Source: Unsplash
Neuralink is starting human trials of Blindsight. Synchron's patients are already browsing the internet using only their thoughts. Apple quietly built a brain-computer interface protocol into its operating system in 2025. The brain-computer interface revolution is not a future event. It is happening right now, in this quarter. And while everyone is talking about the neuroscience, nobody is asking the harder question: who is designing the actual user experience?
Brain-computer interfaces (BCIs) are the most extreme UX design challenge ever attempted. The BCI market reached $3.75 billion in 2026 and is growing at a 16.7% CAGR toward $15 billion by 2035, with over $1.6 billion in VC and institutional funding raised just in 2025-2026. Neuralink, Synchron, Merge Labs, and Science Corporation are all racing toward commercial availability by 2028. Product teams have roughly two years to figure out what it means to design an interface where the medium is not a screen, a voice, or a gesture. It is a human brain.
"Neuralink will begin high-volume production of its brain-computer interface devices and transition to an entirely automated surgical procedure in 2026."
— Elon Musk, CEO of Neuralink, January 2026
What's Actually Happening in BCIs Right Now
Let me give you a fast status report, because the pace of development here is easy to underestimate.
Neuralink's Blindsight implant is heading into its first human clinical trials in 2026. The concept is this: a camera captures images, sends the data wirelessly to an implant embedded in the brain's visual cortex, and the implant stimulates neurons to generate the perception of sight. Initially the resolution will be low, closer to a pixelated grid than natural vision. But Neuralink's new surgical robot inserts threads in just 1.5 seconds, down from 17 seconds with the older system, making the procedure dramatically faster and safer.
Meanwhile, Synchron, which takes the less invasive route of threading its Stentrode device through blood vessels rather than cutting directly into brain tissue, has now implanted devices in over 50 patients with paralysis. These people are browsing the internet, sending messages, and controlling computers using nothing but their thoughts. In August 2025, Synchron demonstrated a patient controlling an Apple Vision Pro purely by thought. Apple's response was not slow. The company announced a formal BCI Human Interface Device input protocol in May 2025, meaning BCI support is now a first-class citizen in Apple's ecosystem, sitting alongside touch, voice, and gaze.
The investment numbers tell you everything about where this is headed. Neuralink raised a $650 million Series E at a $9 billion valuation in June 2025. Synchron closed a $200 million Series D in November 2025. Merge Labs raised a $252 million seed round in January 2026, and Science Corporation secured a $230 million Series C in March 2026. That is not fringe biotech money. That is serious capital signaling a serious commercial timeline.
And yet, if you look at the product design community right now, the conversation is basically silent. We are debating button radius in design systems while the entire paradigm of what an "interface" is gets dismantled at the hardware level.
The design community needs to start treating BCIs as a product problem, not a medical one. That shift has to happen now, not in 2028.
The Three UX Problems Nobody Is Solving
I have spent the last few years designing AI-native interfaces, thinking through what happens when products need to interpret user intent rather than just respond to user action. I wrote about this on Medium when working through the mindset shifts product teams need for AI-first experiences. BCIs take every one of those challenges and multiply them by a factor I cannot fully calculate. Here are the three design problems that nobody in the BCI space is talking about seriously.
Problem 1: The Learning Curve Is Not a UI Problem. It Is a Neural Plasticity Problem.
Every UX designer is trained to think about onboarding. Reduce friction, give contextual help, build in progressive disclosure. With BCIs, this entire mental model collapses. When Neuralink's Blindsight delivers low-resolution visual signals to the visual cortex of someone who has been blind their entire life, the patient's brain does not just "see" a pixelated image. The brain has to learn what those signals mean. Neural plasticity, the brain's ability to rewire itself in response to new input, is not a fast process. It happens over weeks and months, not minutes and sessions.
This means the concept of a "user onboarding flow" doesn't apply here. The product is not teaching someone a new app. It is participating in the long-term rewiring of a human brain. The feedback loops, the mental models, the error recovery, all of it operates on a completely different timescale than anything we have designed before. When I think about how we handle first-run experiences even in complex enterprise SaaS, we are talking about getting someone productive in minutes or hours. BCI onboarding is measured in months of neurological calibration.
Nobody is designing for that timescale. There are no established frameworks, no playbooks, no design patterns. This is a blank canvas, and it requires a completely new discipline that does not exist yet.
Problem 2: What Is an Error State When the Input Is a Thought?
Here is a question I want every product designer to sit with for a moment: How do you design an undo function for a thought that was acted on?
In every interface I have ever designed, there is an implied contract between the user and the system. The user performs an action, the system responds, the user confirms or corrects. The action is discrete. It is a click, a tap, a voice command. There is a clear moment of intention and a clear moment of execution.
BCIs break this contract. Neural signals are continuous, messy, and highly contextual. The brain does not produce clean "I intend to do X" signals. It produces a constant stream of electrochemical activity that the BCI system has to interpret through statistical models. Early systems like Synchron's are trained to detect specific motor intentions, like imagining moving a hand, to trigger a cursor click. But as these systems get more sophisticated, the margin between a conscious intention and an involuntary neural pattern narrows dangerously. At what point did the user "mean" to do the thing the system acted on?
This is not a minor usability problem. It is a fundamental question about consent in interface design. And it becomes even more loaded when you consider that the same device interpreting your control intentions could theoretically intercept emotional states, stress patterns, or other neural data the user never consented to share. The European Union has already started framing cognitive data as a new category of protected personal information. The privacy design challenge for BCIs is unlike anything the tech industry has faced before.
Problem 3: The Interface Itself Has an Accessibility Problem
This one is subtle and important. BCIs are being developed primarily to restore function to people with disabilities: vision loss, paralysis, ALS. The accessibility mission is real and meaningful. But there is an accessibility problem built into the interface model itself that nobody wants to say out loud.
Not all brains produce the same signals with the same clarity. Neural signal variability between individuals is significant. Factors like age, existing neurological conditions, stress levels, sleep deprivation, and medication can all affect the quality of BCI signal interpretation. A system calibrated for one user may perform completely differently for another user with similar medical needs but different neural patterns. And unlike a touchscreen where you can redesign the tap target size or increase contrast ratios, you cannot redesign the human brain.
IEEE Spectrum noted that experts expect Neuralink's Blindsight to "almost certainly disappoint" when measured against Elon Musk's public claims, not because the science is fraudulent, but because individual neural variability makes universal performance guarantees impossible. The product will deliver. Just not uniformly, not predictably, and not at the level the marketing suggests.
That gap between the promise and the actual, personalized, variable experience is a design problem. Managing expectations, designing for the full range of real outcomes, communicating progress in a way that does not devastate patients who respond more slowly than average, all of this requires a level of UX empathy and ethical design thinking that goes far beyond what most product teams are currently trained to deliver.
What Apple's BCI Protocol Tells Us About the Real Timeline
Here is the signal that made me take this seriously at a product level. When Apple, a company that moves with extreme deliberateness on platform changes, bakes a BCI input protocol into its operating system, that is not a research experiment. That is a platform bet. Apple does not add hardware protocols for technology that is a decade out from relevance.
Synchron's patients are already controlling iPhones and the Apple Vision Pro through this protocol today. The infrastructure layer is being built right now. By 2028 to 2030, when commercial BCIs start becoming available for medical and eventually consumer use, the major operating systems will already have the input layer ready. App developers, enterprise software teams, and product organizations will need BCI-compatible interaction models from day one of commercial availability.
At reloadux, we have been writing about the shift from reactive to anticipatory interfaces, products that interpret intent rather than just respond to action. BCIs are the ultimate form of that shift. When the interface reads neural intent directly, the entire model of what an "app" does has to change. Every assumption about navigation hierarchy, discoverability, error recovery, and state management has to be rebuilt from scratch for a world where input is continuous, biological, and deeply personal.
The product teams that start thinking about this now will not be caught flat-footed when the hardware ships commercially. The ones that wait will spend years retrofitting experiences designed for fingers onto an input modality that works nothing like fingers.
Where This Leaves Product Designers in 2026
I do not think most product designers need to become neuroscientists. But I do think we need to start asking fundamentally different questions about the nature of interaction. Here is a short list of things product teams should be examining right now:
- How would your product's core workflow change if the input were thought-based? Identify the moments where your UI depends on precise, discrete user action, and ask whether those moments survive BCI input.
- What cognitive load does your current product place on users? BCIs introduce new cognitive overhead during the learning phase. Products that already max out working memory will be brutal to use with BCI input layered on top.
- What data does your product passively collect, and how would you handle neural signal data? Your privacy model needs to be BCI-ready before the hardware ships, not after.
- What does "accessibility" mean in your product when the interface is the user's nervous system? Universal design takes on a completely different meaning when you cannot standardize the input hardware itself.
- How would you measure success if the user's "session" spans months of neural adaptation? Your metrics framework needs new definitions for retention, engagement, and activation in a BCI context.
None of these are hypothetical questions for some distant future. Apple has the protocol live in production. Synchron has 50 plus active users. Neuralink is in human trials this year. The design work needs to start now, even if the commercial product ships in 2028.
I have spent eight years designing products across 42 products in 30 plus industry verticals. I have watched the industry shift from desktop to mobile, from mobile to voice, from voice to spatial computing. Each shift broke assumptions and rewarded teams who started thinking early. BCIs are not the next incremental shift. They are the most fundamental change to what an interface is since the invention of the graphical user interface itself. The brain is the new canvas. And almost nobody in the design community is picking up the brush yet.
What do you think? Are BCIs a near-term design problem, or is this still too far out for product teams to prioritize? Drop your thoughts in the comments below. I would love to hear how you are seeing this play out in your own work.
Sources:
1. Brain-Computer Interface 2026: Neuralink, Synchron, and Real Progress — 3zebras.com
2. State of BCI: 2026 Annual Industry Report — bciintel.com
3. Neuralink's Blindsight Implant Won't Deliver Natural Sight — IEEE Spectrum (spectrum.ieee.org)
4. Neuralink prepares first human trials of Blindsight implant — thenews.com.pk
5. Brain-Computer Interface Statistics and Facts 2026 — media.market.us
6. Brain-Computer Interfaces: User-Experience Design Principles — coderio.com
7. The intersection of UX and brain-computer interfaces — kryshiggins.com
8. Neuralink to scale brain implants, automate surgery — interestingengineering.com