Angelus said:
Five ways AI might destroy the world... Ok...and in how many ways exactly are humans ALREADY destroying the world? |
Many, but AI is much more guaranteed to destroy the world. Stop shilling for the apocalypse. End of discussion.
Can AI be trusted? | |||
No | 38 | 70.37% | |
Yes | 16 | 29.63% | |
Total: | 54 |
Angelus said:
Five ways AI might destroy the world... Ok...and in how many ways exactly are humans ALREADY destroying the world? |
Many, but AI is much more guaranteed to destroy the world. Stop shilling for the apocalypse. End of discussion.
CaptainExplosion said:
Many, but AI is much more guaranteed to destroy the world. Stop shilling for the apocalypse. End of discussion. |
How am I shilling for anything, exactly?
Angelus said:
How am I shilling for anything, exactly? |
You're inferring that mankind is worse than AI purely because of mankind having more experience nearly destroying the world and that we might as well stick with AI because it hasn't made as big a mess of the world as we have. In short, AI is the lesser threat based on your words. You'll change your mind if, and it's very likely to happen, governments try to rely on AI for defense and the AI ends up launching nukes.
Probably not but I don't really care. All I need is ai sex dolls and my life shall be complete.
PC Specs: CPU: 7800X3D || GPU: Strix 4090 || RAM: 32GB DDR5 6000 || Main SSD: WD 2TB SN850
Where it's at currently?It's harmless in the sense it won't cause malicious intent to anyone. It will cause people to lose their job. Will we ever get to the point AI has the will power to harm others? Potentially, but hopefully not.
Wdym "trusted"? AI in its current state is merely a tool that has no ability to think or act on its own. The biggest threat it poses at this current point in time is the destruction of jobs (as you've pointed out the Hollywood strikes) and the further spread of misinformation (deep fakes for example). But both of these examples require human input to implement, so really you're asking if humans can be trusted to use AI in a responsible and ethical way, and the answer is a resounding no.
CaptainExplosion said:
You're inferring that mankind is worse than AI purely because of mankind having more experience nearly destroying the world and that we might as well stick with AI because it hasn't made as big a mess of the world as we have. In short, AI is the lesser threat based on your words. You'll change your mind if, and it's very likely to happen, governments try to rely on AI for defense and the AI ends up launching nukes. |
That's not actually what I was saying at all, but ok.
Currently - no, not really. I caught Bing chat lying on several occasions.
Eventually - perhaps.
CaptainExplosion said: Imagine if an AI had advanced to the point where it decided that mankind is outdated so it might as well nuke us all. We as a species must band together and destroy AI before it destroys us all. |
This is a really cartoonish outcome and it won't happen.
(edit: Well, at least not literally. Something close to that could happen in the sense that AI might replace most jobs and lead to really bad standard of living for everyone except the ultra rich.)
Can AI be trusted... I feel like that's the wrong question. AI is just a tool created by and used by people. It could be an extremely helpful tool for people of all trades, artists, doctors, architects, whatever you name it, AI could make everyone's jobs easier and more efficient in many ways. But that's not what happens, right? Instead it's used to replace people, because no paid employees means more profit for the higher-ups. So you get companies offering actors to work for a day and then have their likeness replicated by AI forever for no pay (even after death, death gets in the way of profit). Openings of TV shows created entirely by AI because it's just easier and cheaper than hiring human artists. Only made worse when you think that the way AI works is that it's essentially stealing people's work in a way that doesn't quite qualify as theft.
In short the question is, can people be trusted? No.
Last edited by mZuzek - on 23 July 2023