July 24, 2025 • 26 mins
On this week’s In the City, hosts Allegra Stratton and Francine Lacqua are joined by James Kanagasooriam, chief research officer at Focaldata, to unpack how AI is reshaping employment. The conversation explores how AI’s ripple effects are hitting the white-collar sector in the UK, from shrinking graduate job pipelines to the potential collapse of middle-class tax bases.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:02):
Bloomberg Audio Studios, podcasts, radio news.
Speaker 2 (00:17):
Welcome to in the City. Each week we unpack a
story that's crucial to the world's financial capitals. I'm Francis Laqua.
Speaker 1 (00:23):
And I'm Allegra Stratton.
Speaker 3 (00:24):
So Fran.
Speaker 1 (00:25):
It's been going on for years, but it feels like
it's reached something of an apoget, this debate about artificial intelligence,
and it's of course the ability to deliver big productivity gains,
particularly here in the UK, but also the problems with it.
Speaker 2 (00:39):
Yeah, and we've we've even seen the government actually double
down on this commitment to AI. Recently the PM focused
the Cabinet meeting on how it could potentially be used
to improve government services. But us a living standards.
Speaker 1 (00:51):
Right and so for all of us it is unclear
how technology, this technology is going to reshape the economy.
But we've got report on Bloomberg that businesses are already
changing the way they hire. A new study from McKinsey
looked at how UK companies are reducing hiring for jobs
that they think will be impacted by AI. Postings for
(01:12):
jobs were actually already down are a crazy thirty eight
percent in industries that will be affected by AI, So
in our conversation today we're going to get into what
role AI will play is playing in solving the UK's
productivity puzzle guess, but also more immediately, how it is
already changing the UK's job landscape.
Speaker 2 (01:33):
Now joining us is James Kanagasorium. James's chief research officer
at the AI enabled research firm Focal Data, and has
been instrumental in implementing data driven decision making processes across
various sectors, including finance and tex so. James, thank you
so much for being.
Speaker 3 (01:49):
On the show.
Speaker 2 (01:50):
Welcome to the City of London, the city of the
City of London. Please mind the gap between the tram.
Speaker 3 (02:01):
The financial heart of the country the city the city.
Welcome to in the city, stand clear of the dawn.
Speaker 1 (02:16):
We are seeing a debate in America, particularly between big
big tech titans in different points of view, about whether
AI will demolish the white collar employment cohort. We've had
fort CEO say two weeks ago could be as many
as fifty percent of jobs lost for white collar workers
from automation. We've got the anthropic CEO saying he thinks
(02:39):
twenty five an increase of twenty percent unemployment you know,
these guys are kind of bidding to outbid each other
with their apocalyptic record rhetoric, but clearly there is there
is a problem. And then on the other hand, you've
got the boss of Nividia saying, guys, we're going to
create more jobs through AI. Tell us where you stand.
Speaker 3 (02:56):
I would stand more on the former camp, which is
that I can see the immediate upsides, and I can
see the fact that different countries might benefit from it differently.
And I think the UK is potentially not that badly positioned,
given it's kind of human capital stocks, the fact that
it has lots of businesses that are pretty adjacent to
AI or are in AI itself. But the reality is
(03:17):
that if you strip away all the rhetoric, firms that
hire white collar graduates are hiring fewer of those people,
and they're doing so in a degree that you would
typically associate with a mass recession. So again, the numbers
that you've talked about, whether it's twenty five, fifty percent,
those types of numbers in terms of cohorts of people
rolling on is what you'd associate with the mass financial crash.
(03:40):
But the idea that that's a kind of permanent shift
is pretty structural. Sorry to interrupt you.
Speaker 1 (03:47):
James mckimsey showed there was a thirty eight percent reduction
in what they call AI exposed jobs.
Speaker 3 (03:54):
Yeah, exactly, And it depends on who's getting augmented and
who's getting exposed. You know, I'm in one of those
jobs in terms of market research, but lucky enough to
work at a firm that is actually an AI market
research firm, so I'm one of the I work with
a few small band of cohort people who are getting
augmented to hopefully at the forefront of this role out
(04:17):
of this technology. But there are much wider societal impacts,
like what does it mean to literally have a milk
ground that is a third to half the size of
what it is before? What does that mean for the
tax base? What does that mean for the university education
sector and the number of people graduating with the degrees
to expect to go into white collar jobs. Like a
lot of politics and social issues are basically downstream of
(04:38):
the gap between people's expectations and then what is delivered
in terms of society, jobs and pay. And the third
thing is obviously, I mean, what's really interesting about this
is that this is clearly differentially affecting a group of
people that you would not necessarily associate with being left behind.
Right we're talking about graduates who I would have expected
(05:01):
to basically earn good to great salaries over the course
of their careers, who are probably living in cities who
could expect some kind of graduate premium. Then they're suddenly
walking into a job market where none of these things exist.
But they've already crystallized a lot of the downsides, so
they already have, for example, debt from a university. They
have already moved to cities that have high house prices.
Speaker 2 (05:24):
I speak to a lot of CEOs and I mean,
these are big CEOs that should one hundred percent understand
what AI means for their workforce, and frankly they're not.
I mean, you know, you'll have a coffee with them
and they say, I'm looking at AI training, but I
don't one hundred percent understand how it's going to impact
my organization.
Speaker 3 (05:40):
Yet.
Speaker 2 (05:40):
We also spoke to the Accenture CEO and she was saying,
they're training five hundred thousand staffers because a consulting work
in AI is increasing by so much. So do we
actually have a full understanding of the impact of AI
in the next three four five years and what sectors
will be most affected.
Speaker 3 (06:01):
Yeah, I think it's a really good question, like how
much of the analysis is evergreen. One thing that I
when I first started this kind of geographic analysis of
which jobs were more exposed to AI and therefore what
were the kind of community impacts. The thing that really
struck me in a UK context was the government's own
analysis of which jobs were AI exposed is basically the
(06:23):
opposite in twenty twenty three to its initial analysis in
twenty nineteen. In other words, the jobs, the geographies, the
education levels that people four were going to be automated
are not just slightly different, but they're almost the opposite.
You know. Back in twenty nineteen, the Department of Education,
I think, released a paper that basically indicated that a
(06:43):
lot of the areas that were going to be AI
exposed were the same areas that suffered from de industrialization.
That is now the opposite of what the analyses are saying,
which is actually the jobs of the AI exposed far
more white colar. But again these are very broad terms.
The analysis that a lot of these things are based
(07:04):
on is around emerging technologies, their technologies change, the mixed changes,
the applications change. You know. One of the bits of
feedback that I thought was very interesting to my substack
around it was a financier called James Wise, who's one
of the lead partners of Balderton, who invests a lot
in this area, and he said, look, okay, fine, you've
(07:26):
pointed out that white collar jobs are more vulnerable than
blue collar, but look at the advances in robotics. So
he was potentially painted an even a more challenging picture.
But I do think there's obviously going to be a
one off, cohort effect of people who are clearly going
to benefit, right, and we're going to hear a lot
from them, and we're going to hit that they're very
(07:48):
florid about how the ships might increase things like productivity
and efficiencies. But I'm taking a step back and thinking fundamentally,
this is a complete re engineering of the wider structure
of how societies work. We're potentially losing mass seams of
upper middle class citizens who very disproportionately form a massive
(08:08):
part of the tax space, for better or for worse,
and the fact that this is happening at such a
high cadence and pace means policymaking or whether it's corporate's
find it quite hard to distill. For me. The most
interesting thing really about all of this was I was
speaking to an ext the ex chairman of You Girl,
(08:29):
gentleman called Roger Parry. Quite interesting guy. You know, he's
settled many boards, including companies like Uber, and he was saying, look,
there's a degree of parallels to the initial rollout of
of the Internet, which was back in the day when
it was first, when it first happened. You often these
boards would often get the it guy whose job was
(08:49):
hardware to explain the Internet to boards. And maybe we're
a little bit in that phase where we're not quite
sure who the experts are, What is the information that
companies need to do, what are the amateurs? How is
that communicated to workforces?
Speaker 2 (09:03):
But James, it's not showing up in data yet, right.
If what you're explaining is true, then unemployment should go up.
It hasn't gone up yet, So is there going to
be Is it like a small drip feed where you
lose jobs you don't really understand why the industry is
changing and no one really realizes or is it going
to be like a recession where suddenly that's it. You
(09:24):
lay off thousands and thousands of staff in a company.
Speaker 3 (09:28):
So my job, I'm not not an economists. My job
isn't a forecast the labor market. What I can fundamentally see, though,
is the displacement and substitution that's happening is not offset
yet by the creation of new roles. But we can
see new roles right, we see them. I see them
in my own firm. You know, people who are concentrated
(09:51):
around customer success or people who are involved around sales,
or people that are used to basically take these technologies
and deploy them. People are getting augmented. There are new roles,
there is a degree in creativity, but the loss of
thirty to fifty percent of a white collar industry that
is mass displacement. You know, we should be thinking about
(10:12):
this in exactly the same terms as we think about
for example, when oil runs out of a particular area,
what does that look like? So the UK that context,
that would be north sea oil. We should be talking
about this is is like de industrialization equivalent, but on steroids.
Speaker 1 (10:27):
Do well do you see that, James?
Speaker 3 (10:29):
Do?
Speaker 1 (10:29):
What is what are those I mean Obviously it's sort
of difficult to dynamically predict what new jobs spring up
because of AI. But certainly that is Navidia CEO's position.
Isn't it that the jobs are going to be more
and they're going to be better. But as you say,
we're not yet completely take your point that you can
see in your own firm. But are they as many
jobs as the ones that are being displaced?
Speaker 2 (10:50):
Do you think?
Speaker 3 (10:51):
I'm skeptical, is what I would say, But jury is out.
We're just of the foothills of these pretty large ships
in both the labor market and how we understand work.
A lot of companies don't know how they would create
the new jobs that exist. The duo Lingo CEO right saying,
you know, we're not hiring anyone if we think they're
you know, exposed in any kind of way. I think
(11:12):
we're still getting to the stage where we're trying to
understand what are the jobs that could be expanded, what
are the ones that could be augmented. It's clearly not
going to be all negatives, and it's clearly not all
going to be positives from this aspect, but certainly from
what I can see in the most automatable industries. Like
take something really really basic, like a young investment banker
(11:34):
would spend a lot of their time building things like
discal cashlow models in Excel and pulling financials and doing
research into certain sectors. These are jobs that cannot just
be done with a few lines of code, but can
also then be done by a lot of these technologies.
And we're talking. Yeah, I was thinking about to my
first graduate job when I worked in the city, and
(11:55):
I was just thinking how few people would now be
needed to conduct that. I mean, there's a much wider
conversation to be had about skills and training. So how
do people come experts in an era where basically the
whole of society has this big red button that they
can push to basically push the automatically get along pushed someone.
Speaker 1 (12:20):
But that's something This is this is what question I
wanted to ask you. I mean, Neil focus and was
writing a couple of weekends the historian and economists writing
a couple of weekends about setting up a university, and
the university is not going to allow the use of AI.
So how much do you think in the years ahead,
Maybe not immediately, maybe he'll be an outlive for a while,
But how much is some of this going to be about?
The kind of refused nick mentality, which is, yes, AI
(12:43):
can do a whole lot of this, but at some point,
if we think we're looking at thirty to fifty percent unemployment,
we have to choose not to do.
Speaker 3 (12:49):
Some of this. So there's there's there's a lot of
different things baited into that question, so sorry. The first
is what does it know? It's brilliant. I love it,
So there's like what what? The first is what does
it do to people's learning frameworks MIT? I think have
Reas recently released a paper on what the consequences are
for using lllms and different aspects of AI, and basically
(13:12):
it leads to quite a drastic reduction in basically your computation.
In other words, we're basically leasing out bits of our
brains for technology to basically execute. I think we'll start
to see AI three labels like we do with food
with organic and that you might start to see kind
of things like two tier pricing systems where what people
(13:33):
are really trying to passe because when you pay for something,
you've obviously got cost based pricing. How does someone value that?
And it's the idea. Let's say you've got a consultancy
that comes in and does a project for you, and
you know that it's all kind of been done with
human minds versus once it's been augmented. Will people start
to think about that pricing differently. I don't know. Will
(13:54):
people be happy to pay the same half million pounds
for a consulting project if they know that it's basically
being conducted by by one human and ten bots and
one co pilot. Maybe we don't know. This is really
at the kind of frontier of kind of human psychology
and economics. But I do think you'll start to see
the refuse next, as you put at Allegra both in
(14:15):
terms of political parties, by giving away too much of
my yeah, I'm actually more I'm more optimistic than perhaps
my column indicated. I was giving the bare case about
how it's going to change our politics as opposed to
kind of a labor market pieces, which is that all
the people who have done quite well basically out of
(14:36):
a globalized financial system, expanded higher education, and slightly medium
to large immigration levels have done pretty There's a group
of people who've done pretty well. They all live in
certain parts of the UK and the US and other
English speaking countries. Those appeared to be the same geographies
that are most day are exposed, which produces a very
(14:57):
interesting politics of people who's previously very well starting to
feel much more threatened.
Speaker 2 (15:04):
As we say, I mean that feels like the Trumford already,
doesn't it.
Speaker 3 (15:09):
Well for me? One of the really interesting things of
politics over the last twenty five years has been this
decoupling of people's income household income and basically their levels
of education. Because back in say, for example, the early nineties,
really up until about twenty ten, they kind of walked
to lockstep. If you went to an area that had
(15:31):
lots of people who'd gone to university, it was probably richer,
and you know, there were some differences. But I think
what we've seen basically since higher education has been expanded
out to be fifty percent in a lot of countries,
but actually the promise them of that delivering higher incomes
has declined, and also the growth of secure blue collar
employment has meant this really really odd decoupling of the two.
(15:53):
And it was I think Thomas Packetti was really the
guy who's you know, he's the economist who is most
just work on inequality. But he pointed out, actually the
future that we may see is of secure blue collar workers,
which he called the merchant right making a march on society,
versus what he called the Brahman left, which is the
(16:15):
person who's got three degrees but no mortgage and very
very low salary. And I think we haven't quite thought
through what the consequences of that, because the politics of
those two groups are drastically different to classically left and
classically right.
Speaker 1 (16:28):
One question on the just the kind of the politics
of the left and right that we have right now.
You've worked in politics yourself, James, and listeners don't know.
But when I first came across you, as we say,
you were the guy who coined the term the red wall,
didn't you? It was really brilliant analysis and it has
come to live and live and then get smashed down
as well as a political entity. But how do you
(16:52):
think this government? Do you think this government, the labor
government is understanding the scale of what we've been talking about.
And then the country that I've been really about recently, Sweden,
they brought in a furlough type scheme to help people
skill up while in work in order to deal with
these sort of existential threats, but we can't really necessarily afford.
Speaker 3 (17:15):
That right now. Yeah, it's very interesting couple of points.
So the first is the whole political system, where is
the UK or US is currently dealing with we can
only describe as an omni crisis of the arrival of aire,
geopolitical instability, financial market instability. Particularly I think about the
UK and the potential for a debt crisis. There is
not necessarily the bandwidth or the expertise within government to
(17:39):
basically pass and understand these technologies and then work out
what should we do because they're basically trying to survive
day to day governments. As you will note, alegra live
Pole by Pole and this one has lost twenty points
in less than a year and a half. So are
they going to be doing the long term thinking about
what the labor market is going to be looking like
when effectively they've lost the most a man of votes
(18:00):
share for a majority elected government in living memory? Unlikely?
Do they have the MPs? Do they have the MPs
to understand? There are good MPs that sit on the
government benches who really understand this. You see them, you know,
I think there's that the Labor Growth Group. I think
there are a lot a lot of MPs there who
are very interested in the sector. I think MP's like
(18:21):
Jake Richard's conditionnarian who really understands some of these kind
of technologies, who are kind of thinking about what the
long term kind of consequences are. But let's, you know,
will the UK have the appetite for another furlough scheme depends.
I think a lot of the issues that we have
now is that the UK is quite vulnerable in terms
(18:41):
of its its debt pile, it's deficit, the problems it
has with its labor market and the number of people
who are out of work. I think the idea that
there's going to be another labor market collapse that's going
to require unilateral government support might win on voters very
very to see where the markets would absorb it, like
(19:02):
the idea that we still paying for the last one.
Speaker 2 (19:06):
They are are white collar workers, I mean, do they
tend to vote conservative or labor Because one of the
best analysis that I've ever seen was actually the fact
that you know, you start maybe a little bit idealist
at like eighteen, nineteen, twenty years old, and then you
buy a house and you turn a little bit more
to the right. And actually, if you can't afford to
buy a house anymore than it completely changes the politics,
(19:26):
and it changes the politics of this country. I mean,
can you see the same with with actually AI, what's
the analysis?
Speaker 3 (19:33):
Yeah, So in terms of white collar work, how people
split politically is very much down to a couple of acces.
So I'll just run through them. So the first is
basically where is your wealth from? So in general, if
the white collar work is a function of business that
is very government adjacent or is actually government, those voters,
(19:53):
irrespective of their age, tends to tilt to the left.
So that might be for example, working for a NGO
in be working for a you know, a quango or
kind of government body, or it might be working for
a company who the majority of whose contracts are basically
at the pleasure of federal government spending, and where the
work is much more downstream of you know, kind of
(20:14):
pure play private sector that tends to tilt much more
to the right. Then you've got age where that's a
clear effect as you just described. Then of course, if
we're just talking about white collar work, talk about industry.
So industries have hugely different political tilts and the US
we know this from donation data, where you can basically
track who's donated to which party in aggregate. By industry,
(20:38):
certain industries tilt much more strongly to the right. They
tend to be more in kind of computing, parts of finance,
particularly those more to do the kind of asset management
real estate. And then there are other sectors that tilt
very strongly to the left, whether that would be the
education sector, the health sector. Again, the other thing think
(20:59):
about is you unionization. You think of people who belong
to unions, is all blue collar? They're not, you know,
millions of private sector union numbers. That's another very very
strong tilt. So I think the answer is quite quite complicated.
But the effect of AI, I think it's to rob
countries of a of a classic kind of managerial cohorn
(21:22):
of people that's sufficiently large, and I wonder where they're
all going to go is what I think what we're
seeing with a lot of these of these kind of
what ecre scalable companies is there aren't that many people
who work in them versus their size, and I'm not
sure we thought about what that looks like in terms
of in terms of white collar wealth.
Speaker 1 (21:41):
Okay, let's let's let's take a turn for them more optimistic.
If you said, you said halfway through this conversation that
actually your piece was was bearish, but give us the
bull case in two sentences, why is it going to be? Okay, James.
Speaker 3 (21:59):
So, so the thing I'm very bullish on is that
we are going to see a reappraisal the qualities of
creativity and intellectual plasticity and dynamism. Why because actually, if
you look at the analysis, So if you're wondering where
does this AI exposure analysis all come from, it's very boring,
(22:21):
becus I should go through the methodology. But basically, you
can break down human skills into fifty two components A
typologist card, yeah, yes, this is a lot. A guy
called Fleischer break down basically human skills and capabilities into
fifty two dimensions. And then basically most of the AI
exposure analysis is basically taking those fifty two skills and
(22:43):
asking across each one, what is the degree to which
that skill is exposed to a certain type of technology. Now,
for me, what was very interesting when I did the
initial analysis is basically a matrix is the skills the
human skills that are least exposed, our creativity and kind
(23:04):
of problem solving and kind of domain switching in other words,
that kind of intellectual dynamism and being able to kind
of context switch. And for me, that's really interesting because
that's it kind of reminds us what's very human and
what we most value in people. And I'm very interested
to see whether the education systems start to kind of
(23:24):
re engineer around that, Like how much of the processes
and education exams we have to really examine pure intellectual creativity.
Speaker 2 (23:33):
But quite skeptical, doesn't it puts introverts as at a
natural disadvantage.
Speaker 3 (23:38):
It depends on your definition of creativity. You know, someone
could write something that's an incredible memo that explains the
world in a different way that you'd never thought of before,
but they never spoke it. They've just delivered it into
your inbox. Creativity can have all sorts of import I
do think on the kind of what is it the
humans are going to do, it's going to be around uh,
(24:01):
communication connectivity. Still you know, people still sell to people,
people still talk to people, And you're right, there's there's
a bit of humanity there that's much more kind of extrovert,
kind of tilted. That's going to have some value. But
you know, these things come in waves. The key thing
(24:21):
is these technologies keep on changing. The point that James
wise I spoke about earlier in terms of robotics, that
may change things up again. Let's see.
Speaker 1 (24:30):
Okay, all right, well we got there. We got there
to upbeat to an upbeat note. Thank you so much
for joining us, James.
Speaker 2 (24:38):
Thank you, James, Thank you both.
Speaker 3 (24:44):
So we have some news everybody.
Speaker 1 (24:45):
This is actually our last in the City episode.
Speaker 3 (24:49):
Oh I know, the.
Speaker 2 (24:51):
Grand emoji is in full force. In the podcast studio,
We've had so much fun, like huge fun.
Speaker 1 (24:56):
It's been sort of like a little Bloomberg confessional for
our guests occasionally for us. You can't see it at home,
but it's it's a small podcast studio, but it's surrounded
by glass and we can see lots of people working
outside on Bloomberg terminals, so it has a very good
atmosphere in here.
Speaker 2 (25:11):
Yeah, and it's also nice to work with colleagues that
you respect and kind of exchange ideas with and every
week trying to figure out the topic of the week
and actually how we can make a little difference in
explaining things that are not always easy to understand or
understood by everyone.
Speaker 3 (25:27):
Totally.
Speaker 1 (25:27):
It's made me sharper and know and hear about things
that I wouldn't have done otherwise. So thank you from.
Speaker 2 (25:34):
No it's been great. We're missing Dave exactly.
Speaker 1 (25:36):
Thank you and all of the brilliant Bloomberg reporters who
you will all be familiar with, who join us so often.
Speaker 3 (25:44):
It's been a.
Speaker 2 (25:44):
Blast, it has been, and thank you to our producers.
But keep an eye out on the feed because we
have something new. It'll be coming out soon, so more
conversations with great guests and we'll tell you more soon
about that.
Speaker 3 (25:58):
D
Bloomberg Audio Studios, podcasts, radio news.
Speaker 2 (00:17):
Welcome to in the City. Each week we unpack a
story that's crucial to the world's financial capitals. I'm Francis Laqua.
Speaker 1 (00:23):
And I'm Allegra Stratton.
Speaker 3 (00:24):
So Fran.
Speaker 1 (00:25):
It's been going on for years, but it feels like
it's reached something of an apoget, this debate about artificial intelligence,
and it's of course the ability to deliver big productivity gains,
particularly here in the UK, but also the problems with it.
Speaker 2 (00:39):
Yeah, and we've we've even seen the government actually double
down on this commitment to AI. Recently the PM focused
the Cabinet meeting on how it could potentially be used
to improve government services. But us a living standards.
Speaker 1 (00:51):
Right and so for all of us it is unclear
how technology, this technology is going to reshape the economy.
But we've got report on Bloomberg that businesses are already
changing the way they hire. A new study from McKinsey
looked at how UK companies are reducing hiring for jobs
that they think will be impacted by AI. Postings for
(01:12):
jobs were actually already down are a crazy thirty eight
percent in industries that will be affected by AI, So
in our conversation today we're going to get into what
role AI will play is playing in solving the UK's
productivity puzzle guess, but also more immediately, how it is
already changing the UK's job landscape.
Speaker 2 (01:33):
Now joining us is James Kanagasorium. James's chief research officer
at the AI enabled research firm Focal Data, and has
been instrumental in implementing data driven decision making processes across
various sectors, including finance and tex so. James, thank you
so much for being.
Speaker 3 (01:49):
On the show.
Speaker 2 (01:50):
Welcome to the City of London, the city of the
City of London. Please mind the gap between the tram.
Speaker 3 (02:01):
The financial heart of the country the city the city.
Welcome to in the city, stand clear of the dawn.
Speaker 1 (02:16):
We are seeing a debate in America, particularly between big
big tech titans in different points of view, about whether
AI will demolish the white collar employment cohort. We've had
fort CEO say two weeks ago could be as many
as fifty percent of jobs lost for white collar workers
from automation. We've got the anthropic CEO saying he thinks
(02:39):
twenty five an increase of twenty percent unemployment you know,
these guys are kind of bidding to outbid each other
with their apocalyptic record rhetoric, but clearly there is there
is a problem. And then on the other hand, you've
got the boss of Nividia saying, guys, we're going to
create more jobs through AI. Tell us where you stand.
Speaker 3 (02:56):
I would stand more on the former camp, which is
that I can see the immediate upsides, and I can
see the fact that different countries might benefit from it differently.
And I think the UK is potentially not that badly positioned,
given it's kind of human capital stocks, the fact that
it has lots of businesses that are pretty adjacent to
AI or are in AI itself. But the reality is
(03:17):
that if you strip away all the rhetoric, firms that
hire white collar graduates are hiring fewer of those people,
and they're doing so in a degree that you would
typically associate with a mass recession. So again, the numbers
that you've talked about, whether it's twenty five, fifty percent,
those types of numbers in terms of cohorts of people
rolling on is what you'd associate with the mass financial crash.
(03:40):
But the idea that that's a kind of permanent shift
is pretty structural. Sorry to interrupt you.
Speaker 1 (03:47):
James mckimsey showed there was a thirty eight percent reduction
in what they call AI exposed jobs.
Speaker 3 (03:54):
Yeah, exactly, And it depends on who's getting augmented and
who's getting exposed. You know, I'm in one of those
jobs in terms of market research, but lucky enough to
work at a firm that is actually an AI market
research firm, so I'm one of the I work with
a few small band of cohort people who are getting
augmented to hopefully at the forefront of this role out
(04:17):
of this technology. But there are much wider societal impacts,
like what does it mean to literally have a milk
ground that is a third to half the size of
what it is before? What does that mean for the
tax base? What does that mean for the university education
sector and the number of people graduating with the degrees
to expect to go into white collar jobs. Like a
lot of politics and social issues are basically downstream of
(04:38):
the gap between people's expectations and then what is delivered
in terms of society, jobs and pay. And the third
thing is obviously, I mean, what's really interesting about this
is that this is clearly differentially affecting a group of
people that you would not necessarily associate with being left behind.
Right we're talking about graduates who I would have expected
(05:01):
to basically earn good to great salaries over the course
of their careers, who are probably living in cities who
could expect some kind of graduate premium. Then they're suddenly
walking into a job market where none of these things exist.
But they've already crystallized a lot of the downsides, so
they already have, for example, debt from a university. They
have already moved to cities that have high house prices.
Speaker 2 (05:24):
I speak to a lot of CEOs and I mean,
these are big CEOs that should one hundred percent understand
what AI means for their workforce, and frankly they're not.
I mean, you know, you'll have a coffee with them
and they say, I'm looking at AI training, but I
don't one hundred percent understand how it's going to impact
my organization.
Speaker 3 (05:40):
Yet.
Speaker 2 (05:40):
We also spoke to the Accenture CEO and she was saying,
they're training five hundred thousand staffers because a consulting work
in AI is increasing by so much. So do we
actually have a full understanding of the impact of AI
in the next three four five years and what sectors
will be most affected.
Speaker 3 (06:01):
Yeah, I think it's a really good question, like how
much of the analysis is evergreen. One thing that I
when I first started this kind of geographic analysis of
which jobs were more exposed to AI and therefore what
were the kind of community impacts. The thing that really
struck me in a UK context was the government's own
analysis of which jobs were AI exposed is basically the
(06:23):
opposite in twenty twenty three to its initial analysis in
twenty nineteen. In other words, the jobs, the geographies, the
education levels that people four were going to be automated
are not just slightly different, but they're almost the opposite.
You know. Back in twenty nineteen, the Department of Education,
I think, released a paper that basically indicated that a
(06:43):
lot of the areas that were going to be AI
exposed were the same areas that suffered from de industrialization.
That is now the opposite of what the analyses are saying,
which is actually the jobs of the AI exposed far
more white colar. But again these are very broad terms.
The analysis that a lot of these things are based
(07:04):
on is around emerging technologies, their technologies change, the mixed changes,
the applications change. You know. One of the bits of
feedback that I thought was very interesting to my substack
around it was a financier called James Wise, who's one
of the lead partners of Balderton, who invests a lot
in this area, and he said, look, okay, fine, you've
(07:26):
pointed out that white collar jobs are more vulnerable than
blue collar, but look at the advances in robotics. So
he was potentially painted an even a more challenging picture.
But I do think there's obviously going to be a
one off, cohort effect of people who are clearly going
to benefit, right, and we're going to hear a lot
from them, and we're going to hit that they're very
(07:48):
florid about how the ships might increase things like productivity
and efficiencies. But I'm taking a step back and thinking fundamentally,
this is a complete re engineering of the wider structure
of how societies work. We're potentially losing mass seams of
upper middle class citizens who very disproportionately form a massive
(08:08):
part of the tax space, for better or for worse,
and the fact that this is happening at such a
high cadence and pace means policymaking or whether it's corporate's
find it quite hard to distill. For me. The most
interesting thing really about all of this was I was
speaking to an ext the ex chairman of You Girl,
(08:29):
gentleman called Roger Parry. Quite interesting guy. You know, he's
settled many boards, including companies like Uber, and he was saying, look,
there's a degree of parallels to the initial rollout of
of the Internet, which was back in the day when
it was first, when it first happened. You often these
boards would often get the it guy whose job was
(08:49):
hardware to explain the Internet to boards. And maybe we're
a little bit in that phase where we're not quite
sure who the experts are, What is the information that
companies need to do, what are the amateurs? How is
that communicated to workforces?
Speaker 2 (09:03):
But James, it's not showing up in data yet, right.
If what you're explaining is true, then unemployment should go up.
It hasn't gone up yet, So is there going to
be Is it like a small drip feed where you
lose jobs you don't really understand why the industry is
changing and no one really realizes or is it going
to be like a recession where suddenly that's it. You
(09:24):
lay off thousands and thousands of staff in a company.
Speaker 3 (09:28):
So my job, I'm not not an economists. My job
isn't a forecast the labor market. What I can fundamentally see, though,
is the displacement and substitution that's happening is not offset
yet by the creation of new roles. But we can
see new roles right, we see them. I see them
in my own firm. You know, people who are concentrated
(09:51):
around customer success or people who are involved around sales,
or people that are used to basically take these technologies
and deploy them. People are getting augmented. There are new roles,
there is a degree in creativity, but the loss of
thirty to fifty percent of a white collar industry that
is mass displacement. You know, we should be thinking about
(10:12):
this in exactly the same terms as we think about
for example, when oil runs out of a particular area,
what does that look like? So the UK that context,
that would be north sea oil. We should be talking
about this is is like de industrialization equivalent, but on steroids.
Speaker 1 (10:27):
Do well do you see that, James?
Speaker 3 (10:29):
Do?
Speaker 1 (10:29):
What is what are those I mean Obviously it's sort
of difficult to dynamically predict what new jobs spring up
because of AI. But certainly that is Navidia CEO's position.
Isn't it that the jobs are going to be more
and they're going to be better. But as you say,
we're not yet completely take your point that you can
see in your own firm. But are they as many
jobs as the ones that are being displaced?
Speaker 2 (10:50):
Do you think?
Speaker 3 (10:51):
I'm skeptical, is what I would say, But jury is out.
We're just of the foothills of these pretty large ships
in both the labor market and how we understand work.
A lot of companies don't know how they would create
the new jobs that exist. The duo Lingo CEO right saying,
you know, we're not hiring anyone if we think they're
you know, exposed in any kind of way. I think
(11:12):
we're still getting to the stage where we're trying to
understand what are the jobs that could be expanded, what
are the ones that could be augmented. It's clearly not
going to be all negatives, and it's clearly not all
going to be positives from this aspect, but certainly from
what I can see in the most automatable industries. Like
take something really really basic, like a young investment banker
(11:34):
would spend a lot of their time building things like
discal cashlow models in Excel and pulling financials and doing
research into certain sectors. These are jobs that cannot just
be done with a few lines of code, but can
also then be done by a lot of these technologies.
And we're talking. Yeah, I was thinking about to my
first graduate job when I worked in the city, and
(11:55):
I was just thinking how few people would now be
needed to conduct that. I mean, there's a much wider
conversation to be had about skills and training. So how
do people come experts in an era where basically the
whole of society has this big red button that they
can push to basically push the automatically get along pushed someone.
Speaker 1 (12:20):
But that's something This is this is what question I
wanted to ask you. I mean, Neil focus and was
writing a couple of weekends the historian and economists writing
a couple of weekends about setting up a university, and
the university is not going to allow the use of AI.
So how much do you think in the years ahead,
Maybe not immediately, maybe he'll be an outlive for a while,
But how much is some of this going to be about?
The kind of refused nick mentality, which is, yes, AI
(12:43):
can do a whole lot of this, but at some point,
if we think we're looking at thirty to fifty percent unemployment,
we have to choose not to do.
Speaker 3 (12:49):
Some of this. So there's there's there's a lot of
different things baited into that question, so sorry. The first
is what does it know? It's brilliant. I love it,
So there's like what what? The first is what does
it do to people's learning frameworks MIT? I think have
Reas recently released a paper on what the consequences are
for using lllms and different aspects of AI, and basically
(13:12):
it leads to quite a drastic reduction in basically your computation.
In other words, we're basically leasing out bits of our
brains for technology to basically execute. I think we'll start
to see AI three labels like we do with food
with organic and that you might start to see kind
of things like two tier pricing systems where what people
(13:33):
are really trying to passe because when you pay for something,
you've obviously got cost based pricing. How does someone value that?
And it's the idea. Let's say you've got a consultancy
that comes in and does a project for you, and
you know that it's all kind of been done with
human minds versus once it's been augmented. Will people start
to think about that pricing differently. I don't know. Will
(13:54):
people be happy to pay the same half million pounds
for a consulting project if they know that it's basically
being conducted by by one human and ten bots and
one co pilot. Maybe we don't know. This is really
at the kind of frontier of kind of human psychology
and economics. But I do think you'll start to see
the refuse next, as you put at Allegra both in
(14:15):
terms of political parties, by giving away too much of
my yeah, I'm actually more I'm more optimistic than perhaps
my column indicated. I was giving the bare case about
how it's going to change our politics as opposed to
kind of a labor market pieces, which is that all
the people who have done quite well basically out of
(14:36):
a globalized financial system, expanded higher education, and slightly medium
to large immigration levels have done pretty There's a group
of people who've done pretty well. They all live in
certain parts of the UK and the US and other
English speaking countries. Those appeared to be the same geographies
that are most day are exposed, which produces a very
(14:57):
interesting politics of people who's previously very well starting to
feel much more threatened.
Speaker 2 (15:04):
As we say, I mean that feels like the Trumford already,
doesn't it.
Speaker 3 (15:09):
Well for me? One of the really interesting things of
politics over the last twenty five years has been this
decoupling of people's income household income and basically their levels
of education. Because back in say, for example, the early nineties,
really up until about twenty ten, they kind of walked
to lockstep. If you went to an area that had
(15:31):
lots of people who'd gone to university, it was probably richer,
and you know, there were some differences. But I think
what we've seen basically since higher education has been expanded
out to be fifty percent in a lot of countries,
but actually the promise them of that delivering higher incomes
has declined, and also the growth of secure blue collar
employment has meant this really really odd decoupling of the two.
(15:53):
And it was I think Thomas Packetti was really the
guy who's you know, he's the economist who is most
just work on inequality. But he pointed out, actually the
future that we may see is of secure blue collar workers,
which he called the merchant right making a march on society,
versus what he called the Brahman left, which is the
(16:15):
person who's got three degrees but no mortgage and very
very low salary. And I think we haven't quite thought
through what the consequences of that, because the politics of
those two groups are drastically different to classically left and
classically right.
Speaker 1 (16:28):
One question on the just the kind of the politics
of the left and right that we have right now.
You've worked in politics yourself, James, and listeners don't know.
But when I first came across you, as we say,
you were the guy who coined the term the red wall,
didn't you? It was really brilliant analysis and it has
come to live and live and then get smashed down
as well as a political entity. But how do you
(16:52):
think this government? Do you think this government, the labor
government is understanding the scale of what we've been talking about.
And then the country that I've been really about recently, Sweden,
they brought in a furlough type scheme to help people
skill up while in work in order to deal with
these sort of existential threats, but we can't really necessarily afford.
Speaker 3 (17:15):
That right now. Yeah, it's very interesting couple of points.
So the first is the whole political system, where is
the UK or US is currently dealing with we can
only describe as an omni crisis of the arrival of aire,
geopolitical instability, financial market instability. Particularly I think about the
UK and the potential for a debt crisis. There is
not necessarily the bandwidth or the expertise within government to
(17:39):
basically pass and understand these technologies and then work out
what should we do because they're basically trying to survive
day to day governments. As you will note, alegra live
Pole by Pole and this one has lost twenty points
in less than a year and a half. So are
they going to be doing the long term thinking about
what the labor market is going to be looking like
when effectively they've lost the most a man of votes
(18:00):
share for a majority elected government in living memory? Unlikely?
Do they have the MPs? Do they have the MPs
to understand? There are good MPs that sit on the
government benches who really understand this. You see them, you know,
I think there's that the Labor Growth Group. I think
there are a lot a lot of MPs there who
are very interested in the sector. I think MP's like
(18:21):
Jake Richard's conditionnarian who really understands some of these kind
of technologies, who are kind of thinking about what the
long term kind of consequences are. But let's, you know,
will the UK have the appetite for another furlough scheme depends.
I think a lot of the issues that we have
now is that the UK is quite vulnerable in terms
(18:41):
of its its debt pile, it's deficit, the problems it
has with its labor market and the number of people
who are out of work. I think the idea that
there's going to be another labor market collapse that's going
to require unilateral government support might win on voters very
very to see where the markets would absorb it, like
(19:02):
the idea that we still paying for the last one.
Speaker 2 (19:06):
They are are white collar workers, I mean, do they
tend to vote conservative or labor Because one of the
best analysis that I've ever seen was actually the fact
that you know, you start maybe a little bit idealist
at like eighteen, nineteen, twenty years old, and then you
buy a house and you turn a little bit more
to the right. And actually, if you can't afford to
buy a house anymore than it completely changes the politics,
(19:26):
and it changes the politics of this country. I mean,
can you see the same with with actually AI, what's
the analysis?
Speaker 3 (19:33):
Yeah, So in terms of white collar work, how people
split politically is very much down to a couple of acces.
So I'll just run through them. So the first is
basically where is your wealth from? So in general, if
the white collar work is a function of business that
is very government adjacent or is actually government, those voters,
(19:53):
irrespective of their age, tends to tilt to the left.
So that might be for example, working for a NGO
in be working for a you know, a quango or
kind of government body, or it might be working for
a company who the majority of whose contracts are basically
at the pleasure of federal government spending, and where the
work is much more downstream of you know, kind of
(20:14):
pure play private sector that tends to tilt much more
to the right. Then you've got age where that's a
clear effect as you just described. Then of course, if
we're just talking about white collar work, talk about industry.
So industries have hugely different political tilts and the US
we know this from donation data, where you can basically
track who's donated to which party in aggregate. By industry,
(20:38):
certain industries tilt much more strongly to the right. They
tend to be more in kind of computing, parts of finance,
particularly those more to do the kind of asset management
real estate. And then there are other sectors that tilt
very strongly to the left, whether that would be the
education sector, the health sector. Again, the other thing think
(20:59):
about is you unionization. You think of people who belong
to unions, is all blue collar? They're not, you know,
millions of private sector union numbers. That's another very very
strong tilt. So I think the answer is quite quite complicated.
But the effect of AI, I think it's to rob
countries of a of a classic kind of managerial cohorn
(21:22):
of people that's sufficiently large, and I wonder where they're
all going to go is what I think what we're
seeing with a lot of these of these kind of
what ecre scalable companies is there aren't that many people
who work in them versus their size, and I'm not
sure we thought about what that looks like in terms
of in terms of white collar wealth.
Speaker 1 (21:41):
Okay, let's let's let's take a turn for them more optimistic.
If you said, you said halfway through this conversation that
actually your piece was was bearish, but give us the
bull case in two sentences, why is it going to be? Okay, James.
Speaker 3 (21:59):
So, so the thing I'm very bullish on is that
we are going to see a reappraisal the qualities of
creativity and intellectual plasticity and dynamism. Why because actually, if
you look at the analysis, So if you're wondering where
does this AI exposure analysis all come from, it's very boring,
(22:21):
becus I should go through the methodology. But basically, you
can break down human skills into fifty two components A
typologist card, yeah, yes, this is a lot. A guy
called Fleischer break down basically human skills and capabilities into
fifty two dimensions. And then basically most of the AI
exposure analysis is basically taking those fifty two skills and
(22:43):
asking across each one, what is the degree to which
that skill is exposed to a certain type of technology. Now,
for me, what was very interesting when I did the
initial analysis is basically a matrix is the skills the
human skills that are least exposed, our creativity and kind
(23:04):
of problem solving and kind of domain switching in other words,
that kind of intellectual dynamism and being able to kind
of context switch. And for me, that's really interesting because
that's it kind of reminds us what's very human and
what we most value in people. And I'm very interested
to see whether the education systems start to kind of
(23:24):
re engineer around that, Like how much of the processes
and education exams we have to really examine pure intellectual creativity.
Speaker 2 (23:33):
But quite skeptical, doesn't it puts introverts as at a
natural disadvantage.
Speaker 3 (23:38):
It depends on your definition of creativity. You know, someone
could write something that's an incredible memo that explains the
world in a different way that you'd never thought of before,
but they never spoke it. They've just delivered it into
your inbox. Creativity can have all sorts of import I
do think on the kind of what is it the
humans are going to do, it's going to be around uh,
(24:01):
communication connectivity. Still you know, people still sell to people,
people still talk to people, And you're right, there's there's
a bit of humanity there that's much more kind of extrovert,
kind of tilted. That's going to have some value. But
you know, these things come in waves. The key thing
(24:21):
is these technologies keep on changing. The point that James
wise I spoke about earlier in terms of robotics, that
may change things up again. Let's see.
Speaker 1 (24:30):
Okay, all right, well we got there. We got there
to upbeat to an upbeat note. Thank you so much
for joining us, James.
Speaker 2 (24:38):
Thank you, James, Thank you both.
Speaker 3 (24:44):
So we have some news everybody.
Speaker 1 (24:45):
This is actually our last in the City episode.
Speaker 3 (24:49):
Oh I know, the.
Speaker 2 (24:51):
Grand emoji is in full force. In the podcast studio,
We've had so much fun, like huge fun.
Speaker 1 (24:56):
It's been sort of like a little Bloomberg confessional for
our guests occasionally for us. You can't see it at home,
but it's it's a small podcast studio, but it's surrounded
by glass and we can see lots of people working
outside on Bloomberg terminals, so it has a very good
atmosphere in here.
Speaker 2 (25:11):
Yeah, and it's also nice to work with colleagues that
you respect and kind of exchange ideas with and every
week trying to figure out the topic of the week
and actually how we can make a little difference in
explaining things that are not always easy to understand or
understood by everyone.
Speaker 3 (25:27):
Totally.
Speaker 1 (25:27):
It's made me sharper and know and hear about things
that I wouldn't have done otherwise. So thank you from.
Speaker 2 (25:34):
No it's been great. We're missing Dave exactly.
Speaker 1 (25:36):
Thank you and all of the brilliant Bloomberg reporters who
you will all be familiar with, who join us so often.
Speaker 3 (25:44):
It's been a.
Speaker 2 (25:44):
Blast, it has been, and thank you to our producers.
But keep an eye out on the feed because we
have something new. It'll be coming out soon, so more
conversations with great guests and we'll tell you more soon
about that.
Speaker 3 (25:58):
D
Host
Francine Lacqua
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. Special Summer Offer: Exclusively on Apple Podcasts, try our Dateline Premium subscription completely free for one month! With Dateline Premium, you get every episode ad-free plus exclusive bonus content.
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!