with American EAs (including amongst some of the EAs that were at the helm of the SBF political operation--very consequential to EAs net political effect). I'm sure you could find some qualifed EAs to take the counter position.
To give a little context of the severity we are talking about: Berkowitz (as of July last year) thought there was a 50-50 chance of US democracy failing in the next 6 years.
Just read the comment about not necessarily being famous EAs. I would also be willing to debate in favor of the position that US democracy faces a serious chance (~50%) of destabilizing, and that the probability and severity of consequences necessitate that EA treat it as a major X-factor.
Here is piece I wrote arguing that position (i have an unpublished one roughly twice the length)
The best debates I've seen are by Intelligence Squared. John Donovan appears to be a very skilled moderator. Watching him cut off debaters or re-route the conversation is a pleasure, and shows the difference between an excellent moderator and a passable moderator.
I don't know all the factors that contribute to Intelligence Squared debates, but I suspect that the pre- and post-debate audience voting, choice of debaters, and good moderation contribute a lot.
My gut impression is that it would be hard to do this well for EAs because most of the interesting topics seem to be either A) things that you really need to understand the details in order to make a sound judgement on, or B) things that basically come from different judgement calls given the same facts/circumstances. But if there are good topics then I'd love to see it happen.
Very typical EA caveat: my observations are probably not representative of the EA community as a whole, and my impression could be wrong.
I have some thoughts that are more comment than answer, so I'll type those up separately.
Excited abstractly about debates as EAG programming! Though I don't have any particular ideas that aren't either too broad or too in the weeds.
I would love to see people debate the question of how difficult AI alignment really is. This has been argued before in for example the MIRI conversations and other places but more content would still be helpful for people similar to me who are uncertain about the question. Also, at the EAG events I went to, it felt like there was more content by people who had more optimistic views on alignment so would be cool to see the other side.
I'm thinking about arranging some debates at EAGxNYC in August. What questions should be debated? Who would you like to see debate? Would you like to debate? Is this a good idea?
This sounds great! I love the idea of experimenting with new formats.
I like the idea of the debate trying to be judged on different goals e.g. How truth seeking the person is trying to be, how well presented the evidence is etc rather than being right.
(Also potentially bringing back some old ones e.g. poster presentations?)
Debate topics:
Critiques of givewell methodology ( e.g. Wellbys etc) Different ai research agendas (e.g prosaic alignment vs nonprosaic alignment) Economic development vs RCT (pull in non EAs here) Maybe the case for GiveDirectly type interventions (many EAs give to them despite GW trying to find 10x more effective interventions) Maybe debates on more fringe EA causes and the case for and against them? Person affecting views Moral offsetting When to explore vs exploit in your career (maybe focused on uni students)