Post AYbRe5Z7d88Z1g50We by Streakmachine@mastodon.sdf.org
(DIR) More posts by Streakmachine@mastodon.sdf.org
(DIR) Post #AYbFJTvKfm5V2ZsMXg by Streakmachine@mastodon.sdf.org
2023-08-10T20:43:49Z
0 likes, 0 repeats
So wait, let mw get this straight;You’re saying if I don’t add ChatGPT to my ‘robots.txt’, OpenAI will gladly scrape my whole website content? Line, hook, and sinker? Even if my website were, hypothetically, filled with absolute nonsense, gobbledygook, and crud? You know, the kind of stuff that would potentially be poison to “A.I.” model training?Just thinking out loud and hypothetically, of course, but what would happen if LOTS of such websites were to spring up all over the web?
(DIR) Post #AYbFJVTOuavvqWp5Hc by ParadeGrotesque@mastodon.sdf.org
2023-08-10T21:09:55Z
0 likes, 0 repeats
@Streakmachine They will scrape any nonsense. A lot of information in the original chatgpt came from Fox news and other "reputable" sources.There has already been studies about how AI generated web sites will eventually pollute the language data pool so much all the LLM will eventually stop working.
(DIR) Post #AYbRe5Z7d88Z1g50We by Streakmachine@mastodon.sdf.org
2023-08-10T23:27:46Z
0 likes, 0 repeats
@ParadeGrotesque Poisoning the proverbial well could do good, in this particular instance.