Artificial intelligence, diplomacy and democracy: from divergence to convergence

#CriticalThinking

Peace, Security & Defence

Picture of Chris Kremidas-Courtney
Chris Kremidas-Courtney

Senior Advisor at Defend Democracy, Lecturer at the Institute for Security Governance and former Senior Fellow at Friends of Europe.

In current times, our discussions about artificial intelligence (AI) are often centred on global competition, interactive AI search engines, robot weapons and AI possibly making workers redundant. But what about opportunities to make the world safer and more peaceful? Can AI help to prevent war and famine? Can it also strengthen democracy and international cooperation?

To begin, I asked ChatGPT – an AI chatbot – how AI could help to prevent tensions from rising to armed conflict in a scenario set in 2030 among two states with a significant border dispute standing on the brink of war.

This was ChatGPT’s response:

“In an effort to prevent the conflict from escalating further, the United Nations (UN) has deployed a team of mediators to the region to try to negotiate a peaceful resolution to the dispute. The mediators are equipped with a powerful artificial intelligence (AI) system that is able to analyse data from a wide range of sources, including satellite images, social media and government documents.

“The AI system is able to quickly analyse the data and identify key points of contention between the two countries. It also uses machine learning algorithms to generate a range of possible solutions to the dispute and provides the mediators with real-time analysis and recommendations as they engage in negotiations.

“Thanks to the assistance of the AI system, the mediators are able to help both sides reach a peaceful resolution to the dispute and prevent the conflict from escalating into a full-blown war.”

Trying to use AI to in this way to advise a diplomatic process may be helpful but it seems unlikely to produce any major breakthroughs on its own

On the surface, ChatGPT’s response may seem plausible, but it also assumes that we’ve overcome a major obstacle that we face now: diplomats and politicians often rejecting evidence-based approaches and solutions. Just because something may be the right answer does not mean that it will be politically achievable for all the same reasons that science and fact-based approaches to things like climate change are so hard to implement. Trying to use AI to in this way to advise a diplomatic process may be helpful but it seems unlikely to produce any major breakthroughs on its own.

According to experts, we must also temper our expectations of AI in such situations since it can struggle on its own when dealing with uncommon situations – an area where human minds excel.

On the same front, Meta Research announced in November 2022 that their AI agent Cicero had achieved human-level play in the game of Diplomacy by combining language models with strategic reasoning. Diplomacy is a strategy game involving both cooperation and competition that emphasises negotiation and coordination between seven players.

Cicero was able to infer players’ beliefs and intentions from its conversations and generate dialogue in pursuit of its own plans. The results across 40 games in an anonymous online Diplomacy league showed that Cicero achieved more than double the average score of the human players and ranked in the top 10% of participants who played more than one game.

Does this mean AI can replace human diplomats? Not exactly.

Cicero did well in a finite game with a clear set of rules and used a data training set optimised for the game itself. As with all things, the real world is far more vast, complex and nuanced. Furthermore, any training set of data used to prepare an AI is inherently drawn from the past so its ability to understand new developing dynamics can be limited.

AI can be a valuable tool for helping decision-makers to make more informed strategic choices

Looking to the future, we can see several ways that AI can aid in diplomacy and strategic decision-making. The first and most obvious one is data analysis. AI can help analyse large amounts of data quickly, which can be useful for decision-making. For example, AI can be used to analyse economic data to help inform trade negotiations.

In broader terms, AI can be used to automate routine tasks, freeing up time and resources for more strategic capabilities needed for crisis response. AI-powered language translation can facilitate communication and understanding between individuals and organisations that speak different languages, which is always helpful in crisis situations.

AI can also be helpful to monitor social media and other online platforms for early warning signs of a crisis, as well as to track the spread of a crisis and identify areas that are most in need of assistance. This kind of analysis can be especially useful for helping leaders to understand the context and implications of the choices they face.

While AI can be a valuable tool for helping decision-makers to make more informed strategic choices, what else can it do to help diplomats to negotiate and reach agreements on difficult issues?

One answer is to go beyond using AI as an information tool and instead use it to facilitate decision-making among groups of diplomats and subject matter experts. For example, a platform called Swarm® developed by Unanimous AI uses a biologically inspired form of AI called artificial swarm intelligence (ASI) to enable networked groups of stakeholders to quickly combine their knowledge, wisdom, insights and interests, and converge on solutions together in real-time, optimising consensus rather than reinforcing differing positions.

Can these same tools help us to improve democratic processes, making them more inclusive and satisfying for citizens?

For example, ASI has been used by the UN Food and Agriculture Organization (FAO) to assess food insecurity. Deployed as part of the Integrated Food Security Phase Classification (IPC) process, ASI allowed decisions to be efficiently reached by UN, governments, NGOs and other stakeholders regarding the severity of chronic food insecurity and acute malnutrition in various nations.

According to Louis Rosenberg, CEO of Unanimous AI, this new technology can also help diplomats and political leaders to reach consensus on divisive issues:

“I am confident that ASI technology could help stakeholders with conflicting interests more efficiently reach trade agreements, cease-fire agreements, and resolve a wide range of other contentious negotiations. The reason is that the ‘swarming process’ enables groups to quickly find the solutions they can best agree upon, maximizing their collective satisfaction in the outcome. This is very different from votes and polls which do the opposite, highlighting where groups disagree by revealing differences rather than commonality. After all, polls are polarizing – they drive groups apart, while swarms are unifying, bringing groups together. This is not an accident — after hundreds of millions of years of evolution, schools of fish and flocks of birds and swarms of bees have developed methods that optimize the collective satisfaction of their members. What biological swarms never do is polarize, entrench, and stagnate. If they did, they would have gone extinct.”

One additional feature of ASI is that it makes it possible for citizens on both sides of an issue or border to participate in a diplomatic negotiation process alongside government representatives. Even better, they can do so in a way that is non-hierarchical and anonymous, allowing selected citizen stakeholders to have the same level of influence in the swarm as government officials.

Is ASI a solution for improving democracy in the digital age? Given the divisive nature of democracy in the digital age, can these same tools help us to improve democratic processes, making them more inclusive and satisfying for citizens? Can ASI help us to overcome the divergence so prevalent in this era of social media and instead promote convergence so we can find consensus on critical issues that impact our future?

The best use of swarming to support democracy may be its use by elected representatives in parliaments

According to several recent studies, the answer may be yes. In one study on politically divisive questions, ASI-enabled deliberations were preferred by participants over majority voting methods by 10%. Perhaps more critically, the results achieved by ASI methods were 7% more satisfactory for the out group than traditional voting. This is an important distinction since the challenge of implementing consensus solutions often requires buy-in from those with the minority view, lest they be tempted to thwart implementation.

Another study showed that ‘swarming’ was much more effective at enabling groups to converge on socially optimal solutions at a rate 22% higher than traditional voting methods. Further studies have shown how ASI can enable consensus in small groups, as well as enable consensus between two sides polarised over differences on how to approach the COVID-19 pandemic.

But, according to experts, ASI faces an uphill battle in supporting large scale elections and referendums since it’s not yet suited to facilitate such massive numbers of stakeholders.

According to Rosenberg, the best use of swarming to support democracy may be its use by elected representatives in parliaments and other deliberative bodies by shifting to a convergence model:

“Many of today’s problems are driven by how these groups make decisions, often using a simple majority vote which is far from optimal. If these groups used ASI instead, they would empower representatives to more readily find solutions that maximize the collective satisfaction of all stakeholders. I truly believe it would make representative governments work better.”

Democratic governments should find new ways to integrate these developments to strengthen legitimacy, inclusion and citizen satisfaction

Even without ASI, platforms such as pol.is have enabled Taiwan’s world-leading approach to digital democracy starting with vTaiwan in 2016, which enabled citizens to propose and vote on solutions to technology-related issues. Under the leadership of digital minister Audrey Tang, vTaiwan helped break deadlocks on a number of difficult issues and paved the way for the development of Join, a platform that is easier for less tech-savvy citizens to use and which covers a broader range of topics.

Since its inception in 2018, over half of Taiwan’s 24mn citizens have participated in democratic consultations on Join. But shifting power dynamics is seldom easy since even in Taiwan, the results of these digital democratic processes are not binding on governments. At present they must provide responses to these consultations, but they are not bound the enact them.

In Brazil, the mobile app Mudamos allows citizens to propose laws and collect signatures to gain support for their proposals, but adoption so far has been limited to less than 400,000 citizens. Other similar digital democracy efforts have been underway in Iceland, Belgium, Chile, Estonia, the United States, Spain and Ghana, to name a few.

New tools, new way forward. Given the advances and lessons in ASI and digital solutions to enhance democracy, democratic governments should find new ways to integrate these developments to strengthen legitimacy, inclusion and citizen satisfaction.

This can happen at the local, national and EU levels and start with critical groups of issues like decisions on the storage and use of citizens’ data and how to reduce inequality during the green transition.

These new technologies can enable the convergence of ideas and solutions

Of course, we can’t compromise on human rights or the rule of law, but we must find ways to evolve democracy for the digital age lest it continue to erode in the face of new technologies. The real challenge will be finding ways to achieve better democratic convergence without erasing dissenting voices and minority opinions.

Special technical efforts should also be made to enable these new technical enablers to protect democracy and public debate from malign influences which can skew their results and damage societal cohesion.

At the same time, the UN, EU, NATO, African Union and their member states should explore ways to leverage ASI and other technical developments to allow citizens in stakeholder countries to be able to have more meaningful inputs on diplomatic negotiations, such as cease-fires, border disputes, trade agreements and water-sharing agreements.

In an era during which social media has emphasised divergence, these new technologies can enable the convergence of ideas and solutions that are supported by citizens and various sides of a debate or diplomatic negotiation. While we still have much to learn about how and when to apply them while not compromising our democratic and sovereign values, they do provide ways we can use technology to evolve and modernise our societies in positive ways.

Only by evolving and strengthening diplomacy and democracy to meet the digital age can citizen-centred forms of government prevail over authoritarianism, promote just forms of peace and stability, and better meet the needs, expectations and desires of the citizens they serve.


The views expressed in this #CriticalThinking article reflect those of the author(s) and not of Friends of Europe.

Related activities

view all
view all
view all
Track title

Category

00:0000:00
Stop playback
Video title

Category

Close
Africa initiative logo

Dismiss