Boost for artists in AI copyright battle as only 3% back UK active opt-out plan
Liz Kendall faces pressure from campaigners as she tells parliament there is no clear consensus on issue
Kendall is lying: 88% of the public in consultation exercise wanted *stronger* copyright law, and 95% want AI companies to pay for training data.
@cstross LLM companies have inspired me to start a business where I rob banks, and then weaken punishments for robbing.
@codinghorror @jhooper Uh-huh. Downloading huge volumes of pirated content hosted illegally by websites then using them to generate slop in direct competition with the authors of the original works—that's the crime in progress. (The fact that the LLM crawlers are also pounding legit websites like mine into the dirt and disregarding robots.txt is just the icing on the turd cake.)
@codinghorror @cstross @jhooper I can't and don't speak for Charlie or anyone, but my position is:
* LLMs are not AI
* LLMs are bad and 99% of bot slop is bad. Ecologically, for users, for consumers, in general, bad.
* Taking publicly-available data to feed LLMs is bad
Bots can make slop because they've been trained on stuff humans made. Companies running bots are making money off stuff they did not pay for. That is bad.
Searching is OK. Searching connects creators to consumers.
Remixing that and serving it instead is bad.
Crawling for content is bad.
This is not complicated or hard or even subtle. I do not understand your frantic requests for clarification.
No, I have not watched your video. I hate video. I speedread. Is there text anywhere?
@codinghorror are you OK?
@codinghorror good to know!
Intense feelings on both sides. I'd love to see you and @cstross@wandering.shop debate in person.
Am standing in a wet field with a dog at mo tho
@codinghorror yet #openai etc appear to be escaping punishment for breaching the law when went beyond that. Google's indexer doesn't create persistent copies of pirated book content. Unless I'm misunderstanding.
