The invisible hand of AI – collective intelligence underlying the Frictionless Enterprise

Machines and what they were able to perform, led to the First Industrial Revolution. But with the advent of thermodynamics, physics was able to explain the limits of mechanical performance. These limits did not depend on the technical components of the machines in question, but on the transformations of energy allowed – or not – by physics.

Gaining order always costs energy, and nature constantly tends to dissipate order. Disorder cannot be ordered spontaneously without an investment of energy.

Click on the below buttons to download and subscribe to the full magazine:

Download PDF

Subscribe

Business processes and the frictions within

The First Industrial Revolution gave birth to the concept of operational processes – orderly tasks that mark out a chain of production and value. The discipline of business process management appeared, which aimed at streamlining and optimising these processes within enterprises to avoid inefficiency.

However, over time, inefficiency and friction creep into processes, just as they do in nature. In the digital age, emerging technologies can extract and process massive data sets via digital, cloud-based platforms. This data can be collected and leveraged with a premeditated use to deliver frictionless customer service, for example. Traditional organisations whose data legacy has not been collected or planned with intentional objectives in mind often find this a challenge.

When we apply analytics to data generated by the information systems of these traditional organisations through what we call “process mining,” we discover processes that are very different in reality from originally-defined operational processes. Bottlenecks appear that affect the efficiency of the process. This is an example of friction and the movement to disorder that inevitably takes place – and it tends to have a human origin.

In traditional businesses, humans use and supervise machines as tools rather than operators in  processes. The machines do not supervise operations and do not make decisions. When technologies and tools evolve or humans change, processes change in order to circumvent the problems that are created by these changes. And this is an inevitable decline to disorder in operational processes.

However, in a frictionless business, humans and machines work in symbiosis through new and emerging technologies that have enabled the emergence of the platform economy. As we shall see, artificial intelligence (AI) plays a central role by augmenting, not just by replacing, humans with machines.

AI at the heart of frictionless processes

Before a task can be automated, the process it is part of needs to be completely redesigned. This is the first form of order that must be put into place, so when exceptions occur, they can be managed by humans without friction.

AI can orchestrate machines and humans by predicting operations upstream from a particular process that creates exceptions. These are then assigned to a human operator who handles them correctly. In this way, AI becomes part of the actual architecture that keeps the “human in the loop” to perfect the process managed by machines. In essence, processes are managed by machines and augmented by humans.

This intelligent architecture is another form of order established in the process. AI acts as an invisible hand governing operations between humans and machines to improve the overall efficiency of the process.

There are other ways for AI to intervene in the architecture of a process in order to limit friction. One of them is to embed analytics across a process. By leveraging the data manipulated and generated by machines and humans at each node of a process, each operator can access essential insights into operations executed by other operators, regardless of whether they are machines or humans.

A level of coherence then emerges between the operations executed within the same process.

Collective intelligence and the Frictionless Enterprise

In nature, coherence is one of the main features of effective and frictionless collective decision-making processes. This is used, for instance, by living beings to avoid certain threats. This collective intelligence provides the capacity for adaptation and resilience to change.

As an example, look at the spectacular way in which a flock of birds travels and behaves collectively in a seemingly well-choreographed ballet (see Figure 1).

When a bird first sees a threat, it changes its direction of flight, and this information spreads across the flock. As a result, all of the other birds change their direction of flight accordingly. For the flock to move coherently in this way, each bird gathers insights from its nearest neighbors and follows their average direction.

Step by step, all of the birds follow a coherent trajectory. This is an example of analytics inherently embedded in each bird’s brain that enables the flock to be resilient to external threat and change.

In business, an enterprise is composed of processes, each of which is made up of a combination of humans and machines. To make a process more resilient to change and limit the evolution towards disorder, each operator must be fed with insights from the data generated and processed by all operators in the process.

In our example, each bird is analogous to an operator working in a given node of the process. AI creates coherence between all the nodes of a process. This notion of collective intelligence is fundamental for understanding the concept of the Frictionless Enterprise.

AI as the guardian of order

To build a frictionless business, having access to data and digital is not enough. Neither is replacing humans with machines through leveraging AI. The processes that organise an enterprise must be completely reimagined to be able to implement intelligent automation and create a digitally augmented workforce at scale.

AI is no longer only involved in the execution of tasks, but also in the orchestration of operations between machines and humans. By acting like an invisible hand, processes become smarter and operators more coherently.

In short, AI acts as the guardian of order at different levels of the process – providing resilience to change that are the sources of frictions impacting the effectiveness of the process.

Taoufik Amri is a former quantum physicist. He graduated from the Ecole Normale Supérieure in France with a Ph.D. in quantum and statistical physics. As the principal data scientist for Capgemini’s Business Services, Taoufik advises Capgemini’s clients on implementing ready-to-use deep tech and intelligent automation solutions to dramatically accelerate the optimisation of our clients’ business operations. Taoufik also identifies tasks that can be performed better and/or faster with AI, measures the value added by AI with advanced quantitative business process models, and designs human-in-the-loop solutions.