• Question: Is it possible for robotics to overpower humans?

    Asked by mee3megemel to Alex, Claire, Kate, Marcus, Neil on 23 Jun 2014.
    • Photo: Marcus Johns

      Marcus Johns answered on 23 Jun 2014:

      Currently, we’re definitely not at a stage at which robots could overpower us independently. They are increasingly being used on the battlefield but generally have a human controller behind them – like the drones used by the US, the UK and other military powers.

      If we ever developed a computer that had true artificial intelligence, it would theoretically be possible that it could one day decide that we’re not logical enough for it’s liking, or that we aren’t efficient enough, or that we’d make a great power source. There are plenty of examples in literature and films of this happening: HAL in 2001: A Space Odyssey; VIKI in I, Robot; the computers and robots in The Matrix, etc.

      However, I’m not sure that we’d ever be able to create a computer or robot that could think exactly like a human – although I’m sure we’d get pretty close – and I would hope that the researchers that develop the system would develop it such that it would never have the ‘urge’ to take over humanity or have a way to shut the system down if it ever happened.

    • Photo: Neil Dhir

      Neil Dhir answered on 23 Jun 2014:

      Marcus make very many good points about the state of nature when it comes to robotics. Though there are now robots that can ‘think’ to an extent for themselves, but when we say ‘think’ we merely mean that they e.g. can recognise certain things from streaming video (be it patterns or shapes or objects) and the classify these accordingly. Now, Marcus mentioned military robotics in particular. Now from this point of view there are robots which do something called ‘automatic target acquisition’ – which is to say that the robot will look ahead, much like a soldier would, and then pick out what it thinks is a threat and then ask the operator of the robot to take action or not. That way the human is always in control of firing button as it were.

      For what we as humans define as true intelligence, a computer would first have to pass something called the ‘Turing test’ (you can read up about it, it is very interesting). It was designed back in the day by famous British cryptologist Alan Turing. It basically is a test to check the ‘intelligence’ of a proposed algorithm (a set of codified instructions).

      Now for the most important part: could robots ever hurt humans in a malign way? In theory not, not if encoded with certain laws know as Asimov’s three laws of robotics. But, if a robot was truly intelligent, it would by-pass these laws. The future is at best uncertain in this area, since we are not really sure what robots will be capable of if they were to surpass our intelligence, which they surely will some day.

    • Photo: Alex Lyness

      Alex Lyness answered on 23 Jun 2014:

      Hey mee3megemel,

      Sounds like someone is a bit of a sci-fi fan!

      The guys have given great answers to your question and you should definitely look up the Turing Test… particularly as a computer the other day just beat it! A computer convinced some (not all) humans that it was a 13 year old Ukrainian boy, check out the story here: http://www.bbc.co.uk/news/technology-27762088 Bit creepy right?

      I think what you want to know is will there ever be a ‘Technological Singularity’, which is the point where AI (artificial intelligence) and machines show signs of greater-than-human intelligence? This is a scary concept as it would be hard to see machines would bother with having us humans on the planet too! You can read more about it here: http://en.wikipedia.org/wiki/Technological_singularity

      Don’t panic! It is still all still fairly unlikely, although as we develop more and more computing power, you never know, we might stumble on something we wished we hadn’t!

      P.S. If you haven’t seen iRobot you really should check it out all of your question is in there: http://www.imdb.com/title/tt0343818/

    • Photo: Claire Brockett

      Claire Brockett answered on 24 Jun 2014:

      I think the others have answered this really well – at the moment it’s not possible as we need human controlllers of the robots – but they do provide very practical solutions to going into environments where it might be unsafe for a human.
      In medical engineering, there are things like robodoc – which is robot assisted surgery and this helps the surgeon do his job – I can’t really see there being a time where the robot is left to do all the surgery by itself any time soon.
      What do you think about robots? Are they a good thing?