Jibo Regression Test – What Works (and What Doesn’t) | v1.0.18

This is a full regression test run of Jibo using my latest cloud build (v1.0.18). Not everything works yet. Actually, a lot doesn’t. That’s kind of the point. Most AI demos you see are controlled and polished. Real systems aren’t like that. They break. They drift. They behave differently when you actually try to use them end to end. In this video I walk through: what still works what’s broken what’s inconsistent and what needs to be fixed next This is part of a larger effort to bring Jibo back to life and understand what it really takes to make AI systems usable in the real world. A lot of what shows up here isn’t a model problem. It’s an orchestration problem. That’s the layer I’m focused on building with CoffeeBreak. More to come as I keep pushing this forward. ☕