Ch 6: Big Data and Tiny Tasks
Crowdsourcing As a Shortcut to the Technological Singularity (Free E-book)
Next Chapter… (How to Amp Up Crowdsourcing with AR and VR)
Let’s start with a short story:
How Big Data Took You From Programmer to Scrapper
You’re a programmer, and you like your job. But, you always liked your hobby more. Your hobby is collecting scrap-metal.
Every trash-day, early in the morning, you drive around your neighborhood, and you throw whatever discarded appliances or other metallic things you find into the back of your truck. Then you take it all home, disassemble the stuff, and separate the different metals (copper, brass, aluminum, motors, etc.), and then you turn them into a scrap metal yard for the $.
One-day, you’re at work programming an app that uses Google Street-View via the Google Maps API (application programming interface), and you have a thought:
What if you could somehow get all the visual-data you need to scrap effectively, i.e., a view like Google-streetview, but have it show exactly what’s on your neighbors’ curbs on trash day in close to real-time? (That would make you an incredible scrapper!) …Then, you realize, even if you could do that by using a car with a sophisticated camera-rig like Google’s (and by paying someone to drive quickly around your neighborhood), you’d still have to manually look through all the images, which would take you all day.
But, you don’t give up on that thought just yet…You give yourself a moment to ponder the problem.
That’s when you remember the concepts of ‘crowdsourcing’, distributed labor, and ‘microtasking’. You could PAY people (very little $, btw) to look through the pages/images for you in parallel.
That means, within a few hours or even minutes at the beginning of trash-day, you could start generating an exact map that leads only to the places where there is scrap-metal to find. Then, you just have to drive-along the route and follow the map, never going anywhere there isn’t the evidence of good scrap.
Over time, too, you might not even need to send out the camera-mobile each time, you could just use predictive-algorithms, instead. Those could create an intelligent map of where there is most likely to be scrap that day based on what it’s learned from all the data it has collected so far.
Dang! This could be both fun and lucrative.
Hmmm, now you have to figure-out how to make one of those Google-Camera-Mobile thingies next…
What is Big-Data?
For a brief definition, big-data is the resulting, left-over, pile of all digitally-recorded activity in our world. In other words, it’s all that stored information that we don’t always think about. This data is in the form of social-media posts, text, dates/times, numbers, video-footage, pictures, and factoids. It’s all stuff that needs to be sorted so that it can give us information to help us with fact based decision making.
We Use Big-Data in a Number of Interesting Ways:
We use it to find a needle in a haystack… whether that means tracking criminals on the run via phone-call records, or finding crashed planes lost at sea by sorting through satellite-imagery.
We can also use it to classify groups of items in sets of data, such as with a customer groups/types that respond to a certain marketing message, etc.
Lastly, we can use it to find correlations. This is the kind of thing that fuels recommendation-engines when they recommend you products or movies to watch, etc.
Big-Data, Big-Business, and People in Their Underwear Doing Tiny Tasks Online
It’s no secret. It has been the nudge of big-business and big-data that has really driven microtasking forward. The budgets of retail, technology, and information services companies have been the real impetus behind making online-work a substantial marketplace.
It shouldn’t be a surprise either. Who else is motivated or skilled enough to organize billions of tasks for the crowd than big companies trying to make sense out of their vast data-stores?
Your Uncle’s bowling-club probably doesn’t have thousands of product descriptions that need created by the end of the week. Do they?
So, without the intelligence, the budget, and the staff available to take problems, break them up into little-chunks, and make them available online, average people haven’t really jumped into the space.
As a result, however, the benefits of microtasking have been funneled mostly into helping the status-quo of big-businesses getting bigger by becoming more efficient, better informed, etc.
A Raw Deal?
In contrast, what do all these crowd-laborers get from the deal that is microtasking? At best they get a flexible source of mini-income, at worst they get drudgery and low-pay on a platform that kinda treats them like a machine (while avoiding paying them minimum wage in the process).
This is a problem that has made the microtasking field a controversial space, and it is for that reason that I have always thought microtasking should take a different direction, and big-data can play a role in that.
We Can Do Better: Microtasking and Big-Data for Social-Good
There a number of ways to microtask for a good cause. Some of these have already been done, and others have not:
- We can ask people to complete microtasks that raise money for things they care about, their favorite charities, political-campaigns, their friends’ crowdfunding goals as we saw in an earlier chapter, or even for things to be built in their hometown (e.g., parks and libraries).
- People can be engaged in games that are actually simulations of real problems from scientists and others that help them crunch-and-produce data that helps cure diseases or improves people’s lives in other ways.
- Microtaskers could help out in disasters to find critical information and answer important questions quickly.
- Microtaskers can also work to help train ‘benevolent AI’. Many algorithms now-a-days benefit by watching and learning (machine-learning) from the ways humans get things done. An example could be a AI bot that is trained to scan satellite-imagery to view changes in climate, traffic, and rooftops to predict where aid is needed most at any given time.
In the end, crowdsourcing and big-data are growing-up alongside each other. I hope that this relationship helps the workers as much as it does large organizations. Since this chapter is brief and the topics big, I highly recommend reading more about big-data and microtasking for a deeper understanding.
Table of Contents: