WEBVTT

00:00:00.000 --> 00:00:03.423
[TEASER]
[MUSIC PLAYS UNDER DIALOGUE]

00:00:03.423 --> 00:00:10.040
JACKI O’NEILL: I love living in different 
places, and those experiences are what help  

00:00:10.040 --> 00:00:15.240
us innovate better and design things that 
are, like, taking another point of view,  

00:00:15.240 --> 00:00:27.888
more creative, I think. Just sparks things in 
your, in your head. And, I mean, it's so much fun.

00:00:27.888 --> 00:00:27.903
[TEASER ENDS] 

00:00:27.903 --> 00:00:32.160
JOHANNES GEHRKE: Microsoft Research works at 
the cutting edge. But how much do we know about  

00:00:32.160 --> 00:00:36.960
the people behind the science and technology 
that we create? This is What’s Your Story,  

00:00:36.960 --> 00:00:42.320
and I’m Johannes Gehrke. In my 10 years 
with Microsoft, across product and research,  

00:00:42.320 --> 00:00:45.800
I’ve been continuously excited and 
inspired by the people I work with,  

00:00:45.800 --> 00:00:50.760
and I’m curious about how they became the 
talented and passionate people they are today.  

00:00:50.760 --> 00:00:56.440
So I sat down with some of them. Now, I’m sharing 
their stories with you. In this podcast series,  

00:00:56.440 --> 00:01:00.560
you’ll hear from them about how they grew up, 
the critical choices that shaped their lives,  

00:01:00.560 --> 00:01:06.385
and their advice to others 
looking to carve a similar path.

00:01:06.385 --> 00:01:06.400
[MUSIC FADES] 

00:01:06.400 --> 00:01:11.360
In this episode, I’m talking with Jacki 
O'Neill, director of the Microsoft Africa  

00:01:11.360 --> 00:01:18.600
Research Institute—or MARI, for short—in 
Nairobi, Kenya. Jacki’s decadelong career  

00:01:18.600 --> 00:01:22.760
at Microsoft began at the company’s 
India research lab, where she applied  

00:01:22.760 --> 00:01:28.680
her ethnographic and human-computer interaction 
expertise to advancing equity in the country.

00:01:28.680 --> 00:01:34.200
After the opening of two Microsoft software 
engineering centers in Africa, Jacki made the case  

00:01:34.200 --> 00:01:40.240
for a research lab on the continent. She now leads 
the MARI team in making technology more inclusive,  

00:01:40.240 --> 00:01:45.680
a role that allows her to pursue her goal 
of positive local change with global impact.  

00:01:45.680 --> 00:01:50.480
Here’s my conversation with Jacki, beginning 
with her time growing up in Plymouth, England.

00:01:50.480 --> 00:01:54.280
GEHRKE: We just had a discussion 
maybe a couple of years ago, right,  

00:01:54.280 --> 00:01:58.400
when you were just in transition to Africa. 
So it’s really great to have you here and  

00:01:58.400 --> 00:02:01.400
both learn a little bit what’s happening 
there, but also to learn a bit more about  

00:02:01.400 --> 00:02:06.600
your story. Where did you grow up, and 
how did you end up here at Microsoft?

00:02:06.600 --> 00:02:12.760
O’NEILL: Yeah, thanks for asking that. I've had 
a very, well, it's definitely not been a straight  

00:02:12.760 --> 00:02:19.640
road to get here, but the windy roads are the 
most interesting ones. I grew up in Plymouth,  

00:02:19.640 --> 00:02:28.000
which is a dockyard and naval town in the 
southwest of England, so a socially deprived  

00:02:28.000 --> 00:02:32.360
working-class town. So when I was growing 
up, it was a thriving, working-class town,  

00:02:32.360 --> 00:02:38.320
but of course with those industries, you 
know, they didn't, they didn't pass so well  

00:02:38.320 --> 00:02:43.080
through those years. So, you know, by the time 
I was leaving school, it was quite a deprived  

00:02:43.080 --> 00:02:53.440
city and still is. I think that it's really 
important to be in those type of places, though,  

00:02:53.440 --> 00:03:00.632
because you get a very rich view of life, and 
I left them as soon as I could, [LAUGHS] so ...

00:03:00.632 --> 00:03:01.580
GEHRKE: When you went to university?

00:03:01.580 --> 00:03:06.120
O’NEILL: Went to, well, I went and I was 
a cook for a year in the Lake District,  

00:03:06.120 --> 00:03:09.560
which is a very beautiful part of 
the UK, and then went to university.

00:03:09.560 --> 00:03:10.800
GEHRKE: Where's the Lake District?

00:03:10.800 --> 00:03:14.920
O’NEILL: It is northwest, and 
it's all hills. It's, like,  

00:03:14.920 --> 00:03:21.160
Wordsworth Country. It's all hills and 
poetry and beautiful houses. And, yeah,  

00:03:21.160 --> 00:03:26.780
it was a fantastic time working as a cook there. 
And then I went to Manchester to do my degree.

00:03:26.780 --> 00:03:28.720
GEHRKE: OK. And what is your degree in?

00:03:28.720 --> 00:03:34.720
O’NEILL: Ah, so, yes, I had, I did a 
social science degree to start with.  

00:03:34.720 --> 00:03:38.240
I started at the time when you could get 
a degree in anything and get any job at  

00:03:38.240 --> 00:03:41.980
the end of it. But by the time I came 
out of my degree, it was a recession.

00:03:41.980 --> 00:03:46.600
GEHRKE: But did you have, did you have specific 
plans while you were studying of what you want,  

00:03:46.600 --> 00:03:48.640
you know, what profession you wanted to go into?

00:03:48.640 --> 00:03:53.120
O’NEILL: Not really. I didn't. I think 
I'd, I think like many young people,  

00:03:53.120 --> 00:03:57.720
I didn't really know, but I felt that 
I would find something interesting when  

00:03:57.720 --> 00:04:04.440
I came out. And then, you know, I just 
worked lots of different jobs. [LAUGHS]

00:04:04.440 --> 00:04:06.240
GEHRKE: What is your favorite college course?

00:04:06.240 --> 00:04:11.080
O’NEILL: My favorite college 
course—in my degree? Gosh,  

00:04:11.080 --> 00:04:15.272
that's a good question. It 
was all so long ago. [LAUGHS]

00:04:15.272 --> 00:04:16.389
GEHRKE: OK …

00:04:16.389 --> 00:04:22.640
O’NEILL: My favorite, I guess, yeah, no, I, so, 
I did ... my degree was in psychology. I worked,  

00:04:22.640 --> 00:04:28.369
and then I did my master's in computer science 
and then my PhD in human-computer interaction.

00:04:28.369 --> 00:04:30.620
GEHRKE: That's quite a change, right, from 
psychology into computer science, then.

00:04:30.620 --> 00:04:34.880
O’NEILL: Yes, yes. And I just, you know, 
I'd always just wanted to do computing,  

00:04:34.880 --> 00:04:40.280
but when I was at school, it was ... we had one 
computer in the school, and so it was, like,  

00:04:40.280 --> 00:04:46.200
a computer at home or you don't do computer 
science. So, you know, I didn't do it.

00:04:46.200 --> 00:04:46.509
GEHRKE: Right.

00:04:46.509 --> 00:04:52.480
O’NEILL: So then as computers became 
more prominent, more available, you know,  

00:04:52.480 --> 00:04:56.880
I was working in libraries, and they started 
computerizing, and I worked on that project,  

00:04:56.880 --> 00:05:01.120
and then that led me to do a master's. And 
so I was like, hey, this is the opportunity  

00:05:01.120 --> 00:05:07.280
to really get into this area, and I loved it. It 
was fantastic. And Manchester's computer science  

00:05:07.280 --> 00:05:14.080
department is one of the top departments, and 
I had an amazing ... Carole Goble was my thesis  

00:05:14.080 --> 00:05:19.480
supervisor. She was absolutely amazing and strong 
for women in computing. But at the end of it,  

00:05:19.480 --> 00:05:24.080
I was like, OK, so I didn't want to 
do pure social science and I didn't  

00:05:24.080 --> 00:05:30.520
want to do pure computer science. What I 
want to do is do human-computer science,  

00:05:30.520 --> 00:05:35.520
so where you really merge the two. And that's how 
I got into HCI, and I think that's where I started  

00:05:35.520 --> 00:05:42.280
finding my favorite courses. You know, I loved the 
research methods. I loved those types of things.

00:05:42.280 --> 00:05:43.900
GEHRKE: And what is your PhD about?

00:05:43.900 --> 00:05:47.360
O’NEILL: Ooh, it was very boring. [LAUGHTER]  

00:05:47.360 --> 00:05:51.944
My PhD was in computer-supported 
cooperative work [CSCW], and ...

00:05:51.944 --> 00:05:53.560
GEHRKE: OK. Oh, yeah. Very relevant now, right?

00:05:53.560 --> 00:05:59.000
O’NEILL: Yeah, very relevant now. And that 
was a really exciting time for CSCW, as well,  

00:05:59.000 --> 00:06:04.800
because there were so many different labs. 
There were Sun Systems, there was Xerox,  

00:06:04.800 --> 00:06:10.720
there was Microsoft—all doing really cool, like, 
collaborative technologies. So it seemed like a  

00:06:10.720 --> 00:06:19.747
brilliant area to go into. But I was looking at, 
can we support networking events for businesses?

00:06:19.747 --> 00:06:20.509
GEHRKE: Wow. Uh-huh …

00:06:20.509 --> 00:06:24.000
O’NEILL: So it was just at the 
time of the first, you know,  

00:06:24.000 --> 00:06:28.986
things like Webex and things, you know, 
the first collaborative seminar-y ...

00:06:28.986 --> 00:06:31.970
GEHRKE: Yeah, so you're way ahead of the 
social networks, right, and everything, right?

00:06:31.970 --> 00:06:31.992
O’NEILL: Yeah, yeah.

00:06:31.992 --> 00:06:34.000
GEHRKE: And there was a whole 
conference at that point in time,  

00:06:34.000 --> 00:06:36.549
right? CSCW, I think I remember. Wasn't there ...

00:06:36.549 --> 00:06:36.992
O’NEILL: Yes, yes, yes.

00:06:36.992 --> 00:06:39.260
GEHRKE: So it was and still is, 
I think, a really big field.

00:06:39.260 --> 00:06:43.680
O’NEILL: Yes, it's a, it's a, it's really 
interesting. And I think one of the things  

00:06:43.680 --> 00:06:49.160
that's interesting with the foundational models 
now is many of the things that people like me,  

00:06:49.160 --> 00:06:53.680
HCI people, have been wanting to happen—"Oh, 
if only we can enable people to interact with  

00:06:53.680 --> 00:06:59.520
technology like this"—are now suddenly 
possible, which is quite exciting.

00:06:59.520 --> 00:07:03.080
GEHRKE: Yeah, so we'll get to that in a 
little bit because I think, you know, as you  

00:07:03.080 --> 00:07:08.160
said the whole field of HCI is now changing with 
foundational models and what the interfaces are,  

00:07:08.160 --> 00:07:13.080
will be. I think it's a really interesting, deep 
research question right now. So, so, OK, so you  

00:07:13.080 --> 00:07:19.100
got your PhD; you're in Manchester. What's the 
next step in your career? Where did you go next?

00:07:19.100 --> 00:07:24.440
O’NEILL: Yeah, I actually got a job before I 
finished my PhD. So I took quite a long time  

00:07:24.440 --> 00:07:31.640
to do my PhD. I think it was seven years in the 
end, partly because I was teaching. When I was  

00:07:31.640 --> 00:07:39.600
doing—like, lecturing when I was doing my PhD, 
and I also had a job as a consultant occasionally,  

00:07:39.600 --> 00:07:45.920
working with, I think, I worked with the Co-op 
Bank. I worked with some usability companies,  

00:07:45.920 --> 00:07:51.160
and you could, I could make enough 
money to live for a term on, like,  

00:07:51.160 --> 00:07:55.105
two weeks' consultancy because I 
didn't have very high costs. [LAUGHS]

00:07:55.105 --> 00:07:56.420
GEHRKE: Right. You lived as a grad student, right?

00:07:56.420 --> 00:08:00.040
O’NEILL: Yes. Yeah. Yeah. And, actually, 
you know, I was living in Manchester. I  

00:08:00.040 --> 00:08:04.312
was living in a squat, so I wasn't 
paying any rent, [LAUGHS] so ...

00:08:04.312 --> 00:08:04.960
GEHRKE: Oh, really?

00:08:04.960 --> 00:08:07.320
O’NEILL: Yes. So I didn't have very many costs.

00:08:07.320 --> 00:08:07.920
GEHRKE: OK.

00:08:07.920 --> 00:08:15.120
O’NEILL: Which was very handy. So I didn't 
have any real incentive to finish my PhD  

00:08:15.120 --> 00:08:20.280
until I got a job, you know. When I finished my 
master's, I looked at the job market, and with  

00:08:20.280 --> 00:08:27.140
my computer science master's, the main job was 
database manager, [LAUGHS] which didn't appeal.

00:08:27.140 --> 00:08:29.480
GEHRKE: That sounds now 
really interesting. [LAUGHTER]

00:08:29.480 --> 00:08:33.520
O’NEILL: Yeah. So I, actually, that's why 
I ended up doing a PhD, because I was like,  

00:08:33.520 --> 00:08:39.400
I don't want to go back to work yet. You know, 
I've been working for five years before. So, so,  

00:08:39.400 --> 00:08:46.200
yeah, I just was enjoying doing a PhD and 
doing pieces of work here and there. And  

00:08:46.200 --> 00:08:51.560
then I got a job at Xerox in Cambridge, 
and then that's when I got motivated to  

00:08:51.560 --> 00:08:55.840
finish my PhD because working and doing 
a PhD at the same time is not much fun.

00:08:55.840 --> 00:09:00.280
GEHRKE: Right, right. So you got 
your PhD, had your job lined up,  

00:09:00.280 --> 00:09:03.120
and then you're starting at Xerox. 
What were you doing in Xerox?

00:09:03.120 --> 00:09:09.760
O’NEILL: Human-computer interaction. 
Yeah, it was a really exciting time.  

00:09:09.760 --> 00:09:14.800
There was so much going on in the industry. 
I was so delighted. It was like my dream job  

00:09:14.800 --> 00:09:19.920
to be in industry and to maybe create cool 
interfaces and, you know, cool collaborative  

00:09:19.920 --> 00:09:26.580
systems. So ... and then they closed the lab 
[LAUGHS] within six months. It wasn't my fault.

00:09:26.580 --> 00:09:27.280
GEHRKE: So quickly?

00:09:27.280 --> 00:09:28.040
O’NEILL: Mm-hmm.

00:09:28.040 --> 00:09:31.669
GEHRKE: Wow. And what did you do then? I 
mean, this is your first big job, and ...

00:09:31.669 --> 00:09:32.880
O’NEILL: Yes ...
GEHRKE: ... such a quick setback.

00:09:32.880 --> 00:09:39.360
O’NEILL: They offered me a job in their lab in 
France. So I stayed in the UK for a while and  

00:09:39.360 --> 00:09:43.110
worked half in France, half in the UK, 
and then I shifted to France full time.

00:09:43.110 --> 00:09:45.800
GEHRKE: OK. Oh, wow. So do you ... 
where in France did you live then?

00:09:45.800 --> 00:09:46.360
O’NEILL: Grenoble.

00:09:46.360 --> 00:09:48.354
GEHRKE: OK, yeah. In the middle of ...

00:09:48.354 --> 00:09:48.384
O’NEILL: In the French Alps.

00:09:48.384 --> 00:09:50.814
GEHRKE: ... the French Alps. 
Exactly. Beautiful place.

00:09:50.814 --> 00:09:54.440
O’NEILL: Absolutely ... yes. Yeah. 
Skiing, climbing, hiking. So much fun.

00:09:54.440 --> 00:09:59.320
GEHRKE: And, OK, so you're at Xerox PARC 
in the French Alps. What's, what's next?

00:09:59.320 --> 00:10:06.000
O’NEILL: They were opening, Xerox was opening 
a research lab in India. And I'd always wanted  

00:10:06.000 --> 00:10:10.520
to travel. You know, I'd always wanted ... and I 
never really had the money or the opportunity to  

00:10:10.520 --> 00:10:16.080
travel. So when they said they were opening it, I 
just went to my boss and said, hey, I don't know  

00:10:16.080 --> 00:10:20.333
what you'd want me to do, but if there's any 
opportunities for me to do anything to help …

00:10:20.333 --> 00:10:21.400
GEHRKE: Wow.
O’NEILL: … the opening of India,

00:10:21.400 --> 00:10:26.940
I'd love to. And I went out for a month 
and then I went out for three months.

00:10:26.940 --> 00:10:30.320
GEHRKE: I mean, both of these sound like 
really bold steps to me. First of all,  

00:10:30.320 --> 00:10:32.880
I mean, Grenoble is probably pure French speaking,  

00:10:32.880 --> 00:10:35.749
right? And, I don't know, did you have high 
school French or you were good ... [LAUGHS]

00:10:35.749 --> 00:10:40.600
O’NEILL: I had high school French, 
yes, and then we drove, we drove  

00:10:40.600 --> 00:10:45.915
from the UK to Grenoble listening 
to "learn French" tapes [LAUGHS] …

00:10:45.915 --> 00:10:47.872
GEHRKE: OK, wow … [LAUGHS]
O’NEILL: …in the car. Yeah.

00:10:47.872 --> 00:10:51.589
GEHRKE: Wow. And that was enough 
then to get by with a daily ...

00:10:51.589 --> 00:10:55.120
O’NEILL: Actually, so it was great in France 
because they expect you to learn the language,  

00:10:55.120 --> 00:11:01.040
so you have French lessons at work. And 
then, actually, I did an evening class,  

00:11:01.040 --> 00:11:05.760
as well, that was paid for by work, a really 
intensive one-month, like two hours a night,  

00:11:05.760 --> 00:11:09.820
every night of the week. And that really 
helped. Yeah, it was, it's fantastic.

00:11:09.820 --> 00:11:13.000
GEHRKE: Wow, that's really great. And 
then, and then you took the even bigger  

00:11:13.000 --> 00:11:19.300
step to move to India, right. How was that 
like, and what was your experience there?

00:11:19.300 --> 00:11:23.440
O’NEILL: Yeah, India is just magical. You 
know, initially, I just went for one month,  

00:11:23.440 --> 00:11:28.320
then three months, and it was just—the 
people, the culture, the work I was doing,  

00:11:28.320 --> 00:11:35.120
the research I was doing was like no research ... 
you know, I’d spent a lot of time in call centers  

00:11:35.120 --> 00:11:41.000
around Europe doing studies, ethnographic studies, 
and designing technology. Lots of time looking at  

00:11:41.000 --> 00:11:46.840
photocopiers because I was with Xerox. [LAUGHS] 
And then so going to India, suddenly, you know,  

00:11:46.840 --> 00:11:54.160
I'm looking at social enterprises. I'm looking 
at all sorts of businesses and different ways  

00:11:54.160 --> 00:11:59.760
of life and different people. And it was just 
so rich and so amazing that I was like, OK,  

00:11:59.760 --> 00:12:06.120
I really want to do this. And that's actually when 
I applied to Microsoft because Microsoft had the  

00:12:06.120 --> 00:12:11.809
Technology for Emerging Markets group there, 
which is world-class research in that space.

00:12:11.809 --> 00:12:11.829


00:12:11.829 --> 00:12:15.360
So I was like, OK, if I want to 
keep on doing this, then that's what I'm  

00:12:15.360 --> 00:12:19.700
going to apply to. And luckily enough, I got 
the job, and that's how I joined Microsoft.

00:12:19.700 --> 00:12:24.080
GEHRKE: Wow. So, so, OK, so you're now at 
Microsoft in India. That was in Bangalore,  

00:12:24.080 --> 00:12:26.174
right, where our research lab there is?

00:12:26.174 --> 00:12:26.192
O’NEILL: Mm-hmm.

00:12:26.192 --> 00:12:29.160
GEHRKE: And so what, what were you 
working on there for the next few years?

00:12:29.160 --> 00:12:36.280
O’NEILL: Yeah. So initially, I looked at a few 
different things. I joined some existing projects.  

00:12:36.280 --> 00:12:42.240
So I was on MEC, which was the educational 
platform, looking at whether we could bring the  

00:12:42.240 --> 00:12:50.000
power of MOOCs [Massive Open Online Courses] to 
Indian education to improve the level of education  

00:12:50.000 --> 00:12:55.080
because they have amazing colleges at the top, 
but, actually, the vast majority of students go  

00:12:55.080 --> 00:13:01.600
to these intermediate colleges, and the teaching 
level really varies. And so the idea was, can  

00:13:01.600 --> 00:13:07.640
you help with blended learning? Can you help the 
teachers teach better? That turns out to be really  

00:13:07.640 --> 00:13:12.640
challenging. And, actually, the system ended up 
being used by the students to teach themselves.

00:13:12.640 --> 00:13:15.813
GEHRKE: Oh, like for independent learning?

00:13:15.813 --> 00:13:17.680
O’NEILL: Mm-hmm. Mm-hmm. And that 
was really, so that was interesting,  

00:13:17.680 --> 00:13:24.760
doing some studies there. I looked at ... Indrani 
[Medhi Thies] had done an amazing project where  

00:13:24.760 --> 00:13:30.200
they'd built "Facebook for Farmers." So 
I did a study of that, which was really,  

00:13:30.200 --> 00:13:35.240
really fun. And then I worked in financial 
inclusion, one of my big areas. I spent about  

00:13:35.240 --> 00:13:40.480
five years working with auto-rickshaw drivers 
in Bangalore, designing technologies to help  

00:13:40.480 --> 00:13:46.880
them understand the loans they'd taken out, 
which was really, really fun. They're a very  

00:13:46.880 --> 00:13:52.352
great community to work [with]. You don't get any 
nonsense from an auto-rickshaw driver. [LAUGHS]

00:13:52.352 --> 00:13:53.280
GEHRKE: Well, I was just thinking,  

00:13:53.280 --> 00:13:57.640
what was it like to, like, live in India 
and just move there and start out there?

00:13:57.640 --> 00:14:00.760
O’NEILL: Uh, it was, I mean, it 
was fantastic. It's a great place  

00:14:00.760 --> 00:14:06.200
to live. The people are amazing. The food 
is amazing. Moving with Microsoft makes it  

00:14:06.200 --> 00:14:10.920
very easy because Microsoft takes care of 
you when you move so you're not, you know,  

00:14:10.920 --> 00:14:18.880
some of the stresses that you might have around 
the move are taken care of. I had a young family.  

00:14:18.880 --> 00:14:26.080
I had a 2-year-old son when we moved out 
there and within a year had another one,  

00:14:26.080 --> 00:14:30.000
which was not 100 percent planned, because you 
don't usually move to a new company and then  

00:14:30.000 --> 00:14:35.160
have a baby. You're like, oh, sorry. 
[LAUGHS] But that was all fine. Yeah.

00:14:35.160 --> 00:14:40.920
GEHRKE: And, and, you know, you worked with 
all of these different communities in India,  

00:14:40.920 --> 00:14:45.789
right. How did you connect to the 
communities? I mean, these were teachers …

00:14:45.789 --> 00:14:49.920
O’NEILL: Yeah, you need to, you really need 
to go with people so you have to convince some  

00:14:49.920 --> 00:14:55.520
organization that what you're going to do is going 
to be beneficial to them and useful for them. And  

00:14:55.520 --> 00:15:01.600
then if they're trusted by the community, they 
give you access. And that's really great because  

00:15:01.600 --> 00:15:06.400
you do have access that you wouldn't otherwise 
have. You know, if you're really wanting to build  

00:15:06.400 --> 00:15:11.360
technologies to support people, you really need to 
understand what they care about—what do they want  

00:15:11.360 --> 00:15:17.600
help with?—and you only get that if you've got a 
trusted relationship with them. So we worked with,  

00:15:17.600 --> 00:15:21.840
there was one organization that worked with 
the auto-rickshaw drivers' wives. It was about  

00:15:21.840 --> 00:15:26.360
empowering women, and we got access to the 
drivers initially through that organization.

00:15:26.360 --> 00:15:30.000
GEHRKE: That's amazing. I mean, you 
know, I've visited India many times,  

00:15:30.000 --> 00:15:33.360
but I can only imagine how it is to live there, 
actually. So do you have some of the stories  

00:15:33.360 --> 00:15:38.560
of what is, sort of, most surprising 
for you given that you've lived there?

00:15:38.560 --> 00:15:44.200
O’NEILL: Yeah … what's most surprising? 
I think, so one thing is, one thing is  

00:15:46.680 --> 00:15:53.720
people want to tell you what they think 
you want to hear. So if you're lost,  

00:15:53.720 --> 00:16:01.040
you need to ask quite a few people for directions 
and then make some sort of assessment about  

00:16:01.040 --> 00:16:05.680
whether the person was just saying "yes, yes, 
that way" because he knew the way or "yes,  

00:16:05.680 --> 00:16:09.480
yes that way" because he just didn't want to 
tell you that he didn't know. And so you have to,  

00:16:09.480 --> 00:16:13.872
sort of, judge. [LAUGHS] So that's 
one, like, useful piece of ...

00:16:13.872 --> 00:16:16.544
GEHRKE: So the first few times you 
went in the wrong direction? [LAUGHS]

00:16:16.544 --> 00:16:18.920
O’NEILL: Yes, exactly. And then you're like, 
"But they said ..."; you ask someone else,  

00:16:18.920 --> 00:16:24.400
and they're like, "No, it's over there." And 
then someone ... so that's … the most useful  

00:16:24.400 --> 00:16:28.440
piece of advice I could give to anyone who's 
visiting India, is when you cross the road,  

00:16:28.440 --> 00:16:31.780
just find someone else who's already 
crossing the road and cross with them.

00:16:31.780 --> 00:16:34.680
GEHRKE: Because it's so dangerous 
if you go by yourself potentially?

00:16:34.680 --> 00:16:37.920
O’NEILL: Yes, yeah. You get used to it quite 
quickly, and there's obviously something that  

00:16:37.920 --> 00:16:42.840
changes in you when you've been there a 
while. You know, when you first go there,  

00:16:42.840 --> 00:16:46.280
all the auto-rickshaw drivers are going to 
overcharge you and drive around the block  

00:16:46.280 --> 00:16:51.480
twice and all of those things. And I find after 
about four to five weeks when you've been there,  

00:16:51.480 --> 00:16:56.120
they know, like, there must be something 
that changes in your attitude because they  

00:16:56.120 --> 00:17:00.880
actually know that you're there longer term 
and you're not going to take any nonsense.

00:17:00.880 --> 00:17:03.680
GEHRKE: So, so do you behave 
differently? What's the change there?

00:17:03.680 --> 00:17:06.920
O’NEILL: I don't know. That's, I've tried to 
think about this, but I think, I don't know,  

00:17:06.920 --> 00:17:11.200
it must be just an air of confidence or 
an air of certainty or something. But,  

00:17:11.200 --> 00:17:15.529
yeah, it's like something just clicks or changes.

00:17:15.529 --> 00:17:16.520
GEHRKE: That's so interesting. Is it only for the  

00:17:16.520 --> 00:17:19.520
drivers, or is it in other aspects of 
your life, as well, where, sort of,  

00:17:19.520 --> 00:17:23.180
you get treated differently because 
you suddenly have become a native?

00:17:23.180 --> 00:17:26.440
O’NEILL: I think you notice it most 
in the drivers because they're the  

00:17:26.440 --> 00:17:29.360
ones that you're interacting so 
much with to get about, you know,  

00:17:29.360 --> 00:17:33.680
to get ... you're always getting a tuk-tuk to go 
from here to there. And they really do, you know,  

00:17:33.680 --> 00:17:37.267
if they can make extra money out of you, they 
are going to make extra money out of you.

00:17:37.267 --> 00:17:40.617
GEHRKE: They smell it, that you're a tourist.

00:17:40.617 --> 00:17:42.192
O’NEILL: Yeah, yeah, yes. [LAUGHS]

00:17:42.192 --> 00:17:47.080
GEHRKE: And then so you were in India 
and then another opportunity came along.  

00:17:47.080 --> 00:17:50.040
So tell us a little bit about that 
opportunity, where you ended up now.

00:17:50.040 --> 00:17:57.000
O’NEILL: Yes, yes. So when I heard that the 
ADCs were opening—the Africa Development Center,  

00:17:57.000 --> 00:18:03.680
so our software engineering center in Nairobi 
and Lagos—I thought that that was a great time  

00:18:03.680 --> 00:18:11.760
to pitch for research in Africa for Microsoft. 
It seemed like a bit of a hole in our portfolio.  

00:18:11.760 --> 00:18:16.200
I have family connections to Africa. So, 
actually, one of the reasons for joining  

00:18:16.200 --> 00:18:21.440
Microsoft was partly because I thought there 
might be opportunities eventually in Africa  

00:18:21.440 --> 00:18:25.440
because we had a great Africa startup 
program, for example. So, you know,  

00:18:25.440 --> 00:18:32.240
but there wasn't any research there. And so when 
I heard the ADCs were open, I just put together a,  

00:18:32.240 --> 00:18:38.880
like, pitch for setting up research in Africa 
within the ADCs, and, you know, all sorts of  

00:18:38.880 --> 00:18:49.388
people really helped me hone that pitch. And then 
I flew at the end of February 2020. I flew ...

00:18:49.388 --> 00:18:51.261
GEHRKE: Oh, just right before the pandemic.

00:18:51.261 --> 00:18:56.320
O’NEILL: Mm-hmm. I flew to ... I was in Barcelona 
for a Future of Work event, and then I flew to  

00:18:56.320 --> 00:19:02.160
Nairobi and then Lagos to meet the people who 
were running the ADCs and to think about where,  

00:19:02.160 --> 00:19:08.560
which one I would want to set up research in if 
such a thing were to happen. And I did that. I  

00:19:08.560 --> 00:19:15.960
decided that Nairobi was the right one. And 
when I went there, Jack Ngare ran the ADC,  

00:19:15.960 --> 00:19:23.240
and he was so enthusiastic about having 
research there. So I did a pitch and got  

00:19:23.240 --> 00:19:28.280
some funding just—I think if it had been two 
weeks later, I'm not sure. But, you know,  

00:19:28.280 --> 00:19:33.740
it was just before we knew how bad COVID was 
going to be, so I was very lucky with timing.

00:19:33.740 --> 00:19:37.280
GEHRKE: And, I mean, you've made these 
amazing moves throughout your career,  

00:19:37.280 --> 00:19:42.120
right. You, sort of, raised your hand for India 
when the lab was open; now here in Africa. Why,  

00:19:42.120 --> 00:19:46.000
and how? I'm just, I mean, so curious because 
people make the most unexpected turns in their  

00:19:46.000 --> 00:19:50.160
careers from time to time. But it's more like 
because, you know, they lose their current job  

00:19:50.160 --> 00:19:53.640
or they, their manager moves away and they 
really think about their career. But you,  

00:19:53.640 --> 00:19:58.480
like, raise your hand from time to time and 
make these really bold and amazing moves.

00:19:58.480 --> 00:20:01.735
O’NEILL: Yeah, I mean, life's 
meant to be exciting, isn't it?

00:20:01.735 --> 00:20:01.749
GEHRKE: OK …

00:20:01.749 --> 00:20:07.000
O’NEILL: I think. You know, life's meant to be 
exciting. I love living in different places and,  

00:20:07.000 --> 00:20:10.960
you know, as an ethnographer, as a person 
interested in human-computer interaction,  

00:20:10.960 --> 00:20:18.960
it's, like, those experiences are what help us 
innovate better and design things that are, like,  

00:20:18.960 --> 00:20:24.880
taking another point of view, more creative, I 
think. Like, just sparks things in your, in your  

00:20:24.880 --> 00:20:28.640
head. And, I mean, it's so much fun. Like, I don't 
understand why everyone doesn't do it. [LAUGHS]

00:20:28.640 --> 00:20:34.040
GEHRKE: So it's just really amazing. 
So if I think about, you know, India,  

00:20:34.040 --> 00:20:38.320
where you said, right, the experience 
for you was that the drivers were  

00:20:38.320 --> 00:20:41.560
treating you suddenly differently. Did 
you have a similar experience in Africa,  

00:20:41.560 --> 00:20:45.700
or what is one of the or a few of the 
defining experiences and stories there?

00:20:45.700 --> 00:20:53.440
O’NEILL: Yeah, I think ... so the animals are 
amazing in Kenya. They've done such an amazing  

00:20:53.440 --> 00:20:58.600
job at conservation. I imagine that they 
would, you would only see, like, these big  

00:20:58.600 --> 00:21:03.240
animals in the national parks, but—they're 
not everywhere. They're not going to be,  

00:21:03.240 --> 00:21:07.840
you're not going to find a hippo walking down 
the road in Nairobi. But they are all over the  

00:21:07.840 --> 00:21:15.200
place. So you can go camping in Lake Naivasha, 
which is just an hour and a half from Nairobi,  

00:21:15.200 --> 00:21:21.520
and I was camping with a friend, and the kids 
were in their tent, and my friend was in her tent,  

00:21:21.520 --> 00:21:24.720
and I was just sitting by the fire. 
It's about 10 o'clock. I said, yeah,  

00:21:24.720 --> 00:21:31.360
I might go to bed in a minute. And then I just 
heard this snort, and I get up with my torch,  

00:21:31.360 --> 00:21:39.810
and I look, and there's a hippo, [LAUGHS] 
like, probably less than a meter and a half …

00:21:39.810 --> 00:21:42.240
GEHRKE: Wow …
O’NEILL: … away from me. So I carefully went and

00:21:42.240 --> 00:21:46.872
sat back down by the fire and waited 
for a while before I moved. [LAUGHS]

00:21:46.872 --> 00:21:49.800
GEHRKE: So are they dangerous in that 
aspect, if you've startled them or so ... ?

00:21:49.800 --> 00:21:52.800
O’NEILL: Yeah, I think ... they say 
that you should never get between a  

00:21:52.800 --> 00:21:57.040
hippo and the water. So, luckily, I was 
on the other side of the, [LAUGHS] of  

00:21:57.040 --> 00:22:01.360
the hippo and the water. But they are 
big. I mean, they can be very grumpy.

00:22:01.360 --> 00:22:04.840
GEHRKE: And so you should, just, shouldn't startle 
them or ... ? I'm just trying to understand what's  

00:22:04.840 --> 00:22:07.840
the recommended behavior. Don't get 
between the hippo and the water.

00:22:07.840 --> 00:22:11.320
O’NEILL: Yes, that's recommended, 
and don't, yeah, don't startle them,  

00:22:11.320 --> 00:22:15.400
and just, you know, stay very, stay very 
calm. So, actually, when you're camping,  

00:22:15.400 --> 00:22:20.360
if you don't have an electric fence around 
the campsite, then you shouldn't come out of  

00:22:20.360 --> 00:22:28.760
your tent at night. So don't drink too much beer 
before you go to bed, [LAUGHTER] because it's the  

00:22:28.760 --> 00:22:32.800
"zip." When you unzip it, you can really 
startle ... If there's any wild animals,  

00:22:32.800 --> 00:22:36.940
lions, or whatever around, then you can really 
scare them. And you don't want to scare a lion.

00:22:36.940 --> 00:22:39.040
GEHRKE: Yeah, I was thinking, just, 
actually, about the lions or so,  

00:22:39.040 --> 00:22:42.660
right. I mean, they could be probably even more 
dangerous than the hippos or, or not really?

00:22:42.660 --> 00:22:47.360
O’NEILL: Hippos are actually more dangerous 
than lions. Yeah, lions will generally not  

00:22:47.360 --> 00:22:51.800
attack you. And apparently, the thing—I haven't 
had to try this, I'm glad to say—but the thing  

00:22:51.800 --> 00:22:57.000
you should do if you encounter a lion is just 
look them in the eye, and then they'll go off.

00:22:57.000 --> 00:22:57.949
GEHRKE: Stare them down.

00:22:57.949 --> 00:22:58.440
O’NEILL: Mm-hmm.

00:22:58.440 --> 00:22:59.040
GEHRKE: OK.

00:22:59.040 --> 00:23:00.512
O’NEILL: I hope I never have to try that 
because they are quite scary … [LAUGHS]

00:23:00.512 --> 00:23:03.749
GEHRKE: I hope I never have 
to do that but good advice …

00:23:03.749 --> 00:23:07.400
O’NEILL: Yes, yeah, yeah. I think hippos 
are more likely to charge at you. Like,  

00:23:07.400 --> 00:23:10.140
a lion's more likely to go 
off in the other direction.

00:23:10.140 --> 00:23:12.920
GEHRKE: And what's the daily life like, you know,  

00:23:12.920 --> 00:23:17.320
living in Nairobi, right? I mean, 
is it, I mean, it must be very,  

00:23:17.320 --> 00:23:21.560
very different from living in both India, 
as well as, you know, Great Britain or here.

00:23:21.560 --> 00:23:28.240
O’NEILL: Yeah. I mean it is very different. The 
traffic's bad but not as crazy as India. Like,  

00:23:28.240 --> 00:23:32.560
I drive in Kenya. I didn't drive 
in India because it was a bit too  

00:23:32.560 --> 00:23:40.600
scary with the bikes and everything. It's 
a really, it's a really nice pace, I think,  

00:23:40.600 --> 00:23:48.240
in Nairobi. It's a beautiful city. There's 
nightlife, and there's cafes and restaurants,  

00:23:48.240 --> 00:23:51.760
but you've got countryside so close. You 
know, compared to Bangalore, it's quite  

00:23:51.760 --> 00:23:59.360
a small city. And the weather is amazing, 
and the people are really friendly and kind,  

00:23:59.360 --> 00:24:03.180
and, you know, it's just, it's a very 
nice, it's a very nice place to live.

00:24:03.180 --> 00:24:04.080
GEHRKE: That's amazing,  

00:24:04.080 --> 00:24:11.777
and you now are leading the Microsoft 
Africa Research Institute there, right?

00:24:11.777 --> 00:24:11.792
O’NEILL: Yes.

00:24:11.792 --> 00:24:15.581
GEHRKE: What is the focus of the 
institute, and what are you studying there?

00:24:15.581 --> 00:24:19.200
O’NEILL: Mm-hmm. Yeah, we're mainly focused 
on foundational models. It won't be a  

00:24:19.200 --> 00:24:27.560
surprise to anybody. [LAUGHS] Which actually, you 
know, it's worked out very well for us because,  

00:24:27.560 --> 00:24:33.280
you know, we have a mixed disciplinary team. 
We have HCI and AI and ML and data science.

00:24:33.280 --> 00:24:34.300
GEHRKE: And all local?

00:24:34.300 --> 00:24:45.040
O’NEILL: All local. Yeah. And, yeah, we're 
looking at multilingual languages in models. So  

00:24:45.040 --> 00:24:50.480
we're working with MSR [Microsoft Research] India, 
thinking about how can you benchmark these models  

00:24:50.480 --> 00:24:55.640
for different languages. And we're thinking all 
the way along the scale from your high-resource,  

00:24:55.640 --> 00:25:01.400
you know, French and German, to your mid-resource 
Swahili, Hindi, all the way to your low-resource  

00:25:01.400 --> 00:25:07.480
languages because, you know, the vast majority of 
training data is in English. So we've been working  

00:25:07.480 --> 00:25:11.920
a lot. That's nice because we're having, you know, 
in a very short amount of time, you know, four or  

00:25:11.920 --> 00:25:17.440
five months, we're having both scientific impact 
with papers but also product impact, working with  

00:25:17.440 --> 00:25:24.150
the Copilot Language Globalization team as they're 
rolling out Copilot in different languages.

00:25:24.150 --> 00:25:27.240
GEHRKE: I see. So the research that 
you have will go into, let's say,  

00:25:27.240 --> 00:25:32.787
Word or PowerPoint or so to make it available 
in some of the languages from the continent.

00:25:32.787 --> 00:25:37.600
O’NEILL: Yes, exactly. Because it's not just about 
translation. It's also if you think about RAI,  

00:25:37.600 --> 00:25:43.160
responsible AI, you know, a lot of that is 
language based. And so how do ... you can't  

00:25:43.160 --> 00:25:48.120
just translate this to words. You have to find 
the right list of words in those languages. And  

00:25:48.120 --> 00:25:53.080
then what about things like tone and stuff? 
So that's one area. And then related to that,  

00:25:53.080 --> 00:26:00.120
it's in a much bigger space of equity, the 
models and equity. You know, what's going to  

00:26:00.120 --> 00:26:04.440
happen to the digital divide with these models? 
In some ways, you could imagine that they may  

00:26:04.440 --> 00:26:09.600
be flattening it, but in other ways, they could 
be increasing it. So we really are trying to map  

00:26:09.600 --> 00:26:16.840
out how … the different elements of the digital 
divide as it plays out in these models. Because  

00:26:16.840 --> 00:26:22.400
you obviously have your traditional things 
like access to devices, access to, you know,  

00:26:22.400 --> 00:26:27.840
infrastructure, and things like that. But there's 
also the data divide. So not only is most of the  

00:26:27.840 --> 00:26:34.160
training material in English; it's also mostly 
from America and the Global North. So it embodies  

00:26:34.160 --> 00:26:41.400
very particular world views. And if you think 
about data on Africa, data on Africa tends to be  

00:26:41.400 --> 00:26:49.840
collected by particular organizations. So there's 
lots of data on poverty and disease and forced  

00:26:49.840 --> 00:26:58.520
migration and things like that. Not much data 
on, like, the stories, the creativity, wealth,  

00:26:58.520 --> 00:27:04.120
innovation. So what does that mean? Even if the 
models can speak perfectly, which they can't yet,  

00:27:04.120 --> 00:27:10.120
but they'll eventually get quite good at, 
you know, even smaller languages like Luo,  

00:27:10.120 --> 00:27:17.960
if that model is just translating English 
content into Luo, that's not necessarily what  

00:27:17.960 --> 00:27:21.500
we want from a model. So there's some really 
interesting questions there to be answered.

00:27:21.500 --> 00:27:24.040
GEHRKE: Well, it seems to me like 
it's clearly also a question of,  

00:27:24.040 --> 00:27:27.480
like, getting the right kind of data. So where 
do you get the data, and how do you get the data?

00:27:27.480 --> 00:27:31.960
O’NEILL: Yeah, that's a big question. And it 
was already a challenge, you know, before these  

00:27:31.960 --> 00:27:37.560
models. You know, many people have been working 
with Masakhane, which is one of the African NLP  

00:27:37.560 --> 00:27:43.160
communities which is around creating datasets 
in African languages for training the models.  

00:27:43.960 --> 00:27:49.680
So that was, you know, getting good quality 
training data is already a challenge. Sriram  

00:27:49.680 --> 00:27:54.240
[Rajamani] from MSR India, though, was telling 
me of a really interesting project they've got  

00:27:54.240 --> 00:28:00.360
going on in India with the Indian government where 
they are trying to collect data from each region  

00:28:00.360 --> 00:28:06.280
of India so that they can use it to train the 
OpenAI models, which would be really cool. And  

00:28:06.280 --> 00:28:10.680
we should think about, is that what we can do 
for different African countries and contexts?

00:28:10.680 --> 00:28:13.800
GEHRKE: Exactly. It seems to be very 
much like a citizen science project,  

00:28:13.800 --> 00:28:16.280
right, where you, sort of, involve 
the citizens that speak different  

00:28:16.280 --> 00:28:19.903
dialects and then involve them in 
collecting the right kind of data.

00:28:19.903 --> 00:28:22.200
O’NEILL: Yeah, yeah. And maybe 
collecting the stories, you know,  

00:28:22.200 --> 00:28:26.380
and the cultural attributes and 
assets from different places.

00:28:26.380 --> 00:28:28.120
GEHRKE: That'll be really, really exciting  

00:28:28.120 --> 00:28:31.180
probably also about preservation 
of the culture and history, right.

00:28:31.180 --> 00:28:33.060
O’NEILL: Yes, yes. But challenging.

00:28:33.060 --> 00:28:34.360
GEHRKE: But challenging. [LAUGHTER]

00:28:34.360 --> 00:28:35.020
O’NEILL: Yeah.

00:28:35.020 --> 00:28:39.040
GEHRKE: So that's one big aspect of the 
work. Anything else that's happening there?

00:28:39.040 --> 00:28:42.240
O’NEILL: Yeah. So we're doing a lot of 
work, you'll be unsurprised to hear,  

00:28:42.240 --> 00:28:50.800
on Future of Work and AI. And so we've 
got a project on modern work and LLMs,  

00:28:50.800 --> 00:28:57.360
so looking at the work that enterprise workers, 
frontline and knowledge workers, are doing and  

00:28:57.360 --> 00:29:03.920
then what bits of their job they would like to get 
rid of if they could and what bits they would keep  

00:29:03.920 --> 00:29:09.360
and how we can use LLMs to support them. And 
we've also, like, Maxamed [Axmed] on my team,  

00:29:09.360 --> 00:29:15.760
also worked with The Garage to train them up 
in foundational models, both the LLMs and the  

00:29:15.760 --> 00:29:21.413
vision models, and then they've introduced them 
to a whole load of small businesses in Kenya.

00:29:21.413 --> 00:29:21.429
GEHRKE: Oh, wow.

00:29:21.429 --> 00:29:26.880
O’NEILL: So that's really interesting. You got 
everyone from like car salespeople to lawyers  

00:29:26.880 --> 00:29:31.149
who are now using, like, LLMs as part of 
their everyday work, which is amazing.

00:29:31.149 --> 00:29:33.623
GEHRKE: As part of like composing 
messages or part of ... what's ...

00:29:33.623 --> 00:29:39.893
O’NEILL: Yeah. Writing contracts, sales documents 
for cars, all sorts of really interesting things.

00:29:39.893 --> 00:29:39.909
GEHRKE: Oh, wow.

00:29:39.909 --> 00:29:44.920
O’NEILL: So we're going to go out and look 
at what they're doing and think about how,  

00:29:44.920 --> 00:29:48.100
you know, what else is needed, 
what, what more do they need.

00:29:48.100 --> 00:29:50.960
GEHRKE: What's the prevalent form 
factor in terms of if I think about,  

00:29:50.960 --> 00:29:54.960
like, a computer there? Is it my, is 
it a mobile phone? Is it a tablet?

00:29:54.960 --> 00:29:55.920
O’NEILL: Yeah.

00:29:55.920 --> 00:29:57.075
GEHRKE: It's a mobile phone?

00:29:57.075 --> 00:29:57.112
O’NEILL: It's a mobile phone. Yeah.

00:29:57.112 --> 00:29:59.300
GEHRKE: So you have to rethink 
also, probably, all the interfaces.

00:29:59.300 --> 00:30:00.192
O’NEILL: Yes, I mean ... 

00:30:00.192 --> 00:30:01.160
GEHRKE: You mentioned that early on,  

00:30:01.160 --> 00:30:05.800
right, as you think about the next 
generation of HCI with AI in it, right.

00:30:05.800 --> 00:30:09.760
O’NEILL: Yes, yes. I mean conversational 
interfaces. The idea that you can talk to  

00:30:09.760 --> 00:30:15.240
your phone or enter existing text, you 
know. If you look at small businesses,  

00:30:15.240 --> 00:30:19.720
a lot of their interactions with customers 
are on chat. If you can enter that chat  

00:30:19.720 --> 00:30:26.480
into an LLM and extract structured data from 
it, then suddenly you've got all this data  

00:30:26.480 --> 00:30:31.280
that's been lost to the business becomes 
usable. So it's a really exciting space,  

00:30:31.280 --> 00:30:37.000
and I think voice interfaces are going to become 
really, really, really big. And that's why there's  

00:30:37.000 --> 00:30:44.800
opportunities for leapfrogging, because suddenly 
everyone with a mobile phone potentially has a  

00:30:44.800 --> 00:30:49.480
really powerful office productivity tool in 
their hand and can do things ... you know,  

00:30:49.480 --> 00:30:55.160
many of the small businesses, they don't employ a 
designer; they don't employ an accountant. But now  

00:30:55.160 --> 00:30:59.800
they could maybe have an accountant or a designer 
in their pocket, which enables them to do more,  

00:30:59.800 --> 00:31:04.381
which is definitely the more positive side 
of the future of work than some of the ...

00:31:04.381 --> 00:31:07.440
GEHRKE: Right. You know, this whole enablement 
story of people is just really amazing,  

00:31:07.440 --> 00:31:14.520
what you can do with LLMS and especially with 
voice interfaces, as well. Let me conclude maybe  

00:31:14.520 --> 00:31:19.920
with a question about your career. I mean, it 
seems like you've always amazingly managed to  

00:31:19.920 --> 00:31:24.960
somewhat align your career moves with your 
passion. You moved to India because you're  

00:31:24.960 --> 00:31:29.040
just excited to live in India. You moved then 
to, you know, Microsoft Research, but then you  

00:31:29.040 --> 00:31:33.480
moved to Africa again for, what I hear, is 
a little bit the adventure, as well, right?

00:31:33.480 --> 00:31:33.920
O’NEILL: Yes.

00:31:33.920 --> 00:31:36.200
GEHRKE: So what's your advice 
for people who want to, sort of,  

00:31:36.200 --> 00:31:40.200
align these two and who want to not 
only work but also want to work on  

00:31:40.200 --> 00:31:44.120
something they're really passionate about? 
How do you manage to create that alignment?

00:31:44.120 --> 00:31:49.400
O’NEILL: That is a good question. I don't know. It 
just, sort of, happens. I mean, I think you have  

00:31:49.400 --> 00:31:54.960
to, you have to be passionate about it; you have 
to talk about it and decide what you want to do.  

00:31:54.960 --> 00:32:01.000
You know, I never really imagined MARI would 
happen. But I just started talking to people,  

00:32:01.000 --> 00:32:05.680
and people were saying, before I did the pitch, 
people were saying to me, oh, what would you like  

00:32:05.680 --> 00:32:10.400
to do in five years, Jacki? And I was like, oh, 
you know what? If I had my way, I'd love to run  

00:32:10.400 --> 00:32:16.360
a research center in Africa. And then within a 
couple of years … it was nothing more than an  

00:32:16.360 --> 00:32:21.440
idea in my head. So I think that you have to have 
the ideas, verbalize it, and maybe it can happen.

00:32:21.440 --> 00:32:25.760
GEHRKE: And why a research center in 
Africa? What's personal for you there?

00:32:25.760 --> 00:32:30.480
O’NEILL: So my children are African; my 
children are Cameroonian. So I wanted  

00:32:30.480 --> 00:32:35.880
them to grow, spend some time on the 
continent, and, you know, as a family,  

00:32:35.880 --> 00:32:41.200
we'd always had that idea of moving to the 
continent eventually. So that was part,  

00:32:41.200 --> 00:32:47.600
that was a personal motivation in 
there as well as the passion. Yeah.

00:32:47.600 --> 00:32:51.320
GEHRKE: So it's, well, sort of, the confluence 
of, I guess, opportunity but then also drive  

00:32:51.320 --> 00:32:55.920
on your side? Because that's what I've heard very 
often in careers, that it's not only about, well,  

00:32:55.920 --> 00:32:58.860
this is what I finally want to do but 
also watching out for that opportunity.

00:32:58.860 --> 00:32:59.960
O’NEILL: Yes.

00:32:59.960 --> 00:33:05.600
GEHRKE: So it seems like that played a big role 
here, as well. And so when you heard about,  

00:33:05.600 --> 00:33:09.040
you know, that there was an Africa Development 
Center, how did you, what were your next steps  

00:33:09.040 --> 00:33:12.700
then? I mean, you must have been excited, 
but you also had to take some action.

00:33:12.700 --> 00:33:16.720
O’NEILL: Yeah, I mean, I created, 
[LAUGHS] I created a small pitch,  

00:33:16.720 --> 00:33:21.800
a small set of slides, and then I just 
started talking to everybody I knew  

00:33:21.800 --> 00:33:24.985
who was doing anything. I didn't 
have any contact with the ADCs.

00:33:24.985 --> 00:33:26.942
GEHRKE: So you created that 
energy and excitement about it?

00:33:26.942 --> 00:33:29.760
O’NEILL: I just started to, you know, every 
time anyone would come to India, you know,  

00:33:29.760 --> 00:33:36.180
I was just like, oh, this is what I'd like to do. 
And you just almost talk it into being, I think.

00:33:36.180 --> 00:33:39.200
GEHRKE: And were there some setbacks, or 
was it just like a straight line from,  

00:33:39.200 --> 00:33:40.980
sort of, the excitement all 
the way up to realization?

00:33:40.980 --> 00:33:45.640
O’NEILL: No, I mean, I didn't, I don't think 
I ever really imagined it would happen,  

00:33:45.640 --> 00:33:48.120
you know. But you're just doing 
it, and you're plugging away,  

00:33:48.120 --> 00:33:50.820
and then taking the, you know, 
taking the advice of people.

00:33:50.820 --> 00:33:54.440
GEHRKE: Really an awesome story. So maybe 
as a last question, where do you see the  

00:33:54.440 --> 00:33:58.960
center being in like three to five years? I 
mean, you're starting off right now, but I'm  

00:33:58.960 --> 00:34:03.380
sure you have really big ambitions for the center, 
and there's so much to do on the whole continent.

00:34:03.380 --> 00:34:10.240
O’NEILL: No, absolutely. I think that I have a 
few ambitions. So the most important, I think,  

00:34:10.240 --> 00:34:16.360
I want it to be really established as this 
thing that's really beneficial to Microsoft,  

00:34:16.360 --> 00:34:19.840
that Microsoft is like, really, "Yeah, 
the guys at MARI, they're doing great  

00:34:19.840 --> 00:34:24.280
research. We really like them." So 
that it, sort of, exists without me,  

00:34:24.280 --> 00:34:29.712
you know. At the moment, I think 
I'm the driver of it. I would …

00:34:29.712 --> 00:34:31.800
GEHRKE: So you want to grow the next generation  

00:34:31.800 --> 00:34:34.360
that is basically going to be 
the next generation of leaders?

00:34:34.360 --> 00:34:40.760
O’NEILL: Yes, exactly, exactly. And then 
I think also grow, I would love to help  

00:34:40.760 --> 00:34:44.960
in growing Microsoft's market in Africa. 
We don't have a particularly big market  

00:34:44.960 --> 00:34:48.920
in Africa, but I think there's a lot of 
opportunity, especially now with these,  

00:34:48.920 --> 00:34:53.240
with these large language models. I think that we 
... so that would be really exciting, you know,  

00:34:53.240 --> 00:34:57.960
if we can help. I don't see our success 
only being about growing the African market,  

00:34:57.960 --> 00:35:02.960
but I think it's part of what we can do, and if 
we can grow that market, as well as do research  

00:35:02.960 --> 00:35:08.680
that's relevant for Redmond and relevant globally, 
that's really, that's really exciting, I think,  

00:35:08.680 --> 00:35:14.640
you know. So everything we do, I think, has 
to have a relevance globally. And I think,  

00:35:14.640 --> 00:35:18.080
you know, at the beginning I was talking 
about different ways of viewing the world  

00:35:18.080 --> 00:35:23.600
and how that leads to innovation. I think 
by having researchers who are African,  

00:35:23.600 --> 00:35:30.020
based in Africa, doing this great research, 
we can create better products for everyone.

00:35:30.020 --> 00:35:33.880
GEHRKE: That's such a great finishing note. Thank 
you so much for the great conversation, Jacki.

00:35:33.880 --> 00:35:37.440
O’NEILL: Thank you, Johannes. It's been fun.

00:35:37.440 --> 00:35:39.480
[MUSIC]

00:35:39.480 --> 00:35:42.640
To learn more about Jacki or to 
see photos of Jacki living and  

00:35:42.640 --> 00:35:56.480
working abroad, visit aka.ms/ResearcherStories.

00:35:56.480 --> 00:35:56.980
[MUSIC FADES]