Artificial Super Intelligence - are we ready for it?

[broncofan]

Right now no, computers are useful but if the computer that barks at me 'Unexpected item in bagging area' when there is no such item and won't let me leave the store...no, it doesn't give me enough confidence. The solution is in cockpit design, real-time communications between the aeroplane and the nearest air traffic control -real time video streaming?-and safeguards to protect the passengers from loony crews, but is any system of any kind 100% safe?

I agree. I used to have a bad phobia of flying. When I looked up the statistics years ago, among most U.S airlines, the chances of a crash were 1 in several million (maybe 5 million or more). There are specific functions that computers perform very efficiently and perhaps a fail safe mechanism that prevents a nosedive would be useful, but any change to existing protocols could in theory cause a net decrease in safety.

I have an automobile that issues warnings every time I do something it thinks is unsafe. For instance, if the car registers the tire pressure as low, I am not allowed to shift it into a faster gear. However, occasionally it registers low tire pressure even though when I go to the mechanic, the tire pressure is normal. So it is the electronic system (the computer) and not the tire pressure that needs to be reset. I would be concerned about the tradeoff of any attempt to inhibit pilots from being able to perform functions that in their judgment are necessary. Of course, a fail safe that identifies and prevents clearly unsafe maneuvers would be useful, but what are the odds that it registers something as unsafe that could under a rare set of circumstances be necessary?

After all, pilots intentionally crashing commercial airliners is very rare. We've seen 3 or so of them in how many tens of millions and yet we are considering fail safe mechanisms that could have tradeoffs...even the protocol that enabled the crash was intended to prevent another unlikely event....unauthorized people entering the cockpit!

[Stavros]

Robots learn to cook by watching YouTube videos , sponsored by DARPA ! I guess the army needs more cooks?
I posted this earlier in another thread
http://cmns.umd.edu/news-events/features/2708
http://rt.com/news/219687-robot-learns-watching-video/
http://www.homecrux.com/2015/01/22/2...be-videos.html

This phrase leap out at me from your first link:
“We are trying to create a technology so that robots eventually can interact with humans,” said Cornelia Fermuller, an associate research scientist at UMIACS. “So they need to understand what humans are doing. For that, we need tools so that the robots can pick up a human’s actions and track them in real time."
-I don't think of my computer as a robot, indeed in my spite of my age, or because of it, I am forever grateful to see the end of typewriters and tippex, but I am not sure I want to interact with a robot. And I don't want a robot watching my every move. I have seen Haley Joel Osment in AI, and it is wrong, very wrong.

[Odelay]

After all, pilots intentionally crashing commercial airliners is very rare. We've seen 3 or so of them in how many tens of millions and yet we are considering fail safe mechanisms that could have tradeoffs...even the protocol that enabled the crash was intended to prevent another unlikely event....unauthorized people entering the cockpit!

It's funny you should say this because over the last 24 hours or so I was reflecting on my snap opinion and reversed it based on the very point you make. And really, having backup human controls in some remote location or at air traffic towers isn't the answer either. There's enough redundancy, especially with at least 2 pilots in every cockpit - even the small 50 seat jets - that anything more is just overkill. Two pilots, plus the sophisticated computer controls (which I'm sure is far better than what we're seeing in automobiles), is plenty to avoid 99.99% of shenanigans or just plain accidents.

[Stavros]

This ad aired on the tv the other night, I was taken aback and although I now understand what it's for, I find it an eerie mix of the desirable and the disturbing...tempted to buy one -if I live long enough for it to happen!

[Stavros]

Along with 'Humans' which airs in two weeks, I saw Ex Machina the other day, and then read in today's papers this report of a speech by the Astronomer Royal, Martin Rees on the extent to which the future is mechanical:
“There has been just a thin sliver of time when organic beings have existed and billions of years after machines will take over, so they will be the future.”

The link is here:
http://www.telegraph.co.uk/news/scie...-machines.html

[buttslinger]

Artificial intelligence won't happen, because real computers have no motive to walk down a crowed street in New York and people watch, they are complete as they are.
Using jumbo computers to fashion genetic engineering, that's different, you could quite possibly program a certain super computer to weed out all the bad DNA crap, and fashion things the humans like: intelligence, looks, health, longevity.
Evolution hasn't changed the alligator of cockroach over millions of years, but in one or two hundred years we could be popping out little master kinders beyond our local imagination. At the very least knock out some diseases or basic genetic flaws.
Better than a computer that can cook your breakfast or fly your plane is a computer that can make your children living breathing Gods. Cue the Twilight Zone music, humanity is flawed. catastrophe awaits.

[sukumvit boy]

Here's some recent news.
http://www.cnet.com/uk/news/musk-bac...ence-research/



Mr. Musk at it again with his , what I find eccentric , campaign about the dangers of AI.
But hey , I grant the guy his quirks. Remembering that Newton ended his life immersed in Alchemy , Alfred Russel Wallace in spiritualism , and Einstein abhorred quantum mechanics.

[Stavros]

I watched the first episode of Humans, but was not impressed by another version of robots in rebellion, or the heroic individuals fighting the machine...and that's just a tv programme.

What puzzles me about a lot of this AI material is the inability of people to decide if we humans have a soul even when they talk about robots developing initiatives on their own. Humans may be 'machines' in the sense that most of us have a body and internal organs, and need water, and air and food to live, but many scientists cannot explain why one person has creative skills that another does not, or, crucially, why we are all different if our bodies are all the same. Surely the limitation of AI is that it will only have as its historical memory whatever is programmed into it, and that it will not be capable of writing poetry unless it is an imitation of someone else's?

[trish]

Given the vast chaotic complexity of the world it seems perfectly reasonable to me that two identical artificial intelligences placed in distinct but similar environments might develop entirely different behavior patterns which ultimately cannot be explained in any satisfactory detail. One may write completely original poetry in its very own inexplicable style. The other may develop an obsession for money and power. Once the complexity of a dynamic system passes a certain threshold, it’s behaviors become effectively incalculable. At that point it’s useless to attempt to understand it on the level of switches and circuits. It is more readily understood on the more abstract level of it’s patterns of behaviors. Intentions, goals, and souls are higher level abstractions that clearly apply to the behaviors and personalities of the machines we call people. The question is, “Will it ever become appropriate to seriously apply these concepts to other machines?”

For me the worrisome part of AI is the possibility that some machines will have “souls” in the sense that we do; i.e. we feel, we love, we experience the world and are driven to create art, music, poetry that reflects our inner selves in reaction to those experiences. We are also machines. I see no reason other sorts of machines might not also experience the world in similar ways. The moral danger posed by AI is two-fold: 1) there is the possibility that we may refuse to extend our empathy to machines that deserve it; and 2) there is the possibility that we may grant personhood to machines that are not persons but simply passable simulations. (Btw in other circumstances I would find the phrase “passable simulation” somewhat toxic.)

[buttslinger]

If you considered DNA as computer code, then it is conceivable that one day they could construct a computer clone of a human, or a dog, or a tree. You could even network in some raven DNA or dolphin DNA. You could even un-program disease and even death, in a laboratory setting.
I think even a blade of grass has self awareness, it has a desire to live and procreate, it will alter it's own DNA to adapt to it's environment. It has no need to talk to me. I think it feels things, just totally different than humanoids.
95% of yoga is breathing, and the motive is the end of suffering, not super intelligence, so somewhere along the line the stress of constructing a smart toaster might make us re-examine our motives, why would a real life Cmdr Data give a shit about us? What if GORT grows some balls and starts blasting people for kicks?
If a SUPER computer is allowed to reprogram itself, ,,,hmm, I don't think any scientist alive now or in the future would allow that, without a kill switch real close by.
If you gave a computer lungs of some sort, I guess you could manufacture a soul, and since any computer made now can beat me in chess EVERY TIME, in that sense we already have intelligent computers.
One Day in the far future I suppose you could build a huge computer in Lucerne that would be the ultimate ORACLE for mankind, curing cancer, running the IRS, exploring our history and all the galaxies, but to the computer that stuff would be like paying the rent, answering the mysteries of the universe in exchange for electricity and lubricating oil. Housework. Who knows what motives an all-knowing entity would have?