I try rather hard not to dwell on what happened during class on my drive home. I don’t generally want to spend an hour driving home in the dark recounting all the things I have spent tens of hours reading about, and many hours talking about. I try to give my brain a break. I put on music or a podcast. But on Monday, I had to relisten to the podcast episode I had put on during my drive home. I couldn’t remember a thing about it. I got rather distracted by how incredibly life-changing getting a license was, and how world-altering readily available automobiles were. And somehow, I circled right back to the class I was trying to relax from.
There have been very few global changes as impactful as the innovations and connections made available by the internet. I remember my Pops once describing his childhood as “the dark before-times,” as a joke. But that joke was followed by a conversation about how truly, radically different the twenty-first century was from the twentieth century. To hear him describe it, you had to consider what it must have felt like to suddenly have ready access to automobiles, or electricity in homes, or even indoor plumbing. When these things were invented and made accessible, it was life-changing. The internet felt as if all three of those innovations came together, suddenly, and with few clues as to just how massive it would be. The internet would not only make history, but it would communicate its own history.
The internet has come with explorations, innovations, and very difficult lessons. Among these innovations has been the digitization and online access of historical material, just one aspect of the broader digital history I posted about last week. Most of the reading material assigned to me this week discussed the lessons that could be taken from the mass digitization efforts. The two authors I came back to made arguments that, by changing the method of historical research, we lose the face-to-face friction and interaction that aided, even completed, research.
One of these authors, Lara Putnam, has much to say in her article The Transnational and the Text-Searchable.1 She brings awareness to what is lost “when the real-world friction that international research once demanded is radically reduced”(pg. 380). Though she admits that online access to research material allows much quicker research, and that it demolishes many barriers in the way of transnational research, the concern remains. What is lost in digital research, to Punam, is the experiences of fieldwork. On-the-ground research necessarily brings a whole range of unexpected challenges and obstacles to a historian. Putnam argues that the major risk of digitizing everything is that incoming generations of historians will not see the value, or even the option, of fieldwork. This was an argument that I was less concerned with before the class on Monday.
The issue of on-the-ground research has always been a complex issue. I feel digitization solves more problems in this area than it causes. Putnam concedes in her article that the expense of traveling for archival access used to be a barrier to entry. Funding continues to be a problem in academic research. However, I have found in my own academic circles that dedicated researchers, degreed academics and students alike, will find ways to surmount this problem. A friend of mine found numerous Boston-area contacts for undergraduate research, based in Virginia. Through this continued correspondence, she managed to find the support and funding to visit Boston for an extended period. I myself slogged through scholarship and grant applications to be able to study abroad, and it was an experience I will forever be grateful for.
Nothing can replace physically experiencing an art piece you’ve only ever seen on a screen. But, for me, I am less concerned with newer researchers voluntarily forgoing such experiences. A dedicated historian will try to access as much as they can about their topic. Financial, geographic, and legal barriers put in their way are not their fault.
What concerns me more as a student and researcher is the technology discussed by Dan Cohen in The Writing Is on the Wall for Handwriting Recognition,2 a blog post from late 2025. He covers the many advancements in, and forms of, transcription software. Digital handwriting recognition, referred to by Cohen mostly as Handwritten Text Recognition (HTR), is a field experiencing quick and incredible advancement. Cohen covers both internally trained HTR systems, such as Transkribus, and AI-powered software, such as Google’s Gemini. The tone of the newsletter is happily excited about the advancements of HTR. Cohen’s main concerns after testing and comparing HTR systems are similar to Putnam’s: where do we draw the line between the benefits of digital tools and the costs of using them? Cohen points out that jobs will be lost as HTR advances, and that nothing can really replace the ability of a trained expert. He concludes that HTR and AI are tools that need to be carefully contained to aid researchers, not to do research.
I considered Cohen’s argument carefully during our discussion in class. Time is both the currency and enemy of researchers. The time it takes to train a contained HTR system would be immense, but once trained, HTR systems can save many people lots of time. On the other hand, the privacy concerns, oversight issues, ecological concerns, energy and water costs, and quality-of-life concerns for those most impacted by the Data Centers that power AI are very, very real. If the performance of either type of system can reach similar qualities, and if it all has to be checked and confirmed by humans anyway, the decision on what software to use comes down to a question of time and personal ethics.
On my drive home after class, I ultimately returned to considering Putnam’s concerns on digital research, and how her concerns and Cohen’s warnings seem all the more prescient when considered together. Though I found the hardest thought to shake was the conversation with my Pops recounted earlier. The conclusion he came to was that whatever the true “next big thing” would be, we need to get out ahead of it. Once a concern is identified, we need to take steps to prevent it from occurring, instead of repairing the damage once it is done. In the case of digital history and AI, a combination of Putnam and Cohen’s approaches seems the most rational. We need to consider what may be lost with digital history and AI use, and mitigate the damage through recognition, regulation, education, and accountability. I do believe that a dedicated researcher will try to experience in-person what they may only have access to on a screen. However, I have also seen the learned helplessness and lack of critical thinking that comes with grade school students today who consider ChatGPT a search engine, or think that outsourcing their thinking is just “saving time.”
This week’s reading and class gave me a lot to think about. Ultimately, I hope it isn’t too late for the internet, digital history, and AI to be implemented carefully. Maybe we can get out ahead of this one. If we can, it will require that face-to-face friction that Putnam wanted, and more of the hands-on investigation that Cohen produced. Mostly, I think it will require communication. Conversations with your peers, parents, younger siblings, and/or children. Discussions about what is gained, learned, lost, and damaged by the current state of digital history and the tools available to the internet age. And on a personal basis, I think these concerns require thought. It might not be enough to keep someone up at night. But if this is enough of an issue to take up an entire, dark, hour-long drive, it’s enough of an issue to set my own boundaries on.
- Putnam, Lara. “The Transnational and the Text-Searchable: Digitized Sources and the Shadows They Cast.” The American Historical Review Volume 121, no. Issue 2 (April 2016): 377–402. https://academic.oup.com/ahr/article-abstract/121/2/377/2581842?redirectedFrom=fulltext&login=false. ↩︎
- Cohen, Dan. “The Writing Is on the Wall for Handwriting Recognition.” Dan Cohen Newsletter (blog), November 25, 2025. https://newsletter.dancohen.org/archive/the-writing-is-on-the-wall-for-handwriting-recognition/. ↩︎