Source: Unsplash
Imagine trying to send an email by thinking about it. Not metaphorically. Literally composing thoughts into words into sent messages without touching a keyboard. That's what brain-computer interfaces promise. The hardware is getting there. The interaction design? Nowhere close.
Brain-computer interfaces are no longer a science fiction premise. Synchron is targeting its pivotal FDA clinical trial in 2026, the study required for the very first premarket approval of an implantable BCI. Neuralink has expanded its human trials to over a dozen participants, with its N1 chip featuring 1,024 electrodes placed in the motor cortex. Total investment in BCI companies across 2024 and 2025 surpassed $1 billion, with Neuralink alone raising a $650 million Series E in mid-2025 at a $9 billion valuation. We are at the commercialization threshold. The problem is that nobody has seriously worked on what it means for a human being to actually use one of these things day-to-day.
"If we can read from the brain with increasing precision, the boundary between private thought and public expression dissolves. If we can write to the brain, the boundary between self and system dissolves."
— Tech for Impact Summit, BCI Analysis 2026
The Race to FDA Approval Is Accelerating
Two companies are leading the commercial charge right now, and they have taken opposite approaches to getting there.
Neuralink's approach is high-density and surgical: a robotic neurosurgical procedure implants flexible electrode threads directly into brain tissue. The N1 chip reads motor signals with enough precision to let paralyzed patients control cursors, type text, and interact with devices through thought alone. The company is building toward a commercial medical device pending FDA approval.
Synchron's approach is lower risk: their Stentrode device travels through the jugular vein and settles inside a blood vessel near the brain's motor cortex, reading signals through the vessel wall. No open skull surgery. The company raised $200 million in Series D funding in November 2025 specifically to run its pivotal trial and prepare for commercial launch. If the trial succeeds, Synchron becomes the first company in history with an FDA-approved implantable BCI.
Both paths point to the same destination: BCIs becoming real medical products, used by real people, in real daily life. Which is exactly when the design problem becomes unavoidable.
The Part Nobody Is Building: The Interaction Layer
Here's what strikes me as someone who designs enterprise products for a living. The BCI industry has some of the most sophisticated hardware engineering in the world. Flexible electrode arrays, robotic neurosurgery, AI-powered signal processing. Billions in funding. The best neuroscientists and materials engineers money can hire.
And almost nobody is thinking seriously about the human experience of using these devices.
When I look at what has been published about BCI usability from the research community, the framing is almost always clinical. Response accuracy. Signal fidelity. Information transfer rate measured in bits per second. These are the right metrics for a medical device. But they are not the right metrics for a product that a person is supposed to live with, use daily, and rely on for communication or mobility or creative work.
For BCI technology to work for real users, three things have to be true simultaneously: the information transfer rate has to match a natural interactive pace, the error rate has to be low enough that users do not lose trust in the system, and the complexity of the interaction model has to be minimized to the point where people do not have to consciously manage the device. None of these are purely engineering problems. All three are fundamentally design problems.
Why Normal UX Principles Break Down for BCIs
The interaction model for every screen-based product we have ever designed rests on assumptions that BCIs completely upend.
Conventional UX assumes an intentional, physical input: a tap, a click, a swipe, a voice command. The user makes a deliberate choice to perform an action. The system responds. The feedback is immediate and visible. This is the loop every design pattern in the history of digital product design has been built around.
With BCIs, the input is neural activity. And neural activity does not come with a clean on/off switch. Your brain does not distinguish cleanly between thinking about moving your hand to click and idly imagining clicking something. The noise floor is extraordinary. The brain's attentional networks, the Dorsal Attention Network, the Ventral Attention Network, the Default Mode Network, the Executive Control Network, are all generating signals simultaneously. The BCI has to decode intent from a sea of overlapping electrical activity.
What makes this a design problem, not just an engineering problem, is that the cognitive load of using a BCI is entirely invisible. A user has no idea whether their mental state is producing clean, readable signals or noise. There is no interface affordance for signaling scattered attention. There is no visual indicator of signal quality in terms the user can act on. The interaction model for thinking more clearly to make this work better simply does not exist yet.
This is the most fundamental unsolved design problem in BCI: giving users meaningful control over something they have never had to consciously manage before, their own mental state.
What BCI Interaction Design Actually Needs
I have been thinking about this through the lens of what we know works in adaptive interfaces. I wrote about AI-native UX earlier this year on Medium, and some of those principles apply directly here, just with the stakes substantially raised.
Five things BCI interaction design needs that do not exist yet as coherent design systems:
- Real-time signal legibility feedback: Users need to know, in terms they can understand and act on, whether the system is reading them clearly. Not raw electrode data. A simple, intuitive indicator of connection quality that translates into something the user can change about their physical or mental state.
- Intentionality affordances: There needs to be a designed "on" state, something that tells both the user and the system that intentional input is happening. Most BCI systems today use dwell time or specific motor imagery patterns. Neither is intuitive enough for mainstream use.
- Graceful error recovery: When a BCI misinterprets a thought as a command, and it will, the user needs to be able to correct it without frustration spiraling into learned helplessness. The undo/redo mental model does not map cleanly to neural input.
- Adaptive calibration UX: Every brain is different. Every person's neural patterns are unique. The system has to learn from each individual user over time. The design challenge is making that learning process feel natural, not like recalibrating a medical device.
- Privacy boundary design: Users need to trust that incidental mental activity, thoughts they are not intending to act on, is not being captured or misinterpreted. This is both an engineering requirement and a design requirement. The product needs to communicate clearly what it is and is not reading.
What This Means for Product Designers Right Now
I want to be clear about where we are on the timeline. Nobody is designing consumer BCI apps today. The current clinical use cases are mostly mobility restoration and communication assistance for people with paralysis or ALS. These are deeply important applications, and the design work happening in clinical settings is meaningful.
But the industry is moving fast. Synchron is one pivotal trial away from commercial approval. Neuralink has raised enough capital to iterate through multiple product generations. Precision Neuroscience and Kernel are advancing along parallel tracks. Within 3 to 5 years, the question will not be whether BCIs exist as consumer products. It will be whether anyone designed them well enough for ordinary people to actually use.
If the humanoid robot industry is making the mistake of treating UX as an afterthought, the BCI industry is several steps behind even that. At least humanoid robots have visible, legible actions that designers can work with. BCIs ask designers to build interactions around invisible mental states, individual neurological variation, and feedback loops that have never existed in any product category before.
This is the frontier work. It requires neuroscience literacy, clinical context, and the willingness to throw away most of what you know about screen-based interaction design. The designers who get there early are going to define what it means to interact with computing at the most intimate possible layer.
The patterns we are building now in AI-native UX, designing for ambiguous intent, for adaptive systems that learn from individual behavior, for interactions that happen faster than conscious thought, are rehearsal for the BCI design challenge. The skills transfer. The context is just far more extreme. I explored this connection in my AI readiness work at reloadux.com earlier this year, and the thread runs directly from there to here.
The hardware is almost ready. The clinical trials are running. The regulatory pathway is clearing. What is missing is a design community that takes this seriously before the first commercial products ship and define broken patterns for the next decade.
Are you thinking about BCI design? Are there companies in this space actually doing the UX work well? Drop your thoughts in the comments. I would genuinely love to know what people are seeing on the ground.
Sources:
1. TechTimes — Synchron Brain Implant Targets 2026 Pivotal Trial for First FDA-Approved BCI (June 6, 2026)
2. ApplyingAI — Transforming Brain-Computer Interfaces: Neuralink's 2026 Breakthroughs and Market Dynamics (April 2026)
3. Tech for Impact Summit — Brain-Computer Interfaces 2026: Neuralink, Synchron, and the Race to Read Your Mind
4. Built In — How to Design Products for the Brain, the Next Frontier in User Interfaces
5. Coderio — Brain-Computer Interfaces: User-Experience Design Principles
6. Basenor — Neuralink in 2026: Where the Brain-Computer Interface Stands Today