May 28, 2024 • 51 mins
In this classic episode of Stuff to Blow Your Mind, neuroscientist and author David Eagleman drops by the show to discuss his new podcast series “Inner Cosmos” with Robert and Joe, along with some general discussion on the human brain. (Originally published 04/13/2023)
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:06):
Hey, welcome to Stuff to Blow Your Mind. My name
is Robert Lamb.
Speaker 2 (00:09):
And I am Joe McCormick. And today we are bringing
you an episode from the Vault. This one originally published
April thirteenth, twenty twenty three, and this was our conversation
with David Eagleman, neuroscientist and host of the Inner Cosmos podcast.
I remember this was a very interesting chat. We hope
you enjoy it.
Speaker 3 (00:32):
Welcome to Stuff to Blow Your Mind, a production of iHeartRadio.
Speaker 1 (00:41):
Hey, welcome to Stuff to Blow your Mind. My name is.
Speaker 2 (00:44):
Robert Lamb and I'm Joe McCormick, and hey, welcome back, Rob.
You were out sick earlier this week. It's good to
have you back.
Speaker 4 (00:51):
It's good to be back now.
Speaker 2 (00:53):
Because you were outsick, we ended up putting a pause
on an ongoing series we were doing on childhood amnesia.
We ended up running a Vault episode on Tuesday, and
I just wanted to assure people that we will be
coming back to that subject. We will be resuming the series,
probably for next week's core episodes, but because we already
had it scheduled out this way, today's episode is going
(01:15):
to be an interview. So in fact, we are talking
to a return guest, the neuroscientist and author David Eagleman.
This is actually the second time David has been a
guest on the show. In September twenty twenty, Rob, you
spoke to him about his book Live Wired, which is
a popular science book on the subject of brain plasticity.
Speaker 1 (01:36):
Yeah. Absolutely. That was a fun episode. You can find
it in the archives, and the book Live Wired is
an absolute delight. If you're at all interested in anything
you hear us discussing in this episode, pick up a
copy of it.
Speaker 4 (01:48):
It's great.
Speaker 2 (01:49):
So this week we invited David back on the show
because he now has a fantastic, brand new podcast on
our very own network on the iHeart Network, and it
is called Inner Cosmos with David Eagleman. So before we
get started with our interview, I thought we should just
share a bit of background about David. This is from
(02:10):
his website.
Speaker 1 (02:11):
That's right. David Eagleman is a neuroscientist at Stanford University
and an internationally best selling author. He's the co founder
of two venture backed companies, Neosensory and brain Check, and
he also directs the Center for Science and Law, a
national nonprofit Institute. He's best known for his work on
sensory substitution, time, perception, brain plasticity, synesthesia, and neurolaw. He
(02:35):
is the writer and presenter of the international PBS series
The Brain with David Eagleman and the author of the
companion book, The Brain The Story of You. He's also
the writer and presenter of The Creative Brain on Netflix.
Speaker 2 (02:47):
David Eagleman is the author of over one hundred and
twenty academic publications and many many books of popular science.
Eagleman is a TED speaker, a Guggenheim Fellow, and serves
on several boards, including the America Brain Foundation and the
long Now Foundation. He's the Chief Scientific Advisor of the
Mind Science Foundation and the winner of the Claude Shannon
(03:09):
Luminary Award from Bell Labs and the McGovern Award for
Excellence in Biomedical Communication. He serves as the academic editor
for the Journal of Science and Law. Was named Science
Educator of the Year by this Society for Neuroscience and
was featured as one of the quote Brightest Idea Guys
by Italy's Style magazine. He served as the scientific advisor
(03:31):
on several TV shows, including Westworld and Perception and has
been profiled on The Colbert Report, Nova Science Now, The
New Yorker, CNN's Next List, and many other venues. He
appears regularly on radio and television to discuss literature and science,
and I guess now he is going to start having
to add podcasts to the end of his bio here, So, Robin,
(03:54):
unless you have anything else, I think we should jump
right into our conversation with David Eagleman.
Speaker 1 (04:01):
Hi, David, welcome back to the show.
Speaker 4 (04:03):
Great, Thanks Rob for having me again. It's a pleasure
to be here. And hello, Joe, it's great to meet you. David.
Speaker 1 (04:08):
Now, you have a great new podcast series out Inner
Cosmos through iHeart. How did you decide what path to
take with the podcast format?
Speaker 4 (04:18):
You know, it's a great question. The truth is I
had not listened to many podcasts at all. Now I have.
But when I first was putting this together with iHeart,
I thought, look, I want to do a forty five
minute to hour long monologue every week. And that seemed
like a terrific idea at first, and then my wife
said she was going to kill me because it turns
(04:39):
out that's a ton of work. It takes me about
kind of about twelve hours a week to get a
good monologue that's almost an hour long. So that's how
I decided on the format because I thought it would
be something special rather than you know, I've been on
many different podcasts where we're doing interviews just like this,
(05:01):
and it's super fun. But I wanted to do something different.
So that's how I accidentally stumbled into that format.
Speaker 2 (05:08):
So I figure we should give people a taste of
the kind of things you talk about on your show.
I got a chance to listen to the episode you
did about memory and the perception of time, and I
thought it was great, by the way, a really great
way to kick off the show. So your starting premise
in that episode was that many people who have been
through intense or life threatening events, maybe falling off of
(05:30):
a building or seeing a car speeding toward them, report
afterwards that time seemed to have somehow slowed down for
them during the pivotal few seconds, almost as if they
were able to inter a state of slow motion or
bullet time, like from the matrix. Can you talk a
bit about your research on this subject and what you
(05:50):
discovered about this perception.
Speaker 4 (05:52):
Yeah, so my research, of course started off very personal
which is that I fell off of a house and
it seemed like things took a long time, and that's
what got me interested in this. Then when I got older,
it became a neuroscientist. Eventually I realized that I was
hearing this story not uncommonly from people who had been
in a you know, gunfight or some scary situation or
(06:14):
car accident or whatever, and they felt that things took longer.
And so I looked in the literature and there was
not anything on this. So that's when I came to
realize I was going to have to do this myself
and figure out how to run an experiment on this.
So you know, briefly, what I did is I built
a device that I hooked to people's wrists that flashed
information them at them met a certain way so that
(06:36):
I could tell how rapidly their brain was perceiving, and
that way I could test whether they were actually seeing
in slow motion or The whole thing was a trick
of memory, meaning when you were in an intense situation,
you wrote down more memories. So when you said what
just happened, when just happened, that it seemed like it
must have taken longer because you have all these memories.
(06:59):
So what I did then is dropped people from one
hundred and fifty foot tall tower in free fall backwards
into a net below, and I measured their perception of
time on the way down this way. And what I
found after, you know, I didn't sell myself, but then
we dropped twenty three participants. What I found is that
(07:20):
it is in fact a trick of memory, which is
to say, when everything is hitting the fan, your brain
writes down much denser memory, and when you read that
back out, your brain has to make an assumption about
you know, how much memory, how much footage maps onto
how much time, and so it says, oh wow, that
must have been five seconds, even though it was only
(07:40):
one second worth. But the point is people were not
able to see in slow motion, which, by the way,
was disappointing for me because I already had. I was
already talking with the military about building cockpits in a
way that flashed information more rapidly at people when they
be in some intense situation. But it turns out none
of that makes the difference. You can't actually get information
there faster, you can only remember it faster, So.
Speaker 2 (08:04):
There's not actually any any increased ability of perception. It's
just a trick of the memory exactly.
Speaker 4 (08:12):
Now. It is the case that you know a lot.
You can do a lot of things pre consciously, by
which I mean your conscious awareness of something is always
the slowest thing on the ladder to ever get any information.
So by the time your brain puts together all the
signals and says, okay, this is what just happened. You know,
that's at least half a second to a second behind
(08:33):
real time. But the point is your body can react
much faster than that. Your body can get signals and say, whoa,
I got to do something about this right away, And
so you can react, you know, often much faster than
you can be consciously aware. So you know, I don't
know if you've been on a I mean, this is
what happened to me recently. I was on a hike
with a friend and a branch snapped back, and I was,
(08:54):
you know, halfway into the move of ducking out of
the way of the branch before I consciously realized it,
or my foot gets halfway to the break of my
truck before I realized that there's a car pulling out
of the driveway ahead of me. In other words, consciousness
is always the last guy on the ladder to get
any information, and your body can almost always react much faster.
(09:15):
In fact, wait, let me just say one more example
of that, which is when I was younger, I used
to play baseball, and my experience was always that, you know,
I'd be waiting for the pitch, and then I would
realize after it had happened that I had already hit
the ball. And I would consciously realize, oh, I have
just hit the ball. Now throw the bat and run.
(09:38):
But you know, the whole thing, the ball moving from
the mound to the plate, and the swing and the batting,
that's all really fast process and it often happens pre consciously.
Speaker 2 (09:48):
Somewhat related to that, this raises questions about the different
kinds of circumstances that would favor the perception of slow
motion in intense situations or not.
Speaker 3 (10:01):
So.
Speaker 2 (10:01):
My example was there was one night years ago I
was driving under an overpass and there was a sudden
deafening sound and a shudder, And what my wife and
I deduced later was that somehow like a brick had
fallen from above and hit the roof of our car
just above the windshield. As we passed under a highway
speed and I don't know if somebody threw it or
(10:21):
if it somehow just fell, But I not only don't
recall a feeling of stretched time or a greater density
of memories right before and after the impact, I felt
almost a kind of retrospective amnesia, like a real paucity
of detail. And it was like we were suddenly a
good ways down the road and just trying to remember
(10:42):
or figure out what had happened.
Speaker 4 (10:45):
Yeah, that's exactly right. It's because you didn't write down
any memory. And this is generally because as you are
taking a drive down the highway, your brain is writing
down very little stuff going on. In fact, the interesting
part is that although we think about memory as being
like a video recorder or something, in fact it's nothing
like that. You write down very little of what happens
(11:06):
in your life, especially when you're driving on a road
you've been down before. So what happened is there's the
deafening crash and suddenly you're thinking what just happened? What
just happened? And you've got nothing to draw on. There's
just no footage there. And by the way, just as
a very quick side note, I think this is what
happens to people when they are high on marijuana, is
they say, oh, my gosh, how long have I been
(11:28):
standing here? It feels like I'm standing here forever, And
it's because they're not writing down memories in the same way.
So when their brain looks for how long have I
been standing here, what it's looking for is footage in time,
as in, Okay, I remember getting here, I remember this happening,
someone said this. Then someone put the glass down and
blah blah blah, so that it can estimate how long
(11:48):
it's been there. But suddenly it can't grab on to
any memories at all, and so suddenly people are lost
in time anyway, So this is Joe exactly what happened
pins when people suddenly are hit by a car that
they don't see coming, like a car teet bones them
or something. Or I might have mentioned in the podcast,
(12:09):
I can't remember that, you know. I was once riding
my bike and the wheel suddenly dropped in a pothole
and I went flying over the handlebars. But because I
didn't see that coming, I just had the sensation of suddenly,
oh my god, what's is you know?
Speaker 1 (12:23):
Here?
Speaker 4 (12:23):
I'm lying on the asphalt, bloody, and I have no
idea what just happened. And it's precisely because you're not
running down any memories. So when your brain says what
just happened, what just happened, there's nothing to draw on,
as opposed to the brick sliding on ice towards the
brick wall phenomenon, which is where you say, oh my gosh,
I'm predicting what's going to happen and this is really
gonna hurt, this is gonna be bad, And that's when
(12:44):
you're writing down lots of stuff.
Speaker 2 (12:46):
So would you say you're more likely to have this
memory density perception number one if you see the event
coming ahead of time, there's there's expectation of it. But
also if you're generally in a novel or unusual situation.
Speaker 4 (13:03):
Yeah, that's exactly right. Actually, So two aspects of that.
One is that I just mentioned a moment ago that
you write down very little memory, and that's because as
an adult now, your brain has sort of figured out
a pretty good model of the world, meaning you don't
need to write stuff down because you've seen all the
personalities before, you've seen different cities before, you've seen roads,
(13:25):
and people and events and television shows and you sort
of got it. But if something really novel happens, that's
when your brain writes something down and says, WHOA, wait
a minute, I'm surprised, And that's when stuff starts getting
written down. So when you look back at the end
of let's say a novel event. Let's say you go
on some really wild trip on the weekend to Glockos
(13:46):
Islands and you see new things and so on, then
it seems like forever since you were at work on Friday.
But if you just go off for a normal weekend
and you come back to work, you think, oh, I
was just here because you didn't lay down any new
memories over the weekend. So it is true that things
that are novel generally seem to last longer. However, it
should be noted that when things are actually life threatening,
(14:10):
you have essentially an emergency response memory system that kicks
into gear. That is a secondary track on which you
write down memory. And this is underpinned by a part
of the brain called the amigla, and its job is
to say, WHOA, everything is going really bad and scary here,
and I got to write this down because that after
(14:31):
all is the point of memory is to make sure
that you write down stuff that is important and specifically
life threateningly important.
Speaker 2 (14:38):
So if the normal memory system, would that be the
hippocampal memory system exactly? If that's the normal memory system,
and then the amygdala tends to be recruited in intense situations.
Do we know generally if there is any if there
are any characteristic differences between how memories are recorded in
(15:00):
the hippocampus versus the Amygdalah.
Speaker 4 (15:02):
Yeah, And it turns out it's a tragic one, which
is that amygdala memories are uneaseable, whereas hippocampbel memories can
be erased. So let me unpack this because there's two
surprising parts here. So first of all, the fact that
hippocampal memories can be erased is terrifying and weird and wild.
(15:23):
Which is, if I ask you to recall the name
of your fifth grade teacher and then suddenly that brick
drops off the highway bridge and hits you in the head.
God forbid, let's say that happens. You will now have
amnesia for that one fact, You will not remember anything
(15:45):
about your fifth grade teacher anymore about the at least
the fifth grade teacher's name. Why it's because the name
of your fifth grade teacher is stored deep in the
structure of your brain, and when I ask you to
recall it, you're actually transferring it from that structural form
into activity, you know, spikes in the brain. And that's
(16:05):
how you're remembering the name of your teacher. Now, when
you're done remembering it, it has to get reconsolidated back
into its physical form. And if you get hit in
the head during that moment, it's gone. It's now you know,
it's been transferred from the physical to the you know,
activity in spikes. And if you you know, before it
(16:27):
gets transferred back into the physical, it can it is
susceptible to erasure, which is weird and terrifying. And by
the way, this can also be done with protein synthesis inhibitors.
So people do this in rats. They've been doing this
for decades, where you know, you train a rat how
to run different maases and then you put the rat
on a particular maze where the rat has to remember, oh, yeah,
(16:48):
that's this one, and then you just feed the rat
a protein synthesis inhibitor and now it cannot reconsolidate that
memory into physical form. So number one is hippocampal memory
can be erased. The number two point is that amigla
memories cannot be erased, which is to say, when you
recruit the emergency control system to say, wow, this is
(17:10):
really important, write this down. Than those are permanent, which
the reason I say that's unfortunate is because those are
the ones that people want to erase. In other words,
you know, let's say a rape victim or something like that,
that is the one thing that she wants to forget
more than anything, but cannot.
Speaker 1 (17:28):
Now, I was checking out the show as well, and
I was listening to your I believe this is an
episode from just earlier this week on the topic of
animal uplift, which I don't think is a term that
I was familiar with. Can you give us a brief taste,
a brief overview of what animal uplift is.
Speaker 4 (17:45):
Yeah, it's this idea that you know, look, the human
brain is made out of exactly the same stuff that
a mouse brain, the dog brain, the giraft brain. You know,
it's all the same stuff. It's got the same anatomy,
same general structure. We just have more of this wrinkly
outer bit called the cortex. But somehow we are you know,
we've taken over the whole planet as a species. We've
(18:07):
gotten off the planet. We've made vaccines and internet and
quantum mechanics and so like. There's some real difference in
what we are doing versus our neighbors in the animal kingdom.
But the genetic differences, as you know, are not that much.
I mean, we have enormous similarity with almost every Like
if you're building a giraffe, you got to build the
heart and the lungs and the brain and then the
(18:29):
esophagus and all that stuff is really the same stuff.
And so it's just some small algorithmic difference in the
DNA that's making our brain run in a more souped
up way. Okay, the idea of animal uplift is if
we can figure that out. And this won't happen, you know,
for at least a few more decades, but if we
can figure out, ah, here's the sequence of a's and
(18:50):
c's and t's and g's that gives us this high intelligence.
The question is should we give this to animals? Should
we help animals become intelligent? Now, let me just mention
this is an area that bioethicists and philosophers and neuroscientists
have been talking about for a while, and there's you know,
(19:12):
plenty of debate about it. And on one end of
the spectrum you have people say that's a terrible idea,
We wouldn't want to give intelligence to animals, and other
people say it's a moral obligation in the same way
that you know, if we know how to fix some
viral disease or fix a broken leg or something, of
course you should do this for your dog instead of
let your dog, you know, not have the medical advances
(19:33):
that we have made. So anyway, it's a big debate,
but this is the idea of animal uplift. You make
an animal as intelligent as a human. And I just
find this area fascinating. And you know, as I proposed
the podcast, what would the consequences of this be in
terms of, you know, will World War five be fought
(19:55):
by other animal species not just humans? You know, And
the way I sort of introduced the podcast is with
this question of what will my kids look back on?
Are my grandkids Obviously there's lots of things that will
be very different about our world right now and their world,
and let's say fifty years from now. But one of
(20:17):
them is will they look back and say, Wow, I
can't believe there was a time when humans were the
only species on Earth that was really doing anything, and
now we've got all these other you know, crows running universities,
and donkeys' programming computers and whatever, gophers in the Senate
and so on.
Speaker 2 (20:47):
So one idea of yours that I came across because
Rob sent it to me and I found really interesting
was from a paper you published in twenty twenty one,
I think maybe in Frontiers and Neuroscience, offering a hypothesis
about the adaptive function of dreams, which you call the
(21:07):
defensive activation theory, Could you lay out what is the
basic controversy about the biological function of dreams and how
your proposed solution here would answer this question.
Speaker 4 (21:21):
Yes, so it turns out there is no controversy about
the purpose of dreams because nobody knows, right, everyone, I
mean in the sense of everyone's got a little hypothesis
about it. But really it's complicated, and people think, well,
maybe it has something to do with learning and memory.
Maybe it just has to do with you know, energy restoration.
(21:41):
Maybe it has to do with you know, obviously the
Freudians thought that there was some important meaning in the
content of dreams and so on, but no one really knows,
and certainly no one has a quantitative hypothesis that can
make predictions about dreams and how much dream time we have.
But my student and I developed a theory that actually
(22:02):
does make quantitative predictions across animal species about it predicts
actually how much each animal species will dream. And to
explain something to take one step back, which is about
brain plasticity, which is this term that we use to
explain that the brain is very malleable, the human brain
in particular, and it's constantly reconfiguring its own circuitry and
(22:25):
that's how it learns and remembers, and that's how it
learns new skills and all that. So it turns out,
this is what my last book, Live Wired was about,
is the massive flexibility of the brain. It turns out
is probably a lot of people already into it. If
you go blind at a young age, the visual part
(22:47):
of your brain gets taken over, and in fact, if
you're born blind, that takeover is complete. The rest of
the territories in your brain involved in hearing and touch
and other things. These all take over what we would
normally think of as the visual cortex, and it's no
longer visual. It's now, you know, subserving other functions.
Speaker 1 (23:07):
Okay.
Speaker 4 (23:09):
One of the surprises in neuroscience was a study that
came out about a decade ago from some colleagues of
Mind Harvard where they put people in a scanner. These
are normally cited people, but they blindfolded them tightly and
they put them in the scanner and they were looking
at their brain's response to touch or to sound or
things like that. And what they found, to their surprise,
is that after an hour, you could start seeing the
(23:32):
first hints of signals in the visual cortex in response
to touch and sounds. So, in other words, the visual
cortex was starting to get annexed from these other territories
that they've touched and sound after one hour. And so
this was a much more rapid kind of movement than
anyone had expected. And so what my student and I
(23:56):
immediately realized is that this is the basis of dreaming.
It's because we are on a planet that rotates, and
we spend half our time in the darkness, away from
the light of our star, and so in the dark.
You can still hear and touch and taste and smell
just fine, but you can't see. And obviously I'm talking
(24:17):
about evolutionary time, you know, not our modern electricity blessed times.
And so what this means is the visual system in
particular has a real disadvantage, which is it is in
danger of getting taken over from the other senses. And
this is because of the brain's great plasticity, and so
(24:38):
as a result, the visual system needs a way to
defend its territory during the night. And that's what dreaming is.
Dreaming is essentially a screen saver. It's making sure that
at night time, when you're curled up in the corner
of your cave, staying out of trouble sleeping, and sleeping
has other benefits too, in terms of energy restoration and
(24:59):
so on. So when that's happening, you know, you can
still feel if some touches your skinner, if you're smelling
something or whatever. All that can still function in the dark,
but you're not seeing anything at all. And so what
happens is you've got this circuitry that just blows activity
into the visual system to make sure it stays active
during the night. Every ninety minutes, you have this wave
of active, random activity that just gets blown in there.
(25:21):
And because we're visual creatures, we see we have full,
rich visual experience even though our eyes are closed and
it's dark out, and it's because we are just making
sure the brain is making sure that it's keeping this
competition going so the visual system doesn't get taken over. Interestingly,
dream sleep is something we find across the animal kingdom,
(25:41):
but what we were able to demonstrate is that it
correlates with how plastic the animal species is. So some
animals drop out of the womb and they figure out
in thirty minutes how to run, how to walk. Very quickly,
they reach adolescence, they can reproduce, so all kinds, you know,
they're just they're obviously very pre programmed, let's just put
(26:02):
it that way. But other creatures like humans, are extremely plastic.
We take forever to learn how to walk, to wean
to reach reproductive age, things like that, precisely because we're
extremely plastic, and so we have lots of dreaming because
we have to protect our visual cortex at night. But
other animals that are, you know, these pre program types,
(26:26):
they have just a tiny bit of visual dreaming, but
not a lot. And by the way, I'll just mention
that the amount of visual dreaming we have goes down
with age. So it's an infant, you're dreaming all the time,
and as you get older and older, you dream less
and less is a fraction of your sleep. And you
know that's just a correlation. But in theory, what that
(26:46):
suggests is, you know, as an infant, your visual system
is very highly at risk of getting taken over, and
as you get older and things get more cemented into place,
it's less at risk of getting taken over, so you
don't need as much screensaver time and only.
Speaker 2 (27:00):
This kind of reminds me of studies I've read on
a related subject, which is a prolonged blindfolding of normally
cited people who apparently it's very common for people under
those circumstances to experience a lot of visual hallucinations. Does
that have any relationship to what you're talking about here?
Speaker 4 (27:18):
It does, Thank you for asking. That's perfect because this
is all part of the defensive activation theory, which is
to say, if a system is used to having data
coming in and suddenly it's not getting that data anymore,
it fights back from the inside it starts producing that
data itself. So one example of this is let's say blindfolding,
(27:39):
or you also see this for example in you know,
when people get thrown in solitary confinement in the dark,
they start having hallucinations both auditory and visual because they're
not getting that data and they're used to it they're
supposed to get. There's also something called Bonet syndrome Charles
Bone syndrome, which is people start losing their vision, but
they don't realize that they're losing their vision because they
(28:00):
start having hallucinations that essentially fill in for them. This
is all the same issue, though, which is that the
brain is used to getting certain inputs, suddenly it's not
getting it anymore, and so it starts generating in itself.
One more example is tonightis which is ringing in the ears.
This typically comes about because somebody loses hearing in some
(28:22):
frequency or some band of frequencies and the brain says,
wait a minute, I'm not hearing anything at twelve thousand
hurts anymore, so I'm going to start making myself and
it starts making this sound by itself. So this all
falls under the defensive activation theory.
Speaker 2 (28:39):
One of the interesting things I recall about the studies
on prolonged blindfolding was that the hallucinations that were reported
were not entirely random. So it wasn't just you know,
people seeing strange scenes play out in front of them.
That they would often hallucinate stuff that you would expect
to see in that place in the room based on
(29:00):
other senses. So like if they heard someone come to
the door of the room, they would hallucinate the image
of that person in the door.
Speaker 4 (29:08):
Yeah, that's perfect. And by the way, I think this
also has a lot to tell us about dream content
because the thing about dreams, I love the way you
put this, because you could, in theory, dream about anything
at all. You could dream that you are in Cambodia
and that you are in the fourteen hundreds and you're
(29:28):
a magician who's doing something. But you know, you tend
to dream about you know, your work and your spouse
and your drive and whatever, you know, things that are
more local to you. And it's precisely because when you
slam random activity into the visual system, the synapse is
the connections that are essentially hot from the day's work,
(29:52):
you know, those those are the things that tend to
get activated, and the association is very loose in a
sleeping dream state, and so what happens is, you know,
things can go off on weird tangents, but physics still
works fine in a dream. You know, rocks don't float
upwards and stuff like that, and so you know, essentially
you're just rebooting things that were there during the day.
(30:16):
And and this is closely related to what you're saying about.
You know, all the associations that your brain builds up
over a lifetime. So you hear the voice and you're
expecting to see that person, and that's exactly what happens. Actually,
I just want to mention one other thing about the dreams,
which is people will often ask me, well, what about
a blind person, how do they dream? And the answer
(30:38):
is blind people also dream because you have this very
ancient circuitry in your head that's blasting activity into the
back of the brain the occipital cortex, which is normally
the visual cortex. And people but if you're born blind,
it's you know, long taken over by hearing in touch,
and so a blind person's dream is all about hearing
in touch. They don't see anything, but they say, oh,
(31:00):
I was, you know, moving through the living room and
I felt the furniture was rearranged, and then there was
a big dog in the corner and I ran from
it and I was scared. And so, you know, they've
got full, rich dream experience. It's just that it is
not visual because that part of their brain is no
longer visual.
Speaker 2 (31:17):
And would you well, that makes me wonder then if
if your hypothesis about the defensive activity of dreaming is correct,
does that mean that that dreaming is now to protect
still what would normally be used for visual processing that
part of the brain, but its role in processing auditory
and other stimuli.
Speaker 4 (31:37):
A great question. No, that it's that these circuits that
underlie dreaming are extremely ancient, and so they are assuming
that you've got perfectly fine vision. And if you don't
have vision for some reason, then the circuits aren't going
to change. They're just doing a basic architectural job of saying, hey, guys,
(31:58):
every ninety minutes, just last some activity into the back
of the brain.
Speaker 1 (32:02):
There.
Speaker 4 (32:02):
That's all they're doing. And they don't know if you're
blind or not.
Speaker 1 (32:05):
Now, speaking about dreams, I guess it's not too much
of a leap to start talking about consciousness. I was wondering,
where do you think we are in terms of coraling, testing,
and even eliminating various theories concerning the nature of human consciousness. Yeah?
Speaker 4 (32:22):
Boy, this still remains to my mind the central unsolved
mystery of neuroscience. What's interesting, by the way, I wrote
an article, the cover article is Discover magazine back in
something like two thousand and six, called ten Unsolved Questions
of Neuroscience. And what's fascinating to me is it's now
twenty twenty three and they are equally as unsolved. I mean,
(32:42):
it's funny because we're making so much progress in the
field in some ways, and yet in other ways we're
just facing some really tough problems. So the consciousness, you know,
what is consciousness is really I think the central one.
And you know, for any listeners who are wondering what
is the question, the question is how do you take
(33:04):
eighty six billion cells and stick them together and hook
them up in such a way that you have private,
subjective internal experience. So you know, the smell of apple
pie and the taste of feta cheese, and the pain
of pain, and the redness of red and so on
how does that happen? Because you know, my laptop computer
(33:28):
has lots of signals running around, zeros and ones running around,
and it's transistors, but presumably it's not experiencing anything. It
can play a YouTube video for me, but presumably it
doesn't find it funny the way I do. And so
this is really the question of consciousness. We don't know
the answer to it. I can just tell you my
(33:51):
general feeling on this, which is when you look at
the history of science, what you find is that in
every era there were big pieces of information missing, and
yet the scientists were in a position of having to
try to explain everything not knowing some other thing. Here's
an example. You know, when the pump was invented, people
(34:13):
suddenly said, oh, I see the heart is like a pump,
and then it was obvious, oh, click falls into place.
But before that, everyone's trying to figure out what the
heck the heart was doing, but no one had the
concept of a pump. Or you know, before the magnetosphere
of the Earth was discovered, you'd have no way to
explain the northern lights. You'd have to make up some
crazy story about the northern lights and so on. Anyway,
(34:35):
I feel like we're in that situation now with consciousness.
There's something right at the edges. We're all listening for
its whispers. We can sort of feel that there's something there,
but we don't know exactly what it is that we're
missing that will allow us to explain how you take
a bunch of physical stuff and have it experience.
Speaker 1 (35:03):
Now, a topic that's being discussed a lot right now is,
of course, as always artificial intelligence, but specifically generative artificial intelligence,
especially with so many of these text and image creative
tools that are available just to the average person to
experiment with and share the results of. And I was wondering,
(35:23):
what's your take on generative artificial intelligence and how it
relates or doesn't relate to human creativity.
Speaker 4 (35:30):
Yeah, so actually this is my next episode because I'm
fascinated by this. Yeah, I'm just so Okay. So, you
guys may know I'm a neuroscientist, but I'm also a writer,
including of fiction. And so suddenly, when when generatively I
started blowing up really at the end of last year,
I of course, like many artists, thought, oh my gosh,
(35:51):
what does this mean for me? What's the what's the
future for writers? But actually, in my next episode, I
make a four part argument why I think it'll be
an important part of the symbiosis between humans and machines
that eventually comes about.
Speaker 1 (36:08):
But it's not going to.
Speaker 4 (36:09):
Replace writers and artists. There are many reasons. You know,
One thing is it can it can only do short
form stuff, and it does it very nicely. But you know,
if you want to write a little blog post or
a little jingle or poem or whatever, like, it's great
for that, but to actually write a novel is a
completely different sort of thing because what the author is
doing there is planting clues and having let's say, a
(36:31):
cliffhanger that doesn't come back for two or three chapters,
and you know, there's a continuity through time where what
the author knows is what the end of the story
is and then writes towards that. But AI can't even
do things like make up a joke, because to make
up a joke you have to know the punchline first
and then construct the joke to meet it. But it's
(36:51):
doing everything in the forward direction, so there are there
are reasons like that. There's also make the argument that
we as readers, I think, actually really care about the
heartbeat behind the page, which is to say, if you
offered me two books, and one was written by AI
and one was written by you, Rob, and you know,
(37:13):
I would absolutely want the one that's written by a
real human, because I know that you're a human with
all the you know, limitations and anxieties and joys and
ecstasies of a real human. And and that's what I
care about as a fellow human. And you know, part
of my evidence for this is a colleague of mine
(37:35):
here in Silicon Valley announced recently that he'd written a
book that was half by him and half by chat
gpt And and I actually read most of the book
and it's it's actually a good book, but I was
not inspired when I heard that I read it for
other reasons. I thought that sounds terrible, and I was
(37:55):
trying to figure out why why did I feel that way?
Why did I feel that it was interesting to me?
And it has to do with this heartbeat behind behind
the page that that matters. Here's the analogy that I'm
thinking about nowadays, is it was when cameras first came
on this scene, visual painters all panicked and thought we're
(38:19):
done for because why would anyone want me to sit
here and paint something for weeks and weeks when you
can just get a perfect representation of it in a
fraction of a second with a camera. And the answer is,
cameras did not kill visual painting. They just ended up
filling a different neighboring niche and became their own art form.
But visual painting still exists because you could do other
(38:40):
things with it, and I can you know, at least
right the moment, all these text generation programs are extraordinarily
boring in what they come up with, because they get
pushed through reinforcement learning with humans so that humans say, oh,
don't say that, don't say that, that might defend someone,
and so on, which is fine. I mean, I'm not
(39:00):
opposed to that. But the thing is that good literature
is stuff that really challenges us. Any good piece of
literature that you find is something that's full of stuff wethink.
Oh yikes, that sounds like a terrible thing that just happened.
And none of these large language models are even willing
to go near that or touch that. So I think
they're going to be quite a distance from real literature
(39:21):
for the foreseeable future.
Speaker 2 (39:23):
I would tend to think also with literary craft. A
lot of what we really like about literary style is
being surprised. But I wonder if a you know, surprised
by like a strange word choice or a strange metaphor
or something. Those are the things that feel really good.
But can can a generative AI tell the difference between
(39:46):
a comparison or a word choice that is strange and
a pleasing and exciting way versus one that will be
essentially interpreted as a hallucination or an error by the AI.
Speaker 4 (39:58):
That's interesting, I would say, I mean, say something positive
about these ais. I think probably it would be able
to do that, because remember, all us doing is a
statistical game of saying, Okay, what's the most probable thing
to come next, and you can turn up the temperature
on it so that it does things that are increasingly
less probable but somehow makes sense. I had not thought
(40:23):
about that, But I think these things might be great
at making really good metaphors that are surprising, because one
of the things that authors have to deal with all
the time is that they recycle metaphors and it's totally
you know, a soporific to the reader. It puts them
to sleep. But a good author I was just reading
(40:44):
the other day. It was name Frank Herbert who in
Dune he said something about the waves throwing white robes
over the rocks. That's how he was describing the foam
hitting the rocks, which is beautiful because it wakes you
up in that moment you think, oh, what a nice
way of describing that. But if one of these large
language models simply says, hey, I want to make something
(41:06):
not the most probable thing, but less probable, less probable,
I'll bet it could come up with really good stuff
like that that no human author has yet tried. Let
me just give it one example, which is when Alpha
Go beat le so Dole, the Go champion back and
so I think twenty seventeen. You know, here's the best
(41:27):
human in the world of playing the game of Go,
and the AI program beats him, and everybody sort of
watched that and thought, wow, wow, that's the end of that.
But the most interesting part of the story was what
happened next, which is le Sodole ended up then playing
against his his human opponents and took on different sorts
(41:53):
of moves that he had seen Alpha Go play that
no human had played before, like it was just doing
these weird things that were totally in the rules. They
were legal, but no one had thought of doing it before.
So now he started doing this and started beating his
human opponents at a much higher rate. So the point
is we can learn from AI, and I think there's
(42:14):
going to be this really interesting collaboration that happens into
the future where we see new things happening and in
this case, new metaphors that come out in literature and
we think, wow, I would have never thought of that one,
and then we can use them.
Speaker 2 (42:29):
Another element of literary style, though, I think about with
Generative AI is the role of insight in writing, and
it makes me wonder what insight actually is. This is
obviously something we prize, you know, when we read a
novel that we really like and we say that it
is true, you know there's something true in it. Obviously
(42:50):
the story is literally fictional, it didn't happen, but it
observes something about life that we perceive as like deeply
correct and do I don't know, I would have an
intuition that says I would come across insights like that,
or things that feel like insights like that. Less in
generat in something generated by AI. I can't prove that,
(43:11):
but it does raise this question of what insight is.
Speaker 4 (43:15):
Yeah, you know, I tell you, I think I'm signing
with the A on this one. Because what these large
language models are essentially is every human all put together.
So whatever insights people have had previously, this is all
available to the language model, and so there's no reason
that it can't put something together that's very insightful. And
(43:36):
it's not that it's having the insight, it's that it
gets to say, Okay, well, here's a billion people who
have written stuff down, and I've noticed that, you know
a number of these maybe two hundred of these people
have all said the same thing over here. And maybe
Joe's never read that sentence, you know, that paragraph, but
but I there's something going on over here, and it
(43:57):
puts it together and then you say, oh my gosh,
that was really insight because it's not a machine telling
you the story, it's a billion people telling you the story.
Speaker 1 (44:05):
Now, Joe and I are actually currently doing some episodes
of our own podcast on the subject of infantile amnesia.
You know, why we don't remember our earliest childhood or
you know, infancy and birth and so forth. And and
we've heard from some listeners, as we inevitably knew we would,
who say that they do remember their birth, so they
(44:27):
do remember very early childhood. And we were just wondering
what your take is on people who who have that
experience or seem to have that memory. What may be
going on there?
Speaker 4 (44:40):
Yeah, I mean, here's here's what we think in there.
Science generally is that, you know, memory is a is
something that unpacks slowly with time. It's a cognitive development
in some sense. And you know, as you guys know,
it's about three years old for girls and three and
a half years old for boys that they start laying
down their first memories. Here's the interesting thing. Memory is
(45:01):
a myth making machine, and we're constantly reinventing our past.
And so one of the difficult things to assess when
someone says, hey, I remember whatever being born or being
one years old, is it's really difficult to know the
degree to which they think that's true. But it's not
true because we're all told stories by our parents of oh,
(45:23):
and you were an infant, you did this hilarious thing
and blah blah blah, and you hear the story once
or twice, and eventually it becomes a false memory. So
I think it's very difficult to know to be able
to tell this sort of thing. And of course for
someone who has a memory, it's very difficult to tell them, Hey,
that might be false and you just think you remember that.
(45:47):
That makes people angry. But you know, the truth is,
this kind of stuff comes up all the time in
courts of law, in the realm of eyewitness testimony, because
people think that they're memories are like a video recorder,
and they're simply not. There's a giant psychology literature on
(46:07):
this showing all kinds of ways that things get false,
memories get introduced, and so on. You know, a colleague
of mine did a really great study right after September eleventh,
two thousand and one. She was in New York, and
she went and interviewed a bunch of people in downtown
in Midtown, New York about what they had just seen
(46:30):
on September eleventh, and then she was clever enough to
also ask them to describe a memory from September tenth,
the day before, like lunch here, and I did this,
and I did that, And then she went and tracked
all these people down a year later and asked them
to tell their memories again about September eleventh, September tenth,
of the year before, and it turns out that in
(46:50):
both cases the memories drifted. So this comes back to
the beginning of our conversation. Even Amigdala memories, even the
scariest memories that you have, it doesn't mean they're accurate.
And so I mean, especially it doesn't surprise me about
September eleventh, because especially the more you tell a story,
the more you start laying down these ruts in the road,
and that becomes the story that becomes the truth. And
(47:13):
you know, we've all run into these things in our
life where someone suddenly shows us a photograph or something
that says, wait, that's not my memory. Look here's the
thing here, and you go, oh, gosh, I had actually
misremembered that that thing that happened, or where I was
standing or what I was doing. Anyway, So this is
the concern when people say, oh, I remember whatever being
born or this event when I was really young, is
(47:34):
that we know how easy it is to believe memories
that are not true.
Speaker 2 (47:40):
Now correct me if I'm wrong, But I think I
recall reading that. In some of these cases with the
like where there's a big public event and people remember you,
people are like asked to write down their experiences that day,
and then the researchers contact them again later and have
them try it again. You say, what do you remember
about that day? Not only do they often get details wrong,
(48:00):
but don't they often insist that the way they remember
it now is correct and what they wrote at the
time was wrong.
Speaker 4 (48:07):
Exactly. That's exactly right, Yeah, because it's so hard to
disbelieve our own memories about things. You know, this is
obviously at the heart of lots of spousal arguments too.
You know, you have two brains, you have two different
ways of remembering what precisely happened. Yes, that's exactly right.
Speaker 1 (48:26):
All right, Well, you already mentioned that you have the
episode coming up about AI creativity. I'm definitely excited to
check that one out. Is there anything else you want
to tease out for listeners what else they can expect
from future episodes of Inner Cosmos? Yeah?
Speaker 4 (48:41):
Well, okay, so my next one after that is going
to be is a I sentient Because you know, this
is a big question now as these models get larger
and larger, and things are moving at an extraordinary pace. Now,
what does sentients mean? And this is related to the
question you asked me Rob about consciousness and so on.
So I think this actually gives an interesting tool into
(49:02):
studying consciousness that we haven't had before. But I have
other episodes about my One after that is about counterfeiting
money and what it is that we notice about counterfeits
or we do not. I have an episode on will
you perceive the event that kills you? And I find
(49:23):
this is just a topic I've been thinking about for
a long time and have put together a lot of
work on this about you know, if suddenly something, let's
say that the brick from the pedestrian bridge over the
highway fell on your head when you were in a convertible,
The question is, would you perceive dying or would you
be dead before you knew anything happen? And what does
(49:44):
that look like? Does it look like you know, suddenly
the footage just ends but there's no pain, stuff like that.
So I have lots and lots of episodes. Can we
create new senses for humans? Which is a big part
of what I've been doing over the last eight years
with company that I run called Neosensory. Yeah, and I
(50:04):
have forty six episodes this year, all of which I've outlined,
and then it's just a matter of spending the twelve
hours per week of writing the hour long monologue.
Speaker 1 (50:15):
Awesome, it sounds exciting. I'm excited to check out more episodes.
Speaker 4 (50:20):
Great. Thank you guys so much for having me. It's
been a pleasure to see all.
Speaker 1 (50:25):
Yeah, thanks for coming on the show.
Speaker 2 (50:29):
All right, well that was our conversation with David Eagleman.
Once again, much appreciation to David for taking the time
to chat with us today. If you want to check
out his new show, and we do recommend it once again,
it is called Inner Cosmos with David Eagleman. You can
find it on the iHeart app or wherever you get
your podcasts.
Speaker 1 (50:47):
Just a reminder that's Stuff to Blow Your Mind is
a science podcast with core episodes on Tuesdays and Thursdays
in the Stuff to Blow your Mind podcast feed. On
Mondays we do listener mail episodes. On Wednesdays we do
a short form artifact or monster fack episode, and on
Fridays we set aside most serious concerns to just talk
about a weird film on Weird House Cinema.
Speaker 2 (51:06):
Huge thanks to our audio producer JJ Posway. If you
would like to get in touch with us with feedback
on this episode or any other, to suggest a topic
for the future, or just to say hello, you can
email us at contact at stuff to Blow your Mind
dot com.
Speaker 3 (51:26):
Stuff to Blow Your Mind is production of iHeartRadio. For
more podcasts from my Heart Radio, visit the iHeartRadio app,
Apple Podcasts, or wherever you're listening to your favorite shows.
Hey, welcome to Stuff to Blow Your Mind. My name
is Robert Lamb.
Speaker 2 (00:09):
And I am Joe McCormick. And today we are bringing
you an episode from the Vault. This one originally published
April thirteenth, twenty twenty three, and this was our conversation
with David Eagleman, neuroscientist and host of the Inner Cosmos podcast.
I remember this was a very interesting chat. We hope
you enjoy it.
Speaker 3 (00:32):
Welcome to Stuff to Blow Your Mind, a production of iHeartRadio.
Speaker 1 (00:41):
Hey, welcome to Stuff to Blow your Mind. My name is.
Speaker 2 (00:44):
Robert Lamb and I'm Joe McCormick, and hey, welcome back, Rob.
You were out sick earlier this week. It's good to
have you back.
Speaker 4 (00:51):
It's good to be back now.
Speaker 2 (00:53):
Because you were outsick, we ended up putting a pause
on an ongoing series we were doing on childhood amnesia.
We ended up running a Vault episode on Tuesday, and
I just wanted to assure people that we will be
coming back to that subject. We will be resuming the series,
probably for next week's core episodes, but because we already
had it scheduled out this way, today's episode is going
(01:15):
to be an interview. So in fact, we are talking
to a return guest, the neuroscientist and author David Eagleman.
This is actually the second time David has been a
guest on the show. In September twenty twenty, Rob, you
spoke to him about his book Live Wired, which is
a popular science book on the subject of brain plasticity.
Speaker 1 (01:36):
Yeah. Absolutely. That was a fun episode. You can find
it in the archives, and the book Live Wired is
an absolute delight. If you're at all interested in anything
you hear us discussing in this episode, pick up a
copy of it.
Speaker 4 (01:48):
It's great.
Speaker 2 (01:49):
So this week we invited David back on the show
because he now has a fantastic, brand new podcast on
our very own network on the iHeart Network, and it
is called Inner Cosmos with David Eagleman. So before we
get started with our interview, I thought we should just
share a bit of background about David. This is from
(02:10):
his website.
Speaker 1 (02:11):
That's right. David Eagleman is a neuroscientist at Stanford University
and an internationally best selling author. He's the co founder
of two venture backed companies, Neosensory and brain Check, and
he also directs the Center for Science and Law, a
national nonprofit Institute. He's best known for his work on
sensory substitution, time, perception, brain plasticity, synesthesia, and neurolaw. He
(02:35):
is the writer and presenter of the international PBS series
The Brain with David Eagleman and the author of the
companion book, The Brain The Story of You. He's also
the writer and presenter of The Creative Brain on Netflix.
Speaker 2 (02:47):
David Eagleman is the author of over one hundred and
twenty academic publications and many many books of popular science.
Eagleman is a TED speaker, a Guggenheim Fellow, and serves
on several boards, including the America Brain Foundation and the
long Now Foundation. He's the Chief Scientific Advisor of the
Mind Science Foundation and the winner of the Claude Shannon
(03:09):
Luminary Award from Bell Labs and the McGovern Award for
Excellence in Biomedical Communication. He serves as the academic editor
for the Journal of Science and Law. Was named Science
Educator of the Year by this Society for Neuroscience and
was featured as one of the quote Brightest Idea Guys
by Italy's Style magazine. He served as the scientific advisor
(03:31):
on several TV shows, including Westworld and Perception and has
been profiled on The Colbert Report, Nova Science Now, The
New Yorker, CNN's Next List, and many other venues. He
appears regularly on radio and television to discuss literature and science,
and I guess now he is going to start having
to add podcasts to the end of his bio here, So, Robin,
(03:54):
unless you have anything else, I think we should jump
right into our conversation with David Eagleman.
Speaker 1 (04:01):
Hi, David, welcome back to the show.
Speaker 4 (04:03):
Great, Thanks Rob for having me again. It's a pleasure
to be here. And hello, Joe, it's great to meet you. David.
Speaker 1 (04:08):
Now, you have a great new podcast series out Inner
Cosmos through iHeart. How did you decide what path to
take with the podcast format?
Speaker 4 (04:18):
You know, it's a great question. The truth is I
had not listened to many podcasts at all. Now I have.
But when I first was putting this together with iHeart,
I thought, look, I want to do a forty five
minute to hour long monologue every week. And that seemed
like a terrific idea at first, and then my wife
said she was going to kill me because it turns
(04:39):
out that's a ton of work. It takes me about
kind of about twelve hours a week to get a
good monologue that's almost an hour long. So that's how
I decided on the format because I thought it would
be something special rather than you know, I've been on
many different podcasts where we're doing interviews just like this,
(05:01):
and it's super fun. But I wanted to do something different.
So that's how I accidentally stumbled into that format.
Speaker 2 (05:08):
So I figure we should give people a taste of
the kind of things you talk about on your show.
I got a chance to listen to the episode you
did about memory and the perception of time, and I
thought it was great, by the way, a really great
way to kick off the show. So your starting premise
in that episode was that many people who have been
through intense or life threatening events, maybe falling off of
(05:30):
a building or seeing a car speeding toward them, report
afterwards that time seemed to have somehow slowed down for
them during the pivotal few seconds, almost as if they
were able to inter a state of slow motion or
bullet time, like from the matrix. Can you talk a
bit about your research on this subject and what you
(05:50):
discovered about this perception.
Speaker 4 (05:52):
Yeah, so my research, of course started off very personal
which is that I fell off of a house and
it seemed like things took a long time, and that's
what got me interested in this. Then when I got older,
it became a neuroscientist. Eventually I realized that I was
hearing this story not uncommonly from people who had been
in a you know, gunfight or some scary situation or
(06:14):
car accident or whatever, and they felt that things took longer.
And so I looked in the literature and there was
not anything on this. So that's when I came to
realize I was going to have to do this myself
and figure out how to run an experiment on this.
So you know, briefly, what I did is I built
a device that I hooked to people's wrists that flashed
information them at them met a certain way so that
(06:36):
I could tell how rapidly their brain was perceiving, and
that way I could test whether they were actually seeing
in slow motion or The whole thing was a trick
of memory, meaning when you were in an intense situation,
you wrote down more memories. So when you said what
just happened, when just happened, that it seemed like it
must have taken longer because you have all these memories.
(06:59):
So what I did then is dropped people from one
hundred and fifty foot tall tower in free fall backwards
into a net below, and I measured their perception of
time on the way down this way. And what I
found after, you know, I didn't sell myself, but then
we dropped twenty three participants. What I found is that
(07:20):
it is in fact a trick of memory, which is
to say, when everything is hitting the fan, your brain
writes down much denser memory, and when you read that
back out, your brain has to make an assumption about
you know, how much memory, how much footage maps onto
how much time, and so it says, oh wow, that
must have been five seconds, even though it was only
(07:40):
one second worth. But the point is people were not
able to see in slow motion, which, by the way,
was disappointing for me because I already had. I was
already talking with the military about building cockpits in a
way that flashed information more rapidly at people when they
be in some intense situation. But it turns out none
of that makes the difference. You can't actually get information
there faster, you can only remember it faster, So.
Speaker 2 (08:04):
There's not actually any any increased ability of perception. It's
just a trick of the memory exactly.
Speaker 4 (08:12):
Now. It is the case that you know a lot.
You can do a lot of things pre consciously, by
which I mean your conscious awareness of something is always
the slowest thing on the ladder to ever get any information.
So by the time your brain puts together all the
signals and says, okay, this is what just happened. You know,
that's at least half a second to a second behind
(08:33):
real time. But the point is your body can react
much faster than that. Your body can get signals and say, whoa,
I got to do something about this right away, And
so you can react, you know, often much faster than
you can be consciously aware. So you know, I don't
know if you've been on a I mean, this is
what happened to me recently. I was on a hike
with a friend and a branch snapped back, and I was,
(08:54):
you know, halfway into the move of ducking out of
the way of the branch before I consciously realized it,
or my foot gets halfway to the break of my
truck before I realized that there's a car pulling out
of the driveway ahead of me. In other words, consciousness
is always the last guy on the ladder to get
any information, and your body can almost always react much faster.
(09:15):
In fact, wait, let me just say one more example
of that, which is when I was younger, I used
to play baseball, and my experience was always that, you know,
I'd be waiting for the pitch, and then I would
realize after it had happened that I had already hit
the ball. And I would consciously realize, oh, I have
just hit the ball. Now throw the bat and run.
(09:38):
But you know, the whole thing, the ball moving from
the mound to the plate, and the swing and the batting,
that's all really fast process and it often happens pre consciously.
Speaker 2 (09:48):
Somewhat related to that, this raises questions about the different
kinds of circumstances that would favor the perception of slow
motion in intense situations or not.
Speaker 3 (10:01):
So.
Speaker 2 (10:01):
My example was there was one night years ago I
was driving under an overpass and there was a sudden
deafening sound and a shudder, And what my wife and
I deduced later was that somehow like a brick had
fallen from above and hit the roof of our car
just above the windshield. As we passed under a highway
speed and I don't know if somebody threw it or
(10:21):
if it somehow just fell, But I not only don't
recall a feeling of stretched time or a greater density
of memories right before and after the impact, I felt
almost a kind of retrospective amnesia, like a real paucity
of detail. And it was like we were suddenly a
good ways down the road and just trying to remember
(10:42):
or figure out what had happened.
Speaker 4 (10:45):
Yeah, that's exactly right. It's because you didn't write down
any memory. And this is generally because as you are
taking a drive down the highway, your brain is writing
down very little stuff going on. In fact, the interesting
part is that although we think about memory as being
like a video recorder or something, in fact it's nothing
like that. You write down very little of what happens
(11:06):
in your life, especially when you're driving on a road
you've been down before. So what happened is there's the
deafening crash and suddenly you're thinking what just happened? What
just happened? And you've got nothing to draw on. There's
just no footage there. And by the way, just as
a very quick side note, I think this is what
happens to people when they are high on marijuana, is
they say, oh, my gosh, how long have I been
(11:28):
standing here? It feels like I'm standing here forever, And
it's because they're not writing down memories in the same way.
So when their brain looks for how long have I
been standing here, what it's looking for is footage in time,
as in, Okay, I remember getting here, I remember this happening,
someone said this. Then someone put the glass down and
blah blah blah, so that it can estimate how long
(11:48):
it's been there. But suddenly it can't grab on to
any memories at all, and so suddenly people are lost
in time anyway, So this is Joe exactly what happened
pins when people suddenly are hit by a car that
they don't see coming, like a car teet bones them
or something. Or I might have mentioned in the podcast,
(12:09):
I can't remember that, you know. I was once riding
my bike and the wheel suddenly dropped in a pothole
and I went flying over the handlebars. But because I
didn't see that coming, I just had the sensation of suddenly,
oh my god, what's is you know?
Speaker 1 (12:23):
Here?
Speaker 4 (12:23):
I'm lying on the asphalt, bloody, and I have no
idea what just happened. And it's precisely because you're not
running down any memories. So when your brain says what
just happened, what just happened, there's nothing to draw on,
as opposed to the brick sliding on ice towards the
brick wall phenomenon, which is where you say, oh my gosh,
I'm predicting what's going to happen and this is really
gonna hurt, this is gonna be bad, And that's when
(12:44):
you're writing down lots of stuff.
Speaker 2 (12:46):
So would you say you're more likely to have this
memory density perception number one if you see the event
coming ahead of time, there's there's expectation of it. But
also if you're generally in a novel or unusual situation.
Speaker 4 (13:03):
Yeah, that's exactly right. Actually, So two aspects of that.
One is that I just mentioned a moment ago that
you write down very little memory, and that's because as
an adult now, your brain has sort of figured out
a pretty good model of the world, meaning you don't
need to write stuff down because you've seen all the
personalities before, you've seen different cities before, you've seen roads,
(13:25):
and people and events and television shows and you sort
of got it. But if something really novel happens, that's
when your brain writes something down and says, WHOA, wait
a minute, I'm surprised, And that's when stuff starts getting
written down. So when you look back at the end
of let's say a novel event. Let's say you go
on some really wild trip on the weekend to Glockos
(13:46):
Islands and you see new things and so on, then
it seems like forever since you were at work on Friday.
But if you just go off for a normal weekend
and you come back to work, you think, oh, I
was just here because you didn't lay down any new
memories over the weekend. So it is true that things
that are novel generally seem to last longer. However, it
should be noted that when things are actually life threatening,
(14:10):
you have essentially an emergency response memory system that kicks
into gear. That is a secondary track on which you
write down memory. And this is underpinned by a part
of the brain called the amigla, and its job is
to say, WHOA, everything is going really bad and scary here,
and I got to write this down because that after
(14:31):
all is the point of memory is to make sure
that you write down stuff that is important and specifically
life threateningly important.
Speaker 2 (14:38):
So if the normal memory system, would that be the
hippocampal memory system exactly? If that's the normal memory system,
and then the amygdala tends to be recruited in intense situations.
Do we know generally if there is any if there
are any characteristic differences between how memories are recorded in
(15:00):
the hippocampus versus the Amygdalah.
Speaker 4 (15:02):
Yeah, And it turns out it's a tragic one, which
is that amygdala memories are uneaseable, whereas hippocampbel memories can
be erased. So let me unpack this because there's two
surprising parts here. So first of all, the fact that
hippocampal memories can be erased is terrifying and weird and wild.
(15:23):
Which is, if I ask you to recall the name
of your fifth grade teacher and then suddenly that brick
drops off the highway bridge and hits you in the head.
God forbid, let's say that happens. You will now have
amnesia for that one fact, You will not remember anything
(15:45):
about your fifth grade teacher anymore about the at least
the fifth grade teacher's name. Why it's because the name
of your fifth grade teacher is stored deep in the
structure of your brain, and when I ask you to
recall it, you're actually transferring it from that structural form
into activity, you know, spikes in the brain. And that's
(16:05):
how you're remembering the name of your teacher. Now, when
you're done remembering it, it has to get reconsolidated back
into its physical form. And if you get hit in
the head during that moment, it's gone. It's now you know,
it's been transferred from the physical to the you know,
activity in spikes. And if you you know, before it
(16:27):
gets transferred back into the physical, it can it is
susceptible to erasure, which is weird and terrifying. And by
the way, this can also be done with protein synthesis inhibitors.
So people do this in rats. They've been doing this
for decades, where you know, you train a rat how
to run different maases and then you put the rat
on a particular maze where the rat has to remember, oh, yeah,
(16:48):
that's this one, and then you just feed the rat
a protein synthesis inhibitor and now it cannot reconsolidate that
memory into physical form. So number one is hippocampal memory
can be erased. The number two point is that amigla
memories cannot be erased, which is to say, when you
recruit the emergency control system to say, wow, this is
(17:10):
really important, write this down. Than those are permanent, which
the reason I say that's unfortunate is because those are
the ones that people want to erase. In other words,
you know, let's say a rape victim or something like that,
that is the one thing that she wants to forget
more than anything, but cannot.
Speaker 1 (17:28):
Now, I was checking out the show as well, and
I was listening to your I believe this is an
episode from just earlier this week on the topic of
animal uplift, which I don't think is a term that
I was familiar with. Can you give us a brief taste,
a brief overview of what animal uplift is.
Speaker 4 (17:45):
Yeah, it's this idea that you know, look, the human
brain is made out of exactly the same stuff that
a mouse brain, the dog brain, the giraft brain. You know,
it's all the same stuff. It's got the same anatomy,
same general structure. We just have more of this wrinkly
outer bit called the cortex. But somehow we are you know,
we've taken over the whole planet as a species. We've
(18:07):
gotten off the planet. We've made vaccines and internet and
quantum mechanics and so like. There's some real difference in
what we are doing versus our neighbors in the animal kingdom.
But the genetic differences, as you know, are not that much.
I mean, we have enormous similarity with almost every Like
if you're building a giraffe, you got to build the
heart and the lungs and the brain and then the
(18:29):
esophagus and all that stuff is really the same stuff.
And so it's just some small algorithmic difference in the
DNA that's making our brain run in a more souped
up way. Okay, the idea of animal uplift is if
we can figure that out. And this won't happen, you know,
for at least a few more decades, but if we
can figure out, ah, here's the sequence of a's and
(18:50):
c's and t's and g's that gives us this high intelligence.
The question is should we give this to animals? Should
we help animals become intelligent? Now, let me just mention
this is an area that bioethicists and philosophers and neuroscientists
have been talking about for a while, and there's you know,
(19:12):
plenty of debate about it. And on one end of
the spectrum you have people say that's a terrible idea,
We wouldn't want to give intelligence to animals, and other
people say it's a moral obligation in the same way
that you know, if we know how to fix some
viral disease or fix a broken leg or something, of
course you should do this for your dog instead of
let your dog, you know, not have the medical advances
(19:33):
that we have made. So anyway, it's a big debate,
but this is the idea of animal uplift. You make
an animal as intelligent as a human. And I just
find this area fascinating. And you know, as I proposed
the podcast, what would the consequences of this be in
terms of, you know, will World War five be fought
(19:55):
by other animal species not just humans? You know, And
the way I sort of introduced the podcast is with
this question of what will my kids look back on?
Are my grandkids Obviously there's lots of things that will
be very different about our world right now and their world,
and let's say fifty years from now. But one of
(20:17):
them is will they look back and say, Wow, I
can't believe there was a time when humans were the
only species on Earth that was really doing anything, and
now we've got all these other you know, crows running universities,
and donkeys' programming computers and whatever, gophers in the Senate
and so on.
Speaker 2 (20:47):
So one idea of yours that I came across because
Rob sent it to me and I found really interesting
was from a paper you published in twenty twenty one,
I think maybe in Frontiers and Neuroscience, offering a hypothesis
about the adaptive function of dreams, which you call the
(21:07):
defensive activation theory, Could you lay out what is the
basic controversy about the biological function of dreams and how
your proposed solution here would answer this question.
Speaker 4 (21:21):
Yes, so it turns out there is no controversy about
the purpose of dreams because nobody knows, right, everyone, I
mean in the sense of everyone's got a little hypothesis
about it. But really it's complicated, and people think, well,
maybe it has something to do with learning and memory.
Maybe it just has to do with you know, energy restoration.
(21:41):
Maybe it has to do with you know, obviously the
Freudians thought that there was some important meaning in the
content of dreams and so on, but no one really knows,
and certainly no one has a quantitative hypothesis that can
make predictions about dreams and how much dream time we have.
But my student and I developed a theory that actually
(22:02):
does make quantitative predictions across animal species about it predicts
actually how much each animal species will dream. And to
explain something to take one step back, which is about
brain plasticity, which is this term that we use to
explain that the brain is very malleable, the human brain
in particular, and it's constantly reconfiguring its own circuitry and
(22:25):
that's how it learns and remembers, and that's how it
learns new skills and all that. So it turns out,
this is what my last book, Live Wired was about,
is the massive flexibility of the brain. It turns out
is probably a lot of people already into it. If
you go blind at a young age, the visual part
(22:47):
of your brain gets taken over, and in fact, if
you're born blind, that takeover is complete. The rest of
the territories in your brain involved in hearing and touch
and other things. These all take over what we would
normally think of as the visual cortex, and it's no
longer visual. It's now, you know, subserving other functions.
Speaker 1 (23:07):
Okay.
Speaker 4 (23:09):
One of the surprises in neuroscience was a study that
came out about a decade ago from some colleagues of
Mind Harvard where they put people in a scanner. These
are normally cited people, but they blindfolded them tightly and
they put them in the scanner and they were looking
at their brain's response to touch or to sound or
things like that. And what they found, to their surprise,
is that after an hour, you could start seeing the
(23:32):
first hints of signals in the visual cortex in response
to touch and sounds. So, in other words, the visual
cortex was starting to get annexed from these other territories
that they've touched and sound after one hour. And so
this was a much more rapid kind of movement than
anyone had expected. And so what my student and I
(23:56):
immediately realized is that this is the basis of dreaming.
It's because we are on a planet that rotates, and
we spend half our time in the darkness, away from
the light of our star, and so in the dark.
You can still hear and touch and taste and smell
just fine, but you can't see. And obviously I'm talking
(24:17):
about evolutionary time, you know, not our modern electricity blessed times.
And so what this means is the visual system in
particular has a real disadvantage, which is it is in
danger of getting taken over from the other senses. And
this is because of the brain's great plasticity, and so
(24:38):
as a result, the visual system needs a way to
defend its territory during the night. And that's what dreaming is.
Dreaming is essentially a screen saver. It's making sure that
at night time, when you're curled up in the corner
of your cave, staying out of trouble sleeping, and sleeping
has other benefits too, in terms of energy restoration and
(24:59):
so on. So when that's happening, you know, you can
still feel if some touches your skinner, if you're smelling
something or whatever. All that can still function in the dark,
but you're not seeing anything at all. And so what
happens is you've got this circuitry that just blows activity
into the visual system to make sure it stays active
during the night. Every ninety minutes, you have this wave
of active, random activity that just gets blown in there.
(25:21):
And because we're visual creatures, we see we have full,
rich visual experience even though our eyes are closed and
it's dark out, and it's because we are just making
sure the brain is making sure that it's keeping this
competition going so the visual system doesn't get taken over. Interestingly,
dream sleep is something we find across the animal kingdom,
(25:41):
but what we were able to demonstrate is that it
correlates with how plastic the animal species is. So some
animals drop out of the womb and they figure out
in thirty minutes how to run, how to walk. Very quickly,
they reach adolescence, they can reproduce, so all kinds, you know,
they're just they're obviously very pre programmed, let's just put
(26:02):
it that way. But other creatures like humans, are extremely plastic.
We take forever to learn how to walk, to wean
to reach reproductive age, things like that, precisely because we're
extremely plastic, and so we have lots of dreaming because
we have to protect our visual cortex at night. But
other animals that are, you know, these pre program types,
(26:26):
they have just a tiny bit of visual dreaming, but
not a lot. And by the way, I'll just mention
that the amount of visual dreaming we have goes down
with age. So it's an infant, you're dreaming all the time,
and as you get older and older, you dream less
and less is a fraction of your sleep. And you
know that's just a correlation. But in theory, what that
(26:46):
suggests is, you know, as an infant, your visual system
is very highly at risk of getting taken over, and
as you get older and things get more cemented into place,
it's less at risk of getting taken over, so you
don't need as much screensaver time and only.
Speaker 2 (27:00):
This kind of reminds me of studies I've read on
a related subject, which is a prolonged blindfolding of normally
cited people who apparently it's very common for people under
those circumstances to experience a lot of visual hallucinations. Does
that have any relationship to what you're talking about here?
Speaker 4 (27:18):
It does, Thank you for asking. That's perfect because this
is all part of the defensive activation theory, which is
to say, if a system is used to having data
coming in and suddenly it's not getting that data anymore,
it fights back from the inside it starts producing that
data itself. So one example of this is let's say blindfolding,
(27:39):
or you also see this for example in you know,
when people get thrown in solitary confinement in the dark,
they start having hallucinations both auditory and visual because they're
not getting that data and they're used to it they're
supposed to get. There's also something called Bonet syndrome Charles
Bone syndrome, which is people start losing their vision, but
they don't realize that they're losing their vision because they
(28:00):
start having hallucinations that essentially fill in for them. This
is all the same issue, though, which is that the
brain is used to getting certain inputs, suddenly it's not
getting it anymore, and so it starts generating in itself.
One more example is tonightis which is ringing in the ears.
This typically comes about because somebody loses hearing in some
(28:22):
frequency or some band of frequencies and the brain says,
wait a minute, I'm not hearing anything at twelve thousand
hurts anymore, so I'm going to start making myself and
it starts making this sound by itself. So this all
falls under the defensive activation theory.
Speaker 2 (28:39):
One of the interesting things I recall about the studies
on prolonged blindfolding was that the hallucinations that were reported
were not entirely random. So it wasn't just you know,
people seeing strange scenes play out in front of them.
That they would often hallucinate stuff that you would expect
to see in that place in the room based on
(29:00):
other senses. So like if they heard someone come to
the door of the room, they would hallucinate the image
of that person in the door.
Speaker 4 (29:08):
Yeah, that's perfect. And by the way, I think this
also has a lot to tell us about dream content
because the thing about dreams, I love the way you
put this, because you could, in theory, dream about anything
at all. You could dream that you are in Cambodia
and that you are in the fourteen hundreds and you're
(29:28):
a magician who's doing something. But you know, you tend
to dream about you know, your work and your spouse
and your drive and whatever, you know, things that are
more local to you. And it's precisely because when you
slam random activity into the visual system, the synapse is
the connections that are essentially hot from the day's work,
(29:52):
you know, those those are the things that tend to
get activated, and the association is very loose in a
sleeping dream state, and so what happens is, you know,
things can go off on weird tangents, but physics still
works fine in a dream. You know, rocks don't float
upwards and stuff like that, and so you know, essentially
you're just rebooting things that were there during the day.
(30:16):
And and this is closely related to what you're saying about.
You know, all the associations that your brain builds up
over a lifetime. So you hear the voice and you're
expecting to see that person, and that's exactly what happens. Actually,
I just want to mention one other thing about the dreams,
which is people will often ask me, well, what about
a blind person, how do they dream? And the answer
(30:38):
is blind people also dream because you have this very
ancient circuitry in your head that's blasting activity into the
back of the brain the occipital cortex, which is normally
the visual cortex. And people but if you're born blind,
it's you know, long taken over by hearing in touch,
and so a blind person's dream is all about hearing
in touch. They don't see anything, but they say, oh,
(31:00):
I was, you know, moving through the living room and
I felt the furniture was rearranged, and then there was
a big dog in the corner and I ran from
it and I was scared. And so, you know, they've
got full, rich dream experience. It's just that it is
not visual because that part of their brain is no
longer visual.
Speaker 2 (31:17):
And would you well, that makes me wonder then if
if your hypothesis about the defensive activity of dreaming is correct,
does that mean that that dreaming is now to protect
still what would normally be used for visual processing that
part of the brain, but its role in processing auditory
and other stimuli.
Speaker 4 (31:37):
A great question. No, that it's that these circuits that
underlie dreaming are extremely ancient, and so they are assuming
that you've got perfectly fine vision. And if you don't
have vision for some reason, then the circuits aren't going
to change. They're just doing a basic architectural job of saying, hey, guys,
(31:58):
every ninety minutes, just last some activity into the back
of the brain.
Speaker 1 (32:02):
There.
Speaker 4 (32:02):
That's all they're doing. And they don't know if you're
blind or not.
Speaker 1 (32:05):
Now, speaking about dreams, I guess it's not too much
of a leap to start talking about consciousness. I was wondering,
where do you think we are in terms of coraling, testing,
and even eliminating various theories concerning the nature of human consciousness. Yeah?
Speaker 4 (32:22):
Boy, this still remains to my mind the central unsolved
mystery of neuroscience. What's interesting, by the way, I wrote
an article, the cover article is Discover magazine back in
something like two thousand and six, called ten Unsolved Questions
of Neuroscience. And what's fascinating to me is it's now
twenty twenty three and they are equally as unsolved. I mean,
(32:42):
it's funny because we're making so much progress in the
field in some ways, and yet in other ways we're
just facing some really tough problems. So the consciousness, you know,
what is consciousness is really I think the central one.
And you know, for any listeners who are wondering what
is the question, the question is how do you take
(33:04):
eighty six billion cells and stick them together and hook
them up in such a way that you have private,
subjective internal experience. So you know, the smell of apple
pie and the taste of feta cheese, and the pain
of pain, and the redness of red and so on
how does that happen? Because you know, my laptop computer
(33:28):
has lots of signals running around, zeros and ones running around,
and it's transistors, but presumably it's not experiencing anything. It
can play a YouTube video for me, but presumably it
doesn't find it funny the way I do. And so
this is really the question of consciousness. We don't know
the answer to it. I can just tell you my
(33:51):
general feeling on this, which is when you look at
the history of science, what you find is that in
every era there were big pieces of information missing, and
yet the scientists were in a position of having to
try to explain everything not knowing some other thing. Here's
an example. You know, when the pump was invented, people
(34:13):
suddenly said, oh, I see the heart is like a pump,
and then it was obvious, oh, click falls into place.
But before that, everyone's trying to figure out what the
heck the heart was doing, but no one had the
concept of a pump. Or you know, before the magnetosphere
of the Earth was discovered, you'd have no way to
explain the northern lights. You'd have to make up some
crazy story about the northern lights and so on. Anyway,
(34:35):
I feel like we're in that situation now with consciousness.
There's something right at the edges. We're all listening for
its whispers. We can sort of feel that there's something there,
but we don't know exactly what it is that we're
missing that will allow us to explain how you take
a bunch of physical stuff and have it experience.
Speaker 1 (35:03):
Now, a topic that's being discussed a lot right now is,
of course, as always artificial intelligence, but specifically generative artificial intelligence,
especially with so many of these text and image creative
tools that are available just to the average person to
experiment with and share the results of. And I was wondering,
(35:23):
what's your take on generative artificial intelligence and how it
relates or doesn't relate to human creativity.
Speaker 4 (35:30):
Yeah, so actually this is my next episode because I'm
fascinated by this. Yeah, I'm just so Okay. So, you
guys may know I'm a neuroscientist, but I'm also a writer,
including of fiction. And so suddenly, when when generatively I
started blowing up really at the end of last year,
I of course, like many artists, thought, oh my gosh,
(35:51):
what does this mean for me? What's the what's the
future for writers? But actually, in my next episode, I
make a four part argument why I think it'll be
an important part of the symbiosis between humans and machines
that eventually comes about.
Speaker 1 (36:08):
But it's not going to.
Speaker 4 (36:09):
Replace writers and artists. There are many reasons. You know,
One thing is it can it can only do short
form stuff, and it does it very nicely. But you know,
if you want to write a little blog post or
a little jingle or poem or whatever, like, it's great
for that, but to actually write a novel is a
completely different sort of thing because what the author is
doing there is planting clues and having let's say, a
(36:31):
cliffhanger that doesn't come back for two or three chapters,
and you know, there's a continuity through time where what
the author knows is what the end of the story
is and then writes towards that. But AI can't even
do things like make up a joke, because to make
up a joke you have to know the punchline first
and then construct the joke to meet it. But it's
(36:51):
doing everything in the forward direction, so there are there
are reasons like that. There's also make the argument that
we as readers, I think, actually really care about the
heartbeat behind the page, which is to say, if you
offered me two books, and one was written by AI
and one was written by you, Rob, and you know,
(37:13):
I would absolutely want the one that's written by a
real human, because I know that you're a human with
all the you know, limitations and anxieties and joys and
ecstasies of a real human. And and that's what I
care about as a fellow human. And you know, part
of my evidence for this is a colleague of mine
(37:35):
here in Silicon Valley announced recently that he'd written a
book that was half by him and half by chat
gpt And and I actually read most of the book
and it's it's actually a good book, but I was
not inspired when I heard that I read it for
other reasons. I thought that sounds terrible, and I was
(37:55):
trying to figure out why why did I feel that way?
Why did I feel that it was interesting to me?
And it has to do with this heartbeat behind behind
the page that that matters. Here's the analogy that I'm
thinking about nowadays, is it was when cameras first came
on this scene, visual painters all panicked and thought we're
(38:19):
done for because why would anyone want me to sit
here and paint something for weeks and weeks when you
can just get a perfect representation of it in a
fraction of a second with a camera. And the answer is,
cameras did not kill visual painting. They just ended up
filling a different neighboring niche and became their own art form.
But visual painting still exists because you could do other
(38:40):
things with it, and I can you know, at least
right the moment, all these text generation programs are extraordinarily
boring in what they come up with, because they get
pushed through reinforcement learning with humans so that humans say, oh,
don't say that, don't say that, that might defend someone,
and so on, which is fine. I mean, I'm not
(39:00):
opposed to that. But the thing is that good literature
is stuff that really challenges us. Any good piece of
literature that you find is something that's full of stuff wethink.
Oh yikes, that sounds like a terrible thing that just happened.
And none of these large language models are even willing
to go near that or touch that. So I think
they're going to be quite a distance from real literature
(39:21):
for the foreseeable future.
Speaker 2 (39:23):
I would tend to think also with literary craft. A
lot of what we really like about literary style is
being surprised. But I wonder if a you know, surprised
by like a strange word choice or a strange metaphor
or something. Those are the things that feel really good.
But can can a generative AI tell the difference between
(39:46):
a comparison or a word choice that is strange and
a pleasing and exciting way versus one that will be
essentially interpreted as a hallucination or an error by the AI.
Speaker 4 (39:58):
That's interesting, I would say, I mean, say something positive
about these ais. I think probably it would be able
to do that, because remember, all us doing is a
statistical game of saying, Okay, what's the most probable thing
to come next, and you can turn up the temperature
on it so that it does things that are increasingly
less probable but somehow makes sense. I had not thought
(40:23):
about that, But I think these things might be great
at making really good metaphors that are surprising, because one
of the things that authors have to deal with all
the time is that they recycle metaphors and it's totally
you know, a soporific to the reader. It puts them
to sleep. But a good author I was just reading
(40:44):
the other day. It was name Frank Herbert who in
Dune he said something about the waves throwing white robes
over the rocks. That's how he was describing the foam
hitting the rocks, which is beautiful because it wakes you
up in that moment you think, oh, what a nice
way of describing that. But if one of these large
language models simply says, hey, I want to make something
(41:06):
not the most probable thing, but less probable, less probable,
I'll bet it could come up with really good stuff
like that that no human author has yet tried. Let
me just give it one example, which is when Alpha
Go beat le so Dole, the Go champion back and
so I think twenty seventeen. You know, here's the best
(41:27):
human in the world of playing the game of Go,
and the AI program beats him, and everybody sort of
watched that and thought, wow, wow, that's the end of that.
But the most interesting part of the story was what
happened next, which is le Sodole ended up then playing
against his his human opponents and took on different sorts
(41:53):
of moves that he had seen Alpha Go play that
no human had played before, like it was just doing
these weird things that were totally in the rules. They
were legal, but no one had thought of doing it before.
So now he started doing this and started beating his
human opponents at a much higher rate. So the point
is we can learn from AI, and I think there's
(42:14):
going to be this really interesting collaboration that happens into
the future where we see new things happening and in
this case, new metaphors that come out in literature and
we think, wow, I would have never thought of that one,
and then we can use them.
Speaker 2 (42:29):
Another element of literary style, though, I think about with
Generative AI is the role of insight in writing, and
it makes me wonder what insight actually is. This is
obviously something we prize, you know, when we read a
novel that we really like and we say that it
is true, you know there's something true in it. Obviously
(42:50):
the story is literally fictional, it didn't happen, but it
observes something about life that we perceive as like deeply
correct and do I don't know, I would have an
intuition that says I would come across insights like that,
or things that feel like insights like that. Less in
generat in something generated by AI. I can't prove that,
(43:11):
but it does raise this question of what insight is.
Speaker 4 (43:15):
Yeah, you know, I tell you, I think I'm signing
with the A on this one. Because what these large
language models are essentially is every human all put together.
So whatever insights people have had previously, this is all
available to the language model, and so there's no reason
that it can't put something together that's very insightful. And
(43:36):
it's not that it's having the insight, it's that it
gets to say, Okay, well, here's a billion people who
have written stuff down, and I've noticed that, you know
a number of these maybe two hundred of these people
have all said the same thing over here. And maybe
Joe's never read that sentence, you know, that paragraph, but
but I there's something going on over here, and it
(43:57):
puts it together and then you say, oh my gosh,
that was really insight because it's not a machine telling
you the story, it's a billion people telling you the story.
Speaker 1 (44:05):
Now, Joe and I are actually currently doing some episodes
of our own podcast on the subject of infantile amnesia.
You know, why we don't remember our earliest childhood or
you know, infancy and birth and so forth. And and
we've heard from some listeners, as we inevitably knew we would,
who say that they do remember their birth, so they
(44:27):
do remember very early childhood. And we were just wondering
what your take is on people who who have that
experience or seem to have that memory. What may be
going on there?
Speaker 4 (44:40):
Yeah, I mean, here's here's what we think in there.
Science generally is that, you know, memory is a is
something that unpacks slowly with time. It's a cognitive development
in some sense. And you know, as you guys know,
it's about three years old for girls and three and
a half years old for boys that they start laying
down their first memories. Here's the interesting thing. Memory is
(45:01):
a myth making machine, and we're constantly reinventing our past.
And so one of the difficult things to assess when
someone says, hey, I remember whatever being born or being
one years old, is it's really difficult to know the
degree to which they think that's true. But it's not
true because we're all told stories by our parents of oh,
(45:23):
and you were an infant, you did this hilarious thing
and blah blah blah, and you hear the story once
or twice, and eventually it becomes a false memory. So
I think it's very difficult to know to be able
to tell this sort of thing. And of course for
someone who has a memory, it's very difficult to tell them, Hey,
that might be false and you just think you remember that.
(45:47):
That makes people angry. But you know, the truth is,
this kind of stuff comes up all the time in
courts of law, in the realm of eyewitness testimony, because
people think that they're memories are like a video recorder,
and they're simply not. There's a giant psychology literature on
(46:07):
this showing all kinds of ways that things get false,
memories get introduced, and so on. You know, a colleague
of mine did a really great study right after September eleventh,
two thousand and one. She was in New York, and
she went and interviewed a bunch of people in downtown
in Midtown, New York about what they had just seen
(46:30):
on September eleventh, and then she was clever enough to
also ask them to describe a memory from September tenth,
the day before, like lunch here, and I did this,
and I did that, And then she went and tracked
all these people down a year later and asked them
to tell their memories again about September eleventh, September tenth,
of the year before, and it turns out that in
(46:50):
both cases the memories drifted. So this comes back to
the beginning of our conversation. Even Amigdala memories, even the
scariest memories that you have, it doesn't mean they're accurate.
And so I mean, especially it doesn't surprise me about
September eleventh, because especially the more you tell a story,
the more you start laying down these ruts in the road,
and that becomes the story that becomes the truth. And
(47:13):
you know, we've all run into these things in our
life where someone suddenly shows us a photograph or something
that says, wait, that's not my memory. Look here's the
thing here, and you go, oh, gosh, I had actually
misremembered that that thing that happened, or where I was
standing or what I was doing. Anyway, So this is
the concern when people say, oh, I remember whatever being
born or this event when I was really young, is
(47:34):
that we know how easy it is to believe memories
that are not true.
Speaker 2 (47:40):
Now correct me if I'm wrong, But I think I
recall reading that. In some of these cases with the
like where there's a big public event and people remember you,
people are like asked to write down their experiences that day,
and then the researchers contact them again later and have
them try it again. You say, what do you remember
about that day? Not only do they often get details wrong,
(48:00):
but don't they often insist that the way they remember
it now is correct and what they wrote at the
time was wrong.
Speaker 4 (48:07):
Exactly. That's exactly right, Yeah, because it's so hard to
disbelieve our own memories about things. You know, this is
obviously at the heart of lots of spousal arguments too.
You know, you have two brains, you have two different
ways of remembering what precisely happened. Yes, that's exactly right.
Speaker 1 (48:26):
All right, Well, you already mentioned that you have the
episode coming up about AI creativity. I'm definitely excited to
check that one out. Is there anything else you want
to tease out for listeners what else they can expect
from future episodes of Inner Cosmos? Yeah?
Speaker 4 (48:41):
Well, okay, so my next one after that is going
to be is a I sentient Because you know, this
is a big question now as these models get larger
and larger, and things are moving at an extraordinary pace. Now,
what does sentients mean? And this is related to the
question you asked me Rob about consciousness and so on.
So I think this actually gives an interesting tool into
(49:02):
studying consciousness that we haven't had before. But I have
other episodes about my One after that is about counterfeiting
money and what it is that we notice about counterfeits
or we do not. I have an episode on will
you perceive the event that kills you? And I find
(49:23):
this is just a topic I've been thinking about for
a long time and have put together a lot of
work on this about you know, if suddenly something, let's
say that the brick from the pedestrian bridge over the
highway fell on your head when you were in a convertible,
The question is, would you perceive dying or would you
be dead before you knew anything happen? And what does
(49:44):
that look like? Does it look like you know, suddenly
the footage just ends but there's no pain, stuff like that.
So I have lots and lots of episodes. Can we
create new senses for humans? Which is a big part
of what I've been doing over the last eight years
with company that I run called Neosensory. Yeah, and I
(50:04):
have forty six episodes this year, all of which I've outlined,
and then it's just a matter of spending the twelve
hours per week of writing the hour long monologue.
Speaker 1 (50:15):
Awesome, it sounds exciting. I'm excited to check out more episodes.
Speaker 4 (50:20):
Great. Thank you guys so much for having me. It's
been a pleasure to see all.
Speaker 1 (50:25):
Yeah, thanks for coming on the show.
Speaker 2 (50:29):
All right, well that was our conversation with David Eagleman.
Once again, much appreciation to David for taking the time
to chat with us today. If you want to check
out his new show, and we do recommend it once again,
it is called Inner Cosmos with David Eagleman. You can
find it on the iHeart app or wherever you get
your podcasts.
Speaker 1 (50:47):
Just a reminder that's Stuff to Blow Your Mind is
a science podcast with core episodes on Tuesdays and Thursdays
in the Stuff to Blow your Mind podcast feed. On
Mondays we do listener mail episodes. On Wednesdays we do
a short form artifact or monster fack episode, and on
Fridays we set aside most serious concerns to just talk
about a weird film on Weird House Cinema.
Speaker 2 (51:06):
Huge thanks to our audio producer JJ Posway. If you
would like to get in touch with us with feedback
on this episode or any other, to suggest a topic
for the future, or just to say hello, you can
email us at contact at stuff to Blow your Mind
dot com.
Speaker 3 (51:26):
Stuff to Blow Your Mind is production of iHeartRadio. For
more podcasts from my Heart Radio, visit the iHeartRadio app,
Apple Podcasts, or wherever you're listening to your favorite shows.
Stuff To Blow Your Mind News
Hosts And Creators
Robert Lamb
Joe McCormick
Popular Podcasts
1. The Podium
The Podium: An NBC Olympic and Paralympic podcast. Join us for insider coverage during the intense competition at the 2024 Paris Olympic and Paralympic Games. In the run-up to the Opening Ceremony, we’ll bring you deep into the stories and events that have you know and those you'll be hard-pressed to forget.
2. In The Village
In The Village will take you into the most exclusive areas of the 2024 Paris Olympic Games to explore the daily life of athletes, complete with all the funny, mundane and unexpected things you learn off the field of play. Join Elizabeth Beisel as she sits down with Olympians each day in Paris.
3. iHeartOlympics: The Latest
Listen to the latest news from the 2024 Olympics.