00:00:00
Well, with artificial intelligence being
00:00:02
a growing part of our lives, both in
00:00:04
personal and in business, many questions
00:00:06
are being asked about areas like
00:00:08
productivity growth and just how much of
00:00:10
an impact we will see from this area in
00:00:13
the future. A new report out by the Pen
00:00:15
Wharton budget model takes a deeper dive
00:00:17
into these questions and pleasure to be
00:00:19
joined once again by Kent SMEs who is
00:00:21
faculty director of the Pen Wharton
00:00:23
budget model. He's also professor of
00:00:25
business economics and public policy
00:00:26
here at the Wharton School. Hi Kent, how
00:00:28
are you today?
00:00:30
>> Good and great to be back.
00:00:31
>> Thank you. All right, so let's start
00:00:33
with just how important you think it is
00:00:36
to project this productivity path
00:00:39
especially with how AI and technology
00:00:41
are having such a greater role.
00:00:43
>> Yeah, absolutely. And you really see
00:00:45
that with market concentration right
00:00:47
now. Um the bottom 493rd
00:00:51
firms as uh the Apollo chief economist
00:00:54
to Slack is documented that their
00:00:57
earnings projections really haven't
00:00:59
changed but it's almost all the markets
00:01:01
driven by these seven firms that
00:01:03
magnificent seven um and that's you know
00:01:06
mostly an AI play and so AI is having a
00:01:09
very big um kind of uh salient you know
00:01:14
uh impact the real question is how broad
00:01:17
is it going to
00:01:18
across the economy and then um really
00:01:22
can AI help deal with this major debt
00:01:27
problem that we have of of growing debt
00:01:29
and that was the big motivation. And so
00:01:30
what we basically did is we look at a
00:01:32
bunch of different occupation groups and
00:01:35
really based on other experts and their
00:01:38
insights about uh the actual data and
00:01:41
how to interpret things and everything
00:01:43
from office administrative support
00:01:47
functions the business and financial
00:01:49
operations, computer and math
00:01:51
occupations. Those are the ones who are
00:01:53
uh most likely the greatest impact by
00:01:56
AI. and then all the way down to like
00:01:58
farming, fishing, construction and so
00:02:00
forth, buildings and groundskeeping. Uh
00:02:03
those are the occupations are the least
00:02:05
uh likely to be impacted. And so then
00:02:07
the big question is how much impact will
00:02:10
this have in the macro economy and then
00:02:12
how much impact could this potentially
00:02:14
have on government revenue um and then
00:02:17
ultimately potentially even the
00:02:18
government debt over time. And there's
00:02:20
been, you know, this belief by policy
00:02:22
makers that we're in this new era that
00:02:25
AI, we don't have to be physically
00:02:27
responsible because AI is going to solve
00:02:30
everything. And what we're showing is
00:02:32
that um that that's simply not true.
00:02:35
We're not even nearly that's not even
00:02:38
close. Um and so they're making a lot of
00:02:41
you know uh incorrect uh uh assumptions
00:02:46
about the impact of AI. Even with those
00:02:48
incorrect assumptions, they're not then
00:02:50
mapping to the budget deficit problem
00:02:52
correctly.
00:02:53
>> Can you start to estimate then with so
00:02:56
many businesses bringing AI on board and
00:02:58
I think the expectation that many others
00:03:00
will just how much of our GDP will be
00:03:03
impacted by AI.
00:03:05
>> Yeah, it's not a small uh uh amount. So
00:03:08
if you just look at the metric of uh of
00:03:12
jobs that at least 50% of the task and
00:03:15
so we have this taskbased model that is
00:03:19
uh very similar to Don Ace Mloo recent
00:03:21
Nobel Prize winner in economics his
00:03:23
taskbased model that he used for
00:03:25
modeling AI we get bigger results than
00:03:28
he got uh for the impact of AI but
00:03:30
nonetheless uh if you break it down
00:03:33
literally at the task level within each
00:03:35
these occupational grips I basically
00:03:38
about 40% of occupations or at least uh
00:03:44
based on the employment in those
00:03:46
occupations are going to have at least
00:03:48
50% of their task that will be
00:03:51
essentially replaceable in the future uh
00:03:55
by AI. So it's not a small impact by any
00:03:58
measure. um you know the uh primary
00:04:01
author on this report out as the the
00:04:04
supervisor of this report um Alex Arnon
00:04:07
he I think he put it really well uh to
00:04:10
uh recently he says it's it's it's it's
00:04:12
roughly the impact of email um in
00:04:15
particular email definitely had a big
00:04:17
impact um yeah but it's not and if you
00:04:20
look across the occupations lots of
00:04:23
occupations email email like technology
00:04:26
um definitely had an impact grow We used
00:04:29
email this morning to communicate before
00:04:31
this. And so, uh, but at the same time,
00:04:34
it's not one of those things that's like
00:04:35
this magic bullet either. It's not
00:04:38
electricity. It's it's not
00:04:39
refrigeration. It's not that
00:04:42
transformative. And so because of that,
00:04:45
I is there an element of as AI comes in
00:04:49
and the the level of productivity that
00:04:51
we see grow in the years ahead that it
00:04:54
becomes to a degree kind of a known
00:04:56
quantity just how much we're going to
00:04:59
see the growth occur because companies
00:05:02
are going to add it and you know the
00:05:04
technologies will change as we move
00:05:05
forward. But once you install that
00:05:07
technology I think the expectation is
00:05:09
you're going to do your job faster.
00:05:11
you're going to have better
00:05:12
productivity, but it's not going to, you
00:05:15
know, marketly change yeartoear. You're
00:05:17
going to kind of reach a level and
00:05:19
continue to grow maybe incrementally.
00:05:21
>> Yeah, that's right. And so almost like
00:05:24
so one of the things that uh uh we did
00:05:26
in this study is we looked at adoption
00:05:28
of past technologies those kind of
00:05:29
growth curves and rather fascinating um
00:05:32
and we're talking about really the
00:05:35
technologies as opposed to things that
00:05:37
like the invention of the transistor or
00:05:40
you know refrigeration you know these
00:05:42
general purpose big jumps that uh have
00:05:46
much bigger impacts. If you look at
00:05:47
these adoption curves of technology and
00:05:50
everything from the from the personal
00:05:52
computer to the smartphone to the
00:05:54
internet to cloud computing rather
00:05:56
shockingly their adoption curve are very
00:05:59
similar to each other where you get this
00:06:01
what kind of call turnpike theorem like
00:06:04
stuff with pretty quick adoption but you
00:06:07
can only go so far um you know 100% is
00:06:10
is the upper bound and so many uh
00:06:13
occupations are going to be way uh less
00:06:15
than that some occupations are going to
00:06:17
quite a bit more than that than you know
00:06:19
50% you know we have office
00:06:22
administrative support functions at 75%
00:06:24
of uh their their task will eventually
00:06:27
be replaceable by AI whereas building
00:06:30
and grounds clean and maintenance
00:06:31
operation we said it's about 2.6% 6% and
00:06:34
we we literally if you're people are
00:06:36
curious broken that down by over 700
00:06:39
different occupations um and then we
00:06:42
summarize it in big occupational groups
00:06:44
but I think it is uh the case that you
00:06:47
have quick adoption like you do with uh
00:06:50
internet and other technologies but then
00:06:52
yes you get this leveling effect as a
00:06:54
result of that what we find is that
00:06:56
during the first decade it will increase
00:06:59
GDP by about 1 and a half% by you you
00:07:02
know, within 30 years maybe 3% in the
00:07:05
long run by 2075 maybe 3.7%. So you're
00:07:08
getting this leveling effect. Not
00:07:10
everybody's been happy with this. It's
00:07:12
bigger than some you know results
00:07:14
although we're kind of the first to kind
00:07:16
of map it into this kind of this uh type
00:07:18
of change but nonetheless it is
00:07:21
especially with the budget uh issues but
00:07:24
nonetheless it is uh bigger than some
00:07:27
estimates but definitely smaller than
00:07:29
the rhetoric that we are hearing about
00:07:31
AI. And so what kind of impact might you
00:07:34
see play out with the federal budget as
00:07:36
we move forward here in the next
00:07:38
>> and this one is very much a work in
00:07:40
progress because the first step was on
00:07:43
productivity
00:07:45
more generally speaking and then GDP and
00:07:47
wages uh wage growth and so forth. The
00:07:51
reason why it's harder to map that back
00:07:53
to the budget is that there we have to
00:07:55
go category by category. So think about
00:07:58
healthcare. Healthcare can actually be
00:08:00
improved by AI uh for certain things
00:08:02
like scans and reading scans. I call it
00:08:05
the TSA attack. When you go to the
00:08:06
airport, you don't have 50 people at
00:08:09
the, you know, um the TSA uh line
00:08:12
looking at your bags, but it's only
00:08:14
takes a couple people because we have
00:08:16
very modern scanners uh right now. And
00:08:19
so, uh you can imagine some innovations
00:08:22
uh especially for diagnostics, breeding,
00:08:24
and so forth. But at the same time in
00:08:27
healthcare
00:08:28
it's going to AI overall is going to
00:08:30
increase cost simply because it's going
00:08:31
to allow us to cure more things and
00:08:33
that's a good thing but there's costs
00:08:35
and so overall what we're finding for
00:08:37
the first decade I would call this maybe
00:08:40
upper boundish maybe about $400 billion
00:08:44
in uh uh deficit reduction uh through
00:08:48
expended tax base but that is uh that's
00:08:51
a really rough first cut of all this and
00:08:54
we're going to be looking at it going
00:08:55
forward.
00:08:56
>> Kent, great to talk to you as always.
00:08:58
Thanks, sir.
00:08:59
>> Pleasure, sir.
00:09:00
>> Thank you. Kent SME, faculty director of
00:09:02
the Pen Wharton budget model and also
00:09:04
professor of business economics and
00:09:05
public policy here at the Wharton
00:09:07
School.
