Terence Tao has warned that artificial intelligence is now draining the pool of open mathematical problems before researchers can fully explore them. He notes that merely mentioning a difficult problem often triggers a flood of AI-generated attempts to solve it instantly. This rush flattens the work before human mathematicians have time to develop the necessary insights. Tao suggests that these incentives might soon force scholars to withhold promising directions from the wider community. Such secrecy would break centuries of tradition in open science and harm the long-term progress of the field.
The core issue is a shift in how research value is measured and consumed. When machines can quickly produce partial solutions, the motivation to share deep ideas diminishes. This dynamic threatens to isolate top researchers who must now guard their own discoveries. The potential loss of shared knowledge could slow down progress in pure mathematics significantly.
* AI tools flatten problems before human insight can mature
* Incentives may drive researchers to hide promising directions
* Open science traditions face serious long-term damage