[GoLUG] I just solved a Harbour language problem with ChatGPT

[Syeed Ali]

On Wed, 4 Oct 2023 16:19:03 -0400
Steve Litt <slitt at troubleshooters.com> wrote:

> My impression so far is that ChatGPT is just as error prone as all the
> wild and whacky "answers" out on the web, but instead of doing
> searches that take 10 minutes or asking questions online that could
> take hours for somebody to respond, within 1 to 5 seconds I get a
> (might be right, might be wrong) answer to get me a little down the
> trial and error road to the goal. I liked my first experience.
> 
> After what just happened, I highly suggest you guys tune into
> tonight's meeting.

This is exactly the sort of lesson I would end up stressing.  :)

Because you had an open mind it gave you enough expertise just minutes
earlier so it's later guesswork became closer to useful.

It's so easy to iteratively improve.  While you're troubleshooting the
topic, you sort of perform troubleshooting on the meta of ChatGPT
itself.

This is why I say that ChatGPT isn't going to make us dumb, it's going
to necessitate a certain mindset to work with it most effectively.