@Wildduck @ZachWeinersmith I mean, you say that, but this already happened!
https://www.reddit.com/r/AO3/comments/zl72bn/update_ao3s_response_to_sudowritesai_scraping/
@Wildduck @ZachWeinersmith
I'm not sure how much free time you want to spend on this, but let's just say the full controls stopping chatGPT from going horny on main are....not exactly ironclad.
Honestly you can get around pretty much every restriction, especially since it builds on itself within the conversation, so you can take it on a walk to get these things to reply in a certain way. And I got one version to just start spouting porn, basically.
I was bored, okay?
@Wildduck @ZachWeinersmith Fanfic is still under copyright of the author, because it's their creative content in a recorded form and it's a "transformative work". They just down own the characters and so are limited in how they can publish that work.
That's why there's the OTW (https://www.transformativeworks.org/) to protect fan works. Just because they don't own the characters, it doesn't mean that other people (or even the characters' owners) can ignore copyright.
@ZachWeinersmith I am so sick and tired of my BossBot quoting Beowulf and Shakespeare.
“A null pointer by any other name” does not help me…
…but it’s no worse than my last boss.
The movie Metropolis is now in the public domain.
The writers and actors striking over AI use should be on the lookout for any of their kin who look or sound a bit robotic, as the film, television and streaming studio heads will undoubtedly have fed Metropolis into their AI and asked how it can break the strike.
Besides the legal inaccuracy, the #Copyrightmafia is the problem!
https://felixreda.eu/2021/07/github-copilot-is-not-infringing-your-copyright/
@ZachWeinersmith @cstross
I only wish that OpenAI and their ilk would stop scraping copyrighted content.
However, as I shared on the @nytimes, OpenAI has definitely incorporated the entire text of at least one of my (copyright registered in 2010) books on meeting design into the ChatGPT database.
See the replies to my comment to see folks questions about copyright and my responses.
Gift link (click on comments and then Readers Picks to view): https://www.nytimes.com/2023/07/13/technology/chatgpt-investigation-ftc-openai.html?unlocked_article_code=zQStECBrrFhPBhU3blM7qStSiHWVxV47WSBJcRfDYNVb8I1nAydquS_ykDX9gBSzw_MEDdpOfHr54PwAjPGaVKEFXuBp-jTtoQJRviDdDIKMzBe3ZlOZB1_9mux9YP9dHVrDCij2fsLFLge1ZrrqP-kNHrOpgRqMZZeef3v96mcCB8N6ki-0GcbQ9cYZpJaKsgeyLowW_oE_2Ex2JPs3C_Hk_resfb8gxPuAfktwTKMxDLN7MvckgcXl71vtQXfRoG7IdX-rSeVQCalB2_u-A4w0GYjiPicNWU2UO2tApXz4e7gUYFKAb79CoRoQTd-46VpL8zWztkYzB91aovHSNBfeeQQUtcwKdadp&smid=url-share
#ALT4you 5 panel cartoon drawn in a loose style, coloured.
Panel 1: two people coded as scientists because they're wearing white coats, one has glasses, the other busy at something that might be a terminal.
Scientist 1 says:"wait wait so we got sued by all those artists and writers because we scraped their words to train AI.
Scientist 2 responds: " Right"
Panel 2
Scientist 1: "So we're only able to train the new superintelligence on uncopyrighted material"
Scientist 2: "yep"
Panel 3
Scientist 1: "but almost all rights-free content is from over a century ago"
Scientist 2 (looks unexcited) "oh fuckity dickballs."
Panel 4
Huge, grey, blocky robot figure adjusting its tie: "Ladies no swearing or you'll never get a man. Go clean my office"
Panel 5, bigger than the other panels:
All three figures, shown from above.
Scientist 2: "This is going to be a problem, isn't it? "
Robot : "And now to implement world wide colonialism - the best form of government!"