1. About HAKONE
It is not about stopping technology.
The goal is for humans to regain agency.
HAKONE is an AI-driven tool developed in Japan to support the digital autonomy of young people.
The challenge we are addressing is not the smartphone itself, social media, or AI.
The sophisticated algorithms powering modern digital services—such as recommendation engines, notifications, autoplay, and infinite scroll—have dramatically improved human life.
That is precisely why we have no intention of eliminating them.
What is needed
is to create a society where humans can choose for themselves when to start using technology and when to stop, even as it continues to evolve.
HAKONE is designed as a practical digital infrastructure to make this a reality.
At its core is the "Finger Snap"—a brief digital self-regulation intervention that uses the fingertip as an anchor for attention.
It does not involve confiscating smartphones or forcibly cutting off access.
It empowers the person on the other side of the screen
to regain the power to pause,
and the power to say, "I’ll stop here."
HAKONE is an AI designed to support those small yet significant moments.
This project began with an initiative the founder launched for his own children.
However, through more than a decade of direct engagement with young people and their families, it became clear that this issue is not limited to children.
Upon turning 18, young people are freed from parental supervision and school rules, facing the need to decide for themselves how to interact with their smartphones.
At that stage, what is required is not "management," but the skills to navigate the digital society autonomously.
Consequently, HAKONE has evolved into a system primarily aimed at young adults aged 18 and older.
2. Our Mission
Instead of weakening the algorithms,
we empower the people.
Smartphones, social media, videos, games, and generative AI have become the foundation of daily life across the globe.
Reversing this trend is not realistic.
Nor do we believe it needs to be reversed.
Commercial algorithms hold immense economic and cultural value.
The problem is that
while humanity has been taught how to *use* these tools, we have hardly been taught how to *step away* from them.
Bicycles have a way to stop, not just a way to ride.
Cars have brakes, not just an accelerator.
Yet with smartphones, while technology for "seeing more," "watching the next thing," and "opening the app again" has advanced astonishingly,
the technical design for ending the experience of one's own volition has not been sufficiently developed.
HAKONE fills this gap.
From prohibition to autonomy.
From compulsion to choice.
From surveillance to self-regulation.
Not abandoning the digital world, but becoming people who can coexist with it.
That is our mission.
3. Born in Japan. Bound for the world.
Born in Japan. Designed for a Digital World.
HAKONE began its journey on a small scale in Japan.
However, the challenges we face know no borders.
Screens, social media, video platforms, games, and generative AI—
the same digital environment exists in Tokyo, New York, London, Paris, and Singapore alike.
That is why, alongside disseminating information in multiple languages, HAKONE engages in international outreach and policy advocacy across sectors such as education, public health, telecommunications, and digital policy.
Our international outreach extends far beyond North America (including the U.S. and Canada) to encompass Europe, Asia-Oceania, Africa, and Latin America. As outlined in the attached materials, our engagement targets include the Office of the U.S. Surgeon General, the FCC, the UK’s Ofcom, and various national agencies responsible for education, health, and digital policy.
In the Asia-Oceania region, we work with entities such as Australia’s eSafety Commissioner, Singapore’s Ministry of Digital Development and Information and Ministry of Education, as well as health and education agencies in India.
Furthermore, in Latin America, we are focusing on institutions related to education, health, and telecommunications in countries including Brazil, Mexico, Colombia, Argentina, Chile, Peru, and Ecuador.
We do not aim merely to provide a service for improving lifestyle habits within a single nation.
Our goal is
to propose—originating from Japan—a new form of social infrastructure that enables humanity to coexist harmoniously with digital civilization for the long term.
4. Founder
Tokkyu Hosokawa
Digital Public Health Auditor / Algorithmic Ethics Designer
Spanning the diverse fields of law, systems engineering, and biology, I have spent the last decade—since 2014—conducting research and development rooted in direct engagement with young people and their families.
At the heart of this work lies my own life story.
During junior high school, I stepped off the conventional educational track and experienced school non-attendance.
In the early 1980s—before computers had become commonplace in households—I was captivated by electronic devices and synthesizers; I built my own radios and cassette decks and taught myself about electronic circuits.
Since then, I have traversed worlds that might seem unrelated at first glance: working as a sewer pipe installer, in legal practice, at a systems company, on a deep-sea tuna fishing vessel, in research, and in the food and beverage industry.
Yet today, all of these experiences converge in HAKONE.
Understanding machines.
Understanding the law.
Understanding the human body.
And, through my own personal experience, understanding the feelings of young people who have stepped off society's established path.
In 2014, I established a facility where young people who had left the conventional social track could gain real-world work experience.
That marked the beginning of a long-term process of practical engagement and observation involving young people and their families.
Drawing on insights gained from working with 1,500 families over the past decade, I am now designing HAKONE based on a specific philosophy:
Rather than banning or eliminating the commercial algorithms used by massive platforms,
we should further develop them while implementing mechanisms that allow humans to reclaim agency.
The Role of Digital Public Health Auditing
Mechanical failures are monitored by sensors.
Networks have monitoring systems.
Finance has risk management.
Aviation has safety audits.
Yet,
when it comes to the time, attention, and behavior of the humans using these systems, much is left solely to individual self-management.
Norihisa Hosokawa views this domain through the lens of "digital public health auditing."
It is not about casting algorithms as the villain.
Nor is it about seeking solutions solely through regulation.
If algorithms are becoming more sophisticated,
then human self-regulation techniques must evolve in tandem.
HAKONE is an attempt to implement this concept through software.
5. Company Information
COMPANY INFORMATION
Service Name
HAKONE AI
Business Activities
Research, design, and operation of AI for digital independence support for adults
Research and development of digital well-being technologies
Ethical algorithm design
Research and information dissemination regarding digital public health
Policy recommendations and information provision to domestic and international government bodies and public institutions
Operating Company
Synapse Cascade (Gomei Kaisha)
Corporate Registration Number (Japan)
5012803001524
Date of Establishment
November 18, 2016
Project Founder / Product Leader
Tokkyu Hosokawa
Registered Representative
Kyoko Hosokawa
Country
Japan
Official Contact
contact@personal-shaman.com
Tokyo Atelier & Headquarters
102 Sky Heights Hiyoshi
3-2-6 Hiyoshi-cho
Kokubunji-shi, Tokyo 185-0032
Japan
Tel: +81 (0)42 359 4368
Please direct inquiries regarding HAKONE AI, as well as research, policy recommendations, and international collaborations, to the Tokyo headquarters.
6. 12 Major Works and Design Projects by the Founder
01 — HAKONE AI
Personal AI for Human Digital Independence
HAKONE AI is a private AI system designed to help people reclaim human agency in an increasingly algorithm-driven world.
Designed primarily for young adults aged 18 and over, HAKONE supports people living alongside sophisticated digital environments—including smartphones, social media, video platforms, games, and generative AI—while helping them strengthen the ability to decide for themselves when to begin and when to stop.
The goal is not to separate people from technology.
It is not to ban commercial algorithms.
It is not to take smartphones away.
HAKONE begins from the opposite assumption:
technology will continue to become more powerful.
So the human side must also become better equipped.
HAKONE therefore aims to provide each user with a personal system for digital self-regulation.
It brings together the user’s daily rhythms, goals, conversations, emotional inputs, and relevant digital behavior as one personal context, then uses that context to shape individualized support.
HAKONE is not an AI designed to stop people from using technology.
It is an AI designed to support people who want to remain capable of choosing both:
when to use technology,
and when to stop.
HAKONE AI also serves as the central intelligence layer that brings together multiple specialist systems—including Yubizen, emotional interfaces, learning AI, creative AI, and safety systems—into one coherent architecture.
02 — Yubizen / 指禅
60 Seconds to Stop the Finger and Return the Decision to the Human
In modern smartphone use, the finger rarely rests.
Scroll.
Tap.
Next.
Another one.
And another.
Yubizen is HAKONE’s original digital self-regulation system, designed to introduce a deliberate moment of stillness into that continuous stream of action.
The fingertip becomes an anchor for attention.
Sound, subtle physical feedback, stillness, and short interactions are combined to create a small space inside a sequence that might otherwise continue automatically.
The purpose is not to test endurance.
It is not:
“Try harder not to use your phone.”
The purpose is to return the decision to the user before the algorithm presents the next thing.
Yubizen draws inspiration from ideas long associated with Japanese Zen:
stop,
observe,
return attention to the present.
These ideas are not reproduced as a religious practice. They are reinterpreted for HAKONE’s digital environment.
Instead of treating the smartphone as the enemy, Yubizen changes its role.
For 60 seconds, the same device that normally invites continued scrolling becomes a device for stopping.
That is Yubizen.
03 — Acoustic Glass Resonance
Turning Smartphone Glass from a Surface of Stimulation into a Surface for Returning
A smartphone screen is normally used to see, touch, and move toward the next piece of information.
Acoustic Glass Resonance uses that same glass surface differently.
It combines:
natural sound,
subtle vibration generated through the smartphone,
fingertip contact,
and deliberate stillness.
The objective is to shift the experience from:
looking at the screen
to:
feeling the interaction through the fingertip and the body.
This is HAKONE’s audio-haptic design concept.
The important point is that the system does more than play sound.
By shifting part of the user’s attention away from visual information and toward sound, touch, vibration, and bodily sensation, the design creates another route for awareness outside the visual stream.
The smartphone can be a device that enables endless scrolling.
But if the interaction is redesigned, the same smartphone can also become a device that signals:
stop here for a moment.
Acoustic Glass Resonance was born from that reversal.
04 — LADDER 2.0
Haute-Couture AI Designed Around One Individual Life
Most general-purpose AI systems are designed to produce answers that can be useful to many people.
Human lives, however, are not average.
One person is preparing for university.
Another is studying for a professional qualification.
Another has just started a career.
Another wants to devote more time to creative work.
Some people live late at night.
Others begin early in the morning.
And what each person wants to do with time recovered from unnecessary digital use can be completely different.
LADDER 2.0 is HAKONE’s internal haute-couture AI design framework for bringing those differences into the system itself.
It can combine relevant elements such as:
personal values,
daily rhythms,
goals,
learning themes,
conversation patterns,
and the user’s own intentions regarding digital use.
The objective is not to build an experience for:
“the average user.”
It is to build an experience for:
this person.
That is the role of LADDER 2.0.
HAKONE does not ask the human being to adapt to the AI.
The AI is adapted to the human being.
LADDER 2.0 also serves as an important coordinating framework, helping the specialist systems described in Sections 07–12 work together as one consistent experience for a single user rather than as disconnected AI functions.
05 — Digital Public Health Audit
Looking Beyond “Weak Willpower” to Audit the Digital Environment Itself
Modern digital services are deeply connected to human attention, time, choice, and emotion.
Yet when digital use becomes longer than a person intended, the response is often reduced to:
“Just stop.”
or:
“Manage yourself better.”
HAKONE introduces another perspective.
Where does a commercial algorithm capture attention?
Where does interface design encourage continuation?
Where are natural stopping points missing?
And what can be designed on the human side to create another option?
Digital Public Health Audit examines digital problems not only as individual behavioral issues, but also as questions of system design.
Its purpose is not to attack technology companies or commercial algorithms.
HAKONE does not begin from the idea that commercial technology should disappear.
The question is:
Can commercial technology continue to develop while the human side is also equipped with better safety mechanisms?
Instead of saying:
“People must become better at resisting.”
HAKONE asks:
“Can we design an environment in which people do not have to rely on resistance alone?”
Digital Public Health Audit provides a foundation for connecting that idea to research, system design, and public-policy proposals.
06 — Adult Digital Independence Infrastructure
What Comes After Parental Control Ends?
Children often grow up inside structures created by others:
parents,
schools,
household rules,
age restrictions,
and other forms of supervision.
After adulthood begins, that structure changes dramatically.
How many hours should I use my smartphone?
Should I watch another video at midnight?
When should I close social media?
Should I study?
Sleep?
Prepare for tomorrow’s work?
Ultimately, those decisions belong to the adult individual.
Yet society often assumes that the moment a person becomes an adult, they are automatically capable of managing highly sophisticated digital environments independently.
At the same time, there are few systematic opportunities to learn how to disengage from digital systems by choice.
HAKONE sees this as a gap in both policy and technology:
the adult digital independence gap.
The answer is not simply:
ban it,
confiscate it,
block it,
or ask someone else to monitor you.
The alternative is to support the ability to:
notice for yourself,
stop for yourself,
choose the next action for yourself,
and connect with human support when appropriate.
The social and technological foundation for those capabilities is what HAKONE calls:
Adult Digital Independence Infrastructure.
HAKONE’s long-term objective is therefore larger than distributing a single application.
When people learn to drive, they learn not only how to accelerate, but how to stop.
HAKONE believes digital civilization also needs a way for people to learn:
how to get off.
This sixth layer extends that concept across technology, education, safety design, and policy.
07 — HAKONE Sentinel AI
Detecting Unusual Change and Creating a Path to Human Support
HAKONE Sentinel AI is a specialist safety layer designed to identify meaningful changes from a user’s own usual patterns.
Depending on the functions enabled and the information intentionally provided by the user, relevant signals may include:
interaction patterns,
self-reported state,
conversations with HAKONE,
and changes observed over time.
The system is not intended to treat a single word, score, or behavior as proof that something is wrong.
Instead, the design focuses on change relative to the individual user’s own normal pattern.
Where the user has explicitly configured an approved contact or notification pathway, predefined safety conditions may allow HAKONE to surface significant changes for human attention.
HAKONE Sentinel AI is not a medical diagnostic system.
Its purpose is not to determine whether a person has a disease, disorder, or clinical condition.
Its purpose is narrower:
notice meaningful deviation,
and create a path back to a human being when appropriate.
It is a digital safety layer designed around the principle that small changes should not be ignored simply because they do not yet constitute a crisis.
08 — TATOEMON Learning Refuge AI
Turning Time Away from Algorithms into Time Invested in Your Future
Stopping the smartphone is only the beginning.
If nothing meaningful exists beyond the screen, it can be easy to return to the same digital stream.
TATOEMON is HAKONE’s learning AI designed for what comes next.
A user can define goals such as:
professional qualifications,
career development,
specialist knowledge,
languages,
academic subjects,
or other areas of learning.
TATOEMON can then organize an individualized learning route around those goals.
One of its core approaches is the use of analogy—turning difficult ideas into familiar examples that are easier to understand.
The learning experience can change according to the user’s progress and level of understanding.
TATOEMON is an important part of why HAKONE is not designed merely as a system for stopping scrolling.
The goal is not simply to reduce time.
The goal is to redirect time toward the future.
TATOEMON is designed as an intellectual refuge—a place to go after stepping away from the commercial algorithm.
09 — Illustrated Drawing Coach AI
Create. Learn. Reflect.
Illustrated Drawing Coach AI is a specialist creative-learning system designed to support users through their artwork and creative process.
It can help organize learning around areas such as:
drawing,
composition,
form,
materials,
technique,
and methods of expression.
The goal is not simply to judge whether a work is:
good or bad.
Instead, the system focuses on a more useful question:
What has changed compared with your earlier work?
It can help record a user’s development over time and provide structured observations about the use of pencils, paint, pens, composition, and stages of production.
The system is also designed so that relevant information can be organized for a human teacher, reviewer, or evaluator where human involvement is part of the process.
HAKONE does not believe that AI should determine the value of a work of art by itself.
The principle is:
AI observes.
Humans evaluate.
Creative education is designed with both in the loop.
10 — 16-Emotion Slider AI
Understanding “How I Am Today” Before It Becomes a Single Word
A human state cannot always be described accurately with one label such as:
fine,
tired,
sad,
or stressed.
HAKONE uses a system of 16 emotional sliders to allow users to express their daily state across multiple dimensions.
Different combinations of these 16 axes can create a large range of possible states.
Those inputs can then be considered alongside the user’s own previous records and conversations to help shape the style and context of that day’s interaction.
The word “tired” can mean very different things.
Tired with anxiety.
Tired with satisfaction after achieving something.
Tired with loneliness.
Tired from boredom.
Those are not the same experience.
And they may not call for the same conversation.
HAKONE is not designed to force the human being toward an average profile.
Instead, it asks:
Who is this person today?
The 16-Emotion Slider AI is the emotional interface designed around that question.
11 — HAKONE Multilingual Navigation AI
Not Just Translation—Navigation Without Losing Meaning
HAKONE was designed with international use in mind.
Its multilingual layer is therefore not intended simply to replace Japanese words with words in another language.
The goal is broader.
Across:
buttons,
instructions,
safety notices,
AI conversations,
learning functions,
and navigation,
the user should be able to understand in their own language:
where they are,
what the system is doing,
and what they can do next.
As language changes, the basic philosophy and safety-critical meaning of HAKONE should not disappear.
The multilingual layer is therefore designed to adjust communication and navigation while preserving the intent of the system.
HAKONE is not conceived as:
a Japanese AI that is translated later for foreign markets.
It is designed from the beginning for use across multiple languages and cultural environments.
12 — Safety-by-Design AI
AI That Watches the AI—With Humans Retaining Final Oversight
Safety-by-Design AI is a specialist safety layer designed to operate across the HAKONE architecture.
Its role is to help monitor boundaries across HAKONE’s different AI systems.
HAKONE’s internal safety framework may draw on publicly available principles and guidance concerning digital environments, well-being, responsible AI, and the safety of children and young people published by organizations including WHO and UNICEF.
These materials are used as reference inputs for HAKONE’s own safety-design standards.
The system is designed to examine areas such as:
AI-generated content,
user interaction,
defined safety boundaries,
and notification-related logic.
Where behavior may fall outside established criteria, the architecture is designed to make that deviation identifiable.
Most importantly:
HAKONE does not assume that AI should supervise itself without human involvement.
For important areas, a human path remains.
The safety architecture can combine:
automated monitoring,
defined safety rules,
recording and traceability,
human review,
and, where necessary, suspension or correction.
The purpose is not to build:
AI for the sake of using more AI.
It is to build:
AI that helps other AI remain within defined safety boundaries.
References to WHO, UNICEF, or other international organizations do not imply certification, approval, endorsement, partnership, or authorization by those organizations.
Their publicly available principles and guidance are used only as reference materials in HAKONE’s safety-design process.
ONE HAKONE. TWELVE INTELLIGENCES.
HAKONE is not a system that places everything in the hands of one giant AI.
It is an architecture of specialized intelligences.
An AI that helps create a stopping point.
An AI that redirects time toward learning.
An AI that works with emotional inputs.
An AI that supports creative development.
An AI designed to notice unusual change.
An AI that connects languages.
And an AI layer designed to watch the safety boundaries of the others.
Each specialist system has a different role.
Each has a boundary.
And each is designed to work with the others around one human being.
That is the HAKONE architecture.
The technologies described from 01 to 12 are not isolated features.
Together, they form a continuous pathway:
step away from the commercial algorithm,
return attention to yourself,
learn,
create,
move into the real world,
and connect with human support when needed.
What we are building is not:
an AI designed to make people stop using smartphones.
It is:
a twelve-layer infrastructure for human digital independence in an age of increasingly powerful algorithms.
These are not twelve separate projects.
They are twelve answers to one question:
As AI and algorithms become more powerful, how can human beings remain human?
HAKONE begins there.
Our Journey
OUR JOURNEY
2011
Designing the Human Side Instead of Fighting the Algorithm
The origins of HAKONE go back to 2011, before the word “AI” had become part of everyday conversation.
Internet services, search engines, video platforms, social media, and smartphones were rapidly becoming part of daily life.
At the same time, a new era was beginning—one in which a growing number of our “next actions” were being influenced by algorithms.
The first question we asked was not:
How do we stop the algorithm?
Commercial algorithms would continue to advance.
They would become increasingly sophisticated and increasingly important to modern society.
So perhaps the real challenge was not to eliminate them.
Perhaps what humans needed was a system that would allow us to coexist with advanced algorithms without surrendering the final choice.
In 2011, we began exploring that idea through design research.
The earliest thinking that would later become:
05 — Digital Public Health Audit
and
06 — Adult Digital Independence Infrastructure
began here.
2014
Moving into the Real Lives of Young People—and the Beginning of Citizen Science
Human behavior cannot be understood through theory alone.
So in 2014, we created a practical environment where young people who had stepped away from conventional educational or employment paths could experience real work and social participation.
We called it a work-experience environment for young people outside the conventional employment track.
The setting was a restaurant specializing in tuna cuisine.
There, young people could:
work,
talk with other people,
eat,
draw,
and use their smartphones.
We believed that these ordinary moments of everyday life contained important information about what it means to be human in the digital age.
Ten Years with Approximately 1,500 Families
From this work, with the cooperation of young people and their families, we began a long-term process of accumulating observations and data involving approximately 1,500 families.
This was not research confined to a laboratory.
It took place in real life.
What kind of day was it?
What were they watching?
What were they feeling?
What did they do?
What did they draw?
How did they interact with other people?
By accumulating these everyday records over time, the project developed as a form of citizen-participatory research and development.
Several of the technologies that would later become part of HAKONE began to emerge from this work.
Turning Emotion into Data
In 2014, we created an early version of the emotional-input system that would eventually lead to:
10 — 16-Emotion Slider AI
At first, it was not sophisticated AI.
It was an extremely simple system built in Excel to record everyday emotions and personal states and compare changes over time.
But the system continued to evolve.
We repeatedly revised:
the input categories,
classification methods,
time-series comparisons,
and the way emotional information could connect with conversation.
Over time, this developed into one of HAKONE’s central principles:
Never reduce a human being’s state today to a single word.
Looking at Art
In the same environment, we also began paying attention to the illustrations created by young people.
What they drew.
Color.
Composition.
Changes over time.
Their concentration during creation.
And the digital content and everyday experiences surrounding them during the same period.
By observing these elements together, we began exploring how creative activity and the digital environment could be understood within the same context.
This later developed into:
09 — Illustrated Drawing Coach AI
The objective was never for AI to determine the value of a human work of art.
The idea was to record process and change, then return useful information to the creator and, where appropriate, to a human teacher or evaluator.
That principle emerged directly from the work we did alongside these young people.
2016
Synapse Cascade Is Founded
On November 18, 2016, Synapse Cascade Gomei Kaisha was established in Japan.
Synapse represents connection:
between people,
between humans and technology,
and between different forms of intelligence.
Cascade represents the idea that one small change can trigger another, creating a continuing flow toward the future.
After the company was founded, our research expanded beyond human behavior itself.
We began looking more deeply at the commercial algorithms shaping that behavior.
2016–2023
Do Not Criticize the Algorithm Only from the Outside. Test It from Within.
We independently studied the types of recommendation, selection, distribution, and demand-prediction systems used across major technology platforms.
But analysis alone was not enough.
The important question was:
Would our understanding survive contact with the real world?
To test that, we used our own restaurant business as a live commercial environment.
On delivery platforms, we observed:
what was selected,
under what conditions visibility and ordering behavior changed,
how product structure affected outcomes,
how demand shifted,
how timing mattered,
and how users behaved.
Platforms including Uber Eats became practical environments in which hypotheses about algorithmic systems could be tested against real commercial activity.
That work also contributed to strong performance by our restaurant within sushi and Japanese-food categories.
But the most important lesson was not commercial.
It was this:
You cannot fully understand an algorithm simply by observing it from the outside.
You have to use it.
Watch what it does.
And observe how humans respond.
This practical, field-based approach became part of the foundation of:
05 — Digital Public Health Audit
Our purpose was never to attack commercial algorithms.
It was to understand what they could do—and then build an algorithmic capability on the human side within the same technological civilization.
That path eventually led to HAKONE.
2018–2023
The Data Began to Divide into 12 Different Roles
As years of everyday data and field experience accumulated, we began to realize that one AI could not adequately handle everything.
Stopping.
Feeling.
Understanding.
Learning.
Creating.
Protecting.
Each of these requires a different kind of intelligence.
From this realization, specialist functions began to separate and evolve.
This became the foundation of what would later become the HAKONE Architecture.
2024
HAKONE Prototype Zero
After years of accumulating research and technical ideas, we integrated them into:
HAKONE Prototype Zero
For the first time, previously separate areas of research began to converge within one system.
The objective was no longer only to understand human digital behavior.
We began designing an architecture in which the user could:
stop once,
return attention to the body,
receive support adapted to the individual,
and decide what to do next.
At this stage, the foundation layers of today’s HAKONE began to come together:
01 — HAKONE AI
The intelligence that integrates the whole system.
02 — Yubizen
The technology of stopping.
03 — Acoustic Glass Resonance
The technology of returning attention to physical sensation.
04 — LADDER 2.0
The technology of individualization.
2025
HAKONE AI Comes Online
In 2025, HAKONE began operating in earnest as an AI system.
The goal was never to create another screen-time management app.
The goal was to create a private AI that could help a person living inside a highly algorithmic society:
notice,
stop,
understand,
and choose the next action for themselves.
Drawing on insights developed through years of citizen-participatory research involving approximately 1,500 families, we continued testing and refining the system with real-world use in mind.
HAKONE also began evolving from a single AI into an architecture in which multiple specialist AIs could work together.
Designing What Happens After the Stop
One question became increasingly important:
What does a person do after putting the smartphone down?
If there is nothing waiting on the other side of the screen, it is easy to return to scrolling.
That led to the design of:
08 — TATOEMON Learning Refuge AI
Its purpose is to turn time away from algorithms into time invested in:
professional qualifications,
specialist knowledge,
languages,
and personally meaningful fields of study.
Stopping is not the final objective.
The objective is to give recovered time somewhere new to go.
That is the educational design behind HAKONE.
Making Creation a Place to Go
Another destination is creativity.
We developed the experience of observing young people’s illustrations since 2014 into:
09 — Illustrated Drawing Coach AI
The system is designed to support learning around:
drawing technique,
materials,
composition,
creative process,
and the user’s own development over time.
AI can help organize observations and learning support, while relevant information can also be passed to human teachers or evaluators where appropriate.
From time spent consuming what others have created,
to time spent creating something of your own.
This is the role of HAKONE’s creative AI.
Understanding the Person Who Exists Today
The emotional records that began with an Excel sheet in 2014 evolved into:
10 — 16-Emotion Slider AI
Instead of reducing a person to one “mood,” the system uses multiple emotional dimensions alongside past states to help shape the conversation around:
What does this particular user need today?
HAKONE is not designed to push humans toward an average.
Difference is expected.
The AI should be designed around the reality that every person can be in a different state on a different day.
That is the principle behind this layer.
Noticing “This Is Different from Usual”
As we began working with long-term records, another insight emerged.
Sometimes the most important distinction is not:
safe or dangerous.
It is:
different from this person’s normal pattern.
That idea led to the design of:
07 — HAKONE Sentinel AI
Depending on the functions enabled, this safety layer is designed to consider signals such as fingertip interaction, user-provided inputs, conversations, and changes over time.
Its purpose is to identify days or patterns that differ significantly from the user’s own usual baseline and, where predefined conditions and prior user consent are in place, create a pathway toward human attention.
The AI is not there to diagnose.
The system is not designed to end with AI.
Where appropriate, the path continues to a human being.
That is the role of Layer 07.
2026
HAKONE Architecture — Twelve Intelligences
In 2026, HAKONE developed into a digital-independence architecture composed of twelve distinct roles.
01
Integrate the whole system.
02
Stop.
03
Return to the body.
04
Adapt to the individual.
05
Audit society and algorithms.
06
Build the social infrastructure.
07
Protect.
08
Learn.
09
Create.
10
Understand.
11
Connect across languages and cultures.
12
Govern the whole architecture safely.
HAKONE is not an app made by collecting twelve unrelated features.
It is a system designed as one continuous cycle:
a person steps away from the algorithm,
returns to themselves,
and reconnects recovered time with the next part of life.
2026
From Family Support to Adult Digital Independence
For many years, much of our work focused on children and families.
But as the research developed, a larger gap became visible.
Age 18.
Minors often have:
parents,
schools,
age restrictions,
household rules,
and other forms of protection or supervision.
After adulthood begins, much of that structure disappears.
But the algorithms do not.
Social media remains.
Video platforms remain.
Games remain.
Online shopping remains.
Generative AI remains.
Delivery platforms remain.
Advertising remains.
In fact, as a young adult’s world becomes larger, the number of situations in which they encounter algorithmic systems may increase.
And yet there remains relatively little social infrastructure specifically designed to teach an adult:
how to disengage from digital systems by their own choice.
We came to see this as:
the policy gap in adult digital independence.
HAKONE therefore began shifting its primary focus toward young adults aged 18 and over.
2026
From Japan to Public Institutions Around the World
Digital algorithms do not stop at national borders.
An 18-year-old in Tokyo,
an 18-year-old in New York,
an 18-year-old in London,
an 18-year-old in Paris,
and an 18-year-old in Singapore
may all interact with digital environments connected through the same global platforms.
This cannot therefore be understood only as a problem for one country.
In 2026, we began presenting:
06 — Adult Digital Independence Infrastructure
as a broader policy concept.
We began sharing information and proposals with public institutions and government bodies working in areas such as:
education,
public health,
digital policy,
communications,
and youth policy.
The central question was:
What happens after age 18, when formal protection decreases but the algorithmic environment remains?
Our proposal is that a significant gap still exists in adult digital independence—and that the HAKONE Architecture may offer one framework for addressing it.
The Role of Multilingual Navigation
An important part of this international direction is:
11 — HAKONE Multilingual Navigation AI
Its purpose is not simple translation.
The goal is to help users in different language environments understand:
HAKONE’s philosophy,
its safety information,
its operation,
its learning functions,
and the next action available to them.
HAKONE is not intended to be:
a Japanese system translated for export.
It is being designed from the beginning as digital infrastructure intended for multilingual and multicultural environments.
And Then, the Twelfth AI
The Stronger the Technology Becomes, the Stronger Safety Must Become
The final layer of HAKONE is:
12 — Safety-by-Design AI
As Layers 01 through 11 become more capable, the importance of governing AI behavior also increases.
That is why HAKONE does not treat safety as a warning written at the bottom of a Terms of Service page.
Safety is treated as part of the system itself.
Drawing, where appropriate, on publicly available international thinking around safety, responsible technology, and digital well-being, HAKONE combines mechanisms such as:
automated monitoring,
rule-based controls,
logging and traceability,
human review,
and correction or suspension when necessary.
AI can help monitor AI.
But the final responsibility remains human.
That is the principle behind HAKONE’s safety architecture.
2011 → 2026
For 15 Years, the Question Has Been the Same
In 2011, the question was not:
How do we stop the algorithm?
It was:
How can human beings live alongside algorithms without losing the final choice?
From there, the journey moved forward.
In 2014, into the lives of young people.
Into citizen science with approximately 1,500 families.
Into emotion.
Into illustration.
Into the fingertip.
Into commercial algorithms.
Into AI.
And eventually, into global policy discussions.
The technology changed.
The focus expanded—from families to adults aged 18 and over.
But the original question never changed.
Technology does not need to become weaker.
Human beings simply must not surrender control.
HAKONE is building twelve forms of intelligence around that principle—
from Japan to the world.