When we talk about democracy, are we consistently talking about the same concept? How does the majority of us view democracy, and how does the data we produce (e.g., social media, news pages, YouTube, blogs, etc.) influence how artificial intelligence interprets the concept again?

Definition Whenever I research a distinct topic, I typically start by understanding very clearly how the term that my research involves is being defined. My first resort is asking Google for the official definition, which is usually retrieved from Oxford Languages. Here is the commonly understood definition of democracy:

(1) a system of government by the whole population or all the eligible members of a state, typically through elected representatives.

“capitalism and democracy are ascendant in the third world”

(2) control of an organization or group by the majority of its members.

“the intended extension of industrial democracy”

Out of curiosity more than anything else, I consulted two generative AIs about their visual interpretation of democracy. I didn’t provide any further descriptions other than a prompt to visualize democracy. This is everything my original prompt (the text-based descriptive command) consisted of:

democracy

Midjourney was the first AI I consulted. Midjourney is a generative artificial intelligence program that lets users generate images from natural language descriptions.

The second AI I consulted was ChatGPT. ChatGPT is a chatbot developed by OpenAI. Since DALL·E 3, a text-to-image model, was natively integrated into ChatGPT, it lets users generate images in a similar manner to Midjourney.

Since Midjourney is a generative AI specialized in image generation, I received a very immediate result:

According to Midjourney founder David Holz, the open dataset for Midjourney was built by “a big scrape of the Internet” and, therefore, billions of images and their captions. This means, in return, that we can assume that the result for our prompt is representative to a certain degree of everything available on the internet.

In the case of ChatGPT and DALL·E 3, I entered the same prompt, which resulted in a feedback session as follows:

“To create images symbolizing democracy, I would need a bit more context or specific elements you'd like to include that represent democracy to you. Democracy can be represented in many ways, such as people voting, discussions and debates, democratic symbols like a parliament building or a voting box, or more abstract representations of democratic principles. Could you please provide more details on the specific aspects of democracy you would like to be visualized?”

I clearly stated in my response that I am not looking for my version of democracy but rather would like to understand how DALL·E 3 sees the most prominent/ popular version of democracy:

Let both image groups sink in for a few minutes and evaluate by yourself what these interpretations really show. Do they visually align with what you think of when visualizing democracy in your mind? Do these images align with our former definition of democracy according to Oxford Languages? If not, what stands out and why? If yes, question what are your expectations toward democracy!

Design, more than political science or system design, is a great tool/ point of view to dissect problems such as this one. Within the discipline of design, we have one principle that stretches over every type of design output: “Visual Hierarchy.” It translates to how the focus of the consumer/ viewer is being guided by the hierarchy placed within a design. The following example is a good comparison:

The graphic above is easy to decipher since it not only guides visually but also in a text format. If you were to apply the same principle to the generated imagery, the focus would never lie on the group of people who are supposed to define democracy (“whole population”, “by the majority”) but rather on either one central figure or a very small group of people who apparently are at the visual center of democracy.

Research suggests that “anti-politics” (*2) is on the rise since 2008. Anti-politics describes a movement that voices political discontent, mistrust, and cynicism. In the age of social media, the expressed discontent is often published to a wider audience and ends up taking influence on a larger scale, where rumors turn into conspiracies and unproven ideas are circulated as facts. All of this is dangerous and influences how political participation is executed. (3)

The real question, though, is why this is happening. According to the authors of “Need for Chaos” (3), the reason lies primarily in the “rising levels of income inequality and stagnation in real wages” in democracies. This means the political sphere and the economic sphere in democracies are tightly interwoven with each other.

The interesting detail is that it can seem that the feeling of political frustration is present in all groups of political participation when, in fact, and according to many studies (4), the discontent is stronger among citizens who belong to lower social groups within societies. An article by Luigi Droste shares further insight into where higher levels of public corruption and increasing levels of income inequality are the main drivers for such behavior. The reason it looks to us like the discontent might be so omnipresent is that this concerns the majority of people on this planet, as most people live in unfavorable conditions and experience income inequality on a daily basis. This is supported by the World Inequality Report (5) which states that the poorest 50% of the population own just 2%, the middle 40% own 22%, while the richest 10% own 76% of total net wealth. This is inequality based on facts and numbers and should make us aware of how dangerous inequality can become if not addressed.

Equally harmful to anti-politics can be “political apathy” (6), where political frustration can go so far that people disassociate with their own government and/ or their individual relevance within it. As described in Anna Zhelnina’s article, the so-called “apathy syndrome — a combination of emotional mechanisms and cultural norms that produce political apathy.” The syndrome evolves in different stages and eventually.

Both forms of political participation are dangerous to the original idea(l) of democracy and should see more attention.

Going back to the AI-generated imagery, the reasoning behind the results I received becomes more obvious: Democracy has strayed from its original path and definition where neither the “majority” nor the “whole population” can really influence the future of our societies nor how democracy is performed.

Most citizens within democracies have limited actions at their disposal to become “active” political participants by voting, staying informed, engaging in direct dialogue with their local government, engage with local communities that advocate for something specific and similar.

All of these options require not only a lot of dedication but also a solid foundation of education and time. Marginalized groups who are mostly affected by anti-politics or political apathy can’t afford most of the above due to time constraints and sometimes also educational restrictions. This means that the system of how we perform democracy is not built for the majority of people who are supposed to be active in it!

While facing this truth is somewhat frightening and agonizing, it should motivate us to look for solutions and build systems of democracy where the common citizen can actively participate. A secret to unlocking that part of a system is by using methodologies we know well from game design. Many very popular games emulate existing societies. Yet, for many, those games are more engaging than real-life scenarios. Josh A. Learner asks in his book “Making Democracy Fun” if a democracy should be fun. (7) Yes, it definitely should be, and it could be!

This also brings up my last point of this conclusion: Politics isn’t what we should focus on. Politics is an artificial system that enables a few to make decisions for the many. Instead, let’s unlock the next level of democracy together by engaging in research on decentralized governance, and we will be heading in a future-forward direction!

In collaboration with Neo-Cybernetics, I will be publishing more on the fundamentals of decentralized governance, and I hope you guys will stick around to read and share your feedback with me!

Check out Stella on her Social Media Channels here.

References:

• (1), Benmbarek, I. (2022) What is visual hierarchy in design? (explained with examples), RSS. Available at: https://www.creator-fuel.com/blog/what-is-visual-hierarchy-in-design-explained-with-examples (Accessed: 26 February 2024).

• (2), Hochuli, A., Hoare, G. and Cunliffe, P. (2021) The end of the end of history: Politics in the twenty-first century. Winchester: Zero Books.

• (3), PETERSEN MB, OSMUNDSEN M, ARCENEAUX K. The “Need for Chaos” and Motivations to Share Hostile Political Rumors. American Political Science Review. 2023;117(4):1486–1505. doi:10.1017/S0003055422001447

• (4), Droste, L. (2021) Feeling left behind by political decisionmakers: Anti-Establishment sentiment in contemporary democracies: Article, Politics and Governance. Available at: https://doi.org/10.17645/pag.v9i3.3949 (Accessed: 29 February 2024).

• (5), Inequality: Global trends (no date) Development Initiatives. Available at: https://devinit.org/resources/inequality-global-trends/#:~:text=According to the World Inequality,%2457%2C300 per adult in 2021.