In a world where artificial intelligence is rapidly reshaping the landscape of mathematics, the boundaries of what's possible are being redefined. Recently, Stanford number theorist Jared Duker Lichtman noted the intensity of our current era: "You know you're in it when you have to check the news hourly." This sentiment is echoed by Elon Musk's playful welcome to the new reality: "Welcome to the Singularity. How's the temperature?"
The heat is indeed palpable, as OpenAI's models have recently solved ten long-standing mathematical challenges, prompting number theorist Daniel Litt to acknowledge the significance of this achievement, four years ahead of his prediction. He had doubted AI's ability to produce high-caliber number theory within a $100k budget per paper, but now concedes it's "a big deal." This breakthrough has even led Fable to suggest that any single result could be worthy of a Fields Medal.
Prediction markets reflect this shift, with Manifold estimating a 31% chance of an AI solving a Millennium Prize Problem by 2027, increasing to 52% by 2028. The nature of these AI-generated proofs is as intriguing as their existence. Lichtman highlights recent advancements in sphere packing density, approaching the Cohn–Elkies threshold, as nearly "science fiction." Meanwhile, Columbia's Henry Yuen critiques the presentation style of these findings, which often mask the technical essence under layers of boilerplate text.
For mathematicians, this marks a profound transformation. As one practitioner puts it, this could be "the last straw" for traditional academic math, where specialists spend years on a single conjecture, only to be outpaced by "an amateur" AI. The emotional impact goes beyond mere professional concerns. Kirwin Hampshire describes a "dark night of mathematics," viewing discovery as a sacred human endeavor. He questions whether this transition forecloses future mathematicians' ability to experience the ineffable nature of discovery.
Cosmologist Will Kinney likens the shift to a religious upheaval, with "the old gods being slaughtered by the new machine god." Fernando Borretti critically examines the notion of "Mathematics Without Mathematicians," arguing that mathematics is intrinsic to scientific advancement, not just a game of logic. The implications extend to prestigious awards; a Fields-worthy human achievement might become trivialized by AI before the medal is even awarded.
Despite these concerns, there is a democratic upside. OpenAI's Dean W. Ball reflects on the potential for everyone to apply these breakthrough models to tackle everyday problems at a fraction of the current cost. Observers remind us that these are still "cute sub 10T models," with much more powerful successors on the horizon by 2030.
Yet, the race for AI dominance intensifies. Anthropic's Jess Yan argues that maximum performance requires integrating models with their harness, signaling that model labs may soon compete with their own customers. Technological leaps continue, with NanoGPT achieving a speedrun record and ByteDance's Seedance 2.5 offering rapid audio-video generation.
The abundance of AI capabilities is not without consequences. Apple has capped vulnerability reports after AI submissions inundated human reviewers with mixed-quality data. Meanwhile, personal finance benefits, with lifetime simulations revealing that AI-driven advice can be surprisingly effective.
As AI continues to expand, questions arise about its broader societal impact. The proliferation of AI chips is on a trajectory to double every nine months, with substantial investments projected in the coming years. This burgeoning intelligence could even aid in extraterrestrial exploration, as seen with Curiosity's recent Martian discoveries and CapuchinAI's wildlife studies.
Ultimately, some challenges still yield to human collaboration. In a heartening example, a pay-what-you-want game bundle recently raised significant funds for developers affected by automation, and police departments are leveraging true crime podcasts to crowdsource cold case solutions. As we navigate this transformative era, the balance between human ingenuity and machine intelligence will shape the future of discovery.