The case against leaving AI safety systems solely to frontier labs
The Information says Superhuman CEO Shishir Mehrotra questioned trusting toolmakers to handle all of their own safety.
TLDR
The Information says frontier AI labs shouldn't be solely responsible for creating their own safety systems. It quotes Superhuman CEO Shishir Mehrotra: “You don't sort of trust the toolmaker to come up with all of their own safety, right?”
Combined views
10.1K
1 Source, first seen ago
7 likes4 comments1 reposts