@Austin Thanks for clarifying!

Fwiw, without much context on the finances here, I'd be pro not having a 5% fee on smaller donors - my intuition is that it won't raise much money, and the discouragement of a fee would have an outsized effect on reducing donations. While larger donors are more likely to be rational about this kind of thing, and also to be able to judge how much value Manifund is actually adding to the ecosystem.

this looks like a great project!

Also benefited greatly from an 80k advisor call!

@NeelNanda The latter - PauseAI US has signed a specific fiscal sponsorship agreement where all donations made to them are structured as donations to our 501c3, and they pay the fiscal sponsorship fee on that.

Other grantees on Manifund do not pay any fee to our platform, as of today. Instead, we often ask large donors we work with (including the one funding your regranting budget) to pay an extra 5% to help cover Manifund operational costs; smaller donors (generally <50k in total, paying with credit card) we serve at no fee.

(I've been considering changing our structure so that all donations made through Manifund include the 5% fee, to simplify our offering/align incentives/make our costs clear; if we do this we'd announce it and make it very clear in our UI)

@Austin To clarify, do all Manifund grantees pay a 5% fee on donations via Manifund? Or just the grantees who are also fiscally sponsored by Manifund, and they pay it regardless of whether donations come via Manifund or elsewhere?

We’re planning to relaunch the site soon and we’re running a pilot test on Friday at 8pm GMT (3pm ET, 12pm PT).

You can access the new site here: 3p.turingtest.live. The pilot will be accessible here on Friday.

Any feedback you can provide about how the site and your interactions work would be greatly appreciated!

If you have any questions or comments, please let me know at cameron@ucsd.edu.

Approving this grant! I appreciate that the grantees and Neel called out the potential conflict of interest; Manifund is generally happy to have regrantors fund researchers who they know or have mentored, so long as the regrantor does not gain significant private benefits through this grant.

Love that you're doing this. Thank you for putting in so much work and time already! Love that you're aiming for a balanced presentation--I think we'll all learn more that way.

I'm fully funding this. I think that understanding, detecting and potentially mitigating chain of thought unfaithfulness is a very important problem, especially with the rise of o1 models. I think the approach taken here is reasonable. I think Arthur is fairly good at supervising projects, and that under him Jett and Ivan have a decent shot of making progress, and that enabling this to start a month earlier is clearly a good idea. I know Arthur well, and got a moderate amount of data on Jett and Ivan's ability to execute on this project in my MATS training program.

Conflicts of interest: Jett and Ivan were in my MATS training program (but no longer are, and I would not be involved in this project going forwards). I manage Arthur and he works on my interpretability team at GDM (but he is doing MATS in a purely personal capacity, and if anything I'm incentivised to reduce his amount of extra-curricular distractions :P ). Overall I don't feel particularly concerned about conflicts of interest here.

@Greg_Colbourn Thank you so much, Greg!

(This was 1 Bitcoin btw. Austin helped me with the process of routing it to Manifund, allowing me to donate ~32% more, factoring in avoiding capital gains tax in the UK).

I've been impressed with both Holly and Pause AI US, and Joep and Pause AI Global, and intend to donate a similar amount to Pause AI Global.

It's more important than ever that PauseAI is funded. Pretty much the only way we're going to survive the next 5-10 years is by such efforts being successful to the point of getting a global moratorium on further AGI/ASI development. There's no point being rich when the world ends. I encourage others with 7 figures or more of net worth to donate similar amounts. And I'm disappointed that all the big funders in the AI Safety space are still overwhelmingly focused on Alignment/Safety/Control when it seems pretty clear that those aren't going to save us in time (if ever), given the lack of even theoretical progress, let alone practical implementation.

New website for the fund that listed the 25 admitted students' profile and the current graduate program they are attending

https://gradapplicationhelpfund.com/our-services/

@mdickens Thanks for your comment! I realized this project page didn't mention that I will discuss the risks of rapid AI-accelerated AI progress (and am generally spooked by the possibility) so I've updated some sections to reflect that.

That being said, I don't think this video will speed up AI-accelerated AI R&D. Leading AI labs are already aware of and actively pursuing this direction (see Sam Altman's talking about this, Anthropic's RSP and head of RSP talking about it, Google's research on chip design, etc.). Also, other researchers who are closer to AI labs have already published detailed technical analyses of this dynamic (see Epoch paper on this, Situational Awareness section on intelligence explosion, Ajeya article) .

We're kind of in the worst possible world where frontier labs know about and are pursuing this possibility, but the public and policymakers don't understand it or take it seriously enough. This video aims to make this topic more accessible to the people outside of labs who can take steps to prepare for or mitigate the possibility of very fast AI progress—I don't think it'll reveal anything new to labs.

What is the case for impact for this? At a glance my guess is it would increase x-risk—a better explanation of AI accelerating AI R&D would (marginally) make people want to do that more, which shortens timelines / increases the probability of a fast takeoff.

Manifund has now received @cfalls's $10k donation for your project and added it to this page!

@NeelNanda Yes, I think so! ~$5k is good enough for the first semester, but I think I could productively use up to $25k total over the space of 3 semesters. Also, if I had $25k for 3 semesters, I could probably hire some paid part-time help, to help with faster growth (reaching many universities simultaneously etc.)

Obviously, anything between what I have now and $25k total is fine too.

I like the proposal! Do you think you could productively use any more money?

I encouraged Tom to apply on Manifund after first seeing this proposal on Nonlinear Network. As I've previously written:

I don't know Tom, but a couple weeks ago I spoke with Lucie on the importance and neglectedness of AI governance in France. I hadn't realized that the next version of the AI Safety Summit was going to happen in France; this seems like a great reason to invest in the folks doing on-the-ground work there. In that conversation, Lucie made a strong endorsement of Tom's prior work; here I'm delegating my trust to Lucie.

This still holds true, in that my donation is not significantly informed by an inside view on Tom's work, but this seems reasonable to me for a cheap and time-sensitive travel grant.

@Tomohaire replied above