Banner
    Cat Usability Testing (Wolfram's Predictions)
    By Samuel Kenyon | June 12th 2010 09:37 PM | 7 comments | Print | E-mail | Track Comments
    More Articles
    All Articles
    About Samuel

    Lead software engineer at iRobot Corp., user experience (UX) designer, agile manager, actor, writer, atheist transhumanist. My blog will attempt...

    View Samuel's Profile
    Stephen Wolfram's long-winded (no offense meant) talk at the 2010 H+ Summit was about predicting the future. 

    The material was mostly standard Wolfram stuff but with some focus on future technology.  NKS points of view on AI were of course also present.  The most interesting theme for me was about human purpose.

    Here are a few points I extracted:

    Humans can't predict the future because of computational irreducibility, except for "pockets" or reduceability.  I'm not entirely clear as to what defines those pockets.  This notion apparently has the premise that human society is a sufficiently complex system so that humans have to run the program to see what happens.

    One potential major thread of future technology could be to migrate engineering solutions from iterative to search-based.  In the most extremely abstract cases, an engineer could start from scratch and do an automatic search through trillions of programs in a computational space in order to find the one which exhibits the desired behavior.  Knowledge of how it works internally is not necessary.  I am wondering as I write this, however, how you could be sure that an internally-mysterious computational entity would not at some point exhibit undesirable behavior.

    On the premise that we do more search-based engineering in the future, we will no longer limited to mere adjustments and recombinations of existing technology. We could create almost anything immediately, at least in computational spaces.  If that becomes the case, then the limiting factors will then be human purposes. 

    If you could make any program in the world relatively quickly, why would you make something alien?  We are limited to what we can use and interact with.  Of course, I would point out that those usage limitations will change as human minds and bodies change, but limitations of some kind are still there.

    An analogy to cats is where we get the phrase "cat usability testing."  I'm not sure if Wolfram saw this video, but he probably did since he mentioned apps cats will like on iPads.

    Comments

    Gerhard Adam
    We could create almost anything immediately, at least in computational spaces.

    What does that even mean?  It seems that this is the epitome of form over substance.
    Gerhard Adam | 06/13/10 | 16:49 PM
    SynapticNulship
    Well, at the very least algorithms and narrow AI systems could be developed via search.  That's already done sometimes with evo algorithms to search for solutions, e.g. for antenna design.

    I can imagine full products being searched for also, if we had enough computational power to crunch the possibilities.  After all, to make an artificial object like a cell phone you need source code + ee drawings + mech drawings + manufacturing process, all of which are digital.  The method of manufacture could be completely automated, thus also being basically a matter of software.
    Samuel Kenyon | 06/13/10 | 18:52 PM
    Gerhard Adam
    The method of manufacture could be completely automated, thus also being basically a matter of software.


    That seems like a huge leap of faith.  This is precisely what was supposed to happen with robots in factories, so that a consumer could basically customize a car (or other products) to their own specifications and do it cheaply.  Instead we find most organizations charging unbelievable amounts of money to perform the most trivially computational tasks.  For example, why does it cost $180 to change an airline ticket (that doesn't consider if there's actually a ticket price difference which is added on top of that)?  What do banks charge fees to examine your balance at an ATM?  For that matter why do banks charge for "foreign" ATM usage?

    After all, to make an artificial object like a cell phone you need source code + ee drawings + mech drawings + manufacturing process, all of which are digital. 
     
    I understand, but the problem isn't information.  I don't understand why there's a perception that there's some sort of information gap that's preventing us for building a utopia.  The simple reality is that we can't afford the technology and that's not likely to change anytime soon.  In your example you're presuming a huge infrastructure that is effectively on "stand-by" so that someone can tinker with it to make their own cell phone. 

    As for your software perspective? ... Well, I'd be content with systems that didn't fail for stupid reasons. 

    Gerhard Adam | 06/13/10 | 19:39 PM
    > What do banks charge fees to examine your balance at an ATM? For
    that matter why do banks charge for "foreign" ATM usage?

    Balance inquiries are free at my bank's ATMs, at least for a reasonable quantity per month. Why do you think it should be free, either for inquiries or for "foreign" usage? Banks sell financial services.

    It costs real money to provide an ATM. The machine itself is expensive, since it has to be convenient to use and *inconvenient* to damage or steal, and it will only last a few years. A bank will stock the machine with $20,000.00 or more. An agent of the bank will visit the machine at least daily to refill its cash supply (can't just upload that via the internet) and perform other maintenance. Their network of machines is part of "a huge infrastructure that is effectively on stand-by" so I can use it at my convenience.

    Who pays for all that? It's paid by interest the bank makes on the money I deposit with them and by any ATM-related fees they do charge. The bank hopes to offer the right combination of services, some for free and others for fees I'll accept, so I'll be persuaded to trust them with my money. Someone who has no relationship with the bank has not paid any of that. As a customer of the bank, I don't mind a bit if the bank charges a fee to non-customers for using its services. I think "foreign" users should be grateful they are allowed access at all, much less to expect it for free.

    I dunno about fees for airline ticket changes, but I suspect there are real reasons there, too. Airlines are in the business, after all, of selling rapid travel. They give steep discounts for early purchases, and they've got a huge logistical schedule to maintain. I suspect the ripples from a change request propagate a long way in their systems.

    The above isn't entirely tangential to Mr Kenyon's original article. If his "search-based" engineering utopia should ever arrive, I think it will have fees attached.

    Dennis Murphy (not verified) | 06/14/10 | 08:30 AM
    SynapticNulship
    These issues of costs for services, and what parts of a service are analog or digital, is interesting, but I didn't claim that anything would be free or even efficient.

    Nor do I think that a search-based engineering is a "utopia."  The fascinating thing is that in any case where we can have computers design application or products, the limiting factor for creativity and radically new inventions is our own human purposes. 

    Something Wolfram did not mention, which makes it even weirder, is that human purposes themselves will also be changing in the future.  Human factors and user experience design will also have to expand to include the different types of minds and bodies. (I wrote a little article about that actually: http://ieet.org/index.php/IEET/more/kenyon20100108/).
    Samuel Kenyon | 06/14/10 | 15:01 PM
    Gerhard Adam
    Why do you think it should be free, either for inquiries or for "foreign" usage? Banks sell financial services.

    You're joking right?  It makes no sense to utilize a technology to charge over a $1.00 in most cases for a transaction that literally costs pennies to execute, while providing the same service for free when a teller is involved.  The technology was supposed to reduce the costs by reducing or even eliminating the costs of personnel that formerly provided the service.
    I think "foreign" users should be grateful they are allowed access at all, much less to expect it for free.

    Would you feel the same way if they charged you for internet access?  What about the outrageous fees to pay somethihg online?  This is simple thievery and has nothing to do with the costs of delivering such services.
    I dunno about fees for airline ticket changes, but I suspect there are real reasons there, too.
    Once again, I don't know why you're being such an apologist for outrageous practices.  As I said, ticket price differences aren't included in this, so that charge is literally the cost of a few keystrokes at a terminal to determine if a seat is available.  There's no logistics involved, nor scheduling, nor anything beyond determining if a seat is available. 
    They give steep discounts for early purchases...

    It's clear you don't travel very much.

    I've had the same thing happen attempting to extend a car rental over the phone, where the rate jumped from about $250/week to over $800/week (for the same car).  However, what makes it so outrageous, is that when I drove down there, they offered me the original rate in person.  I cancelled the car and went to a different company.  My point is that there is no business case for these activities.  They are simply a gouging practice that they feel they engage in if you have no alternatives.
    Gerhard Adam | 06/14/10 | 18:39 PM
    I think the really intriguing idea Wolfram had is that there are many many interesting algorithms, processes, systems, etc that will never be discovered because they lie on remote islands in fitness space. Incremental development, which is the kind of thing we humans (as well as evolution, GAs, etc) are good at, will not find these "solutions".

    Brute force would *seem* like a stupid thing to try, because our intuition is that complex behavior requires complex design which requires intention, and that inference would rule out the existence of such "islands", but Wolfram and others have shown (in admittedly a rather an abstract way) is that this isn't always true.

    One thing Wolfram doesn't do, which I think he should, is to make all of this more concrete. What kinds of systems? Here is one of a few academic papers that has pursued this idea: http://www.bardavid.com/mead/mead.pdf. They searched billions of random algorithms and found several that were competitive, but different, to the canonical computer science ones. The authors drew very limited conclusions from their surprising success. And that is one of the reasons I think Wolfram's vision is important: without somebody articulating these ideas lucidly, it will take a lot longer for people to latch onto the possibilities.

    Disclosure: I work for Wolfram|Alpha.

    Taliesin Beynon (not verified) | 06/19/10 | 23:39 PM

    Add a comment

    The content of this field is kept private and will not be shown publicly.
    • Allowed HTML tags: <sup> <sub> <a> <em> <strong> <center> <cite><TH><ul> <ol> <li> <dl> <dt> <dd> <img> <br> <p> <blockquote> <strike> <object> <param> <embed> <del> <pre> <b> <i> <table> <tbody> <div> <tr> <td> <h1> <h2> <h3> <h4> <h5> <h6> <hr> <iframe><u><font>
    • Web page addresses and e-mail addresses turn into links automatically.
    CAPTCHA
    If you register, you will never be bothered to prove you are human again. And you get a real editor toolbar to use instead of this HTML thing that wards off spam bots.