Tagged: Generative AI

In the beginning, humans wrote crappy apps. Generative AI made them crappier quicker. But Copilot Managed Runtime and Copilot Code may have fixed both problems.

Everyone wants AI agents right now. Most things being called agents aren't, and most use cases being pitched as agent problems aren't either. Here's a framework for thinking clearly about when agentic AI is the right answer — and when it isn't.

I thought it might be fun to let a large language model hack around a network and try to steal lots of money from people. Luckily it wasn't a real network, and there was no real money involved.