The last interface.

As technology gets smarter, it should also be easier for us to use it, more intuitive, frictionless. Right now, you still have to know exactly what the system can and can’t do, translate your goal into a plan it understands, and babysit it through every step of execution. You start with an intent, but what comes back is an approximation of it, filtered through your own fluency with the tool. AI agents promise to close that gap, planning and executing on your behalf instead of waiting to be told what to do. But they only solve half of the problem: an agent that can execute anything still has to guess what you actually mean. Autonomy without a true understanding of intent doesn’t remove the friction. It just relocates it.

We have always been adapting to tech, tech should adapt to us now. The answer isn’t a smarter agent guessing harder. It’s technology that gives you more control while demanding less of your attention. You set the direction and remain free to steer or stop; the system handles the mechanics. Less operation should not mean less agency. It should mean the opposite. For that to work, AI needs more than an instruction. It needs the context from which the instruction came. Human thought is far richer than the formats our computers accept. It does not fit neatly into text, images, audio, and video. It weaves together intuition, abstract reasoning, emotion, memory, and the full range of human sensation: from touch, taste, and smell to proprioception. It draws on knowledge we use without ever putting it into words. When that context can reach AI, execution will no longer limit creation. What you can make will depend less on what you know how to prompt, code, or operate, and more on what you can imagine. That kind of access should belong to everyone.

Getting there starts by admitting that words are a bottleneck. Your intent is a sea of nuances, yet most foundation models still receive it through the narrow straw of language. So we are moving beyond text-prompted models toward thought-prompted ones. We call it a Brain-AI Interface: not merely a brain-computer interface decoding which button you meant to press, but a seamless extension of yourself, the way your arm already is. You form the intention to grab something; you don’t calculate which muscles to fire. The intention is yours. The interface handles the execution. Realizing this vision requires radical accessibility: a discreet wearable built for daily life, not an expensive lab rig or implant, paired with software that adapts until using it feels instinctive. Communicating intent is not another feature. It is the missing link that makes natural interaction with intelligence possible. But that link cannot be built from existing data. Today’s foundation models learned from an internet of text, images, audio, and video. There is no internet of thoughts. It cannot be scraped; it has to be created: deliberately, consensually, and at scale. Building the first reliable way to do that is the work ahead of us.