I just submitted a letter opposing the so-called California Journalism Preservation Act that is now going through the Senate. Here’s what I said (I’ll skip the opening paragraph with my journalistic bona fides):
Like other well-intentioned media regulation, the CJPA will result in a raft of unintended and damaging consequences. I fear it will support the bottom lines of the rapacious hedge funds and billionaires who are milking California’s once-great newspapers for cash flow without concern for the information needs of California’s communities. I have seen that first-hand, for I was once a member of the digital advisory board for Alden Capital’s Digital First, owner of the Bay Area News Group. For them, any income from any source is fungible and I doubt any money from CJPA will go to actually strengthening journalism.
The best hope for local journalism is not the old newspaper industry and its lobbyists who seek protectionism. It will come instead from startups, some not-for-profit, some tiny, that serve local communities. These are the kinds of journalists we teach in the Entrepreneurial Journalism program I started at my school. These entrepreneurial journalists will not benefit from CJPA and their ventures could be locked out by this nonmarket intervention favoring incumbent competitors. From a policy perspective, I would like to see how California could encourage new competition, not stifle it. I concur with the April letter from LION publishers.
More important, the CJPA and other legislation like it violates the First Amendment and breaks the internet. Links are speech. Editorial choice is speech. No publisher, no platform, no one should be forced to link or not link to content — especially the kinds of extremist content that is ruining American democracy and that could benefit from the CJPA by giving them an opening to force platforms to carry their noxious speech.
Note well that the objects of this legislation, Facebook and Google, would be well within their rights to stop promoting news if forced to pay for the privilege of linking to it. When Spain passed its link tax, Google News pulled out of the country and both publishers and citizens suffered for years as a result. Meta has just announced that it will pull news off its platforms in Canada as a result of its Bill C-18. News is frankly of little value to the platforms. Facebook has said that less than four percent of its content relates to news, Google not much more. Neither makes money from news.
The CJPA could accomplish precisely the opposite of its goal by assuring that less news gets to Californians than today. The just-released Digital News Report from the Reuters Institute for the Study of Journalism at Oxford makes clear that more than ever, citizens start their news journeys not with news brands but end up there via social media and search:
Across markets, only around a fifth of respondents (22%) now say they prefer to start their news journeys with a website or app — that’s down 10 percentage points since 2018…. Younger groups everywhere are showing a weaker connection with news brands’ own websites and apps than previous cohorts — preferring to access news via side-door routes such as social media, search, or mobile aggregators.
Tremendous value accrues to publishers from platforms’ links. By lobbying against the internet platforms that benefit them, news publishers are cutting off their noses to spite their faces, and this legislation hands them the knife.
In a prescient 1998 paper from Santa Monica’s RAND Corporation, “The Information Age and the Printing Press: Looking Backward to See Ahead,” James Dewar argued persuasively for “a) keeping the Internet unregulated, and b) taking a much more experimental approach to information policy. Societies who regulated the printing press suffered and continue to suffer today in comparison with those who didn’t.” In my new book, The Gutenberg Parenthesis, I agree with his conclusion.
I fear that California, its media industry, its journalists, its communities, and its citizens will suffer with the passage of the CJPA.
I attended a show-cause hearing for two attorneys and their firm who submitted nonexistent citations and then entirely fictitious cases manufactured by ChatGPT to federal court, and then tried to blame the machine. “This case is Schadenfreude for any lawyer,” said the attorneys’ attorney, misusing a word as ChatGPT might. “There but for the grace of God go I…. Lawyers have always had difficulty with new technology.”
The judge, P. Kevin Castel, would have none of it. At the end of the two-hour hearing in which he meticulously and patiently questioned each of the attorneys, he said it is “not fair to pick apart people’s words,” but he noted that the actions of the lawyers were “repeatedly described as a mistake.” The mistake might have been the first submission with its nonexistent citations. But “that is the beginning of the narrative, not the end,” as again and again the attorneys failed to do their work, to follow through once the fiction was called to their attention by opposing counsel and the court, to even Google the cases ChatGPT manufactured to verify their existence, let alone to read what “gibberish” — in the judge’s description—ChatGPT fabricated. And ultimately, they failed to fully take responsibility for their own actions.
Over and over again, Steven Schwartz, the attorney who used ChatGPT to do his work, testified to the court that “I just never could imagine that ChatGPT would fabricate cases…. It never occurred to me that it would be making up cases.” He thought it was a search engine — a “super search engine.” And search engines can be trusted, yes? Technology can’t be wrong, right?
Now it’s true that one may fault some large language models’ creators for giving people the impression that generative AI is credible when we know it is not — and especially Microsoft for later connecting ChatGPT with its search engine, Bing, no doubt misleading more people. But Judge Castel’s point stands: It was the lawyer’s responsibility — to themselves, their client, the court, and truth itself — to check the machine’s work. This is not a tale of technology’s failures but of humans’, as most are.
Technology got blamed for much this day. Lawyers faulted their legal search engine, Fastcase, for not giving this personal-injury firm, accustomed to state courts, access to federal cases (a billing screwup). They blamed Microsoft Word for their cut-and-paste of a bolloxed notorization. In a lovely Gutenberg-era moment, Judge Castel questioned them about the odd mix of fonts — Times Roman and something sans serif — in the fake cases, and the lawyer blamed that, too, on computer cut-and-paste. The lawyers’ lawyer said that with ChatGPT, Schwartz “was playing with live ammo. He didn’t know because technology lied to him.” When Schwartz went back to ChatGPT to “find” the cases, “it doubled down. It kept lying to him.” It made them up out of digital ether. “The world now knows about the dangers of ChatGPT,” the lawyers’ lawyer said. “The court has done its job warning the public of these risks.” The judge interrupted: “I did not set out to do that.” For the issue here is not the machine, it is the men who used it.
The courtroom was jammed, sending some to an overflow courtroom to listen. Thereweresomereportersthere, whose presence the lawyers noted as they lamented their public humiliation. The room was also filled with young, dark-suited law students and legal interns. I hope they listened well to the judge (and I hope the journalists did, too) about the real obligations of truth.
ChatGPT is designed to tell you what you want it to say. It is a personal propaganda machine that strings together words to satisfy the ear, with no expectation that it is right. Kevin Roose of The New York Times asked ChatGPT to reveal a dark soul and he was then shocked and disturbed when it did just what he had requested. Same for attorney Schwartz. In his questioning of the lawyer, the judge noted this important nuance: Schwartz did not ask ChatGPT for explanation and case law regarding the somewhat arcane — especially to a personal-injury lawyer usually practicing in state courts — issues of bankruptcy, statutes of limitation, and international treaties in this case of an airline passenger’s knee and an errant snack cart. “You were not asking ChatGPT for an objective analysis,” the judge said. Instead, Schwartz admitted, he asked ChatGPT to give him cases that would bolster his argument. Then, when doubted about the existence of the cases by opposing counsel and judge, he went back to ChatGPT and it produced the cases for him, gibberish and all. And in a flash of apparent incredulity, when he asked ChatGPT “are the other cases you provided fake?”, it responded as he doubtless hoped: “No, the other cases I provided are real.” It instructed that they could be found on reputible legal databases such as LexisNexis and Westlaw, which Schwartz did not consult. The machine did as it was told; the lawyer did not. “It followed your command,” noted the judge. “ChatGPT was not supplementing your research. It was your research.”
Schwartz gave a choked-up apology to the court and his colleagues and his opponents, though as the judge pointedly remarked, he left out of that litany his own ill-served client. Schwartz took responsibility for using the machine to do his work but did not take responsibility for the work he did not do to verify the meaningless strings of words it spat out.
I have some empathy for Schwartz and his colleagues, for they will likely be a long-time punchline in jokes about the firm of Nebbish, Nebbish, & Luddite and the perils of technological progress. All its associates are now undergoing continuing legal education courses in the proper use of artificial intelligence (and there are lots of them already). Schwartz has the ill luck of being the hapless pioneer who came upon this new tool when it was three months in the world, and was merely the first to find a new way to screw up. His lawyers argued to the judge that he and his colleagues should not be sanctioned because they did not operate in bad faith. The judge has taken the case under advisement, but I suspect he might not agree, given their negligence to follow through when their work was doubted.
I also have some anthropomorphic sympathy for ChatGPT, as it is a wronged party in this case: wronged by the lawyers and their blame, wronged by the media and their misrepresentations, wronged by the companies — Microsoft especially — that are trying to tell users just what Schwartz wrongly assumed: that ChatGPT is a search engine that can supply facts. It can’t. It supplies credible-sounding — but not credible — language. That is what it is designed to do. That is what it does, quite amazingly. Its misuse is not its fault.
I have come to believe that journalists should stay away from ChatGPT, et al., for creating that commodity we call content. Yes, AI has long been used to produce stories from structured and limited data: sports games and financial results. That works well, for in these cases, stories are just another form of data visualization. Generative AI is something else again. It picks any word in the language to place after another word based not on facts but on probability. I have said that I do see uses for this technology in journalism: expanding literacy, helping people who are intimidated by writing and illustration to tell their own stories rather than having them extracted and exploited by journalists, for example. We should study and test this technology in our field. We should learn about what it can and cannot do with experience, rather than misrepresenting its capabilities or perils in our reporting. But we must not have it do our work for us.
Besides, the world already has more than enough content. The last thing we need is a machine that spits out yet more. What the world needs from journalism is research, reporting, service, solutions, accountability, empathy, context, history, humanity. I dare tell my journalism students who are learning to write stories that writing stories is not their job; it is merely a useful skill. Their job as journalists is to serve communities and that begins with listening and speaking with people, not machines.
Image: Lady Justice casts off her scale for the machine, by DreamStudio
Ben Smith picked just the right title for his saga of BuzzFeed, Gawker, and The Huffington Post: Traffic (though in the end, he credits the able sensationalist Michael Wolff with the choice). For what Ben chronicles is both the apotheosis and the end of the age of mass media and its obsessive quest for audience attention, for scale, for circulation, ratings, page views, unique users, eyeballs and engagement.
Most everything I write these days — my upcoming books The Gutenberg Parenthesis in June and a next book, an elegy to the magazine in November, and another that I’m working on about the internet — is in the end about the death of the mass, a passing I celebrate. I write in The Gutenberg Parenthesis:
The mass is the child and creation of media, a descendant of Gutenberg, the ultimate extension of treating the public as object — as audience rather than participant. It was the mechanization and industrialization of print with the steam-powered press and Linotype — exploding the circulation of daily newspapers from an average of 4,000 in the late nineteenth century to hundreds of thousands and millions in the next — that brought scale to media. With broadcast, the mass became all-encompassing. Mass is the defining business model of pre-internet capitalism: making as many identical widgets to sell to as many identical people as possible. Content becomes a commodity to attract the attention of the audience, who themselves are sold as a commodity. In the mass, everything and everyone is commodified.
Ben and the anti-heroes of his tale — BuzzFeed founder Jonah Peretti, Gawker Media founder Nick Denton, HuffPost founder Arianna Huffington, investor Kenny Lerer, and a complete dramatis personae of the early players in pure-play digital media — were really no different from the Hearsts, Pulitzers, Newhouses, Luces, Greeleys, Bennetts, Sarnoffs, Paleys, and, yes, Murdochs, the moguls of mass media’s mechanized, industrialized, and corporate age who built their empires on traffic. The only difference, really, was that the digital moguls had new ways to hunt their prey: social, SEO, clickbait, data, listicles, and snark.
Ben tells the story so very well; he is an admirable writer and reporter. His narrative whizzes by like a local train on the express tracks. And it rings true. I had a seat myself on this ride. I was a friend of Nick Denton’s and a member of the board of his company before Gawker, Moreover; president of the online division of Advance (Condé Nast + Newhouse Newspapers); a board member for another pure-play, Plastic (a mashup of Suck et al); a proto-blogger; a writer for HuffPost; and a media critic who occasionally got invited to Nick’s parties and argued alongside Elizabeth Spiers at his kitchen table that he needed to open up to comments (maybe it’s all our fault). So I quite enjoyed Traffic. Because memories.
Traffic is worthwhile as a historical document of an as-it-turns-out-brief chapter in media history and as Ben’s own memoir of his rise from Politico blogger to BuzzFeed News editor to New York Times media critic to co-founder of Semafor. I find it interesting that Ben does not try to separate out the work of his newsroom from the click-factory next door. Passing reference is made to the prestige he and Jonah wanted news to bring to the brand, but Ben does not shy away from association with the viral side of the house.
I saw a much greater separation between the two divisions of BuzzFeed — not just reputationally but also in business models. It took me years to understand the foundation of BuzzFeed’s business. My fellow media blatherers would often scold me: “You don’t understand, Jeff,” one said, “BuzzFeed is the first data-driven newsroom.” So what? Every newsroom and every news organization since the 1850s measured itself by its traffic, whether they called it circulation or reach or MAUs.
No, what separated BuzzFeed’s business from the rest was that it did not sell space or time or even audience. It sold a skill: We know how to make our stuff viral, they said to advertisers. We can make your stuff viral. As a business, it (like Vice) was an ad agency with a giant proof-of-concept attached.
There were two problems. The first was that BuzzFeed depended for four-fifths of its distribution on other platforms: BuzzFeed’s own audience took its content to the larger audience where they were, mostly on Facebook, also YouTube and Twitter. That worked fine until it didn’t — until other, less talented copykittens ruined it for them. The same thing happened years earlier to About.com, where The New York Times Company brought me in to consult after its purchase. About.com had answers to questions people asked in Google search, so Google sent them to About.com, where Google sold the ads. It was a beautiful thing, until crappy content farms like Demand Media came and ruined it for them. In a first major ranking overhaul, Google had to downgrade everything that looked like a content farm, including About. Oh, well. (After learning the skills of SEO and waiting too long, The Times Company finally sold About.com; its remnants labor on in Barry Diller’s content farm, DotDash, where the last survivors of Time Inc. and Meredith toil, mostly post-print.)
The same phenomenon struck BuzzFeed, as social networks became overwhelmed with viral crap because, to use Silicon Valley argot, there was no barrier to entry to making clickbait. In Traffic, Ben reviews the history of Eli Pariser’s well-intentioned but ultimately corrupting startup Upworthy, which ruined the internet and all of media with its invention, the you-won’t-believe-what-happened-next headline. The experience of being bombarded with manipulative ploys for attention was bad for users and the social networks had to downgrade it. Also, as Ben reports, they discovered that many people were more apt to share screeds filled with hate and lies than cute kittens. Enter Breitbart.
BuzzFeed’s second problem was that BuzzFeed News had no sustainable business model other than the unsustainable business model of the rest of news. News isn’t, despite the best efforts of headline writers, terribly clickable. In the early days, BuzzFeed didn’t sell banner ads on its own content and even if it had, advertisers don’t much want to be around news because it is not “brand safe.” Therein lies a terrible commentary on marketing and media, but I’ll leave that for another day.
Ben’s book comes out just as BuzzFeed killed News. In the announcement, Jonah confessed to “overinvesting” in it, which is an admirably candid admission that news didn’t have a business model. Sooner or later, the company’s real bosses — owners of its equity — would demand its death. Ben writes: “I’ve come to regret encouraging Jonah to see our news division as a worthy enterprise that shouldn’t be evaluated solely as a business.” Ain’t that the problem with every newsroom? The truth is that BuzzFeed News was a philanthropic gift to the information ecosystem from Jonah and Ben.
Just as Jonah and company believed that Facebook et al had turned on them, they turned on Facebook and Google and Twitter, joining old, incumbent media in arguing that Silicon Valley somehow owed the news industry. For what? For sending them traffic all these years? Ben tells of meeting with the gray eminence of the true evil empire, News Corp., to discuss strategies to squeeze “protection money” (Ben’s words) from technology companies. That, too, is no business model.
Thus the death of BuzzFeed news says much about the fate of journalism today. In Traffic, Ben tells the tale of the greatest single traffic driver in BuzzFeed’s history: The Dress. You know, this one:
At every journalism conference where I took the stage after that, I would ask the journalists in attendance how many of their news organizations wrote a story about The Dress. Every single hand would go up. And what does that say about the state of journalism today? As we whine and wail about losing reporters and editors at the hands of greedy capitalists, we nonetheless waste tremendous journalistic resource rewriting each other for traffic: everyone had to have their own story to get their own Googlejuice and likes and links and ad impressions and pennies from them. No one added anything of value to BuzzFeed’s own story. The story, certainly BuzzFeed would acknowledge, had no particular social value; it did nothing to inform public discourse. It was fun. It got people talking. It took their attention. It generated traffic.
The virus Ben writes about is one that BuzzFeed — and the every news organization on the internet and the internet as a whole — caught from old, coughing mass media: the insatiable hunger for traffic for its own sake. In the book, Nick Denton plays the role of inscrutable (oh, I can attest to that) philosopher. According to Ben, Nick believed that traffic was the key expression of value: “Traffic, to Nick … was something pure. It was an art, not a science. Traffic meant that what you were doing was working.” Yet Nick also knew where traffic could lead. Ben quotes him telling a journalist in 2014: “It’s not jonah himself I hate, but this stage of internet media for which he is so perfectly optimized. I see an image of his cynical smirk — made you click! — every time a stupid buzzfeed listicle pops on Facebook.”
Nick also believed that transparency was the only ethic that really mattered, for the sake of democracy. Add these two premises, traffic and transparency, together and the sex tape that was the McGuffin that brought down Gawker and Nick at the hands of Peter Thiel was perhaps an inevitability. Ben also credits (or blames?) Nick for his own decision to release the Trump dossier to the public on BuzzFeed. (I still think Ben has a credible argument for doing so: It was being talked about in government and in media and we, the public, had the right to judge for ourselves. Or rather, it’s not our right to decide; it’s a responsibility, which will fall on all of us more and more as our old institutions of trust and authority — editing and publishing — falter in the face of the abundance of talk the net enables.)
The problem in the end is that traffic is a commodity; commodities have no unique value; and commodities in abundance will always decrease in price, toward zero. “Even as the traffic to BuzzFeed, Gawker Media, and other adept digital publishers grew,” Ben writes, “their operators began to feel that they were running on an accelerating treadmill, needing ever more traffic to keep the same dollars flowing in.” Precisely
Traffic is not where the value of the internet lies. No, as I write in The Gutenberg Parenthesis (/plug), the real value of the internet is that it begins to reverse the impact print and mass media have had on public discourse. The internet devalues the notions of content, audience, and traffic in favor of speech. Only it is going to take a long time for society to relearn the conversational skills it has lost and — as with Gutenberg and the Reformation, Counter-Reformation, and Thirty Years’ War that followed — things will be messy in between.
BuzzFeed, Gawker, The Huffington Post, etc. were not new media at all. They were the last gasp of old media, trying to keep the old ways alive with new tricks. What comes next — what is actually new — has yet to be invented. That is what I care about. That is why I teach.
My father died on Saturday, April 8. He lived 97 years. Until struck with COVID, he had never had to stay in a hospital. In the last three months, he mustered all his strength to overcome the virus’ results: internal bleeding, then post-COVID pneumonia, then the side-effects of medication, and finally one more case of pneumonia. I curse the disease and all it does. He passed on peacefully last night after we — his entire family, the five of us — had the blessing of spending his last day with him, affirming our love.
I am writing this only for myself. I’m not writing it for him; he outlived everyone he knew. Neither am I writing it for you; I don’t expect you to read this, for you did not know him. I find such memorials for loved ones, including pets, in social media understandable but difficult, for I never know how to react. I do not expect you to. I simply want to memorialize my father, to leave a trace of his life connected with mine here. As an old newspaperman, I understand the value of the obituary more than the grave.
Darrell V. Jarvis was born in the tiny house on his grandfather’s rocky, dirt-poor farm up the holler behind the Methodist church in Weston, West Virginia. His parents were not much educated, Buck finishing the seventh grade, Vera not a lot more. They worked hard and moved often, following drilling crews to gas fields in West Virginia, Kentucky, and southern Illinois, my father attending eight schools along the way. His parents insisted that he and his brother — my late Uncle Richard, a church musician — attend college. My father graduated from the University of Illinois after completing a first stint of service in the Navy at the end of World War II. At U of I, he met my mother, Joan Welch, the daughter of a country doctor and a nurse in Lewistown, Illinois. They had two children: me and my sister, Cynthia, a Presbyterian minister. We lost our mother almost six years ago.
Darrell studied engineering but didn’t much like it. He preferred people. So, after serving in the Navy again in the Korean War, as a lieutenant on the destroyer USS Renshaw, he moved up as a “peddler” — his word — in the electronic and electrical industries, rising to be a VP of sales. He was on the road constantly, “with an airplane strapped to my backside,” and we moved constantly, from Illinois to Iowa to New Jersey to New York to Illinois and then — after Cindy graduated and I, having also attended eight schools, left the nest — to Seattle, California, and Illinois, then back to New Jersey, and finally to three or four places in Florida. They bought many window treatments.
Our mother was quite shy. Our father was the opposite. They did attract. The joke in the family was that by the time he reached the box office at the movie theater, Darrell would be lifelong friends with whoever was behind and in front of him in line, sharing their stories. Meanwhile, we hid. If they ever make a Mad Men about nice people, my father would be the model: the handsome and charming guy in the gray suit with hat and briefcase, for years puffing a pipe, tending to our suburban lawns, playing golf at every opportunity, going to church, sipping bourbon and later martinis at cocktail parties, voting Republican until Trump. Darrell was middle America.
At age 62, Darrell became the model or cliché of the retired American. He and Joan moved to a town in Florida where you must be 55 years old to live (I so want to write that sitcom). He could golf constantly until he needed to stay near Joan because of the type-1 diabetes he nursed her through for more than 50 years. And his knees gave out.
My parents were wonderful grandparents to our children, Jake and Julia. Oh, how they love their PopPop and Mimi. And oh, how my parents adored my wife, Tammy.
I learned much from my father. He so wanted to teach me golf so we could bond on the course, but after I once hit him in the shin with a driver and he hit me in the head with a stray ball, we gave up. He tried to teach me the handy skills, but while watching me try to build a case for the amplifier for my eighth-grade science-fair project (electronic bongos), he shook his head and declared: “You’re so clumsy you couldn’t stick your finger up your ass with two hands.” He said it with love and laughter as well as exasperation. We gave up that ambition, too.
What I did learn from my father more than anything else was ethics. I saw how he did business. I watched how he treated the staff who worked for him. I listened to him teach me how to deal with office politics: Never lower yourself to their level. I will always be proud of him that when he found himself in a meeting that turned out to be about price-fixing, he got up, protested, and left, risking his career. I tried to learn charm from him — even at the end, his smile and warmth would win over every nurse and aide. I yet wish I could learn to be the father he was.
We tried for years to get him and my mother — then him alone — to move up to be near my sister and us. In the summer of 2021 — fully vaccinated — he caught COVID for the first time in his retirement community and spent 11 days in the hospital and a month in rehab before moving to assisted living. Finally, he agreed to move. Tammy asked him: Why not before? “I’m just stupid,” he said. He came to an assisted living community five miles from our home and said he was living where he should. Thus the perverse blessing of COVID was that it brought him to us for a magnificent year and a half we otherwise would not have had — until it stole him from us. We saw him every day. Sister Cindy would come up with her Scottie, Phoebe, to sit in his lap. My dear Tammy took away all his worries of finance and life; he heeded her. She insisted that we have him over to our house three nights a week, culminating in thrill rides on a ramp after his knees finally gave out and he resorted to a wheelchair. He asked me one day to find him an electric wheelchair and that was a wonder to behold, him hot-rodding to meals and bingo and happy hour, balancing a martini in one hand, steering the wheelchair in the other, miraculously managing not to hit any old ladies.
There has been muchpraiseinhumanchat — Twitter — about Ted Chiang’s New Yorker piece on machine chat — ChatGPT. Because New Yorker; because Ted Chiang. He makes a clever comparison between lossy compression — how JPEGs or MP3s save a good-enough artifact of a thing, with some pieces missing and fudged to save space — and large-language models, which learn from and spit back but do not record the entire web. “Think of ChatGTP as a blurry JPEG of all the text on the Web,” he instructs.
What strikes me about the piece is how unselfaware media are when covering technology.
For what is journalism itself but lossy compression of the world? To save space, the journalist cannot and does not save or report everything known about an issue or event, compressing what is learned into so many available inches of type. For that matter, what is a library or a museum or a curriculum but lossy compression — that which fits? What is culture but lossy compression of creativity? As Umberto Eco said, “Now more than ever, we realize that culture is made up of what remains after everything else has been forgotten.”
Chiang analogizes ChatGPT et al to a computational Xerox machine that made an error because it extrapolated one set of bits for others. Matthew Kirschenbaum quibbles:
Agreed. This reminds me of the sometimes rancorous debate between Elizabeth Eisenstein, credited as the founder of the discipline of book history, and her chief critic, Adrian Johns. Eisenstein valued fixity as a key attribute of print, its authority and thus its culture. “Typographical fixity,” she said, “is a basic prerequisite for the rapid advancement of learning.” Johns dismissed her idea of print culture, arguing that early books were not fixed and authoritative but often sloppy and wrong (which Eisenstein also said). They were both right. Early books were filled with errors and, as Eisenstein pointed out, spread disinformation. “But new forms of scurrilous gossip, erotic fantasy, idle pleasure-seeking, and freethinking were also linked” to printing, she wrote. “Like piety, pornography assumed new forms.” It took time for print to earn its reputation of uniformity, accuracy, and quality and for new institutions — editing and publishing — to imbue the form with authority.
That is precisely the process we are witnessing now with the new technologies of the day. The problem, often, is that we — especially journalists — make assumptions and set expectations about the new based on the analog and presumptions of the old.
Media have been making quite the fuss about ChatGPT, declaring in many a headline that Google better watch out because it could replace its Search. As we all know by now, Microsoft is adding ChatGPT to its Bing and Google is said to have stumbled in its announcements about large-language models and search last week.
But it’s evident that the large-language models we have seen so far are not yet good for search or for factual divination; see the Stochastic Parrots paper that got Tinmit Gebru fired from Google; see also her coauthor Emily Bender’s continuing and cautionary writing on the topic. Then read David Weinberger’s Everyday Chaos, an excellent and slightly ahead of its moment explanation of what artificial intelligence, machine learning, and large language models do. They predict. They take their learnings — whether from the web or some other large set of data — and predict what might happen next or what should come next in a sequence of, say, words. (I wrote about his book here.)
Said Weinberger: “Our new engines of prediction are able to make more accurate predictions and to make predictions in domains that we used to think were impervious to them because this new technology can handle far more data, constrained by fewer human expectations about how that data fits together, with more complex rules, more complex interdependencies, and more sensitivity to starting points.”
To predict the next, best word in a sequence is a different task from finding the correct answer to a math problem or verifying a factual assertion or searching for the best match to a query. This is not to say that these functions cannot be added onto large-language models as rhetorical machines. As Google and Microsoft are about to learn, these functions damned well better be bolted together before LLMs are unleashed on the world with the promise of accuracy.
When media report on these new technologies they too often ignore underlying lessons about what they say about us. They too often set high expectations — ChatGPT can replace search! — and then delight in shooting down those expectations — ChatGPT made mistakes!
Chiang wishes ChatGPT to search and calculate and compose and when it is not good at those tasks, he all but dismisses the utility of LLMs. As a writer, he just might be engaging in wishful thinking. Here I speculate about how ChatGPT might help expand literacy and also devalue the special status of the writer in society. In my upcoming book, The Gutenberg Parenthesis (preorder here /plug), I note that it was not until a century and a half after Gutenberg that major innovation occurred with print: the invention of the essay (Montaigne), the modern novel (Cervantes), and the newspaper. We are early our progression of learning what we can do with new technologies such as large-language models. It may be too early to use them in certain circumstances (e.g., search) but it is also too early to dismiss them.
It is equally important to recognize the faults in these technologies — and the faults that they expose in us — and understand the source of each. Large-language models such as ChatGPT and Google’s LaMDA are trained on, among other things, the web, which is to say society’s sooty exhaust, carrying all the errors, mistakes, conspiracies, biases, bigotries, presumptions, and stupidities — as well as genius — of humanity online. When we blame an algorithm for exhibiting bias we should start with the realization that it is reflecting our own biases. We must fix both: the data it learns from and the underlying corruption in society’s soul.
Chiang’s story is lossy in that he quotes and cites none of the many scientists, researchers, and philosophers who are working in the field, making it as difficult as ChatGPT does to track down the source of his logic and conclusions.
The lossiest algorithm of all is the form of story. Said Weinberger:
Why have we so insisted on turning complex histories into simple stories? Marshall McLuhan was right: the medium is the message. We shrank our ideas to fit on pages sewn in a sequence that we then glued between cardboard stops. Books are good at telling stories and bad at guiding us through knowledge that bursts out in every conceivable direction, as all knowledge does when we let it. But now the medium of our daily experiences — the internet — has the capacity, the connections, and the engine needed to express the richly chaotic nature of the world.
In the end, Chiang prefers the web to an algorithm’s rephrasing of it. Hurrah for the web.
We are only beginning to learn what the net can and cannot do, what is good and bad from it, what we should or should not make of it, what it reflects in us. The institutions created to grant print fixity and authority — editing and publishing — are proving inadequate to cope with the scale of speech (aka content) online. The current, temporary proprietors of the net, the platforms, are also so far not up to the task. We will need to overhaul or invent new institutions to grapple with issues of credibility and quality, to discover and recommend and nurture talent and authority. As with print, that will take time, more time than journalists have to file their next story.
Original painting by Johannes Vermeer; transformed (pixelated) by acagastya., CC0, via Wikimedia Commons