0:00

for the show comes from Roomba Robots.

0:03

Ever since humans first invented bolts

0:05

and dented slabs of metal, we've

0:07

been dreaming about robots. Machines

0:10

that can build our cities, conquer

0:12

new worlds, mine our precious diamonds.

0:15

Stuff like that. Turns out, we can do

0:17

all of that ourselves. What we can't

0:19

do is keep our floors clean. That's

0:22

what we really need help with. But hey,

0:25

stop looking at the stars. Progress is

0:27

right here at your feet. The Roomba

0:29

Robot is here to help. iRobot says

0:32

their reimagined Roomba 205 combo gives you

0:34

60 days of hands -free cleaning, all

0:36

without a dock. And they say the

0:38

Roomba Plus 405 combo can vacuum, mop,

0:41

and then empty its own bin and

0:43

wash its own mopping pads. Now

0:46

that's a good robot. Roomba Robots were made

0:48

for this. Learn more at

0:50

iRobot.com. That's i -r -o -b -o

0:52

-t dot com. Check,

0:58

check. Hello. Hey,

1:00

no, mom. Hello, Meredith. So

1:03

as you know, I have the

1:05

cutest dog of all time. Houdini.

1:07

Houdini. We've met. He's adorable. He

1:09

is the cutest little boy and

1:12

he's a tripod. He's missing his

1:14

back left leg. He lost his

1:16

leg, I think, in a car

1:18

accident and we call him Houdini

1:21

because we say that it's because

1:23

he made his leg disappear. He

1:25

really is like a magic dog.

1:27

Yeah, he just hops around and

1:29

has this big swinging tail and

1:32

it just looks like he's always

1:34

leading a parade. Like

1:36

we bring him to airports

1:38

and just like every child

1:40

in the terminal is just

1:42

magnetically attracted to this adorable

1:44

dog. Incredible. But there

1:46

are sometimes when I wonder like, you

1:49

know, does he ever miss

1:51

having his fourth leg? And

1:53

I think the place where I see

1:55

this the most is when he falls

1:58

asleep and falls into a little puppy

2:00

dream, he's yipping, all his paws are

2:02

twitching. It's so cute. And so I've

2:04

always thought that in his dreams, he

2:06

must have four legs because his little

2:08

stump is twitching along with the rest

2:11

of his legs. So

2:13

it's like, is he dreaming

2:15

being a four -legged dog or is he three

2:17

-legged in his dream? The

2:22

question of exactly what Houdini is

2:25

dreaming about might be impossible to

2:27

answer. But Meredith is

2:29

in very good company in connecting

2:31

these twitches with dreams. We

2:35

have been sleeping for thousands

2:37

of years, yet it was

2:39

not until 1953 that a

2:41

graduate student noticed that while

2:43

you sleep, a dramatic activity

2:45

takes place. During the deepest

2:47

stage of sleep, when we've got these twitches

2:49

in our limbs, We've also got

2:51

twitches in our eyes. Scientists

2:55

are pretty sure that REM

2:57

sleep is when most dreams

2:59

happen, at least in people, because

3:02

they've done experiments where they've literally

3:04

woken people up and asked them.

3:12

I was dreaming about school and

3:14

I was blanking the grades

3:16

that's all I know. But

3:18

for animals, these twitches are some

3:21

of the best evidence we've got. There is

3:23

some scientific theory that when we see

3:25

our cats and dogs dreaming, or what

3:27

looks like dreaming, almost barking in their

3:30

sleep or moving their legs, that they

3:32

probably are. The idea is

3:34

that these twitches are bursts of

3:37

activity that kind of break through

3:39

the paralysis that happens during REM

3:41

sleep, which gives us a glimpse

3:43

of animal dreams. And

3:46

this has been accepted for a pretty

3:48

long time. Darwin's protege actually wrote, quote,

3:51

ferrets dream, as I've frequently seen

3:53

them when fast asleep, moving their

3:55

noses and twitching their claws as

3:57

if in pursuit of rabbits. The

4:02

thing about twitches is it looks like

4:04

obviously it's related to dreams, right? Right.

4:06

Mark Bumber, neuroscientist, University of Iowa. I

4:08

mean, we know we have dreams. We

4:10

know that we are moving around. So

4:13

it just makes sense to think, oh,

4:15

movements, movements, why wouldn't they be connected?

4:19

That's what Mark thought at first.

4:22

But then he started seeing twitches

4:24

in really young animals, like newborns.

4:27

And he was like, a newborn animal has

4:29

had very little waking experiences. What

4:31

the hell are they dreaming about? If twitching

4:33

was really related to dreaming, you'd

4:35

expect that the older you get and the

4:37

more experiences you have, the more you dream,

4:40

the more you twitch. But that's not what

4:42

happens. We move the most when we're young.

4:44

So what we're dreaming the most when we're

4:46

young. In case it's not

4:48

totally obvious by this point, Mark's

4:50

kind of over dreams. I

4:53

mean, can we please just not talk about dreams all

4:55

the time? You know, that's sort of, sorry, I'm a

4:57

little bitter. I

5:00

mean, look, dreams are fascinating, but the

5:02

focus on dreams is kind of a

5:04

distraction from what really matters. So

5:09

Mark decided to take dreams. fully

5:11

out of the picture by experimenting

5:13

on newborn rats. We literally surgically

5:15

disconnected the cortex from the other

5:18

half of the brain. He cut

5:20

off the part of the brain

5:22

responsible for creating dreams. And the

5:24

animals twitched completely normally. If

5:27

these twitches were caused by dreams, they

5:30

should have stopped. But we found no

5:32

effect at all on twitches. And so

5:34

I was like, OK, what is this

5:37

about? You know, this is

5:39

happening, as we counted it up, hundreds of

5:41

thousands of times per day. And

5:43

pups are in the business of growing, not

5:45

wasting energy. And you know, hundreds of thousands

5:47

of movements, that takes up a lot of

5:49

energy. Why would you do it unless it

5:52

had some intrinsic value that had not yet

5:54

been explored? I'm

5:57

Noam Hasenfeld, and this week on Unexplainable,

6:00

how this seemingly small question, why

6:02

do we twitch in our sleep? has

6:05

fundamentally shifted how we understand the

6:07

relationship between the brain and the

6:10

body. All

6:20

right, Mark, just

6:22

to make sure

6:24

before we dive

6:26

in here, when

6:29

I think of sleep twitches, I think of

6:31

those twitches I get right when I'm falling

6:33

asleep. Hypnic jerks, yeah. Is

6:36

that part of this? Is that different? It's

6:38

a separate phenomenon. It's more akin to what's

6:40

called a startle than a twitch. You're

6:43

not in REM sleep when that happens. And

6:45

there a lot of theories about it,

6:47

but the fact is it's an extremely

6:49

hard thing to study. So we don't

6:52

really know what's happening there? No. Dang.

6:54

Okay. Well, if we're just talking about

6:56

these REM sleep twitches then, how

6:58

common are they? Do all kinds of

7:01

animals and people twitch? Yeah, I mean,

7:03

I've got a website that collects all

7:05

these different videos, you know, and what

7:07

you see across. different animals is that

7:09

the parts of the body that the

7:11

animals really, really rely on for bringing

7:13

sensory information into their brain are the

7:15

parts that twitch the most. So

7:18

for us, you know, rapid eye movements are twitches

7:20

of the eyes. We also twitch

7:22

our fingers a lot when we're adults. With

7:24

cats, you see their paws moving a lot.

7:26

Ferrets, you see whisker twitches. Rats, you see

7:29

lots of whisker twitches. They use their whiskers

7:31

to learn about the world just as well

7:33

as we use our eyes. And if all

7:36

of these twitches aren't just... know, enacting dreams,

7:39

how do you start figuring out what they

7:41

actually are? Well, you know, the first thing

7:43

you have to do is try to figure

7:45

out what parts of the brain are producing

7:48

this. I mean, how is this all happening?

7:50

Okay. And what we started to see when

7:52

we were recording brain activity is that the

7:55

brains of neonates, baby rats, were much more

7:57

active during sleep and much more active when

7:59

animals were twitching than when they were awake.

8:02

Huh. Okay. It's one thing to think that

8:04

sleep has brain activity associated with it. That

8:06

was a huge finding 80 years ago. It's

8:09

another thing entirely to see that

8:11

the brain activity is greater. And

8:14

I mean, much greater during REM

8:16

sleep than during wake. And

8:18

I assume it's reasonable to think all of that

8:20

brain activity is connected with these twitches, right? Is

8:24

there a way to actually test it? Yeah.

8:26

I mean, the biggest problem was methodological. How

8:28

do you record brain activity in a very,

8:30

very small. baby rat, which

8:32

was the best animal for doing this sort

8:35

of work. You have to figure out

8:37

how to get them in a stable situation so you

8:39

can drop these very fine electrodes into the brain. And

8:42

so it took years to get the methods

8:44

going. But what we started to see is

8:46

that every time the animal twitches, you know,

8:48

10 milliseconds later, the part of

8:50

the brain that's responsive to sensory input for

8:52

that limb shows a huge burst of activity.

8:55

So twitch activity. twitch activity, not the

8:57

other way around. This is a sensory

8:59

signal. So this timing here matters. If

9:02

you have a twitch and then you get

9:04

a burst of activity in the brain after

9:07

that twitch, then you have a pretty good

9:09

idea that that's a sensory signal that you're

9:11

picking up on. Yeah, so like a signal

9:13

the brain is getting from a nerve or

9:15

a muscle or something? Yeah, the sensory it

9:18

puts. So every time you move a limb,

9:20

you have sensors in your muscle, you have

9:22

sensors in your skin and your joints. And

9:24

those sensors, when you have movement, they produce

9:26

neural signals that flow up into the brain.

9:29

That's how we know when our arms are

9:31

moving or when you touch something. So we

9:33

have sensors all throughout our limbs. And so

9:35

when the limbs were moving, that's when we

9:37

were seeing the brain activity in parts of

9:40

the brain that are responsive to those types

9:42

of sensory signals. And just to be totally

9:44

clear here, the activity that you saw in

9:46

the brain was happening after the twitches. Yes.

9:48

I mean, you can't really see it because

9:51

it's happening so fast. You have to get

9:53

down to milliseconds. So what my student, his

9:55

name is Ryan Glanz, what he did was

9:57

he was recording from the part of the

9:59

cortex that responds to sensory feedback. And for

10:02

every neuron that he was recording from, he

10:04

gave it a different musical note so that

10:06

you could easily see that when the limb

10:08

twitches, there was a burst of activity in

10:10

this part of the cortex. And you have

10:13

a video of that rat experiment, right? Yeah.

10:15

Can we take a look? Sure. Wow.

10:21

It sounds twitchy. So

10:24

what I'm looking at here is this kind of

10:26

little paw. And the

10:29

twitching in the rat paw is

10:31

kind of generating these sounds that

10:33

are mapped onto neurons. Correct. So

10:35

every time the limb twitches, There's

10:37

a really, really clear burst of

10:39

activity. And

10:44

at these ages, all the neurons are firing

10:46

together. And that's why it just sounds like

10:48

a chord and not a lot of like

10:50

little different musical notes happening in sequence. And

10:52

it's just a lot of activity happening simultaneously.

10:57

Wow. OK. So you're essentially flipping

10:59

the traditional hypothesis on its head,

11:02

right? It's not dreams causing twitches.

11:05

It's twitches. causing dreams or twitches?

11:07

I don't know. It's not necessarily

11:09

twitches causing dreams, right? It's twitches

11:11

causing some impact in the brain.

11:14

Yeah. I mean, obviously twitches are not

11:16

going to be the sole source of

11:18

all things in the dreaming brain, but

11:21

that it is at least providing sensory

11:23

input to the brain during sleep that

11:25

we know for a fact. So it

11:27

does flip it on its head and

11:29

it completely changes the calculus of what's

11:32

happening in a dreaming brain. So

11:37

then why would the twitching be happening to

11:39

begin with? Like what's the what's the point

11:41

of all of this twitching? Well, so this

11:44

is where you have to start to think

11:46

about what is it that's special about twitches,

11:48

right? Mm -hmm. The first thing that you

11:50

notice is that the movements are discreet. It

11:53

turns out that discreetness is incredibly

11:55

important. So imagine that you're standing

11:57

at a switchboard with hundreds of

11:59

different switches. Let's just say they're

12:01

neurons and then All the wires

12:03

from all of those switches lead

12:05

to a whole bunch of lights.

12:08

So every switch controls a different light. And

12:11

let's say that those lights are muscles. If

12:14

you're sitting at that switchboard and you want

12:16

to figure out which switches control which lights,

12:18

you don't just start throwing all the switches

12:20

simultaneously. Because if you

12:22

did, you're gaining no information. All you're seeing is a

12:24

bunch of lights turn on and you've thrown a bunch

12:26

of switches. The answer is you throw

12:29

one switch at a time, you see which light

12:31

comes on, and then you make that connection. And

12:34

so that's the difference between wake movements

12:36

and twitches. You know, I'm sitting here

12:38

talking to you, and I'm gesturing, and

12:40

I'm moving all my limbs simultaneously, my

12:42

posture, my neck, everything, my eyes. Everything's

12:44

moving simultaneously, right? That's waking.

12:46

That's one of the characteristics of waking

12:48

movements is that they're continuous, and they're

12:50

simultaneous, and they're highly complex. But

12:53

when you're twitching, one twitch at

12:55

a time, you ping your body.

12:57

and the body pings you back.

13:00

And then you know that the first

13:02

thing is related to the second thing.

13:04

And that's the discreteness of twitching. And

13:06

that explains why these animals are twitching

13:09

so much, you never grow and develop

13:11

more than you do when you are

13:13

young. So

13:16

they're literally, your theory is that

13:18

the power goes out and they're

13:20

flipping switches in a fuse box

13:22

or something to see which switch

13:24

controls which light. because

13:27

there's no other stimuli coming in,

13:29

right? They're in a controlled environment,

13:31

so they're essentially doing sort of

13:33

an experiment to learn their own

13:36

body? Yeah, exactly. They're bootstrapping their

13:38

system. They're self -organizing their

13:40

sensory motor system, and it's done for with

13:42

it. It's a big mystery as to how

13:44

we develop things like our sensory motor system.

13:46

how do you actually learn about your body

13:48

when you're a newborn rat or human and

13:50

you're born you have no idea how your

13:52

body is formed you have no idea how

13:54

it moves and it's going to be changing

13:57

every single day as you grow and figure

13:59

out new things right so how do you

14:01

figure out how to move that body in

14:03

real time through the process of development. You

14:05

can't prescribe this. You can't blueprint this.

14:07

There's no genetic mechanism that can tell

14:09

you exactly how you're going to be

14:12

on day three versus day five. So

14:14

you need to have a system that's highly adaptable. But

14:17

you said something that's really important. You said turn off

14:19

the lights. And that's actually

14:21

a metaphor I've used before because that's

14:23

sort of what sleep paralysis is. Sleep

14:25

paralysis is like turning off all that

14:28

background noise. creating a very low noise

14:30

situation for your body. And now when

14:32

you ping it with a twitch, you

14:34

get a really, really clear signal back.

14:37

So you shut down the lights, you

14:39

turn off all that muscular activity, you

14:41

paralyze the body, and then you just

14:43

allow these individual twitches to go through.

14:45

And then you take that information and

14:47

you put that into the system for

14:49

the purpose of maintaining your circuits, calibrating

14:51

your circuits, so that you end up

14:53

with a finely tuned sensory motor system

14:55

so that we can function in the

14:57

world. And is that... know, if twitching

14:59

is about learning, we would assume younger

15:01

animals would twitch more. Is that the

15:03

case? Absolutely. And

15:06

then, I mean,

15:08

older animals also twitch.

15:10

Yes. Why would they be

15:13

twitching? Good question. First, we don't

15:15

twitch as much when we're older. But

15:17

second, some animals do twitch quite

15:19

a lot. And the part of the body

15:21

that twitches matters. This is

15:23

just a theory because nobody has

15:25

really explored it with the level

15:27

of sophistication that we need. But

15:29

we have to calibrate our systems.

15:32

Over the day, we get tired, we lose control, our

15:34

vision gets worse and worse through the day, and then

15:36

you wake up the next day and you're rejuvenated. I

15:40

think it's possible that twitches continue throughout

15:42

life for some parts of the body

15:44

for that purpose to calibrate. a weary

15:46

system. And there's some hints out there

15:48

in the world that this could be

15:50

happening, including work that was done in

15:52

humans. But they're mostly hints, and it

15:54

needs to be done more systematically. And

16:01

why do you think the scientific

16:03

community missed this for so long,

16:05

missed understanding Twitch's as a developmental

16:07

process? Because when you label something

16:09

as a byproduct of dreams, why

16:11

would anybody spend their time studying

16:13

it? Like it's just closing off

16:15

further inquiry. It's just being like

16:17

done. Yeah, I mean there look

16:19

I don't want to be too

16:21

flippant about it dreams are fascinating

16:23

But they're kind of a red

16:25

herring when it comes to studying

16:27

sleep. There is to my mind

16:29

many many Fascinating things about sleep

16:31

that have nothing to do with

16:33

dreams and the focus on dreams

16:35

is kind of a distraction from

16:37

what really matters Support

16:50

for the show comes from Roomba

16:52

Robots. I got a lot of

16:55

things on my to -do list. Laundry, groceries, dishes,

16:58

recording ads. Something's always getting

17:00

put off. But with a Roomba

17:03

robot, one thing you don't have to put

17:05

off is cleaning your floors. iRobot

17:08

says their brand new Roomba

17:10

Plus 405 combo can vacuum,

17:12

mop, and even do

17:14

both at the same time. It

17:16

can vacuum first and then apparently

17:18

retrace its steps and mop the

17:20

area it vacuumed earlier. Except it

17:22

can't really retrace its steps because

17:24

it doesn't have feet. Can you

17:26

imagine? It uses advanced

17:28

spinning, dual -clean mop pads to

17:31

help your floors shine. And

17:33

iRobot says it can even empty itself automatically

17:36

and wash its own mop pads, all

17:38

without you having to get anywhere near

17:40

your to -do list. You can even give

17:42

your Roomba Robot a fun name and

17:44

a completely reimagined Roomba Home app. Flory,

17:47

mop face, vac attack, have

17:50

at it. Roomba Robots were made for

17:53

this. Learn more at iRobot.com. That's I

17:55

-R -O -B -O -T.com. So,

18:09

Mark, if we take a step back

18:11

from Twitch's here and just talk about

18:13

sleep in general for a second, what

18:15

would you say people get wrong about

18:17

sleep? I think

18:19

one thing that people may get wrong

18:21

about sleep is they think about it

18:23

as a single thing. They think about

18:25

it as just a unitary phenomenon. And

18:28

I think there's the jury still out

18:30

as to what sleep actually is. It's

18:32

very highly variable across the lifespan, highly

18:34

variable across different species. And

18:36

there's no singular definition of sleep that applies. Plus,

18:38

you know, I liken it to wake. I mean,

18:41

there's no singular function for wake. Wake is not

18:43

a singular phenomenon. And we do all kinds of

18:45

things in it. We walk, we talk, we eat,

18:48

we watch TV. We do all kinds of things when we're awake.

18:50

Why do we think that wake would be any different

18:53

than sleep? To my mind,

18:55

sleep is a conglomeration of things, all

18:57

these separate components. And you

18:59

got to think about, well, why are all these different parts there?

19:03

What brings all these things together during sleep?

19:06

isn't that all these different parts of sleep are

19:08

like the place settings at a dinner table? You

19:10

know, so you have your fork, your spoon, and

19:12

your knife. They all serve one purpose, you know,

19:14

to eat. Right. But what if they're

19:16

also more like all the collection of tools

19:18

in a Swiss army knife? You know, you've

19:20

got a fish scaler and a magnifying glass

19:22

and a toothpick and a nail file. What

19:24

the hell do those things have to do

19:26

with each other? Nothing except they're all together

19:28

in one place. So

19:30

is sleep more like a dining room setting

19:33

or is it more like a Swiss army

19:35

knife? And the answer is it's probably a

19:37

little bit of both. And our task is

19:39

to figure out how all these different parts

19:41

of sleep fit together. And that's the big

19:43

question about sleep. Which parts of

19:45

sleep do these different animals exhibit? How does

19:48

it contribute to their survival? How does it

19:50

contribute to their learning? How does it contribute

19:52

to their evolution? To me, that's the bigger

19:54

question about sleep. Can

19:57

you tell me about some of those tools

19:59

in the toolbox aside from twitching? Sure.

20:02

I mean, you know, the more we look...

20:04

the more we discover new things, you

20:07

know, we've only scratched the surface of

20:09

sleep, but there's the lymphatic system, which

20:11

is basically being called like a waste

20:13

clearance system for the brain to remove

20:15

debris and to remove dead cells. And

20:17

you have fundamental brain rhythms, for example,

20:20

you have a rhythm that's involved in

20:22

movement and learning and things of that

20:24

nature. And then there's a whole

20:26

bunch of animal and human work that's been

20:28

done looking at the role of these very

20:30

specific brain rhythms for the consolidation of memory.

20:32

So sleep is playing a major role in

20:34

memory consolidation. And the sleep theory

20:37

of memory consolidation is so powerful because the

20:39

idea is that you just can't consolidate memories

20:41

when you're in the process of learning them.

20:43

So you have to stop learning new memories

20:45

in order to consolidate the ones you already

20:47

have. You know, I've been

20:49

getting back into playing piano again

20:51

and I've been going

20:53

through some Bach inventions, and

20:56

I feel like after I

20:58

sleep sometimes, it's like

21:01

more deeply in me. Does

21:03

that make sense? Is that

21:05

something like memory consolidation? It

21:08

absolutely is. So I have the same thing that

21:10

happens to me. I'm a drummer, and when I

21:12

learn a new pattern. Oh, I'm also a drummer.

21:14

I'm more of a drummer. That's my training. So

21:17

tell me if you can relate to this. So

21:19

you know when you start learning a very complex

21:21

new pattern, it's all very deliberate, slow, and rote,

21:23

right? Boom, boom, every limb doing

21:25

something. And you can't do anything else.

21:27

I mean, you're concentrating on every limb and what they're

21:29

doing. And then all of a sudden, six

21:32

days in, I don't know, it's like...

21:35

Right, everything's happening. And

21:38

at the same time, you can hold a conversation with somebody,

21:40

right? It's like

21:42

gone to a deeper level in

21:44

your understanding. Yes, it's called automaticity.

21:47

And, you know, it's a pet hypothesis. It

21:50

has never been confirmed or tested seriously yet.

21:52

But I just think, wouldn't that be amazing

21:54

if sleep were involved in that kind of

21:56

skill transfer? Yeah. You're taking it that's something

21:59

very conscious, you're highly attentive to it and

22:01

turning it into something that you can do.

22:03

on its own. And you know, maybe that's

22:05

how we learn how to do things like

22:08

walk. I mean, they start

22:10

off as pretty difficult, but eventually we

22:12

don't think about those anymore either. And

22:14

we're actually doing an experiment that's in

22:16

the dancing realm with some colleagues out

22:19

in California. And we want to know

22:21

when people develop these automatic behaviors, does

22:23

that show up somehow in their sleep

22:25

patterns? Because the development of automaticity is

22:27

a very, very natural place to think

22:30

about how twitching and these sorts of

22:32

sleep -related phenomena might be playing a

22:34

role in the consolidation of those motor

22:36

memories that happen, like drumming and playing

22:39

Bach and stuff like that. Yeah, I

22:41

think, tell me if this is a

22:43

wrong way to think about it, but

22:45

I think about the New York subway

22:47

system, which is the biggest 24 -hour

22:50

subway system in the world. And rats

22:52

are very important for the subway system,

22:54

too. Very important. Rats are important. But

22:56

the 24 -hour operation of the subway

22:59

is... kind of one of the reasons

23:01

why the subway might suck as much

23:03

as it does, just because it's harder

23:05

to fix stuff. Like other

23:07

cities where even if you could just work from

23:10

like two to six a .m., you can fix

23:12

some stuff. But in New York, if they really

23:14

want to fix stuff, they have to just do

23:16

major surgery. They have to like shut down a

23:19

line. And so a

23:21

lot of the subway just sort of falls

23:23

into disrepair. That's a wonderful

23:25

metaphor. You know, there are just some things

23:27

we do in Wake that seem Just

23:30

obviously to be incompatible with the things

23:32

that need to happen during sleep And

23:34

so we take our day and we

23:36

break it up into two very different

23:38

States here's our moving about the world

23:40

and being friends and talking and eating

23:42

and doing things That's one part of

23:45

our lives and the other part of

23:47

our lives is somehow connected to the

23:49

first part But is very different and

23:51

they have to be complementary so that

23:53

the sleep is supporting what we

23:55

need to do when we're awake, consolidating

23:57

memories, building proteins,

24:00

repairing systems. All of those things

24:02

need to go on, but they're

24:04

just incompatible with wakefulness. So

24:07

then if I were to bring twitches

24:09

back into the conversation here, are

24:12

they kind of emblematic of

24:14

this larger lights off, recalibrating,

24:16

learning process of sleep? Is

24:18

that what stands out to

24:21

you? I'm

24:23

a behaviorally oriented neuroscientist I want to

24:25

understand the role of behavior how it

24:27

develops and how it influences the developmental

24:29

process and. We have a lot of

24:32

ideas out there right now in the

24:34

world, what I would call very simplistic

24:36

ideas. People think about, well, you know,

24:38

you have a genetic blueprint and the

24:40

animal just develops and then it's like,

24:42

blah, blah, blah, okay? And these are

24:45

just corrosive ideas for thinking about development

24:47

because development is a process. It's

24:50

extremely complicated and it's extremely

24:52

plastic. You know,

24:54

brains are not isolated,

24:57

separate standalone organs. They're

24:59

embodied. There's a reason why. Developing

25:02

robots is a lot more difficult than it

25:04

is just to develop like AI that has

25:07

no body to control. And it's because controlling

25:09

a body is hard. And the

25:11

one thing that robots can't do yet

25:13

is develop. And I think

25:16

that there's something fundamentally important about developing

25:18

in a body that changes and learning

25:20

how to adapt to those changes that

25:22

is. partly why we are so facile

25:24

in terms of how we use our

25:26

bodies on a regular basis. So for

25:29

me, it's an extension of everything I've

25:31

ever believed about the developmental process and

25:33

what we need to do as developmentalists

25:35

to understand that process as opposed to

25:37

the more sort of static ways that

25:39

oftentimes people think about what's happening. Wait,

25:46

so if we want to make robots that

25:48

can develop and learn better, we just have

25:50

to make robots that can sleep? Yes.

25:54

There are people who have

25:56

used twitching in robots to

25:58

produce more adaptable robots. Oh.

26:00

They've mimicked twitches in robots

26:02

and have produced robots that

26:04

were better able to adapt

26:06

to changes in their bodies.

26:08

Okay. Crazy stuff. Now

26:13

going to ask a robot's dream, and

26:16

my head's going to explode. Yeah, to

26:18

do Android's dream of electric sheep. That

26:35

was Professor Mark Blumberg. And fun story,

26:37

this episode came about because Mark listened

26:39

to the episode we did a while

26:41

back on the baseball player who had

26:43

the yips. You know, this

26:46

kind of performance anxiety that people used

26:48

to think was only psychological but is

26:50

now sometimes being connected to involuntary twitches.

26:53

So Mark reached out to me and told

26:55

me the episode had given him some new

26:58

research ideas on twitches. We started talking and

27:00

then we ended up with a whole episode

27:02

on sleep twitches. So if you're a scientist

27:04

out there and you're listening, and you're getting

27:07

an idea for some new research, let us

27:09

know. We're always on the lookout for more

27:11

interesting scientific questions, and we'd love to talk

27:13

to you. This

27:17

episode was produced by me, Noam Hasenfeld.

27:19

We had editing from Meredith Haudenot, who runs the

27:22

show, mixing and sound design from

27:24

Christian Ayala, music from me, and fact checking

27:26

from Katie Pensey Mug. Thomas Lu

27:28

is wondering why things make so much

27:30

sense. And Bird Pinkerton

27:32

headed back to the octopus hospital, knowing

27:35

she had her army to fight the

27:37

birds. But when she got there,

27:39

she saw the door blown out. Windows

27:41

were smashed. Cables were pulled

27:43

out, hanging everywhere. She

27:46

was too late. Thanks

27:50

as always to Brian Resnick for co -creating

27:52

the show. And if you have thoughts

27:54

about the show, send us an email. We're at unexplainable

27:56

at vox.com. And you can also leave us a review

27:58

or a rating wherever you listen.

28:00

It really helps us find

28:02

new listeners. You can

28:04

also support the show and all of

28:06

journalism by joining our membership program today.

28:09

You can go to vox.com slash members to sign

28:11

up. And if you signed up because of

28:13

us, send us a note, We'd really love

28:15

to hear from you. Unexplainable is

28:17

part of the Vox Media Podcast and

28:19

we'll be back next week. Support

28:31

for the show comes from Roomba Robots.

28:33

A lot of people are talking about

28:35

robots. Aren't they gonna take our

28:37

jobs? Are they gonna take over the world? Not

28:40

enough people are talking about our floors because

28:43

the robots have already done it. They

28:45

figured out a way to clean floors

28:47

all on their own. And it's kinda great. Roomba

28:50

Robots work so well because they

28:52

use a four stage cleaning system

28:54

and Clearview LiDAR to navigate your

28:56

home. LiDAR, they aren't

28:58

messing around. Roomba robots were

29:00

made for this. Learn more at...