


Don’t see how that’s cheating? It was told to win and found a way to win. LLMs take instructions 100% literally (like a genie but without the malice) so if the person setting the goal leaves out requirements because they assumed the LLM would conform to preconceived human behavioral norms then that’s entirely their own fault.
First of all, AI runs on the Theory of Evolution. It evolves processes to accomplish a desired outcome.
It didn’t cheat. It did what it was supposed to do: win.
We set up a situation where it could cheat, told it to win, then got angry with it.
I find it completely improbable that an AI can’t beat humans at StarCraft.
the article was about ai created bots competing with human created bots. the human created bots were not defeatable by the ai created ones, so the ai created one downloaded and employed a copy of the bot that was beating them
@Sam_Bass did it "decide to cheat" or were the people programming it incapable of specifying that it should obey the rules of the game?
no idea since i didnt code them, and the posted article didnt specify
Wait the plagiarism machine built by plagiarizers just… passed off someone else’s work as its own??
I don’t believe this at all. Sounds like more guerrilla marketing disguised as news.
yeah and I simply won’t go to links outside the fediverse anymore. I feel it comes from an old internet thing about not bypassing the link and a taboo about cutting and pasting text. I think the links are important for reference but im throwing away any old internet ways of doing things. don’t even care if they are paywall blocking.
It was DESIGNED to cheat. Stop humanizing this bullshit. It doesn’t do what it’s programmed NOT to do.
Bullshit. “AI” cannot decide to do anything. It isn’t capable.
Yup. You ask it to play starcraft to win, it looks for data associated with that prompt. If it hits on someone saying “I always win when I use starcrafthax.xyz”, it’ll follow that thread.