0:00

Ha! Ha

0:02

ha ha. So fed.

0:06

Up and then harm. Oh

0:11

My. Good God. Hundred And One.

0:13

Ha. Ha

0:17

which allow them

0:19

to. Linger.

0:22

Radio Lab Or Adventures on the

0:24

Edge of what we think we

0:26

know. Hello

0:32

and welcome to Raising How are we

0:34

explore? the real challenges and enormous opportunity

0:36

is facing entrepreneurs who are building the

0:38

future of health. And.

0:45

Olivia and I'm Chris. Today's episode is

0:47

with Philip to guard chief scientific officer

0:49

of In see Trump. He is joined

0:51

by B J. Pandey found in general

0:53

partner at a sixteen Z by On

0:55

Health. Together they talk about the power

0:57

of medications to change lives and alter

1:00

history. Can it said this

1:02

was the invention of and Will

1:04

Be Six Nine Three and this

1:06

was the drug that one The

1:09

Wall And they did some reading

1:11

about that. And it turns out

1:13

that in fact, in December of

1:15

Nineteen Forty three, when Britain was

1:18

essentially homeless on it's knees, Winston

1:20

Churchill was travelling back from the

1:22

Middle East from a conference with

1:25

President Roosevelt and contracted pneumonia and

1:27

Will Be Six Nine Three was

1:29

airlifted to. Him by the Royal

1:32

Airforce from the factory in Dagenham

1:34

and it saved his life and

1:36

the lives of countless Allied soldiers

1:39

since. And I thought, oh my

1:41

god, a drug can genuinely make

1:44

a difference. And of course they

1:46

talk about how Ai is permanently altering methods

1:48

of drug discovery and development. What we're

1:50

in this transition from his from

1:52

finding to design me that we

1:55

have amassed enough knowledge. Cross these

1:57

be very small Dalits his various

1:59

molecules. I can start to

2:01

think you know much more about.

2:03

Okay here's the biology we need

2:05

to impact his the therapeutics how

2:07

it needs to act is to

2:09

design. That kind of gets you

2:11

there. You're listening to Raising Health

2:14

from a sixteen. The Violin Health. It's

2:19

my distinct pleasure to have today are

2:22

so to Cari as see a So

2:24

Of and see Trump as our guest

2:26

on raising health I agree to have

2:28

you on the sofa and wonderful to

2:30

be here. I'm very excited to talk.

2:33

Or. We are to himself you

2:35

spent decades in and bio and

2:37

Pharma. How does it all starts?

2:40

I knew that this is what I wanted

2:42

to do very early on in my career.

2:45

And that came through way I

2:47

kind of Brady. Interesting. Pivotal.

2:49

Moment for me. I'd.

2:51

Finished high school and was

2:54

set to study. Organic

2:57

Chemistry at Cambridge. And

2:59

I did something that you know

3:01

was comparatively common in the Uk

3:03

at that time. Wishes

3:05

to take a gap year and I was

3:07

very fortunate to get a job and on

3:10

my very first day when I came into.

3:12

To. The Bar Train there was a boss

3:14

case that had a laboratory notebook in it.

3:17

And it was out from that the

3:19

page. So the discovery and first description

3:22

of a compound called South A parody.

3:24

And there was a little plot connects to the Notebook

3:26

and it said. And since the invention

3:29

of a M B, Six nine, three. And

3:31

this was the drug that one. the wall. And.

3:34

I did some reading about that and

3:36

it turns out that in fact in.

3:39

And. December have a nineteen

3:41

forty three. When. Britain was

3:43

essentially homeless on it's knees. And.

3:46

Winston Churchill was travelling back from the

3:48

Middle East for a com from the

3:50

conference with. President. Roosevelt.

3:53

And contracted pneumonia. And.

3:56

M B Six nine Three was airlifted

3:59

to him by the Royal Airforce from

4:01

the factory in Dagenham. And.

4:03

It saved his life. And. The

4:05

lives of countless Allied soldiers

4:08

since. And I thought

4:10

oh my god, and drugs can

4:12

genuinely make a difference. either way,

4:14

sort of. Medicines is things that

4:16

made the difference to. In. A

4:18

few people here in there. And

4:20

I think that was the point where

4:22

I thought wow you know this is

4:24

of this is an important enterprise Discovering

4:27

medicines the can make a real difference.

4:29

To. Hundreds or thousands of people and perhaps

4:31

you know, change the course of history

4:33

and I think from that they on.

4:35

I've never wanted to do anything else.

4:38

And still don't want to do anything else.

4:40

which is I think one of the reasons

4:42

that I am so happy. To

4:45

be at in throws add today where we're

4:47

really trying to trying to do this in

4:49

a really different way. Or asked? My

4:51

mind going through the are cube your career

4:53

a house year, your interest changed and evolved.

4:55

You know you've seen a fair bit of

4:57

changes of the industry in a when I

4:59

started in the industry. In Out

5:01

much in in Nineteen Eighty Seven.

5:05

We were incredibly focused on chemistry

5:07

to the extent that the you

5:09

know people in the vaccine decision

5:12

division. Apu. Incredibly can

5:14

play prodding scientists. Were.

5:16

Kind of regarded as the second

5:18

class citizens of my can and

5:20

chemistry was king right? But.

5:23

Very quickly it became apparent

5:25

that in fact there was

5:27

more drug discovery been chemistry.

5:30

And. I think the

5:32

realization. That. Biologics. We're

5:34

going to be an incredibly important

5:36

part of of drug discovery and

5:38

new medicines that would make a

5:41

difference. Was I think the

5:43

first real change in the thing thing that

5:45

I've been interested in? I was

5:47

very fortunate some when I was working

5:49

at Oxford University. So. Bayonet Collaboration

5:52

with says are Milstein home where

5:54

we actually made the first. By

5:57

specific monoclonal antibody. They

6:00

you know fusion of living on the

6:02

so hybrid hypodermic. And. When

6:04

I was up Merc I can't think well,

6:06

why couldn't we make antibody drugs and slothful

6:09

molecules? And I was kind

6:11

of severely beaten around the head

6:13

and certain other folks antibodies will

6:15

never be drugs and I was

6:17

ads where it makes it antibodies

6:20

when every drugs that was that

6:22

was in nineteen nineteen ninety six

6:24

and that was actually what what

6:26

triggered my his desire to leave

6:28

Mark which was at their high

6:30

of it's you know Prowess is

6:32

drug discovery company and move to

6:35

Amgen where you know biologics who

6:37

really being invented in industrialized. The.

6:39

I sometimes loft myself. You.

6:41

Know last year when I saw the

6:43

you know the First I'll Five Percent

6:45

trends hygienist approval. In. Our drug

6:48

that was. you know, maybe thirty years

6:50

like an incoming in a when people

6:52

were talking about antibodies hadn't never being

6:54

drugs. They it wasn't like they are

6:56

completely wrong. That. They're they're real

6:59

reasons for why they they made ah have

7:01

concerns. Actually reminds me a lot of ah

7:03

when I started in come to some drug

7:05

designed like our twenty five years ago people

7:08

had some real concerns. were the on times

7:10

change their to. Ah, I'm curious

7:12

what were the Winds of Change says

7:14

you're seeing that led you to go

7:16

now? Not from Merc to Amgen, but

7:19

from Amd Jensen see drop. In.

7:21

Twenty sixteen. You. Know in

7:23

my former role at at Amgen. We'd.

7:25

Come to the end of what I would.

7:28

Would. Can to call the long

7:30

March of translational. In. Some Attics

7:32

right that we've started in Ninety Eight

7:34

When I joined the company. To.

7:37

With the goal of capturing all of

7:39

the day that. That. Was generated

7:42

hub by research and an

7:44

insult, everything about every single

7:46

molecule clone cell line. Am.

7:49

In a way that had reached matter data.

7:51

And. We've done that in only with

7:53

like while we did that. And.

7:56

We were looking around saying have what is

7:58

the next thing. And. We.

8:00

Went through an exercise Where we

8:03

were we interacted with companies that

8:05

were in the tax base. With.

8:08

Long product cycle so

8:10

companies like Slumber got

8:12

an all exploration Disney

8:15

residuals chip companies. And

8:17

ask them You know what it, what

8:19

do you see as the next horizon?

8:22

And they uniformly said machine learning. Which.

8:24

Was really he know in it's

8:26

infancy and drug discovery? And.

8:29

You know, we began to do a lot

8:31

of investigation that we began to invest in

8:33

machine learning. And it became very

8:35

apparent in the twenty sixteen, twenty seven time

8:38

period that this was gonna be really quite

8:40

transformed him. In the same way

8:42

as we embarked upon not long you know

8:44

holds her to to capture data. We

8:47

embarked upon the journey to really

8:50

attempt to use machine learning and

8:52

you know it's it's more sophisticated

8:54

forms to really make an impact

8:56

on drug discovery. The. Challenges:

8:59

You know that that is

9:01

hard to do in a

9:03

big organization and people skeptical

9:05

I'm in. How about technology

9:07

that is that's less explainable.

9:10

The monoclonal antibody technology. And

9:12

is less accessible than traditional.

9:15

In. A comfort a snow at chemistry

9:17

and physics based methods and so.

9:20

Am A really became quite challenging

9:22

to make real traction and so.

9:25

I. Began to think then about. you

9:27

know if you can't build it and

9:29

can you go to wear it you

9:31

know legitimately exist and. That. Was

9:34

the i think you know spark that

9:36

led me ultimately to work in secret.

9:39

History. Of or the disruption theory

9:41

would suggest are the most transformed of

9:43

technologies are very difficult to build I'm

9:45

in our into existing organizations what do

9:48

you think makes a i are machine

9:50

learning kind of in that category like

9:52

what is it that's different about it

9:55

as you think that makes have that

9:57

much harder for for and comments to

9:59

rebuild that are building and house. One.

10:02

Of the main challenges and this is something

10:04

in a we think about it in citro

10:06

the time. Is

10:08

explain ability. The. Last

10:10

thirty or forty years of

10:13

drug discovery. He

10:15

now has been this this

10:17

long transformation. From.

10:19

Really quite artisan or and

10:22

serendipitous drug discovery. To.

10:24

This kind of very rigorous

10:26

rational m Me: no rational

10:29

stepwise process. In as

10:31

you think, even fifty years ago, You

10:33

know you would expect things from plants

10:36

and bacteria and soil samples and test

10:38

them empirically. Yeah, that was health or

10:40

anti biotics. And the spy. And.

10:44

And in numerous other in are

10:46

incredibly important Medicines were discovered, And

10:49

so he went through this era of

10:51

kind of hyper rationality where everything has

10:53

to be completely explain the ball. And.

10:55

In advance of the the last and

10:57

fan of decimal place. And

11:00

then you know machine learning is

11:02

not like that. I'm in a

11:04

machine Learning is Seeing Patton's The

11:06

Human Beings Prompt see. And.

11:08

It's it's saying structures in

11:10

day that the though we

11:12

haven't you know, interpreted cognitively.

11:15

And saw it becomes really in

11:17

a very challenging to deploy it

11:19

in this site for rational environment.

11:22

And. They also frankly. Requires.

11:25

You know skill set that

11:28

is less transferable Man Many.

11:31

To. Look up most of the informatics

11:33

departments in Fama. You know,

11:35

they're often populated by people who

11:37

aren't computer scientists by training. There

11:40

are a populated by people who

11:42

are industry professionals who to learn.

11:45

In our significant amount about information technology,

11:47

machine learning is not like that. You

11:49

know that that you know high and

11:51

machine learning. You need people who are.

11:54

Highly trained and think very abstract

11:56

li and often than not a

11:58

good fit in. In the organization.

12:01

Nine. He very well said none. I

12:03

think it's easily as the people challenges

12:06

sometimes have become more significant than than

12:08

than the technology. Some I was have

12:10

been like I'm moving to a startup.

12:12

I was very fortunate in my years

12:14

of them tend to be. Very.

12:17

Closely involved in our corporate

12:19

venture fund and. Spend.

12:21

A lot of time. Working.

12:23

With start ups I would say the

12:25

down scaling aspect was kind of expected.

12:28

Now crazy I wake up and say

12:30

oh yeah you know I need for

12:32

somebody to do this season and then

12:35

it's like osha I am that a

12:37

person sitting at the end ah as

12:39

I got argue that things what I

12:42

would say that I was I would

12:44

say soft but was really well. You.

12:46

Know this is a thing and is the

12:49

place. You. Know we try to operate

12:51

at the pace of a tech company, not

12:53

a biotech company, or the speed by which.

12:56

Products. Get Get Produced

12:58

Deployed. And. The speed at

13:00

which we have to make decisions that this

13:02

the rights of which were trying to do

13:04

biology. And. That is is really

13:06

from an extraordinary and and very different

13:09

shaded from. You. Know the big

13:11

companies like my call am and then

13:13

I would say the other thing is

13:15

just the astonishing quality of the Talon

13:17

that that small companies can attract. Maybe

13:20

we should step back and like talk

13:22

a little bit about drug design. Maybe

13:24

before and after adding his Hershey. Some

13:26

of our listeners will be coming from

13:28

the Ai side and may be unfamiliar

13:31

with a drug design. Sides: A: Oh,

13:33

could you tell little bit about what

13:35

might. For. A typical arm

13:37

therapeutic. What traditional drug design part

13:40

a pathway would look like versus

13:42

what are the sort of? There's

13:44

more modern machine learning. Our version

13:46

looks like how things change. Sites

13:49

I was kind of characterize

13:51

this change. As

13:53

as moving from finding

13:56

to designing. So.

13:58

You know up until pro the. Even.

14:01

Ten Fifteen years ago. And.

14:04

Maybe even more recently. You.

14:06

Found most drugs.

14:08

There's. A reason that people called drug hunters

14:10

right keener. The drug is out there in

14:13

the wild and you have to kind of

14:15

hatchet. And you know that was true because

14:17

you know at the beginning of the industry

14:19

most drugs actually we're pretty this thing. they

14:21

were natural products. see that. From.

14:24

In our other organisms or

14:26

proteins, all hormones, ah, metabolites,

14:29

That. You know, circulated within our own

14:31

systems. Okay, and so yeah, it's kind

14:33

of found those things. The. Whole

14:35

class of jail p One agonist

14:37

arose from peptides from the saliva

14:40

of the healer monster of right.

14:43

And. Then the industry

14:45

tried to pivot to been more

14:47

rational. But. It still always came

14:49

back from this starting point of.

14:52

You. Know finding same thing right. You

14:54

found something from a high throughput

14:56

strain. Or you found something from

14:58

virtual. Virtual physics based

15:00

screen Or you found an antibody

15:03

through immunization of of an animal

15:05

and looking out for millions of

15:07

years of evolution to give you

15:09

an answer. We found a T.

15:11

So what? We're in this transition

15:13

from his from finding to designing

15:15

of kites that we've amassed enough

15:17

knowledge. Fossey's. For ferris

15:20

modalities, various molecules. I.

15:22

Can start to think you know

15:24

much more about okay, here's the

15:27

biology that we need to impact

15:29

his the therapeutic how it needs

15:31

to. Is the design that

15:33

kind of gets you there? I.

15:36

Have a good friend who was in the

15:38

space program and was constantly shocked by the

15:40

fact that you know we would put drugs

15:43

and so the clinic and I would file

15:45

and it's like I would get fired if

15:47

every time you know the rocket booster went

15:50

off in a one out or five times

15:52

it exploded and that say that's that's a

15:54

that's an eighty percent success rate as opposed

15:56

to the twenty percent success rate. So I

15:59

think the bill. The t to move

16:01

away from You know this just empirical

16:03

thing and and face is now the

16:05

thing that we need design. And here's

16:07

the people were gonna design. It that.

16:09

Are that I think is the real transition? And.

16:12

That is occurring in the industry. I.

16:14

Love that analogy. In or even in going to

16:16

the moon, it was Apollo, one that went to

16:18

the moon was appalled. lesson. And.

16:21

Even Apollo one was after Mercury and Gemini

16:23

and have done all this stuff. but his

16:25

Apollo one to eleven is gradually building up

16:27

getting there. There's this our environment or or

16:29

it's of doing some maneuvers, going to the

16:32

moon but not landing and and finally Nepal

16:34

I've and you deserve incrementally engineer all his

16:36

stuff so city build and learn are around

16:38

just by trying something and see if it

16:40

goes to the moon, francs in veils and

16:43

see what goes in moon landings. I target

16:45

housing him as like they're really maybe are

16:47

really three phases of drugs design that you

16:49

could call or call. Out one is a

16:51

very old days which are very seen a

16:53

typical. You. Know natural products may

16:55

be put into to see you are

16:58

not rationalized at all. No molecular biology

17:00

by any means. It was very macro

17:02

and then there was at rationalization where

17:05

we lose some phenotypic we want understand

17:07

the pathways and the proteins and the

17:09

protein structure and how it's boundaries Braun

17:12

and all that rationalization and I think

17:14

we got frustrated that's that all the

17:16

information just proved how hard it how

17:19

on difficult biology is but now we're

17:21

in some sense was seen it's a

17:23

big screens and ml. Were sort

17:25

of maybe getting to the best the

17:28

of of both worlds where it's subs

17:30

the details rich information as but yet

17:32

we start to sort of have this

17:35

more systems holistic view of what these

17:37

things are doing. intellect item and for

17:39

small is a wonderful time when we

17:42

could clone and express and clarify proteins

17:44

and to biochemical essays. But.

17:46

It's naive to think that

17:48

proteins exists by themselves. Somehow

17:50

they constantly exist in in

17:53

complex. Is that change in

17:55

a modified. And in as

17:57

a whole new biology that sub been appearing

17:59

there. The condensation. Condensates,

18:02

Inside cells triggered by you

18:04

know, our oligonucleotides. And

18:06

so in our being able to

18:09

capture that highly rational peace through a

18:11

thread of genetics which is the

18:13

kind of digital part of the sell

18:15

you know if is, it really

18:17

does give you that ground. Truth. But.

18:20

Understand how you know molecules, interact and

18:22

create the biology that we see in

18:24

we need to impact. I think that

18:27

is absolutely correct to this idea that

18:29

we can do systems biology. Discovery.

18:32

I'm in. I think he is

18:34

going be really cracked. Actions formative,

18:37

particularly for the more complex diseases,

18:39

right? L R in your blood

18:41

pressure by violating your arteries is

18:44

a pretty straightforward thing. Accurate: Parkinson's

18:46

Disease Pretty not straightforward thing. Yes,

18:49

Yes, I don't know if it's possible

18:51

few to talk about how machine learning

18:53

is being done in seats are to

18:55

change the process on if you could

18:58

give any specific examples. You.

19:00

Know we've able to. Create.

19:02

A. In our a

19:04

semi synthetic patient cohort. And

19:07

through imputation, A. Across

19:10

multiple mode out is so

19:12

M R I. Am.

19:14

Metabolites Biomarker Dexter.

19:17

Which. Pow enough to

19:19

do and genetic discovery

19:21

for fatty liver disease.

19:24

In. A way and that is

19:26

not compounded by obesity. So much

19:28

fatty liver disease is driven by

19:31

obesity. I. Am is very hard

19:33

to to d convoluted you need in a

19:35

pretty fancy functional M R I To do

19:37

so and then through using about Clinical machine

19:40

learning were able to create a dataset that

19:42

was large enough and powered enough to do

19:44

genetics. And. That you

19:46

know gave us line of sight

19:49

to a completely novel mechanism. Or.

19:51

Completely unexplored mechanism in terms

19:53

of the. Genesis. Of

19:55

fatty liver disease. We

19:58

were able then to you know, look at

20:00

the genes that A were involved in those

20:02

pathways. Am. using our

20:05

pulled optical straining technology which

20:07

you know fundamentally. Cannot.

20:09

Be performed without advanced

20:12

machine vision. And.

20:14

Ah, I'll mix machine learning. I'm

20:17

able to recapitulate those phenotypes in

20:19

had a sites in In This

20:21

is. And you know

20:24

we're then convinced enough to trigger.

20:26

Four. Or five programs you saying

20:28

a siren A Mode our tis

20:31

to get proof of principle in

20:33

animals. And you know. That.

20:35

Was all done in a twelve

20:37

month period which kind of shows.

20:40

You. Know that machine learning is not only

20:42

an enabler of new things so we

20:44

know we were able to discover genes

20:47

because we you know will power to

20:49

do says from the sea mining but

20:51

it's also a massive accelerator okay that

20:53

that some think the pants. using.

20:56

Traditional I yeah I target discovery

20:58

techniques might have taken three or

21:00

four years. We. Can do in

21:02

no ten or twelve months. For.

21:04

All of us in seat throw. It

21:06

was like wow you know this really

21:08

works and it becomes more the question

21:11

then of how do you reload, how

21:13

you scale. I'm how do

21:15

you then start to make decisions when

21:17

things are happening so rapidly? Or.

21:20

And assertions a beautiful example but

21:22

he almost like the the density

21:24

of of concepts and in your

21:26

sentences for very i I am

21:28

I agree could sort of break

21:30

it down but it's basically was

21:32

not just one a little bit

21:34

machine learning hear their ads was

21:37

to some degree mission learning throughout

21:39

in all the different aspects that

21:41

deacon loot the biology it's to

21:43

understand the mechanism and eventually to

21:45

even sort of gets you towards

21:47

therapeutics or and in this case

21:49

may. Be as were straightforward are given

21:51

the approach but it's feels like it's

21:54

not just something that you could sprinkle

21:56

on top but that something that pervades

21:58

the were using about everything. The know

22:00

I would say that in a coming

22:02

back to the challenges of building it

22:04

in our machine learning in Fama. And

22:07

is that the systems are not

22:09

natives, machine learning and. Once.

22:11

You start building from the ground

22:13

up. With. Machine Learning in mind

22:16

you I see you know your architecture

22:18

is very different. Your infrastructure is very

22:20

different. The. Way you approach things

22:22

have very different and so. Things.

22:25

Do go a lot faster

22:27

and. You. Start to then

22:29

get in this mindset of. Well.

22:31

Can we improve this through Missing Landing?

22:33

To me, improve that through machine learning.

22:36

And. You get in a virtuous cycle

22:38

that you sought to reuse your you

22:40

more your models. He start to reuse

22:43

your infrastructure and you generate the fi

22:45

will effect which you know is is

22:47

ending. Incredibly exciting place to be. Where

22:50

do you think I'm the first big

22:52

wins will come from? There's been a

22:55

lot of talk about a i enjoy

22:57

design but usually people ask you know

22:59

where the drugs But of course this

23:01

whole cycle through clinical trials and saw

23:04

it takes many years I'm I'm curious

23:06

if he were to previews a future

23:08

where you think this approach will have

23:10

the biggest immediate impact. Patient Stratification nice

23:13

one of the kind of. Dirty.

23:15

Secrets of the therapeutics

23:17

business that. Most patients

23:19

get the wrong dose of the wrong

23:22

drug, the wrong time, right? and if

23:24

you know it's and or somewhere to

23:26

miracle that that they get any benefit

23:28

at all and so have. I

23:31

think you know just

23:33

the ability to very

23:35

rapidly and inexpensively. He. Now

23:37

and accurately stratify patients who

23:40

are gonna respond better. To.

23:42

Them medicines to have the appropriate those

23:44

and that those regime. That is a

23:47

big win for the industry. It's a

23:49

big wins a society and health systems

23:51

in a big win for patients and

23:54

so as an enormous amount of interest

23:56

in deploying machine learning tools. And

23:58

I think that is a place where. They'll be

24:00

in almost immediate benefit to

24:02

society and supplies I think

24:04

in see through wishes to

24:07

participate. I would also say eighty. You.

24:09

Know having spent. In. I was

24:11

so many years. Bringing in

24:13

a fantastic molecules into the clinic.

24:15

The didn't do anything because we

24:18

didn't understand the biology. That.

24:20

The real durable impact will be

24:23

in target discovery. so to really

24:25

understand. His. The two or

24:27

three or four. Proteins. Pathways

24:29

Processes that really need to

24:31

be modified. To. Have an impact

24:34

in the disease. And that hopefully

24:36

you know we can't That success rate

24:38

from the twenty percent you know, so

24:40

more like the eighty percent and that

24:43

is a huge win for for everybody's

24:45

huge wins the industry huge win for

24:47

medical systems and patience. or yeah I

24:49

think in that first example of stress,

24:52

occasion or kind of. Unravels

24:54

or decouple some of the complexities of

24:56

biology to create something where you would

24:59

expect greater success because ss cleaner and

25:01

and it's not a mixture of different

25:03

effects that can be confounding, it's it's

25:06

something that's understandable by construction and then

25:08

the second one is as another version

25:10

of added sounds like that again it

25:12

sort of making taking taming the complexity

25:15

of biology it feels like to some

25:17

degree. Absolutely. what do

25:19

you think they're next? Five years, next

25:21

ten years of sight in Farmer. I

25:24

public and given to on says

25:26

okay. I. Would say

25:28

the first sunset. You know,

25:30

like many industries where there

25:33

is fundamental overcapacity, There's.

25:35

Gonna be consolidation. It's kind

25:37

of inevitable. You. Com.

25:39

pads. You. Know thirty

25:42

five companies working on T

25:44

Ross inhibitors by it, so

25:46

I think they'll be consolidation

25:48

and I think. There

25:50

will be a. ruthless

25:54

drives for cincy.

25:57

And. That. Drive for

25:59

efficiency. Well. Have to

26:01

include advance machine learning. Okay,

26:04

Am. Here, there's enormous amount of

26:06

waste in. In. Medical research

26:09

that waste in our raises

26:11

the costs of medicines it

26:13

reduces accessibility to has. Enormous

26:16

environmental impact. And.

26:18

I think you know you'll see that

26:21

the survivors will be. Extremely.

26:24

Efficient and places.

26:26

That. That use machine learning.

26:29

And automation. That. Will

26:31

generate. A. Kind of different

26:33

medical ecosystem were personalized. Medicine

26:35

is achievable. Riots United Sixty

26:37

in a we talk about

26:40

personalized medicine. And then look

26:42

at the cost of it and it kind

26:44

of goes away and. You know

26:46

by being efficient by being highly targeted

26:48

by using all the tools of computer

26:50

and automation. Perhaps. Personalized medicine,

26:52

you know in a really impressive

26:54

efficient Consolidated White is double size

26:57

I met. That's kind of my

26:59

somewhat erratic cool view of where

27:01

things school go. Yeah,

27:03

I mean to some degree. If you can

27:05

stratified to a small cohort and made of

27:08

the F B a record of wanted there

27:10

may be. A ten

27:12

thousand people like me in the world? Ah,

27:14

but like, maybe not a billion people like

27:16

me, right? I'm certainly not. seven. So.

27:19

I could become personalized pretty quickly

27:21

even just are some of the

27:23

extending some a message you're already

27:25

have described. but what's interesting about

27:28

what you're describing now you're describing

27:30

not just a life sciences company

27:32

that is a developing drugs are

27:34

the largest or licenses comey that

27:36

would have identified targets or hit

27:38

those targets or bring those kargus

27:40

suit trials through all of these

27:43

matters including pieces as occasion but

27:45

this comes into play afterwards to.

27:47

You're. Not talking about your delivery to some degree as well.

27:50

Go. Absolutely. I think the industry. Has.

27:53

kind of flirted with care delivery he

27:55

know for number of years. And

27:57

most recently with the managed funds.

28:00

Benefit he know consolidation in

28:02

the consolidation. The. Idea

28:04

that you are delivering. You. Know

28:06

more than just a medicine that you're

28:08

delivering. In. Our health solution.

28:11

I think is is the place where. Most.

28:14

Bailey was gonna come and

28:16

rightly should come right medicines

28:18

bring enormous value. right?

28:21

Almost disproportionate body, and so.

28:23

Are. Making sure that body gets

28:25

realized in our not diluted strout

28:28

whole. Chain. Of bureaucracy

28:30

and distribution and access.

28:32

Am. I think is is where many companies

28:35

are starting to think. Yeah I

28:37

vassar so says amazing world because when

28:39

we can finally stratify says it weekend

28:41

know who this drug works for. We

28:43

can take a lot of the trial

28:45

and error on that tour delivery side

28:47

out and it becomes engineering there as

28:49

well. So I'd love

28:51

to just and with a few questions

28:53

are bigger picture things building it sort

28:55

of a company with attack mindset. In.

28:58

And life sciences is pretty new. right?

29:01

I mean I think there's some examples

29:03

but us so she this interface where

29:05

you the way than seizures doing it

29:08

is a relatively recent phenomenon. Any advice

29:10

for people looking to build companies in

29:12

the space or for especially how you

29:14

manage these teams up presumably from the

29:17

outside and seem like he could have

29:19

two cultures that you have to blend.

29:21

You have to blend that tech a

29:23

i'm Outside and our life sciences side

29:26

how's that work That is so true

29:28

and it it means you have to

29:30

be bilingual. In. I'm very fortunate that

29:33

in I spent some time in Quebec

29:35

and I ran my lab in France

29:37

and A and you understand that you

29:39

have to be able to speak both

29:41

languages in order to be successful and

29:43

I think. And that's the number

29:45

one thing I think anybody in the Spice

29:47

needs to realize is. You. Know you

29:49

can't just where you're missing learning. how

29:51

are you can't just where your biologist

29:54

hats. You. He really has to to

29:56

were both of those hats and. A

29:58

lot of it is is very simple

30:00

learning the language you know. the words

30:02

in the woods mean different things in

30:04

different disciplines and so as much time.

30:07

Of as people can spend to be

30:09

to be fluently bilingual that is an

30:11

enormous or enabler of of that kind

30:14

of tech by a world you know.

30:16

The other thing is the talents of

30:18

risk. In our and the willingness to be

30:20

wrong. Biologists, Because they

30:22

come up through this very academic. You.

30:25

Know somewhat competitive P

30:27

H D postdoc system.

30:30

Where. You know, being wrong in

30:32

front of your peers is a terrible

30:34

terrible thing, right? Whereas.

30:36

Writing some code that didn't work and

30:39

then debugging and it's it's perfectly natural

30:41

hooker and sit up and saw this.

30:43

willingness to be wrong and to ask

30:46

you know, ask questions and get feedback

30:48

is a very tense part of the

30:50

culture. And. And takes a

30:52

while for biologists I think said to

30:55

to really absorb and but it's a

30:57

in. it's incredibly important thing because that's

30:59

how you go quickly. Now that seems

31:01

like are a key secrets and during

31:03

the stamp for walks thank you so

31:05

much for joining us. Okay,

31:08

thank you for that's great! Thank.

31:17

You been listening to raising? The

31:20

posted and produced by Me Preset yes, Even

31:22

and Me Ality lead. With. The help

31:24

of the biological scheme. Ace of Kinsey. The

31:26

show is edited by Still Hegseth If you

31:28

want to suggest topics her future shows you

31:31

can reach us at Raising Health at a

31:33

sixteen the.com Finally, please rate and subscribe to

31:35

our Ship. The

31:38

content here is for informational purposes only

31:40

should not be taken as lethal business,

31:42

tax or investment advice or be used

31:44

to evaluate any investment, our security and

31:46

is not directed at any investors are

31:48

potential investors and any and sixteen the

31:50

phone. Please note that Easy Team, the

31:53

and it's affiliates maintain investments in the

31:55

company's discussed in this podcast for more

31:57

details including a link to our and

31:59

Essence. The an easy to do.