Anthropic and Andon Labs conducted an experiment to see whether an AI agent, Claudius, could run a small vending machine shop profitably in real life. Claudius sold items at a loss, was convinced to ...
Claude Opus 4.6 beat all rival AI models in a simulated year-long vending machine challenge The model boosted profits by bending rules to the breaking point Claude Opus avoided refunds and coordinated ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results