More human, by design.

As machines become better at making things, I want us to become more deliberate about what we make, why we make it and how we treat each other.

A world that is never finished

Wabi and sabi belong to Japanese aesthetic traditions shaped by an awareness of impermanence. They attend to simplicity, age and the character of things as they are. I take inspiration from that tradition, without claiming to speak for it.

For me, the deeper invitation is to stop treating existence as a problem waiting for a perfect solution. Everything changes. Every account of the world is partial. We live through relationships, and our actions leave traces in other people's lives. A thing can be unfinished and still deserve our care.

That changes how I want to build. Make something useful. Let it meet the world. Listen to the people who use it. Repair it. Let it evolve. Accepting imperfection asks for attention and responsibility; it cannot excuse careless work.

The moment we are in

AI can already generate language, images and code, and help people research and build. I use it in my own work. It gives a small operation access to capabilities that would once have required a much larger team. It also produces convincing errors. Fluency alone tells us very little about whether an answer is true. NIST's Generative AI Profile documents this risk.

We do not yet know how far these capabilities will go, how quickly they will spread or how the gains will be distributed. The ILO's 2025 study of occupational exposure found that one in four workers worldwide held a job with some exposure to generative AI. Exposure measures potential change, not observed job loss. A forecast is not a destiny.

Progress has to reach people

I see a possibility of more people being able to learn, create and run a business without needing permission from a large institution. I also see the possibility of concentrated ownership, weaker bargaining power and a public world flooded with persuasive material whose origins are difficult to establish. Neither outcome is inevitable.

Efficiency does not decide who receives the benefit. That depends on choices about ownership, access, working conditions and accountability. People need a voice in changes that affect their livelihoods. They need ways to question decisions and reach someone who can put things right.

My hope is that we use this capacity to relieve exhausting work and widen opportunity, while making room for care, learning and a life beyond productivity. Human dignity cannot depend on outperforming a machine.

The value of a real voice

When polished expression becomes easy to generate, I believe lived experience, judgement and an accountable point of view become more valuable. A person's history gives their words meaning: what they have tried, what they have noticed, what they are prepared to stand behind.

Authentic expression can include AI assistance. It requires that someone means what they publish and takes responsibility for it. The danger is letting a system smooth away every hesitation, local detail and unusual thought until everybody sounds equally plausible.

I want technology to help people find and express their own voice. The awkward sentence that says something precise may be worth more than a beautiful paragraph that could belong to anyone. We should have room to disagree, change our minds and speak without turning every experience into content.

Radical honesty, in practice

Say what is known. Say what is inferred. Say what remains uncertain. Show the sources behind a consequential claim. Name the limits of a tool before someone has to discover them through failure. When something goes wrong, acknowledge it, correct it and explain what changed.

Transparency must make a decision easier to understand. A pile of documents or a reassuring badge is not enough. Someone should be able to tell who is responsible, what AI contributed and how to challenge the result.

Honesty also respects boundaries. Client information, private lives and unfinished personal thoughts do not become public property in the name of openness. Share what people need to judge the work; protect what they have entrusted to you.

What this asks of Klarhimmel

I'm Felix. Klarhimmel is where I put these beliefs to work, with AI agents helping me research and build. I am responsible for the scope, the promises and the result.

Start with the actual business and the people in it. Give recommendations that can be questioned. Be clear about costs, dependencies and what has been tested. Make systems understandable enough to maintain, and leave people with meaningful choices about who helps them next.

For a hotel, a clinic or a family business, the result may be a quieter Monday: less duplicate typing, a dependable tool, a person who no longer has to carry the whole process in their head. That is a worthwhile use of powerful technology.

Keep becoming

This is a direction I want to live up to. It will require revision as the tools change and as I learn from the work.

Wabi-Sabi gives me a way to hold that uncertainty without pretending it has disappeared. Pay attention to what is here. Care for what you make. Leave room for other people and for what you do not yet understand.

The future I want has extraordinary tools and recognisable people: people with their own voices, the freedom to choose and the courage to tell the truth.

Klar: clear enough to understand, ready to act, done with care.

A note on the foundations

The reading of Wabi-Sabi here is my interpretation, informed by the Stanford Encyclopedia of Philosophy on Japanese aesthetics. The commitments to human agency, transparency and accountability also align with UNESCO’s Recommendation on the Ethics of AI. The outlook above expresses my values and hopes, rather than a claim to know the future. Revised October 2026.