Artificial intelligence and human labour, what risks for freedom and creativity

The coexistence of man and artificial intelligence in the world of work after many years of raising doubts and existential fears is taking a definite shape: technology does not replace man's work, it supports it. Artificial intelligence performs the repetitive and standardised tasks, man puts his 'special touch' when his discernment, his sensitivity, his creativity are called upon. It is a pattern that works in so many areas of work, but it has implications for the quality of human labour that will have to be carefully explored.

Recently, I often notice that people around me are asking technology to pre-set emails, documents, presentations. In other words, content is created from increasingly 'high-performance' semi-finished products. The dish is ready, it just needs to be stuffed and garnished as desired and then reheated. This simplification cannot fail to have consequences on the way we work and think. Efficiency is guaranteed, but it is hard to believe that this is a 'free meal', an improvement without a quid pro quo. Even assuming that jobs can always be rigidly divided into standardisable and boring tasks and tasks with high 'human' added value, unlearning repetitive and predictable tasks will certainly negatively affect our ability to handle more sophisticated tasks.

Suppose an intelligent system pre-packages the article of a sports journalist reporting on a football match. An artificial intelligence system reads the data, and presets the story and commentary. If the journalist becomes disaccustomed to setting up the article, over time he will also become disaccustomed to observing critically, establishing priorities and hierarchies, and interpreting episodes. He will lose the capacity for analysis. He will lose language. It will lose skill in interviewing footballers. He will lose effectiveness in constructing a 'peppery' post on social networks or in managing an impromptu technical commentary on television. He will lose all those skills where he is asked to make a difference to an algorithm.

Let us take an even more sophisticated example. Let us assume that artificial intelligence allows a lawyer to record his interview with a client (a relational activity with a high 'human' added value) and to pre-arrange a legal opinion. If that lawyer were to get used to delegating the preparation of his documents to technology, how would his ability to listen to clients, interpret and organise logic change? The banal truth that is known to all but that we are in danger of forgetting is that the human being, in order to perform tasks that cannot be trusted to a machine or a robot, needs 'brain muscles' that are developed by performing precisely those 'basic' activities that we are automating day after day.

The fact that an algorithm can suggest the best response to a customer's request does not mean that responding to a customer is a 'stupid' activity that does not keep the brain active, that does not require us to assess a context, ponder, interpret, execute with precision. Certainly artificial intelligence can 'perform' more reliably than we do, but those mundane tasks that we perform so inefficiently compared to what a robot would do still serve us to build those 'flashes' of intuition, sensitivity, creativity that will be the quintessence of human work in the coming decades. The more we rely on artificial intelligence, the more we lose 'brain muscle' for 'human' value-added work.