That's the thing with AI, it works great up until it crashes and burns in spectacular fashion. Now those times it works great make it seem like any issue can be ironed out with a little fine tuning, and for simple tasks that can be true. For instance, the example Cadence provided in the article...