Investors are racing to fund developments in brain implants, a burgeoning technology that can help people who are paralyzed communicate more naturally. Writing about this field can be challenging, because one leading company — Elon Musk’s Neuralink — has historically been closed to the media.
Over the past seven years or so, Wired Staff Writer Emily Mullin has written several stories about brain-computer interfaces, technology that collects and interprets signals directly from the brains of people unable to communicate.
Mullin’s recent piece, “The Race to Put Brain Implants in People Is Heating Up,” provides an interesting update on what’s happening in clinical trials of the devices at some of the leading tech companies, including Neuralink. It’s a thread she has followed while at MIT Technology Review, Medium’s OneZero and the Pittsburgh Post-Gazette, plus some stints as a freelance writer.
In this “How I Did It,” Mullin explains how she became interested in this area of biotechnology and how she keeps abreast of the latest news.
Responses in this interview have been lightly edited for clarity and brevity.
How did you get interested in this topic?
I started covering brain-computer interfaces in 2017, when I was at MIT Technology Review. If you remember the movie, “The Diving Bell and the Butterfly,” the subject of the film (Jean-Dominique Bauby, editor of French Elle magazine) had a massive stroke that left him paralyzed. He dictated his entire memoir by blinking his eye.
But what if a paralyzed person could communicate more naturally? That’s the promise of brain-computer interfaces. I thought the idea of this direct connection from the brain to a computer was amazing. It’s an easy concept to understand, but I wanted to learn how it worked. It seemed like science fiction.
When I got into the reporting, I found out researchers have been trying to develop these as far back as the ‘60s and ‘70s. There was always this idea of “How can we connect a human’s brain directly to a device or a computer or a robot?”
At Tech Review, we had our annual EmTech MIT event (an emerging technology conference), and I remember there was a brain-computer interface scientist there who was talking about the evolution of these systems. Now there are several startups and companies working on them to commercialize them, but for so long, they were confined to the academic space.
When I actually saw a picture of a setup of these systems, they’re really clunky and require patients to have this really heavy pedestal on their head, and a thick cable that runs from the pedestal to the external device. They have to use this system in a research lab, although a handful of patients have been able to use them at home.
They’re complicated to set up, and patients have to go through weeks of learning how to use them. They have these brief few hours a week when they’re doing the testing and they are able to gain back some of the freedom that they had lost, and then they go home and lose that. That was really powerful.
What explains the momentum happening now?
Elon Musk is a character, but [he] has really catalyzed the brain-computer interface field. Every neurotech company I’ve talked to that has been founded since Neuralink in 2016 has said to me, “I don’t think we would have been able to raise the money that we have if it weren’t for Elon Musk.” One of my sources told me Musk is the kind of person who recognizes big trends and inflection points and pumps a ton of money into that.
The technology is sort of at a tipping point. The science and the technology have matured over the past several decades and now it’s a matter of moving from these demonstrations that the technology works to an actual product on the market. You need that investment, you need big figures, you need companies. I think that’s why there’s now so much activity and so much excitement.
How do you keep pace with what’s happening in the field?
I have a lot of Google Alerts. I also have alerts on bioRxiv.org (a website with preprints of journal articles). I’m on the social media platform X (formerly known as Twitter) looking for neurotech companies. I’m talking to neurotech companies a lot. Just keeping abreast of the literature.
Often when you’re talking to a source — even sometimes when there’s not a clear or immediate story — you find out something about another company or another researcher that you might not have known about.
From your reporting, it appears that Neuralink has not been open in terms of responding to emails or giving interviews. How do you fill out the stories if you can’t get interviews with them directly?
Musk in general is not very open to the media, and that extends to Neuralink. They don’t have a communications or PR department, because why do you need one when you have a billionaire running your company who likes to post on social media?
Neuralink’s live-streamed demonstrations have been a window into what they’re working on. Watching those over the years and reading their blog posts has been interesting to try to figure out what their strategy is and where their technology is. Their news comes out on either their YouTube channel or via Musk himself on X, formerly Twitter.
It’s common for biotech companies to not publish peer-reviewed research and make announcements via press releases instead, but Neuralink doesn’t even issue press releases. So you have to fill in the gaps by talking to other sources in the field.
It’s helpful to form your own peer review committee in a way, reach out to three or four experts and ask them, “OK, is this realistic?” or “How does this technology compare to what other researchers are doing?” just to get a gut check.
Would those sources be from other companies, or people in academia?
It’s good to have a mix of both. Academics are great for talking about the state of science and technology. Talking to companies gives you the perspective of what it’s going to take to bring a product or a device to the market.
It’s one thing to have a device be successful in a clinical trial, but then you look at how many companies now are working on this technology, and realistically, are all of them going to make it to market? Are they all going to perform similarly? Will some of them be better at certain tasks than others? We don’t know yet.
How do you balance the hype and enthusiasm of these technologies with fair reporting?
On one hand, this technology is really exciting, because you’re talking about restoring communication to patients paralyzed from illness or an accident, giving them back some amount of freedom and allowing them to do things that they’ve maybe never thought they would do again. You have to keep in mind that there are certain technological advancements with Neuralink’s device, but fundamentally, what Neuralink is doing isn’t necessarily new.
Before Neuralink came on the scene, researchers were toiling away in academic labs for years working on these. That’s what I always try to remind readers when I write about Neuralink and advances in the field. Musk has catalyzed the field.
He’s trying to bring innovation to brain-computer interfaces, but these aren’t new. He didn’t invent them. There’s a balance of “this is new and exciting,” but he’s also standing on the shoulders of researchers and scientists who came before him that did a lot of the fundamental work already.
Over the course of your reporting, what have you learned about these devices that you find the most interesting?
There are two ways to approach brain-computer interfaces. There’s the approach of going into the brain with an implant like Neuralink’s that actually goes into the tissue and records signals from hundreds or even thousands of individual neurons.
Then there’s the second approach, which is an implant that actually lays on top of the brain’s surface, and is collecting more population-wide data. When I first started reporting on these devices, I thought, well, of course, you would have to record from lots of neurons at once in order to get the assistant to perform certain actions.
But what we’ve seen just in the past couple of years is that you might be able to do certain tasks by just reading that surface-level data. And we might not have to put an implant into the brain, which is less invasive for the patient and probably is going to last longer. It’s going to be interesting to see how these systems evolve.
