Latest / Becker’s Healthcare Podcast / How AI Is Reshaping Healthcare's Next Decade with Manav Sevak, Venkat Mocherla, & Malinka Walaliyadde

Transcript

How AI Is Reshaping Healthcare's Next Decade with Manav Sevak, Venkat Mocherla, & Malinka Walaliyadde · Becker’s Healthcare Podcast

0:02
- There's a small circle of people who truly
0:04
- understand what it takes to lead a health
0:05
- system right now. The decisions are enormous.
0:08
- The variables are constantly shifting, and the stakes
0:11
- of getting a call wrong don't stay contained.
0:14
- That's not a complaint. It's a structural reality.
0:17
- And it's why the most effective c suite
0:19
- leaders are deliberate about where they go to
0:21
- pressure test their thinking.
0:22
- Becker's fourteenth annual CEO and CFO roundtable
0:26
- brings together more than 1,500
0:28
- executive level attendees
0:30
- and nearly 400 speakers for four days built
0:32
- around growth, financial sustainability,
0:35
- workforce strategy, and the leadership questions that don't
0:38
- have easy answers.
0:39
- Join us November second through the fifth in
0:41
- Chicago.
0:42
- For the agenda and event details,
0:45
- visit beckershospitalreview.com
0:47
- and click on the events tab in the
0:49
- upper right.
0:50
- This is Scott Becker with the special episode
0:54
- of the Becker's health care and Becker business
0:57
- and private equity podcast.
0:59
- We're joined today by three remarkable health care
1:02
- technology founders,
1:04
- all at the forefront of sort of technology,
1:07
- artificial intelligence,
1:08
- and health care.
1:10
- We've got the founder of Acasa,
1:12
- the founder of Midstream Health,
1:15
- the the founder of Memorial
1:17
- Health, all with us today. And I'll ask
1:19
- each to take a moment and introduce themselves.
1:21
- I'll start with Melinka from Acasa.
1:24
- Take a second on sort of introducing yourself,
1:27
- the organization,
1:28
- and I'll ask each of you this question
1:30
- about the problem in health care
1:32
- that you've been most passionate about solving or
1:35
- where you're most focused today.
1:37
- Let's do that as part of your introduction,
1:39
- then I'll ask Manav and Venkat to do
1:41
- the same thing, and then we'll we'll we'll
1:43
- we'll pivot to different questions for each of
1:45
- you. Melenka, can you take a moment and
1:48
- introduce yourself?
1:49
- Of course. Delighted to be here, especially with
1:51
- other folks, that I've known for a long
1:53
- time and and respect greatly. So,
1:56
- I'm cofounder Sierra at Acasa,
1:58
- and what we do at Acasa is we
2:00
- develop AI for the health care revenue cycle,
2:03
- right, with the particular focus on the mid
2:05
- cycle where the most complex interactions between clinical
2:08
- and financial data occur.
2:10
- And and the way we think about this
2:11
- is American medicine is the best in the
2:13
- world,
2:14
- but the American health care system is not.
2:17
- And the gap that exists there comes from
2:20
- how complicated we have made paying for health
2:23
- care, in in The US.
2:25
- That's what we call revenue cycle. Right? And,
2:27
- and and if you think about it, revenue
2:29
- cycle, like, at its core, it is how
2:30
- do you capture and communicate the full patient
2:32
- story to a pair.
2:34
- If you do that correctly, claims don't don't
2:36
- get delayed, things don't get denied, health systems
2:39
- get full credit for the care they've delivered,
2:41
- but that's been really, really hard to do
2:44
- in the past because of the complexity of
2:46
- under understanding the clinical record. So what I
2:48
- am passionate about is solving that problem, closing
2:50
- that gap, and having,
2:52
- you know, the way we deliver care be
2:54
- as,
2:55
- world leading as the quality of care itself.
2:57
- And, Malika, let me ask you a question
2:59
- there.
3:00
- Currently, estimates are all over the place, but
3:03
- The US spends about 5,000,000,000,000
3:04
- out of $2,627,000,000,000,000
3:06
- dollars a year,
3:08
- of its GDP on health care. So about
3:09
- 5,000,000,000,000 spent a year on health care.
3:12
- Revenue cycle, I would assume, is a good
3:14
- 10% of that somehow or another.
3:17
- Yes. So that means we're sending 500,000,000,000,000
3:19
- on revenue cycle.
3:21
- Is that inevitable
3:22
- with the amount of different payers and payments
3:24
- that we have going? Is that sort of
3:26
- gonna be the the the case going forward
3:29
- if we've got 50 different Medicaid systems, the
3:31
- Medicare system, the VA,
3:33
- dozens of different payers, including the big four,
3:36
- plus several others that really focus on Medicaid
3:38
- managed care,
3:39
- can that number be improved,
3:42
- or is that is that destined to be
3:44
- that we're gonna spend 10% of 500,000,000,000,000
3:47
- or 5,000,000,000,000, 5,000,000,000,000, not 500,000,000,000,000, 5,000,000,000,000
3:50
- on on revenue cycle?
3:52
- It can and should be lower than that,
3:55
- is the is the quick answer. Right? It
3:57
- does not make sense that we spend so
3:59
- much
4:00
- on the back end
4:02
- to still deliver not a great experience. Right?
4:05
- So that the we're spending a lot and
4:06
- the ROI we're getting is not great. And
4:08
- the incredible thing is with the with these
4:10
- new generative AI capabilities through large language models
4:13
- that we're getting today, you can do things
4:15
- that used to be impossible because of the
4:17
- complexity. You can do things
4:19
- much faster, much more efficiently, much more accurately,
4:21
- and we anticipate and are working towards
4:23
- reducing
4:24
- that cost to collect across the board. Thank
4:27
- you. We'll come back to you in a
4:28
- moment.
4:29
- Manav, you're you're a serial entrepreneur in health
4:32
- care.
4:33
- Started a company when you were in college,
4:36
- exited at some point that company,
4:39
- starting something else. Talk a little bit about
4:41
- yourself. Introduce yourself,
4:43
- and and talk about what you're most passionate
4:45
- about in health care today.
4:47
- Yeah. Thanks so much for having me, Scott,
4:49
- and and great group of folks.
4:51
- So really quick introduction on myself. My name
4:54
- is Manu Savak. I'm a
4:56
- scientist by training and a former founder and
4:59
- CEO of a company called Memorial Health. And
5:01
- big picture, what we do is we build
5:03
- workflow automation software for health systems,
5:06
- to specifically help automate a lot of different
5:09
- follow-up tasks and making sure that care teams
5:12
- are able to manage their patients after they've,
5:14
- left the walls of the hospital. Scaled the
5:16
- company for,
5:18
- over nine years after starting it in college,
5:20
- and it was acquired in 2024
5:21
- by another business in our space called Comir,
5:24
- and now spending some of my time,
5:27
- investing and sort of working with early stage
5:29
- founders and spending some of my time, starting
5:32
- starting something new as well in health care.
5:34
- And
5:35
- the the, I would say, like, you know,
5:38
- distillation of all the work with my prior
5:41
- company and this next one is specifically focused
5:43
- around solving navigation.
5:45
- I'll I'll build a little bit on what
5:47
- Melinka said
5:48
- that,
5:49
- you know, the The US health care system
5:52
- is unbelievable in the sense that for the
5:55
- highest acuity cases,
5:57
- it is able to perform extraordinarily well, and
6:00
- and the breadth of therapies that we're able
6:02
- to give to patients is extremely, extremely high.
6:04
- But there's there's a ton of last mile
6:07
- issues that exist in how people are able
6:08
- to actually access those.
6:10
- And it's very you know, there's there's quite
6:13
- a bit of cognitive dissonance between
6:15
- on one end of the spectrum,
6:17
- people coming from all over the world to
6:19
- The US,
6:20
- taxes very, very rare treatments and therapies, and
6:22
- on the other end of the spectrum,
6:24
- the average American,
6:25
- when they're not feeling well, not being able
6:27
- to get basic information on,
6:29
- finding the right provider and being able to
6:32
- to just understand what's wrong with them. So,
6:35
- spending a lot of time thinking about how,
6:38
- not just AI, but but generally advances in
6:40
- in the world are, are changing that dynamic.
6:44
- With the nation, I'll I'll go I'll ask
6:45
- you a question similar to what I asked
6:48
- Malinka.
6:49
- Is, Manav,
6:50
- with a country of 350,000,000
6:52
- people, again,
6:54
- third largest in the world, is there is
6:56
- are we inevitably gonna see a regression to
6:59
- the norm
7:00
- in terms of health care quality
7:02
- compared to other large nations, or is there
7:04
- something The US can do
7:07
- to continue to or to move in the
7:09
- right direction? I mean, it seems like currently,
7:12
- cost is going in the wrong direction.
7:14
- Access is really challenging.
7:16
- Quality is
7:18
- is good in in certain spots,
7:21
- but it's certainly challenging in other spots just
7:23
- in part due to access problems.
7:26
- But but can we bend that curve? We're
7:28
- not so not even talking about the cost
7:30
- curve right now, but can we do better
7:32
- in getting better health care to more people,
7:35
- or will we regress to the norm and
7:37
- and have more challenges?
7:39
- Yeah. Great question.
7:40
- The the short answer is I I don't
7:42
- think that well, I we're on the path
7:44
- of regressing to the norm, but I don't
7:46
- think that that has to be the case.
7:48
- I think, one of the biggest levers, not
7:50
- just in health care, but in every industry
7:52
- from
7:53
- all the new sort of advancements in artificial
7:55
- intelligence is this concept of abundance.
7:58
- There's you know, not only are there exceptional
8:01
- therapies, but we now have a lot of
8:02
- the tooling
8:03
- that allows us to scale really excellent clinical
8:06
- resources, access to those therapies.
8:09
- And it's really a question of whether or
8:11
- not people in the industry choose choose to
8:13
- use those tools the right way. So,
8:15
- I don't think it's inevitable by any means.
8:17
- I do think that, you know, there's a
8:19
- lot of perspectives that health care over time
8:22
- should become deflationary as,
8:24
- the percent of GDP shrinks, and and
8:27
- I am not entirely sure that that's also
8:29
- what will happen. I think we'll probably get
8:31
- closer to health care growth being in line
8:33
- with inflation.
8:34
- But on on the quality side, I think
8:36
- it's more around using the tools to make
8:39
- sure that we stay ahead of the curve,
8:40
- and and sort of course correct on the
8:42
- direction that we're heading in.
8:44
- Thank you.
8:45
- Venkat,
8:46
- you've had this fascinating career
8:49
- both as an investor
8:50
- where you originally, I think, got to know
8:52
- Melinka at a 16 z,
8:55
- and and now as a founder and entrepreneur,
8:57
- talk a little bit about yourself
8:59
- and and the biggest problem in health care
9:01
- that you're passionate about trying trying to solve.
9:04
- Yeah. Yeah. Absolutely. This is a this is
9:06
- a fun reunion.
9:07
- You know, and,
9:09
- health care is so small, and the world
9:11
- of health care entrepreneurs is even smaller. So
9:13
- it's a joy it's a joy to be,
9:15
- on this podcast with these guys.
9:17
- Yeah. Just to introduce myself, as Scott mentioned,
9:20
- I'm the cofounder of a company called Midstream
9:22
- Health,
9:23
- focused entirely on how do you think about
9:24
- this AI native platform for some of the
9:27
- largest health, health systems in The US
9:29
- all around automating financial actions.
9:32
- And, you know, in this
9:34
- current margin environment,
9:36
- where every dollar every dollar has always mattered,
9:39
- and there's tremendous amount of excitement and opportunity
9:41
- as as, Manav and, Malinka have mentioned using
9:44
- AI.
9:45
- I think, you know, my own career, I
9:47
- I've, one way of looking at is maybe
9:49
- I've not stuck to a continue one job.
9:51
- You know, I've had the privilege of working
9:54
- at, starting new companies inside a Fortune 500
9:56
- company called DaVita. We started new companies in
9:58
- primary care. I've worked I've worked in 13
10:01
- countries, globally for what's now part of Optum.
10:04
- And then I got into AI and ML
10:06
- about twelve years ago on the founding team
10:08
- of a company called Kuventus.
10:10
- And so it's been, it's been a pretty
10:12
- amazing thing. I'll just sort of say, just
10:14
- to reiterate,
10:15
- I I think there's no sort of optimism
10:17
- between the three of us. If people are
10:19
- looking for, pessimistic views on the future of,
10:21
- technologies, probably this is not the right podcast
10:23
- to dial into.
10:25
- You know, Mara mentioned abundance. I just think
10:27
- that, whether it's, Midstream, Acasa,
10:31
- or CommonSpirit, or Houston Methodist, or or New
10:33
- York Presbyterian, or any any of these health
10:35
- systems,
10:36
- I think the excitement now is that these
10:38
- organizations and sectors like health care where productivity
10:41
- has always been a challenge,
10:43
- finally have the framework around a 10 x
10:45
- teammate with a 10 x stack. And I
10:47
- think that,
10:48
- that's the excitement around,
10:51
- all these new tools and superpowers we're seeing
10:53
- and and creating that sort of 100 x
10:55
- advantage for a sector that's been plagued with
10:57
- productivity challenges for a long time with high
10:59
- cost issues, low quality, all the things we
11:00
- know about. So I think, Scott, this is
11:02
- probably the most, you know, I got into
11:04
- the AI space around twelve years ago. It's
11:06
- probably the most exciting and optimistic,
11:08
- I've I've been. If you ask Melincans, not
11:11
- of them are as excited and optimistic they've
11:13
- been. But,
11:14
- but, yeah, that's sort of a little bit
11:16
- of me in a nutshell on my viewpoint.
11:19
- No. Thank you very much. And I and
11:21
- I wanted that optimism, and I see reason
11:23
- and room for optimism in health care more
11:25
- so today than I've seen in the last
11:27
- few years for for a bunch of different
11:29
- reasons.
11:30
- Venkat, let me start this term with you.
11:34
- Health care organizations
11:35
- continue to face tremendous staffing and financial pressures.
11:39
- I know you work with some of the
11:40
- largest systems in the country.
11:43
- Where do you see the biggest opportunities today
11:45
- for AI and automation to create value in
11:47
- health care? Where do you see the biggest
11:49
- opportunities for artificial intelligence?
11:51
- Yeah. I mean, I think for me, I
11:53
- know it sounds
11:54
- very generic, but but,
11:58
- to me, the the American health care is
12:00
- so interesting in that. The most of the
12:02
- health system in The US, including our customers
12:04
- and our our partners like Common Spirit and
12:06
- Mount Sinai and Houston Methodist, others are all
12:09
- not for profit.
12:10
- And it's an interesting dynamic where they all
12:13
- negotiate with for profits. So it's payers, pharma,
12:16
- med device, or all fortune 550 or five,
12:19
- depending on which company, which bucket you choose.
12:22
- And so I think that the excitement for
12:25
- me is sort of twofold. First is internally,
12:28
- some of these large health care systems,
12:31
- they've just not been able to solve around
12:33
- some of the core,
12:35
- optimization issues around how do we, you know,
12:38
- negotiate against these payers the right way, how
12:40
- do we sort of purchase things in a
12:41
- more optimized way, how do we make sure
12:43
- our,
12:44
- costs are in line with our revenue, especially
12:47
- because private equity is taking over all the
12:49
- interesting high profit areas.
12:50
- And so for us, I think the excitement
12:53
- that we see in this era
12:54
- for health systems is not just thinking about
12:57
- areas like the revenue cycle, which are incredibly
12:59
- important,
13:00
- and there's incredible companies like Acosta right on
13:02
- it. But what about everything else? And and
13:05
- it turns out these health systems that are
13:06
- not for profits, which have an amazing mission,
13:08
- are make are making something like a single
13:10
- digit margin and every day make thousands of
13:13
- financial decisions. And so how do you, for
13:15
- the first time, have a consistent way of
13:17
- of helping them make decisions? I think the
13:18
- way of sort of the operating jump I
13:20
- see is,
13:21
- not too long ago, I I you know,
13:23
- I can still see him in the Mazda
13:24
- MPV. My parents arguing about which exit to
13:26
- take because they had MapQuest printed,
13:29
- and, my dad was damn sure about his
13:31
- way of taking directions even though he probably
13:33
- is dead wrong.
13:34
- And then somewhere along the way, we have
13:36
- Google Maps and Waze. And not too long
13:38
- ago, my wife and I went to a
13:40
- remote town in Japan, and I, who don't
13:42
- have a great sense of direction, made it
13:44
- there on time. So, you know, I just
13:45
- think that jump in an operating environment with
13:48
- a single digit margin environment, that to me
13:50
- is the most exciting thing. It's not any
13:51
- one thing, but it's how we make every
13:53
- decision that I think will,
13:55
- be 10 x better, whether it's clinical, operational,
13:57
- or financial.
13:59
- Thank you very much. And I and I
14:00
- love those examples that we see in daily
14:02
- life and trying to see how they ultimately
14:04
- translate into health care. I I I I
14:07
- love that.
14:08
- Manav,
14:09
- when you think about
14:11
- health care leaders and what they're getting wrong
14:13
- about AI adoption,
14:15
- in in in what separates organizations that seem
14:18
- to be getting this right and really making
14:20
- progress
14:21
- and succeeding from those that might be struggling
14:23
- in trying to adopt and work with AI,
14:25
- particularly in an enterprise level versus a one
14:28
- off individual level?
14:30
- So I think,
14:32
- you know, I I think one big trend
14:33
- that I hear from health systems and health
14:36
- plan leaders
14:37
- is
14:38
- essentially how a lot of people have chosen
14:40
- to use AI tools to essentially just become
14:42
- more adversarial. So the number of health systems
14:44
- that are using tools to, you know, essentially
14:47
- better adjudicate against plans and the number of
14:49
- plans that use tools to find ways to
14:51
- to deny more claims has,
14:53
- sort of created this. It's it there's always
14:55
- been somewhat of an arms race, but it
14:57
- sort of enhanced this arms race to some
14:59
- degree. And
15:01
- I think that it's generally the wrong the
15:03
- wrong framework,
15:04
- especially when they're thinking about implementing tools like
15:06
- that reactively.
15:08
- I think that
15:09
- the the
15:11
- best applications of AI that I've seen and
15:14
- the way that people have thought about financial
15:16
- pressure and staffing pressure
15:19
- is, from my perspective, on rightsizing utilization,
15:22
- and giving people a lot of confidence in
15:24
- managing their own health. So when,
15:27
- you know, AI tooling becomes more and more
15:29
- prominent for consumers and it's easier for people
15:32
- to access more health care information,
15:34
- naturally, they have a lot more choice in
15:35
- how they think about navigating the health care
15:37
- system.
15:38
- As a result of that choice, for health
15:40
- systems to stay competitive
15:42
- and also manage staffing shortages that they have
15:46
- and utilize their resources as efficiently as possible.
15:49
- It's important for them to be able to
15:51
- to not only treat people in cases where
15:53
- they actually need to be treated, but also
15:55
- make sure that they can maintain a relationship
15:57
- with them after they leave.
15:58
- So ensuring that we're sort of filling the
16:00
- gaps between visits and managing people longitudinally, I
16:03
- think, is really important. And scaling the capacity
16:05
- of providers,
16:06
- I think, is is incredibly important just because,
16:09
- ultimately, that's that's probably one of the biggest
16:11
- cost levers for making sure that people are
16:12
- using the health system the right way.
16:15
- And you're using these best in class clinical
16:17
- resources
16:18
- in cases where people really need them, and
16:21
- and do it in a way that actually
16:23
- is is manageable from a cost perspective.
16:26
- Thank you.
16:27
- Melanca, let me ask you this question.
16:30
- We're at a cost and you've been at
16:31
- this for how long have you guys been
16:33
- at this now? You were at the forefront
16:35
- of artificial intelligence
16:37
- and revenue cycle before most of us knew
16:39
- what artificial intelligence was.
16:41
- Maybe not Venkat, Manav, and you, but the
16:44
- rest of us.
16:45
- When you look at that today, what are
16:47
- you most focused on right now at Akasa?
16:50
- And where do you see the greatest opportunities
16:51
- for growth and impact over the next twelve
16:53
- to twenty four months in what you're doing
16:55
- with health systems?
16:57
- Yeah. So, yes, we started the company pretty
17:00
- large language model. That's sort of the dividing
17:02
- line for us as a company. And and
17:04
- I say that because that is what has
17:06
- made so much possible, I mean, for I
17:08
- think everyone was called around the world. Because
17:10
- in in in 2023,
17:12
- right, we have the advent of large language
17:14
- models. People call it generative AI. It's just
17:16
- a such a substantially more powerful form of
17:19
- AI that could do
17:21
- way more complex things, and it just made
17:23
- it things that were impossible before
17:26
- are now actually possible and sometimes even easy.
17:29
- Right? And and to to go from a
17:31
- world where,
17:33
- you know, understanding something like a clinical record
17:36
- was impossible for software to do.
17:38
- Now software can understand a clinical record better
17:41
- than a human.
17:42
- Right? And and the opportunities that unlocks is,
17:44
- like, massive.
17:45
- Right? And so that's what we've taken into.
17:48
- Now there's many other unlocks from LLMs, and
17:50
- I think, you know, folks here are doing
17:51
- a lot many other things. For us, though,
17:52
- in our world, the unlock that comes from
17:55
- plugging into the clinical record
17:58
- at scale is the biggest unlock. Right? And
18:01
- everything downstream of that dramatically changes. All the
18:03
- problems downstream of that in, you know, in
18:05
- in in the revenue cycle and documentation and
18:07
- coding and all of these things that were
18:09
- impossible to do with software and AI before.
18:11
- Now you can do,
18:13
- and you can actually do the most complex
18:15
- bits. Right? And so that's what I'm most
18:17
- excited for in this upcoming, that's what we've
18:19
- been working on. That's what we've deployed at
18:22
- scale at our at our customers.
18:24
- And,
18:25
- and and just to go back to one
18:26
- of the questions you were asking earlier, right,
18:28
- like, what are maybe people getting wrong about
18:29
- AI adoption? Sometimes it is it is human
18:32
- nature
18:33
- to apply these things to the simplest tasks.
18:35
- Right? Because that's you sort of think I'll
18:36
- do the simple things. No. No. Just do
18:38
- the you can actually do the very complex
18:40
- things now. Right? And it's actually useful to
18:43
- start there because that's,
18:45
- that's often where you see the biggest ROI.
18:47
- Right? And so you can just directly go
18:50
- and address some of the most complex challenges
18:52
- in the domain. And so for us, right,
18:53
- we focus on, for example, inpatient facility,
18:56
- in the in the revenue cycle. That is
18:58
- the most complex stuff, and we do it
19:01
- at some of the most complex health systems
19:03
- in the world. Some of our customers are
19:04
- folks like Cleveland Clinic and Mayo Clinic, and
19:06
- we have seen remarkable
19:07
- remarkable,
19:08
- outcomes there because what you can do with
19:11
- LMs now is so much greater. And so
19:13
- we're just gonna keep doubling down and and
19:15
- investing, and in particular doing the most complex
19:17
- stuff that was historically unattainable because
19:21
- that is where often you see the most
19:22
- value creation because it is the most complex.
19:25
- I actually just to double down on that,
19:28
- I I think it's so interesting that
19:30
- so much of the way we were making
19:32
- decisions pre LMs
19:33
- is a proto efficiency, like, eighty twenty, and
19:36
- everything is like an eighty twenty. And,
19:38
- and I think that it's interesting. I I
19:40
- think that if if your margin is in
19:42
- the single digits,
19:44
- you should work on the 20. You should
19:46
- work on the really hard,
19:47
- really complicated things that are actually applicable for
19:50
- compute and not humans. And I think of,
19:52
- you know, Malinka, what you said, which is,
19:54
- like, taking it on the most hardest decision
19:56
- is so spot on. Because when I think
19:58
- about how hospitals on the cost side
20:01
- have more SKUs than Walmart and Target combined,
20:05
- and and yet, you know, if you think
20:06
- about functions like supply chain, some of their
20:08
- legacy functions, they've just been not as AI
20:11
- native in the past, and now they are.
20:13
- And so when you think about that jump,
20:15
- I think you're absolutely right. The the focus
20:17
- in the RPA,
20:18
- you know, phase was let's do the simple
20:20
- things because they tend to break down. But
20:22
- this is the actual inverse, and I think
20:24
- it's a it's a spot on,
20:26
- you know, observation.
20:27
- I I I wanna ask you about that.
20:29
- And and because so many of us are
20:31
- prone to
20:33
- take on easy tasks first, get wins. And
20:36
- what I see in sort of the AI
20:38
- world for so many of us is,
20:40
- you know, we use ChatGPT
20:42
- everyday for little things.
20:44
- And you've made a different argument
20:47
- that really working towards enterprise and big solutions
20:50
- is the real payoff.
20:52
- How do you move teams and minds towards
20:54
- that? Because that's more complexity that most of
20:57
- us wanna deal with. Malika, how do you
20:59
- think about that? Because I think you're right,
21:01
- but I think so many of us as
21:02
- humans
21:04
- are engineered or built to think in terms
21:06
- of simple victories versus big enterprise
21:10
- challenges.
21:11
- And and and and how do you
21:13
- manage
21:14
- and I'll talk to Manav and Venkat about
21:16
- this exact same question because both of them
21:18
- are doing this at scale as well and
21:20
- and working with huge systems and enterprise efforts.
21:23
- How do you move people towards that thinking?
21:25
- Because that seems so hard for the average
21:27
- person, including myself.
21:28
- Yeah. It it is a mindset shift, but,
21:31
- again, the capability sets that we've gotten over
21:34
- the last couple of years are so massive.
21:36
- So, I mean, what I would say to
21:37
- you is, like, do you know, not it's
21:39
- it's,
21:40
- easy things first, get wins,
21:42
- I think I think made sense in a
21:43
- pre owned world. Now it's, like, do the
21:45
- hard thing first, get way bigger wins. Right?
21:48
- And just to give you context, with with
21:49
- with with some of our customers, we're literally
21:51
- delivering
21:53
- 9
21:54
- figures of value a year. Right? Like, that
21:56
- is the scale of impact that's not possible
21:58
- when you actually,
22:01
- take on the very complex things. And the
22:03
- nice thing is when you just start there,
22:06
- it becomes so much more straightforward
22:08
- to increase awareness, adoption. Right? Like, I heard
22:12
- this I heard this line, somewhere where someone
22:14
- was saying, you know, when you do the
22:16
- simpler things,
22:18
- you see AI everywhere except in the bottom
22:20
- line. Right? It was like, I don't know,
22:21
- some pity thing I heard somewhere. And I
22:22
- was like, that's only if you do it
22:24
- if you try to scale the old way
22:25
- of doing it.
22:27
- If you actually lean into the new way
22:29
- of doing it and actually take these incredible
22:30
- capabilities and actually take tackle the really complex
22:33
- things head on, you can deliver so much
22:36
- ROI that you will never hear an argument
22:38
- like that from the enterprise. And when you
22:40
- can show wins like that quickly,
22:42
- it becomes it creates this incredible flywheel where
22:45
- people just want more.
22:47
- And so if you that's part of what
22:48
- it is driving. I mean, you've seen the
22:50
- growth of AI be very rapid.
22:52
- Right? And it's because
22:54
- people that didn't believe
22:56
- believe when you tack on a complex challenge
22:58
- delivering credible value, and then you start seeing,
23:01
- more and more people lean in without really
23:03
- needing needing to convince them because they're convinced
23:05
- already.
23:07
- And let me ask this question. Manav, let
23:09
- me
23:10
- spin to you on this question. And then
23:11
- if anybody wants to comment on that last
23:13
- question about going through this in enterprise level
23:15
- versus a one off level, we'd love to
23:18
- have your thoughts.
23:19
- But but, Manav,
23:20
- we talk so much about
23:23
- AI and
23:25
- administrative work
23:27
- in revenue cycle,
23:29
- in logistics,
23:32
- operating management, predictive analytics, which is where I
23:34
- think quite frankly Venkat started his career.
23:37
- What are some of your thoughts on where
23:39
- we're gonna see the most improvement in clinical
23:42
- outcomes
23:43
- or or even in clinical navigation. Where where
23:46
- do you see this in three to five
23:47
- years,
23:48
- and how do you think this will fundamentally
23:50
- change experience
23:51
- clinically and maybe in navigation too?
23:54
- Yeah. Great question. One thing I'd I'd maybe
23:57
- add to what Milenka was mentioning just before
23:59
- I get to your question is
24:01
- I do think there's sort of this interesting
24:03
- question, especially for health systems and health plans
24:05
- of AI native versus AI enabled,
24:07
- just because the delta between
24:10
- thinking of building a system entirely from scratch
24:12
- versus augmenting an existing system is actually so
24:15
- large
24:16
- in in uniquely in those types of businesses
24:19
- that
24:20
- and there's a lot of cases where I
24:21
- think it's actually easier to, like, go through
24:23
- design exercises with leaders at those places and
24:27
- say, look. If you were to, like, redesign
24:29
- an entire prior authorization or utilization management function
24:32
- from scratch, or if you were to redesign
24:34
- a care management function from scratch,
24:36
- you'd probably just make very different decisions on,
24:38
- like, what you did and did not support
24:40
- in inside of that. So,
24:42
- I think that that also, I think, has
24:44
- somewhat changed this
24:46
- concept of what's possible for folks.
24:48
- As far as on the clinical side where
24:50
- I think
24:51
- there's there's the most leverage, I think there's
24:54
- two things that are gonna change dramatically. The
24:56
- first is
24:57
- this this kind of future
24:59
- where people have access to some form of
25:01
- an AI doctor
25:02
- is certainly gonna come one way or another.
25:05
- What that means is something different for everyone.
25:08
- In some cases, that, you know, people think
25:09
- of that very intensely as actually diagnosing and
25:12
- prescribing autonomously.
25:14
- And in other cases, it's,
25:16
- essentially, as lightweight as really high quality triage
25:19
- that's making sure that you're being navigated to
25:21
- the right type of care.
25:23
- And I think that that is is gonna
25:26
- dramatically change how people understand their health care,
25:29
- and it's gonna change a lot of the
25:30
- decision making. I do think that,
25:33
- you know, the the number of health care
25:35
- decisions today that are actually well informed,
25:37
- is still a fraction of what everyone in
25:39
- the industry would want it to be. So
25:41
- that's one area,
25:42
- as people themselves will just become better stewards
25:44
- of their own care as a result of
25:46
- having having more access to information.
25:49
- The second is,
25:50
- obviously, it's gonna give clinicians a lot of
25:52
- superpowers.
25:53
- You're already seeing it somewhat in primary care
25:55
- where the breadth of what a primary care
25:59
- doctor is able to do is starting to
26:01
- seep more and more into any secondary and
26:03
- tertiary care to some degree,
26:04
- just just sort of on the fringes.
26:07
- And it's because they have access to much
26:09
- better clinical decision support tools. It's because they
26:11
- have access to so much more information,
26:14
- in a very, very short period of time
26:16
- that is very easy to query
26:18
- that clinicians are just gonna be able to
26:20
- to do a lot more in terms of
26:22
- breadth of, you know, diagnosis as well as,
26:26
- the the total number of people that they're
26:28
- actually able to manage. So those would be
26:30
- kind of, like, the two broad strokes,
26:33
- changes at least on the clinical side,
26:35
- that that I think we'll start to see.
26:38
- It seems as though
26:39
- we're starting to see almost a who moved
26:41
- our cheese type of moment
26:44
- where so much of, for example, the GLP
26:46
- one delivery
26:47
- diagnosis
26:48
- prescribing
26:50
- is done outside of the traditional health system,
26:52
- outside of a traditional doctor's office, where just
26:54
- a couple years ago, it was it was
26:56
- being done. And it it seemed like twenty,
26:58
- thirty million people are now getting their diagnosis,
27:01
- getting their drugs that way. The price has
27:03
- gone down.
27:04
- There might be a human in the loop
27:06
- there someplace, but god knows how you find
27:08
- that person who's actually doing the final prescribing.
27:11
- But it but it does seem like so
27:12
- much of this has moved towards almost being
27:14
- automated with a little bit of a person
27:16
- in the loop.
27:17
- Is that where a lot of primary care
27:19
- will end up going over the next couple
27:21
- decades or decade
27:23
- to more and more
27:26
- AI self serve with some help with humans
27:28
- in the loop.
27:29
- When I think I I I I almost
27:31
- think that and I'm curious amount of, like
27:33
- I also think it's important, by the way,
27:35
- not to just think about, like, on this,
27:37
- topic, not just from a US centric view,
27:39
- but actually an international view because I think,
27:41
- you know, we've had these conversations, but it
27:42
- it might even go faster in some of
27:43
- these other markets outside The US. But,
27:46
- Utah allowed
27:48
- for, I think, Doctronic,
27:50
- to have, you know, prescriptions with with an
27:52
- AI service. So maybe the genie is out
27:54
- of the bottle on that one. I also
27:56
- think that
27:57
- while, you know, it's funny, enterprise talks about,
27:59
- like, AI adoption, but meanwhile, all your physicians
28:02
- are using open evidence. All your patients are
28:03
- leveraging, you know, all the latest models to
28:05
- figure out if they can actually understand what's
28:08
- happening to them. So in some ways, like,
28:10
- I would argue you're you're very right, Manav,
28:13
- that organizations that are AI native versus
28:16
- new to it, is a delta. But I
28:18
- think everyone is learning how to be AI
28:20
- native just because of the frustration
28:22
- of being a patient in the American health
28:24
- care economy. I think every,
28:27
- board deck and sales deck starts with the
28:29
- same thing. You know, America costs more than
28:31
- the German GDP and, you know, outcomes suck.
28:34
- But the bottom line is we failed a
28:35
- patient every single day, and I think that
28:37
- that frustration has led to a,
28:40
- an incredible
28:42
- resurgence on these sort of consumer apps,
28:45
- whether it's, you know, for physicians or for
28:47
- patients. And so,
28:48
- I I would argue, like, that that argument's
28:50
- kind of done. Like, it's it's
28:52
- it's happening.
28:53
- True? Isn't that true that so much care
28:55
- is being done more and more directly
28:58
- because patients are frustrated? It's why people go
29:00
- back to the ER. The hospital administrators will
29:02
- say you shouldn't go there, but people go
29:04
- there because it's lowest common denominator.
29:06
- And same thing with getting electronic prescribing and
29:09
- and dealing with things
29:10
- in in this way. Venkat, any comment there?
29:13
- Yeah. I mean, it's funny. I'm I'm looking
29:15
- at, you know, Milenka when he was an
29:17
- investor at, a sixteen z. I feel like
29:19
- the trope back in the day was that
29:21
- and this is in venture, not talking about
29:23
- any firm, but, like, the worst way to
29:25
- make, you know, monetize
29:27
- startups in health care was DTC. You never
29:30
- you know, it's almost like the death kiss.
29:32
- Right?
29:33
- But then now if you look at the
29:35
- investments that have gone direct to consumer, it's
29:37
- like one eighty. And so I think if
29:38
- you follow the money and capital and how
29:41
- some of these businesses are being funded at
29:43
- the scale they're funded, I well, like, you
29:45
- could you could almost comment about the past,
29:46
- but I it feels like we've taken a
29:48
- turn on that. Just to respond to that,
29:50
- yes. I I totally remember that. I think,
29:52
- again, what's changed is
29:54
- the products work now. Like, they work so
29:56
- much better now. Right? I mean, if you
29:58
- look at transcription,
29:59
- right, you've seen, like, a massive uptick in,
30:01
- obviously, the all the transcription campaign companies.
30:04
- Many of those companies were, you know, were
30:06
- around pre l m and and, like, the
30:09
- Right. And they were an okay product.
30:11
- And I think this is, like, this goes
30:13
- to why, you know, people in the past
30:15
- thought, you know, health care adoption was low
30:16
- and that that was, like, kind of a
30:18
- thing. But, actually, it turns out the products
30:21
- were just
30:22
- not good enough yet for the complexity of
30:25
- health care, which is just higher than most
30:26
- industries. And now the AI is good enough
30:29
- to tackle that complexity. And when you have
30:31
- products that are good and deliver great value,
30:34
- people do adopt those at a at a
30:36
- rate that is comparable to other industries because
30:38
- it they they finally deliver meaningfully. And but
30:41
- just to go to the DTC point,
30:44
- yes. Like, you you actually we need the
30:46
- technology to catch up to deliver this type
30:48
- of experience that patients want to use it
30:50
- and feel, you know, safe using it,
30:52
- but it has to happen. Right? Like, you
30:54
- know, the the the the demand for health
30:56
- care is so much higher, so much higher
30:59
- than the supply. My dad lives in Sri
31:00
- Lanka. Right? My dad lives in Sri Lanka,
31:02
- and he spends he's like, when I talk
31:05
- to a doctor and she'll actually say, okay.
31:06
- It's not great there. He's like, when I
31:08
- talk to a doctor, they spend, like, three
31:10
- minute five three to five minutes with me.
31:12
- Right? That that's it. That's all the time
31:13
- they get. And he's like, but chat GBT
31:14
- will, like it's infinite. Right? And so it's
31:18
- just going to happen, and I I think
31:19
- it will happen first outside, The US and
31:21
- then will quickly and it's happening here already
31:23
- just because the demand is so, so much
31:24
- higher than the supply. It has to happen.
31:27
- But but your point is so well taken.
31:29
- Because people can access this in certain ways
31:32
- that don't require them to go through the
31:34
- entire system, and, like, you talk about the
31:36
- three to five minute visit, now I could
31:38
- spend as much time as I want diving
31:40
- into symptoms and ideas with GHT or whatever
31:43
- my my choice is. You could do that
31:45
- and it becomes very easy and you're getting
31:47
- the responsiveness
31:48
- in in real time
31:50
- that you couldn't get before without making an
31:52
- appointment, without getting in, without getting out, without
31:54
- getting tested, everything else. It's fascinating to see
31:56
- that evolution. I'm gonna ask each of you
31:59
- the the following question.
32:01
- And and I'll tell you what's fascinating to
32:02
- me and and Venkat and Manav, you could
32:04
- you could you could comment on this if
32:06
- you want. The audience is listening on audio,
32:08
- not video.
32:10
- But it is amazing
32:11
- how much more polished Malinka looks today than
32:14
- eight to ten years ago when I first
32:16
- met him and how professional mature he looks.
32:20
- I mean, it it it
32:22
- There's there's pre LLM Malinka and post LLM
32:24
- Malinka. So Yes. And and Right. And now
32:26
- he looks so polished and professional. He's built
32:29
- this huge company that delivers 9 figure results
32:31
- for clients.
32:32
- It's really remarkable to watch the the the
32:35
- growth and evolution of both LLMs, AI,
32:38
- and the three of you. The the the
32:40
- question I'll ask, and I'll start with you,
32:41
- Malinka, then I'll go to Venkat and Manav.
32:44
- If you were advising a health system CEO
32:46
- today, and each of you spend a lot
32:48
- of time in this, we've gotten some hint
32:50
- of of your thoughts in this. You'd go
32:51
- big versus small problems.
32:53
- Malika, what is the one AI initiative
32:56
- you would prioritize first and why? And then
32:59
- Manav and Venkat will ask you the same
33:01
- question for our final round of questions.
33:04
- So
33:05
- as a CEO of a WebCycle AI company,
33:07
- I'm sure everybody here will be shocked to
33:09
- hear me say that they should they should,
33:12
- look at the midst the the the revenue
33:14
- cycle as one of the core bits. But
33:16
- more seriously,
33:17
- I say that because
33:19
- when you're trying to get people excited about
33:21
- something new and potentially scary like AI, it
33:24
- is
33:25
- helpful to lead with outcomes that are so
33:28
- bulletproof.
33:29
- Right? And then people you can just build
33:31
- a coalition of support around doing everything else
33:34
- that you want to do. And there are
33:35
- definitely a lot of areas where maybe it
33:36
- feels a little,
33:38
- you know, squishy, like the ROI, and you
33:40
- have to kind of make an argument. But
33:42
- in the revenue cycle in particular and in
33:43
- particular in the mid cycle,
33:45
- there is such a clean and obvious ROI,
33:49
- especially in the very complex stuff that you
33:51
- can start with that. No one should ever
33:54
- be saying something like we we don't see
33:56
- we see AI everywhere except the bottom line.
33:57
- That that shouldn't that should never be on
33:59
- people's minds. And you get people excited to
34:02
- do
34:03
- not just, you know, more in revenue cycle
34:04
- where obviously there's major pain points that need
34:06
- solving, but more everywhere. Right? So lead
34:09
- with complex
34:11
- high ROI areas,
34:12
- and it will get the organization so much
34:14
- more excited to do so much more,
34:17
- over
34:17
- time. Yeah. Can I get you to weigh
34:19
- in or or or Venkat first, whichever one
34:21
- is ready to tee up? And what's the
34:24
- initiative that you'd be advising health system CEOs
34:27
- on first, that they should attack first?
34:30
- I mean, I think,
34:32
- genuinely, I think just taking to Melinka's,
34:35
- point, I think that
34:37
- all every persona inside the c suite views
34:39
- success of AI differently, in my opinion.
34:42
- But, ultimately,
34:43
- I think in a in the current and
34:45
- margin environment we're in,
34:47
- I think,
34:48
- the the ROI piece is to double down
34:50
- on that matters. And in fact, I would
34:53
- argue that any AI solution or initiative
34:57
- that is not outcomes oriented in whatever way,
34:59
- shape, or form is not gonna succeed. Because
35:01
- ultimately, you'd need a clear scorecard of is
35:03
- this thing gonna work or not.
35:05
- I think, you know, if you think about,
35:07
- there's there's n number of areas where you
35:09
- can start, but there's two points I'll just
35:12
- make real quick, which is if if Malenika's
35:14
- gonna go on the revenue side, I'll go
35:15
- on the cost side. I'll take the other
35:16
- other side of the bet, which is I
35:18
- think that we put a lot of lip
35:20
- service to the cost of American health care.
35:22
- But if you look at the functions that
35:25
- underline
35:26
- the health care costs into the health system,
35:29
- these institutions have these functions have not had
35:31
- the same level of investment that the revenue
35:33
- functions have had. I think the danger that
35:35
- I see inside of boardrooms and set of
35:37
- health systems is that there's a lot of
35:39
- investment going and and appropriately so, by the
35:41
- way, in thinking about how to get paid
35:42
- for appropriately for what you do. But I
35:45
- think that making sure that
35:47
- AI, you know, native decisions are made on
35:49
- the cost side, I think, will also be
35:51
- incredibly important for the future.
35:53
- The the other thing I'll sort of say
35:55
- is
35:56
- that
35:57
- the the hardest thing inside a a large
35:59
- enterprise and health system is a classic case
36:01
- for this because in most parts of America,
36:03
- the largest employer is actually the health system.
36:06
- Forget the left hand doesn't know what the
36:08
- right hand sometimes is doing. The left finger,
36:11
- you know, right next to that finger is
36:13
- isolated, siloed.
36:14
- And so I think for me, the opportunity
36:17
- to build
36:19
- this second brain using AI where everybody has
36:22
- shared context
36:23
- like, I think a lot about this where
36:25
- everyone thinks a lot of time about the
36:27
- org chart, the human chart, but there's gonna
36:28
- be an agent chart, and agents talk to
36:31
- agents twenty four seven. And so how do
36:33
- you seamlessly orchestrate the level of intelligence that
36:36
- a Fortune 500 company has when they're negotiating
36:38
- against you?
36:40
- That same level of transparency, enterprise transparency, not
36:43
- price transparency,
36:44
- is gonna be critical in thinking about how
36:46
- do you build this organization in the future.
36:47
- And so for me, that that sort of
36:49
- you know, is is it's not it's more
36:50
- esoteric philosophical thing than building a second brain,
36:53
- but I think it's gonna be really critical
36:54
- in how you culturally change these organizations,
36:57
- to make to make the most optimized decisions.
37:01
- Yeah.
37:02
- Those are both of you had great points.
37:05
- I I think the only thing that I
37:07
- would add is
37:09
- the way that people similar to what I
37:11
- mentioned earlier, the way that people access the
37:13
- health care system is just gonna evolve, right,
37:14
- for the same reasons that
37:16
- Melanca was describing of how people will have
37:18
- endlessly long conversations with something like ChatGPT.
37:22
- People will just have so much more choice
37:24
- in how they choose to engage with the
37:25
- health care system,
37:26
- And I think that that's gonna create
37:29
- all sorts of new
37:30
- competition for how health systems actually manage their
37:33
- patients that they're historically haven't had to deal
37:36
- with, and in in many cases, like, may
37:38
- even be hard to predict what sorts of
37:40
- avenues come up.
37:41
- And as a result, I think ensuring that,
37:44
- health systems are spending a lot of time
37:46
- and focus around
37:49
- using all of this super powerful tooling to
37:51
- make sure that patients are leaving visits feeling
37:54
- empowered, make sure patients actually feel as if
37:56
- they're getting the requisite amount of time that
37:58
- they want with their providers, make sure that
38:00
- they actually feel cared for is is gonna
38:02
- be front and center. And making sure that
38:04
- they actually feel as if they are competing
38:06
- at the experience level
38:07
- rather than just having some sort of service
38:09
- that other people don't and and competing whether
38:11
- it be geographically or,
38:13
- on on the basis of some scarcity.
38:15
- And I think the cautionary tale
38:18
- of what Manav said to if if we
38:20
- were if there's a health system CEO or
38:21
- a board listening to this, I think the
38:23
- the, you know, I think there's a Mark
38:25
- Anderson andresanism
38:27
- on which I think is actually a Jim
38:28
- Clark quote, by the way, from the nineties,
38:30
- that all of businesses like bundling and unbundling.
38:33
- I think we're the risk is we're in
38:35
- a great unbundling cycle right now of given
38:38
- all the trends that and the frustrations of
38:40
- the consumer.
38:41
- And so,
38:42
- for the health system to think about, which
38:44
- has been an incredible,
38:46
- is is,
38:48
- ridden the top wave of bundling for a
38:50
- long time of physicians and all the capabilities
38:52
- and technology,
38:53
- for people to access that if we're in
38:55
- a wave of unbundling, I think the health
38:57
- system business model, you know, is at risk.
38:59
- And so I think that's the fundamental question
39:01
- that every boardroom's gotta ask themselves of how
39:04
- they position themselves the next decade.
39:06
- I I wanna take one more second today.
39:08
- I know we're ready to wrap up, but
39:10
- but I think there's so much truth to
39:12
- that. As I watch some of these kinds
39:14
- of care, and I talk about this in
39:16
- the who moved my chase type comment,
39:18
- move outside of the traditional health system. It's
39:21
- not the disruptors that people feared,
39:24
- such as Walgreens,
39:26
- Walmart, CVS that people were so fearful of
39:28
- a couple years ago.
39:30
- It is all this
39:31
- artificial intelligence and digital interchange
39:35
- and different teletype services.
39:37
- How much concern is there that the traditional
39:39
- health system model
39:41
- changes much more dramatically than expected over the
39:44
- over the next few years? Ben, can can
39:46
- you comment on that just as a as
39:48
- a last comment? And then Melanca or Manav,
39:50
- if you wanna comment, and I'd love to
39:51
- hear your thoughts too. I mean I mean,
39:53
- look. I I I think that that, you
39:54
- know, I I think a lot of health
39:56
- systems have, you know, will argue that, and
39:58
- this is true. If any loved one really
39:59
- was in trouble, they would wanna go to
40:01
- any one of these institutions we've mentioned,
40:04
- on the podcast because of quality and because
40:06
- of the brand,
40:07
- etcetera. But I just think, you know, if
40:09
- you read,
40:10
- you know, a chapter of innovator's dilemma,
40:12
- and how things work, which is it's always
40:15
- death by a thousand cuts. And I think
40:16
- that you know, man, I'm I'm, like, just,
40:18
- Internet memes today, from Silicon Valley, but, like,
40:21
- I'm thinking about the paranoid survive
40:23
- from Andy Grove. And I just think, you
40:24
- know, if you're not
40:26
- if you're
40:27
- not evolving
40:28
- with what is possible and partnering, I think
40:31
- that even in Silicon Valley, if you take
40:34
- some of the companies like Salesforce, they have,
40:36
- arguably avoided the walled garden strategy, which I
40:39
- think they tried. And now it's it's the
40:41
- it's the headless model. Right? So these open
40:44
- ecosystems, I think we give a lot of
40:45
- grief to Epic,
40:47
- you know, for the innovative ecosystem. How do
40:49
- you open that up? And I think of
40:51
- same thing of a health system, which is,
40:52
- you know, I think my push for health
40:53
- system CEOs is what does it mean to
40:55
- be much more partnership friendly and partnership
40:58
- forward,
40:59
- than ever before because of the rate at
41:01
- which these these apps are you know, today,
41:05
- they might look laughable and funny and tiny
41:07
- and,
41:08
- niche, but I think that very much quickly,
41:11
- you know, compounds over time. So I I
41:12
- do think it's a thing you have to
41:14
- worry about, and I don't think enough attention
41:15
- is being paid.
41:16
- Yeah. And I think, just
41:18
- to highlight that, the again,
41:21
- the amount
41:23
- of value creation now for any one of
41:24
- these areas is so much higher than it
41:26
- used to be that people should just forget
41:29
- their priors on these things and, like, be
41:31
- willing to explore an area they may have
41:34
- looked at a couple years ago. Like, the
41:35
- rate of change is so high that the
41:38
- thing now might literally be 10 x better
41:40
- than it was when you last thought about
41:42
- it. So and and if you aren't prepared
41:43
- for that, like, you will be surprised and
41:45
- and disrupted by, you know, someone who is
41:47
- looking into that seriously. But I think just
41:50
- just being aware of the rate of change,
41:52
- which is a very difficult thing for humans
41:54
- to do, I think is is something to
41:55
- be really, really, cognizant of.
41:58
- I I I think it's such a good
41:59
- point. You made a point earlier about how
42:01
- much better the tools are today than they
42:03
- were ten years ago. And your point resonates
42:06
- here as well. I mean, if you I'm
42:08
- much older than you folks, which is embarrassing.
42:10
- But years ago,
42:12
- people would try and dictate with Dragon NaturallySpeaking.
42:16
- And it was so bad twenty years ago
42:19
- that it it made you shy about using
42:22
- it again.
42:23
- And then recently, I do more and more
42:25
- things dictating through the phone in other ways,
42:28
- and the pickup today
42:30
- is brilliant compared to that time ago. But
42:33
- it took a long time for me to
42:34
- restart really using it again because it was
42:37
- so bad some time ago. And I love
42:38
- your point and how quickly so much this
42:40
- is accelerating,
42:41
- and and you you have to be open
42:43
- to it in looking at it or you're
42:44
- gonna miss the moment or fall behind. I
42:47
- I think that's so right.
42:49
- Yeah. Your your your son's an engineer,
42:52
- Scott, and I think of engineering as, like,
42:54
- anybody who's and this is probably why,
42:57
- you know, there's a a an arb of
42:59
- paranoia if you go, you know, in into
43:01
- the West Coast. Because if you follow the
43:04
- arc of engineering software engineering,
43:06
- I actually think
43:08
- it is, you know, depending on your point
43:10
- of view, a canary in the coal mine
43:11
- or or or the doomsday scenario, but I
43:13
- actually think it's a canary in the coal
43:14
- mine of how work has changed. You know,
43:16
- I remember two years ago when someone called
43:18
- me and said, you know,
43:19
- what's the best way for job security they're
43:21
- about to graduate? And I was like, you're
43:22
- fine. You're a comp you know, you're you're
43:23
- a software engineer. You're gonna be good forever.
43:25
- And by the way, for all the automation
43:27
- that's happened, actually, there's more jobs now in
43:29
- demand. We're seeing, so all the job scares
43:31
- is overblown.
43:32
- But, you know, I think at one point,
43:35
- you were writing early code. Next point, you
43:37
- were using cursor, then you're using plot code.
43:39
- It feels like every three days you're reborn.
43:41
- Today, you don't even code because you're telling
43:43
- agents to code and do your work. And
43:44
- And if you just that's just in the
43:45
- last two years. Some of this stuff is
43:47
- in the last three months. And so I
43:49
- just think that if you apply that, to
43:51
- Melinda's point, that rate of change to all
43:53
- jobs, not just one particular type of job,
43:57
- I think that that that order of magnitude,
44:00
- people underestimate
44:01
- the change
44:02
- that's gonna occur in the next decade. I
44:04
- and I I just, you know, implore people
44:06
- to look into the arc of engineering, you
44:08
- know, if you're not, you know, fixated like
44:10
- some of us every day in this world.
44:13
- It it is a fascinating discussion. I I
44:15
- wanna thank
44:17
- Manav, Venkat, and Melinka for joining us. Really
44:20
- a pleasure for me. I hope you three
44:22
- enjoyed it as well. And as importantly, I
44:24
- hope our audience loves it too. Thank you
44:27
- all. You're at the forefront of what's going
44:29
- on in the change in health care
44:31
- and just remarkable
44:32
- thinkers and entrepreneurs and people. Thank you so
44:35
- much for joining us. Thank you, Scott. Thanks,
44:38
- Scott.