Neel Nanda

3137Joined Nov 2019


Independent mechanistic interpretability researcher.


A similar version has been done before and this might risk duplicating it. I don't think this is the case because the debate was hard to follow and not explicitly written with the indent of finding a joint belief.

That seems like a terrible attempt at adversarial collaboration, with a bunch of name calling and not much constructive engagement (and thus mostly interesting as a sociological exercise in understanding top AI researcher opinions). I am extremely not concerned about duplicating it!

To me the main issue with this plan will be finding an AI x-risk skeptic who actually cares enough to seriously engage and do this, and is competent enough to represent the opposing position well - my prediction is that the vast majority wouldn't care enough to, and haven't engaged that much with the arguments?

I would personally prefer a separate option to allow personal blog posts but not drama-of-the-day posts - I like seeing personal blog posts, but want to avoid drama.

I was at EAG SF and didn't notice!

If we had incentivised whistle-blowers to come forward around shady things happening at FTX, would we have known about FTX fraud sooner and been less reliant on FTX funding? Very plausibly yes. She says "likely" which is obviously not particularly specific, but this would fit my definition of likely.

Why do you believe this? To me, FTX fits more in the reference class of financial firms than EA orgs, and I don't see how EA whistleblower protections would have helped FTX employees whistleblow (I believe that most FTX employees were not EAs, for example). And it seems much more likely to me that an FTX employee would be able to whistle-blow than an EA at a non-FTX org.

Also, my current best guess is that only the top 4 at FTX/Alameda knew about the fraud, and I have not come across anyone who seems like they might have been a whistleblower (I'd love to be corrected on this though!)

Thanks for sharing! How much is this aimed at highly involved EAs vs people with no EA context? (ie, should I expect there to be a bunch of time explaining things I'm already familiar with/Holden has discussed on other podcasts).

I've been really enjoying some of your other episodes, thanks for making them!

Pascal's Mug(ging) and the Scout Mindset Hat are absolutely incredible. Excited to see more designs!

Huh, I'm curious what people use this for? To me, scheduling a 1-1 is enough of a coordination problem, scheduling it for many more feels like a big headache (at least, if the people have packed schedules)

Hmm, that feels much more annoying to me - I personally think 1-1s tend to be a much better use of conference time, and being restricted from scheduling them at certain times in the app sounds irritating (and the kind of thing that gets me to bail on Swapcard and use Calendly). For me a space is good, because if someone no shows, or I have a break and want low-intensity chat, I can go there.

Plus one to this! It would be great to eg have a designated "no planned 1-1s here" area where it is socially acceptable to just approach people

The first: "I'm confused about [x]."

Could you give an example of this? I'm confused about what you mean.

Load More