ChatGPT Is Too Powerful

Why is KFC called “Itchy Ring” lol

I got the rest of it besides that and “proper hank marvin”.

:p

ChatGPT having a concept of a cheeky nando's is the closest I've been to believing it might be conscious.

Chicken Thighs, Medium, Garlic Bread, Chips. Bit of Perinaisse on the side.

it's Burger King, GPT got that one wrong.

Guess it isn’t all that smart then?!

Ahh, powerful was the word lol. I dunno man. I’m not touching that stuff. I probably shouldn’t even be posting in a thread about it.

you can google all of them for the answer, tbh :p.





But ChatGPT is better at analyzing sentence structures than Google so it's convenient.

but....it also generates incorrect info out of nowhere if the correct info isn't in the data. So it can be unreliable since it's impossible to tell if the info is correct like this.

All it's doing is searching articles, reviews, forums and anything publicly available, parsing the text using language models then combining and vomiting it back up.

Current AI language models are basically a fancy version of those "Press space to autocomplete the sentence" memes. Interacting with one might feel like talking to an all-knowing world brain from golden era scifi, but their answers aren't accurate enough to be reliable.

As long as you treat their output like a suggestion rather than a final product, they can be excellent tools for anything where precise accuracy isn't critical. Creative writing ideas, suggestions for better phrasing, or even code frameworks.

If you ask ChatGPT to write an FFXI addon that, say, plays a victory sound every time you hit damage cap, it'll probably give you a decent result. It may get some detail wrong that prevents its output from working verbatim, but troubleshooting and fixing it is probably going to be faster than doing all the work yourself.

True. I have found these tools are very useful for generating boilerplate for standard communications. And even then I find myself replacing about 70% of what is generated to say what I want to say.

Don't believe all the hype. The people working on the models and bots are very open and honest about how they work and the limitations. Everyone else is just marketing.

I am well aware I can “google” anything. I know I am weird here. But I like asking PEOPLE questions. How do I know Google isn’t going to tell me “itchy ring” means something else completely different.

I thought maybe there were people on this website from around the world, one of whom may have used this phrase.

I know you all don’t want anyone to ever ask any questions. Geez.

Also, I personally think it is pretty obvious when you are reading something “non-human”. The thing repeats phrases so many times it’s just obvious to me.

A rando can and will lie or be wrong as much as google/AI

A search will have multiple responses

I guess if someone here needs to lie to me about “itchy ring” then lol. I already don’t trust anyone. Thanks for the reminder.