Sure, but.. I've been coding for 40 years and I don't know everything. To me, a LOT depends on what the plumber asked chatgpt about. For example: building codes in that city, to figure out what his options are - like, is he allowed to just put in any old toilet, or is there a gpf restriction? What's the replacement part number for faucet XYZ's gasket? Those seem reasonable.

"how do I fix a clogged toilet?" would be bad..

I cling a bit to a prompt i sent a while ago about just tossing a chopped pepper into a recipe for baked ziti. I had a recipe that i followed fairly tightly with slight changes to see how they would work out each time. Instead of prompting "when should i add chopped bell pepper?" the small change of just, "what are my options for when to add chopped bell pepper?" opened up a variety of different methods i could try when returning to that recipe, and decide what i like best based on the outcome.

The first prompt style is I think a way society towards drifts incidentally towards a less interesting one, with less variety in solutions. The second one i think allows people to still exercise their potential to try a variety of things and keep that variety.

>like, is he allowed to just put in any old toilet, or is there a gpf restriction?

And if the LLM gets that wrong? It's his job to know the codes or how to go to a reliable resource to find out the correct codes.

Presumably in his jurisdiction he should know what official resources to consult. But the point about it depending on his question is definitely fair.

[dead]