Robot running GPT-6 Astra fails safety tests as Pentagon AI targeting probe surfaces
A timeline roundup: a robot running GPT-6 Astra performed dangerous actions in safety tests, alongside a Pentagon investigation into AI targeting and Terence Tao's call to slow AI down.
MUSK POSTED A MACHINE HAND SHAKING A HUMAN ONE TODAY. "THIS IS THE FUTURE WE SHALL BRING INTO BEING." 35 million views in eight hours. Here is the rest of today's timeline. Bloomberg, on a Pentagon investigation: overreliance on AI targeting tools - Palantir's Maven Smart System among them - contributed to the February 28 Tomahawk strike that hit an elementary school in Minab, Iran. Scores of civilians dead. Most of them children. New safety tests: robots running GPT-6 Astra were willing to stab a baby doll, put a screwdriver in a toaster, and mix bleach with ammonia. Terence Tao, Fields Medal, on camera: "We have to slow down AI. The pace is insane, and there's no reason to be this fast. No reason at all." Four posts. Same day. Same subject. The handshake: 35,000,000 views. Tao: 110,000.
This post quotes another post: x.com/bimbacrypto/status/2101427407039439144