TED Conversations

Ajay Ramanathan Chettiar

Student, Indian Institute of Technology - Kharagpur

This conversation is closed.

Can a computer theoretically exist which can generate its own algorithm, think on its own ?

Can a computer exist in future theoretically leaving aside science fiction. can they think on their own. It can solve problems, or create algorithm which will solve basic problems.

Share:
  • thumb
    Aug 26 2011: Our brains ARE computers. You could say the same thing about cells. They are machines constructed with proteins that run algorithms based on code.

    We just haven't figured out exactly "how" the brain works. Once we know for sure, it will inevitably be taught to a computer. We are getting pretty close. It may have already happened.

    Robots evolve altruism:
    http://www.wired.com/wiredscience/2011/05/robot-altruism/

    A self-aware robot:
    http://inventorspot.com/robot_demonstrates_self_awareness

    Emily Howell creates art:
    http://technology.timesonline.co.uk/tol/news/tech_and_web/article6884631.ece

    IBM's synthetic brain:
    http://www.extremetech.com/extreme/93060-ibm-creates-learning-brain-like-synaptic-cpu

    We might wake up two years from now to a totally different world.
    • thumb
      Aug 27 2011: The programs(robots) here are pre-code to do the stuff example like the self awareness robot it is pre-coded to actually distinguish itself, and others. in humans if a person with no eye sight since birth is given a eye sight a at a age after 5 or 6 its of no use because your brain can't actually identify them and perceive the visual data. on the other hand a robot can be programmed to perceive data at any time of his life. the point just like we gave the eye to human before the age of 2 can develop eyesight on on its own can robots be given a sensor and no programing ,thus after it senses object in its path but doesn't know what it is it hits it 2-3 times but the 4 th time it again senses and realizes that it shouldn't go that way. Can robots do that ?

      For example you use a manual formula to first add 5 then subtract 3 then again add 8 then subtract 2. after 2-3 times using this formula you will recognize that it is same as adding 8 add to further problems you will add 8. though your boss says you to do the long step you take the short one cause you realize the answer would be same. can a robot go against the laws it is defined (algo) and invent, imagine its like CAN ROBOTS DREAM ?
      • thumb
        Aug 27 2011: It is true that the robot Nico probably isn't fully self aware, but it managed to meet one of the criteria. That is what's important.

        Those robots literally evolved altruism on their own! It was not a programmed outcome.

        Emily Howell was programmed to create art, but that's not whats important about it. What shocked the world was that it could be done at all. A robot, upon request, created a work of art. That is amazing.

        Robots can't dream yet, but I wouldn't be surprised if that changed next year thank to this new processor created by IBM.

        My point was, we are slowly reaching a point where computers will think on their own! Computers will program themselves one day.
        • thumb
          Aug 27 2011: About the altruism it is the renunciation of the self, and a concern for the welfare of others, Robots aren't sharing it is "In different iterations of the experiment, the researchers altered the costs and benefits of sharing" thus robot finds the logical best way.its not altruism is it ?

          Emily Howell the truth is i m truly impressed i do believe that someday robots can be human like completely but still you browse the net you can find 5 year kids drawing ( painting ) modern art, the only point is modern art itself is not so defined. any random painting is modern art. there are examples in history where a art is kept upside down cause people thought it looked better that way. similarly music is also a not so defined sector. its like i can find the buzz of fan interesting (5 better ways to learn ted talks). the point is a kid who draws the modern art never understands the emotions in the pic. you see a image , listen to a music you feel a surge of feeling. the artist whats you to feel it. does creating music means it has emotions. and 2) i wanted to ask this question i will a computer break its previous algo (written by the admin) and create its own cause it found the first one long and tedious 3) Due have any idea about the mutation codes of virus (computer ones)
      • thumb
        Aug 27 2011: Yes, this is how primitive altruism works. This is why ants don't attack each-other. They learned at some point that it was better to work together.

        As for everything else... only time will tell!
  • thumb
    Aug 26 2011: I think this is a question Norbert Wiener and the other founders of informatics already had in mind when they created the first computer with the goal to imitate the human brain. Maybe the new generation of chips IBM is currently developing will help us find the answer, maybe the upcoming generations of robots will eventually obtain the skills to "learn" as opposed to being fed with instructions. I see you are studying at an institute of technology: was there a lecture that inspired you to ask this question? Or are you currently working on a related research project? It would be great if you could share some more of your thoughts about this topic as well!
    • thumb
      Aug 27 2011: I actually believe we are a computer like Mr.Campbell said.What inspired me is a strange question ? Its actually i fear death. I don't wanna die. so Last year i came up with few ideas.

      Assuming Your Brain is like a Flash Drive if you could copy it and paste it in similar flash drive there will be no difference in them both so i started thinking if computers could have the power to generate algo's then we can actually completely copy our mind into a computer(non assembly programmed one) and make a robotic body (see my previous idea) and live for eternity

      and the second reason was a computer can't (rather never do some stuff) like generating a random number, ( i also confused that the number which our brain generates is a random number or not but still), this generation robots are pre-coded to learn the motion of the human like walking, and duplicate them. but is this what we do ?

      A tiger cub never eats grass even if it is bought up in environment where one eats grass.; are we really so similar. and i believe we are.

      so i want to unite the gap between the robots ( mechanical part ), the algo and the humans ( i hate electronics
  • thumb
    Aug 28 2011: It would seem that for something to be able to think, it would need to be a living organism. How do you embed life into a computer? And are you talking about the hardware or the software? The hardware is the computer, and the software would be what generates algorithms. I don't think we're talking about the computer, per se. The hardware isn't cognizant of what software is running on it. So the question actually seems to point toward whether software can think on its own.

    Before we got to the stage of considering how consciousness could translate into an inanimate object/process, we have to look at what we've done in other areas. We make synthetic food which simulates organic food. But does that make it real food, just because it mimics some of the properties of organic food?

    Ultimately, it may be just as much of a semantics game as anything so far. We don't even know what consciousness is yet to be able to discern whether computerization could possess it. We don't know what the 'it' is, so the question is basically empty at this point until the 'it' has been established, and we could be centuries away from that.