It's not the robots--it's the rip-off

My beef with A.I.

Photo by Rock'n Roll Monkey on Unsplash

Do you eyes glaze over when you see “A.I.” in the headline? Are you tired of hearing that Artificial Intelligence and the Robots are Coming for Our Jobs? I confess to extreme ChatGPT chatting fatigue. And I am cynical about the claims that his latest tech miracle will make our lives so much better. But even glancing at headlines and skimming articles has been enough to piss me off. Royally.

MIT Technology Review explains how the heist got organized:

Diffbot is building the biggest-ever knowledge graph by applying image recognition and natural-language processing to billions of web pages.

Like GPT-3, Diffbot’s system learns by vacuuming up vast amounts of human-written text found online. But instead of using that data to train a language model, Diffbot turns what it reads into a series of three-part factoids that relate one thing to another: subject, verb, object. . . .

To collect its facts, Diffbot’s AI reads the web as a human would—but much faster. Using a super-charged version of the Chrome browser, the AI views the raw pixels of a web page and uses image-recognition algorithms to categorize the page as one of 20 different types, including video, image, article, event, and discussion thread. It then identifies key elements on the page, such as headline, author, product description, or price, and uses NLP to extract facts from any text.

In short, AI programs hoover up every published web page in every single language, categorize the pages, parse the information into “three-part factoids,” appropriate visual images and discussions—and use all this material for whatever they want. The programmers use it to develop and test “better” algorithms. The tech companies who employ the programmers use it to develop new products to sell. The corporations who purchase the AI products market them to other corporations interested in reducing labor costs; then they can boost “shareholder value” and further inflate their executive compensation. Who doesn’t make money from this? The writers, editors, artists, photographers, editorial assistants, agents, publishers, sound engineers, and musicians who created all the articles, stories, images, sounds, and videos published on the web. The entire AI endeavor may have started as an interesting intellectual puzzle—but now it exists as a fencing operation: selling stolen goods at a huge markup. It’s pure profit.

The AI advance reminds me of the early days of Uber and AirBnB. In hindsight, it’s easy to perceive the tech bro campaign plan: 1) Develop new technology that appropriates core functions of highly regulated industries like taxis and hotels. 2) Rename the function so it sounds loose and informal: “ride-sharing,” and “home sharing.” 3) Undercut the established industry—taxis, car hires, motels, hotels, or youth hostels—by offering your service at below-market prices in each of the categories. (This is easy because your start-up is flush with VC cash and you don’t comply with existing regulations—those stupid labor laws, taxes, and licenses). 4) Spend money on lobbyists and attorneys. (You know that the politicians don’t understand tech and it will take years for the law to catch up with you.) 5) When the law does catch up and the investors demand to see a profit, no problem! Now that your competitors have been squashed, you can raise your prices. It’s a classic market takeover. But with computers!

AI is following the same pattern. This time the target isn’t the hospitality industry, it’s the culture. This time it’s not the taxi drivers taking the hit; it’s the creatives.

What about copyright, you say? Am I James Patterson? Do I have lawyers on retainer who can spend the next five years suing Open AI and Microsoft? Ahh . . . no. Even for prominent artists and authors a successful copyright defense is difficult. It turns out that AI’s use of copyrighted materials is currently considered a “gray area” of the law. The AI companies argue that their vast scraping of internet content, even when it gathers up copyrighted material, is “fair use.” IANAL* but I think that’s a bit of a stretch.

The “fair use” doctrine is familiar to me because it is commonly applied to copyrighted materials (scholarly articles or artworks) that are used for classroom teaching. Educators must follow copyright laws; universities have procedures to enforce compliance. If I want to assign an article for my class, I submit a form to the reference librarian, she does a fair use assessment (four factors), and if a copyright waiver is allowed, she uploads the article to eReserve for my students. Of course, that type of system wouldn’t work at the scale needed for AI algorithm training. So the tech bros embrace the motto: ask for forgiveness, not for permission.

Even the fashion industry is worried:

. . . some of [AI’S] most cutting-edge types generate art, computer code suggestions, and even music (“generative AI”). To make the predictions, AI must go through machine learning, which involves processing incredible amounts of input training data to identify patterns. The more training data is input into the datasets, the more precise and valuable the output data. To give you an idea of how vast a training dataset can get, LAION-5B consists of 5.85 billion image-text pairs (this dataset is used by Stable Diffusion and Lensa AI, for example).

Often, those massive datasets contain copyrighted materials – photos, paintings, books, or computer source code. Even more often, copyright owners have no idea about (let alone consent to) the use of their material in machine learning. No longer rare are the cases of human artists or programmers discovering that someone used their works to train AI that produces output containing recognizable portions of their work. 

So, not only is the creator not compensated for their writing/music/design/art—they never know where parts of their unique creation will turn up, and how exactly their work will be exploited.

Since our political system is slow and ramshackle, the only effective defense is collective action. To resist this latest encroachment by our corporate overlords, resistance to unregulated AI has to mobilize quickly. In Los Angeles, it is game on.

Hollywood’s major studios would like to use the legal “gray areas” regarding streaming content to limit residual payments to writers and performers; they have a strong financial incentive to automate, for example, to use AI to generate scripts.

At the moment, the writers are on the front line. The Writers Guild of America (WGA) will soon to be joined on the picket line by the Directors Guild of America (DGA) and the Screen Actors Guild (SAG-AFTRA.) This growing labor action by Hollywood’s actors, writers, and directors against the Alliance of Motion Picture and Television Producers (AMPTP) is being described as“a first skirmish in a new war.”

It is especially worth noting that, like the studios themselves, these two unions [DGA and SAG-AFTRA] also risk their representatives being replaced by AI—if you don’t believe me, go and watch . . . the completely AI-generated Great Catspy trailer, or even the AI-created RNC attack ad against Biden. There were no directors, actors, grips, or directors of photography used to create that content. Just a few lines of text that an AI turned into video. In other words, everyone is fucked if they don’t all team up and give this story a Hollywood ending. 

Writers are fighting to maintain the viability of creative writing as a sustainable career. They don’t want to be sucked into the contingent labor force . . . as has happened to academics, musicians, photographers, journalists, and graphic artists over the last three decades.

Everyone has a stake in this fight. The writers are just the tip of the spear.

The potential rise of A.I. has workplace implications for writers, but it’s not only a labor issue. We, too, have a stake in the war with the storybots. A culture that is fed entirely by regurgitating existing ideas is a stagnant one. We need invention, experimentation and, yes, failure, in order to advance and evolve. The logical conclusion of an algorithmicized, “more like what you just watched” entertainment industry is a popular culture that just … stops.

[I]t is a human skill to create a story that surprises, challenges, frustrates, discovers ideas that did not exist before. Whether we care about that — whether we value it over an unlimited supply of reliable, good-enough menu options — is, for now, still our choice.

And, in conclusion, I may have erroneously dismissed part of the AI threat. My bad . . . it’s also the robots.

*Internet slang for “I am not a lawyer . . .”

Keep scrolling down (below Notes) to reach the comments, share, and like buttons

Find me on social media:

Notes:

Diana Bikbaeva, AI Trained on copyrighted works: when is it Fair Use?

Nick Bilton, Why AI should be central in the Writers Strike

Ted Chiang, Will AI become the new McKinsey?

Jake Coyle, Television, motion writers strike after failed negotiations with studios

Will Douglas Heaven, This know-it-all AI learns by reading the entire web non-stop

James Poniewozik, Striking writers are worried about AI. Viewers should be, too. (Unlocked NYT gift article.)