The Future of Happiness: How Communication Technologies Will Change Our World—Or Not

It’s now my pleasure to
introduce tonight’s panelists and I’m actually
just going to list their names so we can
get on and hear from them rather than the long and
amazing CVs that they all have, although the names,
as you will discover, are quite long in and
of them themselves. So, Laura Kubzansky here,
the Lee Kum Kee Professor of Social and
Behavioral Sciences and co-director of the
Lee Kum Sheung Center for Health and Happiness at
the Harvard TH Chan School of Public Health,
JP Onnela, associate professor of biostatistics in
the Department of Biostatistics at the Harvard TH Chan
School of Public Health. And then at the end,
Jonathan Zittrain, who’s the first person I’ve ever
known attached to four schools here at Harvard– George Bemis Professor of
International Law at the John F. Kennedy School of
Government, professor of computer science at
the John A. Paulson School of Engineering and
Applied Sciences, vice dean of library
and information resources at the
Harvard Law School, and faculty co-director at
the Berkman Klein Center for Internet and Society
as part of the Faculty of Arts and Sciences. I’m still working on
the dental school. [LAUGHTER] And last but definitely
not least, Vish Viswanath, Lee Kum Kee Professor
of Health Communication and co-director of the
Lee Kum Sheung Center for Health and Happiness
at the Harvard TH Chan School of Public Health, and
he will now introduce the topic and moderate our discussion. Vish, thank you. So thank you, Jane. Thanks for– let me thank
the Harvard Museums and Jane and her team for inviting us to
this panel and offering other pebbles of wisdom– mine will be Pebbles, Laura’s
will be pearls of wisdom– to talk about happiness. Happiness is a
pretty loaded word and I prefer to have
used something else, but we didn’t have
much of choice, right? So let me start by introducing
the Center and why we are here. And I do have some
notes so that you have an illusion that I took
the assignment seriously and in order to deserve
that big check that Jane’s going to send me later. I wanted a check. Laura wanted those rocks,
beautiful rocks, you know. Great crystals out there. So we’ll see how that goes. So the Lee Kum Sheung Center
for Health and Happiness is about, roughly, a
little more than a year old, funded by the
Lee Kum Kee family in Hong Kong whose
goal was to really look at how happiness
and the movement, social movement for
happiness can be spread. And so they came to us and as it
always happens with academics, they wanted answers. We had more questions. And if you give
away the answers, then what is the purpose of
existence for an academic, right? So we had posed a number
of questions for them and they agreed that
that is a lot that remains to be done to
connect happiness with health and the connection
is not very clear. The Center itself has two foci. One is to really advance
the signs of health and psychological
well-being, a.k.a. happiness, and then to translate
that signs to influence policy and practice. So those are the two
foci of the center and ever since the
center was announced, there has been a lot of
pressure on us to be happy. As soon as the
announcement came, I think every person we
encountered in the school said, are you happy? I didn’t know what I
was smoking at the time, but I did have to force
myself to look happy about it, but it’s really a serious
issue, I think, in terms of– given the social conditions
and the conditions that we are facing all around
us, including what happened yesterday, it remains a
viable question to ask, what are really
the social factors and psychological factors
that influence well-being and how is this well-being
related to health, if at all? Or is it the other way around? So these are really
serious questions that our center aims
to address, but, as you will realize from the
conversations today, it will go beyond that. I think there are a lot
of social factors, what we, in our lingo, call
social determinants, as well as a policy and regulatory
issues that influence a number of these issues. More critical I think is
the development, which is the reason we are here, is
the development in information and communication technologies. The sheer ubiquity
of these technologies is what has compelled
and impelled people to raise some
profound questions about, what role do these communication
technologies or information technologies play in the way
we work, the way we play, the way we converse, the way
we engage in relationships with those around us. I can give you a lot
of statistics on this, but I think the
point is if you just put a hand in your pocket
or your bags and pick it up, everybody has one of these. These are more
powerful, as the cliche goes, than the computers that
sent man to moon in 1960s. So they’re
extraordinarily powerful. The power doesn’t
stem from the fact that it has the bandwidth and
the hardware as much as 90% of the American audience or
American adults go online. About 80% of them are on social
media of one kind or other. A substantial proportion
of them spend their time on social media. So the question,
then, people often ask is, with all this ubiquity,
with all this engagement, with all this exposure to
information or misinformation, what is the impact
on one’s well-being? So those are the kind of
questions that keep coming up. So what we want to do is
raise some of these issues, both in terms of where the
science is as well as some of the regulatory and
legal issues including privacy and other issues that
these technologies engender. So what we want to do is to
have Laura Kubzansky, who has pioneered the issue of
positive science and health in our school and in the
world, to first talk to us, introduce us to this idea
of what is positive health or happiness. And then we will
follow that by– again, I am trying to connect
this issue of happiness to information and
communication technologies. So let me first open it up
and let Dr. Kubzansky talk about what is positive health. And please give us what are the
pills we are supposed to take, and how can we be happy, what
should we smoke, or whatever. Tell us. If I had those answers, I’d
be a very wealthy woman. Well, thank you for the
wonderful introduction and setting the stage
for some of the thinking that we’re going to
talk about here today. I just want to start with
a brief mention of how I got into this work. So a lot of my work
in this area stems from this interest
in breaking down this idea of mind-body
dualism, which is a historical way of
thinking in biomedicine that we still have today with
this idea that somehow, mind and body are separate, and
that mental states can’t really influence physical
health states. And well, I say this,
and I suspect many of you will be thinking, oh,
we know that’s not true. We know that stress influences
heart disease, and so forth. In fact, this has been
a longstanding debate in both public health
and biomedicine. And the debate is ongoing. So if you get on to the American
Heart Association web site, they actually have
a statement that says it has not been
demonstrated convincingly that stress is a risk
factor for heart disease. So there’s still a lot
of discussion about this. And I got into this work
thinking I really genuinely believe that mental health
has a significant influence on physical health. We should look at it. We should understand it. Because if indeed
that is the case, then we are missing
a lot of information about how to influence health
and how to promote well-being, how to promote better
health over the life course. So a lot of the work
we’ve been doing has been focused on the
interplay between mental health and physical health. And this is indeed one of the
focuses for the Lee Kum Sheung Center, which is to say how
does positive mental health influence physical health. And part of that
thinking is also trying to understand
how the social and the physical environment
influence both mental health states and physical health. Now, most of the work in this
area, up until very recently, has really focused on negative,
adverse mental health states. So thinking about
stress, Everybody has something to
say about stress. If I ever get on an
airplane and someone says to me what do you
study, and I say, oh, I study stress and
health, almost inevitably, the next thing they say
is you should study me. And then I’m always like, just– I don’t need to know. Anyway, so there’s
a lot of thinking about how negative
mental states may influence health– things
like depression and stress and anxiety. And there’s a good amount
of evidence now gathering to suggest that, in fact, poor
mental health is a precursor to poor physical health. But as part of that work,
when I was doing it, I started thinking, I wonder
if positive mental health would have an equal
influence on physical health. And would it have an effect
mostly because it just tells you, well,
someone isn’t depressed? Or would it have an effect
because actually, there’s something over and above
not being depressed or not being angry that
confers health states? And so we started
looking at this issue of whether positive
psychological functioning could actually confer health benefits
beyond simply telling you that someone is not depressed. And as part of that work, we
have found pretty consistently that indeed, there is
a health benefit that seems to be associated
with positive psychological functioning over and
above just the absence of negative mental
health states, and that these seem to travel
through both people engaging in healthier behaviors,
and there may also be some biological
alterations that come along with these
positive mental states. And one of the very
strong findings has been that having
positive states signals more than just the
absence of something bad. So the presence of
something positive is very different than saying,
oh, my friend is not depressed, so they’re OK. And one of the things
we’ve been arguing is that most of public
health and medicine has focused on the bottom
half of functioning– for good reason. We look at problems. We’re very interested in
how to solve problems. And people who seem to
be functioning poorly, we really want to help them. But one of the things
we’ve been arguing is that we really focus
on the bottom half. So we say someone is either
depressed or not depressed. And we don’t, as
often, look at what happens if you look
at better functioning, or even optimal functioning,
or you thought about what does it mean to flourish. And if you never look at the
positive half of the spectrum, there may be a
lot of information that you can’t learn about how
to help people function well or how to help people get
onto a healthy trajectory so that they can continue to
function well over the life course. So I think there’s
a lot to learn. And certainly, the
research bears this out– that we gain new insights
if we look not only at what happens when
there’s problems, but what happens when
people are functioning well. And we try to understand
who functions well and what are the circumstances
that enable people to do well. And how come some people seem
to recover relatively quickly from very difficult
circumstances, and other people seem to be
just laid out flat by those circumstances? And as part of this work,
I think one of the factors that we found to be really
important is that emotions matter a lot. so both negative
emotions and positive emotions, and that well-being is
more than simply a state of simple happiness. In fact, there are many
facets of well-being that comprise what
it means to have positive psychological
functioning. And this includes things like
a sense of optimism or hope about the future, being
engaged in your life, and a sense of
satisfaction and purpose, a sense of belongingness. So when you look at surveys
that simply measure happiness, I would say that’s interesting,
but they’re often very limited. For instance, if you ask a
group of parents are they happy, and their kids are small and
running all over the place, they may say the levels of
happiness are not so high. They’re running around. They’re really tired. On the other hand, if you
ask are they very engaged, do they have a sense of purpose
and meaning in their lives, often those answers
are much more positive. And so it’s a little
simplistic in the way that we often think
about happiness as a very simple emotional
state when actually, positive well-being is
comprised of a much more complex and nuanced set
of ways of functioning. So what I wanted to say about
all of these issues around positive well-being and how we
think about it in terms of how it relates to social media– I think there are at least
three ways in which social media interfaces with health. And in fact, people on the
panel may come up with more. But when I was thinking about
it, I was thinking, well, there’s the issue of
health communications. So how do we inform
people and help people to engage in strategies
to enhance their health? So if we think that there are
circumstances or strategies that would enable
people to attain higher levels of
well-being, how would we communicate that most
effectively to people? There’s also this
incredible opportunity right now for data collection
using social media. And I think that’s an
exploding area that, for those of us in research,
is incredibly exciting. I know we’re going to talk a
little bit about that next. You have more information
and more opportunity to see people function
in the day-to-day world in ways we’ve never
really had before. And I think that’s
going to give us a very rich, nuanced
understanding of these kinds of questions of circumstances,
and how how people function, and so forth. And then this other
question is one that I think many people
are very concerned about, which is the effect of social
media on mental health. And then we’re
particularly interested, because our thinking
is if it’s having an effect on mental
health, then down the road, it’s also probably having an
effect on physical health. And so we should be
very concerned about how those things play out. So there are many,
many questions still to think about in terms of
the role of social media in how people function. There are some knowns,
though, that we can take with us to this
research and this thinking. And we know that positive
psychological well-being includes a sense of purpose,
and a sense of belongingness, and a sense of community. And it does go beyond
the simple feeling of pleasure and happiness. So when we’re thinking about
how to look at these things, we should be looking at
a broader set of measures or a broader set
of facets of how people are thinking than just
asking people are you happy. And with that, I’m
going to leave it there. Thanks, Laura. I was ready to cut you off. But I had to be careful. I work with her. So let me– I think that was a
very good introduction to issues of positive health. she Introduced a number
of different terms. Let me just connect that to a
little bit to my area of work in the information and
communication technologies just to introduce
a couple of points and then punt the
question over to JP. So here is the
reaction of people who look at a
particular type of media and said there is massive,
irrefutable evidence that exposure to this media lead
to juvenile delinquency– massive, irrefutable evidence
that exposure to this media lead to juvenile delinquency. They were talking about
movies in the 1930s, the so-called nickelodeons
where people can put in a nickel and then go in and
watch these movies. . And of course, as it so
happens, Herbert Blumer– some of you may know that name. You may not know that name–
the very famous sociologist– actually goes out and
does research, looks at the empirical
evidence, and says, well, it’s a bit more
complicated than that. There’s really no
evidence that everybody who’s watching these
movies, especially the children, the kids, are
becoming juvenile delinquents. So that was one in the 1930s. In the 1950s, people
were concerned about comics and comic books. So a fellow called
Frederick Wharton, who had an MD by his name, started
making a big noise about how exposure and reading of
comics and comic books could potentially
lead children astray. And they wrote a book called
The Seduction of the Innocent with the title, with the
framing, with an M.D. by his name. And of course, it
caused a big human cry. And the government
was even ready– in fact there, was a
proposal by a group of people to form something
called Comics Code Authority. Just think about it. So again, a new medium,
a new technology– people are concerned. People are ready to intervene. And of course, when television
was introduced in the country in the 1950s and ’60s,
again, there were a series of committees in the Congress– The Senate Committee on Juvenile
Delinquency, the hostile– what they were saying is the
very exposure to television, the very use of television
could potentially lead children to imitate acts of aggression. Because they have seen kids
go out and beat the heck out of the Bobo dolls in
these small experiments. And now there were a
series of proposals in the country to really control
the exposure of children to TV. And we know how that worked out. As I said, after the
television entered the bedrooms of children,
all the battle was Lost. But of course, none of us
have become delinquents. At least not publicly
we would admit to it. We all grew up with television. Except me, actually. Television was not in
India for a long time. But the point is every
time we have introduced a new technology, a new
communications technology, this question always comes up. What is the impact it is
having on people, number one? Number two, then there is
a great deal of anxiety to regulate this technology. And then over
time, people become really adept at how to use these
technologies and manage it. So I think we should remember
that as we get very anxious about these information
communications technologies, including social media. The question I have today– for which I have an answer, but
I will let others answer, too– is, is there something
fundamentally different about these
technologies, the information and communication
technologies of today as opposed to the communication
technologies of yesterday? And if that is the
case, then there is a reason for us to
become anxious or celebrate, either way, and take
appropriate action. So that is the fundamental
question that is– if you look at the evidence–
and I’ve been looking at this evidence very closely– at least in terms of happiness,
psychological well-being, depression, mental health, the
evidence is all over the board. You can download any
article you want. Some articles will tell
you that social media are leading to depression. Some articles will
tell you, actually, social media are connecting
people to other people, thereby influencing
psychological well-being in a very positive way. So we do have a lot of
questions to answer. Bart people don’t have
the patience or time. They can’t afford to
wait for these answers. The question, then, is
what do we do given this. Those are some of the
questions that will come up as we discuss this. But Laura said there
are some advantages that these technologies
allow us to test and provide evidence much more adeptly than
has been the case in the past. So Dr. JP Onnela,
who is a faculty member in the Department
of Biostatistics at Harvard and Harvard Chan, has been
doing some very interesting work on social networking science,
but more importantly, a line of work on what he
calls digital phenotyping, how smartphones can be
used to collect data on psychological health. So JP will talk about that. Well, thank you so
much for having me, and thank you for that bridge. For a moment, I wasn’t sure
if I was in the right panel, but I think I am. So my background is– I have a somewhat
different background from some of the
other panel members. I was trained in physics
and network science, and I’m currently faculty in
biostatistics, as Vish said. And I wanted to give you– I wanted to take a step back and
give you the focus of our work. So as a natural
scientist, as somebody who thinks about data and
measurement very deeply, I’d like to argue that
progress in science has always been driven by
the availability of data. But something very dramatic has
happened in the last few years. If we even think about
the last five or 10 years, it’s not only that we have
more data available today about human behavior than
we’ve ever had before, but we have completely different
types of data available now, the kinds of data that
we couldn’t maybe even dream about a few years ago. And the fundamental reason
for this is probably what’s called Moore’s law– the idea that the
number of transistors we can pack into a small
chip approximately doubles every 18 months. So I won’t go into excruciating
detail about that, but why should that be relevant for
what we’re talking about today? And I think there
are two reasons. One is what that implies is
that we can make sensors that are cheap to make,
that are small, and they can be put
almost everywhere. So that’s the data
collection piece. But equally important,
if not more important, is the data analysis piece. So now we have the
combination capacity– we can take
sophisticated models, whether it’s more standard
statistical models, machine learning, AI models,
whatever you like, and try to make sense of these data. So my argument would be that
while collecting research to create data is
still very difficult, I think the main intellectual
challenge is moving away from data collection
towards data analysis. So in our work in this area,
we’ve– you already introduced the term “digital phenotyping.” So we’ve tried to address
what some people call the phenotyping challenge. So a phenotype is a collection
of all different kinds of traits of organisms– so
for example, transcriptomics and imaging and so on. And if we think about
these different classes of phenotypes, behavior
has traditionally been a very difficult
phenotype to handle. And the main reason
behind that is that it’s highly
context-dependent and it’s highly time-dependent. So our solution to
this problem has been what we call
smartphone-based digital phenotyping. So we’ve defined
digital phenotyping as the moment-by-moment
quantification of individual-level human
phenotype in situ, in the wild, using data from personal
digital devices– in particular, smartphones. And when we started this line
of work several years ago, when I would talk
about this stuff, the common objection would be,
but nobody has smartphones. And how many of you
have a smartphone today? Thank you for
validating our work. So now, as it turns
out, 95% of Americans have some kind of a phone, and
almost 80% have a smartphone. And I think what
this enables us to do is to measure social and
behavioral markers at scale at a level of precision that
we simply never had before. If we think about
how do we– again, I know nothing about
happiness, but I’m surrounded by happiness
experts, which makes me happy. We typically would do surveys
to learn about people’s emotions and how happy they
are, and so on. But if you think about what
are the limitations of survey instruments– well,
one thing is it’s difficult to remember things. I just got back from
Europe on Sunday. I’m completely jet lagged. So I can’t even remember what
I had for dinner or a week ago. But typically, what
we do, especially when we study patient
populations, we’re asking patients who
really have trouble remembering the past few days– we’re asking them questions
about how happy have you been in the last two
months, or have you been good about x, y, or z. So the idea is that
we try to move away from using survey instruments
to really trying to objectively measure behaviors in the wild. And when we do this, we can
think about two different kinds of data. So one data type that
we call active data refers to data
that’s only collected if the participant actively
participates in our study. So this would imply things
like taking, for example, a survey on smartphone
screen or giving an audio sample, which can
be diagnostic for example, for depression. The other category
is passive data. And these are data that
are collected continuously in the background. So this would be things like
GPS data, communication logs, accelerometer data, screen
on/off data, and do on. And what these
enable us to do is to learn about
mobility, movement, social networks, communication
habits, and so on over time. And so this obviously
has privacy implications, which I think we will hear a lot
about in a couple of moments. But it’s important to realize
that in this type of work, we always consent
our participants. So we started this work– this was a crazy idea which
we had a few years ago. And we started with
a single study at MGH in a bipolar cohort. Today we are running about
15 different studies. We have studies in depressed
subjects, subjects with bipolar disorder, schizophrenics. We also do stroke. We have brain tumor,
we have spine tumor, and we have a breast
cancer cohort, and so on. So what started as
a simple idea, which is try to capture the
lived experiences of people using devices that we already
carry, most of us carry on us. The idea then spread and seems
to encompass a broad range of different medical cohorts. So if we try to think about– to come back to the idea
of health and well-being and happiness, I think
the opportunity here with these new
communication technologies is to provide better measurement
of social and behavioral markers in the wild. And then if we have the
methods to take those data and try to make
sense of those data, I think then we’ll be
in a better position to try to understand some
of these relationships. So for example, if you want
to think about bidirectional relationships– does happiness
cause something or does something cause happiness– you really need to have
data that’s what’s sometimes said temporally dense. So we have not just one or
two ways of observations, but potentially
hundreds or thousands of measurements per individual. So I’m so jet lagged I
have no sense of time. This feels like
five minutes to me. Is that all right more or less? That’s fine. I’ll not let you go. I have some questions. I’ll come back and ask
you those questions. So thank you. That was very good. The next will be Professor
Jonathan Zittrain. If you haven’t had a chance
to look at his website at Berkman Klein Center,
I urge you to look at. It is very provocative. I have heard him speak about
15 years ago, actually, at a conference. And then it so happens now
that we are both at Harvard and that I have– it’s very fortunate
that we have him here. He has written a number of books
around the issues of internet and privacy, cyber security,
a whole range of issues, including a book
I would recommend. He’s not paying me for this yet. It’s a very provocative book– The Future of Internet
and How to Stop It. And it’s a very– one of the most thoughtful
people in this area. And Jonathan, why are
you causing trouble? Thank you. For a minute, I saw
the title of the panel is “The Future of Happiness
and How to Stop It.” Maybe that turns out to
be the right perspective. Already what’s been
so fascinating to me is thinking about,
consistently through the panel, the idea of the
study of happiness as a scientific and rigorous
endeavor like any other. And to hear the excitement JP’s
description of all the data points he can at last
gather, this totalizing panopticon of happiness– I, too, feel myself excited
about what might be discovered. I don’t know how many
people use FitBits or other quantified
self-mechanisms. And of course, as those
develop, we can not only study individuals but groups. We can ask is Boston happy
tonight and get an answer. And we have to be
careful that sometimes– don’t ask the question unless
you want to know the answer and to get near
instant feedback. As you think about, say,
the presidential debates, if you watch it
on some channels, they have the little line
with people with dials, sentence by sentence. If you see them as you’re
the person speaking, you can see the thing
going down and just change the end of the sentence to
try to get it back up again. It does make you wonder. And Vish mentioned
a number of books. We’re in an era
now where data is able to be gathered
through your Kindles, say, of how your
reading is going. And again, you could
get group data out of that to have a
sense for here’s the paragraph of chapter
2 where everybody stops reading your book. And wouldn’t that make it
tempting to issue an update to your book, pushed
out to the Kindle with just a picture of a happy
cat where that paragraph was? And now we bring you back to
your book without that hump. And it shows that there could
be such a thing as pursuing. It’s why Vish had mentioned
it’s so careful to quantify what is it we’re wanting
to measure and possibly optimize for. You could optimize
for it and then find the only books we
produce are ones that appear to make us happy or
bring us to finish them, but they might not
be the books that we would want to be reading. Now, Laura had
mentioned that when you start studying
something like this, you might start looking for– in a correlation, say,
between mind and body, how unhappiness leads to
unhealth rather than how happiness leads to health. And it’s true that the stories
we tell tend to start– or have as the middle– a snake
in the garden and a problem. And I think of it more
generally as the things we tend to worry
about, including if we’re telling a
story about privacy, classically are Orwellian– invasions of privacy in the
service of another, maybe a government, like
Big Brother, that use the data that they
find to further advance the cause of misery. As I recall, in 1984, when
it’s revealed the aim of this government– why does it exist– I think they just say it’s
a boot forever stamping on the face of the human soul. OK. Well, you’re
maximizing for that. But there’s actually
another worry that is the complement to
the classic Orwellian vision. And that’s Aldous
Huxley’s vision that what we should fear
most is not what harms us, but what we fear most
is what makes us happy. That could put us into a state. And that, of course, is
maybe not a bad introduction to thinking about the impact
of technology upon us. And it calls to mind– it’s a computer scientist named
Luis von Ahn, Carnegie Mellon. He is very interested in
a phenomenon now called gamification and the
ways in which you can get people to do stuff
by turning it into a game. And one example– he
calls it the ESP game. You’re shown an
image, and you type in what you see in the image. And that gets matched
with another player that you’ll never meet. It’s somebody online, maybe
who isn’t even playing contemporaneously, although
he doesn’t make that clear. And when you match the same
word for the same image that you see– there’s a picture of
a kid, and you say kid– you win points. And it turns out the points
are not good for anything. You can’t cash them in. They’re not like
American Airlines points. But people love accruing points. And he found that people
would play this game for as many as 40 hours a week. And when he tells the story, he
says that his academic advisor made him put in a check in
the game that if people were playing it coming from
a domain ending in .edu, it would cut them off after 20
hours and say get back to work on your thesis. His conclusion from that
was the ESP game is fun. I think fun might be
not the right word. I think compelling is
probably the right word. And thinking about
happiness versus compulsion might be really important. Because as we build
systems, either like the ESP game,
which it turns out is an image labeling tool– it gets images labeled through
people playing the game. If it’s in the service
of something else, you might find our
own penchant for fun to be exploited to that end. And we may feel differently
about it once we know that. But even if it’s aimed
to make us happy, as measured through
continued engagement, it may not really
be the right metric for what we’re aiming for. And that’s also
why it’s probably right to call these smartphones
rather than happy phones. And it might be
right to ask what would a happy phone be like. It be one that shuts down
at around 10:00 and is like, go out and play. I want to just say a
word about technology as an end unto itself,
the gateway to knowledge it might represent,
the relationship you feel you’re having with
the world of cyberspace, versus technology as a mediator
to engage with other people. And the latter, I
think, for a while– if you’d asked me
in 2005 or 2006, I would have said it
would be important to see the second aspect
of technology is that which might
bring us together as important and helpful. I think since then,
a lot of realization has happened across many
people studying this that we worry about technology
in connecting people– in impelling us to try to
present our best selves. And if we’re not showing
big smiles on Instagram, that we don’t feel
like we’re leveling up with the number of likes
that we’re receiving. And at the same time,
if we see everybody else having a delightful
time on Instagram, and we’re just like– I don’t know. It’s a Friday night, and I’m
just kind of sitting around, that might not make
us feel so good. And of course, it’s also about
our relationship to ourselves. And there are now a
range of meditation apps so that you can take even
meditation and appify it. And I’m not ready to
say that’s a bad thing. It might turn out
that it’s quite good. It’s just interesting
to see us relying on leaning on our
technology for the things that we normally thought of
as completely technology free. A last note about happiness
versus helplessness. I think in the list that Laura
was saying of qualities that have been discovered to be
associated with, in her words, well-being– and I do like that phrase
more than happiness, if only for the reason
that it’s a gerund. It suggests a verb rather than–
it’s a noun hiding as a verb. It suggests it is a process,
a continuation, a vector, rather than an equilibrium,
a static kind of state. But helplessness does
seem to me a real source of feeling not in well-being. And to the extent that
our technology makes us feel helpless and overpowered
rather than enabling, that might be something
to keep an eye out for. And it may even be that our
previous analog feeling– fighting against
helplessness, we might even do it and seek it,
even as against our own health. I’m reminded over
in Inman Square, there used to be a
restaurant called Jay’s. It was a sushi restaurant. And it said at the top, eat
at Jay’s and live forever. And next door to Jay’s
was the East Coast Grill, the Cajun barbecue place. And it said at the top–
literally right next door– eat at the East Coast
Grill and die happy. And that seemed to point
at a weird form of unhealth as a form of happiness. So I’ll just close
by saying I think it’s worth contemplating not
only the science of happiness– which there really
is such a thing, and it is something
we can study– but the humanities of
happiness and the ways in which we might use that
science to reflect on something like Aristotle’s idea of
the golden mean of virtue and perhaps fulfillment coming
from being in between two extremes. And is there a way to
quantify that operationally is a scientific
question, and inform the kind of philosophizing
that also seems so valuable. And this which makes
us different from inanimate objects– the idea that we have a self,
we have a consciousness, and we experience
ourselves and the world– is a way that could be
less or more fulfilled. Thank you. Thank you, Jonathan. That was great. So the program is that I will
ask the panelists a question, each maybe just a minute or
so, and then we will open it up for the audience. I’m sure you have
a lot of questions. So let me start with Laura. So you study
happiness, well-being, and you have made it
complicated by identifying a number of factors. So here is some evidence on
social media that shows that– so in fact, Jonathan mentioned
this notion of self-disclosure, disclosing one’s emotions. Apparently, in this study,
the large majority of people, if there is a lot
self-disclosure, actually, it increases
a feeling of well-being. On the other hand, the other
two functions of social media– informational and
entertainment– are not necessarily
related to well-being. So the question is what
is it about this idea of relationships and
disclosure of emotions that seems to lead to
well-being and happiness, but not about seeking
information out of the things. Do you have any
speculations on that? Those are interesting questions. So there is a whole
line of research that talks about the
importance of acknowledging and recognizing
your own emotions, particularly when you’ve
had difficult experiences as a way to process and
think about what’s happened, and put order on it, and
make meaning out of it, and that people who are
able to make meaning out of their experiences
tend to have much higher levels
of well-being, even when they’re faced with
very difficult circumstances. So I’m not familiar with
this particular research, but I do think that it probably
speaks to this idea of both having a sense of
connection– so being able to self-disclose
means, in some sense, you feel some trust. You are going to share how
you feel with other people, and someone is going to greet
it in a way that is supportive and make you feel like
you’ve been understood. So there’s a lot
of research that suggests that it’s really
important to feel like someone has heard you, that someone
understands who you are and what you’re feeling,
and even when it’s a really difficult
time, to just feel hurt. In fact, there’s some
really interesting research that when there’s a
natural catastrophe, or when many, many, many
people have experienced a terrible event, this
is one of the things that gets very difficult.
People want to talk about their difficult
experiences and process them, but everyone around them has had
their own difficult experience, and they don’t want to hear it. And in fact, then
you end up with a lot of people really
struggling, because there isn’t that natural way of
sharing those feelings. So there’s some really
interesting research that suggests that
being able to process these emotions,
both good and bad, is a really important outlet. And so that would be
one explanation for why you would find benefits. And one thing I
will say is that I think we do tend to
have a lot of fear about the effects
of social media. It’s not clear how
much we have actually looked for potential
benefits other than this idea of social connectedness. And are you more
connected because you think you have 500,000
friends on Facebook, or are you actually more
isolated because, in fact, you never talk to a human being? You’ve done everything in
your room, sitting in bed. And I think the jury is out. And my bet is going
to be that there is reasons to think that both
sides could have effects. And one thing that will
be really important for us to think
about is are there ways that we can nudge
the use of technology so that it is more health
beneficial and make it harder, like putting these limits– you can’t spend 40 hours
playing this game– that shuts down certain
functions and so forth. So that would be my hunch. That’s a great answer, Laura. I think it also speaks to the
issue that you raised earlier. We always seem to focus
on the bottom half and not the top half, the negativity
rather than the assets. That seems to be the trend
with our technologies, as I said earlier
with comics or movies. But the evidence clearly– at least as clear
as it can get– does show that social
media could potentially be helpful and beneficial. But of course, we
always focus on that one case, the cyberbullying
or whatever it is. So I think it’s
important to know that. So thanks for that response. So JP, I have a
question for you. Again, this comes
out of left field. But hey, you’re jet lagged. So if you don’t know
the answer, make it up. So it’s very exciting to
see you can do this kind of micro-moment data collection
that is very helpful, particularly in small contexts
where you can treat people or help people immediately,
understanding– and collecting data
with a great deal of integrity and reliability. So there’s a two-part question
on this one, all in one minute. And my question will be
longer than your answer. One is how do we
talk about protection of data in terms of– as we collect these data. This is an issue that came
up when we were doing– using smartphones
with urban poor. And most of our research
is with the urban poor. Second– also, most
of our research found out that when it comes
to urban poor, working poor, they have new
challenges in terms of paying bills,
bandwidth as well as the hardware of their phone. So a lot of people
don’t even bother to recruit a number of
people from a certain lower social economic position
into these studies, as a result of which
much of what we know seems to be from a certain class
of people, the same people who have been studied, I think. And our research was
very telling in terms of how difficult it is for
them to hang on to their phones in terms of paying their bills,
having connections, et cetera. So how do we
overcome those things to really get to
the promise that you claim with smartphones? That’s a challenging
question for one minute. So let me give you– I’ll actually answer
three different parts. So I’ll start by you
alluded to the idea that this is a way how we can
study small groups, precisely. But I would actually make
the opposite argument– that this is a way to study
very large groups at scale. So the instrumentation, which
is essentially a smartphone, iOS or Android– these are not measurement
instruments we have just here in Cambridge or Boston. So we can use the
same instrumentation almost everywhere in the world. So we could easily recruit– and have some plans to do so– to have a large
international study where we can recruit people in
Germany, Sweden, India, the United States, and everybody
has identical instrumentation. So I think for me, the
biggest appeal is actually the scalability of
this technology, which is interesting. Now, to your actual
first question on the prediction of data– so of course, this
is nothing new. We’ve always had this
idea that– people have been studying for a long
time how do we protect data. How do we guard against
intrusion of privacy? And I think that these
techniques have also developed as is the nature
of data has developed. This is not really a
one-minute question. But one useful way,
potentially, to think about this is the probability of
re-identifiability. So the idea is not whether
you can re-identify someone in data, because you always
have a tiny probability of doing that. But a more productive
way of framing that question can
sometimes be what is the probability of
re-identifying somebody in the data set that
has been anonymized. And so this is an area where
people are doing a lot of work. And it gets very complicated,
very technical, very quickly. But that’s a great question. The second part
on the idea that– are we only studying a
small subset of individuals? And this certainly used
to be the case when the prevalence of smartphones
was maybe 5% or 10%. But my prediction is that
in the next few years, we’re going to hit maybe
95% or 99% prevalence for smartphone ownership, if for
no other reason than the fact that these are the only kinds
of phones that are made today. So eventually, almost
everybody is going to have one of these phones. But there are still
important differences. So for example, one of my
favorite statistics is that– think about a family where
everybody has an Android phone, and think about a family where
everybody has an iOS phone. Think about the annual
income of these families. What do you think? Is there going to be a
difference between the incomes of these two families? And the answer is,
there is a $40,000 difference in the median
incomes of these families. So this is to say
that although we can use this technology today
to study 80% of all families, there are still huge differences
in what types of phones are owned by people, depending
on their socioeconomic standing, and also how
people use this technology. Because to get the most
out of your phone– well, your smartphone
or your happy phone– there are costs
associated with that. But I think this digital divide
is probably getting less. That’s my sense. Thank you. So Jonathan, last
question for you, I think, before we open it up. So you have written– and I may be wrong, but
I’m making up maybes. And also, in a number
of your writings, one of the themes
that comes through is the notion of control. Who controls these things? And you and others have
argued, in your group, that the control
seems to be lying in a small group of
corporations and companies as opposed to
individuals or families, or whoever it is, groups
controlling the technology. And I think that has
significant implications in the way we
design architecture and systems in place. Do you have some
thoughts on how do we– because that has an
impact on not just psychological well-being,
but I might say even social well-being and how
we use this technologies to our advantage. Yes. And I think that’s
a worthy question to ask to amplify the question
you asked JP about privacy. And if you look at it from the
lens of control and autonomy, first it’s worth noticing,
for example, I have no doubt that JP is extremely sensitive
to privacy issues for the folks from whom he is
collecting data– data that he’s going
to use to have insights about what makes people
happy and how to pursue that. Now compare that to the
usual data and telemetry that’s being collected
through the corporate vendor. And the corporate vendor
has no such obligation to be thinking about the social
good of the experiments they’re running or the interventions
they’re making. And it’s kind of funny to think
that an academic might have to spend a couple
of months dealing with an institutional
review board, an IRB. And it’s like– because of
Stanley Milgram and Philip Zimbardo and the Stanford
Prison Experiment, this is why we can’t
have nice things. But because of that, we
have to be really careful. And Facebook is like, this
is what they do on Wednesday morning before 8:00. And it’s like, on
to the next thing. So it’s a weird imbalance
and asymmetry that really doesn’t make much sense to me. And the weird dividing
line on policy is Facebook is
allowed to do what it does so long as
what it’s learning is not to advance
general knowledge. So long as it’s just for
Facebook, go to town, Facebook. Milgramize yourself. Whereas if it’s to
advanced general knowledge, well, now the
Common Rule applies, and you’ve got to think
about the experiment. It’s a little bit weird. And so as we’re
bolting the front door, it might be worth looking to see
the back door that’s wide open. The other point to make that has
to do with control and respect for people is when you think
about data protection, often the touchstone– with a tip of the hat
towards autonomy– is choice, so
presenting somebody with a list of how they’d
like their stuff to be shared or a question about– who knows? It’s a long question. And at the end, it’s
like, OK or cancel. And so I don’t even– irrespective of what the words
say, how many people click OK? How many people click Cancel? The Cancel people,
it’s like, that’s why Friday night
is Yahtzee night. Because it’s like,
you’re not doing the thing you were going to do. It’s like, I’m out of here. And I think that’s an
illusion of choice. That is itself a form of
nudging people the way you want them to nudge if
you’re the person asking for the consent. And it’s why, ultimately,
I’d push a model– and the word is not the
best word, but fiduciary, a fiduciary model. And it’s a word
that basically means a duty of loyalty and care to
put the interests of the person that you’re in privity with,
and whom you’re learning a ton about, and you’re
filling their news feed with news that might excite
them or depress them– to do that in a way that puts
their interests ahead of yours. And it may be hard to
know when you’re doing it. It’s easier to know when
you’re not doing it. And if you’re doing something
that’s designed only to elicit the people
on Facebook that might vote the way that Mark
Zuckerberg wants them to vote, hypothetically speaking,
and only alerting, say, those folks that
it’s election day, and for the others, you show
them the picture of the cat– that’s not being
a good fiduciary. And I think if we
introduce that concept, it may be a way to start
thinking across the board– not just for academics, but
for everybody in a position of power running a platform– handling this data
stream, thinking about how it’s serving the
person you’re working with. And as we– this is now
to answer your question originally, Vish, about
what’s different– what’s different is the prospect
of what we loosely call AI. And I mean it in the narrow
sense, not the general AI sense of Her or HAL 9000. But I mean in the
narrow sense of an AI that can exquisitely find out
when somebody is feeling really down. And a fiduciary would
say, you’re feeling down. First you should
be aware of that. Just so you know, we’ve noticed
this is not your baseline. And you might not notice that,
too, because you’re in a fog, and here are some options. That’s meant to look
out for the person that you’ve become aware
of how they’re feeling. If, instead, you’re like,
this is the perfect time for a payday loan– we’re just going to
send you an ad right now to unload at 9% interest. And no one else will see it. This is a special
offer just for you. Ralph Nader isn’t there to
vindicate your interests, or the attorney general. That’s not being
a good fiduciary, even though you’re
likely to click. So that’s how I’d think
about autonomy in power. Wonderful. Thank you, Jonathan. That’s great. So I believe microphones
are in the audience. We can take, Jen,
maybe five minutes. Is that OK? Yeah, five or 10. Five, 10 minutes. Great. I was able to buy
five more minutes. Thank you. That’s good. So there are some hands up here. I wonder if the panel members
might comment on the concept that it’s sometimes
perceived that people use social media to not
just share about themselves, but to validate their
own view of the world, and that there’s this duality,
this concept that we’re all going to learn from one another
globally and be so informed, but there’s also this dynamic
where we like to hang out with people who think like us. And we’ve seen a lot
of evidence of that on the micro and
macro scale of late. So how does that play into
well-being and mental states of happiness? Is there this
cognitive dissonance that if we engage
with people who see the world very
differently, is that going to make us unhappy? And does that then result in
self-reinforcing behavior, or not even being
aware of undercurrents that change the world? Anyone? It’s about echo chambers
and the algorithms. And I think that’s
a significant issue. The evidence– again, I
hate repeating myself, but what the heck? I can do that. So the evidence is all
over the board again. I think that there is actually–
the literature on political science which actually looked
at people being exposed to opposing points of view, and
how that actually influences their own ideology and
then voting behavior– I think I don’t recall
exactly all the studies there. But I think increasingly
with social media, this has become the big issue. I think there is
that part where, with these algorithms, that
you are only exposed to people or ideas or opinions that
validate your own position, validate your own self. I think there are both
good and bad about it. On the other hand, it does
raise a significant issue of are we becoming increasingly
so polarized that when the country doesn’t vote
the way we want to vote, then we get shocked, which
affects our well-being. So I think this is a big issue. I don’t know whether there is
a specific answer, easy answers to these questions. But I think– I also study vaccinations
in the area of social media. And I can see– I think the echo
chamber effect actually leads to unhealthy outcomes in
terms of well-being, I think. So that does raise some issues. I think you have
written some on this. I was going to say in the
study of the First Amendment– so we’re talking the local
ordinance for the United States– the general view of it is it’s
about the right of a speaker to speak and a listener
to listen, and maybe to hear what you want to hear. That’s the right. But there’s another view
of the First Amendment that people like Cass
Sunstein talk about, which is from society’s
point of view, trying to grease the skids as much
as possible so that people are confronted with views
they don’t agree with and can’t readily turn to
the law to prevent that. And you see it come
up in cases about if there’s a protest of
a particular activity, like a political convention is
in town, under the second view, giving the people
their own protest pen several miles away is not
conducive to confronting the people at the
convention with speech that they might not
like on first blush, but that could actually
change their minds. And that’s putting society’s
interests to the fore rather, maybe, than the
comfort of the person doing it. And you might ask,
then, if you’re Google or Facebook
or somebody else capable of constructing what
Eli Pariser calls a filter bubble around somebody– would it be helpful
to construct something that really confronts
people with stuff they might want to hear? Is a good fiduciary
somebody who, when you– Google, should I vaccinate
my child, one that’s like, let me show you what
you want to hear, or let me show you what we know? In that case, I think
it’s the second. And how do you reconcile that
with wanting to be empowering to the individual? It’s a dynamic
system for which I think asking people about would
you like to be challenged, would you like to hear from
somebody that might genuinely disagree with you– I think a lot of
people would say yes. I’ve pushed Facebook, in
addition to the Like button, which then means I’d
like to say stuff that earns the most
likes, which means that people are agreeable to– I think there should be–
and I’m not a good ad man, so it shouldn’t be named
this– the Voltaire button, after the idea of “I
completely disagree with everything you say, but I
defend your right to say it.” In this case, it would be I
completely disagree with that, but that was really
interesting and well put. Thanks. And if you create
such a button, it creates an extrinsic
motivating system for people to post Voltaire-like things,
which could then naturally get us into a zone where,
as I think we experience in academia, we cherish the
opportunity to hear from people that don’t agree, and
possibly to change our minds. And I think the
ideal academic is one that’s like, god, I was
completely wrong about that. That’s somewhat embarrassing,
but I’m better now. And that’s good. That’s what humanity
should be aiming for. And I have hope that that– I was just talking to a friend
about friendship, actually. And luckily, the
discussion affirmed that we were, in fact, friends. And his view of
friendship was it’s a relationship
whose aim, mutually, should be to help the other
become their best self and to grow in that. And it would be great to
see the technology aiming for that, too, which generally
would be confronting us with stuff that doesn’t merely
affirm who we already are. Thank you, Jonathan. So maybe 30 seconds,
and then we have– He had a minute and a half. I’m feeling competitive. So just one other– 45 seconds since
I work with Laura. Because you’re going to
hear about it tomorrow. So one other point that
I think is interesting– so I’m not actually aware
of any research that’s looked at this directly in
terms of emotional states and so forth. But there is some interesting
research that suggests that when people are in very negative
mental states– so fear, depression, anxiety– their focus gets very narrow,
and they become very attention deficit. So it’s very hard for
people to see anything from a bigger picture. Whereas when people are in
more positive emotional states, they have a much broader
perspective on the world. They’re more creative. They tend to be better
problem solvers, and so forth. And so in that
sense, it would argue that the more fear and
anxiety you induce in people, the more likely they are to
only seek the echo chamber. They just want the
stuff that’s going to confirm what they think
because they’re so afraid. It’s hard to think big. It’s hard to problem
solve and look outside your own
little world view. Whereas if you have people
in more positive states, they may be more open to
thinking about different ideas and trying to incorporate them
in their current worldview and maybe changing
how they think. So I don’t know of a research
that looks at the question that you ask. But I do think it’s really
interesting to think about when people are trying to influence
other people, the effect of putting people into
negative states versus trying to keep a more positive state
to allow people to think more broadly. Thank you, Laura. That was 50%, by the by. We can take one more
question, right, Jen? So what is the impact
of advertisements that we are all forced to see
daily, Including social media– there are so many on every page,
we almost see on the internet– on well-being, happiness? Anybody want to take
that on, advertising? It is so important. We need that money. Ethan Zuckerman
calls advertising the web’s original sin. And it’s interesting,
because it really is a somewhat Faustian bargain. It’s advertising that provides
the basis of free services. Also calls to mind the quote– I think originating from Hippie
Bear on MetaFilter– that says that if something
is free online, you’re not the customer. You’re the product. And the thing I worry
about most with advertising is that the models are tending
first towards more integration. I don’t know how you would feel,
for example, if it turned out the panel tonight was
sponsored by Twitter. Social media isn’t as
awful as you think. That would give you a different
gloss on some of the findings that have been
presented tonight. And it gets back to– I was talking about
fiduciary as an agent that is acting in your interests. It would be nice
to know, online, when you are communicating
with somebody or experiencing
information, is it at the behest of someone else? Or are you connecting
with another human that has a view about something,
and you’re talking about it? And in our current
situation, we’re thinking about Russian
trolls and pay bots and stuff like that. A nice start would be am
I talking to a bot or not. Am I bot or not would
be a great service. And it’s one that–
even Facebook has, in Messenger, something
called M, and it’s their AI. It’s like, hi, I’m
M. I’m here to help. And it’s completely
unclear if it’s a person that’s just punching
in or if it’s just an AI bot. And if you ask it,
it gets very cagey. And I think it would be
nice to know are you a bot or are you a person. And if you are a person,
is this your job? Because if it’s your job
to say what you’re saying, there’s no persuading you. It’s like a CNN
panel or something. It’s a panel of
surrogates, and it’s like, now say the thing
you’re supposed to say. It would be much better to
have people that present a view and might change it. And advertising is generally
not conducive to that. So that’s a problem. Thank you, Jonathan. This is certainly not like a
CNN panel, I can assure you. I wasn’t meaning to
suggest that it was. That’s great. So we are 15 minutes
past our time. We can have one more question. Great. Yes, please. So I love the idea
of fiduciary, but it sounds like you
would need somebody to actually be sitting
behind all of this technology and deciding what to push out. My understanding,
which is simple, is that this is all
automated, and that really, the ads you see
are the ads that we know are likely to
be products that are likely to be
purchased just by the data that you’ve provided. And so it’s– there’s really
no fiduciary where you’d have to get in there and actually
figure out how to stop the ones that might not be good for
you, which is a real point. Because the payday
loan is a real issue. It’s getting pushed out
to you just because you’re likely to take it. Yes. Well, one crude
answer to start with is to maintain the availability
of so-called incognito mode across platforms. So you can say– I’m asking you,
platform, and you’re committing, if I choose– not to be personalizing stuff. I realize that that means
that if I have a cat, I might still get ads for dog
food, and that would be tragic. But that might help
with my autonomy because of the way in which
advertising is, of course, meant to be
a form of manipulation. So maintaining that
distinction might be helpful. And then more subtly, I think
my example of a payday loan was picking out something that
society has generally agreed– and in fact, so has Google
and Facebook under pressure– they reject those ads now– that however precisely
targeted and successful, they’re not appropriate,
because they are preying on vulnerable people. I would just add to that. There’s also the idea
of educating people on what’s happening digitally. And so one of the, I think,
really underappreciated facets of how we
understand the world is how important emotions are. And so I have a
colleague who is always talking about how ads that want
to make you buy their product– so their cigarettes
or their alcohol– are always showing
really happy people who are leaping into the
air and looking like life is just so much better because
now they’re smoking Winston or whatever it is
they’re smoking. And I love how it still
says at the bottom– actors, dramatization. Whereas the public
health messages are often very fear-based
and very negative. And so we tell
people, if you smoke, your lungs are
going to turn black, and that’s really
bad, and so forth. And so whether or
not it’s appropriate, whichever emotion you’re using
that you’re trying to induce– just having people understand
that these kinds of messages are designed to help them
feel one way or another, and the importance of
those emotional states for driving both behavior and
health and all kinds of things, would be useful to be an
informed consumer of digital of activities and so forth. And certainly, there’s
a lot of discussion about how do we do that. Now, the technology is
always racing ahead, and everybody is trying to
come up with new strategies. But I do think that some
element of understanding how these things work is
also potentially useful. It won’t solve everything,
but it’s a useful strategy. Wonderful. Thank you. Let me thank the panel. [APPLAUSE] That was great. And let me thank Howard
Museums for inviting us, and let me thank all of you
for taking this evening. Thanks. Thank you. [APPLAUSE]

Leave a Reply

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