He added, however, that “the technology is moving fast as well”. He said the institute should put in place standards that other governments and companies can follow, such as “red teaming”, where specialists simulate misuse of an AI model, rather than take on all the work itself.
Warner said the government could find itself in a situation where it was “red teaming everything” and that a backlog could build up “where they don’t have the bandwidth to get to all the models fast enough”
Referring to the institute’s potential as an international standard setter, he said: “They can set really brilliant standards such that other governments, other companies … can red team to those standards. So it’s a much more scalable, long-term vision for how to keep these things safe.”
Leave a reply