AI Lab Watch

Zach has consistently been one of those on top of the safety and security plans, the model cards and other actions of the major labs, both writing up detailed feedback from a skeptical perspective and also compiling the website and its scores in various domains. Zach is definitely in the ‘demand high standards that would actually work and treat everything with skepticism’ school of all this, which I feel is appropriate, and I’ve gotten substantial benefit of his work several times.

However, due to uncertainty about whether this is the best thing for him to work on, and thus not being confident he will have this ball, Zach is not currently accepting funding, but would like people who are interested in donations to contact him via Intercom on the AI Lab Watch website.