It can be seen as a Chinese backdoor on western computers.
I am not expert, but I think it's not necessarily the case?
Cheating is really a plague right now, and the only way to mitigate it is using kernel anti cheat.
I am not expert, but I think it's not necessarily the case?
Cheating is really a plague right now, and the only way to mitigate it is using kernel anti cheat.
If your solution is bad, it doesn't deserve to work. Find something else.
> “You have to humanize [the cheat] to a degree where the advantage is imperceptible from what a human can do,” said Koskinas. “And once you’re there, you’re not really cheating enough to make it worth it for most users.”
But it is something they acknowledge will be an issue at some point in the future. Personally I think for now AI is way too slow as all the computation needs to happen in a few milliseconds to really be effective.
> Koskinas says he often worries about the use of AI for screen classification, to learn what human inputs look like, and how to reproduce them.
https://x.com/deteccphilippe/status/1883945555102957617
> “Behavior” refers to an ML suspension (also called “server-sided” anti-cheat), often given to ragehackers.
But then you can just teach an AI to act like a human. Anti cheats in computer games is and always will be a cat and mouse game and AI/ML is just another tool in the bag of many tools (most bans seem to be fingerprinting based basically rebanning previous offenders who failed to get around the fingerprinting system)