Mainstream WeeklyPicture two doctoral scholars sitting in the same university library. One spends three weeks meticulously sifting through journal archives, cross-referencing footnotes by hand, and slowly wrestling with dense arguments. The other feeds a handful of prompts into tools such as ChatGPT, Elicit, or Connected Papers and walks away with a structured, visual map of the entire field before lunchtime. On paper, both produce an identical literature review. In reality, the academic clock has split in two.
Across universities worldwide, a quiet shift is underway. For the past few years, the public conversation has focused on whether artificial intelligence will replace human scholars. But the more immediate and urgent reality is about pace. AI is not simply changing how research is done; it is changing how quickly scholarship is expected to happen. A survey conducted for Springer Nature revealed that over half of researchers are already using AI tools to read literature, synthesise papers, and draft grant proposals. The result is an emerging hierarchy that operates not through access to information, but through the currency of time.
Traditionally, academic privilege was visible. It meant having a library card to elite archives, access to expensive laboratory equipment, or the institutional wealth to afford paywalled journals. Today, two researchers on opposite sides of the globe can access the exact same open-access paper. Yet, one can summarise fifty such papers in an hour using an automated assistant, while the other takes days of careful reading. Because hiring committees, fellowship panels, and university administrators evaluate the final output rather than the time taken to produce it, speed becomes an invisible benchmark of intelligence and competence.
This compression of time changes the very nature of learning. The literature review has long been the most formative, if slowest, stage of scholarly training. It is where a young researcher learns to sit with confusion, trace forgotten debates, and stumble upon unexpected ideas while browsing unrelated pages. When an algorithm delivers instant summaries and highlights only the most popular citations, that creative friction disappears. Scholars risk developing what might be called synthetic competence: they can present polished overviews of a subject without having internalised the difficult, messy ideas behind it. Speed begins to masquerade as mastery.
The institutional response to this shift has largely been reactive, trying to police the machine rather than questioning the accelerated tempo itself. In India, for example, the University Grants Commission’s plagiarism regulations are increasingly stretched to catch AI-generated text in doctoral theses, imposing strict resubmission penalties and threats of cancellation based on arbitrary similarity thresholds. Yet, as an analysis in The Hindu pointed out, relying on blunt pattern-detection tools like Turnitin or ZeroGPT creates an absurd paradox: detectors flag predictable sentence rhythms, which means multilingual scholars writing in a non-native language or scholars using standard academic phrasing are disproportionately hit with false positives. Honest researchers find themselves deliberately degrading their prose—inserting typos or stripping away punctuation—just to look “human” to a machine. By focusing on automated surveillance rather than intellectual depth, institutions trap scholars in a brutal double bind: demanding faster outputs while penalising the very tools that make such speeds achievable.
The pressure of this new tempo is especially visible in the grant and publishing cycle. Researchers now use AI to draft proposals, refine research questions, and submit applications at record speed. Even platforms like Coursera now offer structured courses on using AI to write funding bids. But rather than giving scholars more free time, this acceleration simply speeds up the treadmill. When everyone can produce applications faster, funding bodies raise their volume expectations. Scholars are pushed to write tactical, safe proposals that cater to algorithmic templates rather than pursuing high-risk, slow-burning ideas that might take years to show results.
The same pressure is spilling over into peer review and editorial desks. As noted in a study on AI in scholarly publishing indexed on PMC, editors and reviewers increasingly rely on AI tools to summarise complex manuscripts, identify weak citations, and speed up turnaround times. While faster feedback can ease backlogs, it also risks creating a closed loop of machine-assisted writing reviewed by machine-assisted readers. Authors are expected to revise dense manuscripts in a few short weeks, leaving little room for the slow reflection that genuine intellectual revision demands.
This temporal divide carries deep global and structural consequences. As highlighted in the Stanford AI Index Report, access to advanced systems remains unevenly distributed. Scholars at well-funded Western institutions often have access to institutional subscriptions, premium AI tool tiers, and dedicated university workshops. Meanwhile, a researcher at an underfunded public university in the Global South may face frequent power cuts, paywall restrictions, and rate-limited free versions of the same software. Furthermore, most leading generative models are trained and optimised in English, creating a built-in linguistic head start for anglophone researchers and pushing vernacular knowledge systems even further to the margins.
When speed becomes the primary measure of scholarly value, the academy risks losing its most essential quality: deep, patient thought. Breakthroughs in history, social theory, and fundamental science have rarely come from working faster; they have come from taking the time to doubt, wander, and rethink old assumptions. If academic evaluation systems continue to reward volume and rapid turnaround, they will favour formulaic, safe outputs over original thought.
The solution is not to banish these technologies, which can legitimately ease bureaucratic burdens and make information more reachable. Rather, universities and regulatory bodies must rethink how they measure worth. Beyond punitive AI checks, evaluation criteria should reward the depth and originality of a small selection of works, rather than the sheer volume of output. Institutions must make space for slow thinking and recognize that the time spent wrestling with an idea is not wasted time. The future of scholarship will not be determined by how quickly we master new machines, but by whether we have the wisdom to keep the academic clock under human control.
(Author: Disha is a PhD Scholar and Senior Research Fellow at Dr K. R. Narayanan Centre for Dalit and Minorities Studies, Jamia Millia Islamia, New Delhi)