Another "agent" kicked off an "intelligence" revolution nearly six hundred years ago...

Why the printing press from the 1450s is the really starting point to understand how agents change the way information manages our lives

(This commentary is one of two connected pieces. It was inspired by an intriguing discussion with a Gen Zeder all “hepped” up on AI’s possibilities and potentials. It soon became clear to me he had no sense at all of the “information transformation arc” over centuries of which AI is the latest manifestation. What follows ideally inspires people to dig deeper into how these technologies have shaped, and continue to shape, the information we consume – and how we consume it. Lots of homework, to be sure. I left the Gen Zeder with that exact homework for our next coffee: to report back on how he sees his life in 2035, with a focus on his “information self.” )

Everyone trying to make sense of the current AI moment reaches for an analogy, and most reach for the nearest one: the dot-com boom and bust.

Funny thing is many of AI’s biggest boosters only have a vague sense of that time in information-and-technology history.

The comparison is not wrong, exactly. The structural “historical rhymes” are real enough. There was massive capital chasing a genuinely transformative technology, valuations detached from revenue, a picks-and-shovels supplier (Nvidia now, Cisco then) capturing outsized value while its customers burn cash, companies bolting “AI-powered” onto their names the way an earlier generation bolted on “.com.”

Remember then? I do. I had just launched Western Canada’s first daily newspaper website in the mid 1990s amid the run-up to the bust.

The attitudinal rhymes are real too: the FOMO, the talent gold rush, the conviction that traditional valuation logic no longer applies, the career risk of not participating, the blurred line between visionary and grifter, the skeptics dismissed as people who simply don’t get it.

And the aftermath of the first boom carries a lesson worth keeping front and centre in our minds. The dot-com believers were largely right about the internet. They were just wrong about timing, and about which companies would win. Amazon and Google emerged from the wreckage; Pets.com and Webvan did not.

If history rhymes, the interesting question about today’s darlings is not whether AI is real but which of them are the Amazons.

But the dot-com analogy has a limit, and the limit matters. The dot-bomb was a financial event with human casualties: roughly a million lost tech jobs, a Nasdaq that fell nearly eighty percent and took fifteen years to recover, retirement accounts gutted, suicides and depression in the Valley, a generation of stock options turned to wallpaper. Painful, but mostly contained to people who had opted into the boom. When the startups failed, the damage largely stopped at the edge of the industry, and a laid-off engineer could find another job because the job categories themselves survived.

What makes AI different in kind is that its deepest disruptions arrive if the technology succeeds, not if it fails. The dot-com crash hurt people because the companies collapsed; AI stands to reorder the work of translators, junior programmers, paralegals, customer-service agents, and illustrators precisely because the companies deliver what they promise.

Add the subtler harms with no 1999 precedent (synthetic media eroding trust, cognitive deskilling, parasocial dependence on chatbots, the psychological weight of choosing a career in a fog) and the picture sharpens: a financial correction and a structural social transformation are not alternatives here. They may well be sequential, and the second happens whether or not the first does.

In the dot.com era, we paid a lot of attention to technological transformation and virtually ignored the impacts on information transformation. I didn’t give much thought to how a digital version of a daily newspaper might change the way people consumed its content from their print-based consumption habits.

We should, then, be very mindful in the AI boom not to lose the opportunity to reflect on how the technology is changing the information. It seems obvious in so many ways, but deeper probing takes us in more intriguing directions.

For that kind of change matters. To understand change in how a civilization produces, distributes, authenticates, and absorbs information, you need a deeper precedent. You need the fifteenth century. And the essential guidebook to that century remains Elizabeth Eisenstein’s The Printing Press as an Agent of Change (1979). Anyone seriously interested in how information regimes shift across time and space should own a copy, read it with a pencil, and keep it within reach for the next decade.

Eisenstein’s great achievement was to treat the printing press not as a mere invention but as an agent. Hmmm. An agent. Starting to connect the dots?

Her real contribution? Asking us to consider the printing press as a force that acted upon institutions, professions, and habits of mind whether or not anyone intended it to. Before her, historians tended to mention Johannes Gutenberg and his 15^th-century printing press – come on down, movable type, in 1455 – in passing and move on to the Reformation or the Scientific Revolution as though those upheavals had generated themselves.

The didn’t. The press did. And it started with Gutenberg’s first effort: a bible.

Eisenstein supplied the connective tissue made possible by that effort: print made texts stable, comparable, and cumulative. A scholar in Kraków and a scholar in Lisbon could, for the first time, argue about the same edition. Errors could be corrected across printings rather than propagated by tired copyists. Maps, tables, and diagrams could be reproduced exactly. Knowledge began to compound. And everything downstream of compounding knowledge, from Copernicus to the modern research university, became possible.

I’ve read and reread her work many times over the years. Each time it resonates differently, depending on the information context of the moment. Read today, three of her themes land with uncomfortable force. First, the destruction of a skilled class: the “scriptoria” and professional scribes who had monopolized textual reproduction were rendered economically obsolete within two generations. That’s an early instance of the pattern now looming over knowledge work.

Second, the authenticity crisis: early print was awash in errors, forgeries, piracies, and unauthorized editions, and it took more than a century to develop the antibodies we now take for granted: things like title pages, editorial standards, citation norms, the very concept of a “reliable source.” Think of that that ubiquitous disclaimer of today: “AI can make mistakes.” Anyone worried about synthetic media should study how long that immune response took to build.

Third, the escape of information from institutional control: cheap pamphlets were the rage-bait of their day, the witch-hunting manual Malleus Maleficarum was an early bestseller, and most historians regard print as a necessary condition for the Reformation. The press triggered a century and a half of religious war that followed, culminating in a conflict that killed perhaps a third of Central Europe. After mass destruction, the press eventually delivered mass literacy and the Scientific Revolution, but the transition was measured in generations and, at times, in blood. Eisenstein does not moralize about any of this; she documents it, which is far more useful. Just like we ought to be doing with the moment with AI, instead of waiting for history’s rearview mirror.

The lesson is not that catastrophe is destined. It is that the gap between an information technology’s arrival and society’s development of norms for living with it is long. Perhaps 150 to 200 years in the print case; and that the payoff era and the turbulence era are not mutually exclusive but braided together. If the pattern holds even loosely, we are standing at roughly 1455, admiring Gutenberg’s first printed Bibles, with Luther’s Reformation still decades away, in terms of the time frames it will take before AI’s real impact becomes apparent.

My young Gen Zed buddy was only vaguely aware of the dot.com boom and bust, but to his credit, was very much intrigued. He promised to research both through a financial impact lens, but also on what happened to “information” in and of its information self. And committed to being more curious about what his “information life” will look like in 2035.

(Eisenstein gives us the history. To turn that history into a working imagination of 2035, we need two theoretical side tours: through Harold Innis and Marshall McLuhan. That is Part II, coming up.)