In my job, I often hear the sentence “technology is not the problem”. I disagree. We mostly say it because we don’t understand technology well enough to detect its limits – it’s a kind of black box.
In the Crossroad project, we are looking at future applications and technologies for collaborative governance and policy making. In this process, a few ideas came to me.
My argument is that all the collaborative tools that we have now, from idea-rating to mash-up and debating tools, work well only at small scale.
We all have problems in leveraging participation: yet we would not be able to deal with large-scale participation.
Most of Obama administration initiatives are very basic in terms of functionalities and performance (ideascale, gModerator, Innovation Jam…). They rely on strong human effort, which is difficult to scale up. And the scalable tools, like quantitative ratings and e-petitions, are very easily manipulated.
Basically, we don’t have the tools for mass conversation and collaboration, so it’s actually good that few people participate!
A few examples:
– the Global Pulse 2010 used IBM Innovation Jam, which as far as I can see is a large-scale forum with sophisticated human-analysis capacities and methodologies, but basic technology (only robust enough to handle thousands of participants
– the Google Moderator tool used by the White House uses basic ratings systems and is prone to hijacking, as the marijuana debate proved
– the recent Commencement Challenge allowed citizens to vote for the best essay and video
– any form of Google Mash-up works well only with small numbers, as we realized when setting up the debate space for the Amsterdam Declaration
– using ideascale or uservoice, as we did in the Open Declaration, people only read and comment on the top proposal, or come out with a new proposal of their own. Free text search is provided by Uservoice, but it’s a very rough method. Nobody reads all the proposals before posting and voting.
– PatientOpinion deals with high human-intensive processing of comments, one by one.
Obviously, I am not criticizing the initiatives. They are great. But “technology isn’t there”.
So my question is: what would be future transparency tech? what applications can we envisage 10 years from now? What basic tools need to be developed? Is RDFa the most futuristic thing we can come out with? Is there anything after things like digress.it, developed within the US e-Rulemaking research programme?
For me, most relevant research field are:
– collaborative filtering
– reputation-management systems
– visual analytics
– natural language processing
May 5, 2010 at 2:23 pm
A good clarification comes from the IBM Innovation Jam paper: http://wwwdim.uqac.ca/~attrembl/8INF848/Cases_HBRP/SMR291-PDF-ENG.pdf
“Senior executives and others spent weeks sifting through tens of thousands of postings — gigabytes of often aimless Jam conversation.”
May 5, 2010 at 2:27 pm
great insight from the same paper: “Technology for “promoting” new ideas (like Salesforce Ideas and Digg) could have promoted more coherent dialogue, but with the danger that many good ideas might be lost.”
June 3, 2010 at 2:04 pm
this is exactly predicting the problem DIGG has now: see http://www.guardian.co.uk/technology/blog/2010/jun/03/digg-dead-falling-visitors
May 5, 2010 at 2:32 pm
So the innovation jam actually makes use of “sophisticated text mining software” for analysing comments and input
May 5, 2010 at 2:35 pm
“IBM has explored using artificial intelligence techniques to reduce these errors, but they did not improve accuracy”
May 5, 2010 at 5:15 pm
David – I agree with you that we don’t currently have the tools particularly in relation to capturing people’s opinions in an aggregated form. However, for me the biggest issue is: how do you make citizen input useable? Or perhaps better: how do you redesign the policy process so that citizen input can easily and usefully be made at all points in it? I think this is more difficult and a higher priority than a new tool for capturing what people think (not that I am suggesting that this is necessarily what you are arguing for).
May 5, 2010 at 9:23 pm
Hmm. You might want to think a bit about what the optimum is. “Gaming the system”, for example, happens when the group who most wants something (like getting a marijuana question on top of a given list) gets it, paying in money, time, and effort. Economists call that “allocative efficiency”, and think it’s great.
In order to design technology that improves the current situation, we would need to understand what we would like to happen. You seem to have in mind tens or hundred of thousands of people discussing issues (I imagine big ones), whereas, say, Beth Novek wants to mobilize “microelites” (the five people who really understand in depth a very specific matter). The case for microelites to me is clear. Would 200K people discussing online, for example, the bailout of Greece be a good thing? In what sense?
May 28, 2010 at 10:36 am
alberto, if we accept your allocative efficiency arguments, then we could refuse all kind of tech advance: street demonstration would then be effective because they select only really interested people. On top of this, they are inclusive as they allow also non-digital savvy people to participate.
May 28, 2010 at 10:57 am
Sorry, I don’t understand the argument. 😦
May 6, 2010 at 7:44 am
Paul, I agree although technology and processes are mutually dependent, so the availability of technology will affect the design of the process, and viceversa. A good example is the eRulemaking research programme in the US – fully embedded in the eRulemaking process.
Alberto, even in the case of microelites we lack the tools. How do you recognize elites? You need interoperable reputation management systems. How do you allow new elites to emerge? You need collaborative filtering, tools for divergent thinking, etc. Peertopatent is pretty rough, people can vote on each other proposal but my experience is this is too time consuming once you have more than 10 proposals. And reputation is only internal to Peertopatent, you cannot export it.
THANKS FOR THE GOOD POINTS!
May 6, 2010 at 8:01 am
Alberto, in fact when dealing with microelites the key issue is finding the experts and the expertise. This is a key functionalities of web2.0 tools by the way. But we need much better and interoperable reputation management system coupled with social networking tools.
May 6, 2010 at 10:11 pm
David, I think information propagates quite well through the Internet (actually it’s a little more than “I think”: the case is too long to argue here, but there are mathematical grounds for saying that in today’s Pageranked Internet relevant information reaches relevant people). Microelites are reached in an almost uncanny way – which is how you, Paul and I met, after all. Then the process is solid in probability, but messy at the individual level: some people have the right skills and expertise but they just are not that interested in participating.
I also disagree with reputation being non-exportable. We all have “digital reputations” that carry across the different environments in which we participate. Which is yet more evidence confirming the former point about effective information diffusion, I guess.
I am all for better tools. But we do have an important piece of technology, which bootstrapped a lot of social dynamics: Pagerank. Socialgraph is another stab at the same thing.
May 7, 2010 at 9:37 am
alberto I forgot to say I agree that hijacking by interested people is in itself a filtering mechanism – but I also think this is too rough to be valid.
When I talk about upscaling, i do not mean only about large amounts of people: I mean can we use gov20 for any government related activity, even at micro-level. could I trust users’ opinion for getting my passport, or could government use it for all its decisions? at the moment, the “many eyes” approach works only for eye-catching ideas.
May 7, 2010 at 9:43 am
David: my vision (outlined in a forthcoming book as you know) is precisely this: many issues, large and small, competing for eyeballs. This does not mean every issue will GET eyeballs: attention will always be a scarce resource. But that’s no bug in e-gov2, it’s prettymuch a law of nature. In fact, David Lane thinks “innovation is a Ponzi scheme”, because the optimal scale of innovative clusters goes to infinity in a finite time. I’m sure we can talk about it in more details
October 1, 2014 at 12:03 am
A fascinating discussion is definitely worth comment. I
believe that you should publish more about this subject matter, it might
not be a taboo matter but usually folks don’t discuss such subjects.
To the next! Cheers!!
December 6, 2016 at 1:03 pm
I read this article completely about the resemblance of most
up-to-date and preceding technologies, it’s remarkable article.