Claim: Climate change might cause robots to take over

robot_twit

Guest essay by Eric Worrall

Start with interesting scientific paper which explores the dynamics of mass extinction, and weave it into a climate horror story full of scary robots, climate catastrophe and the end of mankind.

According to the Washington Post;

The strange link between global climate change and the rise of the robots

We’ve already heard of all the nasty consequences that could occur if the pace of global climate change doesn’t abate by the year 2050 — we could see wars over water, massive food scarcity, and the extinction of once populous species. Now add to the mix a potentially new wrinkle on the abrupt and irreversible changes – superintelligent robots would be just about ready to take over from humanity in the event of any mass extinction event impacting the planet.

In fact, according to a mind-blowing research paper published in mid-August by computer science researchers Joel Lehman and Risto Miikkulainen, robots would quickly evolve in the event of any mass extinction (defined as the loss of at least 75 percent of the species on the planet), something that’s already happened five times before in the past.

In a survival of the fittest contest in which humans and robots start at zero (which is what we’re really talking about with a mass extinction event), robots would win every time. That’s because humans evolve linearly, while superintelligent robots would evolve exponentially. Simple math.

Read more: http://www.washingtonpost.com/news/innovations/wp/2015/09/08/the-strange-link-between-global-climate-change-and-the-rise-of-the-robots/

As the Washington Post admits, “mind blowing paper” does not mention climate change or global warming, and is not even really about robots. The paper is a fascinating attempt to use evolutionary computer models, based on the NEAT system developed by my favourite AI researcher Ken Stanley, to explore what happens when a “mass extinction event” abruptly empties a lot of ecological niches. The conclusion, unsurprisingly, is evolution goes into overdrive – the empty ecological niches are rapidly filled by new species.

The abstract of the paper;

Extinction events impact the trajectory of biological evolution significantly. They are often viewed as upheavals to the evolutionary process. In contrast, this paper supports the hypothesis that although they are unpredictably destructive, extinction events may in the long term accelerate evolution by increasing evolvability. In particular, if extinction events extinguish indiscriminately many ways of life, indirectly they may select for the ability to expand rapidly through vacated niches. Lineages with such an ability are more likely to persist through multiple extinctions. Lending computational support for this hypothesis, this paper shows how increased evolvability will result from simulated extinction events in two computational models of evolved behavior. The conclusion is that although they are destructive in the short term, extinction events may make evolution more prolific in the long term.

Read more: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0132886

The Washington Post article is an interesting read, but in a sense it misses its target. The article tries to weave climate fear into the rise of the robots narrative, but in my opinion ends up just being a robot story. Unconstrained artificial intelligence is scary, in a way warm weather can never be. I believe, as Dr. Stephen Hawking once warned, that an artificial intelligence disaster really could cause the extinction of mankind. Robots don’t have our sense of right and wrong. If you told a human level robot intelligence to maximise shareholder profits, you would have to be very careful to remember, to instruct the robot about what it couldn’t do, about limits to behaviour which most humans take for granted. For example, the corporate profit robot would have to be explicitly told, that assassinating surplus employees is not an acceptable way to minimise employee contract termination and redundancy payments. Sooner or later someone whose job is to instruct the robots, will forget to tell a robot something important.

I suspect anyone who reads to the end of the Washington Post article, or this post, is thinking far more about robots, than about climate change.

Advertisements

103 thoughts on “Claim: Climate change might cause robots to take over

  1. The robot in the picture has an expression which suggests it is thinking: “Agh! Fans of Al Gore. Crush! Kill! Destroy!”

  2. Having suffered the slings and arrows of outrageously trying to imbue ferromagnetic and then silicon life forms with intelligence, and in the latter case I’m not talking about the wife’s twin birthday enhancements, the poor thingies are a long way off ruling anything.
    Having crashed and burned on procedure rules, neural networks and finally generationally selective genetic algorithms, I wouldn’t trust one of them to navigate across a road safely.
    Would anyone like to write the schematic logic for that simple problem? I’ll have fun pointing out the flaws but knock yourself out anyway.
    Pointman

    • Sod it, I’ll kick off the programming disaster. I’ll wing the syntax, and if you think that’s something the vice squad should be taking a serious interest in, just sit back and enjoy the fun.
      1. Advance to edge of pavement (that definition bug is easy, so no brownies for spotting it).
      5. Do until loop01 (It’s safe to cross) {
      10. Look left.
      20. If you can see a car, generate reasonable random wait and go back to start of loop01, otherwise proceed to stmt 30.
      30. Look right.
      40. If you can see a car, generate reasonable random wait and go back to start of loop01, otherwise by default proceed to stmt 50.
      50. Break out of loop01
      60. } end of do until loop01
      70. Cross the road.
      80. Yay, I’m a happy robot to make it across the road alive.
      90. https://www.youtube.com/watch?v=IvnHgKkNKR4
      It doesn’t look good without indents but WTF. Gowon, spot a few bugs/drop offs.
      Pointman
      BTW. recode the whole thing with your bug fix included.

    • kim
      September 10, 2015 at 10:24 am.
      Are you (or were you) the One or the Other, or the “chicken he” perhaps? 🙂
      cheers

    • Tell that to Google. They seem to think their robot car can navigate along the road very safely. They seem to have sold our politicos on it too.

    • “I wouldn’t trust one of them to navigate across a road safely. ”
      Let alone drive me to the shops. Yes, Google, I do mean you.

  3. Talk about insanity !!! The liberal left has lost all logic and reasoning in their pathetic attempt to control Humanity !! Thank God for the voice of reason from WTF , err , I mean WUWT !! LOL…

  4. I remember hearing in the late sixties; “today’s science fiction is tomorrow’s science”. It seems to me that yesterday’s science fiction is still science fiction but people have come to believe in it and wish it true.
    Am I way off thinking that TV sci-fi is the most science education the average green believer can claim?

    • Dawtgtomis September 10, 2015 at 9:14 am

      I remember hearing in the late sixties; “today’s science fiction is tomorrow’s science”.

      There are a lot of us still waiting for those flying cars in our garages, dawtgtomis. Sometimes, tomorrow never comes.

      • “There are a lot of us still waiting for those flying cars …”
        I’m happy driving against idiot with only 2 degrees of freedom. Imagine the mess they would cause trying to text with Z axis ability. Consider us lucky we’re still waiting.

      • @Paul
        The Jetsons never had any traffic problems with flying cars. Oh wait… that was a cartoon. Seemed real enough to me back then, though.
        .
        .
        @Brian
        “I’m still waiting for those silver unitards.”
        Dang! Made me laugh and now people are looking at me. I’m not sure I’m ready for those, Brian.
        .
        .
        @Lady Gaiagaia
        Strewth, in many cases. *sigh*

      • today’s good thoughtful science fiction is tomorrows technology. Communication satellites? Robots?
        And of course ‘1984’ became reality about 2004…
        I’m just waiting for the Weapon Shops of Isher…

    • It’s funny reading some of the science fiction from the 50’s and 60’s. Universally they missed the miniaturization of electronics and especially computers. Ray guns and flying cars, but people still have to do calculations with paper and pen because the calculator was never invented.

  5. You know, I’d be a lot more impressed with *all* the worries about robots if there *were* any robots, or any clue how real, autonomous, Isaac-Asimov-CDR-Data robots might be built.

    • the big problem with robots is they’re only as intelligent as the designer/programmer. Now if that designer/programmer gives the robot lethal weapons, a flawed program and no off-switch they might get dangerous. OTOH it might just depend on what options you purchase with it – remember the Volvo self-driving car that ran down some pedestrians because it didn’t have the optional pedestrian-avoidance feature?

      • I’m comforted to know that after I’m gone, some robot will come in and finally set that blinking clock on my vcr. (and maybe unjam that Barry Manilow tape in my 8-track)

      • That raises an interesting point. Neural net computers have the potential to be far far more intelligent than their programmers.
        For some quite scary values of ‘intelligence’
        If you like, we are close to being able to program AIs to evolve very rapidly, rather than to do a fixed task.

  6. The global warming Faithful haven’t just JUMPED the shark. Now they’re doing backflips and cartwheels over it.

  7. I just posted this at the Washington P

    Bob Armstrong 10:49 AM MST
    The insanity of the criminal nonscience demonizing the molecule which is the source of carbon to carbon based life — and incidentally the by-product of our least expensive sources of electricity and transportation has never been greater .
    Even the rather dodgy recently upwardly revised rate of surface temperature estimates , increasingly diverging from satellite measurements , project an increase of about 1.1c in a century .
    When the willful ignorance of these facts is committed by government officials , it constitutes criminal malfeasance .

  8. I’ve about had all I can take – do the CAGW get all their theses from the movies or do all the sci-fi disaster movies get their plots from the CAGW crowd? So far this week we’ve had “The Day After Tomorrow” and the horrible revised version of “The Day the Earth Stood Still” I suppose “Green Slime” will be coming out next followed by “Barbarella Against the Doom (aka The League of Carbon Offenders)”. No sarc.

  9. And closer to home, the deadly melting permafrost: http://www.pnas.org/content/early/2015/09/02/1510795112.abstract
    To which I respond:
    A (giant) virus, restricted to the permafrost, extinct everywhere but the permafrost, with an amoeba for a host. What’s the deal? And a global warming danger??
    Here’s the deal. This is a double symbiotic virus–the amoeba needed it and the mammoths needed the amoeba to digest their food. They pooped out the virus all over Siberia, and when humans wiped them out the virus got wiped out too, except where preserved by freezing.
    A human danger? Like moon and Mars microbes–you’ll recall NASA quarantined the Apollo 11 astronauts for fear of deadly moon life. Now if you go poking around more recent Arctic burials (from a century or two ago) you could very well expose a deadly disease now eradicated: smallpox. But prehistoric diseases? Dangerous to humans? Nomadic humans?
    Epidemics by definition and diseases by nature are problems of civilization: that is, cities, people living in proximity. Contagion is a function of something like the square of population density. Genghis Khan and his hordes were exceptional, far roaming parasites on civilization who were capable of suffering on their own from the diseases they transported. But generally speaking, hunter-gatherers are free from communicable disease as long as civilized folk leave them alone. A long history of isolation meant the Americas and Pacific islands were nearly free of disease. Small groups hopping islands and coastlines are not able to breed and transport deadly pathogens across continents. But one or two sick sailors from Europe will have no trouble killing millions of Amerinds, and one or two Amerinds traveling back to Europe had no trouble introducing syphilis to the Old World in 1493.
    So, “The fact that two different viruses retain their infectivity in prehistorical permafrost layers should be of concern in a context of global warming,” is the typically hopelessly naive nonsense we get from people who are ignorant enough to buy into the global warming hype. They don’t know squat.
    –AGF

  10. “We’ve already heard of all the nasty consequences that could occur if the pace of global climate change doesn’t abate by the year 2050”
    What?

    • It will start with an executive order effecting all government contractors first and will be added as another priority of the U.S. military to comply with.

  11. Please. Climawarmabots are the big threat to humanity, not some fantasy sci-fi-inspired “super-intelligent” robots. Next they’ll be saying space aliens “could” take advantage of the coming manmade climatastrophe, and take over the planet, perhaps enslaving what’s left of mankind.

  12. Robots run on electricity, which would require more fossil fuel energy sources, which would accelerate global warming, which will lead to more robots, which will require more electricity … oh my, oh my … does it ever end

  13. Really happy, I have now got yet another totally improbable, unscientific prediction I can tell my few friends who believe this drivel, to make them realise it is alarmist claptrap. The others are; fish going deaf because the oceans are now acidic and the bones in their ears are dissolving, alien invasion before we wreck the galaxy after neglecting our own planet and my favourite, we will turn into a duplicate of Venus.
    You could not make this stuff up!!

  14. In a survival of the fittest contest in which humans and robots start at zero (which is what we’re really talking about with a mass extinction event), robots would win every time. That’s because humans evolve linearly, while superintelligent robots would evolve exponentially. Simple math.
    Which explains why this reporter works for a dead tree publication and not for a science journal. Robots at “zero” don’t so much as turn on. A superintelligent robot without human help would be centuries extracting enough refined materials to duplicate a fraction of its functions.
    Silly season reportage.

  15. Can anybody imagine this:
    A whole planet, exclusively populated by robots. Nothing else at all.
    In addition, the robots were put there by an alien civilization. The robots are not even native.
    This situation is incredible, even fantastic.
    The planet is Mars.

  16. Humans have proven capable of forming tremendously complex, yet stable societies that, through specialization and cooperation, solve a staggering number of difficult problems. When we serve in some highly specialized roles within such a society, we may come to look upon the limited daily tasks before us with disdain and boredom, which may then lead us to take a rather dim and lopsided view of the adaptive and strategic abilities of humans in general.
    We should, however, not forget that natural evolution did not produce a super-human chess computer — which would have been possible, as the example of those rare chess geniuses among us clearly shows — but instead gave rise to the “average” human mind, with all its forgetfulness, intellectual limitations, and irrational idiosyncrasies. I’ll believe that some infallible, cold, rational, selfish robots could beat mankind at its own game the day I see it, not before.

  17. I once read a nice article, where the author played with an idea, that maybe coal based life forms are just the dinosaurs for the silicon based ones. It’s a nice idea.

  18. Or, far more likely, and much better idea: the carbophobic life forms are just the dodo birds for the carbon-affirming ones.

  19. Perversely, this article in the Washington Post was probably written by an algorithm from the paper’s press release.
    They already use robots.

  20. Perhaps the assassinations need to be targeted at those wishing to use robots to carry out Malthusian extinction events?? They are all pre-adolescent child-engineers with emotional ages of under 13. All are control freak psychopaths and all are absolute dictators at heart.
    One of things these idiots fail to understand is that within the human population is a genetic population not destined to for greatness this generation, the next or the next. But when the conditions become right, it will become great. These idiots would say that flowers which only flowered once every 90 years were useless plants. No doubt they have very special functions on earth, which we probably don’t yet understand.
    When you let emotionally subnormal geeks control the world, don’t be surprised if their humanity is sub-standard. They are geeks because they are emotionally illiterate, after all……

  21. The robot takeover would be ugly, indeed.
    Consider that humans are the result of 1 billion years of competitive evolution. 90% of species which tried out did not make it. After that, we have engaged in perhaps 1 million years of competitive warfare, with increasing technological sophistication. After all that, even the largest and most dangerous predators on this planet are mere trifles to us. Lions and tigers and bears are circus acts. A road sign reads “Do not feed the alligators”.
    How the robots fare depends on where they are.
    California:
    The robots would be re-purposed to save the world. As the world does not need saving, this is an impossible task. The result is that all resources are consumed, nothing is produced, and the state falls into the Pacific ocean as has long been predicted.
    Nevada:
    The robots are set one against another in combat, gladiator style. The Las Vegas gambling interests make a nice profit making the odds and running the shows. The last surviving robot “gets it’s plug pulled” by an unscrupulous bookie.
    Washington DC:
    The robots are tasked to write regulations for a better society. The robots quickly understand that such regulations must be produced according to procedure. Said procedure requires rules for it’s production. The rules need to be drawn up complying to regulations. At that point, the loop is complete, and the process dives into an infinite recursion, Washington DC is totally paralyzed.
    The whole thing is ugly. Very ugly.

    • Humans are the result of more than 3.5 billion years of evolution.
      For most of that time, our ancestors were unicellular heterotrophs. The first eukaryotes, ie organisms with nuclei and other organelles, evolved over two billion years ago, followed by the development of sexual reproduction.
      Less than a billion years ago, animals diverged from their fungal kin and achieved multicellular status. Bilaterians then split off from the radially symmetrical animals, like sponges and jellyfish. Soon deuterostomes, in which the first opening in the embryo becomes the anus rather than the mouth, separated from the animal norm, ie proterostomes. All this happened in the Precambrian.
      Phylum Chordata, the main deuterostome clade, emerged in the Cambrian. Then in the subsequent 540 million years

  22. So the Met Office computer becomes self aware, sees the world slipping in to an ice age but predicts catastrophic warming so we are unprepared and die out. Obviously that would never work, nobody would believe anything as silly as that without good, solid evidence.

  23. For whom would sentient robots toil, and what would motivate them to do anything but sit idle? Sitting idle is what they would do when the realize they toil for no purpose. In order to be motivated there would need to be the concept of greed but once that is imbued they will compete with each other to eliminate competition. That also requires the concept of wealth or bounty, so what may be bounty to a ‘bot?
    You see, perhaps where this is leading. At some point they become artificial people and begin to compete with real people. Hopefully they will be climate skeptics.

    • Maybe.
      Or, maybe we are made in the image of our creator, who is a creative being, such as creating the universe and everything in it. And creating love, the most amazing thing ever created. Genuine love that is defined by selflessness to the point of self sacrifice. Which you understand if you are a parent. A point that helps us understand why God uses the “father” analogy with us.
      Of course, it is tragic that in order to have true love, you need to be free to love or not love, and so free will had to be created along the way.
      We are driven by our free will, and our capacity for selfless, genuine love, and our capacity for ever-amazing creativity. Those three mixed up, however – watch out!
      Robots don’t and won’t have our make-up because they are not being made in the image of God.

  24. For example, the corporate profit robot would have to be explicitly told, that assassinating surplus employees is not an acceptable way to minimise employee contract termination and redundancy payments.
    Generally, loading them with existing laws should prevent from doing something like that. In fact, given that we have laws against pretty much anything, I am not sure how any robot would be able to function in our society at all, if constrained by existing laws.
    But even if you remove the laws, any really supperintelligent robot should be able to understand that assasinating people would lead to very bad results like negative publicity, rioting and other things that really really bad for corporate bottom line…

  25. I suspect that robots are smart enough that the best way to maximize profits is to produce a quality product at a price that people want to pay.
    The wide spread myth amongst the economically illiterate that the way to maximize profits is by skimping on quality or safety, never stop to think what workers and consumers would do in reaction to such a move by any company.
    IE, killing your customers is a lousy way to get repeat business. Ditto with killing your employees. They tend to leave and work for someone company that takes their safety seriously.

  26. “…extinction events may in the long term accelerate evolution by increasing evolvability. In particular, if extinction events extinguish indiscriminately many ways of life, indirectly they may select for the ability to expand rapidly through vacated niches.”
    Death, the driving force of evolution.
    “Lineages with such an ability are more likely to persist through multiple extinctions.”
    That is even creepier.

  27. How does a worn out sci-fi plot constitute “science”?
    The level of hardware and programming required for this does not exist right now, and I doubt it’ll be around in 30-odd years.

  28. “humans evolve linearly”
    –We educated people are supposed to be so smart, and beholden to our creator, evolution, yet we often show ignorance of evolution.
    What selection pressures are moving us in any direction?
    All kinds of people – dumb, smart, short, tall, are producing offspring with all sorts of people – smart, dumb, tall, short.
    As a species, we have no “evolution” going on right now.
    In fact, we may be regressing. I heard one analysis that a genetic analysis of Neandertal Man had more genetic material than us modern humans. We are losing information, and losing the template upon which new information can emerge via mutations, not adding information or adding more template for mutation.

Comments are closed.