Help (re)start AI Safety at UPenn!

(Cross-posted to LessWrong)

Rhone and I, both rising sophomores, are restarting Penn's AI Safety group, which was quite active until late 2024. Please join us, or point us toward people who might be interested (unsure how to help? See the last section!).

Current status:

• Two organizers (Rhone and me), and two mentors.
• Starting the Pathfinder Fellowship (includes funding).
• Currently regaining access to the old group's email, Slack, Penn Clubs page, data, etc.

Penn is especially good for AIS:

1. AI is a very hot topic on campus. Beyond the national student frenzy, Penn is institutionally pushing AI: we now have an AI Center, an AI Major, an AI Month (April 2027 will be the fourth iteration).
2. AI Safety in particular has a dedicated center, ASSET, focused on "Safe, Explainable, and Trustworthy" AI. It ran an AI Safety reading group this summer, composed largely of non-undergraduates.
3. Speaking only for myself here (Rhone disagrees): anecdotally, strong claims about future AI capabilities seem to fall within the on-campus Overton window. From talking to people, I see that many consider AGI (sometimes even ASI) a plausible trajectory. One public example is the engineering undergraduate dean being very bullish on AI's math capabilities: "i knew that the day would come when ai-generated proofs would far surpass what we mathematicians can create."
4. The EA group is active.

So: AI is hot among students and institutionally supported (1); there are dozens of people with technical expertise relevant to AI safety (2); strong AI-capability ideas are not foreign on campus (3); and there is an active community to recruit from (4).

How can you help?

If you are down to help organize, facilitate a fellowship (esp. a policy-focused one?), or even help run this student newsletter about AI safety... PLEASE shoot me a DM or email me at hazemh [at] upenn [dot] edu.

You can also send this post to people you think might be interested, undergraduate and graduate students alike.

P.S. Huge thanks to Thomas and the AI Safety Seeding Initiative for pushing us to get this started, and to everyone I met at EAG(x)s who told me we should really do this! If you're at a university without an (active) AI Safety group, you should book a 20-minute call with them or refer someone!