Between efficiency and mental overload, the hidden costs of artificial intelligence

TANSTAAFL - There Ain't No Such Thing As A Free Lunch! Thus Milton Friedman popularised a predominantly economic aphorism that emphasises, now as then, that every resource has a cost. I like to recall how Carl Rogers, one of the fathers of humanist psychology and founder of customer-centred therapy (and consequently of 'customer centricity'), also took up this principle in the context of personal growth and development.

Since time immemorial, man has sought to optimise his labours and efforts in order to evolve and progress. Sometimes seeking shortcuts, let us say. The divine condemnation after the expulsion from the Earthly Paradise "you shall eat the fruit of the earth with toil all the days of your life. It shall bring forth thorns and briars, and thou shalt eat the grass of the field; thou shalt eat bread with the sweat of thy face" (Genesis 3:14-24), perhaps it is the drive for redemption to autonomy and the striving for a return to a state of bliss that has most conditioned and emancipated humanity. The quest to optimise energy, resources and time is at the heart of every technical and technological achievement. Progress, however, has often masked the above-mentioned famous costs; they have also become scarcely visible over time because of cumulative effects emerging only in the long run, or because monetising them and making them tangible was difficult due to a lack of KPIs. Artificial intelligence, today, presents itself as an optimiser of time, effort and effort for many professions. The question we can ask ourselves is twofold: on the one hand, whether for it, AI, effort and cost do not exist, and on the other hand, whether it really will relieve us from wasted/costly mental and physical energy.

The answer to the first, deliberately provocative question can be summarised by reminding us that behind every answer there are energy-intensive data centres, complex hardware and global infrastructures. And here I will stop, because this is not an area in which I am an expert. Whereas, as for the second, I can share some thoughts I have gained in the field over the past 18 months and the results of some recent research. They call it "AI brain fry" - literally "brain fried by AI" - and define it as mental fatigue resulting from overuse or supervision of AI tools beyond one's cognitive capacity.

Participants in a Harvard research study (conducted after a series of organisational well-being surveys reported an increase in cognitive overload and burnout) described it as a feeling of 'buzzing' or mental fogginess with difficulty concentrating, slowed decision-making and headaches. The study was conducted on 1,488 full-time employees based in the United States (49% men and 51% women among independent contractors and managers) at large companies in various industries. The research shows that although artificial intelligence promises to make work more efficient, in many cases it is having the opposite effect: instead of simplifying tasks, it increases the cognitive load on workers. This is particularly the case when people have to monitor several AI systems simultaneously or when AI increases the volume of work instead of replacing repetitive tasks. Productivity increases when using one or two AI tools, but decreases significantly if this number rises due to 'trying' multitasking and information overload.

This fatigue also costs the company, concretely: more errors, more 'decision fatigue', hence endless time and communication and a higher propensity to change jobs (related, not causal). Conversely, when AI is used to eliminate repetitive tasks, burnout appears to decrease. In the light of this and countless ongoing research, it is evident how necessary it is to develop new skills to effectively manage work with AI, such as better defining problems, devising new analysis plans and setting strategic priorities. Without these skills, people may risk spending more time managing AI tools than solving real problems. Moreover, a warning for organisations is to treat human attention as a limited resource. Crucial skills such as judgement, decision-making and strategy require focus, which for the human being is not linear or binary, on-off.