Mindreading with Jean Rémi-King

Summary

StarTalk Radio – "Mindreading with Jean Rémi-King"

Host: Neil deGrasse Tyson
Guest: Jean Rémi-King, Senior Research Scientist at Meta (FAIR, Paris)
Date: August 22, 2025

---

Episode Overview

This Special Edition of StarTalk delves into the intersection of neuroscience, artificial intelligence, and the tantalizing future of "mindreading." Neil deGrasse Tyson, with comic co-hosts Gary O'Reilly and Bubba Wallace, welcomes Jean Rémi-King, who leads research at Meta’s Fundamental AI Research (FAIR) lab in Paris. The episode explores how AI is used to decode the brain’s signals, the possibilities and current limits of mindreading technologies, the profound ethical implications of such advances, and how understanding the brain can help improve AI systems, and vice versa.

---

Key Discussion Points & Insights

1. Meet Jean Rémi-King and the Challenge of Mindreading

• Background: Jean Rémi-King is a neuroscientist at Meta FAIR, Paris. His work sits at the frontier of AI and neuroscience, aiming to better understand both biological and artificial "intelligence."
• AI and the Brain: His team explores how AI can decode neural signals and how the brain’s computational principles might inform new AI algorithms.
• Central Question: Can AI truly “read our minds”? What’s technologically possible, and what are the limitations?

“AI will be driving our cars... But surely it’s never going to be able to read our minds, is it? Well, actually, yeah, it can.”
— Gary O’Reilly (02:30)

---

2. How We Measure Brain Activity

• Non-Invasive Tools:
  • EEG (electroencephalography) and MEG (magnetoencephalography) measure fluctuations in electrical/magnetic fields from neuronal activity.
  • FMRI (functional magnetic resonance imaging) detects changes in blood flow, a proxy for neural activity.
• Spatial and Temporal Resolution:
  • EEG/MEG: High temporal (milliseconds), low spatial.
  • fMRI: High spatial, low temporal resolution.
• Intracranial Recordings: Used in clinical settings (e.g., epilepsy), provide direct measurement, but only with individuals who have electrodes implanted for medical reasons.

“When you look at the raw data, it’s very difficult to guess anything... you’d probably need to start to do the very same task again and again to try to average out the noise.”
— Jean Rémi-King (11:05)

---

3. Decoding Perception: How Close Are We to ‘Mindreading’?

• AI Pattern Recognition: AI excels at finding patterns (after extensive training), making it possible to reconstruct (with varying accuracy) what image a person was viewing from brain scans.
• Limits of Interpretation: Measuring imagination or dreaming remains a massive challenge due to lower brain signal-to-noise ratios and less defined neural patterns.
• Across Individuals: Despite individual variability, certain neural regions (like the fusiform gyrus for faces) are shockingly consistent across people—even for culturally learned skills like reading, suggesting deep biological-programmed architecture but cultural adaptability.

“There is a surprisingly common structure across individuals in ways which raise questions... you have an area in the brain called the face fusiform gyrus which responds specifically to faces... But it also is the case for reading... This cannot be genetically programmed, right?”
— Jean Rémi-King (14:44)

---

4. The Boundary between Human & Machine: Comparing Brains and AI

• Model Comparisons: Researchers compare the activation patterns of AI systems (like large language models) step-by-step to brain activity in humans doing similar tasks.
• Emerging Convergence: Surprisingly, AI models trained on human tasks (even with minimal design for biological realism) produce internal representations comparable to those measured in human brains.
• Major Differences Remain: Training is far less efficient in AI—humans learn language from thousands of words, AIs need trillions of data points. Also, transfer to new tasks and learning from few examples (“one-shot learning”) remains much harder for machines.

“This simple task… pushes the algorithm to generate hidden latent representations which resemble those that we have in our own heads. And that suggests something to me which is very profound.”
— Jean Rémi-King (31:03)

“For the first time, we have AI systems... that we trained for a task... that generate representations which are comparable to those of the brain.”
— Jean Rémi-King (31:03)

---

5. Limits and Possibilities: Tech, Privacy, and Ethics

• Current State: True "mindreading" is, for now, limited: we can decode perceptions (what you’re seeing/hearing) under controlled lab conditions, but not personal thoughts, memories, or dreams in real time.
• Ethical Considerations: The potential for future misuse (e.g., privacy invasions, dystopian scenarios) is acknowledged; researchers and policymakers must develop safeguards now, before the tech is widespread.
• Application to Disability: Technologies could restore speech to paralyzed patients (via brain-to-text translation), already working in cases with invasive electrodes; non-invasive applications are the next frontier.

“What is possible today… is really limited to specific cases like perception... as soon as we try to do this in imagination… things become drastically more difficult.”
— Jean Rémi-King (47:58)

“Ethics… that scares people... When do you come up with those guardrails? Because if you come up with them after you’re able to do it, the horse is out of the barn.”
— Bubba Wallace (51:02)

“The risks seem limited [now], but technology continues to evolve and we want to make sure that the risks are limited. This is why we engage in these kind of discussions… not just within the scientific community, but with the rest of the world.”
— Jean Rémi-King (49:41)

---

6. Human Uniqueness: Creativity, Learning, and the ‘Noise’

• Why Can Kids Learn So Quickly? Human brains acquire language (and other skills) from limited data—a marvel neuroscientists are still trying to understand and which far outpaces current AI.
• The Mystery of Imagination: Not everyone can picture an apple (“aphantasia”); the diversity even in basic perception highlights both our cognitive limitations and the mystery of consciousness.
• Creativity within Noise: Neil deGrasse Tyson speculates that true human creativity may reside in the “noisy confusion” of the brain—territory AI may never access.

“Leaving me to wonder whether the true creativity of what it is to be human may actually lurk within the noise that can be never read by a machine... genuinely creating that which is human and can never be machine. I just wonder. That is a cosmic perspective.”
— Neil deGrasse Tyson (58:24)

---

Notable Quotes & Memorable Moments

• On the Ethics of Mindreading
  “It’s the ethics of being able to potentially... decode the brain’s messages and then reverse engineer it so you can read someone’s mind—that’s going to freak people out.”
  — Gary O’Reilly (47:18)

• Scary and Scary-Awesome
  “That is scary AF, okay? I mean, it’s fascinating and it’s really cool, but it’s also kind of scary... it kind of diminishes us as this crowning jewel...”
  — Bubba Wallace (29:47)

• Machines vs. Brains—The Learning Gap
  “If you show us a ball, you can show us one ball... and then you show us a basketball and we’ll say that’s a ball, show us a baseball, we’ll say that’s a ball. But the machine is like, ‘well I have never seen that before’. So that’s the difference.”
  — Bubba Wallace (35:58)

• Brain, AI, and Society
  “We have now some preliminary evidence suggesting that you have similarities between AI systems and the brain... So discovering what those laws are and trying to understand what is missing in AI systems for them to be as intelligent, as efficient as us remains a major topic of research.”
  — Jean Rémi-King (52:36)

---

Timestamps for Key Segments

• [03:34] – Guest intro: Jean Rémi-King, his background, and Meta’s FAIR Lab.
• [04:55] – How do we probe the brain’s information processing?
• [08:09] – FMRI vs EEG: what are we actually measuring?
• [11:57] – Can we reconstruct what people see from brain data? (AI and image decoding)
• [14:44] – Consistency of brain areas across people, and the philosophical implications.
• [21:34] – Decoding perception vs imagination: why imagination remains much harder.
• [29:03] – Comparing human nuance and intuition to AI representations.
• [31:29] – Surprising convergences (and inconsistencies) between deep learning models and brains.
• [33:21] – The speed of processing in brains vs machines.
• [38:04] – Nature versus nurture: are brain region specializations innate?
• [43:23] – Clinical neurotechnology: Restoring communication to paralyzed patients.
• [47:01] – Technology limits and ethical frontiers in mindreading.
• [51:17] – When (and how) should we develop AI neuroscience guardrails?
• [58:24] – Neil deGrasse Tyson’s “cosmic perspective” on human creativity and the limits of machines.

---

Episode Summary

This lively, thought-provoking episode of StarTalk situates listeners at the edge of neuroscience and artificial intelligence. Jean Rémi-King’s work highlights both the breathtaking progress—reconstructing visual perception from neural signals—and the immense hurdles that remain, especially for more abstract mindreading. The hosts probe ethical minefields, societal risks, and grand philosophical questions, all while keeping the tone approachable and often humorous. The state of mindreading AI is less “reading thoughts” and more “pattern matching” in tightly controlled settings; true telepathy remains far off, both technically and ethically. Perhaps, as Neil suggests, our deepest human creativity is still safely tangled in neural noise—at least for now.

---

End of summary.

Transcript

[Kristen Bell]

Hi, I'm Kristen Bell, and if you know my husband, Dax, then you also know he loves shopping for a car. Selling a car, not so much.

[Bubba Wallace]

We're really doing this, huh?

[Kristen Bell]

Thankfully, Carvana makes it easy. Answer a few questions, put in your VIN or license, and done. We sold ours in minutes this morning, and they'll come pick it up and pay us this afternoon.

[Jean Remy King]

Bye. Bye, Truckee.

[Kristen Bell]

Of course, we kept the favorite.

[Neil DeGrasse Tyson]

Hello, other Truckee.

[Kristen Bell]

Sell your car with Carvana today. Terms and conditions apply. Did you know Tide has been upgraded to provide an even better clean in cold water? Tide is specifically designed to fight any stain you throw at it. Even in cold butter? Yep. Chocolate ice cream. Sure thing. Barbecue sauce. Tide's got you covered. You don't need to use warm water. Additionally, Tide pods let you confidently fight tough stains with new coldzyme technology. Just remember, if it's gotta be clean, it's gotta be tied.

[Neil DeGrasse Tyson]

Gary, you're taking us inside the brain again.

[Gary O'Reilly]

I know. It's the inner space, and it's fascinating.

[Neil DeGrasse Tyson]

Is it as fascinating as outer space?

[Gary O'Reilly]

You'll argue it's not.

[Bubba Wallace]

Mm. I knew you were both going to say that.

[Neil DeGrasse Tyson]

Are you reading our mind?

[Bubba Wallace]

Absolutely.

[Neil DeGrasse Tyson]

Okay. An expert tells us the future of AI and reading your mind coming right up on StarTalk Special Edition. Welcome to StarTalk, your place in the universe where science and pop culture collide. Star Talk begins right now. This is Star Talk. Neil Degrasse Tyson, your personal astrophysicist, and I see to my right, Gary O'Reilly. That must mean it's Special Edition.

[Gary O'Reilly]

Yes.

[Bubba Wallace]

Gary.

[Gary O'Reilly]

Hey, Neil.

[Neil DeGrasse Tyson]

How you doing, man?

[Bubba Wallace]

I'm good.

[Neil DeGrasse Tyson]

Former soccer pro.

[Gary O'Reilly]

Allegedly.

[Neil DeGrasse Tyson]

Allegedly.

[Gary O'Reilly]

No.

[Neil DeGrasse Tyson]

Are you better here than you were when you were playing soccer?

[Gary O'Reilly]

As I get older, I. I do get better.

[Neil DeGrasse Tyson]

Good answer.

[Bubba Wallace]

That is.

[Neil DeGrasse Tyson]

Hey, as you get older, you get.

[Bubba Wallace]

I get older. That is what happens to me. That's about it.

[Neil DeGrasse Tyson]

So I'm looking at the title. You propose reading your mind.

[Bubba Wallace]

Ooh, and I thought this was a science show. I know.

[Neil DeGrasse Tyson]

Start the seance now.

[Bubba Wallace]

Right.

[Gary O'Reilly]

Buckle up.

[Neil DeGrasse Tyson]

Okay.

[Bubba Wallace]

I'm getting an M. And there's a. You have a relative somewhere.

[Neil DeGrasse Tyson]

You were in the hemisphere, right?

[Bubba Wallace]

You had a mother. Okay, sorry.

[Gary O'Reilly]

That's just it. You're not all right. AI will be driving our cars, our trucks, our trains soon enough. And probably, if not already, it will help us solve our everyday problem.

[Neil DeGrasse Tyson]

It already is.

[Bubba Wallace]

It has. Exactly.

[Gary O'Reilly]

And it'll probably solve some of our big problems. It may even help us tidy up some of the mess we've made over the years. But surely it's never going to be able to read our minds, is it?

[Bubba Wallace]

Hmm.

[Gary O'Reilly]

Well, actually, yeah, it can. And our guest today leads a research team using AI to decode the language of our brains. But before you start shouting at your devices, stop and think about the positivity that could come with this as a tool. But those who can think but not speak, who will get a voice. So for that, and if that happens, that would be truly amazing.

[Neil DeGrasse Tyson]

And the ethics of that too.

[Gary O'Reilly]

Absolutely. That's what I'm talking about. So if we would introduce our beast.

[Neil DeGrasse Tyson]

I'd be delighted to you.

[Gary O'Reilly]

Thank you.

[Neil DeGrasse Tyson]

Jean Remy King. Ooh, Jean Remy King.

[Gary O'Reilly]

Oh, you're gonna be saying that for hours, aren't you?

[Neil DeGrasse Tyson]

From P. France. Did I say all that right?

[Jean Remy King]

Absolutely. Perfect accent.

[Neil DeGrasse Tyson]

Welcome to StarTalk. Welcome to my office here at the Hayden Planetarium.

[Jean Remy King]

Thank you very much for having me.

[Neil DeGrasse Tyson]

And you work for Meta.

[Jean Remy King]

That's right.

[Neil DeGrasse Tyson]

Facebook, basically.

[Jean Remy King]

But absolutely, yes.

[Neil DeGrasse Tyson]

But Meta, I mean, I think it's.

[Gary O'Reilly]

Not just one singular.

[Neil DeGrasse Tyson]

It's not a thing anymore.

[Bubba Wallace]

It's Meta, right, yeah.

[Neil DeGrasse Tyson]

All right. You work for Meta in Paris. You have a background in neuroscience. I love neuroscience. We have neuroscientists on the show all the time.

[Bubba Wallace]

We really do.

[Neil DeGrasse Tyson]

We're all in the situation when we have a neuroscientist. And describe to us what your goals.

[Bubba Wallace]

Are aside from world domination.

[Jean Remy King]

So we have a lab at Meta which is called Fair for Fundamental AI Research, which is structured as an academic lab, in a sense. The goal is really to understand more about the principles of artificial intelligence. And within that lab, I'm working with a team that interfaces two disciplines, neuroscience on the one hand, and AI on the other hand, try to both better understand how the brain works and also try to perhaps improve AI algorithms. In light of these principles, how do.

[Neil DeGrasse Tyson]

You have any idea at all how the brain is processing information?

[Jean Remy King]

So we have tools for this, of course, in your science.

[Neil DeGrasse Tyson]

Tools, Tools. Interesting tools do you put on people's brains?

[Gary O'Reilly]

Oh, this is not hammers and chisels.

[Bubba Wallace]

Tools.

[Neil DeGrasse Tyson]

That's a euphemism for something. And I want to know what.

[Bubba Wallace]

Sure, yeah.

[Jean Remy King]

You have really a wide battery of tools that you can use. The one that we.

[Neil DeGrasse Tyson]

On the human brains.

[Jean Remy King]

On human brains, yeah. So the one we tend to use the most in the team are non invasive neuroimaging techniques. So from magnetic resonance imaging, like the big scanner you have in hospitals, to electron cephalography, these are the small nets that you can put on people's little caps.

[Bubba Wallace]

That you put on your head with all of the.

[Neil DeGrasse Tyson]

And how does that work? Electrodes, it looks like, for fields. And each of those electromagnetic fields that come through your skull.

[Jean Remy King]

That's right. So each of those work with different principles. So for eeg, for electroencephalography, and meg, the magnetoencephalography, you measure the fluctuations of electric and magnetic fields which are elicited by neuronal activity. So typically, thoughts, the biological instantations of thoughts. Yeah.

[Bubba Wallace]

So is every brain process you were.

[Neil DeGrasse Tyson]

I said something about the biological in. What's the word?

[Jean Remy King]

In sensation.

[Bubba Wallace]

In sensation of thought. So does that mean that every action in the brain has an electrical counterpart or so like the firing of a synapse? Is it actual electrical? You know, actually you have a connection.

[Jean Remy King]

A lot going on in the brain which is not electric or doesn't lead to electric fields. In fact, even the neurons which are firing, not all of them are being measured with EEG or meg, and we tend to only measure those that are spatially aligned. So in the cortex, which is the part of the brain which is folded, you have a lot of neurons, which we call pyramidal cells that tend to be positioned in the same way. So when they discharge electricity, their electric field can build up over space because they actually are aligned spatially. If they.

[Neil DeGrasse Tyson]

So it strengthens it, the signal.

[Jean Remy King]

Yeah. If they were facing any direction.

[Neil DeGrasse Tyson]

Get some canceling.

[Jean Remy King]

Yeah.

[Bubba Wallace]

You have no.

[Jean Remy King]

Average down to zero.

[Bubba Wallace]

Basically. You have no.

[Jean Remy King]

Because they all aligned with one another, then you can measure these electric fields at a macroscopic level, even with electrodes that are positioned on the scalp, so not inside the brain.

[Bubba Wallace]

That's amazing.

[Gary O'Reilly]

So you're in an FMRI and you're offering images functional magnetic.

[Jean Remy King]

Exactly.

[Gary O'Reilly]

Yes.

[Neil DeGrasse Tyson]

So that's like you are actively. You're awake talking to the person while they're messing with your brain.

[Gary O'Reilly]

Well, they're not messing with your brain. They'll offer you an image, and that then gets picked up through the data. But while you're offering an image to a patient, there's other noise.

[Neil DeGrasse Tyson]

When you declared something he hasn't declared yet, can we get him to say it first? Okay, when you read the brain, what do you see?

[Jean Remy King]

We see a lot of noise, but.

[Bubba Wallace]

Maybe just depends on who's. Yeah, I didn't say my brain. I said when you read a regular brain, what do you see? We see a lot of noise, but.

[Jean Remy King]

Just a clarification on the fmri. So FMRI is a different type of technology that does not pick up electric and magnetic fields like eeg and meg. It actually picks up a proxy of neuronal activity, which is the deoxygenation or the blood flow in the brain. So when neurons are active, they consume oxygen, and so you have a change in the vascular flow, which you pick up with ephemera.

[Gary O'Reilly]

So you're getting the geography of the brain as to what's happening and where.

[Jean Remy King]

Absolutely. And this is a very different type of signal that you would measure with EEG and meg, and it's very slow. So, of course, the blood comes only, like, it doesn't change every millisecond, let's say. And so you have a very different type of signal that you would observe depending on the device of choice, whether it's FMRI or EEG or MEG or intracranial recordings, when you can have access to this type of signals.

[Neil DeGrasse Tyson]

Intracranial means you actually have probes inside the brain. Inside the brain, Absolutely.

[Jean Remy King]

So this is very common.

[Neil DeGrasse Tyson]

People say, go ahead and do that.

[Jean Remy King]

No. So.

[Bubba Wallace]

No.

[Jean Remy King]

You do have patients, typically patients who suffer from intractable epilepsy, who needs to have the part of the brain which generates the seizures to be removed. And before doing this, it is common to have a procedure where you. Well, the neurosurgeons and the epileptologist decides to put electrodes inside of the areas which is believed to be pathological, in order to be sure that this is indeed the brain region that should be removed.

[Bubba Wallace]

Right. You don't want to cut out the wrong part of the brain.

[Jean Remy King]

Absolutely. And so these individuals typically would stay about a week in the hospital, during which these signals can be analyzed by neurologists. And during that week, you can ask them whether they would like to participate in, for instance, an experiment that involves, I don't know, story listening or watching a movie.

[Bubba Wallace]

We're already in your brain, so why not? I mean, we're already in here. You know, it's like when the mechanic goes, listen, I got to go in there anyway, so I might as well get the calipers done on the brakes, you know?

[Neil DeGrasse Tyson]

Yeah, yeah, right.

[Gary O'Reilly]

So when you're decoding the brain ways, whether it's blood flow or the magnetic fields, and you said there's noise, how is. How is your algorithm filtering out that, and how is it breaking down? Because you said there's different data. The way the data comes is different from an FMRI than it does from an meg. So how can you explain how the algorithm is reinterpreting that?

[Jean Remy King]

Sure. So maybe just to start with, the reason why I said that when you look at it. It looks like noise is because these signals are impacted not just by neural activity, but by a lot of different factors. So, for instance, magnetic fields are constantly evolving. I shouldn't try to say this in front of you guys, because you know more about this than I do, but we are in a flux of magnetic fields all the time. And the magnetic fluctuation that are being generated in the brain are extremely small, orders of magnitude smaller than the objects that surround us and move around when they have methodic parts. And so the signals that can be picked up are basically contaminated by all of these things. So when you look at the raw data, it's very difficult to guess anything. Actually, you would probably need to start to do the very same task again and again to try to average out the noise and start to see what is the average brain response.

[Bubba Wallace]

I'm really looking for patterns than anything else.

[Gary O'Reilly]

Then what better than to use AI to recognize patterns?

[Bubba Wallace]

That makes perfect sense.

[Neil DeGrasse Tyson]

Wait, so let's back up for a minute. I understand you can look inside someone's brain and see the image that they're seeing, as though you were somehow their eyes behind what the brain processed. Do I understand this correctly?

[Jean Remy King]

The goal is to try to understand how the brain represents perception. In the case of this experiment, you're alluding to, the individuals are typically watching images one at a time. Each image is lasting for about a couple of weeks.

[Neil DeGrasse Tyson]

Did you do this on mice before? You did it on humans, but you only saw big chunks of cheese.

[Jean Remy King]

Are you saying this because I'm a French researcher. We do not work with non human animals in our team. But of course, in neuroscience, you have a wide variety of approaches, and a lot of people are indeed working on the visual system, rather in macaques and mice. Mice are not so great for vision, but yes, they are a lot of.

[Neil DeGrasse Tyson]

Different parts, if I remember correctly.

[Jean Remy King]

Well, I'm not an expert in this, but I think they do see things, but they don't count on vision as much as we do.

[Neil DeGrasse Tyson]

Right, right.

[Bubba Wallace]

So what you're really doing is you're measuring these signals as a person is seeing something. And that what you're measuring, once you filter it, you're able to determine that this is the pattern. And if we match that from person to person, what are you measuring against is really my question.

[Jean Remy King]

So you really have two types of things, right? You have the images that you present to the participants, and you have the brain responses to those images. And the whole goal is to try to find the linking function between the two.

[Bubba Wallace]

Okay, so you could use the same person, actually, and just replicate that over and over again. If you keep seeing the same pattern, then you know, from this pattern that represents a sports car or a this or that, so you don't have to.

[Neil DeGrasse Tyson]

You're mapping the signals.

[Bubba Wallace]

You're mapping the signals.

[Neil DeGrasse Tyson]

Right, right.

[Bubba Wallace]

So wait a minute, here's, here's the real rub though. The human brain varies from person to person in not in its general regional response to stimuli, but it does vary in how we actually perceive things. So how do you make sure that what you're measuring in one person is actually going to be what you're measuring in another person? Like, if I were to lose my sight, my occipital lobe would go like dead. But, but other parts of my brain would take up that activity. And so you would be measuring a completely different data set because I'm blind, but in my mind I would still be seeing stuff.

[Jean Remy King]

So I think you're highlighting something which is actually an open question at the moment, which is the inter individual viability of neural representations.

[Bubba Wallace]

Yeah, that's what I meant.

[Jean Remy King]

So up to recently, most of the human neuroscience research was really trying to focus on what was common across individuals. So typically, the very sort of standard experiment is you take 20 or 40 participants like you and me, and you make them do a task for about an hour in the scanner and then you try to see whether their brain responds similarly to the same stimulus. For instance, if you present half of the images with faces and half of the images with houses, is it the case that the brain areas that respond to faces is similar across individuals? And the result is that there is a surprisingly common structure across individuals in ways which raise questions. For instance, you have an area in the brain called the face fusiform gyrus, which is an area that responds specifically to faces. And this area tends to be located in the similar part of the brain for every individual. And which is fine, you can say, okay, maybe genetically this was pre programmed. We have some neurons in the brain which are specifically tuned for this. But it also is the case for reading, for instance, for orthography. So if you present words, you can find that indeed some part of the brain are specifically responding to letters or the letters that you know or the words that you know. And this is, this tends to be in a brain region which tends to be similar across individuals, but this cannot be genetically programmed. Right. Because words is something that emanates from culture. This is a recent trait. So trying to understand why the same high level representations end up being represented in the same place in the brain is a major question. Now, having said that, the field is shifting towards more and more focus on individuals. And we do realize that indeed the representations are very specific to some extent to individual brains. And that so far we may have emphasized too much the similarity across individuals and not pay enough attention about the individual specificities.

[Neil DeGrasse Tyson]

But if you have to calibrate against the individual for the individual's thoughts, then you can't just come up to a stranger and know anything about them.

[Jean Remy King]

So we would. So for instance, we would know that auditory inputs, the sounds that come into your ear, tend to be processed in the same brain regions at first, right? It's not that the ear is connected to a random part of the cortex. It tends to arrive ultimately in the primary auditory cortex. And this would be common to most people, except if you have brain lesions or a variety of pathologies. And that would be the same for vision, and that would be the same also for the sense of numbers. For instance, if you have a sense of magnitude, this is typically hosted in the parietal cortex. And this tends to be the same across individuals. But as soon as you want to get more specifics, you want to really try to get a more fine grained level of representations, then this becomes really specific to individuals. And it's difficult indeed to transfer the knowledge that we observe from one participant to another. Buying a car in Carvana was so easy.

[Kristen Bell]

I was able to finance it through them.

[Jean Remy King]

I just.

[Kristen Bell]

Whoa, wait, you mean finance? Yeah, finance. Got pre qualified. Qualified for a Carvana auto loan, entered my terms and shot from thousands of.

[Jean Remy King]

Great car options, all within my budget. That's cool.

[Kristen Bell]

But financing through Carvana was so easy. Financed, done.

[Jean Remy King]

And I get to pick up my.

[Kristen Bell]

Car from their Carvana vending machine tomorrow. Financed, right?

[Jean Remy King]

That's what they said.

[Kristen Bell]

You can spend time trying to pronounce financing, or you can actually finance and buy your car.

[Jean Remy King]

Today on Carvana financing, subject to credit approval.

[Kristen Bell]

Additional terms and conditions may apply.

[Bubba Wallace]

Bubba Wallace here with Tyler Reddick. You know what's more nerve wracking than waiting for qualifying results?

[Jean Remy King]

Waiting for the green flag to drop.

[Bubba Wallace]

Instead of pacing, you rev up with Chumba Casino's weekly new releases. It's like a fresh set of tires for your brain. Play for [e-mail] let's Chumba. No purchase NECESSARY VGW Group Void we're prohibited by law.

[Gary O'Reilly]

CTNC's 21+ sponsored by Chumba Casino.

[Kristen Bell]

Big moments or small moments, they all deserve great style. At Maurices, we're here to help you show up feeling your best. No matter what's on your calendar, whether it's a date night, a special event, or just another Tuesday, our over 800 stylists are here to help. From tried and true denim to head turning looks, we've got style that fits your life and your budget. Visit your local maurices to find fashion fit for you, with a little help from us.

[Jean Remy King]

This is Ken the Nerd Neck Zabera from Michigan, and I support StarTalk on Patreon. This is StarTalk Radio with Neil Degrasse Tyson.

[Gary O'Reilly]

If we step back into the office, an image to a patient. How accurate now is your algorithm in terms of replicating as much of that first image? And how much does the algorithm say? Well, I'll take a calculated guess at filling in the blanks.

[Jean Remy King]

That's a very difficult question.

[Neil DeGrasse Tyson]

How many blanks are there to fill in? Yeah.

[Jean Remy King]

Because the metric that we use for evaluating how well we reconstruct the images in this case is not well posed. So if you take, for instance, a pixel level, you want to compare how good your image, the image that you manage to decode from brain activity is compared to the true image, you may get every individual pixel wrong because perhaps, I don't know, the color is slightly off and the objects are slightly to the left or to the right. And so you would have a very bad decoding metric. But if the image has the same content, if it's, I don't know, the true image has a horse, and you also decoded a horse, you don't want to say that this was a terrible reconstruction. You want to say, well, it's maybe not pixel accurate, but it tends to have the right concept. And so there is, for now, a difficulty in even quantifying the quality of the reconstructions. However, what is striking is to see that when you have a lot of hours per participants, typically 20, 40 hours per participants of them just watching images in the scanner, and you have a very good scanning technique, like ultra high field data.

[Gary O'Reilly]

Yes.

[Jean Remy King]

Yeah. This would be a huge amount of data for neuroscience, for physicists, the universe.

[Bubba Wallace]

Is bigger than your brain. I was going to say they're only mapping the entire universe of which your.

[Neil DeGrasse Tyson]

Brain is a part.

[Jean Remy King]

So once you have a lot of data per individuals, then you can really start to reconstruct what they perceive in a surprisingly accurate way. However, going beyond perception currently remains very difficult.

[Gary O'Reilly]

Okay, so if you've offered an image to a patient, you get a certain set of data back depending on the subject matter of that what's the difference if the patient is asked to imagine an image and do you get a.

[Neil DeGrasse Tyson]

Variant seeing it through your senses?

[Gary O'Reilly]

Yes.

[Bubba Wallace]

Yeah.

[Gary O'Reilly]

Which mind's eye, for want of a better term.

[Jean Remy King]

Right. So in the case of perception, this is where the most progress has been made. So when you watch an image or when you hear a sound, it is becoming increasingly easy to decode what the person has seen or has heard. However, when you do the same type of tasks, but on imagination, you can get performances above chance level from a statistical point of view. But frankly it's not very convincing to anyone who don't want just to look at the stats and just want to see the reconstruction. The reason for this is. Well, there are two reasons. The first reason is that the signal to noise ratio in imagination is much lower than in perception. So when you look at the brain signal, on average they are weaker when you try to imagine, let's say, an apple.

[Neil DeGrasse Tyson]

Some people have vivid imaginations though still.

[Jean Remy King]

And I don't think we know this. So I think this trying to evaluate whether the people, for instance, who claim not to have any visual imagination, indeed do not have representation that would be decodable at all.

[Neil DeGrasse Tyson]

Because I just learned days ago that a colleague of mine, he went around the room and said, Matt, picture an apple in your head. Picture an apple. Okay, Picture an apple. He can't picture an apple in his head. Wow, right. Is this some rare.

[Bubba Wallace]

Not even the computer.

[Neil DeGrasse Tyson]

He cannot conjure an image on command in his head. We all thought of apple, red apple, apples. Right.

[Gary O'Reilly]

But any image on demand.

[Neil DeGrasse Tyson]

Well, he used that as a simple one, but I. So I didn't know this was an issue.

[Jean Remy King]

Yeah, I think this is actually quite common. I am not an expert in this, but I think that the term is aphantasia. I think this is something which is more than 5% of the population, I think that claim not to be able to. To imagine visually objects so they don't become artists. I don't think artists are restrained to just imagining objects in the head. You have musicians that may not engage in this monology.

[Gary O'Reilly]

How much more in terms of a percentile do you think your research is going to take to how the brain interprets images?

[Jean Remy King]

This is a very difficult question. Again, the.

[Gary O'Reilly]

Sorry, the question was easy.

[Neil DeGrasse Tyson]

Is your answer that's difficult?

[Jean Remy King]

Probably. I don't know about our research specifically, but what is clear is that there is a huge progress which is being made thanks to AI, but not as a tool like you would see in other sciences. So for instance, in I Don't know. In biology, in cosmology, in sciences where you have a lot of data, you use AI as tools, you have a lot of numbers, you don't know how to crunch them. You train a system to do whatever you want you're looking for, and it helps you process this data. In neuroscience, we also do this and the pattern matching that we discussed earlier, but we also use this as a modeling framework because the AI system, in a sense, is also trying to do something that we do. We train AI system to perceive the world, to try to recognize objects, to reason upon the world, to discuss with us in, in a linguistic form. And so this creates basically systems that can then be used as models of how the brain works. This is really accelerating, I think, the understanding of how the brain functions.

[Gary O'Reilly]

So you talked about linguistics there. If you presented a sentence to a patient, then you're going to have that sort of perceptual stage of where they perceive what the sentence, they see the sentence. Then you go through what they call a lexical stage and then a contextualization stage.

[Bubba Wallace]

That all makes sense.

[Gary O'Reilly]

Good.

[Bubba Wallace]

Are you able, basically how we communicate.

[Gary O'Reilly]

I know, but are you able to.

[Neil DeGrasse Tyson]

Get the algorithm, seven syllable words, though?

[Gary O'Reilly]

Are you able to get the algorithm to feel the nuances of the brain and actually see how that breaks down?

[Neil DeGrasse Tyson]

Is that just the future?

[Bubba Wallace]

I'm waiting for this answer.

[Jean Remy King]

Maybe I can say how we do this in the first place, so we can have individuals like you and me. And I'm often a subject of my own experiments, going to the scanner and reading a sentence. And so you flash a sentence word by word, once upon time and so forth. And for each millisecond you can see, okay, what is the brain activity now? What is the brain activity now? So you end up with an activation pattern associated with each moment of time. And that you can time lock to words or to syllables, phonemes. And then you can do this same approach in the AI algorithms. You can present a sentence. And deep learning networks nowadays have activation patterns inside of them which are known to be difficult to interpret. But nevertheless, we can do the same trick. We can time lock the activations of the deep nets in response to words, syllables, and so forth. And then we can do the comparison between the activations of the AI systems to the activations of the brain. And we don't know what these two things represent, but we can still try to do correspondence to try to see whether they tend to be similar in the geometrical structure that they hold. And what we observe is that this helps Us decompose the stages of processing that you mentioned. So we can, firstly, that you have algorithms that are trained to do visual processing, but know nothing about words, about language, that you can map and correspond to the activations of the perceptual system. And then you can do the same type of comparison with an algorithm which this time is not trained to recognize images or pixels or to transform pixels, but is trying to analyze words and combine them together. And you will see that the activation patterns of these algorithms that are processing things at a language level and not at a perceptual level, they do have activation that corresponds to other brain regions and other time moments. And so we can try to do this sort of one to one correspondence between the model and the brain to try to understand the structure of these representations.

[Bubba Wallace]

And where exactly in that process do you get the language model to, I'll say, mimic perception and the nuance that we have, which is experientially based. So when you look at once upon a time, there is an activation pattern, Right?

[Jean Remy King]

Right.

[Bubba Wallace]

And you can replicate that activation pattern in the AI, but what you can't do in the AI is replicate all the different things that once means to you. I went to the movies once. Really? Only once? You went to the movies once upon a time. I know that as the beginning of all fairy tales. So it brings in a completely different contextual meaning. So where along that line of comparison do you get to interject what we do that machines don't, which is intuit and find nuance?

[Jean Remy King]

Right? That's a great question. And maybe I should emphasize one thing, which is that when we do this comparison, we don't actually train or tune the algorithms to resemble the brain. We don't actually try to inject this knowledge. We just have these AI algorithms that we can use off the shelf open source models, either produced by our colleagues or by the rest of the scientific community. And these algorithms, they're not trained to mimic the brain. They're trained for whatever other purposes, to be chatbots and to recognize cats from dogs in images. But what we observe empirically is that training these algorithms tend to make them generate representations which are comparable to those.

[Bubba Wallace]

Of what we do in our brains. Okay, first of all, that is scary af, Okay? I mean, it's fascinating and it's really cool, but it's also kind of scary.

[Neil DeGrasse Tyson]

Tell them what AF means.

[Bubba Wallace]

It's scary as. But the reason why it's a little scary is because on the one hand, it kind of diminishes us as this crowning jewel in all of creation with the zenith of intellect, that we believe.

[Neil DeGrasse Tyson]

That we're the zenith of anything.

[Bubba Wallace]

Right. That's what. Yeah.

[Gary O'Reilly]

Couldn't we do with a little bit of humbling every now and again?

[Bubba Wallace]

I don't know about you. No, no.

[Kristen Bell]

Here you go.

[Neil DeGrasse Tyson]

Here's how you get out of that. Here's how you get out of it.

[Bubba Wallace]

Go ahead.

[Neil DeGrasse Tyson]

We are so brilliant, right. We created something more brilliant than ourselves.

[Jean Remy King]

So I wouldn't say this quite, quite yet, because AI is really limited in many ways today. In spite of the hype, I understand the emotional reaction, but frankly, I also think that there is a source of marvel here. For the first time, we have AI systems or systems that we trained for a task. The task is surprisingly arbitrary or even mundane. For instance, trying to predict the next word given the preceding words. That sounds like.

[Bubba Wallace]

I mean, that is what all LLMs do.

[Jean Remy King]

Exactly.

[Bubba Wallace]

Yeah.

[Neil DeGrasse Tyson]

Large language models.

[Bubba Wallace]

Thank you. Large language model.

[Jean Remy King]

And this simple task pushes the algorithm to generate hidden latent representations which resembles those that we have in our own heads. And that suggests something to me which is very profound. Right. Which they exist. General principle that push these systems, biological, artificial systems, to generate a similar computational path, a similar set of representations.

[Gary O'Reilly]

So is there a similarity between the brain and how it processes data and its architecture and that of a large language model that it's learning in a very similar way to the human brain?

[Neil DeGrasse Tyson]

Because, as I understand it, the original idea of neural nets as invoked in computers was an attempt to mimic what we thought our brain wiring was doing. And we learned that that's not really how our brains work. So it's just dangling there now as its own thing with its own utility, but it's no longer biologically biological analog.

[Jean Remy King]

Yeah. So the history of AI and neuroscience intertwined quite a lot, but for a long time, it was. These links were metaphorical, like the idea of a neural network was. It wasn't, I think, a useful concept. But the goal was not to be as close to the brain as possible. In fact, it's really a huge simplification, this idea of artificial neurons as compared to what was already known at the time. And you've had these bridges between the two disciplines for over many decades. What is different now is that this comparison is not just conceptual kind of loose, it's very precise. We can't quantify the extent to which the activation patterns in the brain and the activation patterns in this AI systems do look alike or not. And even though these systems are not built for that purpose. Now, having said That I also feel the need to mitigate these results because this is a tendency that we have. But we also see a lot of edge cases where this does not work. So typically, if you take the very best model, the largest model, this similarity tends to break down. So we do have cases where what we call the convergence of representations between AI systems and the brain is not monotonous. It's not systematically the same.

[Gary O'Reilly]

All right. One thing I haven't sort of got to grips with, the speed at which image back of brain and how quickly that is and how quickly you're able to then process that data back through an algorithm.

[Jean Remy King]

In the head? Yeah, in the head. It's quite slow, actually. So when you look at reading, for instance, you flash a word onto your retina. This takes for about 70, 100 milliseconds to really blow up the visual cortex in the occipital lobe. And from there, you'll get another 50 milliseconds for this visual information to be processed.

[Neil DeGrasse Tyson]

The milliseconds? A thousandth of a second.

[Jean Remy King]

A thousandth of a second.

[Neil DeGrasse Tyson]

So 50,000th of a second would be 500ths of a second? Yeah, that's correct. Okay.

[Jean Remy King]

I'm not good with math, especially not in my native language.

[Neil DeGrasse Tyson]

Well, you're investing in my head if you're not good with that.

[Jean Remy King]

But. So, yeah, around 100 milliseconds, this is really when the activity peaks in the visual cortex for the sensory processing, let's say. And then this information is being analyzed into edges that will eventually construct the representations of letters and of morphemes of words. And this is around 200 milliseconds. So 1/5 of a second, and then it takes another 200 milliseconds, so around 400 milliseconds. And does the semantics part of words really rise in the brain and is broadcast to a wide variety of brain regions? And so this process is relatively slow. It takes about half a second for you to analyze the sensory.

[Bubba Wallace]

How fast would machine learning do it? Or if, let's say, an ocr, how fast would it know?

[Jean Remy King]

Yeah, in terms of inference, the machine would be much faster. It would be just a few milliseconds with a gpu.

[Bubba Wallace]

A few milliseconds to do the whole process. And we take like half a full half a second.

[Jean Remy King]

Absolutely.

[Bubba Wallace]

So we're basically like, duh, well.

[Jean Remy King]

At the inference stage.

[Neil DeGrasse Tyson]

So what, we stop giving AI ideas about what to do with us when it becomes our overlord?

[Jean Remy King]

What is fast is what we call the inference stage. Right. So once you Already trained the algorithm using it is actually very fast. What is typically slow is loading the information onto the graphic card, but once it's there, it's actually very fast. However, training these algorithms is ridiculously slow. Right. If you want to train an LLM today, a large language models today you need trillions of words, which represents many, many, many lifetime of just reading all of the text that we created in humanity.

[Bubba Wallace]

That takes us back to what we were talking about earlier. So in order for it to know, it has to see all the words. In order for us to know, we just have to see like a word and then something similar and we're like oh yeah, it's that, you know, so that's what we're for instance, a ball. If you show us a ball, you can show us one ball. We've never seen a ball before in our life and you show us a ball and then you show us a basketball and we'll say that's a ball. And then you show us a baseball and we'll say that's a ball. But the machine is like, well I have never seen that before. So that's the difference.

[Jean Remy King]

Yeah, this is one of the many differences. I mean when we emphasize similarities, we emphasize the similarities because we are in a field of differences. Everything is different. The architecture is different, the type of data that they receive is different. The training, of course the physical instantiation is different. This is also highlighting why we are all the more surprised and interested in the fact that in spite of all of these differences, we can still find similarities in the way they process information.

[Neil DeGrasse Tyson]

Wow.

[Gary O'Reilly]

And you've got, when you've got these sort of caps that you're putting on, they're sensitive enough to be able to operate at that sort of speed. I know you say it's slow, but that for me is really quite fast.

[Jean Remy King]

So for, for it depends on the device. So with functional magnetic resonance imaging fmri, you get a snapshot of brain activity approximately every two seconds. So a lot can go on within two seconds. However, if you take magnetomes and photography, you can get a snapshot every millisecond. So you'll get a much more well resolved signal in time. But the spatial resolution now is much lower. So you tend to have a blurry image, let's say of, of brain, brain activity. So you have a trade off between, between these different technologies.

[Bubba Wallace]

Wow. So the kind of like cosmology, the better your like looking tools get, the.

[Neil DeGrasse Tyson]

Easier general truth of science, it's just.

[Bubba Wallace]

Going to be so much easier for you to figure out anything you need to figure out. It's just a matter of we got able to see it faster and then know more clarity.

[Neil DeGrasse Tyson]

I got a question. When I think of the brain, I think of it as this organ. And there are these parts of the brain that are similar from one person to another, even if there's differences in detail. Are we to believe that the brain knows that in advance how it would divide up its territory? Or are we all just socialized the same way? We all grow up in a civilization, and so we all have the same influence on our developing brain for it to take the shape the way it does?

[Jean Remy King]

So this is a very profound question. There is a tension. I'm not a historian of science, but there is a tension in the field that dates back to philosophy between empiricists, people who think that the structure of representations come from the data to which you are exposed, and rationalist, the people who would rather emphasize the importance of innate representations and innate structures. So you have, for instance, Plato, on the one hand, if we take this back to ancient Greece, there would really be in the rationalist point of view with this idea that there exist innate representations in ourselves and ultimately we can approximate them with reasoning, whereas other people. And I think that the whole study of AI is really on the extreme empiricist side, is just let's take a blank system and just press a lot of data onto it, and ultimately the system will manage to perform a task. And what is interesting nowadays is to see that irrespective of whether the representations are innate or acquired through exposition, not necessarily culture, but even just sensory data, they seem to at least have some similarities. This is what I think is interesting in the case of this comparison between AI and the brain for language. The brain is obviously very structured very differently to these AI algorithms. And obviously there must be some innate structuring in our brain. This is why only human have language, in a sense of being able to combine words together in order to reason and to communicate.

[Neil DeGrasse Tyson]

This is not a New Yorker comic too. Dolphins swimming together, right? They're in a water show, okay?

[Bubba Wallace]

Like the seaworld type.

[Neil DeGrasse Tyson]

Seaworld. They swim together. One says to the other of the humans, right? They face each other and make sounds. But we're not sure they're actually communicating.

[Bubba Wallace]

Right? That's pretty funny.

[Jean Remy King]

But there are a lot of experiments.

[Neil DeGrasse Tyson]

Their brain's bigger than our brain.

[Jean Remy King]

Yeah, for some of them, not all of them, but. So the reason why there is. I mean, there's been a lot of experiments on behavior with dolphins. But also with apes to try to see whether they would be able to combine concepts. And there are some experiments that show that in some edge cases they are able to do this. But for now, we don't have any evidence suggesting that you have any other species that can learn this vast amount of concepts and be able to combine these concepts together in order to produce a sentence or to understand a sentence, a new meaning that they've never heard before. So this ability must be to some extent input in our genome and be an innate, an innate structure.

[Bubba Wallace]

Well, I have to be. I mean, we're the only you. And it's so funny because it's disassociated from everything else that we are and have language. For instance, I can be deaf, dumb and blind and you can teach me any language. I don't have to have an actual reference like everybody else does. So, I mean, we are truly unique in the way that we do communicate. I don't know if, I mean, I'm sure other animals communicate too. I'm not sure. Yeah, all animals communicate. Yeah, all animals communicate. But we're very unique in that if I don't know how to communicate with somebody I meet from halfway across the world, we will find mediums that allow us to know each other's language.

[Jean Remy King]

Right. And this is coming back to the whole empiricist versus rationalist tension. This is why there is something very interesting here. So we established that the human brain must have some genetic or innate properties for it to acquire language. And this is why it differentiates itself in part from other animals. And we also know that these deep learning algorithms, they have very little what we call inductive biases. The architecture that we use in deep learning, they are remarkably blank and versatile. And so it's really the data with which they are trained that pushes them to build the representation that they have. And nevertheless, it seems to be comparable, at least to some extent, to those of the brain. Not in every way, but in some ways. And so that suggests that no matter where you come from, whether you come from this really rationalist type of approach to cognitive science or much more from an empirical empiricist, sorry, approach, there seems to be some sort of convergence between these two approaches. What.

[Gary O'Reilly]

But what I want to get into now is the application of your research, where it could go as we progress with this. Now, I said in the opening about people who can think but can't speak. Is there an opportunity with this research to give them a voice, to have their understandings made public, made aware?

[Jean Remy King]

Right. So you have indeed a lot of patients who suffer from an inability to communicates, typically because of a brain lesion. So either a traumatic brain injury or an anoxia that will lesion the part of the brain which is responsible for for instance motor control. So they will be paralyzed or lose an ability to move their facial movements. And there are now a few team that have shown that it is possible to put a set of electrodes in the motor cortex and to use these neural signals to feed in algorithms that can then be used to do a brain to text translation and allow the individuals to regain communication abilities. So this is already something which is happening with invasive approaches. So with electrodes which are implanted with neurosurgery. One of the goal of course is to try to see whether it would be possible to push this approach with non invasive devices which do not require a brain surgery in order to rehabilitate communications in patients, but also perhaps to diagnose. So sometimes you have patients who do not respond but they are awake. It's a paradoxical state which occurs sometimes after a coma. And in these patients you want to know whether they don't communicate because they're not conscious of the environment, or whether they don't communicate because they are fully paralyzed, for instance, or they just don't like you and they just don't want to, which is actually an issue. Right. See if you have lesions to the part of the brain that are intrinsically linked to motivation, that could also be a cause of, demotivated, of a lack of.

[Bubba Wallace]

I'm tired of talking to you, I'm just done.

[Jean Remy King]

And so for these patients, having devices that would allow us to, well, allow them to communicate, but even to allow us to know whether indeed they are conscious or not of the environments is certainly of a prime use.

[Gary O'Reilly]

We likely to find that sort of ability in the near future. Or are we having to wait for invasive electrodes?

[Jean Remy King]

This is already happening.

[Bubba Wallace]

You know our boy Elon, that's what he wants to do. Neuralink, Yeah, I want to put a link, I want to put a chip in everybody.

[Neil DeGrasse Tyson]

You can do that through the vaccines?

[Bubba Wallace]

Well no, that's Bill Gates. Let's get our billionaires straight, okay? Elon wants to put a chip in your head. I mean an electrode in your head. Bill Gates already did it. Businesses that are selling through the roof, like Untuck it, make selling. And for shoppers, buying simple. With Shopify, home of the number one checkout on the planet. And with shop pay you can boost consumer conversions up to 50% businesses that sell more sell on Shopify. Upgrade your business and get the same checkout untuck it uses. Sign up for your $1 per month trial period at shopify.com podcast free. All lowercase go to shopify.com podcast free to upgrade your selling today. Bubba Wallace here with Tyler Reddick. You know what's more nerve wracking than waiting for qualifying results?

[Jean Remy King]

Waiting for the green flag to drop.

[Bubba Wallace]

Instead of pacing, we rev up with Chumba Casino's weekly new releases. It's like a fresh set of tires for your brain. Play for [e-mail] let's Chumba no purchase necessary. VGW Group Voidwear prohibited by law CTC's.

[Jean Remy King]

21+ sponsored by Jumba Casino Big moments or small moments?

[Kristen Bell]

They all deserve great style. Our over 800 stylists are here to help you look and feel your best. Visit your local Maurices to find fashion fit for you with a little help from us.

[Gary O'Reilly]

I think the limits of the technology, as pointed out, will be reduced because of AI, and they'll find solutions sooner. But it's the ethics of being able to potentially sort of decode the brain's messages and then reverse engineer it so as you can read someone's mind, it's the ethics of that being possible. Because I think that's going to freak, not just chuck out.

[Bubba Wallace]

You know, I'm going to be honest though, isn't that already happening? When you look at metadata that's taken from our phones and our location and other phones that are around us, couldn't you pretty much tell what I'm thinking?

[Jean Remy King]

Well, I don't. I don't know. But what I can say is these are certainly topics that come up very often and there are several things to say. The first thing is that what is possible today in terms of decoding brain activity is really limited to specific cases like perception and motor control. And the reason for this is because we know what the person sees so we can attach the image to the brain patterns. However, as soon as we try to do this in imagination, for instance, as we mentioned before, then things become drastically more difficult. Not just because of an inability of the algorithm to work, but really because a signal is just not there.

[Neil DeGrasse Tyson]

That means it's not likely that you will anytime soon read someone's dreams until.

[Bubba Wallace]

You get a signal booster.

[Jean Remy King]

From a statistical point of view, for fundamental research, there is research on the science of dreams. However, there is all of the evidence point out to the fact that it will be very difficult with the state of Knowledge that we have to have a device which can read your mind in the way that people think, like with your train of thought and all this. And the reason for this is because even with the largest multimillion dollar type of devices that are being used, the signals remain extremely noisy and it's very difficult to, to go beyond this. So the physics of the signal that we pick up is really the main constraining factor, not the AI algorithm part. So the AI algorithms can be used as a useful tool, but in terms of the signal that you can pick.

[Bubba Wallace]

Up, you can't generate the input necessary for them to do a good job.

[Jean Remy King]

Yeah, the data that can be collected with these devices remains extremely, extremely noisy. And so from that point of view, the risks seem limited. Now, this is the current state of affair, but our role here as scientists is also to say what is possible, what is the state of the art, and to share this through the research, through open sourcing and all this. That's the reason why we do this.

[Neil DeGrasse Tyson]

Work openly in science. You're always limited by your signal to noise.

[Jean Remy King]

Absolutely.

[Neil DeGrasse Tyson]

You'd have to add up days, weeks, months of measurements to pull a signal out of that noise.

[Jean Remy King]

But you want to have. Then this only works if you have the same signal that comes up again and again. Whereas when people think of mind reading, they think of reading the mind at a given instant, you don't think of the same thing again and again and just repeat this until your noise average is down. So this is why currently all of the evidence suggests that there is not a systemic risk. However, technology continues to evolve and we want to make sure that the risk are limited. And this is also why we engage in these kind of discussions, of course, to ensure that the discussion does not just happen within the scientific community, but with the rest of the.

[Bubba Wallace]

So when is the time to make that determination? Is it now, before you actually have the equipment to measure this?

[Neil DeGrasse Tyson]

What's the determination?

[Bubba Wallace]

The determination as to the ethics, like codifying the ethics themselves. Guardrails going forward. When do you come up with those guardrails? Because. Because if you come up with them after you're able to do it, it's the, you know, the barn is. The horse is out of the barn, as they say.

[Jean Remy King]

Yeah, absolutely. So this has already started. Right. There is already a lot of regulations on what you can and cannot do. For instance, I work in France and so we have the GDPR in Europe. It's constrained the way the data that is being collected from brain imaging can be used. In France, for instance, you're not even allowed to do neuromarketing. You're not allowed to use use brain data for marketing purposes. So this discussion is obviously already engaged and along the way we need to continue and update these decisions with the state of knowledge that continues to evolve. Yes.

[Gary O'Reilly]

What was the movie Minority Report where they had this sort of precognosis.

[Bubba Wallace]

Precogs.

[Neil DeGrasse Tyson]

Yeah, Precogs.

[Bubba Wallace]

That's a great movie.

[Gary O'Reilly]

That's everyone's default thought as regards this research and where it leads to. And I think that's what scares them. And I think they'd be grabbing not just for the guardrails. I mean, do you look well, precogs?

[Neil DeGrasse Tyson]

You were not digging out of their head what they saw from the past. You were digging out of the head what they foresaw in the future.

[Bubba Wallace]

In the future.

[Neil DeGrasse Tyson]

Right. So that was different. So they would see you committing a crime that you haven't committed yet.

[Bubba Wallace]

You haven't committed yet, but you were definitely going to commit.

[Neil DeGrasse Tyson]

And then they'd just go arrest you.

[Bubba Wallace]

Right.

[Neil DeGrasse Tyson]

They started doing that. The crime rate went to nearest.

[Bubba Wallace]

How about immigration in America right now? Oh, how interesting.

[Neil DeGrasse Tyson]

You pre arrest people. Just pre arrest people. So what is the end game for Meta in this?

[Jean Remy King]

I actually don't know why Meta hired me in the first place. I can only tell you what we're trying to do within our team. So the goal here is. Well, the goal is well posed. Right. We have now some preliminary evidence suggesting that you have similarities between AI systems and the brain. And that suggests something which to me is very intriguing, that there exists these general principles that shape the information processing in AI system and the brain. So discovering what those laws are and also trying to understand what is missing in AI systems. For them to be as intelligent, as efficient as us, remains a major topic of research. So this is why we're pushing on this frontier to better understand the brain and make better AI algorithms.

[Gary O'Reilly]

So if you're able to achieve that, people are going to feel an invasion of privacy. They're going to feel thought security becomes potentially compromised. You said there's discussion over the ethical point of view. Are we looking at those sort of features as well?

[Jean Remy King]

So, yeah, it's the same topic that we briefly discussed before. We have an ongoing discussion to try to see whether AI and neuroscience developments are changing the risks associated with, for instance, mental privacy. As of now, this discussion is ongoing, but I don't think we have a change in the technology that changes the risk. What we observe is that it is possible to decode brain activity in certain cases, typically for motor control or for visual perception. But it is not possible to decode what you are thinking at a given moment or your train of thoughts or to extract your password from brain activity. The reason for this is because the.

[Neil DeGrasse Tyson]

Signal that we have, it just takes the password out of your head.

[Bubba Wallace]

Right. Like the readers that they have now, they steal your credit card from you.

[Gary O'Reilly]

Yes. Radio frequency.

[Jean Remy King]

All of the physics on which we base this analysis prevent us to work outside of the lab. Right. So with an mri, you need to plunge someone into a very high magnetic field. This is not something that can be translated for, I don't know, a consumer product.

[Bubba Wallace]

But what you. You could envision is a dystopic future where the state, who has the power and the money to actually have a machine that could read your brain and during an interrogation, extract information from you that you don't want to give up. You know, basically, like, you know, you violated my mental privates. So, you know, that's actually foreseeable based on just what we talked about today.

[Jean Remy King]

Yeah. I mean, if we go down to the dystopic possibilities, I suspect that the states will not need an MRI to force you to give away your password. But it is an important, good point.

[Bubba Wallace]

I'm just like, I'm not going to your MRI machine. I refuse. Just like, yeah, okay, yeah, this baton says different.

[Jean Remy King]

But still, if the risk does exist, we should try to characterize it to ensure, like, what is the path to this risk? And this is part of the scientific enterprise, too. Yeah.

[Bubba Wallace]

Cool, man.

[Gary O'Reilly]

You've got some new research that you're about to release into the public domain. Can you sort of expand upon that for us, please?

[Jean Remy King]

Sure. Yeah. So far we've done this comparison between AI systems and the brain with adult participants. And to some extent, this is frustrating because there is something which is missing here in the picture, which is the learning process. Right. So in the case of language, we don't just want to understand how the brain process language, but we want to understand what makes it able to acquire it so efficiently. Like, with just a few words, we acquire language. The average number of words that we hear is typically around a few thousands per day, a few dozen of thousands per day. And. And if you compare this amount of data to the data which is input for the training of AI models, this is really a droplet of information compared to the oceans of data that these algorithms use. And so what this means is that fundamentally the architecture or the training principle that we use for AI, they are really, really mediocre we need to understand much better how you can get to a system that learn, learns much more, much more efficiently.

[Bubba Wallace]

So if you train your AI on children, you may end up learning how we actually learn or acquire language. But then you're also going to have AI saying things like, I hate you so much. I hate you. You never let me do anything.

[Jean Remy King]

But certainly it would be important to understand, not necessarily to train AI models with this data, but to understand the principles that allow young children to acquire language so efficiently. This is one of the, of the big marvels of our species and this is certainly what we're trying to do. So this is actually a work with a hospital, the Rothschild Hospital in Paris, that has a unit for epilepsy and young patients, down to two years old patients. So you have these patients who suffer from intractable epilepsy. Again, same patient. As we mentioned before, we have electrodes that are implanted inside the brain in order to identify the location that is generating the seizure. And who can stay for about a week in the hospital and listen in that context to audiobooks. And then we can time lock the brain responses to each individual word to try to understand how the representations of language are processed in these young patients and how this evolves with age.

[Neil DeGrasse Tyson]

Let me see if I can offer a perspective here. I'm as big a champion of AI as the next person, but I still enjoy being human. And whatever I can do to distinguish being human from a machine, I will embrace. Leaving me to wonder whether the true creativity of what it is to be human may actually lurk within the noise that can be never read by a machine. The first person to paint an impressionistic representation of reality. Could a machine have had that first thought? Or is that a human being rummaging within the noisy confusion of our own brains, pulling out something that no one had done before, no one had imagined before, and in the end, genuinely creating that which is human and can never be machine. I just wonder. That is a cosmic perspective.

[Bubba Wallace]

And that was beautifully said, except the first impressionist was just some dude who was nearsighted. Yes. I was just friends. Just how he saw the world, people were like, what an incredible interpretation. He's like, what are you talking about?

[Neil DeGrasse Tyson]

So thank you for visiting.

[Jean Remy King]

Oh, yeah.

[Neil DeGrasse Tyson]

All right. Chuck, always good to have you, man.

[Bubba Wallace]

Always a pleasure.

[Gary O'Reilly]

Gary, Pleasure, Neil, thank you.

[Neil DeGrasse Tyson]

Thanks for. To you and, and Lane and others.

[Gary O'Reilly]

Yeah.

[Neil DeGrasse Tyson]

Coming up with these topics.

[Gary O'Reilly]

They keep coming up with them. So we're going to keep finding them.

[Neil DeGrasse Tyson]

We chase them down.

[Jean Remy King]

Yep.

[Neil DeGrasse Tyson]

All right. This has been startalk Special Edition Neil DeGrasse Tyson, your personal astrophysicist, do keep looking up and in.

[Kristen Bell]

Did you know Tide has been upgraded to provide an even better clean in cold water? Tide is specifically designed to fight any stain you throw at it, even in cold butter. Yep, chocolate ice cream. Sure thing, barbecue sauce. Tide's got you covered. You don't need to use warm water. Additionally, Tide pods let you confidently fight tough stains with new coldzyme technology. Just remember, if it's gotta be clean, it's gotta be Tide.

[Bubba Wallace]

Summer's heating up and so is the action with chumba casino and 23xi racing. Whether you're trackside with Bubba, Riley and Tyler or cooling off at home, the fun never stops at Chumba Casino, the online social casino packed with free to play games like slots, blackjack and more. Jump into [e-mail] and score your free.

[Gary O'Reilly]

Welcome bonus 2 million free gold coins.

[Bubba Wallace]

And 2 free sweeps coins. No purchase necessary.

[Jean Remy King]

VGW Group Void where prohibited by law.

[Bubba Wallace]

CTNC is 21 + sponsored by Chumba Casino.