00:00:02
[Music]
00:00:07
so now we move from patient care to Life
00:00:09
Sciences to understand how the
00:00:11
Enterprises in the Life Sciences sector
00:00:14
what are they doing about AI so we've
00:00:16
picked a panel first of all tun matur is
00:00:18
the CTO of indigene and he's kindly
00:00:22
agreed to moderate the panel we have uh
00:00:25
Jeremy Zang from Gilead he brings the
00:00:28
length of uh the clinical research drug
00:00:32
development to this discussion uh we
00:00:35
have uh Michelle rorer she brings the
00:00:38
regulatory lens you know when I was
00:00:39
asking Michelle what she's most
00:00:42
passionate about she said it takes too
00:00:44
long from the time we realize we have a
00:00:46
medicine to the time we dose the first
00:00:48
patient I want to cut that time and I
00:00:50
want to see how AI can help us cut that
00:00:52
time so that's her passion and finally
00:00:54
priia Priya is a CEO of alive core so
00:00:57
she brings both the Enterprise lens and
00:01:00
as well as what alive core is doing to
00:01:02
incorporate AI into their products and
00:01:05
solutions so uh tarun matur uh will
00:01:09
moderate the session he's right behind
00:01:12
me I'm really excited to be able to
00:01:16
engage with a really distinguished panel
00:01:17
of experts and our topic is of course
00:01:21
one that's on the top of all of our
00:01:22
minds we've seen it throughout the
00:01:23
various sessions these couple of days um
00:01:26
my name is teron Mo I'm the chief
00:01:28
technology officer for indig
00:01:30
and uh so our company is a 25-year-old
00:01:33
organization we exclusively work in the
00:01:36
Life Sciences space Enterprise Life
00:01:37
Sciences to be more specific and we are
00:01:40
primarily a services and solutions
00:01:42
company about 6,000 is employees 40% of
00:01:46
our company are domain experts so
00:01:47
clinicians pharmacists um other uh
00:01:50
business experts and we have heavy
00:01:52
technology expertise as well now with
00:01:55
the Gen revolutions happening we we
00:01:58
recognized a few years ago this shift in
00:02:01
Opportunity that's happening with geni
00:02:03
and I can say from the indigene side we
00:02:06
actually went through a substantial
00:02:07
reorganization within the company to
00:02:09
really capitalize on that intersection
00:02:11
of domain and technology and so we've um
00:02:15
rebuilt our teams we we've changed the
00:02:17
processes we you know sdlc for familiar
00:02:19
with software development life cycle
00:02:21
we've actually rebuilt that thinking
00:02:22
about how do I fold in domain experts
00:02:25
differently and we started from prompts
00:02:28
to we heard retrieval augmentation and
00:02:30
today is all about agents and we've
00:02:31
heard a couple of talks about the power
00:02:33
of Agents so U we've seen a lot of
00:02:36
really amazing opportunities from the
00:02:37
services uh side of the business on what
00:02:40
you can do from Clinical Research
00:02:41
through medical and Regulatory writing
00:02:43
and submission all the way through
00:02:44
commercialization and postc
00:02:46
commercialization activities I'm I'm
00:02:48
honored to have such a fantastic panel
00:02:50
of experts here who really know this
00:02:52
space inside and out and I want to make
00:02:54
sure that we really can get great
00:02:55
advantage of their time so what'll do is
00:02:58
maybe we'll dist off the discussion if
00:02:59
each of you then can introduce yourself
00:03:00
while we talk about the uh The Prompt
00:03:03
but let's let's start off with really
00:03:04
from an entrepreneur and innovator lens
00:03:08
within your functions in your space What
00:03:10
are some of the big challenges that
00:03:12
you're seeing and what are some of the
00:03:14
opportunities and maybe they're the same
00:03:15
thing right opportunities come from
00:03:17
challenges but maybe we can start from
00:03:19
from uh that discussion and maybe you'll
00:03:22
follow the the drug life cycle maybe
00:03:24
Jeremy I'll start with you and look at
00:03:26
from the clinical side of things um
00:03:28
maybe you can tell us bit about yourself
00:03:29
and your function and the opportunities
00:03:31
and challenges that you're seeing in
00:03:32
your your role sure and yeah thank you
00:03:34
so much tan and avanish for inviting me
00:03:37
um so I lead drug uh drug Discovery and
00:03:40
development uh from an AI perspective at
00:03:43
Gilead um we a medium-size Pharma of
00:03:46
about $27 billion of Revenue um and by
00:03:50
far like if you're in this space you
00:03:51
know about eo's law which means that the
00:03:54
ROI of drug R&D is trending towards like
00:03:58
negative so so AI I think is like one of
00:04:01
the things that can potentially turn
00:04:03
this around like there was a blip during
00:04:04
covid where the ROI went up um and there
00:04:07
was a famous story that fiser claimed
00:04:09
that they developed their drug using AI
00:04:11
the reality is um and I think the last
00:04:14
uh speaker said it really well the AI is
00:04:16
quickly not becoming that much of a
00:04:19
differentiator um I think even at Gilead
00:04:21
what we're seeing is that the key to
00:04:23
providing value and one of the biggest
00:04:25
challenges that we face um especially
00:04:28
when I I speak to startups in this space
00:04:30
like pretty much weekly I think the key
00:04:32
differentiator that I see is the ones
00:04:34
that really understand how to address
00:04:36
the challenge of integrating the AI into
00:04:39
how decision- making actually happens
00:04:41
for like these billion dooll drug
00:04:43
development programs are the ones that
00:04:45
really get like a second look from us or
00:04:48
even like a CO a partnership or a
00:04:50
co-design opportunity um the ones that
00:04:53
we tend to skip over are the ones who
00:04:55
try to lead too much with hey we have
00:04:57
like some AI special secret sauce um but
00:05:00
since we partner with like open Ai and
00:05:01
anthropic we know that that's probably
00:05:03
not really the case especially for a
00:05:06
really small and early startup so any
00:05:08
any any you know group of entrepreneurs
00:05:11
that really understands like how to
00:05:12
actually integrate into the clinical
00:05:14
workflows and also decision decision-
00:05:16
making and development is really the
00:05:18
ones that are actually addressing the
00:05:19
challenges so that's so that's
00:05:21
interesting for from a business side of
00:05:24
of the view this technology enablers are
00:05:26
one thing so certainly familiar with the
00:05:27
anthropic and open eye so we're seeing
00:05:29
powerful technology enablers um then the
00:05:32
change management aspect so for example
00:05:35
we can go to maybe gp4 and through
00:05:39
clever prompting and Rag and agents make
00:05:41
it do claims right uh or try to identify
00:05:44
new claims but from a change management
00:05:47
and Regulatory perspective are there
00:05:49
certain challenges or roadblocks in
00:05:50
terms of establishing The credibility or
00:05:52
trust that you see from your function um
00:05:55
along those
00:05:56
lines yeah maybe I can answer it and
00:05:58
actually I think Michelle probably has a
00:06:00
really good followup to this um you're
00:06:02
absolutely right I think one of the
00:06:03
biggest challenges in a really highly
00:06:05
regulated space is um our threshold for
00:06:09
risk is is quite High um so if we see
00:06:13
something that's um potentially or
00:06:15
actually quite low like if we see see
00:06:17
something that is even has a chance of
00:06:19
impacting patient safety it's just
00:06:21
completely unacceptable to use um which
00:06:23
makes it very difficult to trust large
00:06:25
language models in a vacuum for any kind
00:06:27
of um design of a clinical study or
00:06:31
patient you know research angle making
00:06:33
product claims about what drugs can do
00:06:36
um everything is augmentation um and I
00:06:39
think that's another approach like I
00:06:41
highly recommend people take is to
00:06:43
approach it from an augmentation
00:06:44
perspective um and I I really think that
00:06:47
we will almost never reach a state where
00:06:49
a large language model can perfectly
00:06:51
prescribe um like how to run a program
00:06:53
because there just too much risk
00:06:54
associated with patients Michelle yeah
00:06:56
Michelle it's a great segue to you if
00:06:58
you can maybe introduce yourself thanks
00:07:00
hi good morning good afternoon everyone
00:07:02
Michelle roar I come to from jentech Ro
00:07:05
where I head up development regulatory
00:07:09
um we can talk at length about the
00:07:11
challenges of AI use in a large Pharma
00:07:15
and there are also an abundant number of
00:07:18
opportunities but if I could take you
00:07:20
into my world of regulatory for a minute
00:07:23
a key two key challenges that I would
00:07:26
highlight um for you um did you know
00:07:29
know that by and large drugs are
00:07:32
developed and reviewed with Regulators
00:07:35
still in more or less a paper format now
00:07:39
it is an electronic paper format but we
00:07:42
have to create large volumes of um
00:07:46
Consolidated PDFs that are linked of
00:07:50
course electronically and then the
00:07:52
regulatory reviewers at the different
00:07:55
countries and we operate in excess of
00:07:57
150 countries all review these large um
00:08:01
electronic
00:08:03
dossier that's a real reality for us
00:08:07
which means that those dossier that are
00:08:10
electronic paper need to be
00:08:14
written um and now of course we can we
00:08:17
can leverage generative Ai and
00:08:19
structured content management to write
00:08:21
those dossier um but the
00:08:24
opportunity um is is really large um in
00:08:30
part because of a stark
00:08:32
reality and I want to take you into a
00:08:34
world where you all pretend that you are
00:08:38
innovators at a biotech or at a large
00:08:41
Pharma and you have a drug that you're
00:08:43
really excited about that you've been
00:08:45
developing you've been running it
00:08:47
through the phase one phase two and
00:08:48
phase three trials and you're at the
00:08:51
point where you have pivotal trial data
00:08:54
now let's just pretend that this trial
00:08:57
data set is a large data set said maybe
00:09:01
upwards of a th000 patients maybe 2,000
00:09:04
patients and you're operating in a
00:09:06
disease where there is no drug there is
00:09:08
an unmet medical need and you're sitting
00:09:11
in the room and we're the review team
00:09:13
and we're sitting in the room when we
00:09:15
unblind that
00:09:17
data and the data is very clear this
00:09:20
product works and it works like Gang
00:09:24
Busters let's just pretend that it saves
00:09:27
lives people live if you get the treated
00:09:31
treatment or they die if they don't you
00:09:34
know that
00:09:36
day that you have a
00:09:39
drug you are
00:09:42
probably approximately 6 months away
00:09:45
from submitting that data to regulators
00:09:49
and then Regulators will review that
00:09:50
data for another
00:09:52
year but you know that day you have a
00:09:55
drug this is our inefficient process
00:09:59
that Jeremy alluded to the opportunity
00:10:02
is that the day after that unblinding in
00:10:06
the world of using Tech from both the
00:10:09
innovator side as well as the regulator
00:10:12
side that the next day you have a drug
00:10:15
that's available for patients and we
00:10:17
will be there someday but we're not
00:10:20
there today because we are drowning in a
00:10:24
paper process that is regulated and
00:10:28
expected
00:10:31
so there's a a lot of daylight between
00:10:34
the opportunity and the challenge but I
00:10:36
do believe um that together because of
00:10:40
the inefficiencies of the process are
00:10:43
creating bottlenecks both at the
00:10:44
regulator side as well as the sponsor
00:10:48
side in terms of having an business
00:10:51
models that are are profitable and can
00:10:54
sustain The Innovation that the
00:10:56
dissatisfaction is rising to a level
00:10:59
where we will see breakthroughs and that
00:11:01
there will be more experimentation to
00:11:03
lead us to the world of when we know we
00:11:05
have a drug it will be available for
00:11:08
patients in a very short time window and
00:11:11
now I want to introduce you to my
00:11:13
colleague priia who has a really
00:11:15
exciting company with um a patient
00:11:17
centered approach just keep I could just
00:11:20
keep listening to all of you seriously
00:11:22
um hi everybody I'm priia abani CEO of
00:11:24
live core uh been here a little bit over
00:11:27
4 and a half years came from Tech so I
00:11:30
came from the Dark Side um and uh
00:11:33
strongly believe in this intersection of
00:11:35
healthcare and Tech because I think we
00:11:37
all know the challenges that we are
00:11:38
facing including some of what was
00:11:40
mentioned right now and that's not just
00:11:42
for drug delivery and Drug creation it's
00:11:44
also for devices it's also for AI it's
00:11:47
the same process that we all go through
00:11:48
of creating something that we know is
00:11:51
going to change the lives of millions of
00:11:53
people but then having waiting for those
00:11:56
processes to unfold and sometimes many
00:11:58
many years go by before actually
00:12:00
patients can benefit from it so a little
00:12:03
bit about aive cod we're a health tech
00:12:04
company we are devoted to the
00:12:07
advancement of remote cardiological care
00:12:10
and just from our Journey's point of
00:12:12
view we our founder founded the company
00:12:15
like about I think 11 years back I'm
00:12:17
losing track of time now because years
00:12:19
are just flying by this is the longest
00:12:21
I've been in one job by the way um and I
00:12:24
feel like I'm just getting started but
00:12:26
um we started off with the single lead
00:12:27
ECG device we went to a six lead ECG
00:12:30
device we're now in the process of going
00:12:32
through clearance of a 12 lead device
00:12:34
that can be used from home uh but what
00:12:36
is changed in the meantime is we have
00:12:38
been doing AI way before it was school
00:12:41
so um we had like you know three
00:12:43
determinations uh for those of you who
00:12:45
know something about the heart we had
00:12:47
arhythmia so atrial fibrillation Brady
00:12:50
cardia Taki cardia then we basically
00:12:52
looked at another lead and we said we
00:12:54
can do more things so we started doing
00:12:57
cue prolongation detection we started
00:12:59
doing sinus rhythms we are now looking
00:13:01
at atrial flutter and then with the 12
00:13:04
lead of course we want to expand that
00:13:06
the reason I'm saying this is this is
00:13:08
clinical Ai and that's the only way it's
00:13:11
going to be built it's not going to be
00:13:12
algorithmic anymore some things are okay
00:13:14
to be algorithmic but the kinds of uh
00:13:17
data that we can feed to DNN right now
00:13:20
there is no human possibility for
00:13:22
anybody even like a whole set of human
00:13:24
beings to consume that and train
00:13:26
themselves to create that kind of
00:13:28
prediction um and that's something we
00:13:30
all have to understand so from my point
00:13:31
of view I see nothing but opportunity
00:13:34
now of course generative AI is the new
00:13:37
buzz word we all like it we all
00:13:39
understand it but we have been very
00:13:40
measured right so the first thing we did
00:13:42
is we took um I want say chat GPT but we
00:13:46
took open AI you know GPT version
00:13:49
something put a rapper around it trained
00:13:51
it only to address cardiological needs
00:13:55
so it's right now sitting there as a
00:13:57
Bart in some of our device and some of
00:13:59
our subscriptions where the patient can
00:14:01
go and say hey my heart I'm having
00:14:03
palpitations I drank water today and I'm
00:14:06
doing this tomorrow is this okay or what
00:14:08
am I feeling do I have a blood pressure
00:14:10
issue Define atrial tribulation for me
00:14:13
so it's sitting and doing something very
00:14:14
simple right now we're watching how
00:14:17
people are using It ultimately we do
00:14:20
think this will become a coach because
00:14:22
right now in our system we have human
00:14:23
coaches right and we of course have
00:14:25
cardiologists which I frankly I don't
00:14:27
think will'll ever be replaced uh
00:14:29
hopefully they can use all these tools
00:14:31
to augment and just get more efficient
00:14:33
and better uh but I think that human
00:14:35
coach element can probably be augmented
00:14:38
with an AI coach and that's what we are
00:14:41
doing right now so it's it's a step by
00:14:42
step you're not going wild and saying
00:14:44
I'm going to push this out there you
00:14:46
know in its full form because I don't
00:14:48
know what the definition of the full
00:14:49
form is and because we are already
00:14:52
training people to start using devices
00:14:56
in the first place to start monitoring
00:14:58
thems eles you can't throw like an
00:15:01
extreme thing at them just because the
00:15:03
tech industry thinks it's cool and this
00:15:05
is from a person who used to work on
00:15:07
Alexa which was all AI right so even I
00:15:10
understand the measured approach uh and
00:15:12
then how do you train but most
00:15:15
importantly win the trust of your
00:15:17
customer and the patient step by step to
00:15:21
ultimately take them to where you know
00:15:23
we all think they should be at so um I
00:15:26
mean it's a very exciting world right
00:15:28
now Hospital are running at Major you
00:15:30
know really thin margins we don't have
00:15:33
enough resources enough Physicians
00:15:35
there's like one cardiologist per 10,000
00:15:37
population in the United States um we
00:15:39
don't have enough nurses as we all know
00:15:41
we don't have enough administrative
00:15:42
staff U I mean how is this going to work
00:15:45
especially for a aging population so I
00:15:48
think AI is here to stay and it is going
00:15:50
to be a roll out some people will be
00:15:52
fast some people will be slow but the
00:15:54
patient has to accept it the system has
00:15:56
to accept it so that's kind of my
00:15:59
impression great thanks for and what I'm
00:16:01
hearing is and a common theme is that
00:16:03
the significant opportunities we see
00:16:06
were focused on getting AI into the
00:16:08
hands whether it's patients and
00:16:09
caregivers like you're talking about
00:16:10
priia or in the business how do I get
00:16:12
the right medicines to the right
00:16:14
patients faster right it's obviously
00:16:15
good for everybody at the same time
00:16:17
we're managing risks and we're trying to
00:16:18
figure out how to mitigate that I'm
00:16:20
Wonder within your roles have you seen
00:16:23
have there been certain heris or
00:16:24
approaches that okay we're going to be
00:16:26
launching gen we're going to try using
00:16:28
it in some way but here's a a measured
00:16:29
methodology for how we measure risks so
00:16:32
that we can make decisions and I'll tell
00:16:33
you from our our perspective we haven't
00:16:35
found a workbench for building gen
00:16:38
applications that infuses actual prompt
00:16:41
evaluation and testing Frameworks at a
00:16:43
sophisticated level right we've had to
00:16:44
build our own at indene but in each of
00:16:47
your organizations as you've been in
00:16:48
this geni journey have you started
00:16:50
tackling this problems about folding in
00:16:52
issues or un responsible ethical safety
00:16:55
and risk into your workflow and
00:16:56
processes maybe you can speak a bit
00:16:57
about that and Michelle you started off
00:17:00
Jeremy you're about to go the mic so you
00:17:01
start Jeremy yeah I guess we'll follow
00:17:03
the order so your spot on like I think
00:17:06
one of the biggest challenges is you
00:17:07
know you see a whole bunch of very
00:17:10
interesting Solutions um everything from
00:17:13
how to prescreen patients better lead
00:17:15
generation um how to design clinical
00:17:17
trials how to find patients through
00:17:20
terabytes of medical claims and EMR I
00:17:23
think the common challenge is well how
00:17:24
do you evaluate the effectiveness of
00:17:26
that solution and as well as the safety
00:17:28
um so internally we've tried to develop
00:17:31
adversarial challenges to multiple
00:17:33
different types of llms um we Implement
00:17:36
those challenges whether it's for
00:17:37
summarization extraction or whatever the
00:17:40
vendor is trying to do um but you're
00:17:42
right like if somebody could just come
00:17:43
up with something like almost every
00:17:45
single large Enterprise would be on
00:17:46
board immediately um to some sort of
00:17:48
like useful framework for adversarial
00:17:50
challenges for NLP um it would just see
00:17:53
so much usage um in a vacuum I know not
00:17:56
just at Gilead but I've You Know spoken
00:17:58
to colleagues at NCH and uh astroica for
00:18:00
example everybody is trying to develop
00:18:02
those Frameworks due to the risky nature
00:18:04
of what we do right so sure yeah you can
00:18:08
really we have ai being embedded
00:18:10
throughout the drug development
00:18:11
Continuum all the way to drug Discovery
00:18:13
to um protein engineering or small
00:18:17
molecule engineering um to decision
00:18:19
support like how do you make sure that
00:18:22
you're making the right bets in your
00:18:24
portfolio um to designing the protocols
00:18:28
um here AI is competing with Scientists
00:18:31
who always want more
00:18:32
data and so designing a protocol that
00:18:35
just collects enough data um to doing
00:18:38
things that probably you all are quite
00:18:40
well aware of like um translation of
00:18:44
informed consense um a few years ago we
00:18:46
used to have to have a person translate
00:18:48
into the multiple languages where we run
00:18:51
a clinical trial but today we can do
00:18:53
that um using a bot um so there's a the
00:18:57
ho there's a host of bis for for AI or
00:19:00
large language models generative AI or
00:19:03
technology Bots to be used throughout
00:19:05
the Continuum and I I would say that the
00:19:08
interest from the industry is at an
00:19:09
all-time high um because the the ROI
00:19:14
generally is declining and and that's
00:19:16
due to the increasing costs and
00:19:19
complexity of operating in this
00:19:22
space yeah just I I I agree with
00:19:25
everything that's being said but just to
00:19:27
add to that right ethical so security
00:19:30
privacy consent um for us in our case
00:19:34
getting data from a vast majority of
00:19:37
people who look different from each
00:19:40
other um you know making sure we do that
00:19:42
like when we train U you know a DNN on 1
00:19:46
million ECGs you don't want it all to be
00:19:49
from the same zip code from Bay Area
00:19:52
right or uh specific so I I I think
00:19:55
making very deliberate efforts to you
00:19:58
know be
00:19:59
inclusive uh and sometimes I am certain
00:20:02
like all other things in our country
00:20:04
we're going to go to one extreme and
00:20:06
then we'll have to go back to the other
00:20:07
extremes and we'll come back in the
00:20:08
middle but I think every leader um not
00:20:12
just the cxos but I think actually the
00:20:15
head of every discipline has to think
00:20:17
about a how they can use AI in their own
00:20:20
discipline um and then secondly how can
00:20:22
they be more inclusive and then of
00:20:24
course the the whole matter of security
00:20:26
and privacy and you know having an on
00:20:28
onized data like this is just Baseline
00:20:31
in my mind I think companies that breach
00:20:33
that should like should definitely be
00:20:35
reprimanded right because that is just
00:20:37
putting it's it's a it's a question of
00:20:39
putting funding and resources towards
00:20:42
making sure that that Baseline is not uh
00:20:45
breached great so in the couple of
00:20:47
minutes we have remaining I want to do
00:20:49
kind of a quick lightning round here um
00:20:51
so if you put yourself in the shoes
00:20:53
let's say you're leading a startup
00:20:55
incubator and you have this pile of
00:20:57
applicants coming in to to get started
00:20:59
with what would be especially given your
00:21:02
experience your pragmatic reality and
00:21:04
your vision what would be things that
00:21:06
would kind of excite you and are there
00:21:07
anything things maybe you just kind of
00:21:09
throw away saying okay these guys are
00:21:10
just talking as a PowerPoint as
00:21:12
vaporware but potential but is there
00:21:13
anything that you would get excited and
00:21:14
jazzed about that you would look for as
00:21:16
a an incubator lead and maybe we'll go
00:21:18
order again maybe jerem me sure um
00:21:21
finding patients is tough to enroll into
00:21:24
clinical trials extremely hard like in
00:21:27
oncology um you typically only find a
00:21:29
patient as a referral um and then you
00:21:32
know in HIV and he viral hepatitis um
00:21:35
when we try to find patients who are
00:21:37
naive meaning they've never been treated
00:21:39
it's extremely hard um partially due to
00:21:42
our own success but yeah so like what
00:21:44
we've tried to start doing is take the
00:21:46
volume of not just us but Global
00:21:48
electronic medical records and build
00:21:51
massive arrays of predictive models that
00:21:53
will predict in advance a potential
00:21:55
successful screening or successful
00:21:57
diagnosis 6 months to 1 year before a
00:22:00
patient is even diagnosed um but that is
00:22:03
a massive undertaking and it is
00:22:05
something that I think if if a
00:22:07
really well-minded group of people could
00:22:10
solve that challenge to how do we
00:22:11
connect Global electronic medical
00:22:13
records together and make predictive
00:22:15
patient identification a reality I think
00:22:17
it would be fantastic awesome Yep two
00:22:20
two items for me that I would look for
00:22:21
from entrepreneurs there's a a real
00:22:24
desire for clinical research to be
00:22:27
embedded into clinical care today these
00:22:31
worlds operate connected but
00:22:34
separate um the second um would be that
00:22:38
most clinical trials that are run
00:22:42
fail um and it would be wonderful to
00:22:45
have um an assist of some sort um that
00:22:50
would help make it so that most clinical
00:22:53
trials
00:22:55
succeed great yeah yeah I by the way I
00:22:58
wish I an incubator one day I'm so I I
00:23:00
have this tendency of wanting to run
00:23:02
things versus sitting behind so that's a
00:23:04
problem so two things from my side um I
00:23:06
think investing in companies that are
00:23:09
aiming from day one to make things
00:23:11
affordable and scalable across the globe
00:23:13
I think as part of a mission even if it
00:23:15
takes them decades is something I always
00:23:17
look for it's I have something I've
00:23:18
always looked at when I joined my new
00:23:22
companies and the second thing is any
00:23:24
any kind of AI that will make my
00:23:26
teenagers listen to me I think would be
00:23:28
one wonderful so I I I would invest in
00:23:31
that like right now oh yeah I think most
00:23:33
of us in the room may feel along those
00:23:34
lines I'm totally with you thank you so
00:23:36
much I know we're at time uh just as a
00:23:38
quick final takeaway and what I'm
00:23:39
hearing as well is that um gen
00:23:41
opportunities of course we're all
00:23:42
excited about it we've heard that
00:23:44
statement over and over again but we've
00:23:45
gone beyond individual productivity
00:23:48
right the days of just saying chat and
00:23:49
GPT is you know a great productivity
00:23:52
tool we're looking at Value chain
00:23:53
disruption how can I actually change
00:23:56
business processes in a meaningful way
00:23:58
but she prioritize those decisions
00:24:00
thinking about risk in our space
00:24:02
Enterprise Life Sciences risk comes
00:24:04
first and we're we'll take a hit on the
00:24:06
ROI if we're able to manage those risks
00:24:09
and I think that's a powerful place I'll
00:24:11
I'll end it there uh we're at time but
00:24:13
thank you so much U really delighted and
00:24:16
hopefully we can hear more on this topic
00:24:17
throughout the rest of the sessions
00:24:19
thank you
00:24:20
[Applause]
00:24:23
[Music]
00:24:34
every
00:24:37
[Music]
