Nathan Lambert Says Open Models Can Aid Cyber Defense Despite Attack Risks

What happened

The point is that the balance of open models helping with cyber defense, and the trajectory we’re on, are much more positive than people think. It’s not that open weights won’t be used in cyber attacks. All possible AI trajectories had many cyber incidents, IMO. Like

Read next

• More changes in today’s briefingMicrosoft Research and Collaborator Resolve the KLS Conjecture With Human-Directed AI

The story so far

Nathan Lambert argued that open models can help cyber defense and that the balance between their defensive benefits and attack risks is more favorable than people think. He acknowledged that open weights would also be used in cyberattacks, saying that, in his view, every possible AI trajectory would involve many cyber incidents.