SPEAKER 1: Welcome back, folks, to our Stanford AI Health

Podcast.

We're super excited to be joined by Dr. Syed

Mohiuddin, who is a close friend, physician by training.

Formerly at McKinsey, running a lot of AI efforts

at Health and Human services, and is now

the Chief AI Transformation and Strategy

Officer at UnitedHealth Group.

Thank you so much for joining us.

SPEAKER 2: Pleasure to be here.

And great to see you and Matt.

SPEAKER 1: How many of those queries

that used to go out to your own personal consults

to friends no longer go out to those friends?

How much of a reduction based on using the AI systems

are you feeling now, just from a personal capacity?

SPEAKER 2: To me, it's like--

as an internal medicine doc, you are a quarterback to care,

so you're constantly in this dynamic with other people

to keep getting better and better information

and refine your hypothesis, right?

I'd say-- again, this is not practicing.

This is just in my personal life, matters

impacting my daughter, my family, my friends or myself.

I'd say overall volume is up 5 to 10x.

And I'd say, because a lot of the lower hanging stuff now I

can get answers to versus just they linger in the back

of your mind, overall volume, I'd say, is up 5 to 10x.

And I'd say my outreach to specialists is 50% less.

So you could say it's a 10 to 20x productivity gain.

SPEAKER 3: I mean, that's much better quantity.

I think on my side, it's probably in that ballpark.

I would say that I still--

if it's super important, I like to just show up

like I've done some homework, I guess, in that way.

And so I'll do some of the cognitive work on my side

that I wouldn't have done otherwise, because I'm in my--

would I've really gone down a rabbit hole in the literature?

Maybe if it's super important.

But now, OK, I'm going to have this thing do a shotgun pass,

and maybe I'll do some follow-ups,

and then I bring that to a colleague.

But you're totally right.

For the quick, hey, FYI, I've got a neighbor

that mentioned this and just--

I said this, just want to double check I'm right,

that kind of thing, not often anymore.

SPEAKER 2: Not often.

Exactly.

SPEAKER 3: [INAUDIBLE] is lower too, right?

Don't you feel like--

SPEAKER 2: Yeah.

You get less.

SPEAKER 3: They're using the model too, right?

SPEAKER 2: I used to have friends and family reach out

to me all the time for questions that they don't reach out

to me anymore, which is great.

But now, if you do get incoming, you take it like,

all right, I'm going to create space for this.

And this feels like borderline inappropriate,

but I'm going to name drop here at the risk of politics.

What still happens is--

I had a friend reach out to me with an urgent geographic--

like, a local.

It was in the mid-Atlantic slash Northeast.

Medical need, needing to get a specialist

for a certain kind of cancer.

And AI is not going to nail the answer on who's the exact right

doctor at Hopkins or MSC or whatever

for this very rare and unique thing.

There, I picked up the phone and called who we all

know, I know, and said, all right-- this is back

when she was running NCI and said,

all right, here's the situation, what do I got to do?

There's no substitute for that yet.

SPEAKER 3: Yeah.

SPEAKER 1: It's fascinating because this trend, again,

we all feel it.

Again, not in the professional sense but in a personal sense

for how we're managing care.

And I think the interesting thing

is we're starting to see more and more data come

out supporting this.

This was a Bain survey.

Looks like it was 500 people in March 2024 to September 2025.

And there's a few different pieces,

but I'll jump to the bottom, which--

SPEAKER 2: Justin, I love how you bring the ex-McKinsey guy in

and first slide, here's a Bain survey.

Thanks, man.

SPEAKER 1: Yeah.

SPEAKER 2: I love you back.

SPEAKER 1: I had to drill in.

I had to drill in.

But this last one is fascinating.

When there's a McKinsey one to show, I'll show that next.

AI becomes my doctor.

Percentage of consumers comfortable with AI.

And we don't know the exact question or exactly how they

did it, but 11% in 2024 to 2025, almost 30% now.

That's fascinating.

And as we talk about all the other impacts--

and we'll get to value-based care and health systems

and payers and all that soon--

this one trend-- again, we all feel, as we just talked about,

in our personal lives, patients are starting to choose this too.

What happens to this system?

What happens to the system?

And extrapolate one more year.

Maybe this number's 50%.

That is just such a fast pace of change to a healthcare system

that normally moves extremely slow.

And Syed, you've seen this also at the government side as well.

How are you putting these kind of trends

into how you run your job and what

you think is coming for health?

SPEAKER 2: Yeah.

I think it is-- look, jokes aside,

it's obviously a powerful number.

You're talking about a 2 and 1/2X increase in one year,

right?

One year.

And I think just to break it down a bit,

this is a combination of headwinds and tailwinds,

where you have headwinds, people are increasingly

frustrated with the care they get because there is a supply

issue.

There is a physician shortage.

There is a clinician shortage.

And the amount of time you have in your patient visits,

the amount of time-- the availability or access issues

in terms of time from when you want

to schedule an appointment to when you actually

schedule an appointment.

Also the personalization.

Back in the day, you knew your doctor.

Now, increasingly, people don't know their doctor.

So then the substitutability goes up, right?

Now, that's the headwind.

The tailwind is, OK, substitute goes up,

convenience issues, delays.

Why I start doing what Matt and Justin and Syed

are doing, what everyone else is doing-- and you guys,

I know the three of us know the internal numbers

that OpenAI and others have probably

shared with us around the number of millions upon millions

of users who are going to them and submitting clinical queries.

By the way, physicians and non-physicians

alike submitting clinical queries.

And that's true of all these AI applications.

So increasingly, the idea that the information can come

from a non-human has gone up.

Now, the important question is the question you asked, Justin.

What then does that mean and what does that look like?

So we've heard this existential, oh my God, doctors

are going to be gone, blah blah blah.

Calm down.

What this is saying is that massive volume of care that

isn't being done, can we introduce a layer

where care is actually done?

Can we manage care more effectively?

Can we do preventive care?

Can we actually do prevention?

Can we actually answer people's questions

when they have them, not six and 10 and 12 weeks later?

So all of this stuff that is not just the 15 minutes

of the actual patient visit, but the 364.5 days you spend outside

of the clinic, that's what we're able to get at by everyone--

imagine a world where everyone has their own LLM,

individualized LLM, their own--

a chatbot that knows them, that knows everything about them,

all the data, everything that's helping them with their care.

Guess what?

Increasingly, that becomes the quarterback of your care.

So I, Syed Muhammad Mohiuddin, licensed

internal medicine physician, trained

to be a quarterback of care.

Now I have an agent who is my sidekick who's

helping me do two things.

One, actually expand my panel size

because that's what the population needs.

And then, by the way, expand by managing and doing

the care management, doing a lot of the administrative stuff,

taking care of the lower risk questions around the health.

But then two, allows me to come in and do

the reason I fell in love with medicine-- take on harder cases

and figure out, just like I had to consult

the former head of the National Cancer Institute, who

in my friend network and who in my referral network, right?

In my appropriately tiered referral network.

Do I need to pull in to ensure that this person gets

the best possible care at the highest value?

So you're actually getting into a world with AI

where you can take care of people end

to end the whole time, not just a couple of days you see them,

and achieve value in a way that we've not really

seen in the roughly 15 years since we

launched CMMI and started really pushing for value based care.

It's a long answer, but those are my two cents.

Matt, what do you think?

SPEAKER 3: Well, no, I was just nodding my head.

I think your answer is incredibly well articulated.

It does bring up an interesting point

that there is this-- it's like the classic iceberg, right?

We're all thinking about the tip all the time about the work that

we're doing in our day-to-day lives and in the practice

of medicine today, knowing it's there, but not really--

feeling overwhelmed at just trying to address that

under surface mass of unmet needs, access,

the quality of the care.

And frankly, to your point, the time

between the visits, where we have been saying forever,

man, if we could just get patients

to jump in and just be more proactive about there--

I think that there's a willingness to do that,

but the tools I just don't think were there.

The information and asymmetries we've talked about on this show

has been a barrier.

Are we now in a place where that starts to--

again-- is there a true partnership that starts to form?

And to your point, the quarterbacking,

I think the other version of this that we see in the coding

community is like you're elevating

into another meta level of orchestration

of different capable--

whether they're agents or other services.

And now that's a higher cognitive plane, I think,

and maybe even a more satisfactory one

as a practitioner to say, I'm really

elevating, working at the top of my license,

and orchestrating things at, I guess,

a scale that I never could before, but also

makes me feel as though I'm truly partnering

with my patients and their journey

through the health system.

I remain optimistic, but everyone knows I'm a glass 3/4

full guy on a lot of this stuff.

And there will be detractors from some of this, I'm sure.

But on the road there, it feels like it's

a puss-pull, to your point, about the headwinds

and tailwinds.

SPEAKER 2: The amazing thing is, in a traditional world

where you don't have AI agents and you

don't have this level of orchestration,

you unfortunately end up doing what folks often

refer to as sick care, because that's what you can bill for.

In this future world, if you do have AI agents able to engage

individuals on so many things beyond just the quote unquote

"sick care," huh, I'm getting nutrition counseling.

OK, yes, I have diabetes, but I have, at my fingertips,

help navigating what I put in my body that

is responsive to my metabolic demands

and my individualized metabolic state.

I'm able to go far beyond in my activities,

beyond my doctor just saying, hey,

the American Heart Association says work out

x number of minutes y times per week based on your risk factors.

Now, OK, I like biking or I hate biking, I like swimming.

I can get individualized plans and I can get someone

to help me manage my life in a way

that-- man, life is hard for most people,

and it's hard to do all of this on your own.

To get a little bit help, not just

for people who can afford all the coaches and the trainers

and all that, but just in the form

of smart board-certified help at my fingertips,

that's pretty great.

SPEAKER 1: So I'm curious.

Syed, you painted this amazing future vision.

You said a few things.

Amazing future vision about what's possible.

Let's move from sick care, proactive care, health,

all these things.

Typically, when those things come up,

policy is the next question.

Well, the incentives aren't there.

We don't have the right system to pay

for that, which you brought up and, all these things.

My question is, do we think we need policy changes to start

seeing some of the benefits from these tools

that you talked about?

Can we leave the status quo the same from a policy perspective,

and are there enough right circles

where the incentives are lined up

where you think this adoption will happen naturally?

SPEAKER 2: That's a good question.

I will say this, you can always accelerate priorities

and manage risk through policy.

I think the position we took in the last administration

and the position this administration is continuing

to take, building on what we did,

is there's still so much we have to learn so that we don't

overregulate this space, right?

Because that is a risk.

We want to make sure that we're spurring innovation.

At the same time, we want to make sure

that we're continuing to learn, and we're actually

talking to each other.

What the administration-- where they've really

built on what we started doing with things

like our voluntary AI commitments

and whatnot is they've really tried to say you know what?

Let's align on a few priorities, create

a coalition of the willing.

And the CMS pledges, for instance,

and try to drive adoption in responsible ways

by bringing people together and having them talk to each other.

That can only be a start.

And we're not going to go where we need to go.

You're not going to have AI-driven primary care

without meaningful FDA involvement.

So the way it needs to happen, the FDA

doesn't know how to do this.

They still don't have mechanisms in place

for postdeployment surveillance and monitoring,

for instance, which is obviously essential for gen-AI-based

applications, especially, say, you

have something like a pacemaker or something like something

that's in your body.

So in that world, they know, just as we did,

that there are mechanisms not yet established to fully vet

how you're going to do this.

I think this notion of experimental sandboxes where

you have a data sharing and partnership between government

and these entities and you work together

on the right regulatory mechanisms

is the right way to go.

How exactly that plays out and what exactly the policies are,

there's lots of debates being had on that.

I left that game over a year ago,

so it's not my problem anymore.

SPEAKER 1: Well, separate-- one more separate

from the policy side.

SPEAKER 2: It's all of our problems.

But anyway, sorry.

Go ahead.

SPEAKER 1: No, but separate from the policy side.

Let's say there are more changes.

So it's more or less status quo.

Yes, there's lots of good things going on to learn, share data,

but nothing sweeping.

You can't wave your wand and say, hey, single

payer, anything like that.

Is there enough incentive today anyway to see

these things start to move?

And I'll give my--

SPEAKER 2: Yeah, I want to hear.

SPEAKER 1: I think the answer is yes.

As you look at previously, when you

talked about digital therapeutics or other areas,

you really needed a push and a payment mechanism

because there wasn't that natural pull.

There wasn't a natural pull.

And I'll share a little bit more data that we can look at.

Now, what's happening, there is a pull.

There is a pull towards these AI tools because people believe,

health systems believe that there's

a real value in ROI separate from any new payment mechanism,

carrots or sticks that may come down the road.

And so again, where I'm less certain

is how far does that take us.

Does that take us to the medical utopia you talked about,

or do we get somewhere separate?

And so just to talk through a little bit of the data we're

looking at here, there was a recent report by Menlo Ventures

that talks about the state of health AI,

both funding and where adoption is happening by buyers.

And there's a few interesting things

they called out, which I'm not sure all of us

agree with from what we're seeing on the day

to day, but a ton of scribe adoption, a ton of coding

and billing adoption, mostly by health systems,

and a lot more than what they found on the payer side of this.

And so one thing that's interesting versus before

is we're seeing technology adoption without any policy

changes happening naturally.

And so the question is, how far does that

take us, and then also, want to get this understanding of,

do we think these numbers are right?

Does this payer versus provider spend match what

we're seeing in the real world?

SPEAKER 2: All fair questions.

Two or three kind of quick reactions

to the page and just this idea.

What's clear is folks on the payer and provider side,

I'd say led by payers, but we're seeing

increasing activity, especially in the RCM

space among providers, are really

focused on administrative areas of opportunity.

So that could be SG&A opportunities

or OpEx opportunities.

You see some on operations there on the right,

but I think that's more universal than just that little

$50 million narrow sliver.

But then you also see revenue opportunities.

So think coding optimization as a good example.

So that, I think, in a world that

relies on administrative efficiency

or where administrative efficiency is a priority

and you're not directly impacting patients, i.e.,

you're kind of free from a regulatory standpoint

or more free from a regulatory standpoint, as you

get AI agents replacing the task of humans

and able to coordinate with one another

to execute more and more complex tasks,

I think we're, frankly, at the tip

the tip of the iceberg of where that administrative efficiency

space can go without any additional regulation.

I think on the clinical side--

and by the way, even your ambient number,

that ties to a lot of administrative applications,

as you know, in terms of how you can get revenue

and what you can enable, what really actually even beyond just

coding optimization you can enable

by having ambient scribes, including

care management, including a whole bunch of other stuff

that benefit from a higher fidelity, higher integrity

clinical documentation.

So that is the administrative side.

On the clinical side, that's where even

engagement is not necessarily directly a clinical play.

That's like, do you have coverage of benefits clear?

Are you showing up to your appointment, et cetera?

But the real clinical side, I think that's--

revenue models that will actually serve people

effectively that will have the right incentives

to do the right thing and then make systems better

at delivering care.

That, I think, is where it's going to--

at a minimum, you're going to need some guidance

from the federal government in terms

of how far you can lean in, because you start

touching humans and patients.

That's where, I think, the rubber

meets the road and regulation will be required.

What do you guys think?

SPEAKER 3: My comment on the slide--

so I agree with your points.

And I think directionally, this--

and this makes sense, I think, to look at it by spend

and buyer, but I think there's something else there too,

which is, as we increasingly hear from again,

those that are leading in this space in various ways,

the rapid reduction in costs or the quote unquote,

"asymptotically" heading to free intelligence kind of vision--

I do think that this is going to be a poor way

to track the outcomes or the impact of AI

in a variety of areas.

In particular, you brought up the administrative side,

which at some level, a lot of those tasks we all know,

which occupy a lot of humans time today

are probably achievable with where the technology is now

without requiring a great deal of spend.

At least I believe that will increasingly be the case.

And so I guess if you look at some of these spending charts,

what would you predict--

how would you predict they would change over time?

My guess would be that the ambient spend, so to speak,

would go down, right?

And the spend is going to shift towards,

in my view, things that are much more

challenging for the technology.

In other words, some difficult precision

medicine efforts, some clinical trial based matching,

some other more sophisticated aspects.

And a lot of the spend is going to, I think--

I don't want to call it totally deflationary because there

are services and things, but at some level,

I just don't know how this will ultimately

be the best way we track the impact of AI in health.

SPEAKER 2: Yeah, it's not.

You're totally right.

There's entire categories missing from this.

No offense to Menlo Ventures.

And I think it's good in terms of saying, hey,

what are some of the main use cases

and, directionally, where are players putting their resources.

But even if you go back to the page before,

to not have green for payers in some of these buckets,

even if you were to confine yourself

to these kind of core six use cases is an intellectual miss.

And then if you go to the next slide

and you say, OK, where's the puck going,

I couldn't agree more with Matt in terms of categories

that are just not represented.

I think clinical AI is going to be

massive in three to five years.

And it's not just what you're seeing today.

Even from some of the EHR vendors and some of the others

is, how do we improve provider productivity?

Ambient?

Doesn't really do that?

It might do that a little bit, a few percent,

but the in-basket management and some of the simplifying

back office operations and stuff like that, which

is different than payer operations,

I think that's going to be tons of activity and investment

in those spaces.

But then the whole space of clinical innovation

is fundamentally missing here.

And that's what increasingly becomes possible

when you have more data, better data,

and you're able to connect these sources

and have folks working on each individual case

to match precision medicine point in a scalable way.

SPEAKER 1: Yeah.

It's clear and good.

And all the conversations we have is

it is right to start on the administrative side first.

How that spending happens and what comes with clinical AI

is going to change what medicine looks and feels like.

And just to show one more piece of data, this was--

SPEAKER 2: Justin, even before you do that-- sorry--

even the spend and administrative

is going to collapse in interesting ways, right?

Like payment integrity, revenue cycle management,

prior authorization.

As these systems are increasingly

able to talk to each other and you've

got APIs connecting everything, you've

got high fidelity on your data.

Because it's, in part, enabled by ambient, you're connecting.

You have better interoperability.

You have more longitudinal data.

You have deeper data.

Now, you can do transactions in real time.

Now you have transparency in a way you've never

had transparency before.

That's an entirely different category of it's

not just spend going up, but in some ways,

you're shrinking value pools.

In other ways, you're creating new value pools that we also

need to think about.

So it's fine to take snapshots, and those

are snapshots that I would argue aren't entirely accurate,

but what really is the interesting work is,

what does tomorrow look like?

SPEAKER 1: And so you are squarely--

Because it's come up on a number of panels I've been on,

people talking about right now, there's this AI arms race.

You have talking UHD about 1,000 AI systems providers,

at least in Menlo's report, talking about more spend on AI

coding and billing.

So there's the worry.

And I've even talked with some on the administration of,

do these AI fighting AI just increase net health care costs?

But it sounds like you're very much on the opposite side

of maybe that's a short-term blip piece,

but where we're going is no way.

SPEAKER 2: The honest truth is different players

in the ecosystem structurally have

different incentives on what happens to health care costs.

And I'm not going to go and name them.

You can look at trend analyzes and make those conclusions

for yourself.

Where I sit-- and this is not intended

to be an infomercial for UnitedHealth Group--

but where I sit is actually, I feel privileged.

And frankly, the reason I came to this company,

it felt almost like an extension of being in the government.

The goals were identical, which was

how do you serve the maximum amount of this United States

health population to the best of your ability, achieve the health

goals, health outcomes that you want to achieve

while controlling price?

If you think about the function of a payer,

or in the case of Optum Health, fully delegated model,

or in the case of Optum Insight in terms of what they're

bringing to the industry, it's all

centered around doing it in a way that brings price down.

That is-- I will comment on us--

our incentives are to bring price down

while maximizing the footprint and depth of impact

on care delivery.

No matter where you sit-in the organization, philosophically,

that is our focus.

I think there are countervailing forces in the industry

that want to see it go to the other side.

Who wins will be a multifactorial play.

And what that looks like a few years from now,

also, the collaboration model between payers and providers

will be a multifactorial play.

At some point, it can't just be about punching each other

in the face and seeing who blinks last or blinks first,

whatever the analogy is.

At some point, it's got to be sitting together and saying,

hey, let's bring data together.

Let's be transparent about what rules engines we're applying

and make sure that, based on that transparent understanding

of what's there and what's possible

and what we're trying to do, that we execute a plan together.

And then that show starts showing up in contracts

and so on and so forth.

And we're ways away from there.

I think there's probably going to be,

in a way, increasing tension between payers and providers

before it gets to the other side,

and regulators are going to have to decide if where and how they

want to step in.

I won't step on their toes.

I've lost that baton.

And we'll see how it plays out.

But I am very committed for the role

of UnitedHealth Group and my role

in it to be a deflationary force.

SPEAKER 3: And I think as you're talking about this--

and, again, I feel the same way in terms, like, there's always

been this kind of yin and yang, fox and rabbit kind of moments

that's just based on the way the incentives are structured,

but I guess I wonder if there's a third or maybe even

a fourth trend that we started with that also

starts to play a role here.

And I'm thinking about, again, the patient who let's be honest,

we definitely think of them.

All of our mission statements talk

about them and all of our different organizations.

But ultimately, the true empowerment

really wasn't there until I think, recently.

Now I'm starting to see this kind

of new empowered, patient voice leveraging this technology,

and I wonder how that starts to play

a role in some of these traditional incentives.

And frankly, even to the point where

I would also throw physicians in who,

yes, we've had a voice in various ways,

but not maybe at the level that really

impacted some of these broader tensions in the market.

And more physicians than ever are

choosing to work at places that have things like ambiance.

The technology is actually helping physicians

make decisions and potentially even strikeout on their own,

for example, more than I've seen before.

I don't know if there's a way to put

a finger on the pulse of this trend at this point,

but, I guess, it seems like it's another player

on the field that, before now, hasn't really impacted the game.

SPEAKER 2: Here's your hot take for the day,

because I know we're nearing the end.

This industry is going to look drastically different

in five years.

Drastically.

The role of the incumbent, obviously,

as a member of an incumbent, as an employee of an incumbent,

the role of the incumbent is for the incumbent to decide.

But if the incumbent keeps just doing what they've historically

done, they will have a significantly diminished role

for one reason.

The most important people in this entire ecosystem

are patients, the citizens of the United States.

And increasingly, it'll be citizens of the world

as all these systems get more connected.

But the citizens, scope it to the 330 million here,

they will make that decision, and they

will have the options to make that decision.

So if we're going to be in service of them,

then we gotta step into the lead in disrupting the industry.

First, disrupt ourselves, then ultimately disrupt the industry.

That's a choice for all of us to make.

And it's hard because then you're

saying no to legacy business models and legacy practices,

but you're saying yes to the future

and you're saying yes to the most important people

in the ecosystem-- your patients, your members.

SPEAKER 1: I totally love that, and I couldn't agree more.

I think what's fascinating right now

is everyone who has said for their careers in healthcare,

healthcare will look so different

in five years, everyone who has made that statement or that hot

take that you mentioned has been wrong in pretty

big ways for the past decades.

I completely agree with you, but I

mention that because the historical context for most

leaders now, whether it's a big health system, payer, government

organization across the country, it

has been the wrong answer to lean into the new technology.

They got to that point in their career

because they didn't jump too early

at AI in a different iteration.

They didn't jump too early.

But it is coming, and I agree with that force

of the consumer who will force this upon the industry.

And with where we started, that 3x increase now and AI,

I feel comfortable with, as my doctor,

that is what's going to drive change.

SPEAKER 2: You gotta--

let's just-- a couple factors, right?

When you had high tech and meaningful use

in the decades prior, basically spanning over

the last couple of decades, that was a multi-billion dollar

central investment to drive a change

in an industry that ultimately made physicians less satisfied

and patients less satisfied.

Now you get a printout of a discharge summary

and you have no idea what the hell it says--

excuse me-- what it says.

Now, forward looking, you have a technology

that you just showed the stat, the 11 to 28%,

where patients are increasingly using it and increasingly

getting comfortable.

Doctors-- look at the trend in ambient.

Doctors are using it.

They are demanding.

So, to Matt's point, if you don't have this,

then I'm not going to practice here.

So at the end of the day, if you want technological--

if you want to transform an industry,

and it's an overused term, but if you really

do want to transform anything, you

have to show the end user how what you're offering

them is meaningfully better.

You can't just throw money at them.

You can't just dangle rules at them.

Just show them how it's meaningfully better.

If you do that, change will happen super fast.

And I do believe, especially with what's

coming-- we've only talked about what's been,

but what's coming with the ability

to execute intelligent, highly intelligent tasks

in a coordinated and orchestrated way,

essentially being a boots on the ground workforce

that is not human, I think that's going to materially

change what we can do, how it's done,

and potentially, TBD, who does it.

SPEAKER 3: Yeah.

SPEAKER 2: I don't think I'm going to be wrong.

We'll see, Justin.

SPEAKER 3: You're putting your bet on the same place we are.

And again, this goes back, again,

to the beginning where I do--

I just feel it.

I know they always say that.

It feels different this time.

And the reason I say that is that it's

coming from the bottoms up and not the top down.

And I just feel like that is the way

that we've seen the true change happen in legacy industries

or businesses, et cetera.

And so I'm here for it.

And this is why we do this show, because we get to keep tabs

on what's happening.

SPEAKER 2: So are all the major institutional investors

in this country, so fingers crossed.

Bigger issues if we can't crack this.

SPEAKER 1: Well, on that happy note, Syed, thank

you so much for joining us.

SPEAKER 2: Great to see you guys.

Let's make this change happen responsibly.

It's going to be awesome for humans, for patients.

And that is, I know, what the three of us live for.

It is why we're friends, other than hopefully

having decent personalities and the like.

See you guys.

SPEAKER 1: Thank you.