No Permission, no pay: How my Book became AI training fodder

I was at the Leipziger Buchmesse (Leipzig Book Fair) to talk about AI.

First off: I was robbed.

Supposedly, Meta (and probably others) trained its AI using LibGen, a shadow library with millions of books and articles. A book that Christian Stoecker, Konrad Lischka and I wrote 13 years ago is part of the LibGen dataset.

Our little book made it to two editions. It’s available in proper libraries, even the German Bundestag has a copy. Nowadays, you can only purchase it secondhand.

Meta didn’t acquire it. Training happened without permission and without compensation. Christian is calling for the European Union to bare its teeth against tech companies because of this massive book theft. And he wants money. A lot of money.

For a moment, I caught myself thinking it’s kind of comforting that our little book is part of this world machine and hasn’t been forgotten. But of course, we need rules: no training without consent.

Robin Sloan (author of genuinely excellent books like Moonbound, Sourdough, and Mr. Penumbra’s 24-Hour Bookstore) can imagine backing off from this hardline position if AI actually helps, let’s say, cure cancer. But with generative AI that writes texts and searches through company data, we’re not close to that moral threshold yet, are we?

Anyway, back to the Leipzig Book Fair: The Forum Human and AI had an exciting program, and I got to talk with Gizem Celik and Nina Kolleck about the effects of AI on media.

I still harbor this probably naive hope that we’re all competent enough not to spread AI fakes en masse and instead trust more in the media brands that actually deserve it.

But research on media literacy regularly paints a bleak picture, as Nina reminded us. Even more important than fakes was the debate about addictive algorithms like those on TikTok that can contribute to radicalization. Gizem talked about her school days and her “internet driver’s license,” and we all agreed: great idea.

Katharina Heckendorf prepared and moderated us excellently, and thanks to the audience, I was also able to explain how we handle AI at SPIEGEL-Gruppe – as a tool for better journalism, not as a shortcut to create more content faster and cheaper.

Thanks for the invitation, Damaris Schmitz. (This post is also on LinkedIn.)