Let’s take a particularly contentious real-world example: It would be impossible to ask the 6.7 million people in Libya exactly how they feel about the current political situation, because even if you only spent 10 minutes per person, doing so would take 127 years, not to mention that the pollster could be biased, corrupt, or ignorant.
However, an intelligence can pay attention to thousands of data streams simultaneously in real time. Traditional surveys and polls could be eliminated by using artificial intelligence not only to present questions and record responses, but to identify biases and collect and interpret results in real time .
Better democracy
We know that a question can induce a particular answer depending on how it is formulated. The questions can also be limited to questions that do not take into account the whole problem: for example, asking yes or no to the independence of Catalonia can be tortuous in the sense that many people want independence, but why not ask independence this way or this way? Or independence from a different measure? And so on.
However, designing surveys and polls through artificial intelligence would create greater transparency and hold those asking the questions accountable for any misleading, intentional or other questions . Plus, it would give pollsters and scientists a chance to ask follow-up questions, explain confusion, and deal with outlandish data right away.
And it is something that we are already beginning to do. Going back to the first example, when Libyan forces signed a ceasefire last year, it was not the cunning of politicians or the diplomacy of ambassadors that brokered the peace. It was the United Nations and a relatively unknown AI startup. This startup is called Remesh and it creates tools that allow organizations to conduct real-time surveys with more than 1,000 people simultaneously .
Basically, the system allows you to display information from a live audience. For example: if you wanted to poll 100 people on the taste of a new candy, you would traditionally send them a questionnaire and then count the responses to see what the majority thought. But with Remesh, you can ask follow-up questions in near real time to give respondents an opportunity to expand or explain .
And, instead of tasking a poor team of humans with trying to moderate the responses of hundreds or thousands of people at once, AI does all the heavy lifting almost instantly. Basically, Remesh turns the survey results into a conversation with the people who answer it, as they take it . As the Washington Post explains:
Participants in Yemen and Libya were asked to visit a web link, answer open-ended questions, and respond to surveys on their smartphones. They were asked to identify which community they represented or which party they strongly identified with. All the information was shared with local political figures who were able to respond live on television or act according to what the audience says.
This type of strategy could open the way to other types of democracy that are purer, or perhaps less tendentious. After all, the term democracy has become a wild card when we are not even able to define it (as it has innumerable meanings).