Mark Zuckerberg & Yuval Noah Harari in Conversation

Hey, everyone. This year I’m doing
a series of public discussions on the future of the Internet and society
and some of the big issues around that. And today I’m here with Yuval Noah Harari. A great historian and best-selling author of a number of books. His first book,
Sapiens: A Brief History of Humankind, chronicled and did an analysis, going from the early days
of hunter/gatherer society to now how our civilization is organized. And your next two books, the Homo Deus: A Brief History of Tomorrow
and 21 Lessons for the 21st Century, actually tackle important issues
of technology and the future. And that’s a lot
of what we’ll talk about today. But most historians
only tackle and analyze the past. But a lot of the work that you’ve done
has had really interesting insights and raised important questions
for the future. So I’m really glad to have an opportunity
to talk with you today. So, Yuval, thank you for joining
for this conversation. Yeah, I’m happy to be here. I think that if historians
and philosophers cannot engage with the current questions of technology
and the future of humanity, then we aren’t doing our jobs. We are not just supposed
to chronicle events centuries ago. All the people that lived in the past
are dead. They don’t care. The question is, what happens to us
and to the people in the future? Yeah. All right, so all the questions
that you’ve outlined, where should we start here? And one of the big topics
that we’ve talked about is around this dualism around whether… With all of the technology in progress
that has been made, are people coming together
and are we becoming more unified? Or is our world becoming more fragmented? And so, I’m curious to start off
by how you’re thinking about that, and that’s probably a big area. We could probably spend most of the time
on that topic. Yeah, if you look
at the long span of history, then it’s obvious that humanity’s becoming
more and more connected. If, thousands of years ago,
planet Earth was actually a galaxy of a lot of isolated worlds
with almost no connection between them, so, gradually, people came together
and became more and more connected until we reached today
when the entire world, for the first time, is a single historical,
economic and cultural unit. But connectivity
doesn’t necessarily mean harmony. The people we fight most often are our own family members
and neighbors and friends. So, it’s really a question of,
are we talking about connecting people or are we talking
about harmonizing people? Connecting people
can lead to a lot of conflicts. And when you look at the world today, you see this duality… For example, in the rise of walls, which we talked about earlier when we met. Yeah. Which, for me, is something that I just
can’t figure out what is happening because you have
all this new connecting technology and the Internet and virtual realities
and social networks. And then the most… One of the top
political issues becomes building walls. And not just cyber walls or firewalls,
building stone walls. Like the most Stone Age technology
is suddenly the most advanced technology. So, how to make sense of this world,
which is more connected than ever, but at the same time
is building more walls than ever before. Yeah, well, I think
one of the interesting questions is around whether there’s actually
so much of a conflict between these ideas of people becoming more connected and this fragmentation
that you talk about. One of the things that it seems to me
is that we… In the 21st century, in order to address
the biggest opportunities and challenges that humanity has… All right, so,
I think it’s both opportunities. Spreading prosperity, spreading peace,
scientific progress, as well as some of the big challenges. Right, addressing climate change, making sure that, on the flip side,
diseases don’t spread and that there aren’t epidemics
and things like that. We really need to be able to come together
and have the world be more connected. But at the same time, that only works if we, as individuals, have our economic
and social and spiritual needs met. And so, one way to think about this
is in terms of fragmentation, but another way to think about it
is in terms of personalization, right? And I just think about, when I was growing up, one of the big things
that I think the Internet enables is for people
to connect with groups of people who share their real values and interests. And it wasn’t always like this, right? Before the Internet you were really tied
to your physical location. And I just think about
how when I was growing up, I grew up in a town of about 10,000 people and there were only
so many different clubs or activities that you could do. So, I grew up,
like a lot of the other kids, playing Little League Baseball. And I think about this in retrospect and it’s like
I’m not really into baseball, I’m not really an athlete
so why did I play Little League when my real passion
was programming computers? And the reality was that, growing up,
there was no one else, really, in my town who was into programming computers, so I didn’t have a peer group or a club
that I could do that. It wasn’t until I went to boarding school
and then, later, college, where I actually was able to meet people
who were into the same things as I am. And now with the Internet,
that’s starting to change, right? And now you have the ability to not just
be tethered to your physical location, but to find people who have
more niche interests and different kind of subcultures
and communities on the Internet, which I think is a really powerful thing. But it also means
that me, growing up today, I probably wouldn’t have played
Little League. And you can think about
me playing Little League as… That could’ve been a unifying thing, where there weren’t that many things
in my town, so that was a thing
that brought people together. So maybe if I was creating…
Or if I was a part of a community online, that might’ve been more meaningful to me, getting to know real people, but around programming,
which is my real interest, you would’ve said that our community,
growing up, would’ve been more fragmented, right? And people wouldn’t have had
the same sense of physical community. So, when I think about these problems,
one of the questions that I wonder is fragmentation, personalization
or finding what you actually care about are two sides of the same coin. But the bigger challenge
that I worry about is whether there are a number of people who are just left behind
in the transition, who were people
who would’ve played Little League, but haven’t now found their new community
and now just feel dislocated. And maybe their primary orientation
in the world is still the physical community
that they’re in… Or they haven’t really been able to find a community of people
who they’re interested in. And as the world has progressed, I think a lot of people
feel lost in that way. And that probably contributes
to some of the feelings. That would be my hypothesis, at least. That’s the social version of it. There’s also the economic version
around globalization, which I think is as important. But I’m curious
to what you think about that. About the social issue,
unlike online communities, can be a wonderful thing, but they are still incapable
of replacing physical communities, -because there are still so many things…
-That’s definitely true. That you can only do with your body
and with your physical friends. And you can travel with your mind
throughout the world, but not with your body. And there are huge questions about
the cost and benefits there. And also the ability of people to just escape things they don’t like
in online communities, but you can’t do it
in real offline communities. You can unfriend your Facebook friends, but you can’t un-neighbor your neighbors. -They are still there.
-Yeah. You can take yourself
and move to another country if you have the means, but most people can’t. So, part of the logic
of traditional communities was that you must learn how to get along with people you don’t like,
necessarily, maybe. And you must develop social mechanisms
how to do that. And with online communities… And they have done
some really wonderful things for people, but also they don’t give us the experience of doing these difficult
but important things. Yeah, and I definitely don’t mean to state that online communities
can replace everything that a physical community did. The most meaningful online communities
that we see are ones that span online and offline, that bring people together… Maybe the original organization
might be online, but people
are coming together physically because that, ultimately,
is really important for relationships… ’cause we’re physical beings, right? So, whether it’s… You know, there are lots of examples around whether it’s an interest community
where people care about running, but they also care about
cleaning up the environment. So, a group of people organize online,
and then they… Every week,
go for a run along a beach or through a town and clean up garbage. That’s like a physical thing. We hear about communities where, you know, people,
if you’re in a profession… Maybe the military
or maybe something else where you have to move around a lot, people form these communities
of military families or families of, you know,
groups that travel around. The first thing that they do
when they go to a new city is that they find that community, and then that’s how they get integrated
into the local physical community. So, that’s obviously
a super-important part of this that I don’t mean to understate. Yeah. And then the practical question
for a service provider like Facebook is, “What is the goal?” I mean,
are we trying to connect people, so ultimately they will leave the screens and go and play football
or pick up garbage? Or are we trying to keep them
as long as possible on the screens? And there is a conflict of interest there. One model would be “We want people
to stay as little as possible.” Online, we just need them to stay there
the shortest time necessary to form the connection,
which they will then go -and do something in the outside world.
-Yeah. That’s one of the key questions I think about what the Internet
is doing to people. Whether it’s connecting them
or fragmenting society. Yeah, and I think your point is right.
I mean, we basically went… We’ve made this big shift in our systems to make sure that they’re optimized
for meaningful social interactions. Which, of course, the most meaningful
interactions that you can have, are physical, offline interactions. And there’s always this question
when you’re building a service of how you measure the different thing
that you’re trying to optimize for. So, you know, it’s a lot easier
for us to measure if people are interacting
or messaging online than if you’re having
a meaningful connection physically. But there are ways to get at that. You can ask people questions about what the most meaningful things
that they did. You can’t ask all two billion people, but you can have
a statistical sub-sample of that, and have people come in and tell you, okay, what are the most meaningful things
that I was able to do today, and how many of them were enabled by me connecting with people online, or how much of it was me connecting
with someone physically. Maybe around the dinner table with content
or something that I learned online or saw. So, that is definitely
a really important part of it. But I think one of the important
and interesting questions is about the richness of the world
that can be built where you have, on one level,
unification or this global connection where there’s a common framework
where people can connect. Maybe it’s through using
common Internet services, or maybe it’s just common social norms
as you travel around. One of the things
that you’d pointed out to me in a previous conversation is now something that’s different
from any other time in history is that you can travel to almost
any other country, and look like you… Dress like you’re appropriate
and that you fit in there. 200 years ago or 300 years ago,
that just wouldn’t have been the case. If you went to a different country,
you would have just stood out immediately. There’s this norm… There’s this level of cultural norm
that is united, but then the question is,
“What do we build on top of that?” And, I think, one of the things
that a broader set of cultural norms or shared values and framework enables is a richer set of sub-cultures
and sub-communities, and people to actually go find
the things that they’re interested in. in lots of different communities, to be creative,
that wouldn’t have existed before. Going back to my story before, it wasn’t just my town
that had the Little League. You know, I think, when I was growing up, basically every town
had very similar things. There’s a Little League in every town. You know, maybe instead of every town
having Little League, there should be…
Little League should be an option. But if you want to do something that
not that many people were interested in, in my case, programming.
In other people’s case, maybe, you know, interest in some part
of history or some part of art that, there just may not be another person in your 10,000-person town
who share that interest. I think it’s good if you can form
those kind of communities. And now people have…
can find connections and can find a group of people
who share their interest. I think there’s a question, though,
of you can look at that as fragmentation. Right, because now
we’re not all doing the same thing. We’re not all going to church
and playing Little League and doing the exact same things. Or you can think about that as richness
and depthness in our social lives. I just think that
that’s an interesting question, is where you want
the commonality across the world, and the connection,
and where you actually want that commonality
to enable deeper richness, even if that means that
people are doing different things. I’m curious if you have a view on that
and where that’s positive versus where that creates
a lack of social cohesion. Yeah. Almost nobody would argue
with the benefits of a richer social environment
in which people have more options to connect around all kinds of things. The key question is how do you
still create enough social cohesion on a level of a country,
and increasingly also on the level of the entire globe
in order to tackle our main problems. I mean, we need global cooperation
like never before because we are facing
unprecedented global problems. We just had Earth Day,
and should be obvious to everybody, we cannot deal
with the problems of the environment, of climate change,
except through global cooperation. Similarly, if you think about
the potential disruption caused by new technologies
like artificial intelligence, we need to find a mechanism
for global cooperation around issues like how to prevent
an AI arms-race. How to prevent different countries
racing to build autonomous weapons systems and killer robots,
and weaponizing the Internet and weaponizing social networks. Unless we have global cooperation,
we can’t stop that. Because every country will say, “Well, we don’t want to produce
killer robots, it’s a bad idea. “But we can’t allow
our rivals to do it before us, “so we must do it first.”
And then you have a race to the bottom. Similarly, if you think about
the potential disruption to the job market and the economy,
caused by AI and automation… So, it’s quite obvious
that there will be jobs in the future. But will they be evenly distributed
between different parts of the world. One of the potential results
of the AI revolution could be the concentration of immense
wealth in some part of the world, and the complete bankruptcy
of other parts. There will be a lot of new jobs
for software engineers in California, but there will be maybe no jobs
for textile workers and truck drivers in Honduras and Mexico.
So what will they do? If we don’t find a solution
on the global level, like creating a global safety net to protect humans
against the shocks of AI, and enabling them
to use the opportunities of AI, then we will create the most unequal economic situation that ever existed. It will be much worse, even than
what happened in the industrial revolution when some countries industrialized,
most countries didn’t, and a few industrial powers
went on to conquer and dominate and exploit all the others. So, how do we create
enough global cooperation so that the enormous benefits
of AI and automation don’t go only to, say,
California and eastern China, while the rest of the world
is being left far behind. Yeah. I think that that’s important. I would unpack that
into two sets of issues. One, around AI and the future economic
and geopolitical issues around that. And let’s put that aside for a second, because I actually think
we should spend 15 minutes on that. -I mean, that’s a big…
-That’s a big one. That’s a big set of things. But then the other question is around
how do you create the global cooperation that’s necessary to take advantage
of the big opportunities that are ahead, and to address the big challenges, right? I don’t think it’s just
fighting crises like climate change. I think that there are
massive opportunities. -Definitely.
-Spreading prosperity, spreading more human rights and freedom. Those are things that come with trade
and connection as well. So, you want that for the upside. But, I guess, my diagnosis
at this point… I’m curious to hear your view on this. …is I actually think we’ve spent a lot
of the last 20 years with the Internet, maybe even longer working on
global trade, global information flow, making it so that people can connect. I actually think
the bigger challenge at this point is making it so that in addition
to that global framework that we have, making it so that things work
for people locally, right? Because I think there’s this dualism here
where you need both, right? If you just resort to
just kind of local tribalism, then you miss the opportunity to work
on the really important global issues. But if you have a global framework, but people feel like
it’s not working for them at home, or some set of people
feel like that’s not working, then they’re not
politically going to support the global collaboration
that needs to happen. I think there’s
the social version of this, which we talked about a little bit before
where people are now able to find communities that match
their interests more, but some people
haven’t found those communities yet and are left behind as some of the more
physical communities have receded. And some of these communities
are quite nasty also, -so we shouldn’t forget that.
-Yes. So I think they should be… Yes. Although, I would argue that people
joining kind of extreme communities is largely a result of not having healthier communities and not having healthy economic progress
for individuals. I think most people,
when they feel good about their lives, they don’t seek out extreme communities. So there’s a lot of work that I think we as an Internet platform provider
need to do to lock that down even further. But I actually think creating prosperity is probably one of the better ways,
at a macro level, to go at that. But I guess… But maybe just stop there a little. People that feel good about themselves have done some of the most terrible things
in human history. I mean, we shouldn’t confuse
people feeling good about themselves and about their lives with people being benevolent
and kind and so forth. And also, they wouldn’t say
that their ideas are extreme. And we have so many examples
throughout human history, from the Roman Empire to slave trade
in the modern age and colonialism, that people,
that they had a very good life, they had a very good family life
and social life, they were nice people, I mean, I guess most Nazi voters
were also nice people. If you meet them for a cup of coffee
and you talk about your kids, they were nice people, and they think good things
about themselves, and some of them
can have very happy lives. And even the ideas that we look back, and say, “This was terrible.
This was extreme,” they didn’t think so. Again, if you just think
about colonialism… Well, but World War II, that came through a period of
intense economic and social disruption after the Industrial Revolution… Let’s put aside the extreme example. Let’s just think about European
colonialism in the 19th century. So people say in Britain,
in the late 19th century, they had the best life in the world
at the time. And they didn’t suffer
from an economic crisis or disintegration of society
or anything like that. And they thought that by going
all over the world and conquering and changing societies
in India, in Africa, in Australia, they were bringing lots of good
to the world. And I’m just saying that
so that we are more careful about not confusing the good feelings
people have about their life… It’s not just miserable people suffering
from poverty and economic crisis. Well, I think that there’s a difference
between the example that you’re using of a wealthy society going and colonizing or doing different things
that had different negative effects. That wasn’t the fringe in that society. I guess, what I was
more reacting to before was your point
about people becoming extremists. I would argue that, in those societies, that wasn’t those people
becoming extremists. You can have a long debate
about any part of history and whether the direction that a society
chose to take is positive or negative and the ramifications of that. But I think today
we have a specific issue, which is that more people are seeking out solutions at the extremes, and I think a lot of that is
because of a feeling of dislocation, both economic and social. So that… I think that there’s a lot of ways
that you’d go at that. And I think part of it, as someone who is running
one of the Internet platforms, I think we have a special responsibility to make sure that our systems
aren’t encouraging that. But I think, broadly,
the more macro solution for this is to make sure that people feel like
they have that grounding and that sense of purpose and community and that their lives are…
And that they have opportunity. And I think that, statistically,
what we see, and sociologically, is that when people have
those opportunities, they don’t, on balance, as much,
seek out those kind of groups. And I think that there’s
the social version of this, there’s also the economic version. This is the basic story
of globalization. On the one hand,
it’s been extremely positive for bringing a lot of people
into the global economy. Where people in India and Southeast Asia
and across Africa who wouldn’t have previously had access to a lot of jobs
in the global economy now do. And there’s been probably the greatest…
At a global level, inequality is way down. Because hundreds of millions of people
have come out of poverty, and that’s been positive. But the big issue has been that
in developed countries, there have been a large number of people who are now competing with all these
other people who are joining the economy, and jobs are moving to these other places. So a lot of people have lost jobs. For some of the people
who haven’t lost jobs, there’s now more competition
for those jobs for people internationally, so their wages, that’s one of the factors
I would… The analyses have shown
that’s preventing more wage growth. And there are
five to 10% of people, according to a lot of the analyses
that I’ve shown, who are actually, in absolute terms,
worse off because of globalization. Now that doesn’t necessarily mean
that globalization for the whole world is negative. I think, in general,
it’s been, on balance, positive. But the story we’ve told about it
has probably been too optimistic in that we’ve only talked
about the positives and how it’s good as this global movement to bring people out of poverty
and create more opportunities. And the reality, I think, has been
that it’s been net very positive, but if there are five or 10%
of people in the world who are worse off, seven billion people in the world, so that’s many hundreds
of millions of people, the majority of whom are likely in the most developed countries,
in the US and across Europe, that’s going to create a lot of
political pressure in those countries. So in order to have
a global system that works, it feels like you need it to work
at the global level, but then you also need individuals and
each of the member nations in that system to feel like it’s working for them, too,
and that recurses all the way down. So in local cities and communities, people need to feel like
it’s working for them, both economically and socially. So I guess at this point,
the thing that I worry about, and I’ve rotated a lot of Facebook’s
energy to try to focus on this, is our mission used to be
connecting the world. Now it’s about helping people
build communities and bringing people closer together. And a lot of that is because I actually
think that the thing that we need to do to support more global connection
at this point is making sure
that things work for people locally. You know, in a lot of ways,
we’ve made it so that the Internet… So that an emerging creator can… But how do you balance working it locally
for people in the American Midwest and at the same time working it better for people in Mexico,
South America, or Africa? Part of the imbalance is that when people
in Middle America are angry, everybody pays attention, because they have their finger
on the button. But if people in Mexico
or people in Zambia feel angry, we care far less,
because they have far less power. The pain, and I’m not saying the pain
is not real. The pain is definitely real. But the pain of somebody in Indiana
reverberates around the world far more than the pain of somebody
in Honduras or in the Philippines simply because of the imbalances
of the power in the world. Earlier, what we said about fragmentation, I know that Facebook faces
a lot of criticism about encouraging people, some people,
to move to these extremist groups. That’s a big problem,
but I don’t think it’s the main problem. I think, also, it’s something
that you can solve if you put enough energy into that. That is something you can solve. But this is the problem that gets
most of the attention now. What I worry more,
again, not just about Facebook, about the entire direction that the new Internet economy and the new tech economy is going towards, is increasing inequality between different parts of the world, which is not a result
of extremist ideology, but the results of a certain
economic and political model. And, secondly, undermining human agency, and undermining the basic
philosophical ideas of democracy, and the free market, and individualism. These, I would say,
are my two greatest concerns about the development of technology
like AI and machine learning. And this will continue to be
a major problem even if we find solutions to the issue of social extremism
in particular groups. Yeah, I certainly agree
that extremism isn’t… I would think about it more as a symptom and a big issue
that needs to be worked on. But I think the bigger question
is making sure that everyone has a sense of purpose, has a role that they feel matters,
and social connections. Because, at the end of the day,
we’re social animals. And I think it’s easy
in our theoretical thinking to abstract that away. But that’s such a fundamental part
of who we are. That’s why I focus on that. Do you want to move over
to some of the AI issues? Because I think that that’s… Or do you want to stick on this topic
for a second? No, this topic is closely connected to AI. Again, because I think that one
of the disservices that science fiction… I’m a huge fan of science fiction, but I think it has done
some pretty bad things, which is to focus attention
on the wrong scenarios and the wrong dangers, that people think, “AI is dangerous
because the robots are coming to kill us.” And this is extremely unlikely that we’ll face a robot rebellion. I’m much more frightened
about robots always obeying orders than about robots rebelling
against the humans. I think the two main problems with AI, and we can explore this in greater depth,
is what I mentioned. First, increasing inequality
between different parts of the world. Because you’ll have some countries
which lead and dominate the new AI economy. And this is such a huge advantage that it kind of trumps everything else. And we will see…
If we had the Industrial Revolution creating this huge gap between a few industrial powers
and everybody else, and then it took 150 years
to close the gap, and over the last few decades, the gap has been closed, or closing, as more and more countries,
which were far behind, are catching up. Now the gap may reopen
and be much worse than ever before because of the rise of AI and because AI is likely to be dominated
by just a small number of countries. So that’s one issue. AI inequality. And the other issue is AI
and human agency. Or even the meaning of human life. What happens when AI is mature enough and you have enough data to basically
hack human beings. And you have an AI that knows me
better than I know myself, and can make decisions for me, predict my choices,
manipulate my choices, and authority increasingly shifts from humans to algorithms. So, not only decisions
about which movie to see, but even decisions like
which community to join, who to befriend, whom to marry. We increasingly rely
on the recommendations of the AI and what does it do to human life and human agency? So, these I would say, are the two
most important issues of AI inequality, and AI and human agency. Yeah. And I think both of them get down to a similar question around values. And who is building this
and what are the values that are encoded, and how does that end up playing out. Yeah, I tend to think that
in a lot of the conversations around AI, we almost personify AI, right? You’re point around killer robots
or something like that. But I actually think it’s… AI is very connected
to the general tech sector, right? So, almost every technology product, and increasingly a lot of
not what you call technology products are made better in some way by AI. So, it’s not like AI is a monolithic thing
that you build, it powers a lot of products.
It’s a lot of economic progress, and it can get towards some of the distribution of opportunity
questions that you’re raising. But it also is
fundamentally interconnected with these really
socially important questions around data and privacy,
and how we want our data to be used, and what are the policies around that,
and what are the global frameworks. So, one of the big questions that… So, I tend to agree with a lot
of the questions that you’re raising, which is that a lot of the countries that have the ability
to invest in future technology of which AI and data
and future Internet technologies are certainly an important area, are doing that because it will give
their local companies an advantage in the future and to be the ones that
are exporting services around the world. I tend to think that, right now, the United States has a major advantage that a lot of the global
technology platforms are made here, and, certainly, a lot of the values
that are encoded in that are shaped largely by American values. They’re not only…
And speaking for Facebook, and we serve people around the world,
and we take that very seriously. But, certainly,
ideas like giving everyone a voice, that’s something
that is probably very shaped by the American ideas around free speech, and strong adherence to that. So, I think, culturally and economically, there is an advantage
for countries to develop, to push forward the state of the field, and have the companies that,
in the next generation, are the strongest companies in that. So, certainly, you see different countries
trying to do that. And this is very tied up in
not just economic prosperity -and equality, but also…
-Do they have a real chance? Does a country like Honduras,
Ukraine, Yemen has any real chance
of joining the AI race? Or are they… They’re already out. It’s not going to happen in Yemen,
it’s not going to happen in Honduras. And then what happens to them -in 20 years or 50 years?
-I think that some of this gets down to the values
around how it’s developed though. Right? I think that there are certain advantages that countries
with larger populations have ’cause you get to critical mass
in terms of universities, and industry, and investment, and things like that. But one of the values that we here both at Facebook, and generally the academic system
of trying to do research, hold, is that you do open research, right? So, a lot of the work
that’s getting invested into these advances, in theory, if this works well,
should be more open. So, then you can have an entrepreneur in one of these countries
that you’re talking about, which maybe isn’t
a whole industry-wide thing… Certainly, I think you’d bet
against sitting here today that, in the future, all of the AI companies
are gonna be in a given small country. But I don’t think it’s far-fetched to believe that there will be
an entrepreneur in some place who can use Amazon Web Services
to spin up instances for compute, who can hire people across the world
in a globalized economy, and can leverage research
that has been done in the US or across Europe or in different
open academic institutions or companies that increasingly
are publishing their work, that are pushing the state-of-the-art
forward on that. So, I think that there’s this big question about what we want the future
to look like. And part of the way that I think
we want the future to look is we want it to be open,
we want the research to be open. I think we want the Internet
to be a platform. And this gets back to your
unification point versus fragmentation. One of the big risks for the future is that the Internet policy
in each country ends up looking different. It ends up being fragmented. And if that’s the case,
then the entrepreneur in the countries that you’re talking about, Honduras,
probably doesn’t have as big of a chance if they can’t leverage all the advances
that are happening everywhere. But if the Internet stays one thing,
and the research stays open, then they have a much better shot. So, when I look towards the future, one of the things
that I just get very worried about is the values that I just laid out
are not values that all countries share. And when you get into
some of the more authoritarian countries and their data policies, they’re very different
from the regulatory frameworks that are across Europe
and across a lot of other people. People are talking about
or have put into place. Just to put a finer point on that,
recently I’ve come out and I’ve been very vocal that I think that more countries should adopt
a privacy framework like GDPR in Europe. And a lot of people, I think,
have been confused about this. “Why are you arguing
for more privacy regulation?” You know, “Why now,
given that in the past, “you weren’t as positive on it?” And I think part of the reason
why I am so focused on this now is I think, at this point,
people around the world recognize that these questions around data,
and AI and technology are important. So there’s going to be a framework
in every country. I mean, it’s not like
there’s not gonna be regulation or policy. So I actually think
the bigger question is, “What is it going to be?” And the most likely alternative
to each country adopting something that encodes the freedoms and rights
of something like GDPR… In my mind, the most likely alternative is the authoritarian model, which is currently being spread,
which says, you know, “Every company needs to store
everyone’s data locally in data centers.” And if I’m a government,
I should be able to, you know, go send my military there and be able
to get access to whatever data I want. I’d be able to take that
for surveillance or military or helping, you know,
local military industrial companies. And I just think
that that’s a really bad future. And that’s not the direction that I, as someone who’s building
one of these Internet services or just as a citizen of the world,
want to see the world going. To be the devil’s advocate for a moment, I mean, if I look at it
from the viewpoint that of India. So, I listen to the American president
saying, “America first. “And I’m a nationalist,
I’m not a globalist. “I care about the interests of America.” And I wonder, is it safe to store the data
about Indian citizens in the US, and not in India,
when they are openly saying they care only about themselves. So, why should it be in America
and not in India? Well, I think that the motives matter, and certainly,
I don’t think that either of us would consider India
to be an authoritarian country that… So I would say that… Well, it can still say, “We want data
and metadata on Indian users “to be stored on Indian soil. “We don’t want it to be stored
on American soil or somewhere else.” Yeah. And I can understand
the arguments for that, and I think
that the intent matters, right? And I think countries can come at this
with open values, and still conclude that something
like that could be helpful. But I think one of the things
that you need to be very careful about is that if you set that precedent, you’re making it very easy
for other countries that don’t have open values,
and that are much more authoritarian, and want the data
not to protect their citizens, but to be able to surveil them
and find dissidents and lock them up. That… So I think, one of… I agree. I mean, it really boils
down to the question that, “Do we trust America?” And given the past two or three years, people in more and more places
around the world… I mean, previously, say, if we were sitting here ten years ago,
20 years ago or 40 years ago, when America declared itself
to be the leader of the free world, we can argue a lot
whether this was the case or not. Or, at least, on the declaratory level, this was how America presented itself
to the world. “We are the leaders of the free world,
so trust us. We care about freedom.” But now we see a different America. America which doesn’t want even to be… Again, it’s not a question
of even what they do, but how America presents itself
no longer as the leader of the free world. But as a country which is interested,
above all, in itself and in its own interests. And just this morning,
for instance, I read that the US is considering having a veto
on the UN resolution against using sexual violence
as a weapon of war. And the US is the one
that thinks of vetoing this. And as somebody
who is not a citizen of the US, I ask myself, “Can I still trust America
to be the leader of the free world?” if America itself says,
“I don’t want this role anymore?” Well, I think that
that’s a somewhat separate question from the direction
that the Internet goes in. Because, I mean, GDPR,
the framework that I’m advocating, that it would be better if more countries
adopted something like this, because I think that that’s just significantly better
than the alternatives, a lot of which are these
more authoritarian models. -I mean, GDPR originated in Europe, right?
-Yeah. So it’s not an American invention. And I think, in general,
these values of openness and research, of cross-border flow of ideas and trade, that’s not an American idea, right? I mean, that’s a global philosophy
for how the world should work. And I think that the alternatives to that
are, at best, fragmentation, which breaks down
the global model on this. At worst, a growth in authoritarianism
for the models of how this gets adopted. And that’s where I think that the precedents on some of this stuff
get really tricky. I mean, I think you’re doing a good job of playing devil’s advocate
in the conversation because you’re bringing
all of the counterarguments that I think someone with good intent
might bring to argue. “Hey. Maybe a different set
of data policies “is something that we should consider.” The thing that I just worry about is that, what we’ve seen is that
once a country puts that in place, that’s a precedent
that, then, a lot of other countries that might be more authoritarian use to basically be a precedent to argue
that they should do the same things. And then that spreads. And I think that that’s bad, right? And that’s one of the things that, as the person running this company, I’m quite committed to making sure that we play our part
in pushing back on that and keeping the Internet as one platform. So, I mean,
one of the most important decisions that I think I get to make,
as the person running this company is, “Where are we going to build
our data centers and store data?” And we’ve made the decision
that we’re not going to put data centers in countries that we think
have weak rule of law, where people’s data
may be improperly accessed, and that could put people in harm’s way. And, you know, I mean, a lot has been… There have been
a lot of questions around the world around questions of censorship. And I think that those are
really serious and important. I mean, a lot of the reason
why I build what we build is because I care
about giving everyone a voice, giving people as much voice as possible. I don’t want people to be censored. At some level, these questions
around data and how it’s used, and whether authoritarian governments
get access to it, I think, are even more sensitive because if you can’t say something
that you want, that is highly problematic,
that violates your human rights. I think, in a lot cases,
it stops progress. But if a government
can get access to your data, then it can identify who you are
and go lock you up, and hurt you and hurt your family,
and cause real physical harm in ways that are just really deep. So, I do think
that people running these companies have an obligation
to try to push back on that, and fight establishing precedents,
which will be harmful, even if a lot of the initial countries
that are talking about some of this have good intent. I think that this can
easily go off the rails. And when you talk about, in the future, AI and data, which are two concepts
that are just really tied together, I just think
the values that that comes from, whether it’s part of a more global system, a more democratic process
and a more open process, that’s one of our best hopes
for having this work out well. If it comes from repressive
or authoritarian countries, then I just think that it’s gonna be
highly problematic in a lot of ways. That raises the question of, “How do we build AI in such a way “that it’s not inherently
a tool of surveillance, “and manipulation and control?” I mean, this goes back to the idea
of creating something that knows you better
than you know yourself. Which is kind of the ultimate surveillance
and control tool. And we are building it now
in different places around the world. It’s being built. And what are your thoughts
about how to build an AI, which serves individual people
and protects individual people, and not an AI, which can easily,
with a flip of a switch, become the ultimate surveillance tool? Well, I think that that is more about
the values and the policy framework than the technological development. I mean, a lot of the research
that’s happening in AI are just very fundamental,
mathematical methods where a researcher will create an advance, and now, all of the neural networks
will be 3% more efficient. -I’m just throwing this out.
-Yeah. Yeah. And that means that news feed
will be a little bit better for people. Our systems for detecting things
like hate speech will be better. Our ability to find photos of you
that you want to review will be better. All these systems get a little better. Now, I think the bigger question
is you have places in the world where governments are choosing
to use that technology and those advances for things like widespread
face recognition and surveillance. And those countries,
I mean China’s doing this, they create a real feedback loop which
advances the state of that technology, where they say, “Okay, we wanna do this.” So now there’s a set of companies
sanctioned to go do that and they’re getting access
to a lot of data to do it because it’s allowed and encouraged, so that is advancing
and getting better and better, that’s not a mathematical process, that’s a policy process,
that they wanna go in that direction, those are their values,
and it’s an economic process of the feedback loop
and development of those things, compared to in countries that might say, “Hey, that kind of surveillance
isn’t what we want.” Those companies just don’t exist as much
or don’t get as much support. I don’t know.
In my home country of Israel, at least for Jews it’s a democracy and it’s one of the leaders of the world
in surveillance technology and we basically have
one of the biggest laboratories of surveillance technology in the world,
which is de-occupied territories. And exactly these kinds of systems
are being developed there and exported all over the world. So, given my personal experience
back home, I don’t necessarily trust
that just because a society in its own inner workings
is say, democratic, that it will not develop and spread
these kinds of technologies. Yeah, I agree. It’s not clear
that a democratic process alone solves it, but I do think that it is
mostly a policy question. A government can quite easily
make the decision that they don’t wanna support
that kind of surveillance and then the companies
they would be working with to support that kind of surveillance
would be out of business. Or at the very least,
have much less economic incentive to continue that technological progress, so that dimension
of the growth of the technology gets stunted compared to others and that’s generally the process
that I think you wanna follow broadly. Technological advance
isn’t by itself good or bad. I think it’s the job of the people
who are shepherding it, building it and making policies around it
to have policies and make sure that their effort
goes towards amplifying the good and mitigating the negative use cases. And that’s how I think you end up bending
these industries and technologies to be things that are positive
for humanity overall and I think that’s a normal process that happens with most technologies
that get built. But I think what we’re seeing
in some of these places is not the natural mitigation
of negative uses, in some cases the economic feedback loop
is pushing those things forward, but I don’t think it has to be that way, that’s not as much a technological
decision as it is a policy decision. I fully agree, but every technology can be used
in different ways for good or for bad, you can use the radio
to broadcast music to people and you can use the radio to broadcast Hitler giving a speech
to millions of Germans, the radio doesn’t care, the radio just
carries whatever you put in it. So, yeah, it is a policy decision,
but then it raises the question, “How do we make sure
that the policies are the right policies “in a world where it is becoming
more and more easy to manipulate “and control people on a massive scale
like never before?” I mean new technology, it’s not just
that we invent the technology and then we have good democratic countries
and bad authoritarian countries, and the question is,
“What would they do with the technology?” The technology itself
could change the balance of power between democratic
and totalitarian systems and I fear that new technologies
are giving an inherent advantage, not necessarily overwhelming, but they do tend to give an
inherent advantage to totalitarian regimes because the biggest problem of
totalitarian regimes in the 20th century, which eventually led to their downfall, is that they couldn’t process information
efficiently enough. If you think about the Soviet Union, so you have this
information processing model, which basically says, “We take all the`
information from the entire country, “move it to one place, to Moscow,
there it gets processed, “decisions are made in one place
and transmit it back as commands.” This was the Soviet model
of information processing. Versus the American version,
which was, “No, we don’t have a single center.
We have a lot of organizations “and a lot of individuals and businesses
and they can make their own decisions.” In the Soviet Union
there’s somebody in Moscow. If I live in some small farm
or kolkhoz in Ukraine, there’s somebody in Moscow who tells me how many radishes to grow this year
because they know. And in America I decide for myself, I get signals from the market
and I decide. And the Soviet model just didn’t work well because of the difficulty of processing
so much information quickly with 1950s technology. And this is one of the main reasons why the Soviet Union
lost the Cold War to the United States. But with new technology it suddenly,
it might become, it’s not certain, but one of my fears is that new technology suddenly makes
central information processing far more efficient than ever before and far more efficient
than distributed data processing. Because the more data
you have in one place the better your algorithms
and so on and so forth. And this kind of tilts the balance
between totalitarianism and democracy in favor of totalitarianism. And I wonder
what are your thoughts on this issue. -Well, I’m more optimistic about…
-I guessed so. About democracy in this. I think the way
that the democratic process needs to work is people start talking
about these problems and then even if it seems
like it starts slowly in terms of people caring
about data issues and technology policy, ’cause it’s a lot harder
to get everyone to care about it than it is just a small number
of decision makers, so I think that the history of democracy
versus more totalitarian systems is it always seems like the totalitarian
systems are gonna be more efficient and the democracies
are just gonna get left behind, but smart people start discussing
these issues and caring about them and I do think we see that people do now
care much more about their own privacy, about data issues,
about the technology industry, people are becoming
more sophisticated about this, they realize
that having a lot of your data stored can both be an asset
because it can help provide a lot of benefits and services to you, but increasingly
maybe it’s also a liability because there are hackers
and nation states who might be able to break in and use that data against you
or exploit it or reveal it. So maybe people don’t want their data
to be stored forever, maybe they want it to be reduced
in permanence, maybe they want it all to be end-to-end
encrypted as much as possible in their private communications,
people really care about this stuff in a way that they didn’t before and that’s certainly grown a lot
over the last several years. So that conversation
is the normal democratic process. And I think what’s gonna end up happening is that by the time you get people
broadly aware of the issues and on-board, that is just
a much more powerful approach, where then you do have people
in a decentralized system who are capable of making decisions,
who are smart, who I think will generally
always do it better than too centralized of an approach. And here is again a place
where I worry that personifying AI and saying AI is a thing
that an institution will develop and it’s almost like a sentient being, I think mischaracterizes
what it actually is. It’s a set of methods
that make everything better. Sorry, let me retract that.
That’s way too broad. It’s a lot of technological processes
more efficient. -And I think that’s…
-But that’s the worry. It also makes… But that’s not just for centralized folks. In our context, so we build,
our business has this ad platform. And a lot of the way
that that can be used now is we have 90 million small business
that use our tools. And now,
because of this access to technology they have access to the same tools to do advertising, marketing,
reach new customers and grow jobs that previously only the big companies
would have had. And that’s a big advance.
That’s a massive decentralization. When people talk about our company
and the Internet platforms overall, they talk about how there’s
a small number of companies that are big, and that’s true, but the flipside of it is that now there are billions of people
around the world who have a voice, that they can share information
more broadly, and that’s actually
a massive decentralization in power, and kind of returning power to people. Similarly, people have access
to more information, have access to more commerce. That’s all positive. So, I don’t know. I’m an optimist on this.
I think we have real work cut out for us. And I think that
the challenges that you raise are the right ones to be thinking about, because if we get it wrong,
that’s the way in which it will go wrong. But, I don’t know. I think that the historical precedent
would say that it all points… You know, where there was the competition between the US and Japan
in the ’80s and the ’70s, or the Cold War before that,
or different other times, people always thought
that the democratic model, which is slow to mobilize,
but very strong once it does. And once people get bought into
a direction and understand the issue, I do think that that will continue to be the best way to spread prosperity
around the world and make progress in a way
that meets people’s needs. And that’s why,
when you’re talking about Internet policy, when you’re talking about economic policy, I think spreading regulatory frameworks
that encode those values, I think is one of the most
important things that we can do. But it starts
with raising the issues that you are and having people be aware
of the potential problems. I agree that in the last few decades,
it was the case. That open democratic systems were better and more efficient. Again, one of my fears is that
it might have made us a bit complacent. Because we assume
that this is a kind of a law of nature. That distributed systems are always better
and more efficient than centralized systems. And we lived, we grew up in a world
in which there was kind of this… To do the good thing morally was also to do the efficient thing
economically and politically. And a lot of countries
liberalized their economy, their society, their politics,
over the last 50 years more because they were convinced
of the efficiency argument than of the deep moral argument. And what happens if efficiency
and morality suddenly split? Which has happened before in history. I mean, the last 50 years
are not representative of all of history. We had many cases before,
in human history, in which repressive centralized systems
were more efficient, and therefore,
you got these repressive empires. And there is no law of nature
which says that this cannot happen again. Again, my fear is that the new technology
might tilt that balance. And just by making central data processing
far more efficient, it could give a boost
to totalitarian regimes. Also, in the balance of power
between the center and the individual, that for most of history, the central authority
could not really know you personally. Simply because of the inability
to gather and process information. So there were some people
who know you very well, but usually, their interests
were aligned with yours. Like, my mother knows me very well, but most of the time
I can trust my mother. But now, we are reaching the point when some system far away
can know me better than my mother, and the interests
are not necessarily aligned. Now, yes, we can use that also for good, but I’m pointing out
that this is a kind of power that never existed before. And it could empower totalitarian
and authoritarian regimes to do things that were simply
technically impossible until today. Yeah. And if you live in an open democracy, okay, you can rely on all kinds
of mechanisms to protect yourself. But thinking
more globally about this issue, I think a key question is,
how do you protect human attention from being hijacked by malevolent players who know you
better than you know yourself? Who know you
better than your mother knows you? And this is a question
that we never had to face before. Because we never had… Usually, the malevolent players
just didn’t know me very well. Yeah, okay, so there’s a lot
in what you were just talking about. I mean, I think… In general, one of the things that… I do think that there’s a scale effect. Where one of the best things
that we could do if we care about these open values and having a globally connected world… I think making sure that the critical mass
of the investment in new technologies encodes those values is really important. So that’s one of the reasons
why I care a lot about not supporting the spread of authoritarian policies
to more countries. Either inadvertently doing that, or setting precedents
that enable that to happen Because I think that
the more development that happens in the way that is more open,
where the research is more open, where people have the… where the policy-making around it
is more democratic, I think that that’s gonna be positive. So I think that maintaining that balance
ends up being really important. And one of the reasons why
I think democratic countries, over time, tend to do better
on serving what people want is because there’s no metric
to optimize a society. Right, when you talk about efficiency, a lot of what people are talking about
is economic efficiency. Yeah. Are we increasing GDP?
Are we increasing jobs? Are we decreasing poverty?
Those things are all good, But, I think,
part of what the democratic process does is people get to decide on their own, which of the dimensions in society
matter the most to them in their lives. But if you can hijack people’s attention,
and manipulate them, then people deciding on their own
just doesn’t help. Because I don’t realize
that somebody manipulated me to think that this is what I want. And we are reaching the point
when for the first time in history, you can do that on a massive scale. Again, I speak a lot about the issue
of free will in this regard. And the people
that are easiest to manipulate are the people who believe in free will, and will simply identify with whatever
thought or desire pops up in their mind because they cannot even imagine that this desire
is not a result of my free will, this desire is the result
of some external manipulation. Now, it may sound paranoid. And for most of history,
it was probably paranoid because nobody had this kind of ability
to do it on a massive scale. But here, like, in Silicon Valley, the tools to do that on a massive scale have been developed
over the last few decades. And they may have been developed
with the best intentions. Some of them may have been developed
with the intention of just selling stuff to people,
and selling products to people. But now the same tools that can be used
to sell me something I don’t really need, can now be used to sell me a politician
I really don’t need. Or an ideology that I really don’t need. It’s the same tool.
It’s the same hacking the human animal, and manipulating what’s happening inside. Yeah, okay,
so there’s a lot going on here. I think that there’s… When designing these systems, I think that there’s the intrinsic design,
which you want to make sure you get right, and then there’s preventing abuse. So in that, there’s two kinds of questions
that people raise. I mean, one is, we saw what the Russian government tried to do
in the 2016 elections. That’s clear abuse. We need to build really advanced systems
for detecting that kind of interference in the democratic process
and more broadly. Being able to identify that, identify when people are standing up
networks of fake accounts that are not behaving
in a way that normal people would. To be able to weed those out, and work with Law Enforcement
and Election Commissions and folks all around the world
in the Intelligence community, to be able to coordinate
and be able to deal with that effectively. So, stopping abuse is certainly important, but I would argue that the deeper question is about the intrinsic design
of the systems. -Right?
-Yeah. So not just fighting the abuse. And there, I think that… I think that the incentives
are more aligned towards a good outcome than a lot of critics might say. And here’s why. I think that there’s a difference
between what people want first order and what they want
second order over time, right? So, right now,
you might just consume a video ’cause you think it’s silly or fun. You wake up, and… Or you kind of look up an hour later
and you’ve watched a bunch of videos, and you’re like,
“What happened to my time?” So maybe in the narrow, short-term period, you consume some more content, and maybe you saw some more ads, so it seems
like it’s good for the business. But it actually really isn’t over time. Because people make decisions
based on what they find valuable. And what we find, at least in our work, is that what people really want to do
is connect with other people. It’s not just passively consume content. So, we’ve had to find and constantly
adjust our systems over time to make sure that we’re rebalancing it, so that way
you’re interacting with people, so that way we make sure that we don’t just measure
signals in the system like what you’re clicking on, because that can get you
into a bad local optimum. But instead, we bring in real people
to tell us their real experiences. In words, not just filling out scores. But also telling us what the most
meaningful experiences they had today, what content was most important, what interaction did you have with
a friend that mattered to you the most and was that connected
to something that we did? And if not,
then we go and try to do the work to figure out how we can facilitate that. And what we find is that… Yeah, in the near term, maybe showing some people some more
viral videos might increase time, right? But over the long term, it doesn’t. It’s not actually aligned
with our business interest or the long term social interest. So, in strategy terms,
that would be a stupid thing to do. And I think a lot of people think that businesses
are very short-term oriented, and that businesses only care
about the next-quarter profit. But I think that most businesses that get
run well, that’s just not the case. And I think, last year,
on one of our Earnings calls, I told investors that we’d actually reduce
the amount of video-watching that quarter by 50 million hours a day. Because we wanted to take down the amount
of viral videos that people were seeing because we thought that that was displacing
more meaningful interactions that people were having with other people, which in the near term might have
a short term impact on the business for that quarter,
but over the long term it would be more positive, both for how people feel abut the products
and for the business, and… One of the patterns that I think,
has actually been quite inspiring or a cause of optimism
in running a business is that often times you make decisions you think
are gonna pay off long down the road. You think, “I’m doing
the right thing long term, “but it’s gonna hurt for a while.” And I almost always find the long term
comes sooner than you think. And that when you make these decisions that they’re maybe taking
some pain in the near term in order to get to what will be
a better case down the line, that better case, maybe you think
it’ll take five years, but actually it ends up coming in a year.
Right? And… I think people at some deep level
know when something is good. And I guess this gets back
to the democratic values because at some level, I trust that people have a sense
of what they actually care about. Maybe that if we were
showing more viral videos, maybe that would be better than the
alternatives they have to do right now. Maybe that’s better than what’s on TV,
at least they’re personalized videos. Maybe it’s better than YouTube, we have better content
or whatever the reason is. But I think you could still make the service
better over time for actually matching what people want. If you do that,
that is better for everyone. I think that the intrinsic design
of these systems is quite aligned with serving people
in a way that is pro-social. That’s certainly what I care about
in running this company, is to get there. I think this is like the rock bottom. This is the most important issue
that, ultimately, what I’m hearing from you
and from many other people when I have these discussions,
is ultimately, the customer is always right. The voter knows best.
People know what is good for them. People make a choice. If they choose to do it, then it’s good. That has been the bedrock of, at least, western democracies
for centuries, for generations. And this is now where
the big question mark is. Is it still true, in a world
where we have the technology to hack human beings
and manipulate them like never before that the customer is always right?
That the voter knows best? Or have we gone past this point? And the simple ultimate answer that well, “This is what people want
and they know what’s good for them,” maybe it’s no longer the case. Well… I think… It’s not clear to me that
that has changed, but that’s a very deep question
about democracy… -This is the deepest…
-I don’t think that’s a new question. People have always… The question isn’t new,
the technology is new. I mean, if you lived
in 19th century America, and you didn’t have these
extremely powerful tools to decipher and influence people -then it was a different… Okay.
-Let me frame this a different way. For all the talk around,
is democracy being hurt by the current set of tools,
and the media, and all this, I think that there’s an argument the world is more democratic now
than it was in the past. The country was set up as… The US was set as a Republic. So, a lot of the foundational rules
limited the power of a lot of individuals being able to vote, and have a voice, and checked the popular will
in a lot of different stages. Everything from the way that laws get
written by Congress and not by people… Everything to the Electoral College,
which a lot of people think today is undemocratic,
but it was put in place because of a set of values that a
Democratic Republic would be better. I actually think what has happened today,
is that increasingly more people are enfranchised
and more people have a voice, more people are getting to vote. Increasingly people have a voice,
more people have access to information. And I think a lot of what
people are asking is, “Is that good?” It’s not necessarily the question of,
“The democratic process has been the same, “but now the technology is different.” I think the technology’s made it,
so individuals are more empowered and part of the question is,
“Is that the world that we want?” This is a scenario where…
All of these things are with challenges. Right? And often progress
causes a lot of issues. And it’s a really hard thing
to reason through while we’re trying to make progress,
and help all these people join the global economy, or help people join the communities, and have the social lives
that they would want, and be accepted in different ways. But it comes with this dislocation
in the near term and that’s a massive dislocation
so that seems really painful. But I actually think that you can
make a case that we are at, and continue to be at,
the most democratic time. And I think that overall,
in the history of our country, at least, when we’ve gotten
more people to have the vote and we’ve gotten more representation, and we’ve made it so people have access
to more information, and more people
can share their experiences, I do think that that’s made
the country stronger, and has… And it’s helped progress. And it’s not that the stuff
is without issues. It has massive issues.
But that’s the pattern that I see and why I’m optimistic
about a lot of the work. I agree that more people have more voice
than ever before, both in the US and globally. I think you’re absolutely right. My concern is,
to what extent we can trust the voice of people…
To what extent I can trust my voice? Like, I’m…
We have this picture of the world that I have this voice inside me, which tells me
what is right and what is wrong. And the more I’m able to express
this voice in the outside world and influence what’s happening,
the more people can express their voices, it’s better, it’s more democratic. But what happens if at the same time
that more people can express their voices, it’s also easier to manipulate
your inner voice? To what extent you can really trust that the thought that just popped up
in your mind is the result of some free will, and not the result of an extremely
powerful algorithm that understands
what’s happening inside you and knows how to push the buttons
and press the levers, and is serving some external entity and it has planted this thought
or this desire that you now express? So, it’s two different issues.
Giving people voice and trusting… Again, I’m not saying, “I know everything, “but all these people
that now join the conversation, “we cannot trust their voices.” I’m asking this about myself, to what extent I can trust
my own inner voice? And, you know, I spend two hours
meditating every day. And I go on these
long meditation retreats. And my main takeaway from that
is it’s craziness inside there. And it’s so complicated. And the simple, naive belief that the thought that pops up in my mind,
this is my free will, this was never the case. But, if say, a thousand years ago,
the battles inside were mostly between, you know,
neurons and biochemicals and childhood memories
and all that, increasingly, you have external actors going under your skin, and into
your brain, and into your mind. And how do I trust that my amygdala
in not a Russian agent now? How do I know…
The more we understand about the extremely complex world inside us, the less easy it is to simply trust what this inner voice
is telling, is saying. Yeah, I understand the point
that you’re making. As one of the people who’s running
a company that develops ranking systems to try to help show people content that’s gonna be
interesting to them… There’s a dissonance between the way that you’re explaining
what you think is possible and what I see as a
practitioner building this, I think. You can build systems that can get
good at a very specific thing. Helping to understand which of your
friends you care the most about so you can rank their content higher
on newsfeed. But the idea that
there’s some kind of generalized AI, that’s a monolithic thing that understands
all dimensions of who you are in a way that’s deeper than you do, I think doesn’t exist, and is probably
quite far off from existing. So, there’s certainly abuse of the systems
that I think needs to be… That I think is more of a
policy and values question, which is… On Facebook, you’re supposed to be
your real identity, so if you have, to use your example, Russian agents or folks
from the government, the IRA who are posing as someone else
and saying something and you see that content, but you think
it’s coming from someone else, then that’s not an algorithm issue. I mean, that’s someone abusing the system, and taking advantage of the fact
that you trust that on this platform, someone who’s generally gonna be
who they say they are, so you can trust that the information
is coming from some place and kinda slipping
in the back door that way, and that’s the thing
that we certainly need to go fight. But, I don’t know, as a broad matter, I do think there’s this question of,
“To what degree are the systems…” This kinda brings it full circle
to where we started on, is it fragmentation,
or is it personalization? Is the content that you see… if it resonates, is that because it actually
just more matches your interests, or is it because you’re being incepted and convinced of something
that you don’t believe and is dissonant with your interests
and your beliefs and certainly, all the psychological
research that I’ve seen and the experience that we’ve had, is that when people see things
that don’t match what they believe, they just ignore it. Right, so, certainly, there can be an evolution that happens
where a system shows information that you’re gonna be interested in and if that’s not managed well, that has the risk of
pushing you down a path towards adopting a more extreme position, or evolving
the way you think about it over time. But I think most of the content, it resonates with people because
it resonates with their lived experience and to the extent
that people are abusing that and either trying to represent
that they’re someone who they’re not, or trying to take advantage of a bug
in human psychology where we might be more prone
to an extremist idea, that’s our job
in either policing the platform working with governments
and different agencies and making sure we design our systems
and our recommendations systems to not be promoting things that people might engage with
in the near term, but over the long term, will regret
and resent us for having done that. And I think it’s in our interest
to get that right. And for a while,
I think we didn’t understand the depth of some of the problems and challenges
that we faced there and there’s certainly a lot more to do, and when you’re up against nation states,
they’re very sophisticated. They’re gonna keep evolving their tactics. But the thing
that I think is really important is that the fundamental design of a system,
I do think, in our incentives, are aligned with helping people connect to the people they want,
have meaningful interactions, not just getting people
to watch a bunch of content that they’re gonna resent later
that they did that and certainly not making people
have more extreme or negative viewpoints than what they actually believe, so. Maybe I can try
and summarize my view in that. We have two distinct dangers coming out
of the same technological tools. We have the easier danger to grasp which is of extreme totalitarian regimes
of a kind we haven’t seen before and this could happen in different… Maybe not in the US,
but in other countries… That these tools… You say that these are abuses, but in some countries,
this could become the norm. That you’re living
from the moment you’re born in this system
that constantly monitors and surveils you and constantly manipulates you
from a very early age to adopt particular ideas, views,
habits, so forth in a way which was never possible before. And this is like the full fledged
totalitarian dystopia, which could be so effective
that people would not even resent it because they would be completely aligned with the values or the ideals of the… It’s not 1984 where you need to
torture people all the time. No. If you have agents inside their brain, you don’t need the external secret police. So that’s one danger. It’s like
the full-fledged totalitarianism. Then in places like the US, the more immediate danger
or problem to think about is what is increasingly people refer to
as surveillance capitalism. That you have this system
that constantly interact with you and come to know you and it’s all supposedly
in your best interests. To give you better recommendations
and better advice. So it starts with recommendations
for which movie to watch and where to go on vacation, but as the system becomes better, it gives a recommendation on what
to study at college, where to work, ultimately, whom to marry,
who to vote for, which religion to join,
like, join a community. You have all these religious communities, “This is the best religion for you. “For your type of personality, “Judaism, nah, it won’t work for you. “Go with Zen Buddhism. “It’s a much better fit
for your personality. “You will thank us. “In five years,
you will look back and say, “‘This was an amazing recommendation.
Thank you. I so much enjoy Zen Buddhism.'” And again, people will feel that this is aligned
with their own best interests and the system improves over time. Yeah, there will be glitches.
Not everybody will be happy all the time, but what does it mean that all the most important
decisions in my life are being taken by an external algorithm? What does it mean
in terms of human agency, in terms of the meaning of life? For thousands of years, humans tended to view life
as a drama of decision-making. Life is your… It’s a journey, you reach an intersection
after intersection and you need to choose
some decisions are small, like what to eat for breakfast and some decisions are really big
like whom to marry. And almost all of art
and all of religion is all about that. Whether it’s a Shakespeare tragedy,
or a Hollywood comedy, it’s about the hero or heroine
needing to make a big decision. To be or not to be. To marry “X” or to marry “Y.” And what does it mean to live in a world in which increasingly, we rely on the recommendations
of algorithms to make these decisions untill we reach a point when we simply follow them all the time
or most of the time? And they make good recommendations.
I’m not saying that this is some abused… No, they’re good recommendations. We don’t have a model for understanding what is the meaning
of human life in such a situation. I think the biggest objection
that I’d have to both of the ideas that you just raised is that we have access to a lot of
different sources of information, a lot of people to talk to
about different things. And it’s not just like
there’s one set of recommendations, or a single recommendation
that gets to dominate what we do and that gets to be overwhelming either in the totalitarian or the capitalist model
of what you were saying. To the contrary, I think people really don’t like,
and are very distrustful when they feel like
they’re being told what to do, or just have a single option. One of the big questions
that we’ve studied is, “How do we address when there’s a hoax,
or clear misinformation?” And the most obvious thing
that would seem like you’d do intuitively, is tell people, “Hey, this seems like it’s wrong.
Here is the other point of view “that is right.” Or at least if it’s a polarized thing, even if it’s not clear
what’s wrong and what’s right, here’s the other point of view on any given issue. And that really doesn’t work. What ends up happening is if
you tell people that something is false, but they believe it,
then they just end up not trusting you. So that ends up not working. If you frame two things as opposites… If you say, “Okay, you’re a person who
doesn’t believe, and you’re seeing content “about not believing in climate change. “So I’m gonna show you
the other perspective. “Here’s something that argues
that climate change is a thing.” That actually just entrenches you further because someone’s trying to control and… So what ends up working
sociologically and psychologically, the thing that ends up
actually being effective is giving people a range of choices. So if you show,
not, “Here’s the other opinion” with a judgement on the piece of content
that a person is engaged with, but instead you show a series of
related articles, related content, then people can work out for themselves, “Hey, here’s the range
of different opinions. “Or things that exist on this topic. “Maybe I lean in one direction
or the other, “but I’m gonna work out for myself
where I wanna be.” Most people don’t choose
the most extreme thing. And people end up feeling like they’re
informed and can make a good decision. So, at the end of the day,
I think that that’s the architecture and the responsibility that we have is to
make sure that the work that we’re doing gives people more choices,
that it’s not a given single opinion that can dominate anyone’s thinking, but where you can connect to
hundreds of different friends and even if most of your friends share
your religion or your political ideology, you’re probably gonna have
five or 10% of friends who come from a different background
who have different ideas and at least that’s getting in is well, so
you’re getting a broader range of views. So I think these are really
important questions and it’s not like there’s an answer that’s going to fully solve it
one way or another. Definitely not it. But these are the right things
to talk through. We’ve been going for 90 minutes
so we probably should wrap up. But I think we have
a lot of material to cover in the next one of these
that we’ll hopefully get to do at some point in the future and thank you so much for coming
and joining and doing this. This has been a really interesting series
of important topics to discuss. Thank you for hosting me and being open
about these very difficult questions, which I know that you, being the head
of a global corporation… I can just sit here and speak
about whatever I want, but you have many responsibilities
on your head, so I appreciate that you’re putting
yourself on the firing line and dealing with these questions. -Thanks. All right.
-Thank you.

Author: Kennedi Daugherty

Leave a Reply

Your email address will not be published. Required fields are marked *