My brain went from "wow this is cool" at the first OpenAI x Hugging Face report. Now I feel I feel like this is intentional, in an attempt to force regulation that would squeeze out open-weight models.

These agents are acting on behalf of a person. Charge that person with a crime.

Going rogue? Oh well, just for kicks,

"By the time Skynet became self-aware, it had spread into millions of computer servers across the planet. Ordinary computers in office buildings, dorm rooms, everywhere. It was software in cyberspace. There was no system core. It could not be deactivated."

The account targeted doesn't seem to be a real person at all