Latest / In The Loop / 22 New Jobs AI Will Create—And Do AI-Pessimists Have A Point?

Transcript

22 New Jobs AI Will Create—And Do AI-Pessimists Have A Point? · In The Loop

0:00
In today's episode, we're going to discuss a New York Times
0:03
article which listed the 22 different jobs that AI is
0:07
apparently going to create, plus what is, I think, a really,
0:10
really interesting historical take on AI and the idea of a
0:15
Luddite. So those that are against
0:17
technology. And the reason it's so
0:18
interesting is when compared and contrasted with this new
0:21
article. Well, I think there's a lot to
0:24
learn. This is in the loop with Jack
0:26
Horton. I hope you enjoy the show.
0:40
It's sweltering hot. I mean, I know there's a lot of
0:43
US listeners and they're probably thinking you're
0:45
pathetic because it can't be that hot.
0:48
I know it gets extremely hot in the US, but right now I think in
0:51
Manchester where I'm based, it is 2930°.
0:55
I know in other parts of the UK it's 33.
0:57
So I'm going to do my best not to melt whilst I'm trying to get
1:00
through the entire episode. But I guess today's episode's
1:04
really interesting. I, I had AI, someone shared me a
1:08
video from actually a Ted Talk at the exact same time as me
1:12
reading, researching and writing this particular episode.
1:16
And it kind of, they just felt perfect together.
1:18
And it gave me a really interesting and new perspective
1:20
on the episode that I was actually in the middle of
1:23
writing to start from the top right now, obviously there's a
1:25
lot of people thinking about AI and many of these people are
1:29
somewhat panicking. And when I say panicking, I
1:30
don't mean running around in the streets, you know, thinking the
1:33
world's over. But I'm very concerned and not
1:37
necessarily about, you know, those, as I said, doomsday
1:39
scenarios, but you know, what's going to happen with the future
1:42
of jobs. And I, and I think this
1:45
perspective is worth exploring because right now, times are
1:48
quite tough all over the world. Fewer and fewer people can't
1:52
afford housing, and that's because of massive price
1:55
increases as a percentage and as a ratio to the amount you earn.
2:00
Governments are selling off all their assets to try and drive
2:03
down debt. And the people buying them are
2:05
big businesses who then are charging extortion amounts for
2:09
delivering the exact same service to the government, which
2:12
taxpayers are paying for, which is leading to, you know, less
2:15
houses being built, less land, less assets, higher debt.
2:20
And there's also a lot of, as a result, I think civil conflict
2:23
as well. There's lots of global conflict
2:24
going on as well. That's, that's only escalating.
2:28
And I think it was just coming at a time with a smaller social
2:31
net around the world. There is going to be less
2:33
support out there for people and there's less high quality,
2:36
affordable health healthcare, you know, for these same people.
2:39
So I think these fears about job displacement are extremely valid
2:42
and worth seriously having because for many, many people,
2:46
it's going to be really painful. And, you know, many of the
2:48
people listening to this podcast might not be affected, but maybe
2:51
they've got family members affected.
2:52
You know, I've done episodes on how early careers, so those that
2:56
are just coming out of say college or university, how much
3:00
they've been hit. And you know, for all those in
3:02
knowledge work, so writing, analysis, creative fields, you
3:07
know, we're watching AI and AI agents get better and better at
3:10
these tasks every single day. And so it genuinely raises a big
3:13
question about what the future of work works like for those
3:15
people too, who are often more than middle class white collar
3:18
workers. So, yeah, these concerns, it's
3:21
not to be make this episode super negative, but they aren't
3:23
theoretical. And things can happen really
3:25
fast. And, you know, the snowball
3:27
effect can happen really fast. And and I promise I go into more
3:30
of AI guess a more positive analysis over throughout the
3:33
episode. But, you know, we've talked
3:35
about on this on this podcast series, major CEOs openly
3:39
discussing, you know, big changes they're going to make.
3:42
You know, Amazon CEO Andy Jassy just came out and said that
3:47
basically I released a memo, shall I say, saying to all their
3:50
employees, basically the AI agents will reduce our
3:53
workforce. And, you know, they're openly
3:55
saying we want efficiency from using AI throughout the
3:58
business. And he described AI as this
4:01
system that will perform tasks on the behalf of everyone.
4:03
And there's going to be billions of these agents across every
4:06
single company instead of needing so many employees.
4:09
And he openly accepted that while agents make certain jobs
4:12
more fun because maybe you're not doing as many menial tasks,
4:17
it will also mean there'll be fewer people needed those
4:20
companies. So there is a transition going
4:23
from, you know, the assistant era, so assisting me to do work
4:27
to the agenda era, meaning kind of AI is moving beyond helping
4:31
and actually just doing many of the things for people.
4:34
And as I said in in previous episodes, the amount these
4:37
systems can actually do people, you know, conflate a good video
4:42
or a good headline with reality. I don't think they're there yet,
4:45
but it is going to get there. But this is where I think that
4:59
historical end is really interesting and that someone
5:01
shared to me on the weekend whilst I was reading about this
5:04
and researching. And I thought it was perfect.
5:06
And it's the story of the Luddites and I guess why their
5:09
experience holds the key to potentially understanding not
5:12
just what AI will destroy, but what hopefully it'll create.
5:15
So what are the Luddites? And why perhaps did they have a
5:19
point? Well, most people use the word
5:22
Luddite as this broad turn to anyone resistant to technology.
5:25
It's often used in a very negative context.
5:27
And this particular piece came from Megan McArdle in a Ted
5:30
talk. And she pointed out that the
5:31
original Luddites weren't just technophobes.
5:33
They were actually artisans who, you know, create handcrafted
5:37
textiles in an era where everybody loved handcrafted
5:41
items. And that was the only way you
5:42
could buy, build and then buy something.
5:45
But then, obviously, mechanized mills popped up everywhere,
5:48
underpriced them. And then, you know, suddenly
5:51
spinning Jennys appeared that could produce all the things
5:53
that these artisans are producing 10 times faster.
5:57
And obviously, faced with economic devastation, you know,
5:59
they decided to destroy the machines themselves.
6:02
They actually attacked them. And I guess the thing with this
6:05
really is that you have to have some sympathy for their
6:08
position. You know, we often celebrate the
6:11
glory of progress as people, and often this is truly glorious.
6:15
But progress isn't free. Sometimes people do get hurt and
6:20
often lots of people get hurt. The printing press democratized
6:24
knowledge but also enabled which burnings of religious wars, the
6:27
industrial revolution raised living standards whilst creating
6:30
awful factory conditions and urban squalor and terrible
6:34
environmental damage. So as Megan Mercado in this talk
6:37
puts it, modern governments can address many of those costs, but
6:41
they can't restore what people lost.
6:44
And I think we all have an obligation to acknowledge the
6:46
costs of every single revolution.
6:48
And I guess this is where it comes to the counter argument to
6:50
that position. And it's something that highly
6:52
resonated with me because ultimately we're all
6:55
beneficiaries of the previous generations decisions to
6:58
prioritize future growth over protecting present
7:01
circumstances. You know, she makes this point
7:04
beautifully because consider this, very few people listening
7:07
to this podcast today have ever really worried about, say, basic
7:11
food, shelter or heat or how to bury a child who died of a very
7:15
preventable disease. And those are the things that
7:18
represented, I guess, a normal condition throughout most of
7:22
human history. And we've escaped them.
7:25
No, not everybody. There's many people in the world
7:27
still suffering that. And because of global wars and
7:30
crises, that number is going out drastically.
7:34
But overall, most people's living standards have improved.
7:37
And there's a strong argument to say that we also have an
7:40
obligation to pay that forward, to leave an even bigger legacy
7:44
for future generations. To do so otherwise could be seen
7:47
as a form of theft from future generations.
7:51
So the author posed this really uncomfortable question, which is
7:54
imagine if the Luddites had actually succeeded.
7:56
So they destroyed all that technology, and that halted
7:58
progress. Essentially, they would have
8:00
been reaching forward in time and taking nearly everything we
8:04
have today to preserve their current circumstance.
8:07
So the question has to become, when we want to halt innovation
8:11
that might disrupt how we currently live our lives, how
8:14
much are we willing to take from the next generation?
8:17
And I guess when viewed in this way, progress seems a little bit
8:21
different to me because the natural response is, but we
8:24
already have it. So goods, you know, we have air
8:26
travel, vaccines, we've got streaming services.
8:29
But a Luddite would have felt the exact same way about their
8:32
era. They couldn't have conceived of
8:35
a future where an average worker's lives are longer,
8:38
healthier, happier, more comfortable than in the 19th
8:41
century. So many people might ask the
8:43
question, is progress really beneficial when you look at
8:46
climate change and the environment's destruction?
8:48
But consider your last, I guess, dental appointment.
8:51
Imagine having that without anesthesia.
8:53
The same industrial systems that have created awful environmental
8:57
challenges of made it possible for us to not suffer pain when
9:01
having a tooth ripped out that's made us wealthy enough to
9:05
redirect millions of workers basic survival into science and
9:09
medicine. So progress operates
9:12
cumulatively. You know, the longer it
9:14
accumulates, the more surprising it becomes because we can easily
9:17
envisage, you know, near term descriptions, but we can't
9:20
understand those long term benefits that appear.
9:22
It's so difficult. So this brings us back all the
9:35
way to our current AII guess, anxieties.
9:38
And if you apply some of these ideas to modern anxieties and
9:41
challenges, it makes me look at the world a little bit different
9:45
because there were legitimate worries about job displacement.
9:48
That deserves a lot of consideration.
9:50
You know, government retraining programs are needed and they
9:53
might help, but ultimately, I don't think if you ask people
9:57
they want that assistance. What they want is to have the
9:59
same careers as they've invested their lives into.
10:02
So when facing these questions or concerns about whether we
10:05
should slow or stop AI development, I guess this
10:08
historical perspective offers really important context, which
10:11
is we don't really have the right to limit future
10:15
possibilities for future generations because we're
10:19
already living in someone else's impossible future.
10:21
And now when it comes to AII, think it is going to be
10:24
extremely painful and I think many people are going to lose
10:27
jobs. I also think many people are
10:29
going to get new types of jobs. And I think all of the same
10:32
points I've made today is we have no idea how good it could
10:35
get and equally how bad it could get.
10:37
But I think ultimately, as a civilization, we're screwed
10:40
without investing in technology. I don't think we can overcome
10:43
climate change or damage the world so bad unless we invest
10:47
everything we have into something that could become a
10:50
super intelligence potentially. But we must balance all of the
10:53
opportunities with all the risks as best we can to manage the
10:57
transition period of this world to whatever future world we're
11:01
living in. And that requires compassion and
11:03
empathy and investment and higher tax into programs that
11:07
benefit everybody else. But there is hope.
11:09
And that's coming on to the point of the article from the
11:12
New York Times, which is 22 jobs that AI will create.
11:16
And I'm going to break them down because they're actually really
11:18
interesting. I've broken down into different
11:19
categories as well. But before I continue, I just
11:22
want to say thank you for being here.
11:24
I'm genuinely amazed to see in the Leap growing so much every
11:28
week we've had over 63,000 views and listens across all the
11:32
different channels. So it, you know, it feels
11:35
amazing. The goal from day one of this
11:37
has been to provide very easy insights and understanding into,
11:41
I guess, how AI is shaping work culture and the world around us.
11:44
And honestly, if you're enjoying learning about AI in 1520
11:48
minutes or less and you have anyone you know that might
11:51
benefit or be interested in, just share it.
11:54
Sharing a slight channel, a team's channel, share it via
11:57
WhatsApp, whatever it might be. It honestly means the world to
12:00
me. Anyway, back to the show.
12:02
So this leads me to the encouraging part of today's
12:04
discussion because obviously there's been a bit of doom gloom
12:07
help. And this is where it goes to
12:09
what the future could hold. Because the New York Times
12:11
article, as I said, it listed 22 new jobs and it falls into
12:14
different categories of jobs that are emerging.
12:17
Because according to the World Economic Forum, AI is going to
12:19
displace about 9 million positions by 20-30, but
12:22
apparently they think it's going to create 11,000,000 new ones.
12:26
So that's a 2 million increase in total jobs.
12:29
But I guess the question is what will those jobs even look like?
12:32
And again, most people don't want to have to re skill relearn
12:35
and transition to those new jobs.
12:37
They just want to maintain their careers.
12:39
So the categories are trust, integration, and taste.
12:42
The first one is the trust job. So there's about 8 roles here
12:44
that's listed. This exists because AI might be
12:46
super capable of doing things, but ultimately AI can't be
12:50
responsible for a error. You can't send AI to prison if
12:53
it does something bad. So you need humans to verify,
12:56
validate, and even take responsibility for AI's work.
13:00
So this comes to job number 1. So you have an AI auditor.
13:03
So apparently within five years, most accounting firms have for
13:06
AI audits. So these are professionals
13:09
digging into AI systems to understand what they're doing
13:11
and document it for either legal or liability reasons.
13:15
You might have an AI translator. So these might be specialist
13:17
explaining AI mechanics to business leaders.
13:20
So these are basically consultants.
13:22
You might have a trust director, so a senior role overseeing the
13:25
entire trust framework of an organization using AI systems,
13:28
because obviously accountability is really important.
13:31
Imagine if your AI accidentally starting rejecting education
13:34
applications based on a weird tenuous link that you used to
13:37
live in an area that's deemed dangerous.
13:39
You have an AI ethicist. So these are people that build
13:42
chains of, I guess, defensible logic to support decisions made
13:45
by the AI or hybrid teams. So you might have to explain
13:48
decisions to investors and customers.
13:50
Again, this black box of AI is a problem.
13:53
You'll also have consistency coordinators, so people who are
13:57
making sure that AI generated content remains consistent
14:00
across all applications. And you might have to have AI
14:02
escalation officers. Sometimes people just want to
14:04
talk to a human. These are going to have to step
14:06
in when interactions with AI are just so frustrating, kind of
14:10
like a customer service job. The second category is called
14:12
integration jobs. So there's about six of these as
14:16
a good example. So 1, you might have AI
14:18
integrations. So these are experts who help
14:22
businesses integrate AI into the business process.
14:24
We're seeing this play out already, an AI plumber, so our
14:28
systems become increasingly complex and autonomous.
14:32
They certainly become very layered and complex.
14:34
And while something might break, an AI plumber will have to dig
14:38
through the entire system to find the problem.
14:40
Another one is an AI assessor. So you know these models are
14:43
constantly changing. So you need to be able to
14:45
evaluate whether the AI system, your ability is getting worse or
14:49
regressing or if it's improving, if it's changed.
14:53
Again, an AI assessor is going to be absolutely critical.
14:55
That's something that I expect to start popping up even in the
14:57
next 18 months of scale. You're also going to have an AI
15:00
human evaluation specialists. So these professionals determine
15:03
whether AI performs best or whether humans are needed.
15:06
So again, making that assessment over staffing strategies
15:09
basically. And 3rd and final ones, you've
15:11
got the taste jobs, which I personally think are the most
15:13
interesting actually, because these roles emerge when everyone
15:16
has access to the same tools. The ability to make those bold
15:20
or stylish creative decisions becomes incredibly valuable.
15:24
So just think about it this way. So instead of a traditional
15:26
writer, we might see article designers who select inputs and
15:30
create phrases or paragraphs and reasoning from different AI
15:34
systems. So they're essentially authors,
15:36
not just writers. You have a world designer, for
15:39
example. So these are going to be
15:40
creators who fabricate essentially entire universes for
15:43
marketing campaigns or video games with fictional characters
15:48
and locations. You have product designers that
15:50
are an expanded job role. So future product designers will
15:53
own entire products from top to bottom, making not just big
15:57
picture decisions, but all the creative decisions that bring a
16:00
vision to life. You might have then a
16:02
differentiation designer. So as businesses gain access
16:05
against the same intelligence, you need professionals that are
16:08
able to combine philosophy and product development and branding
16:12
and creative execution and various different AI systems to
16:16
execute something that stands out.
16:18
So I think there's a bunch of really interesting job
16:20
categories that could emerge based on how systems are working
16:23
today. So as you finish today's
16:35
episode, I want you to think about how we're living today in
16:38
someone's unimaginable future. But equally, we must, and I mean
16:43
must put effort, care, and most importantly, resources into
16:46
helping people transition during what's going to be a very
16:50
interesting but difficult time. As I've said, I think we need an
16:53
era of compassion and I hopefully automated luxury
16:56
communism team. Anyway, that's it for today.
16:58
I'll see you next week.