@pozsgaybalazs

Recently Sol made significant progress in one of my research projects. The idea is simple... yet I don't think it would have crossed my mind. Earlier openAI models never had this idea, in fact, they produced completely mistaken proof attempts. So strange, it needs an immense computing power to come up with something really simple. This feels like creativity. Many human innovations in computations seem simple once they are there... but it took time until someone found them. Now I am digesting the computation, and I am instructing Sol Ultra to solve the most general case. It is not there yet. This is about brickwork quantum circuits with "Yang-Baxter gates". I am learning a lot. The experience is like a strange combination of being a PI and a student. A new experience for sure. How will the future look like? Everyone is talking about this. No idea. This is fun for the moment. But I don't expect it will last very long. --- ##### Comments > **NeuSatz @NeuSatz1sw** · [2026-08-15](https://x.com/NeuSatz1sw/status/2088588412077130228) > > May I ask what do you mean by “I don’t expect it will last very long”? > > > **Balázs Pozsgay @pozsgaybalazs** · [2026-08-15](https://x.com/pozsgaybalazs/status/2088646611467411922) > > > > Right now humans are better in chosing the questions, and AI helps with problem solving. However, soon AI could become better also in asking the right questions, and the comes the era of autonomous AI research. Then we will become spectators and interpreters.