Technology allows us to do more with less. But if efficiency becomes the measure of everything, we risk losing sight of what we are trying to achieve.
Efficiency is one of the most persuasive arguments for technological progress. When a new technology allows us to accomplish something faster, more accurately or with fewer resources, we tend to regard it as an improvement. The logic appears straightforward: if we can achieve the same result with less effort, we have created value.
Artificial intelligence has given this logic renewed force. Organisations can automate administrative processes, accelerate research, generate content and analyse information at a scale that would previously have required considerable human effort. Activities that once occupied entire departments can increasingly be supported by software, while individuals can complete tasks in minutes that previously took hours.
These developments offer substantial opportunities. Greater efficiency can make services more affordable, reduce unnecessary work and free resources for activities that matter more. In sectors facing financial pressure or shortages of qualified employees, technological improvements may be essential to maintaining or expanding services.
Yet efficiency also has a peculiar ability to dominate the conversation about technological change. Once we begin measuring success through time saved, costs reduced or output increased, other questions can become secondary. We become very good at asking whether a process can be made faster without necessarily asking whether it produces the outcomes we actually want.
This is not a problem with efficiency itself. It is a problem with allowing one value to stand in for all the others.
A school can become more efficient without providing a better education. A hospital can treat more patients without improving their experience of care. A business can increase productivity while making work less meaningful for its employees. And a public institution can reduce administrative costs while making its services more difficult for some citizens to access.
None of these outcomes is inevitable, and efficiency improvements can contribute to better results in all these settings. But the examples illustrate an important distinction: producing more output is not necessarily the same as creating more value.
As technology becomes capable of transforming an increasing range of human activities, that distinction deserves greater attention.
When the measure becomes the objective
Organisations need ways to evaluate performance. Budgets, productivity figures, processing times and other measurable indicators help decision-makers understand how resources are being used and whether activities are producing the intended results. Without such measures, it would be difficult to manage complex institutions or assess the consequences of technological investments.
The challenge arises when indicators that were originally intended to support decision-making begin to define what success means.
Consider a public service that introduces an automated system to handle citizen enquiries. The organisation may measure its performance through the number of enquiries resolved, the average time required to process them and the proportion handled without human assistance. These indicators can provide useful information about operational efficiency.
But they do not necessarily reveal whether citizens understand the information they receive, whether complex situations are handled appropriately or whether people who need additional assistance can obtain it. A system may perform exceptionally well according to its operational targets while creating difficulties that remain invisible to those targets.
The problem is not that the measurements are incorrect. It is that they describe only part of the value the service is supposed to create.
Similar dynamics can emerge in education. If the performance of an educational institution is assessed primarily through examination results, completion rates or the number of students taught per employee, activities that contribute to learning in less immediately measurable ways may receive less attention. Developing confidence, curiosity, judgement and the ability to work with others can be essential educational outcomes, even when they are difficult to capture in a single performance indicator.
In business, productivity measurements can encourage improvements that genuinely benefit employees and customers. But they can also reward the production of greater quantities of work without sufficient consideration of quality, relevance or long-term consequences. A team may become more productive according to its internal measurements while producing outputs that contribute little to the organisation’s wider objectives.
This creates a risk of confusing what is measurable with what matters. The more influential a performance indicator becomes, the more likely it is that decisions will be organised around improving it. Activities that contribute to the indicator are prioritised, while other forms of value may gradually lose visibility.
Artificial intelligence could intensify this tendency because it makes many forms of output easier to produce and measure. Organisations may be able to generate more reports, respond to more enquiries and complete more administrative tasks than ever before. But an increase in output does not, by itself, tell us whether those activities have become more valuable.
The important question is not simply whether technology improves our performance against existing measures. It is whether those measures reflect what we ultimately want to achieve.
The value of what takes time
Some of the activities we consider most valuable are difficult to accelerate without changing their character.
Learning is one example. Acquiring knowledge involves more than receiving information. It often requires sustained attention, uncertainty, repetition and the effort of connecting unfamiliar ideas. Technology can support these processes by making information more accessible and providing new opportunities for explanation and feedback. But making the production of an answer faster is not necessarily the same as making learning more effective.
A student who uses artificial intelligence to generate a sophisticated argument may produce a better-looking assignment in less time. Whether the student has developed a deeper understanding of the subject is a different question. If the educational purpose of the activity is to develop reasoning and judgement, the time spent struggling with ideas may be part of its value rather than an inefficiency to eliminate.
Professional work presents similar dilemmas. A doctor may benefit from technology that reduces documentation and provides relevant information more quickly. Yet the time required to listen to a patient, understand their circumstances and establish trust cannot always be treated as unnecessary overhead. A social worker may become more productive when administrative tasks are automated, but the quality of their work may still depend on conversations and relationships that cannot be reduced to a standardised process.
These examples do not imply that slower processes are inherently better. Waiting times, unnecessary bureaucracy and repetitive tasks can create frustration and prevent people from receiving the attention they need. The point is that different activities create value in different ways, and some of that value depends on the process rather than merely its outcome.
An organisation that recognises this distinction may make very different decisions about how to use technology. Instead of automating every activity that consumes time, it might prioritise those that prevent people from spending time on more valuable work. Rather than treating human interaction as a cost to be minimised, it might use automation to make meaningful interaction more accessible.
In this sense, efficiency and other values do not necessarily compete. Greater efficiency in one part of a system can create opportunities to invest more time and attention elsewhere.
But that outcome depends on deliberate choices. Saving time does not automatically mean that the time will be used to create greater value.
Who benefits from greater efficiency?
Efficiency is often discussed as though its benefits are self-evident and universally shared. A process becomes faster, an organisation reduces costs or a service requires fewer resources, and the improvement is presented as a collective gain.
In practice, the consequences of efficiency improvements depend on whose perspective we consider.
An automated customer service system may reduce an organisation’s operating costs while allowing customers to resolve routine problems more quickly. That can be a genuine improvement for both parties. But customers with unusual or complex problems may experience longer and more frustrating interactions if access to human assistance becomes more difficult.
A workplace system may help managers coordinate activities more efficiently while increasing the demands placed on employees. Technology that reduces the time required to complete individual tasks may create expectations that workers should complete more tasks rather than enjoy greater flexibility or opportunities for professional development.
Conversely, automation may be particularly valuable for people who have previously encountered barriers. A digital service that can be accessed independently may eliminate the need to rely on assistance from others. Speech recognition, automated transcription and other assistive technologies can reduce the effort required to perform activities that many people take for granted.
The same technology can therefore produce very different forms of value depending on the circumstances of those using it.
This makes efficiency a question of distribution as well as productivity. When an organisation saves time, whose time is being saved? When costs are reduced, who receives the benefit? When a process is simplified, does it become easier for everyone involved, or is complexity transferred from one group to another?
These questions are particularly important when technology is introduced into public services and other essential systems. An efficiency improvement that benefits the institution may still be undesirable if it creates significant barriers for people who depend on the service. Equally, maintaining an inefficient process in the name of personal service may disadvantage people who would benefit from greater independence and flexibility.
There is no universal rule that determines which outcome is preferable. The appropriate balance depends on the purpose of the service, the people affected and the alternatives available.
What matters is that these perspectives are considered before efficiency is treated as proof of improvement.
Artificial intelligence and the temptation to optimise everything
Artificial intelligence introduces a new dimension to the pursuit of efficiency because it can support activities that were previously difficult to standardise or automate.
Traditional automation has often focused on repetitive physical or administrative processes. AI increasingly extends technological assistance into analysis, communication, planning, creativity and decision-making. These are activities in which value is not always easily separated from the judgement and experience involved in producing it.
As AI systems become more capable, organisations will encounter growing opportunities to accelerate processes that have historically depended on human expertise. Reports can be drafted more quickly, information can be summarised automatically, and recommendations can be generated from large volumes of data.
These capabilities can improve the quality and accessibility of work. A professional may gain access to information that would otherwise have been difficult to obtain. A small organisation may be able to perform analyses that previously required specialist resources. People who struggle with conventional interfaces may find new ways to communicate, learn and participate.
But the availability of these capabilities can also influence how organisations define their objectives. If technology makes it possible to produce reports at a fraction of the previous cost, the immediate response may be to produce more reports. If AI enables employees to complete routine analyses more quickly, the expectation may become that they should process more cases or manage larger workloads.
This approach assumes that the existing activity is worth expanding. It treats technological progress primarily as a way to increase the volume of familiar outputs rather than an opportunity to reconsider whether those outputs are the most valuable use of people’s time and resources.
An organisation that can generate twice as many reports should perhaps ask whether it needs all those reports. A public institution that can process applications more quickly might also consider whether the application process could be simplified. A business that can automate parts of its customer service might ask whether the technology could help it understand customers better rather than simply reduce the cost of interacting with them.
These are questions about purpose rather than performance.
AI also raises questions about the development and maintenance of human capabilities. When technology takes over activities through which people learn, practise judgement or develop expertise, immediate productivity gains may be accompanied by longer-term changes in what individuals and organisations are capable of doing independently.
This does not mean that AI should be excluded from such activities. It means that technological assistance should be designed with an understanding of what people need to learn, decide and influence.
The objective should not be to preserve every existing task, but to ensure that the pursuit of efficiency does not unintentionally undermine the capabilities on which future value depends.
What if we measured possibility instead?
Efficiency is relatively easy to understand because it compares inputs and outputs. If a process produces the same result with fewer resources, it has become more efficient. The concept provides a useful basis for evaluation, even when the measurements involved are imperfect.
Possibility is more difficult to measure.
How do we evaluate whether a technology has expanded someone’s independence, strengthened their ability to make decisions or created opportunities that did not previously exist? How do we assess the value of improved relationships, greater participation or the ability to explore ideas that were once beyond an organisation’s reach?
Some of these outcomes can be measured. Others require qualitative assessment, longer-term observation or direct engagement with the people affected. They may also involve competing priorities that cannot be reduced to a single indicator.
This complexity is not a reason to ignore them.
Consider an organisation introducing AI to support employees with administrative work. A conventional efficiency assessment might focus on hours saved, costs reduced and the number of tasks completed. A broader evaluation could also examine whether employees have greater opportunities to exercise professional judgement, whether the quality of services improves and whether people with different abilities can participate more effectively in the workplace.
Similarly, a public institution adopting automated services could measure processing times and operating costs alongside accessibility, user understanding, opportunities for meaningful assistance and the experiences of people with complex needs.
A company developing an AI-powered product might assess technical performance and commercial potential while also exploring whether the product enables customers to do things that were previously difficult or impossible.
Such evaluations will not always produce simple answers. Different outcomes may conflict, and organisations will still need to make decisions about priorities and resources. But the process can reveal opportunities that remain invisible when efficiency is the dominant measure of success.
It can also change the questions asked during technological development. Instead of beginning with how an existing process can be accelerated, an organisation might ask which capabilities it wants to strengthen, which barriers it wants to remove or which new forms of value it hopes to create.
The distinction matters because the questions we ask influence the solutions we consider.
If efficiency is the objective, the most attractive solution may be the one that produces familiar outputs at the lowest cost. If greater human capability is the objective, the most attractive solution may be one that enables people to make better decisions, participate more independently or pursue opportunities that were previously unavailable.
Neither perspective should automatically exclude the other. But recognising that they are different creates room for more ambitious technological choices.
Better decisions require different perspectives
Deciding what constitutes value is rarely a purely technical question. It involves assumptions about people’s needs, organisational priorities and the outcomes society considers desirable.
These assumptions are influenced by the experiences of the people making decisions. A manager may understand the financial pressures facing an organisation, while an employee may recognise how a proposed system will change everyday work. A technology developer may identify new capabilities, while a person who depends on a service may understand barriers that are invisible in technical specifications.
Different generations, professional backgrounds and lived experiences can reveal aspects of technological change that would otherwise be overlooked. This is particularly relevant when decisions concern accessibility, human agency or the organisation of essential services.
A system that appears efficient to its developers may be difficult for some people to use. A solution designed to reduce workload may remove activities through which employees develop expertise. An innovation intended to increase productivity may create opportunities for independence that were not part of the original business case.
These consequences cannot always be predicted through technical analysis alone. They require an understanding of how technology interacts with people’s circumstances, capabilities and expectations.
Bringing different perspectives into technological decisions is therefore not simply a matter of consultation or representation. It can help organisations develop a more complete understanding of the value they are trying to create.
That requires participation early enough to influence the questions being asked. If people are invited to comment only after a technology has been selected and its objectives defined, their ability to challenge the underlying assumptions may be limited.
The more useful approach is to involve relevant perspectives while meaningful alternatives remain available. Different experiences can then contribute not only to identifying problems with a proposed solution, but also to imagining possibilities that were not initially considered.
This does not guarantee agreement or eliminate difficult trade-offs. People may legitimately value different outcomes, and some priorities will remain in tension. But acknowledging those differences is preferable to allowing a single performance measure to determine what counts as progress.
Progress should mean more than doing the same things faster
Technological progress has repeatedly transformed what people and organisations are capable of achieving. Activities that once demanded substantial time and resources have become accessible, affordable or almost effortless. These developments have contributed to improvements in living conditions, communication, knowledge and economic opportunity.
Efficiency has played an important role in that history and will continue to do so. There is little reason to preserve unnecessary bureaucracy, repetitive work or avoidable costs simply because they are familiar. When technology can remove such burdens, it can create genuine value.
But the most significant opportunities may emerge when we move beyond the ambition of performing existing activities more efficiently.
Artificial intelligence could allow organisations to provide services that were previously too expensive or complex to offer. It could help people with different abilities participate more independently in education, work and society. It could give professionals access to knowledge and analytical capabilities that expand what they can accomplish. It could also create opportunities to rethink how institutions operate and how people collaborate.
Realising these possibilities requires more than adopting technology and measuring its immediate performance. It requires decisions about which outcomes deserve attention, whose experiences should influence development and how the benefits of technological change should be distributed.
An organisation might use efficiency gains to reduce costs, improve services, strengthen employee capabilities or develop entirely new activities. These are different choices, even when they begin with the same technological improvement.
The technology does not determine which choice is made.
That is why efficiency should be understood as a means rather than an end. It can help us use resources more effectively, but it cannot tell us which purposes are worth pursuing. It can demonstrate that a process has become faster, but not necessarily that the result is more meaningful. And it can reveal how much effort has been saved without explaining what we should do with the time and capabilities that have become available.
As artificial intelligence expands the range of activities we can automate, accelerate and optimise, these distinctions will become increasingly important.
We should welcome technologies that reduce unnecessary work and make valuable activities easier to perform. But we should also be willing to question whether the activities we are optimising are the right ones, whether our measures of success capture what matters and whether technological progress is expanding the possibilities available to people.
The ambition should not simply be to create a world in which everything happens faster and more efficiently. It should be to use technological progress to make more valuable things possible.
Efficiency is a value. It isn’t the only value. And if we want technology to create a better future, we need to become just as ambitious about what we achieve as we are about how efficiently we achieve it.
A question worth discussing
Think about a technological improvement in your workplace or everyday life that has made an activity faster or more efficient. Has it also made the experience better, strengthened your capabilities or created new opportunities? And if not, what would it take to turn that efficiency gain into something more valuable?
