Anybody interacted with this yet? I've tried it once in the early days and servers were at capacity.

Really impressively general capability from what I see, although still buggy in some cases and simply regurgitating without creativity in others, it's still a step change IMO that has made me revise my outlook on the speed of AI development (towards the faster!).

I think now is very much the time to consider the far ranging economic, cultural and political impact of these technologies and think about how the governments of the world are going to contend with them. Whilst I don't buy into the doomsaying of superintelligent AI (at least not for say 50-100y), I do think the economic and workforce changes from more capable AI are going to hit hard and quite soon. ChatGPT can already pump out essays at easily high school level on any topic you choose, or fully functional code for tasks you define. Just those two applications are huge and this isn't even commercialised yet.

I think its almost scary how well it seems to work. And with how much confidence it has. Sometimes it is completely wrong, but you can't tell unless you know ahead of time.


I had it translate a scripting code for me from one platform to another, it did very well. Having it write entire sections of code to do simple things is easy and it works surprisingly well.

It is going to change things if it becomes commonplace to use, which I think is probable if it isn't behind a paywall.

@IronGoober It’s even more scary to think that this is the worst that AI will be. It will only get more sophisticated in time

Perhaps unfortunately, there's a lot of money to be made with this stuff and if it's not developed in the west then China will be doing it anyway. So I can't see progress being halted for long if at all.

I've used it before, but I take its answers with a grain of salt, as it gives answers with extreme certainty, even if it is wrong. I don't know what it really does or understands, sort of like the Chinese Room thought experiment (https://plato.stanford.edu/entries/chinese-room/#Over:~:text=Searle%20imagines%2...).

It does know something, though, as when I asked for 3 sources on slinging, it gave Slinging.org as the first source!

Chat GPT seems to always use a lot of words and filler, and gives a ton of disclaimers. When I asked it about the optimal slinging style, at the end it said, "It's important to remember that throwing rocks can be dangerous, and you should always practice in a safe location with proper equipment and supervision." well.... yeah, of course its dangerous...

All that aside, I think that Chat GPT can be used effectively as a tool, such as writing code as IronGoober and NooneOfConsequence did, or using it to get ideas on something.