Yes, the local LLM model has been pretty decent at making decisions and definitely moves the needle in the simulation and CoreAI should be a great comparison running on the other team. What surprised me was that FM just STOPPED working a few weeks back when the game classification restricted that forcing me to test as an entertainment app now. Any workaround for that so I could ship with LLM as a game on os26?

J, I am finishing up an AI version of my simulation game and would like to use FM for decisions and to train my CoreAI implementation but as we know, those are not a valid combo. Sounds like 27 will allow us to use our own model to work around this right? Best, Robert

Same issue here. Confirmed AppKit issue with storyboards in Tahoe.

This is still an issue. iCloud subscription valid but services/tokens not refreshed properly .