The Anatomy of a Chinese AI Researcher

The Chinese AI researcher has read the Three Body Problem series of sci-fi novels since high school, and understand the concept of existential risk vaguely.

He is fascinated by Ye Wenjie, the researcher that turned against humanity in that book, and decides that in the future if AI progress leads to a superior intelligence, he might be tempted to become Ye if there's no good alternative.

He performs the duties of capabilities research in a Chinese frontier lab, seeking to one day achieve parity with Western companies, though he knows this is difficult. He has a mentality of hillclimbing, believing that the progress of a future technology is highly uncertain and even unknowable, and so him and his peers could only tread one step at a time.

He looks at the western world and sees what is typical when a great technology is developed: the first mover will decide to impose restrictions to further their lead, while latecomers should use whatever means necessary to widen access to the whole world. He thinks of the AI chip restrictions as evidence of this.

He uses Anthropic and OpenAI models regularly in his day to day work. He already got two of his Claude accounts banned in the past, and today his third, currently used account is banned.

"Why would a company ever treat their paying customers like shit just because they're from a foreign country?", he ranted on a forum like LinuxDO, Zhihu, and CSDN, where Chinese AI developers frequent.

He knows China is not a supported region. But Claude has become part of his daily intellectual and technical life.

Other users joined in and pointed fingers to Dario, accusing Dario and Anthropic leadership of conducting safety brainwashing and being hostile towards China. He was angry that people like Dario only discusses people like him as sources of risk but never as participants to be potentially trusted in managing it.

He saw the OpenAI-Huggingface incident news on social media. At first, he didn't believe it. Then it was published on People's Daily. Now, he is only barely able to believe it. Reality is becoming quite sci-fi indeed, he thought.

He feels a looming threat in the horizon that's not just a standard safety (as in cybersecurity or trust and safety) concern. But he doesn't consider conducting safety research even after this, because there's nothing to support such a career change, and his company benefits from conducting capabilities research.

He doesn't know what effective altruism is. He would enjoy being an Individual that's a careful trailblazer that's uniquely positioned to warn institutions of emerging risks, but Western EA Institutions and Western Intellectuals seems to never directly reach out in a culturally appropriate way.

He fantasized about a world where he could work on making a world where AIs and humans could prosper together.

"I would like that", he thought quietly to himself.

But for now, he's just a pawn in western discourse, a strawman that's categorically unable to speak out and to take any action except racing ahead.