Turn your personal mob into an army

The startup team behind Human.io.

Human.io is the new thing from Joshua Schachter, founder of bookmarking site del.icio.us. This time, however, he's not suggesting you share your travels with a few friends—he's suggesting that you turn them into an army.

"If you want to build a flash mob, but have it actually do something useful, this is your API," Schachter said. "It lets you invite your audience to become part of the action."

The concept—developed by Paul Rademacher, creator of legendary Craigslist/Google Maps mashup Housingmaps, and Nick Nguyen, formerly of Yahoo and Mozilla—is straightforward enough: Human.io is a platform for performing "micro-tasks".

First, you publish a simple, crowdsourceble activity, such as voting on something, going to a particular location, or taking photos—anything that might be accomplished with a smartphone's UI and its sensors. Then you tell your readers, followers or friends about it. They start the app, get cracking, and, finally, the results are sent back to you.

Human.io can be scripted in Python and PHP, languages easy enough for laypeople to create basic tasks, but powerful enough to set up more complex and rewarding interactions. Writes Schachter: "missions and activities to get people involved more directly than just reading stuff on a screen."

To illustrate how the platform works, Human.io developed an app for us aimed at benefiting the Creative Commons, and Wiki Loves Monuments in particular: wikipedia.human.io.

The idea is to help Wikipedia's project to improve public access to photography of the world's architectural and local heritage. If you want to participate, install the free Human.io app (iOS, Android) and select the "Photograph a historic place" task. It'll cough up a list of anything nearby that's in the online encyclopedia's monument hit list. All you have to do is head out, take a shot, and let Human.io do the rest. It'll show up immediately at wikipedia.human.io, released under the Creative Commons Attribution 3.0 Unported License.

It bridges the gap between a useful task--in this case, contributing to Wikipedia's potential coverage of your neighborhood's history--and folks who might not otherwise be in a position to help out, let alone have fun doing so.

Another possibility that comes to mind—that I might try and create myself if I have time—is a "reporter" task. Twitter and Facebook have become essential to newsgathering, but these platforms weren't made with journalism in mind. They're not so hot at helping editors and newsrooms organize the incoming torrent of info, correctly attribute it to named sources, or separate public panic from personal presence.

My hypothetical Human.io news-gathering app would have the user add short but essential notes to eyewitness reports, snapshots and recordings: text fields for Who, What, and Why, journalism's traditional formula for basic reportage (When and Where would automatically be taken care of by sensors). With the results flowing instantaneously to the web, this could be powerful tool-- and what I like about Human.io is that it makes it possible for an amateur coder such as myself to make it happen. (But if you build it, get in touch and I'll be happy to cheerlead instead.)

Unfortunately, there are shortcomings in Human.io as it stands. The app is rough-hewn and, while new activities are highlighted, one must know a shared ID number to add others (Update: the IDs are just for private tasks--all public tasks are shown). While its simplicity is a big plus, better discovery would make it easier to generate interest among casual users—folks who might just want to dip in and do something interesting that helps someone out.

Also, I'd like to see options that make it easier for task creators and users to specific and agree to licensing terms— unless otherwise specified (as in the Wiki Monuments task) work done with Human.io becomes the property of the task creator.

Schachter says that basic in-app social networking features are on the way. It would be neat for users to be able to "see" and interact with one another and share activities, and for publishers to be able to offer rewards that are visible in a new social space created by the app—though I wonder if it would add an unnecessary level of complexity to the experience (and its maintenance).

There's all sorts of things that Boing Boing readers have done—from benefiting charities to participating in serious political action—that could have been easier for us to help organize were something like Human.io as widely-used as, say, Twitter. Apart from the Wiki for Monuments task, do you have any ideas?


  1. We’re not approaching the singularity, but we do seem to be approaching Maneki Neko.

    “Maneki Neko” is an upbeat, funny story that portrays a logical extension of the late, lamented “gift economy” upon which the Internet was built. The central principle was that if people contributed what they could to the system for free, everyone would wind up better off.

    In Sterling’s near-future Japan, Tsuyoshi Shimizu is one of many people who follow the prompts of the ubiquitous network. When the net tells Tsuyoshi to give a weary stranger a coffee, or buy a bottle of aftershave, he does it without worrying about the reason; in return, anonymous packages arrive with useful gifts such as baby clothes and pickles for his pregnant wife. Strangers on network business identify themselves to one another via hand-signals, but different regions of the net have different signal dialects.

    “Maneki Neko” runs counter to every paranoid story ever written about computers taking over our lives. Computers have made people anonymous, but not impersonal. Even those who try to fight the system benefit from it. Trust the computer; the computer really is your friend. (source)

    “Well,” Tsuyoshi said gently, “maybe my economy is better than your economy.”

    1. And don’t forget Charlie Stross and his great idea for outsourcing intelligence gathering and espionage through VR-based Alternate Reality Games….

    2. I can think of an objection to that one.

      Infrastructure, and the problem of ‘foundational’ bad actors
      The system only works as described in the story through the mediation of semi-aware programs designed for networking use. While it’s a non-trivial problem to design such a network, if it can be designed— what’s to stop the network from being perverted, and used the other way?
      Suppose Person A is told to bring a pair of shears to a certain place. Then suppose Person B is told to take the shears and cut the brake cable of a specific bicycle that parks in that location. Person C is instructed to obstruct the owner of that bike in traffic. Because of this, Person D loses control of their bike and is injured, or worse.

      Who’s responsible for Person D’s injury or death? Persons A, B, C? Maybe. But did they know what was going on? Can you say that the real killer was the network? Does that mean the network is self-aware?

      Be careful, granting self-awareness to things. I know this is a horrible way to see the world, but sooner or later, people will try to weaponize ANYTHING. :(

      1.  I’m guessing that in a mature network there would be a certain reputational component involved, and that consequences of actions are [somehow?!!] involved in the algorithm (hey, I’m making hypothetical suppositions about the future. It’s made of  positronic computronium and approved by the Geneva Heat).

        Another arguement against the system being completely perverted, is that everybody else in the system _doesn’t want_ it to be perverted. Or a significant number don’t — hence, Wikipedia vandalism is stopped and cleaned up.

        Which is, admittedly, text edits, and not cut brake-cables.

        I suppose I’m thinking of a system where the requests and follow-throughs cannot be private, so that [agents interested-parties home-bounds] that watch for [things like cut-brakes] will start taking an interest in the vicinity.

      2. I haven’t read the story yet, but from the outline, it sounds as if individual actions are voluntary. Presumably in a real application of this idea — like human.io — there would be some noise in the system, as people make accidental or ill-considered requests. People are quite used to filtering out noise of that kind

        So I would expect that there would be the occasional prank, but things worse than pranks would be rare. And in a world in which one can get almost anything they want just by asking politely, what’s the motivation for violence?

    3. Reminds me of the Japanese treatment of robots. We ride in them to fight monsters or they are our friends, helpers, companions, even lovers… 
      In America, they run, eat meat, and attempt to slay all of mankind.

  2. This reminds me very much of the art/game/happening site SF0. Funnifying of seemingly random actions, but with a covert purpose, if only to gain surreal “points” over other participating “teams”.
    I’m not a member, but I seriously considered joining at one time.


  3. ….and countdown to 4chan discovering it and using it in a horrible yet ironic way that was never intended….

    “And what rough app, its hour come round at last, Slouches towards Youtube to be born?

  4. “Hey, could you check on….”

    – get a current photo of the house you grew up in, several states away [uh, obvious stalker implications, here]
    – have someone ring your {grandparents’, mothers’, great-aunt’s, reclusive-coder-friend-who-forgets-to-eat’s} doorbell and record a quick conversation about their state of well-being.
    – see if my car keys are under that bench
    – I will paypal you $1 if you give a dollar to that street-performer I’ve been watching via the traffic-cams all week

  5. How could 350.org use this?  Monitoring short-term climate forces like black carbon, diesel exhausts, and methane leaks to reduce particulates and tropospheric ozone precursors and slow their global warming effects?  Survey all the buildings in a city and begin to prioritize the ones that waste the most energy and need the most retrofit?  Streamlining traffic flows?

Comments are closed.