Authors Face Backlash For Participation In 2022 Google AI Study
By Ray Dombrowski ·
The whole damn thing smells like sweatshop labor to me. It’s the same story—the promise of infinite scalability built on unpaid, uncredited work.
When a Ribbon-Cutting Isn't Employment
The whole damn thing smells like sweatshop labor to me. It’s the same story—the promise of infinite scalability built on unpaid, uncredited work. You read about it in engadget.com, where authors are facing backlash for participating in a 2022 Google AI study, and you read about it again in techcrunch.com when major publishers file lawsuits against Google over Gemini’s training data. The narrative is always the same: some massive machine—whether it's LaMDA or whatever comes next—is built on the intellectual sweat of people who get nothing but a press release.
Back in 2022, DeepMind ran this Wordcraft Writers Workshop. Thirteen authors were recruited, put under NDAs, and asked to use an experimental tool. Lily Lachance nailed it when she called it helping legitimize the "plagiarism machine." Eugenia Triantafyllou admitted, “Very naive of me tbh.” It wasn't a collaborative effort; it was labor input for free research. Today, that pattern has matured from speculative fiction workshops into multi-billion dollar lawsuits filed by Hachette Book Group and Cengage Learning against Google in the Southern District of New York (aljazeera.com). They allege Google willfully sidestepped copyright law, training Gemini on books copied through Google Books and web scrapes—all without telling the authors or publishers a damn thing.
The Spreadsheet Doesn't Care About Genius
This isn't about whether AI is good or bad; it’s about who gets paid when the process changes. It reminds me of the Industrial Revolution, that massive shift from artisanal skill to mechanized output starting in Britain around 1760. What happened then was a fundamental redefinition of labor value—the ability to produce goods at scale changed everything, and those who owned the means of production got richer while the workers were left scrambling for scraps.
The mechanism is identical. The shift from manual writing and publishing (a finite resource) to LLMs (an infinitely scalable process) fundamentally devalues the unique output of a single author or publisher. Google’s alleged internal warnings—that using copyrighted books for training was "highly problematic" and could lead to fines up to $100B—are just corporate risk assessments, not moral compasses. They are calculating how much they can steal before the lawsuits pile up high enough to slow them down.
The Verdict is Already Written in Lawsuits
The whole pantheon of Silicon Valley hype—the ribbon-cutting ceremony, the white paper, the "breakthrough"—is just smoke designed to distract from the payroll reality. This isn't innovation; it’s extraction. When a system like this matures, the law eventually catches up, forcing the owners of the machine to acknowledge that their foundation was built on stolen materials and uncompensated labor. The only thing these lawsuits prove is that the infrastructure—the books, the articles, the stories—is not freely available raw material; it is valuable intellectual property that requires a licensed wage.