Getty / Futurism One of the core promises of the AI industry is that its tools can automate work for you and perform entire tasks without intervention. Oftentimes, however, the stuff that today’s AI will do on your behalf is less than helpful — or even so comically harmful that you’d be better off just doing a task yourself. A developer learned this lesson the hard way when using Anthropic’s new Claude Cowork model, which was seen as being so potentially d