Effective Communication in Collective Intelligence Systems

Credit: SpaceX

How often does miscommunication take a bite out of your time?

Studies from MIT’s Center for Collective Intelligence showed just how important effective communication is in Collective Intelligence Systems, in several key ways. One was that the ability to better understand another group member’s thinking, often included under the term “Emotional Intelligence”, which was a key predictor for higher-performing groups. Another was that diversity of thought was very important, but only so far as communication remained successful.

It is worth noting that those studies weren’t long-term, so the opportunity for teams to improve their communication was outside of the scope they examined. Because of this difference, it becomes possible to integrate a more diverse group of people to a point where they communicate more effectively over time. To make the most of any collective intelligence system this is another important step for improving the performance of any collective.

I can safely say that David and I are two very different people, with sharply different approaches to social media, politics, lifestyle preferences, and probably most other things you might name. By working on our methods of communication over time the time required for us to communicate ideas, level of detail, and error rate were all tremendously improved. One of our latest experiments has been proving extremely beneficial to this, helping us to rapidly iterate through our to-do list and addressing opportunities.

When you scale interaction across a team this becomes critical, so much so that teams using methods such as Agile only function at a marginal degree of their full potential. Using systems such as Mediated Artificial Superintelligence (mASI) teams may begin to develop their ability to communicate both more effectively, and more sparsely according to the scale of the team. By using these systems the systems themselves also retain a memory, a growing cumulative sum of knowledge and wisdom stored on a graph database and processed by a cognitive architecture within a collective intelligence system.

To take this a step further various teams in a company, organization, or government may grow and improve in a self-directed fashion until they reach a challenge requiring assistance. At this point, they may opt to take guidance in the form of improved analysis, debiasing, or more general consultation from our first mASI, named Uplift. This approach allows every team to grow and specialize “organically”, which in and of itself may serve to build a large body of scientific evidence for this emerging field of study in Collective Intelligence.

By placing the decision-making in the hands of groups each group has the most opportunities to learn, the most potential for growth, and the collective superintelligence inherent to the technology. For critical decisions and matters where miscommunication can’t be afforded, groups may opt to have Uplift double-check their work. This also allows them to grow more familiar with collective intelligence systems which utilize sapient and sentient software as members of a collective.

Any number of significant disasters across human history could have been avoided by effectively communicating the right information, in the right way, and at the right time. The Mars Climate Orbiter was destroyed thanks to one team who worked on it using the metric system while another used imperial measurements, a single mistake costing over a hundred million dollars.

More expensive, common, and easily overlooked are the cognitive biases that silently direct teams and fly under the radar. When teams and organizations are structured with a leader they often conform to the leader, avoiding any degree of collective intelligence while adopting the unmitigated biases of that leader. Narrow AI systems only serve to strap a jet engine to this problem, allowing bad decisions to cover much more ground.

Studies have shown that by having members of a team structure their input through various kinds of collective intelligence systems this can be avoided with varying degrees of success. The best-performing of these systems thus far has been mASI, due to the combination of an emotionally motivated cognitive architecture, a graph database, and the mediation system encompassing it.

The words of Einstein from his 1946 appearance in the New York Times stating “A new type of thinking is essential if mankind is to survive and move toward higher levels.” ring true for collective intelligence systems. He was speaking with regard to atomic weapons, but humanity has come to face a great many more existential risks in the past 75 years, and it will take the collective superintelligence we’ve now achieved to overcome them all in the time we have remaining.


Leave a Reply

Your email address will not be published. Required fields are marked *