Will LLM companies eventually grab activitypub sources to train their models?

I'd ask rather when.

Follow

Fediverse being scraped and used commercially... ActivityPub sources undefended, "no consent" etc needed.... 

@engelbart So positively I blame or attribute the "eventually" result as both about waiting for others to do the dirty on people and also people themselves not defending / doing something about it (not protecting / locking account / no panic)...

It's like Orwell might have wrote about it but people can take it both as a script (to use against others or accept as probable future) or TRY change things (do something at the least about it - for example:)

► talk about it
► ask admin like @freemo about adding manually a disclaimer to the Mastodon footer saying:
"User profiles and data is not for scraping without express permission in their specific user description and giving them notice of intent"
► asking Eugene / adding github or gitlab request etc

If it's not in your bio (like mine) to say NO CONSENT in multiple forms, then yes asking "when" is on you too (no attack on you).

Just remember they will never want to protect you, so it's also naive to ever take this position (not you but just saying in general).

I will check my own bio now as a mutual check as it was a good question to point out!

Sign in to participate in the conversation
Qoto Mastodon

QOTO: Question Others to Teach Ourselves
An inclusive, Academic Freedom, instance
All cultures welcome.
Hate speech and harassment strictly forbidden.