The Impacts of Civic Technology conference aims to bring together researchers and practicioners to qualify and quantify the impacts of modern technology in the civic and political sphere. The conference was hosted by nonprofit group mySociety, and saw a number of high profile presenters including Facebook and Google alongside several smaller organizations, both technical and nontechnical.
It seems clear in the spring of 2018 that the use and misuse of technical platforms like Facebook can have incredibly widespread negative impacts on society and democracy. At TICTec I was expecting to see Silicon Valley skewered for their detached arrogance and carelessness, a call for decentralization of power and separation of our online and political lives. But what I found was an overwhelmingly optimistic tone among the presenters and attendees - a strong resolve that technical solutions still had their place in making modern democracy more effective around the world - and especially solutions led by the largest tech companies in Silicon Valley, who have the greatest reach and power.
A (highly abridged) look back might help set this perhaps surprising attitude in context. For a long time both the mainstream media and the (western) Left saw the networked, extra-governmental nature of social media as having massive potential for positive social change. The favorite example is the Arab Spring, in which Facebook and Twitter played a role hailed as an essential vehicle for the sparking of a revolutionary wave across North Africa and the Middle East. Meanwhile in stable democracies projects like Politifact, change.org, and Facebook’s voter registration efforts seemed poised to bring a new era to western democracy - the same process that revolutionized the way we interact with our financial institutions would bring the speed and personalization that only the internet could provide to our political institutions as well. Of course, we now see that the civic tech world was a touch too optimistic - and that these platforms bring great potential for negative as well as positive change.
So the question is then for me, and for everyone else at this conference - what now? Silicon Valley created a world in which data leakages, fake accounts, targeted political ads, and fake news swing elections and cause real world harm. While it’s ever so tempting to denounce Facebook and Google, throw my smartphone away and spend all my time climbing rocks in the desert, another part of me resonates with Justin Wang’s suggestion that for a tech worker there is another path - to engage constructively with the problems of this industry head on.
This is the bargain that I suspect many at TICTeC have figured for themselves, and why for me it seems so important to figure out just exactly how much we can actually fix, and how much is fundamentally, structurally compromised.
Some highlights and quick takes
Facebook gave a presentation exploring how AI models dramatically increased the amount of inflammatory content flagged and removed in the run up to the 2017 Kenyan Election, how it’s planning on debuting a political ad archive to make targeted political ads less effective, how different UI features increase women’s comfort participating in online civic discourse, and initial steps in measuring political polarization.
Facebook has lately been very excited about the possibilities for AI to complete the superhuman task of categorizing and flagging dangerous content as quickly as it gets produced. The Kenyan case study seems to warrant celebration, but at a high level I am extremely skeptical that AI will be able to handle this problem easily in a general sense. You still need training data, and that means you need cultural context that is not only up to date but accurate on an extremely granular level. This is still a very human job that is going to be very difficult to scale across a global community.
The political ad archive is an unambiguous step forward. However, I do have a few quibbles on details which bubble up to some higher level questions. Details include questions like what are the criteria for an advertiser to be excluded from the list (Facebook cites sensitive minority groups as one example), what criteria there are for ads to be considered political, and so forth which really sum up to a bigger question on transparency. How far is Facebook willing to go seeing as how transparency in advertising seems to be directly opposed to their financial success? If we find out that the public ad archive does nothing or very little to decrease the effectiveness of the ads it’s basically undebateable that it had any effect at all. I’ll be curious to see how this experiment goes…
Martha Lane Fox, of the UK House of Lords delivered a great talk on the idea of responsible technology. While light on implementation details, the comparison to free trade clothing and food items is an interesting one. I’m curious if the project will be able to develop a cohesive set of norms surrounding what makes a piece of technology ethically designed. It seems to me that the raw power of tech to influence society so quickly and at such a large scale means that a slow proliferation of new, ethical services won’t be quite enough to turn the current tide against the issues facing the Big 4.
No one has come up with a good application for blockchain yet
DesignFix described a product model that delivered a 19% increase in self reported political efficacy in a small pilot in Northern Ireland. The key finding is aggregating citizen’s messages into a manageable amount of concerns for a representative to handle, and finding a way to complete the communication loop even if the representative chooses not to directly address the citizen’s concerns.
Many solutions are still country, if not region specific
Sign up for the mailing list