Writing this isn't easy, as I grapple with my emotions. This piece holds great significance for me; it's been on my mind for quite some time, and I take pride in sharing it. It encapsulates my thesis and philosophy of life, reflecting my beliefs about the essence of creation.

This piece is crafted using my newly launched editor, Creader, which embodies my vision for creating novels and other literary works. With AI support and element nodes for everything, I believe that creation in this new era is not reliant on AI alone, but enhanced by it. As model evolution accelerates this year, we've witnessed vibrant discussions on generative music and art. Yet, when it comes to text-based creation and storytelling, humans remain the ultimate creators; it's the power of sentence crafting that truly conveys emotion and the essence of the heart. Therefore, I see AI in writing not as a substitute but as a collaborator, assisting in crafting a complete narrative.

Moreover, I posit that "vibing" is the future. Historically, many celebrated works have emerged from amateur authors, individuals with rich and complex life experiences. These unique stories deserve to be shared. Often, we humans grapple with vague, unformed thoughts. With Creader, I aspire to provide an opportunity for people to express themselves freely, sharing their narratives as effortlessly as we vibe while coding with tools like Cursor and Claude.

-----------------

I have been thinking about this idea for more than ten years.

Not just about building a product, but about creation itself, about why writing feels difficult, why some stories endure, and why others disappear. This piece represents a thread that has followed me from reading web novels in a foreign language, to studying computer science, to experimenting with blockchain and AI.

As AI tools become more powerful, many believe creation will become easier. I am not sure that ease is the goal.

Writing has never been difficult because we lack tools. It is difficult because ideas are fragile. They begin vague and unformed, and the act of shaping them requires time, tension, and restraint.

The question I keep returning to is simple:

How do we build tools that support creation without replacing the thinking that makes it meaningful?

The start

Reading has always been integral to me. I was drawn to stories that created entire worlds, rules, characters, and consequences that felt larger than the page. I immersed myself, imagining not just the text but what lay beyond it.

This idea has been with me for more than a decade. When I first moved to the UK, I could barely speak or write English. Ironically, it was stories that kept me grounded. Web novels became both an escape and a source of motivation. Through them, I began thinking about what I wanted to build in the future.

One day, I wondered whether the publishing mechanisms popular in Asia could exist in Western markets. Whether the structure of online storytelling could be redesigned. That question led me to computer science. I started learning to code, trying to build small sites around the idea.

I was not very good at it then, and university exams eventually pulled me away. But the question never disappeared.

The evolution

Time pasted too quickly, where I lost a bit, but eventually found new a angle for the idea

During both my undergraduate and postgraduate years, the idea never left me. I kept returning to it quietly, even when I did not know how to carry it out.

I tried to approach it practically. I considered starting a company during undergrad and began learning the necessary skills, from technical foundations to business thinking. In a world before AI assistance, those lessons forced me to understand structure, trade-offs, and constraints. For my final academic project, I built a small reading platform. It was not ambitious, but it was the first concrete step.

After graduating, I stepped outside academia. I spent a year working in operations at a large Chinese technology company. The pace was intense. I expected to struggle, but instead I adapted. That experience taught me discipline and execution under pressure, but it also clarified something else. Building systems at scale is different from building meaning.

Around the same time, I began exploring blockchain with two co-founders while preparing for my Master’s degree. It felt like a new frontier. For the first time, I thought infrastructure might solve the problem I had been circling for years.

The Blockchain

Tokenisation is not a disaster, it is a new way to enhance liquidity to static assets

I cannot clearly remember the first moment I encountered blockchain. What I remember is the sense of possibility. It felt like infrastructure could solve problems that traditional systems could not.

With two co-founders, I began exploring whether ownership and intellectual property could be redesigned through on-chain mechanisms. If stories could be registered, split, and distributed transparently, perhaps we could create a new economic model for publishing. We studied cases such as the Chinese copyright chain and the use of blockchain as legal evidence. We built a pitch deck. We entered competitions. For the first time, I spoke publicly about the idea.

The feedback was encouraging. Technically, much of it was feasible. We could design parent and child NFTs. We could model revenue splits. We could write a white paper and deploy demos.

But something felt misaligned.

The real obstacle was not technical infrastructure. It was human friction. Reading and writing are among the most natural actions in the world. Introducing wallets, tokens, and onboarding complexity made the experience heavier, not lighter. I kept asking myself a simple question: Does adding this layer truly help someone create or read better?

My answer was no.

That realisation became the foundation of my master’s thesis and, more importantly, a shift in direction. Infrastructure alone could not protect creativity. Ownership was not the missing piece. The problem I had been circling for years was not economic. It was structural and cognitive.

The AI era

AI has made such a huge change to our life, it is surreal to see myself as a mid-level developer, building this whole thing out.

Time passed, and the idea remained in the background. I kept thinking about it but never fully committed. Something still felt unresolved.

That changed with the rise of AI coding tools. Watching non-technical builders create functioning products in weeks forced me to rethink everything. It was not just about speed. It was about the possibility. For the first time, individuals could prototype complex systems without large teams or funding.

When I joined Story, I was working at the intersection of IP and infrastructure again. The experience sharpened my understanding of systems, constraints, and long-term design. But as models improved, a different realisation began to form.

AI was not solving creation. It was accelerating output.

That difference mattered.

By late 2025, I stopped treating the idea as something abstract. I began building again, this time with new tools and a clearer question in mind. If AI could amplify capability so dramatically, then the real responsibility was not to replace thinking, but to protect it.

Creader emerged from that shift.

The Creation

Creation happens in the delay between confusion and clarity.

My path in creation has taken many forms, from storytelling and games to writing and coding. The medium changed, but the impulse did not. I kept returning to the same question: what does it actually mean to create?

Creation is often described as inspiration or self-expression. In practice, it is neither. Creation is the act of shaping ambiguity into form without resolving it too early. Language makes this visible. A vague idea can survive in the mind indefinitely, but once it is written down, its weaknesses surface. Writing is where thought becomes testable.

This is why writing feels difficult. The difficulty does not come from a lack of vocabulary or tools. It comes from consolidation. Ideas arrive scattered and expansive. Writing forces them into structure. That friction is not a flaw in the process. It is the process.

Long-form storytelling makes this even clearer. As a narrative grows, maintaining alignment between events, motivations, and consequences becomes increasingly complex. The problem is not carelessness. It is structural. Human memory is not designed to hold a large narrative system intact over years of iteration. As George R. R. Martin has explained, the later books in his series became harder to write not because he forgot the story, but because every new decision had to respect the accumulated weight of earlier ones.

With the rise of large language models and agents, many tools now present AI as the solution to this difficulty. They promise clarity, completion, and acceleration. But creation does not suffer from a lack of fluency. It suffers from premature coherence.

An LLM can resolve ambiguity instantly. It can produce text that sounds complete before the underlying thinking is complete. Fluency, however, is not clarity. When coherence arrives too early, alternative possibilities collapse. The space where originality might emerge narrows.

In this era, originality and taste matter more, not less. AI can assist, but it cannot replace the primary act of judgment. It should extend memory and surface connections, not decide for the creator.

After trying many existing tools, I noticed something else. Ideas that should remain in tension were scattered across documents and applications. Structure and prose lived in separate spaces. The act of switching between tools often forced premature decisions just to maintain momentum. The friction was not productive. It was fragmenting.

Agentic Interaction and the Writing Lifecycle

The way we interact with AI is changing. Early tools were prompt-based. You asked a question, the model responded. Now, we are entering an era of agentic systems. These systems do not simply respond. They plan, execute, revise, and iterate across multiple steps.

This shift has important implications for creative tools. If AI is no longer a single-response engine but an ongoing collaborator, then the design of writing platforms must adapt. The question is no longer how to generate text, but how an agent should interact with structure over time.

Writing is not a single action. It has a lifecycle, much like software development. In traditional development, a product moves from concept to design, from architecture to implementation, then testing and revision. Each stage serves a different purpose.

Writing follows a similar pattern. An author begins with setting and structure. Plot and world rules are defined before chapters are written. Drafting is followed by revision and cross-checking. Ideas evolve across stages.

If we look at both processes at a high level, we can distinguish between mechanical labour and cognitive judgment. AI is increasingly capable of handling mechanical tasks such as formatting, summarising, recalling details, or identifying inconsistencies. What remains distinctly human is taste, direction, and thematic coherence.

Agentic systems will amplify this separation. The challenge is not to let agents decide the work, but to design systems where agents support each stage of the creative lifecycle without overriding human intent.

That tension led me back to the central question. If uncertainty is essential to creation, then the tools we use must be designed to hold it rather than eliminate it.

The Creader

If a tool makes your writing feel smarter than you are, be suspicious.

If uncertainty is essential to creation, then a writing tool cannot be designed around completion. It must be designed around containment. It must be able to hold unresolved structure without forcing it into finished prose.

That constraint shaped Creader.

The goal is not to automate thinking, but to protect it. AI should not become the primary thinker. It should extend memory, surface relationships, and assist exploration without collapsing possibilities too early. Judgment, taste, and direction remain human responsibilities.

Most creative difficulty does not come from a lack of ideas. It comes from the inability to manage them. As a story grows, characters, events, rules, and motivations multiply. The writer is forced to rely on intuition where structure is required. Over time, memory becomes the bottleneck.

Creader separates the story's reality from the sentences that describe it. Before prose, there is structure. Stories can be decomposed into elements such as characters, events, settings, and relationships. These elements exist independently of how they are written. They need to remain accessible and consistent across time.

The knowledge base exists for this reason. It serves as a source of truth for a story's internal logic. Events, character arcs, and world rules are stored as structured nodes that can be referenced and revised without rewriting entire chapters. The system is readable by both humans and machines, not to replace authorship, but to support coherence.

AI in this context does not generate a book for you. It helps you navigate the structure you are building. It sparks connections, recalls forgotten details, and highlights inconsistencies. It assists without deciding.

The objective is simple. Enable individuals to explore their own worlds without losing them in fragmentation or premature closure.

Closing thoughts

We are living in the best era, not just as a developer, but as a normal person who want to connect deeper with the world and exploring new things

Looking back, it is clear that I have been on this path for a long time. From childhood experiences with Pokémon, Fire Emblem, and Golden Sun to assembling Gunpla, I have always been drawn to structured worlds. Settings, rules, and interconnected narratives shaped how I learned to think. Abstraction and system building were not academic concepts to me. They were part of growing up.

Years ago, when I first imagined building something like this, I used to say that the goal was to find and support the next great story. Not by predicting it. Not by generating it. But by making space for it to exist.

We are entering an era where individuals can build, write, and publish at a scale that once required entire teams. As this power accelerates, the risk is no longer a lack of tools, but a lack of reflection. Speed will continue to increase. Output will become abundant. Meaning will not.

Creation does not need to be pushed harder. It needs to be handled more carefully.

Creader is not a machine for producing stories. It is a piece of land.

Stories begin as fragile ideas. They are seeds, uncertain and unfinished. They require time, structure, and attention. They cannot be forced into existence. They must be cultivated.

If the soil is right, growth follows.

Creader exists to provide that soil. A place where ideas can remain incomplete long enough to take root. A place where structure supports growth instead of constraining it. A place where one story may become many.

We may not know which seed will grow into a forest. But the responsibility is to prepare the ground.

私の考えた最強の世界。それを描くために私は絵を描いているので設定が命なんです。 -- 浅草みどり, <映像研には手を出すな！>