Isn't that *exactly* the type of solution you'd expect from AI?
Move 37 comes to mind.
Only if you understand nothing about the difference between LLMs and AlphaGo.