The AI war
Sam Altman's answer to Gemini 3!
Thought you'd seen it all?
OpenAI breaks the mold and releases GPT-5.2 Pro.
A model that completely changes the rules of the game in the technology market and the global economy.
Superhuman genius: 147 IQ points
We are talking about a cognitive level that is 3 standard deviations above the average.
Fewer than 1 in 1,000 people (the top 0.1%!) reach such a level of intelligence.
This machine is smarter than most people you know – and it's empirically proven.
The data shaking the industry comes from the GDPval index (the most brutal index for evaluating business output).
For the first time in history, an AI model defeats human experts in a knockout:
GPT-5.2 Thinking model: a 70.9% win rate against experts.
Claude Opus 4.5 model: stands at 59.6%.
Gemini 3 Pro model: stands at 53.5%.
GPT-5 model (previous generation) – only 38.8% (intern level).
The meaning?
Within one generation, we moved from a "talented intern" to a "senior manager" who knows how to make complex decisions.
GPT-5.2 changes the return on investment equation from end to end:
It is 11 times faster than a human expert, and costs less than 1% of their salary.
Think about it for a moment...
A financial model that takes an analyst three working days?
The model finishes in minutes – at the cost of a cup of coffee.
And perhaps the most important point – the end of the era of hallucinations?
In the prestigious AIME 2025 mathematics test, the model achieved a perfect score of 100.0%.
This is a game-changer for industries where accuracy is critical – financial engineering, engineering design, and medicine.
The technology is mature, accurate, and ready for immediate implementation.
The business world is about to change at a speed we don't know.
The question is not "whether to use it," but "who will adopt it first and gain the competitive advantage."
What will Google's response be?
We are following and will update...