00:00:00
generative AI can simultaneously help
00:00:03
businesses and customers
00:00:06
create more value
00:00:08
at lower cost the effects are going to
00:00:12
be more exponential than later and I
00:00:16
think for for me that last part
00:00:18
suggests the theory about you know is
00:00:22
this a new form of disruption that we
00:00:24
have not seen
00:00:26
it it starts looking more promising
00:00:30
welcome to the ripple effect the podcast
00:00:32
that takes you on a journey through the
00:00:34
minds of work and faculty I'm your host
00:00:36
Dan Loney and in each episode we'll be
00:00:38
diving deep into the inspiration behind
00:00:41
the groundbreaking research that Wharton
00:00:43
professors have conducted and exploring
00:00:45
how their findings resonate with the
00:00:47
world today we'll be covering a diverse
00:00:50
range of topics bringing you the latest
00:00:52
insights and knowledge that you can
00:00:54
apply to your life into work so get
00:00:56
ready to dive into new ideas with the
00:00:59
ripple effect
00:01:01
Rahul there's such a a large
00:01:03
conversation going on right now around
00:01:06
open AI chat GPT and when you think
00:01:10
about
00:01:11
management and and companies
00:01:14
a lot is still to be determined but it
00:01:17
seems like there is the potential for a
00:01:20
lot of thought to go on right now about
00:01:22
what what might what the impact might be
00:01:25
in the in the years ahead
00:01:27
absolutely Dan and uh and great to
00:01:30
connect with you on this very important
00:01:31
contemporary topic
00:01:33
um
00:01:34
yeah you know as somebody who has done
00:01:36
research on deceptive Technologies and
00:01:38
business models for almost 20 years
00:01:40
uh chat GPT generative AI more broadly
00:01:45
uh is truly emerging as what I would
00:01:48
call a general purpose technology like
00:01:51
semiconductors in the 50s and the 60s
00:01:53
like the internet in the 1990s
00:01:58
um and uh clearly they are you know
00:02:02
emerging across multiple use cases
00:02:05
multiple applications
00:02:07
and so there is a big what I call an era
00:02:11
ferment that all of us are in where they
00:02:14
are multiple use cases multiple business
00:02:18
cases
00:02:20
multiple
00:02:21
application domains uh with enormous
00:02:24
potential uh for growth and disruption
00:02:27
across the economy
00:02:30
so would it be safe to say uh talking
00:02:32
about the history you just laid out
00:02:34
there with semiconductors uh and the
00:02:37
internet is this kind of a big pivot
00:02:40
moment then for a lot of these companies
00:02:42
when you think about how those
00:02:44
components Changed History they they had
00:02:47
such a great impact
00:02:49
yeah and and I think uh if you look at
00:02:51
you know semiconductors and internet as
00:02:54
two of those uh disruptive forces that
00:02:58
we have lived through
00:03:00
um what is similar between those and
00:03:04
generative AI is that uh they are
00:03:08
solving two types of problems right one
00:03:12
is a problem around
00:03:14
productivity slash efficiency you know
00:03:17
can we do tasks more efficiently whether
00:03:21
it's time or cost as you know internet
00:03:23
and semiconductors you know really boost
00:03:25
the frontier around those issues and the
00:03:29
second is can we find new ways to create
00:03:32
value whether it's in Enterprise side of
00:03:35
things or it's in the consumer side of
00:03:37
things
00:03:38
and generative AI seems to check both of
00:03:41
those boxes if you think of many markets
00:03:45
and professions uh generative AI is
00:03:48
going to make things more efficient
00:03:50
uh at the same time it's going to
00:03:53
present many cases of additional value
00:03:56
being created through those and I do
00:03:58
think
00:04:00
um these parallels suggest that uh this
00:04:03
is not a not a three year or a five-year
00:04:06
phenomenon and this is a generation wide
00:04:09
phenomenon that we will all be living
00:04:12
through for the next 20 30 years
00:04:14
so what are the challenges then when you
00:04:16
have those those massive changes from
00:04:20
these different entities how do does
00:04:22
management and how do corporations react
00:04:25
to that and and and kind of does it
00:04:28
change the process to a degree of how
00:04:30
these companies organize and run yeah no
00:04:33
great great question you know it's it's
00:04:35
a commentary that I'm I'm seeing more
00:04:38
and more coming through in the popular
00:04:40
and the business price uh and there's
00:04:42
clearly discussion about you know
00:04:44
different types of use cases different
00:04:45
types of Industries different types of
00:04:47
professions who are getting impacted or
00:04:51
going to be impacted uh very
00:04:53
significantly right things like customer
00:04:55
service content creation as you know
00:04:58
very well software development uh
00:05:01
internet search data analytics
00:05:04
um and there's enough evidence that they
00:05:06
are going to be
00:05:08
impacting them uh in a pretty short
00:05:12
order and then you have you know what I
00:05:14
call on recombinations so these are not
00:05:19
uh
00:05:20
markets where there's an existing way of
00:05:23
doing things and this new technology can
00:05:25
do it better
00:05:26
but recombinations is where new use
00:05:29
cases can emerge through those new
00:05:31
technologies as building blocks you know
00:05:34
so think about
00:05:35
chant gbt a combination of virtual
00:05:39
augmented reality as a completely new
00:05:42
use case that you and I have not really
00:05:44
witnessed before
00:05:46
so I think from my perspective uh you
00:05:50
know when it when it gets down to what
00:05:53
leaders and and managers can do is to is
00:05:58
to recognize that these paradigmatic
00:06:00
shifts uh don't take place in two years
00:06:04
three years five years right internet
00:06:06
emerged in the 1990s
00:06:08
Amazon as the poster child of the
00:06:11
success of Internet uh was successful 20
00:06:15
years later you know semiconductors
00:06:17
emerge in the 60s Intel as opposed to a
00:06:20
child for a semiconductor leader was
00:06:22
only
00:06:23
visible in the 1990s right so
00:06:27
so so the good news is that this is not
00:06:30
going to be a disruptive change that's
00:06:33
going to happen overnight
00:06:34
and that means there is no need to panic
00:06:37
you know if you are a leader if you're a
00:06:38
manager uh but there is a clear and
00:06:42
critical need to understand
00:06:44
what generative AI or chat GPT uh can do
00:06:49
in terms of your business and I think
00:06:51
this is where most of my research and
00:06:54
teaching at the Wharton School
00:06:56
um
00:06:57
engages in terms of a framework that
00:07:00
leaders and managers can use and I'll
00:07:02
just give you a a very short
00:07:05
um aspect of that framework right so if
00:07:07
you are a leader or even an entrepreneur
00:07:10
in an emerging business
00:07:12
um you have to think about the Assets in
00:07:14
the business model that your that your
00:07:16
firm is is engaging with and then you
00:07:20
have to think about these shifts taking
00:07:22
place in the technology landscape chat
00:07:25
GPT for example
00:07:27
and what aspect of your assets and
00:07:29
business models uh could complement uh
00:07:33
what shared GPT could do so if you are
00:07:36
in the healthcare service uh maybe there
00:07:38
are certain things that would continue
00:07:40
to be done the way they are done but
00:07:43
then chat GPT or generative AI becomes
00:07:46
an additional value-added service for
00:07:49
for the patients for for uh for your
00:07:53
clients and there might be other cases
00:07:55
where uh it could replace right so if
00:07:58
you are in the content creation business
00:08:00
and there's certain aspects of content
00:08:03
creation that could be fundamentally
00:08:04
replaced through generative AI models
00:08:07
and and I think that's a starting point
00:08:09
is how your existing assets and business
00:08:12
model are going to be shape and not
00:08:15
framing it as a negative threat but
00:08:17
think about where it can complement and
00:08:20
where it might replace
00:08:22
a very parallel thought process that I
00:08:26
engage with Dan uh with students and
00:08:29
Executives is thinking on the customer
00:08:32
side as well you know you have existing
00:08:35
customers and then you're looking for
00:08:36
growth as new markets and new customers
00:08:39
and existing customers are part of an
00:08:41
existing ecosystem and new customers may
00:08:43
require you to build a new ecosystem
00:08:46
so once you start thinking about the
00:08:48
ecosystems you think about how is a new
00:08:52
technology like generative AI
00:08:54
affects your existing ecosystem or your
00:08:57
partners are your customers better off
00:08:59
or you need to invest in the ecosystem
00:09:02
to create a higher value for your
00:09:05
customers do you need to create new
00:09:07
ecosystems like electric cars you know
00:09:09
what Tesla has done and what uh the
00:09:13
success story there is it's not just
00:09:15
about the electric car it's about
00:09:16
building the whole ecosystem so I think
00:09:18
from my perspective uh this is a process
00:09:22
of disruption that is not going to get
00:09:24
resolved in the next three to five years
00:09:27
the organizations and the leaders who
00:09:31
are going to do well are the ones who
00:09:33
are going to have a very thoughtful
00:09:34
approach
00:09:35
to how this shift is affecting their
00:09:38
business both in terms of the positives
00:09:40
but also in terms of the challenges
00:09:42
thinking about it not in terms of just
00:09:46
existing customers but New Market
00:09:48
opportunities
00:09:49
and not just focusing on the existing
00:09:52
value chain but the broader ecosystem of
00:09:54
how these companies create value
00:09:56
so if I can have you talk about that
00:09:59
customer component again for a second
00:10:00
because it's interesting there's been so
00:10:03
much conversation in the last few years
00:10:04
about the relationship between the
00:10:07
company and the customer about how the
00:10:11
company the industry can do a better job
00:10:14
of making that connection and keeping
00:10:16
that consumer longer term that this is
00:10:19
seemingly that next step in that process
00:10:22
isn't it
00:10:23
yeah no I think you know we've talked
00:10:25
about customer lifetime value and as we
00:10:29
have embraced more personalization uh
00:10:32
much of this is coming through data and
00:10:34
data analytics we are explicitly
00:10:37
thinking about you know how we can
00:10:40
enhance and maximize the value that we
00:10:43
can create for every customer and it
00:10:47
doesn't require a one-size-fits-all
00:10:49
approach right so we can generate things
00:10:51
that are more personalized we can create
00:10:53
and capture value for each customer or
00:10:56
entity over time and I do think you know
00:10:59
a generative AI uh would enhance that
00:11:02
you know it will broaden the scope
00:11:05
of what personalization might mean for
00:11:08
you personalization might be let's say
00:11:11
from a from a Content creation
00:11:13
perspective yeah for me it would also
00:11:16
include research perspective as well and
00:11:19
so I can imagine scenarios where
00:11:23
companies are able to create a suite of
00:11:27
personalizations and bundles through
00:11:30
generative AI
00:11:32
add a level of a customer at a level of
00:11:34
an Enterprise that was not possible
00:11:37
before primarily based on data analytics
00:11:40
so I think once we include
00:11:42
these new algorithms and these new
00:11:45
learning processes that are algorithmic
00:11:48
the scope of personalization
00:11:51
becomes much broader and that has
00:11:54
exponential effects in terms of success
00:11:56
of businesses I think it's issue when
00:11:59
you talk about this of how it will
00:12:01
impact companies but I think you can
00:12:03
even take it larger scale and think
00:12:06
about how it's going to potentially
00:12:07
impact Industries as a whole correct
00:12:11
absolutely absolutely in fact I was uh
00:12:13
on a panel this morning with fellow
00:12:17
academics from from other institutions
00:12:19
and that was our question uh how can we
00:12:23
think about the impact of a journey
00:12:25
purpose technology like generative Ai
00:12:28
and you know the conversation that I
00:12:32
strongly believe is true is it's gonna
00:12:34
have at least four different
00:12:37
points of impact
00:12:39
right the most obvious one as a strategy
00:12:41
Professor is for businesses and for
00:12:45
Industries
00:12:46
but it will also start having an effect
00:12:48
on professions
00:12:50
and societies
00:12:53
might a much broader level right so
00:12:55
think about what social media has done
00:12:57
to our societies to think about what
00:13:00
internet has done to our society so I
00:13:03
think each of these levels starting from
00:13:05
organizations to Industries to broader
00:13:09
professions and even the social fabric
00:13:12
are likely to be impacted through these
00:13:15
technology emergence
00:13:18
some of this is going to be positive and
00:13:21
value creating and some of it we have to
00:13:22
be careful yeah especially with such a
00:13:25
strong human interface and a modality
00:13:28
that uh these these Technologies uh are
00:13:33
human-like
00:13:35
um I think there has to be a lot of care
00:13:36
that needs to be exercised in terms of
00:13:38
how professions and societies are going
00:13:41
to be impacted by these emergencies so
00:13:43
part of that it would seemingly from
00:13:46
what you said there is you know how
00:13:48
companies think about their structure
00:13:50
the levels of jobs the types of jobs
00:13:53
that they will have what jobs you know
00:13:55
have the opportunity to have open AI
00:13:58
chat GPT benefit from uh from that
00:14:02
technology what in some cases may even
00:14:05
replace uh certain jobs along the way as
00:14:07
well
00:14:08
absolutely and and you there is a lot of
00:14:11
research it's very early stage so we
00:14:13
don't have a definitive statement uh on
00:14:16
that issue uh but when you think about
00:14:20
every new technology right the model
00:14:22
that we all prescribe to as as
00:14:24
researchers of of technology is every
00:14:28
technology follows an s-shaped curve
00:14:31
right early stages of a technology it's
00:14:35
imperfect it's not cost effective we
00:14:38
have to invest in resources and it takes
00:14:41
time for it to improve but once the
00:14:45
technology gets well understood and the
00:14:47
experimentation allows us to fine-tune
00:14:50
how the technology actually works then
00:14:54
we have a very fast robust trajectory of
00:14:57
improvement kind of what we saw on
00:14:59
semiconductors in the 70s or the 80s
00:15:02
what we saw on the internet era post
00:15:05
2000 what we're seeing with electric
00:15:07
cars in the last three five years
00:15:10
and so I think from my perspective
00:15:13
many different professions are going to
00:15:16
be affected I think it's uh it's a
00:15:18
bygone conclusion
00:15:19
the rate and scale of what that would
00:15:24
look like would be a function of those S
00:15:26
squared trajectories across the
00:15:29
different professions you know if you
00:15:31
think about
00:15:33
you know let's think about uh the
00:15:36
professions that are tied up in customer
00:15:37
service versus professions that are tied
00:15:41
up in Creative Industries you're likely
00:15:44
to see different models of s and
00:15:48
different rates of of growth depending
00:15:50
on how the technology is going to
00:15:52
improve and achieve the performance
00:15:54
requirement for those professions to be
00:15:58
substituted by these emerging modalities
00:16:00
we will obviously see this play out in
00:16:03
the years ahead but as you were
00:16:04
describing that
00:16:06
my thought is is that the level and the
00:16:10
the sheer volume of companies and
00:16:13
industries that could be impacted this
00:16:15
it feels like it is extremely large and
00:16:19
it will be realistically on the company
00:16:21
themselves or maybe the industry to
00:16:24
understand how open AI chat GPT will be
00:16:27
able to be incorporated within the
00:16:29
structure of their firm moving forward
00:16:32
yeah and absolutely and and just the way
00:16:35
you describe the problem it's it's it's
00:16:37
uh it's a very complex problem right and
00:16:40
and
00:16:41
um you know when uh when I talk about
00:16:43
how companies and leaders can manage
00:16:46
disruption of a skill like what we're
00:16:48
talking about today we generative AI uh
00:16:52
there there are three impediments
00:16:55
that companies have to deal with in the
00:16:58
shot right the first is uh there is
00:17:02
significant uncertainty in terms of
00:17:05
what the dominant use cases are going to
00:17:08
be what Solutions are going to satisfy
00:17:11
those use cases what business model
00:17:14
would allow for companies embracing
00:17:17
generative AI to actually do it in an
00:17:19
economically viable right so there is
00:17:21
there's many sources of uncertainty that
00:17:25
every era of disruption seems to present
00:17:27
so you are you are in an environment
00:17:29
where there are more unknowns than nodes
00:17:33
secondly you still have to run your
00:17:36
existing business that's not going away
00:17:38
tomorrow often this is going to be a
00:17:41
slow process of evolution three five
00:17:45
many cases longer years
00:17:48
um so you have to think about how do you
00:17:51
want to still focus on the old business
00:17:53
or existing business why
00:17:56
exploring these new possibilities
00:17:59
and the third is the adjustment cost
00:18:03
that you are making in terms of building
00:18:05
new skills building new capabilities uh
00:18:08
dealing with new competitors that you
00:18:10
may have to deal with so once you
00:18:13
introduce inertia which is how you
00:18:16
change and the adjustment cost once you
00:18:19
introduce uncertainty and once you
00:18:21
introduce the balance between the
00:18:23
existing and the new uh it is a very
00:18:27
very hard nut to crack
00:18:30
I'll I'll tell you what I've seen in in
00:18:33
my research
00:18:34
the best time right to explore something
00:18:38
that's uncertain and disruptive
00:18:41
is when you're not desperate for it to
00:18:45
be adopted right the best time to invest
00:18:47
in a new disruptive solution or
00:18:49
opportunity is when you are running a
00:18:51
healthy business
00:18:53
and that's where the pressure is going
00:18:55
to be the lowest for you to take more
00:18:57
risk and explore more broadly
00:18:59
you know the second aspect of the
00:19:01
management uh or leadership that I see
00:19:04
that generates successful outcomes is to
00:19:07
think about what sort of an
00:19:09
organizational structure but allow for
00:19:12
companies to take advantage of these
00:19:15
disruptive opportunities and you know
00:19:18
it's it's often talked about that they
00:19:19
need to be managed as separate
00:19:21
businesses or separate units
00:19:23
um if you see what Google is doing with
00:19:25
alphabet there's a Google business and
00:19:27
then there is a WeMo business which is a
00:19:29
self-driving technology
00:19:31
um I think that's one way to do it I
00:19:33
find Partnerships alliances with many
00:19:36
companies to share risk to share cost to
00:19:40
learn from each other is a model that I
00:19:43
think is a very effective model
00:19:44
especially when it's still uncertain
00:19:48
um and then third I think we need
00:19:51
patience why we need to be disciplined
00:19:55
in terms of how we experiment and
00:19:57
innovate as well so I think a
00:19:59
combination of
00:20:01
how do we balance existing with Neo
00:20:04
how do we engage in organizational
00:20:06
configuration that allows us to maintain
00:20:10
that balance and then having a
00:20:13
disciplined approach to how we are
00:20:15
experimenting around these new
00:20:17
technologies and business models and
00:20:19
giving them enough of a Runway to take
00:20:21
off as opposed to switching back and
00:20:24
forth between something new versus
00:20:27
something else every year every two
00:20:29
years that generates more confusion more
00:20:32
inertia more cost and tends to be
00:20:35
counterproductive let me finish up with
00:20:37
with one question is it it seems like we
00:20:41
have had disruption in our corporate
00:20:44
structure for such a long period of time
00:20:46
at various levels with various impacts
00:20:49
but this feels like that this is a
00:20:51
disruption
00:20:53
that is kind of taking us down A New
00:20:55
Path and it's almost disrupting what we
00:20:58
think the idea of disruption is that
00:21:00
we've kind of made this massive shift
00:21:03
potentially as we move forward
00:21:06
yeah I think the
00:21:07
thing that the juries out there in my
00:21:09
view Dan uh you know is this uh is this
00:21:12
a new wine if I might use that analogy
00:21:14
or this is an old wine in a new bottle
00:21:17
and um you know whenever there is a
00:21:21
widespread uh paradigmatic change like
00:21:25
the internet like the semiconductors
00:21:29
um those conversations emerge and I see
00:21:32
with generative AI or chat GPT more
00:21:34
specifically I'm seeing there are those
00:21:37
two camps within Academia as well
00:21:40
uh many people believe that uh yes the
00:21:45
the the technology is different uh the
00:21:49
the models of value creation are
00:21:51
different compared to what we saw with
00:21:52
internet or semiconductors but the the
00:21:55
nature of shifts in terms of assets and
00:21:58
business models and capabilities uh they
00:22:02
have been well understood and studied so
00:22:04
it is a wine that we understand and we
00:22:07
can we can understand what the
00:22:10
ramifications are
00:22:12
um and others believe that this is
00:22:14
fundamentally different it's a it's
00:22:16
disrupting the way we have thought about
00:22:18
disruption
00:22:20
and I think
00:22:21
and I think how to parse these two
00:22:24
perspectives in my view uh depends on
00:22:27
thinking through what the impact that
00:22:30
generative AI is having right in a
00:22:33
business context
00:22:34
uh it's clear that this is not just a
00:22:37
technology problem you know it is a
00:22:39
broader issue than just thinking about
00:22:41
the technology it's how it impacts
00:22:43
organizations uh in terms of the jobs
00:22:46
but also in terms of how organizations
00:22:48
create value it impacts the business
00:22:50
models uh it can completely replace a
00:22:54
given business model
00:22:55
and find a completely new business model
00:22:57
at the same time right and it also
00:23:00
impacts the ecosystems the broader
00:23:02
ecosystems that many companies are part
00:23:04
of in terms of requiring a major
00:23:07
reconfiguration of the ecosystem or
00:23:10
meeting many old ecosystems completely
00:23:12
redundant
00:23:13
then on top of it the the generative AI
00:23:17
technology is allowing us to do two
00:23:20
things simultaneously
00:23:23
uh that has been very hard to do
00:23:25
historically which is
00:23:27
we can use simulative AI to find use
00:23:31
cases that create more value for the
00:23:35
specific customer
00:23:37
or a market at the same time we can do
00:23:41
so at a much lower cost yeah and I think
00:23:44
because generative AI can simultaneously
00:23:48
help businesses and customers
00:23:51
create more value
00:23:54
at lower cost the effects are going to
00:23:58
be more exponential than linear
00:24:01
and I think for for me that last part
00:24:04
suggest the theory about you know is
00:24:07
this a new form of disruption that we
00:24:10
have not seen
00:24:11
it it starts looking more promising not
00:24:15
so much that we have not seen a flavor
00:24:17
of that before I think internet had
00:24:19
flavors of that in very specific markets
00:24:22
like social media and others but I think
00:24:25
it's not happened at the scale
00:24:28
across markets across Industries across
00:24:32
professions
00:24:34
across societies
00:24:36
at the pace that we think a generative
00:24:38
is going to uh going to go into
00:24:41
propagate so I think for me uh the scale
00:24:45
of the effect with exponentiality and
00:24:49
the linkages between technology
00:24:51
organization business models ecosystems
00:24:55
and professions
00:24:56
is going to create a much larger
00:24:58
Multiplex effect some of it we have
00:25:01
understood and seen before but some of
00:25:04
it remains to be studied and remains to
00:25:06
be understood
00:25:08
thank you for listening to the ripple
00:25:10
effect we hope you found this episode
00:25:11
informative and engaging don't forget to
00:25:14
subscribe and leave us a review so that
00:25:16
we can continue to bring you the best
00:25:18
Insight from the Wharton School
