The traditional foundations of international diplomacy have been on negotiation power, secret channels of communication, and personal chemistry between leaders. According to the co-founder and first international centre for science and diplomacy, a new era has begun in which dispassionate insights from AI algorithms and mathematical methods such as game theory will play an increasing role in the negotiation of deals between nations.
Professor of negotiation and conflict management Michael Ambhl and ex-chief Swiss-EU negotiator said that recent advancements in AI and machine-learning mean these technologies have a significant role to play in international diplomacy. This includes at the Cop26 summit later in the month, and in post-Brexit trade deals.
These technologies have been partially used, and Ambhl stated that they will be used more. We are interested in everything related to data science, artificial intelligence and machine learning.
He said that AI is still in its infancy and AI will be used in international negotiations only when it becomes more widespread. Machine learning can be used to detect fake news and assess the integrity data to ensure that the diplomatic process is solid. These technologies may be used in the future to detect patterns in economic data that underpin free trade agreements and standardise certain aspects of negotiations.
The Lab for Science in Diplomacy is a collaboration between ETH Zrich, Ambhl's base, and the University of Geneva. It will also be focused on negotiation engineering. This involves using existing mathematical techniques like game theory to frame a discussion or to simulate different scenarios before engaging in negotiations.
These tools are not new. John Von Neumann, a Hungarian-American mathematician, developed game theory in the 1920s to formalize the idea of bluffing during poker. Later, he used it to calculate nuclear strike scenarios during the Cold War. According to Ambhl, such techniques were not lost to technology, but because of a lack in knowledge. Diplomats aren't so familiar with it.
As the world becomes more data-savvy and tech-savvy, those who don't use quantitative methods may be putting themselves at risk. Ambhl stated that he was Switzerland's chief EU negotiator and ran a simulation of game theory in preparation for talks that resulted in Switzerland joining the Shengen region and several agreements with the EU regarding tax, trade, and security. It was in Switzerland's best interest that negotiations take place in a package and not sequentially. The Swiss government insists on this basis for talks.
Was the EU able to do its own analysis? Ambhl said, "I don't think so." They didn't know that we were interested in game theory.
According to Ambhl, a mathematical approach can help de-emotionalise the root causes of conflicts. Ambhl cites the 2005 Geneva talks between Iran, the P5+1 nations and where he developed a mathematical formula to calculate the Iranian reduction in nuclear centrifuges. He said, "Now let's talk about the size and gradient of the alpha," when we presented the idea. It is between 0 to 1. It is more technical.
Is it possible to distill deeply-rooted political issues into a gradient along a curve? Ambhl stated that this is missing the point. He said that it is not possible to distill deep-rooted political issues into a gradient on a curve. He said that it is not about reaching a technical agreement. You have to break it down. It is divided into sub-problems, sub-problems, and sub-subproblems.
A more scientific approach does not mean ditching traditional methods. He said that I'm not saying you can only negotiate well if it is done this way. Other factors such as how strong your bargaining power is, how charming you are at negotiations, how supportive you PM are, and how prepared you are for tough negotiations all play a part.
Is there a risk that these new approaches could backfire? Could rival AIs escalate conflicts or arrive at diplomatic solutions that are mathematically perfect but have terrible real-world consequences?
You are not going to war because some blind algorithm decides that it is. This would be absurd, stated Ambhl. It is always a decision tool.
He said that you cannot blindly trust it and that you cannot rely solely on the gut feelings of these politicians. It is necessary to combine new technologies with political analysis.