A conversation with ChatGPT

[Michael GilliganParticipant @michaelgilligan61133]

Whilst sitting in the Hospital car-park, between appointments, I have just had an astonishingly useful discussion with ChatGPT

The content doesn’t really matter, but our closing exchange says a lot.

I wrote:

The most helpful conversation I have had for a long while … With Man or Machine !

Thank You

… and it replied:

 

.

.

 

MichaelG.

 

[Michael GilliganParticipant @michaelgilligan61133]

Grrr!!

The almost inevitable forum failure …the screen-shot is not visible 🙁

I will try adding it when I’m at home.

MichaelG.

.

Edit: … and as if by magic … it has suddenly loaded

[Dave SParticipant @daves59043]

Recently I got Claude (Anthropics AI) to write me some G code for an incrementing Sin wave cut. Got it right and worked as expected.

These things have come a long way in a short space of time, but currently like all tools are only as good as the person using them. I suspect that will become less the case, though I’m not sure what that means

Dave

[Speedy Builder5Participant @speedybuilder5]

What was the original question ?  ” Is my appointment still on-time ? “

[Dave SParticipant @daves59043]

This is a flat piece of copper:

Code by claude

[Dave S]

On Dave S Said:

though I’m not sure what that means

Same here. I have a lot of opinions on AI, but nobody knows anything for sure. What worries me the most isn’t so much the AI but the people selling it. These people are psychopaths at best, and are actively out to make life worse for everybody but themselves.

[Speedy Builder5]

On Speedy Builder5 Said:

What was the original question ?  ” Is my appointment still on-time ? “

If I thought that [rather than just playing the Jester] you were interested, I could post the full transcript. … But I won’t waste the effort.

MichaelG.

[Michael Gilligan]

On 5 March 2026 at 13:25 Michael Gilligan Said:

Whilst sitting in the Hospital car-park, between appointments, I have just had an astonishingly useful discussion with ChatGPT

The content doesn’t really matter, but our closing exchange says a lot.

ChatGPT’s ingratiating closing remarks are presumably designed to make interacting with the thing as addictive as possible. I’ve also been playing with it – although I’m having to strictly limit the time I spend/waste ‘talking’ with it, feeling only too aware of the addiction risk.

Recently, I sought information on Linkwitz-Riley filters, and the component values in the specific circuitry of one example. I don’t know much about these things, but I’ve no reason to suspect that ChatGPT’s information was anything but spot-on – and it was as helpful as a patient human mentor. It also anticipated my questions, to some extent. Worryingly human, in a perfect-servant sort of way. And so now I can go and produce something I don’t really understand, but believe ChatGPT does, and it will probably work (but I may not know). So expertise has just taken a hit. The ignorant and inexpert are facilitated in their mission to create confusion and havoc.

Anyway, the hopefully helpful point of this post is to note that the version of ChatGPT one can access without paying is dynamically allocated by ChatGPT. I thought that the free service was V 3.5, and ChatGPT agreed, but I was accessing V 4, and latterly V 5. It’s worth asking ChatGPT to identify itself – you may be pleasantly surprised. Or perhaps this was a cunning sweetener to encourage me to keep using it, because I’d been trying to get it to tie itself in logical knots – and succeeding.

[Author]

Posts