Dawn of The Robot Leaders: Replacement or Collaboration? (2)

Complementary Human-AI brainpower combination

Although we might conclude previous considerations about AI by stating that its flaws are ultimately similar to human flaws, being the result of human making, its bias towards computable truths also allows for an alternative kind of irrationality. And therefore, collaboration between irrational humans and irrational machines might as well lead to increased rational decision-making. An example of such a collaboration might be found in swarm intelligence technology (Swarm AI®), developed by the tech company Unanimous AI. 

In a recent Extend Limits blog on AI Human-Hive mind, Swarm AI® was discussed in relation to medical diagnostics. Originally, Swarm AI® was developed as a predictive tool, and has been successful in predicting various outcomes of events, such as the Oscars, presidential elections, Kentucky Derby and World Series. Swarm AI amplifies the intelligence of human groups, using algorithms that are modeled after natural swarm behavior. By working together as a unified system, the group outperforms individual group members in decision-making and problem solving, which is called “swarm intelligence”. 

Participants using Swarm AI® have to take the decision simultaneously by controlling a small icon on the screen that enables them to influence the group consensus by pushing it toward their own opinion. The AI algorithms monitors the behavior of the participants during the decision-making process by keeping track of the movement of their icons over time. In order to generate the best outcome, it is not enough to record participant opinions. Instead, Swarm AI® also measures the different levels of confidence, which is a crucial piece of information to achieve the best output. For instance, some participants will immediately move the icon to their specific choice, while other participants take longer to come to a decision, or change their opinion one or multiple times. In the end, all preferences are combined into a specific choice. 

A particularly interesting feature of the technology is the democratizing effect on the decision-making process and outcome. As the opinion of every member of the decision-making group is weighed equally, organizational politics and personality traits have no influence on the decision-making process or its outcome. Besides that, decision outcomes become more transparent, as the decision-making process is digitally logged, can be replayed at convenience and provides insight into how the decision came to be. 

Swarm intelligence technology in politics

 In complexly composited groups, as is often the case at various political levels in which representatives of different parties with varying political and ideological backgrounds need to make collective policy decisions, the application of Swarm AI® can be very useful in decision-making processes.  Not seldomly, political decision-making comes to an impasse or to sub-optimal compromises when voting on subjects in which their political ideology differs to too great an extent. By applying swarm intelligence technology, a decision-making process that has come to a standstill can be revived and result in a democratically deliberated decision. 

Swarm intelligence technology may also benefit the various institutions of the European Union that have legislative functions. These institutions consist of representatives of 28 member states with different cultural backgrounds, political ideologies, and values. It may not come as a surprise that a group which exhibits such differences does experience difficulties in coming to an agreement on policy matters. But when the what, why and why not of a range of policy proposals is clearly mapped out and weighed during a policy debate, by making use of Swarm AI® technology, such socio-ideological barriers to the political deliberation process might be more successfully overcome.  

Also, we could look at the European Council, where the direction and priorities of the EU are deliberated by the political leaders of EU member states in cooperation with the president of the European Council and the president of the European Parliament. Obviously, their agenda’s will not always coincide, but through identifying and acknowledging each other’s concerns, they will get closer to reaching a consensus. And it is exactly with the acknowledgement of these concerns that Swarm AI® might help: through real-time identification of which concerns are deemed most credible, or to whose concerns the various leaders have a more lenient attitude, the direction of a future consensus can be more readily identified. In this particular example, Swarm AI® would thus be a means to speed up the decision-making process.

Not the holy grail of decision-making

The application of Swarm AI® technology in political decision-making processes seems very promising. Provided that swarm intelligence technology is applied with democratic intent and in a constructive manner, the mentioned benefits may enhance the overall level of democracy. However, the use of swarm intelligence technology, like any other instrument, can be used for good as well as bad. Thus, the application of the technology may also have the exact opposite effect, undermining the level of democracy. Authoritarian regimes as well as private interests and foreign governments, for example, might seek to add a bias to AI-guided decision-making processes, thus secretly influencing political processes and outcomes. As the risk of hacking government systems poses a real threat, the implementation of swarm intelligence technology as well as any other form of “Robogov” assisting or overtaking human leadership needs to ensure that it adheres to impeccable security standards.