This visualization is difficult for a casual viewer to decipher,and this is necessarily so, because the visualization is itself aresearch tool. It is a way of analyzing information more than afinished representation of the conclusion.

allows us to merge things createdat dramatically different times to create what are in effect newmodern images which potentially reveal things about the pastthat the original artifacts did not.11

This brings me to new technologies, which are by no meansuncontroversial. Historians like myself who advocate the use ofthese technologies can be accused of having deserted the forces oflight and embraced the forces of reaction.

They are not separate from spatial practice because in largemeasure they are what guide the human labor that creates kitchens,bathrooms, living rooms, roads, train stations, airports, air trafficcontrol and entire landscapes.

In other words, we need smart big data because it can not only adequately represent a sufficient number of relevant features of humanistic objects of inquiry to enable the level of precision and nuance scholars in the humanities need, but it can also provide us with a sufficient amount of data to enable quantitative methods of inquiry that help us transgress the limitations inherent in methods based on close reading strategies.

Basically, there are two possible ways to do this: one is automatic annotation, the other is crowdsourcing. Automation refers to various heuristics of discovering implicit units, structures, patterns and relations, and of making them explicit in the data. Crowdsourcing, on the other hand, relies on breaking down a large task into such small units that each of these little tasks can be performed in a distributed way by a large number of volunteers.

Even without using a new term, we can now redefine what we mean by data in the humanities. Data in the humanities could be considered a digital, selectively constructed, machine-actionable abstraction representing some aspects of a given object of humanistic inquiry.

But rarely would they consider their objects of study to be “data.” However, in the humanities just as in other areas of research, we are increasingly dealing with “data.”

All three of the areas I’ve described here - multimodal generation, Deep Research, and handwriting transcription - have changed dramatically in the past twelve months.

This year also saw the rise of reasoning and agentic models. To oversimplify, reasoning models break their responses down into smaller sequential steps (called “chain of thought”) and spend more time and computing power on generating an answer.

Any historian interested in public engagement needs to pay attention to this development.

Whether or not you guessed correctly, this exercise illustrates a larger point: generative AI has gotten really good, really fast, at generating multimodal content, or stuff that isn’t just text (i.e. images, video, and audio).

We hope this overview helps explain why digitization within archival institutions proceeds the way it does – and why we may never, in fact, digitize everything.

Because digitization involves an investment of time and resources, we need to make sure we get it right – that the electronic files we produce are adequately representing the archival originals. That means our process will need to incorporate quality control checks.

At best it results in a digital “surrogate,” an approximation (even if a very good one) of a dimension of the record.

As archivists we like these questions because they tell us that people are eager for access to archival records. They also show that people realize that not everything is digitized. Indeed only a tiny fraction of the world’s primary resources are available digitally.

ne development far outstripped the impact of any other during the 1990s. This was the arrival of the Internet, but more especially the World Wide Web. The first graphical browser, Mosaic, appeared on the scene in 1993. Now the use of the Internet is a vital part of any academic activity. A generation of students has grown up with it and naturally looks to it as the first source of any information.

If any single-word term can be used to describe this period, it would almost certainly be "consolidation." More people were using methodologies developed during the early period. More electronic texts were being created and more projects using the same applications were started. Knowledge of what is possible had gradually spread through normal scholarly channels of communication, and more and more people had come across computers in their everyday life and had begun to think about what computers might do for their research and teaching.

The other widely used citation scheme was more dependent on punched card format. In this scheme, often called "fixed format", every line began with a coded sequence of characters giving citation information. Each unit within the citation was positioned in specific columns across the line, for example the title in columns 1–3, verse number in columns 5–6, and line number in columns 7–9. The entry of this information was speeded up by functions on the punched card machine, but the information also occupied more space within the computer file.

At this time much attention was paid to the limitations of the technology. Data to be analyzed were either texts or numbers. They were input laboriously by hand either on punched cards, with each card holding up to eighty characters or one line of text (uppercase letters only), or on paper tape, where lower-case letters were perhaps possible but which could not be read in any way at all by a human being.