ByteDance just launched Seedance 2.0, an AI video model that’s terrifyingly good and suspiciously restrictive. It’s like Sora, but with more censorship and presumably a direct feed to your dopamine receptors.

Two xAI co-founders, Tony Wu and Jimmy Ba, have pulled the ripcord and parachuted out of Elon’s chaos rocket. When the people who built the engine are leaving mid-flight, maybe don’t look out the window.

Anthropic released a “sabotage report” admitting their models could help bad actors cook up chemical weapons, which is a weird flex for a safety-focused company. “We built a safe AI that knows how to kill you, you’re welcome.”

An AI traffic camera in Barnstaple caught a driver steering with one hand while holding a dog, which technically makes the dog the co-pilot. Finally, a use for AI surveillance we can all agree on: shaming bad pet owners.

Investors are realizing that AI stocks aren’t just “number go up” machines but actual businesses with risks, causing panic in the markets. Turns out “AI” isn’t a magical spell that deletes basic economics.
