Source: Unsplash
Imagine the screen goes away. No phone, no monitor, no AR headset. The display is now inside your skull, wired directly into your visual cortex. That's not science fiction anymore. In 2026, it's a clinical trial.
Neuralink's Blindsight implant began its first human trials this year, targeting people with complete blindness, including those blind from birth. A camera mounted on a pair of glasses captures the world and sends that signal through more than 1,000 ultra-thin electrode threads directly into the brain's visual cortex, bypassing the eyes and optic nerve entirely. The BCI market is now valued at $3.75 billion in 2026 and growing at a 16.77% CAGR toward $13.86 billion by 2035. But here's what almost no one in tech is talking about: this isn't primarily an engineering problem. It's the hardest user experience design problem ever attempted, and the design community is almost entirely absent from the conversation.
"The Blindsight device will enable even those who have lost both eyes and their optic nerve to see. Provided the visual cortex is intact, it will even enable those who have been blind from birth to see for the first time."
— Elon Musk, Neuralink Founder, 2025
What Patients Will Actually See (It's Not What You Think)
Neuralink's own communications are refreshingly candid about early resolution. Their words: "early Nintendo" or "Atari-level graphics." What patients will perceive are phosphenes, which are discrete points of artificial light that the brain gradually learns to interpret as meaningful spatial information. Not a coherent image. Not a video feed. Dots. Bright and dark dots arranged in a grid that the visual cortex slowly, over weeks of training, begins to translate into something resembling edges, shapes, and movement.
The current implant uses over 1,000 electrodes. Neuralink's roadmap calls for moving to 3,000 electrodes in a next-generation device later in 2026, tripling the effective resolution. For context, the human retina processes signals from roughly 125 million photoreceptors. Even at 3,000 electrode resolution, you're working with a fraction of a fraction of natural vision.
And yet. For someone who has never seen a single photon, detecting that a door is to the left or that someone is standing two feet away? That changes everything. The question I keep coming back to as a product designer is: who is responsible for the quality of that perceptual experience? Who owns the UX of what a brain learns to see?
The Unprecedented UX Problem Nobody Is Talking About
In every interface I've designed over the past eight years, whether fintech dashboards, healthcare apps, or AI-native enterprise tools, there has always been a screen between the user and the product. The screen is the contract. You see a button, you press it, something happens. The feedback is visual, immediate, and contained within a defined display boundary.
Blindsight eliminates that contract entirely.
When the "display" is the visual cortex, the fundamental questions of UX design become:
- What is a "pixel" in neural terms? A phosphene generated by electrode stimulation isn't a pixel. It has varying brightness, persistence, fade time, and spatial bleed depending on the individual's neural anatomy. No two patients will have the same perceptual output from the same signal.
- How do you test usability? Traditional usability testing assumes a shared external reference. A researcher can watch a user navigate a screen. They cannot watch what a user perceives when their brain is generating images from electrode arrays.
- What does "error" mean? If a face is processed incorrectly and the patient perceives a distorted shape, is that a signal processing error, a calibration error, a cognitive adaptation problem, or a fundamental hardware limitation? The debugging loop is entirely invisible.
- Who is the user? The brain learns. Neuralink's trials involve training periods where patients literally teach their visual cortex to interpret new artificial signals. The "user" is not static. They are changing, adapting, and building new neural pathways in response to the product. No existing UX framework has a model for a user whose perceptual system is actively being rewritten by the product they're using.
- How do you iterate? Traditional product iteration: ship, observe, adjust. Neural interface iteration: stimulate, wait for cortical adaptation (weeks), observe behavioral outcomes, adjust signal parameters, wait again. The feedback loop is measured in months, not sprints.
The design problem here isn't how to display information. It's how to teach a brain what information even looks like when it has never seen any before.
The "BCI Designer" Role Is Emerging and Product Teams Aren't Ready
I wrote about this earlier this year in my piece on AI-native experiences on Medium: the biggest mistake product teams make is treating new interface paradigms as skin-deep changes. They bolt conversational UI on top of old workflows. They slap an AI assistant onto a legacy dashboard. They assume the interaction model from one era just transfers to the next.
BCI design doesn't allow that shortcut. A new category of designer is genuinely emerging. Research from the Interaction Design Foundation describes it plainly: BCI designers must work without intermediate input tools like hands, voice, or screens. They are designing relationships between human cognition and computation, at the level of neural signals, with no familiar metaphors to fall back on.
The Blindsight case makes this concrete. The "onboarding" for a blind patient receiving their first visual signals is weeks of training. The first "feature" they experience is a set of phosphene dots. The "interface improvement" from v1 to v2 is moving from 1,000 to 3,000 electrodes. And the "success metric" is whether, over time, a person's brain successfully learns to navigate space using a signal that has never existed in human evolution before.
That's a product problem. It's also a design problem. And the discipline is not close to having the frameworks to handle it.
The Counterpoint: When a Company Writes Directly to Your Brain
There's a part of this that should make every designer uncomfortable, and I think it's worth saying plainly.
Every other interface paradigm in history has been about presenting information to a user who then decides what to do with it. Blindsight doesn't present information to the brain. It writes perception directly into the visual cortex. The distinction is enormous.
As of early 2026, 21 people have enrolled as Neuralink "Neuralnauts" across all Telepathy and Blindsight trials combined. Neuralink reports zero serious device-related adverse events so far. The trials are happening in the UAE in collaboration with Cleveland Clinic Abu Dhabi, after the FDA granted Breakthrough Device Designation in September 2024.
Those are encouraging numbers for safety. But the design ethics questions are only beginning.
What happens when the signal quality degrades and the user's brain has already adapted to a specific perceptual baseline? What happens if Neuralink discontinues the product? What happens when the camera captures sensitive environments, like a courthouse or a private home, and that visual data is also being processed through Neuralink's systems? Who owns the perception data? The patient? Neuralink? The FDA?
We covered the ethics dimension of AI-native interfaces at reloadux earlier this year when writing about AI readiness frameworks. The pattern is consistent: the infrastructure races ahead, the ethical and design frameworks chase it by years. With BCIs, the lag between technical capability and design/ethics governance could be measured in decades, and the stakes are someone's lived perceptual reality.
What Product Designers Should Actually Do with This Information Right Now
I'm not suggesting you pivot to neural interface design tomorrow. That's not realistic for 99% of product teams. But here's what I think this moment demands from anyone serious about product and UX work:
Follow the constraint, not the feature. Blindsight is forcing designers to strip away every assumption about interfaces. No screen. No touch. No voice. No standardized output. What's left? Pure cognitive intent and perceptual feedback. That constraint is philosophically useful. It forces a clarity about what "interaction" actually means at its core.
Study adaptive interfaces now. The Blindsight onboarding is the most extreme version of adaptive UX ever attempted. The interface literally adapts by rewiring the user's perception. Every product team working on AI-native, personalized, or context-aware interfaces is on a dial that points in this direction. Understanding the far end of that dial will sharpen your thinking about the dial you're actually working on.
Get literate in neuroethics. The questions Blindsight raises about consent, data ownership, and perceptual manipulation are coming to mainstream consumer products faster than most teams realize. The conversation about what companies are "allowed" to write into user perception is already beginning in regulatory circles in the EU and US. Designers who understand the philosophical foundations of these debates will be the ones shaping how the industry responds.
The BCI market is real and it's growing fast. $3.75 billion in 2026. $13.86 billion by 2035. Over $1.6 billion in disclosed VC and institutional funding in 2025 to 2026 alone. Neuralink's last round valued the company at $9 billion. The capital is flowing. The design talent pipeline is not keeping up. That's a gap worth paying attention to if you're thinking about where to position your career over the next decade.
The Hardest Design Brief Ever Written
Someone at Neuralink or at one of the clinical research teams is right now working on how to make those phosphene dots as useful as possible for a human being who has never processed visual information in their life. They are designing without a screen, without a standard user, without established success metrics, and without the ability to observe the output directly.
That is either the most terrifying design brief in history or the most interesting one ever written. Probably both.
The thing I keep thinking about is this: the history of design is a history of reducing the distance between human intent and system response. From physical buttons to touchscreens to voice to gesture to gaze-tracking. Each step, the interface gets thinner and more invisible. Blindsight is the logical endpoint of that trajectory. The interface doesn't just become invisible. It becomes the inside of your head.
That's where we're going. The design community needs to catch up, fast.
What do you think? Is this purely a neuroscience and engineering problem, or is there a real design discipline emerging here? Are product teams paying enough attention to what BCI development means for how we think about interfaces? Drop your thoughts in the comments below. I'd genuinely love to hear how practitioners are thinking about this.
Sources:
1. Neuralink Blindsight — tesorb.com/neuralink-blindsight-implant-2026
2. Neuralink Blindsight Human Trials — neurapod.com/blog/neuralink-blindsight-human-trials-what-to-expect
3. Elon Musk Blindsight Update — mobihealthnews.com/news/elon-musk-provides-update-blindsight-implant-trials
4. BCI Market Statistics 2026 — media.market.us/brain-computer-interface-statistics
5. BCI Market Growth to $13.86 Billion — precedenceresearch.com/brain-computer-interface-market
6. State of BCI 2026 Report — bciintel.com/state-of-bci-2026
7. BCI UX Design Principles — coderio.com/blog/biz-tech/brain-computer-ux-design-principles
8. Brain Computer Interface Design Principles — IxDF — ixdf.org/literature/topics/human-robot-interaction