Someone posted this x post but erroneously said it was about arcade games when the entire post was actually complaining about how useless it is to test these frontier models on the same arcade games. Instead this post runs a one-shot test to determine how well Grok 3 fared against other frontier models in creating a 3D game with room for it to come up with gameplay and aesthetics. Tested: Grok 3, O1, Sonnet 3,5, Llama4, DeepSeek, and Gemini using the following prompt.
https://x.com/lishali88/status/1893197683185328188