Latest / The Advancing Healthcare Innovation Show / E170 | From MySpace to Healthcare AI: Aber Whitcomb, CEO of Salt AI

Transcript

E170 | From MySpace to Healthcare AI: Aber Whitcomb, CEO of Salt AI · The Advancing Healthcare Innovation Show

0:11
Hello and welcome back to the Advancing
0:13
Health Innovation Show. I am your host,
0:15
Michael Stantinos, and if you're new to
0:17
us, welcome. We're super excited to have
0:19
you here. So, we're having some
0:21
incredible conversations with real
0:23
leaders that are painstakingly reshaping
0:25
the future of health. And we're really
0:28
passionate about a few different areas.
0:29
because we're passionate about access.
0:30
We're passionate about innovation,
0:32
leadership, and adoption of those
0:33
innovations. And if it's your first time
0:35
tuning in, let me just sort of take a
0:37
quick step back and tell you just a
0:38
little bit more about what we're trying
0:40
to do here. So, our focus is on bringing
0:42
real stories of innovation to life. Not
0:46
just things that you're reading, but
0:47
actually real people that are really
0:49
doing some incredible things and giving
0:51
them the opportunity to share their
0:53
stories, their wins, their learnings
0:55
with our community and ultimately with
0:57
the world. So, we've got an
0:59
action-packed year of just incredible
1:00
guests. If you haven't checked out any
1:02
of our historical interviews, wherever
1:04
you are, if you're watching this, if
1:05
you're listening to this, wherever you
1:07
are, just hit the subscribe button now
1:09
so that any of the new interviews come
1:12
directly to you. So, today's guest is
1:14
actually one of those rare people who
1:18
doesn't just see where the world is
1:20
going, he builds the rails uh for the
1:23
rest of us to actually get there. So,
1:25
Aber Witcom is the CEO of Salt AI, but
1:30
most people
1:32
know him for something a little wild. He
1:34
co-founded MySpace. He helped architect
1:37
one of the largest social platforms on
1:39
the planet during a time when no one
1:41
actually had that playbook. And instead
1:43
of coasting on that legacy, he went
1:45
right back into the arena. So, Salt AI
1:48
is his latest chapter. You're taking one
1:51
of the most painful problems in
1:52
healthcare and turning it into something
1:54
that actually works. Real-time
1:56
intelligence, precision matching, a
1:59
level of speed that health systems and
2:01
and pharma companies and all the likes
2:03
are using, but they desperately
2:06
desperately need it. And Aber is doing
2:09
some of this with a very similar mixture
2:11
of curiosity, engineering rigor, and
2:14
just sort of a calm focus and intensity
2:16
that's made him one of the most
2:17
respected CTOs of his generation. But
2:20
what really stands out to me is Abra's
2:22
not trying to build the loudest company.
2:24
He's actually trying to build the most
2:26
useful one. And in a moment now where
2:28
healthcare, there's just a lot of noise,
2:30
drowning in hype of of AI. And I think
2:33
people are actually really starving for
2:36
substance and that matters. So we're
2:38
going to kind of go into his journey,
2:40
what he's learned about building things
2:42
at massive scale, how he thinks about
2:45
product velocity, um you know, why he
2:47
jumped into healthcare and you know what
2:49
he's seeing across the ecosystem and
2:51
what he believes AI can actually create
2:54
uh lift instead of actual noise. And
2:56
yes, uh we're we're absolutely going to
2:59
talk about MySpace. How can we not? Um,
3:02
so let's let's jump into it. Aber, great
3:05
to see you. Welcome to the show.
3:07
>> Good to see you, Michael. Great to be
3:09
on. Thanks for having
3:10
>> Yeah, man. Of course. So, before you
3:13
know, big companies and and big stages,
3:16
uh, there was just Aber when you when
3:19
you kind of look back, um, what were
3:22
some of your like early clues that you
3:26
were kind of wired to think about
3:29
problems differently?
3:31
um t talk about some of that from your
3:33
like sort of strategic pattern
3:34
recognition. I'd love to kind of get
3:36
into your mind and how you see and look
3:38
at big problems.
3:39
>> Yeah, it's interesting. I think uh at my
3:43
age, I was one of the last generations
3:47
to go through school without cell
3:49
phones, you know. So, and and really,
3:52
you know, when I was in high school,
3:54
>> that's when PCs were really starting to
3:57
become a thing. Windows was blowing up
3:59
and it and and broad distribution of
4:02
PCs. So,
4:04
>> uh I was I think I was lucky in the
4:06
sense that I was able to get a computer
4:09
probably when I was a sophomore in high
4:11
school and I was dialing up, you know,
4:15
into AOL for internet connectivity and
4:19
and uh
4:20
>> yeah,
4:20
>> it just fascinated me early on that you
4:23
could communicate with other people in
4:25
this way. Before that, the the the most
4:28
amount of technology that I had was like
4:29
a Nintendo, you know.
4:32
>> Wow.
4:34
Yeah, I'm right there with you. I
4:35
remember just being able to dial up on
4:37
America Online as is a very very vivid
4:41
picture of that.
4:43
>> Yeah.
4:43
>> Being on instant messenger. It's just uh
4:47
it's wild.
4:48
>> And it was it was a pretty simple life
4:50
before that, you know. I mean there
4:53
wasn't a lot of technology to engage
4:54
with but once it came then I was just
4:58
hooked you know and and at that point I
5:01
when I decided to go to university it
5:04
was pretty obvious to me to study
5:06
computer science because that's what cap
5:08
you know that that's where my interest
5:10
was um captive. So
5:14
>> now look, you know, having kind of done
5:16
what you've done, I I'm imagining that
5:19
there are some environments that were
5:22
pretty in intense, you know, over the
5:24
years. What's like a a story from those
5:29
early days that kind of taught you about
5:32
pressure and dealing with pressure that
5:34
you still kind of carry with you like
5:37
now?
5:37
>> Oh man. Well, you know, if we jump ahead
5:40
to the MySpace days, we uh started
5:43
MySpace in 2003.
5:46
And in those days, that was pretty much
5:48
before open-source was even really a big
5:52
thing. And
5:53
>> yeah,
5:53
>> so you know, the the framework for for
5:56
building uh large scale infrastructure
5:59
just didn't really exist. There was a
6:01
handful of companies that
6:03
>> that had some technology that was mainly
6:05
built inhouse,
6:07
>> you know, and so when we started
6:09
MySpace, it was just we had to do
6:11
everything ourselves, including hosting
6:13
your own infrastructure, right? So I had
6:16
to have data centers all over the the
6:19
country. So, not only are you building
6:21
software that you know at a at a global
6:25
scale, but you also need to manage all
6:27
of your your servers. And I remember
6:31
being,
6:34
you know, sitting at my desk at MySpace
6:36
and some massive array of discs just
6:41
completely fails, you know, and
6:44
>> you're you're sitting there like, what
6:46
do I do? you know, I everyone's
6:48
everyone's looking at me, you know, I I
6:50
to try to figure this out.
6:53
>> Otherwise, we're down and the business
6:55
is gone if I don't figure this out.
6:57
>> So, obviously, that's a huge amount of
6:59
pressure. And I there's nothing you
7:01
could do except get up, get in your car,
7:03
and drive down to the data center and
7:05
start figuring out what's going on,
7:08
calling all the experts that I know,
7:10
replacing hardware to try to get the
7:12
site back on. And that happened dozens
7:15
of times at at at MySpace. And you know,
7:18
it's I guess it's a it's a tale of how
7:21
things have changed so much because
7:23
these days you just use the public
7:25
cloud. You have to architect things
7:27
properly, but you don't have that
7:29
problem. If something fails, you just
7:32
>> you know, you spin up a new uh EC2
7:35
machine on Amazon and you don't even
7:38
think about it.
7:39
>> Wow. So, do you do you kind of have like
7:41
now is that sort of etched into your
7:43
mind? Are there I'm just thinking about
7:46
the behind thescenes story of that?
7:50
>> It's it's serious trauma. Yeah. I mean,
7:52
it's that's trauma. That's
7:54
>> PTSD, man.
7:55
>> Yeah.
7:57
uh you know I and it's even happened to
7:59
me like when I was maybe out on a date
8:01
you know with with a girl and my phone's
8:03
blowing up that the sight's down and I
8:06
had to leave and and go maybe drive down
8:09
to the data center again. So yeah, I I
8:12
got used to that pressure. I got I got
8:14
used to um effectively being on call and
8:19
my my baby being, you know, keeping the
8:22
site up and keeping, you know, that that
8:25
kind of 247 consistency for for all of
8:28
our users.
8:30
>> I'm curious when when MySpace started
8:33
growing at I mean it was a rocket ship,
8:36
right? Like at a pace that no one had
8:38
ever seen.
8:40
What did that uh feel like for you, you
8:44
know, on the inside? What was like the
8:46
dayto-day
8:47
energy uh of running and being in that
8:50
kind of a fast-paced environment like?
8:53
>> Yeah, it's um it's tough. It's only for
8:56
for certain types of people. I I think
8:59
MySpace was probably the first
9:02
internet property that had nonlinear
9:04
growth. And so it it was unheard of that
9:08
you not only would grow in in a straight
9:11
line, but that that straight line would
9:13
keep getting steeper and steeper and
9:15
steeper every day. So So
9:18
uh you you know any type of projections
9:22
that we were making, we would try to you
9:25
know maybe start with a straight line or
9:27
maybe start with some type of you know
9:30
parabola that would that would grow. But
9:32
it was we we underestimated every time
9:35
because we you you just that that viral
9:39
that virality is so powerful, right?
9:42
You've got the next layer of people that
9:44
are then then inviting a bigger layer of
9:47
people. And so yeah, it's it's an insane
9:51
amount of pressure. I mean that I I look
9:53
back on those days fondly because
9:56
everyone had to work together perfectly
10:00
and
10:01
>> if if we if we weren't working all the
10:05
time and and passionate about what we
10:09
were doing, it it just wouldn't wouldn't
10:11
have worked. Um, I mean, luckily it was
10:14
easy to be passionate about it because
10:16
nobody had ever done anything like this
10:17
before or seen anything grow to this
10:20
scale and you could really tell that you
10:22
were making an impact on on the world by
10:26
building something.
10:26
>> Yeah. I mean, literally, you were like,
10:29
I'm just looking at it from my from a
10:31
user standpoint. It's like I was
10:32
watching seeing this unfold and I'm
10:35
like, dude, the the world is
10:36
fundamentally different and it's like
10:39
watching it in real time. It was pretty
10:42
gnarly.
10:43
>> Yeah, it really it really was. And yeah,
10:47
and it was all over the news. I mean, I
10:50
I could go into a coffee shop just to
10:53
order a cup of coffee and and there
10:55
would be a guy sitting down browsing
10:57
MySpace and you know, when you start
10:59
seeing that out in the wild, you know,
11:02
something special is happening there.
11:05
>> Yeah. So, I mean, if you had to go back
11:06
and sort of maybe even just sort of
11:08
teach a class on building for scale, you
11:12
know, before you're ready, what's like
11:14
the first thing you'd tell everyone in
11:17
the room?
11:20
>> Yeah. I mean, if if you're building
11:22
something for for scale, the you have to
11:25
you have to think beyond
11:27
the the number of users that you have
11:30
today, obviously, and and and what even
11:33
what you think is is going to be in a
11:36
year. You have to have um you have to
11:39
think in terms of compute units and unit
11:42
economics, right? if if
11:45
>> if I this amount of resources for this
11:48
amount of users and that has to scale
11:50
horizontally forever and and you can you
11:55
can build a product very very quickly
11:58
these days, but to architect something
12:01
that actually scales is probably 10x the
12:05
time, right? And so a lot of that work
12:06
is under the hood
12:09
>> and
12:10
you know it's you don't you don't get a
12:13
lot of recognition for it too. You know
12:16
it's it's you know people see what they
12:18
see on the front end but most of the the
12:21
the intellectual property is is actually
12:24
on down below the waterline. You know,
12:27
it's like a you see a nice pretty duck
12:29
floating above the water line and then
12:31
you look under and they're
12:33
>> they're just paddling as hard as they
12:34
can under there. That's kind of what
12:36
it's like.
12:38
>> Yeah. And I'm just sort of thinking
12:40
because you've you've built in a number
12:41
of different industries. What
12:45
what what gravitated you towards
12:47
healthcare? You know, um to to me it
12:49
might be, you know, the toughest arena
12:51
of all, but I'm just sort of curious
12:53
what
12:54
>> what
12:56
Yeah. Starting with what you said, I I I
12:57
like challenges and I I do think this is
13:02
uh one of the toughest industries to to
13:05
break into.
13:06
>> Yeah.
13:06
>> Um but it's it's one of the most
13:09
important, right? This is uh this
13:11
impacts our health and I think if you
13:14
ask anyone that's thinking rationally
13:16
what their number one priority is is is
13:19
pro it should be their health, right? I
13:21
mean so
13:22
>> absolutely
13:22
>> you know and because if you don't have
13:24
your health you can't take care of your
13:26
family you can't make money you can't do
13:28
whatever you need to do so um I I like
13:32
that challenge I also see what what is
13:36
happening with AI so AI is being built
13:40
by hardcore engineers um people that
13:43
that code and so naturally
13:48
a lot of that that work is going into
13:51
things like coding, right? Like like
13:54
let's build a model that's really good
13:55
at coding because that's what these
13:57
engineers know.
13:59
>> And so that makes sense and and it is a
14:02
you know it is an incredible use case.
14:05
But um you know healthcare and life
14:07
sciences is is a tick behind that that
14:10
innovation. They're not benefiting as
14:12
much from that. So I I believe that
14:15
bringing some of this really hardcore AI
14:20
engineering into healthcare and life
14:22
sciences is going to make a huge impact
14:25
on on the space.
14:27
>> Um and you know I've had the benefit of
14:30
working with like Dr. Agus at the
14:32
Ellison Medical Institute and he's
14:34
taught me a lot about oncology and and
14:36
and health care and and we're working
14:39
closely with them on a prostate cancer
14:41
therapy. So, you know, having kind of
14:44
mentors like that helped me break into
14:46
this space has has um gotten us a lot
14:50
further.
14:51
>> Yeah, the relationships matter.
14:52
Healthcare is definitely a relationship
14:54
sport, you know, it's it's uh when I
14:57
don't even want to say outsiders, but
14:59
you know, when you have folks that
15:01
aren't native to healthcare, they don't
15:03
know the environment as well,
15:05
>> they're viewed fundamentally
15:06
differently. But when you have someone
15:07
that can lend their credibility, their
15:12
their reputation like, hey, you know,
15:14
and and bring you into those types of
15:16
conversations, it just it changes the
15:19
trajectory.
15:21
I'm just sort of curious, you having
15:22
come from big tech, what was the moment
15:26
where you said,
15:28
okay, you know, healthc care works very
15:32
differently than I expected.
15:35
>> Yeah. Uh
15:37
so I had a yeah I had a very big culture
15:41
shock when we started bringing our
15:44
technology into pharma and these are you
15:47
know top
15:48
>> 20 pharma companies
15:50
>> and they're they're very much interested
15:51
they they like the technology and they
15:54
want to move forward. So once you start
15:56
moving forward with them, you realize
15:59
that wow, this sales cycle could
16:01
actually take a year, you know, and and
16:05
I'm not I'm not used to moving at that
16:07
that pace. I'm you know, I'm an
16:09
entrepreneur.
16:10
>> What's your what's your norm? What's
16:11
your norm? You all right, let's move
16:13
fast. Like what's break neck speed for
16:14
you?
16:16
>> Well, I mean, I I I like to make
16:19
progress every single day. And and you
16:21
know, I would it's like, okay, you like
16:23
it, then let's buy it. let's close the
16:25
deal right now, you know. So, uh um you
16:30
know that they just need they didn't
16:31
need a lot of time to get used to
16:33
something. But, you know, they're if you
16:35
look at the the this industry, there's a
16:39
lot of compliance involved. There's a
16:41
lot of bureaucracy involved. They're
16:43
they're making an incredible amount of
16:46
money. So, they have more to lose and
16:48
more to protect. Um and that's also why
16:53
what what also motivates me too is
16:55
because I think if we can crack this
16:58
industry
16:59
>> Yeah.
17:00
>> and and make kind of a meaningful impact
17:03
on it. I think then we can say okay
17:05
we've solved these problems for one of
17:07
the hardest industries. Maybe let's
17:08
start moving into some other industries
17:10
as well
17:12
>> since uh and and so we're we're starting
17:14
to think about that.
17:16
>> Nice. So because for most that are
17:20
listening to this maybe this this might
17:22
be the first time that they've heard
17:24
about salt. What is sort of the main
17:27
core problem that you're solving that
17:32
keeps many of these
17:34
uh either care delivery leaders and or
17:38
you know life sciences leaders up at
17:41
night.
17:42
>> Yeah. So, Salt is a you can think of it
17:45
as a kind of complete AI operating
17:47
system for for the enterprise. And what
17:50
that means is
17:51
>> we can connect securely to many data
17:56
sources within an organization. And
17:58
these these types of pharma companies
18:00
for example, they have huge amounts of
18:03
proprietary data.
18:04
>> Yes. and they want to connect that data
18:06
to models and and we provide the
18:09
framework for doing that in this visual
18:12
first environment but also providing
18:14
full access to the the code and the
18:16
infrastructure
18:18
and because of that that kind of
18:22
information security requirement you
18:25
have to you have to architect a product
18:28
like Salt a little differently. It's
18:30
it's not not necessarily a SAS tool.
18:33
What we need to do is we need to uh
18:36
install salt deep within their
18:39
infrastructure to get close to the data
18:42
and make sure that that data is not
18:44
leaving their firewall
18:46
>> and can connect to many kind of life
18:49
science-based models, not just LMS but
18:52
many models that have been trained on
18:55
evolution and biology. Um, and so
18:59
providing that kind of very secure
19:03
infrastructure for the enterprise is
19:05
really what what Salt is about.
19:09
>> And and you're you're known for like
19:12
product velocity, you know, what what is
19:15
what does that look like inside of Salt?
19:19
I mean, how do you build a culture that
19:21
can
19:22
ship a product fast without breaking
19:26
trust given that in healthcare trust is
19:29
a really really big thing?
19:32
>> Yeah, absolutely. I I think uh you know
19:35
one of the advantages that I have being
19:37
around for so long is that I can put
19:40
together a team of people that I've I've
19:43
worked with before and I you know I can
19:45
I can cherrypick
19:47
all the best people from from my past
19:49
and that's really what we've done at
19:51
Salt is is built
19:53
>> that team of of you know A-list
19:57
engineers and and go to market
19:59
specialists. So that's really where it
20:01
starts, right? Having the best team
20:03
because if if you're developing this
20:06
type of software for something that's so
20:08
important, it's got to be good. It just,
20:10
you know, it has to be done by
20:12
experienced engineers. You that doesn't
20:15
mean you can't move fast. You can still
20:16
move fast. It just has to be very
20:19
quality code. And we use a lot of AI
20:21
tools to ensure that. I mean, we're not
20:23
we're not using AI to to write all of
20:25
our code, but you can use AI to write a
20:28
lot things like testing harnesses and
20:30
unit tests and make sure that the
20:33
quality is extremely high.
20:35
>> And so when we go and install, we can
20:37
have confidence that things are working
20:39
as expected.
20:41
>> But that being said, it's sorry, that
20:44
being said, it's also pretty hightouch,
20:46
right? when we're we're we've got four
20:48
deployed engineers. Well, we'll we'll
20:50
send engineers on site if if we have to
20:52
to make sure that it works.
20:54
>> Yeah. Because I'm just thinking like,
20:55
you know, you've you you've coached
20:56
you've mentored a lot of brilliant
20:58
engineers.
21:00
Um I'm just think like what like what's
21:03
the difference between someone who's
21:05
like really good versus someone who's
21:08
like, man, this person is special?
21:13
>> Yeah. I mean, you know, have you heard
21:16
about the 10xer? You know, like they're
21:18
they're they're like these unicorn
21:20
engineers that somehow have the ability
21:23
to to get done. They they literally do
21:26
10 times more work than your your
21:29
average engineer. And they're they're
21:30
very hard to find. Um they're not always
21:33
motivated by by money. They're motivated
21:36
by uh solving extremely tough problems.
21:40
And it's it's just another gear that
21:43
some people have to to not only be
21:47
creative with with what they're doing,
21:50
but also when when something's done, it
21:52
actually works as expected. And I I just
21:56
try to I I try to amass as many of those
22:00
types of people as possible without
22:03
compromising and and saying, "Oh, well,
22:05
we need some engineers. We're just going
22:07
to hire these guys right here." That's
22:09
>> right. That's not how it works, you
22:10
know, you really have to find that type
22:12
of talent.
22:14
>> Yeah. And when you sort of zoom out, you
22:17
know, look beyond the curve. I'm just
22:20
going to pick, you know, five years out,
22:22
what what is
22:24
what is success look like for Salt and
22:27
and not like, you know, what you're
22:28
telling your investors, but like the
22:30
actual leader, the human version of the
22:33
purpose of what you're building. What
22:35
does that look like?
22:36
>> Yeah. I mean, you know, the the use
22:39
cases for our customers,
22:43
um, which are things like discovering
22:45
denovo molecules, you know, proteins
22:48
that have never been seen before in
22:50
nature.
22:51
>> Um, you know, use cases on maybe
22:57
creating a a cancer therapy. You know,
23:00
my my goal is to make them successful at
23:03
at what they do and compress these
23:05
timelines that have become the the norm
23:09
and create a a new norm which might be
23:12
10 times faster with with the ultimate
23:14
goal of curing disease, I think, is is
23:18
the the that's the trajectory that we're
23:20
on right now.
23:22
>> Yeah, that's that's the bet that you're
23:23
making in my mind. It's like, hey, we
23:25
feel like we can cure these diseases
23:28
better, faster, more efficiently with a
23:30
higher degree of precision using using
23:33
our technology. Like that that that's
23:36
fundamentally when in the research I was
23:37
like, man, this is this is gamechanging.
23:41
>> Are there any
23:43
frustration points of frustration where
23:46
you see it so clearly and yet maybe
23:48
you're around others that are in the
23:50
industry that don't see the vision that
23:52
you see? Where do you How do you sort of
23:55
deal with that just knowing that you can
23:56
see so far ahead and others are just
23:58
they're just not there yet?
24:01
>> Yeah, there's there's skeptics um you
24:04
know and there's people that have been
24:07
designing proteins for 30 years for
24:10
example and they see something that
24:12
comes out of AI and they say well that's
24:14
that that's never going to work. Um
24:18
which is great. I I think that helps you
24:20
you you need those types of people and
24:23
that constantly kind of maybe raise the
24:26
bar and move the requirements uh higher
24:28
and higher. Um so it's it doesn't really
24:33
frustrate me. You know, I I I embrace
24:36
that type of doubt. Um you know, I don't
24:40
know exactly what's going to happen with
24:41
AI. Nobody does. Um, but I'm very
24:44
confident that this is a powerful
24:46
technology that is changing the paradigm
24:49
here.
24:50
>> Yeah. Yeah. Is that I mean I'm just sort
24:52
of curious what uh is that sort of the
24:55
hill you're willing to die on when it
24:57
comes to AI like hey that like no one
24:58
knows what's going to happen in the
25:00
future and we're just kind of we're
25:03
riding the wave just like everyone else
25:04
is.
25:06
>> Yeah. I I think uh you know the the AI
25:09
singularity is is real and it is very
25:12
difficult to predict what's going to
25:15
happen. Um I mean I'm surprised every
25:19
day about what's happen what's
25:20
happening. I mean you've we're now
25:22
talking about data centers in space um
25:25
which I don't think anyone would have
25:27
even predicted like a year ago, right?
25:30
>> Yeah. No, for sure.
25:32
It's uh I'm I'm excited about it and I'm
25:35
I'm really embracing it and I I like to
25:38
I like to live in the present as much as
25:40
possible or in the near future. It's
25:43
very clear to us what we need to build
25:45
and and build quickly and and then we'll
25:48
take the next phase and the next phase
25:50
from there. But you know making I think
25:53
making long range predictions is uh is
25:57
is difficult right now.
25:59
>> Yeah. So when you when you think about
26:01
the level of access that you have now
26:04
seeing kind of behind the curtain
26:06
especially on the life sciences even on
26:07
the care delivery side is there
26:10
something that sticks out in your mind
26:12
that you feel like and this is this
26:15
should absolutely never be automated and
26:18
what what is that one thing
26:25
>> you know I I think uh drugs are a
26:30
slippery slope, right? I I think uh you
26:33
can you can kind of see this future
26:37
where AI can create some type of
26:42
compound
26:43
so quickly that could maybe even be
26:46
customized for every individual. And in
26:50
that case, you might not need to even go
26:52
through the FDA, right? Somebody can
26:54
take whatever they want personally. Um
26:56
so
26:58
you know I that that is I think that's a
27:02
pretty interesting
27:05
uh dilemma you know where where we could
27:09
we AI is so powerful that it can create
27:11
almost any drug and can make you know
27:14
pills and therapy almost uh ubiquitous
27:16
for everyone. I don't know it seems like
27:19
it could be a good thing but it also it
27:21
could be a bad thing in the future too.
27:24
>> Yeah. Yeah. Yeah. Even just thinking
27:27
about how
27:29
there are other innovators that are out
27:31
there, I guess. Is there sort of a a
27:35
message for those that are tuning in
27:37
that, you know, maybe they can take with
27:38
them as they they're also trying to
27:40
build something that matters? Um is
27:42
there a message that you you want to
27:44
share with them?
27:46
Yeah, I mean I I think uh believing in
27:48
in your vision there there's going to be
27:51
times where you doubt your vision and
27:55
maybe are are ready to give up. So when
27:59
you're in that state, know that that's
28:01
completely normal and and the people
28:04
that win are the people that break
28:05
through
28:07
the the you know negative thoughts like
28:08
that or when it feels like everything's
28:10
against you.
28:12
>> Yeah, absolutely. Before we wrap, I want
28:15
to make sure that you at least kind of
28:16
give like a commercial segment. Where
28:18
can people go to follow all the work um
28:22
around Salt and if there's anything new
28:24
that's coming out that you kind of want
28:26
to maybe broadcast and share like the
28:28
the floor is yours.
28:30
>> Yeah, thank you. Yeah, if you go to uh
28:32
salt.ai, you can take a look at the at
28:36
everything that we're doing. Um you can
28:38
you can book a demo directly from the
28:40
website and we'd be happy to uh show you
28:42
Salt.
28:44
Um, we're also we'll be coming out with
28:48
something called Salt OS, which will be
28:53
um kind of that that that complete
28:54
operating system that I was talking
28:57
about. And
28:58
>> yeah,
28:59
>> maybe a little bit more self-service,
29:02
right? Where right today when we engage,
29:04
we engage with, you know, big big
29:06
companies and we make sure that they're
29:07
successful with a lot of effort. But the
29:09
the our product is getting good enough
29:11
to where I think we can um we can have
29:14
more people actively engaging in it.
29:18
>> That's really really exciting. When you
29:20
say more people actively engaging in it,
29:21
are you thinking
29:23
>> smaller enterprises? Are you thinking
29:25
like physician practices? like how small
29:28
are you thinking or how
29:30
>> any size anyone that values their their
29:34
data their proprietary data and um wants
29:38
to enable AI for themselves right that's
29:42
and that's really the principle that we
29:44
have you know if you've got some
29:46
information that uh powers your business
29:49
you probably shouldn't be sharing that
29:52
with these big model developers right
29:54
you should be owning the solutions
29:56
yourself and that's what salt does.
30:00
>> Amazing. Yeah, this has been a lot of
30:02
fun just even
30:04
just even seeing how you think about
30:06
problems, how you um think about
30:09
pressure building in scale when you know
30:12
when you mentioned about having data
30:14
centers in space that that was actually
30:15
the first time that I heard that. Uh so
30:17
I'm I'm kind of like wow that's actually
30:20
that's actually really interesting.
30:22
>> Yeah, it's fascinating. Yeah. And I I
30:25
think Elon has figured it out. So, uh,
30:28
stay tuned.
30:29
>> Yeah.
30:30
>> Yeah. I I Yeah. So, I mean, you've built
30:32
at a level that I think most people
30:34
never get close to. Um, so it's just
30:39
sort of really cool to see how you're
30:41
thinking about how you see the world and
30:44
how you're approaching things that that
30:46
make it feel like it's actually
30:47
possible.
30:49
Um, I just think what you and the team
30:51
at Salt are doing, it's it's actually
30:54
really it's really clever. And, you
30:57
know, for everyone that's tuning in
30:58
right now, if if there's something that
31:00
you, you know, are going to take from
31:02
this, I think it's pay close attention
31:05
to the to the problems that actually
31:07
matter, you know, and and move towards
31:09
work that actually really stretches you.
31:12
when you when you spoke about a about
31:14
like hey
31:16
actually cultivate and find like that
31:18
world class talent that 10x talent think
31:20
about what that does um and don't
31:23
underestimate what can happen you know
31:25
when you sort of mix in a bit of
31:27
curiosity and execution when they come
31:29
together in the same room and and
31:30
obviously having the right relationships
31:33
in the right context you know with the
31:34
right alignment that can create a lot of
31:36
momentum appreciate you bringing your
31:38
story here and just being very
31:40
rigorously honest and for all those that
31:42
have been tuning in, appreciate uh you
31:44
just wanting us to sort of explore and
31:46
go deeper. So this is a platform that is
31:49
about community. It's about, you know,
31:52
actually making things happen. So thanks
31:54
for listening. Keep building with
31:56
purpose and uh we'll see you next time
31:59
on the next episode of the Advancing
32:00
Health Innovation Show.