AI won’t only enable professional researchers to supercharge their research efforts, but it could get interested amateurs involved too.
That’s according to Terrance Tao, one of the world’s most celebrated mathematicians and a Fields Medal recipient often described as the “Mozart of Math.” In a recent discussion about the future of his field, Tao painted an unexpectedly democratic vision: artificial intelligence could finally bring “citizen mathematics” into existence, breaking down barriers that have long kept amateur enthusiasts on the sidelines of mathematical discovery.
“I think there’ll be a lot more role for the broader public in mathematical research than there is currently,” Tao observed. “There’s almost no citizen mathematics. There’s citizen science in other disciplines. In astronomy there are people who are amateurs who discover comets, and in biology there are people who identify butterflies and so forth.”
The contrast is stark. While amateur astronomers regularly contribute to professional research and citizen scientists help classify galaxies or track wildlife, mathematics has remained largely inaccessible to non-professionals. Tao identified two key reasons for this gap.
“In mathematics, it’s only the fringes that we’re able to use the amateur math community on,” he explained. “Partly because again, it’s an issue of trust. A lot of the output that they have is unfortunately not at the professional level. But also just because the technical difficulty in speaking the language, methods, language. But with both AI and these formal proof assistants, these problems could become solvable.”
Tao’s vision extends beyond just opening mathematics to amateurs. He sees AI as a bridge between disciplines entirely. “I also envisage broader collaborations with other sciences. AI also enables scientists to talk to each other. If I want to talk to a biologist, I don’t know how to express a gene or anything. But AI may provide a translator.”
If AI can serve as both a quality-control mechanism and a translator of technical language, it could democratize not just mathematics but scientific collaboration more broadly. Formal proof assistants—software that can verify mathematical proofs with absolute certainty—address the trust problem Tao mentioned, while AI language models could help bridge the communication gap between experts and enthusiasts, or between researchers in different fields.
This vision aligns with emerging trends in AI-assisted research. Several Erdos problems have reportedly been solved by AI systems, and Google and OpenAI have won gold medals at the International Math Olympiad. Tools like Claude Code, Codex and Antigravity lowering barriers to programming and technical work. If Tao’s prediction holds, we may be approaching an era where the walls between professional and amateur science become more permeable—not by lowering standards, but by providing tools that help anyone with curiosity and dedication contribute meaningfully to humanity’s knowledge frontier.