Designing Human Capability

If you strip the technology away from everything I want to make, what remains is always the same question, asked in a different domain: how do people grow? How do they organize knowledge, truly learn a language, keep practicing a hard skill, internalize what they read, choose where to put their attention, understand each other without shields.

These are not technical problems. They are problems of human development, and computation is one material for approaching them — alongside cognitive science, behavioral design, and lived practice. The question is never can this be built? but should this be built, and if so, what kind of people will it help become?

Accessibility

Almost any human capability — from drawing to programming to a new language to system design — is learnable, and learning it is an engineering problem that can be optimized. Yet most of these capabilities remain gated: by credentials, by jargon, by access to experts, by the residue of an education system that is largely anti-human and anti-learning. If we design the right platforms and frameworks, the capabilities that look deep and distant can become non-monopolistic and accessible to anyone.

Learning should respond to people's core nature — curiosity and playfulness — rather than suppressing it. It should awaken subjectivity, not dull it.

Agency

We are entering an era where AI will do more and more of what humans used to do. The danger isn't only job displacement — it's the quiet erosion of subjectivity. When machines decide, recommend, and generate, humans risk becoming consumers of capability rather than makers of it. My mission is to make sure people in the AI era still learn, grow, create, and retain agency.

This means designing infrastructure that puts the human in the loop of their own growth: feedback that's visible enough to keep going, knowledge systems that reward internalization over accumulation, practice loops that make improvement felt rather than assumed.

Technology must face reality

Tech circles often carry a fantasy of technical omnipotence — the belief that any sufficiently advanced tool will fix things by existing. But technology lives in a real world trending toward closure: platforms walling in, algorithms shaping attention, precarious work normalizing, nihilism spreading among the people who grew up inside these systems. A tool that doesn't face this reality is, at best, a distraction.

The question for any creator shouldn't only be "what can this achieve for us?" but "how do we use it to respond to this world?"

This is why I care about behavior design, about dialogue spaces stripped of jargon, about infrastructure that respects the whole person. Tools shape people, and people shape the world. The responsibility runs all the way through.

On knowledge and its forms

A limit of current human development is that knowledge must be expressed in human-comprehensible forms to count as knowledge — a mathematical proof no one else can parse cannot be acknowledged. Our cognitive horizon is bounded by the forms and senses we have. As we extend AI into new "modalities," we're still mapping its intelligence onto human-sensed categories: text, image, audio.

But what if there are knowledge forms native to AI that fall outside human sense entirely — forms we can't currently imagine, only accept by correctness? Can we imagine knowledge we cannot recognize? This isn't a distant abstraction; it touches how we design learning systems, knowledge tools, and the boundary between human and machine cognition. The question stays open.

What I'm building toward

Each project in Practice is infrastructure for one facet of human growth: knowledge intake, knowledge organization, skill building, attention and agency, knowledge internalization. None is a product in the SaaS sense. They are pieces of a longer argument — that human growth itself deserves infrastructure, the way code has Git, computing has Linux, and thinking has Obsidian.

If you're working on adjacent edges, let's talk.