AI, acceleration, productivity, work... fracture ?
- David Gateau
- Feb 20
- 4 min read
During this week of celebrations for Artificial Intelligence, when ‘decision-makers’, politicians and entrepreneurs, are gushing about its benefits and the spectacular progress we can expect from it in all areas (economy, health, industry, etc.), other voices are being raised to warn us about a number of risks: impoverishment of thought (Gaspard Koenig), replacement of certain jobs (economists), increased energy consumption, etc.
Among the positive reflections on AI, I often read ‘AI does not replace humans, it enhances them: used wisely, it frees up time for higher added-value tasks’.
Based on this assertion, and far from being sceptical or averse to technology in any way, I'd like to share with you some pragmatic thoughts on the subject, and in particular on the consequences for work:
Let's imagine that you work in the service sector and that you currently produce 1000 units of value without help.
In the first year, the AI relieves you of 10% of your work and therefore takes over 10% of the value produced (100 units of value).
Assuming that the AI's phenomenal progress relieves you of 10% of your tasks each year, the machine will gradually take on some of the value you produce until, in year 10, it potentially takes on 90% of the value.
Only 10% of the value of the work you did 10 years ago will be attributable to you !!!!!
Several possible and logical consequences emerge:
We will soon realise that the machine has considerably increased productivity. The 10 people who, like you, produced 1000 units of value will be reduced to 1 person whose role will be to control the machine. The value ‘produced and attributable’ for this person will remain at 1000. Your remuneration will remain the same, while the other 9 people will have to find another activity.
To compensate for the 90% that will no longer be attributable to you, you will also be able to concentrate on so-called higher added value tasks, we're told, i.e. the 10% you have left! You're going to have to be pretty inventive to justify the 90% of your tasks that have been taken away from you by the machine being compensated for by brain juice, or else by a 10-fold increase in your ability to generate value.
The probability that you will be able to convert the 90% taken by the machine into a reduction in working time for your leisure activities will remain marginal, let's be aware of that in the world we live in.
If we look back at the history of work, and in particular the evolution of the primary sector, the parallels that can be drawn are striking. Having worked in a cooperative for some time, the farmer's job has been transformed into that of a company director connected to world prices for maize or wheat to sell his harvest at the best time. It's a tall order in terms of skills development, and not for everyone. The farming sector already accounts for just 2% of the population, and only a few happy few who have managed to turn the corner are making a go of it.
A second effect is the devaluation of jobs, since a large part of the value placed on the human being will be replaced by the value placed on the machine. The recognition of gestures, know-how and the ability to fetch and sort will be transformed and impoverished in our representation of the value provided.
For example, in a post dated 14/01/2025, François-Xavier de Vaujany said of consultants: ‘Can we pay for hours or a fixed fee at the same level as in the past, in a world where generative AI is multiplying? Since they use AI, consultants should invoice their services on 2 lines: One line for human work, one line for AI! On the one hand, this recognises the value of skills, but on the other hand it devalues research and other acts assisted by machines. Increased productivity in the service sector will also mean lower prices for the value produced. Colleagues, we're going to have to produce more to compensate for lower turnover, just as we do in the product world (cf. the famous price curve for video recorders)! The same reasoning could also apply to researchers, teachers, etc. in preparing their lectures, marking papers, writing articles, etc. in order to make more teaching hours! No ?
And that's the third effect: increased ‘production’. If prices fall, we have to compensate with more production to generate and maintain at least the same value base. Acceleration once again.
AI redefines work
AI is not just a productivity tool; it is a powerful engine in a dynamic of continuous acceleration, as Hartmut Rosa describes so well [Acceleration - a critique of modern times]. Past industrial revolutions have reshaped our societies, and AI represents a new groundswell before the next one, already announced (the advent of quantum computers), pushes these dynamics to unprecedented levels.
The major problem lies in the fact that these accelerations impose an increasingly rapid pace to which humans are struggling to adapt. The growing gap between technological innovation and human capacity to adapt will create a social divide, with individuals unable to keep pace.
AI and thinking about the future of work
AI has immense potential to improve productivity, but it also poses major challenges. While technological acceleration will enable a spectacular increase in human performance, it will also bring with it a series of social and economic consequences: a transformation of jobs, a redefinition of the value of human work, and growing inequalities between those who can adapt to this rapid change and those who cannot.
It is crucial that we begin (calmly?) to think seriously about the implications of this transformation. How can we redistribute productivity gains? How can we support individuals and communities in the face of this constant acceleration? And above all, how can we preserve the human element in a world where ‘machines’ will be increasingly central?
It is becoming essential to take these transformations into account if we are to build a more equitable and sustainable future, and that is the purpose of the few voices that are being raised to alert us to this.
Comments