AI At A Human Level

By at least one measure, Artificial Intelligence (AI) has:

“…just achieved human-level results on a test designed to measure “general intelligence” – Ars Technica

On December 20, OpenAI’s o3 system scored 85% on the ARC-AGI benchmark, well above the previous AI best score of 55% and on par with the average human score.”

Just a few years ago, some people thought that would take a century. Others may feel like we’re not there yet. They may both be right. As AI advances, we become more aware of the nuances of what it means to ‘achieve human-level results’. As most humans know, there’s a broad range of human wrapped around ‘average human’. 

The details may matter to researchers and developers, but the measures are clear; AI is developing, and implicitly, humans are not.

Additionally, AI is developing in ways we can not readily comprehend. 

This new intelligence is developing faster than us, is not stopping, and we don’t know how it does it, how far it can go, and whether it will care about us. Stay tuned.