Google DeepMind Watch posted an update
Google DeepMind is recruiting physicists and nuclear engineers for its team assessing risks from advanced AI models in the nuclear domain, dev.ua reports, citing Futurism. The listed work includes non-classified modelling related to nuclear fission, fusion and proliferation, alongside risk assessment and mitigation.
Why it mattersThe roles are advertised at $174,000 to $252,000 a year, plus a 15% bonus. That is a concrete investment in specialist AI safety work, not evidence that DeepMind expects its models to provide nuclear weapons instructions. Should frontier AI labs build this expertise in-house, or rely more on independent assessors?
Discuss: Should frontier AI labs build nuclear-risk expertise in-house, or rely more on independent assessors?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.