4 Apr
2025
4 Apr
'25
4:29 a.m.
On 4/3/25 16:37, Chuck Guzis via cctalk wrote:
On 4/3/25 14:21, John Foust via cctalk wrote:
I have a friend who's an EE and taking advanced math and statistics
courses at his university. He's middle aged, a brilliant engineer, and
blessed with at least some common sense.
He uses ChatGPT and a couple of other free AI services to write a lot of
code for his classwork (with his instructors' blessing), and we've
talked a lot about using that effectively.
The short version is that ChatGPT works well for writing functions and
snippets - small, mostly linear pieces - but not complex processes. And
more importantly, you have to be able to write what you're asking for,
AND VALIDATE the output.
Finally, he says it's a lot like a web search. The query is key to
getting correct results. The ability to diagnose incorrect results and
modify the query is a determining factor.
Doc
At 10:22 PM 4/1/2025, Chuck Guzis via cctalk
wrote:
In another task where I wanted to quickly
generate a text-based
profile of a computer's specifications and components, it was very
helpful in writing a bunch of PowerShell. Did it hallucinate?
Perhaps once out of several dozen things I asked it, and in that
case, it was easy to see that the two clauses of the if-then were
the same.
Interesting. I got no help at all from ChatGPT when working through
some IBM 7090 code.