The Contours of "Usable by Everyone"
A person is trying to record their voice with TokiStorage. They pick up a smartphone, open a browser, press the record button, speak for thirty seconds, and save. For most people, this sequence appears to be a task that takes only a few minutes.
But who, exactly, are "most people"?
The process assumes the user can see the screen. It assumes they can tap a button precisely with their finger. It assumes they can read Japanese text. It assumes they can produce voice. It assumes they can maintain a stable posture for thirty seconds. It assumes they own an internet-connected device.
"Usable by everyone" typically refers to "everyone" within the range of a designer's imagination. Those outside that imagination were never part of "everyone" to begin with.
TokiStorage declares its mission as the "democratization of proof of existence." If that word—democratization—is meant seriously, its reach must not remain confined to the inside of a designer's imagination.
The Assumption of Able-Bodiedness
In Japanese, the phrase gotai-manzoku (five-body satisfaction) describes a state in which both arms, both legs, and all senses function properly. The word carries an implicit meaning: this state is "complete." By extension, any other state is "incomplete."
Product design proceeds, often silently, on the assumption of this "complete body." Contrast ratios are decided based on standard vision. Button sizes are determined by standard finger size and motor precision. Voice guidance is designed for standard hearing. Within the word "standard," a boundary has already been drawn.
But this boundary is not binary. It is not a choice between seeing and not seeing—there is a gradient of diminished sight. It is not a choice between hearing and not hearing—hearing loss varies by frequency range. It is not a choice between mobile and immobile—a person may have good grip strength today and none tomorrow.
Aging moves every human being along this gradient. Text that was effortless to read at thirty becomes difficult at sixty. Temporary injury also shifts the boundary. A person with a broken dominant hand must operate a smartphone with one hand. So must a parent holding a child. On a noisy train platform, everyone effectively has a hearing impairment.
If TokiStorage promises preservation for a thousand years, then within that time frame, every user's body will inevitably change. The person who recorded their voice at thirty may try to play it back at eighty—and the same interface may no longer work for them. Accessibility is not only a present-day consideration. It is a promise to the future user.
What Governs Usability
As long as accessibility is framed as "accommodation for people with disabilities," the full picture of what affects usability remains invisible. Usability is determined by multiple overlapping layers.
First, there are physical factors: vision, hearing, motor function, speech. These may be congenital or acquired, permanent or fluctuating.
Second, there are cognitive factors: literacy, digital literacy, language barriers. For someone whose native language is not Japanese, a Japanese-only interface is an obstacle. For a user in their seventies, "scan the QR code" is not self-evident.
Third, there are environmental factors: connectivity, lighting conditions, noise levels, temperature. In direct sunlight, screens become hard to read. In cold climates, users must operate devices while wearing gloves. In areas with slow connections, uploading large files is impractical.
Fourth, there are economic factors: device cost, data plan fees, access to electricity. Roughly thirty percent of the world's population still does not own a smartphone. Among those who do, inexpensive devices may not support the latest browser features.
Fifth, there are temporal factors. This is a problem unique to TokiStorage. A service that promises thousand-year preservation must withstand a thousand years of bodily change, technological change, and social change. Today's accessibility standards may be obsolete in a hundred years. But a voice recorded today must still be playable a hundred years from now.
Accessibility is not accommodation for disability. It is design that begins from the premise that every human being changes over time.
Devices, Software, Operations—The Strata of Modern Consideration
The concrete implementation of accessibility is layered like geological strata.
At the deepest layer lies semantic HTML. Headings marked as headings, buttons marked as buttons, links marked as links. When this structure is correct, screen readers can vocalize the content of the page. Conversely, a page styled purely with div tags may look beautiful visually but appears as a meaningless string of characters to a screen reader.
Above that layer sits WAI-ARIA—attributes that communicate states HTML alone cannot express: expanded or collapsed, selected or unselected, error notifications. Used correctly, ARIA becomes a bridge. Misused, it becomes a wall.
Above that, visual considerations: contrast ratios, font sizing, conveying information without relying solely on color, reducing motion. The Web Content Accessibility Guidelines (WCAG) systematize these standards. Yet meeting the standard and being genuinely usable are not the same thing.
Then the operational layer: Can all functions be accessed by keyboard alone? Are touch targets large enough? Is switch access supported? Can the interface be operated by voice? These are not individual features to be added later—they should be built in from the start as design premises.
TokiStorage's technical choices carry accessibility implications. Adopting a PWA means the service can be used directly from a browser without going through an app store. This lowers the installation barrier, but it may also limit access to OS-native assistive features that native apps can leverage. Publishing on GitHub Pages provides transparency—anyone can inspect the source code—but it also means dependency on a specific platform. QR codes bridge the physical and digital worlds, but they do not function on devices without cameras or in environments where camera access is restricted.
Consideration is not addition. Whether it was included in the original blueprint or appended afterward changes the reach entirely.
The design philosophy of "completing everything with a single smartphone" provides powerful convenience for those who own a smartphone. But for those who cannot own one, that very convenience becomes a barrier. Every design choice has both an inclusive and an exclusive side. The issues discussed in "Device, Browser & Organizational Dependency" take on even more serious implications in the context of accessibility.
Between "They Can Probably Use It" and "They Cannot"
Designers judge "ease of use" using their own bodies as the baseline. They project their own body—one that can see a screen and press a button—onto the user's body. The judgment "they can probably use this" is an act of imagination originating from the designer's own physical experience.
Between this imagination and the lived experience of the actual user, there is a structural disconnect.
For a person with color vision diversity, information differentiated by red and green is indistinguishable. What the designer thought was "easy to understand because it's color-coded" appears to the user as "both look exactly the same." The designer added color with good intentions. But those intentions were premised on the designer's own color vision.
For a person with hand tremors, tapping a small button precisely is difficult. What the designer thought was "easy to press at this size" becomes, for the user, "I keep hitting the wrong button no matter how many times I try." The designer tested with their own finger. But that test was premised on the designer's own motor precision.
For a person sensitive to cognitive load, an information-dense screen is overwhelming. What the designer thought was "comprehensive—all the necessary information is here" becomes, for the user, "I don't know where to look." The designer valued thoroughness. But for the user, reducing choices is what matters.
"It's probably easy to use" is an imagination projected from the designer's body. The lived experience of the actual person lies outside that imagination.
There is only one way to bridge this disconnect: ask the person. Test with the person. But "the person" is not a monolith. Among people with visual impairments, there are those who are completely blind, those with low vision, and those with color vision diversity. Each experience is entirely different. What one person reports may not apply to another.
This is precisely why accessibility is never finished. It is not something that concludes with a single design cycle. It is a continuous process of observation and correction. The "exclusion through good intentions" discussed in the previous essay "What Stands Opposite Ethics" holds the same structural pattern in the context of accessibility. The possibility always exists that "making something easier to use" has excluded someone else.
The Tectonic Shift AI Brings
The evolution of AI is fundamentally reshaping the terrain of accessibility.
Real-time transcription has opened a path for people with hearing constraints to access audio content. Speech synthesis has made it possible for people with visual constraints to receive text as sound. Image recognition can now describe in words what appears in a photograph. Intent prediction infers what a user wants to do from incomplete input and supplements the operation.
These technologies function as "translators" standing between human capability and digital interfaces. Support that once required specialized hardware or expensive software is now built into smartphones as standard. VoiceOver and TalkBack are available at no additional cost.
AI is also opening the possibility of adaptive interfaces—learning a user's interaction patterns to optimize button placement, analyzing input tendencies to improve predictive text accuracy. A world in which the interface reshapes itself for each individual user is no longer experimental.
However, accessibility that depends on AI carries structural risks.
First, when an AI service goes down, accessibility goes down with it. If a transcription service experiences an outage, people with hearing constraints are once again cut off from audio content. TokiStorage's principle of "not increasing external dependencies" exists because of this very vulnerability. If accessibility depends on AI, then when AI disappears, accessibility disappears too. For a service that promises thousand-year preservation, this is an unacceptable risk.
Second, if an AI's training data is biased, the accessibility it provides will be biased too. English speech recognition achieves high accuracy, but recognition of minority languages remains poor. Standard speech patterns are recognized precisely, while speech from people with articulation disorders is often misrecognized. AI can become a tool for everyone—but only if its training data includes everyone. Otherwise, AI too becomes an instrument of exclusion.
Third, when AI "does it for you," there is a risk that investment in foundational technology is neglected. If AI can interpret a screen regardless, developers may stop caring about semantic HTML structure. But without a foundation that functions without AI, nothing works in environments where AI is unavailable.
So what can a designer do?
Observe. Continuously observe who your design includes and who it excludes. Then apply what you observe to the design. Satisfying a checklist is not the end. Meeting WCAG criteria is not the end. When you find someone outside the criteria, update the criteria themselves.
"Observe and apply" is not a one-time act but a posture of design. If TokiStorage genuinely commits to thousand-year preservation, it must be prepared for a thousand years of observation and application. Democratizing proof of existence means closing the right to record to no one. To close it to no one, we must first observe where it is closed.
Democratizing proof of existence means closing the right to record to no one. To close it to no one, we must first observe where it is closed.