[quickthrowman]

I sell and run electrical work and I don’t see any good use for LLMs for what I do on a day to day basis.

LLMs don’t understand constructions drawings so they’re no help when it comes to doing a takeoff. Construction specifications are already well-organized and able to be searched so LLMs don’t help there.

They can’t synthesize information from different sources (words from a person spoken on a phone conversation, napkin sketches, information that is embedded in an electricians head about a specific facility, etc) or coordinate multiple parties through a variety of communication methods (email, text, phone calls, RFIs, in-person meetings, etc)

About the only use I’ve found for them in my line of work is cleaning up data from outside sources. YMMV. Construction is a very relationship and trust based business that has been around longer than almost every other profession.

[tharmas]

Yes, totally agree. LLMs are largely pattern recognition. They cant reason.

[orwin]

That's arguably the most interesting about the discussion on LLMs. Can't they reason? If they do, reason is an emerging fonction. 2 years ago, before inference was truly added to LLMs,a Honk Kong university paper wrote that ChatGPT 3.5 reasoning was correct 64% of the time on a specific task (you asked it to make the same reasonning 100 time, 36% of the time it would be wrong for no reason). I'd like to see how modern LLMs with added inference matrices and a lot of helpers before the actual transformers perform on this test.

If a consensus arrives in 5 years and we decided that yes, LLMs can in fact reason, reasonning would be an emerging capacity, and that would be incredibly interesting.

[tsunamifury]

“The internet is a passing fad”

[quickthrowman]

I don’t think LLMs have zero applications, I just haven’t found any for my specific use cases. If someone is able to figure out a way for LLMs to interpret construction drawings, it would be immensely useful, I’d be able to price substantially more work.

[sleepyguy]

I think back to a 1991-1992 when someone stood over my shoulder watching and telling me what a waste of time it was and who they hell is ever going to use it :)). I remind them from time to time and we have a good laugh.

[do_not_redeem]

Ah yes, the same quote often used by cryptocurrency enthusiasts.

"One time a new tech succeeded despite skepticism, therefore all new tech will succeed despite skepticism"

[tharmas]

Distributed ledger/blockchain is an interesting idea but far too energy intensive to be of practical use.

And then theres the whole concept of what currency actually is.

[eiishiiiou]

"[Distributed ledger/blockchain] | [LLM] is an interesting idea but far too energy intensive to be of practical use."

[do_not_redeem]

"And then theres the whole concept of what intelligence actually is."

Thank you, I'm glad someone understood my comment at least.