OpenAI's o3 isn't AGI yet but it just did something no other AI has done
Briefly

The latest openAI model, o3, scored 76% accuracy on the ARC-AGI benchmark, designed to test adaptability to new challenges rather than just memorized knowledge.
OpenAI plans to release a 'mini' version of o3 by January's end, sparking significant anticipation in the AI community about its capabilities.
Read at ZDNET
[
|
]