Gruber, Copyright and LLM scraping.

John Gruber, noted Apple access journalist, is back defending the company after its launch of Applebot-powered Apple Intelligence. There's some interesting treatment of Apple Intelligence and more specifically Applebot crawling the 'public web' and then 'transforming' that crawled content, via this post from Louie Mantia:

From John Gruber today:
It’s fair for public data to be excluded on an opt-out basis, rather than included on an opt-in one [...]
No, no it’s not. This is a critical thing about ownership and copyright in the world. We own what we make the moment we make it. Publishing text or images on the web does not make it fair game to train AI on. The “public” in “public web” means free to access; it does not mean it's free to use.

The idea that being on the public web means that content can be actually taken and recreated with no human contribution is bizarre. If a music video is published on YouTube, does that suddenly mean it is no longer a Copyrighted piece of work? Absolutely not. Does a paywall on a site, like Netflix for example, suddenly mean that the content is different? No.

It's insulting that Apple now claims its crawler is opt-out, after already training its AI on content without warning.

I guess Gloss republishing a quote from Mantia's blog above is also a recreation, maybe some might even say theft or copyright infringement. But I'm adding to it, linking back, and the original article is still its own viable piece. You might want to read the rest, in which case the original creator can receive pageviews.

Training AI