I was approached last summer to take part in an online interview about AI and history (with a lot of big data thrown in). The questions were great, and the piece was published in Chinese last month: https://www.cssn.cn/skgz/bwyc/202602/t20260225_5974090.shtml; https://epaper.csstoday.cn/epaper/read.do?m=i&iid=7326&eid=53730&sid=250669&idate=12_2026-02-25_A03
Below is the English text as submitted. Thanks to Mirong Chen for setting this up and overseeing the translation and publication. Despite six months of radical development in AI, I am not totally embarrassed by my then views.
Email Interview: Tim Hitchcock. September 2025
1. Big data and AI appear to be driving historical research from traditional “close reading” toward “distant reading” and macro-analysis. In your view, does this constitute a fundamental paradigm shift? Could it potentially “dehumanize” the field, causing us to lose the nuanced context and individual experiences within history? Could the conclusions drawn from AI-driven macro-analysis conflict with traditional, archive-based historical interpretations?
History as a discipline is headed in both directions – towards more distant reading AND more – better contextualised – close reading. The assumption behind the question is partially correct. ‘Distant reading’ is revitalising forms of social science history that depend on charting large scale change. I am very hopeful that this trend will help ensure the growth in historical studies that seek to explain as well as describe historical processes. Arguably since the rise of post-modernism in the 1980s historians have been increasingly timorous about proposing broad models of historical development – to the detriment of the field as a whole. The ability to aggregate massive bodies of historical material; and to seek out patterns within that resource via big data via AI promises something quite new. This does not, however, imply a ‘dehumanisation’ of history writing. At its best, good social science history uses individual experiences to illustrate broader change; while all forms of history writing demand examples and empathy.
But! The question ignores the possibilities created by big data and AI for close reading. The underlying vector mathematics behind much of AI relies on locating the meaning and import of a single word, in the context of all its many bed-fellows. Undertaking the close reading of a line of text, or a short document, in light of every word published that year – or by that author – or in that genre, promises to re-invigorate and transform close reading as an analytical practise. I look forward to the moment when every word I read, in every document, is accompanied by detailed information about its changing meaning and context. Is the writer using a neologism, or vocabulary popular in their youth? How is that word’s use distributed across genre, and between authors? Many of the practises associated with close reading can be made more powerful and more illuminating through AI.
2. Historical archives are themselves filled with biases (e.g., the lack of records on women, minorities, and ordinary people). When we train AI models on these incomplete and biased datasets, will we repeat—or even amplify—historical structural biases? What strategies can we employ to identify and correct these biases?
Yes! The very essence of academic history writing is founded in a Western-centric notion of the ‘archive’ that excludes as much as it reveals. And the development of the internet has re-enforced all those biases in multiple ways. To take a single example, most early digital history sites were built on the basis of microfilm collections created between the 1930s to the 1980s in Europe and North America. These collections were created to give authority to the most conservative of historical materials, and this conservatism (indeed racist bigotry) has been simply reproduced online. As a result, the modern ‘object of study’ for most historians – texts and data that can be found online – already pushes us in specific retrograde directions. And this before we factor in the inherent biases created by issues of survival across the many cultures of the globe. AI takes the biases of the ‘archive’ and gives it new legs. It is not a new problem, but it is one that historians need to address urgently.
At the same time there is room for optimism. Much of the response to this issue has taken the form of the analysis of the ‘silences’ in the archive – frequently replacing detailed evidence with imaginative engagement. But there is another approach available. The rise of mass digitisation allows us to model the archive to expose what is there, and what is not. The life of your average 19th century working-class man from Manchester can be traced through twenty different records. The equivalent number for a person in much of the global South would be a fraction of a single record (and half the number again if we changed the gender). We could create a statistical model of the relationship between the historical population of the world, and the representation of that population in the archives, to allow us to measure of the relative importance of each document – creating a measure of how much each should contribute to a new historical understanding. If just a handful of documents stand in for the lives of hundreds of thousands of people, they deserve to be thought of differently. As important as this would be for historical understanding, an AI that builds in this relationship would help turn it from a system that largely reproduces the biases of inherited texts, into something that genuinely advances our understanding. As much as it would assist historical understanding, it would also form part of the process of reforming AI more generally. As it stands, just because the internet is Western and English centric, most AIs reproduce a Western and English centric understanding of the world. Incorporating a precise relationship between text (both online and inherited) and people, would help transform LLMs into something much more useful, and less prone to the reproduction of banal platitudes.
3. Historical scholarship is built on evidence, sourcing, and verifiable interpretation. But when a deep learning model reveals an unexpected pattern or correlation (for instance, that shifts in literary style preceded political upheaval), how do we “explain” the AI’s discovery? How can we differentiate between genuine historical insight and statistical coincidence or fallacy?
To explain evidence and patterns, however discovered, is simply the job of good history writing. The last 250 years have witnessed (with the rise of the bureaucratic state) the development of ever more sophisticated aggregations of data, that historians have grown old arguing about. The current state of the technology does not feel substantially different to me. The essence of the historical discipline is sifting through the alternative correlations to create an argument – an explanation – that is useful in the present. Whether the correlation was identified via an AI, a spreadsheet, or a census return is irrelevant.
Much more problematic is the possibility of automating the writing process – of eliminating the historian (however AI assisted) from the equation. Unlike our treatment of evidence and correlation, this transformation in ‘writing as thinking’ fundamentally undermines the historical process. It is only when we turn data and correlation into evidenced argument, that we create history. Outsourcing this process would both fundamentally undermine the intellectual project, and invalidate history as an academic training.
4. Generative AI can create highly realistic historical texts, images, and even simulate conversations with historical figures. How do you think of this technology being used for historical research? How should we address the risk of “deepfaking history” that it invites?
For the most part AI generated historical texts, images, etc, feel more relevant to public history and the presentation of the past to a broad interested community, than it does to historical research as a professional and academic discipline. I am put in mind of the ‘Cast Courts’ or reproductions gallery at the Victoria and Albert Museum in London. Opened in 1873, it was filled with detailed reproductions of classical and renaissance architecture and sculpture that would supposedly allow every museum visitor to experience at least a faint reflection of the original. They were deep fakes. More recently we have witnessed endless, largely failed, attempts to create historical evidence via dioramas, anamatronics, 3d modelling of archaeological sites, and historical re-enactments. As each form of presentation gets better, it demands greater critical attention, and perhaps new tools of analysis. This is an ongoing challenge for historians – but not a new one. The development of ‘deep fakes’ certainly implies the need to re-formulate the undergraduate curriculum to include critical engagement with digitally produced materials but does not change the nature of the challenge. My frustration is that this clear need has not been answered by innovative thinking about history teaching. For the most part, both the methods and content of a history degree have not changed substantially since the mid-twentieth century. Until we reform our teaching to incorporate the tools needed to assess new forms of evidence and fakery, we are both undermining the discipline, and making our students ever less effective as trained participants in other, non-academic, fields.
5. Successful digital history projects usually require close collaboration between historians, data scientists, and computer experts. What have been the greatest challenges in this interdisciplinary undertaking? Is it the difference in language and terminology, divergent methodological priorities, or the incompatibility of academic incentive structures?
In my now thirty years of leading digital history projects, it has always been the structures of the academy that are most irksome. In part this is about disciplinary arrogance – the tendency of a specialist, whether historian or computer scientist , to underestimate the depth of knowledge and disciplinary skills of the counterpart. More than this, there is a tendency for historians to simply want computer scientists to build websites; while computer scientists simply want historians to provide data. Holding the right kind of constructive conversation is predicated on the ability to understand what each discipline brings, and what each individual wants. There is also a deeper structural problem. Most universities – despite the rhetoric – are not set up for, and do not reward, inter-disciplinary or multi-disciplinary work. Appointments, promotions and grants are all apportioned by discipline. And while in the abstract, university leaders push for inter-disciplinary cooperation, they manage their institutions via competitive systems that set each discipline against the others.
6. Looking ahead, what is the most exciting prospect for big data and AI in historical research? And simultaneously, what is your most significant warning or caveat for the field?
The innovation I am most excited about is the development of pipeline systems for accessing historical materials in different languages and scripts. To take a single example, there are millions of pages of early modern Ottoman legal records preserved in Türkiye that have not been used by historians because they don’t have the language and paleography skills. We are rapidly coming to the point where creating a digital image of these sources would allow Handwritten Text Recognition (HTR) to generate a transcription in the original Arabic, allowing for an automated translation into another language. This translation could then be used to extract relevant text from the formulaic. The abstracted content could then be marked up for statistical analysis via automated semantic and geographical tagging, leading in turn to a resource that would open all these records to historians working anywhere, in any language. This sort of pipeline would make it possible for a second-year undergraduate to work with recondite sources in a way that has hitherto been unimaginable. If you want one answer to the biases of the archive, it lies in this sort of possibility.
My anxieties focus more on the historical profession than on the technology. At its best, the historical profession takes a self-conscious approach to evidence and argument – endlessly re-thinking how we know things about the past, and how we use the past in the present. But for the most part, in recent decades the profession has not risen to the challenge of interrogating the changing character of evidence when it is put on line. Our endless footnotes cite hard copy archives and physical books, when for the most part we read books online, and visit archives via the web. My concern is that this self-delusion will extend to AI; and in the process history as a discipline will be undermined, and increasingly marginalized. There is a dystopia out there in which AI generated ‘history’ is allowed to take the place of the debate and discussion that animate the discipline and justifies its role in a wider culture. As a profession we are well placed to prevent this happening, but it requires a much clearer sense of how history works, and our role in its creation.
Tim Hitchcock
Professor Emeritus of Digital History
University of Sussex
Hitchcock.t@gmail.com
1. Would conclusions drawn from AI-driven analysis conflict with those derived from traditional archival-based historical studies? If so, how should we reconcile interpretations from these vastly different scales and methodologies? Might this give rise to new historical theories?
AI-driven analysis is not fundamentally different from other forms. It is perhaps a subsection likely to privilege measurable change in large scale datasets, or changes in language use. But historians and corpus linguists having been doing this for generations. There is the possibility that AI-driven analysis will allow different source types to be integrated in novel ways that traditional historians have not yet explored. In an abstract world of data I can imagine an AI pulling together every dataset that includes a specific year into a single pool of information – including geological and weather records, ships logs and tree rings. This might produce something entirely novel, but I doubt it. Underlying the question is perhaps an assumption that unsupervised, AI might generate a new ‘explanation’ or model of social change, and that we would be forced to accept it, just because there are newly discovered correlations between different data-types. But I don’t believe this is how history works – when you strip out the hyperbole, history is just an evidenced argument with the present. You (or AI) can provide compelling ‘evidence’ of some ‘fact’, but it will still need to be argued; and in my experience humans are remarkably resistant to logic and argument.
2. About public history you mentioned in the "deepfake" question, big data and AI are indeed transforming how the public engages with history—through personalized historical recommendations, immersive VR experiences, and so forth. While these technologies make history more "democratic" and "accessible," how can we prevent them from creating new "filter bubbles" or oversimplified historical narratives?
I believe this is a real problem – though again, not a new one. The rise of ‘family’ history, of DNA ancestry tests, and so on, have allowed millions of people to curate the history of their choice; and AI may well facilitate this trend further. But this is not so different from the ways in which nation states traditionally created self-serving histories designed to re-enforce national identities. History has always been used as fodder for division and contest. The real question is how we can harness AI to complicate histories designed to divide. My hope is that the sheer volume of data from which AI draws will allow counterpoint and correction. As I mentioned in a previous answer, I look forward to the moment when every word I read (and fact I am presented with) is accompanied by access to a comprehensive context drawn from endless related data. The closed minded and self-serving will always find a way to read evidence to justify their own prejudices. It feels to me that AI might make that slightly more difficult.
3. With regard to challenges in interdisciplinary collaboration, can you offer your solutions to existing problems?
All the solutions have been tried and more or less failed. Free standing interdisciplinary research groups, institutes, grants requiring interdisciplinary teams, and cross disciplinary doctoral programmes have all been tried repeatedly over the last seventy years (and longer). The problem is endemic to the ‘University’ as a global model of education. As a result, you can reform one institution, but the weight all the world’s other universities will blunt its impact. In other words, we are pretty much stuck with the relatively ineffectual tools we have.
No comments:
Post a Comment