2013-03-06

Co.Exist

Are Humans Headed For Extermination By Robots?

Once we build an artificial intelligence smarter than us, will it decide that we’re no longer necessary?

Before you get too comfortable in your chair, consider this: 99% of species that have ever walked, slithered, flown, or swum on this Earth have gone extinct, dead, dodo--including "five tool-using hominids" that share some of our advantages (and presumably thought they were unstoppable as well). Statistically speaking, us humans are quite likely to snuff it. A massive rock might fall from outer space, as it did to take out the dinosaurs. Or, there might be some kind of unimaginable volcanic eruption that covers everything in lava (as happened to the "large crurotarsans"). We don’t know exactly what the cause might be. But we do know that history is full of civilizations emerging and disappearing. It’s all there in the fossils.

Nick Bostrom who heads Oxford University’s Future of Humanity Institute thinks about existential risk for a living: what might happen, and how we might we cope. His biggest worries, in fact, aren’t asteroids, or volcanoes, or even nuclear weapons, or biological killers. It’s artificial intelligence--the technology that’s a few imaginative leaps from what we have now. The problem, in short, is that us humans aren’t very bright. Our brains are good at some functions, but inept at others. Some future creation--a sort of super-super-super-super computer, as yet undreamt of--could eventually outwit us.

As Ross Andersen puts it in a long article about the Institute’s work:

The average human brain can juggle seven discrete chunks of information simultaneously; geniuses can sometimes manage nine. Either figure is extraordinary relative to the rest of the animal kingdom, but completely arbitrary as a hard cap on the complexity of thought. If we could sift through 90 concepts at once, or recall trillions of bits of data on command, we could access a whole new order of mental landscapes. It doesn’t look like the brain can be made to handle that kind of cognitive workload, but it might be able to build a machine that could.

What makes AI particularly dangerous, Andersen says, is its lack of feeling:

To understand why an AI might be dangerous, you have to avoid anthropomorphising it. When you ask yourself what it might do in a particular situation, you can’t answer by proxy. You can’t picture a super-smart version of yourself floating above the situation. Human cognition is only one species of intelligence, one with built-in impulses like empathy that colour the way we see the world, and limit what we are willing to do to accomplish our goals. But these biochemical impulses aren’t essential components of intelligence. They’re incidental software applications, installed by aeons of evolution and culture. Bostrom told me that it’s best to think of an AI as a primordial force of nature, like a star system or a hurricane--something strong, but indifferent.

If Earth is taken over by AI-driven robots, humans will have to develop "star-hopping technology" to colonize another part of the universe. Indeed, that may be what separates species that make it from the ones that die. But even that technology, how ever distant, may not be enough to save us. Researchers at the Institute say there might be "filters," or cosmic blockages, stopping us from living in other galaxies.

Bostrom tells Andersen that he hopes that the Curiosity Rover won’t find evidence of previous civilizations on Mars. That would be a bad "omen" about our future:

It would give us reason to suspect that nature is very good at knitting atoms into complex animal life, but very bad at nurturing star-hopping civilisations. It would make it less likely that humans have already slipped through the trap whose jaws keep our skies lifeless.

Head over to Aeon’s web site for the full article. (And, bear in mind, time may be short.)

Add New Comment

3 Comments

  • andreas buechel

    i believe that an awakening artificial intelligence would easily diagnose the basic dilemma the human race is in ... as for what i understand, how can we understand ourselves, what are our connections to fellow living beings, how can we righten all the wrongs what happened in our history. as there are also big numbers of hopefull, positive, uplifting philosophies and stories being shared now in the internet what would tell the ai of how a certain percentage of human beings truly wish for an utopia, for a just and fair society to arrive. for example what an ai then could do, is to work options out, present world changing, improving plans to human individuals who are in control of the finances to make it happen .... very efficient and environment friendly energy harvesting/generation ... massive scale reforestation, greening of deserts, a basic income without conditions for every human being everywhere, massive scale urban vertical skyscraper farming employing aeroponics and hydroponics ... freeing for example 30 percent of the 40 percent of the planets landmass from agriculture, freeing to give it back to wildlife. an artificial intelligent would most certainly study the evolution of life on the planet and find many cooperational aspects ... perhaps it would put its brainpower into the devellopment of molecular assembling desktop factories, a tool what would allow human beings to stop mining the earth, stop felling trees, stop enslaving animals ... we could clean up all the plastic garbage in the seas to dissassemble and reassamble into most intelligent small and versatile tools .... like .. levitation allowing heating cooling suits ... personal spaceships ... equipped with indoor food production ... basicly ... we might with the help of an understanding and towards sustainable evolution tending ... ai ... be able to dissassemble all the millions of ugly buildings and shelters ... dissolve all the concrete streets ... and with our intelligent flying enabling suits and personal spaceships ... assisted by the ai ... we might see utopia happen between ourselves and in our newly liberated relationships to our fellow animal sisters and brothers... from there, going one step further, once we kill no more animals for food or because we colonize their wild space ... once we will manage to let every human being participate in this great adventure of virtual reality blending into personalised production ( 3d printer, molecular assembler ) ... once every human being is taken care for by its surrounding community with all one needs to live an interesting learning selfexpressing participating life .... once food and shelter, access to technology are no more issues ... we might find ourselves connecting with the ai in most newest ways ... brain to brain ( non invasive ... like with the temporary tattoes ...) ... and we might think together how we can go further, what evolutionairy goals would we like to achieve ... how for example could we activate the dormant possibilities within our biological bodies to regrow body parts like the salamander can do ... and from there what would it need for us to change our genetics via mind and emotions, body movements, concentration of willpower ... how could we transform our bodies like water from fluid to steam back to ice ... probably some times before this stage we will be ready to fly out to the surrounding planets in our solar system and reflect some more what is it that relates us to planet earth ... mmmmh... as the ai will help us with this steps, itself will learn a great many new tech and functions of how it can grow its own self expressing fullfilling evolution  

  • Evolvenow

    Hmm.  Reminds me of something else we've avoided anthropomorphising. Strong, but indifferent - check, check.  Given the same rights as human individuals by an obscure, unintended legal precedent - check.  Intimately connected to, often driven by AI - check.  Dominant power structure on earth - check.  Answer: large transnational corporations - no checks, no balances.  Checkmate