Beyond the Frontier: Advancing, Circulating, and Sustaining Knowledge in the Age of AI

Two Academic Responses to AI — Looking Inwards and Looking Outwards

On 6 October 2026, OpenAI published a document titled "Sharing AI Progress in Mathematics", announcing a broad collection of mathematical results produced by its internal frontier model. The announcement, concerning progress on numerous frontier mathematical problems, attracted considerable attention. Reactions on X ranged from excitement about the apparent acceleration of mathematical discovery to apprehension about the implications for mathematicians, mathematical research, and academic institutions. Some questioned the reliability and significance of the results; others saw an opportunity to challenge established research practices and the institutions that sustain them. Western news media subsequently brought these developments and the contrasting reactions to a wider audience.

I found the reactions particularly interesting because they went beyond the technical question of whether AI had actually solved the mathematical problems in question. They raised questions about the future of mathematical research, the professional identity of mathematicians, the role of academic institutions, and the relationship between human researchers and increasingly capable AI systems.

A few weeks earlier, on 16-17 September 2026, the London School of Economics had hosted the inaugural Global Forum for AI and the Social Sciences. Its focus was quite different: how AI is transforming labour markets, employment, the quality of work, inequality, and institutions, and how governments, businesses, workers, and international organisations should respond.

This contrast caught my attention. The mathematical community's reactions to OpenAI's announcement appeared to be directed predominantly inwards, towards what AI might do to the discipline and its academic institutions. The social-science community, at least in the case of the LSE Forum, appeared to be looking predominantly outwards, examining what AI might do to society and the economy.

Neither orientation is misplaced. But shouldn't both academic streams be looking in both directions? Mathematics must examine how AI is changing the wider world in which mathematical knowledge is applied. Social sciences must examine how AI is changing their own research methods, knowledge-production processes, and academic institutions.

The question, therefore, extends beyond the contrasting approaches of two academic communities. Every discipline must consider not only how AI is transforming the world it studies, but also how AI is transforming the discipline's own capacity to understand that world.

Comments

Post a Comment

Popular posts from this blog

The Age of the Sub-City: How Indian Municipalities Can Accelerate Economic Growth

India’s real estate sector is booming, again. Across states — from Telangana and Andhra Pradesh to Tamil Nadu, Maharashtra, Gujarat, and Uttar Pradesh — governments are announcing new industrial cities, capitalising on private investment and rising land values. These new cities promise jobs, growth, and industrial dynamism. But, there an invisible structural flaw. I argue that these state-led industrial city projects, while grand in ambition, neither strengthen India’s structural democracy nor ensure structural sustainability. What I suggest, instead, is India's next urban leap should not be another round of state-created industrial hubs — but the creation of what I call “sub-cities”: autonomous, self-contained ecosystems within large cities, governed under the broader sovereignty of existing municipal bodies. The Real Problem: Disempowered Municipalities India’s municipalities remain among the weakest in the world. Despite the 74th Constitutional Amendment, which promised greater ...
Read more

When the Monsoon Moves In: Reflections from the Uttarakhand Himalayas

The 'Announcement' in the Night Every year in early July, news platforms announce the advance of the southwest monsoon into North India, as meteorologists explain its progress with weather maps and rainfall forecasts. Those reports, no doubt, are scientifically precise, but they say very little about what the arrival of the monsoon actually feels like in a Himalayan town. I was reminded of that around a week ago. It was around 2 am. It had already been raining heavily for a few hours. Unlike the scattered pre-monsoon showers of previous weeks, this was continuous, determined rain. The kind that settles in rather than merely passing through. Then came the siren.  Living close to the Ganga, we occasionally hear warning sirens whenever the upstream UJVN dam is about to release excess accumulated water. Its purpose is to warn people to stay away from downstream river-banks. That night, however, the experience was unlike anything I'd had before. At around 2 am, the first long si...
Read more

From Targets to Systems: Rethinking India’s Ethanol Strategy

A Familiar Pump, An Unfamiliar Future The Indian fuel pump is beginning to look like a quiet site of transition. The nozzle is the same, the queue is the same—but the story around it is changing. Targets are sharper, alternatives are multiplying, and the idea of a “future without petrol” is no longer abstract. It is being signalled, incentivised, and—crucially—pre-built. And yet, for all the certainty in rhetoric, there is a gap in reality. The future is being announced faster than it is being assembled. The Push: Ethanol as Policy Workhorse India has already reached E20—20% ethanol blending in petrol—a milestone that would have seemed ambitious not long ago. The next horizon being invoked is far more dramatic: E85, even E100, supported by flex-fuel vehicles and a broader multi-fuel ecosystem. The rationale is compelling. India imports most of its crude oil, exposing itself to volatile prices and geopolitical risk. Ethanol, by contrast, is domestic, renewable, and politically attractiv...
Read more