Ryan Bridge: Is this really AI theft?

IP lawyers will be licking their lips with all these AI cases landing on their desks.  

The latest, we spoke about it yesterday, was Sony and Warner going after Anthropic.  

Their argument is basically that Anthropic's AI models, the ones we use to get answers to questions, scrape and bottom trawl the internet for information.  

That information, including song lyrics, is often IP owned and produced by humans or companies. So, scraping it and using it to sell as a product is theft.  

Media companies in New Zealand take a similar view.  

If the robot is scraping content produced, at cost, by a news business, then how is that not also theft? 

But if you think about it, humans do this all the time. I do it every day. It's actually what I'm paid to do. I ingest large amounts of information and then use it to inform interviews and content.  

Lawyers do it for legal cases. Politicians do it. Businesses do it. We all do it in one way or another.  

And we all use the information the internet provides for our own purposes. Usually, to make money.  

Knowledge is power and content is king.  

So, the question then becomes, what is the difference between us doing this and using this information and an AI model doing the same? 

Surely Claude is committing no greater crime than the rest of us.  

Of course, how they get their hands on the information is key.  

If you're ripping it off bit torrent, then you might have other problems. But if you're getting it from a paid subscription, isn't that what we do?  

In the news business you credit your sources. But not always. There's a point where the collective weight of the information you've consumed becomes your own thoughts and your own IP.  

Are robots any different? 

See omnystudio.com/listener for privacy information.