Citizen Twain: Jeff Jarvis on Hot Type, the Magnificent Machine That Gave Birth to Mass Media

00:00:32 Andrew Keen: Hello, everybody. It's Wednesday, August 19, 2026. It's turning into quite a Luddite summer, at least according to The Wall Street Journal. In fact, it's the summer that America became a nation of Luddites. We're all concerned about the impact of technology on our politics, on our economics, and above all else, on our culture. The Journal also ran a piece about how AI is plunging the book industry, the book publishing industry, into utter chaos. And, as my guest today on the show, an old friend, frenemy in many ways, Jeff Jarvis wrote recently, words matter. Damn it. And they matter to him on lots of fronts, particularly because he's got a new book coming out tomorrow, Hot Type, The Magnificent Machine That Gave Birth to Mass Media. Jeff is joining us from New Jersey. Jeff, congratulations on the book.

00:01:33 Jeff Jarvis: Thank you, friend. Good to see you.

00:01:35 Andrew Keen: So, Jeff, this summer of our Luddite discontent, what do you make of it, and how is it connected to this new book? You're always very timely in terms of your publications.

00:01:51 Jeff Jarvis: I cover in my books the confluence of technology and media and thus culture. And so we've certainly seen times before when technology enabled a flood of more text in society, which caused all kinds of panic from people thinking that novels were gonna ruin women and children. Going back further to Gutenberg, obviously, the Reformation, and on up to today, what we saw was the mechanization and industrialization of print and thus media and culture in the nineteenth century. For those of you watching on video, I'm holding up a [unclear] type. And pretty much every word in every language [unclear] was set this way for four hundred fifty years. And then came the steam-powered press, which enabled, scale unforeseen before. Before the steam-powered press, the average circulation of a daily paper in the US was only 4,000. It was a small Substack newsletter. And cheap paper made from wood pulp and stereotyping and the telegraph and so on. The one thing that was standing in the way of scale in media was the typesetter standing there at the — with the composing stick, one damn letter at a time. And there were more than 300 efforts to invent the typesetting machine. The one that was successful was the Linotype, which molded type instead of one letter at a time, one line at a time. And I thought I was gonna try to write a biography of Ottmar Mergenthaler, the inventor, but he died young, which didn't help him or me. But then I found myself going down —

00:03:30 Andrew Keen: You do have a thing, Jeff, about German inventors. You're —

00:03:34 Jeff Jarvis: As it turns out, yes, between Gutenberg and Mergenthaler. But Mergenthaler worked in Baltimore, so he was — we'll claim this as an American

00:03:42 Andrew Keen: Well, shall we say, inventors with German-sounding names?

00:03:45 Jeff Jarvis: Very much so. And, so there were a whole bunch of great rabbit holes here. The invention of the typewriter, which was the necessity of the keyboard. Mark Twain went bankrupt. He threw all of his fortune and his wife's fortune into a competitive machine called the Paige Compositor, which was just a complete cluster f—. The story of labor and the Linotype is fascinating, Andrew, because you would think that they would have — the typesetters would have fought this machine called the typesetter, but they realized based on the experience of the Luddites that it was a losing battle to fight the technology. And based on the experience of pressmen with the steam-powered press, they saw that the industry increased. So they said, okay. You could — you publishers, you could have this new machine, but, you gotta put us in charge. We have to be the ones to operate it. They lost by one count 30,000 jobs, but within five years, they were back to full strength and then grew from there for seventy-five years until faced with the next technologies. Fascinating stories about the foundries that used to make type and their sad decline. The whole story ends with the invention of PostScript and desktop publishing, which eliminated the inventions of both Gutenberg and Mergenthaler and changed everything. And all these inventions not only changed society and its relationship to text, but it changed my own relationship. At the end of the book, I write an autobiography — a typographical autobiography about how typewriters and computers and desktop publishing all changed the way that I wrote and thus that I thought. I don't know exactly — when did you learn how to type, Andrew?

00:05:24 Andrew Keen: I taught myself.

00:05:26 Jeff Jarvis: How old?

00:05:29 Andrew Keen: I guess in my early to mid twenties.

00:05:32 Jeff Jarvis: Oh, that long. Woah. Woah. So you were a scribe until then.

00:05:38 Andrew Keen: Yeah. And, I mean, I still write. I still physically write. I mean, I don't write everything, but I do — I mean, probably because I'm more of a reactionary than you, I always assume that I do my best writing by hand rather than on the keyboard, which may or may not be true.

00:05:56 Jeff Jarvis: Yeah. I have to think on the keyboard. I cannot think otherwise. I my handwriting has gone awful, but that's just the way that I process.

00:06:05 Andrew Keen: So let's go back to Hot Type, Jeff. You've written a series of books about technology, our relationship with technology, particularly in terms of journalism and the printing press. As I said, you're forever associated with the great Johannes Gutenberg. You wrote, a wonderful book, The Gutenberg Parenthesis. Did that — you've always been someone, I think, and correct me if I'm wrong, who's argued that we've lived through this before when it comes to technology and culture, and your historical analysis has calmed down the skeptics, the hysteria around media, which we're once again living through in our age of AI. Did the research and writing of Hot Type, did it change your mind in any way, or did it confirm everything you thought in the first place?

00:06:57 Jeff Jarvis: It's a good question. I think that I still essentially believe that we figure it out, that technology at first, confounds us and frightens us as a society. But what happens inevitably, in my view, is that before long, the technology and the technologist fade into the background. The typewriter wasn't mysterious anymore. The typesetting machine wasn't mysterious anymore. The computer was not mysterious anymore. Then, I think the Internet is less mysterious than it was. Right now, AI is the province of priests with very weird religions, which we can talk about. But I think that because AI is meant for anyone and everyone to use it merely with our own language, in this case, I think the technologist will fade into the background, faster than with other changes. So the question becomes, what do we do with these technologies? Do we control them properly? And there's no technological determinism in this. I've been researching the next phase of media now, which was electronic with the amplifier. And I learned that, for example, in the UK, of course, you had the BBC and in Europe, you had state, supported media in the US. We didn't in great measure because the Department of the Navy in the US insisted that the patents could not go to you Brits because it would, be a security problem, and they insisted on the creation of the company called RCA, Radio Corporation of America. And that's how we got broadcast the way we did as a corporate commercial entity. It didn't have to happen this way. We could have had the same model that the UK did with BBC. So what I'm trying to say is that the technology causes change. There's not much hope in trying to forestall that change, but there is every reason to learn from history so that we can try to take control of this technology and its impact, and there are choices to be had.

00:08:50 Andrew Keen: Jeff, given the massive conservative romantic reaction against the industrial revolution in the nineteenth century, particularly in the United Kingdom, what was the reaction against hot type? Was that the same general hysteria that we're living through now when it comes to AI?

00:09:10 Jeff Jarvis: No. Oddly, because I think it came late enough. I think people have lived through steam. They had the very beginnings of electricity with the telegraph. And, again, the typesetters themselves, they'd lived through 300 machines that failed. Bad inventions. At first, they didn't care. And then the first machine was installed in the New York Tribune in, July 3, 1886, under the tutelage of Whitelaw Reid, who's kind of the villain of this tale, the ultimate capitalist, the first capitalist media mogul in a way. And, he had a nonunion shop. He used the machine to get rid of the union, but there were union spies in the composing room. And they saw that it worked. And so they reported back, and the International Typographical Union sent people to see how the machine operated. They realized it was inevitable. They realized that the only choice was to take charge of it. So no one was really fighting against the Linotype. What's fascinating was that then in my research, I found a lot of work by the US government with special commissions to understand the impact of industrialization on many industries, but especially in our industry of media. And they found that media grew. And we saw a lot more at first, a lot more publications because it was cheaper to start a publication, but then quickly, there was a consolidation, and we had fewer bigger publications. They were thicker. They had more advertising. They had more revenue. They had larger circulations. It was a real flowering of media in general. As the various media kind of jockeyed for position, it wasn't clear who was gonna win, newspapers, magazines, or books. They were each jealous of each other. Newspapers, started putting magazines in themselves, and they sold themselves for 5¢ on a Sunday, which almost killed magazines as a genre. Magazines started running books. Harper's was a book company that owned a magazine. It circled around until they finally kind of figured out where the musical chairs ended. So there was a — there was an amazing flourishing then of media. This comes in the time of the Gilded Age, the era —

00:11:13 Andrew Keen: Right. So but let me ask you. I wanna come to the Gilded Age and how in many ways in our AI, it seems as if history is repeating itself. But, explain in terms of this new machine that, I mean, would it be right to say that it essentially scaled publishing? Is that the way —

00:11:33 Jeff Jarvis: Yes.

00:11:34 Andrew Keen: It simply made the act of publishing incredibly more efficient.

00:11:40 Jeff Jarvis: Exactly right.

00:11:41 Andrew Keen: To borrow a word from Silicon Valley, a much abused and controversial word scale. Did it make — was it — did it have the same impact? You mentioned magazines, newspapers, and books. Did it have the same impact on all these industries, or was it better for some and worse for others?

00:11:58 Jeff Jarvis: It grew first and fastest in, newspapers, but it was soon in magazines and also in books. There was really in the end one competitor called the Monotype, which became more popular in the UK and Europe, and it was used more in books. But Linotypes were used in all across all media. And, even in small newspapers, there was a Linotype Junior that worked as well. What all the inventors had tried to do before was they tried to replicate the tasks of typesetting. There was making the type, which was a very expensive and artistic process. There was composing the type, one letter after another to put letters into words. There was hyphenation and justification of type to even up the columns. And then when you're done, you had to redistribute the type back. And what the inventors all tried to do was to replicate the past because that's what they knew. They were printers. Mergenthaler was a watchmaker. He didn't know anything. He had a guy named James O. Clephane, who was his muse, who came to his offices, who was a stenographer in Lincoln's cabinet, who wanted to, find more efficient ways to record events. So he came with this idea. Mergenthaler knew nothing about printing, and he had the rare eureka where he realized the way to do this was to create molds for every single letter, multiple molds, and have them come down in place, and then you could mold a whole line of type at a time. And that not only made typesetting a lot cheaper, it also made it faster, which meant that, you could catch up with the telegraph and the speed of news. It meant that you could produce more content and more type, from one shift, from a day, from a typesetter. You could do it cheaper. So, yes, that had huge impact on our industry.

00:13:42 Andrew Keen: I can understand from the productions, from the supply side, it must have been revolutionary. What about from the reader's point of view? You remind us that words matter, and, of course, they matter in many ways, particularly how we read them. Did it — was it self-evident from this revolution, from the reader's point of view, that this technology had changed, or did it look the same when they picked up their morning newspaper or their weekly magazine or one of their books?

00:14:18 Jeff Jarvis: At first, the effort was to make it seem as if there was no change. Just as with Gutenberg, the first fonts looked like scribal handwriting. In that first edition of the New York Tribune that had Linotype type in it, it used the same font. It was on the same page. You had to study the page to find — and it was very difficult to find what was set by machine versus by hand. So at first, it was very much meant to duplicate the experience you had. However, in short order, it became evident that the type was so much better because the type got reused. It was very expensive to buy. So you put it back and reuse it again and again and again. It got inky and messy and bent, and it was miserable in form. If you read, the paper started by, or run by Mark Twain's brother, Orion, it was a mess of typography. It was absolutely completely hard to read. But now with the Linotype, you were making fresh type with every edition. And at the end of that edition, you take the type and just melt it down and then melt new lines. So, yes, I think in the long run, the reader saw that it was much better to read these publications. And then you also got — you know, the other factor that came in here was advertising changed, not using the Linotype so much, but using art fonts. They wanted more space in publications, and so they broke open this invention of white space and design, and things became much airier. We also had new ways to publish, drawings, not wood prints anymore, but also engravings. And so that changed publication. So it wasn't the Linotype alone, but media changed, in many ways around that period.

00:15:53 Andrew Keen: The subtitle of the book is The Magnificent Machine That Gave Birth to Mass Media. It's by definition — you couldn't, I'm guessing, Jeff, you couldn't have had mass media without this new invention. Could you? I mean, it wouldn't be possible.

00:16:09 Jeff Jarvis: No. The —

00:16:10 Andrew Keen: Or an invention like it, whether it's —

00:16:12 Jeff Jarvis: Like it. Yes.

00:16:13 Andrew Keen: Like it's this machine or another machine. But by definition, this represents the birth of mass media.

00:16:20 Jeff Jarvis: So Whitelaw Reid, my villain, the publisher of the New York Tribune, who stole it out from underneath Horace Greeley, its founder —

00:16:26 Andrew Keen: I'm guessing, given your, at least from what I remember, your dislike of Rupert Murdoch that there is a Murdochian quality to him, is there?

00:16:36 Jeff Jarvis: Yes. There is. He ran for vice president. He was a bit of a jerk. He forced Mergenthaler out of the company with his name. Yeah. But he was basically just a capitalist who understood the value of capital and of equity. So in 1879, the decade before the Linotype, Whitelaw Reid gave a speech to a bunch of editors, and he predicted that the newspaper would never expand past eight pages, that it would not be supported by advertising at all because you couldn't cram the ads in. And as papers increased in size because of steam-powered presses, the advertisers wouldn't be able to support the audience. He was completely wrong.

00:17:11 Andrew Keen: It was like the guy who said — who was the guy who said the Internet would never catch on?

00:17:15 Jeff Jarvis: That was somebody I used to work for at Time Inc. who called it a black hole. The guy who invented the business model for, media, for mass media, I wrote about in my book Magazine, is named Frank Munsey, who ran a magazine called, Munsey's quite egotistically. And he was losing money selling the magazine, 25, 35 cents a copy [ed.: figures as spoken], so he decided one day to lose money on every copy, selling it for a dime. Clearly, he made that up on advertising. Thus, the product was no longer the content. The content was the honey to draw in the audience, and the audience was the product sold to advertisers. Here was the beginnings of the attention economy, which ruled mass media and I think infected the Internet today.

00:17:55 Andrew Keen: And I know you're working on the end of mass media. Maybe that will be, another conversation for another time. You're a keen McLuhanite, Jeff Jarvis, very well versed. And, of course, McLuhan's so famous for his medium is as important — or, the medium is the message. To what extent is this new medium? If we indeed can call it a medium, is that the message? I know in the book, you suggest that, writers as diverse as Nietzsche, Hemingway, and Hermann Hesse, who I know you're all familiar with, were in one way or the other, their work and maybe their professional lives were impacted by this new technology. So did it shape or reshape a new kind of writing, a new kind of art?

00:18:47 Jeff Jarvis: I think altogether these technologies did. Nietzsche used a ball typewriter when he was going blind, and it didn't last very long. But Friedrich Kittler, the critic, said that it changed Nietzsche's way to write and the way he thought. The argument was that Hemingway was created by the typewriter. Hermann Hesse found that typed pages felt so different to him from handwritten pages, from scribal pages. That had a difference on how people, created text because the typewritten page alone, the idea of writing with type, changed our sense of perfection of acceptable quality. And it made things look more acceptable, looked like it was almost printed. And I think that affected how people wrote as a process. It certainly did for me. And then come the computer. I was very — I'm old, and so I was very, very young, reporter on the Chicago Tribune in the nineteen seventies when we got our first computers. And I found that, I was a fast writer because I was on deadline because I was a journalist. When I got that computer, I now could hold on to the type the whole time to [unclear] deadline, and I would write as fast as I could, get a draft done so I know I had it done, and then I would use every minute I had to edit. So the technology made me more of an editor than a writer. And then come desktop publishing when I started Entertainment Weekly where my wife, Tammy Westpark [ed.: surname as spoken; unverified], did all the technology behind it, we were now editing pages in Quark exactly as how the reader would see it. So that made us all into typesetters, and it changed again how I could craft something in the way the reader would see it. In the end, it was all still me. I was just using different technologies. I think this is what we have to grapple with now with AI.

00:20:29 Andrew Keen: Right. That's what I was thinking of. Your piece on words matter, damn it, is, a piece about Anthropic's decision to watermark their, quote, unquote, intelligence. The way you described how you first used, the computer in the seventies is actually rather like how I use Claude. It gives me an outline, which I fundamentally change. So whether or not I'm really borrowing from Claude is debatable. But is anything then in that sense, Jeff, really changed with AI?

00:21:02 Jeff Jarvis: I think it does in nuance. Yes. Potentially. I'm working on the next book. Just yesterday, I've been reading for months and months and months tons of clips about the present business state of mass media. I put them in a huge file, a huge folder. And last night, I gave it to Gemini, and I said, just pull out the facts for me. And it did a good job. Now I'm gonna have to check every one of them because I don't trust it. I told it specifically, don't write for me because I don't wanna find myself influenced by it.

00:21:27 Andrew Keen: And don't say anything nice that these things tend to be way too complimentary.

00:21:31 Jeff Jarvis: They are that. Though in the process of writing Hot Type, I, at one point, wanted to test out, a new model. And so I had an idea about, I won't bore you with it, but paper tape, had a huge influence, after the Morse code in changing text into digits. So I had a thought about this. I'd already written my paragraph, and I thought I'll test out this model. And I threw in, the question that I had just grappled with into the model, and it came back with this wonderful phrase that I wanted to use. And I thought, well, how do I do this? Do I say as observed? I can't do that. I had I did a long footnote confessing that I had brought that in from AI into the writing because I think it's important to be, transparent. And I see that happening more and more. And I don't think there's anything wrong with that as long as what I express, I wanna express, what you express, you wanna express. And there's other things that it can do. I can't draw worth a damn. If I wanna illustrate something on my blog, I can now create a drawing for it. I'm not great at making video. We could now make video explanations. These are just tools in our hands. If we let them take over, we're gonna end up with pap and crap. But if we use them to our own ends, I think that there's opportunity here.

00:22:45 Andrew Keen: As I said, the publishing industry is in utter chaos. Oh, ever since I've started publishing books, I have to admit, it always seems to be an utter chaos. But who knows? Maybe it's in particularly utter chaos these days, particularly when it comes to the employment of writers and, of course, editors. And I know your book addresses this in a way in your treatment of the International Typographical Union, the labor organization that responded or maybe didn't respond to this revolutionary new technology. What are the lessons, Jeff, from how we should respond in economic terms and particularly when it comes to labor of this new technology?

00:23:29 Jeff Jarvis: It's a great question, and there are kind of two phases to this. In the first phase, as I said, the ITU realized that this machine worked and it was inevitable, and so they said, we must run it. We are typesetters. You need us to run it. Mergenthaler, the inventor, agreed that just any typist off the street shouldn't run this thing. People who understood type and the requirements of good typesetting needed to run this machine. So the publishers wanted the machine, so they said, okay. And they lost many jobs in the first few years. But after only five years, they were back to full strength and then they grew tremendously. And the ITU, which was America's first major union, was also, in some ways, the most powerful for years until we get to the nineteen sixties. And then we get to the second lesson. So the first lesson is, if you can take charge of this technology, then do. Then that's the way to usher it in, and that's the way to take advantage of it. However, then came new technologies. Paper tape could run a Linotype as if it were a player piano or a robot. Then came computers, which could hyphenate and justify text, which is a very complicated thing that required brains to operate. Now a computer could do it. And, then came cold type, phototypesetting that rather than setting the type in lead, it was set by photo. Then finally came PostScript. And so in that process, the ITU at first said, we are the all-powerful union. They said in New York publishers, you must, give us the benefits of savings. You must not have any attrition. However, the only way to get benefits from the technology was through attrition. But the union was so powerful it fought. Six of nine New York newspapers died in the sixties and the seventies. And it was after that bad strike, which killed the newspapers, that Bert Powers, the all-powerful head of the New York local Big Six, saw the computer for the first time hyphenating and justifying type at a rate that he said two computers could set every word in New York all the time. And if he'd seen that before the strike, it might have changed his, negotiating. After the strike, he changed his strategy and said, now you can have the automation. I have one key demand, which is that all my present members must have lifetime job guarantees. And the last guy with a lifetime job guarantee from that contract retired from the New York Times about 2012. A different story occurred in the UK with Rupert Murdoch and Wapping. Yeah. There, the unions had kept — yes. There was some automation, but they had kept full strength. They'd kept it out. Murdoch secretly built or fairly secretly built a new plant, supposedly to print a new tabloid, which never existed, in Wapping. And then one day just dropped it on the unions, and he was all ready to turn the switch. And he had Atex computers, and he had new presses, and everything new. And the unions tried to fight this for about a year. And it was under Thatcherite labor rules where when they went on strike, they can all be fired. And so that was the end of the unions. And the other papers, even the Guardian, my liberal friends at the Guardian, once this was done said, you know, hallelujah. We're moving too. And they also moved out of Fleet Street, and that was the end of Fleet Street. So that was also the end. It was one century after that first Linotype was installed at the New York Tribune came the final end of hot type, of typeset in metal. And instead, everything from then on was computerized. So the lessons, finally back to your point, is if you can take charge of this technology, if you can benefit from it, do. But if you recognize that you have no hope, that you will be replaced, that your job is simply not needed anymore because of it, you have little choice but to find something else to do [ed.: as spoken, "but that it doesn't find"].

00:27:06 Andrew Keen: So what should people do in the face of AI then?

00:27:09 Jeff Jarvis: I don't think we know what all it can do yet. You know, there were — how many — two years ago, we thought that there was gonna be a whole new —

00:27:14 Andrew Keen: — you and I do as writers? Should we behave towards Sam Altman and Elon Musk? I mean, neither of us, I think, are particularly big fans of either of them or probably Dario at Anthropic either. Are they the new Murdochs? Are they just another version of the gilded class that you write about in the late nineteenth century?

00:27:37 Jeff Jarvis: I think they are to a great measure. And so the last few weeks, I've seen some hope in the movement toward, and the Chinese are doing this, open-weight models. Not fully open source, but much more open that you can run locally, that you can modify, and that create competition for the frontier companies for OpenAI and Anthropic, and that also bring more control in the hand. And, you know, one of the regrets I have, Andrew, about the Internet age is that I didn't support open source technologies enough. I had my blog on WordPress. It's open source. I pat myself on the back for that. But come Twitter, I said, swirl [unclear]. Okay. I'll move a lot of my public discourse there, and, of course, it could get taken over by the malign narcissist who is Elon Musk. And I think this time around, we need to, try to support open technologies that we could have more control over. Now we're just writers. We're not technologists. But I think that the more that we can use these tools to our benefit to help us organize content, I think NotebookLM — or now called Gemini Notebook [ed.: as spoken, "GeminiLM"] — at Google is an amazing tool. I think we can use this to make other media. We can use this to I mean, I just recorded the audio book for Hot Type, but I can see very soon that you can mimic my voice, come up with a markup language to do it more efficiently than I did because I drove the producer nuts talking this fast. I think we can use these technologies to our benefit. But if we hand over too much to those new big companies, I think we will regret it again. Also, because these companies are run by jerks. I don't much like Altman. I certainly don't like Musk. And Amodei, acts as if he is the paragon of safety, but he has his own skewed definition of what safety means in AI. They're all aiming for this Valhalla of AGI, of artificial general intelligence, which I think is pure BS. These are amazing technologies. They can do great things. We can control them and do great things with them, but I don't think we wanna be hauled along by others' visions for them.

00:29:38 Andrew Keen: Yeah. I agree with a lot of what you're saying. Although, our old friend, I know you used to be rather keen on him. I don't think you are anymore. Mark Zuckerberg has also embraced or in theory, at least embraced open source. So sometimes and, clearly, I don't think he's committed to the philosophy of open source. It just benefits him from a business point of view because it's his way of trying to catch up with the dominant players. So open source is sometimes a little more complicated than it seems. It's not some sort of straightforward good and evil distinction, is it, Jeff?

00:30:12 Jeff Jarvis: No. I think you're absolutely right. It depends on what's done with it. And one of the fears about open-weight, open source, models is that, oh my god, anybody can do anything with them. They'll have no guardrails. Well, I argue that the guardrails are pretty much a fool's errand because no one can predict everything that any malign actor or accidental fool could do with a model. The printing press was a general machine. You couldn't say to Gutenberg, well, you can have this thing, but you can't ever let that guy named, Martin Luther use it in seventy years because that'll mess up the world. Anybody could use the machine in any way. AI is the same. It is a general machine. Yes. You should try to build in some guardrails against some things and dealing with children and so on, but you're never gonna succeed fully. So we as a society have to grapple with the fact that these models are gonna be around. There are some who say that open source and open-weight should be banned because you can't put guardrails in it. I think that's foolish. I think that's a form of regulatory capture for the existing big companies. I think we need open-weight and open source again to build competition, to give us more control, and we're just gonna have to deal with the fact that people could use it for good ends and bad.

00:31:23 Andrew Keen: Jeff, the older you get, the more you look like for people watching, the more you look like Mark Twain. Is there some sort of convergence in?

00:31:32 Jeff Jarvis: No. Nature did this to me with my hair. I don't smoke cigars. But I did find — so, Americans — I went to a Twain conference at Elmira College two weeks ago. These are all Twain scholars unlike me. It was fascinating to be there. And they asked people where they first came across Mark Twain. For people of my age, there was a traveling show with Hal Holbrook called Mark Twain Tonight, and he would get up there on stage, a famous actor in a white suit with a cigar, and he would have a certain voice to imitate what Twain was. And it wasn't totally fictional in its structure, but Holbrook really set that tone. So as I read the audiobook for Hot Type, when I had a block quote of Twain, language, I couldn't stop myself from sounding like Hal Holbrook, which to a lot of readers won't mean a damn thing. But it sounds like somebody from the Midwest talking with a slight weird voice. And, yeah. So maybe I am coming to be like him.

00:32:32 Andrew Keen: And the subtitle of the book is The Magnificent Machine That Gave Birth to Mass Media and Drove Mark Twain Mad, which obviously has a humorous element, and Twain was a humorist. But I'm guessing there was a darker side to that. Did it, in some ways, ruin Twain's life?

00:32:50 Jeff Jarvis: It did. So he fancied himself an investor and inventor. He wanted to become filthy rich and stop writing, actually. And into his life came this guy, James W. Paige, with the Paige Compositor, which was a magnificent machine. If you ever go to Hartford, Connecticut, go to the Mark Twain House there. It's a really wonderful place to go. And in the basement is the only extant version of this machine. There were only two ever built. And so it's this gigantic thing that looks like a railroad with nowhere to go. And it was very complicated, brilliant, but very limited, very expensive. It never could work fully. Twain had constant hope. He threw money after money after money after money into it until he went through his fortune and his wife's family coal fortune. And at the end of it, he went bankrupt, to his shame. His wife insisted that he had to pay back all debtors. Yeah.

00:33:41 Andrew Keen: And bank — and bankruptcy, yeah, at the end of the nineteenth century was a much more shameful affair than it is today.

00:33:47 Jeff Jarvis: But he had a buddy in an oil magnate who convinced him it was the right thing to do, and it was. And it — and because he paid back his debtors, it, improved his reputation, and he ended up life with that. But what's fascinating, Andrew, is — how you see a change in his attitude in his writing at the end of his life. As he was going through all this mess with the Paige Compositor, he was writing A Connecticut Yankee in King Arthur's Court. And real Twain scholars will debate, how much this affected his view. I'm not a Twain scholar, but I think it absolutely did. And you see at the end of that novel, not to have a spoiler here, there's a technological Armageddon, where I think you see Twain having been enamored of technology, having brought technologies to Camelot, recognized the danger of them uncontrolled, and he had his own technologies destroy his technology. His last unfinished book was called No. 44, The Mysterious Stranger. It's a really odd book about, a medieval, castle where printers operated, and there were these strange beasts who could duplicate themselves and duplicate pages. And so this whole idea of scale and publishing, of being able to make multiplications of things, I think freaked him out a little bit at the end. And he saw an unharnessed power to this. So it said that the Paige Compositor robbed him of a sense of humor. I don't think that's quite the case. But it did rob him of his love of technology, I think, without doubt.

00:35:17 Andrew Keen: Someone was gonna make a movie of this rather sad chapter in Twain's life. Maybe they should call it Citizen Twain.

00:35:25 Jeff Jarvis: There was a movie made of his life, and they actually filmed the real Paige Compositor. But the filmmakers from Hollywood found it just too boring and too inscrutable, so they made up a machine with wooden things going on all over and beaks that would pick up type and then ended up biting the inventor, and, they had to make it Hollywood.

00:35:44 Andrew Keen: So, of course, with AI now, anyone, supposedly, in theory, can be Mark Twain or acquire his voice, his humor, his writing skills, maybe in the long run if we get to this much vaunted, much promised AGI. You're editing a new series, with Bloomsbury — Intelligence: AI and Humanity. Are you concerned, Jeff? You and I have crossed swords for many years about the democratizing nature of technology, of the Internet. You were one of the biggest critics of my first book, Cult of the Amateur.

00:36:21 Jeff Jarvis: And you of mine.

00:36:22 Andrew Keen: Yes. And me of yours, and we're still very good friends. Should we be concerned about the democratization of intelligence that AI promises, maybe in theory rather than in practice? But are we going through all this again with this, democratization that will get people to defend one kind of elitism or another or justify?

00:36:51 Jeff Jarvis: You ask the question differently from most people how they ask it, and then you ask it wisely, I think. And then there's two pieces to that. One is the democratization, and, generally, I'm going to be in favor of spreading the power of tools into more hands. I wish we were better educated. I wish we supported education. I wish the Enlightenment weren't being attacked today. I wish technology wouldn't go into so many idiotic hands, but I think democratization is, a worthy goal. But the other word you use, which I think is very important here is intelligence. It's the name of the series. The first book out of the series by Rumman Chowdhury, will be asking what is intelligence. And intelligence is a, construct that has been used differently, used sometimes for eugenics and for exclusion and for, education. And so what is intelligence is, I think, a key question of the age. You have Sam Altman out there acting as if he's gonna basically create vending machines for intelligence, that he will become the vendor of intelligence, that he will sell you intelligence. I think that's BS. I think what he'll sell is faster, better, smarter, compute, that can do, again, amazing things. I don't take anything away from it in that sense, but I don't think that defines intelligence. I think that we have to reconsider. The reason we're doing the series is because I think AI forces us to reconsider, much about our worldview, our Weltanschauung. So it makes us ask again, what is intelligence? Another book is Matthew Kirschenbaum, from the University of Virginia, is writing about the Textpocalypse and how it changes our relationship with text. Another book by, Charlton McIlwain is gonna write about race and AI and oddly, or counterintuitively say that there is an opportunity for Black technologists to take charge and do better than was done by them with the Internet. I'm looking for books about what happens to Ronald Coase's theory of the firm now that we have AI. Does everything need to be done inside companies? Can it be done outside? I certainly wanna get — try to get books about creativity in AI and what our definition is when we use this as a tool of creativity. What about God in AI? When we had people like —

00:39:04 Andrew Keen: All right. Well, I was thinking of God and AI. Pope Leo has, thrown his hat, so to speak, into that debate. Did you get to read his piece? I was — I wasn't struck. I know you probably were perhaps a bit more sympathetic than I am. Do you think it's a good contribution? Well, I can't tell you this. Did you call the Vatican, Jeff, and see if, Leo could add a volume to your —

00:39:32 Jeff Jarvis: I wouldn't mind having him as an author of this. I have tried to track down one of his advisors, and I'm trying to find a way to get closer. I think it would — the encyclical was very good and surprising in some ways. So you have the AI guys who say that, oh, we're so powerful. We're gonna eliminate work, and we have to do, universal income. And the pope said, no. We don't want that. That work brings us dignity. And I think that's the kind of perspective that we need is to remain human, and that's the bottom line of all of this. I — the post you showed about words matter, damn it, is about, Anthropic's efforts to watermark, AI output because

00:40:12 Andrew Keen: Yeah. Keith Teare and I actually do — and That Was The Week show on the weekend we discussed that. I don't have as much of a problem as Keith or you do with it. I mean, if you — it doesn't seem to me as if it — [unclear] Keith seems to think that it's shameful if — or he's fearful that it's a — it will be used as a mark of shame, but I'm not sure if that's how people will use it.

00:40:36 Jeff Jarvis: I think it's possible if you're a student and a machine wrongly accuses you of making this up or cheating or, just because you used it to edit or to translate, I think that's a problem. But when Anthropic revealed the methodology, what they said is that they're gonna it's a word prediction machine so that it will change some words based on a key that it has so that it can, which we won't know it exists, but which they say can tell them whether Anthropic likely created this. So they said in the context of weather, the words overcast and gray, no difference. Well, that's the problem I have with it is there is a difference. And, as humans, as writers and editors, we choose words very carefully for a reason. I'm not suggesting that AI is gonna ever do a good job of that, but for them — for Anthropic and the EU, by extension — to devalue that word choice, to say that all language is fungible, that bothers me. That's a cultural statement that I think, is damaging to our craft.

00:41:39 Andrew Keen: Right. And but coming back to Leo's encyclical. What can we learn then from the International Typographical Union in terms of the value of labor? Is Leo's kind of, in some ways, his Christian socialism. Is that the vision that we need to have, Jeff, in our new age? I know you also pointed out in another interesting piece you're very familiar with, the German media. There's an interesting essay by Georg Diez in [ed.: Die Zeit] about we need a vision now of how to shape this technology. Is Leo our visionary? His Christian socialism, is that the best thing we have these days to deal with a voice?

00:42:25 Jeff Jarvis: I think he's a voice that's important to have in this. He's not being a Luddite. He's recognizing the value it has, but he recognizes the value first of humanity because he's a pope. And I think that what this all forces us to do is to ask the question of ourselves, what does it mean to be human, versus machine?

00:42:45 Andrew Keen: Easy question. Right?

00:42:46 Jeff Jarvis: I think that's a good debate to have constantly.

00:42:49 Andrew Keen: Do you think the word human's valuable given how mixed up we are now with machines? We are screwing —

00:42:55 Jeff Jarvis: — it up, pretty badly. And in terms of just the writing style, I was reading something today that says that, we could find that right now AI mimics us, but then we could end up mimicking the AI more where it mushes in. This is Matthew Kirschenbaum's fear in the book he's writing in the essay he wrote for The Atlantic about the Textpocalypse, is we end up with kind of a extruded gray goo of text in the end. And my hope is that we continue to value poetry and, meaning and rhyme and rhythm and sense and feeling in our language. Our language is valuable, and we've gotta hold on to that. That's the most human thing we have.

00:43:38 Andrew Keen: In other words, words matter, damn it. When it comes to gray goo, there's no one less gray or gooey than my old friend Jeff Jarvis. His book is out tomorrow, Hot Type, and it's not pornography. It's a book about, the mass media, The Magnificent Machine That Gave Birth to Mass Media and Drove Mark Twain Mad. Oh, always nice to have Jeff on the show. One of the —

00:44:01 Jeff Jarvis: Always good to see you, my friend.

00:44:02 Andrew Keen: One of the wisest men in, media. He's lived through, like I have, many of these revolutions. So he has historical insight, and that's what he brings with Hot Type. Jeff, as always, good to talk, and we'll see you again very soon. Thanks so much.