News for hackers. Stuff that matters.
Re:ARC-AGI without pretraining by butlike on Wednesday March 05, 2025 @10:47AM (#43268135)
So you're saying AI can be incompetent at a grander scale. Got it.
Re:ARC-AGI without pretraining by FrustratedMonky on Wednesday March 05, 2025 @01:15PM (#43270174)
Yes. It can be as good and bad as a human. Humans also make up BS to answers.