Q&A: How DARPA hacked a science writer’s brain and turned her into a sharpshooter

By Sara Goudarzi | May 29, 2023

Brain with electrical currentsCredit: Inna/Adobe

In 2007, science writer and editor Sally Adee attended a technology conference for DARPA—The Defense Advanced Research Projects Agency—when she learned about a brain surface electrical stimulation method that allegedly cut in half the time it takes a person to go from novice to expert in sharpshooting. Known as trans cranial direct current stimulation (tDCS), the brain stimulation technique was also presented as a way to potentially speed up language learning and mathematical aptitude.

Adee was struck by the idea of surface brain stimulation and for several years worked on convincing the Department of Defense agency to grant her permission to try it out for herself. In 2011, she flew from London to California to participate in an army training simulation. With a device on her head delivering electrical currents, Adee went from a frustrated shooter to a skilled sniper.

At the time, the idea of transcranial direct current stimulation had been circulating for about a decade, and scientists wanted to know whether the concerted firing of neurons, by strengthening their connection through currents, would affect a person’s ability to learn. It turned out it did, at least for Adee. But did it work for others, and how long would the effects last?

To better understand the experience, Adee set out to learn more about the electrical currents that run through all living organisms—the electrical biome, or electrome—and details the findings in her book We Are Electric. The Bulletin’s Sara Goudarzi spoke to Adee about her encounter with transcranial direct current stimulation and how the technique can be used and potentially misused. The resulting discussion has been edited for length and clarity.

Goudarzi: Can you talk about the experience of trying transcranial direct current stimulation at the DARPA facility?

Adee: The research that I was interested in ostensibly improved focus to let people enter sort of a flow state, which amplifies learning ability. Just so you know, what I did was just a Gonzo stunt. Basically, it was just “here’s how it feels to try this,” and it so happened to work really well for me.

I was very frustrated and had made a lot of bad decisions in the run up to the simulation. After chasing the story for four years, I had done the dumbest thing imaginable: I booked myself on an 11-hour flight from London to California and not given myself even a day to recover, which was so stupid. Then I tried to save New Scientist [magazine where Adee was an editor] $100 for the hotel room. So, I stayed with a friend who lived in the mountains and got altitude sickness. I hadn’t slept, had jetlag, and was running on this wild amount of coffee. When I showed up to the simulation, I thought what am I doing? how can I report a story in this state? I was so freaked out and they [the Darpa contractors] said “okay, we’re going to put you into this simulation.” Absolutely wired from the coffee, exhausted, and my circadian rhythm messed up, I did such a bad job. It’s not like I’ve ever been any good at first-person shooter games, but I was somehow even worse.

They kept putting me through these training exercises and I kept failing and being really overwhelmed. Everything was happening all at once—the fake people were running at me, and I didn’t know what order to dispatch them in. It was incredibly frustrating, and I had this entire script of self-recrimination running through my head the entire time. It was just vicious. Then I started just having really negative thoughts. I was thinking this is all garbage, my whole job is garbage—just catastrophizing.

Then the tech came in and switched the current on, and it was like somebody had put noise cancelling headphones on me. I couldn’t hear any of that [negative thoughts] anymore. I did the training again and thought this is completely fine; I know exactly what I’m doing; what’s all the fuss? I was actually really enjoying it. After three minutes, the tech came in (normally, the runs were taking about 20 minutes to get through), and I had dispatched all the bad guys and was waiting the next wave, and said, “it’s done.” I was really surprised—actually, really annoyed—because I was thinking so now, I get it right for the first time and you stop it early. I said, “but I’ve only been here for a couple of minutes.” She said “no,” and I looked at the clock and was shocked to find that 20 minutes had passed. The researcher [in charge] said it was a common response that time starts to fly quite quickly.

Many years later, I was talking to another researcher in this field who studies different phenotypes of depression and the effects of tDCS on them. She said that for people who have the angry little angry waspy voices in their head, the particular placement of the electrode that I had for that part of the cortex found great success. She said those people have this overactive angry cortex guy just sitting there. When you apply the electric field, it shuts him right up and those people find that their depressive symptoms are alleviated.

I was thinking of conceptualizing this little angry voice as the guy from the “Inside Out” Disney movie, whose head blows up as a volcano. Those were the sorts of angry little voices that I hear frequently as my background elevator music. I never thought of that as being related to being depressed, but I certainly sometimes have a pretty hard time with that. The experience of having that turned off was really interesting. It made me think about what my inner soundscape is daily, and afterwards I really started investigating how to manage that a little bit more deliberately.

Goudarzi: Did the inner negative voices go away immediately or after the currents were flowing for some time?

Adee: I didn’t really notice it go away initially. It wasn’t dramatic. It’s just that when I looked back, I noticed it. After, when I was sort of going about my day, I felt this very deep sense of calm. And it wasn’t only for the duration of the treatment; it was a very slow come down for about the next three days. I just wasn’t this sort of white-knuckling, anxious person—that was really dramatic.

Goudarzi: Can you talk about the science behind your experience—what was happening when they ran that current through your head, and what was it that led to that focus?

Adee: That is very much a matter of continued research and speculation. During learning, there’s this axiom, “neurons that fire together, wire together,” and it’s why if you keep practicing the same move in say tennis or sharpshooting over and over again, those brain circuits learn to function in concert. At first you have to do it consciously, and it’s really cumbersome and feels awkward. But the more you practice, the more the neurons all fire at the same time, and then it becomes automatic. That’s that automaticity of expertise. That’s what characterizes it. At the time, researchers thought that applying the electric field during the learning of a task would hasten that, and there are still people who think that is true, particularly in stroke recovery. They said that when someone recovering from a stroke gets electrical stimulation while they are relearning a task, it instructs the neurons to work together better. If you just give someone tDCS a few times a day randomly, while they’re recovering from a stroke, then it doesn’t work. The thinking is you need electrical stimulation at the actual intervention point of relearning—be it to walk or move a hand.

Goudarzi: So, you were receiving current while shooting.

Adee:  During my last trial while I was shooting. But there’s still so much we don’t know. Around 2014, someone else had done a meta-analysis and said it turns out all of this is bunk, none of it works; if you average out all the effects, you get absolutely nothing. Other people tried to put tDCS on a cadaver and said that there wasn’t enough electricity that was penetrating the skull to make any difference in terms of action potentials. But the next meta-analysis found some effectiveness for depression and anxiety. The study, published in 2017, was one of the best out there.

I got interested in this because I think a lot of the people who think it’s garbage are basing their assessment on two things. The first one is that there’s a long history of electro quackery, people using really dubious electrical devices. In my book, that was one of the most fun chapters to research, because you get to talk about people saying electricity will cure hysteria and other Victorian maladies that we know don’t exist. The second thing is that we don’t really know how the brain works in any meaningful way.

We can intervene in ways that are really targeted, like deep brain stimulation for Parkinson’s disease, and they can be good; nobody’s calling that garbage because that works amazingly well. But the thing is, people think they know how the brain works because we’ve had this idea of the brain as a computer, and it’s all about neurons with their action potentials. But bioelectricity is way more complicated than that. They’re still trying to figure out what variables to take into account when they’re doing transcranial direct current stimulation. For example, you have to account for the thickness of a person’s skull. Because living bone is a conductive material, you need to account for placement on the precise area of the cortex that you’re trying to stimulate to know exactly how you want the current to flow. I think we’re not going to have verified clinical treatments for surface stimulation for a long time. But that doesn’t mean it’s never going to happen. It just means that we really need to do a lot of basic science and get standards.

Goudarzi: Why would DARPA disclose this research?

Adee: They’ve always had an interesting relationship with the public because the military obviously can’t really get that cozy with journalists. But I think there’s something about blue sky research that gets tech journalists really excited, especially when it’s not directly weapons related. I can raise my hand and say I also was a bit guilty of that. I don’t want to say something ill-considered, but if the army had a PR wing, it would be DARPA. This is the agency that developed GPS, for example, and many other technologies society now relies on. It’s really well funded, which is something you can’t really say for a lot of basic science; people don’t throw that much money at stuff that doesn’t have immediate and obvious benefits. For example, in the clinical trial of metformin, a diabetes drug, there was a control group of people who were not diabetic and ostensibly healthier than type two diabetic people who were being given the drug. Supposedly, the control group was the measure against which this drug was meant to work. After being on metformin, all the people who were taking this drug ended up being way healthier than the control group. So, researchers wondered if this is an anti-aging super drug. Everybody got really excited about metformin, and you would think people would be just lining up to throw money at a proper clinical trial. But that’s not true. It was approved by the FDA for a proper clinical trial in 2015, and it has been impossible to fund the trial because clinical trials are wildly expensive. I think this year, a Saudi oligarch gave the money to finally get it started but that’s the kind of lottery-winning lightning bolt from the sky you have to depend on. So, when the military gives a lot of money to something like neural prosthetics, or something that could alleviate the symptoms of depression or brain machine interfaces, people get really excited, because where else is that money going to come from to pursue the basic science that is needed to push it forward? I think DARPA knows that’s something that’ll make people optimistic, and so they will report on it.

Goudarzi: You talked about some of the upsides of manipulating the electrome, at least in some cases such as treating certain types of depression. Can you first explain what the electrome is, and then talk about some of the downsides of manipulating it?

Adee: The electrome is quite a new term: It was put in a paper for the first time in 2016, and not everybody shares the same definition. I’ve been trying to ask many researchers for their opinion about the definition. The thing that we’ve sort of converged on is the electrical properties and dimensions of cells, sub-cellular compartments, at the level of organs and at the level of organisms and how that interacts with other aspects like endocrinology hormones and gene expression. It sounds complicated, but when you break it down to the basics, you end up finding, for example, that cellular voltage is something useful that cells use in order to communicate, not just in the nervous system. That’s basically the electrome—I want the word “electrome” to do for bio-electricity research what the word “genome” did for molecular biology.

There are a lot of upsides to manipulating it, but there are also a lot of downsides, and I see it as parceling out into three rough areas. One of them is near term issues to look out for. There’s a bioethicist at Duke University, Nita A. Farahany, who has written a book called The Battle for your Brain: Defending the Right to Think Freely in the Age of Neurotechnology that lays out well something that I just skim over in my book, which is this idea that people see these brainwaves and conclude that they say something definitive about the person’s inner mind. In cases where there have been years of rigorous research, that can be true. So, for example, sleep scientists use brain waves, like the delta rhythm, to determine where someone is in their sleep cycle, such as REM, and non-REM sleep. This kind of research can be used to diagnose sleep disorders. Back in 1924 when Hans Berger invented the EEG, he actually was able to see waveform differences between when someone was alert and when they were sort of drifting off and not totally paying attention. Now we’re getting to the point where people have a lot of faith in technology to tell them about their internal state in ways that maybe overreach. And I would say it’s way too much faith because we have this two-dimensional view of the brain in terms of its electrical rhythms, but we mistake that for a really precise picture of what’s happening.

Dr. Farahany goes deeply into this in her book, but the danger is that corporations are going to want to measure this and surveil it. They’re already doing it for long haul truckers. They’re trying to make sure that they’re awake—that kind of monitoring is very hard to argue against. But when you get other types of monitoring, [then we run into issues]. For example, say your employer wants you to wear a headset, and maybe it’s really cute like some kind of stylish headset, and it’ll measure if you are really paying attention on the job, or if you are really sluggish today. Management gurus are always trying to come up with some new bullshit; they’re always trying to find some area in which their expertise can give them a unique selling point, and this is just tailor made for it. Suddenly everybody’s going to have to wear these headsets. Does it matter if it’s giving you a real picture of what’s going on in the brain? Or is it going to become a mechanism of control? And are you going to have to figure out how to get your brainwaves into the right state so that your employer isn’t going to get mad at you and dock your pay? So, I think Dr. Farahany is correct, that’s the most imminent danger because those headsets don’t even have to work perfectly for them to be used as an instrument of control.

It’s a similar vibe to the algorithmic policing story, where they measured certain things and thereby decided where they should deploy more police. Well, it turned out that they were measuring all the wrong things and making the problem almost worse. I think that flavor of problem is a possibility for this.

The second potential problem is the cult of personality around neural implants, neurohacking, that may be emerging. Deep brain penetrating electrodes are probably a very poor idea for anyone who doesn’t have a really profound disorder that prevents them from speaking, completely locks them into their body, or makes them unable to move any part of themselves.

With a Utah Array, it’s been possible to decode speech from the brain’s electrical signals. This is a penetrating implant that requires brain surgery. Used with machine learning and predictive signal processing, it can decode up to 65 words per minute, which is amazing. But I think there’s a lot of excitement right now around new generations of neural implants. And I worry that people want to try these things—and I’m not even sure if it would be illegal to have a private company start doing trials. It wasn’t illegal to have private “trials” in the young blood clinics in the mid 2010s. They claimed that they were doing trials of young blood transfusions, from young people to old people. But it was a pay to play trial. It wasn’t registered with the FDA, but it wasn’t criminal. I think if you were someone with loads of money, you could just have your own volunteers. Maybe brain surgery is different, but I do worry that people will believe the claims of charismatic billionaires and line up to be trial participants.

The third worry is super deep into the future and the military aspects of this. There’s a lot of work looking at the bioelectric differences between animals that can regenerate limbs when they are lopped off—like starfish and certain lizards—and animals that can’t. They’re trying to figure out whether changing the electrical pattern could, in very specific ways, cause the body’s initial programming of limb formation to basically kickstart again. This is very early days, they’ve had really promising results in frogs, and some interesting results in mice. But if it works in humans (and we don’t know yet, and it’s a long way out), what are some of the psychological effects of that? One of my side gigs is being the science fiction reviewer for New Scientist, and I was just reading a book called Rubicon by J. S. Dewes. It’s about people on the battlefield whose bodies get chewed up and can be perfectly regenerated after being killed in action, and the body is in completely pristine condition after these treatments, but the mind just gets more and more frayed. We have this tradition of trying to figure out how to get soldiers back into the battlefield. After World War I, there was something called torpillage, a kind of shock therapy performed on soldiers to rid them of shell shock so that they could be rushed back onto the battlefield. That’s the sort of thing I would worry about, like how would people use capacities like that in order to treat people more like robots and less than humans.

Goudarzi: Is the United States military or any other government currently conducting this research?

Adee:  This is super preliminary. It’s not regeneration, but it is wound healing. There’s a program called DARPA BETR (Bioelectronics for Tissue Regeneration), and they are looking at whether battlefield wounds can be healed more quickly by, among other things, amplifying or otherwise manipulating the natural electric field that underpins a lot of healing. It’s very preliminary. But I think they’ve done really interesting work. It’s very well-funded: They awarded a $16 million grant to Tufts University, University of California at Santa Cruz, and University of California at Davis.

Goudarzi:  How about transcranial direct current stimulation for being a better shooter; is the United States military using that in the battlefield?

Adee: If it is, I haven’t heard of it. Either it wasn’t working in a way that they could replicate, which I think is the most likely possibility; or if it did work, they’re not advertising it. I think people got quite grumpy about this idea that you can take people and sort of remove those inner voices because the inner voices aren’t just there to torture you. There’s a lot of uses for inner voices that maybe act as a conscience or a sort of a handbrake on things that don’t feel right.

Goudarzi: What happens if this works in a replicable way and governments start using this technology? How would it work, say, in the hands of the wrong groups, like an authoritarian government?

Adee: Well, it’s issues of free will, and right now I’m not sure that’s the biggest worry. The bigger worry is more in terms of the extraordinary levels of propaganda that any government feeds soldiers to set them off against each other. I’m not sure you need electrical stimulation to make that any worse. And then also there’s the whole thing about AI-controlled weapons. That might put soldiers out of a job eventually, so I’m not sure tDCS is the worry to focus on when it comes to how the military functions.

Goudarzi: It appears that people can build systems to give themselves transcranial direct current stimulation: There are instruction videos on YouTube, and the materials are readily available to purchase for relatively small sums. There are also companies that sell kits online. How would this therapy or experiment work in the hands of a regular person? How could it harm individuals, if not used in a studied and controlled way?

Adee: Well, that’s an issue of democratization. As soon as you have a network that lets people download the blueprints, there’s always going to be people who really want to [try it]. The stuff that’s being sold, I’m not sure. Like I said, they’re still really working out those the issues that surround how to make tDCS actually work in a replicable way, including bone density and precise location and time of stimulation—so many variables. And that’s in the lab, with people who’ve been working on this for 20 years. It sounds like a really good idea to build your own kit, except that’s self-experimentation. In 2016, a group of neuroscientists wrote an open letter, basically begging people who were building these things to stop experimenting on themselves because you can temporarily blind yourself. I’ve seen some pictures of really bad burns that people gave themselves because they misjudged how good of an electrician they were. But it’s a free country, and if you want to download the blueprint and make one of these, it’s not going to be possible to stop a person. I think people have gotten maybe slightly more disappointed in this as it’s revealed that it’s not like the “Matrix” or a silver bullet that makes you into a weapons-grade assassin. It has been subject to the hype cycle and is now in the trough of disillusionment, and people aren’t probably doing it very much anymore.

Goudarzi: That’s interesting because I was wondering if people with destructive tendencies, like a mass shooter, will use this as an aid.

Adee:  Sadly enough, I don’t think people need a tDCS kit to do mass shootings.

Goudarzi: You mentioned business owners who might want more productivity out of their workers. Is that a serious worry?

Adee:  As soon as they can show something really works, then I think that is a danger. If you can really show that something makes people—I don’t know how you measure it—way more productive, like no one is chit-chatting, and people are hyper focused on getting their work done during the day, then, of course, companies are going to start being like, “are you willing to wear one of these?” Then that’s going to turn into well, “you better be prepared to wear one of these,” and that’s going to turn into “if you don’t wear it, then you’re fired.” So yeah, it’s a very familiar slope. But first they would have to prove that this actually works. At that point, that’s a problem.

There’s a reporter called Alex Hutchinson, who wrote about tDCS being used in sports because for a while, a lot of elite athletes—like the U.S. Olympic ski team, the Golden State Warriors, and cyclists—were using tDCS kits. He was talking to ethicists about the psychology of this and they told him, don’t kid yourself, once one person uses it and gets results, it’s no longer optional. It’s no longer a matter of free will saying I’m not going to use it. Now, compulsory use is the new baseline.

I think if you use these things once it’s fine, but if it works, then you’re doing something to the brain’s metabolism. If you use it for eight hours a day, that’s going to either just deplete you or have long-term effects. So, again, we just need so much basic science around this. Studying the electrome would be really useful for this because there’s electrical stuff going on in the brain that doesn’t have to do with action potentials, and we just don’t really understand it yet.

Goudarzi: With regards to athletes using transcranial direct current stimulation, what are some ethical concerns?

Adee: There was an article in Nature about this, but basically the people on the Olympic ski team who were using it were raising questions about brain doping. There was a mini controversy for a while asking how this is different from using other types of enhancers, and how the electrical stimulation of the brain is defined. I think all of that seems to have sort of died away because again, there are so many variables. The kits these guys were using were from Silicon Valley companies that had sprouted up in the wake of a bit of media hype who were like we can make one of these headsets ourselves, but it’s a one-size-fits-all deal. And tDCS is about affecting your brain, which is the least one-size-fits-all thing I can think of. So, it’s not that tDCS isn’t going to work, it’s just that we don’t yet understand the conditions under which it does. It definitely works for some people, same as deep brain stimulation. (They’re trying to trial that for depression right now, and it’s been really good for Parkinson’s.) But with depression, it’s one of these things where for some people it absolutely doesn’t work; for other people, they wake up like, look at the clinician, and say, “what did you do?” It’s a feeling similar to, by orders of magnitude more than, what I experienced, as in what happened to all the angry voices in my head? This deep brain stimulation is so dramatic for some people that suddenly they’re better. And the people who respond stay better; it doesn’t fade, it doesn’t go away. But how do you figure out which particular circuit got turned on and how to turn that circuit on in somebody else when you’re messing with the most complex organ in the universe? So, if tDCS worked perfectly, it would be brain doping. Absolutely. And it would be something that should be forbidden because you’d either force everyone to do it, which is unethical, or some people would have an advantage, which is also unethical.

Goudarzi: Anything else that you wanted to add, especially in the context of disruptive tech?

Adee: I wanted to talk about—it’s a bit meta—a fear of a fear. One of the interesting projects in bioelectricity research is the creation of these robots which are not really robots, but they’re not really alive. Called xenobots, they are cells taken from frog skin and removed from the bio electric governance of the frog individual. When you take these cells and you put them in a little petri dish, they start glomming together and form little blobs that move around. There were three really big papers about this. One of them detailed that they learned to reproduce. Not by sexual reproduction. They just, like ran around with their little PAC-MAN mouths gobbling up more frog cells and making them into new little blobs that could also autonomously ran around. I read some really alarmed coverage of this where they were talking about how this could be the precursor to the gray goo threat that could kill us all and would turn everything into frog xenobots and then the world ends, or that it’s an environmental catastrophe. But forcing this existing disaster narrative onto new research can be really damaging. Because this xenobot thing shows that you can make robotic organisms out of living materials, which is a an incredibly positive development. These are skin cells. They don’t eat, and they don’t reproduce uncontrollably, because they die, like all living things. So, they’re not going to be these silicon monstrosities that go out of control. Things like this could be the foundation of the sorts of organisms that we need to clean up some of the environmental catastrophe that we’ve created. And I think, when it comes from a new discipline, new research gets people panicky, and it’s worth remembering that everything needs to be looked at on its own terms. The gray goo narrative fit so nicely into it because they were little and autonomous, but when you look more closely, that’s not actually what anyone is building here. But it was so tempting to slot them into this idea that the xenobots are going to take over the world and that it’s research with inevitable bad consequences. I think with a lot of bioelectricity research, people should look at it with clear eyes, not through old frameworks.

Goudarzi: And by bioelectricity, you’re also referring transcranial direct current stimulation research?

Adee:  Yes. Because some of the stuff around how xenobots are created are about manipulating their ion channels. So, bioelectricity underpins some of their creation. I just think, this field has been so affected by people thinking this fits into a preexisting narrative and bias. So that is how they will view it, and it’s really hard to achieve escape velocity from existing narratives to try to look at something directly. You always have to understand if something is going to pose a threat or be a problem, that’s really important. But I think sometimes people overindex on old narratives and try to shoehorn in new stuff so that it seems similar. That’s just a just a caution against that because it can be an opportunity cost.


Together, we make the world safer.

The Bulletin elevates expert voices above the noise. But as an independent nonprofit organization, our operations depend on the support of readers like you. Help us continue to deliver quality journalism that holds leaders accountable. Your support of our work at any level is important. In return, we promise our coverage will be understandable, influential, vigilant, solution-oriented, and fair-minded. Together we can make a difference.

Get alerts about this thread
Notify of
guest
0 Comments
Inline Feedbacks
View all comments