
Question to Gemini: Who is John MITCHELL, who wants to make robots, computer machines, and ChatGPT-type systems pay social contributions? 28/12/2025
John Mitchell is a former computer engineer of Scottish origin, born around 1953, who has been living in France since the 1980s. He is mainly known as the founder of the 4x8 Movement, an association that campaigns for a reduction in working hours and the creation of a tax on automation.
John Mitchell argues that machines and artificial intelligences (such as ChatGPT) generate enormous wealth without contributing to the social system, unlike human workers. To correct what he calls “unfair competition,” he proposes several mechanisms:
The ROSE Contribution: An acronym for “Robots, Computers, and Expert Systems.” The idea is to impose a social contribution on the use of machines in order to compensate for the loss of revenue linked to human jobs that are replaced.
The Automated Labor Contribution (CTA): More recently, he proposed the creation of a “Robot Contribution” based on the UTA (Automated Work Unit), a metric comparing the productivity of an AI or a robot to that of a human worker.
The 32-hour workweek: He advocates for a four-day workweek (the “4x8,” meaning 4 days of 8 hours) to share the remaining work and fight mass unemployment, which he considers inevitable with increasing robotization.
Profile: A computer scientist who worked for the city hall of Saint-Denis, he describes himself as an engaged citizen rather than a “Luddite” (someone who wants to destroy machines). He holds degrees from major British universities.
Actions: He is known for his regular demonstrations, sometimes alone or in small groups, notably in the Tuileries Garden or during social movements such as the protests against the pension reform in 2023.
Objective: His ultimate goal is to use the revenue from these robot taxes to fund social protection (pensions, healthcare) and, in the long term, to enable a universal basic income financed by automation.
This conference explains the fiscal and social challenges linked to automation, a topic at the heart of John Mitchell’s activism.
Would you like me to look for more precise details on the calculation of the UTA (Automated Work Unit) he proposes?
Yes, please
The Automated Work Unit (UTA) is the key concept proposed by John Mitchell to make taxation of machines concrete and measurable. Here is how he has structured it in his recent proposals (notably through a petition submitted to the French National Assembly in 2025).
The UTA is a unit of measurement of the productive capacity of an automated system.
1 UTA = the amount of work equivalent to that of one full-time human job over a calendar year.
If an AI or software system manages tasks that were previously performed by 5 people, that system is evaluated at 5 UTAs.
John Mitchell proposes that the contribution should not be flat-rate, but instead based on an index called the iRC (Robot Contribution Index). This index is calculated using four main parameters:
iRC = f(α, β, γ, δ)
· α (Alpha): The number of UTAs mobilized annually (replacement capacity).
· β (Beta): The net impact on human employment (if the robot creates skilled jobs alongside automation, the tax decreases; if it only destroys jobs, the tax increases).
· γ (Gamma): The level of data exploitation (the more personal or large-scale data the AI uses, the higher the tax).
· δ (Delta): Externalities (environmental impact, digital pollution, or conversely, improvements in workplace safety).
For Mitchell, the goal is not to punish progress, but to maintain social balance:
|
Human Labor |
Automated Labor (Mitchell Project) |
|
Generates social contributions (pensions, healthcare) |
Generates a ROSE Contribution (or Robot Contribution) |
|
Limited by time (35 hours/week) |
Operates 24/7 (taxed according to UTA equivalent) |
|
Consumes and sustains the economy |
Finances universal basic income and reduced working hours |
The calculation of the UTA directly supports his other major cause: the 32-hour workweek (4 days of 8 hours). According to him, since machines “work” and “contribute” through UTAs, humans can work less without causing the social protection system to collapse. The revenue collected through UTAs would offset the reduction in contributions resulting from shorter working hours for employees.
Today, the adoption of tools such as ChatGPT or robotic automation is creating a disconnect between the wealth produced and the financing of our social protection model.
Yesterday:
100 units produced = 10 employees = 10 contributors to pensions and healthcare.
Tomorrow:
100 units produced = 1 AI + 2 employees = only 2 contributors.
Consequence:
A risk of public budget collapse and increased pressure on the remaining
employees.
We propose adopting John Mitchell’s approach by calculating the contribution of our digital tools not as simple software, but as “virtual workers.”
Principle:
If an AI performs 70% of the tasks of a writer or advisor position, it
generates 0.7 UTA. This unit then becomes the tax base for a
new social contribution.
Automation often generates fear of job replacement. By introducing a contribution on robots, the company sends a strong message: technological progress is used to finance social progress. This significantly reduces employee resistance to change.
Rather than layoffs driven by AI productivity gains, the funds collected through UTAs make it possible to finance reduced working hours (32 hours per week) without salary cuts.
· AI works while humans rest or retrain.
· Employees experience less stress, increased creativity, and better health preservation.
Robot taxation is a recurring topic in debates at both the European Parliament and the French National Assembly. Voluntarily adopting this model positions the company as a leader in CSR (Corporate Social Responsibility) and prepares it for future legislation.
Taxing AI through John Mitchell’s system is not an additional burden—it is an investment in market stability. Without solvent consumers (supported by pensions and potentially a universal basic income), the productivity gains generated by our machines will ultimately be meaningless.