ANDREW HUBERMAN: Welcome to
the Huberman Lab podcast, where we discuss science
and science-based tools for everyday life. [MUSIC PLAYING] I'm Andrew Huberman. And I'm a professor of
neurobiology and ophthalmology at Stanford School of Medicine. My guests today are Mark
Zuckerberg and Dr. Priscilla Chan. Mark Zuckerberg,
as everybody knows, founded the company Facebook. He is now the CEO of Meta, which
includes Facebook, Instagram, WhatsApp, and other
technology platforms. Dr. Priscilla Chan
grad
uated from Harvard and went on to do her medical
degree at the University of California San Francisco. Mark Zuckerberg and
Dr. Priscilla Chan are married and the
co-founders of the CZI, or Chan Zuckerberg Initiative,
a philanthropic organization whose stated goal is to
cure all human diseases. The Chan Zuckerberg Initiative
is accomplishing that by providing critical funding
not available elsewhere, as well as a novel
framework for discovery of the basic
functioning of cells, cataloging all the
different human cell types, as well as providing
AI, or artificial intelligence, platforms to mine
all of that data to discover new pathways and
cures for all human diseases. The first hour of
today's discussion is held with both Dr. Priscilla
Chan and Mark Zuckerberg, during which we discuss
the CZI and what it really means to try and cure
all human diseases. We talk about the motivational
backbone for the CZI that extends well into each
of their personal histories. Indeed, you'll learn quite a
lot
about Dr. Priscilla Chan, who has, I must say, an absolutely
incredible family story leading up to her role as a
physician and her motivations for the CZI and beyond. And you'll learn from Mark, how
he is bringing an engineering and AI perspective
to the discovery of new cures for human disease. The second half of
today's discussion is just between Mark Zuckerberg
and me, during which we discuss various Meta Platforms,
including, of course, social media platforms, and
their effects on menta
l health in children and adults. We also discuss VR,
Virtual Reality, as well as augmented and mixed reality. And we discuss AI,
Artificial Intelligence, and how it stands to transform
not just our online experiences with social media and
other technologies, but how it stands to
potentially transform every aspect of everyday life. Before we begin, I'd
like to emphasize that this podcast is separate
from my teaching and research roles at Stanford. It is, however, part
of my desire and effort to b
ring zero cost to
consumer information about science and
science-related tools to the general public. In keeping with
that theme, I'd like to thank the sponsors
of today's podcast. Our first sponsor
is Eight Sleep Eight Sleep makes smart mattress
covers with cooling, heating, and sleep tracking capacity. I've spoken many times before
on this podcast about the fact that getting a
great night's sleep really is the foundation of
mental health, physical health and performance. One of the key things
to
getting a great night's sleep is to make sure that the
temperature of your sleeping environment is correct. And that's because in order to
fall and stay deeply asleep, your body temperature
actually has to drop by about 1 to 3 degrees. And in order to wake up feeling
refreshed and energized, your body temperature
actually has to increase by about
1 to 3 degrees. With Eight Sleep, you can
program the temperature of your sleeping environment
in the beginning, middle, and end of your night. It h
as a number
of other features, like tracking the amount
of rapid eye movement and slow wave
sleep that you get, things that are essential
to really dialing in the perfect
night's sleep for you. I've been sleeping on
an Eight Sleep mattress cover for well
over two years now. And it has greatly
improved my sleep. I fall asleep far more quickly. I wake up far less often
in the middle of the night. And I wake up feeling
far more refreshed than I ever did prior to using
an Eight Sleep mattress cover.
If you'd like to
try Eight Sleep, you can go to
eightsleep.com/huberman to save $150 off their Pod 3 cover. Eight Sleep currently
ships to the USA, Canada, UK, select countries
in the EU, and Australia. Again, that's
eightsleep.com/huberman. Today's episode is also
brought to us by LMNT. LMNT is an electrolyte drink
that has everything you need and nothing you don't. That means plenty of
electrolytes-- sodium, magnesium and
potassium-- and no sugar. The electrolytes are absolutely
essential for
the functioning of every cell in your body. And your neurons,
your nerve cells, rely on sodium,
magnesium and potassium in order to communicate with
one another electrically and chemically. LMNT contains the optimal
ratio of electrolytes for the functioning of
neurons and the other cells of your body. Every morning, I drink a
packet of LMNT dissolved in about 32 ounces of water. I do that just for
general hydration and to make sure that I
have adequate electrolytes for any activities that day.
I'll often also have an LMNT
packet, or even two packets, in 32 to 60 ounces of water
if I'm exercising very hard and certainly if I'm
sweating a lot, in order to make sure that I
replace those electrolytes. If you'd like to
try LMNT, you can go to drinklmnt.com/huberman to
get a free sample pack with your purchase. Again, that's
drinklmnt.com/huberman. I'm pleased to
announce that we will be hosting four live events
in Australia, each of which is entitled The Brain Body
Contract, during which I
will share science and
science-related tools for mental health, physical
health, and performance. There will also be a live
question and answer session. We have limited
tickets still available for the event in
Melbourne on February 10, as well as the event in
Brisbane on February 24. Our event in Sydney, at
the Sydney Opera House, sold out very quickly. So as a consequence,
we've now scheduled a second event in Sydney
at the Aware Super Theatre on February 18. To access tickets to
any of these
events, you can go to
hubermanlab.com/events and use the code Huberman at checkout. I hope to see you there. And as always, thank you for
your interest in science. And now, for my discussion
with Mark Zuckerberg and Dr. Priscilla Chan. Priscilla, Mark, so
great to meet you. And thank you for having
me here in your home. MARK ZUCKERBERG: Oh, Thanks
for having us on the podcast. PRISCILLA CHAN: Yeah. ANDREW HUBERMAN: I'd like to
talk about the CZI, the Chan Zuckerberg Initiative. I learned about t
his
a few years ago, when my lab was-- and
still is now-- at Stanford, as a very exciting
philanthropic effort that has a truly big mission. I can't imagine
a bigger mission. So maybe you could tell us
what that big mission is. And then we can get into
some of the mechanics of how that big mission can
become a reality. PRISCILLA CHAN: So like
you're mentioning, in 2015, we launched the Chan
Zuckerberg Initiative. And what we were
hoping to do at CZI was think about how do we build
a better futur
e for everyone and looking for ways
where we can contribute the resources that we have
to bring philanthropically and the experiences that
Mark and I have had, for me as a physician
and educator, for Mark as an
engineer, and then our ability to bring teams
together to build the builders. Mark has been a builder
throughout his career. And what could we
do if we actually put together a team to build
tools, do great science? And so within our
science portfolio, we've really been focused
on what som
e people think is either an incredibly
audacious goal or an inevitable goal. But I think about
it as something that will happen if we
continue focusing on it, which is to be able to cure,
prevent, or manage all disease by the
end of the century. ANDREW HUBERMAN: All disease? PRISCILLA CHAN: All disease. So that's important, right? And so a lot of times, people
ask like, which disease? And the whole point is that
there is not one disease. And it's really about taking
a step back to where I always
found the most hope
as a physician, which is new discoveries
and new opportunities and new ways of understanding
how to keep people well come from basic science. So our strategy at CZI is really
to build tools, fund science, change the way basic
scientists can see the world and how they can move
quickly in their discoveries. And so that's what
we launched in 2015. We do work in three ways. We fund great scientists. We build tools-- right
now, software tools to help move science along and
make i
t easier for scientists to do their work. And we do science. You mentioned Stanford
being an important pillar for our science work. We've built what we call
biohubs, institutes where teams can take on grand
challenges to do work that wouldn't be possible
in a single lab or within a single discipline. And our first
biohub was launched in San Francisco, a
collaboration between Stanford, UC Berkeley, and UCSF. ANDREW HUBERMAN: Amazing. Curing all diseases implies
that there will either be a ton of
knowledge gleaned
from this effort, which I'm certain there will be--
and there already has been. We can talk about some of those
early successes in a moment. But it also sort of implies
that if we can understand some basic operations
of diseases and cells that transcend autism,
Huntington's, Parkinson's, cancer and any other
disease that perhaps there are some core principles that
would make the big mission a real reality, so to speak. What I'm basically saying is,
how are you attacking this? M
y belief is that the cell sits
at the center of all discussion about disease, given that
our body is made up of cells and different types of cells. So maybe you could
just illuminate for us a little bit of what the
cell is, in your mind, as it relates to disease and
how one goes about understanding disease in the context of cells
because, ultimately, that's what we're made up of. MARK ZUCKERBERG: Yeah. Well, let's get to the
cell thing in a moment. But just even taking
a step back from that, we
don't think,
at CZI, that we're going to cure, prevent
or manage all diseases. The goal is to basically
give the scientific community and scientists around
the world the tools to accelerate the
pace of science. And we spent a lot
of time, when we were getting started
with this, looking at the history of science and
trying to understand the trends and how they've
played out over time. And if you look over
this very long-term arc, most large-scale
discoveries are preceded by the invention of a new
tool
or a new way to see something. And it's not just
in biology, right? It's like having
a telescope came before a lot of discoveries
in astronomy and astrophysics. But similarly, the microscope
and just different ways to observe things or
different platforms, like the ability to do
vaccines preceded the ability to cure a lot of
different things. So this is the engineering part
that you were talking about, about building tools. We view our goal is to
try to bring together some scientific and e
ngineering
knowledge to build tools that empower the whole field. And that's the big arc
and a lot of the things that we're focused on, including
the work in single cell and cell understanding,
which you can jump in and get into that if you want. But yeah, I think I
think we generally agree with the
premise that if you want to understand this
stuff from first principles-- people study organs a lot right. You study how things
present across the body. But there's not a very
widespread understandin
g of how each cell operates. And this is a big part of
some of the initial work that we tried to do on the Human
Cell Atlas and understanding what are the different cells. And there's a bunch
more work that we want to do to carry that forward. But overall, I think, when we
think about the next 10 years here of this long arc to
try to empower the community to be able to cure, prevent
or manage all diseases, we think that the next
10 years should really be primarily about being
able to measure and
observe more things in human biology. There are a lot
of limits to that. It's like you want to look at
something through a microscope, you can't usually
see living tissues because it's hard to see through
skin or things like that. So there are a lot of
different techniques that will help us
observe different things. And this is where the
engineering background comes in a bit because-- I mean, when I think about this
is from the perspective of how you'd write code or
something, the idea of tryin
g to debug or fix a code base,
but not be able to step through the code
line by line, it's not going to happen, right? And at the beginning of any
big project that we do at Meta, we like to spend a bunch of
the time up front just trying to instrument things
and understand what are we going to
look at and how are we going to measure things so
we know we're making progress and know what to optimize. And this is such a
long-term journey that we think that it actually
makes sense to take the next 10
years to build those
kinds of tools for biology and understanding just how the
human body works in action. And a big part of
that is, cells. I don't know. Do you want to jump and talk
about some of the efforts? PRISCILLA CHAN: Sure. ANDREW HUBERMAN: Could I just
interrupt briefly and just ask about the different
interventions, so to speak, that CZI is in a unique
position to bring to the quest to cure all diseases? So I can think of-- I mean, I know, as a scientist,
that money is necessary but
not sufficient, right? When you have money, you
can hire more people. You can try different things. So that's critical. But a lot of philanthropy
includes money. The other component is you
want to be able to see things, as you pointed out. So you want to know that
normal disease process-- like, what is a healthy cell? What's a diseased cell? Are cells constantly being
bombarded with challenges and then repairing those? And then what we
call cancer is just a runaway train of
those challenges not
being met by the cell
itself or something like that? So better imaging tools. And then it sounds like there's
not just a hardware component, but a software component. This is where AI comes in. So maybe, at some point,
we can break this up into two, three
different avenues. One is understanding
disease processes and healthy processes. We'll lump those together. Then there's hardware--
so microscopes, lenses, digital
deconvolution, ways of seeing things in bolder
relief and more precision. And th
en there's how
to manage all the data. And then I love the
idea that maybe AI could do what human
brains can't do alone, like manage
understanding of the data because it's one thing
to organize data. It's another to say, oh,
this as you point out in the analogy with code,
that this particular gene and that particular gene
are potentially interesting, whereas a human
being would never make that potential connection. MARK ZUCKERBERG: Yeah. PRISCILLA CHAN: So
the tools that CZI can bring to the tab
le-- we fund science, like
you're talking about. There's lots of ways
to fund science. And just to be
clear, what we fund is a tiny fraction of what
the NIH funds, for instance. ANDREW HUBERMAN: So you guys
have been generous enough that it definitely holds
wait to NIH's contribution. PRISCILLA CHAN: Yeah. But I think every funder has
its own role in the ecosystem. And for us, it's
really, how do we incentivize new points of view? How do we incentivize
collaboration? How do we incentivize
open s
cience? And so a lot of our grants
include inviting people to look at different fields. Our first neuroscience RFA was
aimed towards incentivizing people from different
backgrounds-- immunologists, microbiologists--
to come and look at how our nervous system works
and how to keep it healthy. Or we ask that our
grantees participate in the pre-print
movement to accelerate the rate of sharing knowledge
and actually others being able to build upon science. So that's the
funding that we do. In terms
of building, we
build software and hardware, like you mentioned. We put together
teams that can build tools that are more durable
and scalable than someone in a single lab might
be incentivized to do. There's a ton of great ideas. And nowadays, most scientists
can tinker and build something useful for their lab. But it's really
hard for them to be able to share that
tool sometimes beyond their own laptop
or forget the next Lab over or across the globe. So we partner with scientists
to see what i
s useful, what kinds of tools. In imaging, Napari, it's
a useful image annotation tool that is born from
an open source community. And how can we
contribute to that? Or a CELLxGENE, which works
on single cell data sets. And how can we make it build a
useful tool so that scientists can share data sets,
analyze their own and contribute to a larger
corpus of information? So we have software teams that
are building, collaborating with scientists to make
sure that we're building easy to use, durable,
translatable tools across the scientific community
in the areas that we work in. We also have institutes-- this
is where the imaging work comes in-- where we are proud owners
of an electron microscope right now. It's going to be installed
at our imaging institute. And that will really
contribute to the way where we can see
work differently. But more hardware does
need to be developed. We're partnering with
the fantastic scientists in the biohub network to build
a mini-phase plate to increase to
align the electrons through
the electron microscope to be able to increase
the resolution, so we can see in sharper detail. So there's a lot of innovative
work within the network that's happening. And these institutes
have grand challenges that they're working on. Back to your
question about cells, cells are just the smallest
unit that are alive. And your body,
all of our bodies, have many, many, many cells. Some estimate of like
37 trillion cells, different cells in your body. And what are the
y all doing? And what do they look
like when you're healthy? What do they look
like when you're sick? And where we're at right now
with our understanding of cells and what happens
when you get sick is basically we've gotten pretty
good at, from the Human Genome Project, looking at
how different mutations in your genetic
code lead for you to be more susceptible
to get sick or directly cause you to get sick. So we go from a mutation
in your DNA to, wow, you now have Huntington's
disease, for insta
nce. And there's a lot that
happens in the middle. And that's one of the questions
that we're going after at CZI, is what actually happens. So an analogy that I like to
use to share with my friends is, right now, say we
have a recipe for a cake. We know there's a
typo in the recipe. And then the cake is awful. That's all we know. We don't know how the
chef interprets the typo. We don't know what
happens in the oven. And we don't actually
know how it's exactly connected to how the
cake didn't tur
n out or how you had expected it. A lot of that is unknown. But we can actually
systematically try to break this down. And one segment of that
journey that we're looking at is how that mutation
gets translated and acted upon in your cells. And all of your cells
have what's called mRNA. mRNA are the actual instructions
that are taken from the DNA. And our work in
Single-Cell is looking at how every cell in your
body is actually interpreting your DNA slightly
differently and what happens when heal
thy cells
are interpreting the DNA instructions and
when sick cells are interpreting those directions. And that is a ton of data. I just told you, there's
37 trillion cells. There's different large
sets of mRNA in each cell. But the work that we've been
funding is looking at how-- first of all, gathering
that information. We've been incredibly
lucky to be part of a very fast-moving
field where we've gone from, in 2017, funding some
methods work to now having really not complete,
but nearly compl
ete atlases of how the human body
works, how flies work, how mice work at the single-cell
level and being able to then try
to piece together how does that all come
together when you're healthy and when you're sick. And the neat thing about
the inflection point where we're at in AI is that
I can't look at this data and make sense of it. There's just too much of it. And biology is complex. Human bodies are complex. We need this much information. But the use of large
language models can help us act
ually
look at that data and gain insights,
look at what trends are consistent with health and
what trends are unsuspected. And eventually, our
hope, through the use of these data sets that
we've helped curate and the application of
large language models, is to be able to formulate a
virtual cell, a cell that's completely built off of
the data sets of what we know about the human body,
but allows us to manipulate, and learn faster and
try new things to help move science and
then medicine along. A
NDREW HUBERMAN:
Do you think we've cataloged the total number
of different cell types? Every week, I look
at great journals like Cell Nature and Science. And for instance, I saw
recently that, using single cell sequencing, they've categorized
18 plus different types of fat cells. We always think of like a fat
cell versus a muscle cell. So now, you've got 18 types. Each one is going to express
many, many different genes and mRNAs. And perhaps one of
them is responsible for what we see in
advanced
type 2 diabetes, or in other forms of obesity,
or where people can't lay down fat cells, which turns out
to be just as detrimental in those extreme cases. So now, you've got all
these lists of genes. But I always thought of single
cell sequencing as necessary, but not sufficient, right? You need the information, but
it doesn't resolve the problem. And I think of it more as
a hypothesis-generating experiment. OK, so you have all these genes. And you can say, well,
this gene is particularly eleva
ted in the diabetic
cell type of, let's say, one of these fat cells or
muscle cells for that matter, whereas it's not
in non-diabetics. So then of the millions
of different cells, maybe only five of them
differ dramatically. So then you generate
a hypothesis. Oh, it's the ones that
differ dramatically that are important. But maybe one of those genes,
when it's only 50% changed, has a huge effect because of
some network biology effect. And so I guess what I'm
trying to get to here is how does one
meet that challenge. And can AI help
resolve that challenge by essentially placing
those lists of genes into 10,000 hypotheses? Because I'll tell you
that the graduate students and postdocs in my lab
get a chance to test one hypothesis at a time. PRISCILLA CHAN: I know. ANDREW HUBERMAN: And that's
really the challenge, let alone one lab. And so for those
that are listening to this-- and
hopefully, it's not getting outside the scope
of standard understanding or the understanding
we've generated
here. But what I'm
basically saying is, you have to pick at some point. More data always sounds great. But then how do you
decide what to test? PRISCILLA CHAN: So no, we
don't know all the cell types. I think one thing that was
really exciting when we first launched this work
was cystic fibrosis. Cystic fibrosis is caused
by mutation in CFTR. That's pretty well known. It affects a certain channel
that makes it hard for mucus to be cleared. That's the basics
of cystic fibrosis. When I went to med
ical
school, it was taught as fact. ANDREW HUBERMAN: So their
lungs fill up with fluid. These are people who
are carrying around sacks of fluid filling up. PRISCILLA CHAN: Yep. ANDREW HUBERMAN: I've worked
with people like that. And they have to literally
dump the fluid out. PRISCILLA CHAN: Exactly. ANDREW HUBERMAN: They can't
run or do intense exercise. Life is shorter. PRISCILLA CHAN: Life is shorter. And when we applied single-cell
methodologies to the lungs, they discovered an
entirely new c
ell type that actually is affected by
a mutation in the CF mutation, in cystic fibrosis
mutation, that actually changes
the paradigm of how we think about cystic fibrosis. ANDREW HUBERMAN: Amazing. PRISCILLA CHAN: [? Just ?]
[? unknown. ?] So I don't think we know all the cell types. I think we'll continue
to discover them. And we'll continue to discover
new relationships between cell and disease, which leads me
to the second example I want to bring up, is
this large data set that the entire
sci
entific community has built around single cell. It's starting to allow us to
say this mutation, where is it expressed? What types of cell
types it's expressed in? And we actually
have built a tool at CZI called CELLxGENE, where
you can put in the mutation that you're interested in. And it gives you a heat
map of cross cell types of which cell types are
expressing the gene that you're interested in. And so then you can
start looking at, OK, if I look at gene X and I know
it's related to heart dis
ease-- but if you look at
the heat map, it's also spiking in the pancreas. That allows you to
generate a hypothesis. Why? And what happens when
this gene is mutated and the function
of your pancreas? Really exciting way to look
and ask questions differently. And you can also
imagine a world where if you're trying to develop a
therapy, a drug, and the goal is to treat the
function in the heart, but you know that
it's also really active in the pancreas again. So is there going to be
an unexpected
side effect that you should think
about as you're bringing this drug to clinical trials? So it's an incredibly
exciting tool and one that's only
going to get better as we get more and
more sophisticated ways to analyze the data. ANDREW HUBERMAN:
I must say, I love that because if I look at
the advances in neuroscience over the last 15
years, most of them didn't necessarily come from
looking at the nervous system. They came from the understanding
that the immune system impacts the brain. Everyone
prior to that
talked about the brain as an immune-privileged organ. What you just said
also bridges the divide between single cells,
organs and systems, right? Because ultimately,
cells make up organs. Organs make up systems. And they're all
talking to one another. And everyone nowadays is
familiar with gut-brain axis or the microbiome
being so important. But rarely is the discussion
between organs discussed, so to speak. So I think it's wonderful. So that tool was
generated by CZI. Or CCI fund
ed that tool? MARK ZUCKERBERG: We built that. PRISCILLA CHAN: We built it. ANDREW HUBERMAN: You built it. So is it built by Meta? Is this Meta? MARK ZUCKERBERG: No, no,
it has its own engineers. ANDREW HUBERMAN: Got it. MARK ZUCKERBERG: Yeah. They're completely
different organizations. ANDREW HUBERMAN: Incredible. And so a graduate
student or postdoc who's interested in
a particular mutation could put this mutation
into this database. That graduate student
or postdoc might be in a laboratory kno
wn
for working on heart, but suddenly find that
they're collaborating with other scientists that work
on the pancreas, which also is wonderful because
it bridges the divide between these fields. Fields are so
siloed in science-- not just different
buildings, but people rarely talk, unless things
like this are happening. PRISCILLA CHAN: I mean, the
graduate student is someone that we want to empower
because, one, they're the future of
science, as you know. And within CELLxGENE,
if you put in the
gene you're interested in and
it shows you the heat map, we also will pull up the most
relevant papers to that gene. And so read these things. ANDREW HUBERMAN:
That's fantastic. As we all know,
quality nutrition influences, of course, our
physical health, but also our mental health and our
cognitive functioning-- our memory, our ability to
learn new things and to focus. And we know that one of
the most important features of high quality
nutrition is making sure that we get enough vitamins
and mi
nerals from high quality, unprocessed, or
minimally processed sources, as well as enough
probiotics, and prebiotics and fiber to support
basically all the cellular
functions in our body, including the gut microbiome. Now, I, like most everybody
try to get optimal nutrition from whole foods, ideally mostly
from minimally processed or non processed foods. However, one of the challenges
that I and so many other people face is getting enough
servings of high quality fruits and vegetables per
day, as
well as fiber and probiotics that
often accompany those fruits and vegetables. That's why, way back in
2012, long before I ever had a podcast, I
started drinking AG1. And so I'm delighted that AG1
is sponsoring the Huberman Lab podcast. The reason I started taking
AG1 and the reason I still drink AG1 once or
twice a day is that it provides all of my
foundational nutritional needs. That is, it provides
insurance that I get the proper amounts of those
vitamins, minerals, probiotics and fiber to e
nsure optimal
mental health, physical health and performance. If you'd like to try AG1, you
can go to drinkag1.com/huberman to claim a special offer. They're giving away
five free travel packs plus a year's
supply of vitamin D3 K2. Again, that's
drinkag1.com/huberman to claim that special offer. MARK ZUCKERBERG: I just think
going back to your question from before are there going
to be more cell types that get discovered? I mean, I assume so, right? I mean, no catalog of
this stuff is ever-- it
doesn't seem like
we're ever done. we keep on finding more. But I think that that
gets to one of the things that I think are the
strengths of modern LLMs, is the ability to imagine
different states that things can be in. So from all the work that
we've done and funded on the Human Cell Atlas, there
is a large corpus of data that you can now train a
kind of large-scale model on. And one of the things
that we're doing at CZI, which I think is
pretty exciting, is building what we think is one
of th
e largest non-profit life sciences AI clusters. It's on the order of 1,000 GPUs. And it's larger than what
most people have access to in academia that you can do
serious engineering work on. And by basically
training a model with all of the
Human Cell Atlas Data and a bunch of other
inputs as well, we think you'll be able
to basically imagine all of the different
types of cells and all the different states that they
can be in, and when they're healthy and diseased,
and how they'll interact with
different-- interact with each
other, interact with different potential drugs. But I think the state
of LLMs, I think this is where it's
helpful to understand-- have a good understanding
and be grounded in the modern state of AI. I mean, these things
are not foolproof. I mean, one of the
flaws of modern LLMs is they hallucinate. So the question is,
how do you make it so that that can be an advantage
rather than a disadvantage? And I think the way that it
ends up being an advantage is when they h
elp you
imagine a bunch of states that someone could be in, but
then you, as the scientist or engineer, go and validate
that those are true, whether they're solutions
to how a protein can be folded or possible
states that a cell could be in when it's interacting
with other things. But we're not yet
at the state with AI that you can just take the
outputs of these things as gospel and run from there. But they are very good,
I think as you said, hypothesis generators or
possible solution generators
that then you can go validate. So I think that that's
a very powerful thing that we can basically-- building on the first
five years of science work around the Human Cell Atlas
and all the data that's been built out-- carry
that forward into something that I think is going to be a
very novel tool going forward. And that's the type
of thing that I think we're set up to do well. I mean, you had this exchange a
little while back about funding levels and how CZI is just a
drop in the bucket compare
d to NIH. The thing that I think we
can do that's different is funding some of these
longer term, bigger projects. It is hard to galvanize
the and pull together the energy to do that. And it's a lot of what most
science funding is, relatively small projects
that are exploring things over relatively
short time horizons. And one of the things
that we try to do is build these tools over
5, 10, 15-year periods. They're often
projects that require hundreds of millions
of dollars of funding and world-
class engineering
teams and infrastructure to do. And that, I think, is a pretty
cool contribution to the field that I think is-- there aren't as
many other folks who are doing that kind of thing. But that's one of
the reasons why I'm personally excited
about the virtual cell stuff because it just this perfect
intersection of all the stuff that we've
done in single cell, the previous collaborations
that we've done with the field and bringing together
the industry and AI expertise around this. AN
DREW HUBERMAN:
Yeah, I completely agree that the model of science
that you're putting together with CZI isn't just
unique from NIH, but it's extremely
important that the independent
investigator model is what's driven the progression of
Science in this country and, to some extent, in Northern
Europe for the last 100 years. And it's wonderful,
on the one hand, because it allows for that
image we have of a scientist tinkering away or the people
in their lab, and then the eurekas. And that hopefull
y translates
to better human health. But I think, in my opinion,
we've moved past that model as the most effective model
or the only model that should be explored. MARK ZUCKERBERG: Yeah, I
just think it's a balance. You want that. But you want to
empower those people. I think that that's these
tools empower those folks. ANDREW HUBERMAN: Sure. And there are mechanisms
to do that, like NIH. But it's hard to do
collaborative science. It's interesting that we're
sitting here not far-- because I grew
up right
near here as well. I'm not far from the garage
model of tech, right? The Hewlett-Packard model,
not far from here at all. And the idea was the tinkerer
in the garage, the inventor. And then people often
forget that to implement all the technologies
they discovered took enormous factories
and warehouses. So there's a similarity there
to Facebook, Meta, et cetera. But I think, in
science, we imagine that the scientists
alone in their laboratory and those eureka moments. But I think, nowa
days, the
big questions really require extensive collaboration and
certainly tool development. And one of the tools that
you keep coming back to is these LLMs, these
large language models. And maybe you could
just elaborate, for those that aren't familiar. What is a large language model? For the uninformed, what is it? And what does it allow us to
do that different, other types of AI don't allow? Or more importantly,
perhaps what does it allow us to do that a
bunch of really smart people, highly
informed in a
given area of science, staring at the data-- what can it do
that they can't do? MARK ZUCKERBERG: Sure. So I think a lot of the
progression of machine learning has been about building systems,
neural networks or otherwise, that can basically make sense
and find patterns in larger and larger amounts of data. And there was a breakthrough
a number of years back that some folks
at Google actually made called this transformer
model architecture. And it was this
huge breakthrough because
before then there
was somewhat of a cap where if you fed more
data into a Neural Network past some point,
it didn't really glean more insights from
it, whereas transformers just-- we haven't seen
the end of how big that can scale to yet. I mean, I think that
there's a chance that we run into some ceiling. ANDREW HUBERMAN: So
it never asymptotes? MARK ZUCKERBERG: We
haven't observed it yet. But we just haven't built
big enough systems yet. So I would guess that-- I don't know. I think that this
is actually one of the big questions
in the AI field today, is basically, are transformers
and are the current model architectures sufficient? If you just build larger
and larger clusters, do you eventually
get something that's like human intelligence
or super intelligence? Or is there some kind
of fundamental limit to this architecture that
we just haven't reached yet? And once we get a little bit
further in building them out, then we'll reach that. And then we'll need
a few more leaps before w
e get to the
level of AI that I think will unlock
a ton of really futuristic and amazing things. But there's no doubt
that even just being able to process
the amount of data that we can now with
this model architecture has unlocked a lot
of new use cases. And the reason why they're
called large language models is because one of the first uses
of them is people basically feed in all of the language
from, basically, the world wide web. And you can think about them as
basically prediction machines.
You put in a prompt. And it can basically
predict a version of what should come next. So you type in a headline
for a news story. And it can predict what it
thinks the story should be. Or you could train
it so that it could be a chat, bot
where, OK, if you're prompted with this question,
you, can get this response. But one of the
interesting things is it turns out that there's
actually nothing specific to using human language in it. So if instead of feeding
it human language, if you use that mo
del architecture
for a network and instead you feed it all of the
Human Cell Atlas Data, then if you prompt it
with a state of a cell, it can spit out
different versions of how that cell can
interact or different states that the cell could be
in next when it interacts with different things. ANDREW HUBERMAN: Does it have
to take a genetics class? So for instance, if you give
it a bunch of genetics data, do you have to say, hey,
by the way, and then you give it a genetics class so
it understands t
hat you've got DNA, RNA, mRNA, and proteins? MARK ZUCKERBERG: No, I think
that the basic nature of all these machine learning
techniques is they're basically pattern
recognition systems. So there are these very
deep statistical machines that are very efficient
at finding patterns. So it's not actually-- you don't need to teach
a language model that's trying to speak a language
a lot of specific things about that language either. You just feed it in
a bunch of examples. And then let's say you tea
ch
it about something in English, but then you also give
it a bunch of examples of people speaking Italian. It'll actually be able to
explain the thing that it learned in English in Italian. So the crossover and just
the pattern recognition is the thing that is pretty
profound and powerful about this. But it really does apply to
a lot of different things. Another example in the
scientific community has been the work
that AlphaFold, basically the folks at DeepMind,
have done on protein folding. I
t's just basically a lot of
the same model architecture. But instead of
language, there they fold they fed in all
of these protein data. And you can give it a state. And it can spit out solutions to
how those proteins get folded. So it's very powerful. I don't think we know
yet, as an industry, what the natural limits of it are. I think that that's one
of the things that's pretty exciting about
the current state. But it's certainly allows
you to solve problems that just weren't solved with
the g
eneration of machine learning that came before it. ANDREW HUBERMAN:
It sounds like CZI is moving a lot of work that was
just done in vitro, in dishes, and in vivo, in
living organisms, model organisms are humans,
to in silico, as we say. So do you foresee a future where
a lot of biomedical research, certainly the work of CZI
included, is done by machines? I mean, obviously,
it's much lower cost. And you can run millions
of experiments, which, of course, is not to
say that humans are not going to
be involved. But I love the idea that we
can run experiments in silico en masse. PRISCILLA CHAN: I think
in silico experiments are going to be incredibly helpful
to test things quickly, cheaply and just unleash
a lot of creativity. I do think you need to be
very careful about making sure it still translates
and matches the humans. One thing that's
funny in basic science is we've basically cured
every single disease in mice. We know what's going on when
they have a number of diseases because the
y're used
as a model organism. But they are not humans. And a lot of times,
that research is relevant, but not
directly one-to-one translatable to humans. So you just have to be really
careful about making sure that it actually
works for humans. ANDREW HUBERMAN: Sounds
like what CZI is doing is actually creating
a new field. As I'm hearing all of
this, I'm thinking, OK, this transcends immunology
department, cardiothoracic surgery, I mean neuroscience. I mean, the idea of a new field,
where you
certainly embrace the realities of
universities and laboratories because that's where most of
the work that you're funding is done. Is that right? MARK ZUCKERBERG: Mm-hmm. ANDREW HUBERMAN:
So maybe we need to think about what it means
to do science differently. And I think that's one of the
things that's most exciting. Along those lines, it seems
that bringing together a lot of different
types of people at different major
institutions is going to be especially important. So I know that the initi
al
CZI Biohub, gratefully, included Stanford. We'll put that
first in the list, but also UCSF, forgive me. I have many friends at
UCSF and also Berkeley. But there are now some
additional institutions involved. So maybe you could
talk about that, and what motivated the decision
to branch outside the Bay Area and why you selected those
particular additional institutions to be included. MARK ZUCKERBERG: Well,
I'll just say it. A big part of why we wanted
to create additional biohubs is we were jus
t so
impressed by the work that the folks who were
running the first biohub did. PRISCILLA CHAN: Yeah. And you should walk
through the work of the Chicago Biohub
and the New York Biohub that we just announced. But I think it's actually an
interesting set of examples that balance the
limits of what you want to do with physical
material engineering and where things are
purely biological because the Chicago team
is really building more sensors to be able to understand
what's going on in your body.
But that's more of a physical
kind of engineering challenge, whereas the New York
team-- we basically talk about this as like a
cellular endoscope of being able to have an immune
cell or something that can go and understand,
what's the thing that's going on in your body? But it's not a physical
piece of hardware. It's a cell that you can
basically have just go report out on different things that
are happening inside the body. ANDREW HUBERMAN: Oh, so making
the cell the the microscope. PRISCILLA
CHAN: Totally. MARK ZUCKERBERG: And
then eventually actually being able to act on it. But I mean, you should go
into more detail on all this. PRISCILLA CHAN: So
a core principle of how we think about biohubs
is that it has to be-- when we invited
proposals, it has to be at least
three institutions, so really breaking down the
barrier of a single university, oftentimes asking for the
people designing the research aim to come from all different
backgrounds and to explain why that the problem that
they want to solve requires interdisciplinary,
inter-university, institution collaboration to
actually make happen. We just put that
request for proposal out there with our
San Francisco Biohub as an example,
where they've done incredible work in single cell
biology and infectious disease. And we got-- I want to say--
like 57 proposals from over 150 institutions. A lot of ideas came together. And we were so, so
excited that we've been able to launch
Chicago and New York. Chicago is a collaborati
on
between UIUC, University of Illinois
Urbana-Champaign, and University of
Chicago and Northwestern. Obviously, these universities
are multifaceted. But if I were to describe
them by their stereotypical strength, Northwestern has
an incredible medical system and hospital system. University of Chicago
brings to the table incredible basic
science strengths. University of Illinois is
a computing powerhouse. And so they came
together and proposed that they were going
to start thinking about cells i
n tissue,
so one of the layers that you just alluded to. So how do the cells that we know
behave and act differently when they come together as a tissue? And one of the first tissues
that they're starting with is skin. So they've already been
able to, as a collaboration under the leadership, of
Shana Kelly design engineered skin tissue. The architecture looks the
same as what's in you and I. And what they've done is
built these super, super thin sensors. And they embed these sensors
throughout t
he layers of this engineered tissue. And they read out the data. They want to see what
these cells are secreting, how these cells
talk to each other and what happens when
these cells get inflamed. Inflammation is an
incredibly important process that drives 50% of all deaths. And so this is another
disease-agnostic approach. We want to understand
inflammation. And they're going to
get a ton of information out from these sensors that tell
you what happens when something goes awry because
right now
we can say, when you have an
allergic reaction, your skin gets red and puffy. But what is the
earliest signal of that? And these sensors can
look at the behaviors of these cells over time. And then you can apply
a large language model to look at the earliest
statistically significant changes that can allow you to
intervene as early as possible. So that's what Chicago's doing. They're starting
in the skin cells. They're also looking at the
neuromuscular junction, which is the connection between
where
a neuron attaches to a muscle and tells the muscle
how to behave-- super important in things
like ALS, but also in aging. The slowed transmission
of information across that
neuromuscular junction is what causes old
people to fall. Their brain cannot trigger their
muscles to react fast enough. And so we want to
be able to embed these sensors to understand how
these different, interconnected systems within our
bodies work together. In New York, they're doing a
related, but equally exciting p
roject where they're
engineering individual cells to be able to go in and identify
changes in a human body. So what they'll do is-- they're calling it-- ANDREW HUBERMAN: It's wild. I mean, I love that. I mean, this is-- I don't want to go on a tangent. But for those that want to
look it up adaptive optics, there's a lot of
distortion and interference when you try and look
at something really small or really far away. And really smart
physicists figured out, well, use the interference
as part of
the microscope. Make those actually
lenses of the microscope. MARK ZUCKERBERG: We
should talk about imaging separately after you talk
about the New York Biohub. ANDREW HUBERMAN: It's extremely
clever, along those lines. It's not intuitive. But then when you hear it, it's
like it makes so much sense. It's not immediately intuitive. Make the cells that already
can navigate to tissues or embed themselves in
tissues be the microscope within that tissue. I love it. PRISCILLA CHAN: Totally. The way th
at I explain
this to my friends and my family is this
is Fantastic Voyage, but real life. We are going into
the human body. And we're using the immune
cells, which are privileged and already working to
keep your body healthy, and being able to target them
to examine certain things. So you can engineer an immune
cell to go in your body and look inside your
coronary arteries and say, are these arteries healthy? Or are there plaques? Because plaques
lead to blockage, which lead to heart attacks. An
d the cell can then
record that information and report it back out. That's the first half
of what the New York Biohub is going to do. ANDREW HUBERMAN: Fantastic. PRISCILLA CHAN: The
second half is can you then engineer the cells to
go do something about it. Can I then tell
a different cell, immune cell that is able
to transport in your body to go in and clean that
up in a targeted way? And so it's incredibly exciting. They're going to
study things that are immune privilege, that
your immune syst
em normally doesn't have access to-- things like ovarian
and pancreatic cancer. They'll also look at a number
of neurodegenerative diseases, since the immune system doesn't
presently have a ton of access into the nervous system. But it's both mind blowing
and it feels like sci-fi. But science is
actually in a place where if you really push
a group of incredibly qualified scientists
say, could you do this if given the chance, the
answer is like probably. Give us enough time, the
bright team and r
esources. It's doable. MARK ZUCKERBERG: Yeah. I mean, it's a 10
to 15-year project. But it's awesome,
engineered cells, yeah. ANDREW HUBERMAN: I
love the optimism. And the moment you said make
the cell the microscope, so to speak, I was
like yes, yes and yes. It just makes so much sense. What motivated the decision
to do the work of CZI in the context of existing
universities as opposed to-- there's still some real
estate up in Redwood City where there's a bunch of
space to put biotech companies
and just hiring people
from all backgrounds and saying, hey, have at it and
doing this stuff from scratch? I mean, it's a very
interesting decision to do this in the
context of an existing framework of graduate students
that need to do their thesis and get a first author
paper because there's a whole set of structures
within academia that I think both
facilitate, but also limit the progression of science. That independent
investigator model that we talked about
a little bit earlier, it's so cor
e to the way
science has been done. This is very different
and frankly sounds far more efficient, if I'm
to be completely honest. And we'll see if I renew my
NIH funding after saying that. But I think we all
want the same thing. As scientists and
as humans, we want to understand the way we work. And we want healthy people
to persist to be healthy. And we want sick
people to get healthy. I mean, that's really
ultimately the goal. It's not super complicated. It's just hard to do. PRISCILLA CHAN: S
o the
teams at the biohub are actually independent
of the universities. ANDREW HUBERMAN: Got it. PRISCILLA CHAN: So each
biohub will probably have in total maybe 50 people
working on deep efforts. However, it's an acknowledgment
that not all of the best scientists who can
contribute to this area are actually going to, one,
want to leave a university or want to take on the
full-time scope of this project. So it's the ability to
partner with universities and to have the faculty
at all the universi
ties be able to contribute
to the overall project, is how the biohub is structured. ANDREW HUBERMAN: Got it. MARK ZUCKERBERG: But a lot of
the way that we're approaching CZI is this long-term,
iterative project to figure out-- try a
bunch of different things, figure out which things produce
the most interesting results, and then double down on those
in the next five-year push. So we just went
through this period where we wrapped
up the first five years of the science program. And we tried a lot
of different models, all kinds of different things. And it's not that
the biohub model-- we don't think it's
the best or only model. But we found that it was
a really interesting way to unlock a bunch
of collaboration and bring some
technical resources that allow for this longer
term development. And it's not something that
is widely being pursued across the rest of the field. So we figured, OK, this
is an interesting thing that we can help push on. But I mean, yeah, we do
believe in the collabo
ration. But I also think that
we come at this with-- we don't think that the way
that we're pursuing this is the only way to
do this or the way that everyone should do it. We're pretty aware of what
is the rest of the ecosystem and how we can play
a unique role in it. ANDREW HUBERMAN: It
feels very synergistic with the way science
is already done and also fills an incredibly
important niche that, frankly, wasn't filled before. Along the lines of
implementation-- so let's say your large language
models combined with imaging tools reveal that a
particular set of genes acting in a cluster-- I don't know-- set
up an organ crash. Let's say the pancreas
crashes at a particular stage of pancreatic cancer. I mean, it's still one of the
most deadliest of the cancers. And there are others that you
certainly wouldn't want to get. But that's among the ones you
wouldn't want to get the most. So you discover that. And then and the
idea is that, OK, then AI reveals
some potential drug targets that th
en bear
out in vitro, in a dish and in a mouse model. How is the actual implementation
to drug discovery? Or maybe this target is
druggable, maybe it's not. Maybe it requires
some other approach-- laser ablation
approach or something. We don't know. But ultimately,
is CZI going to be involved in the implementation
of new therapeutics? Is that the idea? MARK ZUCKERBERG: Less so. PRISCILLA CHAN: Less so. This is where it's important
to work in an ecosystem and to know your
own limitations. There a
re groups, and
startups and companies that take that and bring it to
translation very effectively. I would say the
place where we have a small window into
that world is actually our work with rare
disease groups. We have, through our
Rare As One portfolio, funded patient advocates
to create rare disease organizations where patients
come together and actually pool their collective experience. They build
bioregistries, registries of their natural history. And they both partner
with researchers to
do the research
about their disease and with drug developers to
incentivize drug developers to focus on what they may
need for their disease. And one thing that's
important to point out is that rare
diseases aren't rare. There are over
7,000 rare diseases and collectively impact
many, many individuals. And I think the thing
that's, from a basic science perspective, the incredibly
fascinating thing about rare diseases is that
they're actually windows to how the body normally should work. And so t
here are often
mutations that when genes that when they're mutated
cause very specific diseases, but that tell you how the
normal biology works as well. ANDREW HUBERMAN: Got it. So you discussed basically the
major goals and initiatives of the CZI for the next,
say, 5 to 10 years. And then beyond
that, the targets will be explored by
biotech companies. They'll grab those targets, and
test them and implement them. MARK ZUCKERBERG:
There's also, I think, been a couple of teams from
the initial bio
hub that were interested in spinning
out ideas into startups. So even though it's
not a thing that we're going to pursue because
we're a philanthropy, we want to enable
the work that gets done to be able to get turned
into companies and things that other people
go take and run towards building
ultimately therapeutics. So that's another zone. But that's not a thing
that we're going to do. ANDREW HUBERMAN: Got it. I gather you're both optimists. Yeah? Is that part of what
brought you together? For
give me for switching
to a personal question. But I love the
optimism that seems to sit at the root of the CZI. PRISCILLA CHAN: I
will say that we are incredibly hopeful people. But it manifests in different
ways between the two of us. MARK ZUCKERBERG: Yeah. PRISCILLA CHAN: How
would you describe your optimism versus mine? It's not a loaded question. MARK ZUCKERBERG: I don't know. Huh. I mean, I think I'm more
probably technologically optimistic about
what can be built. And I think you, because
of
your focus as an actual doctor, have more of a
sense of how that's going to affect actual
people in their lives, whereas, for me, it's like-- I mean, a lot of my
work is we touch a lot of people around the world. And the scale is immense. And I think, for
you, it's like being able to improve the
lives of individuals, whether it's students at any of
the schools that you've started or any of the stuff that we've
supported through the education work, which isn't the
goal here, or just being able
to improve people's
lives in that way I think is the thing that I've seen
be super passionate about. I don't know. Do you agree with
that characterization? I'm trying I'm trying to-- PRISCILLA CHAN: Yeah,
I agree with that. I think that's very fair. And I'm sort of
giggling to myself because in day-to-day
life, as life partners, our relative optimism
comes through as Mark just is overly
optimistic about his time management and will get
engrossed in interesting ideas. MARK ZUCKERBERG: I'm late.
PRISCILLA CHAN: And he's late. ANDREW HUBERMAN: Physicians
are very punctual, yeah. PRISCILLA CHAN: And
because he's late, I have to channel Mark
is an optimist whenever I'm waiting for him. MARK ZUCKERBERG: That's
such a nice way of-- OK, I'll start using that. PRISCILLA CHAN:
That's what I think when I'm in the driveway with
the kids waiting for you. I'm like, Mark is an optimist. And so his optimism
translates to some tardiness, whereas I'm a how is this
going to happen like. I'm going to ope
n a spreadsheet. I'm going to start
putting together a plan and pulling together
all the pieces, calling people to bring
something to life. MARK ZUCKERBERG: But it is one
of my favorite quotes, that is optimists tend
to be successful and pessimists tend to be right. And yeah, I mean, I
think it's true in a lot of different aspects of life. ANDREW HUBERMAN: Who said that? Did you say that,
Mark Zuckerberg? MARK ZUCKERBERG: No, I did not. PRISCILLA CHAN: Absolutely not. MARK ZUCKERBERG: No, no, no
. I like it. I did not invent it. ANDREW HUBERMAN:
We'll give it to you. We'll put it out there. MARK ZUCKERBERG: No, no, no. ANDREW HUBERMAN: Just
kidding, just kidding. MARK ZUCKERBERG: But I do
think that there's really something to it, right? I mean, if you're
discussing any idea, there's all these reasons
why it might not work. And those reasons
are probably true. The people who are stating them
probably have some validity to it. But the question is, is that
the most productive way to view
the world? Across the board,
I think the people who tend to be the
most productive and get the most done-- you kind of need
to be optimistic because if you don't believe
that something can get done, then why would
you go work on it? ANDREW HUBERMAN:
The reason I ask the question is that these days
we hear a lot about the future is looking so dark in
these various ways. And you have children. So you have families. And you are a family, excuse me. And you also have
families independently that are
now merged. But I love the
optimism behind the CZI because, behind
all this, there's a set of big
statements on the wall. One, the future can be
better than the present, in terms of treating disease,
maybe even, you said, eliminating diseases,
all diseases. I love that optimism. And there's a tractable
path to do it. We're going to put literally
money, and time, and energy, and people, and technology
and AI behind that. And so I have to ask,
was having children a significant modifier in terms
of
your view of the future? Like wow, you hear all
this doom and gloom. What's the future going
to be like for them? Did you sit back and
think, what would it look like if there was a
future with no diseases? Is that the future, we
want our children in? I mean, I'm voting a big yes. So we're not we're not
going to debate that at all. But was having
children an inspiration for the CZI in some way? MARK ZUCKERBERG: Yeah. So PRISCILLA CHAN: I think
my answer to that-- I would dial backwards for me. A
nd I'll just tell a very
brief story about my family. I'm the daughter of
Chinese-Vietnamese refugees. My parents and grandparents
were boat people, if you remember
people left Vietnam during the war in these small
boats into the South China Sea. And there were stories about
how these boats would sink with whole families on them. And so my
grandparents, both sets of grandparents who
knew each other, decided that there was a
better future out there. And they were willing
to take risks for it. But
they were afraid of
losing all of their kids. My dad is one of six. My mom is one of 10. And so they decided
that there was something out there in this bleak time. And they paired up their
kids, one from each family, and sent them out on
these little boats before the internet, before
cell phones, and just said, we'll see you on the other side. ANDREW HUBERMAN: Wow. PRISCILLA CHAN:
And the kids were between the ages of like
10 to 25, so young kids. My mom was a teenager, early
teen when this hap
pened. And everyone made it. And I get to sit
here and talk to you. So how could I not believe
that better is possible? And like I hope that that's
in my epigenetics somewhere and that I carry on. ANDREW HUBERMAN: That
is a spectacular story. PRISCILLA CHAN: Isn't that wild? ANDREW HUBERMAN:
It is spectacular. PRISCILLA CHAN: How can I
be a pessimist with that? ANDREW HUBERMAN: I love it. And I so appreciate that
you became a physician because you're now
bringing that optimism, and that epigenet
ic
understanding, and cognitive understanding
and emotional understanding to the field of medicine. So I'm grateful to the people
that made that decision. PRISCILLA CHAN: Yeah. I've always known that story. But you don't understand
how wild that feels until you have your own child. And you're like,
well, I can't even-- I refuse to let her use glass
bottles only or something like that. And you're like, oh my God,
the risk and the willingness of my grandparents to believe
in something bigger and b
etter is just astounding. And our own children give
it a sense of urgency. ANDREW HUBERMAN: Again,
a spectacular story. And you're sending knowledge
out into the fields of science and bringing knowledge
into the fields of science. And I love this. We'll see you on the other side. I'm confident that it
will all come back. Well, thank you
so much for that. Mark, you have the
opportunity to talk about-- did having kids
change your worldview? MARK ZUCKERBERG: It's really
tough to beat that story. AN
DREW HUBERMAN: It is
tough to beat that story. And they are also your children. So in this case, you get two for
the price of one, so to speak. MARK ZUCKERBERG: Having
children definitely changes your time horizon. So I think that
that's one thing. There are all these things that
I think we had talked about, for as long as we've known
each other, that you eventually want to go do. But then it's like,
oh, we're having kids. We need to get on this, right? So I think that there's-- PRISCILLA CHAN:
That was actually one of the checklists, the baby
checklist before the first. MARK ZUCKERBERG: It was
like, the baby's coming. We have to start CZI. PRISCILLA CHAN: Truly. MARK ZUCKERBERG: I'm like
sitting in the hospital delivery room finishing
editing the letter that we were going to publish
to announce the work. PRISCILLA CHAN: Some people
think that is an exaggeration. It was not. We really were editing
the final draft. ANDREW HUBERMAN:
Birthed CZI before you birthed the human child. Well, i
t's an
incredible Initiative. I've been following it
since its inception. And it's already been
tremendously successful. And everyone in the
field of science-- and I have a lot of
communication with those folks-- feels the same way. And the future is even
brighter for it, it's clear. And thank you for expanding
to the Midwest and New York. And we're all very excited to
see where all of this goes. I share in your optimism. And thank you for
your time today. PRISCILLA CHAN: Yeah, thank you. MARK Z
UCKERBERG: Thank you. A lot more to do. ANDREW HUBERMAN: I'd like
to take a quick break and thank our sponsor,
InsideTracker. InsideTracker is a
personalized nutrition platform that analyzes data
from your blood and DNA to help you better
understand your body and help you reach
your health goals. I've long been a believer in
getting regular blood work done for the simple reason that
many of the factors that impact your immediate and
long-term health can only be analyzed from
a quality blood test
. Now, a major problem
with a lot of blood tests out there, however,
is that you get information back about metabolic
factors, lipids, and hormones and so forth. But you don't know what to
do with that information. With InsideTracker,
they make it very easy because they have a
personalized platform that allows you to see the
levels of all those things-- metabolic factors, lipids,
hormones, et cetera. But it gives you
specific directives that you can follow that
relate to nutrition, behavioral mo
dification,
supplements, et cetera that
can help you bring those numbers into the ranges
that are optimal for you. If you'd like to
try InsideTracker, you can go to
insidetracker.com/huberman to get 20% off any of
InsideTracker's plans. Again, that's
insidetracker.com/huberman. And now for my discussion
with Mark Zuckerberg. Slight shift of topic here-- you're extremely
well-known for your role in technology development. But by virtue of your
personal interests and also where Meta
technology int
erfaces with mental health
and physical health, you're starting to become
synonymous with health, whether you realize it or not. Part of that is because
there's posts, footage of you rolling jiu jitsu. You won a jiu jitsu
competition recently. You're doing other forms of
martial arts, water sports, including surfing,
and on and on. So you're doing it yourself. But maybe we could just
start off with technology and get this issue out
of the way first, which is that I think many people
assume that
technology, especially technology that
involves a screen, excuse me, of any kind is going to
be detrimental to our health. But that doesn't necessarily
have to be the case. So could you explain
how you see technology meshing with, inhibiting,
or maybe even promoting physical and mental health? MARK ZUCKERBERG: Sure. I mean, I think this is
a really important topic. The research that we've
done suggests that it's not all good or all bad. I think how you're
using the technology has a big impact on
whether it is basically a positive experience for you. And even within technology,
even within social media, there's not one type of
thing that people do. I think, at its best, you're
forming meaningful connections with other people. And there's a lot of research
that basically suggests that it's the
relationships that we have and the friendships that bring
the most happiness in our lives and, at some level,
end up even correlating with living a longer
and healthier life because that grounding
that you have in community ends up being
important for that. So I think that aspect
of social media, which is the ability to connect
with people, to understand what's going on
in people's lives, have empathy for them,
communicate what's going on with your life,
express that, that's generally positive. There are ways that
it can be negative, in terms of bad interactions,
things like bullying, which we can talk about because
there's a lot that we've done to basically make sure that
people can be s
afe from that and give people tools and
give kids the ability to have the right parental controls. Their parents can oversee that. But that's the interacting
with people side. There's another
side of all of this, which I think of as just
passive consumption, which, at its best, is entertainment. And entertainment is an
important human thing, too. But I don't think that that
has quite the same association with the long-term well-being
and health benefits as being able to help people
connect with
other people does. And I think, at its worst, some
of the stuff we see online-- I think, these days,
a lot of the news is just so relentlessly
negative that it's just hard to come away
from an experience where looking at the
news for half an hour and feel better about the world. So I think that
there's a mix on this. I think the more
that social media is about connecting
with people and the more that when you're consuming
and using the media part of social media to
learn about things that enrich
you and can provide
inspiration or education as opposed to things that
just leave you with a more toxic feeling, that's the
balance that we try to get right across our products. And I think we're pretty
aligned with the community because, at the end of
the day, I mean, people don't want to use a product
and come away feeling bad. There's a lot that
people talk about-- evaluate a lot of
these products in terms of information and utility. But I think it's
as important, when you're designing a
pro
duct, to think about what kind of
feeling you're creating with the people who
use it, whether that's an aesthetic sense when
you're designing hardware, or just what do you
make people feel. And generally, people don't
want to feel bad, right? That doesn't mean that
we want to shelter people from bad things that are
happening in the world. But I don't really think that-- it's not what people
want for us to just be just showing all this super
negative stuff all day long. So we work hard on all the
se
different problems-- making sure that we're helping connect
people as best as possible, helping make sure that
we give people good tools to block people who
might be bullying them, or harass them, or
especially for younger folks, anyone under the age of 16
defaults into an experience where their
experience is private. We have all these
parental tools. So that way, parents can
understand what their children are up to in a good balance. And then on the
other side, we try to give people tools
to
understand how they're spending their time. We try to give people tools
so that if you're a teen and you're stuck in some
loop of just looking at one type of content,
we'll nudge you and say, hey, you've been looking at content
of this type for a while. How about something else? And here's a bunch
of other examples. So I think that there
are things that you can do to push this in
a positive direction. But I think it just
starts with having a more nuanced view of this
isn't all good or all bad.
And the more that you
can make it a positive thing, the better this
will be for all the people who use our products. ANDREW HUBERMAN: That
makes really good sense. In terms of the negative
experience, I agree. I don't think anyone wants
a negative experience in the moment. I think where some people
get concerned perhaps-- and I think about my own
interactions with, say, Instagram, which I use all the
time for getting information out, but also
consuming information. And I happen to love it. It's
where I
essentially launched the non-podcast segment of
my podcast and continue to. I can think of experiences
that are a little bit like highly
processed food, where it tastes good at the time. It's highly engrossing. But it it's not
necessarily nutritious. And you don't feel
very good afterwards. So for me, that would
be the little collage of default options to
click on in Instagram. Occasionally, I
notice-- and this just reflects my failure, not
Instagram's, that there are a lot of street
fig
ht things, like people beating
people up on the street. And I have to say, these have a
very strong gravitational pull. I'm not somebody that enjoys
seeing violence, per se. But you know I find myself-- I'll click on one of
these, like what happened? And I'll see someone get hit. And there's a little melee
on the street or something. And those seem to be
offered to me a lot lately. And again, this is
my fault. It reflects my prior searching experience. But I noticed that it has a bit
of a gravit
ational pull, where I didn't learn anything. It's not teaching me any
useful street self-defense skills of any kind. And at the same time,
I also really enjoy some of the cute animal stuff. And so I get a
lot of those also. So there's this
polarized collage that's offered to me that
reflects my prior search behavior. You could argue that the
cute animal stuff is just entertainment. But actually, it fills
me with a feeling, in some cases, that
truly delights me. I delight in animals. And we're no
t just
talking about kittens. I mean, animals I've
never seen before, interactions between
animals I've never seen before that truly delight me. They energize me
in a positive way that when I leave Instagram,
I do think I'm better off. So I'm grateful for the
algorithm in that sense. But I guess, the direct question
is, is the algorithm just reflective of what one
has been looking at a lot prior to that moment
where they log on? Or is it also trying to do
exactly what you described, which is try
ing to give people
a good-feeling experience that leads to more good feelings? MARK ZUCKERBERG: Yeah. I mean, I think we try to
do this in a long-term way. I think one simple
example of this is we had this issue
a number of years back about clickbait
news, so articles that would have basically
a headline that grabbed your attention,
that made you feel like, oh, I need
to click on this. And then you click on it. And then the article is
actually about something that's somewhat tangential to it. Bu
t people clicked on it. So the naive version of this
stuff, the 10-year-old version was like, oh, people seem
to be clicking on this. Maybe that's good. But it's actually a pretty
straightforward exercise to instrument the system to
realize that, hey, people click on this, and
then they don't really spend a lot of time reading
the news after clicking on it. And after they do
this a few times, it doesn't really correlate
with them saying that they're having a good experience. Some of how we
measu
re this is just by looking at how
people use the services. But I think it's also
important to balance that by having real people
come in and tell us, OK-- we show them, here are
the stories that we could have showed you, which of these
are most meaningful to you, or would make it so that you
have the best experience, and just mapping the
algorithm and what we do to that ground truth of
what people say that they want. So I think that, through
a set of things like that, we really have made large
s
teps to minimize things like clickbait over time. It's not like gone
from the internet. But I think we've done a
good job of minimizing it on our services. Within that though,
I do think that we need to be pretty
careful about not being paternalistic about what
makes different people feel good. So I mean, I don't
know that everyone feels good about cute animals. I mean, I can't
imagine that people would feel really bad about it. But maybe they don't have as
profound of a positive reaction to it
as you just expressed. And I don't know. Maybe people who are
more into fighting would look at the
street fighting videos-- assuming that they're within
our community standards. I think that there's
a level of violence that we just don't want
to be showing at all. But that's a separate question. But if they are, I
mean, then it's like-- I mean, I'm pretty into MMA. I don't get a lot of
street fighting videos. But if I did, maybe I'd feel
like I was learning something from that. I think at variou
s times
in the company's history, we've been a little bit too
paternalistic about saying, this is good content, this
is bad, you should like this, this is unhealthy for you. And I think that we want to
look at the long-term effects. You don't want to get
stuck in a short term loop of like, OK,
just because you did this today doesn't
mean it's what you aspire for yourself over time. But I think, as long as you
look at the long-term of what people both say they want and
what they do, giving people
a fair amount of latitude to
like the things that they like, I just think feels like
the right set of values to bring to this. Now, of course, that
doesn't go for everything. There are things that are truly
off limits and things that-- like bullying, for example, or
things that are really inciting violence, things like that. I mean, we have the
whole community standards around this. But I think, except
for those things which I would hope that
most people can agree, OK, bullying is bad-- I hope
that 100% of
people agree with that. And not 100%, maybe 99%. Except for the things that
kind of get that very-- that feel pretty extreme
and bad like that, I think you want to
give people space to like what they want to like. ANDREW HUBERMAN: Yesterday, I
had the very good experience of learning from the Meta team
about safety protections that are in place for kids who
are using Meta Platforms. And frankly, I was really
positively surprised at the huge number of
filter-based tools and just abil
ity to customize the
experience so that it can stand the best chance of enriching--
not just remaining neutral, but enriching their
mental health status. One thing that came about
in that conversation, however, was I realized
there are all these tools. But do people really know
that these tools exist? And I think about my own
experience with Instagram. I love watching Adam Mosseri's
Friday Q&As because he explains a lot of the tools that
I didn't know existed. And if people haven't
seen that, I
highly recommend they watch that. I think he takes
questions on Thursdays and answers them
most every Fridays. So if I'm not aware of the tools
without watching that, that exists for adults,
how does Meta look at the challenge of making sure
that people know that there are all these tools-- I mean, dozens and dozens
of very useful tools? But I think most of us just
know the hashtag, the tag, the click, stories versus feed. We now know that-- I also post to Threads. I mean, so we know the
major c
hannels and tools. But this is like
owning a vehicle that has incredible features
that one doesn't realize can take you off road,
can allow your vehicle to fly. I mean, there's a lot there. So what do you
think could be done to get that information out? Maybe this conversation could
cue people to [INAUDIBLE].. MARK ZUCKERBERG: I mean, that's
part of the reason why I wanted to talk to you about this. I mean, I think most of the
narrative around social media is not, OK, all of
the different tools
that people have to
control their experience. It's the narrative of
is this just negative for teens or something. And I think, again,
a lot of this comes down to how is the
experience being tuned. Are people using it to
connect in positive ways? And if so, I think
it's really positive. So yeah, I mean, I
think part of this is we probably just need to
get out and talk to people more about it. And then there's an
in-product aspect, which is if you're a
teen and you sign up, we take you through a p
retty
extensive experience that tries to outline some of this. But that has limits, too,
because when you sign up for a new thing, if you're
bombarded with here's a list of features, you're like,
OK, I just signed up for this. I don't really understand much
about what the service is. Let me go find some
people to follow who are my friends on
here before I learn about controls to prevent people
from harassing me or something. That's why I think it's
really important to also show a bunch of these
tools in context. So if you're
looking at comments, and if you go to
delete a comment, or you go to edit something, try
to give people prompts in line. It's like, hey, did that
you can manage things in these ways around that? Or when you're in the inbox
and you're filtering something, remind people in line. So just because of
the number of people who use the products
and the level of nuance around each of the controls,
I think the vast majority of that education, I think,
needs to happen in the
product. But I do think that through
conversations like this and others that we
need to be doing, I think we can create a broader
awareness that those things exist so that way at
least people are primed so that way when those things
pop up in the product people, they're like, oh yeah, I knew
that there was this control. And here's how I would use that. ANDREW HUBERMAN: I find
the restrict function to be very useful, more than the
block function in most cases. I do sometimes have
to block people.
But the restrict
function is really useful that you could filter
specific comments. You might recognize that
someone has a tendency to be a little aggressive. And I should point out that
I actually don't really mind what people say to me. But I try and maintain
what I call classroom rules in my comment section, where
I don't like people attacking other people because I
would never tolerate that in the university classroom. I'm not going to tolerate
that in the comments section, for instance. MA
RK ZUCKERBERG: Yeah. And I think that the example
that you just used about restrict versus block gets to
something about product design that's important, too, which
is that block is this very powerful tool that if someone
is giving you a hard time and you just want them to
disappear from the experience, you can do it. But the design trade-off with
that is that in order to make it so that the person is
just gone from the experience and that you don't
show up to them, they don't show up to you-- i
nherent to that is
that they will have a sense that you blocked them. And that's why I think some
stuff like restrict or just filtering, like
I just don't want to see as much stuff
about this topic-- people like using different
tools for very subtle reasons. I mean, maybe you want the
content to not show up, but you don't want
the person who's posting the content to know that
you don't want it to show up. Maybe you don't want to get the
messages in your main inbox, but you don't want to tell the
person actually that you're not friends or something like that. You actually need to give
people different tools that have different levels
of power and nuance around how the social
dynamics around using them play out in order
to really allow people to tailor the experience
in the ways that they want. ANDREW HUBERMAN:
In terms of trying to limit total amount
of time on social media, I couldn't find really
good data on this. How much time is too much? I mean, I think
it's going to depend on what
one is looking at, the
age of the user, et cetera. MARK ZUCKERBERG: I agree. ANDREW HUBERMAN: I
know that you have tools that cue the
user to how long they've been on
a given platform. Are there tools
to self-regulate-- I'm thinking about the Greek
myth of the sirens and people tying themselves to the
mast and covering their eyes so that they're not
drawn in by the sirens. Is there a function aside from
deleting the app temporarily and then reinstalling it every
time you want to use it again? I
s there a true lockout,
self-lockout function where one can lock themselves
out of access to the app? MARK ZUCKERBERG: Well, I
think we give people tools that let them manage this. And there's the tools
that you get to use. And then there's the
tools that the parents get to use to basically
see how usage works. But yeah, I think that
there's different-- I think, for now,
we've mostly focused on helping people
understand this, and then give people reminders
and things like that. It's tough, thoug
h, to
answer the question that you were talking about before. Is there an amount of
time which is too much? Because it does really
get to what you're doing. If you fast forward
beyond just the apps that we have today
to an experience that is like a social
experience in the future of the augmented reality
glasses or something that we're building,
a lot of this is going to be you're
interacting with people in the way that you
would physically as if you were like
hanging out with friends or working
with people. But now, they can
show up as holograms. And you can feel like you're
present right there with them, no matter where
they actually are. And the question is,
is there too much time to spend interacting
with people like that? Well, at the limit,
if we can get that experience to be
as rich and giving you as good of a sense of presence
as you would have if you were physically there
with someone, then I don't see why you would want to
restrict the amount that people use that technology
t
o any less than what would be the amount of time
that you'd be comfortable interacting with
people physically, which obviously is not
going to be 24 hours a day. You have to do other stuff. You have work. You need to sleep. But I think it really gets to
how you're using these things, whereas if what you're
primarily using the services for is you're getting stuck in loops
reading news or something that is really getting you into
a negative mental state, then I don't know. I mean, I think that
the
re's probably a relatively short
period of time that maybe that's a good thing
that you want to be doing. But again, even
then it's not zero because just because news
might make you unhappy doesn't mean that
the answer is to be unaware of negative things that
are happening in the world. I just think that
different people have different tolerances for
what they can take on that. And I think it's
generally having some awareness is probably
good, as long as it's not more than you're constitutionall
y
able to take. So I don't know. I try not be too paternalistic
about this as our approach. But we want to empower
people by giving them the tools, both people and,
if you're a teen, your parents to have tools to understand
what you're experiencing and how you're using these
things, and then go from there. ANDREW HUBERMAN: Yeah. I think it requires of all of us
some degree of self-regulation. I like this idea of not
being too paternalistic. I mean, it seems like
the right way to go. I find mysel
f
occasionally having to make sure that I'm not
just passively scrolling, that I'm learning. I like foraging for, organizing
and dispersing information. That's been my life's career. So I've learned so
much from social media. I find great
papers, great ideas. I think comments are a
great source of feedback. And I'm not just saying that
because you're sitting here. I mean, Instagram in particular,
but other Meta platforms have been tremendously
helpful for me to get science and health information
out. One of the things that
I'm really excited about, which I only had the chance to
try for the first time today, is your new VR platform,
the newest Oculus. And then we can talk about
the glasses, the Ray-Bans. MARK ZUCKERBERG: Sure. ANDREW HUBERMAN:
Those two experiences are still kind of blowing
my mind, especially the Ray-Ban glasses. And I have so many
questions about this. So I'll resist. But-- MARK ZUCKERBERG: We
can get into that. ANDREW HUBERMAN: OK. Well, yeah, I have some
experience
with VR. My Lab has used VR. Jeremy Bailenson's
Lab at Stanford is one of the pioneering
labs of VR and mixed reality. I guess they used to call it
augmented reality, but now mixed reality. I think what's so
striking about the VR that you guys had me try today
is how well it interfaces with the real room, let's
call it, the physical room. MARK ZUCKERBERG: Physical. ANDREW HUBERMAN: I
could still see people. I could see where
the furniture was. So I wasn't going to
bump into anything. I could se
e people's smiles. I could see my
water on the table while I was doing this what
felt like a real martial arts experience, except I
wasn't getting hit. Well, I was getting
hit virtually. But it's extremely engaging. And yet, on the
good side of things, it really bypasses a lot
of the early concerns that Bailenson Lab-- again, Jeremy's Lab-- was
early to say that, oh, there's a limit to how much VR one
can or should use each day, even for the adult brain
because it can really disrupt your vestibu
lar
system, your sense of balance. All of that seems
to have been dealt with in this new
iteration of VR. I didn't come out of it
feeling dizzy at all. I didn't feel like I was
reentering the room in a way that was really jarring. Going into it is
obviously, Whoa, this is a different world. But you can look to your left
and say, oh, someone just came in the door. Hey, how's it going? Hold on, I'm playing
this game, just as it was when I was a
kid playing in Nintendo and someone would walk in. It
's fully engrossing. But you'd be like, hold on. And you see they're there. So first of all,
bravo, incredible. And then the next question
is, what do we even call this experience? Because it is
truly really mixed. It's a truly mixed
reality experience. MARK ZUCKERBERG: Yeah. I mean, mixed reality
is the umbrella term that refers to the
combined experience of virtual and
augmented reality. So augmented reality is
what you're eventually going to get with some future
version of the smart glasses,
where you're primarily
seeing the world, but you can put holograms in it. So we'll have a
future where you're going to walk into a room. And there are going to
be as many holograms as physical objects. If you just think about all the
paper, the art, physical games, media, your workstation-- ANDREW HUBERMAN: If
we refer to, let's say, an MMA fight, we could just
draw it up on the table right here and just see it repeat
as opposed to us turning and looking at a screen. MARK ZUCKERBERG: Yeah. I mea
n, pretty much
any screen that exists could be a hologram in the
future with smart glasses. There's nothing that
actually physically needs to be there for that
when you have glasses that can put a hologram there. And it's an interesting
thought experiment to just go around and think
about, OK, what of the things that are physical in the world
need to actually be physical. Your chair does, right? Because you're sitting on it. A hologram isn't
going to support you. But like that art
on the wall, I
mean, that doesn't need to
physically be there. So I think that that's the
augmented reality experience that we're moving towards. And then we've had these
headsets that historically we think about as VR. And that has been something
that is like a fully immersive experience. But now, we're getting
something that's a hybrid in between
the two and capable of both, which is a headset
that can do both virtual reality and some of these augmented
reality experiences. And I think that
that's really po
werful, both because you're going to
get new applications that allow people to collaborate together. And maybe the two of
us are here physically, but someone joins us and
it's their avatar there. Or maybe it's some
version in the future. You're having a team meeting. And you have some
people there physically. And you have some
people dialing in. And they're basically like
a hologram, there virtually. But then you also
have some AI personas that are on your
team that are helping you do different
things. And they can be embodied as
avatars and around the table meeting with you. ANDREW HUBERMAN:
Are people are going to be doing first dates that
are physically separated? I could imagine that
some people would-- is it even worth leaving
the house type date? And then they find out. And then they meet
for the first time. MARK ZUCKERBERG: I mean, maybe. I think dating has physical
aspects to it, too. ANDREW HUBERMAN: Right. Some people might
not be-- they want to know whether
or not it's worth
the effort to head out or not. They want to bridge
the divide, right? MARK ZUCKERBERG: It is possible. I mean, I know
some of my friends who are dating basically
say that in order to make sure that they have
a safe experience, if they're going on a first
date, they'll schedule something that's shorter and
maybe in the middle of the day. So maybe it's coffee. So that way, if they
don't like the person, they can just get out
before going and scheduling a dinner or a real, full date. So I don't kn
ow. Maybe in the future,
people will have that experience where
you can feel like you're kind of sitting there. And it's and it's even easier,
and lighter weight and safer. And if you're not having
a good experience, you can just teleport
out of there and be gone. But yeah, I think that this
will be an interesting question in the future. There are clearly a lot of
things that are only possible physically that-- or are so much
better physically. And then there are
all these things that we're buil
ding up that
can be digital experiences. But it's this weird
artifact of how this stuff has been developed
that the digital world and the physical world
exist in these completely different planes. When you want to interact
with the digital world-- we do it all the time. But we pull out a small screen. Or we have a big screen. And just basically,
we're interacting with the digital world
through these screens. But I think if we
fast forward a decade or more, I think one of the
really interesting q
uestions about what is the
world that we're going to live in, I think
it's going to increasingly be this mesh of the
physical and digital worlds that will allow us to feel, A,
that the world that we're in is just a lot richer
because there can be all these things that people create
that are just so much easier to do digitally than physically. But B, you're going to have a
real physical sense of presence with these things and
not feel like interacting in the digital world
is taking you away from
the physical world,
which today is just so much viscerally
richer and more powerful. I think the digital world
will be embedded in that and will feel just as
vivid in a lot of ways. So that's why I
always think-- when you were saying before, you
felt like you could look around and see the real room. I actually think there's
an interesting kind of philosophical distinction
between the real room and the physical room,
which historically I think people would have said
those are the same thing. But
I actually
think, in the future, the real room is going
to be the combination of the physical world with
all the digital artifacts and objects that are in there
that you can interact with them and feel present, whereas the
physical world is just the part that's physically there. And I think it's possible
to build a real world that's the sum of these two
that will actually be more profound experience
than what we have today. ANDREW HUBERMAN:
Well, I was struck by the smoothness of the
interface b
etween the VR and the physical room. Your team had me try a-- I guess it was an exercise
class in the [INAUDIBLE].. But it was essentially
like hitting mitts boxing, so hitting targets boxing. MARK ZUCKERBERG:
Yeah, super natural. ANDREW HUBERMAN: Yeah, and it
comes at a fairly fast pace that then picks up. It's got some tutorial. It's very easy to use. And it certainly got
my heart rate up. And I'm in at
least decent shape. And I have to be
honest, I've never once desired to do any of
these on-
screen fitness things. I mean, I can't think of
anything more aversive than a-- I don't want to insult
any particular products, but riding a stationary
bike while looking at a screen pretending
I'm on a road outside. I can't think of
anything worse for me. MARK ZUCKERBERG: I do
like the leaderboard. Maybe I'm just a very
competitive person. If you're going to be
running on a treadmill, at least give me a
leaderboard so I can beat the people who are ahead of me. ANDREW HUBERMAN: I like
moving out
side and certainly an exercise class
or aerobics class, as they used to call them. But the experience I tried
today was extremely engaging. And I've done enough
boxing to at least know how to do a little bit of it. And I really enjoyed it. It gets your heart rate up. And I completely
forgot that I was doing an on-screen experience
in part because, I believe, I was still in
that physical room. And I think there's
something about the mesh of the physical room and
the virtual experience that makes
it neither of
one world or the other. I mean, I really felt at
the interface of those. And I certainly got
presence, this feeling of forgetting that I was
in a virtual experience and got my heart rate
up pretty quickly. We had to stop because we
were going to start recording. But I would do that for a good
45 minutes in the morning. And there's no amount of
money you could pay me truly to look at a screen
while pedaling on a bike or running on a treadmill. So again, bravo, I think
it's going to
be very useful. It's going to get people
moving their bodies more, which certainly-- social media, up until now,
and a lot of technologies have been accused of limiting
the amount of physical activity that both children and
adults are engaged in. And we know we need
physical activity. You're a big proponent
of and practitioner of physical activity. So is this a major goal
of Meta, to get people moving their bodies more
and getting their heart rates up and so on? MARK ZUCKERBERG: I think
we want
to enable it. And I think it's good. But I think it comes more from a
philosophical view of the world than it is necessarily-- I mean, I don't go
into building products to try to shape
people's behavior. I believe in empowering
people to do what they want and be the best version of
themselves that they can be. ANDREW HUBERMAN: So no agenda? MARK ZUCKERBERG: That said,
I do believe that there's the previous
generation of computers were devices for your mind. And I think that we are
not brains and
tanks. I think that there's a
philosophical view of people of like, OK, you are
primarily what you think about or your values or something. It's like, no, you
are that and you are a physical manifestation. And people were very physical. And I think building a computer
for your whole body and not just for your mind is very
fitting with this worldview that the actual essence
of you, if you want to be present with
another person, if you want to be fully engaged
in experience is not just-- it's not
just a video conference
call that looks at your face and where you can share ideas. It's something that you
can engage your whole body. So, yeah I mean, I
think being physical is very important to me. I mean, that's a lot of the
most fun stuff that I get to do. It's a really important
part of how I personally balance my energy
levels and just get a diversity of experiences
because I could spend all my time running the company. But I think it's good for people
to do some different things and com
pete in different areas
or learn different things. And all of that is good. If people want to do really
intense workouts with the work that we're doing with Quest
or with eventual AR glasses, great. But even if you don't want to
do a really intense workout, I think just having a computing
environment and platform which is inherently physical captures
more of the essence of what we are as people than any of
the previous computing platforms that we've had to date. ANDREW HUBERMAN: I
was even think
ing just of the simple task of getting
better range of motion a.k.a. flexibility. I could imagine, inside
of the VR experience, leaning into a stretch, standard
type of lunge-type stretch, but actually seeing a
meter of are you are you approaching new
levels of flexibility in that moment
where it's actually measuring some
kinesthetic elements on the body in the joints,
whereas normally, you might have to do that in front
of a camera, which then would give you the data on a screen
that you'd look
at afterwards or hire an expensive coach or
looking at form and resistance training. So you're actually
lifting physical weights. But it's telling you whether
or not you're breaking form. I mean, there's just
so much that could be done inside of there. And then my mind
just starts to spiral into, wow, this is very
likely to transform what we think of as,
quote unquote, "exercise." MARK ZUCKERBERG:
Yeah, I think so. I think there's still
a bunch of questions that need to get answered. I don't th
ink most people
are going to necessarily want to install a lot of
sensors or cameras to track their whole body. So we're just over
time getting better from the sensors that are on
the headsets of being able to do very good hand tracking. So we have this
research demo where you now, just with the hand
tracking from the headset, you can type. It just projects a little
keyboard onto your table. And you can type. And people type like 100
words a minute with that. ANDREW HUBERMAN: With
a virtual keyb
oard? MARK ZUCKERBERG: Yeah. We're starting to be able to-- using some modern AI
techniques, be able to simulate and understand where
your torso's position is. Even though you
can't always see it, you can see it a
bunch of the time. And if you fuse
together what you do see with the accelerometer
and understanding how the thing is
moving, you can kind of understand what the body
position is going to be. But some things are
still going to be hard. So you mentioned boxing. That one works pretty wel
l
because we understand your head position. We understand your hands. And now, we're increasingly
understanding your body position. But let's say you
want to expand that to Muay Thai or kickboxing. OK. So legs, that's a
different part of tracking. That's harder because that's
out of the field of view more of the time. But there's also the
element of resistance. So you can throw a
punch, and retract it, and shadow box and do
that without upsetting your physical balance that much. But if you want
to
throw a roundhouse kick and there's no
one there, then, I mean, the standard way that you
do it when you're shadowboxing is you basically
do a little 360. But I don't know. Is that going to feel great? I mean, I think there's
a question about what that experience should be. And then if you want
to go even further, if you want to get
grappling to work, I'm not even sure
how you would do that without having resistance
of understanding what the force is applied to you would be. And then you get
into, OK, maybe you're going to have some
kind of bodysuit that can apply haptics. But I'm not even sure that even
a pretty advanced haptic system is going to be able to be
quite good enough to simulate the actual forces that would be
applied to you in a grappling scenario. So this is part of what's
fun about technology, though, is you keep on
getting new capabilities. And then you need to
figure out what things you can do with them. So I think it's really
neat that we can do boxing. And we can
do the
supernatural thing. And there's a bunch
of awesome cardio, and dancing and
things like that. And then there's also
still so much more to do that I'm excited
to get to over time. But it's a long journey. ANDREW HUBERMAN: And what
about things like painting, and art and music? I imagine-- of course,
different mediums-- I like to draw with
pen and pencil. But I could imagine trying to
learn how to paint virtually. And of course, you could
print out a physical version of that at the end. This
doesn't have to depart
from the physical world. It could end in
the physical world. MARK ZUCKERBERG: Did
you see the demo, the piano demo where you-- either you're there
with a physical keyboard or it could be a
virtual keyboard. But the app basically
highlights what keys you need to press in
order to play the song. So it's basically like
you're looking at your piano. And it's teaching you how to
play a song that you choose. ANDREW HUBERMAN:
An actual piano? MARK ZUCKERBERG: Yeah. ANDREW HUBERM
AN: But it's
illuminating certain keys in the virtual space. MARK ZUCKERBERG: Yeah. And it could either be a
virtual piano or a keyboard if you don't have a
piano or keyboard. Or it could use your
actual keyboard. So yeah, I think
stuff like that is going to be really fascinating
for education and expression. ANDREW HUBERMAN: And excuse
me, but for broadening access to expensive equipment. I mean, a piano is
no small expense. MARK ZUCKERBERG: Exactly. ANDREW HUBERMAN: And it
takes up a lot of sp
ace and needs to be tuned. You can think of all
these things, the kid that has very little
income or their family has very little
income could learn to play a virtual piano
at a much lower cost. MARK ZUCKERBERG: Yeah. And it gets back
to the question I was asking before about this
thought experiment of how many of the things
that we physically have today actually need
to be physical. The piano doesn't. Maybe there's some
premium where-- maybe it's a somewhat better,
more tactile experience to ha
ve a physical one. But for people who don't
have the space for it, or who can't afford
to buy a piano, or just aren't sure
that they would want to make that investment at
the beginning of learning how to play piano, I
think, in the future, you'll have the option
of just buying an app or a hologram piano which
will be a lot more affordable. And I think that's going to
unlock a ton of creativity too because instead of the
market for piano makers being constrained to like a
relatively small set of
experts who have perfected
that craft, you're going to have kids or developers
all around the world designing crazy designs for potential
keyboards and pianos that look nothing like
what we've seen before, but maybe bring even
more joy or even more fun into the world
where you have fewer of these physical constraints. So I think there's going to be
a lot of wild stuff to explore. ANDREW HUBERMAN:
There's definitely going to be a lot of
wild stuff to explore. I just had this
idea/image in my mind
of what you were talking
about merged with our earlier conversation when
Priscilla was here. I could imagine a time
not too long from now where you're using mixed reality
to run experiments in the lab, literally mixing
virtual solutions, getting potential outcomes,
and then picking the best one to then go actually do
in the real world, which is very both financially
costly and time-wise costly. MARK ZUCKERBERG: Yeah. I mean, people are already using
VR for surgery and education on it. And there
's some study that
was done that basically tried to do a controlled experiment
of people who learned how to do a specific surgery
through just the normal textbook and lecture method
versus you show the knee and you have it be a
large, blown-up model. And people can manipulate
it and practice where they would make the cuts. And like the people in
that class did better. Yeah, I think that it's
going to be profound for a lot of different areas. ANDREW HUBERMAN: And the last
example that leaps to mi
nd-- I think social media
and online culture has been accused of creating
a lot of real world-- let's call it physical world
social anxiety for people. But I could imagine practicing
a social interaction. Or a kid that has a
lot of social anxiety or that needs to advocate
for themselves better learning how to do
that progressively through a virtual
interaction, and then taking that to the real world because,
in my very recent experience today, it's so blended
now with real experience that the ki
d that
feels terrified of advocating for
themselves, or just talking to another human
being, or an adult, or being in a new circumstance
of a room full of kids, you could really experience
that in silico first and get comfortable,
let the nervous system attenuate a bit, and then take
it into the, quote unquote, "physical world." MARK ZUCKERBERG:
Yeah, I think we'll see experiences like that. I mean, I also think that
some of the social dynamics around how people
interact in this kind of blended
digital world will
be more nuanced in other ways. So I'm sure that there will
be new anxieties that people develop too, just like teens
today need to navigate dynamics around texting
constantly that we just didn't have when we were kids. So I think it will
help with some things. I think that there will be new
issues that hopefully we can help people work through too. But overall, yeah, I
think it's going to be really powerful and positive. ANDREW HUBERMAN: Let's
talk about the glasses. MARK ZUCK
ERBERG: Sure. ANDREW HUBERMAN: This was wild. Put on a Ray-Bans-- I like the way they look. They're clear. They look like any
other Ray-Ban glasses, except that I could
call out to the glasses. I could just say,
hey Meta, I want to listen to the
Bach variations-- the Goldberg Variations of Bach. And Meta responded. And no one around me could hear. But I could hear with
exquisite clarity. And by the way, I'm not getting
paid to say any of this. I'm just still
blown away by this. Folks, I want a
t
hese very badly. I could hear, OK, I'm
selecting those now-- or that music now. And then I could hear
it in the background. But then I could still
have a conversation. So this was neither headphones
in nor headphones out. And I could say,
wait, pause the music. And it would pause. And the best part was I
didn't have to, quote unquote, "leave the room" mentally. I didn't even have
to take out a phone. It was all interfaced through
this very local environment in and around the head. And as a neuro
scientist,
I'm fascinated by this because, of course, all of
our perceptions-- auditory, visual et cetera-- are occurring inside the casing
of this thing we call a skull. But maybe you could
comment on the origin of that design for you,
the ideas behind that, and where you think it
could go because I'm sure I'm just scratching the surface. MARK ZUCKERBERG:
The real product that we want to
eventually get to is this full augmented
reality product in a stylish and comfortable
normal glasses form fa
ctor. ANDREW HUBERMAN: Not a dorky
VR headset, so to speak? MARK ZUCKERBERG: No, I mean-- ANDREW HUBERMAN: Because
the VR headset does feel kind of big on the face. MARK ZUCKERBERG: There's
going to be a place for that, too, just like you
have your laptop and you have your workstation. Or maybe the better analogy
is you have your phone and you have your workstation. These AR glasses are going to
be like your phone in that you have something on your face. And you will, I think,
be able to, if you
want, wear it for a lot of
the day and interact with it very frequently. I don't think that
people are going to be walking around the
world wearing VR headsets. ANDREW HUBERMAN: Let's hope. MARK ZUCKERBERG: But yeah,
that's certainly not the future that I'm hoping we get to. But I do think that there
is a place for having-- because it's a
bigger form factor, it has more compute power. So just like your workstation
or your bigger computer can do more than
your phone can do, there's a place for
t
hat when you want to settle into an intense task. If you have a doctor
who's doing a surgery, I would want them doing
it through the headset not through the phone equivalent
or the lower powered glasses. But just like phones
are powerful enough to do a lot of things, I think
the glasses will eventually get there, too. Now, that said, there's a
bunch of really hard technology problems to address in order
to be able to get to this point where you can put full
holograms in the world. You're basical
ly
miniaturizing a supercomputer and putting it into a glasses
so that the glasses still look stylish and normal. And that's a really
hard technology problem. Making things small
is really hard. A holographic display
is different from what our industry has optimized
for for 30 or 40 years now, building screens. There's a whole
industrial process around that goes into phones,
and TVs, and computers, and increasingly so many things
that have different screens. There's a whole pipeline
that's gotte
n very good at making that kind of screen. And the holographic
displays are just a completely different thing
because it's not a screen. It's a thing that
you can shoot light into through a laser or some
other kind of projector. And it can place that as
an object in the world. So that's going to need to be
this whole other industrial process that gets built up to
doing that in an efficient way. So all that said,
we're basically taking two different approaches
towards building this at once. One i
s we are trying to keep
in mind what is the long-term thing that-- it's not super far off. Within a few years,
I think we'll have something that's a first
version of this full vision that I'm talking about. I mean, we have something
that's working internally that we use as a dev kit. But that one, that's
a big challenge. It's going to be more expensive. And it's harder to get
all the pieces working. The other approach has
been, all right, let's start with what
we know we can put into a pair of
s
tylish sunglasses today and just make
them as smart as we can. So for the first
version, we worked with-- we did this collaboration
with Ray-Ban because that's well-accepted. These are well-designed glasses. They're classic. People have used
them for decades. For the first version, we
got a sensor on the front, so you could capture moments
without having to take your phone out of your pocket. So you got photos and videos. You had the speaker
and the microphone, so you can listen to music. You co
uld communicate with it. But that was the
first version of it. We had a lot of
the basics there. But we saw how people used it. And we tuned it. We made the camera twice as
good for this new version that we made. The audio is a lot
crisper for the use cases that we saw that people actually
used, which is-- some of it is listening to music. But a lot of it is people want
to take calls on their glasses. They want to listen to podcasts. But the biggest thing that
I think is interesting is the abili
ty to get AI running
on it, which it doesn't just run on the glasses. It also kind of proxies
through your phone. But I mean, with all
the advances in LLMs-- we talked about this a
bit in the first part of the conversation. Having the ability to have
your Meta AI assistant that you can just
talk to and basically ask any question
throughout the day is-- I think it'd be
really fascinating. And like you were
saying about how we process the world as
people, eventually, I think you're going
to want y
our AI assistant to be able to see what
you see and hear what you hear. Maybe not all the time. But you're going to
want to be able to tell it to go into a mode where it
can see what you see and hear what you hear. And what's the
device design that best positions an
AI assistant to be able to see what
you see and hear what you hear so it
can best help you? Well, that's glasses,
where it basically has a sensor to be able
to see what you see and a microphone that is
close to your ears that can hea
r what you hear. The other design goal
is, like you said, to keep you present
in the world. So I think one of the
issues with phones is they pull you away from
what's physically happening around you. And I don't think that the
next generation of computing will do that. ANDREW HUBERMAN: I'm
chuckling to myself because I have a friend. He's a very well
known photographer. And he was laughing about
how people go to a concert. And everyone's filming
the concert on their phone so that they can be the
person that posts the thing. But there are literally
millions of other people who posted the exact same thing. But somehow, it feels important
to post our unique experience. With glasses, that
would essentially smooth that gap completely. You could just worry about
it later, download it then. There are issues, I
realize, with glasses because they are so seamless
with everyday experience, even though you and I
aren't wearing them now. It's very common for
people to wear glasses-- issues of recor
ding and consent. [INTERPOSING VOICES] ANDREW HUBERMAN: Like if I go
to a locker room at my gym, I'm assuming that the people
with glasses aren't filming. Whereas right now, because
there's a sharp transition when there's a phone in the room
and someone's pointing it, people generally say, no phones
in locker rooms and recording. So that's just one instance. I mean, there are
other instances. MARK ZUCKERBERG: We have
the whole privacy light. Did you get-- ANDREW HUBERMAN: I didn't
get a chance t
o explore that. MARK ZUCKERBERG: Yeah. So anytime that it's
active, that the camera sensor is active, it's basically
pulsing a white bright light. ANDREW HUBERMAN: Got it. MARK ZUCKERBERG: Which is, by
the way, more than cameras do. ANDREW HUBERMAN: Right. Someone could be
holding a phone. MARK ZUCKERBERG: Yeah. I mean, phones aren't showing
a light, bright sensor when you're taking a photo. ANDREW HUBERMAN:
People oftentimes will pretend they're texting
and they're actually recording. I actuall
y saw an instance
of this in a barber shop once, where someone
was recording and they were pretending
that they were texting. And it was interesting. There was a pretty intense
interaction that ensued. And it was like, wow, it's
pretty easy for people to feign texting while
actually recording. MARK ZUCKERBERG: Yeah. So I think when
you're evaluating a risk with a new technology,
the bar shouldn't be is it possible to do anything bad. It's does this new
technology make it easier to do something b
ad than
what people already had. And I think because you have
this privacy light that is just broadcasting to everyone
around you, hey, this thing is recording now-- I think that makes it
actually less discreet to do it through the
glasses than what you could do with a phone already, which
I think is basically the bar that we wanted to get over
from a design perspective. ANDREW HUBERMAN: Thank
you for pointing out that it has the privacy light. I didn't get long
enough in the experience to explo
re all the features. But again, I can think
of a lot of uses-- being able to look at a
restaurant from the outside and see the menu, get a
status on how crowded it is. As much as I love-- I don't want to call
out-- let's just say app-based map functions
that allow you to navigate and the audio is OK. It's nice to have a conversation
with somebody on the phone or in the vehicle. And it'd be great if the road
was traced where I should turn. MARK ZUCKERBERG:
Yeah, absolutely. ANDREW HUBERMAN:
These
kinds of things seem like it's going to be
straightforward for Meta engineers to create. MARK ZUCKERBERG: Yeah, in
a version, we'll have it so it'll also have the
holographic display, where it can show you the directions. But I think that there
will basically just be different price points
that pack different amounts of technology. The holographic
display part, I think, is going to be more
expensive than doing one that just has the AI, but
is primarily communicating with you through audio. So I
mean, the current
Ray-Ban Meta glasses are $299. I think when we have one
that has a display in it, it'll probably be some
amount more than that. But it'll also be more powerful. So I think that
people will choose what they want to use based
on what the capabilities are that they want and
what they can afford. But a lot of our goal
in building things is we try to make things that
can be accessible to everyone. Our game as a company isn't to
build things and then charge a premium price for it. W
e try to build things that
then everyone can use, and then become more useful because a
very large number of people are using them. So it's just a very
different approach. We're not like Apple or some
of these companies that just try to make something and
then sell it for as much as they can, which, I mean,
they're a great company. So I mean, I think that
model is fine, too. But our approach
is going to be we want stuff that
can be affordable so that way everyone in
the world can use it. ANDREW
HUBERMAN:
Long lines of health, I think the glasses will
also potentially solve a major problem in
a real way, which is the following for
both children and adults. It's very clear that viewing
objects in particular screens up close for too many hours
per day leads to myopia. It literally changes the
length of length of the eyeball and nearsightedness. And on the positive
side, we know, based on some really
large clinical trials, that kids who spend-- and adults who spend two hours
a day or more
out of doors don't experience that and maybe
even reverse their myopia. And it has something to do
with exposure to sunlight. But it has a lot to do
with long view, viewing things at a distance greater
than three or four feet away. And with the glasses,
I realize, one could actually do digital
work out of doors. It could measure and
tell you how much time you've spent looking at things
up close versus far away. I mean, this is just another
example that leaps to mind. But in accessing
the visual
system, you're effectively
accessing the whole brain because it's the only
two bits of brain that are outside the cranial
vault. So it just seems like putting technology
right at the level of the eyes, seeing what the
eyes see, has just got to be the best way to go. MARK ZUCKERBERG: Yeah. Well, multimodal, I think, is--
you want the visual sensation. But you also want
text or language. ANDREW HUBERMAN: Sure. That all can be brought to
the level of the eyes, right? MARK ZUCKERBERG: What
do you me
an by that? ANDREW HUBERMAN:
Well, I mean, I think what we're describing
here is essentially taking the phone, the
computer, and bringing it all to the level of the eyes. And of course, one would like-- MARK ZUCKERBERG: Oh,
Physically at your eyes? ANDREW HUBERMAN: Physically
at your eyes, right? MARK ZUCKERBERG: Yeah. ANDREW HUBERMAN: And one
would like more kinesthetic information, as you mentioned
before-- where the legs are, maybe even lung function. Hey, have you taken
enough steps today? B
ut that all can be-- if it can
be figured out on the phone, it can be-- by the phone, it can
be figured out by the glasses. But there's additional
information there, such as what are you
focusing on in your world. How much of your time is spent
looking at things far away versus up close? How much social time
did you have today? It's really tricky to
get that with a phone. If my phone were
right in front of us as if we were at
a standard lunch nowadays, certainly
in Silicon Valley, and then we're
peering
at our phones, I mean, how much real direct attention
and was in the conversation at hand versus something else? You can get issues
of where are you placing your attention
by virtue of where you're placing your eyes. And I think that information
is not accessible with a phone in your
pocket or in front of you. Yeah, I mean, a little bit, but
not nearly as rich and complete information as one
gets when you're really pulling the data from
the level of vision and what kids and
adults are a
ctually looking at and attending to. MARK ZUCKERBERG: Yeah, yeah. ANDREW HUBERMAN: It
seems extremely valuable. You get autonomic information,
size of the pupils. So you get information
about internal states. MARK ZUCKERBERG: I mean, there's
internal sensor and outside. So the sensor on the Ray-Ban
Meta glasses is external. So it basically allows
you to see what you see-- sorry, the AI system to
see what you're seeing. There's a separate
set of things which are eye tracking, which are
also very
powerful for enabling a lot of interfaces. So if you want to
just look at something and select it by looking
at it with your eyes rather than having to drag
a controller over or pick up a hologram or
anything like that, you can do that
with eye tracking. So that's a pretty profound and
cool experience, too, as well as just understanding
what you're looking at so that way you're
not wasting compute power drawing pixels and
high resolution in a part of the world
that's going to be in your peripher
al vision. So yeah, all of
these things, there are interesting design
and technology trade-offs, where if you want the external
sensor, that's one thing. If you also want
the eye tracking, now that's a different
set of sensors. Each one of these
consumes compute, which consumes battery. They take up more space. So it's like, where are the eye
tracking sensors going to be? It's like, well, you
want to make sure that the rim of the glasses is
actually quite thin because-- I mean, there's a varianc
e
of how thick can glasses be before they look more
like goggles than glasses. So I think that there's
this whole space. And I think people are going
to end up choosing what product makes sense for them. Maybe they want something
that's more powerful, that has more of the
sensors, but it's going to be a little
more expensive, maybe like slightly thicker. Or maybe you want
a more basic thing that just looks very similar
to what Ray-Ban glasses are that people have been wearing
for decades but has
AI in it and you can capture
moments without having to take your phone out
and send them to people. In the latest version, we got
the ability in to live stream. I think that that's pretty
crazy, that now you can be-- going back to your concert case
or whatever else you're doing, you can be doing
sports or watching your kids play something. And you can be watching. And you can be live streaming
it to your family group, so people can see it. I think that stuff is-- I think that's pretty
cool, tha
t you basically have a normal looking glasses at
this point that can live stream and has an AI assistant. So the stuff is making
a lot faster progress in a lot of ways than
I would have thought. And I don't know. I think people are going
to like this version. But there's a lot
more still to do. ANDREW HUBERMAN: I think
it's super exciting. And I see a lot of technologies. This one's particularly
exciting to me because of how smooth
the interface is and for all the reasons
that you just mentioned
. What's happening with and
what can we expect around AI interfaces and
maybe even avatars of people within social media? Are we not far off
from a day where there are multiple
versions of me and you on the
internet or people? For instance, I get
asked a lot of questions. I don't have the opportunity to
respond to all those questions. But with things
like ChatGPT, people are trying to generate
answers to those questions on other platforms. Will I have the
opportunity to soon have an AI version o
f
myself where people can ask me questions about
what I recommend for sleep and circadian rhythm, fitness,
mental health, et cetera based on content I've
already generated that will be accurate so they
could just ask my avatar? MARK ZUCKERBERG: Yeah,
this is something that I think a lot
of creators are going to want that we're
trying to build and I think we'll probably
have a version of next year. But there's a bunch
of constraints that I think we need to
make sure that we get right. So for one,
I think it's
really important that-- it's not that there's a
bunch of versions of you. It's that if anyone is creating
an AI assistant version of you, it should be something
that you control. I think there are some platforms
that are out there today that just let people like make-- I don't know-- an AI bought
of me or other figures. And it's like, I don't know. I mean, we have
platform policies for-- and for decades,
since the beginning of the company at this point,
which is almost 20 years, th
at basically don't
allow impersonation. Real identity is like
one of the core aspects that our company was started on. You want to authentically
be yourself. So yeah, I think if
you're almost any creator, being able to engage
your community-- and there's just going
to be more demand to interact with you than
you have hours in the day. So there are both
people who out there who would benefit from
being able to talk to an AI version of you. And I think you,
and other creators, would benefit from b
eing able
to keep your community engaged and service that demand that
people have to engage with you. But you're going to want to
know that that AI version of you or assistant is going
to represent you the way that you would want. And there are a
lot of things that are awesome about
these modern LLMs. But having perfect
predictability about how it's going
to represent something is not one of the
current strengths. So I think that
there's some work that needs to get done there. I don't think it n
eeds to be
100% perfect all the time. But you need to have very
good confidence, I would say, that it's going to represent
you the way that you'd want for you to
want to turn it on, which, again, you
should have control over whether you turn it on. So we wanted to start in
a different place, which I think is a somewhat easier
problem, which is creating new characters for AI personas. So that way, it's not-- we built one of the
AIs is like a chef. And they can help you
come up with things that yo
u could cook and
can help you cook them. There's a couple
of people that are interested in different
types of fitness that can help you plan
out your workouts or help with recovery or
different things like that. There's an AI that's
focused on DIY crafts. There's somebody who's
a travel expert that can help you make travel
plans or give you ideas. But the key thing
about all of these is they're not modeled
off of existing people. So they don't have to have
100% fidelity to making sure that they
never say something
that a real person who they're modeled after would never say
because they're just made up characters. So I think that that's a
somewhat easier problem. And we actually got a bunch
of different well-known people to play those characters
because we thought that would make it more fun. So there's like Snoop Dogg
is the dungeon master. So you can drop
him into a thread and play text-based games. And I do this with my daughter
when I tuck her in at night. And she just loves storyt
elling. And it's like Snoop Dogg,
as the dungeon master, will come up with here's
what's happening next. And she's like, OK, I
turn into a mermaid. And then I like
swim across the bay. And I go and find the
treasure chest and unlock it. And it's like, and then
Snoop Dogg just always will have a next
version of the-- a next iteration on the story. So I mean, it's
stuff that's fun. But it's not
actually Snoop Dogg. He's just the actor who's
playing the dungeon master, which makes it more fun. So I
think that's probably
the right place to start, is you can build versions
of these characters that people can interact
with doing different things. But I think where
you want to get over time is to the place where any
creator or any small business can very easily just create an
AI assistant that can represent them and interact with your
community or customers, if you're a business,
and basically just help you grow your enterprise. So I think that's
going to be cool. It's a long-term project. I
think we'll have more progress
on it to report on next year. But I think that's coming. ANDREW HUBERMAN: I'm
super excited about it because we hear a lot
about the downsides of AI. I mean, I think people are now
coming around to the reality that AI is neither good nor bad. It can be used for good or bad. And there are a lot of
life-enhancing spaces that it's going to show
up and really, really improve the way that we engage
socially, what we learn, and that mental health
and physical health don'
t have to
suffer and, in fact, can be enhanced by the
sorts of technologies we've been talking about. So I know you're extremely busy. I so appreciate the
large amount of time you've given me today to sort
through all these things. MARK ZUCKERBERG:
Yeah, it's been fun. ANDREW HUBERMAN: And to
talk with you and Priscilla and to hear what's happening
and where things are headed, the future certainly is bright. I share in your optimism. And it's been only strengthened
by today's conversation. So th
ank you so much. And keep doing
what you're doing. And on behalf of myself
and everyone listening, thank you because, regardless
of what people say, we all use these
platforms excitedly. And it's clear that
there's a ton of intention, and care, and thought about what
could be in the positive sense. And that's really
worth highlighting. MARK ZUCKERBERG:
Awesome, thank you. I appreciate it. ANDREW HUBERMAN: Thank
you for joining me for today's discussion with Mark
Zuckerberg and Dr. Priscilla Chan
. If you're learning from
and/or enjoying this podcast, please subscribe to
our YouTube channel. That's a terrific zero-cost
way to support us. In addition, please
subscribe to the podcast on both Spotify and Apple. And on both Spotify
and Apple, you can leave us up to
a five star review. Please also check
out the sponsors mentioned at the beginning and
throughout today's episode. That's the best way to
support this podcast. If you have questions for me,
or comments about the podcast, or guests
that you'd
like me to consider hosting on the
Huberman Lab podcast, please put those in the
comment section on YouTube. I do read all the comments. Not during today's episode,
but on many previous episodes of the Huberman Lab podcast,
we discuss supplements. While supplements aren't
necessary for everybody, many people derive tremendous
benefit from them for things like enhancing sleep, hormone
support, and improving focus. If you'd like to learn more
about the supplements discussed on the Huber
man Lab podcast,
you can go to live momentous-- spelled O-U-S-- so livemomentous.com/huberman. If you're not already
following me on social media, it's hubermanlab on all
social media platforms. So that's Instagram,
Twitter-- now called X-- Threads, Facebook, LinkedIn. And on all those
places, I discuss science and science-related
tools, some of which overlaps with the content
of the Huberman Lab podcast, but much of which is
distinct from the content on the Huberman Lab podcast. So again, it's
hubermanlab on
all social media platforms. If you haven't
already subscribed to our monthly Neural
Network Newsletter, the Neural Network Newsletter
is a completely zero-cost newsletter that gives
you podcast summaries as well as toolkits in
the form of brief PDFs. We have toolkits related to
optimizing sleep, regulating dopamine, deliberate
cold exposure, fitness, mental health,
learning, and neuroplasticity and much more. Again, it's completely
zero-cost to sign up. You simply go to
hubermanla
b.com, go over to the Menu tab, scroll
down to newsletter and supply your email. I should emphasize
that we do not share your email with anybody. Thank you once again for joining
me for today's discussion with Mark Zuckerberg
and Dr. Priscilla Chan. And last but
certainly not least, thank you for your
interest in science. [MUSIC PLAYING]
Comments
Mark Zuckerberg talking about human health is a bit ridiculous. His company is responsible for a massive decline in human mental wellbeing and a massive rise in stress.
Professor Huberman needs to seriously consider and pay attention to the vast majority of the comments below bcoz that is exactly what he needed before he engaged with and reached out to Mr. and Mrs. Zuckerberg for mental health discussion 😢
What interesting timing for this interview, as I just read this headline: “Dozens of states sue Meta over addictive features harming kids.”
Of all the people in the world, Mark Zuckerberg is among the last that I would want to have anything to do with my healthcare.
I love you to death Andrew but talking with Mark Zuckerberg about negative impact of social media is like talking with Richard Sackler about negative impact of opioid epidemic.
Andrew: people should go outside Mark: why if they just stayed inside with the metaverse?
Having Mark and his wife on the podcast is like having the CEO of Marlboro talking about health issues without mentioning the harmful effects of tobacco consumption. It really does not work does it 🤣🤣
Great timing for this episode. Just days after, Meta is being sued by US for making its products addictive and destroying mental health among young people.
“Technology is useful depending on how you use it” —> as long as you constantly keep clear of its traps developed by Mark and alikes
Social media use (addiction) is directly correlated with the increase in depression, anxiety and eating disorders. Teenagers spend over six hours a day on social media kids no longer move. It has been called the greatest health risk of our time. It is ironic to have Mark Zuckerberg and his wife on the show talking about the future of health and creating tools when they are responsible for one of the biggest health crisis of today. If they wanted to help, they would do more than tools that nobody uses, especially teenagers. It’s not surprising that the medical doctor was not on for the second part of the podcast as she would not be able to lie about the science.
Hi Andrew! First and foremost, I want to express my immense gratitude for the incredible work you've been doing with your podcast. Your content has touched the lives of hundreds of thousands of people, including mine, and it's been truly transformative. However, I couldn't help but have some mixed feelings about this interview. While I appreciate the importance of discussing such topics, it does strike me as a bit hypocritical. Facebook and Instagram, have been repeatedly criticized for their role in promoting various mental health issues like anxiety, depression, anorexia, addiction and the list goes on. It's disheartening to see the tech giants aware of these problems yet not taking more meaningful action to address them. I hope your podcast can continue to shed light on these issues, perhaps even by inviting more experts who can offer solutions to these problems. Keep up the great work, and thank you for being a platform for these essential discussions.
Also, we don’t need cure for diseases, we need prevention! Quitting social media is one of the best actions to stay sane and be in touch with reality a
Quit Instagram. Quit Facebook. Life will be better.
This is the first Huberman Lab podcast episode that I have been disappointed by. Unlike some others, I am not against getting an individual like Mark Zuckerberg on the podcast, because there is the chance to question him and create a thought-provoking discussion. However, there should absolutely have been more pressing questioning surrounding the issues of social media addiction and screen time. This podcast seemed more like an advertisement for the Meta products and services. Please take these comments into consideration, Andrew.
What I love about your content is how practical and proven your and your guests’ advice is. I couldn’t listen for more than 20 minutes because this feels like sales pitch full of empty promises :/
Fantastic! I've always wanted to hear the story about his discovery of penicillin, and the time he cured cancer, and the way he managed to improve health data management.
Sorry but the best mental health tool is to take a break from Instagram and Facebook. I don't think Zuck really cares about others well being
When someone says: “we’re trying to cure ALL diseases”, I get very cautious….
Former Huberman Lab Premium Member, here. Harkening back to the first episodes - with our beloved 'Costello' snoring in the background, my dog 'Buck' (a robust, 100 lb. yellow lab) would listen on a bluetooth speaker. It seemed 'Buck' found Costello's snoring soothing. Loved the episode where you addressed those complaining about Costello's snoring and you said - and I quote - "Sorry - not sorry. Costello stays." Loved the 90-minute Ultradian Cycle-length episodes of just you speaking to the camera. They were clear, concise, and straight to the point. They honored the listener's time and attention. Told everyone I know about The HLP and how - the only way we'd be able to hear a lecture from someone of your stature, would be to enroll in Stanford University Medical School. Those were great episodes, and I was grateful to offer my humble $100 bucks to be a Premium Member of The HLP. It was the least I could do to repay such stellar content. The past year that changed, and I found the format longer and longer, and the guest lists spotty. With that, I decided not to renew my Premium Membership. Such is the price of success in any enterprise. A machine forms around a winning model, and that machine needs to be fed more and more. Your format was once unique, Dr. Huberman - and I'll always be grateful for it. Keep up your stellar work. My very best regards in all your endeavors.
I’m not interested in this episode.. I just came here to say how proud I am of my fellow Andrew Huberman followers. Reading these comments actually healed my face a bit from the slap it got. 👊🏽❤️