There is so much to parse in this Times column inspired by a paper examining alleged political leanings of large language models.
First, the myth of a “center” is imposed on the machine as it is on journalism. That is an impossibility, especially when extremists weigh down the equation & move “center” by gravity downhill, towards them.
Second, in its raw state the model reflects the collected corpus of digital content from those who had the power to publish. Thus, it will reflect that worldview; it is a reflection of that power. Imposing left/right/center on that says little about the machine, much about the that imposition.
But, third, given that — as the Stochastic Parrots paper preaches — the models are too huge to audit. It is impossible to judge the effect of the choices made in training them. That is the problem with size-matters, macho model-making.
Fourth, when, as the article says, models are “fine-tuned,” some of that effort comes in reaction to fears of political and media pressure on the model-makers, to compensate for what is lacking in the material used to train the model. Thus, the tuning says more about that pressure and those fears than it does about the technology per se. See: the fuss over Gemini’s images. So when a model puts a Black person at the US Constitutional Convention, its makers are trying to account for society’s biases; when it is called “woke” by the right, that reveals their further biases. The issues are all human.
All this is why I argue that we must study technology in the context of humanity, examining not the software as if it had a worldview but instead understanding the conflicting worldviews imposed on it and what that reveals not about the machine but about us.
The tl;dr of all this is that just as there is no mythical center in politics, there is no neutrality in technology (AI or social media) and there is no objectivity in journalism. Each is an attempt to impose a given view as a norm.
In my upcoming book (coming not soon enough), The Web We Weave: Why We Must Reclaim the Internet from Moguls, Misanthropes, and Moral Panic, I quote Terrence Sejnowski on his theory of the Reverse Turing Test in the context of Kevin Roose’s affair with ChatGPT. The episode says more about the reporter than the machine he reported on . That snippet:
In my book, I argue we should understand the internet not as a technology but as a human network and enterprise. Similarly, we should not try to analyze the biases in AI so much as we should endeavor to \understand human biases imposed upon it in design or reaction.
AI as it exists could be understood as a reflection on society, in what it summarizes of the collection of human text and image it is trained on, in how our reaction to it exposes our fears or dreams, in how we choose to command it.
If we saw generative AI as a creative tool, then I’d say imagining the founding of America with women & Black people at the table and the Catholic Church headed by Black women and Native Americans is a proper revision of history the way it should have been.
The reaction to #WokeGemini says more about society than the tool itself; that’s what fascinates me about AI: its reflections. Right-wing columnists fear the anti-white machine programmed by commisars of what we used to call political correctness.
And the extremist Murdoch media and pols hunt for enemies: technologists who dare to recognize “white privilege.” Their strategy of projection works in media today: calling anti-racism racism to deflect from their deep racism, just as calling Biden old and doddering deflects from the truly doddering Trump. In today’s credulous media, it works.
Ultimately, this little episode shows the folly of “guardrails” in AI, for a tool can be made to do anything and what it is proscribed from doing — and reaction to that — can be more revealing and risky than the tool alone. It is like telling Gutenberg what movable type must never say. The Church tried to do just that with his successors as, at first, the technology was held responsible for what it produced: printers were beheaded and behanded for what their machines produced. Then booksellers were held liable and controlled via licensing and the Stationers Company in Britain. Finally, authors were the responsible parties — and Foucault says that is the birth of the author.
We debate responsibility over the technology of AI today: Some want the model makers to be responsible for everything that could (or now could not) be produced from their machine. Some try to blame an application of the model (looking at you, Air Canada). Some want to blame the technology for making just what it was asked to make (looking at a certain lawyer and a certain reporter).
The episode also reveals the fraud in associating generative AI with truth. If it were seen instead as a concordance of all society’s biased text and images and a creative tool that can be told to remake that, then there’d be no story here, only interesting reflections of ourselves, our aspirations, our faults, and our fears.
I am coming to a conclusion I have avoided for my last three decades working on the internet and news: It may finally be time to give up on old journalism and its legacy industry.
I say this with no joy, no satisfaction at having tried to get newspapers and magazines to change, and much empathy for the journalists and others caught working in a dying sector and those who count on them. But the old news industry is gasping for air. I’m not suggesting performing euthenasia on what is left. Nor do I dance on the grave. In my time running a Center for Entrepreneurial Journalism, now ending, I have tried to balance support for startups and legacy companies. But I wonder whether it is time stop throwing good money and effort after bad.
The old news industry has failed at adapting to the internet and every one of their would-be saviors — from tablets to paywalls to programmatic ads to consolidation to billionnaires — has failed them. Hedge funds have bought up chains and papers, selling everything not bolted down, cutting every possible cost, and taking every penny of cash flow home with them. The one thing the old companies are still investing in is lobbying.
In my testimony in the Senate last week, I engaged in the wishful if futile act of urging the legislators not to enact protectionist legislation written with the industry lobbyists who sat beside me, but instead to support the emergent reinvention of journalism occurring in communities everywhere. Not likely.
The bad news for news is constant. Just in the last month, the Los Angeles Times islaying off 115 people, throwing its newsroom into “chaos” and “mayhem.”Time magazine lays off 15 percent of its unionized editorial staff. Meanwhile, The New York Timeschronicles the pains of billionnaires losing fortunes trying to save old news. I didn’t think it was possible for hedge funds to torture journalism more, but Alden just did something even worse than buying a paper: sellingThe Baltimore Sun to the mini-Murdoch, David Smith, chair of right-wing Sinclair, hater of news and newspapers. In the UK, the Mirror’s circulation has fallen from 5 million to a quarter million, its local papers are sputtering, and the company predicts print will be unsustainble — something I’ve been warning about for two decades. Once-grand Sports Illustrated is being murdered in plain sight. The FCC just announced it is trying to support local TV news, nevermind that audience for local broadcast news is small, old, and dying … and more and more made up of people watching the faux Fox, Sinclair.
Meanwhile, trust in journalism falls to ever-lower records. The Reuters Institute at Oxford tells us that a third of people actively avoid news, and who can blame them? I myself am fed up with old news’ wishful doomsaying, its credulous coverage of politics as sport, its bothsidesing and normalization of the rise of populist fascism, its refusal to call racism racism, its chronic lack of diversity, its dependence on access to power, its moral panic about technology, and the resource it wastes on copying and clickbait. Semafor and Gallup report that trust in journalists is falling now among Democrats, too.
And now here comes artificial intelligence to manufacture and devalue that thing we call content, robbing the old news industry of its sense of value and purpose in making the commodity. I’ve been trying to convince news organizations that they are not, or should not be, in content business, but that journalism is instead a service built on conversation, community, and collaboration. I have failed.
Of course, there are exceptions. The Boston Globe and StarTribune seem to be surviving or better. My old colleagues at Advance are innovating in Alabama, living on past print. (In his lengthy lamentation on death in news, Ezra Klein lists Alabama going out of print as a loss when I say it is a victory: life after the death of the press.) The Times is growing on the backs of games and food. The National Trust for Local News is saving papers here and there.
But then there’s Scranton, its paper now in the clutches of Alden. The Washington Post has chroncled their pain. On Feb. 9, I’ll be speaking at the University of Scranton’s Schemel Forum about what to do now. What should I tell them?
I will warn them to expect cutbacks and no investment or innovation at their dear old Times-Tribune. I’ve seen how Alden operates. As a member of a Digital First advisory board a decade ago, I saw the company innovate under John Paton and Jim Brady, but when that didn’t yield a sale in 2015, both of them left and the hedgies proceeded to cut to the marrow.
I come with no solution, no salvation; nothing’s that simple. There are many examples of people trying to find new futures for news. In my Senate testimony, I spoke of the 450 members of the New Jersey News Commons, which I’m proud to have helped start a decade ago at Montclair State University; and the 475 members of LION, the Local Independent Online News Publishers; and the 425 members of INN, the Institute for Nonprofit News. See also today’s news that The 19th is starting a new network for sharing news (something I tried in New Jersey years ago). This is where innovation in news is occurring: bottom-up, grass-roots efforts emergent in communities.
But as my old friend and colleague Peter Bhatia said when he made the controversial decision of dismissing the editor of the new Houston Landing, “We’re basically putting out a newspaper on the web. And that’s not a recipe for success for us for the long term, nor is it a recipe for sustainability.” I don’t know Houston Landing well enough to comment but I do worry that some of the efforts at new news still emulate and aspire to the form and function of old news.
I think we need to be more radical than that, much more radical than I have been.
I note with optimism Mike Masnick’s just released report, The Sky is Rising, about the impact of the internet on media writ large — reading, watching, listening, and playing. It concludes, “More creative content is being produced that ever before. More people are able to create content than ever before, and more people are able to make money doing so…. And almost all of this is thanks to the power of the internet.” The report is talking mostly about entertainment but also notes that according to Census Bureau data, “it appears that internet publishing jobs more than replaced the jobs lost in newspapers and periodicals.”
There can be life after legacy. There will be roles for journalists. But journalism schools must expand their horizon to teach more than making content. How do we serve many publics in a networked world?
For the last two decades at least, I have told newspaper editors and publishers that they must imagine a day when print is no longer sustainable, and if they are not profitable digitally by then, they will die. Now I will tell the good people of Scranton to imagine a day when their paper dies, or is as good as dead.
What then? Citizens will have to come together to understand their needs as a community: for information, yes, and also for understanding, collaboration, accountability, repair, and service. They will need to decide what is best for Scranton and its many communities.
They might find some help, though never enough. Press Forward is bringing $500 million to the effort, but that can stretch only so far. The New Jersey Civic Information Consortium is doing interesting work granting state funds to bolster innovation. Perhaps Pennsylvania could do likewise. (Though I worry about what equivalent efforts in Texas, Florida, or Oklahoma would support.) California, like the US Congress with its JCPA, are talking about helping news — but what they’re actually doing is looking to blackmail tech companies on behalf of legacy news companies and their hedge-fund owners. JCPA specifically excludes news enterprises making less than $100,000 — which is to say most of those hundreds of innovators I listed above. No thank you.
The way out of this will be to educate and empower our next generation, not in so-called media literacy, but in media leadership, in taking responsibility for the health of their communities and their public discourse. That is a big, complex, nuanced, unsure order that will require marshalling the wisdom of disciplines far beyond journalism: history, anthropology, sociology, psychology, community studies, ethics, design, and the arts.
I am afraid to say that the newspaper and TV and commercial radio station of today are inadequate to the task. Their news was invented in the long century of mass media, which began (as I recount in Magazine) when Frank Munsey realized he could sell his eponymous periodical at a dime and a loss, but profit by selling his audience’s attention to advertisers. Thus was born the attention economy that now corrupts not only old media but new. The internet isn’t killing news. It is killing the mass and the myth that kept media alive all these years: that our attention is a commodity to be owned, bought, and sold.
I say this with reluctance and sadness but also with hope, for I am priviliged to watch some of my alumni try to create a new journalism at human scale, built on listening and serving communities, not nostalgia. How might Scranton do that? That will be up to Scranton, not to the heartless hedge fund — the Dunder-Mifflin of newspapering — that has come to town.
I wrote an op-ed for NJ.com and the Star-Ledger in New Jersey proposing that the soon-empty Bell Labs should become a Museum and School of the Internet. Here, for those outside the Garden State, is the text:
Bell Labs, the historic headwaters of so many inventions that now define our digital age, is closing in Murray Hill, its latest owners moving to more modern headquarters in New Brunswick. The Labs should be preserved as a historic site and more. I propose that Bell Labs be opened to the public as a museum and school of the internet.
The internet would not be possible without the technologies forged at Bell Labs: the transistor, the laser, information theory, Unix, communications satellites, fiber optics, advances in chip design, cellular phones, compression, microphones, talkies, the first digital art, and artificial intelligence — not to mention, of course, many advances in networks and the telephone, including the precursor to the device we all carry and communicate with today: the Picturephone, displayed as a futuristic fantasy at the 1964 World’s Fair.
There is no museum of the internet. Silicon Valley has its Computer History Museum. New York has museums for television and the moving image. Massachusetts boasts a charming Museum of Printing. Search Google for a museum of the internet and you’ll find amusingdigitalartifacts, but nowhere to immerse oneself in and study this immensely impactful institution in society.
Where better to house a museum devoted to the internet than New Jersey, home not only of Bell Labs but also at one time the headquarters of the communications empire, AT&T, our Ma Bell?
I remember taking a field trip to Bell Labs soon after this web site, NJ.com, started in 1995. I was an executive of NJ.com’s parent company, Advance. My fellow editors and I felt we were on the sharp edge of the future in bringing news online.
We thought that earned us kinship with the invention of that future that went on at Bell Labs, so we arranged a visit to the awe-inspiring building designed by Stephen F. Voorhees and opened in 1941. The halls were haunted with genius: lab after lab with benches and blackboards and history within. We must not lose that history.
We also must not lose the history of the internet as it passes us by in present tense. In researching my book, “The Gutenberg Parenthesis: The Age of Print and its Lessons for the Age of the Internet,” I was shocked to discover that there was not a discipline devoted to studying the history and influence of print and the book until Elizabeth Eisenstein wrote her seminal work, “The Printing Press as an Agent of Change,” in 1979, a half-millennium after Gutenberg. We must not wait so long to preserve memories and study the importance of the net in our lives.
The old Bell Labs could be more than a museum, preserving and explaining the advances that led to the internet. It could be a school. After leaving Advance in 2006, I became a journalism professor at CUNY’s Newmark School of Journalism, from which I am retiring.
I am less interested now in studying journalism than in the greater, all-enveloping subject: the internet. My dream is to start a new educational program in Internet Studies, to bring the humanities and social sciences to research the internet, for it is much more than a technology; it is a human network that reflects both human accomplishment and human failure.
Imagine if Bell Labs were a place where scholars and students in many disciplines — technologies, yes, but also anthropology, sociology, psychology, history, ethics, economics, community studies, design — could gather to teach and learn, discuss and research.
Imagine, too, if a New Jersey university could use the space for classes and events.
There is a model for this in New Jersey in what Montclair State University is doing in Paterson, developing and operating a museum devoted to the history of Negro League baseball in the historic Hinchcliffe Stadium. This is the kind of university-community collaboration that could enrich the space of Bell Labs with energy and life.
There is some delicious irony in proposing that the internet be memorialized in what was once an AT&T facility, for the old telephone company resisted the arrival of the internet, hoping we would pay by the minute for long-distance calls forever.
In 1997, David Isenberg, a 12-year veteran of Bell Labs, wrote an infamous memo telling his bosses they were wrong to build intelligent networks and should instead learn the value of the stupid network that anyone could connect to: the internet.
Isenberg’s web site says the memo “was received with acclaim everywhere in the global telecommunications community with one exception — at AT&T itself! So Isenberg left AT&T in 1998.”
How wonderful if, in the end, Bell Labs could claim to become a forever home for that network that has changed the world.
Well, that was surreal. I testified in a hearing about AI and the future of journalism held by the Senate Judiciary Subcommittee on Privacy, Technology, and the Law. Here is my written testimony and here’s the Reader’s Digest version in my opening remarks:
It was a privilege and honor to be invited to air my views on technology and the news. I went in knowing I had a role to play, as the odd man out. The other witnesses were lobbyists for the newspaper/magazine and broadcast industries and the CEO of a major magazine company. The staff knew I would present an alternative perspective. My fellow panelists noted before we sat down — nicely — that they disagreed with my written testimony. Job done. There was little opportunity to disagree in the hearing, for one speaks only when spoken to.
What struck me about the experience is not surprising: They call the internet an echo chamber. But, of course, there’s no greater echo chamber than Congress: lobbyists and legislators agreeing with each other about the laws they write and promote together. That’s what I witnessed in the hearing in a few key areas:
Licensing: The industry people and the politicians all took as gospel the idea that AI companies should have to license and pay for every bit of media content they use.
I disagree. I draw the analogy to what happened when radio started. Newspapers tried everything to keep radio out of news. In the end, to this day, radio rips and reads newspapers, taking in and repurposing information. That’s to the benefit of an informed society.
Why shouldn’t AI have the same right? I ask. Some have objected to my metaphor: Yes, I know, AI is a program and the machine doesn’t read or learn or have rights any more than a broadcast tower can listen and speak and vote. I spoke metaphorically, for if I had instead argued that, say, Google or Meta has a right to read and learn, that would have opened up a whole can of PR worms. The point is obvious, though: If AI creators would be required by law to license *everything* they use, that grants them lesser rights than media — including journalists, who, let’s be clear, read, learn from, and repurpose information from each other and from sources every day.
I think there’s a difference in using content to train a model versus producing output. It’s one matter for large language models to be taught the relationship of, say, the words “White” and “House.” I say that is fair and transformative use. But it’s a fair discussion to separate out questions of proper acquisition and terms of use when an application quotes from copyrighted material from behind a paywall in its output. The magazine executive cleverly conflated training and output, saying *any* use required licensing and payment. I believe that sets a dangerous precedent for news media itself.
If licensing and payment is required for all use of all content, then I say the doctrine of fair use could be eviscerated. The senators argued just the opposite, saying that if fair use is expanded, copyright becomes meaningless. We disagree.
JCPA: The so-called Journalism Competition and Preservation Act is a darling of many members of the committee. Like Canada’s disastrous Bill C-18 and Australia’s corrupt News Media Bargaining Code — which the senators and the lobbyists think are wonderful — the JCPA would allow large news organizations (those that earn more than $100,000 a year, leaving out countless small, local enterprises) to sidestep antitrust and gang together and force platforms to “negotiate” for the right to link to their content. It’s legislated blackmail. I didn’t have the chance to say that. Instead, the lobbyists and legislators all agreed how much they love the bill and can’t wait to try again to pass it.
Section 230: Members of the committee also want to pass legislation to exclude generative AI from the protections of Section 230, which enables public discourse online by protecting platforms from liability for what users say there while also allowing companies to moderate what is said. The chair said no witness in this series of hearings on AI has disagreed. I had the opportunity to say that he has found his first disagreement.
I always worry about attempts to slice away Section 230’s protections like a deli balogna. But more to the point, I tried to explain that there is nuance in deciding where liability should lie. In the beginning of print, printers were held liable — burned, beheaded, and behanded — for what came off their presses; then booksellers were responsible for what they sold; until ultimately authors were held responsible — which, some say, was the birth of the idea of authorship.
When I attended a World Economic Forum AI governance summit, there was much discussion about these questions in relation to AI. Holding the models liable for everything that could be done with them would, in my view, be like blaming the printing press for what is put on and what comes off it. At the event, some said responsibility should lie at the application level. That could be true if, for example, Michael Cohen was misled by Google when it placed Bard next to search, letting him believe it would act like search and giving him bogus case citations instead. I would say that responsiblity generally lies with the user, the person who instructs the program to say something bad or who uses the program’s output without checking it, as Cohen did. There is nuance.
Deep fakery: There was also some discussion of the machine being used to fool people and whether, in the example used, Meta should be held responsible and expected to verify and take down a fake video of someone made with AI — or else be sued. As ever, I caution against legislating official truth.
The most amusing moment in the hearing was when the senator from Tennessee complained that media are liberal and AI is liberal and for proof she said that if one asks ChatGPT to write a poem praising Donald Trump, it will refuse. But it would write a poem praising Joe Biden and she proceeded to read it to me. I said it was bad poetry. (BTW, she’s right: both ChatGPT and Bard won’t sing the praises of Trump but will say nice things about Biden. I’ll leave the discussion about so-called guardrails to another day.)
It was a fascinating experience. I was honored to be included.
For the sake of contrast, in the morning before the hearing, I called Sven Størmer Thaulow, chief data and technology officer for Schibsted, the much-admired (and properly so) news and media company of Scandinavia. Last summer, Thaulow called for Norwegian media companies to contribute their content freely to make a Norwegian-language large language model. “The response,” the company said, “was overwhelmingly positive.” I wanted to hear more.
Thaulow explained that they are examining the opportunities for a native-language LLM in two phases: first research, then commercialization. In the research phase now, working with universities, they want to see whether a native model beats an English-language adaptation, and in their benchmark tests, it does. As a media company, Schibsted has also experimented with using generative AI to allow readers to query its database of gadget reviews in conversation, rather than just searching — something I wish US news organizations would do: Instead of complaining about the technology, use it to explore new opportunities.
Media companies contribute their content to the research. A national organization is making a blanket deal and individual companies are free to opt out. Norway being Norway — sane and smart — 90 percent of its books are already digitized and the project may test whether adding them will improve the model’s performance. If it does, they and government will deal with compensation then.
All of this is before the commercial phase. When that comes, they will have to grapple with fair shares of value.
How much more sensible this approach is to what we see in the US, where technology companies and media companies face off, with Capitol Hill as as their field of play, each side trying to play the refs there. The AI companies, to my mind, rushed their services to market without sufficient research about impact and harm, misleading users (like hapless Michael Cohen) about their capabilities. Media companies rushed their lobbyists to Congress to cash in the political capital earned through journalism to seek protectionism and favors from the politicians their journalists are supposed to cover, independently. Politicians use legislation to curry favor in turn with powerful and rich industries.
