00:00:00.000 --> 00:00:00.067
[MUSIC]
00:00:00.067 --> 00:00:06.880
CHRIS WHITE: I have seen a lot of really bad
humanity. You know, I've seen terrorism and
00:00:06.880 --> 00:00:13.680
trafficking and sophisticated organized
crime, you know, designed to take things
00:00:13.680 --> 00:00:19.120
from people. Designed to hurt people. And I
just know that I can do something about it,
00:00:19.120 --> 00:00:25.280
you know, and that I have and that we can.
And so I think that that light, you know,
00:00:25.280 --> 00:00:37.563
the light of making a difference, you know, allows
me to hold and handle, you know, the dark parts.
00:00:37.563 --> 00:00:37.577
[THEME MUSIC]
00:00:37.577 --> 00:00:39.760
STANDARD INTRODUCTION: This is
the Microsoft Research Podcast,
00:00:39.760 --> 00:00:43.360
where Microsoft researchers—driving
advancement through fundamental science
00:00:43.360 --> 00:00:50.240
and technology research—explore the who,
how, and what’s next in computing and AI.
00:00:50.240 --> 00:00:51.126
[MUSIC ENDS]
00:00:51.126 --> 00:00:54.480
WEISHUNG LIU: Hi, and welcome. I'm Weishung Liu,
00:00:54.480 --> 00:00:58.320
a principal technical program
manager here at Microsoft Research,
00:00:58.320 --> 00:01:04.640
and today I'm talking to Dr. Christopher White,
who leads Microsoft Research's Catalyst Lab.
00:01:04.640 --> 00:01:09.520
Chris has one of those careers that
sounds almost made up: a Harvard postdoc,
00:01:09.520 --> 00:01:15.360
on-the-ground advanced tech lead during the
war in Afghanistan, a DARPA program manager,
00:01:15.360 --> 00:01:20.000
a dark web search pioneer, and a
public-good technologist. And his
00:01:20.000 --> 00:01:25.760
work has been featured in Popular Science as
well as recognized with honors, including the
00:01:25.760 --> 00:01:31.600
2016 Presidential Award for Extraordinary
Efforts to Combat Trafficking in Persons.
00:01:31.600 --> 00:01:35.920
But what I'm most interested in today
is the person behind that résumé. How
00:01:35.920 --> 00:01:40.240
someone who has seen some of the worst
as part of his work still comes to the
00:01:40.240 --> 00:01:45.680
office with a smile and a commitment to
building tools that make the world better.
00:01:45.680 --> 00:01:47.120
Chris, thanks for joining me.
00:01:47.120 --> 00:01:49.960
WHITE: Thanks for having me, Wei. Nice to see you.
00:01:49.960 --> 00:01:55.920
LIU: Yeah, good to see you, too. And so let's
get started. So I've known you for, I think,
00:01:55.920 --> 00:02:02.080
maybe around five or maybe six years, and now I
work with you in Catalyst Lab. I love your story,
00:02:02.080 --> 00:02:06.240
and when I tell people about you, I say
that you lead the newly formed Cat Lab,
00:02:06.240 --> 00:02:13.040
or Catalyst Lab, in Microsoft Research, you were
a senior executive at DARPA, and you are married
00:02:13.040 --> 00:02:19.120
to an incredibly accomplished woman and scholar.
And I personally think that your life story would
00:02:19.120 --> 00:02:25.920
make a very watchable Hollywood film. So when you
tell your own story, though, where do you start?
00:02:25.920 --> 00:02:32.400
WHITE: Well, you start from where you are.
And where I am today is leading research at
00:02:32.400 --> 00:02:39.360
Microsoft during a period of great fear and
hope about changing technology and society.
00:02:39.360 --> 00:02:46.720
LIU: But, like, that's now. There's clearly
lots of time from now into before, you know,
00:02:46.720 --> 00:02:51.600
the before times. So, like, tell me a little
bit about the before times. Like, where were you
00:02:51.600 --> 00:02:58.280
at—I think you grew up in Oklahoma maybe. Tell me
about that. Tell me about Chris White, the child.
00:02:58.280 --> 00:03:04.720
WHITE: Well, I was nerdy. I was a nerdy
Eagle Scout. [LAUGHTER] I liked to build
00:03:04.720 --> 00:03:11.120
my own computers and have local area network [LAN]
parties with my friends. But I also really liked
00:03:11.120 --> 00:03:18.000
being outside. I liked being in nature and camping
and rock climbing. Those were all important to me.
00:03:18.000 --> 00:03:22.960
LIU: OK, you mentioned LAN parties. Tell me
about … what did we play with our friends?
00:03:22.960 --> 00:03:27.120
WHITE: Well, you know, we played
everything. I preferred playing
00:03:27.120 --> 00:03:33.260
sports games, but I lost a lot of the other ones.
00:03:33.260 --> 00:03:37.360
LIU: [LAUGHS] Wait, so you played
like LAN party sports games, like,
00:03:37.360 --> 00:03:41.884
in basements? Like, with your
… were we all in the same … ?
00:03:41.884 --> 00:03:44.720
WHITE: Yeah. Attics. It was attics. Yeah,
not basements. [LAUGHTER] Mm-hmm. Yeah.
00:03:44.720 --> 00:03:49.440
Everybody had to bring their own computer or
bring their own Ethernet cable. And the more
00:03:49.440 --> 00:03:54.500
that you had your own machine that was higher
powered, the better you were going to play.
00:03:54.500 --> 00:03:56.000
LIU: So did you have the best machine?
00:03:56.000 --> 00:03:59.440
WHITE: I had the best machine.
I built the best machine. Until
00:03:59.440 --> 00:04:02.320
someone built a better machine.
But it was the best for a while.
00:04:02.320 --> 00:04:06.880
And I had a budget for it. And my family
didn't understand that the budget also
00:04:06.880 --> 00:04:13.440
included two giant speakers, like the size
of a computer tower, like subwoofer speakers.
00:04:13.440 --> 00:04:17.520
They didn't know that was part of
the computer building process. But
00:04:17.520 --> 00:04:22.600
it turns out that excellent audio-visual is a
critical part of a well-functioning machine.
00:04:22.600 --> 00:04:26.320
LIU: I love it. But, OK, we're going
to focus a little bit … continue
00:04:26.320 --> 00:04:30.640
focusing on childhood. So you mentioned nerdy kid,
00:04:30.640 --> 00:04:38.080
which means—I'm going to generalize here—debate
team or math competition team or chess team?
00:04:38.080 --> 00:04:43.680
WHITE: Oh, it was the debate team. Yeah.
Debate was a really wonderful activity.
00:04:43.680 --> 00:04:47.760
It taught me how to understand many
sides of an issue and taught me how
00:04:47.760 --> 00:04:51.560
to think about evidence and research in that way.
00:04:51.560 --> 00:04:58.240
LIU: So, Chris, take me back again. Like,
we're talking a little bit about Oklahoma,
00:04:58.240 --> 00:05:03.920
and, you know, at some point, you chose
researcher. Like, I—for example—when I was a kid,
00:05:03.920 --> 00:05:09.680
I was convinced I would be an architect. I am
not an architect as you can see. But did you
00:05:09.680 --> 00:05:14.800
always think you were going to be a researcher?
Was there an epiphany there? Like … or was it
00:05:14.800 --> 00:05:18.713
always kind of like a slow hunch that you
were going to pursue research as a career?
00:05:18.713 --> 00:05:24.080
WHITE: Yeah, I kind of always felt like
it. You know, as I was applying for school,
00:05:24.080 --> 00:05:28.720
those were the kinds of jobs I was interested
in. You know, the kinds of things that you do,
00:05:28.720 --> 00:05:33.280
you know, once you finish degrees that have
research in them. Those were the people,
00:05:33.280 --> 00:05:38.800
when I visited their work, I was inspired by.
The people that were discovering, you know,
00:05:38.800 --> 00:05:43.040
that were inventing, that were innovating,
that were … they're pushing the boundaries
00:05:43.040 --> 00:05:47.440
that were kind of in that unknown, you
know, space and making sense of it.
00:05:47.440 --> 00:05:49.600
And so I've always had that curiosity, you know,
00:05:49.600 --> 00:05:57.840
and that approach to discovery. And so as I
increased my education, it just changed the
00:05:57.840 --> 00:06:05.600
nature of what research meant and the ways
in which I imagined that process evolving.
00:06:05.600 --> 00:06:11.120
LIU: Was a doctorate always on the
cards then? Or did that come, like,
00:06:11.120 --> 00:06:15.120
as you were in college and then gradually learned
and thought, you know, I could keep going?
00:06:15.120 --> 00:06:18.960
WHITE: Yeah, you know, I didn't know
about doctorates really when I went to
00:06:18.960 --> 00:06:24.960
college. I wasn't as familiar with higher
education. And so I knew that I wanted,
00:06:24.960 --> 00:06:28.960
you know, advanced degrees. And I knew that
I wanted more than was being taught in the
00:06:28.960 --> 00:06:34.000
undergraduate level. But I didn't know, you know,
for example, that when you apply to PhD programs,
00:06:34.000 --> 00:06:38.480
that often you get paid to go. You know,
you get admitted and you get a stipend.
00:06:38.480 --> 00:06:44.000
You know, I was familiar with law and medicine
and that you take out loans and you have to
00:06:44.000 --> 00:06:49.360
do it this way. And so for me, it was a
surprise to know that I could apply to PhD
00:06:49.360 --> 00:06:55.200
programs directly from college. And from
then, I just really was excited, you know,
00:06:55.200 --> 00:07:02.080
that I could explore different universities
and different areas of focus and build on,
00:07:02.080 --> 00:07:06.640
you know, strong technical foundations
of math and science and physics.
00:07:06.640 --> 00:07:13.520
And it was really amazing mentors that helped
me and opened doors and inspired me. You know,
00:07:13.520 --> 00:07:19.200
mentors at Oklahoma State University, Johns
Hopkins, and Harvard, really people that
00:07:19.200 --> 00:07:23.800
took their time and helped me. I'm
really grateful for my teachers.
00:07:23.800 --> 00:07:32.420
LIU: So let's talk about your first mentor
of your life, and I'm going to assume … Dad?
00:07:32.420 --> 00:07:34.520
WHITE: Yeah, Dad’s a big influence.
00:07:34.520 --> 00:07:40.800
LIU: OK, Dad's a big influence. And so we're going
to, I'm going to go down this path. So Catalyst
00:07:40.800 --> 00:07:46.240
Lab landed a huge commitment from our executive,
kind of, leadership team on a project that has
00:07:46.240 --> 00:07:52.320
been, I think, like a decade in the making. And
then … so we all, I know we all came together on
00:07:52.320 --> 00:07:57.440
a call. We're like, “Wow, let's debrief. Let's
talk about it.” Like, oh, energy's up. And
00:07:57.440 --> 00:08:03.243
then at the very end of the meeting, you said,
“OK, now I'm going to go clean some toilets.”
00:08:03.243 --> 00:08:03.275
WHITE: [LAUGHS] Mm-hmm. Mm-hmm.
00:08:03.275 --> 00:08:08.640
LIU: And I think all of us kind of
thought, like, wha-what does that
00:08:09.280 --> 00:08:14.560
mean? So before we get kind of more into
the specifics of mentorship and the,
00:08:14.560 --> 00:08:22.400
like, follow-on in your career, what— please tell
me—does cleaning toilets have to do with success?
00:08:22.400 --> 00:08:26.400
WHITE: Well, you know, like I said, Dad's a
big influence, and he taught me so many things,
00:08:26.400 --> 00:08:29.920
and he taught me about respecting
people. And most importantly,
00:08:29.920 --> 00:08:34.480
he taught me about humility and
service. And so his way of making
00:08:34.480 --> 00:08:37.920
sure I understood that was every time
something really good happened to me,
00:08:37.920 --> 00:08:44.560
he'd remind me to go clean some toilets and,
sort of, embody that. And so I did. I did it
00:08:44.560 --> 00:08:50.240
last weekend actually. I went to a community
event and volunteered to do bathroom clean.
00:08:50.240 --> 00:08:56.400
LIU: Is this like … are you … so in, like, the
household chores, are you, like, the bathroom
00:08:56.400 --> 00:08:59.109
cleaner? Like, is this just always your job
when we're dividing up chores in the house?
00:08:59.109 --> 00:09:04.320
WHITE: No, no, no. We trade off, and also we
have a complicated household, and so we have
00:09:04.320 --> 00:09:10.240
a lot of help that also helps with cleaning.
And so this is something that I choose to do,
00:09:10.240 --> 00:09:14.400
especially if I can do it in a community
context, because it again reminds me,
00:09:14.400 --> 00:09:20.160
you know, about humility and about service and
of respect for other people and other jobs.
00:09:20.160 --> 00:09:26.880
LIU: So I think this … there's some Eagle Scout
vibes that I'm getting from here, as well,
00:09:26.880 --> 00:09:35.280
based on sort of humility and service. What
is the Boy Scout badge you are most proud of?
00:09:35.280 --> 00:09:43.120
WHITE: Oh, that's really hard. Each badge is
something that you earn, including the Eagle
00:09:43.120 --> 00:09:48.320
Scout badge. And the Eagle Scout badge for me
was maybe the most difficult to earn because
00:09:48.320 --> 00:09:54.320
it requires an Eagle project. And I was really
close to my Eagle project. I was, I was … had
00:09:54.320 --> 00:10:01.440
it set up and it was ready to go. And it was in
this church in downtown Oklahoma City in 1995.
00:10:01.440 --> 00:10:08.378
And right as I was about to do this project, the
Murrah Building blew up in a terrorist incident, …
00:10:08.378 --> 00:10:08.389
LIU: Whoa.
00:10:08.389 --> 00:10:14.480
WHITE: … and it blew up part of the church. And,
therefore, my Eagle project was not realistic, and
00:10:14.480 --> 00:10:21.760
it was not a priority, and it got shelved. And it
took me several years after that to then come back
00:10:21.760 --> 00:10:28.538
to plan and execute an ambitious Eagle project
right before I turned 18. And so that badge, …
00:10:28.538 --> 00:10:30.640
LIU: Wow.
WHITE: … you know, completing the project,
00:10:30.640 --> 00:10:36.400
overcoming significant setbacks, including
being a teenager and having other interests,
00:10:36.400 --> 00:10:41.000
to come through and finish it,
I'm really proud of finishing it.
00:10:41.000 --> 00:10:48.080
LIU: That's awesome. So you … so that was, so
there were a few years then in between when
00:10:48.080 --> 00:10:55.200
you ideated, had the idea for the project,
and then … I appreciate the story around
00:10:55.200 --> 00:11:02.880
resilience because of when you completed the
project. Like, what is the biggest takeaway
00:11:02.880 --> 00:11:08.160
that you took from there? Like what is that
big lesson? You turned 18, you finished it,
00:11:08.160 --> 00:11:11.600
you celebrate. Actually, you know, maybe
I'll ask you, how did you celebrate?
00:11:11.600 --> 00:11:17.760
WHITE: Oh, with my family. You know, the Boy
Scouts is a long process. It's many years, many
00:11:17.760 --> 00:11:23.440
kinds of roles, many kinds of learning, and to
finish it—I finished it with another of my fellow
00:11:23.440 --> 00:11:28.240
scouts that had been together for a long time.
And so our families, we had a ceremony, and it
00:11:28.240 --> 00:11:32.480
was something we could celebrate. And every time I
think about it, I think about the things that you
00:11:32.480 --> 00:11:37.840
keep repeating when you attend the scout things,
like the motto, “being prepared,” and the, kind
00:11:37.840 --> 00:11:42.720
of, laws about being respectful and courteous.
And those are things that continue to be,
00:11:42.720 --> 00:11:49.960
you know, principles that influence my priorities
and also inspire me and help me in my job.
00:11:49.960 --> 00:11:55.520
LIU: Before we get to your job then, we're
going to talk about this interim period
00:11:55.520 --> 00:12:04.000
where I'll just say you chose to go to war and a
few other things in between. Let's jump to 2010,
00:12:04.720 --> 00:12:09.360
when you joined DARPA, which is the
Defense Advanced Research Projects Agency,
00:12:09.360 --> 00:12:13.840
within the Department of War. And then, I
think, very quickly afterwards you were sent
00:12:13.840 --> 00:12:19.360
to Afghanistan to help implement a new
approach to wartime analysis. And then
00:12:19.360 --> 00:12:29.040
… but before that, I think you were a Harvard
postdoc, and you had actually mentioned it was
00:12:29.040 --> 00:12:35.360
a mentor who encouraged you to attend a meeting
with DARPA that resulted in this course change.
00:12:35.360 --> 00:12:43.280
And so now I'm super curious, how did
this conversation happen where you were,
00:12:43.280 --> 00:12:46.880
you know, promising Harvard postdoc,
which became a mentor going, “Hey,
00:12:46.880 --> 00:12:50.960
there's something happening. You should
go attend.” Like, what happened there?
00:12:50.960 --> 00:12:57.760
WHITE: Yeah. DARPA invests taxpayer
money, and they have a record of
00:12:57.760 --> 00:13:01.360
really creating innovations, massive
innovations. You know, the internet,
00:13:01.360 --> 00:13:07.920
GPS, stealth, UAVs. And at the time, there
was different ways of building programs,
00:13:07.920 --> 00:13:14.080
and one of the kind of tested ways was to
bring together different fields of expertise.
00:13:14.080 --> 00:13:21.520
And so we were in wartime in Afghanistan.
DARPA was not involved very much. And there
00:13:21.520 --> 00:13:28.080
was a question: you know, could we bring
together social scientists with math people,
00:13:28.080 --> 00:13:34.640
computer science and defense experts, and
together could they help us make sense of
00:13:34.640 --> 00:13:41.600
what's happening in the data? Could that be useful
to prevent violence and to protect our soldiers?
00:13:41.600 --> 00:13:46.080
And it was an experiment because it required
all these different areas of expertise coming
00:13:46.080 --> 00:13:53.200
together—from sensors through analysis
through social behavior—and a culture
00:13:53.200 --> 00:14:00.960
change. And so the summer was a Skunk Works to
understand, was there something to contribute?
00:14:00.960 --> 00:14:04.480
And at the end of it, they asked
me, they said, “Well, you know,
00:14:04.480 --> 00:14:08.720
do you believe in your work?” And I was
like, “Yeah.” You know, like, “Well,
00:14:08.720 --> 00:14:15.200
do you … can you prove it? You know, can you
go see if it can help protect our soldiers
00:14:15.200 --> 00:14:21.280
and help prevent violence?” And so they
asked me to be of service, and I said yeah.
00:14:21.280 --> 00:14:28.000
And it was something that I was really unprepared
for. I was supposed to go for nine days the first
00:14:28.000 --> 00:14:32.880
time, and on the third day, I got a battlefield
promotion and was put in charge and then ended
00:14:32.880 --> 00:14:39.920
up staying for 90 days and then went back a few
times after that even. And so, you know, our work
00:14:39.920 --> 00:14:44.560
was towards … oriented towards protecting, you
know, soldiers and toward preventing violence
00:14:44.560 --> 00:14:53.840
and using AI and advanced technology to do
that. And it was a, really, a big opportunity.
00:14:53.840 --> 00:15:02.480
LIU: So you attended a meeting, you contributed
some work, and then, you know, “Do you believe
00:15:02.480 --> 00:15:09.360
in your work? Prove it.” What was the time frame
between, you know, like, showing up at meeting,
00:15:09.360 --> 00:15:15.009
participating in some of the research before,
like, “prove it”? Was it years, or was it months?
00:15:15.009 --> 00:15:19.840
WHITE: The … no, no, the … so the Skunk Works
was a summer. It was like, you know, 40 days of,
00:15:19.840 --> 00:15:25.040
kind of, like focused effort. But at the end
of it, I was expecting to go back to Harvard
00:15:25.040 --> 00:15:31.120
and continue my postdoc. But instead, you
know, they sat us down; they said, “Look,
00:15:31.120 --> 00:15:37.520
you know, we think this can make a difference.
You know, do you think so?” And when I said yeah,
00:15:37.520 --> 00:15:42.400
from then to getting over there was like
a week. It was an extremely fast ...
00:15:42.400 --> 00:15:43.029
LIU: What?
00:15:43.029 --> 00:15:49.960
WHITE: … turnaround. It was extremely
urgent and, you know, unexpected.
00:15:49.960 --> 00:15:56.000
LIU: OK, so, like, how … so you
said, yes, like I believe in my
00:15:56.000 --> 00:16:02.560
work. They gave you … they essentially
said, OK, then go next week. And how,
00:16:02.560 --> 00:16:07.480
how does somebody prepare for an experience
like that? If you can even prepare.
00:16:07.480 --> 00:16:11.440
WHITE: Yeah, I don't know. You know,
this was an experiment. You know, sending
00:16:11.440 --> 00:16:18.560
researchers—even having a presence from DARPA—in
the war zone, it was new. And the idea was there
00:16:18.560 --> 00:16:23.840
could be a difference made because technology
was advancing quickly. And that meant that many
00:16:23.840 --> 00:16:28.640
of the mechanisms in place for other ways of
getting prepared were not available. And there
00:16:28.640 --> 00:16:33.600
was a bunch of training, like crash courses in
training on what to do in certain situations,
00:16:33.600 --> 00:16:40.560
how things work, you know, how to handle yourself
in different ways. But it was … it was very brief.
00:16:40.560 --> 00:16:44.960
And so there wasn't a lot of preparation
because also we were planning to be there
00:16:44.960 --> 00:16:49.360
for nine days. But because there was
an opportunity to do something good,
00:16:49.360 --> 00:16:56.320
that extended, and there was a lot of learning
on the job and a lot of support in theater,
00:16:56.320 --> 00:17:02.520
a lot of people really willing to
help us because we were helping them.
00:17:02.520 --> 00:17:09.920
LIU: So, OK, so you talked about lots of people
there, lots of people supporting. I'm not going to
00:17:09.920 --> 00:17:14.640
focus too much on your time in Afghanistan because
that was covered really well in a Popular Science
00:17:14.640 --> 00:17:23.504
article. But I do have to ask, can you tell me the
origin story of the nickname Dr. Spaghetti Man?
00:17:23.504 --> 00:17:30.000
WHITE: [LAUGHS] Sure. Sure. You know, I'm
a long-time yoga practitioner. You know,
00:17:30.000 --> 00:17:39.440
20-plus years. It's how I keep myself calibrated.
And in Kabul, I was practicing yoga in the gym,
00:17:39.440 --> 00:17:44.480
and all these people were staring at me. You know,
they were staring at me before their … they had
00:17:44.480 --> 00:17:49.840
a class; they had, like, a combatives class, a
kind of different judo-mixed-martial-arts class.
00:17:49.840 --> 00:17:52.720
And, you know, I had a little bit of
training, and so after a couple of times,
00:17:52.720 --> 00:17:59.840
I decided I'd just join in. And I joined in,
and I won a lot. I won a lot of the matches.
00:17:59.840 --> 00:18:05.600
And one of the matches I won, turns out, it
was against the Marine who was guarding the
00:18:05.600 --> 00:18:11.760
entrance to the building I was working
in across the headquarters. And he told
00:18:11.760 --> 00:18:19.840
everyone. He told everyone, you know, that
this, like, lanky spaghetti guy, you know,
00:18:19.840 --> 00:18:25.280
like, got out of things and beat him, and
he did it in such a way that everyone in
00:18:25.280 --> 00:18:33.280
the building knew. And so I got all this
respect, and I got that cool nickname.
00:18:33.280 --> 00:18:39.280
LIU: Did other folks challenge you as a
result of this Marine telling your story?
00:18:39.280 --> 00:18:44.480
WHITE: I mean, everyone that came to that
class was game, you know. And I didn't,
00:18:44.480 --> 00:18:48.800
you know, I couldn't beat everybody, but I
held my own, and that was really surprising
00:18:48.800 --> 00:18:53.520
to a lot of people. And again, once you
establish yourself in different ways,
00:18:53.520 --> 00:18:57.960
you know, there is a lot of
respect and, um, camaraderie.
00:18:57.960 --> 00:19:05.200
LIU: So, you know, tell me, what did
that whole experience teach you about,
00:19:05.200 --> 00:19:09.600
like, a life lived, and how do those
lessons show up in your work today,
00:19:09.600 --> 00:19:13.920
including the, you know, Spaghetti Man experience?
00:19:13.920 --> 00:19:18.320
WHITE: You know, the whole thing,
it taught me about, you know,
00:19:18.320 --> 00:19:25.280
service and what it means to serve,
and it taught me about the reality of
00:19:25.280 --> 00:19:31.720
both peace and violence in the world. And that
continued to motivate me to make a difference.
00:19:31.720 --> 00:19:38.320
LIU: After your time in Afghanistan, you
continued on with DARPA as a program manager,
00:19:38.320 --> 00:19:43.840
and in that role, you developed Memex,
which I believe is a set of tools that
00:19:43.840 --> 00:19:49.200
could be mixed and matched to help uncover
some really shady and dark parts of the web,
00:19:49.200 --> 00:19:53.760
which from what I understand was
really hard to do up until that point.
00:19:54.320 --> 00:20:03.200
I think this work also is what helped you … what
you received your 2016 presidential award on. And,
00:20:03.200 --> 00:20:09.760
you know, like, can you … how do you explain
the Memex project and this dark web work to
00:20:09.760 --> 00:20:16.080
someone outside tech? And what about this project
inspired or appealed to you to even pursue it?
00:20:16.080 --> 00:20:22.480
WHITE: Yeah, well, the inspiration started when I
was overseas because I saw the benefit of evidence
00:20:22.480 --> 00:20:29.840
when it came to stopping bad things from happening
and prosecuting them. And coming back to the US,
00:20:29.840 --> 00:20:35.280
I saw more and more opportunities where
collecting evidence at large scale
00:20:35.280 --> 00:20:40.080
was a really useful mechanism. And so, you
know, people think of search engines a lot,
00:20:40.080 --> 00:20:45.920
and search engines are a great tool for finding
something, but when you need to connect the dots,
00:20:45.920 --> 00:20:51.520
there wasn't a great tool that could work at
the scale of the web and connect the dots,
00:20:51.520 --> 00:20:56.400
especially some of the dots being in
difficult places that were not very visible.
00:20:56.400 --> 00:21:01.520
And so by creating a technology that would
connect the dots across parts of the internet,
00:21:01.520 --> 00:21:06.000
it allowed the prosecution of these kinds
of really awful [trafficking] crimes,
00:21:06.000 --> 00:21:11.760
and so for example, in New York City, it allowed
them to increase their prosecutions by 10 times
00:21:11.760 --> 00:21:15.680
because it gave them new evidence and different
kinds of evidence than they had before,
00:21:15.680 --> 00:21:21.160
and they could use that in a way to
really deal with the issue differently.
00:21:21.160 --> 00:21:26.480
LIU: OK, from what I understand, the
dark web is a pretty—ha-ha—dark place,
00:21:26.480 --> 00:21:32.720
[LAUGHS] and so, like, you've worked in
some very dark spaces and close to some
00:21:32.720 --> 00:21:38.160
very dark uses of the technology. So, like,
how again, like, you and I interact a lot,
00:21:38.160 --> 00:21:47.040
and so it's always curious to me, how has that
not made you cynical and how do you stay hopeful?
00:21:47.040 --> 00:21:52.880
WHITE: You know, I think what it's done is
it's really increased my motivation. You know,
00:21:52.880 --> 00:21:55.840
my motivation to make a contribution. You know,
00:21:55.840 --> 00:22:03.440
my motivation to stay aware and be realistic.
And by being realistic and motivated to
00:22:04.880 --> 00:22:10.800
address some of these very challenging issues,
you know, around violence, around harm,
00:22:10.800 --> 00:22:18.640
and instead to focus on empowerment and to
focus on justice, it gave me something to do.
00:22:18.640 --> 00:22:28.240
And so I have seen a lot of really bad humanity,
you know. I've seen terrorism and trafficking and
00:22:29.120 --> 00:22:33.600
sophisticated organized crime, you know,
designed to take things from people.
00:22:33.600 --> 00:22:38.720
Designed to hurt people. And I just know
that I can do something about it, you know,
00:22:38.720 --> 00:22:44.720
and that I have and that we can. And so it's
really been inspiring to, you know, for example,
00:22:44.720 --> 00:22:49.680
come to Microsoft, where there is a mission
of empowerment and there is that perspective
00:22:49.680 --> 00:22:56.320
that technology can make a difference.
And so I think that light, you know,
00:22:56.320 --> 00:23:05.040
the light of making a difference, you know, allows
me to hold and handle, you know, the dark parts.
00:23:05.040 --> 00:23:10.240
LIU: I mean, that's a very powerful message,
right? Because even with the darkest parts of
00:23:10.240 --> 00:23:17.920
humanity, there is light, or, you know, you
cannot have darkness without light, as well.
00:23:17.920 --> 00:23:22.320
I think there may have been a 60 Minutes
interview that included a short snippet
00:23:22.320 --> 00:23:27.440
of you talking about your work, and you
had shared with me that—which, by the way,
00:23:27.440 --> 00:23:34.480
I cannot find the 60 Minutes episode,
and I desperately want to. But, you know,
00:23:34.480 --> 00:23:41.920
I heard perhaps through you, Chris, that that
60 Minutes episode actually contributed to you
00:23:41.920 --> 00:23:48.960
coming to Microsoft maybe or, you know, getting a
job at Microsoft, and that made me curious. Like,
00:23:48.960 --> 00:23:57.160
after some really incredible experiences, after
all of that, why did you choose Microsoft?
00:23:57.160 --> 00:24:05.280
WHITE: Oh, it was a … it was a really good
choice. And at the time, I had a lot of options,
00:24:05.280 --> 00:24:13.040
and I was looking around, and this mix of the
mission of the company—this mission of empowering
00:24:13.040 --> 00:24:19.760
everyone—and the scope of the company, you know,
being, you know, IT for the world, you know,
00:24:19.760 --> 00:24:24.880
platforms and infrastructure for the world,
and my confidence in its leader. You know,
00:24:24.880 --> 00:24:30.000
the CEO was part of my interview process.
And, you know, that together gave me,
00:24:30.880 --> 00:24:33.600
you know, this perspective that I
could do something here. You know,
00:24:33.600 --> 00:24:40.400
that there … I could start to envision moving
from, you know, for example, very high-powered,
00:24:40.400 --> 00:24:47.760
expensive lasers to table knives. They
both cut, and the world needs both of them.
00:24:47.760 --> 00:24:53.520
LIU: One of the things that you talk a lot
about is problem selection. And I think
00:24:53.520 --> 00:24:58.960
sometimes that also gets referred to as taste
these days. So how do you think about problem
00:24:58.960 --> 00:25:05.760
selection in Microsoft Research, in Catalyst
Lab, given the breadth of what Microsoft does?
00:25:05.760 --> 00:25:10.480
WHITE: Oh, I think it's just the most
important part of research. You know,
00:25:10.480 --> 00:25:16.960
it's the place where the angle of the whole
thing changes, and it allows you to start to
00:25:16.960 --> 00:25:21.680
imagine a different future, you know, or to work
backward from that future that you want to see
00:25:21.680 --> 00:25:27.600
and see what you can do to contribute to it now.
And the amazing thing about Microsoft Research
00:25:27.600 --> 00:25:33.840
is there's tremendous freedom, and so there's
tremendous creativity in problem selection.
00:25:33.840 --> 00:25:39.760
And it's the kind of thing where you have
that latitude to go from idea to evidence
00:25:39.760 --> 00:25:45.040
to this potential for growth and thriving
and along the way get feedback about value
00:25:45.040 --> 00:25:54.000
and usability and benefit or policy or
adaptability, integration. And so its
00:25:54.000 --> 00:26:01.200
scope provides so many vehicles for growth
and utility of advanced research ideas.
00:26:01.200 --> 00:26:07.360
And its mission, you know, allows the focus of
those to address absolutely critical problems. You
00:26:07.360 --> 00:26:14.400
know, problems of, you know, impending security
challenges due to AI, problems of advanced,
00:26:14.400 --> 00:26:21.520
you know, memory and storage; it allows for, you
know, personalization automation, just, you know,
00:26:21.520 --> 00:26:30.640
for example. And I think that this mix of freedom
combined with responsibility and mechanisms,
00:26:30.640 --> 00:26:37.760
you know, to prove out and demonstrate value,
that that combination is just really remarkable
00:26:37.760 --> 00:26:44.320
and it really motivates me to
take the opportunity seriously.
00:26:44.320 --> 00:26:52.000
LIU: So it's almost going to be like asking you
about your favorite kids now. What are some, what
00:26:52.000 --> 00:27:00.080
are some accomplishments in Microsoft Research or
Catalyst Lab that you are particularly proud of?
00:27:00.080 --> 00:27:07.760
WHITE: Oh, you know, I think that, you know, right
now, for example—again, focusing on kind of where
00:27:07.760 --> 00:27:16.240
we are—I think that we have a real opportunity to
contribute something in the cyber defense space.
00:27:16.240 --> 00:27:25.280
I think our work on security and AI
and hacking, it's really in, you know,
00:27:25.280 --> 00:27:31.760
based on such a deep expertise and also based
on, you know, people with a lot of experience,
00:27:31.760 --> 00:27:36.640
and I think it's growing in need, you know,
the need to protect our digital infrastructure,
00:27:36.640 --> 00:27:44.800
the need to deter and interdict people that would
like to break systems or cause harm, that's really
00:27:44.800 --> 00:27:51.840
growing right now. And so being able to show up
in a way that creates breakthrough that can scale
00:27:51.840 --> 00:27:58.000
with relevance to address that growing need, it's
really something that's, you know, awesome to me.
00:27:58.720 --> 00:28:03.200
And also I think that again kind of extending
from the, sort of, Memex conversation,
00:28:03.200 --> 00:28:08.640
where we're trying to connect the dots, you know,
these days, trying to analyze lots of content as
00:28:08.640 --> 00:28:12.960
part of being more efficient, as part of being
more creative, as part of being more capable,
00:28:12.960 --> 00:28:19.040
you know, using tools to augment that human
taste and creativity and ingenuity, that's
00:28:19.040 --> 00:28:24.640
really never been as powerful as it is before
and continues to grow. And so I think making
00:28:24.640 --> 00:28:29.360
contributions in these areas, you know, where
we have state-of-the-art technology and we're
00:28:29.360 --> 00:28:35.200
creating changes in the way systems are getting
built and used, to me, it shows that, you know,
00:28:35.200 --> 00:28:43.040
despite the fast pace of change and despite the
extreme competitiveness in the technology sector,
00:28:43.040 --> 00:28:49.760
that specialized expertise with good problem
selection and an environment to grow and get
00:28:49.760 --> 00:28:55.600
feedback can create the kind of breakthroughs
that matter. And so I really like that.
00:28:55.600 --> 00:29:00.480
Maybe to give one more example from the past.
Just maybe something emblematic of the group
00:29:00.480 --> 00:29:08.800
and some of my predecessors was this underwater
datacenter called Project Natick, [LAUGHTER] where
00:29:08.800 --> 00:29:13.280
this experiment, you know, put a datacenter in a
submersible on the bottom of the ocean connected
00:29:13.280 --> 00:29:18.560
to a wind farm, and it tested this design about,
could you build something that didn't require
00:29:18.560 --> 00:29:24.080
human intervention? Could it be more efficient
with power? Could it be more efficient with heat?
00:29:24.080 --> 00:29:29.360
And the program manager that created that is
one of the best I've ever worked with. And it
00:29:29.360 --> 00:29:35.920
just performed incredibly well as this
huge surprise, and it led to us being,
00:29:35.920 --> 00:29:38.400
you know, named the most innovative company
[World Changing Company of the Year] in 2021
00:29:38.400 --> 00:29:44.160
and other awards. And it's just an example
of really out-there thinking, imagining a
00:29:44.160 --> 00:29:51.360
world that had challenges with water access
and power access and land but where we need,
00:29:51.360 --> 00:29:57.680
you know, computation and data storage
and technical capability available.
00:29:57.680 --> 00:30:00.960
So a little bit out there, but
I really … I feel like we have
00:30:00.960 --> 00:30:04.640
to be imagining these different
futures. We have to be radical
00:30:04.640 --> 00:30:11.200
in that way so that we can start to
work back and make contributions now.
00:30:11.200 --> 00:30:19.600
LIU: I love the Natick example because I feel
like Natick is the gift that keeps on giving. And,
00:30:20.480 --> 00:30:27.120
you know, if I can digress for a minute, I
remember that I think one of the first TikTok
00:30:27.120 --> 00:30:35.600
videos that the Microsoft account posted
was of Natick, but it was using, like,
00:30:35.600 --> 00:30:41.920
a meme at the time which had, like, human eyes
and a mouth being lowered into, uh, and it was
00:30:41.920 --> 00:30:47.680
of Natick being lowered into the ocean, including
like a Clippy, like a real-life Clippy swimming
00:30:47.680 --> 00:30:55.760
across the … it was … I still love that video for
reasons why I don't understand. [LAUGHTER] But if
00:30:55.760 --> 00:31:00.214
you haven't seen it, highly recommend. Maybe we
can link it [LAUGHS] if the video still exists.
00:31:00.214 --> 00:31:04.703
WHITE: Definitely. Look, look, one
of the highlights was the fish cam.
00:31:04.703 --> 00:31:04.713
LIU: Yes!
00:31:04.713 --> 00:31:09.200
WHITE: You know, people were concerned that it
might disrupt the marine ecosystem. And we did
00:31:09.200 --> 00:31:15.983
a lot of preparation and study for that. And just
to be sure, we put cameras and watched it happen.
00:31:15.983 --> 00:31:15.993
LIU: Yep.
00:31:15.993 --> 00:31:20.480
WHITE: And all these fish kept coming around
and making it home. And there was, like,
00:31:20.480 --> 00:31:23.840
computer vision that was analyzing
which fishes and their movement,
00:31:23.840 --> 00:31:29.595
and it became like an always-on thing that people
would just tune in to online to watch the fish.
00:31:29.595 --> 00:31:33.737
LIU: There was a seal. There was a seal.
There were a couple seals that came by, …
00:31:33.737 --> 00:31:33.751
WHITE: Yeah.
00:31:33.751 --> 00:31:38.953
LIU: … too, which, yeah, I remember that.
That's really cool. Thanks for sharing.
00:31:38.953 --> 00:31:46.543
WHITE: Yeah, so you have to go not just
out of the box but out of the datacenter.
00:31:46.543 --> 00:31:50.320
LIU: [LAUGHS] Oh no. All right. All right.
00:31:50.320 --> 00:31:55.120
So, Chris, before letting you go, it
would be a missed opportunity if I did
00:31:55.120 --> 00:32:00.960
not ask you to share some practical advice and
reflections given the breadth of your experience
00:32:00.960 --> 00:32:04.880
and lessons that you learned. So I'm going to
do … maybe it doesn't have to be a super-fast
00:32:04.880 --> 00:32:09.102
lightning round, but it's going to be a
lightning round of a couple of questions.
00:32:09.102 --> 00:32:09.115
WHITE: Sure.
00:32:09.115 --> 00:32:13.440
LIU: All right. One, what is the
advice that you have received in
00:32:13.440 --> 00:32:16.880
your life that blew your mind
or changed your perspective?
00:32:16.880 --> 00:32:23.760
WHITE: You’ve got to look out for the right place
and the right time with the right people because
00:32:23.760 --> 00:32:29.120
it takes all those things to come together
for magic to happen and to make a difference.
00:32:29.120 --> 00:32:33.760
LIU: OK, I'm going to have to ask. This is not …
I'm breaking my own lightning round. How do you
00:32:33.760 --> 00:32:37.040
know? Like, how do you know if it's the right
place or the right time or the right people?
00:32:37.040 --> 00:32:39.840
WHITE: The universe gives a
lot of feedback. You know,
00:32:39.840 --> 00:32:45.840
it's something that with
good listening, you can hear.
00:32:48.560 --> 00:32:51.680
LIU: What did … what did the universe tell you?
00:32:51.680 --> 00:32:57.179
WHITE: Told me to marry my wife.
00:32:57.179 --> 00:33:00.067
LIU: [LAUGHS] Ohhhh. Good choice. Good
universe. Good universe. OK, next.
00:33:00.067 --> 00:33:02.778
WHITE: Yeah. Told me to have my kids. Absolutely.
00:33:02.778 --> 00:33:05.960
LIU: All right, next. Lesson you learned the
hard way or the misstep that humbled you?
00:33:05.960 --> 00:33:11.760
WHITE: I’ve got to slow down. My tendency
is to push too fast and too hard,
00:33:11.760 --> 00:33:18.000
and that can really lead to problems.
And slowing down, it doesn't in the end
00:33:18.000 --> 00:33:23.440
slow everything down. It just actually
makes it just right and just enough.
00:33:23.440 --> 00:33:26.400
LIU: What's an example of slow down?
00:33:26.400 --> 00:33:33.120
WHITE: Well, you know, I think there’s slowing
down in making decisions, especially about people,
00:33:33.120 --> 00:33:40.080
compared to other kinds of decisions.
Or even in my yoga practice. Slowing
00:33:40.080 --> 00:33:48.080
down the way that I use my body so that it
can be healthy and strong for a long time.
00:33:48.080 --> 00:33:52.000
LIU: I do believe that one
of your … the advice that
00:33:52.000 --> 00:33:55.280
you often give us in Cat Lab is take a vacation.
00:33:55.280 --> 00:33:56.000
WHITE: Absolutely.
00:33:56.000 --> 00:34:00.080
LIU: So slow down definitely fits that.
00:34:00.080 --> 00:34:01.560
WHITE: Get some perspective.
00:34:01.560 --> 00:34:07.120
LIU: So next question: talent or
accomplishment you're most proud
00:34:07.120 --> 00:34:12.480
of if given the chance to brag. And you have
stated humility many times, and I'm going to
00:34:12.480 --> 00:34:16.960
ask you to just turn that bit off for a
little bit and brag a little bit instead.
00:34:16.960 --> 00:34:23.840
WHITE: Mm-hmm. Well, I think it would
be putting my leg behind my head and
00:34:23.840 --> 00:34:27.720
saying [LAUGHTER] the alphabet backwards
while playing a percussion instrument.
00:34:27.720 --> 00:34:33.259
LIU: Hold on. Like, this is for realsies, and this
is, like, this is not an AI-generated statement …
00:34:33.259 --> 00:34:33.271
WHITE: No.
00:34:33.271 --> 00:34:36.000
LIU: … or prompt? This is …
And you've done this before?
00:34:36.000 --> 00:34:41.026
WHITE: I have. I've done
it in a leadership offsite.
00:34:41.026 --> 00:34:41.800
LIU: [LAUGHS] You did not.
00:34:41.800 --> 00:34:43.880
WHITE: I did.
00:34:43.880 --> 00:34:49.040
LIU: I am going to need to ask
some of our colleagues about
00:34:49.040 --> 00:34:51.667
that particular experience, but
there was a percussion instrument?
00:34:51.667 --> 00:34:56.960
WHITE: Yeah. There was. But, you know, all kidding
aside, you know, the 2016 presidential medal,
00:34:56.960 --> 00:35:02.800
you know, being able to make such a difference
in a way for so many people that were affected
00:35:02.800 --> 00:35:08.000
and to also change the way technology was used
from the data collection evidence, you know,
00:35:08.000 --> 00:35:14.400
through judicial proceedings, that process
and arc, you know, is definitely something I'm
00:35:14.400 --> 00:35:19.480
extremely proud of, was extremely difficult,
and I would definitely go for it again.
00:35:19.480 --> 00:35:26.880
LIU: I love it. OK, I love asking this
one. You know I've asked people this a
00:35:26.880 --> 00:35:30.240
lot. Career path you would have chosen if you did
00:35:30.240 --> 00:35:34.944
not become a researcher—and you're
not allowed to say yoga teacher.
00:35:34.944 --> 00:35:38.040
WHITE: [LAUGHS] No, I would just
be a different kind of researcher.
00:35:38.040 --> 00:35:39.907
LIU: No!
00:35:39.907 --> 00:35:40.635
WHITE: Yeah, definitely. Yeah.
00:35:40.635 --> 00:35:44.240
LIU: Really? OK, so if not an AI
or computer science researcher,
00:35:44.240 --> 00:35:45.520
what else would we be researching?
00:35:45.520 --> 00:35:48.955
WHITE: Oh, the brain. You know, the …
00:35:48.955 --> 00:35:53.440
LIU: OK, how about a non-STEM
research topic? [LAUGHS]
00:35:53.440 --> 00:35:55.540
WHITE: Anthropology.
00:35:55.540 --> 00:36:01.600
LIU: OK, OK, fine. So, like, no, there's no
scenario where research is not in the picture.
00:36:01.600 --> 00:36:07.200
WHITE: No, it just seems to be part of me,
you know, part of my drive, you know, part of
00:36:07.200 --> 00:36:14.160
what gives me energy, keeps me motivated
and curious. And this, you know, frontier,
00:36:14.160 --> 00:36:17.680
it allows that evolution to keep unfolding.
00:36:17.680 --> 00:36:23.920
LIU: OK. All right. Good to know. So not
yoga teacher. Always some form of research.
00:36:23.920 --> 00:36:28.160
WHITE: Well, I mean, yoga teacher is a
kind of researcher into the universe.
00:36:28.160 --> 00:36:32.800
LIU: What?! [LAUGHS] WHITE: You know, it's a researcher
into the states of consciousness and being.
00:36:32.800 --> 00:36:38.800
Into the alignment between the mind and the body.
It's full of research. Full of personal research.
00:36:38.800 --> 00:36:48.560
LIU: This is why I believe when once upon a time
at a leadership offsite you were asked to prompt
00:36:48.560 --> 00:36:54.640
AI to demonstrate, to build, like, a mascot for
you, and I think that mascot was also doing yoga.
00:36:54.640 --> 00:36:59.600
WHITE: It was. I asked my … I asked one
of my copilots to generate a picture of
00:36:59.600 --> 00:37:04.520
me based on everything it knows,
and it created a cat doing yoga.
00:37:04.520 --> 00:37:11.520
LIU: [LAUGHS] I love it. OK, so lastly,
and this is less “lightning round” and
00:37:11.520 --> 00:37:16.880
more like “take your time.” What would you
like listeners of this podcast to take away
00:37:16.880 --> 00:37:21.360
from this conversation? You know,
about you, about professional life,
00:37:21.360 --> 00:37:25.640
about research, about making … or
maybe even just making choices.
00:37:25.640 --> 00:37:32.800
WHITE: You know, I think it's about the
power of adaptability and the diversity
00:37:32.800 --> 00:37:40.000
of experience. You know, it's about
being open to what wants to happen,
00:37:40.000 --> 00:37:47.640
what needs to happen. You know, being open,
called to serve, willingness to change.
00:37:47.640 --> 00:37:50.000
LIU: I like that.
00:37:50.000 --> 00:37:56.000
So thank you, Chris—Dr. White—for joining me,
and to our audience, thank you for listening.
00:37:56.000 --> 00:37:58.383
WHITE: Thanks, Wei.
00:37:58.383 --> 00:37:59.360
[MUSIC]
STANDARD OUTRO: You’ve been listening to the
00:37:59.360 --> 00:38:02.480
Microsoft Research Podcast.
Check out more episodes of
00:38:02.480 --> 00:38:11.680
the podcast at aka.ms/researchpodcast or
on YouTube and major podcast platforms.
00:38:11.680 --> 00:38:12.204
[MUSIC ENDS]
[MUSIC]
00:00:00.067 --> 00:00:06.880
CHRIS WHITE: I have seen a lot of really bad
humanity. You know, I've seen terrorism and
00:00:06.880 --> 00:00:13.680
trafficking and sophisticated organized
crime, you know, designed to take things
00:00:13.680 --> 00:00:19.120
from people. Designed to hurt people. And I
just know that I can do something about it,
00:00:19.120 --> 00:00:25.280
you know, and that I have and that we can.
And so I think that that light, you know,
00:00:25.280 --> 00:00:37.563
the light of making a difference, you know, allows
me to hold and handle, you know, the dark parts.
00:00:37.563 --> 00:00:37.577
[THEME MUSIC]
00:00:37.577 --> 00:00:39.760
STANDARD INTRODUCTION: This is
the Microsoft Research Podcast,
00:00:39.760 --> 00:00:43.360
where Microsoft researchers—driving
advancement through fundamental science
00:00:43.360 --> 00:00:50.240
and technology research—explore the who,
how, and what’s next in computing and AI.
00:00:50.240 --> 00:00:51.126
[MUSIC ENDS]
00:00:51.126 --> 00:00:54.480
WEISHUNG LIU: Hi, and welcome. I'm Weishung Liu,
00:00:54.480 --> 00:00:58.320
a principal technical program
manager here at Microsoft Research,
00:00:58.320 --> 00:01:04.640
and today I'm talking to Dr. Christopher White,
who leads Microsoft Research's Catalyst Lab.
00:01:04.640 --> 00:01:09.520
Chris has one of those careers that
sounds almost made up: a Harvard postdoc,
00:01:09.520 --> 00:01:15.360
on-the-ground advanced tech lead during the
war in Afghanistan, a DARPA program manager,
00:01:15.360 --> 00:01:20.000
a dark web search pioneer, and a
public-good technologist. And his
00:01:20.000 --> 00:01:25.760
work has been featured in Popular Science as
well as recognized with honors, including the
00:01:25.760 --> 00:01:31.600
2016 Presidential Award for Extraordinary
Efforts to Combat Trafficking in Persons.
00:01:31.600 --> 00:01:35.920
But what I'm most interested in today
is the person behind that résumé. How
00:01:35.920 --> 00:01:40.240
someone who has seen some of the worst
as part of his work still comes to the
00:01:40.240 --> 00:01:45.680
office with a smile and a commitment to
building tools that make the world better.
00:01:45.680 --> 00:01:47.120
Chris, thanks for joining me.
00:01:47.120 --> 00:01:49.960
WHITE: Thanks for having me, Wei. Nice to see you.
00:01:49.960 --> 00:01:55.920
LIU: Yeah, good to see you, too. And so let's
get started. So I've known you for, I think,
00:01:55.920 --> 00:02:02.080
maybe around five or maybe six years, and now I
work with you in Catalyst Lab. I love your story,
00:02:02.080 --> 00:02:06.240
and when I tell people about you, I say
that you lead the newly formed Cat Lab,
00:02:06.240 --> 00:02:13.040
or Catalyst Lab, in Microsoft Research, you were
a senior executive at DARPA, and you are married
00:02:13.040 --> 00:02:19.120
to an incredibly accomplished woman and scholar.
And I personally think that your life story would
00:02:19.120 --> 00:02:25.920
make a very watchable Hollywood film. So when you
tell your own story, though, where do you start?
00:02:25.920 --> 00:02:32.400
WHITE: Well, you start from where you are.
And where I am today is leading research at
00:02:32.400 --> 00:02:39.360
Microsoft during a period of great fear and
hope about changing technology and society.
00:02:39.360 --> 00:02:46.720
LIU: But, like, that's now. There's clearly
lots of time from now into before, you know,
00:02:46.720 --> 00:02:51.600
the before times. So, like, tell me a little
bit about the before times. Like, where were you
00:02:51.600 --> 00:02:58.280
at—I think you grew up in Oklahoma maybe. Tell me
about that. Tell me about Chris White, the child.
00:02:58.280 --> 00:03:04.720
WHITE: Well, I was nerdy. I was a nerdy
Eagle Scout. [LAUGHTER] I liked to build
00:03:04.720 --> 00:03:11.120
my own computers and have local area network [LAN]
parties with my friends. But I also really liked
00:03:11.120 --> 00:03:18.000
being outside. I liked being in nature and camping
and rock climbing. Those were all important to me.
00:03:18.000 --> 00:03:22.960
LIU: OK, you mentioned LAN parties. Tell me
about … what did we play with our friends?
00:03:22.960 --> 00:03:27.120
WHITE: Well, you know, we played
everything. I preferred playing
00:03:27.120 --> 00:03:33.260
sports games, but I lost a lot of the other ones.
00:03:33.260 --> 00:03:37.360
LIU: [LAUGHS] Wait, so you played
like LAN party sports games, like,
00:03:37.360 --> 00:03:41.884
in basements? Like, with your
… were we all in the same … ?
00:03:41.884 --> 00:03:44.720
WHITE: Yeah. Attics. It was attics. Yeah,
not basements. [LAUGHTER] Mm-hmm. Yeah.
00:03:44.720 --> 00:03:49.440
Everybody had to bring their own computer or
bring their own Ethernet cable. And the more
00:03:49.440 --> 00:03:54.500
that you had your own machine that was higher
powered, the better you were going to play.
00:03:54.500 --> 00:03:56.000
LIU: So did you have the best machine?
00:03:56.000 --> 00:03:59.440
WHITE: I had the best machine.
I built the best machine. Until
00:03:59.440 --> 00:04:02.320
someone built a better machine.
But it was the best for a while.
00:04:02.320 --> 00:04:06.880
And I had a budget for it. And my family
didn't understand that the budget also
00:04:06.880 --> 00:04:13.440
included two giant speakers, like the size
of a computer tower, like subwoofer speakers.
00:04:13.440 --> 00:04:17.520
They didn't know that was part of
the computer building process. But
00:04:17.520 --> 00:04:22.600
it turns out that excellent audio-visual is a
critical part of a well-functioning machine.
00:04:22.600 --> 00:04:26.320
LIU: I love it. But, OK, we're going
to focus a little bit … continue
00:04:26.320 --> 00:04:30.640
focusing on childhood. So you mentioned nerdy kid,
00:04:30.640 --> 00:04:38.080
which means—I'm going to generalize here—debate
team or math competition team or chess team?
00:04:38.080 --> 00:04:43.680
WHITE: Oh, it was the debate team. Yeah.
Debate was a really wonderful activity.
00:04:43.680 --> 00:04:47.760
It taught me how to understand many
sides of an issue and taught me how
00:04:47.760 --> 00:04:51.560
to think about evidence and research in that way.
00:04:51.560 --> 00:04:58.240
LIU: So, Chris, take me back again. Like,
we're talking a little bit about Oklahoma,
00:04:58.240 --> 00:05:03.920
and, you know, at some point, you chose
researcher. Like, I—for example—when I was a kid,
00:05:03.920 --> 00:05:09.680
I was convinced I would be an architect. I am
not an architect as you can see. But did you
00:05:09.680 --> 00:05:14.800
always think you were going to be a researcher?
Was there an epiphany there? Like … or was it
00:05:14.800 --> 00:05:18.713
always kind of like a slow hunch that you
were going to pursue research as a career?
00:05:18.713 --> 00:05:24.080
WHITE: Yeah, I kind of always felt like
it. You know, as I was applying for school,
00:05:24.080 --> 00:05:28.720
those were the kinds of jobs I was interested
in. You know, the kinds of things that you do,
00:05:28.720 --> 00:05:33.280
you know, once you finish degrees that have
research in them. Those were the people,
00:05:33.280 --> 00:05:38.800
when I visited their work, I was inspired by.
The people that were discovering, you know,
00:05:38.800 --> 00:05:43.040
that were inventing, that were innovating,
that were … they're pushing the boundaries
00:05:43.040 --> 00:05:47.440
that were kind of in that unknown, you
know, space and making sense of it.
00:05:47.440 --> 00:05:49.600
And so I've always had that curiosity, you know,
00:05:49.600 --> 00:05:57.840
and that approach to discovery. And so as I
increased my education, it just changed the
00:05:57.840 --> 00:06:05.600
nature of what research meant and the ways
in which I imagined that process evolving.
00:06:05.600 --> 00:06:11.120
LIU: Was a doctorate always on the
cards then? Or did that come, like,
00:06:11.120 --> 00:06:15.120
as you were in college and then gradually learned
and thought, you know, I could keep going?
00:06:15.120 --> 00:06:18.960
WHITE: Yeah, you know, I didn't know
about doctorates really when I went to
00:06:18.960 --> 00:06:24.960
college. I wasn't as familiar with higher
education. And so I knew that I wanted,
00:06:24.960 --> 00:06:28.960
you know, advanced degrees. And I knew that
I wanted more than was being taught in the
00:06:28.960 --> 00:06:34.000
undergraduate level. But I didn't know, you know,
for example, that when you apply to PhD programs,
00:06:34.000 --> 00:06:38.480
that often you get paid to go. You know,
you get admitted and you get a stipend.
00:06:38.480 --> 00:06:44.000
You know, I was familiar with law and medicine
and that you take out loans and you have to
00:06:44.000 --> 00:06:49.360
do it this way. And so for me, it was a
surprise to know that I could apply to PhD
00:06:49.360 --> 00:06:55.200
programs directly from college. And from
then, I just really was excited, you know,
00:06:55.200 --> 00:07:02.080
that I could explore different universities
and different areas of focus and build on,
00:07:02.080 --> 00:07:06.640
you know, strong technical foundations
of math and science and physics.
00:07:06.640 --> 00:07:13.520
And it was really amazing mentors that helped
me and opened doors and inspired me. You know,
00:07:13.520 --> 00:07:19.200
mentors at Oklahoma State University, Johns
Hopkins, and Harvard, really people that
00:07:19.200 --> 00:07:23.800
took their time and helped me. I'm
really grateful for my teachers.
00:07:23.800 --> 00:07:32.420
LIU: So let's talk about your first mentor
of your life, and I'm going to assume … Dad?
00:07:32.420 --> 00:07:34.520
WHITE: Yeah, Dad’s a big influence.
00:07:34.520 --> 00:07:40.800
LIU: OK, Dad's a big influence. And so we're going
to, I'm going to go down this path. So Catalyst
00:07:40.800 --> 00:07:46.240
Lab landed a huge commitment from our executive,
kind of, leadership team on a project that has
00:07:46.240 --> 00:07:52.320
been, I think, like a decade in the making. And
then … so we all, I know we all came together on
00:07:52.320 --> 00:07:57.440
a call. We're like, “Wow, let's debrief. Let's
talk about it.” Like, oh, energy's up. And
00:07:57.440 --> 00:08:03.243
then at the very end of the meeting, you said,
“OK, now I'm going to go clean some toilets.”
00:08:03.243 --> 00:08:03.275
WHITE: [LAUGHS] Mm-hmm. Mm-hmm.
00:08:03.275 --> 00:08:08.640
LIU: And I think all of us kind of
thought, like, wha-what does that
00:08:09.280 --> 00:08:14.560
mean? So before we get kind of more into
the specifics of mentorship and the,
00:08:14.560 --> 00:08:22.400
like, follow-on in your career, what— please tell
me—does cleaning toilets have to do with success?
00:08:22.400 --> 00:08:26.400
WHITE: Well, you know, like I said, Dad's a
big influence, and he taught me so many things,
00:08:26.400 --> 00:08:29.920
and he taught me about respecting
people. And most importantly,
00:08:29.920 --> 00:08:34.480
he taught me about humility and
service. And so his way of making
00:08:34.480 --> 00:08:37.920
sure I understood that was every time
something really good happened to me,
00:08:37.920 --> 00:08:44.560
he'd remind me to go clean some toilets and,
sort of, embody that. And so I did. I did it
00:08:44.560 --> 00:08:50.240
last weekend actually. I went to a community
event and volunteered to do bathroom clean.
00:08:50.240 --> 00:08:56.400
LIU: Is this like … are you … so in, like, the
household chores, are you, like, the bathroom
00:08:56.400 --> 00:08:59.109
cleaner? Like, is this just always your job
when we're dividing up chores in the house?
00:08:59.109 --> 00:09:04.320
WHITE: No, no, no. We trade off, and also we
have a complicated household, and so we have
00:09:04.320 --> 00:09:10.240
a lot of help that also helps with cleaning.
And so this is something that I choose to do,
00:09:10.240 --> 00:09:14.400
especially if I can do it in a community
context, because it again reminds me,
00:09:14.400 --> 00:09:20.160
you know, about humility and about service and
of respect for other people and other jobs.
00:09:20.160 --> 00:09:26.880
LIU: So I think this … there's some Eagle Scout
vibes that I'm getting from here, as well,
00:09:26.880 --> 00:09:35.280
based on sort of humility and service. What
is the Boy Scout badge you are most proud of?
00:09:35.280 --> 00:09:43.120
WHITE: Oh, that's really hard. Each badge is
something that you earn, including the Eagle
00:09:43.120 --> 00:09:48.320
Scout badge. And the Eagle Scout badge for me
was maybe the most difficult to earn because
00:09:48.320 --> 00:09:54.320
it requires an Eagle project. And I was really
close to my Eagle project. I was, I was … had
00:09:54.320 --> 00:10:01.440
it set up and it was ready to go. And it was in
this church in downtown Oklahoma City in 1995.
00:10:01.440 --> 00:10:08.378
And right as I was about to do this project, the
Murrah Building blew up in a terrorist incident, …
00:10:08.378 --> 00:10:08.389
LIU: Whoa.
00:10:08.389 --> 00:10:14.480
WHITE: … and it blew up part of the church. And,
therefore, my Eagle project was not realistic, and
00:10:14.480 --> 00:10:21.760
it was not a priority, and it got shelved. And it
took me several years after that to then come back
00:10:21.760 --> 00:10:28.538
to plan and execute an ambitious Eagle project
right before I turned 18. And so that badge, …
00:10:28.538 --> 00:10:30.640
LIU: Wow.
WHITE: … you know, completing the project,
00:10:30.640 --> 00:10:36.400
overcoming significant setbacks, including
being a teenager and having other interests,
00:10:36.400 --> 00:10:41.000
to come through and finish it,
I'm really proud of finishing it.
00:10:41.000 --> 00:10:48.080
LIU: That's awesome. So you … so that was, so
there were a few years then in between when
00:10:48.080 --> 00:10:55.200
you ideated, had the idea for the project,
and then … I appreciate the story around
00:10:55.200 --> 00:11:02.880
resilience because of when you completed the
project. Like, what is the biggest takeaway
00:11:02.880 --> 00:11:08.160
that you took from there? Like what is that
big lesson? You turned 18, you finished it,
00:11:08.160 --> 00:11:11.600
you celebrate. Actually, you know, maybe
I'll ask you, how did you celebrate?
00:11:11.600 --> 00:11:17.760
WHITE: Oh, with my family. You know, the Boy
Scouts is a long process. It's many years, many
00:11:17.760 --> 00:11:23.440
kinds of roles, many kinds of learning, and to
finish it—I finished it with another of my fellow
00:11:23.440 --> 00:11:28.240
scouts that had been together for a long time.
And so our families, we had a ceremony, and it
00:11:28.240 --> 00:11:32.480
was something we could celebrate. And every time I
think about it, I think about the things that you
00:11:32.480 --> 00:11:37.840
keep repeating when you attend the scout things,
like the motto, “being prepared,” and the, kind
00:11:37.840 --> 00:11:42.720
of, laws about being respectful and courteous.
And those are things that continue to be,
00:11:42.720 --> 00:11:49.960
you know, principles that influence my priorities
and also inspire me and help me in my job.
00:11:49.960 --> 00:11:55.520
LIU: Before we get to your job then, we're
going to talk about this interim period
00:11:55.520 --> 00:12:04.000
where I'll just say you chose to go to war and a
few other things in between. Let's jump to 2010,
00:12:04.720 --> 00:12:09.360
when you joined DARPA, which is the
Defense Advanced Research Projects Agency,
00:12:09.360 --> 00:12:13.840
within the Department of War. And then, I
think, very quickly afterwards you were sent
00:12:13.840 --> 00:12:19.360
to Afghanistan to help implement a new
approach to wartime analysis. And then
00:12:19.360 --> 00:12:29.040
… but before that, I think you were a Harvard
postdoc, and you had actually mentioned it was
00:12:29.040 --> 00:12:35.360
a mentor who encouraged you to attend a meeting
with DARPA that resulted in this course change.
00:12:35.360 --> 00:12:43.280
And so now I'm super curious, how did
this conversation happen where you were,
00:12:43.280 --> 00:12:46.880
you know, promising Harvard postdoc,
which became a mentor going, “Hey,
00:12:46.880 --> 00:12:50.960
there's something happening. You should
go attend.” Like, what happened there?
00:12:50.960 --> 00:12:57.760
WHITE: Yeah. DARPA invests taxpayer
money, and they have a record of
00:12:57.760 --> 00:13:01.360
really creating innovations, massive
innovations. You know, the internet,
00:13:01.360 --> 00:13:07.920
GPS, stealth, UAVs. And at the time, there
was different ways of building programs,
00:13:07.920 --> 00:13:14.080
and one of the kind of tested ways was to
bring together different fields of expertise.
00:13:14.080 --> 00:13:21.520
And so we were in wartime in Afghanistan.
DARPA was not involved very much. And there
00:13:21.520 --> 00:13:28.080
was a question: you know, could we bring
together social scientists with math people,
00:13:28.080 --> 00:13:34.640
computer science and defense experts, and
together could they help us make sense of
00:13:34.640 --> 00:13:41.600
what's happening in the data? Could that be useful
to prevent violence and to protect our soldiers?
00:13:41.600 --> 00:13:46.080
And it was an experiment because it required
all these different areas of expertise coming
00:13:46.080 --> 00:13:53.200
together—from sensors through analysis
through social behavior—and a culture
00:13:53.200 --> 00:14:00.960
change. And so the summer was a Skunk Works to
understand, was there something to contribute?
00:14:00.960 --> 00:14:04.480
And at the end of it, they asked
me, they said, “Well, you know,
00:14:04.480 --> 00:14:08.720
do you believe in your work?” And I was
like, “Yeah.” You know, like, “Well,
00:14:08.720 --> 00:14:15.200
do you … can you prove it? You know, can you
go see if it can help protect our soldiers
00:14:15.200 --> 00:14:21.280
and help prevent violence?” And so they
asked me to be of service, and I said yeah.
00:14:21.280 --> 00:14:28.000
And it was something that I was really unprepared
for. I was supposed to go for nine days the first
00:14:28.000 --> 00:14:32.880
time, and on the third day, I got a battlefield
promotion and was put in charge and then ended
00:14:32.880 --> 00:14:39.920
up staying for 90 days and then went back a few
times after that even. And so, you know, our work
00:14:39.920 --> 00:14:44.560
was towards … oriented towards protecting, you
know, soldiers and toward preventing violence
00:14:44.560 --> 00:14:53.840
and using AI and advanced technology to do
that. And it was a, really, a big opportunity.
00:14:53.840 --> 00:15:02.480
LIU: So you attended a meeting, you contributed
some work, and then, you know, “Do you believe
00:15:02.480 --> 00:15:09.360
in your work? Prove it.” What was the time frame
between, you know, like, showing up at meeting,
00:15:09.360 --> 00:15:15.009
participating in some of the research before,
like, “prove it”? Was it years, or was it months?
00:15:15.009 --> 00:15:19.840
WHITE: The … no, no, the … so the Skunk Works
was a summer. It was like, you know, 40 days of,
00:15:19.840 --> 00:15:25.040
kind of, like focused effort. But at the end
of it, I was expecting to go back to Harvard
00:15:25.040 --> 00:15:31.120
and continue my postdoc. But instead, you
know, they sat us down; they said, “Look,
00:15:31.120 --> 00:15:37.520
you know, we think this can make a difference.
You know, do you think so?” And when I said yeah,
00:15:37.520 --> 00:15:42.400
from then to getting over there was like
a week. It was an extremely fast ...
00:15:42.400 --> 00:15:43.029
LIU: What?
00:15:43.029 --> 00:15:49.960
WHITE: … turnaround. It was extremely
urgent and, you know, unexpected.
00:15:49.960 --> 00:15:56.000
LIU: OK, so, like, how … so you
said, yes, like I believe in my
00:15:56.000 --> 00:16:02.560
work. They gave you … they essentially
said, OK, then go next week. And how,
00:16:02.560 --> 00:16:07.480
how does somebody prepare for an experience
like that? If you can even prepare.
00:16:07.480 --> 00:16:11.440
WHITE: Yeah, I don't know. You know,
this was an experiment. You know, sending
00:16:11.440 --> 00:16:18.560
researchers—even having a presence from DARPA—in
the war zone, it was new. And the idea was there
00:16:18.560 --> 00:16:23.840
could be a difference made because technology
was advancing quickly. And that meant that many
00:16:23.840 --> 00:16:28.640
of the mechanisms in place for other ways of
getting prepared were not available. And there
00:16:28.640 --> 00:16:33.600
was a bunch of training, like crash courses in
training on what to do in certain situations,
00:16:33.600 --> 00:16:40.560
how things work, you know, how to handle yourself
in different ways. But it was … it was very brief.
00:16:40.560 --> 00:16:44.960
And so there wasn't a lot of preparation
because also we were planning to be there
00:16:44.960 --> 00:16:49.360
for nine days. But because there was
an opportunity to do something good,
00:16:49.360 --> 00:16:56.320
that extended, and there was a lot of learning
on the job and a lot of support in theater,
00:16:56.320 --> 00:17:02.520
a lot of people really willing to
help us because we were helping them.
00:17:02.520 --> 00:17:09.920
LIU: So, OK, so you talked about lots of people
there, lots of people supporting. I'm not going to
00:17:09.920 --> 00:17:14.640
focus too much on your time in Afghanistan because
that was covered really well in a Popular Science
00:17:14.640 --> 00:17:23.504
article. But I do have to ask, can you tell me the
origin story of the nickname Dr. Spaghetti Man?
00:17:23.504 --> 00:17:30.000
WHITE: [LAUGHS] Sure. Sure. You know, I'm
a long-time yoga practitioner. You know,
00:17:30.000 --> 00:17:39.440
20-plus years. It's how I keep myself calibrated.
And in Kabul, I was practicing yoga in the gym,
00:17:39.440 --> 00:17:44.480
and all these people were staring at me. You know,
they were staring at me before their … they had
00:17:44.480 --> 00:17:49.840
a class; they had, like, a combatives class, a
kind of different judo-mixed-martial-arts class.
00:17:49.840 --> 00:17:52.720
And, you know, I had a little bit of
training, and so after a couple of times,
00:17:52.720 --> 00:17:59.840
I decided I'd just join in. And I joined in,
and I won a lot. I won a lot of the matches.
00:17:59.840 --> 00:18:05.600
And one of the matches I won, turns out, it
was against the Marine who was guarding the
00:18:05.600 --> 00:18:11.760
entrance to the building I was working
in across the headquarters. And he told
00:18:11.760 --> 00:18:19.840
everyone. He told everyone, you know, that
this, like, lanky spaghetti guy, you know,
00:18:19.840 --> 00:18:25.280
like, got out of things and beat him, and
he did it in such a way that everyone in
00:18:25.280 --> 00:18:33.280
the building knew. And so I got all this
respect, and I got that cool nickname.
00:18:33.280 --> 00:18:39.280
LIU: Did other folks challenge you as a
result of this Marine telling your story?
00:18:39.280 --> 00:18:44.480
WHITE: I mean, everyone that came to that
class was game, you know. And I didn't,
00:18:44.480 --> 00:18:48.800
you know, I couldn't beat everybody, but I
held my own, and that was really surprising
00:18:48.800 --> 00:18:53.520
to a lot of people. And again, once you
establish yourself in different ways,
00:18:53.520 --> 00:18:57.960
you know, there is a lot of
respect and, um, camaraderie.
00:18:57.960 --> 00:19:05.200
LIU: So, you know, tell me, what did
that whole experience teach you about,
00:19:05.200 --> 00:19:09.600
like, a life lived, and how do those
lessons show up in your work today,
00:19:09.600 --> 00:19:13.920
including the, you know, Spaghetti Man experience?
00:19:13.920 --> 00:19:18.320
WHITE: You know, the whole thing,
it taught me about, you know,
00:19:18.320 --> 00:19:25.280
service and what it means to serve,
and it taught me about the reality of
00:19:25.280 --> 00:19:31.720
both peace and violence in the world. And that
continued to motivate me to make a difference.
00:19:31.720 --> 00:19:38.320
LIU: After your time in Afghanistan, you
continued on with DARPA as a program manager,
00:19:38.320 --> 00:19:43.840
and in that role, you developed Memex,
which I believe is a set of tools that
00:19:43.840 --> 00:19:49.200
could be mixed and matched to help uncover
some really shady and dark parts of the web,
00:19:49.200 --> 00:19:53.760
which from what I understand was
really hard to do up until that point.
00:19:54.320 --> 00:20:03.200
I think this work also is what helped you … what
you received your 2016 presidential award on. And,
00:20:03.200 --> 00:20:09.760
you know, like, can you … how do you explain
the Memex project and this dark web work to
00:20:09.760 --> 00:20:16.080
someone outside tech? And what about this project
inspired or appealed to you to even pursue it?
00:20:16.080 --> 00:20:22.480
WHITE: Yeah, well, the inspiration started when I
was overseas because I saw the benefit of evidence
00:20:22.480 --> 00:20:29.840
when it came to stopping bad things from happening
and prosecuting them. And coming back to the US,
00:20:29.840 --> 00:20:35.280
I saw more and more opportunities where
collecting evidence at large scale
00:20:35.280 --> 00:20:40.080
was a really useful mechanism. And so, you
know, people think of search engines a lot,
00:20:40.080 --> 00:20:45.920
and search engines are a great tool for finding
something, but when you need to connect the dots,
00:20:45.920 --> 00:20:51.520
there wasn't a great tool that could work at
the scale of the web and connect the dots,
00:20:51.520 --> 00:20:56.400
especially some of the dots being in
difficult places that were not very visible.
00:20:56.400 --> 00:21:01.520
And so by creating a technology that would
connect the dots across parts of the internet,
00:21:01.520 --> 00:21:06.000
it allowed the prosecution of these kinds
of really awful [trafficking] crimes,
00:21:06.000 --> 00:21:11.760
and so for example, in New York City, it allowed
them to increase their prosecutions by 10 times
00:21:11.760 --> 00:21:15.680
because it gave them new evidence and different
kinds of evidence than they had before,
00:21:15.680 --> 00:21:21.160
and they could use that in a way to
really deal with the issue differently.
00:21:21.160 --> 00:21:26.480
LIU: OK, from what I understand, the
dark web is a pretty—ha-ha—dark place,
00:21:26.480 --> 00:21:32.720
[LAUGHS] and so, like, you've worked in
some very dark spaces and close to some
00:21:32.720 --> 00:21:38.160
very dark uses of the technology. So, like,
how again, like, you and I interact a lot,
00:21:38.160 --> 00:21:47.040
and so it's always curious to me, how has that
not made you cynical and how do you stay hopeful?
00:21:47.040 --> 00:21:52.880
WHITE: You know, I think what it's done is
it's really increased my motivation. You know,
00:21:52.880 --> 00:21:55.840
my motivation to make a contribution. You know,
00:21:55.840 --> 00:22:03.440
my motivation to stay aware and be realistic.
And by being realistic and motivated to
00:22:04.880 --> 00:22:10.800
address some of these very challenging issues,
you know, around violence, around harm,
00:22:10.800 --> 00:22:18.640
and instead to focus on empowerment and to
focus on justice, it gave me something to do.
00:22:18.640 --> 00:22:28.240
And so I have seen a lot of really bad humanity,
you know. I've seen terrorism and trafficking and
00:22:29.120 --> 00:22:33.600
sophisticated organized crime, you know,
designed to take things from people.
00:22:33.600 --> 00:22:38.720
Designed to hurt people. And I just know
that I can do something about it, you know,
00:22:38.720 --> 00:22:44.720
and that I have and that we can. And so it's
really been inspiring to, you know, for example,
00:22:44.720 --> 00:22:49.680
come to Microsoft, where there is a mission
of empowerment and there is that perspective
00:22:49.680 --> 00:22:56.320
that technology can make a difference.
And so I think that light, you know,
00:22:56.320 --> 00:23:05.040
the light of making a difference, you know, allows
me to hold and handle, you know, the dark parts.
00:23:05.040 --> 00:23:10.240
LIU: I mean, that's a very powerful message,
right? Because even with the darkest parts of
00:23:10.240 --> 00:23:17.920
humanity, there is light, or, you know, you
cannot have darkness without light, as well.
00:23:17.920 --> 00:23:22.320
I think there may have been a 60 Minutes
interview that included a short snippet
00:23:22.320 --> 00:23:27.440
of you talking about your work, and you
had shared with me that—which, by the way,
00:23:27.440 --> 00:23:34.480
I cannot find the 60 Minutes episode,
and I desperately want to. But, you know,
00:23:34.480 --> 00:23:41.920
I heard perhaps through you, Chris, that that
60 Minutes episode actually contributed to you
00:23:41.920 --> 00:23:48.960
coming to Microsoft maybe or, you know, getting a
job at Microsoft, and that made me curious. Like,
00:23:48.960 --> 00:23:57.160
after some really incredible experiences, after
all of that, why did you choose Microsoft?
00:23:57.160 --> 00:24:05.280
WHITE: Oh, it was a … it was a really good
choice. And at the time, I had a lot of options,
00:24:05.280 --> 00:24:13.040
and I was looking around, and this mix of the
mission of the company—this mission of empowering
00:24:13.040 --> 00:24:19.760
everyone—and the scope of the company, you know,
being, you know, IT for the world, you know,
00:24:19.760 --> 00:24:24.880
platforms and infrastructure for the world,
and my confidence in its leader. You know,
00:24:24.880 --> 00:24:30.000
the CEO was part of my interview process.
And, you know, that together gave me,
00:24:30.880 --> 00:24:33.600
you know, this perspective that I
could do something here. You know,
00:24:33.600 --> 00:24:40.400
that there … I could start to envision moving
from, you know, for example, very high-powered,
00:24:40.400 --> 00:24:47.760
expensive lasers to table knives. They
both cut, and the world needs both of them.
00:24:47.760 --> 00:24:53.520
LIU: One of the things that you talk a lot
about is problem selection. And I think
00:24:53.520 --> 00:24:58.960
sometimes that also gets referred to as taste
these days. So how do you think about problem
00:24:58.960 --> 00:25:05.760
selection in Microsoft Research, in Catalyst
Lab, given the breadth of what Microsoft does?
00:25:05.760 --> 00:25:10.480
WHITE: Oh, I think it's just the most
important part of research. You know,
00:25:10.480 --> 00:25:16.960
it's the place where the angle of the whole
thing changes, and it allows you to start to
00:25:16.960 --> 00:25:21.680
imagine a different future, you know, or to work
backward from that future that you want to see
00:25:21.680 --> 00:25:27.600
and see what you can do to contribute to it now.
And the amazing thing about Microsoft Research
00:25:27.600 --> 00:25:33.840
is there's tremendous freedom, and so there's
tremendous creativity in problem selection.
00:25:33.840 --> 00:25:39.760
And it's the kind of thing where you have
that latitude to go from idea to evidence
00:25:39.760 --> 00:25:45.040
to this potential for growth and thriving
and along the way get feedback about value
00:25:45.040 --> 00:25:54.000
and usability and benefit or policy or
adaptability, integration. And so its
00:25:54.000 --> 00:26:01.200
scope provides so many vehicles for growth
and utility of advanced research ideas.
00:26:01.200 --> 00:26:07.360
And its mission, you know, allows the focus of
those to address absolutely critical problems. You
00:26:07.360 --> 00:26:14.400
know, problems of, you know, impending security
challenges due to AI, problems of advanced,
00:26:14.400 --> 00:26:21.520
you know, memory and storage; it allows for, you
know, personalization automation, just, you know,
00:26:21.520 --> 00:26:30.640
for example. And I think that this mix of freedom
combined with responsibility and mechanisms,
00:26:30.640 --> 00:26:37.760
you know, to prove out and demonstrate value,
that that combination is just really remarkable
00:26:37.760 --> 00:26:44.320
and it really motivates me to
take the opportunity seriously.
00:26:44.320 --> 00:26:52.000
LIU: So it's almost going to be like asking you
about your favorite kids now. What are some, what
00:26:52.000 --> 00:27:00.080
are some accomplishments in Microsoft Research or
Catalyst Lab that you are particularly proud of?
00:27:00.080 --> 00:27:07.760
WHITE: Oh, you know, I think that, you know, right
now, for example—again, focusing on kind of where
00:27:07.760 --> 00:27:16.240
we are—I think that we have a real opportunity to
contribute something in the cyber defense space.
00:27:16.240 --> 00:27:25.280
I think our work on security and AI
and hacking, it's really in, you know,
00:27:25.280 --> 00:27:31.760
based on such a deep expertise and also based
on, you know, people with a lot of experience,
00:27:31.760 --> 00:27:36.640
and I think it's growing in need, you know,
the need to protect our digital infrastructure,
00:27:36.640 --> 00:27:44.800
the need to deter and interdict people that would
like to break systems or cause harm, that's really
00:27:44.800 --> 00:27:51.840
growing right now. And so being able to show up
in a way that creates breakthrough that can scale
00:27:51.840 --> 00:27:58.000
with relevance to address that growing need, it's
really something that's, you know, awesome to me.
00:27:58.720 --> 00:28:03.200
And also I think that again kind of extending
from the, sort of, Memex conversation,
00:28:03.200 --> 00:28:08.640
where we're trying to connect the dots, you know,
these days, trying to analyze lots of content as
00:28:08.640 --> 00:28:12.960
part of being more efficient, as part of being
more creative, as part of being more capable,
00:28:12.960 --> 00:28:19.040
you know, using tools to augment that human
taste and creativity and ingenuity, that's
00:28:19.040 --> 00:28:24.640
really never been as powerful as it is before
and continues to grow. And so I think making
00:28:24.640 --> 00:28:29.360
contributions in these areas, you know, where
we have state-of-the-art technology and we're
00:28:29.360 --> 00:28:35.200
creating changes in the way systems are getting
built and used, to me, it shows that, you know,
00:28:35.200 --> 00:28:43.040
despite the fast pace of change and despite the
extreme competitiveness in the technology sector,
00:28:43.040 --> 00:28:49.760
that specialized expertise with good problem
selection and an environment to grow and get
00:28:49.760 --> 00:28:55.600
feedback can create the kind of breakthroughs
that matter. And so I really like that.
00:28:55.600 --> 00:29:00.480
Maybe to give one more example from the past.
Just maybe something emblematic of the group
00:29:00.480 --> 00:29:08.800
and some of my predecessors was this underwater
datacenter called Project Natick, [LAUGHTER] where
00:29:08.800 --> 00:29:13.280
this experiment, you know, put a datacenter in a
submersible on the bottom of the ocean connected
00:29:13.280 --> 00:29:18.560
to a wind farm, and it tested this design about,
could you build something that didn't require
00:29:18.560 --> 00:29:24.080
human intervention? Could it be more efficient
with power? Could it be more efficient with heat?
00:29:24.080 --> 00:29:29.360
And the program manager that created that is
one of the best I've ever worked with. And it
00:29:29.360 --> 00:29:35.920
just performed incredibly well as this
huge surprise, and it led to us being,
00:29:35.920 --> 00:29:38.400
you know, named the most innovative company
[World Changing Company of the Year] in 2021
00:29:38.400 --> 00:29:44.160
and other awards. And it's just an example
of really out-there thinking, imagining a
00:29:44.160 --> 00:29:51.360
world that had challenges with water access
and power access and land but where we need,
00:29:51.360 --> 00:29:57.680
you know, computation and data storage
and technical capability available.
00:29:57.680 --> 00:30:00.960
So a little bit out there, but
I really … I feel like we have
00:30:00.960 --> 00:30:04.640
to be imagining these different
futures. We have to be radical
00:30:04.640 --> 00:30:11.200
in that way so that we can start to
work back and make contributions now.
00:30:11.200 --> 00:30:19.600
LIU: I love the Natick example because I feel
like Natick is the gift that keeps on giving. And,
00:30:20.480 --> 00:30:27.120
you know, if I can digress for a minute, I
remember that I think one of the first TikTok
00:30:27.120 --> 00:30:35.600
videos that the Microsoft account posted
was of Natick, but it was using, like,
00:30:35.600 --> 00:30:41.920
a meme at the time which had, like, human eyes
and a mouth being lowered into, uh, and it was
00:30:41.920 --> 00:30:47.680
of Natick being lowered into the ocean, including
like a Clippy, like a real-life Clippy swimming
00:30:47.680 --> 00:30:55.760
across the … it was … I still love that video for
reasons why I don't understand. [LAUGHTER] But if
00:30:55.760 --> 00:31:00.214
you haven't seen it, highly recommend. Maybe we
can link it [LAUGHS] if the video still exists.
00:31:00.214 --> 00:31:04.703
WHITE: Definitely. Look, look, one
of the highlights was the fish cam.
00:31:04.703 --> 00:31:04.713
LIU: Yes!
00:31:04.713 --> 00:31:09.200
WHITE: You know, people were concerned that it
might disrupt the marine ecosystem. And we did
00:31:09.200 --> 00:31:15.983
a lot of preparation and study for that. And just
to be sure, we put cameras and watched it happen.
00:31:15.983 --> 00:31:15.993
LIU: Yep.
00:31:15.993 --> 00:31:20.480
WHITE: And all these fish kept coming around
and making it home. And there was, like,
00:31:20.480 --> 00:31:23.840
computer vision that was analyzing
which fishes and their movement,
00:31:23.840 --> 00:31:29.595
and it became like an always-on thing that people
would just tune in to online to watch the fish.
00:31:29.595 --> 00:31:33.737
LIU: There was a seal. There was a seal.
There were a couple seals that came by, …
00:31:33.737 --> 00:31:33.751
WHITE: Yeah.
00:31:33.751 --> 00:31:38.953
LIU: … too, which, yeah, I remember that.
That's really cool. Thanks for sharing.
00:31:38.953 --> 00:31:46.543
WHITE: Yeah, so you have to go not just
out of the box but out of the datacenter.
00:31:46.543 --> 00:31:50.320
LIU: [LAUGHS] Oh no. All right. All right.
00:31:50.320 --> 00:31:55.120
So, Chris, before letting you go, it
would be a missed opportunity if I did
00:31:55.120 --> 00:32:00.960
not ask you to share some practical advice and
reflections given the breadth of your experience
00:32:00.960 --> 00:32:04.880
and lessons that you learned. So I'm going to
do … maybe it doesn't have to be a super-fast
00:32:04.880 --> 00:32:09.102
lightning round, but it's going to be a
lightning round of a couple of questions.
00:32:09.102 --> 00:32:09.115
WHITE: Sure.
00:32:09.115 --> 00:32:13.440
LIU: All right. One, what is the
advice that you have received in
00:32:13.440 --> 00:32:16.880
your life that blew your mind
or changed your perspective?
00:32:16.880 --> 00:32:23.760
WHITE: You’ve got to look out for the right place
and the right time with the right people because
00:32:23.760 --> 00:32:29.120
it takes all those things to come together
for magic to happen and to make a difference.
00:32:29.120 --> 00:32:33.760
LIU: OK, I'm going to have to ask. This is not …
I'm breaking my own lightning round. How do you
00:32:33.760 --> 00:32:37.040
know? Like, how do you know if it's the right
place or the right time or the right people?
00:32:37.040 --> 00:32:39.840
WHITE: The universe gives a
lot of feedback. You know,
00:32:39.840 --> 00:32:45.840
it's something that with
good listening, you can hear.
00:32:48.560 --> 00:32:51.680
LIU: What did … what did the universe tell you?
00:32:51.680 --> 00:32:57.179
WHITE: Told me to marry my wife.
00:32:57.179 --> 00:33:00.067
LIU: [LAUGHS] Ohhhh. Good choice. Good
universe. Good universe. OK, next.
00:33:00.067 --> 00:33:02.778
WHITE: Yeah. Told me to have my kids. Absolutely.
00:33:02.778 --> 00:33:05.960
LIU: All right, next. Lesson you learned the
hard way or the misstep that humbled you?
00:33:05.960 --> 00:33:11.760
WHITE: I’ve got to slow down. My tendency
is to push too fast and too hard,
00:33:11.760 --> 00:33:18.000
and that can really lead to problems.
And slowing down, it doesn't in the end
00:33:18.000 --> 00:33:23.440
slow everything down. It just actually
makes it just right and just enough.
00:33:23.440 --> 00:33:26.400
LIU: What's an example of slow down?
00:33:26.400 --> 00:33:33.120
WHITE: Well, you know, I think there’s slowing
down in making decisions, especially about people,
00:33:33.120 --> 00:33:40.080
compared to other kinds of decisions.
Or even in my yoga practice. Slowing
00:33:40.080 --> 00:33:48.080
down the way that I use my body so that it
can be healthy and strong for a long time.
00:33:48.080 --> 00:33:52.000
LIU: I do believe that one
of your … the advice that
00:33:52.000 --> 00:33:55.280
you often give us in Cat Lab is take a vacation.
00:33:55.280 --> 00:33:56.000
WHITE: Absolutely.
00:33:56.000 --> 00:34:00.080
LIU: So slow down definitely fits that.
00:34:00.080 --> 00:34:01.560
WHITE: Get some perspective.
00:34:01.560 --> 00:34:07.120
LIU: So next question: talent or
accomplishment you're most proud
00:34:07.120 --> 00:34:12.480
of if given the chance to brag. And you have
stated humility many times, and I'm going to
00:34:12.480 --> 00:34:16.960
ask you to just turn that bit off for a
little bit and brag a little bit instead.
00:34:16.960 --> 00:34:23.840
WHITE: Mm-hmm. Well, I think it would
be putting my leg behind my head and
00:34:23.840 --> 00:34:27.720
saying [LAUGHTER] the alphabet backwards
while playing a percussion instrument.
00:34:27.720 --> 00:34:33.259
LIU: Hold on. Like, this is for realsies, and this
is, like, this is not an AI-generated statement …
00:34:33.259 --> 00:34:33.271
WHITE: No.
00:34:33.271 --> 00:34:36.000
LIU: … or prompt? This is …
And you've done this before?
00:34:36.000 --> 00:34:41.026
WHITE: I have. I've done
it in a leadership offsite.
00:34:41.026 --> 00:34:41.800
LIU: [LAUGHS] You did not.
00:34:41.800 --> 00:34:43.880
WHITE: I did.
00:34:43.880 --> 00:34:49.040
LIU: I am going to need to ask
some of our colleagues about
00:34:49.040 --> 00:34:51.667
that particular experience, but
there was a percussion instrument?
00:34:51.667 --> 00:34:56.960
WHITE: Yeah. There was. But, you know, all kidding
aside, you know, the 2016 presidential medal,
00:34:56.960 --> 00:35:02.800
you know, being able to make such a difference
in a way for so many people that were affected
00:35:02.800 --> 00:35:08.000
and to also change the way technology was used
from the data collection evidence, you know,
00:35:08.000 --> 00:35:14.400
through judicial proceedings, that process
and arc, you know, is definitely something I'm
00:35:14.400 --> 00:35:19.480
extremely proud of, was extremely difficult,
and I would definitely go for it again.
00:35:19.480 --> 00:35:26.880
LIU: I love it. OK, I love asking this
one. You know I've asked people this a
00:35:26.880 --> 00:35:30.240
lot. Career path you would have chosen if you did
00:35:30.240 --> 00:35:34.944
not become a researcher—and you're
not allowed to say yoga teacher.
00:35:34.944 --> 00:35:38.040
WHITE: [LAUGHS] No, I would just
be a different kind of researcher.
00:35:38.040 --> 00:35:39.907
LIU: No!
00:35:39.907 --> 00:35:40.635
WHITE: Yeah, definitely. Yeah.
00:35:40.635 --> 00:35:44.240
LIU: Really? OK, so if not an AI
or computer science researcher,
00:35:44.240 --> 00:35:45.520
what else would we be researching?
00:35:45.520 --> 00:35:48.955
WHITE: Oh, the brain. You know, the …
00:35:48.955 --> 00:35:53.440
LIU: OK, how about a non-STEM
research topic? [LAUGHS]
00:35:53.440 --> 00:35:55.540
WHITE: Anthropology.
00:35:55.540 --> 00:36:01.600
LIU: OK, OK, fine. So, like, no, there's no
scenario where research is not in the picture.
00:36:01.600 --> 00:36:07.200
WHITE: No, it just seems to be part of me,
you know, part of my drive, you know, part of
00:36:07.200 --> 00:36:14.160
what gives me energy, keeps me motivated
and curious. And this, you know, frontier,
00:36:14.160 --> 00:36:17.680
it allows that evolution to keep unfolding.
00:36:17.680 --> 00:36:23.920
LIU: OK. All right. Good to know. So not
yoga teacher. Always some form of research.
00:36:23.920 --> 00:36:28.160
WHITE: Well, I mean, yoga teacher is a
kind of researcher into the universe.
00:36:28.160 --> 00:36:32.800
LIU: What?! [LAUGHS] WHITE: You know, it's a researcher
into the states of consciousness and being.
00:36:32.800 --> 00:36:38.800
Into the alignment between the mind and the body.
It's full of research. Full of personal research.
00:36:38.800 --> 00:36:48.560
LIU: This is why I believe when once upon a time
at a leadership offsite you were asked to prompt
00:36:48.560 --> 00:36:54.640
AI to demonstrate, to build, like, a mascot for
you, and I think that mascot was also doing yoga.
00:36:54.640 --> 00:36:59.600
WHITE: It was. I asked my … I asked one
of my copilots to generate a picture of
00:36:59.600 --> 00:37:04.520
me based on everything it knows,
and it created a cat doing yoga.
00:37:04.520 --> 00:37:11.520
LIU: [LAUGHS] I love it. OK, so lastly,
and this is less “lightning round” and
00:37:11.520 --> 00:37:16.880
more like “take your time.” What would you
like listeners of this podcast to take away
00:37:16.880 --> 00:37:21.360
from this conversation? You know,
about you, about professional life,
00:37:21.360 --> 00:37:25.640
about research, about making … or
maybe even just making choices.
00:37:25.640 --> 00:37:32.800
WHITE: You know, I think it's about the
power of adaptability and the diversity
00:37:32.800 --> 00:37:40.000
of experience. You know, it's about
being open to what wants to happen,
00:37:40.000 --> 00:37:47.640
what needs to happen. You know, being open,
called to serve, willingness to change.
00:37:47.640 --> 00:37:50.000
LIU: I like that.
00:37:50.000 --> 00:37:56.000
So thank you, Chris—Dr. White—for joining me,
and to our audience, thank you for listening.
00:37:56.000 --> 00:37:58.383
WHITE: Thanks, Wei.
00:37:58.383 --> 00:37:59.360
[MUSIC]
STANDARD OUTRO: You’ve been listening to the
00:37:59.360 --> 00:38:02.480
Microsoft Research Podcast.
Check out more episodes of
00:38:02.480 --> 00:38:11.680
the podcast at aka.ms/researchpodcast or
on YouTube and major podcast platforms.
00:38:11.680 --> 00:38:12.204
[MUSIC ENDS]
