On a recent rainy afternoon, I wander around an industrial area of Brooklyn with an unusual mission: finding a brain implant startup that doesn’t want to be found.
Synchron does not publicly disclose the address of its headquarters and sent me one that turned out to be imprecise. I spend 15 minutes wandering past smokestacks and a mysterious steel door—with 10 locks and as many handles—before stumbling upon a Synchron employee who guides me and my colleague to the office. On the elevator ride up, she tells us that the address misdirect was intentional, but doesn’t explain why.
Through the office doors, a golden retriever named Murphy greets us. There is a wall of gold-plated patent plaques boasting titles like “control using a neural signal” and “stimulating tissue.” Synchron makes the Stentrode, a brain-computer interface (BCI) implanted in 10 people since 2019. That’s three more than have an implant from Elon Musk’s Neuralink.
We wait on a leather loveseat next to a large, green-lit glass case that houses a model brain with the Stentrode on top. This quarter-sized piece of metal mesh enables the wearer to control computers, phones, and other devices with only their thoughts.
CEO Tom Oxley says Synchron is competing with eight other firms to develop such devices. The company differentiates itself with its minimally invasive implantation process: The Stentrode does not require drilling into the skull, like the Neuralink N1. Instead, it inserts into a vein above the motor cortex, so it’s near the brain but not directly on it.
The Synchron Stentrode is a mesh-like sensor that inserts into a vein above the brain, connected to an electronics pack near the column bone (Credit: Joseph Maldonado/PCMag)
The biggest technical challenge with this approach is achieving the same performance as an N1 since the Stentrode is further from the brain’s electrical signals. But Synchron doesn’t seem concerned. “We think Neuralink is over-engineering,” says Kurt Haagstrom, Synchron’s COO. “I have no doubt this will work.”
By the way, Reuters reported in 2022 that Musk had reached out to Oxley to discuss a potential deal. (Neuralink did not respond to multiple requests for comment.)
Brain implants are a burgeoning field that is little understood but is flush with cash. Neuralink has more than $1 billion in funding and raised a fresh round of $650 million in June. Synchron has a $385 million valuation and counts Bill Gates and Jeff Bezos as investors. The Stentrode is now an official “breakthrough device,” a Food and Drug Administration (FDA) designation that could help fast-track it to market for people with severe disabilities.
At the same time, these implants are among the most uncomfortable topics in tech, equal parts miracle and dystopic Sci-Fi fodder. Once a month, Oxley receives an email from an unwell person convinced Synchron has already implanted the device in their head. They demand that the company take it out, sometimes aggressively, he says. Synchron doesn’t want them to show up at its offices in person—thus the secrecy surrounding the company’s address.
Brain Implants for All by 2040?
When Oxley is ready, an employee takes us into his corner office, a bright room with cases full of books, an easel with handwritten notes, and a Basquiat-style painting above his desk. A 44-year-old with a polite, controlled demeanor, Oxley answers questions methodically and in detail, easily flipping between medical references and more imaginative, philosophical tangents. He’s thought deeply about these issues, which feels appropriate for someone implanting hardware in people’s heads.
Oxley says he’s often imagined a world in which everyone has a brain implant, not just those with severe disabilities. He has predicted that by 2028, Stentrode will receive pre-market approval—a license to begin advertising and selling it—from the FDA. By the 2040s, Oxley expects the Stentrode will be available for all, as long as they have an extra $40,000 to $50,000 lying around. (In this scenario, insurance won’t cover the technology since it would be elective.)
Synchron CEO Tom Oxley (Credit: Joseph Maldonado/PCMag)
“Mass market is not what we’re building for right now,” Oxley says. “But if it’s safe, not that expensive, removable, doesn’t deform your body, and lets you control technology better than what you can do with your body, I think there will start to be early adopters. I know it sounds weird.”
Recent developments make this more likely: Two of the biggest companies in the world, Apple and Nvidia, are developing technology for brain implants. This year, Apple enabled control of its devices through BCIs, hooking them up directly to iPhones, iPads, and Macs via Bluetooth. Synchron will become the first to offer the capability to patients, though Apple designed it to work with all brain implants.
Meanwhile, Synchron is providing data harvested from its implants to see if Nvidia can train large language models on “digitized brain waves,” as Nvidia CEO Jensen Huang put it in his speech at the company’s 2024 GTC AI conference.
If giving powerful corporations access to your brain through an implant gives you pause, you’re not alone. It’s a legitimate concern that nefarious companies could use them for harm, Oxley says.
So, like a real-world episode of Black Mirror, I offer. “Yes,” he says. “I’m thinking about autonomy, privacy, and discrimination, which are the three things I think could be problematic with this technology. It’s not a today problem, but the way that we talk about it in the starting period matters.”
How Rodney Gorham Flirts Using His Stentrode
Today, BCIs are still so limited in scope that the average person would have little reason to get one. But for patients with severe paralysis, they are life-changing.
Rodney Gorham did not hesitate to get the Stentrode in 2020, and he’s now had it for more than 1,600 days—a world record for a fully implantable brain device. Once a world traveler and IT worker at IBM, he’s now 65, barely able to move and unable to speak due to the neurodegenerative condition ALS.
“His body’s disintegrating, but his brain is staying the same,” his wife, Caroline, says on a video call from their home in Melbourne, Australia. “He’s exactly the same person, just his shell is not working. There are thousands and thousands of people across the world in the same situation.”
Rodney Gorham uses his Stentrode to operate a Microsoft tablet via Bluetooth (Credit: Synchron)
Rodney, sitting next to Caroline, uses his BCI to relay “hello” to me on WhatsApp. “Hi, friend!” I reply. “This is so cool!” It might be the most futuristic thing I’ve ever experienced, though I am sure Rodney was thinking much more than a simple “hello.”
A wire connects Rodney’s Stentrode to an electronics pack smaller than a deck of cards, which Synchron embedded near his collarbone via a shallow incision. Rodney “types” through a combination of eye tracking and his BCI, which is connected via Bluetooth to a Microsoft tablet in his living room. He can send texts, write an email, browse YouTube and the rest of the internet, and control the TV.
“I stopped him from being able to control the lights because he flickers them if he wants me to pay attention to him,” Caroline says, eliciting a flirtatious grin from Rodney. “He can still express his personality. He can be Rodney.”
A Stentrode sits inside a model brain where it would on a human patient (Credit: Joseph Maldonado/PCMag)
Rodney moves the cursor with the eye tracker and presses “select” through his BCI. He does so by thinking, “move my foot.” Synchron has programmed that thought pattern like a hotkey on a computer. “Move my foot” now means “select,” essentially.
Eye tracking is exhausting for the patient, so Synchron is exploring cursor control through the implant. They’ve given Rodney exercises to test the process, which he enjoys. The Synchron researchers ask him to think “move left hand left” and “move right hand right” and then the cursor moves in the corresponding direction. “Like any physical skill, you have to repeat it,” Caroline says. The tech is improving, but not yet perfect.
Neuralink patients can already use the company’s N1 implant to control the cursor with their thoughts. This helps Neuralink’s first patient, Noland Arbaugh, a quadriplegic from Yuma, Ariz., use his implant for longer periods of time without exhaustion. “I’m literally just thinking, ‘I want the cursor to go over this part of the keyboard and select the key,'” he tells me.
Noland Arbaugh, Neuralink’s first human trial (Credit: Noland Arbaugh)
But there are limitations. “Cursor control is at 90% of where I want it to be,” Arbaugh says. He can use his implant for many tasks, including “a lot of emails, website editing, writing stuff up, researching, banking, housekeeping—just being an adult trying to find some way to make it through life.”
The Apple integration should open more doors for BCI patients. Gorham was part of testing it, although technically, he and other BCI patients could already connect to Apple devices through an informal, nonstandard connection. The new experience creates a standard for all BCIs, as if they were any other device, like a keyboard or mouse. “So now when you connect, [the device] immediately recognizes that profile and flips into brain control mode,” Oxley says, “It’s like, ‘I know it’s a brain.'”
Arbaugh configures his implant on the Neuralink app on his Macbook (Credit: Noland Arbaugh)
Apple built on an existing accessibility feature called Switch Control, which allows disabled people to operate the company’s devices with external joysticks or hardware that’s more comfortable for them. Apple’s goal is to help people with life-changing disabilities; sources tell me the company has no ambitions to make BCIs and is not working on mass-market adoption of the tech.
Get Our Best Stories!
Your Daily Dose of Our Top Tech News
By clicking Sign Me Up, you confirm you are 16+ and agree to our Terms of Use and Privacy Policy.
Thanks for signing up!
Your subscription has been confirmed. Keep an eye on your inbox!
Like Going to ‘Space or the Deep Sea’
From the tech industry perspective, it’s surprisingly easy to imagine a world where BCIs could one day be the new normal. So many of us are already on our phones and computers all day. What if we could do some of those tasks without hunching over a device and getting carpal tunnel?
Smart glasses could also be an early predecessor to implants. Meta, Google, and likely Apple are developing specs with AI assistants embedded in them that can take actions without requiring us to put our hands on a screen or mouse. Is it possible to move them from the face to the brain in the long term?
Oxley says that may be too much power for a tech company, though he expects the “free market” will correct for this. If people aren’t comfortable with BCIs, they won’t get them. In the meantime, we should heed warnings from “algorithms like TikTok that can [make us addicted] with devices that are outside of the brain,” he says. “If you’re inside the brain at a subconscious level, and have algorithms preying on human vulnerabilities, then you have a recipe for human subversion and subjugation.”
Synchron COO Kurt Haagstrom demonstrates the process of inserting the Stentrode on a replica (Credit: Joseph Maldonado/PCMag)
If BCIs are ever to become mass market, the user would need a “kill switch,” he says. All of these requirements and regulations would likely come from the FDA. And to happen, public trust in the FDA would also have to be extremely high, Oxley points out. A 2022 survey found that just 27% of respondents trusted the FDA “a great deal,” and 23% did not trust the FDA “very much” or “at all.” According to a 2024 survey, the agency is still working to rebuild trust after the COVID-19 pandemic.
At the same time, there are compelling reasons someone without a disability may want a brain implant in the future, assuming the technology keeps progressing. They could enable next-level multitasking, such as typing on a keyboard while moving the cursor with their thoughts. The implantee could tell the computer to correct the spelling on the last word, while at the same time continuing to type the next ones without having to go back. Multiple actions can occur in tandem, limited only by how many of our thoughts the hardware can understand and convert to action.
How we interact with products and services could also change. I pose a hypothetical example to Haagstrom (definitely not based on personal experience): Let’s say you’ve come home after having a few drinks at a party. When you start thinking “I’m hungry,” is it possible that the implant could detect your blood alcohol level and automatically order you a burrito? Yes, he says, adding that if you had the foresight to input a list of post-party foods, your delivery app would know what kind of burrito—steak or chicken?—to get you once the implant gives it the go-ahead.
For software developers, it’s a question of “How do you turn a 2D environment into 3D?” Haagstrom says. “How do you design differently if someone’s using a BCI? I look at this like going somewhere no one’s gone before, like space or the deep sea.”
Elon Musk or Tom Oxley: Whose Implant Would You Put in Your Brain?
The infrastructure for this uncertain future is quietly being built by big thinkers such as Musk and Oxley. Like Musk, Oxley is drawn to some of the hardest problems imaginable. “It was my father”—Alan Oxley, an Australian diplomat and trade expert—”pushing me toward asking questions about ‘What are things that will be impactful over our lifetime?'” Oxley says. “‘What are things we don’t understand? What are the biggest questions?'” Unlike Musk, Oxley comes has a medical background, having practiced and researched neurosurgery for over a decade.
(Credit: Joseph Maldonado/PCMag)
Growing up in Canberra, Australia, Oxley found himself intellectually drawn to neurological questions, largely because so many of them remain unanswered. “I became very romantically obsessed with the brain,” he says. He traveled the world after high school, and along the way sent “hundreds of cold emails” to professors and experts about “an idea swirling around” in his head regarding brain implants.
Recommended by Our Editors
One night, while staying in a “moldy backpack hostel” in New York, Oxley received an email response from Geoffrey Ling, a neurologist and colonel in the US Army working with the Defense Advanced Research Projects Agency (DARPA). Finally, someone was interested in his idea.
Ling invited Oxley, then 29, to Walter Reed in 2010, where they discussed Oxley’s idea for a less-invasive BCI that would slip into a major blood vessel above the brain’s motor cortex. The thought was that because it didn’t require brain surgery, more patients could be open to it. Also, the body would be more likely to accept the foreign object long-term. Healing agents inside blood vessels would calm inflammation.
After medical school, Oxley practiced medicine at Mount Sinai in New York and researched neurological interventions, ultimately leading him to co-found Synchron in 2012. The implant’s design has evolved since then, but the concept has remained the same.
I ask Oxley if he would get one of his company’s implants. “I would,” he says after a moment’s hesitation. “I’m intrigued.”
Despite the fact that Synchron takes an “in-vein” (as opposed to “in-brain”) approach, there are still risks. “Going in the blood vessel doesn’t mean there are no problems,” says Dr. Bradley Greger, an associate professor at the Arizona State University School of Biological and Health Systems Engineering. He’s an expert in this field and works in the same building as the Neuralink trial. “The blood vessels supply the blood to the brain, and if those are compromised, that’s what we call a stroke. Now, it’s rare and unlikely, but it does happen.”
Oh, and while putting Synchron’s device in might only take a 20-minute procedure, it’s not designed to be taken out. Getting the Stentrode is an act of faith.
I ask Oxley if he would get a Stentrode. “I would,” he says after a moment’s hesitation. “I’m intrigued. I think there are going to be some people who are early adopters who want to explore how they can use their brain to do things. I’m an early adopter, so would I get it one day? Quite possibly.” Musk has said he plans to get a Neuralink implant when the technology is ready.
Our Brains Are an Untapped Motherlode of Data
There may be no bigger privacy issue than forking over your brain data, potentially with all your thoughts and desires, to a corporation. Yet the future of the field relies on our willingness to do that and implant companies’ ability to understand our brains’ electrical signals. The concept makes collecting data about us over the internet sound like child’s play.
Brain data is one big soup of infinite combinations that need to be decoded, unlike a keyboard and mouse, which have a fixed set of inputs. This process is an “art,” as Haagstrom describes it, and a formidable data science challenge. It’s like creating a new, shared language between the implant wearer’s thoughts, the implant, and Apple devices.
I decided I had to know what brain data looks like, as long as there’s a possibility it may be extracted from my skull and sent to a server during my lifetime. I ask Synchron in as many ways as I can to see the data without success, and then realize my request is like begging for the code to their safe. It’s their most valuable trade secret, and, outside of the company, for the FDA’s eyes only.
Brain wave pattern of a patient repeating the word “yes,” and then hearing “good job” (Credit: PCMag Composite; NIH Public Author Manuscript)
To explain the basic concept, ASU’s Greger points me to a study with an example of how BCI researchers interpret brain waves. They sat with a patient and asked them to repeat the word “yes” over and over again—the part in the red box in the photo above. There’s a consistent pattern, with signals going up and down. Then, the researchers told them “good job” (where the arrow is). The electrical signals look different when the patient hears that positive affirmation. Over time, the researchers are able to map these brain signals to words and actions, but the progress can be slow, Greger says.
That’s the main reason Greger is not as convinced as Oxley that mass-market BCIs are in the foreseeable future. “Everybody hypes it up,” he says. “They’re all trying to make it into a commercial business. I love this technology, but some companies are trying to promote themselves, as they should, and they kind of obfuscate the risks that are involved.” He adds, “Will people accept that risk with some potentially pretty serious consequences for the benefit of being able to interface with Apple virtual reality device?”
Other academics I spoke to had mixed opinions on mass-market availability. “It is entirely possible that the less invasive types of BCIs will be more available to more people,” Dr. Gregory B. Gogan from Duke University says, “but I still think we’re a ways away.”
AI may help speed up that process by recognizing patterns and adding a predictive element. For instance, if the wearer looks at a Roomba while sitting in a dusty room, a message may pop up on a tablet and ask if they want to turn the vacuum on remotely. They would think, “yes” or “no,” to silently respond.
In this scenario, “the BCI is making a prediction of what you want to do and how to engage those systems, assuming you want to,” Oxley says. “You’ll represent your will through the BCI. It will offer what you want to do, and then you have to approve it. That’s important.”
After a Peek at the Secret Lab, I’m Back on the Subway With a New Perspective
On our way out, we pet Murphy one last time and ask if we can see the lab where Synchron does undisclosed on-site testing. (In late 2022, Reuters reported that Synchron had killed about 80 sheep as part of its research, although Synchron declined, and continues to decline, to comment. That year, Neuralink faced a federal probe for allegedly killing over 1,500 sheep, pigs, and monkeys in its testing. Musk later claimed Neuralink had never caused the death of a monkey.)
Murphy, the office dog, sends us off (Credit: Joseph Maldonado/PCMag)
Haagstrom is reluctant, but gives me the briefest of glances inside. In the two seconds I have, I see fluorescent lights and workstations, along with a boxy, photobooth-type structure in the back. The room looks more like a woodshop than a sterile laboratory, not particularly medical.
Afterward, my colleague and I step outside in a bit of a daze as our innocent, plain old brains try to piece together where society is going. Riding the subway back, we take in our fellow passengers with new eyes. “Look at all this thought data,” my coworker says to me. Our fellow riders are all staring, necks bent, at their phones, unaware that one day they may swap those screens for implantable devices—and, who knows, maybe they’ll even prefer it.
About Emily Forlini
Senior Reporter
