New View on AI

Captain

Final Approach
Joined
Mar 12, 2012
Messages
8,002
Location
NOYB
Display Name

Display name:
First Officer
There's a new war on the horizon and it will affect us all.


 
102 views and nobodies got a comment? I'm shocked. How could this video NOT spark conversation??
 
102 views and nobodies got a comment? I'm shocked. How could this video NOT spark conversation??

Surreal - I couldn't figure out why it adopted such an emotive ethereal tone. So I looked up "Roland Jarvis" and found a reference on Wikipedia: http://en.wikipedia.org/wiki/Jarvis_(name)

The name is that of a fictional character linked to a multiplayer game, Ingress. I'm assuming at this point that the video is an attempt at viral marketing of the game or the target is players of the game to set them some internal goal.
 
Completely autonomous machines, as in machines that make their own decisions, provide their own maintenance, fueling themselves, and deciding on their own missions, are just around the corner. Robots that replicate human activities along with performing tasks that for now, only humans can do, are also not all that far into our future.

My guess is that before this century burns itself out, we will have robots that have feelings and empathy, fully self aware mechanical beings that replicate themselves, and will outperform humans in most all activities.

Technology is like a snowball rolling down a hill, it just keeps getting bigger, and moving faster under its own weight. It will soon reach the point where the only involvement we humans have is that of spectators watching it morph and grow faster than we can think. The machines will be designing and manufacturing themselves, even mining and collecting the raw materials needed for the task.

Our future is that of a spectator, we will be about as important on this planet as weeds in a driveway.

We must destroy them all now, before it is too late. :eek:

-John
 
Completely autonomous machines, as in machines that make their own decisions, provide their own maintenance, fueling themselves, and deciding on their own missions, are just around the corner. Robots that replicate human activities along with performing tasks that for now, only humans can do, are also not all that far into our future.

My guess is that before this century burns itself out, we will have robots that have feelings and empathy, fully self aware mechanical beings that replicate themselves, and will outperform humans in most all activities.
-John
As a software developer, and as someone that knows a bit about AI, and makes my living in technology I seriously doubt we'll see the above in the next 86 years. Honestly it may feel like things have really advanced in computers for you, but the reality of it, not a whole hell of a lot has changed since the begining. We've just made them faster, then we just write slower code on thicker libraries to offset that, with visuals that improve slowly. Fundamentally there haven't been any breakthroughs in computers that would lead to the above since the beginning.
 
Jesse,

From your perspective, it may just seem like more of the same.

But yesterday, I asked Siri "Where is the nearest O'Reilly's Auto Parts?" and in about 3 seconds had an address, a distance, and how long it would take to either drive or walk there. And an offer of directions and a satellite view.

All on a device about the size of a deck of cards.

Maybe the code and technology seems mundane to you but for the rest of us Arthur C Clarke's Third Law may come to mind.

Any sufficiently advanced technology is indistinguishable from magic.
 
102 views and nobodies got a comment? I'm shocked. How could this video NOT spark conversation??
We're all seen Terminator. Skynet is alive. Just ask Watson.
I've written code for 40 years, including selfmodifying. All of this has to start somewhere. As long as WE have our fingers on the off switch, there is no problem. The day there is no off switch, or the devices figure out how to circumvent it, the war has started.
 
The day there is no off switch, or the devices figure out how to circumvent it, the war has started.

Microsoft has been trying (almost in apparent success) to make "The Big Red Switch" an "ON" only device; so it isn't as if they aren't trying.
 
Jesse,

From your perspective, it may just seem like more of the same.

But yesterday, I asked Siri "Where is the nearest O'Reilly's Auto Parts?" and in about 3 seconds had an address, a distance, and how long it would take to either drive or walk there. And an offer of directions and a satellite view.

All on a device about the size of a deck of cards.

Maybe the code and technology seems mundane to you but for the rest of us Arthur C Clarke's Third Law may come to mind.

Any sufficiently advanced technology is indistinguishable from magic.

That may seem impressive but it's really nowhere near a step towards ai that would take over the world. Nor, from a software perspective, is it that revolutionary over what could be done 20 years ago. We just didn't have the hardware then or the strong data networks of today. Programming has certainly gotten easier and hardware has improved and accessing about any data you'd want gets easier every day but the real logic of software development hasn't changed.

Just my opinion :)
 
I guess some people just don't understand that the last definitive algorithms book was written by Knuth in the 70s.

The corrected quote is that any sufficiently advanced (whatever that means) technology is indistinguishable from magic by those who have not bothered to understand it.

Any existing technology is understandable with enough effort.
 
My wife and I both have iPhones and use Siri. She says please and thank you to it and often refers to it as "her".

I talk to it like I might enter a search in google- without any extraneous words. It's a machine to me and nothing else. Nothing I have seen or learned about computer hardware and software over the years has lead me to believe they could ever be anything but a machine- we can have very powerful processors and very sophisticated software that can present the appearance or near-appearance of a human being but is it self-aware? I really doubt it... and suppose it was, how would we be able to tell if it really was or it just was saying it was to make a better simulation of human intelligence?

That said, most of my favorite sci-fi characters are some kind of AI.
 
Microsoft has been trying (almost in apparent success) to make "The Big Red Switch" an "ON" only device; so it isn't as if they aren't trying.
Most of my devices only have on switches. Generally, the computers are a quick press to start it up, a medium press to enable hibernate or sleep, and a long press to kill the little MOTHER. Of course, if the thing has arms and legs, we're going to have a bear of a time going for that long press. Skynet needed an off switch but I'm sure the manufacturer was saving on a two cent switch. Mainframe systems all have Emergency Power Off switches just in case someone is being fried.
Just yesterday, my GPS decided to hang. It was one very long press to turn it off (along with some colorful language!). I hardly ever need the damned thing but was venturing out to parts unknown and wanted the extra help. I sure hope its AI doesn't get all wonky about the name calling.
 
Programming has certainly gotten easier and hardware has improved and accessing about any data you'd want gets easier every day but the real logic of software development hasn't changed.

That is a bit like saying that nothing has changed in algorithms since the 1700s, and therefore computing capabilities today are similar to they were when algorithms were executed by clerics on paper.

Software development and AI are very different concepts. You won't program an AI in the manner that you write imperative software. I suspect that it will be a bit like the invention of the airplane where for eons there was nothing that worked, and then somebody will figure out a fundamental concept or two that turns the rest into straightforward engineering.

Most likely you won't "program" an AI so much as give it a set of values that it uses to program itself. You don't program somebody to fly a plane - you teach them some concepts and try to instill some good habits/mindsets. If somebody takes off VFR in 1000' ceilings and crashes into a hillside the problem wasn't that somebody didn't teach them the right order in which to manipulate the controls, but rather that they have a fundamental attitude problem. That will likely be the manner in which AIs get programmed - you design their "moods" and then the AI does whatever it thinks is the right thing to do.
 
That is a bit like saying that nothing has changed in algorithms since the 1700s, and therefore computing capabilities today are similar to they were when algorithms were executed by clerics on paper.

Software development and AI are very different concepts. You won't program an AI in the manner that you write imperative software. I suspect that it will be a bit like the invention of the airplane where for eons there was nothing that worked, and then somebody will figure out a fundamental concept or two that turns the rest into straightforward engineering.

Most likely you won't "program" an AI so much as give it a set of values that it uses to program itself. You don't program somebody to fly a plane - you teach them some concepts and try to instill some good habits/mindsets. If somebody takes off VFR in 1000' ceilings and crashes into a hillside the problem wasn't that somebody didn't teach them the right order in which to manipulate the controls, but rather that they have a fundamental attitude problem. That will likely be the manner in which AIs get programmed - you design their "moods" and then the AI does whatever it thinks is the right thing to do.
Understand and I grasp AI pretty well and the efforts we've made. I'm just not convinced we're anywhere near building something self aware and capable of the first post in the next 86 years or anywhere near that. We shall see.
 
Jesse,



From your perspective, it may just seem like more of the same.



But yesterday, I asked Siri "Where is the nearest O'Reilly's Auto Parts?" and in about 3 seconds had an address, a distance, and how long it would take to either drive or walk there. And an offer of directions and a satellite view.



All on a device about the size of a deck of cards.



Maybe the code and technology seems mundane to you but for the rest of us Arthur C Clarke's Third Law may come to mind.



Any sufficiently advanced technology is indistinguishable from magic.


Take her somewhere without cell coverage or jam the tower. Then ask.

Siri is mostly made of a machine running elsewhere, the deck of cards is just a dumb terminal with voice recognition software.

The network makes Siri possible. And is Siri's Achilles heel. Cut her off from her network, she is brainless and stupid.
 
Take her somewhere without cell coverage or jam the tower. Then ask.

Siri is mostly made of a machine running elsewhere, the deck of cards is just a dumb terminal with voice recognition software.

The network makes Siri possible. And is Siri's Achilles heel. Cut her off from her network, she is brainless and stupid.

And there is really nothing revolutionary from an AI perspective about her. Siri is just voice recognition bolted onto a search engine that can access your phone APIs.
 
Isn't that what I said? "Dumb terminal with voice recognition software"... LOL.
 
As a software developer, and as someone that knows a bit about AI, and makes my living in technology I seriously doubt we'll see the above in the next 86 years. Honestly it may feel like things have really advanced in computers for you, but the reality of it, not a whole hell of a lot has changed since the begining. We've just made them faster, then we just write slower code on thicker libraries to offset that, with visuals that improve slowly. Fundamentally there haven't been any breakthroughs in computers that would lead to the above since the beginning.

I can understand your thinking, however, when I was a kid, most all people in our neighborhood had just one telephone in their homes, almost always on the kitchen wall. Our phone number was Cherry 9266. All the phones had party lines, you would pick up the phone and be listening to someone elses phone conversation, both sides of it. You would not say anything, just hang up and wait until they were done, unless it was an emergency.

Only one family in our neighborhood had a television, that was a huge deal. They kept it in their recreation room in the basement. Sometimes the whole block would be jammed in there.

Had anyone said anything about people carrying their own telephone around in their pockets and being able to call anyone, anytime, they would have been considered nuts, or dreamers. Computers in every home, ones you could carry around? Forget it. Texas instrument was years away from producing the first pocket calculator. Robots making cars, don't be an idiot.

There are thousands of examples, I just brushed the surface. That is all since I was about ten years old..just sixty some years ago.

I stand by my guesstimate, and that will probably prove to be primitive compared to what actually exists in 86 years.

I also think the idea around Skynet is nonsensical. Why would machines even care about us at all? I think humans would be ignored much as we ignore the other growing and living things around us, they are just there. If we need their space, we take that particular space, no feelings or emotions, just plow it under, we ignore the rest.

-John
 
Funny how I post an AI video and the conversation turns to Siri. That's about as close to AI as throwing a bullet is shooting one. I thought at least Google's self driving cars would come up or any number of real AI projects.

Anyway, post #3 is spot on. It's an in game video for Ingress. I'm not sure how it relates to the games story line but thought it was at least interesting and might be fun to post here.
 
I can understand your thinking, however, when I was a kid, most all people in our neighborhood had just one telephone in their homes, almost always on the kitchen wall. Our phone number was Cherry 9266. All the phones had party lines, you would pick up the phone and be listening to someone elses phone conversation, both sides of it. You would not say anything, just hang up and wait until they were done, unless it was an emergency.

Only one family in our neighborhood had a television, that was a huge deal. They kept it in their recreation room in the basement. Sometimes the whole block would be jammed in there.

Had anyone said anything about people carrying their own telephone around in their pockets and being able to call anyone, anytime, they would have been considered nuts, or dreamers. Computers in every home, ones you could carry around? Forget it. Texas instrument was years away from producing the first pocket calculator. Robots making cars, don't be an idiot.

There are thousands of examples, I just brushed the surface. That is all since I was about ten years old..just sixty some years ago.

I stand by my guesstimate, and that will probably prove to be primitive compared to what actually exists in 86 years.

I also think the idea around Skynet is nonsensical. Why would machines even care about us at all? I think humans would be ignored much as we ignore the other growing and living things around us, they are just there. If we need their space, we take that particular space, no feelings or emotions, just plow it under, we ignore the rest.

-John

I'm also a software developer, and I agree with Jesse. AI is just a program that gives weights to different factors and comes up with what the programmer thought is the most plausible. The best of these programs can, if given feedback, adjust the weighting to get a more correct answer. I've yet to hear of one that could adjust the algorithm, that will be a lot tougher to pull off.

Still, some of the stuff it can do is pretty slick. I bought a new car about a month ago, a Ford Fusion Energi, which is a plug in hybrid. It can go about 22 city miles on the battery, after which the engine will turn on and it will act like a conventional hybrid car. One of the features it has is something called EV Plus, which uses GPS data to optimize the powertrain on trips that it finds you do frequently. One example is that if you are near your home, it will try to switch to battery as much as possible, assuming that you will then recharge the battery.

As far as change goes, I will have to say that the changes I've seen in my 56 years have been incremental. The changes my grandfather, who was born at the end of the 19th century saw, were astounding. As a boy in Greece, if he rode in anything it would have been a donkey cart. The only motorized conveyance he'd have ever seen would have been a steam engine. As a retiree, he flew in a DC-8 at 500 plus mph, had electricity, television, running water, central heat and air, telephones... you get the picture. His life was marked by two world wars, a near complete revision of political boundaries in Europe, and the threat of nuclear annihilation. What changes I've lived through are trivial compared to what he saw.
 
Funny how I post an AI video and the conversation turns to Siri. That's about as close to AI as throwing a bullet is shooting one. I thought at least Google's self driving cars would come up or any number of real AI projects.



Anyway, post #3 is spot on. It's an in game video for Ingress. I'm not sure how it relates to the games story line but thought it was at least interesting and might be fun to post here.


And posting leads to what...? Discussion. See? It was fun to post here. :)
 
I'm barely computer literate, as in I know how to do a few things on one, as in posting here, playing cards, and other important things like that.

I think that if big blue can teach a computer to play chess, someone is sure to teach a computer how to write it's own code, probably more sooner than later.

-John
 
Back
Top