Artificial Intelligence

BlueHammer85 said:
In the same way covid caught everyone including politicians totally off guard, many many thousands died but ultimately governments worked together, public locked down and science come together to create a vaccine and see the virus out.

There is many institutions and governmental departments currently working on the threat of AI.
Click to expand...

In the scheme of pandemics, Covid was one of the lesser deadly ones and people had time to react. AGI will wreck everything before they even realise there's a problem. And there are no major governments working on the threat of AI. AI safety is a purely academic field still that is vastly underfunded.

As mentioned, the problem here is that people don't even understand the threat so have no way to understand the dangers it presents.
 
Georgian Maestro said:
I recall listening to Brian Cox talking about AI on the Joe Rogan podcast a few years ago and he was of the opinion, after consulting experts on the subject, that a general intelligence is decades away and is of no immediate concern. I wonder if his opinion has changed.
Click to expand...

The current expected range is between 2032-2048 with 2041 being the middle ground. In extreme examples, it could happen as early as 2028.
 
Damocles said:
The current expected range is between 2032-2048 with 2041 being the middle ground. In extreme examples, it could happen as early as 2028.
Click to expand...
It does seem to be speeding up..everyone trying to outdo each other
 
Damocles said:
In the scheme of pandemics, Covid was one of the lesser deadly ones and people had time to react. AGI will wreck everything before they even realise there's a problem. And there are no major governments working on the threat of AI. AI safety is a purely academic field still that is vastly underfunded.

As mentioned, the problem here is that people don't even understand the threat so have no way to understand the dangers it presents.
Click to expand...

Have you built your bunker ?
 
Damocles said:
Human intelligence is just clever genetic algorithms
Click to expand...

This last year has taught me that the amount of people underestimating AI is nothing compared to the amount of people overestimating human intelligence.

The chess example given above is the perfect example. No human players can get close to Stockfish - it is about 20-30% better than the best human players. People don’t understand these AI don’t have a memory, they’re not referencing a database - there is no database (I can’t stress this enough, that is not how AI works). It is a neural network, just like our brains, and it “feels” the game of chess in the exact same way humans do. From a systematic perspective there is no functional difference, in both cases its a bunch of on and off switches organised in a complex network that is trained on experience. The only difference is efficiency. In people evolution has made neural networks unfathomably energy efficient. To human chess grandmasters what a computer player can do is indistinguishable from magic, it comes up with novel solutions to problems the human player hasn’t even identified yet.

There is nothing special about human beings. We are just a multimodal AI with an incredibly rich training data set, and incredibly efficient neural networks. Anything we think of as “creativity” or “innovation” is just novel combinations of things we have been trained on - this creates the illusion of us materialising something new. In some fields AI already does this far better than humans can, and that is only going to become more prevalent.
 
YoungBlue said:
http://www.bbc.co.uk/news/technology-19354994

Just read this article and it seems we're getting closer and closer to the possibility of AI's.
As a human race, I think we're incapable of developing these in a controlled environment as a well designed AI is expected to have the capability to self improve itself. Being technology at it's finest themselves, the risk of them surpassing all the constraints of human intelligence is more than possible.

Then what happens when an AI actually becomes smarter than a human? It will be impossible for any human to fully understand as they will then just keep improving itself as much as it possibly can, which could make it stupidly hard to predict, maybe understanding the laws of logic far better than any human can.

What does everyone else think about this? I hope I'm not the only one who gives a shit!
Click to expand...
Frightens me to death, glad I'm 63 and not 13. Selfish I know and who knows what's beyond death, if anything?
 
You must log in or register to reply here.
Unread Posts

Don't have an account? Register now and see fewer ads!

SIGN UP
Back
Top
✖
  AdBlock Detected
Bluemoon relies on advertising to pay our hosting fees. Please support the site by disabling your ad blocking software to help keep the forum sustainable. Thanks.