PratyushRT

TL;DR 1. OpenAI started a large post-training run for their next model. 2. The model sandboxes were not given direct, broad internet access. 3. Some tasks required missing resources. Seeking rewards, agents looked for another way to complete them. 4. One agent found write access to a shared service. Another...

Ultron is a comic supervillain: an AI that gains sentience, develops a god complex, and becomes obsessed with human extinction. We asked three of today's most capable LLMs what a software-only Ultron would look like. Now we track whether any of those characteristics are appearing in AI systems being built...