It trains based off the mistakes it made with the solution you gave OR what it got. Thank you for reading Comments URL: https://news.ycombinator.com/item?id=49139461 Points: 4 # Comments: 0