On Collective Intelligence Optimization
Do you ever think about collective intelligence (CI)? In the near future I believe a key function of enterprise AI teams will be collective intelligence optimization.
It's a relatively small field of study that produces some thought-provoking findings, especially so with the boom of AI. Here's an abstract from a paper called "Quantifying collective intelligence in human groups" that I feel is pretty important:"CI, in turn, is most strongly predicted by group collaboration process, followed by individual skill and group composition. The proportion of women in a group is a significant predictor of group performance, mediated by social perceptiveness."
AI experts will immediately recognize in this chart that Process, Skill, Group Size, and Composition are all factors that can be heavily modified through the implementation of synthetic team members. It hasn't been remotely possible until very recently that we could introduce a more stable variable into the highly abstract process of measuring collective intelligence. But with AI models, we can. The rise of AI gives us the tools to optimize around this complex equation of process, skill, size, and composition, and find the most potent mix for supplementing the intelligence of human individuals and groups. Model gardens and AI workflow suites give us the measure of consistency we need to start implementing AI as a collective IQ amplifier across organizations. (This is where I'm building software, and it continues to blow my mind)
The best teams in the world will soon be intelligently augmented with AI in a range of capacities. And their collective intelligence (and impact) is going to absolutely dwarf teams that are only comprised of humans, no matter how brilliant they may be.
The battle for supremacy in this space will revolve around implementing collective intelligence amplification suites that can serve to significantly boost the CI of any human team enterprise-wide.
Source: "Quantifying collective intelligence in human groups" https://www.pnas.org/doi/10.1073/pnas.2005737118
