The Strange Origin of AI’s ‘Reasoning’ Abilities It involves 4chan, of all places.

Let's give credit where it's due. www.theatlantic.com/technology/2...

AI’s Memorization Crisis Large language models don’t “learn”—they copy. And that could change everything for the tech industry.

AI doesn't "learn" like a human, it makes copies of books, art, music, and videos. Which could be a big problem for Silicon Valley. My latest investigation for the The Atlantic presents research that the AI industry has been trying to keep hidden. www.theatlantic.com/technology/2...

The Nonprofit Doing the AI Industry’s Dirty Work The web archive Common Crawl has been quietly funneling paywalled articles to AI companies—and lying to publishers about it.

A little-known nonprofit has been lying to news publishers while funneling millions of paywalled articles to tech companies for AI training. Read my investigation in The Atlantic. www.theatlantic.com/technology/2...

The End of Publishing as We Know It Inside Silicon Valley’s assault on the media

AI is doing to publishers what Uber did to taxi companies. Books and journalism may increasingly be controlled by Silicon Valley. See my latest piece in The Atlantic for details. www.theatlantic.com/technology/a...

ChatGPT Turned Into a Studio Ghibli Machine. How Is That Legal? Three possible arguments against the tech company

Style imitation isn't copyright infringement, but it's potentially disastrous for artists and studios. Can anything stop AI companies from doing it? I wrote about this for The Atlantic. www.theatlantic.com/technology/a...

Chatbots Are Cheating on Their Benchmark Tests AI programs train on questions they’re later tested on. So how do we know if they’re getting smarter?

If AI chatbots are getting better, why can't tech companies measure their progress? I wrote about a serious problem with the industry's benchmark testing. www.theatlantic.com/technology/a...

There’s No Longer Any Doubt That Hollywood Writing Is Powering AI Dialogue from these movies and TV shows has been used by companies such as Apple and Anthropic to train AI systems.

Dialogue from more than 139,000 movies and TV shows has been used to train generative AI. I wrote about it for The Atlantic and built a search tool for screenwriters, directors, and actors to find their work. www.theatlantic.com/technology/a...

Generative AI Is Challenging a 234-Year-Old Law The technology might finally bend copyright past the breaking point, upending what it means to have a creative society in the process.

Is generative AI a good thing for our culture? Does it help spread knowledge or does it impede creativity and collaboration? These are central questions in the ongoing lawsuits against AI companies, which I just wrote about for
@theatlantic.bsky.social. www.theatlantic.com/technology/a...

These 183,000 Books Are Fueling the Biggest Fight in Publishing and Tech Use our new search tool to see which authors have been used to train the machines.

You can now search 183,000+ books that are being used without permission as training data for generative AI: www.theatlantic.com/technology/a...

Murdered by My Replica? Margaret Atwood responds to the revelation that pirated copies of her books are being used to train AI.

Great response from Margaret Atwood to my piece in @theatlantic.bsky.social
. "...they could at least buy me a coffee." www.theatlantic.com/books/archiv...

Revealed: The Authors Whose Pirated Books Are Powering Generative AI Stephen King, Zadie Smith, and Michael Pollan are among thousands of writers whose copyrighted works are being used to train large language models.

I did some hacking and wrote a piece for @theatlantic.bsky.social about the complex realities of AI training data.