My guest for this week’s Interviews with Innovators is John Buckman, a serial entrepeneur with a passion for the world-changing possibilities of online communication. I was a customer of his first company, Lyris, whose email list manager I once deployed for a client. A few years later we met at SXSW, where John showed me his new project, Magnatune, and explained why he created it.
In this interview we discuss Lyris and Magnatune, plus some recent interests: BookMooch, an online book exchange, and Wreck a Movie, a “Web platform that is designed to harness the power of passionate Internet communities for creating short films, documentaries, music videos, Internet flicks, full length features, mobile films and more.”
This platform for crowdsourced film production, which is an outgrowth of the collaboration that led to Star Wreck, is now being used to create a new film called Iron Sky. You gotta love the tagline: “In 1945 the Nazis fled to the moon. In 2018 they are coming back.” As John says, who wouldn’t want to see Nazi invaders from the moon?
His ideas about how to do crowdsourcing closely parallel those of Yochai Benkler, as channeled by Bruce Sterling in his SXSW 2007 keynote, which includes a digression on this theme that begins thusly:
Socially motivated commons-based peer production: How to do it. You don’t just open up a website and invite comments. It actually has to be engineered with some thought and care. And I want to explain how this guy [Yochai Benkler] thinks it’s done. First you have to divvy up the work, because there’s a lot of it, and nobody wants to do it, and you’re not paying them, and you’re not ordering them, because you’re not the market and you’re not the state, so you can’t pay them and you can’t draft them. So you have to divide it up. It has to be granular, modular, and integratable. Granular means that even if I contribute for 5 minutes, I’m going to contribute something of merit.
Software is part of the answer. You need a framework within which to define tasks, parcel them out, and gather back results. Amazon’s Mechanical Turk is the best-known commercial example, but there are noncommercial precedents like distributed proofreaders and LibriVox. It’ll be interesting to see how Wreck a Movie follows and extends those examples.
John affirms Yochai Benkler’s point about granularity, and stresses the importance of precise task definitions.
If you say, nebulously, “Help us get press,” then nobody does anything. If you very specifically say this is the message, post to these press outlets, and name three URLs, that’s a well-defined thing. You get much stronger response. People need direction.
Also, I read recently that membership in volunteer organizations is down, and has been for a decade, but the amount of time people put into volunteer tasks is up. Why? People are less interested in open-ended commitments, instead they’re interested in knocking off specific tasks, and feeling a sense of accomplishment for finishing them.
As Bruce Sterling says, you have to engineer that scenario with thought and care. It’s a kind of engineering that is both social and technical, and some of the leading practitioners are game designers. For more on this theme, listen to Ned Gulley’s reflections on the MATLAB programming contest he’s been running for a number of years.