4 ms·
I hereby set a terms of service for everything I post on the internet from now on. OpenAI may not train future GPT models on my words or my code without my expr
by Jevon23 4y ago
I hereby set a terms of service for everything I post on the internet from now on. OpenAI may not train future GPT models on my words or my code without my express written permission.
…
Somehow, I don’t think they’ll care.
- nico 4y agoSure. If you can get everyone to create an account and agree to those terms before reading your comments, you might have a case. Otherwise, it will be considered public information, at which point it is free to be scraped by anyone (see the precedent set by the LinkedIn/hiQ case).
- verdverm 4y agoLinkedIn won that case on appeal, HiQ waas found to be violating the ToS, common misconception I was pointed at a link explaining the case here on HN, after trying to make a similar point, but cannot find the link currently edit, not the one I was pointed at, but similar https://www.fbm.com/publications/what-recent-rulings-in-hiq-v-linkedin-and-other-cases-say-about-the-legality-of-data-scraping/ https://www.fbm.com/publications/what-recent-rulings-in-hiq-...
- sebzim4500 4y agoThat's just because they made accounts and so agreed to the terms right? From your link: >These rulings suggest that courts are much more comfortable restricting scraping activity where the parties have agreed by contract (whether directly or through agents) not to scrape. But courts remain wary of applying the CFAA and the potential criminal consequences it carries to scraping. The apparent exception is when a company engages in a pattern of intentionally creating fake accounts to collect logged-in data.
- verdverm 4y agoNo, the case did not decide anything, no precedent was set. The point is that you cannot use this case to argue that you can scrape public data free of consequence