Tuesday, November 10, 2009

Stuff Internet-savvy People Like: Pranav Mistry

There's a limit to the amount of adulation I can tolerate for raw ideas. Sure, wearable computers are the future; sure, querying the Cloud is Tomorrow's way of just asking someone if that brand of garlic bread tastes good; sure, information that you are now forced to get from slow Facebook stalking can come right from a few discreet thumb-presses, and you can impress that girl with a amazingly coincidental set of interests. ZOMG! She just can't imagine there's someone else in the world who has  cats with the same names!

All that granted. But the collective shagging on "the genius behind the innovative SixthSense application", Pranav Mistry, is nauseating. It is just a concept, an idea, like a million other ideas! The video that Pattie Maes showed in the TED Talk is, in my opinion, a reasonably well made funding-agency demo. Nearly everyone I know in grad school who is working on hands-on research has made a video like that. You get a funky piece of technology that is extremely common among some circles, but is surprisingly rare in general. In this case, it's a matchbox-size projector and a gesture recognition system. 90% of the people who see one in action are hypnotized just by the fact that something is being projected out of that small box, but every lab has tens of them lying around. Next, hook it up to a nice GUI, in this case a simple Cocoa-based one. Choose a sequence of examples that are of interest to the target agency - in this case, shopping. Hard code an application to follow a certain procedure to demonstrate a concept - in this case, look up something on the web and project it. And voila, instant internet recognition!

The demo comfortably sits inside the continuum ranging from completely hardcoded demos (everything is scripted) to completely open demos (like if instead of showing a video, a member of the audience was asked to come up, wear it, and do something in realtime). It's much more closer to the hardcoded demos - sure, the realtime web downloading is real, the projection is real, the keypad identification is real, even the gesture identification is real (he claims that it took 50,000 lines of code to get gestures+the computer vision part working - that seems about right for code written from ground up, but no way in hell is this new or even innovative [*]). But the entire thing together is scripted! A few typical scenarios are chosen, and a sequence of events of using those technologies is orchestrated. Getting a few parts of the demo to work would certainly have been challenging, and it does take several weeks to even get a simple computer vision code working for a perfectly defined object if we just vary the lighting and the angles. But it is impossible for that to be considered a completely general product!

The slashdot summary, on a story that says he graciously decided to use open source after having "put paid to the canard that open source and innovation are incompatible" reads:
Mistry’s decision has meaning beyond Sixth Sense. The desire of inventors is always to get their work into the market as quickly as possible. Usually this means waiting for it to be turned into a useful, profitable invention. Mistry is bypassing this by going straight to open source. (Italics mine)

Honestly, I think people are far too consumed in imagining elaborate visions of future technology and heralding the next victory for open source to really look hard and notice that this is just a slick frickin' demo! 
I've seen far too many demos and done enough of them myself to believe this is a real, open demo. If indeed it is a completely open demo and anybody today is free to try out the product if they wanted to, then I'm completely wrong and Pranav Mistry has achieved something that I believe is impossible  (in a 2nd law sense) with the time and resources available as a grad student.


[*] - Here's a product review of a $130 projected keyboard.
And a product demo of gesture recognition with video, on a mobile phone.


Reactions to this post, mainly from critiquing myself and some from others, have included 'Cynic!', 'Overly spiteful critical old crone!' , 'Pessimist!', 'Indian crab!', 'Unbeliever!', 'What have you got against the poor guy?', 'You're just jealous!', 'You try doing something like that!', 'What have you achieved?', 'You're going against public opinion just to seem cool', 'Inverted watermelon!' (Red outside, Green inside (Anger, jealousy)),  and I quite amusedly actually agree. I don't know what's possessing me to get so irritated at this.


Sameer said...

What a coincidence. My thoughts were along very similar lines when I read the /. post yesterday and revisited the TED lecture, Mistry's website, etc. The standard defense to the "all this is scripted" argument is that all technologies start off small, it takes time for a technology to mature, etc. However, I don't see how something that's fundamentally grounded in hard-coding/scripting can ever generalize, how much ever you want it. The pragmatic side of this argument, however, is that Mistry et al are doing exactly the right thing from a "how to get funded" perspective.

Anonymous said...

I can't agree more on this topic with you.

-Pops said...

I completely agree ! Most of my last semester was spent working on one such demo :) It is impossible with the current state of research in computer vision, that lots of what Mistry shows in his video could work in anything other than a very controlled setting. And in a controlled setting, it's actually quite unimpressive. But like Sameer says above, a good demo goes a long way in getting funded :)

dileep said...

What was your demo about??

tejesh kinariwala said...

I am a die hard fan of TED videos ,in fact lined up after this feed from your blog is a feed from TED videos. I have downloaded and seen over 4 GB of these videos on genius inventions,jaw dropping feats and what not.I love them.
Now when I started reading your post, I thought I would be irritated by the way you cynically killed the whole Pranav Mistry hawwaa one layer at a time. But surprisingly I found myself agreeing with you on many lines.I understand being a elec grad (to be) myself, that I have lost the ignorance to be blissful about such crap thrown on me.
But then again all said and done, looking at it from another perspective, mistry is no different than other sellers of cutting edge technology.comeon the "life changing iPod that will make the world a better place to live" as projected by Jobs is nothing but an mp3 player! :)

Mohan K.V said...

@Sameer and Pops, thanks for chipping in, it's gratifying to know I'm not insane.

@Sameer: Funding, ahh, now we're talking :)

@jatkesha: Super nick guru! How are the noble Southpark translation efforts coming?

@Tejesh: I love this line: "I have lost the ignorance to be blissful about such crap"!!

And about Mistry being no worse than other tech salesmen, absolutely. The poor chap has to live a life, too. If my words have brought down his stature from being Genius Inventor of Sixth Sense to Just Another Salesman, they have not been in vain :)

Oooh, I love the smell of The Referendum in the morning. It smells like..victory! :)

Mahesh Mahadevan said...

Mistry's work had a Flash-y Demo-graphic :-P . I had referred to part of it in a post, but more so in the context of how I already refer to a lot of people as tag clouds (StumbleUpon hasn't made it any easier).
The Krishna Iyer excerpt has a brilliance I can (grossly insufficiently) best describe this way - my console-mate has a penchant for -v.
PS: Your posts deserve a "See Also" section, much like some landmark Wiki articles.

karthikv4u said...

Totally agree! Mistry has done good for himself as a grad student, but our 'new age' internet starters have been so awed by the video that even after months, we get mails n facebook/orkut updates - Astounding, Amazing, You wouldnt have seen it, Totally Unbelievable, Indian Invention, (of course all this with lots of !!!!!!!!!!!!!!!!!!) - so much hyped to the point that 90% of the Indians thought and still feel the recent TEDIndia was all about Pranav Mistry if you discount Usha Uthup and Shivmani from the scene. Probably Maes too wouldnt have thought the Indian netizens would make it soooooooooo sooooooooo popular! Unbelievable!!!!!!!!!!!!!!!!!!!!!!!!!

DJ said...

You are not alone dude !! Every engineer feels miserable when he sees this. He feels "Shit.. what is so great about this? I know how to make it.. there is actually nothing new he has found.. last time, I had told this idea to my friend .. Y is he getting all the recognition ? "

Well these are some of the less aggressive thoughts..

One thing Mr Mistry has been successful is to make all the self proclaimed PhD's/Engineers feel miserable.

I agree with you man.. shit I should have been the one standing there! well sadly, I screwed up !

Anonymous said...

1 q..is ALL wat he shwd possible at all?..

Swapna said...

i kind of totally agree with you. i am a patent attorney and because of my wokr, i come across such agglomerations of well known technologies, where the sum of the total is equal to the sum of the individual parts.