Writing — Brain Computer Interfaces and Cerebral OS

The Next Interface Is Your Nervous System

November 9, 20265 min readBy Milan Cheeks

Brain computer interfaces will not replace screens overnight, but Cerebral OS taught me they will change who gets to use a computer at all.

The Next Interface Is Your Nervous System

Every computer I have ever used asked me to translate myself first.

I think in intent. I want the file open, the message sent, the game world built. But the machine does not take intent. It takes keystrokes, taps, swipes, and clicks. So I learned its language instead of it learning mine. We all did.

That is the quiet story of every interface in computing history. The keyboard translated thought into letters. The mouse translated thought into position. Touch translated thought into gesture. Voice, which my work on Harmony is built around, translates thought into speech. Each step removed a layer between what a person means and what a machine does.

Brain computer interfaces are the next step in that same story, and they are the reason I built Cerebral OS.

The Interface Has Always Been a Translation Problem

When people hear the phrase brain computer interface, most picture science fiction. A person thinks a command and a machine obeys instantly. The reality today is much slower and much more interesting than that.

A brain computer interface reads signals from the nervous system and turns patterns in those signals into input a computer can use. The signals are weak and noisy. They change with fatigue and attention. The software has to learn what a pattern means for one specific person, and that learning has to happen again as the signals drift.

There is no mind reading here. What exists is pattern recognition applied to biology. A system notices that a certain signal tends to show up when a person focuses on a certain choice, and it learns to treat that signal as a selection. It is closer to teaching a very patient instrument than to opening a window into thought. Magic would not need builders. This does.

Why an Operating System, and Why I Call It Cerebral OS

I founded SELF Labs to build systems, not demos. SELF OS grew out of that instinct, a personal operating system that integrates goals, habits, and AI assistance. Cerebral OS comes from the same instinct applied to a harder problem.

If brain signals are ever going to be a real input method, they cannot live inside one app built for one experiment. They need a layer underneath, one that handles calibration, interpretation, and permission. Calibration so the system learns a person. Interpretation so raw signal becomes intent a program can act on. Permission so a person always knows what is being read, when, and what it will be used for.

Cerebral OS is my research effort toward that layer. It is research, and I want to be plain about that. It is not a finished product and it does not promise abilities that do not exist yet.

Every interface is a promise about who gets to participate. A better interface is a wider door.

The Hard Problems Are Not the Ones in the Movies

The hard problems in this field are not about control. They are about trust.

First, the data itself. Neural data is about as personal as data gets. So the questions that matter most are simple ones asked early, before the technology gets impressive. Who owns this data? Where does it live? Can a person delete it completely? If a builder cannot answer those questions clearly, the builder should not be collecting the data at all.

Second, consent has to be continuous, and accuracy has to be honest. Agreeing once, on a screen full of legal text, is not enough for a system that reads signals from your body. And a system that guesses at intent will guess wrong, so it has to show its uncertainty and let the person confirm what matters. Speed is worthless if the person cannot trust what got sent in their name.

I completed a UNESCO Global MOOC on the Ethics of AI in August 2026, and the lesson that stayed with me applies here directly. Principles are cheap at the start of a field and expensive to retrofit later. Brain computer interfaces are still at the start. That is exactly when the principles need to be built in.

Who This Is Really For

The first people this technology should serve are not people looking for a faster way to check messages. They are people the current interfaces have left out.

Think about what a keyboard assumes. It assumes hands that work in a certain way. A mouse assumes the same. Voice assumes speech that a recognizer can parse. For many people, one or more of those assumptions fails. For them, an interface that reads intent from signals the body still produces is not a novelty. It is the difference between using a computer with help and using one on their own terms.

That order matters to me. Accessibility first, convenience later. If the technology proves itself with the people who need it most, the broader uses will follow honestly.

Where This Goes From Here

I do not think brain computer interfaces replace screens or keyboards on any schedule worth predicting. The direction of computing has pointed the same way for fifty years, toward less translation between a person and a machine. This work points the same direction. It will not reverse.

That is also why I went back to school. I am currently enrolled in the computer science master program at the University of Colorado Boulder, running 2027 to 2028, alongside six CU Boulder algorithms certifications I completed in August 2026. Models get stronger every year. The question that decides who benefits is how a person reaches that strength, and on whose terms.

The next interface is your nervous system. Not because it sounds like the future, but because it is the shortest path between meaning something and making it happen, and building that path responsibly is work worth doing in the open.

If this work speaks to your audience, you can book me for keynotes via https://www.milancheeks.com/#contact

← All writing Book Milan →
The Newsletter

The frontier, in your inbox.

One email when a new essay drops — the Kardashev roadmap, AI-native building, and coexistence, written as it happens. No noise, unsubscribe anytime.

You're on the list.