Crowdsourcing thoughts

Standard

On Wednesday, I’m attending Remixology 2, an event put together by Fresh Media, on the topic of crowdsourcing.  In particular, I’ll somehow be the representative of the entire open web perspective on crowdsourcing (!), Alfred Hermida will be talking about the journalist’s perspective, and Leigh Christie will be there representing artists.  I’m hoping that the audience doesn’t expect any one of us to speak authoritatively on any topic, and that we can instead have a conversation.  Since talking to Hanna Cho about the event, I’ve had a couple of thoughts on the topic that I’m hoping I’ll be able to fit into that conversation.

Crowdsourcing, like most buzzwords, is loaded with too many meanings, and I rarely use it.  I’m more interested in figuring out how to leverage the internet to enable collaboration on a grand scale.  Everyone has experience with 1-1 collaboration, whether through email, shared writings spaces, voice calls, etc.  The internet has provided the technologies to make such collaboration radically cheaper and faster than before, and the biggest challenge it has brought have been widely discussed: we’re always connected, for better or worse; we’re always interrupted; the world is smaller; nobody knows you’re a dog.  All of which is old hat to anyone who’se spent any time online in the last couple of decades.

The advent of mass instant collaboration and mass participation is made possible by the same technologies, but I think we’re still in the earliest stages of figuring out both how to do it well, and what the societal impact will be.  I’m hoping we can talk about that a bit.

It’s easier than ever to spread a meme, and to recruit a population the size of a small army who are all interested or even passionate about your meme.  With ubiquitous communication systems (phones, laptops, cheap broadband, internet cafes, etc), social “viral” media  (twitter, facebook, chain letters, etc.), rich media production models (video on phones and youtube), it seems that viral messages spread like wildfire (of course there’s a massive selection bias: deliberately starting a wildfire is incredibly hard in practice).  Let’s grant that getting the word out is easy.  Depending on the topic, one can get the attention of a cohort of like-minded folks fairly easy (that’s 500 soldiers, if the roman army is a guide).  If any one of them has an hour or two to contribute, pretty soon we’re talking a person-year or more of effort, which can be a potent resource if focused!

The cost of building and running web sites has also plumetted, and the number of people capable of doing so is skyrocketing, which makes it easy (in theory) for anyone to create a place for these people to gather, discuss, coordinate, work, agitate, whatever.  Some will build dedicated websites, others will use shared tools like Facebook groups, mailing lists, etc.  In most countries, such gatherings are undetected, let alone regulated.  We now have mechanisms for coordination of group action.  The potential is seemingly unbounded.

Many online activities are virtually free.  Interestingly, even when there are real (or forecasted) costs to a project, the last few years have seen the maturation of many interesting micropayment systems.  The trendiest is Kickstarter, which somehow gathered the mindshare in the “let’s get together and fund X” world, and its most famous success is Diaspora, who raised $200k, which was 20 times what they asked for, just because they said they’d take on Facebook.  So even in the treacherous arena of cash, there are now funding models which seem to work (at least for small-scale efforts).  Thus, to the sheer hours of invested time, we can now add a few thousand dollars.

So now we have a few hundred people, excited about some idea.  There’s a website, and even a modest bank balance. To use the techy jargon, we’ve got scalable models for meme propagation, recruitment, coordination & communication, advocacy, marketing & PR, and funraising.  Awesome.

Now it’s time to actually do stuff.  In particular, it’s time to plan, schedule, prioritize, make decisions, commit some code, commit to something.  In my experience, that’s the part that we still don’t know how to scale.  Everyone in the army of volunteers has ideas about what should be done (but only a small percentage will actually have relevant skills or experience).  Everyone will have opinions about what words should be used, but only a small number will actually really listen to the other’s opinions.  If we’re not careful, we now have a large group of people who think share a goal, but who are not organized.  And that can be really hard to deal with, especially given that we’ve made it really easy for them to shout at each other.

Which leads to my main point, which is that the next challenge for mass collaboration and coordination over the internet isn’t going to be technological, but human.  Specifically, what will differentiate important projects from the rest are the people who can help groups of people achieve common goals.  That’s not a new task, but the cybernetic setting will require to adapt old skills and create new cultural norms.  Three skills at least are needed to facilitate that kind of coordination:

The first is some form of leadership.  Quite often, the initiator of the meme didn’t really intend to start a micro-movement.  She just tweeted something, or uploaded a ranty video, or wrote a scathing blog post.  And all of a sudden she is the center of attention from a bunch of strangers who “agree” and want to “do something about it”.  In that kind of situation, converting emotional energy into effective action will (I claim) depend on the emergence of a leader of some kind.  Which doesn’t mean a spokesperson, or a dictator (benevolent or not).  It just means someone who, using whatever means are appropriate for that group, can get the group focused, moving in a roughly consistent direction towards some vague approximation of a common goal.  Different groups of people will respond to different types of leadership, but I’m pretty sure all large groups need at least one individual they can anchor to.

The second is organizing.  The style of organization needed will vary wildly depending on the group, from simply taking notes to gardening a wiki to tweeting a lot, nagging, proofreading, testing.  But there is a yin to the leadership yang, and the people who are good at getting people excited are rarely the same who can remind them to uphold their commitments.

The third is what my friend David Eaves refers to as negotiation, or the process of seeking common interests among a set of potential collaborators, and building commitments and mutual trust along the way.  This skill is rarely explicitly discussed in many organizations, because most organizations have built-in power structures which have well understood tie-breakers (“the senior person decides”, “the client decides”) as well as clear consequences to disagreement (“you’re fired/demoted/etc.”, “this contract isn’t renewed”, “you’re not invited next time”, etc.).  Neither of these are as clear in a setting where peering and fraternity are assumed over hierarchy and management.  If I show up at your virtual event expecting to be treated like a peer, but it so happens that I misunderstood what your goal was, the odds are pretty good that one of us will frustrate or disappoint the other.  If we both care about our own visions, the odds of a flame war are high.  To avoid that, we need to clarify the goals up front and review them often.  We need to really explore everyone’s interests and both detect overlap and explore differences.  And we need to keep in mind everyone’s BATNA.  It’s work, but it’s the only way to actually draw from everyone’s strengths.  I think the open source / open web world is still a beginner in this arena, but I’m glad that we’re working on those muscles.

Of course, the technologist and UX thinker in me is keen to figure out whether we can design systems that help with these all-to-human (and all-too-fragile) tasks, build digital prostheses of a sort.  You can see baby steps emerging among the more “social” web apps of the day: the indicators of mood on support forums like getsatisfaction.com for example, let people emote quietly, and provide non-verbal cues to emotional state, which are all to often lost in textual communications.  Building interfaces that surface the people behind the comments leads, I think, to more humane conversations (one of Facebook’s brilliant early moves was to encourage/require “real names, real photos”).  There are also simple tricks: at Mozilla, we’ve also found that if one detects conflict, it’s usually a good idea to try and resolve it using private voice calls rather than prolonged, public, painful email discussions.

I’m sure that by Wednesday I’ll have other thoughts in my head which will push these out of the way, but I’m curious to see whether these thoughts resonate with people in other disciplines, or whether different cultures lead to radically different world views.

hotel recommendation in NYC?

Standard

I’m going over to Manhattan for the IORG meeting. I’d love a recommendation for a place to stay. My favorite hotels are small boutique places — I have a strong dislike for the huge chains, and I’ll pick a safe hostel over a hilton. The conference is midtown (30 W 44th Street), so anywhere within walking or cab distance from there would be ideal.

FOSSCoach, OSCON

Standard

At the last eLiberatica, I was talking to some of the speakers, and several of us reflected that while we really enjoyed giving the standard, “speak up and monopolize everyone’s attention for 20 minutes, then take 5 minutes of questions”, given the energy in the crowd, we’d be really keen, in general, to have less cathedral and more bazaar at open source conferences as a rule.

OSCON is coming up, and of the mainstream conferences it’s been quite friendly to the less commercial “unconferences”, while still being a very worthwhile event in its own, conferency way. A few points to note:

  • Dan Mosedale and I will be giving a talk about Thunderbird, and we expect you all to be there, and to walk away with confidence in the future of Thunderbird, because you will also walk away with some action items!
  • I’m on the program committee for OSCON, so I get to share this 15% off discount code on my blog. How cool is that? os08pgm is the word of the day.
  • Zak Greant is leading an unconference at OSCON on the topic of how do we transfer knowledge about community building and open source, just before we turn senile. I’ve told him I’d help lead a discussion, but I’m not sure what topic to cover. Anyway, it’s free, it’ll be fun, and you’ll get a chance to ask a bunch of experienced people

Romania ahead

Standard

I’ll be heading out to Romania next week, to talk in Bucharest at eLiberatica about Mozilla and open source, and to learn about everything from open source in eastern Europe to food (always). I’ll also be visiting my grandfather’s hometown (Braşov) for a bit of personal root-digging. Should be fun, especially if I get rid of this grogging (new word!) cold I’ve just picked up.

As a side note: it’s interesting to see that an open source conference is now sponsored by everyone from the FSF Europe to Microsoft. What is the world coming to?

Computers, sigh…

Standard

Sigh. Things aren’t quite right in the computer department. Thunderbird is using 170,394Kb of memory after running for 10 minutes. My hard drive is flaking out while at a conference. VPN doesn’t make it out of the conference LAN. Luckily, people are more reliable, friends are still friends and chefs can still cook.

PyCon

Standard

First day at PyCon 2005. It’s, as usual, interesting. Random bits:

  • Crowded! It’s bigger than ever, clocking in over 400. It’s caused some headaches of the good kind (catering more expensive than planned, not enough t-shirts, rooms are packed).
  • Not too surprisingly given the buzz around Python, there are big names (although we’ve had big names at Python conferences for years). This year it’s Jim Hugunin from Microsoft, Greg Stein from Google, and we’ll see who else).
  • More interesting to me, a lot of old friends, including many who used to come at conferences, then stopped, and are now back. I don’t know if it has anything to do with Python itself, the economy in general, or it’s just random — but it’s nice to touch base again.
  • It’s only been three hours, so it’s a bit hard to know for sure yet, but it seems as though there’s more money around — more startups, some VC influences, and a greater proportion of people who do Python for their jobs, not just for love

Washington DC is still a great city, the Rouge Hotel still has attitude and free broadband, and I’m looking forward to good dinners with old friends.

Python and OpenGL on Nokia Phones

Standard

Sitting here at ETech, just after Erik Smartt, product manager for the Python on Nokia product, gave the first real public demo of the Symbian/Series 60 port of Python. The highlight was a PyOpenGL demo (the code isn’t available yet, and the author is not public either). Click on the picture for the Quicktime movie.

As someone who worked on PyOpenGL a long, long time ago on high end Silicon Graphics workstations, it’s fairly stunning to see it fit into a pocket.

Also notable was a demo of the FlashLite stuff, which is equally cool, but probably available elsewhere (if not, I’ll post a movie later).

Cal Henderson Quotes

Standard

Cal Henderson, from flickr, at the PHPWest conference:

“Avoid code reuse (just a nice typo)”

and, talking about phpxpath:

“phpxpath is great. It’s on Sourceforge, but it’s still great”

and about UTF8:

“Can’t remember what UTF-8 stands for. Something something something eight!”

a bit that made Rasmus and me feel old:

“I think UUencode stands for unix to unix encode — it’s pretty old” (it’s true, but …)

About users:

“We have a lot of Mac users, because OS X users tend to be early adopters and have a lot of free time”

And then, when Cal said that PHP isn’t great for a daemon, Rasmus’ answer was:

“Write it in C, you sissy!”

Overall, a nice, contentful presentation from the guy responsible for the PHP (and, it appears, a lot of the DB work) code behind flickr, with good lessons learned on how to scale a web app to accomodate massive growth. I look forward to the ETech presentation.

Lightweight Languages Workshop in Boston

Standard

The program for the lightweight languages workshop is set (just in time, too — it’s on Dec 4!). I got to read the abstracts as I’m on the PC, and I wish I could be there to hear the talks. Go there and blog about it please! I’m particularly interested in the “Debugging without Programming” and “Frink” talks, but a lot of the conversations are bound to be a lot of fun.