March 17, 2026 • 70 mins
Dr. Christopher Lamont is Professor of International Relations at Tokyo International University and Deputy Head of Program for AI and Global Governance at the Global Governance Institute in Brussels. His research career has been built around transitional justice, international criminal law, and how societies rebuild after institutional collapse. In this conversation, Brittany brings that experience to the AI age.
This is a conversation about trust. What destroys it, how societies rebuild it, and why it may be the most consequential word in the AI era that nobody is taking seriously enough.
IN THIS CONVERSATION
→ Whether societal change really has a before and after, and what that means for how we approach AI transformation
→ What digital sovereignty actually means in 2026, and why it is often being pursued with the wrong tools
→ Why which foundation model your company adopts is a values decision
→ What thriving post-crisis societies have in common, and how that applies to building AI-ready organisations
→ The accountability question that policymakers and business leaders are not asking
FIND US
→ Instagram: @businesskaraokepodcast
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (00:00):
In in societies that have experienced significant
traumatic events or dislocationssuch as conflict is the complete
breakdown of trust.
Not just trust between kind of athe person and a state or some
governing authority that mighthave abused that trust through
(00:23):
committing serious human rightsabuses or doing other things.
But also at a very interpersonallevel as well.
SPEAKER_01 (00:31):
Today we're joined by Dr.
Christopher Lamont.
Dr.
Le Mont is a professor ofinternational relations at Tokyo
University, Deputy Head ofProgram for AI and Global
Governance at the GlobalGovernance Institute in
Brussels, and in his free time,the co-host of the Age of AI
podcast.
Chris is also my friend, so Idid my best to call him Dr.
Lamont, but I think I actuallyfailed miserably and I called
him Chris most of the time.
(00:53):
His research career began aroundtransitional justice,
international criminal law, andhow societies rebuild after
institutional collapse.
This conversation really doescenter around trust.
What destroys it, how societiesrebuild, and why it may be very
much the most consequential wordin our AI era.
I always so much enjoy uhreading your comments and your
(01:14):
thoughts.
So please go ahead and reach outwith your thoughts about this
episode.
Okay, everyone, let's jump in.
So Christopher Lamont, welcometo the Business Charity Podcast
for our human civic AI ClubSeries.
Welcome.
Thank you.
SPEAKER_00 (01:25):
It's a pleasure to be here.
SPEAKER_01 (01:26):
So uh I wanted to share a little bit about why I
was really interested in yourspecific views and your lens of
the world.
The as I was preparing for thisum for our interview, uh my um
AI and I decided to call you theLamont lens.
This was the Lamont lens.
SPEAKER_00 (01:42):
Interesting.
Oh wow.
Okay.
Well hopefully I live up to thatlens, or the lens is one that
makes things clearer, I suppose.
SPEAKER_01 (01:49):
So the Lamont lens.
So let me share a little bitabout as I was kind of chatting
and saying this is why I'mreally interested um for um
Chris to come on.
So um for those of you that arethinking why I'm not calling Dr.
Chris Filmont Dr.
Le Mont, because Chris and Ialso are friends.
So I can't I'm allowed to callyou Chris.
SPEAKER_00 (02:04):
Yes, so I prefer to be called Chris also.
It's it's it's kind of yeah.
SPEAKER_01 (02:07):
I'm really interested when it comes to AI.
So, as you know at DCJ, we'rehuman-centered AI, which means
don't begin begin with thetechnology, begin with the
human.
Right.
And I'm gonna share with you astory that I heard recently that
was told to me recently, andthen kind of how that um then
informs kind of our thinking.
So one was I was chatting withone of the professors from Grips
University, so at the GraduateUniversity for Policy Studies
(02:30):
here uh in Japan, and we werechatting about AI and
cybersecurity, security.
And he said, I don't know whyeveryone is kind of jumping up
and down about security now.
And I was like, that's reallyinteresting because I think AI
and security is probably like areally big issue.
Right.
And I said, Do you want toexpand more?
Like, what do you mean by that?
And he said, a hundred yearsago, when you wanted to send a
(02:52):
letter, you would send somebodyand they would jump on a boat
and go somewhere, but thisperson could equally be
assassinated or murdered orkidnapped.
Right.
Yeah.
And not that any of this isfunny, but what it kind of
highlighted was that yes, we arein a new time, being with uh,
you know, the AI age, but theissues of security, the issues
(03:14):
of unforeseen circumstances havealways been with us.
And I'd really like to beginwith that with that um as how
it's connected to kind of you,is that I've always found it so
interesting how you have seensocieties go through immense
change.
Immense change.
(03:34):
So, you know, we can start niceand light and talk about how,
you know, you've studied, youknow, war crimes in the Balkans
and things like that.
I mean, we're not three minutesinto the episode, but I, you
know, we can talk about warcrimes.
And I'm really interested from abefore we jump into the
technology and the people, I'dreally love to hear your
thoughts or your views on, evenbefore jumping to AI, when
(03:55):
they're society and they'regoing through immense change,
what are some of the things thatusually lead up to those kinds
of immense changes?
Is there kind of like then anabsolute like switch where it's
like before and after thismoment type thing?
And then the rebuilding.
So with that, um I thoughtthere's no one better than you
to come and talk to us throughthe society evolution when it
(04:20):
comes through, obviously, umwhat you know, and then
comparing that to where we arewith AI.
So that is my thinking around umour episode for today.
SPEAKER_00 (04:28):
Okay, well, lots of questions to um to respond to
coming out of that.
I think that what you kind ofended with, and that's the
question of like, okay, the thelike transformation as a switch,
the before and after.
Um in most types of change,right, processes or society or
(04:48):
where societies have experienceduh some sort of massive um
disrupture, it actually isdifficult to identify that like
one point in time, right?
So in countries that haveexperienced, for example, civil
conflict, right?
So even a question like when didthat conflict start?
Was it when kind of the shootingstarted, or was it when there
(05:08):
was a political breakdown, orwas it when there was some sort
of incident that happenedpreceding that?
And then also when does it end,right?
Is it when kind of a peaceagreement is signed, but there's
still high levels of societalkind of suspicion, distress,
walls separating communities,for example?
Um and and that fluidity interms of how we experience
(05:32):
change cuts against how we talkabout change, because we tend to
assume that there's like abeginning and end, right, um, to
any major kind of historicalepisode, and then apply that to
technology and what youmentioned early on is what makes
AI different from other securitychallenges or other types of
governance challenges that we'veencountered in the past.
(05:54):
Right?
So if you think about AI andsurveillance, right, that states
have had the capacity to survey.
And um, often you could usepeople, you would use people to
do that decades ago to survey.
But there are a coupledifferences that have that have
kind of been creeping up on usthat we might not see without
(06:15):
kind of perfect hindsight.
And and one, and this issomething that gets to the kind
of broader governanceconversation, is just how much
data is out there about all ofus, right?
That we are, of course, notnecessarily kind of live
streaming our lives as such, butas a figure of speech, right, in
(06:36):
a way we are, right?
All this data is out there aboutus.
And then you look at that andyou aggregate that.
And what you can do with thattypes of that type of
information, what that enablesin terms of um not just the
private sector, but also thepublic sector in terms of
governance for good and bad, issomething that um previous
(06:59):
generations did not have accessto when thinking about
governance or thinking aboutsociety.
And I think that's that's that'san important difference.
And so if you think about alsothis question of security, what
type of security vulnerabilitiesdo we have?
Well, um if an individual wassomehow compromised, who was
(07:22):
carrying a message in the past,right, to go back to the the
analogy, um you would, if you ifyou identified that, be able to
limit the exposure you have.
But if your kind of data systemsare compromised, right, and you
don't know it, right?
Or even if you do know it, butyou don't know how it's being or
(07:43):
where it's being or what to doabout it, um the scale of the
challenge and the potential harmthat it can bring about has
grown significantly.
Right.
And so I think that part of thisseemingly like, oh, everybody
now is talking about securitywhen it comes to AI or thinking
(08:05):
about some of the kind ofcybersecurity implications.
I think that's natural,particularly when it comes to
developments like agentic AI.
But I also think, yes, there is,it's helpful to look to the past
for parallels and see kind ofhow we have addressed similar
challenges, but to be carefulnot to apply the tools of the
(08:29):
past to the challenges andproblems that we are facing
today, right?
Because the um the nature of, orso the society that we're
seeking to govern has changedhow it communicates, right?
I'll use a very kind of like ifwe go back 100 years, for
example, um, I also teach on akind of international program at
(08:51):
a university.
And one of the things that'sinteresting to observe, or just
like you think about, is that,well, a hundred years ago, for
example, you could not haveinternational programs the way
you do today, where with peoplefrom all these different
continents coming together,right?
Like if you were um going totravel intercontinentally, it
(09:12):
would probably be like a once ina life, or maybe twice in a
life, or or maybe if you'resuper, super wealthy, right?
A few times, right?
But these the the these thesejourneys take a lot of time.
Um and then once you gosomeplace else, you need to like
learn the language, adapt life,right?
Um it it's it's a it's it's alifetime commitment.
Um but in the 2020s, of course,um, it's something that for
(09:36):
young people in a lot ofdifferent parts of the world,
that's kind of natural.
Well, think about like whichcountry do you want to study in,
right?
Um so on the one hand, right,mobility has increased, but also
if you think about the world inwhich they live.
If you come to a country likeJapan, for example, you're um
you remain connected on socialmedia with all your friends back
home, right?
You can have friends also makenew friends, like virtually who
(10:00):
are on who are on differentcontinents.
Um and also you can remain inyour like own media environment
from back home.
You can kind of keep watchingyour favorite series, you can
keep doing all this, right?
And so the the world, right, thethe the the the kind of the
experiences that people havetoday are much more
(10:23):
interconnected than they were inthe past.
And if you think about that interms of AI and governance,
right, that that's also true,right?
Because what what we're seeingat a very individual level, your
own kind of personalexperiences, how you are
experiencing the world, has alsochanged.
(10:45):
For example, business, right?
Think about supply chains, thinkabout all these other elements
of interconnectivity.
And so it's an interesting timeto be talking about that in
2026, of course, because itseems that, well, um, a lot of
those um institutions thatfacilitated that
interconnectedness are underincreasing stress.
(11:07):
But there's there's no way, Imean, right now, right, to think
about it to kind of pull back,right, the um turn back the
clock when it comes to thesetypes of societal and individual
level changes and how technologyhas changed our lives, right?
So the the policy responses tothat would reflect the world we
(11:30):
live in today, not 100 yearsago.
SPEAKER_01 (11:32):
That's so interesting.
I think the the the point that Iwanted um dive a little bit more
into is the uh when youmentioned that the tools of
yesterday can certainly forinform, but shouldn't
necessarily form our approach totomorrow, right?
So use them as for aninformative rather than a
formative um uh approach.
(11:52):
And I think this is interestingand almost an opportunity to
maybe even um break down or evendefine the idea of sovereignty
and digital sovereignty.
So I remember, so I'm from anisland, um, even though I have a
very odd accent, it's still anAustralian accent to some
extent.
Um I remember my grandmothertelling me um during the Second
(12:16):
World War, they in the papers,they basically told, sort of
they being the government ofAustralia, told everybody, don't
worry, the war will not reachour borders because we are an
island and the Japanese havebicycles.
And how are they gonna get fromJapan to here?
(12:36):
Honestly, there was like you canthere was a front page cover
with a Japanese person on abicycle and the Australians on
land kind of looking safe andkind of laughing.
SPEAKER_00 (12:48):
Right.
SPEAKER_01 (12:48):
And then I said, Well, how did that end up?
She goes, Well, Darwin wasbombed, you know.
SPEAKER_00 (12:52):
Right.
Aviation.
SPEAKER_01 (12:54):
Aviation.
And so the idea that asovereignty used to be this is
my land, you're in your land,that that's very much not what
sovereign means anymore.
SPEAKER_00 (13:06):
Yes.
Um so this is a very goodquestion, right?
So we're talking aboutsovereignty itself.
I'd like to focus on, though, II think you mentioned digital
sovereignty there, because Ithink that's that's something
that is increasingly talkedabout and um is something that
um a lot of states have begun toembrace.
(13:27):
Now there's sovereignty as weknew it in the the past century,
and there's and and how we seethe world is actually very much
informed by sovereignty, right?
You if you were to imagine, forexample, a world map, you would
imagine a world map with stateborders, right?
That's that's that's like you'veclosed your eyes, right?
Yeah, flags all the in in allthe countries.
(13:49):
And of course, kind of the thereality of geography is
different, but nonetheless,that's that's what informed our
how we see sovereignty.
It's it's territorial, it'slinked to kind of a population
and political authority, and allthese things were nicely
aligned.
And in this world that Idescribed that had emerged with
(14:12):
kind of this deepinterconnectedness and also
fluidity and mobility, um, whatsovereignty kind of means in
practice has has begun to shift.
And it's shifted so much nowthat policymakers are beginning
to seek to find, look for tools,right?
(14:34):
Like, okay, so um, if you thinkabout other kind of strategic
industries in the 20th century,like the steel industry, right?
So this was something that waskind of closely associated with
national power, it was somethingthat you could um subsidize,
prop up, and even the smallestof countries could have their
own steel mill and locallyproduced steel, right?
(14:56):
And so that mindset isinforming, unfortunately, how
some people think about digitalsovereignty.
So if we just, if we just buildthe data centers, right, or if
we um just have a nationalchampion, like a kind of a kind
of a foundational model, right,that is, that is developed
domestically within our country,right, we have our own digital
(15:20):
ecosystem that we can stick ourflag to and feel better about
ourselves.
But as you know, as somebody whokind of works in this in this
field, that it's actually muchmore complex than that.
If you think about, for example,where tech companies get
funding, right?
This this often crosses borders.
(15:41):
And if you also think about datacenters themselves, the question
of like, well, who owns the datacenter, right?
And you think about data centersthat are built in different
parts of the world as part ofthese initiatives to um increase
a state's data sovereignty,well, most of the know-how
operators kind of and also kindof the rules governing them come
(16:03):
from outside, right?
So you might have, on the onehand, um, increased a particular
capacity, but that doesn'treally fit your understanding of
sovereignty, but is being usedas an indicator to measure
sovereignty, right?
And so in this particularcontext, then the the the
(16:25):
traditional tools to bring aboutsovereignty or to achieve
sovereignty are tools that werehoned in an era where you were
governing something that wasless internet connected and less
complex.
And also when I sayinterconnectedness, I'm not just
(16:47):
talking about interconnectednessacross borders, but I'm saying
interconnectedness within the AIecosystem itself, right?
So if you think about, okay, youhave these different regulatory
instruments that are governingdifferent parts of it, like you
might be thinking about kind ofthe services aspect, or you
might be thinking about the kindof the kind of design aspect
(17:08):
with through the EU AI Act, orum you might be thinking about
like platforms.
Um all of these areinterconnected, meaning that if
you are thinking aboutregulating one particular
aspect, you might begin to feellike you're playing
whack-a-mole, right?
(17:28):
Um in terms of of governance.
And also, as we mentioned, likethe elements of of change,
right, that is taking place isone that it's it seems obvious
when you talk about it in aconversation like this, um, but
uh in terms of kind ofgroupthink, right?
(17:49):
When you have people in a roomthinking about, okay, so so what
can we do to achieve digitalsovereignty, they tend to fall
back on some of the same kind ofindicators that that I mentioned
before.
SPEAKER_01 (18:03):
And if we think about digital sovereignty as,
for example, sovereignty, maybewe don't even need the word
digital, like maybe it's justsovereignty in 2026.
If we think about traditionallylike sovereignty and what that
what that means, it's deeplyemotive to people.
If you were to take somebody'sflag and burn it versus fly it,
(18:26):
right?
There's people have deeplyemotional reactions to this.
But if you were to ask somebody,for example, how they feel about
the data center or or their owndata, for example, I um have not
seen such a natural humanresponse to that.
(18:47):
And I'm and I'm wondering, youknow, the symbols of flag, the
symbols of nation, who are you,who you are not?
And there's even almost um, youknow, many people who listen um
to our podcast are similar tous, so either global
professionals in Japan orJapanese professionals globally.
And you know, we always get thequestion, where are you from?
Where are you from from?
(19:08):
Right.
You know, this kind now, youknow, as you kind of mentioned,
there's this switch of like whendid change kind of happen?
When did this borderless youknow kind of change?
Well, uh perhaps there's alsoelements of the fact that you
could, you know, for example, myhusband is is is now is Mexican.
I have a an Australian-Mexicanchild.
So there's no the borders are,for example, now in many
(19:32):
circumstances are kind ofsurpassed.
And so I'm thinking it's quiteinteresting if we go back to
like the symbolism of a flag.
And I'm wondering, will we getto a point where people feel so
deeply passionate about theirdigital sovereignty as they do
kind of national sovereignty?
Um and if they did, like whatwould that what would that look
like?
SPEAKER_00 (19:53):
This is this is this is interesting.
I think this is this gets to oneof the reasons why this
conversation is beginning tohave this kind of outsized
presence.
Um because it resonates withpeople's sense of identity,
(20:14):
right?
So if you look at digitalsovereignty conversations in
different contexts, they meandifferent things, right?
So if you were to take someAmerican con uh understandings
of this term, um, of which thereare many, um, we can get into
that maybe a little bit later.
But but one would see this as,well, digital sovereignty means
(20:34):
that American tech firms shouldnot be subjected to foreign
regulatory oversight, right?
Um and and that's very emotive,right?
So it's talking about like,well, the these are US companies
and these are foreignregulators, right?
Which which creates kind of thisvery zero-sum image of kind of
digital sovereignty and and thedigital space.
(20:56):
Um for others, it's a it's adevelopmental goal, right?
It's it's well, um our countryis facing developmental
challenges, right?
You see there's this remarkablenew technology.
Does it have the potential tolevel the playing field?
Can countries skip certainstages of of development?
Um and therefore we won't ask alot of questions about well, we
(21:17):
don't data centers, right, whatthey're actually doing, right?
What the what kind of individualum data protections at at that
level, right?
And so I wanted to like justlike put a caveat to what I said
earlier, simply because um whileour experiences are much more
interconnected and while kind ofthe technological infrastructure
(21:38):
is much more interinterconnected, um, we still
live in a world of nationstates.
Yeah.
Right.
Um perhaps the most importantum, I guess, identifying marker
that people possess would betheir passport, right?
And be kind of the country ofthe from, and as you mentioned
also.
Um it's also a very emotive kindof Identity as well.
(22:02):
And you do have models ofdigital sovereignty where
countries have sought toessentially develop their own
ecosystems domestically, right?
And so China being kind of thebiggest example of that.
But of course, China with alarge population, and China with
kind of very top-down governancewhere national leaders establish
(22:23):
very ambitious targets when itcomes to, even before AI, other
digital goals.
And then also as a countrythat's seeking to export its
technologies through mechanismslike the digital Silk Road,
right, you can the the flag isvery visible, right, in that
(22:44):
sense.
But most countries in the worldare not the US or China, right?
Obviously.
And how do those countries thensee digital sovereignty, right?
So if you think aboutinterconnectedness, I think one
of the places where this is mostvisible, where you're seeing
(23:05):
conversations of governance, butalso challenges associated with
governance coming to a head isEurope, which is not too
surprising because the EuropeanUnion was an early mover in
terms of attempting to regulateartificial intelligence.
But you are seeing now in 2026kind of a um significant
(23:28):
distance between the ambitionsthat originally surrounded early
discussions around the EU AI Actto now the EU omnibus, right?
Which is arguably taking a stepback from some of the larger
ambitions that the EuropeanUnion once had when it came to
(23:52):
governance in this particularrealm.
In large part also becausedifferent EU member states now
see digital sovereigntydifferently.
So if we are to look to France,right, France is one that's
seeking create nationalchampionship.
And have, yeah.
And have, right, and havecreated national champions.
But but these national championsalso exist with this very very
(24:14):
kind of deeply interdependentkind of network of either
funding or kind of bringingtheir products to market, which
means that it's not like thekind of a French-owned steel
mill, right?
Um, in terms of actually beingan indicator or be being a kind
(24:40):
of sovereign entity, the way Iguess paradoxically, some of the
kind of Chinese counterpartsmight be in a system where you
have a top-down kind of statepicking um those um national
champions within the technologysector.
Um immense challenges, right?
(25:01):
And I think this is one of theinteresting um aspects about
this conversation is it's umit's very easy to move to kind
of almost caricature extremes ofsovereignty and interdependence,
like this kind of world where,okay, well, nation-states matter
less.
(25:23):
I would argue that very fewpeople in 2026 are seriously
making that claim thatnation-states matter less today,
in the 2020s, than they did inthe past, given the level of
geopolitical tension that youare seeing in the world today.
(25:44):
But the combination of kind oftechnological change and
societal change means that howwe address problems of
government-to-governmentrelations in the 2020s is going
to change, right?
And so I think this is one ofthe big questions in terms of
(26:08):
where this technological spacewill lead to.
Are we going to be in a worldwhere it's possible to have many
different digital sovereigns?
Right.
Um, or in a world where thereare maybe two digital
sovereigns, right?
And then you're kind of in thesedifferent camps between the two,
(26:32):
right?
And so like if you think aboutlike what does digital
sovereignty mean, um indifferent countries, it like I
said, it means different things,right?
So some countries might seedigital sovereignty as saying,
okay, well, let's um distanceourselves from the US tech
ecosystem, right?
Whereas others might have anopposite view.
(26:53):
Um and some are also thinkingabout kind of re um rethinking,
as you mentioned at the verybeginning, how we think of
sovereignty in a world where youhave a few very large big tech
players, right?
And interesting here, they'rethey're historical parallels,
right?
(27:13):
Um as well.
unknown (27:15):
Yeah.
SPEAKER_01 (27:16):
That is um also definitely terribly interesting.
I'd like to unpack a few thingsthere.
I think the first thing is, asyou mentioned, um the team there
seem there's two players, right?
So it's either the US or China.
Now, this is I I reflect a loton this, not being from the US
nor being from China.
(27:37):
Right.
Um and for a couple of reasons.
One is, does that just meanthere's like a haves and a haves
nots?
Like there's an AI haves and anAI have-nots, and that's kind of
like it.
Is that really all we have?
Um and then another interestingapproach is as you said, there's
like the US approach, which isalso quite culturally informed
in a way.
Um so, for example, if you takeum the foundation model of Grok,
(28:01):
now Grok's like flag in theground is truth.
We you, you know, the Grok's itlike the number one factor is
grog should be truth seeking.
Right.
Um and then, for example, youhave uh on the other hand, you
perhaps, for example, China'smany foundation models, one of
which, for example, is DeepSeek, and is an open or they've
(28:22):
taken a different differentapproach and had it be an open
model.
And so Malaysia, as of 2026, Imean things are always changing,
so but it's 2026 February.
Um Malaysia has adopted DeeksDeep Seek as their national
model, which I think is reallyinteresting.
And then does, for example, inthis moment, do you need to
have, you know, what what wouldthe theatrics of a of a an
(28:47):
American delegation to Malaysiabe, where it's like, oh coming
and yes, let's, you know, haveour good relationships and you
know, you're spending ice massand all that kind of stuff.
Um that's one thing.
But then if a country is using afoundation model of like another
country, for example, that youof which you uh don't see eye to
eye on many things, this is theninforming not only the way that
(29:09):
uh this entire country isoperating, but also thinking and
creating.
Um this would just the simpleadoption of a foundation model
would means more than a thousanddelegation visits to a country.
SPEAKER_00 (29:23):
Um yeah, and exactly, and and what you know
about a country.
Right.
I mean, because if you'rethinking about this in the
context of public services, umand you're thinking about, okay,
the need to um digitize allthese records that a state might
have, right, so as to make thesemodels useful.
(29:46):
And then you think about who hasaccess to that data, right?
Um you're in a world wherepotentially um the tech players
from either side know way moreabout not just what's happening
in your country, but in othercountries as well through access
(30:09):
to that data.
And so this is where if you'rethinking about um adoption and
diffusion of technology, thatChina has a slightly different
perspective here.
That if you look at therationale behind um DeepSeek and
(30:31):
making that open, is that thebet is that China can diffuse
its models faster and easierthan the US can, because the US
tech ecosystem is embedded inkind of a highly monetized
system in which you needcountries to buy into, and also
(30:53):
um you don't want these modelsto to open, right, because
they're proprietary and you haveinvested or somebody has
invested billions of dollars inin that.
And in some cases, and this iswhere if you think about what
conversations are happening, umthat that that could be
(31:16):
something that could be harmfulto disproportionately harmful to
US tech, right?
As there is a perception that umstates are leveraging types of
vulnerabilities that you havedescribed um in order to achieve
(31:37):
different policy outcomes,right?
And so there is a significantnervousness about this.
And this is where um so OscarGestrain, who I had a
conversation with um on kind ofdigital sovereignty, where he
likes to use the term um digitalautonomy.
SPEAKER_02 (31:56):
Right.
SPEAKER_00 (31:56):
So as opposed to sovereignty, because what people
are saying is digitalsovereignty, but when you push
them on it, they don't actuallymean sovereignty.
They're just using that becauseit's a it's it's a word that
everybody kind of understands.
But the the actual meaning ofsovereignty, as we talked about
earlier, is something that'svery difficult to apply in the
(32:19):
context of the technologies thatwe are talking about, and also
the type of technologicalcompetition that we're talking
about.
So if you think about autonomyas a state, that's actually what
you you want.
You want the ability toautonomously act, right?
And that's the feature kind ofof sovereignty that you're
(32:41):
really seeking to get at, notnecessarily putting your flag on
everything, right?
And then you're gonna startasking a different set of
questions.
And so um, as you mentioned,there's differences in culture
kind of between the US andChina, but there's also
differences of cultureinternally, right, um, within
states and their and theirfoundation models, right?
(33:02):
Even kind of between Grok andkind of OpenAI or Anthropic,
right, you're noticing umdifferences in terms of the ways
these models make sense of andrespond to and interact with
users.
And so if you're thinking aboutautonomy, you might be thinking
(33:22):
about kind of value alignment.
Now, of course, there arelimitations to that, and those
limitations are most visiblewhen you begin to um discuss
sanctions, right?
Um, and um these companiescannot divorce themselves from
the country that they are theyare based in, and whatever that
(33:44):
country's you know foreignpolicy's goals are, whether
that's like you're talking aboutBeijing or you're talking about
Washington, um, right.
Um and in that sense, right, Ithink that a lot of the
conversations about digitalsovereignty are actually a
response to that nervousnessthat um if we are reliant on um
(34:06):
a company that is flagged inanother state to provide kind of
very basic things that we needand that we cannot do without,
um what's to prevent that kindof country from leveraging that
dependency over us to get us todo things that we otherwise
(34:27):
don't or would never do, right?
Um and again, I think there whatyou're trying, what you're
looking at is this question ofautonomy.
And autonomy might actually comefrom an interdependent network,
but of a different constellationof countries, different
constellation of actors,different constellation of
companies, right?
(34:47):
Um that is not as narrow as thiskind of idea of absolute
sovereignty.
So you could have countries orcompanies in countries with kind
of similar sets of kind ofvalues, right?
That we might think about kindof data privacy as one.
Um, and then aligning to kind ofwork together um to um ensure
(35:13):
that kind of basic level of dataprivacy protection across the
board and kind of allow foreasier interoperability across
cross-borders.
But so perversely, converconversely, right?
Like um what thisinterdependence means is that we
it's it's not that countriesmatter less, right?
(35:34):
But it's that we um need to havea greater awareness of and a
greater understanding of who weare working with to provide us
with something that we need.
Right.
Um and what does that mean forour autonomy going forward,
whether you're a government or aprivate firm?
SPEAKER_01 (35:56):
That that's very interesting.
Um, as someone who's Australian,I can speak to like fake
autonomy.
So fake autonomy is what we havein Australia, which is like
well, yes, Australia, you're acountry, but if there's ever a
war, we're the Commonwealth.
And all of a sudden we're theCommonwealth.
(36:17):
No, Australia, you're your owncountry, so you need a visa to
work in the UK, but our royalspaces are all over your money.
So like also like a very oddkind of um situation.
Um, but just to kind ofunderstand how deeply valuable
this is, Eric Schmidt is a CEOor the ex-CEO of Google who's
deeply invested in the idea ofperhaps AI autonomy.
(36:39):
Um and that, you know, he he wasmentioning in a in a recent
keynote, you know, under nocircumstances can we can we risk
just a few billion dollars forlosing the race, losing the AI
race.
SPEAKER_02 (36:55):
Right.
SPEAKER_01 (36:56):
And so this is where you're saying, okay, as you
mentioned, the AI, the AIsystems of the US are deeply um
embedded in in financial models,as in how who's making money
from this, how we're makingmoney from this, what's the
business model.
But he was saying, just to likeunderstand how valuable this is,
he's like, who cares about a fewbillion dollars if we lose the
AI race?
(37:16):
And I thought this is superinteresting because I'm kind of
seeing a little bit of a curve.
So that we go from individual,the middle is kind of like a
little bit gray and very lots oforganizations, and then there's
the end user.
So I'll tell you a little bitwhat I mean.
So I find and perhaps this mighthave always been the case, um,
but for me personally, I feelthat um politics is incredibly
(37:37):
individual at the moment.
So for example, it's not the US,it's Trump, it's not Japan, it's
Takaichi, it's not France, it'sMacron.
Like they're individual people.
Then there's this kind of likegrey mass in the middle, which
is like uh Washington or Beijingor some kind of you know uh
committees that sit somewhere,and then there's like the end
(37:58):
person, me as a person.
So it's like, how do you connectwhat's going on from like that
individual like countryleadership?
How do you navigate the gray?
And then what does that thenmean to someone that's getting
up every day and just trying torun their business?
SPEAKER_00 (38:15):
Oh, this is a this is a great question because I
think you're you're touching onkind of a couple different
aspects here, right, um, thatthat I can go into.
And I think first you mentionedthe AI race, right, in this
framing.
It's like, okay, traditionally,if you want to kind of mobilize
investment in something, right,you need to present a business
(38:35):
model case that investors willinvest in.
And with the framing of the AIrace, right, you it's it's being
framed as, well, this is um thisis existential for the country.
Yes.
Right.
And this is this is this isinteresting because this gets
back to kind of Americanunderstandings of digital
sovereignty and sovereigntyitself, right?
That if if somehow in thisframing of the race, the US is
(39:00):
not out front, does not win it,right, then that's it, right,
for kind of for the US.
And so therefore, you need kindof an all-of-society kind of
response to this, where kind ofthe government mobilizes private
capital um into making sure thatthe US comes out in front.
(39:24):
This is actually not though,because this is interesting.
So you mentioned that that thatthis is this this language um
people might associate with, oh,this sounds like all kind of the
the kind of the tech techleaders who were surrounding
Trump and who attended hisinauguration, right?
This is this is kind of theirmessaging.
SPEAKER_01 (39:40):
Yes, exactly.
SPEAKER_00 (39:41):
The the the messaging.
But but the kind of the US iskind of the it was the Biden
administration that came outwith a AI strategy, right?
And interestingly, um in theforward framed the AI race using
the exact same language, andalso some of the exact same
people who you're talking aboutwere involved in the kind of the
crafting of that commission thatcame out of a Biden
(40:03):
administration presidentialcommission's report, right?
Um So there's there there hasbeen a significant amount of
continuity in terms of this kindof this framing of the AI race.
Um but you mentioned kind of theTrump administration is doing a
lot of things differently, notnecessarily here, but in terms
(40:28):
of AI governance, how itunderstands AI risks, kind of AI
safety institutes, deconstructalso kind of international
cooperation when it comes todigital technologies.
Uh one of the things that theprevious US administration was
supportive of was, for example,the UN Digital Compact, right,
which was a high watermark forinternational cooperation when
(40:50):
it comes to AI and sustainabledevelopment, and also some of
the more ambitious goals thatframed AI as a catalyst in kind
of a win-win technological racethat can decrease global
inequality rather than increaseglobal inequality.
And also in terms of AIgovernance, right more broadly,
the previous administration hada very different approach to
(41:12):
risk and safety and wasembedding the US into this
growing network of AI safetyinstitutes, and also was part of
a whole lot of conversationsthat also went beyond kind of
civil governance of AI tolooking at military applications
of AI and the such.
And so there the Trumpadministration has taken a very
(41:33):
different approach, right?
The US US policy is completelykind of reversed on these
issues.
And policy is less predictable,right?
The policy process of theprevious administration has not
been replicated, right?
Um that's kind of along-standing kind of policy
(41:54):
process that then means that,well, US policy could be
something one day, but then Iguess I don't know what the word
uh the the president truthssomething out.
Um and then something is isradically changed.
And this is interesting in asense that this is not just,
again, not just kind of theTrump administration in the US,
(42:16):
but this increasingly kind of umpersonalistic uh approach to
governance in work with worldleaders today, right?
So as mentioned kind of TakaiChi's election in in Japan,
right, a lot of focus on theprime minister as a person, but
then forgetting there's like acabinet office and there's a
political party and there'sministries, right, that are all
(42:38):
influential in these processes.
Um and then also if you thinkabout how kind of course she is
very different in in China,right, um, discussing um, but
but but there are all theseconstraints, right, regardless
of um what leader we are talkingabout.
And that's the gray area of ofthe process that you were
(43:00):
talking about.
And so for people like lookingat this from the outside, what
you'd want to do is look, okay,what do you identify as the
policy problem, or what does itwhat are leaders talking about
as the problem?
Um is a good indicator that willtell you a lot about like what
the subsequent policy processwill will look like, right?
And that's why I thinkconversations about
(43:21):
conversations about digitalsovereignty are very important,
because that's at the stage of,okay, so what's the problem?
Right, what what is it we aretrying to achieve, right?
And they're different ways ofdifferent visions of that, but
some of them might actually beharmful in the sense that, yes,
the the things that could havemade you more sovereign in the
past are actually going toincrease your vulnerabilities
(43:42):
and dependencies in the presenttoday.
And kind of this what what doesthis mean for the average
person?
Well, I think this is veryimportant.
I think this is something thatpeople don't think about
sometimes um enough.
Uh that that often When it comesto issues that we might think
(44:03):
are very important, for example,like data privacy or
cybersecurity, individual usersmight think, oh, well, what's
what's wrong?
So just to use like another kindof parallels, if you think about
like open open AI and ChatGPTs,kind of subscription models,
right?
People upload all sorts ofpotential, like personal
(44:26):
identifying information,potential propriety information
for all sorts of reasons, right?
And they don't think a lot aboutwhat that means for them.
And in fact, for most people,there's probably not going to be
an individual like hugerepercussions.
But if you think about this atan aggregate level, um well,
(44:49):
this provides one particularkind of firm with a whole lot of
data that otherwise they wouldnever have had had access to.
And you mentioned Malaysiaearlier and other countries are
thinking about, well, if youhave all that very kind of
cultural specific knowledgeabout our society, um then going
(45:16):
into training something that wehave no insight into or no
control over, that potentiallyone day if relations break down
between our countries could beused in all sorts of ways that
we're not thinking about, right,then that that is that is a
concern, right?
And I think that for those whoare thinking about these
(45:39):
conversations, to think aboutkind of well, there's different
levels of of individual um Iguess, stakes that are that are
at play in in this conversation.
So it's it's it's much moreimportant, right, than um
perhaps some other policydebates, right, that that you
(46:02):
might or might not be befollowing.
Um because of its knock-oneffects, and when I say these
knock-on effects, right, if youthink about um if you look at
policy responses, that mightactually make it more difficult
to work with collaborators indifferent countries.
(46:24):
Um a lot of these co-creationspaces on the internet, you
don't really know a lot aboutthe other people who are kind of
helping to refine something,right?
And so you might find yourselfin a situation where um you're
just somebody kind of in the USor Australia working on
something collaboratively in acollaborative space, but you've
(46:46):
been working with somebody who'skind of in China, right?
And then you might find that,oh, you've contributed to,
right, or you might find thatyou've fallen foul of some
domestic legislation that isaimed at enhancing your
country's own digitalsovereignty.
Now, what would the effect be?
Well, the effect would beprobably rather chilling, right,
(47:09):
on all of those creative spaces,like your your own people's
ability to engage in thosespaces.
Um people just like simply like,okay, well, I I don't want to
risk this, right?
Um, and tune out, but then thesespaces will still exist, right?
And so it's for those who arekind of involved in the cutting
(47:32):
edge of this um sector, right,or the sphere, these are really
important conversations becausethey might impact upon your
ability to um to go aboutcreating what it is you have
been creating um in a way thatyou might not have thought
(47:56):
about, right?
Um and it might also just likecreep up on you out of the blue,
right?
Like, oh my gosh, I didn'trealize that this was even
something that I shouldn't orsomebody should not be be doing.
And so that's why these broaderconversations are incredibly
valuable.
So I think in in terms of that,the gray space that to what to
look out for is well look at howthe problem is being formulated,
(48:18):
right?
How it's how it's beingarticulated.
Of course, you won't get insightinto like the back room, right?
But but that that will tell youa lot about the kind of the
direction of travel.
So like look at how differentcountries um talk about digital
sovereignty.
Yeah.
Right.
Yeah.
SPEAKER_01 (48:35):
And and if we think about um an A AI sovereignty or
even just AI or a country'sapproach to AI, um it requires
one fundamental thing, which isit requires the society to get
behind it.
Yes.
It requires every single personin the society to say, yes, this
is this is for us, we're goingto use it.
So it's not just you can developit, but people have to adopt it.
The actual adoption of thetechnology is also key.
(48:58):
And we've seen, you know, as aswe've kind of mentioned um in
this past a few in this podcasta few times, that you can
certainly have the pastinformed, but it can't certainly
form, you know, where we'regoing to tomorrow.
So I think the informative thingthat we can take from the past,
from getting people, you know,behind systems, is that um, you
know, what why did peoplebelieve in their country?
(49:18):
Let's take, for example,America.
Um, Americans love their flag,they love their flag, they love
the American dream because theysaw the results.
They saw if I do these things, Ithen get there, right?
This was a there was a verylinear path.
They achieved it, people aroundthem they achieved it, people on
(49:38):
the other side of the countryachieved it.
It became like this is a thingthat we can do together.
And I trust if I get involved inthis system, it'll take me
there.
Right.
Um, I think when it comes tothat, what would the digital
American kind of dream looklike, right?
And then you can replace Americafor whichever country you like.
Um, but how are you going to getyour society to adopt the
(50:01):
technology?
And then I think the resultsfactor, I think, is also quite
interesting because like theAmericans, it's like, you know,
yes, there is that kind of ragsto riches story that the
Americans love, and they've seenit multiple, multiple times.
Right.
Um and I think people haven'tseen that yet with AI.
Right.
They've seen big companies geteven more rich from through
(50:26):
investment from other bigcountry companies, right?
Right.
But we I don't think we've seenthe rags to riches story as much
yet that helps people believeand trust in a system that they
have in the past.
SPEAKER_00 (50:38):
Right.
This is interesting.
So there there are there are acouple elements to this
question.
And and one relates toessentially kind of the future
of work and in AI, right?
Like will anybody like me have ajob, right?
Um and and the anxieties thatthat creates.
So as I mentioned, right, thatthere's a large push to kind of
(50:59):
center AI in the broaderconversation of US technological
competition, right?
For example, the space race,right, where all the tremendous
amount of kind of governmentmoney was put in to um making
sure that while the US wasn'tthe first kind of in space, the
US would be the first on themoon, right?
(51:19):
Um and similarly, um you havethat mobilization here.
I think the problem of aquestion of adoption for a
country like the US is less of aum challenge.
In part because uh the USdominates kind of platforms that
can bring things to market,meaning that, for example, if
(51:40):
you were a country that waspushing your own foundational
model and you had adversarialrelations with the US and you're
a small country, you're notgonna be on the Apple store.
unknown (51:51):
Right.
SPEAKER_02 (51:52):
Right, right.
SPEAKER_00 (51:52):
Right.
So like a Cuban foundationalmodel, for example, um, if one
right would not would not end upthere, right?
Um and and because that kind ofbroader kind of platform
ecosystem is right now like soUS dominated that that that's
less of a a challenge.
(52:14):
But the second part is what youmentioned is going to be a
problem, right?
Um as people perceive, okay,well, um there's this AI
technological race, there's AIis creating fabulous amounts of
of wealth for those big playersin this industry.
(52:36):
But from the perspective of auniversity computer science
graduate in 2026, there are nojobs for me, right, um that
would have existed just fiveyears ago.
Right.
Um and from the perspective ofthose already working in um
(52:59):
these fields, um companies aregetting smaller in terms of
workforce, not larger, right?
I think, I mean, I'd have to goback and double check with this,
but even like these large datacenters that consume vast
amounts of electricity andprocess vast amounts of data, um
it's not like going to visit afactory, in a sense.
(53:19):
Like there are very few humanbeings who need to work there,
right?
So you're you're using a largeamount of land, right?
You're using a large amount ofresources, but comparatively,
you're not really creating jobsthrough bringing data centers,
for example, to your kind ofstate, if you're if you're in
the US.
Um, you're bringing differentkinds of investment, but but not
(53:42):
necessarily creating jobs andcreating wealth at the lower
end.
And in a country where you dohave um significantly more than
Japan, right, great, greaterinequality.
Um this is going to be achallenge for um for how we move
forward.
Of course, this is not the firsttime that the US has experienced
(54:03):
this kind of like economicdislocation.
We were talking about steelearlier, right?
Go to steel, formersteel-producing parts of the US
today, right?
Um a lot of these are long-termeconomically depressed
communities.
Um what is AI going to do forthe average person in the
(54:26):
context, as you mentioned, thatkind of that broad American
dream that you had kind ofdescribed so eloquently earlier,
right?
Um, well, that that reallyremains to be seen.
And in order to continue tomobilize this kind of consistent
support, um one would need toaddress that.
And this is what's interesting.
Um there are, despite kind of atthe kind of that you mentioned
(54:48):
that we we tend to focus onleaders in the world today,
despite kind of the Trumpadministration's um reflexive
distrust of regulation in thissector, um, at the state level
in the US, you had um you have awhole host of initiatives,
(55:08):
right, to kind of mitigate riskor to do things as simple as,
well, um, if AI is interactingwith a customer, then it should
be kind of disclosed, right,that you are you are interacting
kind of with with AI.
At that level, um there seems tobe a growing kind of consensus
(55:29):
that this type of kind ofgovernance uh at is is
necessary.
And and interestingly, whenthere was an attempt to kind of
put this into um uh anotherpiece of legislation, um, there
were opponents from DonaldTrump's own party, right?
Who were like, well, wait asecond, now this is this
(55:51):
something that would helppeople, right?
Um and and so there there is agrowing conversation about this.
And it's something thatdefinitely um will kind of
increase, right?
It's only it it it there's onlygonna be more people talking
about this as the real effectsof AI begin to be felt more
(56:13):
broadly in in the economy ineconomy.
SPEAKER_01 (56:17):
We're we're big on AI literacy, right?
But your mi minimum viableliteracy.
What's the minimum viableliteracy that you need in AI in
order to be a contributingmember of society or to thrive?
This is something that we often.
SPEAKER_00 (56:30):
Okay, I would say there's uh so one thing that I
that that in conversations umthat sometimes is surprising,
and this gets back a little bitto the beginning of our
conversation about kind ofsocietal breakdowns or societal
transformation, is that um forusers with very little literacy,
(56:50):
meaning that this is justsomething else that they've
downloaded on their phone, justanother app, right?
Um there there isn't anunderstanding of how their data
that they're sharing is beingused, or kind of what this or
how this model thinks.
(57:11):
Right.
SPEAKER_01 (57:12):
Well, either end, right?
There's a there's a lack ofunderstanding as to how good it
can be or how risky it can be.
It's just like on both ends ofit.
SPEAKER_00 (57:18):
Yes, exactly.
So it's like kind of this ideaof just um you're using it for
relatively kind of simple andstraightforward tasks.
You don't understand itspotential, but then you're
putting this likedisproportionate trust into
something that you don't reallyunderstand, but you're not
really using it the right way.
SPEAKER_02 (57:37):
Either, yeah.
SPEAKER_00 (57:38):
Right?
Because you don't understandthat, well it's not going to
really be able to help you if ithasn't been trained on any kind
of data that that is relevant towhatever you are are talking
about or whatever kind of uhconversation that you're seeking
to elicit from it.
(57:59):
And so um I think justunderstanding um and I think
this kind of lifting the veil uma little bit um without getting
too technical behind um LLMs,for example.
And I think this is actuallyanother element, minimum level
of of AI literacy is peopletend, when you say AI, to
(58:23):
immediately think about largelanguage models.
SPEAKER_01 (58:26):
Right.
Only.
SPEAKER_00 (58:27):
Only, exclusively.
SPEAKER_01 (58:28):
Yeah.
SPEAKER_00 (58:28):
Without understanding kind of larger
ecosystem, like my thing aboutcomputer vision or kind of all
the other kind of elements.
And and then when people arelike, oh well, um I don't really
need to worry about AI because Iasked Chat GPT to create a map
of my country and it put thecities in the wrong place.
Oh, this is so stupid.
(58:49):
Stupid.
Right.
So why are we even talking aboutsecurity, or why are we having
all these other kind ofconversations about it?
Um that in itself is a a markerof somebody who's just being
exposed, like trying to learn,but doesn't really understand
the technology, but then comingto this like conclusion, this
broad sweeping conclusion thatwe don't need to worry about.
SPEAKER_01 (59:10):
Yeah, it's like you know, go back to either the late
90s or early 2000s and sayingthis internet thing is so
stupid, it takes so long to youknow dial up, for example, not
seeing that you know thepotential.
Um you want to do a quick check?
Okay.
Uh not so, for example, with theinternet and then not not
seeing, for example, thepotential.
Um what I would like to umclose, uh start clo c closing up
(59:34):
with is reflecting on kind ofwhere you have come from, which
is the idea of internationalrelations and seeing seeing
societies rebuild and rebuild.
And sometimes obviously there'syou know horrific things that a
society can go through, butwe've we've seen time and time
again in humanity that evenafter a fundamental shift, um,
we've seen societies build backand thrive.
(59:57):
Um and we've even seen somesocieties miss entire blocks of
what we would say is kind oflike l- like um linear
development.
So, for example, many countriesin Africa or even many countries
in Southeast Asia didn't do dialup internet, they have no
memories of what it's like to goto bed and wait for two songs to
download on a internet that wasnever their life.
They went from nothing torunning an entire business
(01:00:19):
through WhatsApp, right?
Right, yeah.
So super interesting.
So not everyone has to have thesame journey.
Right.
And so I'd love to think fromwhat we have seen in in the past
and what you've seen societiesgo through as they've
re-rebuilt, what does a realthriving society have in common?
As in, they've gone throughmassive change and they've now
set themselves up for you know areal element of success, um,
(01:00:42):
whatever that might look like.
W what are some lessons that wemight learn from and then apply
for obviously some of thechanges that society is is
really going through now?
SPEAKER_00 (01:00:51):
Okay, so um broadly speaking, I would say that one
thing that you do see in commonin places or in in societies
that have experiencedsignificant traumatic events or
dislocations such as conflict isthe complete breakdown of trust.
(01:01:11):
Not just trust between kind of athe person and a state or some
governing authority that mighthave abused that trust through
committing serious human rightsabuses or doing other things,
um, but also at a veryinterpersonal level as well, not
(01:01:32):
trusting your neighbors.
Aaron Powell So I lived in EastGermany for a while.
I didn't know.
SPEAKER_01 (01:01:35):
So I lived in but I lived in Berlin for like 10
years.
SPEAKER_00 (01:01:38):
Yes, right, exactly.
And so this will this this thisthis will resonate.
SPEAKER_01 (01:01:40):
That was like even you felt it in the in the
staircases, like still.
SPEAKER_00 (01:01:44):
Yes, exactly.
It's a who is this person?
Why are they talking to me?
Yeah, right.
This kind of immediate kind ofsense, or what do they want from
me, or how are they going tokind of leverage whatever this
is?
Or all of these um are symptomsof a breakdown kind of in trust.
And one of the things that whenkind of thinking about
(01:02:08):
transitions, um that thatsometimes is overlooked is that
that element of trust needs tobe rebuilt first, right?
So you can create a kind ofgovernmental commission or
public inquiry, or you can doall of these things, right?
But if but if nobody trusts thestate, nobody wants to talk to
the state, right?
Then that's going to limit theeffectiveness of whatever you're
(01:02:31):
trying to do at kind of a macropolicy level.
And here, right, getting back tolike these, these, these
changes, and you're mentioningthat different countries have
different experiences.
And this also highlights umdifferent levels of trust in
technology, right?
So you might think about youlived in Germany.
Why are Germans so distrustfulof technology, right?
(01:02:54):
Um why why are um conversationsabout, for example, data privacy
or kind of cameras in publicplaces so much more visceral in
a German context than they wouldbe in Australia or anywhere in
Southeast Asia, right?
Um where people might just belike, well, yeah, it's just kind
(01:03:16):
of a public space, so why whynot?
Um this there there aredifferent um equilibrium points,
right, of of what cultures andsocieties will will tolerate,
and it's not one size fits all,right?
So in the conversation on AIgovernance and also AI futures,
(01:03:39):
right, the um while the world isinterconnected and while these
ecosystems are interdependent,kind of what AI governance might
look like in one place can onthe margins and will very likely
differ another place, meaningthat this could this could mean
that there's a greatertolerance, whereas some
(01:04:00):
countries might see datacenters, even if they're fully
privately owned andforeign-owned, as some sort of
marker of linear development, asyou as you mentioned, right?
Or as a way to skip different umdevelopmental kind of milestones
that we tend to associate.
(01:04:21):
And and therefore, right, we arein a world where lots of
conversations are focused ondigital sovereignty.
There are lots of differentvisions of digital sovereignty,
we are very interconnected, butum that landscape is shifting
and changing constantly beneathour feet, which means we need to
(01:04:44):
have just a better awareness ofthis, right?
And that awareness will help usbetter understand policy shifts
that are happening and hopefullywill allow for um changes when
they do happen to not liketotally come out in blue and
also um erode our trust in umthese processes.
(01:05:05):
Although, of course, um thereare that that there probably
will be some developments thatwill, right, and and should lead
us to question our trust.
SPEAKER_01 (01:05:16):
I think trust is actually um perhaps the word
that was connecting perhapseverything.
So, for example, even if we'rekind of reflecting on the um
like the American dream, I guessone of the the reasons, you
know, if you're thinking aboutshall I participate in this new
system or in this new economy ornot, it's because people trust
their participation is going tomean something, it's going to
result in something.
(01:05:37):
Um so, all right, so thinkingabout an AI economy and an AI
economy to to thrive, I mean,one thing, obviously, I mean, at
DTJ we always are talking aboutliteracy.
Um, however, I think we're goingto have to look at trust now,
um, perhaps a little bit moredeeply.
I think it's always been there,but I don't know if we've kind
of like put a stamp on it.
Like putting words to it, kindof like as you have today.
SPEAKER_00 (01:05:58):
It can be interesting, also, just like um
in conversations.
Just like are there are and andare there different like models
that people just intrinsicallyfor no good reason trust more
than another one?
And then ask why that is likeand and also maybe um what at
what level would your trust likestop?
Right.
Right.
Right.
SPEAKER_01 (01:06:17):
So like um when is okay, when is not okay.
Yeah.
SPEAKER_00 (01:06:20):
Right, exactly.
Like recently um had aconversation with a commercial
airline pilot um who who workson certifying uh or who could
does research on certificationof kind of um of digital flight
instruments, right?
And of course you can't havekind of autonomous instruments
right now in in civil aviation.
But the question is is that wellit really boils down to is just
(01:06:42):
will people trust getting on aplane one day without human
pilots?
Right?
Probably not for for near timewhen you would think about like
it's it's there's all thesesocial interactions that happen
during flight or all thesethings that might happen.
For example a medical emergencyor a disruptive passenger um
whereas if you have like a humanbeing who is the captain right
(01:07:04):
um who makes decisions onwhether to land the plane or
not.
SPEAKER_01 (01:07:07):
That's interesting without opening up like a whole
nother kind of worms I thinkthat's so interesting because
the way in which societies alsoapproached us is really
interesting.
So for example I remember we youknow we had a hypothetical
situation we're running um witha few policymakers, but they
were not just Japanesepolicymakers, they were also
from abroad.
And it was interesting most ofthe participants from either
Southeast Asia or Africa saidthat they wouldn't use AI for
(01:07:30):
invoices.
And I was like that'sinteresting.
I wonder why why would you not?
Right.
And then their reasoning was ifsomething goes wrong who who do
I who's responsible and Ithought oh that would never kind
of like came across my kind ofvalue stack.
Right.
That didn't kind of I was likeyes invoices AI yes thank you.
SPEAKER_00 (01:07:47):
Yes but they were like no who who's gonna
apologize who's gonna apologizethat was a thing who's going to
say sorry and I thought wow sointeresting yes yeah so uh in in
that sense right like we don'thave until you're able to like
right think think in those termsright and and also different as
I mentioned different culturalperspectives on um what trust
(01:08:10):
means how trust is achieved howtrust can be lost um all of
these these vary.
And I think this is one of thechallenges you might think about
in terms of the the broaderconversation about a technology
that tends to focus on um learnfrom data at an aggregate level
(01:08:30):
might miss some of these nuancesand therefore you might find
something well it's kind of verywell suited for like an American
consumer or an Australianconsumer but then is is very
distrusted in a different partof the world, even though
nothing is different in terms ofhow it operates.
SPEAKER_01 (01:08:48):
Chris thank you so much for joining I feel like
this has to be a part one I feelI feel part two.
Thank you thank you so much.
I think certainly um I mean myuh reflections of of today
certainly we can use the thepast to uh inform but definitely
not form for tomorrow we have tothink very seriously about where
we are um and then and and thenagain I mean as society we've
(01:09:10):
kind of like done this all thetime which is like just try to
be better right just try to bebetter than we were yesterday
and to take that um to thefuture but then to also very
deeply think about um your ownsystems of trust what does that
look like um and then how toapply that in your own teams in
your own business because as youmentioned I mean if we look
about societies that arethriving societies they do trust
(01:09:31):
their neighbors.
Right.
And so what kind of steps canyou take in your own environment
to literally trust yourneighbor?
SPEAKER_00 (01:09:39):
Absolutely right so I think these are this is
definitely a conversation we weshould continue at at some
point.
Thank you very soon.
Thank you thanks so much thankyou.
SPEAKER_01 (01:09:48):
The Business Carakue Podcast is brought to you by DTJ
Design Think in Japan.
In a world where you can buildanything with AI now knowing
what to build is really thework.
And so there's two ways that youcan come and work with us at
DTJ.
One is you can come and join ouracademy and build a real level
of executive fluency.
Work with us in our AI buildswhere in 12 weeks we can have a
custom AI solution for you upand running for your business.
(01:10:10):
But the number one thing iseveryone go and download our
2026 signals report it has someof the most meaningful trends in
AI that you'll be seeing in2026.
In in societies that have experienced significant
traumatic events or dislocationssuch as conflict is the complete
breakdown of trust.
Not just trust between kind of athe person and a state or some
governing authority that mighthave abused that trust through
(00:23):
committing serious human rightsabuses or doing other things.
But also at a very interpersonallevel as well.
SPEAKER_01 (00:31):
Today we're joined by Dr.
Christopher Lamont.
Dr.
Le Mont is a professor ofinternational relations at Tokyo
University, Deputy Head ofProgram for AI and Global
Governance at the GlobalGovernance Institute in
Brussels, and in his free time,the co-host of the Age of AI
podcast.
Chris is also my friend, so Idid my best to call him Dr.
Lamont, but I think I actuallyfailed miserably and I called
him Chris most of the time.
(00:53):
His research career began aroundtransitional justice,
international criminal law, andhow societies rebuild after
institutional collapse.
This conversation really doescenter around trust.
What destroys it, how societiesrebuild, and why it may be very
much the most consequential wordin our AI era.
I always so much enjoy uhreading your comments and your
(01:14):
thoughts.
So please go ahead and reach outwith your thoughts about this
episode.
Okay, everyone, let's jump in.
So Christopher Lamont, welcometo the Business Charity Podcast
for our human civic AI ClubSeries.
Welcome.
Thank you.
SPEAKER_00 (01:25):
It's a pleasure to be here.
SPEAKER_01 (01:26):
So uh I wanted to share a little bit about why I
was really interested in yourspecific views and your lens of
the world.
The as I was preparing for thisum for our interview, uh my um
AI and I decided to call you theLamont lens.
This was the Lamont lens.
SPEAKER_00 (01:42):
Interesting.
Oh wow.
Okay.
Well hopefully I live up to thatlens, or the lens is one that
makes things clearer, I suppose.
SPEAKER_01 (01:49):
So the Lamont lens.
So let me share a little bitabout as I was kind of chatting
and saying this is why I'mreally interested um for um
Chris to come on.
So um for those of you that arethinking why I'm not calling Dr.
Chris Filmont Dr.
Le Mont, because Chris and Ialso are friends.
So I can't I'm allowed to callyou Chris.
SPEAKER_00 (02:04):
Yes, so I prefer to be called Chris also.
It's it's it's kind of yeah.
SPEAKER_01 (02:07):
I'm really interested when it comes to AI.
So, as you know at DCJ, we'rehuman-centered AI, which means
don't begin begin with thetechnology, begin with the
human.
Right.
And I'm gonna share with you astory that I heard recently that
was told to me recently, andthen kind of how that um then
informs kind of our thinking.
So one was I was chatting withone of the professors from Grips
University, so at the GraduateUniversity for Policy Studies
(02:30):
here uh in Japan, and we werechatting about AI and
cybersecurity, security.
And he said, I don't know whyeveryone is kind of jumping up
and down about security now.
And I was like, that's reallyinteresting because I think AI
and security is probably like areally big issue.
Right.
And I said, Do you want toexpand more?
Like, what do you mean by that?
And he said, a hundred yearsago, when you wanted to send a
(02:52):
letter, you would send somebodyand they would jump on a boat
and go somewhere, but thisperson could equally be
assassinated or murdered orkidnapped.
Right.
Yeah.
And not that any of this isfunny, but what it kind of
highlighted was that yes, we arein a new time, being with uh,
you know, the AI age, but theissues of security, the issues
(03:14):
of unforeseen circumstances havealways been with us.
And I'd really like to beginwith that with that um as how
it's connected to kind of you,is that I've always found it so
interesting how you have seensocieties go through immense
change.
Immense change.
(03:34):
So, you know, we can start niceand light and talk about how,
you know, you've studied, youknow, war crimes in the Balkans
and things like that.
I mean, we're not three minutesinto the episode, but I, you
know, we can talk about warcrimes.
And I'm really interested from abefore we jump into the
technology and the people, I'dreally love to hear your
thoughts or your views on, evenbefore jumping to AI, when
(03:55):
they're society and they'regoing through immense change,
what are some of the things thatusually lead up to those kinds
of immense changes?
Is there kind of like then anabsolute like switch where it's
like before and after thismoment type thing?
And then the rebuilding.
So with that, um I thoughtthere's no one better than you
to come and talk to us throughthe society evolution when it
(04:20):
comes through, obviously, umwhat you know, and then
comparing that to where we arewith AI.
So that is my thinking around umour episode for today.
SPEAKER_00 (04:28):
Okay, well, lots of questions to um to respond to
coming out of that.
I think that what you kind ofended with, and that's the
question of like, okay, the thelike transformation as a switch,
the before and after.
Um in most types of change,right, processes or society or
(04:48):
where societies have experienceduh some sort of massive um
disrupture, it actually isdifficult to identify that like
one point in time, right?
So in countries that haveexperienced, for example, civil
conflict, right?
So even a question like when didthat conflict start?
Was it when kind of the shootingstarted, or was it when there
(05:08):
was a political breakdown, orwas it when there was some sort
of incident that happenedpreceding that?
And then also when does it end,right?
Is it when kind of a peaceagreement is signed, but there's
still high levels of societalkind of suspicion, distress,
walls separating communities,for example?
Um and and that fluidity interms of how we experience
(05:32):
change cuts against how we talkabout change, because we tend to
assume that there's like abeginning and end, right, um, to
any major kind of historicalepisode, and then apply that to
technology and what youmentioned early on is what makes
AI different from other securitychallenges or other types of
governance challenges that we'veencountered in the past.
(05:54):
Right?
So if you think about AI andsurveillance, right, that states
have had the capacity to survey.
And um, often you could usepeople, you would use people to
do that decades ago to survey.
But there are a coupledifferences that have that have
kind of been creeping up on usthat we might not see without
(06:15):
kind of perfect hindsight.
And and one, and this issomething that gets to the kind
of broader governanceconversation, is just how much
data is out there about all ofus, right?
That we are, of course, notnecessarily kind of live
streaming our lives as such, butas a figure of speech, right, in
(06:36):
a way we are, right?
All this data is out there aboutus.
And then you look at that andyou aggregate that.
And what you can do with thattypes of that type of
information, what that enablesin terms of um not just the
private sector, but also thepublic sector in terms of
governance for good and bad, issomething that um previous
(06:59):
generations did not have accessto when thinking about
governance or thinking aboutsociety.
And I think that's that's that'san important difference.
And so if you think about alsothis question of security, what
type of security vulnerabilitiesdo we have?
Well, um if an individual wassomehow compromised, who was
(07:22):
carrying a message in the past,right, to go back to the the
analogy, um you would, if you ifyou identified that, be able to
limit the exposure you have.
But if your kind of data systemsare compromised, right, and you
don't know it, right?
Or even if you do know it, butyou don't know how it's being or
(07:43):
where it's being or what to doabout it, um the scale of the
challenge and the potential harmthat it can bring about has
grown significantly.
Right.
And so I think that part of thisseemingly like, oh, everybody
now is talking about securitywhen it comes to AI or thinking
(08:05):
about some of the kind ofcybersecurity implications.
I think that's natural,particularly when it comes to
developments like agentic AI.
But I also think, yes, there is,it's helpful to look to the past
for parallels and see kind ofhow we have addressed similar
challenges, but to be carefulnot to apply the tools of the
(08:29):
past to the challenges andproblems that we are facing
today, right?
Because the um the nature of, orso the society that we're
seeking to govern has changedhow it communicates, right?
I'll use a very kind of like ifwe go back 100 years, for
example, um, I also teach on akind of international program at
(08:51):
a university.
And one of the things that'sinteresting to observe, or just
like you think about, is that,well, a hundred years ago, for
example, you could not haveinternational programs the way
you do today, where with peoplefrom all these different
continents coming together,right?
Like if you were um going totravel intercontinentally, it
(09:12):
would probably be like a once ina life, or maybe twice in a
life, or or maybe if you'resuper, super wealthy, right?
A few times, right?
But these the the these thesejourneys take a lot of time.
Um and then once you gosomeplace else, you need to like
learn the language, adapt life,right?
Um it it's it's a it's it's alifetime commitment.
Um but in the 2020s, of course,um, it's something that for
(09:36):
young people in a lot ofdifferent parts of the world,
that's kind of natural.
Well, think about like whichcountry do you want to study in,
right?
Um so on the one hand, right,mobility has increased, but also
if you think about the world inwhich they live.
If you come to a country likeJapan, for example, you're um
you remain connected on socialmedia with all your friends back
home, right?
You can have friends also makenew friends, like virtually who
(10:00):
are on who are on differentcontinents.
Um and also you can remain inyour like own media environment
from back home.
You can kind of keep watchingyour favorite series, you can
keep doing all this, right?
And so the the world, right, thethe the the the kind of the
experiences that people havetoday are much more
(10:23):
interconnected than they were inthe past.
And if you think about that interms of AI and governance,
right, that that's also true,right?
Because what what we're seeingat a very individual level, your
own kind of personalexperiences, how you are
experiencing the world, has alsochanged.
(10:45):
For example, business, right?
Think about supply chains, thinkabout all these other elements
of interconnectivity.
And so it's an interesting timeto be talking about that in
2026, of course, because itseems that, well, um, a lot of
those um institutions thatfacilitated that
interconnectedness are underincreasing stress.
(11:07):
But there's there's no way, Imean, right now, right, to think
about it to kind of pull back,right, the um turn back the
clock when it comes to thesetypes of societal and individual
level changes and how technologyhas changed our lives, right?
So the the policy responses tothat would reflect the world we
(11:30):
live in today, not 100 yearsago.
SPEAKER_01 (11:32):
That's so interesting.
I think the the the point that Iwanted um dive a little bit more
into is the uh when youmentioned that the tools of
yesterday can certainly forinform, but shouldn't
necessarily form our approach totomorrow, right?
So use them as for aninformative rather than a
formative um uh approach.
(11:52):
And I think this is interestingand almost an opportunity to
maybe even um break down or evendefine the idea of sovereignty
and digital sovereignty.
So I remember, so I'm from anisland, um, even though I have a
very odd accent, it's still anAustralian accent to some
extent.
Um I remember my grandmothertelling me um during the Second
(12:16):
World War, they in the papers,they basically told, sort of
they being the government ofAustralia, told everybody, don't
worry, the war will not reachour borders because we are an
island and the Japanese havebicycles.
And how are they gonna get fromJapan to here?
(12:36):
Honestly, there was like you canthere was a front page cover
with a Japanese person on abicycle and the Australians on
land kind of looking safe andkind of laughing.
SPEAKER_00 (12:48):
Right.
SPEAKER_01 (12:48):
And then I said, Well, how did that end up?
She goes, Well, Darwin wasbombed, you know.
SPEAKER_00 (12:52):
Right.
Aviation.
SPEAKER_01 (12:54):
Aviation.
And so the idea that asovereignty used to be this is
my land, you're in your land,that that's very much not what
sovereign means anymore.
SPEAKER_00 (13:06):
Yes.
Um so this is a very goodquestion, right?
So we're talking aboutsovereignty itself.
I'd like to focus on, though, II think you mentioned digital
sovereignty there, because Ithink that's that's something
that is increasingly talkedabout and um is something that
um a lot of states have begun toembrace.
(13:27):
Now there's sovereignty as weknew it in the the past century,
and there's and and how we seethe world is actually very much
informed by sovereignty, right?
You if you were to imagine, forexample, a world map, you would
imagine a world map with stateborders, right?
That's that's that's like you'veclosed your eyes, right?
Yeah, flags all the in in allthe countries.
(13:49):
And of course, kind of the thereality of geography is
different, but nonetheless,that's that's what informed our
how we see sovereignty.
It's it's territorial, it'slinked to kind of a population
and political authority, and allthese things were nicely
aligned.
And in this world that Idescribed that had emerged with
(14:12):
kind of this deepinterconnectedness and also
fluidity and mobility, um, whatsovereignty kind of means in
practice has has begun to shift.
And it's shifted so much nowthat policymakers are beginning
to seek to find, look for tools,right?
(14:34):
Like, okay, so um, if you thinkabout other kind of strategic
industries in the 20th century,like the steel industry, right?
So this was something that waskind of closely associated with
national power, it was somethingthat you could um subsidize,
prop up, and even the smallestof countries could have their
own steel mill and locallyproduced steel, right?
(14:56):
And so that mindset isinforming, unfortunately, how
some people think about digitalsovereignty.
So if we just, if we just buildthe data centers, right, or if
we um just have a nationalchampion, like a kind of a kind
of a foundational model, right,that is, that is developed
domestically within our country,right, we have our own digital
(15:20):
ecosystem that we can stick ourflag to and feel better about
ourselves.
But as you know, as somebody whokind of works in this in this
field, that it's actually muchmore complex than that.
If you think about, for example,where tech companies get
funding, right?
This this often crosses borders.
(15:41):
And if you also think about datacenters themselves, the question
of like, well, who owns the datacenter, right?
And you think about data centersthat are built in different
parts of the world as part ofthese initiatives to um increase
a state's data sovereignty,well, most of the know-how
operators kind of and also kindof the rules governing them come
(16:03):
from outside, right?
So you might have, on the onehand, um, increased a particular
capacity, but that doesn'treally fit your understanding of
sovereignty, but is being usedas an indicator to measure
sovereignty, right?
And so in this particularcontext, then the the the
(16:25):
traditional tools to bring aboutsovereignty or to achieve
sovereignty are tools that werehoned in an era where you were
governing something that wasless internet connected and less
complex.
And also when I sayinterconnectedness, I'm not just
(16:47):
talking about interconnectednessacross borders, but I'm saying
interconnectedness within the AIecosystem itself, right?
So if you think about, okay, youhave these different regulatory
instruments that are governingdifferent parts of it, like you
might be thinking about kind ofthe services aspect, or you
might be thinking about the kindof the kind of design aspect
(17:08):
with through the EU AI Act, orum you might be thinking about
like platforms.
Um all of these areinterconnected, meaning that if
you are thinking aboutregulating one particular
aspect, you might begin to feellike you're playing
whack-a-mole, right?
(17:28):
Um in terms of of governance.
And also, as we mentioned, likethe elements of of change,
right, that is taking place isone that it's it seems obvious
when you talk about it in aconversation like this, um, but
uh in terms of kind ofgroupthink, right?
(17:49):
When you have people in a roomthinking about, okay, so so what
can we do to achieve digitalsovereignty, they tend to fall
back on some of the same kind ofindicators that that I mentioned
before.
SPEAKER_01 (18:03):
And if we think about digital sovereignty as,
for example, sovereignty, maybewe don't even need the word
digital, like maybe it's justsovereignty in 2026.
If we think about traditionallylike sovereignty and what that
what that means, it's deeplyemotive to people.
If you were to take somebody'sflag and burn it versus fly it,
(18:26):
right?
There's people have deeplyemotional reactions to this.
But if you were to ask somebody,for example, how they feel about
the data center or or their owndata, for example, I um have not
seen such a natural humanresponse to that.
(18:47):
And I'm and I'm wondering, youknow, the symbols of flag, the
symbols of nation, who are you,who you are not?
And there's even almost um, youknow, many people who listen um
to our podcast are similar tous, so either global
professionals in Japan orJapanese professionals globally.
And you know, we always get thequestion, where are you from?
Where are you from from?
(19:08):
Right.
You know, this kind now, youknow, as you kind of mentioned,
there's this switch of like whendid change kind of happen?
When did this borderless youknow kind of change?
Well, uh perhaps there's alsoelements of the fact that you
could, you know, for example, myhusband is is is now is Mexican.
I have a an Australian-Mexicanchild.
So there's no the borders are,for example, now in many
(19:32):
circumstances are kind ofsurpassed.
And so I'm thinking it's quiteinteresting if we go back to
like the symbolism of a flag.
And I'm wondering, will we getto a point where people feel so
deeply passionate about theirdigital sovereignty as they do
kind of national sovereignty?
Um and if they did, like whatwould that what would that look
like?
SPEAKER_00 (19:53):
This is this is this is interesting.
I think this is this gets to oneof the reasons why this
conversation is beginning tohave this kind of outsized
presence.
Um because it resonates withpeople's sense of identity,
(20:14):
right?
So if you look at digitalsovereignty conversations in
different contexts, they meandifferent things, right?
So if you were to take someAmerican con uh understandings
of this term, um, of which thereare many, um, we can get into
that maybe a little bit later.
But but one would see this as,well, digital sovereignty means
(20:34):
that American tech firms shouldnot be subjected to foreign
regulatory oversight, right?
Um and and that's very emotive,right?
So it's talking about like,well, the these are US companies
and these are foreignregulators, right?
Which which creates kind of thisvery zero-sum image of kind of
digital sovereignty and and thedigital space.
(20:56):
Um for others, it's a it's adevelopmental goal, right?
It's it's well, um our countryis facing developmental
challenges, right?
You see there's this remarkablenew technology.
Does it have the potential tolevel the playing field?
Can countries skip certainstages of of development?
Um and therefore we won't ask alot of questions about well, we
(21:17):
don't data centers, right, whatthey're actually doing, right?
What the what kind of individualum data protections at at that
level, right?
And so I wanted to like justlike put a caveat to what I said
earlier, simply because um whileour experiences are much more
interconnected and while kind ofthe technological infrastructure
(21:38):
is much more interinterconnected, um, we still
live in a world of nationstates.
Yeah.
Right.
Um perhaps the most importantum, I guess, identifying marker
that people possess would betheir passport, right?
And be kind of the country ofthe from, and as you mentioned
also.
Um it's also a very emotive kindof Identity as well.
(22:02):
And you do have models ofdigital sovereignty where
countries have sought toessentially develop their own
ecosystems domestically, right?
And so China being kind of thebiggest example of that.
But of course, China with alarge population, and China with
kind of very top-down governancewhere national leaders establish
(22:23):
very ambitious targets when itcomes to, even before AI, other
digital goals.
And then also as a countrythat's seeking to export its
technologies through mechanismslike the digital Silk Road,
right, you can the the flag isvery visible, right, in that
(22:44):
sense.
But most countries in the worldare not the US or China, right?
Obviously.
And how do those countries thensee digital sovereignty, right?
So if you think aboutinterconnectedness, I think one
of the places where this is mostvisible, where you're seeing
(23:05):
conversations of governance, butalso challenges associated with
governance coming to a head isEurope, which is not too
surprising because the EuropeanUnion was an early mover in
terms of attempting to regulateartificial intelligence.
But you are seeing now in 2026kind of a um significant
(23:28):
distance between the ambitionsthat originally surrounded early
discussions around the EU AI Actto now the EU omnibus, right?
Which is arguably taking a stepback from some of the larger
ambitions that the EuropeanUnion once had when it came to
(23:52):
governance in this particularrealm.
In large part also becausedifferent EU member states now
see digital sovereigntydifferently.
So if we are to look to France,right, France is one that's
seeking create nationalchampionship.
And have, yeah.
And have, right, and havecreated national champions.
But but these national championsalso exist with this very very
(24:14):
kind of deeply interdependentkind of network of either
funding or kind of bringingtheir products to market, which
means that it's not like thekind of a French-owned steel
mill, right?
Um, in terms of actually beingan indicator or be being a kind
(24:40):
of sovereign entity, the way Iguess paradoxically, some of the
kind of Chinese counterpartsmight be in a system where you
have a top-down kind of statepicking um those um national
champions within the technologysector.
Um immense challenges, right?
(25:01):
And I think this is one of theinteresting um aspects about
this conversation is it's umit's very easy to move to kind
of almost caricature extremes ofsovereignty and interdependence,
like this kind of world where,okay, well, nation-states matter
less.
(25:23):
I would argue that very fewpeople in 2026 are seriously
making that claim thatnation-states matter less today,
in the 2020s, than they did inthe past, given the level of
geopolitical tension that youare seeing in the world today.
(25:44):
But the combination of kind oftechnological change and
societal change means that howwe address problems of
government-to-governmentrelations in the 2020s is going
to change, right?
And so I think this is one ofthe big questions in terms of
(26:08):
where this technological spacewill lead to.
Are we going to be in a worldwhere it's possible to have many
different digital sovereigns?
Right.
Um, or in a world where thereare maybe two digital
sovereigns, right?
And then you're kind of in thesedifferent camps between the two,
(26:32):
right?
And so like if you think aboutlike what does digital
sovereignty mean, um indifferent countries, it like I
said, it means different things,right?
So some countries might seedigital sovereignty as saying,
okay, well, let's um distanceourselves from the US tech
ecosystem, right?
Whereas others might have anopposite view.
(26:53):
Um and some are also thinkingabout kind of re um rethinking,
as you mentioned at the verybeginning, how we think of
sovereignty in a world where youhave a few very large big tech
players, right?
And interesting here, they'rethey're historical parallels,
right?
(27:13):
Um as well.
unknown (27:15):
Yeah.
SPEAKER_01 (27:16):
That is um also definitely terribly interesting.
I'd like to unpack a few thingsthere.
I think the first thing is, asyou mentioned, um the team there
seem there's two players, right?
So it's either the US or China.
Now, this is I I reflect a loton this, not being from the US
nor being from China.
(27:37):
Right.
Um and for a couple of reasons.
One is, does that just meanthere's like a haves and a haves
nots?
Like there's an AI haves and anAI have-nots, and that's kind of
like it.
Is that really all we have?
Um and then another interestingapproach is as you said, there's
like the US approach, which isalso quite culturally informed
in a way.
Um so, for example, if you takeum the foundation model of Grok,
(28:01):
now Grok's like flag in theground is truth.
We you, you know, the Grok's itlike the number one factor is
grog should be truth seeking.
Right.
Um and then, for example, youhave uh on the other hand, you
perhaps, for example, China'smany foundation models, one of
which, for example, is DeepSeek, and is an open or they've
(28:22):
taken a different differentapproach and had it be an open
model.
And so Malaysia, as of 2026, Imean things are always changing,
so but it's 2026 February.
Um Malaysia has adopted DeeksDeep Seek as their national
model, which I think is reallyinteresting.
And then does, for example, inthis moment, do you need to
have, you know, what what wouldthe theatrics of a of a an
(28:47):
American delegation to Malaysiabe, where it's like, oh coming
and yes, let's, you know, haveour good relationships and you
know, you're spending ice massand all that kind of stuff.
Um that's one thing.
But then if a country is using afoundation model of like another
country, for example, that youof which you uh don't see eye to
eye on many things, this is theninforming not only the way that
(29:09):
uh this entire country isoperating, but also thinking and
creating.
Um this would just the simpleadoption of a foundation model
would means more than a thousanddelegation visits to a country.
SPEAKER_00 (29:23):
Um yeah, and exactly, and and what you know
about a country.
Right.
I mean, because if you'rethinking about this in the
context of public services, umand you're thinking about, okay,
the need to um digitize allthese records that a state might
have, right, so as to make thesemodels useful.
(29:46):
And then you think about who hasaccess to that data, right?
Um you're in a world wherepotentially um the tech players
from either side know way moreabout not just what's happening
in your country, but in othercountries as well through access
(30:09):
to that data.
And so this is where if you'rethinking about um adoption and
diffusion of technology, thatChina has a slightly different
perspective here.
That if you look at therationale behind um DeepSeek and
(30:31):
making that open, is that thebet is that China can diffuse
its models faster and easierthan the US can, because the US
tech ecosystem is embedded inkind of a highly monetized
system in which you needcountries to buy into, and also
(30:53):
um you don't want these modelsto to open, right, because
they're proprietary and you haveinvested or somebody has
invested billions of dollars inin that.
And in some cases, and this iswhere if you think about what
conversations are happening, umthat that that could be
(31:16):
something that could be harmfulto disproportionately harmful to
US tech, right?
As there is a perception that umstates are leveraging types of
vulnerabilities that you havedescribed um in order to achieve
(31:37):
different policy outcomes,right?
And so there is a significantnervousness about this.
And this is where um so OscarGestrain, who I had a
conversation with um on kind ofdigital sovereignty, where he
likes to use the term um digitalautonomy.
SPEAKER_02 (31:56):
Right.
SPEAKER_00 (31:56):
So as opposed to sovereignty, because what people
are saying is digitalsovereignty, but when you push
them on it, they don't actuallymean sovereignty.
They're just using that becauseit's a it's it's a word that
everybody kind of understands.
But the the actual meaning ofsovereignty, as we talked about
earlier, is something that'svery difficult to apply in the
(32:19):
context of the technologies thatwe are talking about, and also
the type of technologicalcompetition that we're talking
about.
So if you think about autonomyas a state, that's actually what
you you want.
You want the ability toautonomously act, right?
And that's the feature kind ofof sovereignty that you're
(32:41):
really seeking to get at, notnecessarily putting your flag on
everything, right?
And then you're gonna startasking a different set of
questions.
And so um, as you mentioned,there's differences in culture
kind of between the US andChina, but there's also
differences of cultureinternally, right, um, within
states and their and theirfoundation models, right?
(33:02):
Even kind of between Grok andkind of OpenAI or Anthropic,
right, you're noticing umdifferences in terms of the ways
these models make sense of andrespond to and interact with
users.
And so if you're thinking aboutautonomy, you might be thinking
(33:22):
about kind of value alignment.
Now, of course, there arelimitations to that, and those
limitations are most visiblewhen you begin to um discuss
sanctions, right?
Um, and um these companiescannot divorce themselves from
the country that they are theyare based in, and whatever that
(33:44):
country's you know foreignpolicy's goals are, whether
that's like you're talking aboutBeijing or you're talking about
Washington, um, right.
Um and in that sense, right, Ithink that a lot of the
conversations about digitalsovereignty are actually a
response to that nervousnessthat um if we are reliant on um
(34:06):
a company that is flagged inanother state to provide kind of
very basic things that we needand that we cannot do without,
um what's to prevent that kindof country from leveraging that
dependency over us to get us todo things that we otherwise
(34:27):
don't or would never do, right?
Um and again, I think there whatyou're trying, what you're
looking at is this question ofautonomy.
And autonomy might actually comefrom an interdependent network,
but of a different constellationof countries, different
constellation of actors,different constellation of
companies, right?
(34:47):
Um that is not as narrow as thiskind of idea of absolute
sovereignty.
So you could have countries orcompanies in countries with kind
of similar sets of kind ofvalues, right?
That we might think about kindof data privacy as one.
Um, and then aligning to kind ofwork together um to um ensure
(35:13):
that kind of basic level of dataprivacy protection across the
board and kind of allow foreasier interoperability across
cross-borders.
But so perversely, converconversely, right?
Like um what thisinterdependence means is that we
it's it's not that countriesmatter less, right?
(35:34):
But it's that we um need to havea greater awareness of and a
greater understanding of who weare working with to provide us
with something that we need.
Right.
Um and what does that mean forour autonomy going forward,
whether you're a government or aprivate firm?
SPEAKER_01 (35:56):
That that's very interesting.
Um, as someone who's Australian,I can speak to like fake
autonomy.
So fake autonomy is what we havein Australia, which is like
well, yes, Australia, you're acountry, but if there's ever a
war, we're the Commonwealth.
And all of a sudden we're theCommonwealth.
(36:17):
No, Australia, you're your owncountry, so you need a visa to
work in the UK, but our royalspaces are all over your money.
So like also like a very oddkind of um situation.
Um, but just to kind ofunderstand how deeply valuable
this is, Eric Schmidt is a CEOor the ex-CEO of Google who's
deeply invested in the idea ofperhaps AI autonomy.
(36:39):
Um and that, you know, he he wasmentioning in a in a recent
keynote, you know, under nocircumstances can we can we risk
just a few billion dollars forlosing the race, losing the AI
race.
SPEAKER_02 (36:55):
Right.
SPEAKER_01 (36:56):
And so this is where you're saying, okay, as you
mentioned, the AI, the AIsystems of the US are deeply um
embedded in in financial models,as in how who's making money
from this, how we're makingmoney from this, what's the
business model.
But he was saying, just to likeunderstand how valuable this is,
he's like, who cares about a fewbillion dollars if we lose the
AI race?
(37:16):
And I thought this is superinteresting because I'm kind of
seeing a little bit of a curve.
So that we go from individual,the middle is kind of like a
little bit gray and very lots oforganizations, and then there's
the end user.
So I'll tell you a little bitwhat I mean.
So I find and perhaps this mighthave always been the case, um,
but for me personally, I feelthat um politics is incredibly
(37:37):
individual at the moment.
So for example, it's not the US,it's Trump, it's not Japan, it's
Takaichi, it's not France, it'sMacron.
Like they're individual people.
Then there's this kind of likegrey mass in the middle, which
is like uh Washington or Beijingor some kind of you know uh
committees that sit somewhere,and then there's like the end
(37:58):
person, me as a person.
So it's like, how do you connectwhat's going on from like that
individual like countryleadership?
How do you navigate the gray?
And then what does that thenmean to someone that's getting
up every day and just trying torun their business?
SPEAKER_00 (38:15):
Oh, this is a this is a great question because I
think you're you're touching onkind of a couple different
aspects here, right, um, thatthat I can go into.
And I think first you mentionedthe AI race, right, in this
framing.
It's like, okay, traditionally,if you want to kind of mobilize
investment in something, right,you need to present a business
(38:35):
model case that investors willinvest in.
And with the framing of the AIrace, right, you it's it's being
framed as, well, this is um thisis existential for the country.
Yes.
Right.
And this is this is this isinteresting because this gets
back to kind of Americanunderstandings of digital
sovereignty and sovereigntyitself, right?
That if if somehow in thisframing of the race, the US is
(39:00):
not out front, does not win it,right, then that's it, right,
for kind of for the US.
And so therefore, you need kindof an all-of-society kind of
response to this, where kind ofthe government mobilizes private
capital um into making sure thatthe US comes out in front.
(39:24):
This is actually not though,because this is interesting.
So you mentioned that that thatthis is this this language um
people might associate with, oh,this sounds like all kind of the
the kind of the tech techleaders who were surrounding
Trump and who attended hisinauguration, right?
This is this is kind of theirmessaging.
SPEAKER_01 (39:40):
Yes, exactly.
SPEAKER_00 (39:41):
The the the messaging.
But but the kind of the US iskind of the it was the Biden
administration that came outwith a AI strategy, right?
And interestingly, um in theforward framed the AI race using
the exact same language, andalso some of the exact same
people who you're talking aboutwere involved in the kind of the
crafting of that commission thatcame out of a Biden
(40:03):
administration presidentialcommission's report, right?
Um So there's there there hasbeen a significant amount of
continuity in terms of this kindof this framing of the AI race.
Um but you mentioned kind of theTrump administration is doing a
lot of things differently, notnecessarily here, but in terms
(40:28):
of AI governance, how itunderstands AI risks, kind of AI
safety institutes, deconstructalso kind of international
cooperation when it comes todigital technologies.
Uh one of the things that theprevious US administration was
supportive of was, for example,the UN Digital Compact, right,
which was a high watermark forinternational cooperation when
(40:50):
it comes to AI and sustainabledevelopment, and also some of
the more ambitious goals thatframed AI as a catalyst in kind
of a win-win technological racethat can decrease global
inequality rather than increaseglobal inequality.
And also in terms of AIgovernance, right more broadly,
the previous administration hada very different approach to
(41:12):
risk and safety and wasembedding the US into this
growing network of AI safetyinstitutes, and also was part of
a whole lot of conversationsthat also went beyond kind of
civil governance of AI tolooking at military applications
of AI and the such.
And so there the Trumpadministration has taken a very
(41:33):
different approach, right?
The US US policy is completelykind of reversed on these
issues.
And policy is less predictable,right?
The policy process of theprevious administration has not
been replicated, right?
Um that's kind of along-standing kind of policy
(41:54):
process that then means that,well, US policy could be
something one day, but then Iguess I don't know what the word
uh the the president truthssomething out.
Um and then something is isradically changed.
And this is interesting in asense that this is not just,
again, not just kind of theTrump administration in the US,
(42:16):
but this increasingly kind of umpersonalistic uh approach to
governance in work with worldleaders today, right?
So as mentioned kind of TakaiChi's election in in Japan,
right, a lot of focus on theprime minister as a person, but
then forgetting there's like acabinet office and there's a
political party and there'sministries, right, that are all
(42:38):
influential in these processes.
Um and then also if you thinkabout how kind of course she is
very different in in China,right, um, discussing um, but
but but there are all theseconstraints, right, regardless
of um what leader we are talkingabout.
And that's the gray area of ofthe process that you were
(43:00):
talking about.
And so for people like lookingat this from the outside, what
you'd want to do is look, okay,what do you identify as the
policy problem, or what does itwhat are leaders talking about
as the problem?
Um is a good indicator that willtell you a lot about like what
the subsequent policy processwill will look like, right?
And that's why I thinkconversations about
(43:21):
conversations about digitalsovereignty are very important,
because that's at the stage of,okay, so what's the problem?
Right, what what is it we aretrying to achieve, right?
And they're different ways ofdifferent visions of that, but
some of them might actually beharmful in the sense that, yes,
the the things that could havemade you more sovereign in the
past are actually going toincrease your vulnerabilities
(43:42):
and dependencies in the presenttoday.
And kind of this what what doesthis mean for the average
person?
Well, I think this is veryimportant.
I think this is something thatpeople don't think about
sometimes um enough.
Uh that that often When it comesto issues that we might think
(44:03):
are very important, for example,like data privacy or
cybersecurity, individual usersmight think, oh, well, what's
what's wrong?
So just to use like another kindof parallels, if you think about
like open open AI and ChatGPTs,kind of subscription models,
right?
People upload all sorts ofpotential, like personal
(44:26):
identifying information,potential propriety information
for all sorts of reasons, right?
And they don't think a lot aboutwhat that means for them.
And in fact, for most people,there's probably not going to be
an individual like hugerepercussions.
But if you think about this atan aggregate level, um well,
(44:49):
this provides one particularkind of firm with a whole lot of
data that otherwise they wouldnever have had had access to.
And you mentioned Malaysiaearlier and other countries are
thinking about, well, if youhave all that very kind of
cultural specific knowledgeabout our society, um then going
(45:16):
into training something that wehave no insight into or no
control over, that potentiallyone day if relations break down
between our countries could beused in all sorts of ways that
we're not thinking about, right,then that that is that is a
concern, right?
And I think that for those whoare thinking about these
(45:39):
conversations, to think aboutkind of well, there's different
levels of of individual um Iguess, stakes that are that are
at play in in this conversation.
So it's it's it's much moreimportant, right, than um
perhaps some other policydebates, right, that that you
(46:02):
might or might not be befollowing.
Um because of its knock-oneffects, and when I say these
knock-on effects, right, if youthink about um if you look at
policy responses, that mightactually make it more difficult
to work with collaborators indifferent countries.
(46:24):
Um a lot of these co-creationspaces on the internet, you
don't really know a lot aboutthe other people who are kind of
helping to refine something,right?
And so you might find yourselfin a situation where um you're
just somebody kind of in the USor Australia working on
something collaboratively in acollaborative space, but you've
(46:46):
been working with somebody who'skind of in China, right?
And then you might find that,oh, you've contributed to,
right, or you might find thatyou've fallen foul of some
domestic legislation that isaimed at enhancing your
country's own digitalsovereignty.
Now, what would the effect be?
Well, the effect would beprobably rather chilling, right,
(47:09):
on all of those creative spaces,like your your own people's
ability to engage in thosespaces.
Um people just like simply like,okay, well, I I don't want to
risk this, right?
Um, and tune out, but then thesespaces will still exist, right?
And so it's for those who arekind of involved in the cutting
(47:32):
edge of this um sector, right,or the sphere, these are really
important conversations becausethey might impact upon your
ability to um to go aboutcreating what it is you have
been creating um in a way thatyou might not have thought
(47:56):
about, right?
Um and it might also just likecreep up on you out of the blue,
right?
Like, oh my gosh, I didn'trealize that this was even
something that I shouldn't orsomebody should not be be doing.
And so that's why these broaderconversations are incredibly
valuable.
So I think in in terms of that,the gray space that to what to
look out for is well look at howthe problem is being formulated,
(48:18):
right?
How it's how it's beingarticulated.
Of course, you won't get insightinto like the back room, right?
But but that that will tell youa lot about the kind of the
direction of travel.
So like look at how differentcountries um talk about digital
sovereignty.
Yeah.
Right.
Yeah.
SPEAKER_01 (48:35):
And and if we think about um an A AI sovereignty or
even just AI or a country'sapproach to AI, um it requires
one fundamental thing, which isit requires the society to get
behind it.
Yes.
It requires every single personin the society to say, yes, this
is this is for us, we're goingto use it.
So it's not just you can developit, but people have to adopt it.
The actual adoption of thetechnology is also key.
(48:58):
And we've seen, you know, as aswe've kind of mentioned um in
this past a few in this podcasta few times, that you can
certainly have the pastinformed, but it can't certainly
form, you know, where we'regoing to tomorrow.
So I think the informative thingthat we can take from the past,
from getting people, you know,behind systems, is that um, you
know, what why did peoplebelieve in their country?
(49:18):
Let's take, for example,America.
Um, Americans love their flag,they love their flag, they love
the American dream because theysaw the results.
They saw if I do these things, Ithen get there, right?
This was a there was a verylinear path.
They achieved it, people aroundthem they achieved it, people on
(49:38):
the other side of the countryachieved it.
It became like this is a thingthat we can do together.
And I trust if I get involved inthis system, it'll take me
there.
Right.
Um, I think when it comes tothat, what would the digital
American kind of dream looklike, right?
And then you can replace Americafor whichever country you like.
Um, but how are you going to getyour society to adopt the
(50:01):
technology?
And then I think the resultsfactor, I think, is also quite
interesting because like theAmericans, it's like, you know,
yes, there is that kind of ragsto riches story that the
Americans love, and they've seenit multiple, multiple times.
Right.
Um and I think people haven'tseen that yet with AI.
Right.
They've seen big companies geteven more rich from through
(50:26):
investment from other bigcountry companies, right?
Right.
But we I don't think we've seenthe rags to riches story as much
yet that helps people believeand trust in a system that they
have in the past.
SPEAKER_00 (50:38):
Right.
This is interesting.
So there there are there are acouple elements to this
question.
And and one relates toessentially kind of the future
of work and in AI, right?
Like will anybody like me have ajob, right?
Um and and the anxieties thatthat creates.
So as I mentioned, right, thatthere's a large push to kind of
(50:59):
center AI in the broaderconversation of US technological
competition, right?
For example, the space race,right, where all the tremendous
amount of kind of governmentmoney was put in to um making
sure that while the US wasn'tthe first kind of in space, the
US would be the first on themoon, right?
(51:19):
Um and similarly, um you havethat mobilization here.
I think the problem of aquestion of adoption for a
country like the US is less of aum challenge.
In part because uh the USdominates kind of platforms that
can bring things to market,meaning that, for example, if
(51:40):
you were a country that waspushing your own foundational
model and you had adversarialrelations with the US and you're
a small country, you're notgonna be on the Apple store.
unknown (51:51):
Right.
SPEAKER_02 (51:52):
Right, right.
SPEAKER_00 (51:52):
Right.
So like a Cuban foundationalmodel, for example, um, if one
right would not would not end upthere, right?
Um and and because that kind ofbroader kind of platform
ecosystem is right now like soUS dominated that that that's
less of a a challenge.
(52:14):
But the second part is what youmentioned is going to be a
problem, right?
Um as people perceive, okay,well, um there's this AI
technological race, there's AIis creating fabulous amounts of
of wealth for those big playersin this industry.
(52:36):
But from the perspective of auniversity computer science
graduate in 2026, there are nojobs for me, right, um that
would have existed just fiveyears ago.
Right.
Um and from the perspective ofthose already working in um
(52:59):
these fields, um companies aregetting smaller in terms of
workforce, not larger, right?
I think, I mean, I'd have to goback and double check with this,
but even like these large datacenters that consume vast
amounts of electricity andprocess vast amounts of data, um
it's not like going to visit afactory, in a sense.
(53:19):
Like there are very few humanbeings who need to work there,
right?
So you're you're using a largeamount of land, right?
You're using a large amount ofresources, but comparatively,
you're not really creating jobsthrough bringing data centers,
for example, to your kind ofstate, if you're if you're in
the US.
Um, you're bringing differentkinds of investment, but but not
(53:42):
necessarily creating jobs andcreating wealth at the lower
end.
And in a country where you dohave um significantly more than
Japan, right, great, greaterinequality.
Um this is going to be achallenge for um for how we move
forward.
Of course, this is not the firsttime that the US has experienced
(54:03):
this kind of like economicdislocation.
We were talking about steelearlier, right?
Go to steel, formersteel-producing parts of the US
today, right?
Um a lot of these are long-termeconomically depressed
communities.
Um what is AI going to do forthe average person in the
(54:26):
context, as you mentioned, thatkind of that broad American
dream that you had kind ofdescribed so eloquently earlier,
right?
Um, well, that that reallyremains to be seen.
And in order to continue tomobilize this kind of consistent
support, um one would need toaddress that.
And this is what's interesting.
Um there are, despite kind of atthe kind of that you mentioned
(54:48):
that we we tend to focus onleaders in the world today,
despite kind of the Trumpadministration's um reflexive
distrust of regulation in thissector, um, at the state level
in the US, you had um you have awhole host of initiatives,
(55:08):
right, to kind of mitigate riskor to do things as simple as,
well, um, if AI is interactingwith a customer, then it should
be kind of disclosed, right,that you are you are interacting
kind of with with AI.
At that level, um there seems tobe a growing kind of consensus
(55:29):
that this type of kind ofgovernance uh at is is
necessary.
And and interestingly, whenthere was an attempt to kind of
put this into um uh anotherpiece of legislation, um, there
were opponents from DonaldTrump's own party, right?
Who were like, well, wait asecond, now this is this
(55:51):
something that would helppeople, right?
Um and and so there there is agrowing conversation about this.
And it's something thatdefinitely um will kind of
increase, right?
It's only it it it there's onlygonna be more people talking
about this as the real effectsof AI begin to be felt more
(56:13):
broadly in in the economy ineconomy.
SPEAKER_01 (56:17):
We're we're big on AI literacy, right?
But your mi minimum viableliteracy.
What's the minimum viableliteracy that you need in AI in
order to be a contributingmember of society or to thrive?
This is something that we often.
SPEAKER_00 (56:30):
Okay, I would say there's uh so one thing that I
that that in conversations umthat sometimes is surprising,
and this gets back a little bitto the beginning of our
conversation about kind ofsocietal breakdowns or societal
transformation, is that um forusers with very little literacy,
(56:50):
meaning that this is justsomething else that they've
downloaded on their phone, justanother app, right?
Um there there isn't anunderstanding of how their data
that they're sharing is beingused, or kind of what this or
how this model thinks.
(57:11):
Right.
SPEAKER_01 (57:12):
Well, either end, right?
There's a there's a lack ofunderstanding as to how good it
can be or how risky it can be.
It's just like on both ends ofit.
SPEAKER_00 (57:18):
Yes, exactly.
So it's like kind of this ideaof just um you're using it for
relatively kind of simple andstraightforward tasks.
You don't understand itspotential, but then you're
putting this likedisproportionate trust into
something that you don't reallyunderstand, but you're not
really using it the right way.
SPEAKER_02 (57:37):
Either, yeah.
SPEAKER_00 (57:38):
Right?
Because you don't understandthat, well it's not going to
really be able to help you if ithasn't been trained on any kind
of data that that is relevant towhatever you are are talking
about or whatever kind of uhconversation that you're seeking
to elicit from it.
(57:59):
And so um I think justunderstanding um and I think
this kind of lifting the veil uma little bit um without getting
too technical behind um LLMs,for example.
And I think this is actuallyanother element, minimum level
of of AI literacy is peopletend, when you say AI, to
(58:23):
immediately think about largelanguage models.
SPEAKER_01 (58:26):
Right.
Only.
SPEAKER_00 (58:27):
Only, exclusively.
SPEAKER_01 (58:28):
Yeah.
SPEAKER_00 (58:28):
Without understanding kind of larger
ecosystem, like my thing aboutcomputer vision or kind of all
the other kind of elements.
And and then when people arelike, oh well, um I don't really
need to worry about AI because Iasked Chat GPT to create a map
of my country and it put thecities in the wrong place.
Oh, this is so stupid.
(58:49):
Stupid.
Right.
So why are we even talking aboutsecurity, or why are we having
all these other kind ofconversations about it?
Um that in itself is a a markerof somebody who's just being
exposed, like trying to learn,but doesn't really understand
the technology, but then comingto this like conclusion, this
broad sweeping conclusion thatwe don't need to worry about.
SPEAKER_01 (59:10):
Yeah, it's like you know, go back to either the late
90s or early 2000s and sayingthis internet thing is so
stupid, it takes so long to youknow dial up, for example, not
seeing that you know thepotential.
Um you want to do a quick check?
Okay.
Uh not so, for example, with theinternet and then not not
seeing, for example, thepotential.
Um what I would like to umclose, uh start clo c closing up
(59:34):
with is reflecting on kind ofwhere you have come from, which
is the idea of internationalrelations and seeing seeing
societies rebuild and rebuild.
And sometimes obviously there'syou know horrific things that a
society can go through, butwe've we've seen time and time
again in humanity that evenafter a fundamental shift, um,
we've seen societies build backand thrive.
(59:57):
Um and we've even seen somesocieties miss entire blocks of
what we would say is kind oflike l- like um linear
development.
So, for example, many countriesin Africa or even many countries
in Southeast Asia didn't do dialup internet, they have no
memories of what it's like to goto bed and wait for two songs to
download on a internet that wasnever their life.
They went from nothing torunning an entire business
(01:00:19):
through WhatsApp, right?
Right, yeah.
So super interesting.
So not everyone has to have thesame journey.
Right.
And so I'd love to think fromwhat we have seen in in the past
and what you've seen societiesgo through as they've
re-rebuilt, what does a realthriving society have in common?
As in, they've gone throughmassive change and they've now
set themselves up for you know areal element of success, um,
(01:00:42):
whatever that might look like.
W what are some lessons that wemight learn from and then apply
for obviously some of thechanges that society is is
really going through now?
SPEAKER_00 (01:00:51):
Okay, so um broadly speaking, I would say that one
thing that you do see in commonin places or in in societies
that have experiencedsignificant traumatic events or
dislocations such as conflict isthe complete breakdown of trust.
(01:01:11):
Not just trust between kind of athe person and a state or some
governing authority that mighthave abused that trust through
committing serious human rightsabuses or doing other things,
um, but also at a veryinterpersonal level as well, not
(01:01:32):
trusting your neighbors.
Aaron Powell So I lived in EastGermany for a while.
I didn't know.
SPEAKER_01 (01:01:35):
So I lived in but I lived in Berlin for like 10
years.
SPEAKER_00 (01:01:38):
Yes, right, exactly.
And so this will this this thisthis will resonate.
SPEAKER_01 (01:01:40):
That was like even you felt it in the in the
staircases, like still.
SPEAKER_00 (01:01:44):
Yes, exactly.
It's a who is this person?
Why are they talking to me?
Yeah, right.
This kind of immediate kind ofsense, or what do they want from
me, or how are they going tokind of leverage whatever this
is?
Or all of these um are symptomsof a breakdown kind of in trust.
And one of the things that whenkind of thinking about
(01:02:08):
transitions, um that thatsometimes is overlooked is that
that element of trust needs tobe rebuilt first, right?
So you can create a kind ofgovernmental commission or
public inquiry, or you can doall of these things, right?
But if but if nobody trusts thestate, nobody wants to talk to
the state, right?
Then that's going to limit theeffectiveness of whatever you're
(01:02:31):
trying to do at kind of a macropolicy level.
And here, right, getting back tolike these, these, these
changes, and you're mentioningthat different countries have
different experiences.
And this also highlights umdifferent levels of trust in
technology, right?
So you might think about youlived in Germany.
Why are Germans so distrustfulof technology, right?
(01:02:54):
Um why why are um conversationsabout, for example, data privacy
or kind of cameras in publicplaces so much more visceral in
a German context than they wouldbe in Australia or anywhere in
Southeast Asia, right?
Um where people might just belike, well, yeah, it's just kind
(01:03:16):
of a public space, so why whynot?
Um this there there aredifferent um equilibrium points,
right, of of what cultures andsocieties will will tolerate,
and it's not one size fits all,right?
So in the conversation on AIgovernance and also AI futures,
(01:03:39):
right, the um while the world isinterconnected and while these
ecosystems are interdependent,kind of what AI governance might
look like in one place can onthe margins and will very likely
differ another place, meaningthat this could this could mean
that there's a greatertolerance, whereas some
(01:04:00):
countries might see datacenters, even if they're fully
privately owned andforeign-owned, as some sort of
marker of linear development, asyou as you mentioned, right?
Or as a way to skip different umdevelopmental kind of milestones
that we tend to associate.
(01:04:21):
And and therefore, right, we arein a world where lots of
conversations are focused ondigital sovereignty.
There are lots of differentvisions of digital sovereignty,
we are very interconnected, butum that landscape is shifting
and changing constantly beneathour feet, which means we need to
(01:04:44):
have just a better awareness ofthis, right?
And that awareness will help usbetter understand policy shifts
that are happening and hopefullywill allow for um changes when
they do happen to not liketotally come out in blue and
also um erode our trust in umthese processes.
(01:05:05):
Although, of course, um thereare that that there probably
will be some developments thatwill, right, and and should lead
us to question our trust.
SPEAKER_01 (01:05:16):
I think trust is actually um perhaps the word
that was connecting perhapseverything.
So, for example, even if we'rekind of reflecting on the um
like the American dream, I guessone of the the reasons, you
know, if you're thinking aboutshall I participate in this new
system or in this new economy ornot, it's because people trust
their participation is going tomean something, it's going to
result in something.
(01:05:37):
Um so, all right, so thinkingabout an AI economy and an AI
economy to to thrive, I mean,one thing, obviously, I mean, at
DTJ we always are talking aboutliteracy.
Um, however, I think we're goingto have to look at trust now,
um, perhaps a little bit moredeeply.
I think it's always been there,but I don't know if we've kind
of like put a stamp on it.
Like putting words to it, kindof like as you have today.
SPEAKER_00 (01:05:58):
It can be interesting, also, just like um
in conversations.
Just like are there are and andare there different like models
that people just intrinsicallyfor no good reason trust more
than another one?
And then ask why that is likeand and also maybe um what at
what level would your trust likestop?
Right.
Right.
Right.
SPEAKER_01 (01:06:17):
So like um when is okay, when is not okay.
Yeah.
SPEAKER_00 (01:06:20):
Right, exactly.
Like recently um had aconversation with a commercial
airline pilot um who who workson certifying uh or who could
does research on certificationof kind of um of digital flight
instruments, right?
And of course you can't havekind of autonomous instruments
right now in in civil aviation.
But the question is is that wellit really boils down to is just
(01:06:42):
will people trust getting on aplane one day without human
pilots?
Right?
Probably not for for near timewhen you would think about like
it's it's there's all thesesocial interactions that happen
during flight or all thesethings that might happen.
For example a medical emergencyor a disruptive passenger um
whereas if you have like a humanbeing who is the captain right
(01:07:04):
um who makes decisions onwhether to land the plane or
not.
SPEAKER_01 (01:07:07):
That's interesting without opening up like a whole
nother kind of worms I thinkthat's so interesting because
the way in which societies alsoapproached us is really
interesting.
So for example I remember we youknow we had a hypothetical
situation we're running um witha few policymakers, but they
were not just Japanesepolicymakers, they were also
from abroad.
And it was interesting most ofthe participants from either
Southeast Asia or Africa saidthat they wouldn't use AI for
(01:07:30):
invoices.
And I was like that'sinteresting.
I wonder why why would you not?
Right.
And then their reasoning was ifsomething goes wrong who who do
I who's responsible and Ithought oh that would never kind
of like came across my kind ofvalue stack.
Right.
That didn't kind of I was likeyes invoices AI yes thank you.
SPEAKER_00 (01:07:47):
Yes but they were like no who who's gonna
apologize who's gonna apologizethat was a thing who's going to
say sorry and I thought wow sointeresting yes yeah so uh in in
that sense right like we don'thave until you're able to like
right think think in those termsright and and also different as
I mentioned different culturalperspectives on um what trust
(01:08:10):
means how trust is achieved howtrust can be lost um all of
these these vary.
And I think this is one of thechallenges you might think about
in terms of the the broaderconversation about a technology
that tends to focus on um learnfrom data at an aggregate level
(01:08:30):
might miss some of these nuancesand therefore you might find
something well it's kind of verywell suited for like an American
consumer or an Australianconsumer but then is is very
distrusted in a different partof the world, even though
nothing is different in terms ofhow it operates.
SPEAKER_01 (01:08:48):
Chris thank you so much for joining I feel like
this has to be a part one I feelI feel part two.
Thank you thank you so much.
I think certainly um I mean myuh reflections of of today
certainly we can use the thepast to uh inform but definitely
not form for tomorrow we have tothink very seriously about where
we are um and then and and thenagain I mean as society we've
(01:09:10):
kind of like done this all thetime which is like just try to
be better right just try to bebetter than we were yesterday
and to take that um to thefuture but then to also very
deeply think about um your ownsystems of trust what does that
look like um and then how toapply that in your own teams in
your own business because as youmentioned I mean if we look
about societies that arethriving societies they do trust
(01:09:31):
their neighbors.
Right.
And so what kind of steps canyou take in your own environment
to literally trust yourneighbor?
SPEAKER_00 (01:09:39):
Absolutely right so I think these are this is
definitely a conversation we weshould continue at at some
point.
Thank you very soon.
Thank you thanks so much thankyou.
SPEAKER_01 (01:09:48):
The Business Carakue Podcast is brought to you by DTJ
Design Think in Japan.
In a world where you can buildanything with AI now knowing
what to build is really thework.
And so there's two ways that youcan come and work with us at
DTJ.
One is you can come and join ouracademy and build a real level
of executive fluency.
Work with us in our AI buildswhere in 12 weeks we can have a
custom AI solution for you upand running for your business.
(01:10:10):
But the number one thing iseveryone go and download our
2026 signals report it has someof the most meaningful trends in
AI that you'll be seeing in2026.
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.
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
The Breakfast Club
The World's Most Dangerous Morning Show, The Breakfast Club, With DJ Envy, Jess Hilarious, And Charlamagne Tha God!