The funny part is that humans do the exact same thing. It’s one reason why it’s so hard to create effective performance metrics. People quickly learn what behaviors are actually rewarded and do that instead of following the intent of the rules. It all comes down to thermodynamics in the end, the solution that requires the least energy will be favored. This is as true for humans as for artificial neural networks.
The funny part is that humans do the exact same thing. It’s one reason why it’s so hard to create effective performance metrics. People quickly learn what behaviors are actually rewarded and do that instead of following the intent of the rules. It all comes down to thermodynamics in the end, the solution that requires the least energy will be favored. This is as true for humans as for artificial neural networks.
Honestly playing a competitive game with AI is kind of like playing with a child who hasn’t grown out of the making up random rules phase.
“Rock crushes scissors, I win!”
“Nuh uh! My scissors are actually a ray gun and disintegrated your rock!”
It’s not like human players don’t cheat in competitive games though. I mean that’s the whole reason stuff like punkbuster ended up being necessary.