ishpreet.singh's Techdirt Profile

ishpreet.singh

About ishpreet.singh

Posted on Techdirt - 15 November 2021 @ 10:05pm

Digital Democracies: How Liberal Governments Can Adapt In The Technological Age

At the turn of the last millennium, there was a wave of optimism surrounding new technologies and the empowerment of the modern digital citizen. A decade later, protestors across North Africa and the Middle East leveraged platforms such as Facebook and Twitter to bring down authoritarian regimes during the revolutions of the Arab Spring and it was believed these technologies would bring about a new flourishing of the worldwide liberal democratic order.

Unfortunately, the emancipatory potential of the open internet has been undermined by the latest in a long line of authoritarian regimes hijacking the technology. Autocracies adapted, with China leading the way. Towards the end of the 20th century, the CCP introduced a new political-economic model revolving around centralized rule and a controlled market economy. With this new model, China has successfully broken common political-economic orthodoxy by limiting domestic desire for democracy while maintaining a sizable middle class. A key driver of this is a “comprehensive system of state repression, bolstered by the latest digital technologies.”

China has turned applications of key technologies such as artificial intelligence, facial recognition, and data analytics to usher in a new age of digital dictatorships which can spy on its citizens, predict dissent, and censure unwanted information, increasing regime resiliency through a virtually manufactured world safe for autocracy. Many leaders of the developing world have taken advantage of this, as China has exported its model to other countries, such as Venezuela, to restrict the freedoms of the modern citizen.

As authoritarian governments increase regime resiliency by leveraging digital technologies, citizens of liberal democracies have trouble trusting these same technologies in no small part due to their exploitation by authoritarians and their agents at home and abroad. This understandable hesitancy may not only lead to a technology gap, but leave democratic institutions vulnerable to threats by these same technologies; one can simply look to Russia’s meddling attempts in recent U.S. elections. The pendulum has swung towards autocracy as digital technologies seem to have had an asymmetrical effect, bolstering authoritarian regimes as the tides turn against the previous post-Cold War wave of liberalism. Pessimists are looking backwards, opening that perhaps digital technologies are in tension with–or even pose a threat to–liberal democratic governments. This does not need to be the case.

As pointed out by Mike Masnick, when charting the future of the digital landscape, one should not automatically assume “progress towards a ‘good’ outcome is inevitable and easy,” but nor is the path towards a technological dystopia. Not all progress is good, but if we can’t move forward and can’t stay still, that leaves only one option. By rethinking how we apply these technologies, democracies around the world can find methods by which to both enhance the liberal democratic ideological values, and protect against weaknesses inherent to the political ideology, creating resilient “digital democracies” to stand against the rising autocratic tide and reinforce their own struggling institutions.

Transparency through Technology

The main method by which democracies can enhance liberal principles is by creating more transparent governments. Transparency can act to limit the powers of government, promoting freedom and individualism, while making policy and officials more effective and accountable, respectively. An informed, educated citizenry is the cornerstone of any robust liberal democracy, and there’s no reason the technologies of the Third Industrial Revolution shouldn’t be employed in service of this goal. To do this, digital technologies should be leveraged to better illuminate the wants and needs of citizens for more informed policy making, create reliable metrics so constituents can directly measure the effects of politician’s policy decisions, and generally increase the transparency of actions of government officials, law enforcement agencies, and lobbyists.

Though digital technologies make it possible for countless data points to be gathered on individual citizens, the potential for these processes goes beyond delivering targeted ads without the development of a CCP-style social credit score. Data analysis allows for the ability to create policy options that better fit citizen’s needs, for example customizability. If insurance companies allow customers to select policy options based on their needs, why can national governments not do the same based on a person’s location, age, family size, economic situation, etc? We already see this in tax policy, but with digital technologies, this can be applied more broadly. For example, the U.S. welfare state is notoriously difficult to navigate and runs on systems almost old enough to qualify for Social Security. An embrace of new developments could not only make programs more accessible to those who qualify, but more tailored to their specific needs. Policymaking can be reimagined to create options that incorporate different citizens’ wants and needs, and advances in digital technologies allow for a larger, faster, and more diverse analysis of the data required to design and implement these complex solutions.

One of the main benefits of artificial intelligence, machine learning, and “big data” is the ability to uncover causal directions between different variables. By using these technologies, it will be easier to understand the direct effects of policy decisions across a number of variables. This level of transparency would create for a more informed citizen as they could see how their representatives’ voting habits are directly affecting their bank accounts, the job market, their access to healthcare, etc. For example, governments could use the abundance of financial data to publish reports showcasing how specific tax laws actually affect different demographic groups, taking the guesswork out of evaluating policy decisions. The use of granular data to analyze mistakes in the designation of opportunity zones created by the Tax Cuts and Jobs Act is one example. Digital technologies could also allow for simulations of different policy proposals to see how they might affect an individual, city, or state. Digital twin technologies created by companies like Deloitte and Dassault are already being leveraged by governments to understand how certain decisions on energy and logistics will change how cities operate. Another possible use case of these technologies is policy trials that would allow governments to study the effects of lawmaking for a given geographic area and time period, similar to how businesses run trials on different marketing plans. A “low fidelity” version of this could be seen in 2013 when the Obama administration allowed Colorado and Washington to experiment with legalizing recreational marijuana.

While reasonable steps should be taken to anonymize such information, it’s possible to publish it in a manner that’s transparent and easily digestible by the press, policy analysts, and public officials to make it harder for pie-in-the-sky policy proposals to be introduced and adopted without just taking the word of their proponents.

Lastly, digital democracy can shed light on the action of government and adjacent officials. The idea is similar to that of China’s surveillance system, but in reverse. If China can use digital technologies to monitor citizens, thereby dissuading populations from making certain choices, why can citizens not do the same to governments? Indeed the adoption of a social credit system was a deliberate, top-down choice by the CCP–not the natural evolution of “big data.” Who’s to say that a liberal democracy can’t flip the script? By capturing and sharing instances of government corruption, police abuse, and lobbying malpractice, society’s officials will be dissuaded from making decisions against the public’s interests. Apart from Facebook, and Twitter, a number of specialized platforms are being deployed for that very purpose. Guatemalans have experimented with a social platform that allows users to share examples of police corruption. German made LobbyControl provides transparency on lobbying at the local and EU level. Working together, liberal democracies can share these platforms to create an international system of government accountability.

Defense through Digitization

Liberal democratic systems are not without their weaknesses. One such vulnerability is the slowness, as can be seen by America’s sedated and haphazard response to the COVID-19 pandemic. As they can be used to enhance tenets of liberal democracies, digital technologies can also protect against inherent weaknesses by accelerating government response in the face of crisis, preventing the spread of propaganda and polarization, and protecting citizen’s rights to freedom and privacy.

AI, ML, and big data can be leveraged in three key ways to help liberal democratic governments with crisis response. First before a crisis strikes, algorithms can analyze data to uncover vulnerabilities in a system before they take hold. This could have been useful in predicting the collapse of the housing market prior to the 2008 financial crisis. Once a crisis has struck, these technologies can ascertain the principal drivers of a crisis so resources can be deployed accordingly. Tools like this could have been useful during the ongoing hunger crisis in Burkina Faso, where the government may have chosen not to close key resources in the food supply chain had they realized that malnutrition has been a larger cause of death than COVID-19. Lastly, these tools can be used in the post-crisis period to understand which policies had the most beneficial impact, helping to prepare for future events, such as the next pandemic.

And again, openness with this information makes it possible for more parties to cry foul when something isn’t right. In the case of the 2008 financial crisis, there was a vocal minority sounding the alarm. Still, the talking heads and smartest guys in the room maintained a rosy view and were able to dismiss those critics as Cassandras. Open access by a larger swath of the public to warning signs from reliable sources makes it less likely that those who should know better can take a “nothing-to-see-here” line to be repeated by pundits, public intellectuals, and policymakers.

Sometimes weaknesses and vulnerabilities are driven by our own applications of technology and require course correction. In democracies across the globe, it seems as though the public is becoming more polarized. One driver of this is the ease of leveraging technology to sharpen divides in societies to subsequently weaponize public opinion. But it is not technology that is inherently at fault, but the applications of these algorithms to maximize views and profits. Studies show that a large percentage of citizens are less polarized than previously believed. Unfortunately, these moderates may choose to stay away from certain social platforms in order to avoid inflammatory media. But what if the algorithms were rewritten to prop up neutral voices rather than to spread inflammatory content? Moderates may be more willing to use these platforms, and populations would more readily see muted perspectives on an issue. By redesigning the algorithms we use to spread information, digital technologies may be able to turn the tide against tribalization, and subsequently polarization. There’s no neutral design, and the amplification of those who can bring more light than heat and turn down the temperature of online discourse more broadly deserve promotion.

As discussed in the previous section, recommendation systems operate on algorithms that we do not quite understand. These algorithms are not inherently undemocratic, but their applications can lead to unwanted side effects infringing on our freedoms and privacy. By understanding how to game these algorithmic recommendation engines, outside actors are able to create media which can influence perspectives and subsequently our decision-making process, in effect limiting our freedoms by breaking the integrity of our autonomy. What makes this even more dangerous is that we are seldom aware of this occurrence, as we scroll through videos, posts, and tweets on autopilot. A solution to this infringement on our online freedoms can be found by assessing these algorithms and redesigning them to serve the purposes we require.

Looking to privacy, a number of tools have been created that can provide noise to data, making it difficult for digital entities to uncover insights. As an example, a program on your computer could randomly jump to different websites during your downtime to prevent unwanted AI systems from making accurate recommendations based on your browsing history. Likewise, AI software can add similar noise to online pictures by changing a few pixels’ colors. One could apply this noise to their Facebook or Instagram posts to prevent facial recognition software from recognizing the images, while allowing friends and family to see the pictures largely unchanged. These sorts of systems could be used in places like China to confuse digital surveillance technology. One key note to remember: if and when “digital democracies” start to appear, it is important to not cross thresholds into the authoritarian regime — the goal is to increase resiliency without further infringing on our rights to freedom and privacy through digital technologies.

Increasing Resiliency through Trust

The examples above serve as a start to the discussion of means through which democracies can become resilient through the usage of digital technologies. But “digital democracies” are not inevitable, as western liberal society has a certain mistrust towards big tech companies who are vital in driving such a transformation. In order to evolve into “digital democracies”, three main societal changes must occur.

The first is an establishing of trust between big, Western tech conglomerates and governments. In the US, mistrust of companies like Google, Facebook, and Amazon have led some to call for a breaking up of these giants, but this is not the solution. Without these companies on our side, liberal democracies may not be able to keep up with the same advancements made by Eastern tech conglomerates such as Baidu, Alibaba, Tencent, and Xaiomi. Instead, governments must partner with tech firms in order to more clearly define rules and regulations, as well as responsibilities between the two groups. Without an open, non-overreaching dialogue, the situation will remain hostile, making it difficult to establish technological resiliency.

Second is the trust between governments and the general public, in regards to the usage of digital technologies. Partnerships between big tech and governments, as discussed above, may lead to greater issues, as can be seen by China’s use of big tech to create a state-controlled market and social economy. This partnership itself would also be antithetical to principles of liberalism by placing too much power in the hands of the government. Just as the media was once seen as a watchdog over governments and politicians, there must be an independent body that serves as a watchdog over governments and their use of tech. There are nonprofits, such as the Center for Humane Technology that serve to advocate for mission-focused tech development. Similar organization will be necessary to serve as a guardian between governments and the abuse of digital technologies. It is with the existence of independent bodies such as this that populations may begin to trust governments to use technologies to only further ideals of liberalism.

The last piece is to establish trust between big tech and the general public. This ties back to transparency. As stakeholders in our own data, people should have a say, or at minimum an understanding of how our information is used. But many of the new AI, ML, and data models utilized by big tech are often seen as “black boxes.” Our data goes in, and a result in the form of a product recommendation, news story, or social media post comes out, without a clear understanding of how the outcome was reached. By opening up algorithms and making them fair, accountable, and transparent, people would feel more comfortable by understanding how their data is truly acquired, assessed, and leveraged. This could be a key step in making digital technologies democratic — it would allow citizens to claim a stake in technology, just as big tech has claimed a stake in our data.

Trust between citizens and governments is a fundamental principle of liberalism and democracy, but in today’s ever-polarizing society, this can be hard to come by. The situation becomes even more complex when adding tech titans to the mix. Organizations exist to help establish this trust by guiding governments and big tech in more “humane” directions, but it will take cooperation by all stakeholders, along with NGO partners to increase outreach and communication for a more transparent relationship. This is the first step towards increasing resiliency of democracy, a necessary lever to swing the pendulum back to the people.

Ishpreet Singh is a recent engineering grad currently working as a strategy consultant.

More posts from ishpreet.singh >>