iiinova — where your digital self keeps evolving

Open on Bluesky →

Replies

seidtgeist @seidt.quest · 10/09/2026

i am 1e9% sure that any “opt out of training” settings are decorative. *how* all user data is mined and derived is an ongoing technicality for very expensive in-house lawyers. soylent green, i mean, synthetic training data is made of people!!11elf

mr. TIM@timkellogg.me

another researcher accuses OpenAI of training on conversations and then claiming a breakthrough

seidtgeist @seidt.quest · 15/09/2026

guhhh imagine needing sausage brain to classify and summarize god-tier AI parlance bsky.app/profile/jose...

Joseph Cox@josephcox.bsky.social

There is a lot of new info in this article, based on documents, Slack chats, instruction guides, and me seeing real ChatGPT user prompts in the review system. Firstly, a worker we spoke to thinks ChatGPT users have no idea their chats may be read by someone else www.404media.co/inside-proje...

000

Peter van der Zee @pvdz.ee · 16/09/2026

Look man, you know what they say. If you're not a paying enterprise user with ZDR, you are the product.

Tobi @toh.bsky.social · 10/09/2026

As I understand it, even if you opt-out, the whole conversation is used as training data if you give a 👍🏻 or 👎🏻.

000