I've tried a similar experiment with ChatGPT at the end of last year [0]. The solution wasn't all bad, but contained a subtle error in the initialization of th

I've tried a similar experiment with ChatGPT at the end of last year [0]. The solution wasn't all bad, but contained a subtle error in the initialization of the board. Pointing it out resulted in multiple incorrect fixes. I'm wondering if tic-tac-toe is complex enough to be a good example. You can compute all possible path through the game ahead of time and make a printed book [2] that plays an optimal strategy... [0] http://holzer.online/articles/chatgpt-tic-tac-toe/ http://holzer.online/articles/chatgpt-tic-tac-toe/ [1] https://www.goodreads.com/book/show/18490871-tic-tac-tome https://www.goodreads.com/book/show/18490871-tic-tac-tome