The debate around artificial intelligence has become flat. One side sees the beginning of an economic revolution. The other sees the beginning of social descruction. Every breakthrough is immediately translated into questions about jobs, copyright, surveillance or existential risk. Those questions deserve serious attention, but they have also shaped the way we measure the technology itself. We keep asking how much smarter AI will become, how much work it will automate, or how much value it will create. I suspect we are asking the wrong question.
Technology has usually been judged by the amount of human effort it removes. Computers reduced computational labour. The internet reduced the friction of distance. AI is expected to continue that pattern by reducing cognitive labour. It is a logical progression, but it also assumes productivity is the only measure worth caring about.
There is another possibility. AI may become important not because it outperforms people, but because it accommodates them.
For decades we have built digital systems on a remarkably consistent assumption. The person must adapt to the system. If you want access to a service, you learn its language. If the process is confusing, you read the guidance. If the information is fragmented, you are expected to assemble it. If the form rejects your application, you work out why and start again. We have become so accustomed to these expectations that they barely register as design decisions anymore. They simply feel like the normal cost of participating in modern society.
Working in service design changes the way you see those interactions. You stop looking for what people are doing wrong and start asking what the system assumes about them. Does it assume everyone reads comfortably, interprets ambiguity in the same way or has the time and energy to persist through repeated failures? Those assumptions quietly determine who succeeds and who struggles.
Consider how we usually approach accessibility. We improve the wording of a document, redesign a form or introduce another support channel. All of those changes are worthwhile, but they are still attempts to produce a better standard experience. AI offers something different. The experience no longer has to be standard at all. Two people can begin in the same place and finish with completely different conversations because each needs something different to reach the same understanding.
Someone with dyslexia may need information expressed differently. Someone with ADHD may need help turning scattered thoughts into a coherent message. An older person navigating aged care may simply need enough time and patience to make sense of unfamiliar terminology. These situations look unrelated until you step back. Each reflects a mismatch between the person and the assumptions built into the system they are trying to use.
That mismatch exists everywhere. We see it in government, healthcare, banking, education and employment, to name a few. We have spent years trying to make systems easier for people to navigate. AI introduces another option. Instead of endlessly refining the pathway, we can begin building technology that adjusts itself to the individual walking it.
None of this removes the need for good design. In fact, it raises the standard. AI cannot rescue poor policy, fragmented services or inaccessible information. If the underlying system is incoherent, the technology inherits those weaknesses. It can make an interaction more conversational without making it more truthful. It can make poor decisions sound convincing. It can reflect bias already present in the data it learns from. Those are genuine risks, and they deserve the attention they receive.
Across disability, neurodiversity, ageing and chronic illness, people are quietly using AI to overcome barriers that most of us have simply accepted as part of everyday life. These aren’t dramatic demonstrations that attract headlines. None of these moments will appear in productivity statistics, yet each represents a person participating more fully than they could before.
Perhaps that is why this contribution is so easy to overlook. It doesn’t produce spectacular demonstrations or billion-dollar valuations. It doesn’t fit neatly into the language of disruption that dominates conversations about technology. It is incremental, personal and often invisible to everyone except the individual experiencing it.
History has a habit of simplifying technological change. We remember the calculator, the personal computer and the smartphone through the industries they transformed. We rarely remember the quieter ways they changed everyday life. AI will almost certainly reshape economies and professions, but I wonder whether that will be the story that matters most in hindsight.
If future generations look back on this period, they may conclude that artificial intelligence did something previous technologies rarely managed. It stopped treating human difference as an exception to be accommodated after the fact. Instead, it made adaptation part of the interaction itself. If that happens, inclusion will not be an incidental benefit of AI. It will be one of the reasons the technology mattered at all.
