Intellectual property can’t be “stolen”. That’s not a thing.

It can be copied, though but that doesn’t sound nearly as nefarious.

Remember: When you “steal” something, that’s “theft”. It means the original owner doesn’t have it anymore. If the AI companies were actually stealing the NYT’s IP, they’d head into the office and find all their computers/servers were missing and when we search the Internet Archive for older articles they’d be gone. Stolen!

…but all that is neither here nor there because what this guy is actually bitching about is AI summaries of the news. Whether you think AI can do a good job doing that doesn’t matter. It’s the fact that AI can read the whole article on behalf of the user and then present the content in any form that doesn’t include the original advertisements.

Hypothetical: Imagine you’ve got a 100% free and open source AI model installed and setup on your local computer (e.g. in a chat window, as an extension in your browser, integrated in the OS, etc… doesn’t matter). You ask it:

What's today's news about beans?

The AI model then calls curl a few dozen times (or just uses its internal URL loading capabilities), hitting the New York Times website. It then generates some summaries based on what it found along with links to the original articles (so you can verify or dive deeper into any given story).

The NYT chief is saying is that should be illegal. It’s “theft”!

To that I say, “bullshit!” I will run whatever TF AI model I want and use it however I want, thank you very much!

It’s important to note that the Big AI companies paid the New York Times to load all their historical articles into their training data. So that aspect of AI is not at issue.

SLAMS

He might also want to add that the media has “been too quiet, too passive and too fragmented in the face of abuses by” our own government and big business.