WEBVTT

00:00:00.000 --> 00:00:02.477
Satya, what brings you here to Germany?

00:00:02.560 --> 00:00:07.477
You know, it's our AI tour today in Munich
and I'm excited to be here. Is it

00:00:07.560 --> 00:00:10.317
more difficult these days to come here and

00:00:10.400 --> 00:00:15.397
be in Germany and answer questions? Not
really. I love it. I mean, it feels like

00:00:15.480 --> 00:00:17.037
the atmosphere has changed a little bit.

00:00:17.120 --> 00:00:21.557
There's more skepticism. There's some
governments in Europe even thinking about

00:00:21.640 --> 00:00:25.477
only allowing European tools.
There's even a supermarket giant that's

00:00:25.560 --> 00:00:29.277
developing their own cloud. Well,
I mean, look, you know, first of all, I

00:00:29.360 --> 00:00:34.197
think there is real reason for why
there is competition always. So, there is

00:00:34.280 --> 00:00:37.917
local competition, there is multinational
competition and we welcome

00:00:38.000 --> 00:00:40.717
that and I think what Germany and what

00:00:40.800 --> 00:00:45.757
Europe deserves is great innovation both
coming out of this region and from the

00:00:45.840 --> 00:00:49.877
rest of the world because in fact, when
I look at Europe and look at Germany, in

00:00:49.960 --> 00:00:52.477
fact, I I always describe it, you know,

00:00:52.560 --> 00:00:57.237
whenever I'm in a jeweler in the United
States or when I'm at the dentist, I see

00:00:57.320 --> 00:01:01.237
the German Mittelstand and its excellence.
And so, the world I think

00:01:01.320 --> 00:01:05.357
will remain a connected
world where the innovation

00:01:05.440 --> 00:01:09.037
from any region reaching the other
regions is going to be the true sign of

00:01:09.120 --> 00:01:12.917
progress. But you but you sympathize,
you can understand that some German or

00:01:13.000 --> 00:01:16.517
European companies are a little worried
these days? I mean, I first of all, I

00:01:16.600 --> 00:01:20.197
think that any country should sort
of think about how should they think about

00:01:20.280 --> 00:01:24.157
their own innovation agenda, their own

00:01:24.240 --> 00:01:26.477
competitiveness, their own sovereignty.

00:01:26.560 --> 00:01:28.757
That's the topic of the day today. And I

00:01:28.840 --> 00:01:33.597
talked about this even today in my talk,
which is how one can take all the

00:01:33.680 --> 00:01:37.237
options. So, for example, if you look at
it, right? We are definitely an American

00:01:37.320 --> 00:01:38.957
company, we're a multinational company,

00:01:39.040 --> 00:01:41.797
but we have significant capital that we

00:01:41.880 --> 00:01:44.877
have invested in Europe, in Germany,

00:01:44.960 --> 00:01:49.117
putting our capital at risk so
the German companies can use it and then

00:01:49.200 --> 00:01:52.117
create more value, right? So,
it's not like we're coming here with just

00:01:52.200 --> 00:01:55.197
software. We have both innovation and

00:01:55.280 --> 00:01:59.637
investment here and locally. And we have
all the sovereign options, right? You

00:01:59.720 --> 00:02:04.037
have the ability to manage your own keys,
you can encrypt your data, you can

00:02:04.120 --> 00:02:08.237
have confidential computing, you can
even have private cloud options. And we

00:02:08.320 --> 00:02:12.477
have even local partners. And so, in some
sense, I think as long as one takes

00:02:12.560 --> 00:02:16.037
the approach that, "Look, at the end of
the day, if you want to manage risk, you

00:02:16.120 --> 00:02:19.517
manage it using a portfolio
like everything else in life."

00:02:19.600 --> 00:02:23.357
Um and then to be able to think about
it in clear ways and say, "At the end of

00:02:23.440 --> 00:02:28.037
the day, Germany needs to have
sovereignty and also produce things out

00:02:28.120 --> 00:02:32.077
of Germany that get exported around the
world, and we can be a great partner for

00:02:32.160 --> 00:02:34.517
that." All right. Um maybe briefly on

00:02:34.600 --> 00:02:35.877
your career. It's amazing. I mean, you

00:02:35.960 --> 00:02:39.437
when you started as CEO 2014, um

00:02:39.520 --> 00:02:41.997
the company was 10 times less valuable

00:02:42.080 --> 00:02:44.237
than today, and you I think increased

00:02:44.320 --> 00:02:46.757
the revenue by factor of five or six.

00:02:46.840 --> 00:02:49.757
Amazing numbers. How?

00:02:49.840 --> 00:02:54.437
Look, look, at the end of the day,
um what matters the most is what happens

00:02:54.520 --> 00:02:56.637
next as opposed to what happened in the

00:02:56.720 --> 00:02:59.517
past. And if I sort of had to recount

00:02:59.600 --> 00:03:03.877
how is Microsoft, not just in my
tenure or whatever, 12 years, but

00:03:03.960 --> 00:03:05.637
what's the real question is, what is it

00:03:05.720 --> 00:03:08.757
that after 50 years after our formation,

00:03:08.840 --> 00:03:11.157
here we are in 2026

00:03:11.240 --> 00:03:13.877
uh and going at it. Um and the

00:03:13.960 --> 00:03:16.677
fundamental thing to me is having that

00:03:16.760 --> 00:03:19.477
sense of purpose and mission that allows

00:03:19.560 --> 00:03:22.997
us uh to be relevant going forward and a

00:03:23.080 --> 00:03:26.997
culture that allows us to build
a capability, right? Even take the the

00:03:27.080 --> 00:03:31.917
thing that we're talking about in AI.
It was back in what, 2019, 2018, when we

00:03:32.000 --> 00:03:35.877
decided that there was going
to be essentially a new regime

00:03:35.960 --> 00:03:40.357
uh of AI innovation and see it long
before it's conventional wisdom.

00:03:40.440 --> 00:03:44.997
Is it true that you invested in OpenAI
in 2019, I think, for the first time?

00:03:45.080 --> 00:03:46.997
And Bill Gates was not a big fan of that.

00:03:47.080 --> 00:03:49.477
>> I mean, I think we've
Bill has talked about it,

00:03:49.560 --> 00:03:53.077
I've talked about it uh you know,
externally. And it is I mean, after all

00:03:53.160 --> 00:03:56.877
that, you know, who would have thought
in 2019 that it is a smart idea to go

00:03:56.960 --> 00:03:59.917
put a billion dollars, right? Everybody
thought we were just burning a billion

00:04:00.000 --> 00:04:03.717
dollars. And so, you know, it turns
out some of these things when they work,

00:04:03.800 --> 00:04:07.637
they are celebrated and when they don't
work, nobody knows about it. And so,

00:04:07.720 --> 00:04:09.477
therefore, look, at the end of the day,

00:04:09.560 --> 00:04:12.557
I'm more like the thing that we were

00:04:12.640 --> 00:04:16.677
obsessed about, and Bill in particular
was obsessed about, was natural

00:04:16.760 --> 00:04:20.517
language, right? I mean, throughout
our history from the beginning days of

00:04:20.600 --> 00:04:24.957
Microsoft Research was can we really
have AI breakthroughs in natural

00:04:25.040 --> 00:04:29.437
language? And so, when the OpenAI
team started pivoting from

00:04:29.520 --> 00:04:33.437
what at that time they were doing
a lot more RL and and what have you, it

00:04:33.520 --> 00:04:37.797
interestingly enough they switched
to natural language and scaling laws.

00:04:37.880 --> 00:04:42.717
And that's when we sort of picked it up
and and rest is history. I mean, that's

00:04:42.800 --> 00:04:46.477
the legend that you met Sam Altman
at Sun Valley at the conference.

00:04:46.560 --> 00:04:50.957
What did you see in him after the first
meeting that in the end made you would

00:04:51.040 --> 00:04:53.557
make this investment? You know,
I've I've known Sam for a long lot longer

00:04:53.640 --> 00:04:56.997
than that. I met him, you know, maybe
in you know, when he was in his first

00:04:57.080 --> 00:05:01.077
startup. And so, it's not that I ran
into him in Sun Valley for the first

00:05:01.160 --> 00:05:04.877
time. So, I've known
Sam for a long time. And

00:05:04.960 --> 00:05:08.437
and quite frankly, some of the other
founders of OpenAI and what have you.

00:05:08.520 --> 00:05:11.837
And I think I mean, fundamentally,
interestingly enough, I think the paper

00:05:11.920 --> 00:05:15.677
that was originally written there
was even written by I think Dario

00:05:15.760 --> 00:05:18.317
when he was at OpenAI on scaling laws.

00:05:18.400 --> 00:05:21.717
And so, you know, look, I think it's
fantastic team they assembled. They had

00:05:21.800 --> 00:05:24.157
real belief in scaling.

00:05:24.240 --> 00:05:26.957
They And then we thought that we can be

00:05:27.040 --> 00:05:30.437
great partners. And so, and we went all

00:05:30.520 --> 00:05:33.237
in and it's been great for them and for

00:05:33.320 --> 00:05:37.477
us. Are you sometimes concerned we
are still in an AI bubble, though?

00:05:37.560 --> 00:05:41.637
You know, look, I think it's
at the end of the day, there's only

00:05:41.720 --> 00:05:43.277
thing one thing that matters, right?

00:05:43.360 --> 00:05:47.157
Which is it's not about talking
endlessly about the technology. It's

00:05:47.240 --> 00:05:50.077
about seeing that technology in the real

00:05:50.160 --> 00:05:52.277
world having real impact, right? I come

00:05:52.360 --> 00:05:55.077
here to Munich and I sit and sort of

00:05:55.160 --> 00:05:59.997
listen to a startup out of Munich
that's building one of the world's best

00:06:00.080 --> 00:06:03.677
customer experience. Perloa,
right? Which is fantastic to see

00:06:03.760 --> 00:06:06.397
they're building they're winning share,

00:06:06.480 --> 00:06:10.757
right? That's great to see talking about
German innovation which is not obviously

00:06:10.840 --> 00:06:12.917
born in Germany but winning world over

00:06:13.000 --> 00:06:15.797
and they are partnering with us. Or they

00:06:15.880 --> 00:06:18.757
go the other extreme and say

00:06:18.840 --> 00:06:22.957
you see the Munich fire department using

00:06:23.040 --> 00:06:25.877
AI to get better at their emergency

00:06:25.960 --> 00:06:30.117
transfers. Right? To me the bubble is a

00:06:30.200 --> 00:06:34.637
bubble if you don't have these examples
of diffusion. I give you an example of

00:06:34.720 --> 00:06:38.157
when I thought it might be
a little bit over hyped. Um

00:06:38.240 --> 00:06:41.037
the third most the three most important

00:06:41.120 --> 00:06:43.157
German companies SAP one of them

00:06:43.240 --> 00:06:46.157
Deutsche Telekom and Siemens they're all

00:06:46.240 --> 00:06:49.277
together less valuable than open AI's

00:06:49.360 --> 00:06:53.597
these days. Is that real? Well, I mean
I think at the end of the day there is

00:06:53.680 --> 00:06:56.237
going to be more of a broadening

00:06:56.320 --> 00:06:58.997
of the economy. In fact, if anything the

00:06:59.080 --> 00:07:02.717
in the fullness of time right if what is

00:07:02.800 --> 00:07:06.237
AI going to do? AI is going to do two
things. One is if you think about it

00:07:06.320 --> 00:07:08.157
it's a real physical investment, right?

00:07:08.240 --> 00:07:12.677
After all you brought up Siemens, right?
I think Siemens is going to be one of the

00:07:12.760 --> 00:07:16.357
largest suppliers of a lot of the
equipment that goes into data centers.

00:07:16.440 --> 00:07:19.917
So, I think Siemens is going
to be a major beneficiary

00:07:20.000 --> 00:07:23.917
of what is going to be the build out
of AI not just in Germany but world over,

00:07:24.000 --> 00:07:27.677
right? And so there is going to be
and energy is going to be a massive

00:07:27.760 --> 00:07:31.477
requirement. So, I think there's going
to be a broadening of the supply side of

00:07:31.560 --> 00:07:35.957
it. Then on the demand side, right?
As it gets used it's going to create

00:07:36.040 --> 00:07:38.757
surplus. It's going to create local
surplus whether it's in the health

00:07:38.840 --> 00:07:42.997
outcome or whether that in the public
sector efficiency or what have you. So,

00:07:43.080 --> 00:07:47.277
ultimately there should be and will be
GDP growth, right? That's the real test

00:07:47.360 --> 00:07:50.557
for any of this. >> valuations
are staggering. Valuations

00:07:50.640 --> 00:07:54.117
are look, at the end of the day, you know,
stock markets work because people

00:07:54.200 --> 00:07:57.877
are really looking at what the forward
earnings potential of anything is and

00:07:57.960 --> 00:08:01.117
it's sort of like, you know, in the
short run it is what is it that Warren

00:08:01.200 --> 00:08:05.117
Buffett thing, right? Which is it's a uh
voting machine and in the long run it's

00:08:05.200 --> 00:08:08.877
a weighing machine and I'm not going
to comment on uh stock prices and

00:08:08.960 --> 00:08:14.077
valuations. That's for investors to make
sense of, but what I think of is an all

00:08:14.160 --> 00:08:17.517
up economy benefits from new general

00:08:17.600 --> 00:08:21.957
purpose technologies lifting all boats
when it comes to productivity and this

00:08:22.040 --> 00:08:25.037
is what's happening. Do you think
there's like some new breakthrough

00:08:25.120 --> 00:08:29.037
innovations in AI? Is there something
after the transformer models on the

00:08:29.120 --> 00:08:32.277
frontier already? Yeah, I mean it's it's
a good great question, right? When I I

00:08:32.360 --> 00:08:36.557
think about even it's
not no longer just one

00:08:36.640 --> 00:08:40.037
transformer model. I mean, if you look
at you know, where we've got it is all

00:08:40.120 --> 00:08:43.757
about pre-training scale, then we it
was about post-training, then we came up

00:08:43.840 --> 00:08:48.917
with reasoning, uh then we said, "Oh,
there's RL." Uh and so there's constant

00:08:49.000 --> 00:08:52.517
innovation um that's
happening even in what is

00:08:52.600 --> 00:08:57.077
considered transformers and transformer
architectures. Uh there is obviously

00:08:57.160 --> 00:09:01.237
going to be I always say that we are one
sort of innovation away from the entire

00:09:01.320 --> 00:09:05.037
regime changing even uh
in terms of what's a new model

00:09:05.120 --> 00:09:09.437
architecture that could even be more
efficient uh in its performance. And so,

00:09:09.520 --> 00:09:11.157
I think we have to be very open-minded

00:09:11.240 --> 00:09:14.437
about it. I brought three 5 years out

00:09:14.520 --> 00:09:18.477
questions. The first one is, in 5
years out, will there be vast amounts of

00:09:18.560 --> 00:09:21.357
people without a job? I think

00:09:21.440 --> 00:09:22.957
it's going to be

00:09:23.040 --> 00:09:26.197
one of the biggest questions for us is

00:09:26.280 --> 00:09:30.677
how does work work artifact and workflow

00:09:30.760 --> 00:09:34.517
change? So, let's take software
development, right? Because after all

00:09:34.600 --> 00:09:39.117
this is the place where it all started,
right? In fact, my own belief in this

00:09:39.200 --> 00:09:43.557
particular transformer models and its
scaling properties was, you know,

00:09:43.640 --> 00:09:48.197
reinforced when I saw next edited
suggestions in GitHub Copilot.

00:09:48.280 --> 00:09:52.597
Um if I look at what's happening
with software development, right?

00:09:52.680 --> 00:09:56.277
A- Any If In fact, if you look at even
the history of software development,

00:09:56.360 --> 00:10:00.397
right? You know, at least, you know,
I started learning with assembly and then

00:10:00.480 --> 00:10:03.757
compilers came and then we went
to higher-level languages, we went to

00:10:03.840 --> 00:10:06.437
interpreted languages, and we've had,

00:10:06.520 --> 00:10:10.957
you know, many, many tool chain changes
and levels of abstraction that made us

00:10:11.040 --> 00:10:13.077
more productive. This one is a another

00:10:13.160 --> 00:10:16.317
big leap forward in making it possible

00:10:16.400 --> 00:10:19.957
for both reducing the floor and raising

00:10:20.040 --> 00:10:23.837
the ceiling for anyone who's a software
developer, right? What does it mean

00:10:23.920 --> 00:10:27.757
reducing the floor? Anyone can be
a software developer, just like how Excel

00:10:27.840 --> 00:10:31.837
reduced the floor for anyone to be
an analyst. Now, anyone can be a software

00:10:31.920 --> 00:10:33.917
developer. But, it's also raising the

00:10:34.000 --> 00:10:36.077
ceiling on Oh, what is the new

00:10:36.160 --> 00:10:39.197
sophistication you need in order to be

00:10:39.280 --> 00:10:43.357
productive with these new tools,
so that these code bases that are getting

00:10:43.440 --> 00:10:47.117
generated are not black boxes. That's
going to be what all of us will have to

00:10:47.200 --> 00:10:52.277
reskill ourselves on. So, to me, what
I think is more important in the short

00:10:52.360 --> 00:10:55.437
run. I'm not saying that there is not
going to be displacement. We have to be

00:10:55.520 --> 00:10:57.477
clear-eyed about it. What's the best

00:10:57.560 --> 00:11:00.317
protection against displacement? It's to

00:11:00.400 --> 00:11:03.277
one understand the new medium, the new

00:11:03.360 --> 00:11:08.997
tool, the new skills required,
and transform yourself in job. And that is

00:11:09.080 --> 00:11:12.077
what's happening in software
development, and you can see it where

00:11:12.160 --> 00:11:15.877
software developers are getting more
productive, the IT backlog is huge.

00:11:15.960 --> 00:11:18.997
>> I I I totally buy it
for for the software developers

00:11:19.080 --> 00:11:22.397
bubble or the that industry.
But, I mean, there's talk about like

00:11:22.480 --> 00:11:26.437
whole economies. But, let me say this,
right? If somebody in in the early '80s

00:11:26.520 --> 00:11:30.237
had come to us and said, "3 billion
people, 4 billion people are going to

00:11:30.320 --> 00:11:32.477
get up every morning and type.

00:11:32.560 --> 00:11:34.757
And we would have said, "Why?" Right?

00:11:34.840 --> 00:11:36.677
You know, we have a typist pool and we

00:11:36.760 --> 00:11:39.357
have a slide pool. Why would 4 billion

00:11:39.440 --> 00:11:41.997
people type? We invented this entire

00:11:42.080 --> 00:11:47.077
thing called knowledge work. Uh and so,
I think at some level, I'm not like

00:11:47.160 --> 00:11:48.997
being Pollyannic about this, but I'm

00:11:49.080 --> 00:11:52.317
saying let's at least have a bit of

00:11:52.400 --> 00:11:56.157
optimism in our ability as humans and

00:11:56.240 --> 00:12:00.237
human societies uh and as political
economies, right? One thing that is

00:12:00.320 --> 00:12:05.077
discounted is like we have control.
It's called political economies and

00:12:05.160 --> 00:12:07.157
elections, and people are not going to

00:12:07.240 --> 00:12:09.237
tolerate anything that is not going to

00:12:09.320 --> 00:12:12.797
be beneficial at society scale. And so,

00:12:12.880 --> 00:12:16.877
I'm a big believer that the what is the
real governor? It's just not going to

00:12:16.960 --> 00:12:19.957
happen to us. We're going
to shape what happens to us.

00:12:20.040 --> 00:12:22.637
>> No, it's it's going to be difficult.
I mean, if this if the disruption happens

00:12:22.720 --> 00:12:26.357
fast, I mean, we're up for huge
challenges. I mean, if this unemployment

00:12:26.440 --> 00:12:29.477
really comes, it's going
to be challenging. For sure,

00:12:29.560 --> 00:12:32.317
if unemployment suddenly comes, it's
going to be challenging. And so, that's

00:12:32.400 --> 00:12:34.037
where I would say let us all focus on

00:12:34.120 --> 00:12:36.677
making sure that we are doing all the

00:12:36.760 --> 00:12:39.597
things uh to help us skill ourselves

00:12:39.680 --> 00:12:44.357
such that we have more economic
opportunity with what is going to be uh

00:12:44.440 --> 00:12:47.517
the new productivity curve. >> So,
but you do not afraid of like huge

00:12:47.600 --> 00:12:50.317
waves of unemployment. I I I am afraid

00:12:50.400 --> 00:12:52.757
of anything that is going to not help

00:12:52.840 --> 00:12:57.557
human societies uh really make progress,
right? So, anytime anyone gives me a

00:12:57.640 --> 00:13:00.877
dystopian worldview that we are all
going to be unemployed, of course, I'm

00:13:00.960 --> 00:13:03.717
afraid of it. Uh but I'm also saying I'm

00:13:03.800 --> 00:13:06.437
not so out of control of it. Like we all

00:13:06.520 --> 00:13:09.677
as a you know, as a society get to sort

00:13:09.760 --> 00:13:11.517
of make sure that the technology we

00:13:11.600 --> 00:13:13.957
build and how it diffuses, what happens

00:13:14.040 --> 00:13:17.117
to wages, how we litigate what expertise

00:13:17.200 --> 00:13:19.077
we will reward and not reward. There's

00:13:19.160 --> 00:13:24.157
so many things uh that we are have
control over that we should not abdicate

00:13:24.240 --> 00:13:27.357
all that. Um next five
years question. Um do you

00:13:27.440 --> 00:13:31.477
think Germany will still be a top three
market for Microsoft in the next 5 years?

00:13:31.560 --> 00:13:36.637
I mean, I I definitely hope so. I mean,
to me, you know, what has stood out and

00:13:36.720 --> 00:13:40.477
made Germany unbelievably
successful in the world is

00:13:40.560 --> 00:13:44.397
the German ingenuity, the German
engineering, and its ability to produce

00:13:44.480 --> 00:13:45.837
things that the world needs and the

00:13:45.920 --> 00:13:49.037
world wants. And so, if anything, I

00:13:49.120 --> 00:13:53.677
think that industrial base that Germany
has now coupled with even the digital

00:13:53.760 --> 00:13:55.677
capability. In fact, I always feel,

00:13:55.760 --> 00:13:58.877
again, to my middle start example, why

00:13:58.960 --> 00:14:01.517
not think of, you know, now all this

00:14:01.600 --> 00:14:06.317
digital data, right? And the exhaust
of digital data coming back from

00:14:06.400 --> 00:14:10.237
everywhere, being,
again, translated into the

00:14:10.320 --> 00:14:13.957
intelligence layer that goes with the
Siemens equipment, right? One of the

00:14:14.040 --> 00:14:17.317
things that we're working with Siemens
are all the digital twins and the

00:14:17.400 --> 00:14:19.477
intelligence layer in the digital twins.

00:14:19.560 --> 00:14:21.357
Do you talk to German leaders regularly?

00:14:21.440 --> 00:14:25.397
Absolutely. Like, who do you talk to?
I mean, I talk to Ronald at Siemens, I

00:14:25.480 --> 00:14:29.757
talk to Ola at Mercedes, I talk to Ingo,
I'm going to talk to many play people

00:14:29.840 --> 00:14:33.637
today. And so, yeah,
I'm definitely in contact with

00:14:33.720 --> 00:14:36.637
uh, German CEOs all the time. Do you

00:14:36.720 --> 00:14:38.677
think in the next 5 years, um, Microsoft

00:14:38.760 --> 00:14:44.517
will still rely heavily
on revenue from Office and Windows?

00:14:44.600 --> 00:14:49.037
You know, when we say
rely on Windows Office and

00:14:49.120 --> 00:14:53.277
and Windows, it's fascinating. What
is Office, right? I mean, if I asked you,

00:14:53.360 --> 00:14:58.277
right, the question, Office,
when we started, people bought

00:14:58.360 --> 00:15:02.517
Office clients. Then they
bought servers. And when it

00:15:02.600 --> 00:15:04.677
came to the cloud, they literally kind

00:15:04.760 --> 00:15:07.277
of bought really data,

00:15:07.360 --> 00:15:11.877
uh, and the data storage and the sharing
and collaboration. And so, there is no

00:15:11.960 --> 00:15:16.397
such thing as even what is Office.
It has gone through

00:15:16.480 --> 00:15:18.277
both business model transformation,

00:15:18.360 --> 00:15:20.397
value transformation, right? And so, I

00:15:20.480 --> 00:15:23.677
look at even our, quote-unquote

00:15:23.760 --> 00:15:28.037
in fact in the agent world. Like the
first thing people are doing for agents

00:15:28.120 --> 00:15:32.837
is provision office for it.
Because agents need to collaborate, need

00:15:32.920 --> 00:15:36.157
to be able to collaborate with humans.

00:15:36.240 --> 00:15:40.477
And what's the best way to do it in a
Teams channel to have access to email.

00:15:40.560 --> 00:15:45.317
Like when I want to have a co-worker,
my co-worker that takes my identity needs

00:15:45.400 --> 00:15:47.477
in some sense a separate office

00:15:47.560 --> 00:15:52.677
instance. And so, I think that one way
to think about is if it remains static

00:15:52.760 --> 00:15:56.597
in anything, of course that business
goes away. But when you think about what

00:15:56.680 --> 00:15:59.477
happens next, and this happened
to the cloud, right? We thought we had a

00:15:59.560 --> 00:16:02.917
fantastic server business, and we
thought the cloud will destroy the

00:16:03.000 --> 00:16:04.557
margins. It of course destroyed the

00:16:04.640 --> 00:16:10.197
cloud margins, I mean server margins,
but it expanded the market so massively.

00:16:10.280 --> 00:16:13.357
And I look at what's going to happen
with agents, and I look at that and say,

00:16:13.440 --> 00:16:16.517
"Wow, this is the best news if you
really are in the office business,

00:16:16.600 --> 00:16:18.637
because every agent will need office."

00:16:18.720 --> 00:16:22.357
You always seems like a such a >>
positive sunny guy and what you touch

00:16:22.440 --> 00:16:26.357
turns into gold, but I'm sure you had
to make some huge and and some very tough

00:16:26.440 --> 00:16:30.397
decisions in your life, in your business
career. What was the toughest one?

00:16:30.480 --> 00:16:33.797
Well, I mean at the end of the day,
you know, toughest calls are when you have

00:16:33.880 --> 00:16:37.237
to make redeployment calls that impact
people. I mean at the end, you know,

00:16:37.320 --> 00:16:38.877
that that's I think the hardest stuff.

00:16:38.960 --> 00:16:42.677
But in our case, when it comes
to business strategy, you know, let's, you

00:16:42.760 --> 00:16:45.517
know, call it what it is,
right? When we realized that

00:16:45.600 --> 00:16:49.517
when it came to the phone market,
we were too far behind, and we needed to

00:16:49.600 --> 00:16:53.717
pick where we focused our efforts,
and we picked the cloud. We always thought

00:16:53.800 --> 00:16:58.557
that the cloud as a control plane to go
across devices was more important, and

00:16:58.640 --> 00:17:01.477
that's why I mean even, you know,
making sure that Office was available

00:17:01.560 --> 00:17:05.117
everywhere and what have you. And so,
you know, those were tough decisions,

00:17:05.200 --> 00:17:08.477
but at the end of the day, I I I look
at this and say, "Look, at the end of the

00:17:08.560 --> 00:17:11.957
day for us, making sure we are looking

00:17:12.040 --> 00:17:16.477
at the future, irrespective of,
you know, what happens

00:17:16.560 --> 00:17:18.957
to us, we have to lean into it

00:17:19.040 --> 00:17:24.157
because we can't deny what is going
to be a secular shift. Any tech innovation

00:17:24.240 --> 00:17:26.797
that you'd still like
to see in your lifetime?

00:17:26.880 --> 00:17:28.717
You know, to me,

00:17:28.800 --> 00:17:31.597
my sort of dream would be the coming

00:17:31.680 --> 00:17:34.077
together of these um

00:17:34.160 --> 00:17:35.757
two things that we're working on, right?

00:17:35.840 --> 00:17:39.157
Quantum plus AI. And in fact, the

00:17:39.240 --> 00:17:43.037
juxtaposition of the two. Right?
One of the things we showed even

00:17:43.120 --> 00:17:45.117
at my keynote today was this model we

00:17:45.200 --> 00:17:50.237
built called GigaTime, which
basically, you know, does um

00:17:50.320 --> 00:17:54.637
a simulation uh of spatial
proteomics, right? So,

00:17:54.720 --> 00:17:57.037
it's it's it's a test that is used in

00:17:57.120 --> 00:17:58.957
immunothera- to check whether a

00:17:59.040 --> 00:18:01.877
particular immunotherapy can work for a

00:18:01.960 --> 00:18:04.277
particular cancer tumor. And today, it's

00:18:04.360 --> 00:18:06.757
an expensive test, and

00:18:06.840 --> 00:18:11.637
and so therefore, if we have a model
that really is in silico able to do that

00:18:11.720 --> 00:18:15.517
test, it's then going to be available
in every secondary hospital. Now, think

00:18:15.600 --> 00:18:17.477
about training that model, right?

00:18:17.560 --> 00:18:21.757
Training that model, a quantum computer
can do the best simulation. After all,

00:18:21.840 --> 00:18:25.357
we're trying to learn
the language of nature.

00:18:25.440 --> 00:18:28.317
And nature is quantum. And so therefore,

00:18:28.400 --> 00:18:30.637
a quantum computer will allow us to get

00:18:30.720 --> 00:18:33.237
training data to train AI models that

00:18:33.320 --> 00:18:36.677
then in silico can be much better at

00:18:36.760 --> 00:18:40.437
simulation of the real world, and that
will allow us to get more predictive

00:18:40.520 --> 00:18:42.877
power. Last question, how does a

00:18:42.960 --> 00:18:47.517
Microsoft CEO think about
longevity and how do you work out?

00:18:47.600 --> 00:18:50.917
Well, I get up try to get up every
morning irrespective of the jet lag and

00:18:51.000 --> 00:18:53.757
get to hit the gym. That's >> Every
morning? Every morning. For how

00:18:53.840 --> 00:18:55.917
long? About half hour.
Half hour, and that's it?

00:18:56.000 --> 00:18:59.237
>> That's it. And then you do pull-ups
and push-ups? Well, I mostly nowadays I'm

00:18:59.320 --> 00:19:03.557
like since my knees are no longer
in shape, I'm mostly doing incline walks.

00:19:03.640 --> 00:19:05.477
>> [laughter] >> Okay. Okay.
Thank you very much, Satya.

00:19:05.560 --> 00:19:06.837
Thank you so much. That's a pleasure.

00:19:06.920 --> 00:19:10.120
Let's go. LET'S GO.

00:19:15.640 --> 00:19:18.760
>> LET'S GO. >> [music]