Elon Musk claims humanity will only survive so-called AI Revolution with 5-10% chance

I think there is a point at which intelligence either maxes out or no longer matters, and humanity is pretty close to that level. Neither Einstein (if he were alive) nor a super-intelligent AI will consistently beat me at tic-tac-toe or chutes and ladders.

T-1000 wrote:

I like those odds. We have a zero % chance of surviving ourselves.

This

Racket wrote:

Weak AIs. I find the whole AI fanfare a bit difficult to take seriously when we haven't even resolved the hard problem of consciousness. Also, Kurt Godel didn't think AI was possible either if I recall correctly

One distinction to make is that I'm not convinced AI needs consciousness to be dangerous. Narrow AIs could be used just like nuclear weapons to bring civilization down. China and Russia are already hacking us, and I assume we are hacking them. Narrow AIs could be weaponized to target opposing states in a multitude of ways.

DeepMind researchers have developed a neural network loosely modeled on mammalian brains to craft an artificially intelligent program capable of navigating through mazes.

The results were published in a paper in the journal Nature on Wednesday. DeepMind seems to think its work lifts the lid on how brains really work.

sportscar wrote:

audacity wrote:

What AI revolution? Even multi-million dollar ones can't answer these questions, but many 10 year olds can:

- Invent a game that is somewhat original, easy to learn, and fun to play

- Tell me about an exciting thing you did when it was raining outside

- A group of friends is standing on the side of a lake. The lake just froze last night, and one person walks out onto it. What happens next?

And people worry about a being that can't answer those things? There's a sucker born every minute, and some are just richer and/or more well known than others.

To be fair it took evolution millions of years to develop a species capable of answering those questions. It's taken computers about 30.

People who think AI isn't coming are amazingly short-sighted.. as sportscar wrote, it took homo sapiens a VERY long time to get to where we are today. We are really just at the foot of the mountain in terms of our technological development and look where we are already... you can ask your PHONE a question, and it will answer you back. Sure, it's just fuzzy logic now, but consider how remarkable this is given where we were only say 10 years ago (a blink of an eye...). The other aspect to the people who say it will 'Never happen' is the idea that somehow this intelligence has to be like us....we don't even truly understand the nature of our own sentience yet we reject the idea any other entity might equal or even surpass us? We're just slightly more intelligent than apes really; just smart enough to complete sh*t in our own nest to the point we may not even be able to live here in 100 or so years. We aren't that great and certainly not the apex of evolution...

The ingredients in the AI stew are 1) Pattern Recognition, 2) Repetition of Recognized Patterns, on a 3) Massive Scale, over (most importantly ) 4, TIME. It is definitely going to happen; what we don't know is will AI be relatively benign, desire to work in our service, regard us as an enemy, not regard us at all, what?? No idea.

it takes time wrote:

look where we are already... you can ask your PHONE a question, and it will answer you back. Sure, it's just fuzzy logic now, but consider how remarkable this is given where we were only say 10 years ago (a blink of an eye...).

How was this much different than 10 years ago? In 2008, you typed in a question on Google, and the answers you'd get would be almost identical to what you'd hear today from "asking your phone". The only differences would be for current world news and events.

Get over AI. 1st word- artificial. Can not compare to RI. 1st word-real.

HA!HA!HA! the picture shows an evil robot with TEETH. As if it needs to eat food. Who is the dope who invented that one? Is it programed to brush and floss 3 times a day or just in between destroying humans? LMAO!!!!

It is a picture from the movie Terminator, you complete imbecile.

Letsrun is Skynet

This is about humanity and AI advancing together in a symbiotic manner. A super intelligent organic machine.

Imagine walking down the street and looking at faces and seeing stats- height/weight/Facebook/twitter/socialmedia

Being able to view your surrounding world with different background skins, telepathy, dash cam recording vision, Memories uploaded to the cloud, possible existence in the cloud, download consciousness into different organic beings.

This place is full of complete morons.

Remember how simple computer viruses were 10, 20 years ago, compared to now? Notice how those viruses are essentially allowed and legal now, in the form of algos that track you for advertising gain (Facebook)?

Notice how robots are getting stronger and smarter to the tune of years, decades, EACH DAY?

Notice that science has developed microbrains that will undoubtedly find their way into robots? Yes, this is a cyborg, if you remember Terminator.

I'll stop now, since you think I'm crazy and I think you're naive. Just worry a bit, though. It will be trivial for a large number of robots to spin off a "gray goo" substance. Do the math. Suddenly, that gray goo covers all land on earth. Killing all breathing organisms. Robots don't breathe, and they can assimilate the goo once we're gone. Have a nice day.

Superhuman AI will eventually happen if we don't nuke ourselves or get killed off by an asteroid or a pandemic.

But the time frame for this is long. 100 years minimum. Doubt anyone here will be alive by then.

Much as you're all morons, we deserve this end wrote:

Notice how robots are getting stronger and smarter to the tune of years, decades, EACH DAY?

Wake me up when we finally have a group of robots that can play better soccer than a group of 5 year olds.

Good night.

My fleshlight gained self-awareness last night when I tried cunnilingus on it

Why is it assumed that robots will become hostile to humans due to some sort of evolution of the desire for self preservation or evil world dominance? Wouldn't it be just as likely that AI would evolve to be pacifist? If you program your war robots to kill one side in a conflict and protect your side, wouldn't it be possible for the robot's AI to evolve to understand that killing is bad and then refuse to fight? Why is it more likely for a robot to conclude that killing is good and want to kill all humans? Maybe the robots will take away all the weapons and bring about world peace?

From the former MIT AI director:

https://www.technologyreview.com/s/609048/the-seven-deadly-sins-of-ai-predictions/

Silicon valley billionaires don't tend to be all that smart (outside of the Google founders); they're basically successful salesmen. Don't take things they say seriously.

It runs XC now. Get ready, you inferior humans.

http://digg.com/video/boston-dynamics-atlas-robot-running