Comment by rapind

I wonder if this is a case similar to chess, where LLMs kinda suck, but other models might be viable.