00:00:00
Dan Loney: Well, it seems that we
may be hitting somewhat of a unique and potentially
concerning part of the job
00:00:06
marketplace. We are seeing
layoffs as companies somewhat right size their operations. And
at the same time, you are seeing
00:00:14
recent college graduates find it
harder to get that first job. So what does this all mean for
them, and especially, what does
00:00:22
it mean for the importance and
the value of a degree, or even going for your masters?
Pleasure to be joined right now
00:00:29
by Manav Raj, who's an Assistant
Professor of Management here at the Wharton School. Manav, how
are you? Great to talk to you again.
00:00:35
I'm doing well, Dan. Thanks for
having me. And so you're right in the
middle of this, probably seeing
00:00:40
and hearing from some of your
students, as— you know, some of the issues that they are dealing
with right now.
00:00:46
Yes, that's correct. I mean, I
see this on both sides of things. My research studies
technology change, and has
00:00:52
focused a lot on AI. So, obviously
thinking about the job market effects of that technology. And
certainly in the last few years,
00:00:59
teaching tech strategy and
interacting both with undergraduates and with the MBA
students here, there's a lot of
00:01:06
conversation and concern and
thought about, you know, what the job market might look like
and how students should be
00:01:11
preparing themselves best for
the future. So let's start with the AI
component, because it's
00:01:16
obviously one of the biggest
questions out there right now, and a lot of discussion as to
just how much AI is going to
00:01:23
impact the workforce in the
years ahead. Let me ask you this then, from two perspectives.
One, what are you seeing right
00:01:30
now? And two, what is your
expectation 10 or 20 years out in terms of how AI will be
impacting the marketplace?
00:01:39
Yeah. I mean, I'll
start actually, with the second question. You know, 10 to 15
years out, I think I expect it
00:01:44
to have large impacts. As many
of us know, especially those of us who are using these kinds of
technologies, you know, the
00:01:50
capabilities are fairly
impressive, given, you know, how rapid the progress has been, and
our ability to do things have
00:01:58
changed, right? Like I use it
day to day in my work as well. We— it's altered, kind of, the
ways that I carry out many of my
00:02:05
tasks. And I'm sure that's the
case in many white collar occupations, even at the early
stage of the technology is
00:02:11
relative to where it'll be in 15
years. You know, when we think about the job market impacts
right now, it's something that
00:02:17
me and my colleagues, you know,
are very curious about, because there's just a lot of
uncertainty. I mean, we have
00:02:23
uncertainty in the macro
environment generally right now, I think, but especially with
this technology. You know, we—
00:02:29
the tools are developing
rapidly. There is a big learning curve in figuring out how best
we can deploy them. And, you
00:02:36
know, even as we know that they
improve our individual productivity, there's a
organizational component of how
00:02:43
do firms and organizations
actually reallocate tasks and move things around in a way that
the firm is able to also
00:02:49
increase the productivity of
the aggregate. Is that— at least
in the short term, the important
00:02:54
question right now is just how
much it can have an impact. How does it change somebody's job on
a week to week, month to month
00:03:02
basis, in terms of, you know,
obviously the help they can get from things like ChatGPT, but
how it may change some of the
00:03:09
dynamics of what their job
actually becomes? Yeah, absolutely. I mean, we
think about jobs, in my field,
00:03:16
oftentimes, as like a bundle of
tasks or skills, right? And what we have is this new technology
coming in, and you're changing
00:03:23
the value or in changing how we
do some of these different tasks, right? So if, for
example, you're a copywriter,
00:03:30
right, perhaps some of the— the
tasks you've done previously about actually writing some of
these lines becomes automated,
00:03:36
and you have to reallocate your
tasks and your time in a different way, right? And so
some of those are done at the
00:03:41
individual level. Right? Again,
when I'm playing around with these tools and trying to figure
out how I can best use them for
00:03:47
my research, I'm reallocating on
my end. And then again, you know, when we're thinking about
the job market, a big question
00:03:54
mark right now is, how do firms
move things around? And how are they understanding where they
want to be reallocating their
00:03:59
workers and changing their
hiring. What does this mean, then, for the
students moving forward? And I
00:04:06
kind of open it by saying, you
know, what is this going to mean for the type of degree that
they're going to want to get?
00:04:12
Are they going to want to pursue
a master's degree, if there is— I mean, it's kind of that
balance between— and maybe we
00:04:19
don't know yet, how much of this
is still going to be human and how much is going to be
technology?
00:04:27
Absolutely. I mean,
I think there is a lot of uncertainty here. Obviously, as
a business school professor, you
00:04:32
know, I want to be bullish on
master's education in general, and I am. But I think it's— you
know, it's a challenge for us as
00:04:39
educators to think about how we
can retain the value in the degree. You know, just as
with other technology changes in
00:04:49
the past, you know, the kind of
work will shift. I know when I talk to my colleagues, we may
care a little bit less about the
00:04:57
case write-ups that the students
are doing right now. Because, you know maybe it's pretty
trivial to come up with a
00:05:01
reasonably high-quality case
write up with ChatGPT. But, you know, I want to understand how
the students are thinking
00:05:07
creatively. I want them to work
on their communication skills. The kinds of things that I think
are a little bit more protected,
00:05:13
I guess, at this point, from
these technologies. Whereas there may be other skills that
are a little bit more exposed.
00:05:20
Is this, then, the beginning of a
path where we will see change continuing to occur? And it
always does. But when you think
00:05:30
about how AI is going to factor
in. And then, you know, there are probably, you know, a
variety of jobs where quantum
00:05:35
computing, you know, in the next
several years, is going to be impacting those jobs as well.
And it feels like we're starting
00:05:41
to get into a pattern where
technology is going to play a larger role in everything we do.
00:05:47
I think that seems to be the
case. I mean, when we think about previous generations of
technology change that may have
00:05:55
led to automation, we oftentimes
think of things like manufacturing, right? And I
think what's different here,
00:06:00
perhaps, is that, you know, this
is affecting a lot of what we think of as high skill work. You
know, white collar occupations
00:06:08
where cognitive skills and
communications come into play. And that obviously opens up, you
know, a much larger swath
00:06:14
of the economy, in the labor
market, to be thinking about these effects manifesting in.
00:06:19
Are you seeing sectors right now that maybe
aren't as affected by AI as others right now, and whether or
not they're just playing catch
00:06:31
up at the moment? Or maybe they
are going to be a little bit devoid of having that serious AI
impact moving forward?
00:06:41
Yeah. As of right now, you know,
I think one of the things that people have talked about a lot
with these new technologies is
00:06:47
that they seem to be general
purpose technologies in the sense that they can have impacts
in many sectors and many
00:06:52
settings. And I suspect over the
long run, we'll see that. In the short run, you know, definitely
there are some industries and
00:06:58
sectors paying more attention to
this, and that have just figured out how to use these tools
better. Or perhaps, maybe
00:07:04
another way to put it is, the
tools are better suited for the task, right? And so programming
is, like, a really easy one to
00:07:09
think about, right? Both in
terms of people who are in these technology industries are very
aware of what's happening here,
00:07:15
and these tools are very good at
many of these coding kinds of tasks. So I think we could— you
know, if we're imagining where
00:07:22
we might start seeing some of
these effects pop up first, I wouldn't be surprised if we see
them, you know, start showing up
00:07:28
in kind of coding and
programming jobs, and then kind of slowly diffusing out into
other areas. But again, even
00:07:35
there, I think there's this—
there is this question of, like, trying to figure out the
organizational implications. And
00:07:40
one example I'll just toss out
is that, you know, there's— there's some research that has
been done, or evidence that can
00:07:46
suggest that, you know, many of
these AI tools can carry out tasks that can be done by entry
level consultants or things like
00:07:52
that. But the problem is, if you
automate those tasks, you're not helping your entry level
consultants develop the skills
00:07:58
that they need to kind of move
up in the organization and do things that the AI can't do,
right? And so I think these are
00:08:04
the kinds of questions that
firms are thinking through. And you know, I suspect we're going
to see a little bit of a— some
00:08:09
time before we have good answers
to them. I'll finish up on this. What do
you think, then, is the longer
00:08:15
term message that needs to be
sent to people heading into the workforce, whether you're a
college graduate or you're— you
00:08:22
know, looking for a blue collar
job, whatever it is. And on top of that, how, then, potentially,
does that impact how you and
00:08:28
your colleagues are thinking
about the type of research that you're going to conduct?
00:08:32
Yeah. I mean, so the advice that
I give my students, and maybe this feels like a very basic
piece of advice, is, you know,
00:08:40
if you're worried about the
concerns of AI, just use these tools a lot, first. And I
think that's— that's kind of the
00:08:45
first order condition to
understanding how to respond to them, is just understanding them.
What are they good at? You know,
00:08:51
where can they help me? Where
are they going to make my skills redundant? And then try and, you
know, push towards the things
00:08:56
that you think you can actually
bring value on. For our research, you know, it's certainly
something that I think people in
00:09:03
my field are also interested in
and concerned about. There was an interesting presentation I
saw recently where, you know,
00:09:10
they were analyzing working
papers, research in progress. And they're saying, you know, a
large majority of papers now, or
00:09:17
there's a big increase in influx
of research that may use these tools too. And obviously there
are questions around quality
00:09:24
and, you know, kind of the
provenance of what comes out of this. So I'm trying to follow
the same advice I'm giving my
00:09:30
students. I'm using these a lot.
I'm trying to figure out, you know, where can I be effective
on using these tools to
00:09:36
improve the quality of the
product and improve my productivity? But it's still a work in
progress, certainly on our end too.
00:09:42
Manav, appreciate your time and
your insight. All the best. Appreciate it. Thanks, Dan.
00:09:46
Thank you. Manav Raj, Assistant Professor of Management here at
the Wharton School.