Envision that every one of us—entire society—has arrived on a foreign planet and must establish a governance system: a blank canvas. We possess no historical frameworks from the United States or any other nation. We are devoid of any particular interests that could cloud our judgment. How would we manage ourselves? It is improbable that we would adopt the systems in place today. Contemporary representative democracy was the most effective governance method that the technology of the eighteenth century could devise. The twenty-first century presents a vastly different landscape: scientifically, technically, and philosophically. For instance, democracy in the eighteenth century was conceived based on the premise that travel and communication were both arduous.

Truly, the very concept of representative governance was a workaround to overcome technological constraints. Voting has become simpler now. Does it still seem logical for all of us residing in the same vicinity to assemble every few years and select one of us to journey to a distant, large hall to legislate on our behalf? Representative districts were structured around geography because that was the only feasible model over two hundred years ago. However, we no longer need to proceed in this manner. We could arrange representation by age: one representative for individuals in their thirties, another for those in their forties, and so on. We might structure representation randomly: possibly based on birthdays. We can arrange representation however we choose. American citizens now elect individuals to federal roles for terms varying from two to six years. Would ten years serve better for certain positions? Would ten days be more advantageous for others? Numerous possibilities exist. Perhaps we can increasingly embrace direct democracy through plebiscites. Certainly, we do not wish for each of us, individually, to cast votes on every amendment to every bill; thus, what is the ideal equilibrium between votes executed in our name and ballot measures on which we all vote?

Over the past three years, I have coordinated a series of annual two-day workshops to delve into these topics and others of a similar nature.¹ For each session, I assembled fifty individuals from across the globe: political scientists, economists, law educators, specialists in artificial intelligence, activists, governmental figures, historians, science-fiction authors, and more. We did not arrive at any definitive answers to our inquiries—and I would have been surprised if we had—but several recurring themes surfaced throughout the event. Misinformation and propaganda was, of course, a prominent theme, alongside the challenge of engaging in rational policy dialogues when a consensus on factual information is lacking. The detrimental consequences of tailoring a political framework solely for economic outcomes was another vital theme. If given the chance to start anew, would anyone construct a governance system focusing on the short-term financial benefit of the wealthiest individuals? Another point of discussion revolved around capitalism and its relationship, or lack thereof, with democracy. While the contemporary market economy made sense during the industrial era, it is beginning to unravel in the information age. What succeeds capitalism, and how will it influence our self-governance?

Numerous participants investigated the ramifications of technology, particularly artificial intelligence (AI). We explored whether—and when—we might feel at ease with transferring authority to an AI system. At times, decision-making is straightforward. I am comfortable allowing an AI system to determine the optimal timing of traffic signals to ensure the smoothest movement of vehicles throughout my city. When will we be able to express the same confidence regarding the determination of interest rates? Or taxation? How would we respond to an AI device in our pocket that votes on our behalf thousands of times daily, based on preferences inferred from our actions? Or how would we react if an AI system could propose optimal policy solutions that align with every voter’s preferences: Would it still hold relevance to have a legislature and representatives? It may be more logical to vote on ideas and objectives directly, leaving the minutiae to the computers.

These discussions grew more pointed in the second and third years of our workshop, following the explosive emergence of generative AI on the internet. Large language models are poised to draft laws, enforce various regulations, serve as legal advisers and judges, and devise political strategies. How this ability will measure up to human expertise and capacity remains uncertain, but the technology is evolving swiftly and dramatically. While AI legislators are not imminent, just as we currently accept that all political orations are professionally crafted by speechwriters, will we come to accept that future political addresses will be composed entirely by AI systems? Will legislators embrace AI-generated legislation, particularly when such legislation encompasses a level of intricacy that human-created laws typically lack? Furthermore, how will such changes influence the balance of power between the legislative body and the administrative state? Most intriguingly, what occurs when the AI tools we employ to draft and enforce laws start to propose policy alternatives that exceed human comprehension? Will we embrace them because they yield results? Or will we reject a governance system where humans hold only nominal authority?

The concept of scale was another vital theme discussed during the workshops. The dimensions of contemporary governments mirror the technology available at their inception. European nations and the early American states occupy a specific size because that was a manageable scale in the eighteenth and nineteenth centuries. Larger governments—like that of the United States as a whole and the European Union—reflect a reality wherein travel and communication have become more accessible. Today, however, the challenges we face are either localized, at the level of cities and towns, or of a global nature. Do we genuinely require a political entity the size of France or Virginia? Or is it a blend of scales that we truly need, one that functions effectively between local and global contexts?

Regarding alternative forms of democracy, we examined one from history and another enabled by contemporary technology. Sortition is a method of selecting political officials randomly. We employ this today during jury selections, yet both the ancient Greeks and specific cities in Renaissance Italy used it to appoint significant political figures. In modern times, numerous nations—primarily in Europe—are utilizing this method to address policy on intricate issues. We could randomly select a few hundred individuals, representative of the population, to spend a few weeks receiving briefings from experts, engaging in debates on the issues, and subsequently deciding on environmental regulations, budgets, or virtually anything.

“Liquid democracy” proposes the elimination of elections altogether. The concept is that each individual possesses a vote and can delegate it to anyone they prefer. A representative collects the proxies assigned to them and can either cast votes directly on issues or transfer all proxies to another individual. Perhaps proxies could be allocated differently: one person for economic concerns, another for health issues, a third for national defense, and so on. In the more unadulterated forms of this system, individuals might reassess their vote transfer to another person at any moment. There would be no designated election days: vote totals might fluctuate daily.

Finally, there is the matter of participation and, more broadly, whose interests are acknowledged. Early democracies were not genuinely democratic; they restricted involvement based on gender, race, and land ownership. Nowadays, to cultivate a more inclusive electorate, we could lower the voting age. Yet, even very young children possess rights, and in some instances, so do non-human entities. Should future generations have a “voice,” however that may be defined? What about nonhumans or entire ecosystems? Should every individual have equal volume and type of influence? Presently, in the United States, the wealthy wield significantly more sway than others. Should we formalize that advantage explicitly? Perhaps younger individuals should carry a weightier vote than the rest. Or conversely, perhaps older individuals should.

During the workshops,

those inquiries prompted further reflections concerning the boundaries of democracy. All democracies possess limits restricting the decisions the majority can enact. For instance, we cannot cast a vote to abolish Common Knowledge, yet we can typically oversee speech to a certain extent. We are unable to vote, during an election, to incarcerate an individual, but we can formulate laws that render a specific act unlawful. Each of us has entitlements that must remain inviolable. In the society of our future, what should our individual rights entail? What should the rights of the community encompass, taking precedence over those of individuals?

Personally, I found it particularly intriguing, at each of the three workshops, to examine how political frameworks fail. As a security technologist, my focus is on how sophisticated systems are undermined—hacked, as I would term it—for the advantage of a select few at the detriment of the many. Consider tax loopholes or maneuvers to circumvent government oversight. These exploits are prevalent today, and AI instruments will facilitate their identification—and even their creation—in the future. I would desire any governmental framework to be resilient against deception. Or, to express it differently: I aspire for the interests of every individual to correspond with the interests of the collective at all tiers. We have never experienced a governing system with this attribute, but—in an era of existential threats such as climate change—it is crucial that we cultivate one.

Would this novel governing system even be termed “democracy”? I genuinely do not have the answer.

This sort of conjecture may not be pragmatic, yet it remains significant. Our workshops did not yield conclusive responses and were not designed for that purpose. Our discussions were rich with ideas on how to mend our political structure where it is in distress. Individuals routinely debate modifications to the US Electoral College, or the methodology of defining voting districts, or the establishment of term limits. However, those are merely incremental alterations. It is challenging to locate individuals contemplating more radical reforms: looking beyond the immediate future—not at what is achievable presently but at what could potentially be realized eventually. Incremental thought is crucial, yet it can also be shortsighted. It reflects a hill-climbing approach of constant but considerably restricted enhancements. We also need to consider transformative changes that cannot be easily accessed from our current position; otherwise, we may find ourselves perpetually confined at local peaks. While true progression in governance is considerably more challenging than advancement in technology, particularly absent a drastic upheaval enforcing modifications upon us, it is something we, as a species, must master, one way or another.

Our workshop will reconvene for a fourth session in December 2025.