The TED AI Show: Can AI read your mind? The battle for your brain w/ Nita Farahany

[00:00:00] [SPEAKER_01]: TED Audio Collective

[00:00:07] [SPEAKER_03]: Scene 1, interior.

[00:00:09] [SPEAKER_03]: An office space.

[00:00:11] [SPEAKER_03]: Fluorescent.

[00:00:12] [SPEAKER_03]: There.

[00:00:13] [SPEAKER_03]: The year is 2035, and neurotechnology is the new normal.

[00:00:19] [SPEAKER_03]: An employee sits at their desk.

[00:00:21] [SPEAKER_03]: They're wearing earbuds that their workplace has issued them.

[00:00:24] [SPEAKER_03]: But these earbuds are doing much more than playing music, because they have brain sensors

[00:00:29] [SPEAKER_03]: embedded inside them.

[00:00:31] [SPEAKER_01]: It's also tracking their stress levels while they're interacting with their screen and

[00:00:35] [SPEAKER_01]: typing a memo.

[00:00:36] [SPEAKER_01]: And then they're sitting back relaxing and they're reviewing their brain data over the

[00:00:41] [SPEAKER_01]: past couple of months, and they notice that there's something unusual going on when

[00:00:45] [SPEAKER_01]: they're sleeping.

[00:00:46] [SPEAKER_01]: And so they pull up a new email, type it through their brain sensing earbuds,

[00:00:52] [SPEAKER_01]: and they send off a quick message to their doctor and say like, hey, notice that there's

[00:00:56] [SPEAKER_01]: something unusual going on here.

[00:00:58] [SPEAKER_01]: Could you take a look and let me know what you think?

[00:01:00] [SPEAKER_03]: There are no physical screens or keyboards here.

[00:01:03] [SPEAKER_03]: Only desks where the workers sit and stare straight ahead, clicking around office software

[00:01:08] [SPEAKER_03]: with their minds.

[00:01:10] [SPEAKER_03]: Our worker is sitting with her arms folded at a desk.

[00:01:13] [SPEAKER_03]: As she waits for her doctor to respond, her thoughts wander.

[00:01:16] [SPEAKER_01]: She starts to fantasize about one of her colleagues, and then suddenly starts to

[00:01:22] [SPEAKER_01]: worry realizing that her employer has access to all of her brain data.

[00:01:26] [SPEAKER_01]: And she notices that a little message pops up on her screen warning about interoffice

[00:01:31] [SPEAKER_01]: romances and she doesn't want to get in trouble in that context.

[00:01:34] [SPEAKER_01]: She's relieved when later in the day she gets an email from her boss that tells her

[00:01:38] [SPEAKER_01]: that she's getting a performance bonus, and the reason is her brain metrics show

[00:01:44] [SPEAKER_01]: that she's really just been on it.

[00:01:47] [SPEAKER_01]: She leaves work, she's still jamming to the music with her brain sensing earbuds in.

[00:01:52] [SPEAKER_01]: Playlists will become increasingly responsive to brain activity.

[00:01:55] [SPEAKER_01]: That's kind of a given.

[00:01:57] [SPEAKER_01]: She gets home, has her brain sensing earbuds in while she sleeps at night.

[00:02:01] [SPEAKER_01]: That allows her to track brain activity during the night.

[00:02:04] [SPEAKER_01]: It also potentially could allow people to do things like market to her while

[00:02:08] [SPEAKER_01]: she's sleeping.

[00:02:10] [SPEAKER_01]: And then she comes into work the next day.

[00:02:12] [SPEAKER_01]: There's a somber cloud that's fallen over the office and the reason is that

[00:02:16] [SPEAKER_01]: one of her colleagues has been arrested under suspicion of engaging in some kind

[00:02:23] [SPEAKER_01]: And she's really worried because she's been secretly working with that person on

[00:02:29] [SPEAKER_01]: a startup.

[00:02:30] [SPEAKER_01]: But it turns out that as you work more closely with someone, you can start to

[00:02:33] [SPEAKER_01]: see synchronization of brain data between two people.

[00:02:37] [SPEAKER_01]: And she's worried that that synchronization data is going to be used by

[00:02:41] [SPEAKER_01]: the authorities to implicate her in his wrongdoings.

[00:02:46] [SPEAKER_03]: A world overrun by neurotechnology.

[00:02:48] [SPEAKER_03]: It all sounds very sci-fi.

[00:02:50] [SPEAKER_03]: But our guest today, Nita Farahani, says that many of our devices already have

[00:02:55] [SPEAKER_03]: these sensing capabilities.

[00:02:57] [SPEAKER_03]: And this kind of workplace experience is just within reach.

[00:03:00] [SPEAKER_01]: There's already workplaces that are issuing these earbuds that have brain

[00:03:04] [SPEAKER_01]: sensing devices in them could really both benefit and be used by an

[00:03:09] [SPEAKER_01]: individual and be used against them.

[00:03:14] [SPEAKER_03]: I'm Bilal Vosadou and this is the TED AI Show, where we figure out how to

[00:03:18] [SPEAKER_03]: live and thrive in a world where AI is changing everything.

[00:03:26] [SPEAKER_02]: Imagine this.

[00:03:28] [SPEAKER_02]: In 2030, the CFO of a Fortune 100 company is a bot.

[00:03:33] [SPEAKER_02]: I'm Paul Michaelman and on Imagine This, we'll be exploring possible futures and

[00:03:37] [SPEAKER_02]: the implications they hold for organizations.

[00:03:40] [SPEAKER_02]: Joining me will be BCG's top experts as well as my co-host Jean, BCG's

[00:03:44] [SPEAKER_02]: conversational gen AI agent.

[00:03:46] [SPEAKER_02]: Blending human creativity with AI innovation, this podcast promises an

[00:03:51] [SPEAKER_02]: unmatched listening journey.

[00:03:53] [SPEAKER_02]: Join us on Imagine This from BCG.

[00:03:57] [SPEAKER_00]: Hey, TED Tech listeners.

[00:03:58] [SPEAKER_00]: If you enjoy this show, we think you'll enjoy a new show about how tech is

[00:04:02] [SPEAKER_00]: changing our politics.

[00:04:03] [SPEAKER_00]: It's called Wired Politics Lab, hosted by Wired senior politics editor Leah

[00:04:08] [SPEAKER_00]: Feiger and a rotating cast of Wired's political reporters.

[00:04:12] [SPEAKER_00]: Each week, you'll be guided through the exciting, challenging and sometimes

[00:04:16] [SPEAKER_00]: entertaining vortex of Internet extremism, conspiracies and

[00:04:19] [SPEAKER_00]: disinformation.

[00:04:21] [SPEAKER_00]: From AI chat bots running for political office to deep fakes

[00:04:24] [SPEAKER_00]: influencing voters, you'll hear in-depth analysis and conversations based on facts

[00:04:28] [SPEAKER_00]: and research.

[00:04:29] [SPEAKER_00]: Find and follow Wired Politics Lab wherever you listen.

[00:04:37] [SPEAKER_03]: Today, we're going to dive into AI enhanced neurotechnology with ethicist

[00:04:41] [SPEAKER_03]: and legal scholar, Nita Farahani.

[00:04:44] [SPEAKER_03]: Nita is the author of The Battle for Your Brain, Defending the Right to

[00:04:48] [SPEAKER_03]: Think Freely in the Age of Neurotechnology, a book I found both

[00:04:52] [SPEAKER_03]: fascinating and terrifying.

[00:04:54] [SPEAKER_03]: It's about how neurotechnology without regulation has the power to infringe upon

[00:04:59] [SPEAKER_03]: our last bastion of privacy, our privacy of thought.

[00:05:03] [SPEAKER_03]: But is this kind of technology inevitable?

[00:05:06] [SPEAKER_03]: And if so, how do we preserve our cognitive liberty, a term that Nita

[00:05:11] [SPEAKER_03]: has used that I find myself thinking about a lot lately?

[00:05:14] [SPEAKER_03]: I followed up with Nita because I had so many questions.

[00:05:20] [SPEAKER_03]: Nita, welcome to the TED AI Show.

[00:05:22] [SPEAKER_01]: Thanks for having me.

[00:05:23] [SPEAKER_03]: First, I have to ask, can you start by giving us a brief overview of

[00:05:27] [SPEAKER_03]: neurotechnology?

[00:05:28] [SPEAKER_01]: Sure.

[00:05:29] [SPEAKER_03]: What it is, how does it work at a high level and what are some of the

[00:05:33] [SPEAKER_03]: dominant use cases that companies are pursuing?

[00:05:36] [SPEAKER_01]: Up until now, if somebody hears neurotechnology, they associate it with

[00:05:41] [SPEAKER_01]: maybe something like what Elon Musk is doing, which is with his

[00:05:44] [SPEAKER_01]: Neuralink company, drilling a small hole into the skull,

[00:05:49] [SPEAKER_01]: implanting electrodes deep into the brain and then enabling somebody who

[00:05:54] [SPEAKER_01]: maybe has lost the ability to communicate or has lost the ability to move to

[00:05:59] [SPEAKER_01]: regain some of that functionality.

[00:06:01] [SPEAKER_01]: That is also neurotechnology, but it is not the neurotechnology that's

[00:06:04] [SPEAKER_01]: going to impact most of us.

[00:06:06] [SPEAKER_01]: What I'm focused on is the Internet of Things and wearable technology.

[00:06:10] [SPEAKER_01]: So the fact that people are increasingly wearing devices like a

[00:06:15] [SPEAKER_01]: smartwatch or earbuds or XR devices like a virtual reality and augmented

[00:06:20] [SPEAKER_01]: reality glasses that are packed with sensors and those sensors are picking

[00:06:24] [SPEAKER_01]: up different aspects of our bodily function.

[00:06:28] [SPEAKER_01]: Anytime you think, anytime you do anything, neurons are firing in your

[00:06:31] [SPEAKER_01]: brain.

[00:06:32] [SPEAKER_01]: They give off tiny electrical discharge.

[00:06:34] [SPEAKER_01]: Hundreds of thousands of neurons may be firing at any given moment

[00:06:37] [SPEAKER_01]: that reflect different mental states.

[00:06:40] [SPEAKER_01]: So when you're happy, when you're sad, when your mind is

[00:06:43] [SPEAKER_01]: wandering, or when you think up, down, left, right and you want to navigate

[00:06:46] [SPEAKER_01]: around a screen, those are electrical signals that can be picked up by these

[00:06:51] [SPEAKER_01]: sensors and then AI enables the decoding of that into commands.

[00:06:56] [SPEAKER_01]: Most people are used to, for example, heart rate or the number of

[00:07:00] [SPEAKER_01]: steps that they're taking or even temperature being something that's

[00:07:03] [SPEAKER_01]: tracked.

[00:07:03] [SPEAKER_01]: And what's already on the market but starting to come in a much more

[00:07:07] [SPEAKER_01]: widespread fashion is embedding brain sensors into those devices.

[00:07:12] [SPEAKER_01]: Mostly electroencephalography sensors, EEG sensors, and those sensors can be

[00:07:18] [SPEAKER_01]: put into earbuds or headphones or even small wearable tattoos behind the ear

[00:07:23] [SPEAKER_01]: that pick up electrical activity in the brain.

[00:07:27] [SPEAKER_01]: And so really the way to think about neurotechnology is not some separate

[00:07:32] [SPEAKER_01]: device although there are many of those that already exist on the

[00:07:35] [SPEAKER_01]: marketplace like a forehead band or sensors that can be put into a hard

[00:07:39] [SPEAKER_01]: hat.

[00:07:40] [SPEAKER_01]: But instead to think about our wearable devices we're already wearing with new

[00:07:45] [SPEAKER_01]: sensors that have the capability of picking up brain activity.

[00:07:50] [SPEAKER_03]: Even recently with the release of the Apple Vision Pro or the MetaQuest 3,

[00:07:54] [SPEAKER_03]: the way I'm thinking about it is on one hand it's a VR headset,

[00:07:57] [SPEAKER_03]: on the other hand it's a biometric recorder on your face.

[00:08:00] [SPEAKER_03]: That's right.

[00:08:01] [SPEAKER_03]: What is currently possible with this technology as far as mind reading

[00:08:05] [SPEAKER_03]: goes and what is not quite possible yet but is on the near horizon?

[00:08:09] [SPEAKER_01]: Yeah, it's a good question.

[00:08:10] [SPEAKER_01]: XR is in many ways from the ground up new technology.

[00:08:15] [SPEAKER_01]: It's computing on your face but it doesn't have to be constrained by the same old

[00:08:19] [SPEAKER_01]: rules and the same old rules being something like a keyboard and a mouse or a joystick

[00:08:24] [SPEAKER_01]: that doesn't have to be how we navigate through that.

[00:08:27] [SPEAKER_01]: And so as you build a new class of technology it's possible to reimagine what

[00:08:33] [SPEAKER_01]: interacting with that technology looks like.

[00:08:35] [SPEAKER_01]: So that's what these companies have done is they've packed them full of biometric sensors

[00:08:40] [SPEAKER_01]: that whether those are cameras or facial recognition or eye gaze or brain sensors,

[00:08:45] [SPEAKER_01]: all of them are being trained on the ability to be able to make

[00:08:49] [SPEAKER_01]: inferences about brain and mental state.

[00:08:51] [SPEAKER_01]: So are you happy?

[00:08:52] [SPEAKER_01]: Are you sad?

[00:08:53] [SPEAKER_01]: Are you tired?

[00:08:55] [SPEAKER_01]: Are you awake?

[00:08:56] [SPEAKER_01]: Is your mind wandering or are you paying attention?

[00:08:59] [SPEAKER_01]: Some of these kind of basic emotional states.

[00:09:02] [SPEAKER_01]: Also most of them are pretty good at being trained on brain activity to figure out

[00:09:07] [SPEAKER_01]: cursor activity.

[00:09:08] [SPEAKER_01]: So up, down, left, right.

[00:09:10] [SPEAKER_01]: So a lot of the things that you could do with your mouse

[00:09:12] [SPEAKER_01]: you could do with a brain computer interface device.

[00:09:16] [SPEAKER_01]: And then they're getting better at being able to decode an intention to type.

[00:09:20] [SPEAKER_01]: What's not there yet from a true mind reading capacity for wearable devices

[00:09:26] [SPEAKER_01]: is literally what you're thinking or continuous language in your brain.

[00:09:30] [SPEAKER_01]: You can't quite at this point think, oh, I need to send a quick text message

[00:09:34] [SPEAKER_01]: to my husband and then have that somehow decoded from your brain sent to your

[00:09:39] [SPEAKER_01]: mobile device and then sending off a message to him by thinking about doing so.

[00:09:45] [SPEAKER_01]: But that's coming.

[00:09:47] [SPEAKER_01]: So that's the kind of thing where the intention to communicate is something

[00:09:51] [SPEAKER_01]: that can increasingly be decoded by these devices.

[00:09:54] [SPEAKER_01]: And that's where generative AI really can be a game changer because generative AI

[00:09:59] [SPEAKER_01]: trained on natural language and conversations becomes much more powerful

[00:10:04] [SPEAKER_01]: at being able to predict the next word that it is that you want to type.

[00:10:08] [SPEAKER_01]: And so that kind of auto complete feature of generative AI paired together with brain

[00:10:13] [SPEAKER_01]: activity and the increasing capability of being able to put large language models

[00:10:19] [SPEAKER_01]: on a device mean that the devices and the brain sensors can co-evolve with the

[00:10:25] [SPEAKER_01]: individual so that it becomes increasingly more precise at more and more mind reading.

[00:10:29] [SPEAKER_03]: It's really powerful.

[00:10:30] [SPEAKER_03]: And so it sounds like this technology can kind of coarsely infer your brain states,

[00:10:36] [SPEAKER_03]: not exactly read your internal monologue, but your point as you introduce new modalities,

[00:10:40] [SPEAKER_03]: you're looking at what is the camera feed see?

[00:10:42] [SPEAKER_03]: What is like, what is the user's eye gaze focused on and then throw in generative

[00:10:47] [SPEAKER_03]: AI to throw a layer of prediction on top of that.

[00:10:50] [SPEAKER_03]: Even with these very core sensing capabilities,

[00:10:52] [SPEAKER_03]: you can actually start making very accurate predictions.

[00:10:55] [SPEAKER_01]: That's right.

[00:10:55] [SPEAKER_03]: So I want to understand why is this something so many companies are racing towards and

[00:11:00] [SPEAKER_03]: investing in heavily? What are some of the most positive use cases that you're excited about?

[00:11:05] [SPEAKER_03]: How is this actually going to help users?

[00:11:07] [SPEAKER_01]: Yeah, I think there's a lot of ways it can help users.

[00:11:11] [SPEAKER_01]: And I don't think about neurotechnology as just those brain sensors that are put into

[00:11:16] [SPEAKER_01]: an earbud. I think about this as an entire class of cognitive biometrics,

[00:11:21] [SPEAKER_01]: which allow predictions about what a person is thinking and feeling.

[00:11:24] [SPEAKER_01]: And that can have a lot of really positive use cases, the more we can actually gain

[00:11:28] [SPEAKER_01]: really accurate insights about what's happening inside of our brains.

[00:11:33] [SPEAKER_01]: So the companies who are racing to this space are all of the major tech companies, right?

[00:11:38] [SPEAKER_01]: Apple has all kinds of patents and investments in this space.

[00:11:40] [SPEAKER_01]: And you see the same thing at Meta and Google and others, because if XR technology

[00:11:45] [SPEAKER_01]: really takes off, which I think eventually it will, it doesn't make sense to have it powered

[00:11:51] [SPEAKER_01]: or have us tethered to a keyboard and a mouse or a joystick. There has to be a more natural

[00:11:56] [SPEAKER_01]: and seamless way of interacting with these devices. The approach has not been like,

[00:12:02] [SPEAKER_01]: let's figure out how to commodify all the brain data, but it's like, how do we build

[00:12:05] [SPEAKER_01]: a new class of technology from the ground up and have a new way of thinking about

[00:12:10] [SPEAKER_01]: interacting with that technology? But then there's also a huge amount of investment

[00:12:17] [SPEAKER_01]: that's been happening to the side, which is on mental health and recognizing that the brain

[00:12:22] [SPEAKER_01]: is an untapped potential of areas and products that could be targeted at mental health.

[00:12:29] [SPEAKER_01]: And so this is things like seeing a huge number of apps that are focused on

[00:12:33] [SPEAKER_01]: whether it's AI-based mental health or if it's journaling or meditation apps.

[00:12:40] [SPEAKER_01]: There you have billions to trillions of dollar industry potentially around brain

[00:12:45] [SPEAKER_01]: health and wellness. And then they start to converge because suddenly if you have

[00:12:49] [SPEAKER_01]: this new class of devices that have been focusing on neural interface and you have

[00:12:53] [SPEAKER_01]: this untapped potential of brain health and wellness, suddenly you have access to much

[00:12:58] [SPEAKER_01]: better insights and much better ways of being able to actually interact with the brain and

[00:13:02] [SPEAKER_01]: to gather data that could be used for much more targeted products. So I think it's like

[00:13:08] [SPEAKER_01]: the biggest untapped market if you think of it that way.

[00:13:11] [SPEAKER_03]: That's well said. I believe Zuckerberg called neural interfaces the holy grail of VR, right?

[00:13:17] [SPEAKER_01]: That's right. Yeah. The Cordy Heber doesn't make any sense when you're thinking about

[00:13:21] [SPEAKER_01]: XR, right? It just doesn't make sense. Nor does the mouse, right?

[00:13:27] [SPEAKER_01]: Totally.

[00:13:28] [SPEAKER_01]: It's become second nature to us, but the idea that I have to work on a mouse to

[00:13:32] [SPEAKER_01]: navigate around my screen, touch screen was a good innovation that way, but that's still

[00:13:37] [SPEAKER_01]: awkward. And we just were inefficient in our interactions with technology.

[00:13:47] [SPEAKER_03]: This world you painted sounds very efficient, if that's the right word for it, but there

[00:13:53] [SPEAKER_03]: is something uncanny about it, right? And I want to get into what makes it uncanny.

[00:13:58] [SPEAKER_03]: In evaluating some of the risks of this technology, you brought a lot of awareness

[00:14:02] [SPEAKER_03]: to this idea of cognitive liberty. So how do you define cognitive liberty and why do you

[00:14:08] [SPEAKER_03]: define it to be a useful term when you're looking at this new wave of neurotechnology?

[00:14:12] [SPEAKER_01]: I came on the term cognitive liberty around 2010 or something, and it resonates with me well,

[00:14:17] [SPEAKER_01]: because what it reflects from my perspective is this right to self-determination over

[00:14:23] [SPEAKER_01]: our brain and mental experiences as a fundamental matter. And what does that even

[00:14:27] [SPEAKER_01]: mean to have self-determination? There's this huge amount of literature that has developed

[00:14:32] [SPEAKER_01]: the past couple of decades around what self-determination means and why self-determination

[00:14:37] [SPEAKER_01]: is fundamental to human self-actualization. You need the basic autonomy, the competence,

[00:14:44] [SPEAKER_01]: and the capacity for relatedness to other people. And so for me, it's about that.

[00:14:48] [SPEAKER_01]: It's about those pillars of self-determination, the ways in which technology can both enable it,

[00:14:54] [SPEAKER_01]: but also have increasingly come to interfere with the capacity for self-determination

[00:14:59] [SPEAKER_01]: and how cognitive liberty would give us the goalpost to say, like, what is it that we're

[00:15:04] [SPEAKER_01]: trying to achieve in technological design or in technological regulation? We're trying to

[00:15:09] [SPEAKER_01]: preserve this space of self-determination for individuals to enable them to form their identity,

[00:15:14] [SPEAKER_01]: to have a space of mental privacy, to have the capacity of freedom of thought,

[00:15:18] [SPEAKER_01]: the preconditions to be able to become a fully self-actualized human. So it resonates

[00:15:24] [SPEAKER_01]: really well for me as a liberty interest, as a kind of fundamental right that individuals have

[00:15:30] [SPEAKER_03]: as a precondition to being able to flourish. It's like our minds are this sanctum sanctorum,

[00:15:37] [SPEAKER_03]: right? We feel like we have complete dominion over it, though already technology is influencing

[00:15:44] [SPEAKER_03]: us at a very deep visceral level without us even knowing. And now we're creating these

[00:15:48] [SPEAKER_03]: higher bandwidth forms of sensing. And I'm kind of curious, is there an interesting analogy here

[00:15:55] [SPEAKER_03]: where many folks got their DNA sequenced over the past decade or two without really thinking

[00:16:01] [SPEAKER_03]: about a scenario where maybe a data breach would occur? And it made me wonder from a medical

[00:16:06] [SPEAKER_03]: data perspective, what are going to be the effects of increasing brain and neurological

[00:16:11] [SPEAKER_03]: health transparency without adequate privacy regulations? And obviously America is very

[00:16:16] [SPEAKER_03]: unique in that there is no federal law on internet privacy. Could this medical data

[00:16:22] [SPEAKER_01]: kind of be weaponized against users? Yeah, I mean, I think very much so. So

[00:16:28] [SPEAKER_01]: part of it is people went into direct to consumer genetic testing and could never have

[00:16:34] [SPEAKER_01]: imagined it would eventually be used to solve cold cases and for law enforcement agencies to

[00:16:41] [SPEAKER_01]: collect all of that data and put us all into warrantless searches for here on out, right?

[00:16:48] [SPEAKER_01]: And then some people are like, well, that's okay. I don't mind that. I didn't commit a crime.

[00:16:52] [SPEAKER_01]: And so if it helps find that person, that's great. But as you start to imagine every

[00:16:59] [SPEAKER_01]: possible use case, there are so many use cases that people just don't even contemplate

[00:17:03] [SPEAKER_01]: about the ways in which data can be used or misused against them. I think with brain data,

[00:17:08] [SPEAKER_01]: even more fundamental than that. I don't think it's just let's point to how law enforcement might

[00:17:13] [SPEAKER_01]: use it one day or how others might use it one day. I think it's about the importance of having

[00:17:19] [SPEAKER_01]: a space, like the inner sanctum, the mental privacy, the space we need to even just be us.

[00:17:28] [SPEAKER_01]: And that's almost impossible for us to even grapple with or imagine a world in which we

[00:17:33] [SPEAKER_01]: don't have that. But that's where I spend a lot of my mental energy is imagining that

[00:17:37] [SPEAKER_01]: world where we don't have that. And think about like as a kid, all of the thoughts where you're

[00:17:43] [SPEAKER_01]: like, maybe I'm weird. I have different gender identity or preferences and sexual orientation,

[00:17:51] [SPEAKER_01]: or maybe I don't want to be a doctor and my parents really want me to be a doctor.

[00:17:57] [SPEAKER_01]: And all of these little thoughts that you have every day. If you don't have a space

[00:18:01] [SPEAKER_01]: you can do that, where you don't, where you feel safe to just be that person that figures

[00:18:09] [SPEAKER_01]: out who you are. What does that world look like for humans to become? Right? This like

[00:18:14] [SPEAKER_01]: act of becoming that we're constantly engaged in. That's the world that I think we're

[00:18:19] [SPEAKER_01]: entering into without realizing it. It's a world where what we've taken for granted is

[00:18:25] [SPEAKER_01]: most fundamental aspect of being human that suddenly may no longer exist. And we're not

[00:18:32] [SPEAKER_01]: putting into place the right protections to ensure this linchpin of humanity is still safe.

[00:18:40] [SPEAKER_03]: This is what it has come to. We're talking about

[00:18:44] [SPEAKER_03]: neurosurveillance. Let's go even further down the rabbit hole and like imagine this future

[00:18:49] [SPEAKER_03]: a bit more. Do you think we'll get to a place where our brain states, our inner thoughts

[00:18:53] [SPEAKER_03]: are fully transparent to one another? How would this impact our personal relationships?

[00:18:59] [SPEAKER_03]: How's this going to impact society? It kind of feels like Twitter on steroids,

[00:19:03] [SPEAKER_03]: like without the crude thumb typing to express your thoughts. Just

[00:19:07] [SPEAKER_03]: knowing what everyone is thinking at all times just feels wild.

[00:19:11] [SPEAKER_01]: So I taught a class at Duke called Let's Talk About Digital You. And it was a class

[00:19:16] [SPEAKER_01]: for undergraduates where the goal was to have them think critically about digital

[00:19:20] [SPEAKER_01]: technologies and how they interact with them. Then also think about what does it mean for them

[00:19:25] [SPEAKER_01]: and who they are as a person. And one of the things I learned in that class when we were

[00:19:29] [SPEAKER_01]: talking about privacy was how almost every kid in the class shared their location data on their

[00:19:35] [SPEAKER_01]: phone with every one of their friends, like hundreds of friends who were tracking them at

[00:19:40] [SPEAKER_01]: all times. I was shocked by this. And then I think about how in the writing of the book,

[00:19:45] [SPEAKER_01]: I was interviewing this person who leads a meditation class that uses neural devices

[00:19:51] [SPEAKER_01]: and how they've created Facebook groups where they're sharing readouts of their

[00:19:58] [SPEAKER_01]: meditation sessions. They're like, oh, look at my gamma activity there. Look at what's happening

[00:20:01] [SPEAKER_01]: with my alpha activity here. And I think is this going to be the latest status update where

[00:20:06] [SPEAKER_01]: you're like, Nita's in a bad mood. Don't talk to her right now. Or like, Nita is

[00:20:12] [SPEAKER_01]: thinking about food. And my friends reach out and say, like, hey, I'm thinking about food too.

[00:20:15] [SPEAKER_01]: Like, let's go get a bite to eat. Like it just becomes something that we decide like we're

[00:20:21] [SPEAKER_01]: going to share all this data with. And it's not hard to imagine that we get to that place

[00:20:27] [SPEAKER_01]: where it becomes something that is much more transparent. Is that all bad? I don't know.

[00:20:31] [SPEAKER_01]: Right. There are some people who believe total transparency is better. I can buy into

[00:20:37] [SPEAKER_01]: argument up until you get to mental privacy. And then I really think this act of becoming

[00:20:43] [SPEAKER_01]: requires that we have much greater control over what, if anything, is shared from our brain data.

[00:20:49] [SPEAKER_03]: You're bringing up a point about this technology that is very much a consent issue.

[00:20:53] [SPEAKER_03]: And sometimes I feel like utility with technology almost ends up being a Trojan horse.

[00:20:58] [SPEAKER_03]: Like we do get a lot of benefit from this stuff. But by consenting to these things,

[00:21:02] [SPEAKER_03]: all these other things happen that we did not expect and we may not even be aware of.

[00:21:08] [SPEAKER_01]: Right. And I think that's what I'm trying to really highlight for people. There's been a

[00:21:14] [SPEAKER_01]: lot that's happened over the past year. And part of it has been me worrying about

[00:21:21] [SPEAKER_01]: normalizing neural surveillance, like the risks become invisible to us. And we accept this

[00:21:27] [SPEAKER_01]: new technology without even stopping to recognize all of the implications of what it is that we're

[00:21:32] [SPEAKER_01]: adopting. And part of it is interesting. It's how the technology is introduced to us. A lot of

[00:21:37] [SPEAKER_01]: times it's with, as you put it, utility. Right? There's some utility that we buy into

[00:21:43] [SPEAKER_01]: or there's some experience where it's like the only way you get this fun experience is by

[00:21:49] [SPEAKER_01]: opting into this new technology without then sensitizing us to all of the risks of doing so

[00:21:56] [SPEAKER_01]: and really contemplating like we're crossing a new barrier here. We're crossing some

[00:22:01] [SPEAKER_01]: threshold that we've never passed before and it has profound implications. It is this space of

[00:22:07] [SPEAKER_01]: what it means to be human. Part of how we define vulnerability and intimacy with each other

[00:22:13] [SPEAKER_01]: is what we choose to share and what we choose to hold back. And I don't think that's just a

[00:22:18] [SPEAKER_01]: consent issue, right? Because like at some point there's enough peer pressure and social pressure

[00:22:23] [SPEAKER_01]: that like if you're not sharing your location data, like you're weird, you know, why aren't

[00:22:26] [SPEAKER_01]: you sharing it with everybody else? So there's this coercive pressure toward the norm where we

[00:22:32] [SPEAKER_01]: have this collective action problem. If each of us individually decides like, oh, I'm okay

[00:22:37] [SPEAKER_01]: with sharing my neural data and I don't care if I don't have mental privacy, I don't have

[00:22:40] [SPEAKER_01]: anything to hide. Suddenly we have the collective having not recognized that they've given up

[00:22:47] [SPEAKER_01]: this space of what it means to be human without us ever having really talked about it,

[00:22:53] [SPEAKER_01]: recognized how profound it is of a thing to actually see to other people or to companies.

[00:22:59] [SPEAKER_03]: Yeah, it's like we like targeted ads. Oh, that's fine. I like using Instagram.

[00:23:04] [SPEAKER_03]: These algorithms have the coarsest view into our hopes, wishes and desires

[00:23:10] [SPEAKER_03]: and are still doing such a great job. It feels like we're going into this future where

[00:23:14] [SPEAKER_03]: people are like, oh yeah, I'll get that cheap Alexa and then suddenly they're having dreams

[00:23:19] [SPEAKER_03]: about Bud Light or Coars Light or whatever, but they're like, oh, that's cool with me.

[00:23:23] [SPEAKER_01]: And like I'm asleep. I don't really care. Yeah. I mean, that example is a real one,

[00:23:27] [SPEAKER_01]: right? It's like I write about this in my book where, you know, Coars was really

[00:23:32] [SPEAKER_01]: frustrated. They've been locked out of the halftime show for years for the Super Bowl.

[00:23:37] [SPEAKER_01]: So they got together with dream researchers, figured out that it's possible to incubate in

[00:23:43] [SPEAKER_01]: people's minds when they're in their most suggestible state because their conscious

[00:23:47] [SPEAKER_01]: brain is basically checked out as they're falling asleep. Associations between Coars and

[00:23:53] [SPEAKER_01]: mountains and lakes and streams and like, so that you have this idea that like Coars

[00:23:58] [SPEAKER_01]: is crisp and refreshing. But then I imagine this dystopian future of you're wearing your

[00:24:03] [SPEAKER_01]: sleep earbuds to track your sleep activity. The entire economic system of these companies

[00:24:10] [SPEAKER_01]: is based on targeted advertisement and real time bidding of, you know, here's new real

[00:24:16] [SPEAKER_01]: time state to be able to access Nita, whether she's on her search engine or the app store.

[00:24:21] [SPEAKER_01]: And it's like, Hey, we've never targeted when Nita is asleep, but here's her most

[00:24:25] [SPEAKER_01]: suggestible sleep state. She's got a Amazon Echo in her bedroom, like prime time to

[00:24:32] [SPEAKER_01]: advertise and to play a little jingle is right now. And so like what stops that?

[00:24:36] [SPEAKER_01]: What prevents targeted advertisements to us while we are sleeping? Nothing, right?

[00:24:42] [SPEAKER_01]: I mean, other than like the company shouldn't do that. Okay. What's going to stop them from

[00:24:47] [SPEAKER_01]: doing that if that becomes the most effective way to advertise to us and becomes the most

[00:24:51] [SPEAKER_01]: expensive ad real estate that they could sell to ad brokers? Oh, good Lord. Oh, good Lord.

[00:24:58] [SPEAKER_03]: I mean, this studies is basically this targeted dream incubation demonstrates what companies

[00:25:05] [SPEAKER_03]: and social media algorithms can do and also what neurotechnology is capable of already.

[00:25:10] [SPEAKER_03]: And so it's wild because social algorithms not only extract or infer our way of looking at the

[00:25:17] [SPEAKER_03]: world, but they are actively shaping how we perceive the world. That's right. They predict

[00:25:22] [SPEAKER_03]: what we want, but as these platforms become such an intricate part of our lives, they can almost

[00:25:27] [SPEAKER_03]: define what we want. So when it comes to this threat of neuro surveillance, how do you think

[00:25:34] [SPEAKER_03]: the rise of this tech could influence what we feel safe to think on a conscious level

[00:25:39] [SPEAKER_03]: or even more subliminally, like what we're capable of thinking? That's the question I struggle with

[00:25:44] [SPEAKER_01]: the most these days is, you know, as I try to unpack, like what does self-determination mean

[00:25:51] [SPEAKER_01]: anymore in a world where we're being steered constantly. And, you know, the easiest way

[00:25:57] [SPEAKER_01]: people can really understand that is if you sit down to watch a single episode of a show

[00:26:03] [SPEAKER_01]: you end up watching four, it's by design, you are less likely to get up and leave if it's just

[00:26:11] [SPEAKER_01]: automatic. Like if you don't have to make a self-control choice and we can see these

[00:26:15] [SPEAKER_01]: differences. Like if you look at the studies, there's a difference between when you are making

[00:26:20] [SPEAKER_01]: a choice about what video to watch next versus an algorithm is deciding for you what video

[00:26:26] [SPEAKER_01]: to watch next in that the parts of your brain that are responsible for self-control

[00:26:30] [SPEAKER_01]: are basically just turned off. They're silent when you're just being fed content over and

[00:26:34] [SPEAKER_01]: over again. And when you see that, right, self-control is turned off, you are being fed

[00:26:41] [SPEAKER_01]: information and that's changing how you feel. That's changing what you believe.

[00:26:46] [SPEAKER_01]: We are being steered. And is there such a thing as self-determination in a world that's

[00:26:52] [SPEAKER_01]: increasingly connected to technology and a world where that steering will become much, much more

[00:26:57] [SPEAKER_01]: precise? If it's not just the crude interpretation of how many seconds or milliseconds you spent on

[00:27:04] [SPEAKER_01]: a video to try to make an interpretation of you, but literally a direct measurement of your

[00:27:09] [SPEAKER_01]: reaction to information and then you're in an immersive environment. You're not even in the

[00:27:13] [SPEAKER_01]: real world anymore where at least there are some things that are static and that immersive

[00:27:16] [SPEAKER_01]: environment can continuously change in response to your brain activity. It feels like we're

[00:27:21] [SPEAKER_01]: approaching the matrix pretty quickly. And so then I struggle with like, okay, well, like

[00:27:27] [SPEAKER_01]: is like, is there such a thing as self-determination in that world of self-actualization in the world

[00:27:32] [SPEAKER_01]: where there's true autonomy, even if it's relational autonomy, where there's true

[00:27:37] [SPEAKER_01]: competence, where we're exercising self-control and critical thinking skills, where we're

[00:27:41] [SPEAKER_01]: fostering relationships with each other and relatedness with ourselves and with other

[00:27:46] [SPEAKER_01]: people. I still believe there is. I think that we have cognitive liberty as a guiding principle

[00:27:50] [SPEAKER_01]: that points us to different ways of designing technology that align with self-determination

[00:27:57] [SPEAKER_03]: rather than align with human diminishment. I love the phrasing that you're using of

[00:28:02] [SPEAKER_03]: technologies steering us and really the question is who is at the steering wheel,

[00:28:07] [SPEAKER_03]: right? Like who controls the objective function. And so this future starts sounding rather

[00:28:13] [SPEAKER_03]: scary. So I want to zoom back out just a bit here and ask you, you know, you've been a

[00:28:20] [SPEAKER_03]: leader in the world of self-determination, right? And so you've been working to

[00:28:27] [SPEAKER_03]: encode cognitive liberty as a legal right. I'm curious, what would that look like in

[00:28:32] [SPEAKER_01]: regulation and how would it be enforced? First, I'll just emphasize, I think cognitive

[00:28:37] [SPEAKER_01]: liberty, to your point, it's systemic change, right? It's in part about encoding it into

[00:28:42] [SPEAKER_01]: law, but it's also about changing our incentives, changing, you know, economic

[00:28:46] [SPEAKER_01]: alignment with cognitive liberty, commercial design and redesign. And so I think that's

[00:28:50] [SPEAKER_01]: from a legal perspective, I think it starts with a human rights approach. And this is so that

[00:28:55] [SPEAKER_01]: there's a universal global norm around a right to cognitive liberty and recognizing that as an

[00:29:01] [SPEAKER_01]: organizing principle for how we interpret existing human rights law. It's recognizing

[00:29:05] [SPEAKER_01]: privacy includes a right explicitly to mental privacy, which also safeguards from

[00:29:10] [SPEAKER_01]: interference and interception with the automatic processes in our brain. And so much

[00:29:15] [SPEAKER_01]: of this technology is really about that. It's not about robust thought. It's about these

[00:29:19] [SPEAKER_01]: automatic ways that our brain reacts and interfering with them, hijacking them,

[00:29:23] [SPEAKER_01]: manipulating them. And then there's freedom of thought. As long as it's been a right,

[00:29:27] [SPEAKER_01]: it has been recognized as an absolute human right, but it's pretty narrowly constructed

[00:29:31] [SPEAKER_01]: because it's an absolute right. And so that's really around protecting the interception,

[00:29:37] [SPEAKER_01]: manipulation and punishment of thought. And here we're really in the mind reading realm,

[00:29:41] [SPEAKER_01]: right? Which is protecting this space of, you know, what we would really commonly in common

[00:29:47] [SPEAKER_01]: language think of as thought and images in our mind. That's the human rights perspective.

[00:29:53] [SPEAKER_01]: Any good critic of human rights law will say, well, that's only as good as like,

[00:29:57] [SPEAKER_01]: you know, whether people adhere to it and how it's implemented at a national level.

[00:30:02] [SPEAKER_01]: So I also think it's important to recognize what that looks like, you know, at nation state

[00:30:06] [SPEAKER_01]: levels. And part of that is doing things as part of privacy laws to, you know, create

[00:30:12] [SPEAKER_01]: robust rights around cognitive biometric data. Employees have a right to mental privacy,

[00:30:19] [SPEAKER_01]: giving children and students in the educational system a right to not be surveilled for their

[00:30:25] [SPEAKER_01]: mental activity and giving them that space, right? So it's taking these high principles

[00:30:30] [SPEAKER_01]: and then context by context creating robust laws that actually implement those high concepts

[00:30:37] [SPEAKER_01]: into law at different, you know, nation and nation state levels.

[00:30:42] [SPEAKER_03]: So where are we currently at there? I'm curious if there's been any notable progress on this subject

[00:30:47] [SPEAKER_01]: since your book came out. A lot. Yeah. So in the US,

[00:30:51] [SPEAKER_01]: there's been finally some momentum. I'd say I'm not that excited about what's happened here

[00:30:56] [SPEAKER_01]: yet, but I am appreciative of the fact that there have been conversations. So Colorado

[00:31:00] [SPEAKER_01]: passed a new law that provides some protections around neural data when it's used for

[00:31:06] [SPEAKER_01]: identification purposes. That's a very narrow subset, but it's something. California has a

[00:31:11] [SPEAKER_01]: broader law that's currently pending that would make it a sensitive category of data.

[00:31:16] [SPEAKER_01]: The Uniform Law Commission here in the United States, which has appointed commissioners from

[00:31:20] [SPEAKER_01]: every state, has just agreed to create a study committee that might lead to a drafting

[00:31:26] [SPEAKER_01]: committee to have model legislation for all of the states about how to protect this category

[00:31:31] [SPEAKER_01]: data around cognitive biometrics, which is exciting. That launches the summer. UNESCO has

[00:31:37] [SPEAKER_01]: had a major process underway. 194 member countries voted to move toward adopting a global standard

[00:31:43] [SPEAKER_01]: around the ethics of neurotechnology. The first draft of that was published in April.

[00:31:49] [SPEAKER_01]: I'm part of that process. So I was appointed by the US to that process and the co-chair

[00:31:54] [SPEAKER_01]: of the expert committee. And the second draft will come out at the end of August.

[00:31:58] [SPEAKER_01]: And then internationally, there's been a lot of conversations happening around this. The concept

[00:32:04] [SPEAKER_01]: of cognitive liberty, the concept of whether or not there need to be special rights,

[00:32:09] [SPEAKER_01]: broadening the conversation beyond neurotechnology to understand this is a

[00:32:12] [SPEAKER_01]: new class is moving forward. So it's encouraging, but what you find is trying to claw back

[00:32:17] [SPEAKER_01]: rights is much harder than from the get-go having a set of rights in place. So I think we

[00:32:27] [SPEAKER_01]: need to real change before these products are wide scale across society.

[00:32:33] [SPEAKER_03]: This technology is coming fast and furious. And as you've alluded to, and actually as you've

[00:32:37] [SPEAKER_03]: outlined, it's being infused into tech that we use every single day. People are just going to

[00:32:41] [SPEAKER_03]: go buy the next-gen AirPods. One question I have is for the listeners that are perhaps

[00:32:46] [SPEAKER_03]: being exposed to this idea of neurosurveillance and neurotechnology, what advice would you have

[00:32:52] [SPEAKER_03]: as they go about navigating the world and become exposed to these technologies,

[00:32:59] [SPEAKER_03]: both in their personal life but also in the workplace?

[00:33:01] [SPEAKER_01]: We still have time to make choices individually and collectively. We should demand that if

[00:33:08] [SPEAKER_01]: Apple launches EEG sensors, that they be very clear about what their data privacy policies are

[00:33:13] [SPEAKER_01]: with respect to that data. And they've done that with Apple Pro Vision, right? They've

[00:33:16] [SPEAKER_01]: said like, they've been incredibly specific to say, here's what's happening with the eye

[00:33:20] [SPEAKER_01]: tracking data. Here's what lives on device. Here are the inferences that leave the device.

[00:33:25] [SPEAKER_01]: That kind of transparency is great, and I applaud them for doing that. We should only

[00:33:30] [SPEAKER_01]: buy the products that are both transparent with respect to how they're thinking about each

[00:33:34] [SPEAKER_01]: stream of sensor data and not buy products from companies that do not offer those same

[00:33:39] [SPEAKER_01]: assurances. That's a space we have a right to protect and that we should demand as a set

[00:33:46] [SPEAKER_01]: of protections. And so advocate for the rights that are being called for, be part of the process

[00:33:52] [SPEAKER_01]: of advocating for UNESCO to move forward with this process for states, for countries to move

[00:33:56] [SPEAKER_01]: forward with a robust set of rights around cognitive liberty and be part of the change

[00:34:01] [SPEAKER_03]: that makes that happen. That's beautifully put. Thank you so much for joining us.

[00:34:06] [SPEAKER_03]: Thanks for having me. This tech is amazing. And it's not hyperbole to say that we're

[00:34:17] [SPEAKER_03]: mind reading here. One of the central benefits of this technology is gaining visibility into our

[00:34:23] [SPEAKER_03]: own minds, quantifying that endless stream of activity that really influences our day-to-day

[00:34:29] [SPEAKER_03]: perception. And through this new tech, what seems to be as opaque as emotional experiences

[00:34:35] [SPEAKER_03]: can be translated into measurable neurobiological patterns. This means that

[00:34:41] [SPEAKER_03]: we can better understand what's happening in our heads. And it also means that we can more

[00:34:46] [SPEAKER_03]: change what's happening in our heads. If you understand what's going on in a person's mind,

[00:34:51] [SPEAKER_03]: you can create a stimulus to influencing whether that's Orwellian workplace monitoring or the

[00:34:57] [SPEAKER_03]: most persuasive advertisement you've ever seen. It's giving away read, write access to the

[00:35:03] [SPEAKER_03]: core of who we are. So what happens to a society where there are no secrets?

[00:35:09] [SPEAKER_03]: We're all open books. Maybe this is an inevitable transition. We already share so

[00:35:14] [SPEAKER_03]: much of ourselves via social media and do so proactively. That said, we need to place a lot

[00:35:21] [SPEAKER_03]: more value in our personal data, data that we currently view to be innocuous and often sign

[00:35:27] [SPEAKER_03]: away without even thinking about it. But when it comes to tech that has both read and write

[00:35:32] [SPEAKER_03]: access into our minds, we have to be more proactive than we had been about social media.

[00:35:38] [SPEAKER_03]: And we have to value the data before it becomes table stakes to function in society.

[00:35:44] [SPEAKER_03]: After all, the stakes here are high. We're talking about our hearts,

[00:35:48] [SPEAKER_03]: minds and how we view the world. The TED AI show is a part of the TED Audio Collective

[00:35:57] [SPEAKER_03]: and is produced by TED with Cosmic Standard. Our producers are Ella Federer and Sarah McCray.

[00:36:04] [SPEAKER_03]: Our editors are Ben Ben Chang and Alejandra Salazar. Our showrunner is Ivana Tucker and

[00:36:09] [SPEAKER_03]: our associate producer is Ben Montoya. Our engineer is Asia Pilar Simpson. Our technical

[00:36:15] [SPEAKER_03]: director is Jacob Winnink and our executive producer is Eliza Smith.

[00:36:20] [SPEAKER_03]: Our fact checker is Christiana Parta and I'm your host, Bilal Vosadou. See y'all in the next one.