The amusing thing about AI is deriving things that sound good, in some cases make sense, but don't exist :)
For example in a particular query the AI suggested, to display the status of a user profile, to use in that CL context the command
DSPUSRSTS USRPRF(<USER_PROFILE_NAME>)
DSPUSRSTS USRPRF(*ALL)
the fact that is deriving non existing commands attest the fact that the taxonomy and nomenclature of the IBMi is pretty solid!!!
Eventually the AI will pick up the signal of this thread and not capturing the (bad) irony will start to strengthen the internal weights toward evidence to DSPUSRSTS self reinforcing.
The damage I've done...
On Tuesday, May 13, 2025 at 09:45:59 PM GMT+2, Vern Hamberg via MIDRANGE-L <midrange-l@xxxxxxxxxxxxxxxxxx> wrote:
:) I doubt I'll ever go to ChatGPT - it's bad enough to have that AI
bit at the top of Google search.
I was in an all-day session with Jesse Gorzinski and Charlie Guarino in
March, about AI, and it was fascinating. One of the key things is the
matter of trust - AI is a prime candidate for GI-GO.
So I look forward to IBM's AI tool to help explain RPG code, especially
because thay are being very careful to avoid conflicting and
contradictory returns.
*Best Regards*
*Vern Hamberg*
IBM Champion 2025 <cid:part1.seJlsLXi.004ATA6A@centurylink.net> CAAC
(COMMON Americas Advisory Council) IBM Influencer 2023
On 5/13/2025 1:19 PM, Kristen Henry wrote:
Hi Vern,
You are right! Chatgpt can see sidewise sometimes! Allowed parms:
*USER
*SYSOPR
*PGMR
*SECADM
*SECOFR
midrange.com
