All of Max Nadeau's Comments + Replies
Disclaimer: I joined OP two weeks ago in the Program Associate role on the Technical AI Safety team. I'm leaving some comments describing questions I wanted to know to assess whether I should take the job (which, obviously, I ended up doing).
What sorts of personal/career development does the PA role provide? What are the pros and cons of this path over e.g. technical research (which has relatively clear professional development in the form of published papers, academic degrees, high-status job titles that bring public credibility)?
For me personally, research and then grantmaking at Open Phil has been excellent for my career development, and it's pretty implausible that grad school in ML or CS, or an ML engineering role at an AI company, or any other path I can easily think of, would have been comparably useful.
If I had pursued an academic path, then assuming I was successful on that path, I would be in my first or maybe second year as an assistant professor right about now (or maybe I'd just be starting to apply for such a role). Instead, at Open Phil, I wrote less-academic re...
Disclaimer: I joined OP two weeks ago in the Program Associate role on the Technical AI Safety team. I'm leaving some comments describing questions I wanted to know to assess whether I should take the job (which, obviously, I ended up doing).
How inclined are you/would the OP grantmaking strategy be towards technical research with theories of impact that aren’t “researcher discovers technique that makes the AI internally pursue human values” -> “labs adopt this technique”. Some examples of other theories of change that technical research might have:
- Provi
Disclaimer: I joined OP two weeks ago in the Program Associate role on the Technical AI Safety team. I'm leaving some comments describing questions I wanted to know to assess whether I should take the job (which, obviously, I ended up doing).
How much do the roles on the TAIS team involve engagement with technical topics? How do the depth and breadth of “keeping up with” AI safety research compare to being an AI safety researcher?
Disclaimer: I joined OP two weeks ago in the Program Associate role on the Technical AI Safety team. I'm leaving some comments describing questions I wanted to know to assess whether I should take the job (which, obviously, I ended up doing).
What does OP’s TAIS funding go to? Don’t professors’ salaries already get paid by their universities? Can (or can't) PhD students in AI get no-strings-attached funding (at least, can PhD students at prestigious universities)?
Disclaimer: I joined OP two weeks ago in the Program Associate role on the Technical AI Safety team. I'm leaving some comments describing questions I wanted to know to assess whether I should take the job (which, obviously, I ended up doing).
Is it way easier for researchers to do AI safety research within AI scaling labs (due to: more capable/diverse AI models, easier access to them (i.e. no rate limits/usage caps), better infra for running experiments, maybe some network effects from the other researchers at those labs, not having to deal with all the log...
Sampled from my areas of personal interest, and not intended to be at all thorough or comprehensive:
AI researchers (in no particular order):
- Prof. Jacob Steinhardt: author of multiple fascinating pieces on forecasting AI progress and contributor/research lead on numerous AI safety-relevant papers.
- Dan Hendrycks: director of the multi-faceted and hard-to-summarize research and field-building non-profit CAIS.
- Prof. Sam Bowman: has worked on many varieties of AI safety research at Anthropic and NYU
- Ethan Perez: researcher doing fascinating work to display and add
Best of luck with your new gig; excited to hear about it! Also, I really appreciate the honesty and specificity in this post.
From the post: "We plan to have some researchers arrive early, with some people starting as soon as possible. The majority of researchers will likely participate during the months of December and/or January."
Artir Kel (aka José Luis Ricón Fernández de la Puente) at Nintil wrote an essay broadly sympathetic to AI risk scenarios but doubtful of a particular step in the power-seeking stories Cotra, Gwern, and others have told. In particular, he has a hard time believing that a scaled-up version of present systems (e.g. Gato) would learn facts about itself (e.g. that it is an AI in a training process, what its trainers motivations would be, etc) and incorporate those facts into its planning (Cotra calls this "situational awareness"). Some AI safety researchers I'v...
It is possible but unlikely that such a person would be a TA. Someone with little prior ML experience would be a better fit as a participant.
We intended that sentence to be read as: "In addition to people who plan on doing technical alignment, MLAB can be valuable to other sorts of people (e.g. theoretical researchers)".
I'd love to hear his thoughts on defensive measures for "fuzzier" threats from advanced AI, e.g. manipulation, persuasion, "distortion of epistemics", etc. Since it seems difficult to delineate when these sorts of harms are occuring (as opposed to benign forms of advertising/rhetoric/expression), it seems hard to construct defenses.
This is a related concept mechanisms for collective epistemics like prediction markets or community notes, which Vitalik praises here. But the harms from manipulation are broader, and could route through "superstimuli", addictiv... (read more)