Forgot your password?
typodupeerror
Hardware

How Machine Learning Will Change Augmented Reality 101

Posted by CmdrTaco
from the i-can't-even-people-learn dept.
An anonymous reader writes "Augmented reality is already adding digital information to the world around us — but the next step to making it truly useful will be when it starts to use elements of machine learning to understand the real world, Mike Lynch, boss of machine learning software specialist Autonomy told silicon.com — also explaining machine learnings links with the theorems devised by 18th century cleric Thomas Bayes."
This discussion has been archived. No new comments can be posted.

How Machine Learning Will Change Augmented Reality

Comments Filter:
  • by Toe, The (545098) on Tuesday February 08, 2011 @05:25PM (#35143150)

    (Disclaimer: I have nothing to do with this site, and it is non-commercial as far as I can tell.)

    This leads me to pimp my favorite new site/game/lesson... what is this? It's cool, that's all. Check out this neat implementation of a genetic algorithm to produce a cool demonstration computer-generated evolution: http://www.boxcar2d.com/ [boxcar2d.com]

    • What the fuck is that shit? How many millions of years should I wait unit I get a working bicicle? I'll run out of battery!
    • by Superken7 (893292)

      Wow, I already knew about that 2D car experiment but just re-discovered it with a far more interactive design, thanks!

      Someone should make an app for that (not kidding) that lets you be the designer of a car and makes your car compete against other user's designs, producing an online top ranking, friends ranking (you have been ousted!), etc... Maybe there is even something more sophisticated than building 2D cars that would make a great game...
      (Yes i know the website lets you design a car, thats where I got

      • by Toe, The (545098)

        Try turning max wheels down to zero. In the comments, people are reporting that they end up with spoke-like-contraptions that evolve to roll without wheels!

    • This reminds me of an old simulation like this that was done in 3d. At the time it had to be run on a uni mainframe, but this was years ago... would love to find that and see what could be done with my desktop rig.

  • One day some fool will ask a machine to figure how to rescue the environment and fix it for us, and the machines will figure out that humans are the ones who destroy it... let's hope that Newton's Laws come standard by that point, otherwise we'll be in deep deep trouble.
    • Newton's Laws? (Score:5, Insightful)

      by VirginMary (123020) on Tuesday February 08, 2011 @05:35PM (#35143270)

      I think you meant Asimov's Laws of robotics! I doubt classical physics has anything to do with it.

      • Re: (Score:3, Funny)

        by Toe, The (545098)

        No, no. Learn your history. It's Asimov's version of Newton's Laws.

        The full form goes something like:

        1. Every robot must remain in a state of constantly not injuring humans or causing them to become injured through the robot's state of rest.

        2. Any robot, subject to a force in the form of an order by a human undergoes an acceleration in the form of obeying the order as long as it does not contradict the first law.

        3. The mutual forces of action and reaction between a robot and another object must not allow th

    • Asimov's Laws? Unless, is there something about motion, gravity and robotics that has passed me by all these years?

      • by Anonymous Coward

        Conservation of Isaac.

  • Look, all I want is my AR glasses to overlay the world into an MMORPG/FPS sim and I'll be good, okay? Call it a reparation for the future not providing me with my own jetpack and/or flying car yet.
  • by jfengel (409917) on Tuesday February 08, 2011 @05:33PM (#35143250) Homepage Journal

    Nonexistent product will change your life! Film at 11.

    (Eleven years from now, that is. We think. Maybe the schedule will slip a bit.)

  • ... when it's done, you bastards! I wasted 5 seconds of porn reading the summary.
  • by Un pobre guey (593801) on Tuesday February 08, 2011 @05:49PM (#35143408) Homepage
    If you can get a computer to do [those tasks] then that's a phenomenal saving, and it frees up the human to do something more interesting.

    Right. That's what's been happening. Humans have been freed up to do more interesting things, and for more pay, too. Uh huh.
    So, the more we make machines do more of the work people do, the more interesting work there is for the rest of us? Those of us who don't own the machines? Those of us who need to make a decent living? Does this guy live on planet earth? Can it be that in 2011 there are still people in decision-making positions who still believe that?
    • by retchdog (1319261)

      well, struggling to live on the street and squatting in flophouses can be considered "interesting;" he didn't say "fruitful."

      an ayn rand quote would be all too easy to find, so here's one from the radical left: ""Down with a world in which the guarantee that we will not die of starvation has been purchased with the guarantee that we will die of boredom."

    • This is a complaint about how wealth is distributed, not a complaint against progress.

      • This is a complaint about how wealth is distributed, not a complaint against progress.

        No, I'm serious, and not being snarky -- for many people already in positions of power, "progress" means them getting more [desirable noun]. So while the recent global financial meltdown set many of us back considerably, it has still been deemed as "progress" by the financial elite, at least as I've been reading in the media. For that matter, I've been reading and hearing for over a year now about how the economy is supposedly doing better and better, i.e. "progressing", but I have yet to see my personal

      • Are you equating the elimination of human labor with progress?
    • Your point started off being that 'more interesting = more pay' is not sustainable. That's true. If money is based on any sort of 'scarce' standard, then you'll run out if you increase the pay-grade of everyone phased out by machines. What's really supposed to happen however is that the minimum wage jobs shift, so everyone goes DOWN a pay-grade.

      The 'idea' is supposed to be that when machines are doing the menial tasks, like 'farming', then the cost of living will go down for everyone. After removing

      • You're simply repeating the standard naive argument from 60 or more years ago. Eliminating "menial" labor, more commonly called "blue collar jobs," is neither scalable nor survivable. Those people will not become engineers, scientists, professionals, or "white collar" employees as your model will effectively require. While many products and services may diminish in price, a great many people will become under- or unemployed. The poverty line will go up, not down. Beware of simply accepting pop-culture notio
        • How is automation any different than what's already happening with outsourcing? 70-80% of the GDP of many western countries already comes from the services sector.

        • by tehcyder (746570)
          You are overlooking the fact that under a proper socialist system the wealth created by the worker robots would be shared amongst the entire population, not reserved for rich people who privately owned the means of production.

          This is because you can only conceive of a capitalist system where you are defined by a combination of the money you have and the "productive work" you do.

          If most people no longer have jobs, so fucking what? Here, you are biased by some version of the protestant work ethic, where

          • You are so off the mark it is almost cute. First and foremost, the trend towards replacing human labor with robots is occurring under a corporatist plutocratic regime, not some ill-defined "socialist system." The fruits of the robotic labor serve the plutocracy, who own the means of production. This will not change in the foreseeable future.

            This is because you can only conceive of a capitalist system where you are defined by a combination of the money you have and the "productive work" you do.
            Now that's
        • You're simply repeating the standard naive argument from 60 or more years ago. Eliminating "menial" labor, more commonly called "blue collar jobs," is neither scalable nor survivable. Those people will not become engineers, scientists, professionals, or "white collar" employees as your model will effectively require. While many products and services may diminish in price, a great many people will become under- or unemployed. The poverty line will go up, not down. Beware of simply accepting pop-culture notions of capitalism, they are wrong. Many counter-intuitive results will come from making machines do all the work. Those who don't own robots will be increasingly unable to participate in the economy.

          You haven't given me any reason to think that his 'naive' argument is not correct other than you saying so. Is life worse now than 60 years ago because automation has replaced people in a number of menial tasks? I don't see it. I can buy an iPod for a small amount of money because the factories that create them are largely automated, and the ships that transport them from there to here are largely automated, and the packaging and delivery system is largely automated. The same applies to food, and clothi

          • You have in effect answered your own question with an attempt to qualify your claims:

            If, and this is a huge if, almost all tasks that required human intervention for 'menial' tasks was taken over instantly by robots, then yes we would have a problem. A huge percentage of our workforce would suddenly not have a job, we'd have social unrest, etc. But the way that it's been happening for the past couple hundred years is that the automation has been creeping, slowly replacing tasks. Yes, people who used to wo
    • I think we need to think about radically reforming the economy as automation becomes more and more common. Eventually only creative work will be available to humans, and while creative work is great I doubt it can provide enough jobs for the entire population. If we don't do something radical to make sure everyone shares in the fruits of the increased productivity of society, we will have a huge permanently unemployed underclass, some middle to upper class workers and a few massively wealthy owners of the a

      • Why has this not happened in the past? There has already been an enormous shift in the types of work that people do. Previously, most people did menials tasks on farms, and now they do not. After farming was manufacturing, and that is largely automated (though not entirely obviously, cf China). What are people doing now that they did not do then? How did that shift happen such that we did not have 50% unemployment? What does that imply for the future? I think that it means that people will still be p
        • Why hasn't it occurred? Because powerful computing hardware has never been so cheap and abundant. That is the new, disruptive change. It still growing by leaps and bounds. You can already buy cards with 64 cores running linux [tilera.com] and put them in your PC or robot. Mobile devices are already going multicore [linux-mag.com]. Distributed machine learning [ieee.org] is already a reality. Those things did not exist before. and that is why there hasn't been 50% unemployment due directly to automation. Forget Asimov and Bradbury, they did not fo
  • If I could go back and do it all over again I think I would spend my entire life trying to figure out how a mosquito's brain works. There must be research along these lines happening somewhere but you never hear about it - they are always trying to map out mouse brains, or some other small mammal.

    Why so ambitious? Start small - if a computer program could be made that perfectly imitates a mosquito it would be a huge breakthrough.
    • by vlm (69642)

      If I could go back and do it all over again I think I would spend my entire life trying to figure out how a mosquito's brain works. There must be research along these lines happening somewhere but you never hear about it - they are always trying to map out mouse brains, or some other small mammal.

      If the assumption is once you're done with this project, you'll move on to human brains, then a mosquito has some pretty severe I/O differences compared to a typical mammal human brain.

      • by jomama717 (779243)
        I'd go from there to an ant brain. Maybe a mosquito is even too ambitious - I wonder if anyone has tried to simulate the nervous system of a sea sponge, or an earth worm...
        • The best work I know of is on the sea slug, Aplysia californica. Do a google search for neural system, simulations or models of it. A lot of work has gone into determining the types and connections for every neuron in the organism, where they came from developmentally, what they do and how they work. There are about 19,000 neurons. We do not have a complete model for it yet. So, we're a really, really long way from doing a mosquito brain, though I'm not sure how many neurons they have. Honeybees have
        • by tabrnaker (741668)
          Could be wrong, but i do recall that the they have mapped out an earth worms nervous system, a bit of a stretch to call it a brain though.

          Ant's aren't easy because you can't really view them as individuals.

    • I swear I remember reading that IBM had already simulated an AI "Intelligence" with sophistication on par with a cat's brain, albeit not at full speed.

      Yep, here it is (one of many articles on the subject I picked at random)
      http://www.technewsworld.com/story/68678.html [technewsworld.com]

      • What they did was a simulation of a neural network with same number of neurons as a cat cortex. The cortex is only part of the brain and just simulating a bunch of neuron isn't the same as simulating the functionality. That's still a long, long way from being on par with an actual cat brain.
      • by jomama717 (779243)
        I remember this story - the catch was that they simply (ha) set up a software brain simulation which had enough numbers of neurons and synapses (from the article: 760 million, 6 trillion, respectively) to put it above cat-scale, but as far as I can tell no actual attempt was made to virtually render a living brain in a computer.

        The brain of an ant contains a mere 250K neurons, seems like it would be a cake walk after the cat-scale exercise :)

        More animal neuron counts [wikipedia.org]
        • Correct, running the simulation of the ant brain would be a cake walk, assuming you knew what to put in the simulation :-)

          And determining what the model should be, ah..there's the rub. How do all of those 250k neurons connect? What do they do? That's a hugely hard problem.
          • by jomama717 (779243)
            Thanks for both of your replies - interesting stuff. Someone else posted this link [ieee.org] in reply, it goes into pretty fine grained detail about efforts to model a fruit fly brain, pretty fascinating. The article points out that in addition to the extreme complexity of the actual connections, it is even more complex in that "firings" of the neurons aren't simple on/off firings, they can each fire at at different percentages. It also goes into the storage space required to store what they find - it's staggering
    • you must have never heard of Janelia Farm, the home base of HHMI foundation, where they have a huge project of reconstructing fly brain http://spectrum.ieee.org/biomedical/ethics/reverse-engineering-the-brain [ieee.org]
    • I believe what you're looking for are The Fly Papers [princeton.edu]; research by Bill Bialek and various co-authors which date back almost a decade. For a great overview, check his book, Spikes.

      • by jomama717 (779243)
        Wow, thank you. I should have been a little more diligent in my googling on this topic, or at the very least posted something about it on slashdot earlier :)

        In all of the links to various attempts at modeling "simple" brains that I've been given it becomes clear that even the most rudimentary of brains requires an incredible amount of time and effort - in addition to massive amounts of storage space. I still believe that this path is the one that will eventually lead us to true AI and an understanding of
  • I easily say what I look forward to, and it will come from a combination of machine learning, human input, structured and unstructured information: the ability to look at something and know how it works, what it's made of, where it came from, who's involved with it. I mean, not having to google/wikipedia every interesting aspect, but having it show up translucently in front of what you're looking at.

    This would be especially interesting for complex things like computers, electrical devices, organisms.

    I'm loo

2.4 statute miles of surgical tubing at Yale U. = 1 I.V.League

Working...