Source: Quantum Bit
The prodigy from the University of Science and Technology of China’s Young Talent Program left OpenAI; the youngest Chinese full professor in Harvard’s history, after only five months.
OpenAI, you can't hold onto people either, I see.
Just now, the youngest Chinese tenured professor in Harvard's history, a recipient of the Sloan Prize and the New Horizons in Physics Prize, and a star scholar in the field of string theory—
Yin Xi has officially left OpenAI.
From the rumor of joining at the end of May this year to personally announcing departure, it has only been about half a year～
Regarding this departure, he stated on his personal Twitter:
This has been an amazing journey, but I’m planning to get off early and explore new directions.
Yin Xi also revealed that this is his first corporate job.
He especially thanked Alex Lupsasca and Boris Alexeev for helping him adapt to the role, and Dan Roberts for his ongoing support. He ended with an encouraging note to his colleagues: “I’ll keep cheering you on!”
As for the next step everyone is most curious about, Yin Xi has not yet revealed it. On the official website of Harvard’s Physics Department, his professorship is still listed.
Perhaps, maybe, it’s really just a sabbatical to experience life.
Of course, some netizens directly named Discovery Loop, the AI company founded by former Google chief scientist Jeff Dean and others.
Where exactly to go? Let the bullets fly for a bit longer.
Who is Yin Xi?
Yin Xi entered the University of Science and Technology of China’s gifted youth program at age 12, earned his Ph.D. in physics from Harvard University in 2006, and was promoted to full professor at Harvard in 2015 at the age of 31, earning him the nickname “Harvard’s youngest Chinese full professor” from many media outlets.
He has long researched string theory, quantum field theory, and quantum gravity, receiving the Sloan Research Fellowship in 2013, and later, together with Simone Giombi, winning the 2017 Physics Frontiers Prize for their work on higher-spin gravity and its holographic correspondence.
Beyond an extraordinary genius background, this physicist also holds a very clear stance on AI.
In April this year, before the news of his affiliation was announced, Yin Xi told The Harvard Crimson directly:
I don't believe there is any human intellectual ability that AI cannot replicate.
This optimism about technology stems in part from his experience using AI.
Yin Xi estimates that, even with significant human oversight, AI has accelerated his research coding work by at least 100 times—
The code that AI wrote in a few weeks would take him ten years to write himself.
Some research projects that were previously too labor-intensive to even consider are now feasible to try.
By September this year, the academic community had once again sparked a debate over what AI would ultimately bring to mathematics.
Terence Tao and 25 other Fields Medalists have signed a joint statement expressing concern that AI companies are treating the solving of mathematical problems as a competition metric: while answers are increasing, the processes of understanding, organizing, and transmitting these achievements are falling behind.
Yin Xie clearly sided with those refuting this view.
He shared mathematician Timothy Nguyen's response and said:
This is the viewpoint I fully agree with and support among all the responses I've seen in the past few days.
Nguyen opened by directly stating: these Fields Medalists' objections to AI will not stand the test of time.
His core point is that the way we understand mathematics has always been changing—from geometric constructions and symbolic computation to computer-assisted proofs, mathematics has continually expanded its toolkit.
Verified machine proofs, even if temporarily beyond the understanding of any single individual, can serve as a foundation for further research. If AI makes previously difficult problems easier, humans will continue to explore even farther.
Looking at it this way, although he is a human physicist, Yin Xi is not fixated on placing humans at the center of scientific discovery; as long as the answer stands up to verification, he is happy to see AI take the lead.
Basic scientists join model companies
When Yin Xi previously joined OpenAI, he was already surrounded by a group of top experts in mathematics and physics.
He particularly thanked black hole physicist Alex Lupsasca in his farewell post, who had already announced his joining of OpenAI for Science in October 2025.
Lupsasca said at the time: This team aims to build AI systems that advance scientific reasoning and accelerate discoveries in mathematics and physics.
In addition, alongside the news of Yin Xi joining, attention has also been drawn to Su Weijie, one of the “golden generation” members of Peking University’s School of Mathematical Sciences.
In May of this year, Su Weijie announced that he joined OpenAI on academic leave from the Wharton School to participate in model training. At the same time, he was officially promoted to full professor at the Wharton School.
Earlier, at the beginning of this year, Chen Lijie, a Tsinghua Yao Class alumnus and theoretical computer scientist, also joined OpenAI, focusing on mathematical reasoning.
Scientists bring research questions and the expertise to evaluate the quality of answers; model companies provide models, computing power, and engineering environments.
The two sides came together with a clear goal: to have AI truly participate in scientific discovery.
As for Yi Xi's views on the relationship between AI and researchers, he gave a very clear answer in that April interview:
As long as the AI can find an answer and there’s a way to verify it, he’ll be very happy. Whether the answer was generated on his own is secondary.
Twitter: https://twitter.com/BitpushNewsCN
BitPush Telegram community: https://t.me/BitPushCommunity
BitPush TG subscription: https://t.me/bitpush