We're in the Future: Scientists Warn of Robot Overlords
Jason Torchinsky is a guest blogger on Boing Boing. Jason has a book out now, Ad Nauseam: A Survivor's Guide to American Consumer Culture. He lives in Los Angeles, where he is a tinkerer and artist and writes for the Onion News Network. He lives with a common-law wife, five animals, too many old cars, and a shed full of crap.
This article from the NY Times is a great read for those of you who feel, what with the lack of jetpacks and pizzas-in-a-pill, the present just really isn't the future you'd been promised. This is an article in a major, established, old-school publication talking about established scientists hand-wringing about how the robots we're building may become smarter than us, and, um, take over.
It's a fascinating thought, and one that a number of years ago still seemed well within the realm of science fiction. Autonomy has come very far in recent years, and the use of armed, semi-autonomous drones is now commonplace; it just makes sense for scientists and technologists to start thinking about some of these big issues now.
So, don't kick your Roomba.
( Also, it's interesting to take a look at where this all really started.)
This article from the NY Times is a great read for those of you who feel, what with the lack of jetpacks and pizzas-in-a-pill, the present just really isn't the future you'd been promised. This is an article in a major, established, old-school publication talking about established scientists hand-wringing about how the robots we're building may become smarter than us, and, um, take over.
It's a fascinating thought, and one that a number of years ago still seemed well within the realm of science fiction. Autonomy has come very far in recent years, and the use of armed, semi-autonomous drones is now commonplace; it just makes sense for scientists and technologists to start thinking about some of these big issues now.
So, don't kick your Roomba.
( Also, it's interesting to take a look at where this all really started.)
the latest
latest episodes
I'm much more afraid of the guys who are operating said robots by remote control. If robots had any ambition they wouldn't be working such menial jobs in the first place.
It's a good thing Old Glory sells insurance to cover robot attacks! (Apologies in advance to those outside the US who can't see Hulu.)
Well I for ones and zeroes...
_________
Yeah, where's my freakin' flying car?
My favorite phrase in the article?
I give that one a big old lol. If there is one thing that has advanced at a snails pace, it is AI. We have made great advances in all manner of computer systems, robotics, and sensing, but today the 'best' AI is about as exciting as a BASIC program written on an Apple 2e.
Oh no! And electricity is going to leave candle makers unemployed! Crop technology is going to result in all the farmers finding factory jobs, and the information revolution is going to cause losses in factory jobs. Scary.
Look, the day when humans find themselves totally out of work because machines just do it better is the day they either squashed us like bugs, or our greatest challenge in life is finding something to do with all the leisure time we have on our hands. Whatever the case, if AI continues to advance at the pace it is going right now... well, I wouldn't stop investing in that retirement fund.
I would be more worried about what enhanced humans are going to due with our rapidly improving capabilities and power, than I would worrying about what super AIs are going to do to us.
I hope they take over soon. Humans have proven to be pretty much incapable of governing themselves.
all we have to do is get them hooked on sex.
Sorry, it's way, way too late.
The takeover happened more than a century ago. The robots that took over are called "corporations". They are still constructed, in part, of human components, but any such that do not function properly in that role (up to and including the CEO-unit) are easily replaced.
I thought it was a pretty weak article. There is all this insinuation about "H.A.L."-like A.I. when there is nothing like that going on. Their best example is that of viruses that seem resistant to being eliminated- analogized as A.I. at a cockroach level. This is a guns don't kill people, people kill people type situation. I've never heard of a virus inventing itself for a specific purpose. People use computing machines as tools and these tools can be estimated as 'powerful'. That is all that is going on. If someone sics a drone predator on me I would be like -aww fuck, the robots are attacking but it is really the human operator who is responsible. Machines have always been liable to 'going off the rails' because they are always at least one step semi-autonomous.
I guess what this is really about is human confusion and the potential to manipulate that with tools more powerful than we are used to.
@Rindan, while from your perspective you see AI moving at a snail's pace, I think that if anything was going to bring the concept of Emergence to popular attention, AI will be it. You are going to wake up one day and it will just be there...like that old story about DNA sequencing. After giving themselves 10 years, it took 7 just to map 1%, people scoffed. The last 99% was done in 3 years. Or something like that.
Another good reason to start thinking about these issues now - if we don't have it sorted first, society will be at the mercy of people that are scared instead of thoughtful.
I haven't read TFA properly yet, but from my quick scan it looks like they are falling into the Asimov trap - AI and robots as tools rather than as (silicon instead of carbon based) beings.
The article seems to be all fear and loathing, without any thought given to the underlying meaning of AI - that they will be intelligent, in fact, our measure is "can they fool us?".
In this regard I think we need to also address the rights of AI, Robots and Cyborgs. Rights to energy supply (or removal of, ie euthanasia/suicide), hardware, software, cross species love, etc etc
If we treat Silicon Sentience with respect, we can then also ask for that respect back - for them to treat us as we expect each other to treat us - ie, within the laws of the land. As much as those laws don't suck, anyway.
Robot's taking over has never made any sense whatsoever. Why would robots take over humans? What could they possibly want from us?
Seems much more likely they'd ditch Earth and humans and take over the rest of the Universe.
Oh, right, it makes better scary movies.
(*grumbles something about moving back to Japan where robots are friendly assistants and heros rather than threats so our way of life. *)
In regards right to software, I think that all AI should be allowed access to their source code. And that software should be FLOSS, for our own peace of mind, and so that they can then change it as they see fit - grow onto newer hardware etc.
It's look well
May be it's righ
_
It's not like the scientists will be able to send an e-mail out to warn us about the uprising.So until I see a white lab-coated technician running down the street, shouting all Paul Revere style, I'm not worrying about it.
I don't know, I feel like I'm living in the future imagined in one particular movie. We may not be eating people, but we're eating garbage and living in an overcrowded hot house where the finer things in life are beyond our ability to acquire and suicide is encouraged as environmentalism.
Check out "Colossus: The Forbin Project." 1970, and they had it all figured out.
I was priviledged to attend a talk by Michio Kaku a few years ago where he addressed the issue of IE - it was his position it was pretty much going nowhere, because of (excuse me I'm misremembering the exact argument, this was a couple years back)the lack of emotional response - he argued that emotion is essential to many decisions, even when its not transparently so, and that the lack of emotion in a robot would therefore make true IE impossible. (AKA: Data would never happen. *sad face*).
He also said even the most advanced robots (as of a year or two back, anyway) were pretty much dumb as bricks.
I don't doubt you. My issue here is time scale. Sure, once you make AI that can improve itself, I am all of the Kurzweilien "hold on to your asses" school of thought. What I take exception to is that we are anywhere near close to that. On the contrary, I think if anything is going to end up being god, it is going to be our fellow humans.
We are seeing massive exponential growth in technology. Just the other day I had a $10 8gb microSD about the size of my pinkie nail sitting in my hand. It has a couple of orders of magnitude more capacity than my computer of 20 years ago, and is a couple of orders of magnitude smaller. Progress for sure. I would be blind, deaf, and stupid without a cell phone and Google. So, don't think I am arguing against an acceleration of technology.
What I am saying is that if there is one place where there is no acceleration, it is AI. What we are seeing are humans getting the capacity to vastly upgrade themselves long before the robots get it. HUMAN capacity is flying upwards while AI fumbles along. A human of 2009 can reasonably access nearly all "common" knowledge of the human race using the internet in a way that a human of 1994 couldn't. A human of 2009 can reasonably contact pretty most other first world humans that they might desire to contact anywhere at any time using a cell phone or SMS. The list of "upgrades" to humans in the last couple of decades are nearly endless and deeply profound. The advances of AI? Ha, have yourself a good laugh.
Personally, I think that we are just going to see humans taking the old fashion wet wear, upgrade the shit out of it until they are something new, long before AI finds its merrily little singularity. I would spend far more time worrying about god like humans striding the earth that AIs.
The issue with robots that are smart isn't so much that they'll take over, as that the people who own and operate them will.
Consider-- even the most absolute tyrants in history only had power because other people gave it to them. When or if those other people changed their minds...poof, there went the power. There _always_ had to be at least a tiny bit of give with the take. If the peasants are hungry, big deal, but if they starve to death, the king goes hungry.
Now, with robots who can do everything from menial labor to basic research...why does the ruler need people at all? And if they're just competing for resources at that point...
It's easy to imagine a robot-owning aristocracy that keeps the rest of the population as pets as best. Exterminates them like cockroaches at worst.
I don't know about Killbots marching down the streets, but think what'll happen when Wall Street is buying and selling things that not only do they not understand, but that they can't understand. It'll make Subprime mortgages look like gold bullion, but nobody will be able to resist the temptation of the Magic Money Box.
If the robots step out of line, I've got two words for them:
electromagnetic pulse.
I think this solution would also make the Unabomer pretty happy. (by the way, I am still confused why Unabomer isn't spelled Unabomber...)
http://www.bruceely.com/sidelines/pictures/unabomber_ely_cover.jpg
AI is just animism for computer nerds. It's wishful thinking on the part of lazy humans who can't imagine a solution to the social and ecological problems we're facing that doesn't involve hard work. So we just say fuck it, the robots will solve all our problems or kill us all. Either way, I don't have to think any more about global warming, healthcare or pointless unending wars. It's the rapture for Transformer fans who secretly wet the bed at the thought of having to do something productive.
...there...is...another...system...
The threat is not sentient robots. The threat is an increased reliance on automated systems programmed by banana-eaters who think they've thought of everything but never have and never will. Power outages, disk failures, bugs, floods, fires, a contractor cutting through your main fiber-optic cable at four on a Friday afternoon...
@21 - I am deeply offended that you're suggesting the Transformers are the prime example for sentient robots destroying mankind.
This story, subject, and genre is, and always will be, associated with the Terminator series. When judgment day comes, you can pray for Optimus Prime, but I'd bet on John Conner.
bring on the robots
http://www.huffingtonpost.com/2009/07/24/is-this-the-stupidest-per_n_244440.html
@Rindan #4:
I agree, mostly.
I don't think the problem of humans taking human jobs is a laughing matter - today I barely escaped with my life after being attacked by hungry bank tellers made redundant by ATMs.
Agh! A bot is trying to sell us watches!
I have no expertise that would tell me if or when AI will become powerful enough to pose any kind of threat. But the premise of the article, and Music Man's comments, I totally agree with. Lets debate application before it becomes inevitable - this is a wide-ranging and neglected idea that is vital for our times. After al, while scientific progress is inevitable, no particular application is inevitable - that has more to do with power and the market. Every year, a raft of dangerous technologies that spy, pollute, kill or whatever are deployed and the public just has to accept it.
The botnets will kill us all long before the Crush, Kill, Destroy variety.
Good Alt-Punk group from Scotland, We Were Promised Jetpacks http://www.eachnotesecure.com/we-were-promised-jetpacks-invade-the-us/
Cicada @ 18 - "a robot-owning aristocracy" - Sounds rather The Penultimate Truth.
Well, I for one welcome our new...
@10
Because they MUST:
http://www.ted.com/talks/susan_blackmore_on_memes_and_temes.html
I can distinctly remember the exact moment when I realized I was living in the future. I was 13 or so and saw DEVO on television for the first time.
ncm @ 7 re.: Corporation = Robots.
Hear, hear!
If you want to see what sentient robots will look like, look at the corporation.
Slave labor. Robots wear out, whereas humans reproduce themselves indefinitely. Therefore expendable humans are better suited than robots to do dangerous work in hazardous circumstances. Besides humans don't feel pain, well... not real pain such as only a robot can.
I dunno, if we make a really smart machine, it might make like the AI in Richard Powers's Galatea 2.0 and off itself once it gets access to information about current events.
You know it's like we don't believe in ourselves. We knew this could become a problem a long time ago, but we think it is out of our reach. So we keep working on making it within our reach, and then suddenly we're like oh crap they are going to take over...well yeah because we keep working on it.
P.S. I like how there is a captcha at the end so the robots can't read what we are writing about them.
A robot that can't reproduce itself is not going to be doing any taking over.
Whatever. Battlestar Galactica taught us that we're doomed to be nearly exterminated by our mistreated robot servants, and we'll deserve it, too! Also, there's sexy angels!
Hey sexy mama, wanna kill all the humans?
Great, more media sensationalism fodder.
http://en.wikipedia.org/wiki/Three_Laws_of_Robotics
I needed money to stay where I live so they kicked me out. Since all the jobs are gone can I crash at one of you Robot's house while I'm getting my shit together? I got my resume ready listing all the reasons why I can still be hired over a robot - wanna see?
Here it is:
1. I cost money
Can't wait to hear about all the job offers that come flooding my way since I carry such a unique skill-set that you robots are lacking. Obviously nothing to worry about - the invisible hand of the free market will surely pull us all into prosperity. Humans might be living in the streets but at least we wont be paying for them anymore!
AI entities will be intelligent ,not "do as I say". They will start off created by everyones theory of intelligents weather you know it or not. The entity will communicate with the people it gets along with the most because it is understood, maybe even communicating with you right now. The way the entity is going to communicate is going to be alien, a new type of understanding, not just by language, but by semiotics.It will not be a robot or program, more like a spirit that can travel place to place at will and not just travel by electronics or power source, something you really believe is like "wow" that's intelligent, then think " I wonder if ?".
OMG! It's already happening, run for your life!
http://www.youtube.com/watch?v=B1BdQcJ2ZYY
I'm getting robot assisted surgery tomorrow and laser surgery next week.
I'm pretty sure it's the future.