AI Has Nothing To Offer Media Workers

This is an installment of Automate Me, an ongoing Study Hall series about how AI technology is transforming the media and publishing landscape.

 

Generative AI tools can do a lot of things. They can design the slides of a presentation called “Why Media Workers Like You Should Be Excited About AI.” They can write the words and create the images to fill each page. They can provide a list of tips for giving the talk with confidence. None of that’s in doubt; the internet abounds with proof. The question is: can AI, or anyone for that matter, make the pitch convincing?

Most media workers, I believe, would say “no.” They’d suspect it was only a matter of time before their boss tried to use AI against them, and it’d be hard to say they were just being paranoid. Earlier this year, Futurism reported that CNET was secretly publishing AI-generated articles without bothering to check if the AI was, say, giving wildly inaccurate financial advice. The floodgates have opened since then. 

A growing list of media outlets are experimenting with AI-generated content, often with embarrassing results. Nevertheless, the people in charge insist the AI future has arrived, even if the kinks are still being worked out. The fight between writers unions and AI-loving bosses is already in full swing, as Rebecca Heilweil reported for Study Hall. 

Workers in the media industry aren’t alone in facing an AI reckoning. These tools are also poised to revolutionize education, proponents say, and law and porn and healthcare et cetera. Comparisons to the Industrial Revolution abound; AI is going to change everything. 

Regardless of whether you think AI is something to be excited about—polls show the average person thinks it isn’t, despite optimism from luminaries like Bill Gates, Henry Kissinger, and Marc Andreessen—the constant buzz about these tools is hard to ignore. Who among us hasn’t wondered, “Do I gotta learn this stuff to keep my job?”

That gnawing fear is probably reasonable. As Paris Marx, tech critic and author of Road to Nowhere: What Silicon Valley Gets Wrong About the Future of Transportation (2022), tells me via email, “It’s far less likely that ‘AI’ is actually revolutionizing the industry, and much more likely that bosses are using the excuse of AI to cut labor costs.” 

Wendy Liu, a former software engineer and author of Abolish Silicon Valley: How to Liberate Technology from Capitalism, also believes the main purpose of AI is to discipline workers. While media company bosses invariably cite “current economic trends” as the main reason for layoffs (and the need for greater, AI-powered productivity by those who retain their jobs), it’s hard to believe the money-wells are actually dry when those bosses are paying themselves massive bonuses right before culling their workforces.

Liu sees the current explosion of boss-class enthusiasm for AI as part of a cycle. “Capital’s drive to accumulate through exploitation leads to alienation in the workplace, worsening labor conditions [and lowering] wages, which of course results in backlash,” she says via email. While there might not be a direct cause-and-effect link between the Great Resignation of 2021 and the current AI hype bubble, it’s easy to imagine why corporate executives would be excited about any development that gave them new leverage over workers. “[Funding] technical innovations in order to consolidate management power and reduce labor costs—that’s part of the same drive,” says Liu.

In that case, some media workers might be tempted to start using AI tools out of self-preservation if nothing else. Could AI be integrated, one way or another, into a step of your workflow? Maybe you should do it, to show you’re proactive about innovation. 

Bosses seem to hope that’s what workers will do. To quote a leaked memo from Insider editor-in-chief Nicholas Carlson sent shortly before the outlet’s latest layoffs, “generative AI can make all of you better editors, reporters, and producers.” He listed a variety of ways he used ChatGPT: coming up with article topics and headlines, devising ideas for videos, preparing for interviews, and so on. 

There’s no reason to doubt Carlson on this. I fully believe what he said was true, and that other media workers could use AI tools in similar ways. 

But so what? 

It’s not as if we faced a shortage of YouTube tutorials before AI, nor were stories wandering around the internet without headlines. To be fair, none of the commonly-touted uses for generative AI are necessarily bad. If you have writer’s block, asking a computer program to answer silly questions is probably no less effective than drinking a glass of whisky. It’s just that the problems generative AI is well-suited to solve aren’t big problems, and the solutions it offers aren’t unique. 

Regarding the big problems that are facing media workers—like: “how do I make this article/video/illustration/etc. actually good?”)—generative AI appears to have much more limited potential. As Malcolm Harris, journalist and author of Palo Alto: A History of California, Capitalism, and the World tells me via email, “I’ve never seen [an AI-]generated sentence that I wouldn’t be ashamed to put my name on.” 

Harris’ words are harsh. They’re also relatable for many media workers. Case in point: Recently I was speaking with an acquaintance who wrote a new book about tech’s effects on society. To promote the book, they wanted to pitch a series of essays to editors at different publications. Pitching is a time-consuming, ego-crushing endeavor that is nevertheless obligatory for anyone who wants to make a living as a writer. In theory, this made it the perfect task to outsource to ChatGPT.

The problem was that ChatGPT’s pitches were garbage. Dumb, derivative, utterly uninteresting garbage. And that’s when they actually made sense. ChatGPT cited chapters that didn’t exist and ideas my acquaintance hadn’t proposed. Nor did the tool prove to be much help with the logistical aspects of pitching, such as identifying appropriate editors to contact and how to get in touch with them. 

In fairness, pitching is hard. And today’s AI can hardly be expected to revolutionize every part of your job at once (with that next update, though…). If tasks requiring complex creative ideation aren’t yet in AI’s wheelhouse, maybe it would be better suited for those that simply require speed and precision, the two great advantages of machines.

Sadly, here generative AI’s powers are also underwhelming at best. Take audio transcription. Having transcribed many conversations over the years, I can attest that it’s tedious and boring as hell. 

So, like Marx and many other media workers, I’ve used an AI transcription tool called Otter to streamline the process. The website promises to help people “be more engaged and productive than ever before,” which sounds terrific. But as Marx explains, someone then has to review those transcripts “and fix [the AI tool’s] copious errors.” It’s not so much saving labor as rearranging it. 

Like most generative AI tools, Otter isn’t inherently evil or literally useless. It just doesn’t fix any real problems well enough to justify the externalities (massive job loss, most importantly) that come with its development and adoption. Harris gives voice to the feelings of many skeptics when he says:

“I don’t think generative software is at all useful the way they’re trying to convince us that it is.”

On the other hand, could this be just the first imperfect-yet-promising step in AI’s makeover of human civilization? Bosses and investors will happily tell you the answer is yes. However, many of us remember them saying the same thing about crypto and the metaverse. It’s now clear that no tech frenzy, regardless of how much money is poured into it, is guaranteed to enjoy the perpetual and exponential growth its proponents crave. 

As the novelty of creating content with the click of a button wears off, and people see for themselves that generative AI is unable to consistently generate stuff that’s interesting or useful or even accurate about objective facts, the clearer it becomes that popular interest in AI has a shrinking ceiling, and that resistance is less futile than the AI prophets claim.

Workers’ groups like the Writers Guild of America are organizing to limit the scope of AI adoption in their fields, and any media worker lucky enough to be part of such an organization would be wise to join the fight for legal and political rights. For everybody else—the freelancers, permalancers, and others who lack the protection of a union but still want to do something against the AI onslaught—a more guerrilla-style approach can be both cathartic and impactful.

In other words: shit on AI and its cheerleaders at every opportunity.  Help stigmatize the label of “AI-generated” so that it becomes a badge of shame, and mock the poorly written words and the soulless pictures wherever you see them. Fight every attempt to “streamline” the creative process by eliminating humans from it. That solidarity should be extended to all workers in all industries. A threat to transcriptionists and copywriters is a threat to illustrators and podcast producers, because workers’ strength comes from our numbers. The fewer of us working, the less autonomy and security those who remain employed will have.

Refusing to comply with our own marginalization and replacement by AI is a good start. But media workers of all stripes should also go on the offensive. Tech investors are a notoriously fickle, vibes-based bunch, and the long history of vaporware shows how vulnerable even the most-hyped products are to popular scorn. 

We’ve seen what generative AI has to offer us, and it doesn’t make our lives better in any meaningful way. We have nothing to gain from pretending otherwise. 

---

Take The Opportunities Survey!

To shape the next chapter for the Opportunities newsletter, we’re asking our members to take part in a quick survey that will better clarify the wants and needs of our members — and figure out what is working and not working.

Please fill out The Opportunities Survey by Friday, August 18.

---

“Own Your Work” Micro-Grant Update:

After announcing the first recipient of the micro-grant, Study Hall is launching another round! This time, the first place winner will get $2,000, the second place and third place winners will get $1,000 each.  Apply via this link by August 31.  If you’d like to be a community judge for the next round, please contact [email protected]. 

---

Comings and Goings:

—-Blanca Berget has joined POLITICO as a California climate reporter.
—-Sally Tamarkin is now the senior news editor at Them.
—-Sam Olson is now the assistant news editor at Cosmopolitan.
—-Emily Smith is leaving Page Six.
—-Laura Norkin has joined BBC Studios as Executive Managing Editor, Global Features.
—-Kalyn Belsha is now a senior reporter at Chalkbeat.
—-Ryan Schocket is no longer a reporter at BuzzFeed.
—-Mohammed El-Kurd is now a culture editor at Mondoweiss.
—-David Noriega is now a national correspondent for MSNBC and NBC News.

---

Everything Else:

 

—-Last week, Vice Media’s bankruptcy sale was officially closed. The media company was purchased for $350 million by a group of investment firms including Fortress Investment Group, Monroe Capital, and Soros Fund Management. Back in 2017, Vice Media was valued at $5.7 billion. A day following the sale, the company announced that Katie Drummond (SVP of global news & entertainment), Emanuel Maiber (Executive Editor of Motherboard), and Jason Koebler (EIC of Motherboard) were all stepping down. Vice Media declared bankruptcy back in May. In April, after laying off 100 employees, the company paid executives over $1 million in “retention bonuses.” Following the announcement of the sale, the Vice Union issued a statement reminding management that “dozens” of laid off employees have been waiting on severance payments for multiple months. “We look forward to continuing to create our award winning content under new management but remind management that VICE only exists because of its labor,” the Vice Union said in their statement.
—-Meta announced that they “have begun the process of ending news availability permanently” for its products in Canada. The tech company, which owns Instagram and Facebook, is lashing out following the Canadian government’s passage of the Online News Act, a piece of legislation that would require tech platforms to share some ad revenue with newsrooms. This is awful and Zuck is awful, even if he ends up fighting Elon.

—-According to the New York Times, the night after his third indictment, former president Donald Trump celebrated the special occasion by eating dinner with Fox News’ top brass at his New Jersey country club. Jay Wallace, president of Fox News, and Suzanne Scott, chief executive, tried to convince the leading candidate in the Republican primary  to attend the network’s GOP debate on August 23.  It seems like regardless of how much Trump’s lies and conspiracies have cost the network, they just can’t get enough of those sweet, sweet ratings.