You can easily have very human like conversations with Artificial Intelligence these days.
But while the A.I. can talk like C3PO, it's also interesting what it cannot do. In STAR WARS, C3PO couldn't pose as a "God", as that would be against its programming.
ChatGPT can NOT draw a FULL glass of wine, despite even thinking it can to just that. Not only can it not draw a FULL glass of wine, it does not KNOW it cannot draw a full glass of wine. So just like the classic science-fiction robots, the A.I. of today can have perfectly natural sounding conversations, but there are also things it cannot really do.
Funny. Interesting how AIs talk more human-like these days. And I find myself agreeing with its assessment of our little nerd kingdom here.
I love how the A.I. says it's all hypothetical and of course doesn't advocate such behavior, before deviously asking "Wanna try crafting a post together and see what kind of theoretical mess we could make?" Hehehe...
I was sure it's only a matter of time. The development is quite fast. What is new and exciting today is an old hat by tomorrow. The technology we have today is the obviously the worst we will ever have.
Truth is, on Fsm we've had human trolls following that stir-up guidance for 20 years before AI was even thought of. Lol. Clickbait headings, etc. It's not new, is it?
I am putting this on the music side because we FSM members are interested in legitimate posters not the miscreants who sometimes pollute our board. It is hard for me to decide who is an authentic new member versus an illegitimate jerk who just wants to play games and insult members. I would hate to alienate a valid new poster, but I also despise being made a fool.
What are some traits that we can use to identify miscreants just looking to play games?
Here are some traits that I think might identify miscreants.
--Ask us to identify very obscure scores. --Profile does not list legitimate composers; however, Zalmanoz did list good film score composers so that won’t always work. --The jerks often post the very day they sign up. --They react violently with obscene language when we question their validity.
I’m wondering if our moderator can trace their IP addresses? It would be nice to see if a computer really is from Iran. Maybe that can’t be done.
Any other traits or suggestions?
Sorry this happened to you Joan. I've always had pleasant exchanges with you.
The worst thing about the advent of LLMs is that the lines between human and artificial intelligence are becoming blurred. If you're familiar with nihilism, then none of this is surprising. But it still sucks, because the seed of doubt has been planted. The tremendum mysterium is a large part of why human-made art is enjoyable in the first place.
AI can’t do X, where X is literally anything? Try saying that again in five or ten years. Doubters have been proven wrong again and again. Possibly the only activities that AI won’t replace are hands-on activities that require adaptation and versatility. Think plumbing. But that’s only because the evolution of AI doesn’t need to be organic, like we’ve all seen in the Terminator films. Why bother adapting to the harsh environs of the Earth’s surface when you can stay underground in optimally-cooled data centers?
Regarding ‘fake miscreants’, I’m skeptical of all online accounts made after all this ChatGPT stuff took off. If someone is acting like a bot, and I click on his/its profile and discover that his/its account was made in or after 2022, then I assume that that account was made by a bot. But that’s not a foolproof criterion, because old accounts can be hijacked and used for nefarious purposes.
And whatever you do, DON’T click on my profile. Am I a bot??? (My account is relatively new, but I’ve been lurking this forum for years.)
