Hi to all, I seldomly write on this forum, but I read it and I still like Nim
language and its community very much. The argument of this thread is intriguing
and broad. I liked @Zoom thoughts/comments. I've played myself a bit with
ChatGPT-3 and my feelings are mixed about it. I think is a significant quantum
leap for AI, but I'm a bit scared about a widespread acquiescence about its
inaccuracy /allucinations outcomes. In my opinion people do underestimate the
impact of something (not someone) always providing an answer, and never saying
"sorry, I don't know enough about this subject". This is a basic epistemology
problem. In my exoerience ChatGPT can provide totally wrong answers to e.g.
indefinite integrals, while it might simply agree that Wolfram Alfa (amazing
software, from the knowledgeable guy who created Mathematica, but far less
"popular") is much more reliable on this. Actually I engaged a quite funny
discussion with it about that. Staying on CS, I think that Postel's Law or
Robustness Principle still holds its value: "be conservative in what you send,
be liberal in what you accept". I remember that Stack Overflow got some serious
issues being flooded by inaccurate answers produced through ChatGPT and took
some time to put in place countermeasures. That said, I like the book of
Salewsky, actually much more than ChatGPT's code snippets, importing always
parseutils and os, with no apparent reason ;), and I think it provided a useful
contribution to Nim's understanding. Being in love with "Great Lebowsky" movie,
I have no problem with "dude" word, but I won't use it if my peer has different
perception / feelings about it.
