AivexaNewsSearch
AI news for builders and product teamsChecked every hour

3 Questions: What is the best path forward for AI in academia?

Collected Oct 13, 2026

MIT Statistics and Data Science Center director Sasha Rakhlin has published a new essay synthesizing recent discussions and readings about how artificial intelligence is changing academia, with a focus on mathematics, statistics, machine learning, and engineering. The essay, framed around three questions, concentrates on graduate research and education. It follows an MIT report on AI and education published this summer and arrives as universities weigh both the challenges and opportunities of rapidly expanding AI capabilities. Rakhlin holds the Distinguished Professorship in Data, Systems, and Society across the Institute for Data, Systems, and Society and Brain and Cognitive Sciences.

On research, Rakhlin uses mathematics to show how fast capabilities are moving. A model reached gold-medal level at the International Mathematical Olympiad last year, and only a year later models are producing new research results, including a recently proposed solution to one of the Millennium Prize Problems. Tasks that once demonstrated advanced mathematical ability, he notes, can increasingly be performed by AI. He points to verification speed and reliability as a central factor determining how fast AI advances in any discipline: formalized proofs can be checked automatically, and programs can be run and tested. Where evaluation is fast and reliable, systems can generate candidates, learn from outcomes, and improve. That dynamic applies to AI research itself, since better models contribute to the next round of development, creating a compounding process. For experienced researchers, Rakhlin writes, this opens questions whose technical demands previously put them out of reach, while sharpening the distinction between getting a solution and understanding why it works, what generalizes, and what to ask next. He cautions against assuming abstraction, judgment, or problem formulation will stay beyond AI, and says universities should prepare for a near future in which AI is smarter than all of us in many, perhaps most, aspects of intellectual work.

On academic credit and graduate training, Rakhlin argues a polished paper is becoming a weaker signal of individual expertise in a growing number of fields. Departments, he says, must start now reconsidering what they reward, without defining valuable work simply as whatever AI cannot yet do. He lists asking good questions, replication, synthesis of ideas, informative negative results, and shared datasets as candidates for greater recognition. Evaluation, in his view, should establish what a researcher contributed and takes intellectual responsibility for, including when substantial parts of the work were performed by AI, and those expectations should guide hiring, promotion, and funding and be made explicit for current and incoming PhD students.

Training is the harder problem in his account. Intellectual muscles develop through exercise, and routine calculations, coding, failed approaches, and small discoveries have long helped students acquire intuition and judgment. Delegating that work can remove formative experiences even as AI lets students attempt more ambitious projects, so departments need to distinguish unnecessary friction from the work through which expertise develops. Students should learn to formulate problems, audit model outputs, reproduce results, and defend their choices, and fundamentals may become more valuable as the basis for using these tools well. This is a familiar tension in professional education: tools that remove low-level labor also remove the practice ground where tacit skill forms, and the essay's framing suggests the answer is targeted retention of exercises that build judgment rather than blanket bans or blanket delegation.

On what universities should build now, Rakhlin identifies the clearest opportunity. The goal, he writes, is for universities to pursue questions over long horizons, share results openly, and evaluate claims independently. Industry partnerships will be essential, but he warns against assuming commercial priorities will cover the breadth of science or stay aligned with it over time, so some degree of technological independence will be necessary. Academia's most important strategic asset, in his view, may be the accumulated knowledge and experience inside its laboratories. Published papers are a selective record: failed experiments, abandoned directions, and the reasons an approach did not work often remain unpublished. Scientists and engineers hold tacit knowledge, gained through experience, about which interventions are likely to fail and why. That missing context may help explain why models currently struggle in some domains, especially the empirical sciences, to anticipate consequences that are obvious to an expert. Rakhlin notes these limitations may be temporary, and that capturing negative results and scientists' interpretations could make models much better at scientific exploration.

He sketches a future in which MIT's laboratories function more like a single, living scientific organism connected through shared AI research infrastructure. In his example, a neuroscience laboratory needs better methods for segmenting neurons in microscopy images; an AI agent could recognize a relevant advance from a computer vision group, connect the researchers, propose benchmarks, and help them iterate. It could surface unresolved questions to students and faculty and make lessons from one laboratory available to others. Enabling that requires workflows that capture hypotheses, interventions, outcomes, failures, and interpretations, plus shared systems that connect those workflows so agents can use tools and information across laboratories with appropriate permissions. Rakhlin calls for substantial public and institutional investment now in compute, secure data systems, and expertise in adapting and post-training AI models. He argues that tracing the process could preserve the lineage of ideas and make contributions, including those of graduate students, easier to recognize, and that with agreed rules for consent and credit, researchers could collaborate more openly with confidence their work would be acknowledged — which loops back to the question of how to recognize and reward intellectual contributions.

Why it matters: This is aimed at university administrators, department chairs, and PhD programs, but the credit and training questions apply equally to industry research teams that hire from academia. For developers and technical leaders, the essay's emphasis on verification speed as the engine of AI progress and on capturing negative results and expert interpretation points to concrete infrastructure work: provenance, experiment tracking, and permissioned shared tooling. The likely trade-off, on my reading, is that openness and richer process records improve both credit and model usefulness while adding cost and privacy overhead that institutions must fund up front.

Read at MIT News · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

MIT Statistics and Data Science Center Director Alexander (Sasha) Rakhlin shares important considerations for departments and institutions.