This article was written by **Isao Toyama**, reflecting on why design responsibility must go beyond the user, especially in the age of AI. We’re sharing this perspective to highlight the transition from human-centered design to consequence-aware design, a turning point emerging from our community conversations.
Why Design Responsibility Goes Beyond the User, Especially in the Age of AI
The recent discussions at DDX revealed something important.
Not a minor methodological debate.
Not a branding nuance.
A philosophical turning point.
For decades, Human-Centered Design (HCD) taught us to focus on the user:
Understand needs.
Reduce friction.
Improve usability.
Design better experiences.
It was revolutionary in its time.
It humanized technology.
But something became increasingly clear:
**Once the product is sold, responsibility often **ends. And that is the unresolved gap.
The Uncomfortable Question
At DDX, a deeper question surfaced:
**Who is responsible for what happens after the product leaves the designer?**
If a smartphone alters posture, attention span, sleep patterns, and social interaction…
If a device becomes obsolete in three years and ends up in a landfill…
If a platform amplifies addiction or polarization…
If AI systems quietly shift power structures and labor dynamics…
Were those outcomes accidental?
Or were they designed consequences?
They may not have been intentionally harmful.
But they were structurally enabled.
And if design decisions made these outcomes possible,
Then they are design outcomes.
That realization changes everything.
AI Intensifies the Question
Artificial intelligence magnifies this issue dramatically.
When we introduce AI systems into society, we are not simply improving efficiency.
We are:
● Automating judgment
● Redistributing power
● Restructuring labor markets
● Replacing cognitive tasks
● Collecting unprecedented amounts of personal data
● Normalizing surveillance
● Changing how people think, learn, and decide
More jobs are displaced.
More privacy is surrendered.
More decision-making shifts from human deliberation to algorithmic systems.
The question becomes unavoidable:
**How much of our humanity are we willing to exchange for convenience, speed, and optimization?**
AI is not neutral.
It reshapes:
● Autonomy
● Trust
● Agency
● Skill development
● Institutional control
● Economic distribution
The Limits of “Centered”
Don Norman’s move from Human-Centered Design (HCD) to Humanity-Centered Design (HCD+) was an important expansion.
HCD focused on the individual user.
HCD+ widened the lens to society, culture, and the planet.
But even HCD+ still carries the word “centered.”
To center something means making it the main focus, a perspective, a priority.
But what if the real issue is not what we focus on, But what do we account for?
The landfill is not “centered.”
The displaced worker is not “centered.”
The data-mined citizen is not “centered.”
The weakened democratic structure is not “centered.”
There are consequences.
And consequences do not sit at the center.
They unfold across systems.
This is the next philosophical layer:
We need to move from user-centered to humanity-centered, and then to consequence-aware design. This is not a rejection of Don.
It is the logical continuation of his systemic thinking.
The Myth of Limited Responsibility
The design industry has long operated under an invisible boundary: Designers are responsible for:
● Usability
● Experience
● Interface
● Adoption
After scale?
After automation?
After job displacement?
After environmental accumulation?
That belongs to business.
To policymakers.
To “the market.”
But this boundary is artificial.
Design determines:
● What is automated
● What is measured
● What is incentivized
● What becomes obsolete
● What becomes addictive
● What becomes the default
● What becomes invisible
Design shapes behavior more powerfully than regulation does.
If design shapes systems,
Then responsibility cannot end at interaction.
It extends through the lifecycle, governance, and long-term consequences.
Designing AI: The Humanity Question
When creating AI systems, the stakes are even higher.
We must ask:
● Are we designing augmentation or replacement?
