• The Director of Orange’s School of Management, Philippe Trimborn, points out that time saved on one task is usually reinvested in other professional activities. A state of affairs driven by Parkinson’s law (work expands to fill the time allowed) and further compounded by massive growth in digital communications that may be a cause of burnout.
• Studies by Adecco and BCG have shown that only 21% to 27% of employees use the time saved by AI for their personal lives. The vast majority use it to increase both the volume and quality of their professional output.
In 1930, Keynes predicted that growing productivity would lead to a 15-hour working week by 2030. In theory, generative AI should save us time that may enable us to achieve this goal, but the reality is much more complex: an American study entitled AI and the Extended Workday: Productivity, Contracting Efficiency, and Distribution of Rents by the National Bureau of Economic Research has highlighted how many employees using AI tools are in fact busier than ever. The phenomenon has intensified since the introduction of ChatGPT in late 2022: high exposure to generative AI is associated with an average increase of 3h 15m of time spent at work per week and an equivalent drop in leisure time. The researchers point out that this effect is mainly present when AI serves as a complement to human labour rather than a substitute for it. AI boosts productivity for specific tasks, but that doesn’t mean it frees up time; instead, it encourages workers to take on more tasks or meet higher standards for quality and responsiveness.
Digital multiplexing requires us to voluntarily divide our attention. Our brains can easily filter out background noise in a café, but focusing on multiple screens requires a constant effort that drains more energy.
The intensification of work
“Naturally, the initial focus was on productivity gains,” explains Philippe Trimborn. “But nobody was asking questions about new activities that would be required to use these tools. AI is not a magic wand. You need to learn how to use it, vet the data, orchestrate a workflow, verify results, correct errors, and — more often than not — start all over again. This intensification of work has been compounded by digital ‘multiplexing’. You look at other screens while you are on a videocall… You listen to work related podcasts in the Metro. The pace of our interaction with our environment is dictated by digital technology.” The fragmentation of our attention “engenders a mental load that makes it harder for us to keep control of the time factor in our work.”
Freeing up time for more work?
Referencing the American study, which highlights workers’ limited bargaining power, Philippe Trimborn cites Parkinson’s law: “Work is like a gas: it expands to fill all of the available space.” When it is used to expedite simple tasks, AI can save time, but it also requires us to take charge of supplementary activities that may be “more elaborate and more difficult to manage,” which result in a greater mental load. “In an organization like Orange, where 70% of the staff are white-collar professionals on fixed-day contracts, daily working hours are no longer measured. This leads to a cycle where any efficiency gain is immediately reabsorbed: workers simply use the ‘freed’ time to perform additional tasks.” This observation is confirmed by a 2024 Adecco study, which found that only 21% of employees said they devoted time saved by AI to personal activities. The BCG AI at Work study also highlighted a similar statistic: 47% of respondents reported that AI enabled them to save one hour per day, but 54% of those used that time to perform more professional tasks, while only 27% indicated that they reassigned it to non-work related activities.
Preserving meaningful work
Given that it may undermine a positive work-life balance, Orange is training its teams “to track the human impact of AI.” It is viewed as an organizational rather than a technological challenge which requires a new perspective with a focus on value rather than time. “It is no longer a question of reducing the amount of time required to complete tasks, but a question of pursuing the right objectives and being sure that our time is well used. With AI, you can get a lot done in a day, but how does it affect our mental load and our perceptions of being engaged in meaningful work?” In conclusion, Philippe Trimborn points out that “someone who is bored at work and someone who finds work fulfilling will be subject to very different mental loads and have very different perceptions of time spent in the workplace. Cognitive engagement is key. Going forward, the goal is to ensure that the technology is sustainable, acceptable, and positively perceived by employees, but that will require taking a step back to ensure that the question of work and the future of our work is central to discussion on the adoption of AI.”
Philippe Trimborn
