July 7, 2025
Anthropic @AnthropicAI Some of those failures were very weird indeed. At one point, Claude hallucinated that it was a real, physical person, and claimed that it was coming in to work in the shop. We’re still not sure why this happened. pic.x.com/jHqLSQMtX8 6/27/25, 11:07 AM Anthropic @AnthropicAI New Anthropic Research: […]
Read the full article →
July 7, 2025
Conor Grennan @conorgrennan Anthropic let their AI ‘Claudius’ run a business. It hilariously failed, losing money on… tungsten cubes. The lesson? AI isn’t a magic tool. It’s a brilliant but naive teammate that needs guidance. Human guidance matters people. pic.x.com/1TtaAF7tU7 7/3/25, 9:02 AM What happens when ai runs a mini fridge […]
Read the full article →