ATL Bear said:
nein51 said:
4th and Inches said:
nein51 said:
AI will be the end of humanity. There's no doubt about it.
we are headed to somewhere between Hal and Skynet
I think Armageddon is more likely. Someone is going to create the wrong deep fake and start a war. It's only a matter of time
The modern version of the WOPR from War Games, right? The theoretical negatives of AGI are concerning. But advanced AGI is difficult. The ironic reason is algorithmic solutioning relies on a basis of order, whereas while human behavior prefers order, we have the capacity and adaptation for chaotic reasoning.
The larger problem is one of basic database…garbage in, garbage out. So the AI can "learn" but it can only learn what it's taught. It can't actually think it can only regurgitate.
This is wildly problematic as humans are flawed and the temptation to feed bad information will always be there just to see the outcome.
Nevermind the further devolution to tribalism whereby the AI is taught to block some information while forwarding other. Think every news outlet at 5pm with the exact same cold open using the exact same words (like CNN and MSNBC do now only every outlet is the same).
If you don't think governments will control AI to their benefit then you're crazy.
What happens when an extremist group fakes a large scale attack using AI with the goal of forcing quick retaliation? What happens when humans can no longer decipher real from deep fake (just saw an AOC video on FB that is hilarious and obviously fake but the person who posted it didn't know it was fake)?
To me, there is no end for AI that is a net positive.