It is claimed that OpenAI’s ChatGPT 4o model “got absolutely wrecked on the beginner level” of Atari Chess on an Atari 2600 ...

A cycle-exact emulation of the Atari 2600 CPU running at a meager 1.19 MHz is more than enough to utterly humiliate ChatGPT in a game of ...

The Arati 2600 from the late 1970s is not exactly cutting-edge hardware, but it's apparently powerful enough to beat ChatGPT ...

But Rozman was clearly troubled by its performance. “Not only can ChatGPT play chess now,” Rozman said, “it nearly defeated me.” Ironically, it may very well be Rozman—and his many ...

Now, a more recent experiment shows that ChatGPT o1 decided to hack a chess game on its own, without being explicitly told to do so, just to beat an AI opponent that was a stronger player.

Throughout the evaluation, ChatGPT o1 demonstrated a deep understanding of the tasks at hand and the underlying principles governing them. The model’s performance was particularly notable in the ...

OpenAI’s problem here, as Goedecke points out, isn’t that ChatGPT turned into a yes-man. It’s that its performance became too obvious. This is a big deal. AI discourse tends to focus on ...

o3 and o4-mini are out for Pro, Plus and Team users now, free users can try o4-mini, too They can combine and use every tool in ChatGPT's arsenal o3 and o4-mini add reasoning using images to ...