History has a habit of celebrating breakthroughs while overlooking the assumptions that made them possible. Every generation remembers the invention that transformed an industry, the technology that accelerated productivity or the platform that connected millions. Yet far less attention is paid to a quieter question that often determines whether progress truly benefits everyone in society: Who was this system actually built for?
Modern societies increasingly rely on systems that promise speed, efficiency and scale. From Artificial Intelligence (AI) and digital platforms to health care, education and financial services, success is frequently measured through numbers that are easy to quantify—higher engagement, faster delivery, lower costs and greater productivity. These metrics matter, but they can also create the illusion that if a system is performing well, it must be working well for everyone. In reality, efficiency and inclusion are not always the same thing.
Every system begins with assumptions. Designers imagine a typical user, policymakers anticipate common circumstances and engineers optimise for expected behaviours. These assumptions are often practical rather than discriminating, but they shape who finds the system intuitive and who struggles to participate. When those assumptions remain unchallenged, they quietly become barriers for anyone whose circumstances fall outside the imagined norm.
This is particularly visible in the digital economy. Recommendation algorithms, for example, are designed to identify patterns that maximise engagement. They often reward consistency, production quality and audience behaviour that has previously proven successful. On the surface, such systems appear impartial because they process vast quantities of data rather than human opinion. Yet algorithms can only recognise the signals they have been trained to value. If those signals primarily reflect urban, English-speaking or digitally well-connected communities, creators from different linguistic, cultural or economic backgrounds may find themselves overlooked despite producing equally compelling work. The system has not malfunctioned; it has simply reproduced the priorities embedded within its design.
{{/usCountry}}This is particularly visible in the digital economy. Recommendation algorithms, for example, are designed to identify patterns that maximise engagement. They often reward consistency, production quality and audience behaviour that has previously proven successful. On the surface, such systems appear impartial because they process vast quantities of data rather than human opinion. Yet algorithms can only recognise the signals they have been trained to value. If those signals primarily reflect urban, English-speaking or digitally well-connected communities, creators from different linguistic, cultural or economic backgrounds may find themselves overlooked despite producing equally compelling work. The system has not malfunctioned; it has simply reproduced the priorities embedded within its design.
{{/usCountry}}The same pattern extends far beyond technology. Agriculture offers a compelling illustration. Advances such as precision farming, improved irrigation methods and hybrid seeds have significantly increased productivity across many regions. However, these innovations often presume access to stable electricity, financial capital, reliable infrastructure and predictable supply chains. For smallholder farmers working on fragmented landholdings or in areas with inconsistent resources, adopting these technologies can prove far more difficult. Progress exists, but its benefits are unevenly distributed because the system was designed around conditions that are far from universal.
Even the physical spaces people inhabit reveal similar assumptions. Buildings, workplaces, transport systems and recreational facilities communicate who belongs long before anyone speaks. A poorly designed public space can discourage participation just as effectively as an explicit rule. Inclusion is therefore not solely a matter of policy; it is also a consequence of design. The arrangement of a room, the language used on a website or the accessibility of an interface all send subtle signals about whose presence was anticipated.
This distinction is increasingly important as AI becomes embedded in everyday decision-making. AI systems influence recruitment, health care, education, finance and public administration at a scale unimaginable only a decade ago. Their ability to process information rapidly makes them valuable, but it also means that any assumptions encoded into their design can be amplified across millions of decisions. Small oversights during development may evolve into widespread structural inequalities once these systems become integral to society.
The challenge, therefore, is not to reject innovation but to broaden the perspective from which innovation emerges. Inclusive design requires asking difficult questions before a product, policy or technology reaches scale. Who is missing from the conversation? Whose lived experiences have not been considered? What resources are being taken for granted? Most importantly, what happens to those who do not fit the profile of the imagined user?
These questions should not be viewed as obstacles to progress but as essential components of it. A system that accommodates diverse realities is ultimately more resilient, adaptable and effective than one designed for a narrow segment of society. Businesses gain access to wider markets, governments deliver better public services and communities benefit from opportunities that might otherwise remain inaccessible.
Perhaps the greatest misconception about inclusion is that it requires complicated solutions. Often, meaningful change begins with recognising that no system is truly neutral. Every design choice reflects priorities, and every priority carries consequences. Expanding the range of voices involved in those decisions is not an act of charity; it is an investment in better outcomes.
As societies continue to embrace automation, digital platforms and intelligent technologies, the question is no longer whether systems can deliver progress. They undoubtedly can. The more pressing question is whether that progress is broad enough to include the people who have historically remained invisible. A successful system should not merely function according to its original design. It should evolve to reflect the diversity of the people it ultimately serves, because progress that leaves people behind may be efficient, but it can never be considered complete.
(The views expressed are personal)
This article is authored by Riya Kamat, founder, Shadows of Progress.