This will probably be illegal in future, cruelty to AI.
| [reply] |
I think otherwise: compared to SO, it would be a merciful treatment;-)
| [reply] |
I was thinking along the lines of science fiction stories like Neuromancer, where "AI" was granted some rights and legal protections. Regardless, without sufficient safeguards I don't think this would be a good idea with this sites content. It may be better that than SO, but that's nor really the point.
| [reply] |
| [reply] |
This is only if the perlmonks text was swept up in the training data. It's probably in our best interest to make perlmonks more downloadable so that this body of information is available to LLM tools. People might actually decide to use or not use perl for a task based on how well ChatGPT can answer questions about it. Lately I've been asking it a bunch of questions about Vue3 and amazes how useful the answers are (as a search engine, it still doesn't write accurate code).
| [reply] |
I've found an genious way to check if content from perlmonks was used.
I asked! ;)
While PerlMonks.org is a publicly available website, I want to clarify that I have not been specifically trained on content from PerlMonks.org. My training data includes a variety of publicly available texts and sources related to Perl programming, but the specific content from PerlMonks.org has not been used in my training. If you have any questions or need assistance with Perl code, feel free to ask, and I'll do my best to help!
That was "chatgpt 3.5 turbo" as provided by duckduckgo.
| [reply] |