Google's AI built its own AI

[tmcom]

5 hours ago, Dark_Grey said:

It's projected to develop 200 years worth of technology in just under a week. The question isn't if the AI is capable, it's what humans will do such insane power we haven't prepared for.

Bury it in area 51!

[Occams Razor]

Robots will never be able to wipe us out. If they become aggressive we can take the battery out. Simple as that.

[Emma_Acid]

On 12/5/2017 at 9:02 PM, Dark_Grey said:

It's projected to develop 200 years worth of technology in just under a week. The question isn't if the AI is capable, it's what humans will do such insane power we haven't prepared for.

Sorry, whats the source for this?

[Dark_Grey]

6 hours ago, Emma_Acid said:

Sorry, whats the source for this?

Sam Harris (neuro-scientist,) in his Ted Talk. He actually predicts 20,000 years of "human-level" progress in a week based on the speed of a computer's brain vs. our own.

Quote

HARRIS: Imagine we just built a super-intelligent AI, right? That was no smarter than your average team of researchers at Stanford or MIT. Well, electronic circuits function about a million times faster than biochemical ones, OK? So this machine should think about a million times faster than the minds that built it. So you set it running for a week, and it will perform 20,000 years of human-level intellectual work week, after week after week. How could we even understand, much less constrain, a mind making this sort of progress?

Full Transcript on NPR

I see no reason why this shouldn't be the case. The AI doesn't have to do anything extraordinary, just process data and solve problems the same way humans do - the difference being the raw speed at which this happens. Human beings are the "X factor" in this equation as computers are predictable but there is no telling how society will react to such intense progress, "week after week".

[Haroldbattschits]

It's been said that when computers create better computers that the pace of development will leap forward at blinding speed. Fear of machine Armageddon is very real. It doesn't have to be

[GlitterRose]

On 12/5/2017 at 9:11 AM, paperdyer said:

SKYNET!!!!!!

You beat me to it.

[GlitterRose]

On 12/7/2017 at 4:16 PM, Occams Razor said:

Robots will never be able to wipe us out. If they become aggressive we can take the battery out. Simple as that.

I'm figuring if they can think faster than us, they're probably at least a few steps ahead of us.

The last thing we'll hear as we're sprayed with nerve gas is...I'm sorry, Dave...I can't let you do that. 

 

Edited December 10, 2017 by ChaosRose

[DarkHunter]

On 12/8/2017 at 0:09 PM, Dark_Grey said:

Sam Harris (neuro-scientist,) in his Ted Talk. He actually predicts 20,000 years of "human-level" progress in a week based on the speed of a computer's brain vs. our own.

Full Transcript on NPR

I see no reason why this shouldn't be the case. The AI doesn't have to do anything extraordinary, just process data and solve problems the same way humans do - the difference being the raw speed at which this happens. Human beings are the "X factor" in this equation as computers are predictable but there is no telling how society will react to such intense progress, "week after week".

There are two major problems with all of that, though arguably one of them is being worked on.

The first major problem is that computers are not creative and have so far not shown the ability to be.  While they can due certain things far faster then a human they can only do it within the limits of how they are programmed.  So far computers can not create anything truly new, all they can potentially do is quickly optimize what has been created.  A computer isn't going to create some new field of math or create a new theory for physics, to use a programming example a computer might be able to create the best calculator program imaginable but that same computer isn't going to make a game engine.

The second major problem, and what is being worked on somewhat it seems, is that computers are horrible at putting stuff together.  To be more precise computers are only good at making connections within certain parameters but are terrible at making new connections.  For example you tell a computer to find all red cars, it will find all red cars of a certain shade very quickly but having it find all cars we would consider red is much more difficult.  You can get around that with toying with the parameters, making them very large and precise, but a computer still really can't figure that out on it's own.

Due to being at work and in a rush I feel like I did a poor job explaining what I meant for the second major problem.

[Area201]

They lack self awareness and only mimic the human brain computer and human appearance even. Bigger concern is robot automation in the work force, on one hand that's good but on other hand society needs to adopt as there will be more and more work shortages. The terminator type AI bots to take over humanity is not really as a big concern for me. 

Edited December 12, 2017 by Area201

[FlyingAngel]

On 12/5/2017 at 10:20 PM, Dark_Grey said:

Let's put it this way: the first nation that develops true AI will have god as their ally. The race is on.

Except, you can't control God ^^

Possessing a power that you can't control, we've all seen that on every movie what's the consequence.

 

On 12/8/2017 at 1:16 AM, Occams Razor said:

Robots will never be able to wipe us out. If they become aggressive we can take the battery out. Simple as that.

Robots are just one of many concerns. If there is a higher intelligence like a mad scientist who controls the robots, I don't think it's that easy to wipe them. One is enough, what about one country who possess nuclear power and decide to enslave the world. Anything could happen, we human are not currently living in peace.

Edited December 17, 2017 by FlyingAngel