Since launching the Digital Cognition and Democracy Initiative (DCDI) in January 2022, our interdisciplinary coalition of scientists, technologists, academics, and policy experts has researched and debated key questions around how we might devise models for which cognitive functions are necessary for a healthy democracy and how they are impacted by digital technologies. Our ultimate goal is to understand the relationship between the digital technologies that we trust and use daily and our cognition so we can then identify research gaps, propose potential policy and technical interventions, and explore possible mitigation strategies. We want to share a few of our key take-aways as we near completion of this first, exploratory phase of the project.

This initiative grew out of our 2019-2020 Future Digital Threats to Democracy (FDTD) project, a collaboration with the Center for a New American Security, that identified nine emerging digital threats to democracy that we believe will arise over the next decade. A key underlining principle we identified was digital technology’s effects on human cognition as a growing threat to democracy. By the time we published our FDTD backgrounders, we were 6 months into the pandemic, and the increased digitization that came with it made painfully apparent the need for additional research into what we call “digitally influenced cognition”.

While disinformation, affective polarization, and anti-democratic behavior have always existed, the growing scale of these problems, aided in part by the evolving Internet landscape, poses novel threats to democracy. It is our belief that as our reliance on digital technologies grows, we become increasingly susceptible to the first two issues, which in the most extreme cases can lead to the third. While we may never be able to fully combat the existence and spread of malign information, we do believe that understanding factors that make humans increasingly susceptible will empower us to devise policies and technical solutions to foster a more resilient democratic society.
The Digital Cognition and Democracy Initiative seeks to answer this key question:

What about digital technologies in particular are affecting our cognitive capabilities in a way that makes us more susceptible to these threats?

We are exploring the underpinnings of this susceptibility to mis- and disinformation by examining the different interactions between digital technologies and human cognition, identifying the source of that vulnerability, and conceiving of potential mitigation strategies.

Specifically, our DCDI coalition has spent the past five months meeting regularly and examining the effects digital technologies have on trust, emotions, critical thinking, problem solving and reasoning, metacognition, and other critical cognitive functions.

Throughout that process, we have started to think about two distinct types of digital technologies:

1. Digital technologies that allow us to outsource our cognition
2. Digital technologies that affect, and in some cases even manipulate, our cognition

From Our Minds to Our Machines: Outsourced Cognition

It is well understood that some digital technologies allow us to outsource cognitive operations, potentially freeing up cognitive faculties for other work. This follows the pattern of technological evolution: humanity develops a new way to do things faster and more efficiently, and that new approach allows us to focus elsewhere. For example, the calculator freed us from basic math, cell phones from memorizing phone numbers, and google maps from memorizing routes and directions.

DCDI has been delving into potential second and third order effects of this phenomenon. Through our research and coalition conversations, we have been exploring this key question: Do these systems affect or change our cognitive faculties, and if yes, how?

With cognitive outsourcing comes changes in our brain structure and cognitive processes, as we no longer need to perform certain routine tasks. We are not only outsourcing basic math to digital technologies, but memory storage, elements of critical thinking, information sourcing, and often reasoning skills as well, often without realizing it.

Some cognitive outsourcing we have found in our research and coalition conversations include:

• The use of the internet as an external memory bank, with people remembering less information, as they feel confident that they can look it up online whenever necessary.
• Outsourcing our conscious decisions about the content we intake to algorithms. 
• Instructions and directions exist online and are always accessible, lessening the need for us to learn processes and procedures.
• We are increasingly confident in how much we know, confusing what we actually know with what we know we can google.

This cognitive outsourcing, while perhaps more efficient, appears to shallow out the depth of our learning, understanding, and ability to not only process information but be prepared for critical thinking and decision making tasks, having reduced the amount of cognitive processes we can perform on our own. This reinforces our dependence on digital technologies, as they are no longer simple communication systems, but an external cognitive resource—for example a bank of knowledge and memories—without which we would now struggle to function. While there are certainly benefits to technological advancements that free up our time and energy for less mundane tasks, we are especially interested in how the outsourcing process negatively impacts other cognitive operations, like reasoning and problem solving, which are themselves dependent on memory, learning, emotion, and other processes that are often outsourcing. One of our coalition members has portrayed cognitive functions as a pyramid, with the more simple processes at the bottom. 

Learning, emotion, and memory are foundation blocks in this model, as we are concerned about what happens when they become less and less innate to our cognition and more and more dependent on digital accessories.  

Changing Our Minds: Affected and Manipulated Cognition

Digital technologies, either on their own or if intentionally exploited, can also manipulate our cognitive operations. In these cases, digital technologies affect how we engage with and process information. The key questions DCDI is exploring here are: How do digital technologies affect how we view the world and how we trust in systems, information and each other? And equally important, how do these systems exploit and amplify our existing cognitive biases?

Engagement-based social media is the most obvious example today of digital technologies that influence how we view and engage with the world, both by design and even worse, when intentionally exploited by bad actors. 

Examples from existing research and our coalition conversations of how digital technologies can manipulate cognition and amplify our already ingrained cognitive biases include:

• Algorithms that reinforce our biases by sending us confirming information, learning our biases and reflecting those back in information feeds, and creating information silos.
• Encouraging outrage and anger to boost engagement on digital platforms. 
• Encouraging multitasking across one or multiple platforms, reducing our attention span and discouraging deep focus.

Like cognitive outsourcing, these effects and manipulations are irresistible in part because they are efficient, like recommender algorithms that push us content that we like. They also appeal to our paleolithic brains and reward systems, be it feeling a sense of community, seeing content that keeps us engaged, providing a worldview that mirrors ours, or stirring up emotions that can crowd out critical thinking.  

A Reinforcing Loop

These two categories are not meant to represent a dichotomy, rather, they illustrate the two core areas of concerns we have about technology’s intersection with human cognition. Besides co-existing, outsourcing and cognitive manipulation reinforce each other, each one accelerating the other. Examples of this reinforcement include:

• Because we outsource cognitive functions to digital technologies, we think less critically about their manipulation of our information flows and our emotions. 
• By outsourcing our memory, we are more susceptible to false narratives that pop up in manipulative information flows, and when outsourcing the vetting of those narratives to the internet, we can find whatever information we need to prove it right.
• Our increased multitasking, encouraged by digital technologies, leads us to outsource more processes for greater efficiency, compounding the negative effect on our ability to reason, think critically, and make informed decisions.

What does any of this have to do with Democracy?

The reason we are so concerned with all of this is because beyond impacting us as individuals, the cognitive vulnerabilities these systems create or exploit have broader implications for democracy and national security. The question of whether digital technologies are overwhelming human cognitive abilities—and in turn, how that might affect how citizens engage and participate in democracies—is at the heart of why we are digging so deeply into this problem set. 

These vulnerabilities provide companies, malign actors, and even everyday private citizens an increased opportunity to divide, shape, confuse, and disillusion broad swathes of the population. Understanding the human cognitive vulnerabilities to these technological problems is critical to protecting against those ill effects.

Our goal is to help governments, private companies, and civil society understand this threat so we can devise plans, policies and technologies to bolster our resilience.

What’s next?

We spent a year researching the available literature and speaking with experts from across relevant disciplines, followed by five months of deep dives with our coalition into these issues. As we transition into the “so what can we do about it” phase of the project, our first goal is to elucidate the critical gaps in the research that we have identified; propose a research agenda for funders, governments and research institutions; and begin crafting policy, technology, and civil society strategy recommendations to strengthen our cognitive resilience in the hopes of fostering a healthier democracy.