James,
So, step one: I think historians should conduct research on the history of AI as specific topics of investigation, as much work has to be done here;
That is undoubtedly correct. It isn't occult history, however. I recently read [1] which is among other things a history of AI, tracing it back at least to the formative McCulloch and Pitts paper of 1943 and of course to perceptrons. (The book is mainly about the mathematics behind LLMs, which means that it neglects heuristics and expert systems, but it does explore the relationship between pattern recognition and AI.) [1] Why Machines Learn: The Elegant Math Behind Modern AI, Anil Ananthaswamy, 2025. https://mitpressbookstore.mit.edu/book/9780593185766 Regards/Ngā mihi Brian Carpenter On 06-Nov-26 03:28, James Cortada via Members wrote:
Good news, Paul. Reacting as an historian, your self-critique about not discussing AI more is a bit harsh on yourself. Your role as an historian is not to predict future historiography; there is enough work just trying to deal with what already happened. The AI experience we are undergoing now evolved over a long period of time in what was clearly a far more modest form, e.g., remember "expert systems"? What I do suspect will happen is that future general histories of IT will probably overreact to AI's early history at the expense of other IT developments that occurred prior to Nov 2022 that were, in fact, more consequential in those earlier years. So, step one: I think historians should conduct research on the history of AI as specific topics of investigation, as much work has to be done here; step two, then to integrate those findings into the general history of IT or possibly more tactically into the general history of software and computer architectures (not sure if the latter is right here). Step three then becomes the integration of AI's relative inclusion into the kind of broad histories of IT that you and your co-authors do so well. Apologies for this long response, but with so many talented scholars now looking at the history of computing, a bit of perspective from some of the "grey heads" might be useful here. I wish you and I could be around to write such a book in 15 to 20 years. Jim
On Thu, Nov 5, 2026 at 8:13 AM Ceruzzi, Paul via Members <members@lists.sigcis.org <mailto:members@lists.sigcis.org>> wrote:
I just learned that on November 13, Amazon will be offering a Kindle edition of my 2014 book /Computing, a Concise History/ for $1.99. It is part of Amazon's "Daily Deal" marketing program, (which I had not known about until now). https://www.amazon.com/b?ie=UTF8&node=6165851011 <https://www.amazon.com/b?ie=UTF8&node=6165851011>
After 14 years (a lifetime in "computer years"), I think the book has held up well. With one exception: the only mention of AI in the book is on p. 114, where I mention it in the context of the PDP-10 computer and the LISP programming language. LISP! I am not alone in being blindsided by the advent of the "new" AI. (To be fair, I once was told that LISP was used to develop software for the Roomba robot vacuum cleaner. If anyone has first-hand knowledge of this, let us know.)
I have a somewhat complex relationship with Amazon, as I suspect many of you also have, but for $1.99, it's worth a look. https://www.amazon.com/Computing-Concise-History-Essential-Knowledge-ebook/d... <https://www.amazon.com/Computing-Concise-History-Essential-Knowledge-ebook/dp/B08BT2J6D9/ref=tmm_kin_swatch_0> Best, Paul Ceruzzi