We're a community of UCSB students meeting weekly to question whether powerful AI systems will benefit humanity, and what we can do to help.
* and making sure what we want is good.
Modern AI systems are more capable, and improving more quickly, than most people expected a few years ago. That has moved a set of questions that used to be speculative into the present tense: how do we make sure this goes well?
The questions are partly technical, like interpretability, alignment, robustness, and evaluations. They are also partly political: governance, coordination, and what kind of oversight powerful systems should have.
We are a student group that takes these questions seriously. We read together, argue in good faith, and try to understand the arguments on their own terms. Members range from "this is the most important problem" to "the risks are overstated but interesting," and both are welcome.
You do not need a technical background to join. You do need to be willing to engage with technical ideas when they come up.
We don't always meet every week. Check the calendar below for what's actually happening, or drop into our Discord via Linktree.