Alex Reisner is a Contributing Writer at The Atlantic, working to improve public discussion about artificial intelligence. In August 2023, before it was widely known how AI models were trained, he broke the story of Books3, a collection of 192,000 pirated e-books used by Apple, Meta, Nvidia, Bloomberg, and other companies. Since then, he's investigated and revealed the contents of other AI training datasets, including dialogue from thousands of movies and TV shows, and millions of books and research papers. He's also covered the technical and legal aspects of generative AI, including the ongoing copyright lawsuits and the under-reported phenomenon of memorization in large language models.

 

Prior to being a journalist, he spent 20 years as a programmer, statistician, and CTO for startups in New York City. He uses his technical background to cut through the anthropomorphizing language that AI companies use to mystify the technology and misrepresent its potential.

​