ChatGPT - don’t rely on it

[dspur]

Just to check accuracy and for kicks and giggles, I experimented with some fender tag info on ChatGPT. Most all of it was wrong. I suspect other AI platforms are as well.

Very useful for a lot of things, just not fender tag decoding on Mopars.

 

[PurpleBeeper]

Just to check accuracy and for kicks and giggles, I experimented with some fender tag info on ChatGPT. Most all of it was wrong. I suspect other AI platforms are as well.

Very useful for a lot of things, just not fender tag decoding on Mopars.

True for now. Chat GPT searches the internet. With little or no information available (or with conflicting information) it makes mistakes. As it gains better information, it gets smarter (every second)

 

[Sam69sat]

Yep, i notice it gives me the wrong answer or questionable answers sometimes. Then i'll ask it, "are you sure that's correct" and it will come back and say "Yes, you are right, the answer i gave you before was wrong" or something like that.

 

[kiwigtx]

I asked Co-Pilot .... similar tech ...

 

[OneWheelPeel]

True for now. Chat GPT searches the internet. With little or no information available (or with conflicting information) it makes mistakes. As it gains better information, it gets smarter (every second)

My understanding is it is also trained on data besides just searching, but the odds of the engineers thinking to include that kind of mopar data aren’t good. I’ve run into it with other things and sometimes filled its blind spots by uploading a doc with the details it’s missing.

 

[WileE]

This is my GO-TO guy >

 

[Hilljack68]

Call me an old fart, but they can stick that AI crap. Throw it on the heap with Tic-Tok, Instagram and all that other bullshit.

 

[SlinktRR]

AI is a tool. Stuff like Alpha Fold is very useful for specific tasks, but offloading your brain to AI completely like chatgpt is a recipe for disaster.

Human critical thinking is still required because AI at the moment relies on the information built into the models, and tends to get worse when it makes its own inferences. Think echo chamber. I don't think AI will ever replace the creativity and agency of a person but who knows what will come in the future.