Assuming the same AIs play against each other; white should always evaluate the same first move as best, right? So, black than comes up with the best response for it which should also always be the same, and so on.
Computers are really good, but there is no, “perfect game”. No one knows exactly what move is the best move after all, there is 16 pieces and 64 squares haha. So computers arent even perfected yet.
Hello people of the past, chess AI's have a list of positions and moves that they can choose from in the start to add some variety. Also, as they search for a limited time, not depth, there is a chance they do not search the same depth and therefore find a different move.
Assuming the same AIs play against each other; white should always evaluate the same first move as best, right? So, black than comes up with the best response for it which should also always be the same, and so on.