Contents
2012 – THE MAYA CALENDAR AND THE “END OF TIME”
Background
The Maya civilization, which flourished in southern Mexico, Guatemala, and Belize from about 2000 BC to 1000 AD, had a system of three calendars (fully described for the cover picture of Astronomical Calendar 2010):
- The Long Count, or numbering of days from (probably) 3114 BC Sep. 7: 20 kin (days) make a winal; 18 winal, a tun (360 days); 20 tun, a katun; 20 katun, a baktun (144,000 days, about 394 years). Even larger units were not actually used in the expression of dates: 20 baktun make a piktun (2,880,000 days); 20 piktun, a kalabtun; 20 kalabtun, a kinchiltun; 20 kinchiltun, an alautun (23,040,000,000 days, or around 63 million years).
- Haab, in which a 365-day year was divided into 18 20-day units with names (Pop, Wo, Zip, Zotz, Tzek, Xul, Yaxkin, Mol, Chen, Yax, Zak, Keh, Mak, Kankin, Muwan, Pax, Kayab, Kunku) and a 5-day 19th (Wayeb).
- Tzolkin, in which each day had one of 13 numbers, and also one of 20 names (Imix, Ik, Akbal, Kan, Chikchan, Kimi, Manik, Lamat, Muluk, Ok, Chuwen, Eb, Ben, Ix, Men, Kib, Kaban, Etznab, Kawak, Ahau), so that any one combination, such as 1 Imix, came around after a cycle of 260 days.
Some of these numberings started not with 1 but with zero; thus katun 19 is really the 20th in its baktun, and is followed by katun 0, which is the first in the next baktun.
The Scare
Chances are you have heard that the Maya predicted the end of the world on December 21, 2012. This was the day when the Maya Long Count calendar cycle came to completion. You may have also heard that the world was predicted to be destroyed by an earthly or cosmic catastrophe. We know these predictions didn’t come true – but were any based on fact, or total fiction?
December 21, 2012 marked the end of an important cycle in the Maya Long Count calendar. This cycle is composed of 13 periods, called baktun, of 144,000 days each. When the 13 baktuns are grouped together, they form a Great Cycle that lasts approximately 5,125 years. The Maya used a sophisticated system called the Long Count to track these massive stretches of time.
- This 13-baktun cycle began on the Long Count calendar date 13.0.0.0.0 4 Ajaw 8 Kumk’u, and spans 5,125.366 solar years. Stela C (left) records this date, considered by the Maya to be the creation date of the current or 4th era.
- The monument is at the archaeological site of Quiriguá, Guatemala.
- This creation date corresponds to August 11, 3114 BCE. Monument 6 (right), from the archaeological site of Tortuguero in Tabasco, México, records the only known Maya inscription of the end date of the 13-baktun cycle.
- This end date, 13.0.0.0.0 4 Ajaw 3 Kank’in, corresponds to December 21, 2012.
- If you want to find out what any date in our time would be in the Maya calendar, simply type in your chosen date and click the convert button using the Date Conversion Calculator.
There is no evidence in these inscriptions, or in any other record, that the ancient Maya thought that the Long Count calendar would imply some kind of catastrophic “end.” These predictions were unfounded and are not shared by the Maya people.
What does this mean? The Long Count rolls into “a string of zeros, like the odometer turning over on your car,” as E.C. Krupp put it in his amusing article about “The Great 2012 Scare,” for the November 2009 issue of Sky & Telescope. This is the end of one baktun and the beginning of another; so its component katuns, tuns, winals and kins all start again at 0. According to some authorities, the first baktun was 0, and the baktun now ending is 12; according to others it is 13. Either way, it is the 13th since the start-date of the Long Count.
References
“2012 Phenomenon.” Wikipedia, February 5, 2026. https://en.wikipedia.org/wiki/2012_phenomenon.
Davies, Diane. “Maya Mathematics.” April 6, 2020. https://www.mayaarchaeologist.co.uk/public-resources/maya-world/maya-mathematics-resources/.
Davies, Diane. “The Maya Calendar Explained.” December 31, 2016. https://www.mayaarchaeologist.co.uk/public-resources/maya-world/maya-calendar-system/.
Grandon, Brittny, and Betty Morfin, Vincent Trang and Jackilynn Sterba. “Mayan Science.” SlideServe. Accessed February 7, 2026. https://www.slideserve.com/kolton/mayan-science.
Johnson, Ryan, and Cristen Conger. “How the Mayan Calendar Works.” HowStuffWorks, December 27, 2007. https://people.howstuffworks.com/mayan-calendar.htm.
Living Maya Time. “Maya Calendar Converter.” Accessed February 7, 2026. https://maya.nmai.si.edu/calendar/maya-calendar-converter.
Living Maya Time. “The Meaning of 2012.” Accessed February 7, 2026. https://maya.nmai.si.edu/2012-resetting-count/meaning-of-2012.
Maya Decipherment. “Bak’tuns and More Bak’tuns.” December 19, 2012. https://mayadecipherment.com/2012/12/19/baktuns-and-more-baktuns/.
Universal Workshop. “THE MAYA CALENDAR AND THE ‘END OF TIME.’” September 22, 2017. https://www.universalworkshop.com/maya-calendar-end-time/.
Notes
4th Era: In Maya mythology, we are living in the Fourth Creation of the Universe, which began on August 11, 3114 BCE.
Baktun: One of the cycles of the Maya Long Count calendar containing 144,000 days.
Creation Date: The creation date of the Maya Long Count calendar, corresponding to 11 August 3114 BCE in the Gregorian calendar.
Great Cycle: Each great cycle lasted 5128 years and it repeated indefinitely. The first great cycle was to end on 21 December 2012. This led to the popular idea that the Maya prophesied the world was to end on that date. However, this is completely a modern invention, time was not lineal for the Maya, but cyclical and ever repeating.
- The Building Blocks of the “Great Cycle”
- The Long Count is based on a “vigesimal” (base-20) system. Each unit is a multiple of the one before it:
- K’in: 1 day.
- Winal: 20 days.
- Tun: 360 days (approx. 1 solar year).
- K’atun: 7,200 days (approx. 20 years).
- B’ak’tun: 144,000 days (approx. 394 years).
- The “Great Cycle” (13 B’ak’tuns)
- For the Maya, the number 13 was sacred. They believed the current world began on August 11, 3114 B.C., marking the start of a 13-b’ak’tun era.
- Total Days: 1,872,000 days (13 x 144,000).
- Total Years: Roughly 5,125.36 solar years.
- The Big Reset: This specific era concluded on December 21, 2012.
- Cultural Significance
- The completion of a b’ak’tun or a Great Cycle wasn’t a “doomsday” event. Instead, it was viewed as:
- A “New Year’s” Reset: Like an odometer flipping from 99,999 back to 00,000, the calendar simply restarted a new 5,125-year period.
- A Time of Transformation
- Many modern Maya and scholars view these transitions as the beginning of a new era of spiritual or physical transformation rather than destruction. Cosmic Renewal: According to the Popol Vuh (the Mayan creation story), we are currently living in the “Fourth World,” and these cycles represent the ongoing relationship between humanity, the gods, and the cosmos.
Long Count: One of the calendars of the ancient Maya, spanning cycles of 5,215 years.
Solar Year: A solar, or tropical year, is the length of time that the Sun takes to return to the same position in the cycle of seasons, as seen from Earth. The mean solar year is 365.242 days.
Stela: A monument shaped like a column, usually monolithic, inscribed with a commemorative, funerary, or ceremonial function.
Videos
2026 – Artificial Intelligence (AI)
History
Artificial intelligence has evolved through a long, human-centered journey shaped by myth, philosophy, engineering, and scientific discovery. As Muhammad Faheem notes, “The tale of artificial intelligence (AI) is, in numerous aspects, a tale about humanity itself — our curiosity, our inventiveness…” Early civilizations imagined artificial beings like Hephaestus’s golden robots and the Golem, reflecting humanity’s desire to create intelligence. Philosophers such as Descartes and Hobbes later explored whether thought could be mechanistic, laying conceptual foundations for AI.
The technological roots of AI emerged through mechanical automatons of the 18th and 19th centuries and the groundbreaking work of Charles Babbage and Ada Lovelace, who envisioned machines capable of symbolic reasoning. The 20th century accelerated progress: Alan Turing introduced the Turing Machine and the Turing Test, asking whether machines could think. AI formally became a field in 1956 at the Dartmouth Conference, where researchers imagined machines capable of human-like reasoning and learning.
Despite early optimism, AI faced major setbacks during the “AI Winters” of the 1970s and 1980s, when limited computing power and unrealistic expectations stalled progress. Yet important foundations—such as expert systems and knowledge representation—kept the field alive. The 1980s saw a revival through expert systems like MYCIN, proving AI’s practical value. The 1990s and 2000s marked a shift toward machine learning, enabling systems to learn from data and achieve milestones like IBM’s Deep Blue defeating Garry Kasparov.
The 2010s ushered in the deep learning revolution, powered by neural networks and massive datasets. As Muhammad Faheem explains, deep learning “made a significant contribution to the fields of computer vision, natural language processing, and autonomous systems.” AI became embedded in everyday life through voice assistants, recommendation systems, medical imaging, and autonomous vehicles. Generative AI further expanded capabilities, producing text, images, music, and software.
Alongside these advances, ethical concerns grew—bias, privacy, accountability, and the societal impact of automation. Muhammad Faheem emphasizes that AI’s evolution is inseparable from human values, noting that “AI is ultimately a reflection of human curiosity and creativity.” Today, AI stands at a transformative moment, with emerging fields like explainable AI, quantum computing, and brain–computer interfaces shaping its future. The history of AI reveals a story of ambition, setbacks, breakthroughs, and ongoing responsibility as humanity continues to shape intelligent machines—and be shaped by them.
Controversies
Schools
Pros of AI in Education
- Widespread student use and clear academic benefits: Most students (83%) regularly use AI and report gains in efficiency, idea generation, personalized learning, reduced stress, and improved work quality. “Many students recognise the benefits of AI integration in education, including increased efficiency, improved quality of work, idea generation, personalised learning…”
- Personalized learning experiences: AI adapts content, pace, and assessments to individual needs, helping students master material at their own speed.
- Intelligent tutoring and 24/7 support: AI tutors provide scalable, always‑available help, often improving learning outcomes compared to traditional instruction.
- Administrative efficiency for educators: AI automates grading, attendance, and curriculum planning, reducing burnout and freeing time for teaching.
- Accessibility for diverse learners: Tools like speech‑to‑text, translation, and simplified content support students with disabilities and multilingual learners.
- Equitable tutoring and empowerment: AI offers nonjudgmental, on‑demand help, especially valuable for students with anxiety or limited access to human tutors.
- Supports responsible AI literacy: When used intentionally, AI helps students learn how to evaluate tools, think critically, and prepare for an AI‑driven workforce.
Cons of AI in Education
- Reduced critical thinking and social interaction: A significant portion of students report declines in communication, collaboration, and analytical skills. “Between a quarter and a third of students report reduced critical thinking and less communication, interaction, and collaboration…”
- Overreliance and academic complacency: Many students admit to inappropriate use, rule violations, and dependence on AI for completing work. “Over three-quarters report that they cannot complete their work without the help of AI…”
- Equity and access gaps: Students without reliable devices, internet, or digital literacy are left behind, deepening existing inequalities.
- Reduced human empathy and connection: AI cannot replace emotional support, social development, or nuanced human understanding.
- Data privacy concerns: AI systems collect extensive student data, raising issues around surveillance, consent, and third‑party risks.
- Dependence on fragile tech infrastructure: Outages, cyberattacks, maintenance costs, and inflexible systems can disrupt learning.
- Academic dishonesty risks: AI makes it easier to cheat, plagiarize, or bypass learning, with educators reporting widespread misuse.
- AI hallucinations and inaccuracies: Models often generate false or biased information, making them unreliable for research or assignments.
- Deepening digital divides: Both access gaps and biased training data disproportionately harm low‑income students and students of color.
Workplace
AI is transforming the global workforce faster than any previous technology, automating not only physical and routine tasks but increasingly cognitive and professional work once considered safe from machines. This shift is disrupting industries from manufacturing and customer service to transportation and white‑collar professions, raising the risk of mass unemployment, widening inequality, and leaving many workers’ skills obsolete. While AI offers major efficiency gains, its ability to learn and adapt makes it fundamentally different from past technological revolutions, demanding widespread reskilling, stronger government policies, and ethical corporate practices to protect workers. At the same time, a paradox is emerging: as AI makes productivity abundant, uniquely human qualities—empathy, trust, creativity, and community—become more valuable, shaping a future in which humans and AI collaborate rather than compete. Ultimately, the future of work will depend on how societies prepare for this shift, redefine roles, and ensure that technology enhances human meaning rather than replaces it.
Mathematics
Historically, Al research was constrained by a reliance on mathematical proofs and logic-based expert systems. Moving beyond this “data science” obsession with marginal accuracy improvements, the author proposes a “Real-time Pattern Learning” model inspired by the hippocampus. By simulating biological neurons and feedback loops on specialized optical hardware, this approach aims to achieve human-level intelligence through unsupervised, associative learning, eventually allowing Al to surpass human cognitive capacity and longevity.
Currently, AI systems are advancing through mathematical problems at unprecedented speed, prompting both excitement and alarm within the math community. Over the past year, models from major labs have earned top scores at elite competitions, solved decades‑old research problems, and even produced publishable reasoning, leading many mathematicians to fear that the core purpose of their discipline is being destabilized. In response, more than 3,000 mathematicians—including figures like Terence Tao and Peter Scholze—signed the Leiden Declaration, a 23‑point framework urging transparency, human oversight, and stronger regulation of AI in mathematics. The field is now split between those eager to embrace AI‑accelerated discovery and those worried that opaque, proprietary systems could produce proofs humans cannot understand, distort research priorities, or compromise academic independence. As mathematicians grapple with how to collaborate with powerful AI labs while preserving human‑driven inquiry, many believe their struggle foreshadows similar disruptions across other scientific disciplines, making mathematics a test case for how academia will adapt to increasingly capable AI.
Navier-Stokes Millennium Prize Problem: OpenAI’s claim that it solved the Navier–Stokes existence and smoothness problem with an 88‑hour, 10,000‑agent proof has ignited significant controversy, as the proof relies on a modified equation that leaves the official $1 million Millennium Prize problem unresolved and unverified by the Clay Mathematics Institute. The situation escalated when mathematician Tristan Buckmaster suspected that private drafts of his recent work—stored in Codex—may have influenced OpenAI’s results, raising broader concerns about data privacy and the risks researchers face when uploading unpublished ideas to commercial AI systems. Despite these disputes, the sheer scale of OpenAI’s effort—producing a 130‑billion‑token mathematical argument in under four days—signals a potentially transformative shift in how advanced mathematical research may be conducted with AI.
References
∑ “’In Mathematics the art of proposing a question to AI must be held of higher value than AI solving it.’” Mathematical Mysteries, August 8, 2026. https://mathematicalmysteries.org/2026/08/08/in-mathematics-the-art-of-proposing-a-question-to-ai-must-be-held-of-higher-value-than-ai-solving-it/.
∑ “Mathematics in the Age of AI.” Mathematical Mysteries, September 17, 2026. https://mathematicalmysteries.org/mathematics-in-the-age-of-ai/.
∑ “Navier-Stokes Equations Explained.” Mathematical Mysteries, September 21, 2026. https://mathematicalmysteries.org/navier-stokes-equations-explained/.
∑ “Navier-Stokes Equations – Instagram.” Mathematical Mysteries, September 13, 2026. https://mathematicalmysteries.org/navier-stokes-equations-instagram/.
History
[ ℰ ] Rehman, Rehman. “History of Artificial Intelligence.” Medium, August 3, 2025. https://medium.com/@rehman.12142/history-of-artificial-intelligence-cc32d70aa883.
[ ℰ ] Liu, Jacky. “A 15-Minute Journey Through the History of Artificial Intelligence.” Artificial Intelligence in Plain English, September 2, 2025. https://ai.plainenglish.io/a-15-minute-journey-through-the-history-of-artificial-intelligence-750d9ef15228.
Balanspen. “A Short History of Artificial Intelligence.” Medium, October 27, 2025. https://medium.com/@balanspen/a-short-history-of-artificial-intelligence-9e32bea68cff.
[1] Faheem, Muhammad. “The History of Artificial Intelligence: A Human Journey Into Machine Minds.” Write A Catalyst, January 12, 2026. https://medium.com/write-a-catalyst/the-history-of-artificial-intelligence-a-human-journey-into-machine-minds-552e3ddf8e81.
AI Timeline Contributors. “AI Timeline — A Crowdsourced History of Artificial Intelligence.” AI Timeline. Accessed September 10, 2026. https://aitimeline.live/.
“History of AI.” GeeksforGeeks, April 2, 2024. https://www.geeksforgeeks.org/artificial-intelligence/evolution-of-ai/.
Controversies
Brumfiel, Geoff. “AI Solved One of Math’s Hardest Problems. Humanity Learned Nothing (so Far).” NPR, September 22, 2026. https://www.npr.org/2026/09/22/nx-s1-5968588/openai-navier-stokes-problem-mathematicians-learn-little.
OpenAI recently announced its AI model solved the Navier-Stokes problem, a major Millennium Prize challenge. While the 166-page proof is technically verified via Lean code, mathematicians like Tristan Buckmaster criticize its lack of human-readable insight. The release sparked controversy, with experts decrying rushed results and OpenAI’s competitive tactics against researchers. Ultimately, the mathematical community emphasizes that true discovery requires human understanding, not just computational outputs or raw proofs.
Ramzimubarak. “Artificial Intelligence and the Loss of Human Jobs.” Medium, August 12, 2025. https://medium.com/@ramzimubarak99/artificial-intelligence-and-the-loss-of-human-jobs-2cea4305bd70.
Wong, Sze(Z). “In the Age of AI, Human Interaction Matters More Than Ever.” Medium, October 6, 2025. https://szewong.medium.com/in-the-age-of-ai-human-interaction-matters-more-than-ever-96b25454a14a.
The Editors of ProCon. “Artificial Intelligence (AI) for Schoolwork.” Encyclopedia Britannica, July 13, 2026. https://www.britannica.com/procon/artificial-intelligence-for-schoolwork-debate.
Vilcarino, Jennifer. “Rising Use of AI in Schools Comes With Big Downsides for Students.” Education Week, October 8, 2025. https://www.edweek.org/technology/rising-use-of-ai-in-schools-comes-with-big-downsides-for-students/2025/10.
Turner, Cory. “The Risks of AI in Schools Outweigh the Benefits, Report Says.” NPR, January 14, 2026. https://www.npr.org/2026/01/14/nx-s1-5674741/ai-schools-education.
Melbourne Business School. “Key Findings on AI at Work and in Education.” Accessed September 9, 2026. https://mbs.edu/faculty-and-research/trust-and-ai/key-findings-on-ai-at-work-and-in-education.
“How AI Is Empowering Creatives with Chronic Illness.” AVIXA. July 31, 2023. https://www.avixa.org/explore/articles/pros-and-cons-of-ai-in-workplace-and-education.
Skuse, Benjamin. “‘Predatory behavior’: Elite mathematicians clash over OpenAI’s ‘solution’ to million-dollar math problem.” Live Science. September 17, 2026. https://www.livescience.com/physics-mathematics/mathematics/predatory-behavior-elite-mathematicians-clash-over-openais-solution-to-million-dollar-math-problem.
So where does all of this leave mathematics and mathematicians? A number of initiatives — such as Proofs and Prompts, Mathematical Discourse, and the Association for Human Mathematics — aim to preserve a kernel of human mathematics so the field can continue to be a home for human culture and understanding.
But for all the panelists, the speed of change and advancement in AI means they cannot offer clear predictions of the impacts of these initiatives or what the field might look like in the future.
“To be honest, whenever anyone talks about 10 years from now, or even five years from now … it just seems unfathomably far away,” Tsimerman said. “The world in two years is going to look so incredibly different than it does now.”
Navier-Stokes Equation
Babar, Sudarshan. “OpenAI Navier-Stokes Equation Solver 2026 Explained.” GetInfoToYou Tech, September 9, 2026. https://tech.getinfotoyou.com/openai-navier-stokes-equation-solver-2026-explained.
To understand the problem, you have to look at how physicists view liquids and gases. They call them fluids. The Navier-Stokes equations use Newton’s second law of motion. You probably remember F=ma from school physics. These equations apply that same logic to fluids. (Which makes sense, actually).
They treat water or air as a continuous medium instead of tracking individual molecules. So instead of working out what a billion separate water molecules are doing, the equations treat the fluid as one smooth mass. I’m not sure exactly why, but Claude-Louis Navier and George Gabriel Stokes came up with this back in the nineteenth century.
But there’s a catch. The math works perfectly for most everyday situations. Engineers use these equations right now. The problem is nobody actually knows if the equations always work under extreme pressure.
“Can Math As We Know It Survive AI?” Aventine. Issue 79. n.d. Accessed September 9, 2026. https://www.aventine.org/ai-math-threat-Leiden-Declaration-Terence-Tao.
[ ℰ ] Hauptfleisch, Wolfgang. “Is OpenAI Taking Everyone for Fools?” Misaligned, September 9, 2026. https://read.misalignedmag.com/is-openai-taking-everyone-for-fools-2481fa851544.
OpenAl faces serious accusations of scientific misconduct after claiming to solve a decades-old mathematical problem nearly simultaneously with researchers Tristan Buckmaster and Levent Alpöge. OpenAl admitted starting its effort after learning of the researchers’ work, raising concerns that it may have misappropriated their private data. Lacking transparency regarding its training sources, OpenAl’s claims remain unverified, casting doubt on the company’s integrity and its commitment to ethical research practices.
Kasanmascheff, Markus. “Navier-Stokes Equations: Did OpenAI Cheat When Solving One of Math’s $1 Million Millennium Prize Problems?” WinBuzzer, September 11, 2026. https://winbuzzer.com/2026/09/11/openai-claims-fluid-equation-breakthrough-credit-dispute-xcxwbn/.
Kwan, Matthew. “AI Is Not Mathematics.” Medium, August 23, 2020. https://medium.com/@matti.kwan/ai-is-not-mathematics-4199438627f.
Math in History. “90-Year-Old Math Problem Solved!” Medium, September 9, 2026. https://medium.com/@ganeshonline6/90-year-old-math-problem-solved-5ae72fd37591.
“Navier-Stokes Equations – Instagram.” Mathematical Mysteries, September 13, 2026. https://mathematicalmysteries.org/navier-stokes-equations-instagram/.
OpenAI. “On the Navier–Stokes Millennium Prize Problem.” September 8, 2026. https://openai.com/index/navier-stokes-solution/.
OpenAl has announced a breakthrough in the Navier—Stokes existence and smoothness problem, one of the seven Millennium Prize Problems. Using a highly capable internal Al model and a system of 10,000 concurrent agents, they produced an analytical proof and Lean formalization showing that three-dimensional fluid motion can develop a singularity in finite time. This achievement highlights the rapid progress of Al models in solving complex, long-standing mathematical challenges.
[ ℰ ] Plevris, Vagelis. “What Happens to Mathematics When AI Solves the Problems?” Medium, September 15, 2026. https://vplevris.medium.com/what-happens-to-mathematics-when-ai-solves-the-problems-2efe847f32ac.
In September 2026, OpenAl claimed to solve the Navier—Stokes Millennium Prize Problem, sparking intense debate. Twenty-five Fields Medalists subsequently released a declaration warning of a “severe misalignment” between Al corporate goals and mathematical practice. They argue that viewing unsolved problems as mere benchmarks rather than “lighthouses” risks prioritizing rapid output over deep conceptual understanding, ultimately challenging the core human value of the mathematical struggle and long-term intellectual development.
[ ℰ ] Plevris, Vagelis. “Lean: The Programming Language Rewriting Mathematics.” Medium, September 13, 2026. https://vplevris.medium.com/lean-the-programming-language-rewriting-mathematics-eca90a4aa167.
Lean is a functional programming language and interactive theorem prover used to formalize and machine-check mathematical proofs. By requiring precise, unambiguous logical steps, Lean allows computers to verify proofs independently of human intuition or authority. Now integrated with Al, it serves as a critical verification layer for complex research, such as the Navier—Stokes problem, ensuring logical validity as Al generates arguments that exceed the human capacity for manual inspection.
Saleem, Rohail. “A Mathematician Working On The Navier-Stokes Millennium Prize Problem Now Wonders If OpenAI Stole His Notes That He Stored In Codex [Updated].” wccftech, September 8, 2026. https://wccftech.com/a-mathematician-working-on-the-navier-stokes-millennium-prize-problem-now-wonders-if-openai-stole-his-notes-that-he-stored-in-codex/.
Shimkus, Ben. “You Should Care about the AI Math Breakthrough Drama Even If You’re Not a Nerd.” Business Insider, September 8, 2026. https://www.businessinsider.com/openai-navier-stokes-math-breakthrough-drama-2026-9.
[ ℰ ] oxford.mathematics. “AI Is Ravenous and Is Eating Mathematicians’ Lunch.” Interview with James Maynard. Instagram. Accessed September 11, 2026. https://www.instagram.com/reels/DdJ_kOXCxe_/.
It’s a very exciting time at the moment, but there’s also grave risks and dangers. The primary goal of mathematics has always been about human understanding of the ideas. When a mathematician solves a problem, then they can explain to others the new concepts that led them to this discovery, this proof, and how they came about it.
People in charge of AI systems seem to be using mathematical problems as a benchmark to test the capabilities of their AI systems, and this is not aligned with the values and the aims of the mathematical community.
Because of the competitive nature of the current leading AI systems, there was a huge rush to get to the final goal of the final solution of the Navier-Stokes problem, and this has meant that we’ve moved away from gaining the understanding that mathematicians were really looking for when they were thinking about this problem.
Is there, however, the argument that you guys are just a bit scared?
There’s certainly some aspects of mathematicians who are concerned about the future. I think lots of mathematicians are scared because AI will be this disruptive technology that will change lots of the practices of mathematicians. However, I am optimistic that there can be a very productive way forwards where AI can really boost mathematicians, help mathematical understanding, and enable mathematicians to go further and come up with very exciting discoveries.
I’m very hopeful that AI can be integrated into society in a disruptive but productive way that will be for the benefit of all of humanity. But if we can’t get this right in the microcosm of mathematics, then I worry about how it will affect many other fields and aspects of human work.
Read the Math and AI Declaration by 25 Fields Medalists.
[ ℰ ] Kakaes, Konstantin. “AI Has Solved One of Math’s $1 Million Millennium Prize Problems.” Quanta Magazine, September 8, 2026. https://www.quantamagazine.org/ai-has-solved-one-of-maths-1-million-millennium-prize-problems-20260908/.
Mathematicians at OpenAI showed that the Navier-Stokes equations, which describe how fluids flow, can sometimes “blow up.” But the massive result is not without controversy.
Videos
[ ℰ ] This exceptional reference is highly recommended for your consideration.