How you can make money and my life easier by posing as a Turkish Tin Man
WHAT IS IT
Computers are really, really good at some things. And really, really bad at others. Computing the square root of 42, not a problem. Telling a funny joke or understanding what someone is saying accurately….not so good.
While we may have to wait another decade or two before we’re doubled over in stitches at, ‘the one about the mouse and the kinematicist’ up at the local Robolaughatorium there are some incredibly smart ways here today where man and machine are able to work together to solve problems that on their own would likely not be possible or economically viable.
To start with a short history lesson, courtesy of Wikipedia
The Mechanical Turk was a chess-playing machine constructed in the late 18th century, and exhibited from 1770 for over 84 years, by various owners, as an automaton but later explained in January 1857 as an elaborate hoax. Constructed and unveiled in 1770 by Wolfgang von Kempelen (1734–1804) to impress the Empress Maria Theresa, the mechanism appeared to be able to play a strong game of chess against a human opponent, as well as perform the knight’s tour, a puzzle that requires the player to move a knight to occupy every square of a chessboard once and only once.
Publicly promoted as an automaton and given its common name based on its appearance, the Turk was in fact a mechanical illusion that allowed a human chess master hiding inside to operate the machine. With a skilled operator, the Turk won most of the games played during its demonstrations around Europe and the Americas for nearly 84 years until its destruction by fire in 1854, playing and defeating many challengers including statesmen such as Napoleon Bonaparte and Benjamin Franklin. Although many had suspected the hidden human operator, the hoax was formally revealed in a series of articles in The Chess Monthly in 1857.
Fascinating Adam…what’s that got to do with anything. Well in 2005 Amazon launched a web service and an ecosystem that would allows humans to help the machines of today to perform tasks they aren’t yet suited for, and named it…yup you guessed it…Mechanical Turk.
So as a developer I can write a program that poses tasks known as HITs (Human Intelligence Tasks) to ‘Workers’ who’ve signed up on the Amazon site. These might be choosing the best among several photographs of a storefront, writing product descriptions, or identifying performers on music CDs. Workers can browse among existing tasks and complete them for a monetary payment. You can see some sample HITs here.
The collective term that has arisen for this kind of activity is ‘crowdsourcing‘ and applications that use this model are cropping up with increased momentum and I’m finding them to be some of the most useful out there.
I won’t go into a ton of detail on each one but there are 3 I would highly recommend you check out yourself and see what you think.
Headquartered in Seattle, WA, Jott Networks operates a voice to text service that makes staying organized and in touch easy. Jott allows consumers to easily and safely send emails and text messages, set reminders, organize lists, and post to web services with their voice.
John Pollard (Formerly of Microsoft / Expedia) made some really smart decisions in creating Jott. They provide multiple ways for you to leave a message via voice (either by calling their free number and leaving a message or using a nifty mobile app) They first use voice to text techniques to do an initial translation and then if it isn’t quite up to scratch they have pre-screened employees who’ll take a second listen of the message and tidy up any errors the voice to text engine may have made. I love this service and have integrated it into my Getting Things Done (GTD) system so that if I have a thought while driving I can pull out the iphone, speak the note into the phone and have it instantly translated and emailed to the excellent Omnifocus application where it’s perfectly converted into a task for me to follow up on…(I know it’s not easy being me).
It’s the 20th century equivalent of having a personal assistant who I can call with a random thought and have it neatly filed and ready for me when I need it. Try it, you won’t regret it.
ChaCha is conversational, fun, and easy to use. Simply ask your question like you are talking to a smart friend and ChaCha’s advanced technology instantly routes it to the most knowledgeable person on that topic in our Guide community. Your answer is then returned to your phone as a text message within a few minutes.
ChaCha does exactly what it says, although your mileage on a good answer might vary. It’s basically a bunch of folks who have passed some basic vetting who monitor incoming questions that are texted in and google up an answer for you in order to get paid. So you could ask something like ‘How often should I baste the Xmas Turkey’ and expect to receive an accurate response. I have a friend who is completely addicted to answering questions just for the pure fun of it and you get paid a nominal sum for your effort as well! Check it out at www.chacha.com or text a free message now to 242242.
Beyond just having a quick, convenient way to search or purchase anything from the huge Amazon catalog, they’ve implemented a crowdsourcing feature named ‘Amazon Remembers’. The premise is similar to the SnapTell application I blogged about last time but uses crowdsourcing as opposed to machine vision algorithms to deliver a result as to what you photographed. It may not be as fast as Snaptell which delivers instantaneous results for most CD’s, Books, Videogames, etc. But if you take a picture of something outside of those categories you’ll likely receive a great match in less than 5 minutes, all courtesy of Debbie in Deluth sitting at home posing as an 18th century automaton.
WHY IS IT RELEVANT?
This is a trend that has really only just begun and is likely to lead to entirely new ‘knowledge worker’ industries and marketplaces. Cheap labor is coming online both abroad as well as the US. You don’t need a PhD to accurately transcribe a voicemail or recognize an object. Given the current economy it’s not too unlikely that we’ll see many people with a low cost laptop and a web connection interested in making a little extra by servicing these knowledge needs.
The lines between what a computer is or isn’t capable of will continue to be blurred. In the examples above the tasks are relatively straightforward and simple but think about the potential when simple and complex tasks by computers, generalists and specialists are brought together. In much the same way logic gates provide inputs, operations and outputs to create complex software, different HITs could be strung together to create very complex and impressive new applications.
One could imagine using an interface something like Yahoo Pipes or Microsoft Popfly, to visually build an application that uses real people and data from around the web to create one of these new classes of programs with different payment tiers for different skill sets.
How about an application where you snap a pic of a leaky roof or dripping pipe along with a shortaudio snippet of the problem. Unskilled labor could transcribe and categorize the problem and route it to skilled plumbers or roofers who would then bid on the job and respond with a quote.
Could you save lives in developing countries at a tiny cost by sending medical students pictures from a cellphone tagged, categorized and forwarded to a specialist who’s able to just focus on the diagnosis as opposed to the overhead? Not so far fetched in light of this story of a British Surgeon who received instructions on how to perform a tricky amputation from a colleague via text message last week.
As with all leaps in technology there are privacy, ethical and social issues to be considered but don’t be surprised if your dental x-rays are being examined in Belarus instead of Bellevue next time you get your checkup.