November 13, 2025 • 34 mins
One of the hottest topics in the world is data centres. Demand for the computing power needed to drive modern economies and societies, especially with artificial intelligence, is climbing steeply. Cloud computing services, often provided by big tech firms known as hyperscalers, supply a lot of this power.
In today’s episode, Microsoft’s regional vice president for corporate external and legal affairs in Asia, Mike Yeh, talks about the strategic importance of cloud computing and of turning critical data into national assets. Increasingly, the ability to access and process data effectively is a significant strategic commodity that will help determine the strength of nations.
Mike discusses the meaning of digital “sovereignty”, the use of spreading risk so that a country doesn’t find itself cut off from computing power or from its own data. He talks about the energy demands of computing, the value of digitising and structuring data, the security considerations of cloud computing, Ukraine’s experience with data, and the relationship between strategic competition and rival cloud services in the Indo-Pacific region. He also covers the cybersecurity challenges of cooperation between China, Russia, Iran and North Korea, and levels of social trust in AI in the Indo-Pacific. Once you’re done listening folks, don’t forget to check out the Sydney Dialogue website.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:01):
Stop the world. Welcome to STOP THE WORLD.
I'm David Rowe. And I'm Olivia Nelson.
Today's conversation boils down to one of the hottest topics in
the world right now, which is data centres and the need for
huge amounts of computing power to drive modern economies and
societies. One way to access this computing
(00:22):
power and also to store and manage your data is by buying
cloud services, which are often provided by big tech firms known
as hyperscalers. And one of those hyperscalers is
Microsoft's. To talk about cloud computing
and its strategic importance, wehave Mike Yay, the regional vice
president for Microsoft's corporate, external and legal
affairs in Asia. Microsoft is a sponsor of SP
(00:42):
Sydney Dialogue, which is comingup on December 4 and five.
So it's great to have Mike here in the studio.
Yeah, sure was. The data centres that provide
the computing power for things like artificial intelligence are
becoming key strategic assets for countries around the world.
And in fact, countries and companies fortunes might
increasingly depend on how they manage these issues.
So countries need to think aboutwhat having sovereignty over
(01:02):
their data means, how they spread their risk around so they
don't find themselves cut off from computing power, even
worse, cut off from their own data.
Mike covers all of these issues in a really accessible way.
He talks about the energy demands of computing, the value
of digitising and structuring all your data so you can use it
more effectively, cloud computing and security.
(01:22):
The questions a country like Ukraine faces in where and how
to store its critical data, the concept of digital sovereignty
and the relationship between strategic competition and rival
cloud services in the Indian Pacific.
Yeah, it's really interesting. Mike is a smart and articulate
guy, very thorough, pleasure to deal with.
And once you're done listening, folks, don't forget to check out
the Sydney Dialogue website, whichistsd.aspy.org dot AU.
(01:47):
Tickets are still available. Let's hear from Mike now.
Mike, yay from Microsoft, welcome to Canberra.
Thanks for coming on Stop THE World.
Thank you. It's great to be here.
So I want to start with a bit ofan overview of what you do and,
and your assessment of the region.
Obviously all countries in our part of the world are looking to
digitise, looking to embrace AI.That includes Australia.
(02:08):
We're advanced economy and yet we've got this sort of piddling
productivity growth rate, which we're trying to solve in a
number of ways, including through the the integration and
adoption of AI as one solution. But obviously Southeast Asian
economies and and Pacific economies as well that are still
developing, but looking to grow fast.
So they're all looking to develop policies that encourage
this, but do so in ways that aresafe, secure, trusted, and
(02:32):
enable them essentially to to protect their own critical
national data. Can you give me an overview of
what you're seeing in the regionat the moment in terms of, you
know, trends on these policies in terms of digital
transformation and use of cloud,which you, I know, are very
involved in? How do you see these evolving
over the next, say, five years? That's a great question.
(02:55):
Let me see if I can do this in aminute or two.
There's so many different aspects when I look across Asia
and just what countries are doing.
But I actually think the most interesting thing right now is
to take a pause, right? I think we've sort of been in
this AI moment, I would say, forthe last two years.
I go back to the launch of ChatGPT and I think it's worth
(03:16):
slowing down and thinking about what can we learn from the last
two years that should inform howwe look forward.
One of the things that is top ofmind for me is energy.
And so that that's become, I think, a key priority for so
many countries, especially as people start to grapple with the
data centre investments that aregoing in worldwide, the
(03:36):
trillions of dollars that have been announced for AI data
centres. I do think this is also an
interesting place to make sure, as much as we talked about the
geopolitics between the US and China in particular, there are
things that I think are worth looking at.
What is China doing in this front?
And so they've rolled out a policy called AI Plus.
(03:58):
They did this a few months ago. And it's interesting because
it's really focused on application.
I think in many ways, there's animplicit acknowledgement that
they may not catch up to the US in terms of creating the most
innovative, the biggest LLMS in the near term.
And so they focus right now on figuring out how did they drive
(04:18):
application of AI. And so AI plus crosses a range
of sectors. One of the most interesting ones
for me is what they're doing in the energy space.
And my summary of their plan is essentially energy for AI and AI
for energy, right? And what China has done on the
energy front. They're, they're pretty openly
(04:39):
agnostic about the type of form of energy that will be dominant
in the future. What they see right now is that
the demand for AI, for compute will continue to grow in a way
that outpaces the supply of energy.
And so their whole approach is to figure out, all right, how do
I get as much energy as I can online, build that
infrastructure today, regardlessof where it's from.
(05:03):
And so you see them investing inall types of energy.
And so I do think there that there's some insights and
learnings there that the rest ofthe world would would benefit
from, right? As, as we think through that,
the AI for energy piece is also really interesting.
This goes back to me is like, how do you focus on
applications, right? And So what China's doing is
figuring out how do you apply AIto existing energy
(05:24):
infrastructure to make it more efficient, right?
And that's absolutely something that would be applicable here in
Australia, right? How do you, how do you take the
grid that exists today, but try to modernise it in small ways
without actually completely overhauling or replacing it?
So that's one learning. I, I just think about
application, but think about energy application of AI and
(05:44):
think about energy. The second observation, as I
look back on the last two, 2-3 years, every government talks
about wanting to use AI, wantingto embrace AI.
Every company is talking about that as well.
What's interesting I think is inthe last two years, many
governments and companies have now come to the conclusion that
they may actually not be ready to use AI.
(06:06):
And, and one of the key learnings is in order to fully
leverage AI, you really need structured data.
The data needs to be digitised. I would argue it needs to be in
the cloud and many companies are, are just not at that point.
And so one of the learnings, I think, and one of the calls to
(06:26):
action is to like start digitising now.
And it was great to see. I think it was the AI for public
sector statement that came out today in Australia talking about
the importance of cloud adoption.
But I think as as basic as that sounds, that is the right
foundational step. Right, that's a lot.
There's a lot in there. I definitely want to run with,
(06:47):
with some of it and, and unpack some of it.
Let's talk about the energy thing first.
I mean, that's an interesting, I, I, I, I can completely see
the logic of what you're talkingabout that, you know, let's just
find cheap energy from all sources because we're going to
need much more than we've currently got to power AI, which
is going to solve a lot of the problems that we face in the
(07:07):
future, including potentially, as you say, like, you know,
actually turning it into a cycleand using that AI then to
produce energy efficiency, whichis, which is a lovely idea.
But I mean, first of all, for a country like Australia that's
committed to net zero, what doesthat actually mean for for
someone like us? Look, I, I think there's still
plenty of opportunities even with a net 00 commitment in and
(07:29):
you look at what China's doing, quite frankly, even in the
carbon free green energy space, massive investments in that
space. I think that, you know, their
assumption is that that actuallywill be the cheaper source of
energy going forward. And so that is, I, I think there
there's that still leaves lots of opportunities for Australia
to figure out what that looks like.
Australia, it's a land is a country that has, you know,
(07:52):
significant land resources. You have the space to figure out
wind, solar, potentially thermalin, in other, other forms.
And so I don't think that precludes a strategy or
precludes Australia from being net zero in the future.
OK. You mentioned I think at the
start of your answer before you mentioned just just taking a
(08:12):
breath and and looking over the last two years.
Do you mean that in a policy sense, rather than, I mean
presumably, adoption, integration, we want to do as
quickly as possible within a sort of sensible way?
Just explain that a little bit further.
Yeah, I would. I just think it, it feels like
we should I, and I don't mean itfrom an actual pause in how we
develop and how we use AI. This really is policy stuff, not
(08:35):
even a policy pause. I would say this is like, as
governments develop a strategy, it's just worth taking the time
to look back. It's like, what are you learning
from the first two years? And I think there's a lot,
right? And you see, even from a
regulatory perspective, the initial conversation was around,
is AI going to be an existentialthreat to us?
(08:58):
And then I think everyone used ChatGPT and realised probably
not. And you can see that even that
regulatory conversation has kindof, you know, it peaked and then
it has died down a little bit. I actually think we we run the
risk that we are not paying enough attention to, you know,
what happens when AI becomes an existential threat to us.
(09:21):
And so that that probably is a conversation that should come
back sooner rather than later, right?
Yeah, I mean, I, I, I, I listen very seriously to the
catastrophic risk people becauseI think there is some logic to
it. And while it might be in the in,
in a minority chance rather thana, you know, majority chance,
(09:45):
we're not going to have safe andtrusted use of AI until we
actually resolve those sorts of,you know, big issues to
everyone's satisfaction. So that they are vanishingly
unlikely rather than, you know, sort of somewhere in the, in the
realm of 20 or 30%, which it just seemed like a, a risky
situation. Tell me a little bit about about
(10:07):
this region Asia and that idea of structured data and what they
need to do and where they sort of need to get to in order to
then be more AI ready than they are now and actually just sort
of harness it usefully for theireconomies.
So some of it for me really is basic digitization, right?
How do governments begin on thatjourney?
(10:28):
And in some cases, they'll need to go back because the data is
so important. How do you actually convert
paper data to digital data so that you have full use, you have
longitudinal data available to you?
I think in some cases that we should be pragmatic.
It's like maybe we don't need togo back and completely digitise
everything. But starting today, let's get
(10:49):
this right. Let's think about the right data
structures. And so I think that is, you
know, that's a digitization journey that I, you know, the
world has been on for about 10 years, but this might be the
right catalyst to really force us to look seriously at how do
we move that forward in a way that allows us to take advantage
of AI more quickly in, in in thein the future.
(11:11):
I'll add one other thought on data though, and this is a good
example from Japan. And so one of the things Japan
realised as soon as Check GPT launch was the model really
wasn't optimised for Japanese culture or language, right?
And their response, I think was the right one, which is, all
right, how do we make more Japanese data available for
training, right? And so thinking through that, I
(11:33):
think there's also some intentionality there that is
important for countries across this region, given the breadth
of languages, but not just languages, right?
But even culture. And Australia may sort of face
that in a unique way, right? Where even though English is the
dominant language, culturally, it can be very different from
the way the rest of the Internetis set up and and represented.
(11:55):
Yeah. Yeah.
I mean, South Korea seems to be doing some interesting things
there. They seem to have recognised
that issue and they've they've been working on their own models
as a consequence. So but you're saying that it
needs a sort of top down, top down encouragement.
Is is is that? What you say like I think the
top down helps. Like, if I think about the Japan
example, absolutely you need government to be able to be
(12:16):
intentional and figure out how do you make more Japanese data
available. But I would also say in terms of
just the basic digitization thatcan be done at a company level,
at, you know, department level, even within government.
And so just getting people to really take action to think.
I have that forward look. I, I think about the report that
we just commissioned with the with Mandela partners that
(12:38):
looked at IT spend in governmentand I think they concluded that
85% of IT spend today is spent on legacy IDIT updating legacy
IT. And so I do think there's an
opportunity to remove that dial.You know, it feels like in an
ideal world, you want 15% focus on new IT spend to really get to
(12:58):
like more than 50% right. As you think about new
technology and the pace of innovation, getting from 15 to
50 seems like a a long journey. There's a part of me that says,
let's just try to double it, right?
So how do you get from 15 to 30%in in the next three years?
Tell me a little bit more about this idea of, of structuring and
digitising your data. Can you give me some examples?
(13:20):
Just so I mean, we're AI supposewe're a, we're a general
audience podcast for people who are interested in strategic
policy and national security. But so not, not a, not a
technical audience per SE, but are we just talking about things
like, you know, say a midsize company taking its paper
documents and, and, you know, putting them into digital form
so that that data is then structured that they can use,
(13:42):
you know, and somehow, you know,sort of contribute that data?
I think, look, that's part of it.
In our experience, one of the most useful applications for
Microsoft Copilot within the company is, is just being able
to navigate our company documents, right.
And so in, in your scenario, yeah, just being able to
digitise everything and being able to have AI basically
(14:02):
grounded across all of your company documents.
So people can find HR policies, they can find the financial
policies. There's definitely value there.
But I also think about, I'll usehealthcare as an example, right,
The ability to take some of takeMRI images, right?
Today, many of those are digitised already, but many of
(14:24):
them just sit in separate files.Maybe it's a patient folder, you
know that the clinic has organised.
But figuring out how you anonymize that data, how do you
structure it, bring it together So it's more, it's easier to use
AI to see patterns, right? To train AI so that it looks at
the next MRI image and it can take the first run.
I'm not suggesting a doctor doesn't need to be involved, but
(14:46):
how do you train models by usinganonymized data in a in a way
that we don't do today? That's a great example.
So tell me then about this, you know, the idea of sort of
critical data and data that needs protecting.
Obviously, you know, we're talking about anonymized health
records for very good reason. That's that's personal, that's
that's needs to be protected andsecured.
(15:07):
Tell me a little bit about how countries and how governments
and how companies in this regionare seeing this idea of critical
data that they need to protect, that they need to work together,
including with companies like Microsoft to protect.
And how is that? How is that evolving, that idea
of critical? Data.
Look, I think that the idea of evolution is the right 1 is a
conversation that has again comein different forms.
(15:31):
I think in the initial reaction and and for many governments in
the region, it is still their current reaction, which is they
want data whether it's critical or not.
Many times there's just a desirefor that, for any government
data to be stored in country, right, or to be stored in a
government data centre. I think what the technology over
(15:53):
time has also evolved. So people are increasingly
recognising that actually there are a lot of benefits to using
cloud services. The most high profile example of
some of the benefits actually came in the context of the
Ukraine conflict, where Microsoft partnered with the
Ukrainian government weeks before the attacks began to move
(16:14):
some of the data for the Ukrainian government out of
Ukraine, out of the government data centre.
What's interesting there is whatwe've also learned is that it
turns out data centres can be literally big targets in a
conflict. They're hard to hide.
Russia knew exactly where the Ukrainian government data centre
was and super clever strategy onUkraine's part.
(16:37):
I was like, why not move data from Ukraine into basically a
NATO country? And then that creates all sorts
of externalities that were helpful to Ukraine.
I think this part of the world is also thinking about some of
those lessons now and learning that.
So I think we are slowly there'sa conversation that is I think
with governments that is more, you know, governments are more
(16:58):
open to moving data outside for strategic reasons.
And so I think that's, that's a,that's a helpful conversation to
have. The other idea that has come up
recently is this idea of digitalsovereignty.
Everyone wants complete control.And it's, it's similar to the
location. It's, I think it's a variant of
that, maybe a more advanced version of the location
(17:20):
requirement. And I think what's important for
customers, for governments to recognise is that it's a trade
off, right? If you optimise for complete
sovereignty, you're probably giving up some of the the cost
benefits, Security benefits and maybe innovation benefits that
you would get from hyperscale cloud.
(17:41):
One example of this that happened recently was there's a
data centre fire in Korea. It impacted the Korean
government data centre that powered the government cloud.
One of the services impacted wascalled G Dr Centrally Storage
for Government Employees. That sounds important.
Yes, it, it was, it was the storage for 190,000 employees
(18:08):
and there are over 800 terabytesof data.
And like Korea went down this path because they wanted that
sovereign solution, right, all run in Korea completely
controlled by the government. It turns out that data was not
backed up And, and actually I, it's, it's a little shocking,
but I also, I, I, I use the Apple iPhone analogy.
(18:28):
I think at least for me and I think many other people, it's
like we all buy a little bit of storage, but at some point the
price is like just too much. And I'm like, whatever that
incremental storage is, it's OK,I'll live with that risk.
I think Korea got to that point.They, they just realised that
with 800 terabytes of data, it would just cost so much money to
keep packing it up. And so somebody made a decision
not to back it up. They lost all that data.
(18:52):
And so for me the interesting question is, so did you really
get sovereignty? Because it's really hard to have
sovereignty if you don't have your data.
And so just understanding that trade off in terms of managing
your own, your own IT and and the risk that may come with
that. So, OK, if you're expecting to
be invaded, obviously there's, there's a lot of merit in, in
(19:13):
storing your data outside of your borders to, to mitigate
against, you know, accidents andrisks like, you know, a data
centre Catching Fire. Or if you're not expecting to be
invaded, but you're in within missile strike of a potentially
rival risk country, then that's going to be a problem as well.
(19:33):
And presumably onshore data storage.
I mean, somebody can break in, steal it and then get out again
using cyber enabled techniques. So I can, I can I start to see a
picture emerging here of, you know, why something like secure
cloud is beneficial across the board for different kind of
(19:56):
strategic environments. But just just explain to me in
late terms, just tell me a little bit about the idea of
secure cloud, how that works. And, you know, how does a
country know that they can actually trust it, for instance,
to, to store, you know, criticalnational data well?
Let me start with how we think about the benefits.
So the main driver for a lot of the benefits for hyperscale
(20:19):
cloud today is all but economiesof scale, right?
So for Microsoft, for Amazon, for Google, we are all running
global clouds. And So what that means and
security is a great example. We're able to invest billions of
dollars into security, which is basically more than any
individual customer could have invested on their own, right?
(20:40):
The other advantage is because we have a global footprint,
we're able to see signals from across our whole data set,
right? So you may see a bad actor in
one part of the world, but then we're quickly able to find ways
to mitigate or to address that threat and then scale it out
across our entire network. So even if you're not facing
(21:02):
that particular threat here in Australia, but we detected in
Europe, you benefit from it, right?
Similarly, with some of the regulatory controls like you
think about European GDPR, many of those controls we've built
for our European customers, but also for multinationals that do
business there. And we've also gotten to the
point now where we've just made those controls available.
(21:23):
They've kind of become the de facto set of privacy controls.
So those are all benefits that, you know, come from leveraging
hyperscale cloud. On the technology side, same
thing. I would say even if you look at
GPU's today, the innovation continues very rapidly.
(21:43):
But what we're able to offer youis we can always offer you the
latest AIGPU and we maintain, weupdate that you use what you
need from us. And so that's the advantage
versus say buying that GPU and being stuck with it for 5-10
years, whatever the, you know, the time required is.
Right. OK.
So it's almost like it's like leasing or outsourcing in a way
(22:05):
that means you've got the newestmodel all the time.
Yeah, fascinating. And, and that goes to the point
that, I mean, we, we, we sometimes think about cloud as,
as data related, but in fact, it's computing power related as
well, isn't it? So just talk me through this,
this sort of relationship, I guess between say a company or
country, country's data and it'sthe, the, the compute that it
(22:30):
wants to access in order to be able to drive it's economy, you
know, an efficient economy, a digitised economy.
And, and is there, is there some, I mean, is there, is there
a complicated relationship between, you know, say your data
and your, your compute access that enables you to sort of, you
know, bring all that stuff together and, you know, achieve
what you want to with your economy?
(22:51):
Yeah. I look, I, I think your
instincts are right on that theycan be separate, right, Right.
Where do you put your data? And then like what?
What compute power do you access?
I do think there are some commonthreads though, where there are
benefits to having the two in the same system or platform.
And for me, it does come back down to cyber security, right?
(23:11):
How do you secure the data when it's at rest?
How do you secure it when it's in process, in transit, being
used in the compute mechanisms? And so I do think there are
synergies to bringing those ideas together.
The obvious one is also latency,right?
So thinking about ensuring that for workloads where you need
(23:32):
quick immediate response you youneed the data to be close to the
compute. Yep, Yep.
OK. So what does Asia as a region
sort of need most in this space?I mean, do they need more data
centres? Do they need more developed
policies and legislation? Do they need harmonisation
across the region, you know, especially among economic
(23:52):
partners? What in terms of sort of
priorities, what would you say are the the key things that that
that Asia needs right now? Yeah, that's a great question.
I, I would argue it needs more interoperability, just enabling
cross border data flows across this region.
I'll use one example. You know, as we think about the
geopolitics of the region and, and like I said, you know, with
(24:17):
the Ukraine case, many governments are realising that
they may want to put some data outside of their country.
And, and just to use the, the, the three countries in East
Asia, Japan, South Korea and Taiwan, they all are open to
putting data outside of their country.
They don't have an obvious neighbouring country to put it
in, right. And so that's actually a really
(24:40):
interesting conversation. I, I think that could be an
opportunity for Australia as well, thinking about how do you
build the resiliency, have the right infrastructure in place so
that some of that data could be backed up in Australia.
And so that could, that could bean opportunity as well.
But that's, those are, I think there's a, there's a sense of
trust around the, the data flows.
Southeast Asia is the same way, right?
(25:01):
I look at Singapore and Malaysia.
Malaysia has the data centres. I think, you know, Singapore
would love to put more data intoMalaysia, but figuring out what
are the right trust mechanisms so that customers, whether
they're in Singapore, elsewhere in the region, are comfortable
putting data into Malaysia or into Australia.
(25:21):
And that's, this is where I think you see countries like
Saudi Arabia start talking aboutthe idea of creating a data
embassy, right? How do you give people the
assurance that you know it's safe to put their data here?
Right. So, so in terms of the, the
opportunity for Australia that that's what we could be doing,
what we need, what sort of legislation in place to create
data embassies for instance, that what, what, what, what does
(25:41):
Australia need to do? Yeah.
So I would say that think about how do you create trust, right?
And it could be legislation. I do think in some of these
cases, it could just be a bilateral government to
government agreement. It's like like we're, we're
going to agree to put some data into Australia.
And you know, what I would expect from, for example, the
government of Japan is to be able to say we just need
(26:03):
assurances that from a law enforcement access, you're not
going to touch it, right. So, so it is sort of that
embassy concept, but I think there are different ways to
accomplish that and to provide the the assurance that that each
company needs. In Taiwan's case, I think we
just wouldn't be allowed to callit an embassy.
We'd have to call it a representative office or
whatever. Absolutely.
Data representative office now strategic competition between US
(26:24):
and China right now. Obviously it, it's playing out
in this part of the world in allsorts of ways.
How are you seeing it play out in terms of I suppose the, the,
you know, the offerings that saythe, the, the US hyperscalers
are making to, to this part of the world versus the ones that,
that China via, you know, Huaweior whoever else is offering sort
of cloud solutions to those countries.
(26:46):
How is that sort of playing out?Yeah, we're, I think, I think
it, it, it depends. I would say where we see China
most active in the region is in Southeast Asia.
And that's where we see Alibaba expanding, building new data
centres, Huawei as well very engaged with governments in that
(27:07):
part of the world. And so I do think there will be
an opportunity to for Microsoft to really differentiate, right,
thinking about how do we providetrusted technology solutions,
provide the governance controls that I think many governments in
this part of the world expect and want.
And so I think there will be ways to differentiate but but we
(27:30):
are definitely seeing a resurgence and the Chinese cloud
providers, especially across Southeast Asia, we've seen some
some expansion in Korea as well,which gives you a sense of you
know the markets they're targeting.
Right. Is it difficult for a country to
use, say, Alibaba and Microsoft in a sense alongside each other?
I mean, is there a, is there a sort of a, a 0 sum kind of
(27:52):
competition here? There may be a 0 sum
competition, but the two are notexclusive, right?
So if I especially at a country level, there could be plenty of
workloads in Malaysia that are using Chinese tech and at the
same time plenty of workloads that are using US tech.
So I think there will be room for both.
(28:13):
But you know, and some of it will just depend on what the
workload is too. Obviously, I would hope that the
more sensitive workloads end up on a more trusted platform.
Yeah. OK.
All right. I just want to circle back to
something you mentioned before and that was the, the advantage
of having a large cloud provider, meaning that you are
able to see cyber threats aroundthe world instead of, you know,
(28:35):
and map those and learn from those and then provide
advantages to someone who might not have been hit yet.
Can you give me an, an overview of what you are seeing on the
cyber security side there? I mean, what, what, how, how
new, how diverse are the sorts of threats that your, your
threat assessment teams are picking up?
How much is AI being used to, todrive those new sort of cyber
(28:59):
capabilities? What are you seeing there?
Yeah, so we put out a report every year, that's the Microsoft
Digital Defence Report, and we try to summarise what we're
seeing across the landscape. AI is definitely being used more
frequently by bad actors in the space.
The other dynamic that we are starting to see is there's
definitely a sense that Russia, Iran, North Korea and China are
(29:24):
working together on cyber threats, on cyber attacks.
And, and that is, that is, that's a red flag in in many
ways because they all come to the table with different sets of
expertise and experience, right?And so once they start comparing
notes, they will be more powerful as a group and
collectively. I, I do think that also creates
(29:45):
pressure on democratically minded countries to figure out
how did they work together better too right in, in the
space, something as basic as just figuring out how do we
share cyber threat information. And this is another example.
We've, we've talked with the government in Japan, but also
here in Australia about how do we build systems.
(30:07):
And in some ways, the idea is how do you create a national
cybersecurity dashboard? But again, this kind of goes
back to where we started the conversation.
You have to start digitising things.
You have to get them set up intoa cloud system.
The reality is data is only going to grow exponentially.
So how do you build systems thatcan scale and then build a
system that's interoperable, right?
(30:28):
I don't think we're there yet, but you can quickly envision a
world where Japan and South Korea, Austria, would want to
share cyber threat data with each other in real time.
And so do we have the IT structure set up to enable that?
Absolutely. That's, I mean, that's a great
idea. I, I, I would endorse it fully,
especially it's a lot of debate about how, how closely the sort
(30:50):
of quartet of authoritarian nations is actually cooperating,
how much of it is just a sort ofopportunistic or rhetorical and
how much of it is actually active.
But I mean, if they are actuallyworking together on cyber
threats, you really want democratic countries to to to
respond? I would argue even if if they
aren't working together. Sure, sure.
It it makes sense for us to worktogether.
Absolutely no, that, that, that that makes a lot of sense.
(31:13):
All right, we, we are running out of time. 11 Last one perhaps
to, to, to wrap things up, what do you see in, in societies
around this part of the world interms of, you know, just trust
in, in digitization, trust in AI?
I mean, it seems to be fairly patchy that like sort of rises
(31:33):
and falls according to to different countries and
different populations. But do you think that the the
sort of the trust building in people's readiness to adopt and
embrace AI, including using all of the sort of the, you know,
the sort of technology means that we've talked about?
Is that heading in a positive direction, do you think?
(31:54):
Yeah. I think your word patchy is
probably the right one. I think it depends who you ask.
It depends, you know, I think about whether it's an enterprise
customer, but even consumer, andthen I would break it down by
age, even among consumers. It is interesting.
I, I do think, you know, I, I'veparticipated in a few round
(32:15):
tables recently with young people talking about AI, how
they think about it from a career perspective, education
context. And that's where I really sense
that it is patchy. You know, the, the antidotes
I've heard was at some schools literally just in the last year
have changed policies from like,don't use AI to look, you have
to use AI. And so I think students feel
(32:36):
that whiplash. The, the other thing is, I, I
actually think we need to think through and Australia is a great
place to have this conversation as, as, as it kind of leads the
world on the social media ban. But I think social media is a
great example where I think we, we see a lot of bad AI being
used. You know, I think there's
(32:56):
increasingly a sense that you think about videos today on
social media, no one really knows what's AI generated
anymore. And so figuring out how do we
create trust in that product where people don't think of it
as just AI slop, but actually it's AI that adds value.
It's AI that I want to see and Iwant to, I want to kind of
consume. And so I, I think it's, it's a
(33:19):
mixed bag right now, but I thinkthat also means that there's
work that we need to do as a society, right?
Yeah, absolutely. I mean, I presume the young,
broadly speaking, that the younger a person is, the more
they're likely to embrace and trust AI.
But perhaps that isn't perfect. But I certainly hear what you're
saying about the whiplash that alot of young people might be
(33:39):
feeling. You know, last year, it might
have been cheating. This year it is, you know,
learning to use it as a tool effectively.
And so, yeah, fascinating. Yeah.
So in, in the workplace, I look,we, we fully embraced it, but I,
I also think everyone now is figuring out like, what's the
next wave? What do we really see the
(34:00):
application beyond generative AIbetween, you know, ChatGPT or
Microsoft Co pilot. And so I think that'll be an
interesting time. But I also think that's that's
where opportunity is created, not just for Australia but for
every country in Asia. Yep, Yep.
All right. Terrific.
All right, Mike. Yay.
Thanks so much for coming on. Stop THE World.
Appreciate it. Thank you.
(34:20):
That's all we have time for today, folks.
Thanks for listening to Stop theWorld.
We'll be back with another episode soon.
Stop the world. Welcome to STOP THE WORLD.
I'm David Rowe. And I'm Olivia Nelson.
Today's conversation boils down to one of the hottest topics in
the world right now, which is data centres and the need for
huge amounts of computing power to drive modern economies and
societies. One way to access this computing
(00:22):
power and also to store and manage your data is by buying
cloud services, which are often provided by big tech firms known
as hyperscalers. And one of those hyperscalers is
Microsoft's. To talk about cloud computing
and its strategic importance, wehave Mike Yay, the regional vice
president for Microsoft's corporate, external and legal
affairs in Asia. Microsoft is a sponsor of SP
(00:42):
Sydney Dialogue, which is comingup on December 4 and five.
So it's great to have Mike here in the studio.
Yeah, sure was. The data centres that provide
the computing power for things like artificial intelligence are
becoming key strategic assets for countries around the world.
And in fact, countries and companies fortunes might
increasingly depend on how they manage these issues.
So countries need to think aboutwhat having sovereignty over
(01:02):
their data means, how they spread their risk around so they
don't find themselves cut off from computing power, even
worse, cut off from their own data.
Mike covers all of these issues in a really accessible way.
He talks about the energy demands of computing, the value
of digitising and structuring all your data so you can use it
more effectively, cloud computing and security.
(01:22):
The questions a country like Ukraine faces in where and how
to store its critical data, the concept of digital sovereignty
and the relationship between strategic competition and rival
cloud services in the Indian Pacific.
Yeah, it's really interesting. Mike is a smart and articulate
guy, very thorough, pleasure to deal with.
And once you're done listening, folks, don't forget to check out
the Sydney Dialogue website, whichistsd.aspy.org dot AU.
(01:47):
Tickets are still available. Let's hear from Mike now.
Mike, yay from Microsoft, welcome to Canberra.
Thanks for coming on Stop THE World.
Thank you. It's great to be here.
So I want to start with a bit ofan overview of what you do and,
and your assessment of the region.
Obviously all countries in our part of the world are looking to
digitise, looking to embrace AI.That includes Australia.
(02:08):
We're advanced economy and yet we've got this sort of piddling
productivity growth rate, which we're trying to solve in a
number of ways, including through the the integration and
adoption of AI as one solution. But obviously Southeast Asian
economies and and Pacific economies as well that are still
developing, but looking to grow fast.
So they're all looking to develop policies that encourage
this, but do so in ways that aresafe, secure, trusted, and
(02:32):
enable them essentially to to protect their own critical
national data. Can you give me an overview of
what you're seeing in the regionat the moment in terms of, you
know, trends on these policies in terms of digital
transformation and use of cloud,which you, I know, are very
involved in? How do you see these evolving
over the next, say, five years? That's a great question.
(02:55):
Let me see if I can do this in aminute or two.
There's so many different aspects when I look across Asia
and just what countries are doing.
But I actually think the most interesting thing right now is
to take a pause, right? I think we've sort of been in
this AI moment, I would say, forthe last two years.
I go back to the launch of ChatGPT and I think it's worth
(03:16):
slowing down and thinking about what can we learn from the last
two years that should inform howwe look forward.
One of the things that is top ofmind for me is energy.
And so that that's become, I think, a key priority for so
many countries, especially as people start to grapple with the
data centre investments that aregoing in worldwide, the
(03:36):
trillions of dollars that have been announced for AI data
centres. I do think this is also an
interesting place to make sure, as much as we talked about the
geopolitics between the US and China in particular, there are
things that I think are worth looking at.
What is China doing in this front?
And so they've rolled out a policy called AI Plus.
(03:58):
They did this a few months ago. And it's interesting because
it's really focused on application.
I think in many ways, there's animplicit acknowledgement that
they may not catch up to the US in terms of creating the most
innovative, the biggest LLMS in the near term.
And so they focus right now on figuring out how did they drive
(04:18):
application of AI. And so AI plus crosses a range
of sectors. One of the most interesting ones
for me is what they're doing in the energy space.
And my summary of their plan is essentially energy for AI and AI
for energy, right? And what China has done on the
energy front. They're, they're pretty openly
(04:39):
agnostic about the type of form of energy that will be dominant
in the future. What they see right now is that
the demand for AI, for compute will continue to grow in a way
that outpaces the supply of energy.
And so their whole approach is to figure out, all right, how do
I get as much energy as I can online, build that
infrastructure today, regardlessof where it's from.
(05:03):
And so you see them investing inall types of energy.
And so I do think there that there's some insights and
learnings there that the rest ofthe world would would benefit
from, right? As, as we think through that,
the AI for energy piece is also really interesting.
This goes back to me is like, how do you focus on
applications, right? And So what China's doing is
figuring out how do you apply AIto existing energy
(05:24):
infrastructure to make it more efficient, right?
And that's absolutely something that would be applicable here in
Australia, right? How do you, how do you take the
grid that exists today, but try to modernise it in small ways
without actually completely overhauling or replacing it?
So that's one learning. I, I just think about
application, but think about energy application of AI and
(05:44):
think about energy. The second observation, as I
look back on the last two, 2-3 years, every government talks
about wanting to use AI, wantingto embrace AI.
Every company is talking about that as well.
What's interesting I think is inthe last two years, many
governments and companies have now come to the conclusion that
they may actually not be ready to use AI.
(06:06):
And, and one of the key learnings is in order to fully
leverage AI, you really need structured data.
The data needs to be digitised. I would argue it needs to be in
the cloud and many companies are, are just not at that point.
And so one of the learnings, I think, and one of the calls to
(06:26):
action is to like start digitising now.
And it was great to see. I think it was the AI for public
sector statement that came out today in Australia talking about
the importance of cloud adoption.
But I think as as basic as that sounds, that is the right
foundational step. Right, that's a lot.
There's a lot in there. I definitely want to run with,
(06:47):
with some of it and, and unpack some of it.
Let's talk about the energy thing first.
I mean, that's an interesting, I, I, I, I can completely see
the logic of what you're talkingabout that, you know, let's just
find cheap energy from all sources because we're going to
need much more than we've currently got to power AI, which
is going to solve a lot of the problems that we face in the
(07:07):
future, including potentially, as you say, like, you know,
actually turning it into a cycleand using that AI then to
produce energy efficiency, whichis, which is a lovely idea.
But I mean, first of all, for a country like Australia that's
committed to net zero, what doesthat actually mean for for
someone like us? Look, I, I think there's still
plenty of opportunities even with a net 00 commitment in and
(07:29):
you look at what China's doing, quite frankly, even in the
carbon free green energy space, massive investments in that
space. I think that, you know, their
assumption is that that actuallywill be the cheaper source of
energy going forward. And so that is, I, I think there
there's that still leaves lots of opportunities for Australia
to figure out what that looks like.
Australia, it's a land is a country that has, you know,
(07:52):
significant land resources. You have the space to figure out
wind, solar, potentially thermalin, in other, other forms.
And so I don't think that precludes a strategy or
precludes Australia from being net zero in the future.
OK. You mentioned I think at the
start of your answer before you mentioned just just taking a
(08:12):
breath and and looking over the last two years.
Do you mean that in a policy sense, rather than, I mean
presumably, adoption, integration, we want to do as
quickly as possible within a sort of sensible way?
Just explain that a little bit further.
Yeah, I would. I just think it, it feels like
we should I, and I don't mean itfrom an actual pause in how we
develop and how we use AI. This really is policy stuff, not
(08:35):
even a policy pause. I would say this is like, as
governments develop a strategy, it's just worth taking the time
to look back. It's like, what are you learning
from the first two years? And I think there's a lot,
right? And you see, even from a
regulatory perspective, the initial conversation was around,
is AI going to be an existentialthreat to us?
(08:58):
And then I think everyone used ChatGPT and realised probably
not. And you can see that even that
regulatory conversation has kindof, you know, it peaked and then
it has died down a little bit. I actually think we we run the
risk that we are not paying enough attention to, you know,
what happens when AI becomes an existential threat to us.
(09:21):
And so that that probably is a conversation that should come
back sooner rather than later, right?
Yeah, I mean, I, I, I, I listen very seriously to the
catastrophic risk people becauseI think there is some logic to
it. And while it might be in the in,
in a minority chance rather thana, you know, majority chance,
(09:45):
we're not going to have safe andtrusted use of AI until we
actually resolve those sorts of,you know, big issues to
everyone's satisfaction. So that they are vanishingly
unlikely rather than, you know, sort of somewhere in the, in the
realm of 20 or 30%, which it just seemed like a, a risky
situation. Tell me a little bit about about
(10:07):
this region Asia and that idea of structured data and what they
need to do and where they sort of need to get to in order to
then be more AI ready than they are now and actually just sort
of harness it usefully for theireconomies.
So some of it for me really is basic digitization, right?
How do governments begin on thatjourney?
(10:28):
And in some cases, they'll need to go back because the data is
so important. How do you actually convert
paper data to digital data so that you have full use, you have
longitudinal data available to you?
I think in some cases that we should be pragmatic.
It's like maybe we don't need togo back and completely digitise
everything. But starting today, let's get
(10:49):
this right. Let's think about the right data
structures. And so I think that is, you
know, that's a digitization journey that I, you know, the
world has been on for about 10 years, but this might be the
right catalyst to really force us to look seriously at how do
we move that forward in a way that allows us to take advantage
of AI more quickly in, in in thein the future.
(11:11):
I'll add one other thought on data though, and this is a good
example from Japan. And so one of the things Japan
realised as soon as Check GPT launch was the model really
wasn't optimised for Japanese culture or language, right?
And their response, I think was the right one, which is, all
right, how do we make more Japanese data available for
training, right? And so thinking through that, I
(11:33):
think there's also some intentionality there that is
important for countries across this region, given the breadth
of languages, but not just languages, right?
But even culture. And Australia may sort of face
that in a unique way, right? Where even though English is the
dominant language, culturally, it can be very different from
the way the rest of the Internetis set up and and represented.
(11:55):
Yeah. Yeah.
I mean, South Korea seems to be doing some interesting things
there. They seem to have recognised
that issue and they've they've been working on their own models
as a consequence. So but you're saying that it
needs a sort of top down, top down encouragement.
Is is is that? What you say like I think the
top down helps. Like, if I think about the Japan
example, absolutely you need government to be able to be
(12:16):
intentional and figure out how do you make more Japanese data
available. But I would also say in terms of
just the basic digitization thatcan be done at a company level,
at, you know, department level, even within government.
And so just getting people to really take action to think.
I have that forward look. I, I think about the report that
we just commissioned with the with Mandela partners that
(12:38):
looked at IT spend in governmentand I think they concluded that
85% of IT spend today is spent on legacy IDIT updating legacy
IT. And so I do think there's an
opportunity to remove that dial.You know, it feels like in an
ideal world, you want 15% focus on new IT spend to really get to
(12:58):
like more than 50% right. As you think about new
technology and the pace of innovation, getting from 15 to
50 seems like a a long journey. There's a part of me that says,
let's just try to double it, right?
So how do you get from 15 to 30%in in the next three years?
Tell me a little bit more about this idea of, of structuring and
digitising your data. Can you give me some examples?
(13:20):
Just so I mean, we're AI supposewe're a, we're a general
audience podcast for people who are interested in strategic
policy and national security. But so not, not a, not a
technical audience per SE, but are we just talking about things
like, you know, say a midsize company taking its paper
documents and, and, you know, putting them into digital form
so that that data is then structured that they can use,
(13:42):
you know, and somehow, you know,sort of contribute that data?
I think, look, that's part of it.
In our experience, one of the most useful applications for
Microsoft Copilot within the company is, is just being able
to navigate our company documents, right.
And so in, in your scenario, yeah, just being able to
digitise everything and being able to have AI basically
(14:02):
grounded across all of your company documents.
So people can find HR policies, they can find the financial
policies. There's definitely value there.
But I also think about, I'll usehealthcare as an example, right,
The ability to take some of takeMRI images, right?
Today, many of those are digitised already, but many of
(14:24):
them just sit in separate files.Maybe it's a patient folder, you
know that the clinic has organised.
But figuring out how you anonymize that data, how do you
structure it, bring it together So it's more, it's easier to use
AI to see patterns, right? To train AI so that it looks at
the next MRI image and it can take the first run.
I'm not suggesting a doctor doesn't need to be involved, but
(14:46):
how do you train models by usinganonymized data in a in a way
that we don't do today? That's a great example.
So tell me then about this, you know, the idea of sort of
critical data and data that needs protecting.
Obviously, you know, we're talking about anonymized health
records for very good reason. That's that's personal, that's
that's needs to be protected andsecured.
(15:07):
Tell me a little bit about how countries and how governments
and how companies in this regionare seeing this idea of critical
data that they need to protect, that they need to work together,
including with companies like Microsoft to protect.
And how is that? How is that evolving, that idea
of critical? Data.
Look, I think that the idea of evolution is the right 1 is a
conversation that has again comein different forms.
(15:31):
I think in the initial reaction and and for many governments in
the region, it is still their current reaction, which is they
want data whether it's critical or not.
Many times there's just a desirefor that, for any government
data to be stored in country, right, or to be stored in a
government data centre. I think what the technology over
(15:53):
time has also evolved. So people are increasingly
recognising that actually there are a lot of benefits to using
cloud services. The most high profile example of
some of the benefits actually came in the context of the
Ukraine conflict, where Microsoft partnered with the
Ukrainian government weeks before the attacks began to move
(16:14):
some of the data for the Ukrainian government out of
Ukraine, out of the government data centre.
What's interesting there is whatwe've also learned is that it
turns out data centres can be literally big targets in a
conflict. They're hard to hide.
Russia knew exactly where the Ukrainian government data centre
was and super clever strategy onUkraine's part.
(16:37):
I was like, why not move data from Ukraine into basically a
NATO country? And then that creates all sorts
of externalities that were helpful to Ukraine.
I think this part of the world is also thinking about some of
those lessons now and learning that.
So I think we are slowly there'sa conversation that is I think
with governments that is more, you know, governments are more
(16:58):
open to moving data outside for strategic reasons.
And so I think that's, that's a,that's a helpful conversation to
have. The other idea that has come up
recently is this idea of digitalsovereignty.
Everyone wants complete control.And it's, it's similar to the
location. It's, I think it's a variant of
that, maybe a more advanced version of the location
(17:20):
requirement. And I think what's important for
customers, for governments to recognise is that it's a trade
off, right? If you optimise for complete
sovereignty, you're probably giving up some of the the cost
benefits, Security benefits and maybe innovation benefits that
you would get from hyperscale cloud.
(17:41):
One example of this that happened recently was there's a
data centre fire in Korea. It impacted the Korean
government data centre that powered the government cloud.
One of the services impacted wascalled G Dr Centrally Storage
for Government Employees. That sounds important.
Yes, it, it was, it was the storage for 190,000 employees
(18:08):
and there are over 800 terabytesof data.
And like Korea went down this path because they wanted that
sovereign solution, right, all run in Korea completely
controlled by the government. It turns out that data was not
backed up And, and actually I, it's, it's a little shocking,
but I also, I, I, I use the Apple iPhone analogy.
(18:28):
I think at least for me and I think many other people, it's
like we all buy a little bit of storage, but at some point the
price is like just too much. And I'm like, whatever that
incremental storage is, it's OK,I'll live with that risk.
I think Korea got to that point.They, they just realised that
with 800 terabytes of data, it would just cost so much money to
keep packing it up. And so somebody made a decision
not to back it up. They lost all that data.
(18:52):
And so for me the interesting question is, so did you really
get sovereignty? Because it's really hard to have
sovereignty if you don't have your data.
And so just understanding that trade off in terms of managing
your own, your own IT and and the risk that may come with
that. So, OK, if you're expecting to
be invaded, obviously there's, there's a lot of merit in, in
(19:13):
storing your data outside of your borders to, to mitigate
against, you know, accidents andrisks like, you know, a data
centre Catching Fire. Or if you're not expecting to be
invaded, but you're in within missile strike of a potentially
rival risk country, then that's going to be a problem as well.
(19:33):
And presumably onshore data storage.
I mean, somebody can break in, steal it and then get out again
using cyber enabled techniques. So I can, I can I start to see a
picture emerging here of, you know, why something like secure
cloud is beneficial across the board for different kind of
(19:56):
strategic environments. But just just explain to me in
late terms, just tell me a little bit about the idea of
secure cloud, how that works. And, you know, how does a
country know that they can actually trust it, for instance,
to, to store, you know, criticalnational data well?
Let me start with how we think about the benefits.
So the main driver for a lot of the benefits for hyperscale
(20:19):
cloud today is all but economiesof scale, right?
So for Microsoft, for Amazon, for Google, we are all running
global clouds. And So what that means and
security is a great example. We're able to invest billions of
dollars into security, which is basically more than any
individual customer could have invested on their own, right?
(20:40):
The other advantage is because we have a global footprint,
we're able to see signals from across our whole data set,
right? So you may see a bad actor in
one part of the world, but then we're quickly able to find ways
to mitigate or to address that threat and then scale it out
across our entire network. So even if you're not facing
(21:02):
that particular threat here in Australia, but we detected in
Europe, you benefit from it, right?
Similarly, with some of the regulatory controls like you
think about European GDPR, many of those controls we've built
for our European customers, but also for multinationals that do
business there. And we've also gotten to the
point now where we've just made those controls available.
(21:23):
They've kind of become the de facto set of privacy controls.
So those are all benefits that, you know, come from leveraging
hyperscale cloud. On the technology side, same
thing. I would say even if you look at
GPU's today, the innovation continues very rapidly.
(21:43):
But what we're able to offer youis we can always offer you the
latest AIGPU and we maintain, weupdate that you use what you
need from us. And so that's the advantage
versus say buying that GPU and being stuck with it for 5-10
years, whatever the, you know, the time required is.
Right. OK.
So it's almost like it's like leasing or outsourcing in a way
(22:05):
that means you've got the newestmodel all the time.
Yeah, fascinating. And, and that goes to the point
that, I mean, we, we, we sometimes think about cloud as,
as data related, but in fact, it's computing power related as
well, isn't it? So just talk me through this,
this sort of relationship, I guess between say a company or
country, country's data and it'sthe, the, the compute that it
(22:30):
wants to access in order to be able to drive it's economy, you
know, an efficient economy, a digitised economy.
And, and is there, is there some, I mean, is there, is there
a complicated relationship between, you know, say your data
and your, your compute access that enables you to sort of, you
know, bring all that stuff together and, you know, achieve
what you want to with your economy?
(22:51):
Yeah. I look, I, I think your
instincts are right on that theycan be separate, right, Right.
Where do you put your data? And then like what?
What compute power do you access?
I do think there are some commonthreads though, where there are
benefits to having the two in the same system or platform.
And for me, it does come back down to cyber security, right?
(23:11):
How do you secure the data when it's at rest?
How do you secure it when it's in process, in transit, being
used in the compute mechanisms? And so I do think there are
synergies to bringing those ideas together.
The obvious one is also latency,right?
So thinking about ensuring that for workloads where you need
(23:32):
quick immediate response you youneed the data to be close to the
compute. Yep, Yep.
OK. So what does Asia as a region
sort of need most in this space?I mean, do they need more data
centres? Do they need more developed
policies and legislation? Do they need harmonisation
across the region, you know, especially among economic
(23:52):
partners? What in terms of sort of
priorities, what would you say are the the key things that that
that Asia needs right now? Yeah, that's a great question.
I, I would argue it needs more interoperability, just enabling
cross border data flows across this region.
I'll use one example. You know, as we think about the
geopolitics of the region and, and like I said, you know, with
(24:17):
the Ukraine case, many governments are realising that
they may want to put some data outside of their country.
And, and just to use the, the, the three countries in East
Asia, Japan, South Korea and Taiwan, they all are open to
putting data outside of their country.
They don't have an obvious neighbouring country to put it
in, right. And so that's actually a really
(24:40):
interesting conversation. I, I think that could be an
opportunity for Australia as well, thinking about how do you
build the resiliency, have the right infrastructure in place so
that some of that data could be backed up in Australia.
And so that could, that could bean opportunity as well.
But that's, those are, I think there's a, there's a sense of
trust around the, the data flows.
Southeast Asia is the same way, right?
(25:01):
I look at Singapore and Malaysia.
Malaysia has the data centres. I think, you know, Singapore
would love to put more data intoMalaysia, but figuring out what
are the right trust mechanisms so that customers, whether
they're in Singapore, elsewhere in the region, are comfortable
putting data into Malaysia or into Australia.
(25:21):
And that's, this is where I think you see countries like
Saudi Arabia start talking aboutthe idea of creating a data
embassy, right? How do you give people the
assurance that you know it's safe to put their data here?
Right. So, so in terms of the, the
opportunity for Australia that that's what we could be doing,
what we need, what sort of legislation in place to create
data embassies for instance, that what, what, what, what does
(25:41):
Australia need to do? Yeah.
So I would say that think about how do you create trust, right?
And it could be legislation. I do think in some of these
cases, it could just be a bilateral government to
government agreement. It's like like we're, we're
going to agree to put some data into Australia.
And you know, what I would expect from, for example, the
government of Japan is to be able to say we just need
(26:03):
assurances that from a law enforcement access, you're not
going to touch it, right. So, so it is sort of that
embassy concept, but I think there are different ways to
accomplish that and to provide the the assurance that that each
company needs. In Taiwan's case, I think we
just wouldn't be allowed to callit an embassy.
We'd have to call it a representative office or
whatever. Absolutely.
Data representative office now strategic competition between US
(26:24):
and China right now. Obviously it, it's playing out
in this part of the world in allsorts of ways.
How are you seeing it play out in terms of I suppose the, the,
you know, the offerings that saythe, the, the US hyperscalers
are making to, to this part of the world versus the ones that,
that China via, you know, Huaweior whoever else is offering sort
of cloud solutions to those countries.
(26:46):
How is that sort of playing out?Yeah, we're, I think, I think
it, it, it depends. I would say where we see China
most active in the region is in Southeast Asia.
And that's where we see Alibaba expanding, building new data
centres, Huawei as well very engaged with governments in that
(27:07):
part of the world. And so I do think there will be
an opportunity to for Microsoft to really differentiate, right,
thinking about how do we providetrusted technology solutions,
provide the governance controls that I think many governments in
this part of the world expect and want.
And so I think there will be ways to differentiate but but we
(27:30):
are definitely seeing a resurgence and the Chinese cloud
providers, especially across Southeast Asia, we've seen some
some expansion in Korea as well,which gives you a sense of you
know the markets they're targeting.
Right. Is it difficult for a country to
use, say, Alibaba and Microsoft in a sense alongside each other?
I mean, is there a, is there a sort of a, a 0 sum kind of
(27:52):
competition here? There may be a 0 sum
competition, but the two are notexclusive, right?
So if I especially at a country level, there could be plenty of
workloads in Malaysia that are using Chinese tech and at the
same time plenty of workloads that are using US tech.
So I think there will be room for both.
(28:13):
But you know, and some of it will just depend on what the
workload is too. Obviously, I would hope that the
more sensitive workloads end up on a more trusted platform.
Yeah. OK.
All right. I just want to circle back to
something you mentioned before and that was the, the advantage
of having a large cloud provider, meaning that you are
able to see cyber threats aroundthe world instead of, you know,
(28:35):
and map those and learn from those and then provide
advantages to someone who might not have been hit yet.
Can you give me an, an overview of what you are seeing on the
cyber security side there? I mean, what, what, how, how
new, how diverse are the sorts of threats that your, your
threat assessment teams are picking up?
How much is AI being used to, todrive those new sort of cyber
(28:59):
capabilities? What are you seeing there?
Yeah, so we put out a report every year, that's the Microsoft
Digital Defence Report, and we try to summarise what we're
seeing across the landscape. AI is definitely being used more
frequently by bad actors in the space.
The other dynamic that we are starting to see is there's
definitely a sense that Russia, Iran, North Korea and China are
(29:24):
working together on cyber threats, on cyber attacks.
And, and that is, that is, that's a red flag in in many
ways because they all come to the table with different sets of
expertise and experience, right?And so once they start comparing
notes, they will be more powerful as a group and
collectively. I, I do think that also creates
(29:45):
pressure on democratically minded countries to figure out
how did they work together better too right in, in the
space, something as basic as just figuring out how do we
share cyber threat information. And this is another example.
We've, we've talked with the government in Japan, but also
here in Australia about how do we build systems.
(30:07):
And in some ways, the idea is how do you create a national
cybersecurity dashboard? But again, this kind of goes
back to where we started the conversation.
You have to start digitising things.
You have to get them set up intoa cloud system.
The reality is data is only going to grow exponentially.
So how do you build systems thatcan scale and then build a
system that's interoperable, right?
(30:28):
I don't think we're there yet, but you can quickly envision a
world where Japan and South Korea, Austria, would want to
share cyber threat data with each other in real time.
And so do we have the IT structure set up to enable that?
Absolutely. That's, I mean, that's a great
idea. I, I, I would endorse it fully,
especially it's a lot of debate about how, how closely the sort
(30:50):
of quartet of authoritarian nations is actually cooperating,
how much of it is just a sort ofopportunistic or rhetorical and
how much of it is actually active.
But I mean, if they are actuallyworking together on cyber
threats, you really want democratic countries to to to
respond? I would argue even if if they
aren't working together. Sure, sure.
It it makes sense for us to worktogether.
Absolutely no, that, that, that that makes a lot of sense.
(31:13):
All right, we, we are running out of time. 11 Last one perhaps
to, to, to wrap things up, what do you see in, in societies
around this part of the world interms of, you know, just trust
in, in digitization, trust in AI?
I mean, it seems to be fairly patchy that like sort of rises
(31:33):
and falls according to to different countries and
different populations. But do you think that the the
sort of the trust building in people's readiness to adopt and
embrace AI, including using all of the sort of the, you know,
the sort of technology means that we've talked about?
Is that heading in a positive direction, do you think?
(31:54):
Yeah. I think your word patchy is
probably the right one. I think it depends who you ask.
It depends, you know, I think about whether it's an enterprise
customer, but even consumer, andthen I would break it down by
age, even among consumers. It is interesting.
I, I do think, you know, I, I'veparticipated in a few round
(32:15):
tables recently with young people talking about AI, how
they think about it from a career perspective, education
context. And that's where I really sense
that it is patchy. You know, the, the antidotes
I've heard was at some schools literally just in the last year
have changed policies from like,don't use AI to look, you have
to use AI. And so I think students feel
(32:36):
that whiplash. The, the other thing is, I, I
actually think we need to think through and Australia is a great
place to have this conversation as, as, as it kind of leads the
world on the social media ban. But I think social media is a
great example where I think we, we see a lot of bad AI being
used. You know, I think there's
(32:56):
increasingly a sense that you think about videos today on
social media, no one really knows what's AI generated
anymore. And so figuring out how do we
create trust in that product where people don't think of it
as just AI slop, but actually it's AI that adds value.
It's AI that I want to see and Iwant to, I want to kind of
consume. And so I, I think it's, it's a
(33:19):
mixed bag right now, but I thinkthat also means that there's
work that we need to do as a society, right?
Yeah, absolutely. I mean, I presume the young,
broadly speaking, that the younger a person is, the more
they're likely to embrace and trust AI.
But perhaps that isn't perfect. But I certainly hear what you're
saying about the whiplash that alot of young people might be
(33:39):
feeling. You know, last year, it might
have been cheating. This year it is, you know,
learning to use it as a tool effectively.
And so, yeah, fascinating. Yeah.
So in, in the workplace, I look,we, we fully embraced it, but I,
I also think everyone now is figuring out like, what's the
next wave? What do we really see the
(34:00):
application beyond generative AIbetween, you know, ChatGPT or
Microsoft Co pilot. And so I think that'll be an
interesting time. But I also think that's that's
where opportunity is created, not just for Australia but for
every country in Asia. Yep, Yep.
All right. Terrific.
All right, Mike. Yay.
Thanks so much for coming on. Stop THE World.
Appreciate it. Thank you.
(34:20):
That's all we have time for today, folks.
Thanks for listening to Stop theWorld.
We'll be back with another episode soon.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Ruthie's Table 4
For more than 30 years The River Cafe in London, has been the home-from-home of artists, architects, designers, actors, collectors, writers, activists, and politicians. Michael Caine, Glenn Close, JJ Abrams, Steve McQueen, Victoria and David Beckham, and Lily Allen, are just some of the people who love to call The River Cafe home. On River Cafe Table 4, Rogers sits down with her customers—who have become friends—to talk about food memories. Table 4 explores how food impacts every aspect of our lives. “Foods is politics, food is cultural, food is how you express love, food is about your heritage, it defines who you and who you want to be,” says Rogers. Each week, Rogers invites her guest to reminisce about family suppers and first dates, what they cook, how they eat when performing, the restaurants they choose, and what food they seek when they need comfort. And to punctuate each episode of Table 4, guests such as Ralph Fiennes, Emily Blunt, and Alfonso Cuarón, read their favourite recipe from one of the best-selling River Cafe cookbooks. Table 4 itself, is situated near The River Cafe’s open kitchen, close to the bright pink wood-fired oven and next to the glossy yellow pass, where Ruthie oversees the restaurant. You are invited to take a seat at this intimate table and join the conversation. For more information, recipes, and ingredients, go to https://shoptherivercafe.co.uk/ Web: https://rivercafe.co.uk/ Instagram: www.instagram.com/therivercafelondon/ Facebook: https://en-gb.facebook.com/therivercafelondon/ For more podcasts from iHeartRadio, visit the iheartradio app, apple podcasts, or wherever you listen to your favorite shows. Learn more about your ad-choices at https://www.iheartpodcastnetwork.com
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com