HCI and the U.S. Presidential Election: A Few Thoughts on a Research Agenda

Many in the human-computer interaction community are feeling a sense of despondency after Tuesday’s U.S. presidential election. It’s certainly a sentiment I share, especially with respect to what the results may mean for racial and ethnic minorities, the environment, and democratic norms. However, this often pessimistic fellow has an unusually optimistic perspective to offer: I believe that this election affords the HCI community a tremendous opportunity — and a tremendous responsibility — to make the world a better place.

More specifically, if we look at the campaign and the issues it has raised through an HCI lens, a substantial research agenda emerges. Below, I’ve outlined what I believe to be a portion of this agenda. I’m a relatively junior member of our community, and the election results are still sinking in, so please view this note as the beginning of a discussion, not the end of one. That said, I’m hopeful we can take action on some of these items quickly.

#1: Take the economic implications of technology more seriously

This is a big one. To be blunt: those of us in computing are huge winners in today’s technology-infused economy, and we’re not doing enough to help the folks who aren’t winning. Economists have been sounding loud alarm bells about the economic pain that is being caused by the technologies we build. This pain almost certainly played a role in the events of this past week.

HCI is about improving relationships between humans and technology. As such, the displacement of people from their sources of income due to technological change is, at its core, an HCI problem. In fact, one could argue that this may be The HCI Problem of Our Time.

So, let’s saddle up and try to solve this problem! More specifically, I’d like to pose the following research challenge: Let’s use the next four years to build sociotechnical tools that ensure a more equitable distribution of the economic benefits of technology.

This is going to be hard, but hard is fun.

Where do we start? Well, a major new research agenda of mine is to help people (a) understand the economic value of their data and (b) build tools to facilitate collective action around that value. I invite my friends and colleagues to join me in these endeavors.

Of course, this isn’t the only approach to “equitable tech”. To spitball a bit here:

  • In our ongoing discussions about how to design crowdsourcing markets in which “we would want our children to participate”, do much more work on improving wages and benefits.
  • Do more research to economically support workers in the sharing economy, especially in the face of increasing automation.
  • Work to develop predictive models about who is going to be helped and who is going to be hurt by a given technology, and alter research incentive structures accordingly.

#2: More action, less description

In the HCI community, we have done a wonderful job of observation and critique with respect to many interactions between technology and society. Observation and critique are, of course, very important. However, they need to be followed by action, and I think we’ve been quite weak on that front.

My view is that researchers that have observed key societal problems in which technology plays a role — e.g. algorithmic inequality — must also work toward solutions. This is the responsibility of ALL of us who work in this space (yours truly included). The good news is that as technologists, we are equipped with the tools and the skill sets to quickly take action at a substantial scale. If building systems isn’t in one’s wheelhouse, there are a number of equally important means by which one can work toward solutions, e.g. developing policy and supporting protest.

Here are a few concrete steps we can take immediately to move toward the goal of more action-oriented research in HCI:

  • Let’s double down on our efforts to elevate systems research, especially for systems that do verifiable social good.
  • Let’s find ways to incentivize the maintenance of systems that could have real positive impact.
  • Let’s treat implications for policy as coequal to implications for design and policy work (e.g. drafting legislation) as coequal to system building.

#3: Build technologies to improve our democratic infrastructure

Modern advances in data collection, optimization, and communication allow bad actors to abuse age-old democratic processes and norms. For instance, there’s a strong argument that computer-assisted gerrymandering has contributed to much of the polarization behind our current predicament in the United States. To me, this leads to some pretty clear implications for HCI research and design:

  • Can we build technology to help people and/or lawmakers define and/or demand better electoral regions?
  • Can we accelerate research to support radically distributed workplaces such that employment in computing can have a more diverse geographic footprint?
  • Can we construct systems that help people make residential choices that minimize disparities in voting power? Can we design the correct incentive structures to actually enact change using such systems?

#4: Work to address ALL forms of suffering

I’ve noticed that we in the HCI community have a few blind spots in the socially conscious work that we do. Some of these blind spots are associated with real and substantial suffering, often times accelerated by trends in which computing technology plays a role.

For instance, rural America is in terribly bad shape, in part due to tech jobs being located almost exclusively in cities. For those of us in tech, it’s hard to imagine living in places where neighborhoods aren’t magically improving year after year. For the almost sixty million people who live in rural areas in the U.S., the opposite is occurring: economic decline is pervasive and is getting worse.

Reflecting trends in our larger industry, we in the HCI community are also guilty of being mostly ignorant of the plight in rural areas. For instance, almost all research into “location-aware” social technologies assumes urban population densities.

The good news is that we in the HCI community are well-positioned to ensure that technological progress and rural decline are not two sides of the same coin. Given the number of people that live in rural areas in the U.S. and around the world, I would like to see a lot more attention paid to the development of technologies designed to improve rural quality of life. One obvious direction is to find better ways for rural areas to participate in the tech economy (More attention to supporting low-bandwidth contexts? Well-paid crowdwork? The use of VR to support distributed work?). Other approaches might involve finding ways to build social capital in the face of declining populations or to use technology to increase social connections between urban and rural areas.

Helping rural areas is particularly important because the U.S. electoral system (and that of many other countries) gives people in rural areas more voting power than those of us who live in cities. This is the one plank of my soapbox on which I can claim to have done some work, but there’s so much more research to do (esp. on the action side).

#5: Build (radical) technologies to support civil liberties

The world may be entering a period with more social unrest than most of us in the tech industry have seen in our lifetimes (at least in “the West”). Moreover, history teaches us that the victims of this unrest will not be equally distributed among all communities. As a Jewish American, lessons from my own family’s history makes me particularly concerned for my fellow citizens of the Muslim faith.

So, let me pose this challenge: How can we improve upon existing technologies that have been used to fight against political violence? Can we go beyond basic reporting tools and build technologies that help people collect data that lawyers can use to put perpetrators behind bars? Can we build tools that directly increase the physical safety of potential targets? Can we re-double efforts to eliminate the vicious online harassment that was a hallmark of this election cycle?

Radical approaches may be necessary to achieve these goals. Bad political actors play dirty, and we need to be prepared for that.

#6: Work to ensure that HCI is used for good

“To the question, ‘Is atomic energy a force for good or for evil?’ I can only say, ‘As mankind wills it.’” – Leslie R. Groves (Director of the Manhattan Project)

An underlying assumption in most HCI research is that the more technology we have in our lives, the more society will benefit. This election cycle added heaps of evidence that if we’re actually going to reach a technological utopia, we’re going to have to earn it through disciplined, intentional effort. Put in very simple terms some of us might remember from our comic book days, we need to work harder to ensure that our technologies aid the forces of good more than they aid the forces of evil.

In this vein, the 2016 campaign revealed that some of the technologies most central to HCI are missing key rear-guard defenses. For example:

  • There is a reasonable hypothesis that “fake news” websites helped to spread false information to the millions of voters who are active Facebook users. Anecdotally, I saw a large number of fake news stories in my own Facebook feed. As my colleague Eytan Adar noted, if we expect that social media changes purchasing decisions, it is only reasonable to expect that it can change political decisions as well. I would love to see work coming out of the HCI community to address this issue. This definitely seems like a problem that we could at least mitigate with well-designed technical and sociotechnical solutions.
  • Social media may be a tool to empower dangerous populism. During this election, a number of important actors (ab)used social media to disseminate fact-free messages that would have been filtered out by traditional media’s gatekeepers. We need to find ways to ensure that social media empowers positive speech and disempowers speech that destroys social capital (while not restricting free speech).
  • This year’s campaign saw nefarious actors weaponize the poor design of e-mail systems. Can we work to build usable e-mail systems that increase security and privacy and at the same time make it easier to adhere to transparency and record-keeping regulations? Can we build e-mail tools that balance retrieval capabilities with the security risks associated with archival approaches?

#7: Run for office

The HCI community is filled with incredibly intelligent, talented, and good-hearted individuals who would make excellent political leaders. In HCI, we’re also a community that understands both people and technology, a major advantage in 21st century politics.

In particular, some of the chief organizers of our community — I’m looking at you Cliff Lampe (among others) — would make amazing candidates. It’s a lot to ask, but I know you’d have nearly all of us strongly advocating for you. We also might want to find ways to incentivize this type of service within our community to minimize the risk associated with running for office.

Let the discussion continue!

Over the past few years, I’ve felt that HCI — and computing more generally — is having a Jonah and the Whale moment. We have been blessed with tremendous capabilities but have been abdicating some of the responsibility associated with these capabilities. I’m optimistic that this election can be something of a wake-up call for us.

This wake-up call can’t be limited to American researchers. Most of the issues I discussed above are not unique to the United States. Quite the contrary, these issues are present in most of the countries where HCI has thrived (e.g. the UK, France, Germany, and others). We’re all in the same boat (whale?) here.

At this point, I have certainly worn out my welcome on the soapbox! I’m hopeful that this post can further discussion in our community, so if you have additional ideas, feel free to e-mail me and hit me up on Twitter. Or, even better, just do some kick ass research that makes the world a better place!

– Your colleague, Brent.