05-03-2023, 05:55 PM
(05-03-2023, 05:02 PM)madscijr Wrote: Let's keep up the conversations about this stuff, and hopefully contribute to making it better and avoiding the pitfalls!
I agree, and in this sub-forum, if it has to do with programming in QB64. Otherwise this topic should be in Off-Topic, although it would limit who could see it and post.
(Raises hand) I'm also making a post not having to do with QB64 programming but because I needed to get something out. The ChatGPT is already being used for nefarious purposes like I have explained earlier. Enough dishonest people claim to make a living by doing things like that, when they need to sell goods but try to maximize the profits. Which usually makes them balk at paying someone else to handle customer support.
The TuxBot experiment was designed to get a few laughs although some of its advice is surprisingly accurate and clear. Although none of the code examples the bot provides should be employed in a mission-critical setting. But limit the ChatGPT "experiment" to being just that, to having an entertainment value. Otherwise it's boring and immoral, should be condemned and nipped at the bud.
But what a shame the "openai" site now requires an user to have an account. Likely they changed the format, too, for interacting with the pseudo-philosopher. Might be a job for QBJS to extract text that cannot be provided in HTML or XML for security reasons. At least I will never know. If somebody does have an account with that other site, it would be nice if he/she could give a curt explanation of how the interaction is like. Testing like that could help the HTTP functionality of QB64PE get out of $UNSTABLE status.