On the issue of jobs....

[moonbogg]

Yes it is bound by the speed that electricity can travel, by the amount of time it takes to flip a bit, etc.... you know that dirty little thing called physics.....

Jesus Christ. Is there a place somewhere that offers lessons in trolling? Because you suck at it. Seriously, I guess I've been trolled in a way here, but not by the contents of your post. Its the pathetic troll attempt. It makes me angry. Why can't you do better? WTF is wrong with you? You don't realize the likelihood of a troll fail when you see it?

OK, look. Try this next time. Instead of saying something that is so easy to shoot down, say something that is inline with your target's thinking. Agree with everything they've been saying, but add in a little twist. It has to be subtle. Get inline with your target, but add that tiny little exception that you know they won't be able to accept. They will object, and you compound it by agreeing yet again, but you add another little twist. Keep twisting until the circulation in their legs is cut off by their panties bunching up.

 

[glenn1]

I have thought a lot in an uneducated way, about the how machine intelligence might evolve. I think I have seen robots learn to move around by learning from trial and error. That led me to think that if machines can learn from actions performed in the real world, they might learn such things much faster by simulation, learning inside a virtual environment based on the known laws of physics. Instead of say, building a machine that then learns to walk on real ground, a machine might build a virtual model and then learn to walk in a virtual reality. In this way it would seem to me that iteration after iteration of such computational modeling would lead not only to program for walking but a perfection in the design, a machine say that could change its design at will to sprint, jump, climb, roll, fly etc.

Okay, for sake of argument let's say such modeling exists today and could "change its design at will to sprint, jump, climb, roll, fly etc". What organizing purpose would said machine have? After it developed the ability to sprint would it suddenly then develop the intelligence that it should cross the street, go get a pizza, and join the Occupy Wall Street movement?

Hell, the machine you described sounds exactly like the jobless people now - able to walk, run, etc. but no ability to think independently, come up with value-creating work to do absent direction from another, or do much at all except for walk around mindlessly worrying about how the next generation of robot will make his generation of robot obsolete.

 

[bshole]

Jesus Christ. Is there a place somewhere that offers lessons in trolling? .

Um riiiiggghhht...... I have spent my entire life writing embedded software. All my immediate coworkers are embedded software engineers. I have spoken on this topic at various points in my life with many different software engineers. EVERYONE I have ever spoken with about this have thought that it was a laughable impossibility and that we have not moved a single inch closer to real artificial intelligence. In the realm of software engineers, I would be very surprised if you could find anybody who thinks that the technology is doable. Perhaps you would get a different opinion from physicists. I do believe you that are VASTLY underestimating the complexity of the human brain.

Next you will be telling me that we will have colonies on Mars, etc....

Not trolling, just a completely different world view than you.

 

[Londo_Jowo]

That stuff is awesome. Nice post.



Your troll has failed. You've been owned by the above post. Also, your brain seems to be functioning, yet based on your posting your brain is not "logic" based either. You can't explain that.

What troll? Where are the AI brain controlled robots? Are they ready for prime time? How much do they cost?

I have far more ability/experience programming logic controllers (Ladder logic, structured texts, & function block) than your brain can grasp.

 

[moonbogg]

Okay, for sake of argument let's say such modeling exists today and could "change its design at will to sprint, jump, climb, roll, fly etc". What organizing purpose would said machine have? After it developed the ability to sprint would it suddenly then develop the intelligence that it should cross the street, go get a pizza, and join the Occupy Wall Street movement?

Hell, the machine you described sounds exactly like the jobless people now - able to walk, run, etc. but no ability to think independently, come up with value-creating work to do absent direction from another, or do much at all except for walk around mindlessly worrying about how the next generation of robot will make his generation of robot obsolete.

You are comparing a future technology with today's software capabilities. No worse comparison could be made. This is like explaining a Formula One car to someone living in the time of the horse and buggy. I'd expect them to say things like, "But wheels will fly apart at that speed and horses, no matter how strong, will never run that fast. Besides, the roads are too rough. Impossible".
Yes, this is what people are doing when comparing future technology to anything going on today. Baby steps will get us there, then suddenly it takes off.

Um riiiiggghhht...... I have spent my entire life writing embedded software. All my immediate coworkers are embedded software engineers. I have spoken on this topic at various points in my life with many different software engineers. EVERYONE I have ever spoken with about this have thought that it was a laughable impossibility and that we have not moved a single inch closer to real artificial intelligence. In the realm of software engineers, I would be very surprised if you could find anybody who thinks that the technology is doable. Perhaps you would get a different opinion from physicists. I do believe you that are VASTLY underestimating the complexity of the human brain.

Next you will be telling me that we will have colonies on Mars, etc....

Not trolling, just a completely different world view than you.

This was much better. Also, nice hook with the mars comment. You're going to be fine.

But if I can move past the playing around, I'll say this. Just because you can't imagine a future technology using primitive coding from today, does not mean that advanced hardware can't be made that acts like a brain. You can see past your own nose on this subject, I know you can.

 

[moonbogg]

What troll? Where are the AI brain controlled robots? Are they ready for prime time? How much do they cost?

I have far more ability/experience programming logic controllers (Ladder logic, structured texts, & function block) than your brain can grasp.

Incredible. You did the following, exactly:

Doubled down on a bad troll hoping it would somehow work this time.

 

[Londo_Jowo]

Incredible. You did the following, exactly:

Doubled down on a bad troll hoping it would somehow work this time.

And here you are not being able to answer the questions and try to deflect by calling my posts a troll posts.

How soon are these AI bran controlled robots or device going to be ready to go prime time? (ie perform multiple unrelated tasks without human intervention)

How much are they going to cost? What type a maintenance do they need or will they be self maintaining?

What type of power source will they have? If battery, what happens in the event of a power failure?

What if a short circuit erases their memory, will there be programmer robots to reprogram their memory?

 

[Zaap]

You are comparing a future technology with today's software capabilities. No worse comparison could be made. This is like explaining a Formula One car to someone living in the time of the horse and buggy. I'd expect them to say things like, "But wheels will fly apart at that speed and horses, no matter how strong, will never run that fast. Besides, the roads are too rough. Impossible"

Actually a closer comparison would be someone who knew automotive technology in 1914 telling you that by 2014... the changes you might imagine won't really be all that dramatic. You'll still be driving around on the ground, on paved roads, in vehicles with a very similar layout as now (1914) just a lot of improved technology..."

But you insisting, "OH BULLSHIT! We'll have flying saucers and teleportation by 2014!!!"

But then again, this is another one of those P&N subjects where anyone with any real world experience is to be considered a troll, while people raving wild-eyed gloom and doom of advanced AI robots taking over the world and firing all the humans (probably using humans as batteries if we take it far enough) are to be taken seriously.

 

[moonbogg]

And here you are not being able to answer the questions and try to deflect by calling my posts a troll posts.

How soon are these AI bran controlled robots or device going to be ready to go prime time? (ie perform multiple unrelated tasks without human intervention)

Probably within the next 50-100 years for it to really come into focus and become real.

How much are they going to cost? What type a maintenance do they need or will they be self maintaining?

Probably cost about as much as a very large scale, high profile project such as the LHC, if not more to start with.

What type of power source will they have? If battery, what happens in the event of a power failure?

They will likely be stationary AI at first, being "plugged in to the wall" like a large computer at first, with more advanced, mobile units later. Possibly nuclear powered perhaps or a new technology which mimics metabolic function in biological systems.

What if a short circuit erases their memory, will there be programmer robots to reprogram their memory?

If it tends to break, better units will be built. The thing is, once you have a single true AI, plugged into the wall, it will be the first and last that we will have to design. From there the AI becomes the engineer for the next iteration and can design everything from the power source to the mechanical systems. It will do all of it better than we ever could.
The first one could be the size of a football stadium. The second one could be the size of a person and 10x as powerful.

Our conversation here is simple. I say it will happen, and you seem to say it won't. If you want to know why I think that, just do some googling on AI research, transhumanism etc. Someone posted some good links a few posts back that showed the direction research is going. There is great benefit to making something like this, but as with any powerful technology, it has risks. These risks are of an existential nature this time though.

 

[Moonbeam]

Okay, for sake of argument let's say such modeling exists today and could "change its design at will to sprint, jump, climb, roll, fly etc". What organizing purpose would said machine have? After it developed the ability to sprint would it suddenly then develop the intelligence that it should cross the street, go get a pizza, and join the Occupy Wall Street movement?

Hell, the machine you described sounds exactly like the jobless people now - able to walk, run, etc. but no ability to think independently, come up with value-creating work to do absent direction from another, or do much at all except for walk around mindlessly worrying about how the next generation of robot will make his generation of robot obsolete.

The methodology I hypothesized was just that, hypothetical, and would lead to nothing but a very efficient mobile machine. It would have no purpose other than one that was given it. The purpose of my example was to show one possible way machines might evolve via their profound computational capacities. The same methodology might lead to better mechanical vision, speech recognition, and perhaps also to thinking itself. A machine would also not have a four score and seven limit to its learning curve.

The reason you would cross the street to get a pizza lies in your need for food the desire for self preservation. Why do you care if you life or die? Why do you have a will to survive? If you know the answers to those questions and others, say how does self awareness arise out of molecular structures then you may have some real basis to opine on whether electrical impulses born by silicon could not be the same. The question I would ask is if it is possible to know if consciousness is an emergent property of complexity.

Those that claim that conscious machines can't exist should take a good look at themselves. Is the software that runs our consciousness embedded in and a result of the structure of our brains. Would you not say that intelligence is the survival value imparted by self awareness of ones environmental condition, that wherever there are things that continue to exist by replication, consciousness will inevitably emerge? Perhaps purpose arises out of the consciousness of our own death. What purpose would you have if you didn't have to consider that maybe for billions of years. And how long would a consciousness choose to exist if it could exist that long.

 

[Londo_Jowo]

Our conversation here is simple. I say it will happen, and you seem to say it won't. If you want to know why I think that, just do some googling on AI research, transhumanism etc. Someone posted some good links a few posts back that showed the direction research is going. There is great benefit to making something like this, but as with any powerful technology, it has risks. These risks are of an existential nature this time though.

I haven't said it won't, just not in your lifetime, most likely not in your great great grand children's time. Worrying about what affect on the work force at this time is ludicrous and that was the gist of the OP's posts.

 

[Belegost]

What troll? Where are the AI brain controlled robots? Are they ready for prime time? How much do they cost?

I have far more ability/experience programming logic controllers (Ladder logic, structured texts, & function block) than your brain can grasp.

Took me a while to find (for some reason I thought IBM was showing this off, but it's Qualcomm) but this little robot can pick up objects and sort them into bins. When it encounters unknown objects it will ask which bin they should go in. When it encounters objects that are similar to objects it has seen before it places them in the appropriate bin.

Is this going to take over the world tomorrow? No. Is this a good start on say an automated strawberry/orange/tomato picker? Absolutely, it can already navigate to objects, distinguish between different objects (a green strawberry leave alone, but red ones pick), and place them into collection containers.

 

[Londo_Jowo]

Took me a while to find (for some reason I thought IBM was showing this off, but it's Qualcomm) but this little robot can pick up objects and sort them into bins. When it encounters unknown objects it will ask which bin they should go in. When it encounters objects that are similar to objects it has seen before it places them in the appropriate bin.

Is this going to take over the world tomorrow? No. Is this a good start on say an automated strawberry/orange/tomato picker? Absolutely, it can already navigate to objects, distinguish between different objects (a green strawberry leave alone, but red ones pick), and place them into collection containers.

I agree eventually they will be able to build robots that can be self learning and work in environments such as small fruit/vegetable farming providing the terrain is level enough for their use. Same with working in factories sorting parts after manufacture. I look for self loading CNC mills manufacturing components to become common place in factories. I also think they will always be the need for people to program and maintain the robots.