https://chat.openai.com/

Nothing to say more, I regret that chatopenai didn't exist when we were kids and playing this game.

You can also paste your bugging code and ask to fix it.

I would like to use openai but they ask for the phone number to sign out...

Awesome!

When I asked it for a simple GUI script, it gave me completly wrong scripts (using commands like box and ctext).
It corrected itself after I said that it does not work and then used hudtxt! Wow.

I did upload some of my scripts and removed most comments.
ChatGPT was able to tell me what the scripts does exactly - very spooky...

Yeah

You could use https://www.temp-phone-number.com/ or sites alike.

@ Bowlinghead: Yes brother without doubts.

Wow.. okay... It's quite impressive (and yes, also kind of creepy) that it works with CS2D specific Lua scripting because it's not that popular

They really must have used basically the entire internet - including the CS2D Lua script archive here - to train it.

You can't use even a voip number of what I know so yeah gg.

@ DC: You need to regret that this is released in 2023 rofl. Imagine this script being launched in 2010 or 2009 rofl. I'm pretty sure CS2D would be lit.

https://www.youtube.com/watch?v=toL1tXrLA1c

@ Xirot: Wait! Would be!? CS2D IS lit!

Chat GPT is incredibly impressive but it isn't perfect. It sometimes does mistakes. Fun fact: Sometimes it's also able to fix its own mistakes if you tell it to do so.

I think that it's also dangerous to blindly trust its results without checking them carefully. It sometimes makes things up which look legit but aren't true / don't work at all / may cause problems. This is also why I don't think that many jobs are threatened by it. That may change in future but I think making it 100% reliable is pretty impossible. It's trained with data from the internet. Data made by people - and people make mistakes too (and also put random nonsense on the internet intentionally).

That's not the problem with neural networks. A human being asking a question already knows the end result he wants to get, but a neural network does not know this - it seeks an unknowable result. Human scientists who learn new things do the same thing and same errors. Therefore, it is quite logical that code errors and unoptimized code appear. With time and increasing amount of code passed through neural network (which we can influence by making mass queries and/or checking old code), such situations will decrease and improve.
In the future, with the emergence of full-fledged AI, which will also know the desired end result, the situation will become even more ideal.
For example, I translated this text using a similar neural network www.DeepL.com/Translator (free version)
It is significantly better than any other translator already.

So cool
But it has many errors, it didn't make one working lua script for me till now even if i ask for simple things

@ Pagyra: Just to let you know, Google Translator also uses AI.

Maybe not now, but in future it will be possible, If we asked people who lived 1000 years ago about flying planes, cars that go over 200km/h and so on, they would say that's impossible too

@DC

Really funny, he made a script for me when I've noticed that there was a bug in his script he fixed it rofl.