@pluralistic As a digital archivist who has managed a web archiving program, I 100% agree with your take on scraping, and THANK YOU for writing this. Is there any meaningful difference between the fair use you're talking about and what happened with Anna's Archive scraping OCLC metadata records? https://torrentfreak.com/lawsuit-accuses-annas-archive-of-hacking-worldcat-stealing-2-2-tb-data-240207/ It's fair use, right?? (I know you're not a copyright lawyer but interested if you have an opinion!)
My latest nationally bestselling novel is "The Bezzle," an ice-cold revenge story of high-tech finance crime starring the forensic accountant Martin Hench:
My ebooks and audiobooks (from Tor Books, Head of Zeus, McSweeneys, Beacon, Verso and others) are for sale all over the net, but I sell 'em too, and when you buy 'em from me, I earn twice as much and you get books with no DRM and no license "agreements."
If you prefer a newsletter, subscribe to the plura-list, which is ad-/tracker-free, and is utterly unadorned save a single daily emoji. Today's is "🫐". Suggestions solicited for future emojis!
--
You can also get a fulltext RSS feed, licensed CC BY 4.0:
I don't like that I can identify the legal maximum and then create a machine to use just a shave under that amount but from everyone, to create the exact effect the law was designed to prevent, art having no lasting value.
For instance, dump my original artwork into an AI and you can create thousands of conceptual duplicates minutes after my publication. What chance does my piece have in such a market?
@pluralistic The cup was too thin, the lid didn’t fit, and she was scalded terribly and it took more than 3 surgeries to repair the resulting wounds. McD’s had cups so thin that when they handed them to you, they pinched and their lids fell off. I remember those.
@pluralistic There is one option that would be really useful: Force fair access to models. This may even mean that every model should be open source. AI models are a great thing, as long as you control them yourself. Companies gating models behind APIs, charging a large amount of money for something they made from free data, and restricting what users are allowed to do, are not. As of now it looks like models are also not copyrighed, but that does not help as long as companies keep them private.