WEBVTT - Sanofi CEO Paul Hudson Talks AI for BioPharma

0:00:02.400 --> 0:00:06.760
<v Speaker 1>Bloomberg Audio Studios, Podcasts, Radio News.

0:00:07.040 --> 0:00:09.120
<v Speaker 2>Look, we've got to talk about the impact of generative

0:00:09.119 --> 0:00:12.879
<v Speaker 2>AI in real life across industries, Biofarmer being one of them,

0:00:13.119 --> 0:00:14.920
<v Speaker 2>and I'm priessed to welcome with the show Paul Hudson,

0:00:15.040 --> 0:00:18.840
<v Speaker 2>CEO of Sanofi, discuss how your enormous French her Farmer

0:00:18.880 --> 0:00:22.680
<v Speaker 2>company has been leveraging the new technology in its operations.

0:00:23.000 --> 0:00:26.200
<v Speaker 2>And yes, there's drug discovery, but also what's so interesting, Paul,

0:00:26.280 --> 0:00:29.080
<v Speaker 2>is you're integrating it within the functionality of your business

0:00:29.080 --> 0:00:30.680
<v Speaker 2>on a day to day basis.

0:00:31.280 --> 0:00:33.560
<v Speaker 1>Yeah, thank you, and it's a pleasure to be here.

0:00:33.600 --> 0:00:34.440
<v Speaker 1>We're really trying.

0:00:34.280 --> 0:00:36.960
<v Speaker 3>To do incredible things for patients and discovering new medicines

0:00:37.000 --> 0:00:40.320
<v Speaker 3>and you know, with AI being the great disruptor and

0:00:40.400 --> 0:00:43.360
<v Speaker 3>all of the excitement last year, the wow of large

0:00:43.400 --> 0:00:45.680
<v Speaker 3>anguige models, and this year for us, it's all about

0:00:45.680 --> 0:00:48.400
<v Speaker 3>the how and how we can get our you know,

0:00:48.520 --> 0:00:52.760
<v Speaker 3>the large amounts of our organization focused whether you're in

0:00:53.159 --> 0:00:56.840
<v Speaker 3>a lab doing expert AI, whether using generative AI in

0:00:57.480 --> 0:01:02.480
<v Speaker 3>regulatory or text driven of our business and importantly snackable

0:01:02.480 --> 0:01:05.360
<v Speaker 3>AI when we're looking at really trying to give better

0:01:05.400 --> 0:01:09.360
<v Speaker 3>decision intelligence for almost everybody in the organization with the

0:01:09.480 --> 0:01:12.759
<v Speaker 3>sole intent of becoming much more efficient effective in delivering

0:01:12.800 --> 0:01:13.160
<v Speaker 3>new meds.

0:01:13.200 --> 0:01:13.920
<v Speaker 1>And it's amazing.

0:01:13.959 --> 0:01:16.880
<v Speaker 2>Really, I love this snackable AI you're using play it's

0:01:16.920 --> 0:01:19.320
<v Speaker 2>a proprietary app. But I've got to ask sort of,

0:01:19.560 --> 0:01:22.600
<v Speaker 2>how have you thought about which underlying large language model

0:01:22.600 --> 0:01:24.840
<v Speaker 2>you should look to. How you integrate this do you

0:01:24.880 --> 0:01:27.560
<v Speaker 2>go to outsourced do you build it from within.

0:01:27.400 --> 0:01:32.520
<v Speaker 3>Pol Well, we're drug discoverers and developers, we're not coders

0:01:32.600 --> 0:01:35.720
<v Speaker 3>at our heart. So we have this first of its

0:01:35.800 --> 0:01:39.880
<v Speaker 3>kind relationship now with open AI, partnership with Formation bio

0:01:39.959 --> 0:01:42.680
<v Speaker 3>translating that for is to make sure that we can

0:01:42.720 --> 0:01:46.520
<v Speaker 3>be customizing large language models to look at drug discovery,

0:01:46.880 --> 0:01:50.880
<v Speaker 3>to try and drug undruggable diseases. To this point, and

0:01:50.920 --> 0:01:53.720
<v Speaker 3>we're doing things that have really never been done before.

0:01:53.760 --> 0:01:57.559
<v Speaker 3>It's early days, but it's extremely promising. And remember there's

0:01:57.600 --> 0:02:00.400
<v Speaker 3>such a large amount of diseases that don't have medsons

0:02:00.400 --> 0:02:02.640
<v Speaker 3>available for them because we simply have not had the

0:02:02.680 --> 0:02:06.760
<v Speaker 3>computing pairer of the opportunity to deploy ourselves. And now

0:02:07.040 --> 0:02:10.440
<v Speaker 3>just now we're at this next great revolution in science.

0:02:10.520 --> 0:02:12.920
<v Speaker 3>It's such a privilege actually to lead a company at

0:02:12.919 --> 0:02:17.519
<v Speaker 3>this point. As for play, our sort of snackable approach,

0:02:18.040 --> 0:02:22.000
<v Speaker 3>we partnered with a small startup out of Germany, a Ley.

0:02:22.360 --> 0:02:25.200
<v Speaker 3>They've been working withthers, They've grown weathers and that's, by

0:02:25.200 --> 0:02:26.480
<v Speaker 3>the way, open for everybody.

0:02:26.560 --> 0:02:28.560
<v Speaker 1>We don't own it, were having to share it.

0:02:28.600 --> 0:02:31.239
<v Speaker 3>We want all boats to rise win healthcare after all,

0:02:31.560 --> 0:02:34.440
<v Speaker 3>and so the more users, the more fun, the more accuracy,

0:02:34.680 --> 0:02:36.320
<v Speaker 3>the more chance to discovering new medicines.

0:02:36.480 --> 0:02:39.359
<v Speaker 2>So have your talent reacted because there's so much do

0:02:39.600 --> 0:02:41.760
<v Speaker 2>mongering that AI is going to take all our jobs.

0:02:43.520 --> 0:02:45.880
<v Speaker 3>Well, I think our approach has been how to weaponize

0:02:45.880 --> 0:02:47.080
<v Speaker 3>our people to do better.

0:02:47.160 --> 0:02:49.240
<v Speaker 1>Like you know, I don't know whether you're familiar with.

0:02:49.160 --> 0:02:51.840
<v Speaker 3>Ways, the driving app, but I see it very much

0:02:51.960 --> 0:02:53.720
<v Speaker 3>like that. I got to get in the car. I'm

0:02:53.720 --> 0:02:55.360
<v Speaker 3>going to start the engine and drive to the end

0:02:55.400 --> 0:02:58.520
<v Speaker 3>of the street. And if Ways says I'm statistically more

0:02:58.639 --> 0:03:01.320
<v Speaker 3>likely to have an easier journey turning left rather than

0:03:01.400 --> 0:03:05.080
<v Speaker 3>turning right, I want that advice as crowdsource large data,

0:03:05.120 --> 0:03:07.720
<v Speaker 3>real time. I want our people to feel they are

0:03:08.120 --> 0:03:10.120
<v Speaker 3>amplified or multiplied in that event.

0:03:10.600 --> 0:03:14.120
<v Speaker 1>In that event too, The goal is not about less people.

0:03:14.160 --> 0:03:17.079
<v Speaker 3>The goal is about putting our people to be more

0:03:17.120 --> 0:03:20.680
<v Speaker 3>effective and to make and to have better decision intelligence

0:03:20.720 --> 0:03:23.720
<v Speaker 3>that in itself gives us a great opportunity to do

0:03:23.800 --> 0:03:28.480
<v Speaker 3>something incredible. And we've only seen huge Appetite's why on

0:03:28.560 --> 0:03:32.360
<v Speaker 3>any given day we have eighteen thousand people in our

0:03:32.480 --> 0:03:35.320
<v Speaker 3>organization using the EI we make available and I think

0:03:35.360 --> 0:03:36.440
<v Speaker 3>that's pretty fantastic.

0:03:36.880 --> 0:03:40.160
<v Speaker 2>Let's talk about doing the amazing the incredible, because we

0:03:40.240 --> 0:03:42.760
<v Speaker 2>have just seen a push up in your share price

0:03:42.800 --> 0:03:45.680
<v Speaker 2>on the day and you've been reorientating this business of

0:03:45.760 --> 0:03:48.760
<v Speaker 2>Sonofye and thinking about more R and D, about drug

0:03:48.760 --> 0:03:52.440
<v Speaker 2>discovery and applicate and look some costly cases in the

0:03:52.480 --> 0:03:54.800
<v Speaker 2>way that you have to experiment do clinical trials, but

0:03:54.840 --> 0:03:56.880
<v Speaker 2>they've paid off today and we think about the drug

0:03:56.880 --> 0:04:00.480
<v Speaker 2>discovery when it comes to a key lung chronic lung disorder.

0:04:00.520 --> 0:04:01.760
<v Speaker 2>Can you just tell us about what that means to

0:04:01.800 --> 0:04:02.240
<v Speaker 2>the business.

0:04:03.000 --> 0:04:06.680
<v Speaker 3>Well, in Europe, just yesterday today we got approval for

0:04:06.840 --> 0:04:10.320
<v Speaker 3>chronic constructive coon me disease. COPD picks a huge number

0:04:10.320 --> 0:04:13.119
<v Speaker 3>of people third biggest killer. We're ready to go now

0:04:13.160 --> 0:04:15.840
<v Speaker 3>in Europe to be able to provide an advanced therapy

0:04:15.840 --> 0:04:17.920
<v Speaker 3>that will reduce their chance to go to hospital by

0:04:18.160 --> 0:04:20.680
<v Speaker 3>a third, and that's a breakthrough.

0:04:20.200 --> 0:04:21.479
<v Speaker 1>And there's been so little innovation.

0:04:22.120 --> 0:04:24.599
<v Speaker 3>We aspire to be the world's leading immnology company, and

0:04:24.640 --> 0:04:28.200
<v Speaker 3>that's partly driven by moments like that that we live

0:04:28.279 --> 0:04:31.520
<v Speaker 3>for in our industry to do things that transform or

0:04:31.560 --> 0:04:34.159
<v Speaker 3>protect patients lives, will prevent disease.

0:04:34.720 --> 0:04:34.960
<v Speaker 1>You know.

0:04:35.200 --> 0:04:39.480
<v Speaker 3>The reality is we're a very proud, huge, big company,

0:04:39.560 --> 0:04:42.440
<v Speaker 3>but we wanted to be at the forefront of science

0:04:42.440 --> 0:04:45.279
<v Speaker 3>again and we've made some big bets, but appropriately so

0:04:45.520 --> 0:04:47.719
<v Speaker 3>to try and make sure our science is some of

0:04:47.839 --> 0:04:50.760
<v Speaker 3>the most exciting and it's first in class or best

0:04:50.800 --> 0:04:53.880
<v Speaker 3>in class. And to couple it with our conversation of earlier,

0:04:53.880 --> 0:04:56.360
<v Speaker 3>the more we add AI, the more we improve our

0:04:56.400 --> 0:05:00.240
<v Speaker 3>decision intelligence, the more opportunity we have to do more

0:05:00.240 --> 0:05:02.719
<v Speaker 3>in that space. And I think that's what's catching the

0:05:02.760 --> 0:05:07.000
<v Speaker 3>excitement in our organization, and we believe we're quite far ahead.

0:05:07.040 --> 0:05:09.640
<v Speaker 3>I mean, of course, we're very responsible about how we

0:05:10.000 --> 0:05:12.360
<v Speaker 3>use AI, but others are still hesitating.

0:05:12.400 --> 0:05:15.120
<v Speaker 1>We're taking our moment, but sometimes you need.

0:05:15.040 --> 0:05:18.359
<v Speaker 2>Good data to feed in to these large language models. Paul,

0:05:18.680 --> 0:05:21.280
<v Speaker 2>what about the data on viruses such as well bird

0:05:21.320 --> 0:05:24.800
<v Speaker 2>flu At the moment here in the UK and the US. Indeed, Look,

0:05:24.839 --> 0:05:26.680
<v Speaker 2>there is a worry that the data hasn't been good

0:05:26.800 --> 0:05:28.720
<v Speaker 2>enough in the United States. Is that something that worries

0:05:28.839 --> 0:05:30.880
<v Speaker 2>you and you need to be developing a new vaccine

0:05:30.880 --> 0:05:31.080
<v Speaker 2>for that.

0:05:32.279 --> 0:05:34.760
<v Speaker 3>Look, I mean we've been through one pandemic. I think

0:05:34.760 --> 0:05:36.680
<v Speaker 3>we've learned a great deal about trying to have the

0:05:36.720 --> 0:05:39.680
<v Speaker 3>best data, and I think we position ourselves to take

0:05:39.680 --> 0:05:42.600
<v Speaker 3>advantage of those opportunities on behalf of society. Frankly, we

0:05:42.640 --> 0:05:46.119
<v Speaker 3>played a part small part in COVID, but we stand

0:05:46.200 --> 0:05:49.159
<v Speaker 3>ready should we be asked the data and it's legitimacy,

0:05:49.279 --> 0:05:52.040
<v Speaker 3>you know. We often that data comes in through government channels,

0:05:52.040 --> 0:05:54.679
<v Speaker 3>through health systems, you know, and we work at real

0:05:54.960 --> 0:05:58.120
<v Speaker 3>time speed to make sure that we have the very

0:05:58.240 --> 0:06:01.799
<v Speaker 3>best data available in terms of large language models and

0:06:01.960 --> 0:06:04.080
<v Speaker 3>the rest of the data sets.

0:06:04.400 --> 0:06:06.320
<v Speaker 1>You know, we collaborate far and wide.

0:06:06.360 --> 0:06:10.159
<v Speaker 3>We have punishes with oak in with ADOM, wise with excyensia.

0:06:10.320 --> 0:06:12.960
<v Speaker 3>We really are trying to make sure that we have

0:06:13.480 --> 0:06:15.720
<v Speaker 3>at our fingertips huge data sets.

0:06:15.760 --> 0:06:17.680
<v Speaker 1>To mean, our accuracy.

0:06:17.320 --> 0:06:20.240
<v Speaker 3>And our predictive capability is far ahead of the competition.

0:06:20.480 --> 0:06:23.919
<v Speaker 3>We won't get it right everywhere, but the data integrity

0:06:24.120 --> 0:06:27.839
<v Speaker 3>of the absolute fundamental start in place. Our relationship with

0:06:27.920 --> 0:06:30.600
<v Speaker 3>open AI and with Formation by It, we're putting our

0:06:30.640 --> 0:06:34.080
<v Speaker 3>own data at play to make sure that they can

0:06:34.160 --> 0:06:35.000
<v Speaker 3>learn from it too.

0:06:35.080 --> 0:06:37.560
<v Speaker 1>Again, all boats your rights. We're in healthcare to do good.

0:06:38.200 --> 0:06:40.599
<v Speaker 2>You're also selling off bits of healthcare because you're focusing

0:06:40.680 --> 0:06:42.479
<v Speaker 2>in on the R and D, the drug side, and

0:06:42.520 --> 0:06:45.640
<v Speaker 2>maybe the consumer business goes for a hefty price tag.

0:06:45.720 --> 0:06:48.520
<v Speaker 2>You've been thinking about IPO and you've been thinking about sale.

0:06:49.000 --> 0:06:50.320
<v Speaker 2>How is that process going poor?

0:06:51.200 --> 0:06:53.320
<v Speaker 3>Well, we announced that October that we were going to

0:06:53.320 --> 0:06:55.120
<v Speaker 3>become a pure play by a Farmer. You know, we

0:06:55.160 --> 0:06:57.320
<v Speaker 3>want to double down on the things we just talked about,

0:06:57.320 --> 0:07:02.719
<v Speaker 3>and we have an incredibly successful and energize consumer business.

0:07:03.400 --> 0:07:05.800
<v Speaker 3>We feel that they feel in fact that they can

0:07:05.839 --> 0:07:08.040
<v Speaker 3>stand better on their own two feet and be independent,

0:07:08.120 --> 0:07:10.720
<v Speaker 3>raise their own capital, take on their own projects, and

0:07:10.960 --> 0:07:12.920
<v Speaker 3>we want to support them in that it's the right

0:07:12.960 --> 0:07:16.440
<v Speaker 3>thing to do. Our capital should be allocated towards discovering

0:07:16.440 --> 0:07:19.239
<v Speaker 3>new medicines, and there should be about growing their business

0:07:19.240 --> 0:07:21.400
<v Speaker 3>fast in the market, which by the way, they're doing

0:07:21.480 --> 0:07:24.760
<v Speaker 3>very successfully. We've declared that buy as early as Q

0:07:24.840 --> 0:07:28.440
<v Speaker 3>four twenty four will be ready to go to separate

0:07:28.480 --> 0:07:31.200
<v Speaker 3>that business. We haven't shared how we will do it,

0:07:31.240 --> 0:07:32.680
<v Speaker 3>but I think it's quite.

0:07:32.480 --> 0:07:33.240
<v Speaker 1>Well known that.

0:07:33.880 --> 0:07:37.880
<v Speaker 3>You know, we're talking to interesting parties, certainly Pe, and

0:07:37.920 --> 0:07:41.960
<v Speaker 3>we have the capital markets vehicles available to us too,

0:07:42.080 --> 0:07:45.520
<v Speaker 3>So no, we're excited for them. There's a huge amount

0:07:45.560 --> 0:07:48.240
<v Speaker 3>of interest. It's a great business and we hope to

0:07:48.240 --> 0:07:51.520
<v Speaker 3>create value for patients, for stakeholders, and indeed for the

0:07:51.560 --> 0:07:52.440
<v Speaker 3>people in the business.

0:07:52.720 --> 0:07:57.760
<v Speaker 2>Q four twenty four is pretty notable given that we're

0:07:57.800 --> 0:08:01.200
<v Speaker 2>in quite a politically uncertain time. Look, we two brips

0:08:01.200 --> 0:08:03.000
<v Speaker 2>sit here looking that there's going to be an election

0:08:03.320 --> 0:08:05.920
<v Speaker 2>and people voting in the UK tomorrow and then of

0:08:05.960 --> 0:08:08.720
<v Speaker 2>course this weekend is all important for France. How as

0:08:08.720 --> 0:08:11.040
<v Speaker 2>a leader of such a French giant, are you feeling

0:08:11.080 --> 0:08:13.920
<v Speaker 2>about the political risk of doing a deal in this market?

0:08:15.560 --> 0:08:18.320
<v Speaker 3>Well, look, you know, in terms of healthcare and being

0:08:18.320 --> 0:08:21.160
<v Speaker 3>a leader of an incredibly important French company, you know,

0:08:21.200 --> 0:08:24.480
<v Speaker 3>our first priority is to ensure that whoever governs, they

0:08:24.520 --> 0:08:28.040
<v Speaker 3>provide a predictable, stable environment. Because research is a ten

0:08:28.080 --> 0:08:31.520
<v Speaker 3>to fifteen year process, so you know, we need to

0:08:31.560 --> 0:08:34.040
<v Speaker 3>think long term. We hope that whoever leads the country

0:08:34.280 --> 0:08:36.560
<v Speaker 3>will think long term. We know the current government has

0:08:36.600 --> 0:08:40.160
<v Speaker 3>been incredibly supportive of innovation and creating that ecosystem. We

0:08:40.240 --> 0:08:42.120
<v Speaker 3>hope that that would continue.

0:08:42.320 --> 0:08:42.880
<v Speaker 1>Of course.

0:08:43.160 --> 0:08:45.600
<v Speaker 3>You know, there are people trying to understand what it

0:08:45.679 --> 0:08:48.880
<v Speaker 3>means for the economy, much like for the change in

0:08:48.960 --> 0:08:51.880
<v Speaker 3>government if it's to come about in the UK, and

0:08:52.120 --> 0:08:55.040
<v Speaker 3>people want that predictability in the markets too. You know,

0:08:55.200 --> 0:08:57.880
<v Speaker 3>we're long term player, trying to do the right things

0:08:57.880 --> 0:09:00.680
<v Speaker 3>for patients, trying to drug the undruggable, and we hope

0:09:00.760 --> 0:09:03.880
<v Speaker 3>to partner with whoever's in power, of course, because we

0:09:04.040 --> 0:09:06.240
<v Speaker 3>have to be ready to do our part for society

0:09:06.280 --> 0:09:10.120
<v Speaker 3>and and you know, so far we believe that can continue.

0:09:10.440 --> 0:09:13.600
<v Speaker 2>Paul, it's been great talking to you the intersection of

0:09:13.679 --> 0:09:17.400
<v Speaker 2>drug discovery of artificial intelligence slackable AI as you call it.

0:09:17.440 --> 0:09:20.119
<v Speaker 2>Paul Hudson, great to have some time. CEO of Sanofie

0:09:20.400 --> 0:09:20.840
<v Speaker 2>say well,