Aneesh Chopra: “Innovative State: How New Technologies Can Transform Government” | Talks at Google

Aneesh Chopra: “Innovative State: How New Technologies Can Transform Government” | Talks at Google


MALE SPEAKER: Today, our
guest is Aneesh Chopra, who’s the first CTO of
the United States, which is quite an honor. He was appointed
by President Obama. And he’s written a book
called “Innovation State.” And my first
question, of course, is, what motivated you
to write this book? Why did you feel the
need to get it out? ANEESH CHOPRA:
Well, first I have a sense of hope for the
country that may not be the current refrain
about Washington, looking a bit dysfunctional. I have a sense of
hope that we’re going to solve the biggest
challenges of our generation. And we’re going to
do it this decade. I also wanted to take a step
back and look historically. Have we had a
pioneering government? Has government been at pace
with the private sector and the incorporation of new
techniques, new technologies, new approaches to managing
and operating on work? And then last but not
least, I had accumulated a set of techniques
during my service to the President, what I call an
open innovators toolkit, that I thought if more
widely understood might help to arm up a
generation of entrepreneurs and innovators. Whether you were
inside the government or outside the
government, but looking to find ways to join in this
cause to solve problems. So the book is all three–
communicating a sense of hope, putting it in historical context
that we’ve been here before, and then last but not
least, a set of news you can use–
ideas that can help to bring about this
more innovative state. MALE SPEAKER: I’m glad
you’re an optimist. I’m glad there’s some
optimist still out there. I’d like to think I’m
an optimist as well. Do you think things are trending
more recently more negatively? I mean, given
things like Snowden and how that’s impacting
the idea of open data for government, is that a
real challenge right now? ANEESH CHOPRA: The great news
is that in the wake of Snowden, you saw the political system
take a step back and say, what is the right role of
government on the issue of data and privacy? And here’s where it’s
actually a positive if you want to
spin a positive off of what obviously is a
difficult circumstance. The president’s team just
released a big data and privacy report that outlined a model
for how information should flow in the internet economy. And that model starts
with the premise that individuals are
entitled to electronic copies of their own data. Now, the reason why
that’s important is it’s actually creating more
data liquidity in the regulated sectors of the economy. So if you’re a student and you
have your data in the student information system at a
college or a K-12 school, you’re entitled to an electronic
copy– machine readable– of that data that you
might want to then share with someone who can build
a personalized tutoring service for you. Today, you really can’t get an
electronic copy of that data. You get a transcript, basically,
and that’s what you get. Or your doctor. You can say to your doctor,
I’d like an electronic copy of my medical data so that I
might have a digital tutor that might help me
navigate, if you will, the health care system to figure
out what I should be doing that I’m not so that
I can stay healthier. So I think you’re
seeing the negative into a positive asserting
a new framework. I own the data and I want
the ability to share it, especially data held
in the government. Even the IRS, by the way,
now allows every American to download a machine
readable copy of their filing so that if they want to go shop
around for a new filing support service, they can now take
their data with them to do so. MALE SPEAKER: I mean,
if you think about it, before we had blue
buttons and red buttons. People didn’t know if people
even wanted this data. And so you put it out there
and now we kind of assume it, that this should be out
there, we should need it, we want to get access to it. What’s the next forefront? What’s the next
breakthrough on open data? And the reason why I
ask, I feel like there’s a certain amount
of– expectations are always going higher
and higher and faster. Are we actually outstripping
those expectations? Are we able to fulfill
on that promise? ANEESH CHOPRA: So I have
a colleague of mine, Sanju Bansal, who has long
stated that the value– true value, occurs at
the intersection of multiple disciplines. And if you take that
into the context of data, it’s not just the collection
of data that’s useful. It’s the ability to mash
up data from, perhaps, government sources
with private sources. And that really is
very early days. So you could
imagine, for example, a veteran who’s
got a profile, say, on LinkedIn, or Monster, or
any one of the other companies where they’ve got their profile,
and they could combine what they self-entered
as their profile with the official transcript
of all the skills they acquired in the military. We blue buttoned a version
of what’s called the DD214. So any veteran can now
download a soft copy of their military transcript. Imagine combining that with
one of these social profiles, or Google+. And what I could do is unlock
all kinds of opportunities to connect to people who
will help me find a job, pick a college or a university
that’s right for me. But it’s at that intersection. So I think the next
chapter is going to be entrepreneurs and
innovators who are mashing up data sets in health, energy,
even education markets as we’ve just described in new and
creative ways to add value. MALE SPEAKER: Have
you been surprised by the pace, or lack
thereof actually, of the startups in this space? There was this hope that
all this open data would be created, and these
startups would come out there and make all these cool ideas. And make like your own
DMV app, or whatever it is– the equivalent
of that stuff. Has that been slower or
faster than you expected? Do you see it
turning the corner? ANEESH CHOPRA: So I would say
slow and fast with an asterisk. Fast in health care,
slow out of health care. And the reason for
that is my successor. Todd Park came out of the
Department of Health and Human Services. And he had a very
interesting insight. Opening up data
is the equivalent of a tree falling in the woods. It doesn’t necessarily
mean people know about it. And so the key had been to
reach out to the community to inform people that
all of this information is now available. So in health care, there’s
actually now an ecosystem. Thousands of people gather
every year at an initiative he called Health Datapalooza
where– now four or five years running, McKinsey has
estimated over 200 new products and services have been
born on account of the fact that people– many of these data
sets were publicly available. It’s not like they
were hidden from view. It’s just now they’re organized
in an easier-to-find manner, and there’s a proactive
system in place to go inform people about it. I also put an asterisk
in health care because you have to operate
in a revenue model or business model that makes sense for you. In health care today, we
are changing the underlying incentives to reward
people who stay healthy versus rewarding
doctors and hospitals who treat you when you’re sick. In that transition,
there is now a premium on understanding or
predicting someone before they get really
sick because it’s cheaper to treat them when they’re
only slightly sick, or even before they’re sick. So there’s now a
market incentive to develop products and services
born on this more population health database. And that business condition,
in addition to the fact the people now know
about the data, now they can make
money off the data. That’s why health is
really scaling as a place for open data innovation. We’re starting to see
that begin to emerge. There are new business
models in education. Slow, but competency-based
learning models, getting accreditation and
the ability to get reimbursed is going to create
more opportunity for learning technologies. As well as energy efficiency. States like California that
have proactive policies that reward their utility to
help folks save money on their energy
bill will encourage them to open up the
data and create apps. And so you’re seeing that
where the business models are in place and people
are aware, it’s actually scaling it
at a pretty good clip. MALE SPEAKER: What can Google
do to help foster that ecosystem as a partner to the government? ANEESH CHOPRA: Well, I
was sharing the story. Google has already done it. And I want to just make
sure folks know the story because it’s an
important one to realize how we can all play a role. The whole point of innovative
state is that every one of us can contribute as a unit
within Google, or a company, or as individuals. And I shared this story and I’ll
share it briefly if it’s OK. The President handed
me an assignment. It was August of 2011. He had just given a speech or
was preparing to give a speech announcing a challenge
to the private sector– please hire the 100,000-plus
veterans who are coming home from Iraq and Afghanistan. And that this was going to
grow over the coming years. And we wanted more
and more companies to make an explicit
veteran-hiring commitment. And we had a data problem, which
is, how will a veteran know which employers made
this commitment? Is there going to
be one website, www.findveteranfriendlyemployers.gov? Who would maintain it? Where would that be built? Or, I called Vint Cerf,
who’s a friend and a mentor. And I said Vint, I’d love
to find a creative approach to this problem because I’m
not sure I’m thinking about it the right way. And his immediate
reaction was, you have to talk to Guha, who is one
of the fellows here at Google, one of the smartest folks I’ve
had the pleasure of knowing. And he told me, this is
a search problem, not a database problem. And that what we could
do is get the ecosystem to standardize a very simple
piece of metadata that says, if you’re an employer
and you have a job posting and you want to make it
clear that you want to offer a veteran-hiring commitment,
add this little bit of metadata to your job posting page and
we can crawl the internet and find all job
postings associated with veteran-hiring commitment. It took 30 days to get
the schema.org community to embrace this standard. And it took another
30 days to get folks to commit to
adopt the standard. But within 90 days of
the President’s speech, we were in the Rose Garden
at the White House announcing that Google, LinkedIn,
Monster, a variety of other stakeholders,
voluntarily pledged to make this
service available. And today, literally millions
of job postings associated with those employer-hiring
commitments are now discoverable
through this collaborative. So what I mean by that is–
I think of life in threes– light, medium, and heavy. The light is when
you have spare time and you want to
commit to a problem, visit challenge.gov and see if
you or a group of your friends and teammates might bring
creative ideas to a problem that’s now been organized
in a structured way by the government. So think about contributing or
participating in challenge.gov. Medium is actually
thinking about a problem that motivates you. And figure out,
what can I do if I were to ask for data
that’s not today available? Or find ways to tap into
other aspects of this, whether it be data standards,
or prizes and challenges, and find a way to
build my project. That is, something I
want the world to have. Powered in part by open
data, open government. Or three, heavy. I wouldn’t say leave Google,
but take a sabbatical. We encourage folks to sign
up as Presidential Innovation Fellows. We’re disciples of Eric Ries
lean startup methodology. And every year
now, we have dozens of entrepreneurs
coming in to Washington to spend six months
or a year really scrubbing in with
entrepreneurs inside government to go after
well-defined problems that the agency leadership
has said we really want your brain
power to help solve. And we just closed the
round of the current batch of Presidential Fellows, but
we’ll have more opportunity. And so think about
actually dedicating some time would be
my third option. Light, medium, heavy. MALE SPEAKER: That’s great. So one question about
technology outside of government and innovation
outside of government– the role government
can play in that. You talk about the narrative
you’re trying to change, which is not about big or small. It’s about just being smarter,
which sounds very noble. I hope we can do that. But I guess one question
is, when government tries to spark innovation
outside in certain sectors, through prizes or challenges,
whatever, and takes on risk that the market is
unwilling to take on, but it seems very unwilling
to accept those risks when they sometimes fail. We have a venture capital
group within Google and so we know sometimes they’ll
win, sometimes they fail. And we accept that. How does government, with
accountability and so forth, with a pluralist society deal
with that kind of outcome? ANEESH CHOPRA: So let me
start with a framework. In the first year of the
Obama administration, he unveiled the strategy
for American innovation– whitehouse.gov/innovation if
any of you want to read it. And it articulates
the role of government to foster innovation in society. And it starts at the foundation
of our society, which is we typically see
government play a role investing in infrastructure. And what we suggested in
this particular strategy is that roadways, railways, and
runways were the infrastructure of the 20th century economy. And that digital
is, in many ways, an important part of our
infrastructure in the 21st. So whatever we can do to
make it easier and better to build up our
digital infrastructure. The good news is
that’s bipartisan. We had a national wireless
initiative opening up airwaves for more commercial
services for broadband and opening up government
spectrum for commercial use as an example where both sides
of the aisle came together and said, let’s do this. We’re also using
that program to help build a next-generation
capability for cops and firefighters and emergency
services professionals to better communicate. The last recommendation
of the 9/11 Commission to allow for interoperable
communications. So get infrastructure right. Second, we need
rules of the road. So what are the rules of the
road on privacy and security to maximize the opportunity
for at least the digital side of the innovation economy? And again, we
embrace the framework of collaboration,
enforceable codes of conduct. So we’ve invited stakeholders
to think about ways to honor, for example, the Do Not Track
flag, which was not maybe the most successful
program was launched. But at least it was an enabling
part of the policy ledger so we could get more privacy
protection in the existing economy. But last but certainly
not least, we said there are a few sectors
where government needs to play at least a convening
role and maybe an investing role to catalyze breakthroughs
in health, energy, and education. And here, I wrote in the
book, I borrowed heavily from models in the
private sector for what we would call innovation
pipeline management. There’s actually a science to
bringing new ideas to life. You just don’t want
experimentation, throw some money here, see
if it works or doesn’t. There’s a methodology. And some organizations are
good about it and some are not. And I write in the book
about the negative case study of Kodak, which was
particularly awkward for me, because here I am
growing up in an era where Kodak is sort
of second to none. The most innovative
company in the world, invented the VCR, invented
digital photography, commercialized neither of
them because management chose not to invest in
these opportunities. They had a breakdown in their
innovation management pipeline. And what was
particularly awkward for us is that the
First Lady invited Mike Krieger, the
co-founder of Instagram, a Brazilian
immigrant, to her box. And as you know,
the year Instagram gets acquired for
some ungodly sum of money– a billion dollars,
whatever it is by Facebook, is the year Kodak
declares bankruptcy. And largely, on account of their
management failure to innovate. So my suggestion
was that we create the capacity for innovation
management in the same way. So even if we take a risk,
it’s understood and managed. We have, for example, now
in statute the Centers for Medicare and Medicaid
Innovation in the Affordable Care Act. Instantiating the authority
to experiment with new ways to pay doctors and hospitals
to keep people healthy. Not every idea that the
Innovation Center will launch will be successful, but
it’s designed into its DNA. And there’s actually a backstop. If they can prove that their
payment reforms improved quality and lowered cost, then
the actuary of the department, who’s not political, can
certify it met those conditions, now it can be a payment model
that’s available to all doctors and hospitals without having
to go back to Congress. Innovation pipeline management
an important aspect of this. And I think you’re going
to see some of that even with ARPA-E in energy. And we called for an
APRA-ED in education. Similar principles
to DARPA, which offers this kind of
innovation pipeline work for the defense department. MALE SPEAKER: Who
now works at Google. ANEESH CHOPRA: And all of
the DARPA-funded projects are all employees at Google. MALE SPEAKER: Exactly. Well, great. We’re going to open
this up to questions from the audience
in a few minutes. So if you have any questions,
start thinking about them. One question I had
to ask, you allude to a little bit in the book
because it was still hot in the press, which
is healthcare.gov. ANEESH CHOPRA: Yes. MALE SPEAKER: So I want
to get your opinion about what went wrong. And this is more of a
question about the operations of government’s
IT infrastructure and how it works. What do you think went wrong,
from your perspective now? And how we dealt
with the problem, and how we prevent that kind
of thing from happening again. ANEESH CHOPRA: This is
unbelievably frustrating, right? As an American,
as someone who has been a fan of President
Obama’s, like everyone, we were sort of shocked that
despite all the successes politically, the Supreme Court–
how could literally a website be the demise, if you
will, of the program? My early read was
that procurement remains a significant
problem in government. Everyone in this
room, if they wanted to participate in
government as an individual, they could volunteer their time. Like Guha, they
could actually help design a public-private
collaborative. But the minute money’s
involved, if someone wants to get paid for that work,
which is inevitable– you want to get paid for work–
then it triggers a whole machine called
the Federal Acquisition Regulations. And it’s become so
broken in Washington the Defense Department
wrote a memo and said, we almost have to avoid
fair and open procurements. They didn’t use that language,
I want to be careful about it. But that what happens today
is, if there was truly a fair and open competition,
then the culture is anyone who loses
immediately protests, which delays the project about
a year, on average. And the people who
suffer are the people who want the
program to go alive, not the vendors
fighting over each other about who won and
lost the contract. So my instinct– it
hasn’t been reported and I don’t know if the agency
will get that level of detail out. But my instinct is one of
the root cause problems is that they didn’t have a truly
fair and open procurement. They had to go to
a list of a dozen or so pre-qualified vendors who
were already allowed to do work for the government,
irrespective if they had any experience building an
exchange like healthcare.gov. So when you have a
limited pool of people that you can tap into because
of these acquisition problems, you have a bit of a
recipe for problems. And then you can go into all
of the other aspects of testing and so forth, which
really build off of that fundamental problem. But let me flip
the positive side. Healthcare.gov is one of
my favorite case studies in the book as an example
of an innovative state. The version that went
live in July of 2010. How many of you
knew that there was a version that went
live in July of 2010? No one’s hands are up. On the web, I’m sure you’re
going to say the same. Well, one of you did. OK, fine. One of you. MALE SPEAKER: Read the book. ANEESH CHOPRA: Read the book. Here’s the point, the bill
was signed into law March, I think 9 or 10, 2010. And the very first
IT deliverable was a comprehensive catalog of
all public and private health insurance options that had
never been amassed before. A data set that had never been
collected, let alone released. So that’s a little over 90 days. We didn’t have time
for a procurement. My successor, Todd
Park, was CTO at HHS. And we worked it out that
him and my colleague, Macon Phillips, would kind
of lead this startup to create healthcare.gov. Who was their lead designer? During the debate
on healthcare.gov, a young man from
New Jersey called Edward Mullen
thought that it would be better if he could put
his design talent to work to help explain to
the American people that this is what it will mean
for you if this law passes. And he created myhealth.gov,
a nice visual experience. And he just posted
a screenshot of it and tweeted at my colleague
Macon at the White House. Macon sees the at mention,
sees the image, and says, this is pretty good stuff. And long and short of
it, we bring Ed in. He was not a government
contractor, procurement process, just a talented person
passionate about the issue. Ends up becoming the lead
designer on healthcare.gov. They built the system
with openness in mind, so that when the data was
made available on the website, it was also available in machine
readable form for others. To this day, the US News and
World Report Health Insurance Finder is still
powered by the data from that original
healthcare.gov. The point is lots
of shots on goal. There should be millions of
websites that give people information about
their health care and let them find whatever
path is right for them through a broker,
through a website, through their neighbor,
or the government directly, to get the
care they deserve. You will see this
fall– my hope, a return to that principle, opening
up of those data assets so that more and more sites
can be built, not just one, to allow folks to shop. And that’ll be an example
of an innovative state. MALE SPEAKER: So a
different architecture. Do we have any questions
from the audience? While you’re thinking,
I will add one more. So on the international front. So I’d like to think the
US is a leading innovation country in many ways. I don’t know, are we
a leading innovator in state craft, or running
a country, or government operations? Other places we can
at least point to, like maybe ARPA-E or
something like that. ANEESH CHOPRA: Yeah. MALE SPEAKER: What
areas are we not? What countries do you see
doing really innovative things that we should learn from? ANEESH CHOPRA:
President Obama convened on the sidelines of
the UN General Assembly Meeting about 50 heads
of state back in 2010. Well, he issued a
challenge in 2010 and he began the
convenings in 2011, to say, let’s all do this together. Let’s create an international
movement of open government. And the only prerequisite
for signing up was you had the pledge to
commit to an open government plan that’s written
with your civil society. And that you got to publish
that plan in an open format. And countries all over
the world demonstrated their real powerful
case studies. Brazil has their people
participate in their budgeting process, setting priorities. One of my favorite examples
actually has to do with India and it is not IT-specific. They disclosed all
the folks that get paid– they have the equivalent
of a works progress– WPA, a Works Progress
Administration. People get paid to
do work as a way to create jobs in rural
villages throughout the country. And they decided to make an
open data set out of, what is it that these people
are paid to do? So I’m getting paid
to fix the well, or open up a water
fountain, or create something else in
the neighborhood. So they began publishing
the data sets. In villages in India,
there are NGOs, nonprofits, who take it and write with
chalk on a wall, literally, these are the people
in our village who have been paid
to do these things. And the level of accountability
that created– because they go, wait a minute. He was paid to build that,
but it’s not built yet. What are you doing? MALE SPEAKER: Or the
teacher didn’t show up. ANEESH CHOPRA: Or the
teacher didn’t show up. So you have open data even
without a digital interface in that case study. And you’ll see others moving
towards electronic voting and other ways to
strengthen democracy that I think are useful
to learn as case studies. And the good news is– what
I try to say in this book, this is less about
an ideological fight between the left and the
right even around the world. It’s really about a collective
capacity issue, which is, if we can become more
innovative in the delivery of services, let’s
have the debate what government should do. But once that’s
settled, let’s do it in the most efficient and
effective manner, which will inevitably involve a
digital aspect in today’s environment. And better to do that with
the opportunity for others to co-create and
build on that success. And America is leading in
that co-creation, co-building aspect. MALE SPEAKER: That’s great. And I think you give a
great historical context about where we come in this
process and where we’re going. I think the concern
would be that it may not be a bifurcated process. That you don’t first
decide what you want to do and then we always pick the best
way of doing it, because then people will try to game it. They lost the debate
in the first part, so they’ll undermine the
actual implementation. Kind of with healthcare.gov
was a little bit happening. The other side would be
happy that it was failing. Not because they want the
government to fail on its own. They just didn’t want us
to be doing that thing. ANEESH CHOPRA: The
most depressing aspect of that for me, just
to hit the point, there were seven states
early in the process that were awarded about a
quarter of a billion dollars to pre-load or pre-develop
modules of healthcare.gov that they pledged
to make reusable by other states, this
collective benefit opportunity. Two of those states
had an election in the midst of the grant cycle
where Republican leadership took over– Oklahoma and Kansas. The Oklahoma team,
the staffers who wrote the proposal on what
they envisioned the Oklahoma digital experience to be for
shopping for health care– in my humble opinion, I read it. I had no hand in
the selection of it. After they were awarded, I
went through and reviewed it and met the team. One of the most ambitious,
impressive, and bold, exciting examples of how they
were envisioning combining the healthcare.gov
concept with their existing services for the poor. Beautiful vision. But new governor is elected. Can’t touch Obamacare. Says, I am not accepting a
nickel of Obamacare money, which means this brilliant
idea that my staff developed that’s good for my
people, throw it away. And it was like $60 some-odd
million of investment that would’ve been wonderful
to build a better service to their folks. But it had Obamacare
on it and so she turned it away, as did
the governor of Kansas. And so two of the
seven states just really politicized what
professionals in those states had worked really
hard to want to do. And that’s disheartening. That part’s disheartening. MALE SPEAKER: Let me ask you–
I know you’re an optimist, but do you worry
about some paths that things can go in a bad way? I mean, people
always are worried about security and terrorism. ANEESH CHOPRA: Privacy. MALE SPEAKER: And privacy. So what’s an
example of something which may happen
which might undermine the movement towards open
data and more innovation, technology enabled– ANEESH CHOPRA: Health care. We have a longstanding tradition
that both sides of the aisle said, if we put more
digitization into health care, and we open up more data, and
we change the payment models to reward the kinds
of products that you need to create people– more
prevention-oriented mindset, that that’s the
formula for success. One of the challenges
has been our privacy laws across the 50 states are
pretty tight and variable. And the challenge
is, for the folks who want a simple solution
to health care information exchange, they’re
basically saying, avoid the privacy rules. Let’s just create this
one big health care cloud. Let’s put everyone’s
data in the cloud. And then, let’s
allow– so that if you get injured in an emergency
room and you’re unconscious, that that doctor can access
the cloud, find your records, and treat you well. Even though I understand
the value of that, I believe there’s too much
risk on the privacy front. Health care data, your
learning records data, these are really
sensitive data sets. I’m hopeful that–
what I’m fearful of is well-intentioned souls who
say, let’s put this data in the cloud and we’ll give
the data back to the consumer. But it’s still
sitting in the cloud without their
immediate permission. This story doesn’t
end well, with a data breach or a problem. What I’d rather do is have
kind of in these examples, people think about
new architectures. And the one like we said
in healthcare.gov, new architecture to me–
empowering patients with digital copies of their own
data going back to blue button, and then having
doctors do things like ask patients for
that blue button file– are critical to that success. MALE SPEAKER: I think having
some first– not first, but the next few steps
need to be cautious still as we go because there is a
risk of basically poisoning the well if we don’t do that. Well, great. Are there any questions
from the group? Please. Hey. Use the microphone if you would
because there’s an audience. AUDIENCE: Hello. ANEESH CHOPRA: Hello. AUDIENCE: Thanks for coming
in and speaking to us, and writing the
book, and working in this really important area. ANEESH CHOPRA: Thanks for
wearing a Sunlight Foundation shirt. AUDIENCE: It’s total
coincidence, actually. It was the only clean
shirt I had left. ANEESH CHOPRA: Exactly. AUDIENCE: So my
question is actually just your vision about what
the future of democracy looks like in a digital age. And by democracy, I mean
it in its purest form. So a group of people
coming together and trying to make a decision
about some sort of shared issue. And how does that look
different in the future? And as a little
bit of a context, I think it’s pretty
easy to observe as different industries,
as it were, digitize. It’s not just about
moving stuff online. Like the nature of the services,
the nature of the industry has absolutely changed. Just ask anybody who’s in
journalism or music, et cetera. But when we look at our own
democracy in a digital age, basically the way that
the processes work are exactly the same. So Congress is operating
basically the way it used to. Legislatures are the same. So really, how does this kind
of collective decision-making process, ie democracy, look
different in a digital age once it’s sort of disrupted
like other industries have been? ANEESH CHOPRA:
One attribute will be frictionless participation. What I mean by that
is we all have– we’re a nation of
300 million people with 3 million
government employees. And if we can embrace
the spirit that tapping into the 300
million, or 297 beyond the 3, is a critical path
to solving problems. One important way of
doing that is to make it frictionless participation. You might think of
that as government by APIs, as one technical model. But let me give you
a couple of examples because we’re starting to
see this come into play. In the wake of
Wall Street reform, huge political backlash
around big banks. Big fight in Congress
over new regulations to curtail, “too
big to fail,” and do other things– big
political fight. And in the law are
two pretty interesting new government roles that had
completely different outcomes on what happened on the
implementation side. One is the creation of a
Consumer Protection Bureau to advocate for
the American people to make sure that they’re
not swindled, if you will, on what my friend, now
Senator Elizabeth Warren, talks about tricks and
traps in the fine print. The other was the
establishment of a new rule basically banning proprietary
trading inside banks. It’s called the Volcker Rule. These are two of the more
controversial features of the Dodd-Frank bill. Well, what happened? The Consumer Protection Bureau,
because Elizabeth Warren was named as the first
sort of launcher of it, we met very early
in her appointment. And I sort of asked
her the question, would you like to lead the
last agency of the 20th century or the first agency
born in the 21st? And you know what she said. She’s a data queen. So she said, no, I
want to be more open, data driven, et cetera. So I dispatched my deputy
to kind of be an interim CTO as they got that
agency off the ground. And they launched a
crowd-sourcing program called Know Before You Owe. The idea was the
mortgage disclosure forms are complicated. And there are more than one. And one of the reasons people–
one potential reason why they bought bad
mortgages was they didn’t fully appreciate the
risks that they were assuming. So if we did a better
job of informing them through the mortgage forms
that they complete what the risks are that
they’re taking, maybe we could solve
the problem or prevent this happening in the future. 30,000 Americans visited
knowbeforeyouowe.gov and basically gave commentary
and feedback on not one, but two– kind of
like, am I hot or not? Remember that awkward thing? We created an, am I hot
or not for mortgage forms. Which form do you like? Which ones don’t you like? And give us your
feedback as to why. Tons of participation, which
meant long before the official rulemaking process
even happened, the president of
the United States could hold a copy of
the American-made form and say that this will be the
disclosure form going forward, and then eventually made
its way through regulation. In contrast, the
Volcker Rule, back-door secretive society kind of stuff. Technically, the American people
could have given feedback. It was available
on the internet. But like maybe 20 comments
were posted on that rule. Now, I don’t want
to be a cynic, but I think that 30,000 beats 20
on the quality of feedback and the ability to
get rules right. And that’s a metaphor for
all these other aspects of digital service, which is the
president had a review of all of the voting technologies
and voting problems. Why do we have to wait in
line for six hours to vote? If I can use my smartphone to
check-in to my flight, which has all the TSA sensitivity
and I can do that, why can’t I check-in to vote? So in fact, now there’s
an open collaboration to try to build those
technologies to allow folks to check-in to vote. And more will come. And it has a lot more
to do with understanding the potential of mobile,
cloud, and big data in ways that we hadn’t
imagined them in our democracy. So I have hope, but
I also understand that it’s going to
take time to get there. And that Dodd-Frank
bill is a study in those two different worlds. Other questions? Yes, please. AUDIENCE: I know
open data is quite vital for creating
innovation economy concept. Do you see open source software
having a role in that as well? ANEESH CHOPRA: Yes, so this
is an important question. I’m often asked,
should the government make a preference for open
source over proprietary code? And actually, I am on
the side of the folks that say no, we should
not take preference. And the reason is,
I don’t particularly care about who owns
the code base that facilitates social good. I care that they’re
open interfaces. So let’s just take
medical records again. I don’t necessarily think
that every hospital and doctor in America needs to
have an open source electronic medical
records software package. But what I want is whatever
proprietary systems exist, that they should give
every patient the ability to view, download, and
transfer that medical record file to whomever they
want frictionlessly. And that’s a difference, open
interface versus open source. Now, there’s a moment where open
source has its benefits, too. The VA, as a government agency
and sensitive to its cost structure, is
actually doubling down on its open source instance
of its Vista software package to help doctors
and patients have healthier experiences in the VA. And so you’ll see a
lot more opportunity to create modules to plug-in
aspects of new technologies into that environment. And they’re going to make
the code base open source so anyone can write to it. But that’s not something
I’d impose on the society as a whole, that’s a judgment
that that agency made based on its cost benefit analysis. And I think that’s the
right posture generally, when thinking about
how we solve problems. AUDIENCE: So
collaboration in the space of building the
software, but not– but like just to open data totally
separate kind of thing. ANEESH CHOPRA: Right. One of the points of
history in the book I write is that Herbert
Hoover– we’re not far from the Hoover
Institute– not normally cited as like the greatest
president of all time. In fact, quite the opposite. But when he was
Secretary of Commerce, issued a vision of the role
of government in society. Not picking winners and
losers, ie proprietary versus open source,
but to establish a mechanism for pre-competitive
R&D– my language. He called it the associative
state, where we work together. And that example was that
Boeing and Douglas at the time were airline manufacturers. After the end of
World War I, America wasn’t a leader in
airline manufacturing. European countries were
investing and subsidizing competitors. And we had some
fundamental problems. We didn’t know how to make air
foils the right way or engine cowlings. So Hoover says that
that’s pre-competitive R&D. Let’s fund it, but open
up the intellectual property so that both Boeing’s
247 and the DC-3 can compete on
what they do on top of that pre-competitive
component, so that their final product
is most certainly proprietary but it is powered
by this asset which was built in part by a
collaboration with government. So these are the nuances I
try to highlight in the book, at least as to my point of view. AUDIENCE: Yeah, thanks. ANEESH CHOPRA: Yes, sir. AUDIENCE: So I love the idea of
this opendata.gov initiative. Is there an important
difference we should understand between a newly created
government data set and a data set maybe created through
government funding? ANEESH CHOPRA: Great question. And in the spirit
of depression, this is the Aaron Swartz
story, which is to say government funds basic
research and development. And as a matter of
principle, that knowledge should be freely available. But because of the
way publishing works, there’s a private intermediary
that commercializes and funds all of the necessary work to
select the quality articles and to choose which
ones are worth publishing and all of that. And so I don’t bemoan
the fact that we need a part of our economy
to filter and to have that editorial role and to
produce these documents. But the tension is,
how do we open up the data in
government-funded work? And I think we struck
a happy medium, which is the publishers get to
retain a year of rights, at least that’s the
default setting. And that upon that
year coming up, it moves into more of an open
model– open access model. I think individual case examples
of government-funded research will have different
flavors of this. And there’s a beautiful
debate in our society around the degree
to which we want technology transfer
in commercialization. In some cases, agencies
will do it for free. I write the story
of Procter & Gamble that commercialized modeling
and simulation product out of the national labs
to help them remove particle dust from the facility. Procter & Gamble was able
to repurpose that technology into their diaper
manufacturing operation, improving their cash flow
about a billion dollars over the course
of several years. And they’ve gone back
and contributed back to the Department of Energy more
modeling and simulation tools that could be made
available for others. So there’s not a requirement,
but a collaboration mindset that created an opportunity. So you’re going to see different
examples that are there. As long as there’s a path
to have that conversation, I think we’re in safe hands. AUDIENCE: Could I ask
one quick follow-up? ANEESH CHOPRA: Please. AUDIENCE: Have you
guys thought about ways to solicit feedback about
what public data sets people would like to see? ANEESH CHOPRA: 100%. This is the most important
part of the journey because we started it–
to create the culture change, of course you have
President Obama saying, this is a priority. Day one in office sets this. In the midst of the
economic crisis says, I want to open up data. And I’m directing my newly
created CTO and my management and budget team to put rules
in place to make this happen. When we wrote the rules,
one of the key aspects was, you have like 45 days to
publish 3 high-value data sets. We didn’t care what they were. You have to say what you
deem to be high value. Now, it’s a joke, but the
CIA published their cafeteria menus. MALE SPEAKER: Which are very
important at Google, too. ANEESH CHOPRA: Very
important here at Google. We got to know which building
to go to for the Chinese menu and which for the Indian food. MALE SPEAKER: Google Data. ANEESH CHOPRA: Yeah, of course. But the point is we also
said that they must open up channels of participation so
that people can share feedback in. I give that a C on its
execution because there really hasn’t been a lot of demand. What you see is the
Sunlight Foundation, accountability
organizations that want spending data,
lobbyist data, sort of data on the political
side of government. They’ve been the most
aggressive in asking. So we publish the
White House visitor logs in response
to that request. You could see everybody who came
to visit me at the White House, for example. But for commercial use, there
aren’t as many people thinking in their mind, I want to build
a MOOC that has the ability to access records from the
Department of Education or states. And I want to incorporate
that into my algorithm to predict which
module of the lesson plan should I put in front
of which student based on my sense of their profile. I wish more folks said that
because that would open up a new conversation which
says, oh, well, we may not be collecting it in
that format today, but let’s work
together to see how we could exercise that muscle. The reason the book
to me is so critical, not only for the private
sector to understand it but in the government
is, if I work in the Department of Education
and my mission objective is that I want every child to
have access to a great math education, regardless of their
race and their income status. If you came to me
and said, I’m going to create a digital tutor
for every one of those kids, it would be really helpful if
I had access to such and such and such. Boy, that’s a different
conversation than, can I get a government
contract to go build something on your behalf? Or can you put a new
regulation in place where you have a little bit
more of the traditional levers of government? So C moving to B, maybe to
A as more of the ecosystem is aware of the possibilities
of new data sets to be made available. MALE SPEAKER: Any
other questions? It’s one last one. Are you worried about–
with the potential change of administration,
and potentially change of party, who knows,
what will come? Are you worried about some of
these advances being undone? Or is this inevitable,
this movement? ANEESH CHOPRA: So the
reason I’m so hopeful is that this is the one area
where both sides seem to agree. And just to give
you an example, I served as Virginia’s
Secretary of Technology, which was essentially a
cabinet-level position akin to what President Obama
set up at the White House. A Republican
governor of Virginia created the nation’s first
Secretary of Technology, carried on by
Democratic governor. And again, a
Republican governor. And again, a
Democratic governor. And now we have
this culture where that’s a novel idea
that’s instantiated into the DNA of what is deemed
by the objective measures best managed state in
America according to “Governing” magazine. And now at the federal
level, opening up data actually got a unanimous
vote in the Congress– a unanimous vote. Mark Warner and Eric Cantor, a
Republican member of the House, the majority leader
of the House, and a Democratic senator
both from Virginia don’t normally see eye to eye
and hang out and collaborate. They patroned this data
act, which in this case opened up financial transactions
data in the government on a standardized format. And the president
signs this into law. So my hope is
whomever is elected president next will
build on this foundation and we’ll start
to see a new army. I went to the Kennedy
School of Government, and I was never trained
on the power of open data as a policy lever. I was trained on, how do you
get Congress to take an action? How do you move the executive
branch in regulation? But this idea that
there could be ways to collect data
and disseminate it in a more friendly manner for
entrepreneurs and innovators, didn’t have a
single course on it. So I think as a new army
of government employees understand this as a viable
path to achieve their mission objective, you’re going to see
the whole ecosystem respond better. Entrepreneurs see the chance to
commercialize and make money. If you make a billion
dollars selling a startup who closed the
educational achievement gap, I would say thank you on
behalf of our country. We wouldn’t– oh, how dare you
make a billion dollars doing this? No, make two. Close the achievement gap. So we will see a new era of
entrepreneurs succeed in this. We’re going to see more
data sets come to life. And I hope we’re going to see
the political will to treat this as a very important–
and if not equal to, but a very important
lever in addition to the traditional models
of government spending and regulation. MALE SPEAKER: On that
very optimistic note and hopeful as well,
let’s end there. Thank you very much. ANEESH CHOPRA: Thanks for having
me, and I wish you all well. [APPLAUSE]