How Bristol Myers Squibb Is Accelerating Drug Development With AI – Ep. 264

0:00:16 Hello, and welcome to the NVIDIA AI podcast. I’m your host, Noah Kravitz. Before we dive in,
0:00:21 a quick reminder that the AI podcast now comes your way four times a month. Listen to us wherever
0:00:25 you get your podcasts. And if you have a minute to subscribe to the show and even to leave us a
0:00:31 review, we greatly appreciate it. The intersection of leading-edge technology and health and life
0:00:35 sciences is one of the most vital areas of research and development happening in the world
0:00:39 today. Artificial intelligence is fueling advancements in virtually all areas of healthcare
0:00:44 and medicine, as recent episodes of our show can speak to. Today, we’re talking with someone on
0:00:50 the forefront of integrating digital technology with groundbreaking science. Greg Myers is Executive
0:00:55 Vice President and Chief Digital and Technology Officer at Bristol-Myers Squibb, one of the
0:01:00 world’s foremost biopharmaceutical companies. Greg’s here to talk about how Bristol-Myers Squibb is
0:01:05 using AI to accelerate scientific pursuits and innovation, and what the future of technology
0:01:10 in the biopharma industry looks like. Greg, welcome, and thank you so much for taking the time to join
0:01:16 the AI podcast. Thanks so much for having me. So, could you set the stage for us by telling us a bit,
0:01:20 well, telling us a bit about Bristol-Myers Squibb, if you would, for those who might not be aware,
0:01:23 and then about your role and how you came into this role as well.
0:01:29 Yeah, of course. So, we, at Bristol-Myers Squibb, we discover, develop, and deliver innovative new
0:01:33 medicines that help patients prevail over serious diseases. Those are typically in areas like
0:01:40 oncology, hematology, immunology, cardiovascular disease, and neuroscience. Really, as a company,
0:01:44 we’re focused on being one of the fastest-growing companies in our industry by the end of the decade.
0:01:50 In fact, we’re on track to deliver 10 new medicines and 30 new indications by 2030, a lot of which is
0:01:56 being accelerated through rewiring the enterprise around digital data and AI, which is looking forward
0:02:02 to talking about that today. And then your team specifically, what are you sort of overseeing?
0:02:07 What is your team primarily focused on? Yeah, so we have the traditional technology or IT part of
0:02:12 the organization. My team also is responsible for all the data and analytics throughout the company,
0:02:18 so the majority of the data sciences team is within my remit. And we also have a digital health team,
0:02:23 which is a really exciting team that is looking at the forefront of changing the way patients are
0:02:27 diagnosed, treated, or monitored in key areas that we also tend to have therapies in at the point of
0:02:33 care. And so, how does a major company, a major pharma company like Bristol-Myers Squibb,
0:02:39 think about technology’s impact on your industry in general? And then maybe as you’re talking about that,
0:02:45 you could drill down specifically to how AI has changed things in the past, you know, several
0:02:50 years in particular. Yeah, if you think about the last hundred years, we’ve added 25 years of life
0:02:55 expectancy, and that’s primarily through human-led innovation, right? So think antibiotics, vaccines,
0:03:01 chemotherapy, synthetic insulin, immuno-oncology, of which we were pioneers in. But over time, it’s gotten
0:03:06 more and more difficult to find new medicines. It’s become more expensive. In a lot of cases, a lot of the
0:03:10 low-hanging fruit has been picked. And when you really think about the state of biology, you know,
0:03:16 within a human body, there’s about 40 trillion cells. And each of those cells is about one trillion
0:03:21 molecules. If you add it up, that’s seven octillion atoms. And any one of those atoms out of place
0:03:27 potentially could cause, cure, or prevent a disease. And in the chemical space, which is what a lot of
0:03:32 pharmaceuticals are focused on, there’s about 10 to the 60th power possible chemical entities that can be
0:03:37 created, of which only a small fraction has ever been synthesized in a lab by a human.
0:03:43 So when you really think about drug discovery, we see it as a lot of dark data and the ability to use
0:03:48 computation to uncover, you know, what are otherwise hidden patterns in data about biology. And one of
0:03:52 the things that we’re excited about is how you apply that against this vast, both chemical and
0:03:58 biological space to really find new pieces of information. And of course, as things have grown in
0:04:03 terms of compute capacity, we’re very excited about the ability to apply and unlock more and more of
0:04:03 that data.
0:04:10 And so how far back should we think when we’re talking about sort of the change from traditional
0:04:16 methods of drug discovery and development to this new world that we’re in, kind of newly in, but I
0:04:22 think firmly in, with AI-powered drug discovery and, you know, being able to simulate experiments and
0:04:26 do some of these calculations you’re talking about that weren’t previously possible?
0:04:31 Yeah, it might be good to just explain how historically drug development has occurred and sort of how we
0:04:35 see AI being woven in. And I think it is fair to say it’s kind of more of an evolution and being
0:04:40 woven in. So if you look at, probably most people learned at some point in school, the scientific
0:04:44 method, and the first three steps, if you were to summarize them, are more or less follow your gut.
0:04:48 And if you think about it, there are a lot of hunches that we have in drug development,
0:04:52 right? So there’s hunches on the disease side. So scientists may notice an unexplained
0:04:57 pattern. Like we might notice that a certain protein is overactive on a specific cancer
0:05:02 tumor. We then have hunches about a target, which means we would sort of brainstorm whether
0:05:08 a protein receptor or a gene is causably linked to that disease, whether or not that looks what
0:05:13 we call druggable, which means you can find something that can bind to it. And at that point,
0:05:17 you know, you sort of have this qualitative rather than quantitative hunches that you’re following.
0:05:22 So if we block this receptor, this will cause and potentially stop this negative effect.
0:05:28 And then the last hunches we tend to have is around the hypothesis. So we have to formulate a
0:05:34 testable statement like, you know, can we engineer a molecule to bind to a specific receptor? And can
0:05:40 that receptor itself suppress, for example, maybe the scarring in lung tissue? And then that really
0:05:46 drives a lot of the things that we do both in preclinical testing, which as well as what you do once you get
0:05:51 into the clinical trial process. Now, what AI has done is along the way, and I would say this has
0:05:56 happened in sort of bundles of capability, is it helps us to narrow down the hunches, right? So for
0:06:02 example, if you were to do to look at a cancer tumor, we would typically stain it and a human being
0:06:07 would look at it on a microscope and they’d be trying to look for certain aspects that help us
0:06:13 understand more about the biology of that tumor. But now you could take a whole slide, which is like a
0:06:19 whole sample of a cancer biopsy. And instead of staining it, you can have a convolutional neural net
0:06:24 be trained on looking at that entire slide. So rather than just looking at the tissue sample and what’s
0:06:29 happening in the stain versus unstained, you can actually start to look for hidden subtle features and
0:06:34 patterns by using things like different light spectrums that humans can’t see. It may help us with
0:06:41 diagnostic accuracy and potentially even providing us new insights into what the shape of the tissues are over
0:06:46 time. So what you’re really seeing here is not only the ability to start to use it in research, but you’re
0:06:52 already starting to see the first generation of what I call co-designed AI molecules making their way from
0:06:57 discovery into development. In fact, we have several of those ongoing right now. Can you speak to any of
0:07:03 this? Yeah, well, one that we have that’s actually going into what we call phase 1b, which is sort of
0:07:11 forced in human, is that we designed a protein that can basically degrade another protein that is
0:07:17 responsible in sickle cell disease for basically suppressing a natural gene. So when we’re infants,
0:07:23 we have a gene that produces one type of red blood cells. And after we’re born, and obviously because
0:07:28 our lungs aren’t working, so these blood cells are able to absorb a lot more oxygen content, right? So
0:07:33 they’re much more densely rich. Right, right. Once we’re born, those genes naturally turn off and our
0:07:38 adult hemoglobin genes turn on. And that produces different types of genes. Well, in people with
0:07:43 sickle cell disease, they have a mutation of that gene, which actually makes those cells sort of shape
0:07:48 into a sickle. They fold onto themselves and it makes it very difficult to actually get enough oxygen.
0:07:54 So, you know, what AI has helped us to do is to sort of engineer features of that molecule that
0:08:00 allowed us to be able to effectively turn back on the dormant genes you have as infants. And that
0:08:04 actually alleviates some of the symptoms of people with sickle cell disease. And so there’s examples
0:08:09 where we were using AI to do different types of things that actually help us design it. So we already
0:08:12 knew what we wanted to do, but in this case, it really helped us work through a series of engineering
0:08:18 problems that we had and making sure that it would attach itself to where we wanted it to attach to
0:08:21 and not actually attach itself to other proteins that we didn’t want to affect.
0:08:26 Right. How do you go about, and this might be kind of a boil the ocean type question,
0:08:32 so feel free to ram me in here, but how do you go about kind of quantifying or objectifying the impact
0:08:38 that AI is having in drug discovery and all of the things that you’re talking about? Obviously,
0:08:43 you know, getting a drug or some other therapy to market and seeing it help people is, I’m sure,
0:08:47 the ultimate way to quantify these things. But as you’re going through these processes,
0:08:50 what sort of stats or indicators do you look to?
0:08:55 Yeah, that’s a great question. You know, a few years ago, I would say maybe two years ago or so,
0:09:00 we started experimenting with incorporating what we call predictive molecule invention,
0:09:06 but these were basically predictions. So if you think about what you do when you’re discovering a
0:09:11 drug is you basically have a hypothesis and then you have to go run an experiment in a lab to validate
0:09:17 or invalidate that hypothesis. And what we’ve been able to do is to leverage the millions of compounds
0:09:21 we’ve already created historically, as well as the millions of experiments we’ve already done,
0:09:25 being able to bring that together to actually build predictive models about whether or not
0:09:27 what we’re trying to test is likely to succeed or not.
0:09:35 And when we started incorporating that into our early research, we actually saw a noticeable step
0:09:41 change in the percent successful outcomes of the experiments we were running. So much so that our
0:09:46 entire research organization has moved to a world where in about 100 percent of what we call our small
0:09:52 molecule discovery, so those are things that are sort of think of as pills, and 50 percent of our large
0:09:57 molecule, which are these are things think about as being injectable or infusibles, we would not run
0:10:03 experiments in the wet lab until those predictive models in those two areas respectively suggest it would be
0:10:08 worth trying. So that’s kind of an example where we saw data along the way through proofs of concept,
0:10:12 and each of those allow us to get to a scaling piece. Similarly, in drug development, which is the
0:10:18 most expensive, complicated part of our business, it’s not unusual for it to take seven to 10 years
0:10:22 to get a drug through the development process. So, you know, any amount of time we can get to
0:10:26 shorten that is worth a lot to patients and worth a lot to us. You know, we’ve done a number of
0:10:33 experiments around, you know, how we use large language models to help us try to get us started in
0:10:37 writing a lot of the documentation. Clinical trials have over 300 different documents that have to get
0:10:42 created to get them going. And if you really think of what LLMs are great at, it’s reading and
0:10:47 writing and editing words on a page, which is really a lot of what goes on there. And as a result of some
0:10:51 of those works, I mean, we really started and we’re just starting to get on the other end of, for example,
0:10:56 something like an informed consent form, which is a very simple form. It’s what you give patients to let
0:11:01 them know what they’re signing up for a trial. Like we know that we can get 80% of that pre-written
0:11:06 just by using the master document that that comes from, right? Whereas originally that’d be on a clean
0:11:11 sheet of paper. So as we see these experiments, you know, we build more and more conviction and that
0:11:13 allows us to move more towards scaling.
0:11:18 And so that leads me to ask you kind of about the, I’m going to say the other side of this,
0:11:24 but moving from talking about the outcomes, the results of using AI and discovering more drugs
0:11:28 faster and everything you’ve been talking about, and kind of to the other side of the people working
0:11:34 on these things. And you mentioned, it’s a case I can relate to in my own work when I’m writing and
0:11:39 not, or even writing up the show notes, right? That’s what Gen AI is so good at, scanning large chunks of
0:11:45 text, summarizing, helping you rewrite. What are some of the other AI powered tools that your teams
0:11:50 are using in their day-to-day work? And how are they, you know, not just improving, but maybe
0:11:55 reshaping the way that the work gets done and that people think about the work they’re doing?
0:12:00 Yeah. You know, we were really early adopters on this. In fact, I would say it was maybe three months
0:12:04 after ChatGPT 3.5 came on the scene. Okay.
0:12:07 We were the first company that I’m aware of. I’m not saying there aren’t any, but the first that I’m
0:12:12 aware of that actually built an internal chatbot. And we use Microsoft Teams. We actually built a
0:12:20 Teams bot that was connected to ChatGPT 3.5. That’s since primarily running 4.0 now, but that was built
0:12:24 within the Teams chatbot. And then since then, we’ve expanded that out to a whole, what we call a
0:12:28 Gen AI storefront. And what’s interesting about that is you can go to that storefront and you can use
0:12:35 all the frontier models. So as of today, you’ve got 03 and 04 from OpenAI. You can use Anthropic 2.7.
0:12:43 You can use DeepSeq V2. You can use Gemini 2.5. So they’re all available for free to our employees.
0:12:47 And I was telling someone the other day, if you were to pay out of pocket, you’d be paying $300 a month.
0:12:52 And then we also got about close to 2,500 licenses of Microsoft Copilot out there. So our philosophy was
0:12:57 to really, well, I think a lot of companies were thinking about how to be worried about what would
0:13:01 happen giving, you know, tools to our employees. You know, we went out of the gate and really wanting to
0:13:06 make sure that we gave people proper room to experiment to really use those tools.
0:13:06 Yeah.
0:13:11 And when you think about our industry, you know, there’s a couple of tailwinds. One, it’s a very
0:13:17 much a knowledge-intensive business with lots of documents. So that lends itself well. Also, we tend to be
0:13:22 in an industry where a lot of wheels are being continuously reinvented. So your ability to capture and leverage
0:13:26 knowledge about the past is important. But then there’s also headwinds, right? Very highly regulated
0:13:31 industry with a high cost of failure. And so we thought it was important to just get people
0:13:34 accessibility to the tools to be able to do more and more.
0:13:41 When you started rolling out that first chatbot, those first iterations, what was the response like
0:13:46 from your employees, from your teams? Were people excited? Were people hesitant? Were there
0:13:51 specific worries or specific, you know, sort of tasks and experiments that people latched onto
0:13:52 from the get-go?
0:13:55 It won’t surprise you, my answer. I think it was a mixed bag. I think in the beginning,
0:13:59 at the very early adoption, a lot of people didn’t, you know, in the early days of chatbot,
0:14:04 I mean, I think people didn’t know they would look at the blinking cursor and not sure what to do,
0:14:10 right? There was a lot of training that we had to do. There was a surprisingly large number of people
0:14:15 people who kept asking questions like, well, is it okay to put proprietary information in? Because
0:14:21 I think in the information security world, we do such a good job training people to not put anything
0:14:26 into systems that, you know, don’t have the word SAP or whatever on it. And we had to actually make it
0:14:30 clear that, hey, no, the reason we built these tools is we wanted to build safe alternative. We
0:14:35 didn’t want people going out and leaking proprietary information into consumer tools. So that was a little
0:14:40 bit of a hurdle. And then, as you can imagine, you know, the early use cases are, you know,
0:14:45 helping edit a performance review and editing emails. And now they’ve progressively moved more
0:14:51 and more onto analytical skill sets, data science capabilities, RAG-based capabilities. So it’s
0:14:54 certainly gone up to mature. And I would say that we’ve adapted as the technology has matured.
0:14:59 And on the, I don’t want to say the research side of things to imply that this work isn’t part of
0:15:03 the research, but when we’re talking about researching, you know, how the molecules interact with
0:15:07 each other and how the protein receptors work and all of the things you were talking about with drug
0:15:12 discovery and such, are you building and training your own models? Are you fine-tuning off-the-shelf
0:15:18 models? You know, what can you kind of say about the technology and the AI-specific end of that?
0:15:21 Yeah, you’re talking about computational chemistry and biology. That’s what we call them.
0:15:27 And a lot of that work is being done on the NVIDIA clusters. And so what we’re doing there is we
0:15:33 typically have certain things that we’re trying to do. Like, let’s say it’s very common in drug
0:15:38 development or I should say drug discovery where you’re trying to find, hey, here’s a molecule that
0:15:44 we know works, but it’s having trouble crossing the blood-brain barrier. So we have to sort of,
0:15:49 we don’t want to affect the active parts of the molecule, but we need to affect things around it that
0:15:55 affect its ability to either be something that can go through that barrier or can’t, right? So
0:16:02 so lipophilic, lipophobic. So these are features that you can use models that already exist. In the
0:16:07 case of protein structure prediction, we have alpha fold. There are other modalities we’re using where
0:16:12 we’re using large language model sequences to look at protein structure prediction. Those are things like
0:16:18 ESM fold. So a large extent, we’re using mostly off-the-shelf open-source models on high-capacity
0:16:24 compute that helps us do things like physics-based simulations and things like that. So that’s where a lot of
0:16:29 that work is. It’s still probably at the early adopter phase, to be honest, because a lot of
0:16:34 this stuff is all very new. And so a lot of what we’re working on is how we try to find tools that
0:16:39 have abstraction wires because not every scientist knows Python fluently or TensorFlow, right? So we’re
0:16:44 starting to, and I know NVIDIA has also done quite a bit of work with Bionemo and a few other things to
0:16:49 do that. So we’re working on building abstraction wires to kind of create these co-pilots, for lack of a
0:16:53 better word, to help scientists with really specific problems that they face on a day-to-day basis.
0:16:59 I’m speaking with Greg Myers. Greg is executive vice president and chief digital and technology
0:17:04 officer at Bristol-Myers Squibb, one of the world’s leading biopharmaceutical companies. And we’ve been
0:17:12 talking about Bristol-Myers Squibb’s use of AI internally to really to accelerate all parts of their
0:17:17 scientific pursuits and innovative efforts and everything related to bringing more and more
0:17:25 effective therapies to people who need them around the world. Greg, I want to ask you to kind of step
0:17:31 back or sort of to a higher level from, and I say just Bristol-Myers Squibb because obviously an enormous
0:17:37 operation, but to kind of look at the biopharma industry sort of more holistically. You’ve, as you’ve
0:17:42 been saying, you know, you’ve, you’ve had more than a taste of using AI. You guys were early on at your
0:17:48 own work. How do you see the technology affecting the industry as a whole and kind of specifically when
0:17:54 it comes to workforce and workforce productivity and also operations? Well, if you look at healthcare
0:18:01 more broadly, I mean, 30% of all data that exists in the universe is, is healthcare data, scientific and
0:18:06 the like. Right. But the big challenge is it’s all trapped in these, these little silos. So there are
0:18:11 thousands of hospitals, for example, that all have different electronic medical records at all. They
0:18:16 code things a little bit differently. Each pharmaceutical company like ours will have their
0:18:21 own cadre of data. Many of them, many of us are products of mergers and acquisitions over the years.
0:18:25 Those things seem to be trapped in a lot of legacy IT applications. Sure. So really when you think
0:18:29 about, and then of course, insurance companies have a tremendous amount of information as do
0:18:33 governments. When you get outside the U S the primary payers for healthcare are governments.
0:18:38 So when you think about the integration of all that data together, I think it’s going to be really
0:18:43 critical. I’ll give you, give you an example. If you take lung cancer, it is really the deadliest form
0:18:49 of cancer. And the reason why it’s the deadliest form is because it’s mostly caught too late. And the
0:18:53 reason it’s caught too late is the majority of lung nodules when they’re treatable are just not
0:18:59 detectable by a human being looking at the imaging that they get. But we know that as these
0:19:06 technologies progress, you have the ability not only to diagnose patients earlier, but even when you get
0:19:11 back to the point I was talking about earlier around using AI to look at tumors, you have the ability to
0:19:16 start looking at which patient is likely to do better on which therapy versus another. Right, right.
0:19:20 And in lung cancer, the other reason why you have a high mortality rate is the average lung cancer
0:19:24 patient will fail on four or five different therapies before they find one that works.
0:19:30 And if you only have five years to live, you don’t have time to go through four or five. So your
0:19:35 ability to get on the right treatment earlier and not start a treatment that’s not going to work is
0:19:39 good for everyone. It’s good for the patients. The pharmaceutical industry doesn’t want to be
0:19:43 supporting patients that we can’t help. And of course, payers don’t want to pay for medicine that
0:19:44 they don’t need to pay for. Right.
0:19:51 So it’s my base case that in 20 years, as a life sciences innovator, just shipping a molecule alone
0:19:55 won’t be enough. We’re going to need to work with the rest of the industry to bundle together
0:20:00 real-world evidence about how medications work in the real world, data that comes from live patients
0:20:07 about how they metabolize and respond to therapy differently. As these things all come together,
0:20:12 you know, I’m, I’m really hoping for a major breakthrough in the way patients are going to
0:20:17 be treated because in the next 25 years, um, the number of people over the age of 65 will double.
0:20:22 And those same number, the number of people with cancer will double. One out of every three people
0:20:27 will have some neurodegenerative disease. So we have a high incentive across the whole industry to really
0:20:32 find a way to put data together, uh, both about science, about biology and about individual people
0:20:37 to try to find ways to, to achieve better outcomes for patients around the world.
0:20:42 And, you know, given the stats you just cited about the, the population aging and the number
0:20:46 of people with cancer doubling and everything, I thought back to what you said kind of near the
0:20:51 beginning of our conversation about how human life expectancy, I think you said over the past
0:20:57 hundred years, it’s increased by 25 years. That makes me feel optimistic even, you know, in the
0:21:02 face of thinking, Ooh, I’m one of those people who’s going to be over 65 in the next 25 years.
0:21:07 At the same time, what you said about electronic medical record, electronic health records,
0:21:11 medical records in particular, you know, struck a chord with me because as a healthcare consumer,
0:21:17 to put it that way, I faced that frustration, right? Of my data, not getting to, you know,
0:21:24 from my primary care to a specialist or whatever it may be is kind of cooperation and standardizing
0:21:30 the way that medical data is collected and shared. Is that still one of the biggest impediments
0:21:35 to progress? Or is that something that you feel like we’re making headway on and it’s really,
0:21:40 you know, focusing on, I don’t know if the term precision medicine applies here, but, you know,
0:21:45 in thinking about that, well, how do we narrow it down, get it down from the lung cancer patient
0:21:49 needs five therapies to hit one, to getting it, you know, on the first or second try?
0:21:54 Yeah. I mean, so electronic medical records really grew during the Obamacare time. And I think if we’re
0:21:59 honest about what they’ve evolved into, they’re probably more optimized for accounting, billing,
0:22:05 and reimbursement than for actual patient outcomes. So no doubt there’s an opportunity to standardize
0:22:11 and sort of reframe the ability to try to optimize for outpatient outcomes in addition to making sure
0:22:16 people get paid properly. But I’m also really encouraged that AI, especially with new reasoning
0:22:22 models, actually bridges some of this. So if you think about having slightly different ways of
0:22:27 explaining a disease, a reasoning model that is trained on the disease can actually bridge that
0:22:31 gap much better than trying to do, you know, sort of cross-reference tables or however you would do
0:22:38 that traditionally. Yeah. I’m also really excited. You’re really now just starting to see tools getting
0:22:45 put into the point of care that really are changing, not only sort of just specific things that occurred
0:22:49 today, but even helping uncover diagnoses people didn’t even know they had, you know, we had the
0:22:57 benefit last year working with a partner on getting a tool into the hands of cardiologists to be able to
0:23:03 detect a disease. This is a heart disease called hypertrophic cardiomyopathy. The only way that that
0:23:09 is, so there’s oftentimes a patient with that won’t know they have it, will go 10 years being misdiagnosed
0:23:16 because it is actually really hard for a physician to detect and can really only be found by a highly
0:23:22 skilled cardiologist looking at an echocardiogram. Right. And most patients won’t get an echocardiogram
0:23:27 because they’re very expensive to prescribe. Yeah. So what we were working on with this partner is
0:23:33 being able to create the ability to look at a simple 12-lead ECG. So that’s, you know, when you get the
0:23:39 little stickers put on your body and you go into a doctor’s, most general practitioners even have these.
0:23:46 The ability to actually find the signature of that disease on a 12-lead ECG really unlocks the potential
0:23:51 to uncover people who didn’t know they have the disease. And that’s really important because a lot
0:23:56 of times if you hear about high school athletes maybe falling over and dying on the field, this is from a
0:23:57 disease like this. Right, right.
0:24:04 So the ability when someone goes into a clinic to do something like a routine blood workup, the ability
0:24:09 to retrospectively find that there is a detection of that signal and actually signal them to go see a
0:24:15 cardiologist versus waiting for them to evolve to a symptoms is something that, you know, really has a
0:24:20 tremendous amount of opportunity in that you don’t have all these having to go through this labyrinth of
0:24:23 tests that are unnecessary and being able to get people to point of care before they really progress.
0:24:26 And those are the things, you know, I’m really excited about.
0:24:33 It reminds me of an episode we did and listeners forgive and correct me, I’m getting this wrong,
0:24:39 but I believe it was about AI and cardiac care, actually, where the guest was talking about just
0:24:47 the value of data in kind of legacy images, right? Images that have been captured, you know,
0:24:53 previously. And now with these machine learning AI tools, they’ve developed ways to go back and sort
0:24:59 of reinvestigate, you know, re look at these images again, but with the AI tools and just finding all this
0:25:05 information in there. You know, I think they were they were able to find a marker that had previously, you
0:25:11 know, been undiscovered, just by virtue of going back and reexamining all of these old images.
0:25:18 Yeah, that’s exactly this situation, because this is a situation where the it’s clear that the signature of the
0:25:22 disease appears on a 12 lead ECG, even though it’s not obvious.
0:25:22 Yeah.
0:25:28 In fact, we did an experiment a year ago where I was talking to some leading cardiologists and I said,
0:25:33 you know, we can actually predict the patient’s age based on their ECG. And that is just not intuitive
0:25:37 to it. There would be no way a cardiologist, unless you’re an infant, you probably can tell the difference
0:25:41 between a 90 year old and a two year old, but not the difference between a 30 and a 40 year old.
0:25:47 So it really just goes to show that there is so much hidden data that things like neural networks
0:25:50 are able to uncover. And obviously the heart is great. It’s the only digital organ we have. It’s
0:25:54 either polarized or depolarized. And so it lends itself well to this to this area.
0:26:00 Right. So, Greg, I’m going to ask you to look forward as as we kind of start to wrap up our
0:26:05 conversation here. And, you know, we ask everybody this question and obviously it’s impossible to answer
0:26:10 in some respects, especially given the pace of innovation recently in machine learning and AI.
0:26:15 But kind of big picture, where do you see all of this headed in the biopharma industry?
0:26:21 Kind of, you know, maybe two years down the line, what’s on the horizon when it comes to
0:26:26 what technology and biopharma together might be able to achieve?
0:26:28 Well, I think it’s some of the things we talked about before. I mean,
0:26:33 healthcare is a really, I mean, if you take in the U.S. alone, it’s a $4 trillion industry that
0:26:39 everybody has a long list of complaints about. So you have to believe that that is ripe for a lot of
0:26:44 change in terms of how we manage patients, in terms of how drugs are developed. And I think
0:26:48 as the industries come together, there’s a real opportunity to not only accelerate. And obviously,
0:26:53 our hope is that, you know, we can generate new molecular, more new molecular entities faster.
0:26:57 We can get them through the clinical trials faster. And if you take, for example, you know,
0:27:02 we’ve launched a number of new medicines that are actually what are called first in class medicines,
0:27:07 which means there was not really a previous, previously a direct treatment. And so if you can get a drug to
0:27:13 market six months, 12 months, two years faster, and by the way, we’re on track to shave off almost
0:27:18 three years off of our clinical trial timeline as a result of using digital and data and AI and process
0:27:21 changes. Three years off of that, was it a seven to 10 year?
0:27:23 10. Yeah, also 10 years. Yeah, around nine.
0:27:24 That’s huge.
0:27:30 Yeah. Yeah, that is huge. And so obviously, if you’re a patient, and again, if you look at our pipeline of
0:27:34 therapies, what you’ll find is these are patients who are waiting for something because they have nothing. And if you’re
0:27:39 not getting into a clinical trial, you’re just not getting treated. So the ability to get something to patients two to
0:27:46 three years earlier can be transformational. And I think if you look at cancer care, and we are on the precipice of a whole new
0:27:52 set of therapies, we have cell therapy, which is still growing. We have antibody drug conjugates, we have
0:27:58 radioligin, we have many, many other things that are happening, that getting them to the pipeline will really
0:28:04 transform the care of cancer. So this is something that I’m really excited about. Because, you know, when I think
0:28:10 about the future population, where things are, the ability for us to get out and treat these chronic diseases that we
0:28:14 know we’re going to grow naturally is really going to be transformational for patients.
0:28:20 Absolutely. On a personal note, how do you stay on top of everything? You know, from my end, trying to
0:28:27 stay up on what’s happening with just the technology is a job and a half. But applying that to something as
0:28:33 complex and huge as biopharma, how do you stay on top? And what kinds of news when you read about kind of
0:28:37 lights your fire and makes you think like, oh, wait, this is something to keep an eye on, because I think this
0:28:41 might be a game changer. Yeah, you’re right. You know, running a technology function inside of a
0:28:46 big enterprise, I often sort of joke, you know, if you if you’re in accounting, accounting standards
0:28:52 haven’t changed since the 1970s so much. If you’re a lawyer, while there is additional law, we’re still
0:28:58 operating off of laws that are 50, 100 years old. And in technology, really, you have to be almost
0:29:05 prepared for a complete turnover of technology stacks every five years. It is really hard just to
0:29:09 acknowledge what you’re saying to keep up. Personally, what I do is basically read it and
0:29:14 podcasts. Those are the two places I get most of my information from probably read it is a better
0:29:20 source of news and podcasts are a great source to get much deeper into those areas. Of course,
0:29:25 I read a lot of journal articles, and it’s been great to use, you know, notebook LM to convert those
0:29:31 to podcasts to read. But yeah, I think that’s really what I do is I do really believe that a big part
0:29:36 of my job personally is focusing on what’s going on outside the four walls of Bristol Myers Squibb,
0:29:41 because I’ve got a lot of people in the organization thinking deeply about what is going on within our
0:29:46 company day to day. But, you know, I see my job is to sort of try to see around corners a little bit
0:29:51 more and figure out what to pay attention to. Because as you know, there’s a lot of noise intermixed
0:29:54 with a lot of signal and being able to sift through that is challenging.
0:29:59 Absolutely. Well, on a personal note for me, now I can go to the dinner table tonight and defend my
0:30:06 Reddit habit to my kids. So fantastic. Greg, for listeners who would like to know more about what
0:30:11 Bristol Myers Squibb is doing, about biopharma, about any of the things that you’ve been talking about,
0:30:16 is the Bristol Myers Squibb website the best place to start? Is there kind of a sub site,
0:30:20 technical blog, something like that? Where would you direct folks to go learn more?
0:30:25 Yeah, bms.com. You’re going to find there all the portfolio of things that we currently
0:30:29 provide to patients, as well as a number of really exciting therapies we have in development. We have
0:30:34 one of the richest pipelines in the industry. You’ll find a lot of that online and also follow us on
0:30:38 LinkedIn. We’re pretty active there. Fantastic. Greg Myers, thank you again for taking the time
0:30:43 to come speak with us as a person who relies on healthcare every day. Goes without saying,
0:30:46 but all the best of luck to you and your teams on the work you’re doing.
0:30:48 Yeah, thank you very much. It was a pleasure being with you today.
0:31:45 Thank you.