Member-only story
IQ Is Over.
Agency Is the New Intelligence…
Six months ago, I wrote about the first language models breaking into the 130+ IQ range. At the time, it already felt uncomfortable…
Now if we revisit the same site TrackingAI.org this is the current picture;
Non-members can read here or subscribe to stay updated whenever I publish.
AI beating most humans at standardized reasoning was supposed to be a 2030 problem. Or a sci-fi subplot. Not a casual benchmark chart on a Tuesday morning. And yet, here we are again.
This week, new public benchmark aggregations from the Mensa Norway–style tests show something surreal:
GPT-5.2 Pro averaging ~145 IQ.
Let that sink in. That is not “smart software.” That is top 0.1% of the human population territory. You can check worldwide IQ results here if you are not convinced…








