Who Owns the Time and Surplus Value Saved by AI?
Saved time does not belong to anyone automatically
When artificial intelligence reduces a four-hour task to two hours, two hours of productive capacity have been released.
Those hours can become rest, learning, shorter working days, higher wages, or greater autonomy. They can also become additional assignments, faster deadlines, smaller teams, lower prices, higher margins, or more continuous availability.
The time has been saved technically. Its ownership is a social and economic question.
AI does not decide where the time goes. The organization that owns the system, controls the workflow, and sets the targets usually has the strongest influence over its destination.
Marx and the productivity dividend
Karl Marx’s analysis of machinery and surplus value provides a useful framework for this question. A machine can reduce the labor time required to produce a good or service, but the benefit does not automatically return to the people whose work has become more productive.
One possibility is relative surplus value: productivity reduces the labor time required for production while the organization retains control over the working day and the resulting output.
In modern terms, AI may allow a company to produce more with the same workforce, serve more customers without proportional hiring, or reduce labor costs while keeping the productivity gain inside the organization.
The issue is not whether AI creates useful capacity. It is whether that capacity becomes a social benefit or an additional source of private accumulation.
Thompson and the ownership of working time
E. P. Thompson’s account of industrial time helps show why the distribution of saved time is so important.
When work is measured through schedules, deadlines, and performance intervals, time becomes something that can be controlled. A faster process may shorten the time needed for one task without shortening the worker’s day.
AI can intensify this arrangement. A response that once required a day may be expected within an hour. The saved time appears to belong to the organization because it is immediately placed back into the production schedule.
The worker may experience no additional freedom at all. The only visible change is that the next task arrives sooner.
Braverman and the control of released capacity
Harry Braverman’s analysis of management helps explain why ownership over the workflow matters as much as ownership over the machine.
When a system defines how work is organized, it can also define what happens to the capacity released by efficiency. The worker may perform tasks faster, but management decides whether the extra time becomes a break, a training opportunity, a new assignment, or a reason to reduce staffing.
The worker produces the gain, but the organization controls its use.
This is especially significant when AI absorbs parts of professional judgment. The worker may become more capable with the tool while losing authority over the process that determines how that capability is deployed.
Where can the AI productivity dividend go?
| Destination of saved capacity | Immediate result | Who generally benefits most |
|---|---|---|
| Shorter working hours | More time outside production | Workers and their communities |
| Higher wages | Greater income from increased productivity | Workers |
| More output | Greater volume with the same staffing | Customers and owners, with more pressure on workers |
| Smaller teams | Lower labor costs | Owners and investors |
| Lower prices | Cheaper access to goods or services | Customers, if savings are passed through |
| Higher margins | More revenue retained by the organization | Owners and investors |
| Training and resilience | Stronger future capability | Workers, organizations, and society |
Gray, Suri, and Casilli: whose labor supports the saving?
Mary L. Gray and Siddharth Suri’s work on ghost work reminds us that the time saved for one visible user may depend on hidden labor elsewhere.
People prepare data, evaluate outputs, correct errors, moderate content, and handle cases that the system cannot resolve. Antonio Casilli’s research on digital labor shows how these tasks can be fragmented and placed outside the visible product.
A company may describe AI as saving labor while relying on a distributed workforce to make that saving possible. The visible worker receives assistance, but another group may be performing the preparation and correction work behind the system.
The productivity dividend cannot be calculated honestly if the labor required to create the saving is excluded from the account.
Crawford and the costs behind the saved minutes
Kate Crawford’s analysis of AI adds the infrastructure back into the calculation.
The two hours saved by a worker do not appear from nowhere. They depend on computing power, data centers, energy, cooling, hardware, engineering, maintenance, and global supply chains.
If these costs are ignored, the system may appear more efficient than it actually is. The organization counts the time saved at the interface while the material and human costs remain distributed elsewhere.
The productivity dividend may therefore be divided between the organization that gains time, the infrastructure that makes the gain possible, and the workers who absorb the hidden maintenance and support burden.
Pasquinelli and the ownership of collective intelligence
Matteo Pasquinelli’s work helps extend the question beyond individual employment.
AI systems draw on language, cultural knowledge, professional practice, classification, design, and problem-solving accumulated through collective activity. The model may produce an answer in seconds, but its capability has a much longer social history.
If collective intelligence is converted into a privately controlled system, the saved time may be captured by the owner of the model rather than by the communities and workers whose knowledge made the capability possible.
AI does not own the value in a human or legal sense. The relevant owner is the institution that controls the model, the data, the infrastructure, and the terms of access.
When saved time stops belonging to the worker
Saved time stops belonging to the worker when the organization treats it as unused capacity that must immediately be filled.
This can happen through additional assignments, shorter response windows, higher quality requirements, reduced staffing, or new expectations of availability.
The worker may feel that AI has made the job easier while also experiencing greater pressure. The system has reduced the time required per task, but it has not reduced the amount of time the worker must remain available for production.
Efficiency has released capacity. Management has already claimed it.
What would shared ownership look like?
Shared ownership does not require every worker to own every technical component of an AI system. It requires workers to have meaningful influence over how the gains are distributed.
That may include shorter working hours, higher compensation, protected recovery time, participation in deployment decisions, access to training, professional autonomy, and a clear share in the value created by increased productivity.
It also requires accounting for the hidden human and material work that makes AI possible.
A system cannot claim to be efficient by excluding the conditions that sustain its efficiency.
The question behind AI productivity
AI can save time. The political and economic question begins after the saving occurs.
Does the time return to workers as freedom? Does it become better service for customers? Does it strengthen collective capacity? Or does it become more output, fewer employees, and greater returns for owners?
The answer reveals who owns the productivity dividend.
TravelIAQ Smart Tip: Whenever AI saves time, record where that time goes. Compare working hours, wages, workload, staffing, deadlines, prices, margins, training, and autonomy before and after adoption. The destination of the saved time shows who actually owns the gain.