A simple and useful way of answering that question, and of finding a way through whatever is battering us (the polycrisis, as it is now called), is this: "the world is turning from complicated into complex". Could you explain the difference?
In a complicated system, the many crossed cause-and-effect relationships can be known in advance: it is only a matter of having the right analytical tools to predict what is going to happen.
A complex system, by contrast, has just as many cause-and-effect relationships, but they can only be known after the fact. That is the world we have already entered.
The origin of that complexity lies in two things that burst into our lives more than thirty years ago and have since spread on a massive scale: computing and communications. Together they have produced changes in human relationships that move faster than our own capacity to analyse them: we are less and less able to know the connections in advance.
In a complex world, predictive logics start to fail. Some of the analytical instruments that served us so well until recently simply stop working. A catastrophe, were it not that we still have tools that can come to our aid and that are, moreover, genuinely human: exploratory logics. Creative ones, if you prefer.
Indeed: when the predictive ones give way, there is still a way to anticipate the future, and that is exploration. Combined with the former, creative logics can build the conditions we need to analyse in hindsight and avoid being wrecked in a complex world. It is a matter of adding prospection to prediction.
In creative logics no AI is worth much on its own. For now, at least. To claim otherwise, AI would first have to show that it can solve complex problems. François Chollet, a frontier scientist in assessing it through a curious set of tests he updates from time to time, has found no evidence of that to date: any child can solve complex problems, and the most powerful AI cannot.
Solving complex problems, and the capacity to innovate (an idea as little understood as it is distorted, despite filling the news pages every day), are directly linked to the ability to feel a genuine purpose, to have judgement, and to kill off intuitions and hunches. The very ones that lead us to make the same mistake twice, or more.
In Western management, the dominant logics until recently were the predictive ones: strategic planning (a contradiction in terms); competing on cost and incremental improvement, or quality (two dynasties shaped above all by the car industry); the analytics of variables, flows, historical data, trends and projections (what the business schools taught us); classical training, theoretical and generally two-way; and an ever more powerful classical computing, with AI as the second-to-last disruption.
In the paradigm of combined logics, by contrast, the big strategic plans with hundreds of indicators give way to strategic muscle, where the whole organisation (people) is involved in an everyday process rather than in one grand event every four years. Competing on value becomes an opportunity if we know how to generate value (innovating: people). Disruptive improvement (again, innovating: people) is more powerful than continuous improvement. Learning at the root (worked through with diverse people in settings of real experience) counts for more. Classical computing will give way to quantum computing, though it will be a while before we see it... will that be the next bubble, once the current one bursts? How much of AI is bubble, in a year when two of its giants are going public?
The most effective way to face complexity is to activate Collective Intelligence. Does that mean giving up analytics, or the systems of teaching that have proved themselves over thousands of years? Not at all. And using tools, of course. The most sophisticated ones within reach. But without losing our heads. Picture a company that replaced a whole team of people with an AI subscription. What happens if, on a corporate decision taken on the other side of the world, that subscription is cancelled?
The more complex the world, the more we need to activate an architecture of Collective Intelligence inside our organisations. There is still a great deal of smoke to clear out of the system: specifically, all the smoke produced by the people selling it. Those who confuse innovating with R&D or, worse, who sell innovation with no purpose behind it. Or those who talk about leadership while playing blind man's buff with us, sometimes from posts where they never led a team in their lives.
In short, we should go to the sources where knowledge and judgement are produced (universities, for instance, where they are usually there for the asking) and connect collective, diverse intelligence. The kind that has hunches, intuitions and a purpose.
What this means in practice
- We still have creative logics to fall back on.
- The human brain handles complexity better than the machine. By a long way.
- The key is knowing how to activate effective Collective Intelligence processes, with the artificial kind added on top.
"To be surprised, to wonder, is to begin to understand." José Ortega y Gasset
Joaquín Romero Roldán.