Welcome to Episode #255 of CxOTalk. And, we are speaking about the pharmaceutical
industry. We’re talking about healthcare and patients,
and the relationship between drug companies and patients. And, the business model of drug companies,
and the societal, cultural, political pressure. This is going to be a fascinating show. I’m Michael Krigsman. I’m an industry analyst and the host of
CxOTalk. I want to just say a brief “thank you”
to Livestream for supporting us with video infrastructure. And, if you go to Livestream.com/CxOTalk,
they will give you a discount on their plans! So, do that. We have two extraordinary guests on our show
today. And, I want to introduce first Richie Etwaru,
who is the Chief Digital Officer for QuintilesIMS, which is a huge company. Richie Etwaru, how are you? I’m doing fantastic, man! Thanks for having me on the show and, you
know, every time I’m here, I thank you for what you do. You drive a great conversation, and we all
learn from listening to each other. So, thanks for that! Well, thank you! This is your third time on the show. Richie, briefly tell us about QuintilesIMS
and tell us what you do there. Very quickly, QuintileIMS, we are the largest
partner for the life sciences industry, and we are what I like to call the “patron saint”
of taking the silos out of the industry. So, we’ve got services all the way from
molecule to market, and we work with our customers across that journey to make sure that we can
connect the organization horizontally as pharma transforms, which is what we are going to
talk about today. I love that! The patron saint of taking silos out of the
pharmaceutical industry. So, Richie, thank you for being here. And, we are also joined by Milind Kamkolkar,
who is the Chief Data Officer at Sanofi. And Milind, this is your first time. Welcome to CxOTalk! Yes, it is! I can’t wait to go through my initiation
rites on the show. Looking forward to it! So, Milind, tell us about Sanofi and what
does a Chief Data Officer at a pharmaceutical company do? Okay! So, let’s start with Sanofi because that’s
the easier question. Sanofi is a large pharmaceutical company;
biotech company; headquartered in Paris, France which is where I’m actually dialing in from,
today. They focus on a couple of key areas, mainly
vaccines, general medicine, specifically diabetes and cardiovascular, and also, with the acquisition
of Genzyme, we’ve entered the world of specialty in rare disease and oncology which I find
incredibly powerful. If you sort of take these three combinations
of areas past our device and manufacturing divisions for patients with diabetes, you’ve
got a really nice roundup. I should also say, I forgot the consumer health
division as well. So, you’ve got a nice round-up from, let’s
call it chronic care to specialty rare disease, to then supporting patients through devices
down to the consumer level when we’re in the consumer health world. I was going to say, you’re a huge, huge
company. Your revenue last year was what, like $33
billion? Yeah! Something like that. I mean, to be perfectly honest, I don’t track
it all the time. You know, I think being a Chief Data Officer,
it’s my fourth month over here now and I guess it’s sort of helped to turn… Part of the reason why I don’t get to spend
perhaps as much time as I’d like to on the financials… You know, the reason, I think, why Sanofi
decided to hire a Chief Data Officer was for one very clear reason. We have an endemic problem within the industry
and that is really because of the silos that have often been created over a period of time. We really produce three assets: One is product
— compounds; Two is a share price, particularly if you’re a publicly traded organization;
and Three, is information. Now, if you ask yourself what are you really
good at, I don’t think any shareholder’s going to say “You’re doing the best you can”
because who wants … you want to keep shares going up. Likewise, when it comes to product, you can
always produce a better pipeline. But, when it comes to information, that’s
the one critical piece that whilst we produce a lot of it, you can argue that the distance
between data to decision-making is still far than less desirable. So, my job at Sanofi as a Chief Data Officer
is really to help accelerate creating better decisions faster that are more relevant in
context across the spectrum, from R&D all the way through to commercial, and across
our business units. So, Richie Etwaru, when we talk about silos
inside the pharmaceutical industry, what are we actually speaking about and how does this
relate to the broader changes that are going on regarding healthcare and ultimately, the
impact on healthcare for people? For patients? Well, I think most of us understand silos
pretty well, right? These are vertical departments within large
companies that tend to perform a specific function. And then they perform that function well,
but if you look at multiple of those silos together as a broad organization, you’ll
see inconsistencies and gaps to be solved for. I think, with the pharmaceutical industry,
there was a time when it was okay to have silos. Not that you wanted them, but if you had them,
it was not the biggest deal in the world, and we’re seeing what I like to call the
“three waves” of change enter the environment of the pharmaceutical industry that’s creating
the financial reality and quite frankly, the competitive reality, to start to think about
what the business model looks like and take the silos out. These three waves, I think, most people would
recognize. Now, the first wave is what came from the
supply side. So, I think of this as the patent cliff, right? I think we’ve heard enough about the patent
cliff that the supply of discovery of drugs in the pipeline has changed quite a bit. And, the good thing about the patent cliff
is that it was sort of contained within the pharmaceutical industry, right? Yes, there was some implication too of the
stakeholders, but it didn’t radically change the landscape because each pharma company
was suffering from the same strain from a supply side. The second wave is what I like to call the
wave coming from the demand side. This is the influence and the pressures of
reimbursement being changed, payment terms being changed. I think what we see in the United States with
Obamacare and the model of delivery has created a tremendous amount of strain that created
a whole new wave of pressures for the pharmaceutical industry. So, I think that’s the second wave. Now, this was not as self-contained in the
industry. This included the patients; it included the
payers; it also included the government to kind of look at that. Where we are today is what I like to call
the “third wave” of change that’s coming through the pharma industry. And, this is the digital health and technology
paradigms that are entering at the same time. Now, the patent cliff is not completely solved
for. That’s still moving over, right? The changes in the payment are not completely
solved for. That’s still here. And, we have this new wave that’s being driven
by this new stuff that Milind is talking about, which is more data, digital health, some of
the new technology paradigms. What’s interesting about this third wave is
that it is not self-contained in the pharma industry or in healthcare. Now, you’re seeing new entrants start to enter
the competitive landscape. Apple is a good example, Amazon is a recent
good example, and this is creating the type of strain where we, as vendors in the industry,
have to start to look at what type of solutions we provide to our customers because it’s not
just that the competitive landscape is changing, and the pressures are changing, we are seeing
new entrants to the marketplace, which is going to drive disruption. I think Richie brings up a really good point,
there. You know, I often get asked the question:
“But Milind, you’ve only been here a few months, but what’s your observation in the
industry? You know, how do you see us benchmark?” I always have a quiet grin whenever that question
comes up because I reply with the following: “If you were to compare a brontosaurus and
tyrannosaurus rex, you could argue that they’re both dinosaurs. But when the comet hit, does it really matter,
you know?” And, I think what the industry is facing right
now is, in fact, that horizontalization. When we see consulting practices, when we
see folks coming in and often talking to us about their discipline in the pharmaceutical
industry, I almost have to argue “Well, how relevant is that when you’ve already
got most businesses already becoming technology businesses, and most technology businesses
{are] already eating away at our business?” And, what you’ve seen here now is this,
“I’ll take ‘cautious reluctance,’ if in many ways, to understand what the implications
of that are.” And I think the jury is still out. But one thing is clear; this is a comet and
there’s no getting away from it. And, the sooner we are to embrace that, I
think the sooner we are, in terms of really maximizing our business value to our customers
— namely the patients, physicians, payers, and regulators that we work with. And of course, consumers. So, what… Is there an answer where, where does this
need to go and how does the pharma industry; what does the industry need to do to get there? I think the issue here has really changed
in the last five years. When I first entered the industry, for example,
the name of the game was “How do I thrive ahead of my competition?” Right? If I think about some of the problems that
we’re solving for customers today, they’re starting to look like, “How do I survive? Right?” And I love that metaphor, Milind, on the comet
and the dinosaurs that you’re talking about. The crux of the matter today is not so much
how I engage with patients, or how I discover drugs, or how I figure out my pricing, or
my analytics. Those are all individual problems. But at the core of the problem is the notion
that the business model of the pharmaceutical manufacturer is starting to expire. If the three of us were given $3 billion today,
from (I don’t know) KKR, or TPG [Capital] to go start a pharma company to manufacture
treatments for those that are disenfranchised by health, I’m not sure we would pick the
model we have today that takes 8-10 years to discover a molecule and bring it to market. And takes $2.8-3.4 billion to do that. I’m not sure we would take that model. So, there’s a lot of pressure that’s starting
to wake up [to] the realization that the business model is no longer one that is profitable
and we’ve got to go at the business model. Milind, I’d love to hear your thoughts on
that because that’s sort of where I’m starting to anchor most of our investment. Look, I could agree more, Richie. And it’s one of those areas where, at the
crux of it, yes, the business model is changing. Right? Once upon a time, it was a rep-led — really
a rep-led, if you will — commercial model. I think it’s clear to say that model, particularly
for chronic care and general medicine, is, we’ve seen it already, eroding away. It doesn’t mean that sales reps aren’t
important. I’m still a big believer that at the end
of the day, when it comes to sales, relationships matter. Right? But, the nature of those relationships, the
nature of those engagements, the channels under which you take under; the also primary
market research, if you will, (Right?) to understand what do consumers really want. The timelines under which you operate; those
have dramatically shifted. Right? Where once upon a time, you would do a year-long
brand plan. You could argue, shouldn’t they be months
long now, right? Based on what we’re seeing out there. Likewise, in R&D, starting up a biotech company,
you could argue with some of the particularly garage buyer techs that are coming in that
you probably don’t even need $3 billion. You probably just need a cheap sequencing
machine, which is getting cheaper, as we know. You have some knowledge of biology. But, I think what’s emerging even more so
is the nature of technology in that pursuit of achieving a product that can hit the market. Now, in saying all of that, I don’t want to
be disrespectful to the regulators in this instance because clearly, there needs to be
a new ecosystem that’s evolving. And we’re all learning at the same time. So, I think the opportunity for us as an industry
body, if we were to conclude that this is a healthcare and life sciences industry, is
that we need to work with the regulators because this technology’s moving so fast. But yet, the regulators themselves are still
learning. So, what is true and appropriate may not necessarily
always be relevant. For example, I was having a discussion today. Cybersecurity is the new sexy. I mean, you’ve got to know cyber if you plan
to do digital health. Otherwise, the risk at which you put patients
and physicians in is significantly high. You can’t hire a music major like Experian
and hope that it would work. Michael, before you go, I want to touch on
this notion of time for a little bit. There is a construct of time, that not being
as important, that is left over in the life sciences industry, that we’re now starting
to wake up to. I’ll give you really simple examples. I touched on this notion earlier. It takes about eight to twelve years from
the moment you discover a molecule to the time you get it approved to be able to bring
it to a patient. It takes about eight to ten years to do that. You know, if you think about an automotive
company, let’s say, Mercedes, for example, from the day they start to draw the first
car when they do a full model design, all right? From the day they start to the day it rolls
out on the showroom, it takes about three years, okay? Now, that’s probably not as impressive as
my second example that I want to share with you. If you think about the first video of autonomous
driving, you guys remember that first video on the roof, all right? With those young kids with the car, with the
cones, right? When we first saw autonomous driving, to the
day when it was rolled out by Tesla and Google, you know, to be like commercially ready on
the road, granted, it needed permission, it was about five years. It took about five years to go from a couple
of young people on a laptop controlling a car with some cameras, to industrial-grade
autonomous driving. That is an insanely different way of looking
at time, as opposed to the life sciences industry, where we still do things in decades. And that part of it is dead. Yeah. I mean, if I could just add one last thing,
Michael, to that point, I think where the time, perhaps has that risk-averse nature
in the pharmaceutical industry is that perhaps I’m like… banking in other such industries… The reality is that if you get this wrong,
people die, right? And that’s the real crux of it. So, I can fully appreciate and respect the
fact that sometimes, you do want to be a little bit cautious because, of course, who wants
to create a medicine or some kind of patient service that really is not yielding a positive
outcome? So for one, I’m quite grateful for the fact
that we are basing more medication and pricing, and reimbursements, etc. really on health
outcomes. But I think, in many ways, it’s not just the
actual medicine that has to take into play, it’s also the customer experience at point
of treatment that needs to be part of that equation as well. This is a very fascinating discussion where
you talk about customer experience at point of treatment. And, let’s come back to it because we’ve
got some questions on Twitter already. And, I want to remind everybody, go to Twitter,
use the hashtag #cxotalk and you can ask our guests questions and they will respond! It’s a great opportunity to get access to
these folks who are usually hard to reach. So, ask your questions using #cxotalk. And, we’ve got three similar questions from
Wayne Anderson and from Ian Girtler, and from Chris Peterson who are all asking about security,
privacy, the aggregation of data. And Milind, you mentioned security. Thoughts? You guys are in the industry. Look, I think this is perhaps one of the most… You know, when we think about how we organize
our enterprise information management assets… You know, I remember back in the day when
we used to do some of this work. You know, our security compliance legal teams
were often the poor guys left at the last minute to decide on a particular initiative
because the cool factor was all around the app, or it was all around the campaign. But, the reality is that has to change. It is changing, and it’s probably the most
critical ingredient. I think recently you would have heard the
Head of the FDA talking about, you know, building security is part of your design process when
it comes to devices and digital health aids. So, as far as I’m concerned, it is one of
the first things we think about. You know, all of that really stems from the
basic question: What is the intended use of that data, right? If you can’t answer that question, you probably
shouldn’t start that project. I think the other two pieces that are becoming
equally important in the utility of data is the social responsibility aspect of it as
well as the ethical responsibility aspect of it. It’s not that these are new concepts. They’ve always been there, but they’ve
often been there, say, as part of the overall embedded program but not necessarily integrated
programs that make sense. They’re always on the periphery, you know? “Oh, did we check off this piece?” “Yes.” Now, you can’t check it off, now it has
to be, “Did we design data and exploration activities with these in mind thinking ethically,
socially responsibly, and from a privacy perspective as well?” So, certainly for my team, this is something
we take very, very seriously. I think it’s good that security and privacy
are becoming more actionable in the narrative because it was in the narrative for a while,
but it wasn’t necessarily something we were taking action on, right? One of the things that I find interesting,
and this is exactly the way it was in financial services, right; when financial services was
going through a transformational barrel a decade ago. It’s that when you start to have conversations
about security and privacy, the automatic reaction is to think about, “Oh, look at all
these new datasets! They’re dangerous.” “Look at all these new pipes and integrations
that we’ve opened up. They’re very dangerous.” And that’s because it’s an issue of familiarity. Most of the threat is actually in the data. You know, we’ve had for a very long time that
we work with, that we’ve become comfortable with. And, I think we make a mistake of thinking
that when you do new things with new datasets, suddenly that must be more dangerous than
the one we have right now. I think that’s the first thing that we ought
to make sure that as we go about this exercise — to start to unlock and unleash data to
drive the integration to get those data-to-decision sort of outcomes that Milind’s talking about. We don’t sort of mistakenly misappropriate
where the risk really is. Right? Because, there are still a lot of risks internally. But, I think what’s even more interesting
is that we are at a point now where there are technology paradigms that are allowing
us to have data that is protected but shareable at the same time, okay? We used to have to make this binary decision
before whereas if the data’s protected, then it can’t be shared. And if the data is shared, then it’s probably
not protected. And, I’m going to intend which paradigm
that is, if you’ve been following me, you know it’s what it is! But, we have a reality in front of us with
the invention of the blockchain, with asymmetric cryptography and distributed ledgers to truly
start to build business models where data can be protected but shared at the same time. And that, I think, is opening up a lot of
opportunities. So, we have another comment, a really interesting
comment, from Twitter. And, this is from John Nosta, who says… [Laughter]
Hey, John! [Laughter]
[Laughter] All right. So, John Nosta is saying, “Yes, it’s true!” As one of you said earlier, if you fail in
the pharma industry, then people die. I think, Milind, you said that. But, he points out that, quote, “Snail pace
success also results in deaths.” Could not agree more. So, how do you balance? How do you balance this need for speed and
obviously, there are elements of our healthcare system that are just failing miserably. So, what do we do? And, link it back, again, to this notion of
silos that you were both talking about at the beginning. Yeah. I mean, I think a lot of it really comes down
to, yes, there is a historical paradigm which is still very much prevalent. In any pharma, or even, I would argue, most
biotech companies as well, where the knowledge base under which we operate is often underwritten
by risk. And those risk parameters that are often created
are done either protecting liability of an organization and ideally protecting the protection
of the consumer or patient in this instance. And, I think it’s a fine balance, right? I don’t think there’s any sort of golden
key in terms of how do you uncover that. But, what we are seeing, though, is that the
use of technology allows us to accelerate some of those processes quickly where, for
example, the use of machine learning or deep learning in some of these instances allows
you to explore much larger populations and look for non-responders as much as you see
responders. And, if you really think about, you know,
clinical trial management in many regards, this is an area we are often looking to see
the positive clinical outcomes, right; and balance that with the underwritten risk of
a trial. In post-clinical trial, and that’s why I often
find it quite interesting, we have a treasure trove of information already. We don’t need to buy, necessarily, additional
data, if you will. If we just mined what we already had in a
more sophisticated way using algorithms to go faster, we might already uncover new mechanisms
of action, new therapeutic conditions, and so forth that I think… You know, will it be faster? Yes. Will it be, you know, like the bullet train? No. And, I think we need to have a fine balance
between both of those. But, I do agree with John, that there is this
historical basis of working at snail pace. A lot of it is also because the adoption of
technology in this place has been quite limited as well. I think one of the things that are changing
in the industry; and by the way, John’s the best looking guy in the state of New Jersey,
in case you’re wondering; one of the things that are changing in this industry, which
we’re seeing the same pattern coming from other industries, is that the demand side,
right? The demand signals, the patients, the way
they start to work together in communities like patients like me, the way they start
to get involved in understanding therapeutic areas and researching illnesses and treatments,
the patients are waking up, so to speak, metaphorically and are starting to say to the life sciences
industry, “Hey, I need it better and I need it faster.” So, I think there’s a lot of inside-out
debate and discussion that we have around how we change our operating model and how
we move faster, and how we optimize. But, the final say always comes from the customer. And, you’re seeing more and more patients
interested in early clinical trials because they’re becoming aware that this is there. You’re seeing more and more patients have
discussions about pricing and about reimbursement, and about the efficacy of a drug. And, I think it’s the end of the supply
chain. I think it’s on the demand side, where the
customers are actually going to wake up and are going to force us to get out of this snail
pace, as Nosta describes it. Yeah, I think, you know, if I could add to
that, Richie, you’re 100% correct. I really believe that this healthcare industry
is being driven by customers and these customers are not only becoming more digitally-savvy,
but they’re also becoming more information and context-sensitive as well. You know, I’ve worked with the COPD Foundation,
which is such a wonderful organization, in in the past and I can tell you that the patients
that we met in those groups knew more about their disease; knew more about the molecular
mechanisms than the latest research in this area that really, I was truly humbled to really
understand not only what they’re going through contextually, right; just living with that
disease; but also scientifically in their notion to fight for truth, if you will, and
access to information which is so critical. And, we still have some major boundaries to
get over there. I think where the paradigm shift is going
to happen is when you do start having these trusted relationships. If you think about what a farmer probably
is trying to do ultimately, it’s really trying to enrich that relationship between
physician and patient. And, we all know that even with the advances
of digital technologies, the time a physician actually gets to speak face to face with a
patient is getting more and more limited. And, we would think, that’s the disturbing
effect of digital, right? And I actually think there are a lot of elements
within digital, and particularly the digital dystopia, that are probably more advantageous
to being able to provide patient support programs which ultimately then lead to companies like
ours progressing faster to provide these services. And, this is a known pattern, right? The financial services industry has gone on
through this. I’ll tell you what I call the pattern. I call the pattern the phenomenon when an
industry goes through an industry that has to provide answers to an industry that has
to provide evidence, okay? About fifteen years ago, I remembered the
day when financial services went through that sort of change. I was on the phone with my broker and I was
talking to him about some trade that I was trying to make, and I just looked up for a
second. I said, “Wait! I know more about this trade than the person
on the other side of the phone! Why am I calling them, right?” When the financial services industry went
from an industry where you would go to the industry for answers, to an industry where
today, you’ve got all the answers. You’re going from sort of evidence, right? And sort of confirmation. The life sciences industry is going through
that same change. We were at a point where people came to us
for answers, right? “What’s x, y, z? What’s a, b, c?” Today, our customers are not looking for answers,
they’re looking for evidence or confirmation. And as you talk about that constricting sort
of moment there, where a patient and a provider, a doctor, a PA, a nurse, gets to chat, even
the delivery side has to go away from sort of only providing answers, to start to really
contextualize that conversation properly. These are the changes that I think are coming
from the outside of the industry that will marry with some of the great stuff that guys
like Milind are doing, to really get us to where we need to be. I think, in some ways also, they’re really
helping in dislocating the silos that we have today. If you really think about it, right? Typically, you go to a pharma company or a
biotech company, you’ll have therapeutic area base divisions; you know, cardiovascular,
diabetes, neuroscience, etc. Now, oncology, etc. And all of those groups, God bless ‘em,
because they do work very, very hard to be able to provide the right services to patients. But, the reality is they’re often incentivized
and motivated by supporting that one therapeutic area alone. But, you know who doesn’t care about that? Patients. Because, sometimes, patients have different,
you know, unfortunately, different diseases or conditions at the same time and they don’t
care how we’re constructed. They’re looking for solutions. They’re looking for services to help lead
a dignified life. And I, for one, and one of the things I’ve
experienced is really profound here at Sanofi, I’ve seen this demonstrated. I’ve seen this led. I’ve seen the projects that we’re doing right
now and we’re really making a conscientious effort to break down some of those silos and
really tackle these issues from a patient-centric perspective and also, from a physician and
caregiver perspective, as well. It’s really looking at the holistic care package. You know, John Nosta, again, he’s raising
some very, very interesting points on Twitter. And I just want to remind everybody, use the
hashtag #cxotalk. There’s a very vibrant discussion going
on about these issues on Twitter right now and you should definitely be participating
in that. So, he’s talking about clinical trials. He’s saying the future is about genomic-based
AI augmented trials for smaller patient groups. And, he doesn’t want a doctor… He says, “I don’t want a doctor anymore. Give me the power of AI and I’ll adjust
that genius bedside manner.” This is when I love when a Chief Data Officer
and a Chief Digital Officer get trolled by Nosta on Twitter. [Laughter]
[Laughter] [Laughter]
See, this is where you have to balance reckless disregard with responsible action, right? At the end of the day, the statements that
John are making — of course, we all live that every day, right? There’s no one of us affected in the industry
who are driving the change that doesn’t recognize that. But, we have to be responsible in our actions
to recognize that some of these discussions are on the edge. And yes, it’s our responsibility to scale
the edge and bring them to the center of how we change. But at the same time, change in large, commercial
organizations, requires the marrying of that reckless disregard and responsible action
at the same time. So, there’s no disagreement here that, yes,
genomic data is certainly a landscape that will be fruitful for moving particularly from
care to prevention. And implementing artificial intelligence in
certain areas will be beneficial as yet. I would argue that AI is probably a little
bit too much of what we’re overpromised, and we’re being too irresponsible with it. But, we have to recognize that in any given
instance of the history of our species, when we’ve made macrostructural and social changes,
like the ones we’re about to go through where healthcare is about to become demand-driven,
as opposed to supply-driven, there’s a place for the reckless disregard but at the same
time, it has to be married with responsible action. And I think that’s what being in the healthcare
industry is about today. If you’re a Chief Data Officer, transformation
officer, you have to balance those, too. Two points here. Number one is we have a question from Twitter. Victoria Walters has asked twice and she’s
getting annoyed because I have not asked you this, which is value-based contracts; what
do you think of them? Are they truly valuable to the customer, to
customers or patients? And then, I want to come back to AI because
an article just came out in STAT Magazine journal about how Watson is not working. So, can we take a quick detour; value-based
contracts because Victoria Walters is mad at me that I haven’t asked you that? Yeah, Milind, you take Victoria’s question. I want the Watson question! [Laughter]
[Laughter] [Laughter] Well, I’ll cover a little bit
of that but I’ll let you do that as well! All right! Okay, Victoria, so value-based contracting. I think we’re at the early stages of what
actually a value-based contract actually means. Initially, it’s based on specific clinical
endpoints based on a targeted population. And often, these targeted populations have
inclusion and exclusion criteria that don’t necessarily meet you, as an individual, when
it comes to a point of reimbursement, right? That said, I think the science is getting
smarter. And I think the contracting over a period
of time is going to learn how to make that more effective. I want to give you a very simple example. I do think, in many ways, it’s the right way
to go. I just don’t think we’re there yet at the
level where it’s satisfying and benefiting patients at a holistic degree. Let’s take a hospital, for example. And you have a terrible experience in the
hospital. The likelihood, potentially, of you not having
a great outcome is, in fact, higher whereas if you had wonderful customer service all
the way through, you could argue that, yes, the outcome is higher. So, in the contractual language, who then
owns that component of the outcomes-based contract? Is it the patient satisfaction score when
being hospitalized and treated? Is it the clinical evidence that’s coming
from the medication? Or, is it something more? And I think these are the things that we need
to start drilling deeper into. I know there are a lot of watch streams going
on in this space, but Victoria, my apologies, I think we’re just really at an early stage
of that market. There’s a lot more work that needs to be
done to really define what that means. You bring up genomics and AI. And again, John, a lot of the things. But, you know, one thing that I tell the team
internally; one of our primary job descriptions is to know when to triage bullshit versus
when to triage other things that exist out there. Hey! This is a family-oriented show! You can’t say, “triage bullshit” or
not. Anyway, go for it! Well, let me put it this way! BS and buzzwords lead to a lot of wasted pilots,
right? And God knows we have a dead sea of pilots
going on today. Look, I think we’re at a really remarkable
stage at ML, right? I’m not going to say “AI,” because honestly,
AI is a buzzword that’s grossly overused at this point in time. Nobody’s doing AI. We’re doing natural language processing, we’re
doing computer vision, we’re doing other elements that may constitute as machine learning algorithms
supplied within an AI family, including deep learning. AI is a farce at this point in time in terms
of where we are. In healthcare, I think other industries are
probably a bit further ahead, but in healthcare, we’re not there yet. When it comes to genomic data, I fully agree. Genomic data probably has one of the most
wonderful opportunities around the utility of machine learning and deep learning in these
areas. But what’s changed from forty years ago,
when these classification algorithms, these clustering algorithms, are already there in
a naïve Bayes or a random forest, see; they were there many years ago. Computational processing power; cloud; these
are the things that have allowed the utility and application of what we could have done
many years ago a lot easier. But, why is it easier now? Because your resource allocation to provide
the cost necessary to use that infrastructure is a lot cheaper. Back in the day, we had blade environments
to be able to do high-performance computing. But if you asked yourself, well, if that was
really all then there, why is it still taking almost 48 hours to do some of these analyses,
right? That’s what I would argue. The opportunity for most of the work that
we’re doing… Look, I love AI as much as the next person. I mean, just look at my Twitter handle. You’ll know what I mean, right? I do love AI as much as the next person and
I do believe that’s where the world is going, but we need to have ethical responsibility
in the utility of those, and not overuse buzzwords like AI when, in fact, you’re actually talking
about a very narrow field of machine learning solving one specific problem. Now when it comes to Watson, and you need
to be careful here, I do like IBM as a company. First and foremost, I think the one thing
that I really cherish about IBM over the years is their ability to transform their existing
business, time and time again. And, that’s amazing. You really want to look at transformation
industry? They are the gospel, if you will, in many
regards of how that can be done. I also like them because they allowed, through
the cognitive computing environments and their wonderful marketers who I’m sure made a
ton of cash in this process, enabled an extra zero on my budget line item. Because, now, I was being asked a question:
“Milind, what are we doing with cognitive?” “What are we doing with AI?” “What are we doing with machine learning?” So that was great. That was cool. So now, I was able to hire a bit more stuff
to start looking into these things. But you know, I think, Watson’s at an interesting
point where, and of course, the last thing I would say about Watson is that their research
is perhaps some of the most brilliant minds on the face of the planet; I mean truly impeccable
people that work at the Watson research facilities. Where I think the rubbers hit the road is
sometimes the overpromise of what can be done and by when. Now, I know that I’ve had some good use cases. I’ll be the first to say that’s great. And that’s wonderful because it’s trending
in the right direction. I think sometimes, it’s easy to pick on the
big guys because they’re just big, so the target’s bigger. I think they’re learning as a firm. I think they redefine how they approach the
industry. I don’t think it’s all going to be about professional
services anymore while that’s still honestly still their main model, but it’s, you know… It is what it is. Richie, we have about seven minutes left. So, and this conversation has just begun. So, Richie, again, John Nosta, I feel like
John Nosta is part of our conversation here. He’s on Twitter and you guys are here. So, John Nosta is saying, “Well, so is Watson
marketing over technology; over results?” Richie, your thoughts about this AI topic. And then, let’s move on to a couple of other
things quickly before we have to close out. So, I share Milind’s perspective here, right? I have a tremendous amount of respect for
IBM as an organization and the people there, and they’re doing a great job. I think that first and foremost, if you think
about the comments from Elon Musk and others around AI, one of the things we have to do;
this is for the non-expert viewers; is to always separate robotics from artificial intelligence. I think the entertainment industry has sort
of merged those two at our heads, where we think about robotics and AI, it’s hard to
separate them. AI itself, as algorithm or software, has a
lot of different sub-sciences to it, okay? There’s reasoning and problem-solving. There is knowledge representation. There’s planning, learning, natural language
processing, creativity, general intelligence, social intelligence… There are a lot of pieces of that science
that I will call “sub-sciences.” What Watson has done extremely well is to
really move forward in the area of natural language processing. I’m not saying that they haven’t moved
forward in other areas. I’m saying there’s a tremendous amount
of benefit to that. It’s raised the conversation in the enterprises
for us so that we can start to experiment in these areas. There’s a lot of other small startups that
are operating in the small pieces here and there. But I think the hype, if I may, it doesn’t
fall on IBM’s shoulders. It doesn’t. The hype actually falls from the analysts
and the reporters who are the ones that are getting ahead of their skis because they’re
looking for clickbait, right? I don’t hold IBM responsible for the overhyping
of Watson or artificial intelligence. I think it’s a brilliant commercial strategy. There’s market conditioning-led [garbled
speech]. I think the responsibility is on the analysts
and the writers who are getting way ahead of their skis and don’t just understand
the subject matter well enough to write intelligently about it. We have another question and as an industry
analyst, yes. You know, mea culpa I suppose, right? [Laughter]
[Laughter] [Laughter]
Hey, I don’t mind calling a spade a spade here! [Laughter]
We’ll keep talking about it because we’ve still got more stuff to talk about, so fine! But, we have another question from Twitter. And, Wayne Anderson has got this… Tweets have flown by so fast. He is asking, “How does data help make the
decision about when to be recklessly innovative versus to be cautious and balance that risk?” So, the role of data in innovation versus
risk-taking. Okay, so let me clarify a few things. First of all, innovation in itself is a risk,
right? The notion… I mean, the whole idea of innovation; people
often mistake innovation as being blue-sky thinking. It’s very easy to come up with ideas. It’s very difficult to come up with things
that are either commercially or utility, or experience-based, where you’re driving positive
outcomes in innovation. So, I would argue, innovation, by nature,
is a risk-based venture. That’s why we have companies like venture
capital firms that are measuring risk or startup companies who have brilliant ideas as part
of their valuation, right? So, I think when you use data in those instances,
you’re often looking for a couple of different things. Number one, and first and foremost, what is
the problem we’re trying to solve, right? If you don’t start with that, as a hypothesis,
that’s when you can use data to start either creating a null hypothesis or an outcome-based
hypothesis. And, I think the beauty of innovation in this
instance when you start having a data-driven, experimental approach; is that you can very
quickly reach conclusions or pivot accordingly based on where you are in that life cycle
of the experiment. So, I think as a pharma company, for example,
I mean, this is what we’ve done. This is how we’ve grown. We’ve designed experiments and produced products,
but we often don’t apply that same methodology, that rigor, to experimentation when it comes
to innovation. In reality, they’re exactly the same thing. You have an aim, you have methods, you have
materials, you have a hypothesis, and you have results and conclusions that derived
from that. So, I would say, you know, we should not separate
innovation and, let’s call it “day to day operations.” In fact, I tell our company, you know, as
employees, we really have two pools of employees. We have those that want to do things better,
right? So, you can have innovation within that bottom
line factor, if you will, but then, there are those that want to do better things, right;
that break the rules a little bit. But, you can’t have only one type of group. You need to have both, right? And that’s when our data allows you, in
many ways, to be able to resource allocate accordingly, based on what the business outcome
or the projected customer experience outcome, or frankly, what the value is that you’re
trying to achieve. Richie, I know you’ll have thoughts on this. I’ll ask you just to keep it short because
we’re just running out of time. Well, I think we should do it on another show. If we’re running out of time, we should
take the time to say proper goodbyes and sort of closing statements here, so I’ll pass
and I’ll look at our closing statements. Okay! So, as we finish up, I mean, the time has
just flown by here. So, as we finish up, let me ask you for your
kind of summary or your… The distilled wisdom of your broad experience? And if you can come boil that down to about
a tweet-sized statement? [Laughter] Milind, you want to go first on
that one? Believe in data, be responsible. Wow! That was really good! “Believe in data and be responsible.” So, maybe elaborate with another tweet. [Laughter] Think about the ethical and social
implications of the use of information for good. Okay! Fantastic. And, Richie, your final thoughts and, you
know, maybe address that silo issue that we spoke about right at the very beginning? I think the life sciences industry is suffering
from what I would call a trust deficit. I think we do great work in the life sciences
industry. There are smart men and women here that toil
away at improving lives. The issue is that we have silos that prevent
us from being completely transparent and sharing the great things that we’re doing and the
progress that we’re making. And, we have mechanisms in place that lower
our transparency, okay? At the end of the day, you know, when Apple
launches a new product, everybody lines up in front of the Apple store or the Apple campus. But, when Pfizer launches a drug that saves
hundreds of thousands of lives, nobody’s lining up in front of Pfizer headquarters
at 42nd Street in Manhattan, okay? I think the industry is suffering from a lack
of trust and transparency, and I think we’re going to see, within the next five to ten
years, the life sciences industry go from one that is either being chastised or hated,
to one that is being liked or loved. And, I think trust is going to be the factor
there. You know, we’re out of time, but Milind,
I mean, this is such a kind of powerful statement. And so, Milind, any final thoughts on this
notion of trust and transparency? I think it’s so important. Absolutely! I think, again, it comes down to the ethical
use of data to drive meaningful decisions, right? If you derive meaningful decisions and allow
people to understand the communication of those decisions, then you reach a place of
trust or at least, understand what the key issues are in that trust relationship so you
can further advance a more meaningful relationship. Okay! Fantastic. And, I will just have the last word by pointing
out that on Twitter, Bob Egan, who is an industry analyst, says — is agreeing with Richie
Etwaru that analysts don’t write intelligently about AI because they’re more focused on
clickbait. [Laughter]
And so, there you go! What an interesting show and I feel like we’ve
done forty-five minutes and the time has just flown by! You have been watching Episode #255 of CxOTalk
and, a huge, huge “thank you” to our guests. Milind Kamkolkar is the Chief Data Officer
at Sanofi Pharmaceuticals and Richie Etwaru is the Chief Digital Officer of QuintilesIMS. Thank you, everybody, for watching! Be sure to “like” us on Facebook and be sure
to give us a thumbs up… No, even better; you should subscribe on YouTube! Please do that. Go to cxotalk.com/episodes and you’ll see
all of our shows. We have more shows coming up. Thank you, everybody, have a great day. Bye-bye!