Human-Centered Design Is Not Enough

Human-centered design changed what designers were accountable to. Instead of beginning with technical possibility or organizational convenience, it asked us to understand people, their goals, limitations, contexts, and the meaning an experience has in their lives.

That correction still matters. Research, participation, iteration, accessibility, and usability are ways of refusing to treat a human being as an error inside a technical system.

The problem is not that human-centered design cares too much about people. The problem is where we draw the boundary around the human situation. A person never meets a product alone. They meet it through electricity, logistics, labour, materials, institutions, streets, water, soil, animals, weather, and time.

A food-delivery interface can be effortless for the customer while moving risk, heat, waiting, and traffic pressure onto couriers. A cheap device can feel valuable at purchase while its extraction, assembly, repair, and disposal remain invisible. A recommendation system can satisfy an immediate preference while narrowing culture, increasing consumption, or rewarding inflammatory material.

From the screen, these can all look like successful experiences. The task is completed. The conversion happens. The user reports satisfaction. Yet the system may be creating damage outside the measurement frame.

A narrowly human-centered process can optimize the experience of the person closest to the interface while treating everyone and everything else as external. The better the optimization becomes, the more efficiently the system can export its costs.

The first interface is the one a person can see: the screen, object, service encounter, packaging, instruction, payment, notification, and feedback. Designers are trained to make this layer comprehensible.

The second interface connects the product to the world that makes it possible. It includes supply chains, energy systems, data centres, warehouses, roads, farms, mines, repair networks, policy, waste systems, habitats, and atmospheric emissions. This interface is usually fragmented across departments and hidden from the customer.

Life-centered design begins when both interfaces are treated as design material. The question is no longer only whether the button is clear. It is whether the action behind the button produces a system we are willing to maintain.

A one-click return may reduce anxiety for the shopper, while triggering packaging, transport, inspection, liquidation, recycling, or disposal. The visible interaction lasts seconds. Its material interaction can travel thousands of kilometres.

Life-centered design does not remove the human from design. It places human life inside a larger living system. People remain participants, but so do communities, non-human species, ecosystems, infrastructures, and future inhabitants who cannot attend the workshop.

A tree is not interviewed like a customer, and a river cannot approve a prototype. Their interests need representation through ecological research, field observation, scientific evidence, regulation, local knowledge, advocacy, and people who understand how those systems behave.

The point is not to pretend every stakeholder has equal needs or that trade-offs disappear. The point is to stop making absence look like consent. When a stakeholder cannot speak in the room, the process needs another way to keep its conditions visible.

Life-centered design tools such as systems maps, actant maps, product lifecycle maps, and behavioural-impact canvases do not produce perfect moral answers. They expose relationships, knowledge gaps, and consequences that a conventional user journey can miss.

Green Filter began with a consumer problem: sustainable shopping is difficult because information is fragmented, claims are inconsistent, and people have limited time and attention. An AI companion could help someone compare products, understand trade-offs, save money, and connect everyday consumption with longer-term financial choices.

But the research kept moving upstream. Better information at purchase cannot make a badly designed product circular. A sustainability label cannot repair an item that was glued shut. A persuasive interface cannot compensate for a supply chain built around disposability. Consumer choice matters, but producers, platforms, regulators, investors, and infrastructures shape what choices exist.

This is why the thesis moves through eco-design, circular design, regenerative design, multispecies design, design for human rights, and life-centered design. The vocabulary expands because the object of design expands. The shopping interface is only one point inside a larger system of production and consequence.

The useful lesson for HAAM is that interaction design should not end at the interaction. The interface is where a system becomes perceptible, but design also includes changing the system that the interface reveals.

AI systems are often presented as if they live inside a text box. In reality, they depend on data, human labour, chips, minerals, electricity, cooling, networks, policies, and organizations. Their outputs can influence hiring, credit, education, media, purchasing, mobility, and public services.

An AI agent instructed only to optimize a user's goal can become extremely effective at ignoring everything outside that goal. It might find the cheapest trip, fastest delivery, highest-return investment, or most engaging content while externalizing emissions, labour conditions, market manipulation, accessibility, or social harm.

The prompt cannot be the complete value system. Life-centered AI needs boundaries, evidence, permissions, escalation paths, and ways to represent affected parties who are not the person giving the command.

The design challenge is not to make an AI care in a human sense. It is to build systems in which care becomes operational: what data is considered, which harms are prohibited, whose consent is required, what trade-offs are disclosed, and when optimization must stop.

Life-centered design can sound impossibly broad. The answer is not to model the entire planet before changing a checkout flow. The answer is to expand the frame in deliberate steps and make uncertainty visible.

Begin with five circles. Each asks a different version of who or what has to live with this decision. They can be mapped around a feature, service, building, policy, or business model.

A human-centered project might rely on interviews, usability tests, analytics, and service journeys. A life-centered project still uses them, but adds lifecycle assessment, biodiversity data, energy measurement, labour conditions, repair records, supply-chain traceability, environmental thresholds, and long-term scenarios.

Completion rate, satisfaction, retention, and revenue remain useful, but they are incomplete. A system may also need to show whether products last longer, returns are consolidated, workers gain control, emissions fall, habitats remain connected, or a future participant can repair what has been built.

Metrics do not remove conflict. They make conflict harder to hide. Design becomes the practice of deciding which outcomes deserve visibility and which limits should constrain growth.

Design culture celebrates empathy. It helps us imagine another person's situation, but it is selective, emotionally uneven, and difficult to extend across distance and time. We empathize most easily with a visible person, not a dispersed supply chain, an unfamiliar species, or someone living decades from now.

Life-centered design needs more than a larger empathy exercise. It needs accountability: documenting consequences, assigning responsibility, setting thresholds, preserving evidence, and making it possible for affected people to challenge the system.

This is where design connects to governance. A humane interface without institutional responsibility is fragile. A sustainable intention without measurement can become branding. A participatory workshop without power over the final decision can become theatre.

The stronger question is not only whether we understand the stakeholder. It is what the stakeholder can require from us, and how we will know whether we kept the promise.

Widening the design frame can initially make action feel harder. More relationships appear. Trade-offs become less comfortable. The simple solution dissolves into a system.

Yet this complexity is also where possibility lives. When we see the wider system, we discover more places to intervene: the default, material, contract, incentive, repair path, data model, policy, ownership structure, public explanation, and the relationship between them.

Human-centered design taught us that technology should adapt to people. Life-centered design asks us to take the next step: human activity must remain compatible with the living world that supports it.

The user is still at the centre of the interaction. The user is simply no longer the edge of our responsibility.