Latest / Vanguards of Healthcare by Bloomberg Intelligence / Butterfly Leads the Drive on Semiconductor to Make the Most Out of Ultrasound

Transcript

Butterfly Leads the Drive on Semiconductor to Make the Most Out of Ultrasound · Vanguards of Healthcare by Bloomberg Intelligence

0:21
Speaker 1: Welcome to another episode of the Vanguards of Healthcare series.
0:24
Speaker 1: My name is Matt Hendrickson, the medical technology analyst at
0:27
Speaker 1: Bloomberg Intelligence, which is the in house equity research platform
0:30
Speaker 1: of Bloomberg LP. We are pleased to have with us
0:33
Speaker 1: today Joe Devivo, CEO of Butterfly Network, a medical device
0:38
Speaker 1: company that is pioneering portable ultrasound technology. You can dive
0:42
Speaker 1: deeper into the financials by typing in b f l
0:46
Speaker 1: Y equity go on your Bloomberg terminal. Joe, thank you
0:49
Speaker 1: for joining us today.
0:50
Speaker 2: Thank you for having me us. Really great to be here.
0:52
Speaker 1: And I know I'm just dying to jump into this
0:55
Speaker 1: mid journey press release, but I gotta start with, you know,
0:59
Speaker 1: just grounding the listeners with you know, the technology you
1:03
Speaker 1: joined or you know you joined us for an episode
1:06
Speaker 1: two years ago, we were just talking about the IQ
1:09
Speaker 1: three launch. Why don't we just start with how that
1:15
Speaker 1: launch is going, What exactly is IQ three? And I
1:18
Speaker 1: clearly remember from last episode the story about you know,
1:22
Speaker 1: you and your son waiting for an ultrasound appointment. I
1:25
Speaker 1: think that's such a powerful story just to kind of,
1:27
Speaker 1: you know, for the listeners to hear and how this
1:30
Speaker 1: kind of technology is helping that shortcoming.
1:34
Speaker 2: Well, absolutely, thanks for the thanks for that. Yeah, one
1:38
Speaker 2: day we had a call from a visit from the pediatrician,
1:44
Speaker 2: and a pediatrician after seeing my son said you need
1:48
Speaker 2: to go to the emergency room. And that is not
1:50
Speaker 2: anything a parent wants to hear. So we were quite concerned.
1:53
Speaker 2: We went and picked up my son. We went to
1:57
Speaker 2: the emergency room, took us half an hour to get
1:59
Speaker 2: into a room. The attending who came and visited said,
2:05
Speaker 2: you know, I think I know what this is, but
2:09
Speaker 2: we need to get a confirmatory ultrasound, And so we
2:12
Speaker 2: sat there in the emergency room for three hours to
2:15
Speaker 2: wait for radiology to open up. He was wheelchaired over,
2:19
Speaker 2: had a wonderful scan with a great ultrasnographer. When we
2:22
Speaker 2: came back and the diagnosis was exactly what the doctor
2:28
Speaker 2: had thought. And I had even said to him, Hey,
2:30
Speaker 2: I'm CEO, Butterfly, why don't you have your own probe
2:34
Speaker 2: And he said, you know, help us, we don't have
2:36
Speaker 2: that program here yet. But we learned that it was
2:39
Speaker 2: what he thought it was. If they had a device,
2:43
Speaker 2: they would have been able to make a diagnosis right
2:45
Speaker 2: then and there and we would have avoided that three
2:48
Speaker 2: hour wait time and actually, if our pediatrician had a device,
2:52
Speaker 2: we would have avoided this whole episode and cost of
2:54
Speaker 2: the healthcare system. So what we do is develop point
2:59
Speaker 2: of care ultra sound device using a semiconductor which a
3:02
Speaker 2: doctor can see the entire body with one device, identify
3:07
Speaker 2: the preset and then go And it's low cost and
3:11
Speaker 2: our vision is that every doctor and every nurse in
3:14
Speaker 2: the world will have their own device.
3:16
Speaker 1: And so the current version you have at is the
3:18
Speaker 1: IQ three that was your two years into the launch.
3:22
Speaker 1: How do we think about the first how is the
3:25
Speaker 1: launch going? And the second what were kind of the
3:28
Speaker 1: key feature changes from the IQ three from the previous generation.
3:33
Speaker 2: Right, So, Butterfly is a semiconductor based ultrasound company, the
3:37
Speaker 2: only company in the world that successfully has commercialized a
3:42
Speaker 2: semiconductor that can deliver ultrasound. And just like in any
3:46
Speaker 2: semiconductor development, there's more's law and there is that you
3:50
Speaker 2: can double your power every eighteen months and that's been
3:53
Speaker 2: true for the last sixty years. So we are on
3:57
Speaker 2: that journey and our journey is very similar to digital photography.
4:00
Speaker 2: You know, CODAC came out with the first digital camera.
4:02
Speaker 2: In ninety five, there's a one megapixel camera. Everyone thought
4:06
Speaker 2: the idea of digital photographer was fantastic, but they famously
4:09
Speaker 2: shut the program down two years later because image quality
4:13
Speaker 2: wasn't good enough. Now every other great imaging company said,
4:16
Speaker 2: wait a minute, this is a great idea. You just
4:19
Speaker 2: have to stay the course with your with your processors.
4:23
Speaker 2: And one megapixel turned into three, five, seven megapixel, and
4:28
Speaker 2: in two thousand and three digital photography they sold more
4:31
Speaker 2: digital cameras than they did film cameras, and that was
4:34
Speaker 2: this massive revolution of digital. Butterfly is the digital to
4:40
Speaker 2: the analog and ultracund we IQ three is our third
4:45
Speaker 2: generation chip and it has double the processing power of
4:49
Speaker 2: our second generation chip. So our image quality has gone
4:53
Speaker 2: from a wellness type of image quality to an image
4:57
Speaker 2: quality where you know, a deep clinical decision can be
4:59
Speaker 2: made off of it now and we're equivalent to all
5:02
Speaker 2: the other devices on the market. So it's it's a
5:05
Speaker 2: major major event, and we're not done there. Where we are.
5:09
Speaker 2: We've sent our fourth generation chip to TSMC to be manufacturing,
5:14
Speaker 2: and we're now developing our fifth generation so that the
5:18
Speaker 2: future is very bright. And just to continue on that
5:21
Speaker 2: that photography story. You know, Apple just launched an iPhone
5:24
Speaker 2: with forty eight megapixel image. Yeah, so it's it's this
5:28
Speaker 2: inevitability that as each generation of chip comes out, you're
5:32
Speaker 2: going to get better and better and more powerful imaging.
5:35
Speaker 1: And let me just ask because you know, I'm thinking
5:37
Speaker 1: of you know, traditional ultrasound. Like when my wife was
5:40
Speaker 1: pregnant with the girls, we went in the you know,
5:42
Speaker 1: the big machine is the cord and then you have
5:45
Speaker 1: the handheld and you're talking about that's the analog. But
5:48
Speaker 1: there's also i know gehealthcare is in the past highlighted
5:51
Speaker 1: a portable ultrasound from theirs. Is that a semiconductor or
5:56
Speaker 1: is that the analog technology? And what is that kind
5:59
Speaker 1: of technology utilizing?
6:02
Speaker 2: Right, So we all are connected to computers and so
6:06
Speaker 2: we all process the images digitally, but the acquisition of
6:10
Speaker 2: the image and the dynamics of the acquisition are what's different.
6:15
Speaker 2: The legacy ultrasound uses a piezoelectric crystal, and it's they
6:20
Speaker 2: learned a long time ago that if you put energy
6:22
Speaker 2: through a crystal, it'll vibrate, and if you cut the
6:24
Speaker 2: crystal to a certain dimension, it will vibrate to a
6:27
Speaker 2: certain type of array, and they learn that each array
6:31
Speaker 2: is beneficial for different types of imaging through the body.
6:34
Speaker 2: So if you look at a standard ultrasum machine that
6:36
Speaker 2: will be in that ob suite, they'll have about four
6:40
Speaker 2: different handles, and each handle will have a cut crystal
6:43
Speaker 2: lens for different parts of the body and imaging. The
6:47
Speaker 2: difference for us is we have a semiconductor with nine
6:49
Speaker 2: thousand sensors in a rectangle almost like a TV screen,
6:53
Speaker 2: and we can fire any one of those sensors in
6:56
Speaker 2: any protocol that we wish. You only need one device
7:01
Speaker 2: in order to meet what those four devices do. So
7:05
Speaker 2: what those companies did was basically cut the cord on
7:09
Speaker 2: each of those probes and then put a battery on it.
7:12
Speaker 2: And it is that analog way of sending out the
7:17
Speaker 2: vibrations for the crystal. And the reason why I say
7:19
Speaker 2: it's analog is because they can't change it. Yeah, I
7:23
Speaker 2: was about saying, it's like, once this crystal is shaped,
7:25
Speaker 2: you can't that's shade. It's like, you know, if you
7:28
Speaker 2: had a lens for a near you know, a near
7:30
Speaker 2: sighted lens for a camera that didn't have all focus.
7:33
Speaker 2: No matter what you do digitally, no matter how you
7:36
Speaker 2: process the signal. When you go and acquire the image,
7:39
Speaker 2: it's only going to ever see near because you're going
7:42
Speaker 2: through a glass lens that's formed for that near. So
7:45
Speaker 2: the difference for us is we can change the settings
7:49
Speaker 2: on the same probe. So you say, I want to
7:51
Speaker 2: see the heart that I'm going to use this array.
7:53
Speaker 2: I want to see the abdomen, I'll use this array.
7:55
Speaker 2: And with every preset that's identified, uh, the chip dynamically
8:00
Speaker 2: changes it.
8:00
Speaker 1: Yeah, and so you know, this seems like a lot
8:03
Speaker 1: of directions the launch can't go. I mean we've already
8:06
Speaker 1: talked about, you know, in your anecdol story about the
8:10
Speaker 1: primary care physician could have benefited if he or she had,
8:15
Speaker 1: you know, a Butterfly IQ three.
8:16
Speaker 2: Well, we certainly would have benefited.
8:18
Speaker 1: Yeah, this saved us all that all that panic. Absolutely,
8:21
Speaker 1: But then you know you have the hospitals. I mean,
8:23
Speaker 1: my wife and I watch the pit episodes. Yeah, and
8:26
Speaker 1: I was like pointed like, hey, hey, that's the butterfly
8:28
Speaker 1: right there, and yeah. So and then I went through
8:32
Speaker 1: the whole story with her about what it's doing. And
8:35
Speaker 1: but there's so many different directions you can initially have
8:38
Speaker 1: the launch going through. How are you going about that.
8:41
Speaker 1: Are you starting with the primary care physicians, are you
8:43
Speaker 1: going with the hospitals that already have the traditional ultra sound?
8:47
Speaker 1: Which direction have you gone in those first couple of years.
8:51
Speaker 2: So the pointy care ultra sound originated in the emergency
8:55
Speaker 2: room because you imagine, you know, an emergency room doctor
8:59
Speaker 2: is going to encounter so many different types of things.
9:01
Speaker 2: Right if you're a cardiologist and you're just going to
9:04
Speaker 2: do a cardiac echo every day, then you know, then
9:08
Speaker 2: that's your hammer, and then there's you know, a specific nail.
9:12
Speaker 2: But when you're in the emergency room, you have to
9:14
Speaker 2: you know, you might see a bone fracture, you might
9:16
Speaker 2: see fluid in the lungs, you might see in a
9:19
Speaker 2: regular heartbeat, you might see a hematoma. So you want
9:23
Speaker 2: to have one device that can have that ubiquity and
9:26
Speaker 2: the mobility. You don't have to run to a room
9:28
Speaker 2: to find an ultrasound cart. You just pull it out
9:30
Speaker 2: of your pocket and go and and so we've done
9:34
Speaker 2: a marvelous job, and the whole pocus industry has done
9:36
Speaker 2: a marvelous job in emergency care. But yes, primary care
9:40
Speaker 2: is a huge use case and having something that's ubiquitous
9:43
Speaker 2: there also in primary care, you never know what you're
9:45
Speaker 2: going to see. It's being used in intensive care when
9:49
Speaker 2: they're rounding patients in the hospital instead of having to
9:52
Speaker 2: order an ultrasound, and we have a cartwheeled over the
9:56
Speaker 2: day again pull out of pocket just to a quick
9:58
Speaker 2: scan the check. And it's been found that when doctors
10:00
Speaker 2: have that ability, patients get out of the hospital sooner.
10:04
Speaker 2: And those are the areas. And so if our mission
10:06
Speaker 2: with i Q three has simply been let's get our
10:08
Speaker 2: image quality better. You know, when Netflix came out, you know,
10:12
Speaker 2: the streaming into your home, you know, the bandwidth that
10:15
Speaker 2: they required your home internet just wasn't as powerful and
10:19
Speaker 2: so I don't know for those who remember, but you
10:21
Speaker 2: had buffering issues and you had resolution issues. And so
10:24
Speaker 2: while Netflix was an awesome idea, you know it didn't
10:28
Speaker 2: have early on the image quality that you know, going
10:31
Speaker 2: to Blockbuster and renting a DVD did you know, but
10:34
Speaker 2: the moment digital equals analog, digital wins. And so when
10:39
Speaker 2: that streaming was resolved and they compressed and their image
10:43
Speaker 2: was fantastic and you can push a button and get
10:46
Speaker 2: any video you want, and you know, it put Blockbuster
10:49
Speaker 2: out of business and we're on that path.
10:51
Speaker 1: You're on that path. I mean, so you talked about
10:53
Speaker 1: the third generation. You just you you were with TSMC
10:57
Speaker 1: for the fourth generation. Yep in your mind, When do
11:02
Speaker 1: you think we get to that parody in the quality
11:04
Speaker 1: versus we get to that point where there's no longer
11:07
Speaker 1: buffering within for the poor side.
11:10
Speaker 2: Yeah? Right, Well, I think there's an amazing journey that's
11:13
Speaker 2: happening for this digital acquisition. I think today our IQ
11:17
Speaker 2: three is as good as any other handheld device out there.
11:21
Speaker 2: I think our IQ four or our next version, whatever
11:24
Speaker 2: we call it, we might not call it IQ four,
11:26
Speaker 2: it might be called something else, but the next version
11:28
Speaker 2: that comes out in a year will be much better
11:33
Speaker 2: than every other hand held out there. So you won't
11:36
Speaker 2: need to buy the other handhelds. You just need a butterfly.
11:39
Speaker 2: But also that image quality is going to rival some
11:41
Speaker 2: of the carts in the hospital and so, and those
11:44
Speaker 2: carts are thirty forty thousand dollars. We sell our handheld
11:47
Speaker 2: for four thousand dollars. So the economics are going to
11:50
Speaker 2: allow for every doctor having their own probe, and we're
11:54
Speaker 2: going to be marching up the specialty stream by going
11:58
Speaker 2: from low end carts to medium cards. Our fifth generation
12:01
Speaker 2: chip Apollo will have twenty times of processing power. It's
12:06
Speaker 2: going to be insane. And I think actually what's going
12:09
Speaker 2: to happen is it's not just going to be about
12:13
Speaker 2: image quality. It's going to be about image quality and interpretation.
12:19
Speaker 2: I think the image is going to start looking differently,
12:22
Speaker 2: not just a sharp image of black and white. That
12:24
Speaker 2: image might actually start looking closer to what the anatomy
12:27
Speaker 2: looks like. Okay, I might be in color, it might
12:30
Speaker 2: be three D, it might use AI embedded on the device,
12:35
Speaker 2: a physical AI device. Today, there are fantastic AI companies
12:40
Speaker 2: that take an image and process it in the cloud
12:44
Speaker 2: and then push it back to a user. I think
12:48
Speaker 2: the possibility of taking those big algorithms and putting it
12:53
Speaker 2: into a device is very possible.
12:56
Speaker 1: Yeah, and I'm really just dying to jump into this
12:58
Speaker 1: mid journey press release, but I got I feel like
13:00
Speaker 1: I need to ground a couple more things, because it's
13:03
Speaker 1: not just the technology you guys are doing, it's the
13:07
Speaker 1: the embedded feature, it's your compass AI solution. How are
13:12
Speaker 1: all those kind of creating more of an ecosystem than
13:16
Speaker 1: just a handheld technology piece.
13:18
Speaker 2: Yeah, that's a very important question.
13:20
Speaker 1: Today.
13:21
Speaker 2: We look at our customers in two different two different
13:24
Speaker 2: types of customers. One, a couple thousand doctors every quarter
13:29
Speaker 2: go online to our website, pull out their credit card
13:32
Speaker 2: and bioprop and they're buying it for their own use,
13:35
Speaker 2: so they can use it in their practice when they
13:37
Speaker 2: travel and in the hospital, and we sell a and
13:42
Speaker 2: for them. That market is all about ease of use,
13:46
Speaker 2: costs and also AI because those doctors are not classically
13:50
Speaker 2: trained in ultrasound and so new AI tools is going
13:53
Speaker 2: to make it easier for them to either find the
13:56
Speaker 2: anatomy they're looking for or be able to interpret that anatomy.
13:59
Speaker 2: And one very important pathway. The other pathway is into
14:04
Speaker 2: medical institutions, into hospitals. Now, hospitals if they have fifty
14:09
Speaker 2: or one hundred doctors who have butterflies, well, how do
14:12
Speaker 2: they get that data into their system? How do they
14:14
Speaker 2: how do they record it into the patient record, capture
14:17
Speaker 2: the image and their DICOM database and also get reimbursement.
14:21
Speaker 2: That's what Compass is. So we've built out a whole
14:24
Speaker 2: SaaS platform that allows the hospital to aggregate all the
14:27
Speaker 2: data from all the devices into their current systems. It's
14:30
Speaker 2: a middleware that links it all to allow a butterfly
14:34
Speaker 2: to not just be for an individual doctor, but to
14:37
Speaker 2: be a part of every day in hospital medicine.
14:40
Speaker 1: Interesting, I'm almost envisioning that is then you can you
14:45
Speaker 1: have that data of all those images somewhere down the road.
14:49
Speaker 1: Potentially I can just picture you be able to compare
14:53
Speaker 1: previous images to what you're seeing to help with the
14:56
Speaker 1: diagnostic process. Is that is that?
14:58
Speaker 2: Am I?
14:58
Speaker 1: Am I thinking too far ahead? That kind of how
15:00
Speaker 1: you're envisioning that too, Well, that's.
15:02
Speaker 2: A very powerful thought that you have there right now.
15:04
Speaker 2: I think hospitals are looking to capture the record of
15:08
Speaker 2: the encounter with the patient, and a lot of times
15:11
Speaker 2: because ultrasound is so easy to just pick up, look
15:14
Speaker 2: and put down, doctors don't even record that image. What
15:18
Speaker 2: we've done is we've made it really easy for the
15:21
Speaker 2: doctors to record those images. And so those are just
15:23
Speaker 2: the basics of what we're trying to do with the
15:26
Speaker 2: enterprise software. But now, as you talk about AI and
15:29
Speaker 2: you talk about looking, you know, I've lived in nine
15:33
Speaker 2: different places in the United States for my job, I
15:36
Speaker 2: probably have had eight different primary care doctors. I've had
15:39
Speaker 2: surgery in three different hospitals. I've had imaging in probably
15:43
Speaker 2: six different places. My medical records are just a yard
15:47
Speaker 2: sale the United States, and there's no Now if you
15:50
Speaker 2: could imagine if all those records were in one place
15:53
Speaker 2: and all that data was there, then you can run
15:56
Speaker 2: AI off of the progression. Like if I had three
15:58
Speaker 2: different heart scans, you then can run AI and look
16:00
Speaker 2: at hey, what's deteriorating, right or whatever. So that that
16:04
Speaker 2: is the future of healthcare is the ability to acquire
16:08
Speaker 2: your data, capture your data, and let AI run the data.
16:11
Speaker 2: And that's really the next step in healthcare.
16:13
Speaker 1: Yeah, you kind of hit the you know meil on
16:17
Speaker 1: the head there. It's just like you it doesn't make
16:21
Speaker 1: sense that your records are just lost once you move
16:24
Speaker 1: to a different stata or go to a different doctor,
16:26
Speaker 1: And it's like almost like they have to start from
16:28
Speaker 1: square one versus actually having the data that's already been acquired.
16:32
Speaker 2: That's right. Yeah, And you know, like the EMR companies
16:35
Speaker 2: today are doing a good job of being more cloud based.
16:39
Speaker 2: If they have multiple like Epic has a very high
16:41
Speaker 2: market share, and if there are multiple Hopps hospitals on EPIC,
16:45
Speaker 2: they can actually go and pull a record from a
16:48
Speaker 2: different hospital. But it is not as intuitive and easy,
16:52
Speaker 2: and it is not integrated with all the encounters. You know,
16:54
Speaker 2: each medical company wants to own your data. Yeah, you know.
16:58
Speaker 2: And whether it's the hospital wants to own your data,
17:00
Speaker 2: whether it's you know, the payer wants to own your data,
17:04
Speaker 2: or the medical companies want to own the data, it's
17:07
Speaker 2: incredibly fragmented. The future of healthcare is data being aggregated
17:11
Speaker 2: in a single place. Future healthcare is AI being able
17:15
Speaker 2: to identify patterns, identify trends. I probably have had maybe
17:20
Speaker 2: thirty or forty blood tests. I'm fifty nine years old. Yeah,
17:23
Speaker 2: so I've probably had thirty or forty blood tests in
17:25
Speaker 2: my life. Can you imagine if all of those blood
17:27
Speaker 2: tests were in one place and then and then while
17:30
Speaker 2: all the markers AI can look at it and then
17:32
Speaker 2: start creating trends on hey, you know, finding things early
17:36
Speaker 2: is the essence of healthcare. And right now, Butterfly, all
17:42
Speaker 2: of the one hundred and fifty thousand devices that we
17:44
Speaker 2: have out there worldwide or connected, and we're streaming into
17:48
Speaker 2: our cloud about forty thousand new images a day and
17:51
Speaker 2: have about twenty five million images there. So that those
17:55
Speaker 2: images are used for our ability to develop AI. But
17:59
Speaker 2: I think ultimately we're helping hospitals aggregate that data and
18:03
Speaker 2: make it much simpler for the patient, and I think
18:06
Speaker 2: hospitals and companies over time will use the trending of
18:11
Speaker 2: that data for better healthcare. Right, it's just the problem.
18:14
Speaker 2: It's just how fragmented our healthcare system is, and no
18:17
Speaker 2: one wants to give it up to someone else to
18:19
Speaker 2: be the leader. And you can't fully trust a government
18:23
Speaker 2: entity from owning.
18:24
Speaker 1: All the data. Yeah, it's a lot of work to do.
18:26
Speaker 1: It's yeah, but it makes sense in the long run,
18:29
Speaker 1: especially for the patient's perspective.
18:30
Speaker 2: That's right, a patient really, you know, I mean, at
18:34
Speaker 2: this generation now, parents should do everything possible too. In
18:40
Speaker 2: there's a twenty first Centric Cures Act in twenty sixteen,
18:45
Speaker 2: and it codified that the individual owns their own data,
18:51
Speaker 2: and it actually very specifically made it a legal requirement
18:55
Speaker 2: for every healthcare entity to be able to provide the
18:58
Speaker 2: data to the patient within ten days. And I think
19:01
Speaker 2: it's we're going to find that people are going to
19:04
Speaker 2: be proactive about aggregating their own data because it's the
19:07
Speaker 2: most powerful way to maintain their health yeah, absolutely. So
19:11
Speaker 2: we talked about the compass AI ecosystem, but there's also
19:15
Speaker 2: the embedded program that you have and what what exactly
19:20
Speaker 2: what are you how how is that embedded program working?
19:24
Speaker 1: So?
19:25
Speaker 2: Uh, Sometimes companies, as they're developing technology, they're trying to
19:31
Speaker 2: solve problems, right, So you're developing technology to meet a
19:33
Speaker 2: certain objective, and sometimes you solve problems that other people have,
19:41
Speaker 2: like for example, you know, if I'm sure you remember
19:44
Speaker 2: that that old photograph of Jeff Bezos sitting at his
19:46
Speaker 2: computer with the handwritten Amazon on the side, legendary image,
19:51
Speaker 2: you know when when when he was envisioning Amazon, it
19:55
Speaker 2: was you know, a online book retailer that would then
19:58
Speaker 2: build the infrastructure for online retailer. And on that journey,
20:03
Speaker 2: they realized they needed cloud compute. So how do we
20:06
Speaker 2: get all of this in the cloud? And there was
20:08
Speaker 2: no single cloud compute vendor that had the power of
20:13
Speaker 2: reaching what they needed. So they built it themselves. They
20:16
Speaker 2: put data centers over the world, spent billions of dollars
20:19
Speaker 2: with APIs and Internet connectivity. And when they built that
20:22
Speaker 2: for themselves, they were using less than one percent of
20:25
Speaker 2: its capacity. They said, wow, we spend all this money
20:28
Speaker 2: for this, Let's go see if we could sell some
20:30
Speaker 2: of the capacity. Well, that became AWS, which is now
20:33
Speaker 2: the most powerful, largest cloud data server or service in
20:38
Speaker 2: the world and the most profitable part of Amazon today.
20:41
Speaker 2: They solved the problem for themselves, but everyone else would
20:44
Speaker 2: have that problem. It's the same thing if you look
20:47
Speaker 2: at the video from nineteen ninety three to two thousand
20:51
Speaker 2: and five, the video was a video game company. Yeah,
20:54
Speaker 2: you're not going to find during that time a business
20:57
Speaker 2: plan to create data centers or to be the engine
21:01
Speaker 2: for AI. But in two thousand and sixty made a
21:04
Speaker 2: really important decision, which was we are going to open
21:07
Speaker 2: up our platform to third party developers, and they launched
21:10
Speaker 2: something called Kuda, and Kuda allowed a third party to
21:13
Speaker 2: change the settings of those GPUs for their purposes. I
21:19
Speaker 2: just listened to a three hour podcast of Jensen talking
21:22
Speaker 2: to Joe Rogan and he would just walked through the
21:24
Speaker 2: entire journey and how many you know, near death experiences
21:28
Speaker 2: the company had. It's just fascinating. And everything changed when
21:31
Speaker 2: Elon Musk took eight of those GPUs in a box
21:34
Speaker 2: or to open AI and they now started to use
21:37
Speaker 2: those chips that were designed for video games. Remember, yeah,
21:40
Speaker 2: now to network for AI intelligience.
21:42
Speaker 1: But it's also just interesting too that he's willing to
21:45
Speaker 1: let other people make inputs into the technology development. And
21:49
Speaker 1: it's you know, there's a lot of these CEOs, a
21:51
Speaker 1: lot of these founders that could be very stubborn and
21:54
Speaker 1: very stick to their ground and now my idea is
21:57
Speaker 1: the best idea. But here was that I like, let's
21:59
Speaker 1: get some other ideas and there, let's let's diversify those
22:02
Speaker 1: ideas and let's see what we can build out of it.
22:04
Speaker 1: And it took It was that moment that you know,
22:06
Speaker 1: build them, like you said, from the video game cards
22:09
Speaker 1: to where they are today.
22:11
Speaker 2: So prior to me coming to Butterfly, we were approached
22:15
Speaker 2: by a very large consumer products company that you would know.
22:19
Speaker 2: I don't want to reveal their name without their permission,
22:23
Speaker 2: and they asked us, they said, you know this semiconductor
22:26
Speaker 2: sema technology you have because they see our six hundred
22:29
Speaker 2: patents and is you know if you're going to get
22:31
Speaker 2: into it, you know that this is what we have.
22:33
Speaker 2: And they said, we want to create a fish finder
22:36
Speaker 2: with a digital ultrasound device. Would you work with us
22:41
Speaker 2: on the chat and our team said no, we're in
22:44
Speaker 2: the point of CARLTD this, you know, and so when
22:48
Speaker 2: I got there and heard the story, I just I
22:50
Speaker 2: just yelled, you know, yelled saw loudly. Building almost came down.
22:54
Speaker 2: So we we created a semiconductor that can deliver sound,
23:00
Speaker 2: and they can deliver sound through those nine thousand sensors
23:03
Speaker 2: in any way that you wish to program it. And
23:06
Speaker 2: that's what makes it digital because it can change dynamically.
23:09
Speaker 2: So what we decided to do was to make that
23:11
Speaker 2: technology available to other people. And we found that there's
23:15
Speaker 2: a lot of applications of a semiconductor ultrasound device outside
23:19
Speaker 2: of porn of carrelter sound, and so embedded is our program.
23:22
Speaker 2: It's kind of like an Intel inside. Yeah, you know,
23:24
Speaker 2: it's a butterfly embedded same type of vernacular. Is our
23:28
Speaker 2: program to partner with third parties, allow them or give
23:33
Speaker 2: them the software where they can change the settings of
23:36
Speaker 2: our chips, and then allowing them to be able to
23:39
Speaker 2: tune it specifically for their application. And what that does
23:43
Speaker 2: for our investors. I mean, our first semiconductor cost US
23:46
Speaker 2: three hundred million dollars to build the first one, so
23:49
Speaker 2: there's a big upfront investment in this core technology. And
23:52
Speaker 2: I think today our porny careult sound TAM is about
23:55
Speaker 2: twenty million, but by opening up our technology to third
23:59
Speaker 2: party developers, our TAM opens from twenty billion to three
24:02
Speaker 2: hundred and fifty billion. We now have all these different
24:05
Speaker 2: applications in healthcare where a digital ultrasome can be used
24:08
Speaker 2: and they can be a part of someone else's product. There.
24:12
Speaker 2: You know, we've partnered with a vascular robotics company who
24:15
Speaker 2: wants to position a needle using a very sophisticated technology
24:19
Speaker 2: and they want our chip to help them do that.
24:22
Speaker 2: And there is a Forest Neurotech who is a brain
24:25
Speaker 2: computer interface company that wants to put our chip in
24:27
Speaker 2: the brain to be able to communicate to the brain.
24:30
Speaker 2: And so we now have nine different partnerships that we've signed.
24:34
Speaker 2: We are talking to thirty or forty or fifty other
24:37
Speaker 2: companies actively right now about these partnerships, and embedded will
24:41
Speaker 2: be a whole new revenue stream and a whole new
24:43
Speaker 2: opportunity for Butterfly.
24:45
Speaker 1: And so that gets us now to mid Journey. Yeah,
24:48
Speaker 1: and you know the press release in my understanding of
24:53
Speaker 1: mid Journey is it does texts to imaging. I love
24:58
Speaker 1: to hear more about how we can create how what
25:01
Speaker 1: what's the partnership with ultrasound to imaging.
25:05
Speaker 2: Absolutely so what what mid Journey has an extraordinary expertise
25:09
Speaker 2: of doing is taking a small amount of data and
25:13
Speaker 2: being able to extrapolate it into an extraordinary useful form.
25:18
Speaker 2: Uh And in what they what their core business is
25:21
Speaker 2: is taking texts and you know that text will then
25:24
Speaker 2: generate an image based upon what you've asked for and
25:28
Speaker 2: then we'll give you all these different variations. But everything
25:31
Speaker 2: is based on the truth of the of the text
25:34
Speaker 2: that that you initially started with. So they believe that
25:38
Speaker 2: they can use their core engine to take ultrasound imaging
25:42
Speaker 2: from the body and present it in a way that
25:45
Speaker 2: is extremely medically useful and then use the power of
25:49
Speaker 2: their AI to help identify certain changes or irregularities in
25:55
Speaker 2: in you know, the the actual imaging itself. So their
25:58
Speaker 2: vision is to ate a whole body scanner. And you
26:03
Speaker 2: know today there are two different there are two modalities
26:09
Speaker 2: of whole body imaging. There's MRI and MRI is magnetic
26:14
Speaker 2: and while it doesn't use radiation, it uses very high
26:18
Speaker 2: powerful magnetic ways through the body. If anyone has any
26:21
Speaker 2: metal of any sort, it can create tremendous pain and
26:26
Speaker 2: trauma to the body. So it is a very expensive
26:29
Speaker 2: it's five thousand dollars of scan. Payers have to approve
26:34
Speaker 2: a scan, you have to go into a facility. You
26:36
Speaker 2: don't get an MRI unless you're sick and there's a
26:40
Speaker 2: payer paying a lot of money for it, or there's
26:42
Speaker 2: a CT which uses radiation and you can only have
26:45
Speaker 2: a limited amount of radiation in your body, and so
26:48
Speaker 2: it's really reserved for the most consequential type of imaging.
26:54
Speaker 2: Ultrasound doesn't have that artifact. Ultrasound or sound waves to
26:57
Speaker 2: go through the body and don't leave behind a residual
27:00
Speaker 2: and don't and are very safe and low cost. But
27:03
Speaker 2: historically ultra sound is the most difficult image to read.
27:08
Speaker 2: So now if you can have an extrapolation layer and
27:11
Speaker 2: AI layer that can take that data and represent it
27:14
Speaker 2: in a way that is much easier to read, then
27:18
Speaker 2: you're now opening imaging up to everyone. So mid Journey
27:22
Speaker 2: is using many of our chips in a circle and
27:26
Speaker 2: they're using it to scan the entire body to get
27:30
Speaker 2: a massive amount of data and an extremely low cost. Okay,
27:34
Speaker 2: that's interesting because you mentioned just now a circle of
27:37
Speaker 2: the chips. So it is going to be almost an
27:41
Speaker 2: MRI like tube, but the chips are around you in
27:44
Speaker 2: the tube instead of the magnetic technology. Is that the
27:47
Speaker 2: right way to think about it. That's exactly right. The
27:51
Speaker 2: way that the device works is that you stand on
27:54
Speaker 2: a circle okay, and it lowers you into a cylinder
27:58
Speaker 2: of warm water.
27:59
Speaker 1: Oh.
28:00
Speaker 2: As you're being lowered it to that cylinder, it's capturing
28:03
Speaker 2: the images from your feet and your legs and as
28:06
Speaker 2: you move down through the cylinder. This this circular, this
28:09
Speaker 2: circular ultrasound capability with with you know, almost nine hundred
28:15
Speaker 2: thousand sensors is scanning through your body and and and
28:20
Speaker 2: pulling all that data into an image.
28:24
Speaker 1: Okay, I'm may be answering my own question. Okay, so
28:27
Speaker 1: the the warm water cuts like so you're going into water?
28:31
Speaker 2: Yes?
28:31
Speaker 1: Is that similar to like I said, you know, thinking
28:34
Speaker 1: about my wife had her ultrasound appointments during her pregnancies
28:39
Speaker 1: and they put like that warm goo on. Is that
28:42
Speaker 1: is that essentially that the warm water is replacing that
28:45
Speaker 1: warm and.
28:45
Speaker 2: Grewe it's the connectivity for the sound weights. Okay. So
28:48
Speaker 2: they so the sound waves move do not move through air,
28:52
Speaker 2: so they have to move through some level of tissue
28:55
Speaker 2: fluid and and it's designed so yes, when you put
28:59
Speaker 2: that goo on, it is creating the connection between your
29:02
Speaker 2: skin and the probe. But the ultrasound, the first ultrasound
29:06
Speaker 2: device was actually crystals in water. And because like if
29:11
Speaker 2: you like, whales communicate with with their sonar and those
29:16
Speaker 2: sound waves can travel miles and miles and miles underwater.
29:20
Speaker 2: It's the same type of principle. And so you take
29:24
Speaker 2: someone who gets lowered into the water and then that
29:28
Speaker 2: that you're they get you know, they get sent all
29:31
Speaker 2: these sound waves through the body that harmlessly goes through
29:34
Speaker 2: and then listens and extrapolates that data into the image.
29:38
Speaker 1: Wow, I mean, this is just all very fascinating. But
29:43
Speaker 1: here we are, you know, just the press release came out,
29:46
Speaker 1: all this sounds so futuristic. So what are the steps
29:50
Speaker 1: from a product development side and then ultimately a commercial
29:53
Speaker 1: side to get from where we are today to that idea?
29:57
Speaker 2: Right? Well, so, so much of this is is discovery
30:01
Speaker 2: and so the first phase of our partners is a
30:06
Speaker 2: tech kind of a technology acquisition, technology access where we
30:11
Speaker 2: give them a license to our developer software similar to Kuda.
30:15
Speaker 2: We give them access to our chips, We teach them
30:17
Speaker 2: how to do the programming, We help them understand the system,
30:21
Speaker 2: and they go off and they go off in their
30:22
Speaker 2: labs and they try to figure out can they get
30:24
Speaker 2: the chip to do exactly what they wanted to integrate
30:28
Speaker 2: with their systems? Is it given them the type of
30:30
Speaker 2: outcome they're looking for? And we with mid Journey, we
30:34
Speaker 2: had signed a deal. They had access to our technology
30:37
Speaker 2: and they were doing their thing, and then you know,
30:40
Speaker 2: one day they came back to us said we would
30:42
Speaker 2: like to have exclusivity in this one field of us
30:46
Speaker 2: and that exclusivity turned into a seventy four million dollar
30:49
Speaker 2: licensing deal over five years. So I think we have
30:53
Speaker 2: now eight other deals just like that where people are
30:57
Speaker 2: doing their discovery. And that's why r AND is research
31:01
Speaker 2: and development, and so the research phase is you're going
31:03
Speaker 2: to try to do something no one's ever done. Sometimes
31:05
Speaker 2: it doesn't succeed, sometimes it does. That's why it's called
31:08
Speaker 2: research and we have those partners now doing research today.
31:13
Speaker 2: And then they evolve where they pay us a license fee,
31:17
Speaker 2: they buy our chips, and then if they as they progress,
31:20
Speaker 2: they might want us to do work for them, and
31:21
Speaker 2: then we charge for an solw and then if they
31:24
Speaker 2: wish to go commercial, then they'll be a customer of
31:28
Speaker 2: our chips infinitum, and then there'll be some type of
31:30
Speaker 2: loyalty or for them to be able to use our software.
31:33
Speaker 2: So these create a lot of shots on goal. They
31:36
Speaker 2: open us up to whole new markets. We're still a
31:39
Speaker 2: small company developing our point of care market. We can't
31:42
Speaker 2: go develop each of these different applications ourselves, and so
31:47
Speaker 2: we give our investors the ability to participate in all
31:49
Speaker 2: these markets now because we have this unique, unique technology set,
31:54
Speaker 2: and we have the ability to transfer this technology into
31:58
Speaker 2: a lot of other products.
32:00
Speaker 1: And then how's mid journey planning this? You know, when
32:05
Speaker 1: you have this is this sounds more like a preventive
32:08
Speaker 1: care type of technology that they're looking to create. And
32:13
Speaker 1: so you know, you see you're talking about like the
32:15
Speaker 1: MRI and the CT scans are meant for those really
32:18
Speaker 1: really sick patients, but that's already a kind of too late.
32:21
Speaker 1: You want to get it early in the disease progression.
32:23
Speaker 1: So is this the direction that they're going to go
32:25
Speaker 1: into this kind of earlier disease state or even general
32:30
Speaker 1: health stage of the disease platform.
32:34
Speaker 2: I think what mid Journey is doing is going to
32:38
Speaker 2: create a significant impact for humanity. You know, we talked
32:44
Speaker 2: just a moment ago about the healthcare system, the fragment
32:49
Speaker 2: of data that exists, and the fact that, you know,
32:53
Speaker 2: our healthcare system. I've been in healthcare thirty five years,
32:56
Speaker 2: and so I feel I have the right, you know,
32:59
Speaker 2: to to as an observer of healthcare. You know, call
33:03
Speaker 2: it like it is. You know, our healthcare system is
33:05
Speaker 2: a sick care system. It's not doing anything to keep
33:08
Speaker 2: you healthy, it's doing something to deal with the sickness.
33:12
Speaker 2: Then when you're sick. I kind of grew up not
33:15
Speaker 2: to name drop, but I kind of grew up with
33:17
Speaker 2: doctor Ozka, and I was a product manager at a company,
33:21
Speaker 2: and he was early in his career at Columbia Presbyterian.
33:25
Speaker 2: And one of the things that would frustrate him was,
33:28
Speaker 2: you know, patients would smoke their whole life, or patients
33:31
Speaker 2: would you know, take any other bad habit. Yeah, and
33:36
Speaker 2: then they would hit a wall and they would come
33:38
Speaker 2: to him and say, oh my heart, you know, yeah,
33:40
Speaker 2: I don't feel well this and that, and he's like,
33:42
Speaker 2: you know, they've behaved a certain way their whole life,
33:45
Speaker 2: and then they would just want to be fixed well,
33:47
Speaker 2: you know, And so he started telling people he won't
33:49
Speaker 2: operate on them until they stopped smoking, or they won't
33:52
Speaker 2: he won't do this until they change these behaviors. And
33:55
Speaker 2: he would shock them, but then he'd help them. And
33:58
Speaker 2: so because what he really cared for the long term
34:00
Speaker 2: health of the patient. But a lot of times people
34:02
Speaker 2: don't understand how their habits are impacting their health because
34:07
Speaker 2: we don't have access to that data. Now, over the
34:10
Speaker 2: last ten years, we've seen so many of these personal
34:13
Speaker 2: wearable devices that have come out. You fit it being
34:16
Speaker 2: the first down your Apple Watch. You can get all
34:18
Speaker 2: these different sensing modalities, the whoop that there's so many ways,
34:23
Speaker 2: and patients are taking it in people not patients, but
34:26
Speaker 2: we are kind of taking it upon ourselves to go
34:29
Speaker 2: capture that data because we want to learn earlier mid
34:32
Speaker 2: journey is the next evolution or revolution of that because
34:36
Speaker 2: they're not going to hold this technology hostage inside a
34:40
Speaker 2: healthcare system where a payer decides whether or not you
34:44
Speaker 2: get it, whether you know, in the hospital. With the
34:47
Speaker 2: incredible cost that exists, so if you want to get
34:50
Speaker 2: an MRI, you're going to have to pay a significant
34:53
Speaker 2: amount out of pocket or they're going to tell you no.
34:56
Speaker 2: But mid Journey is going to make this more accessible.
34:59
Speaker 2: They're going to make this things similar to a gym membership,
35:02
Speaker 2: and it's a part of you going in and a
35:04
Speaker 2: part of your health. Now, imagine as a part of
35:06
Speaker 2: this gym membership, you go get a scan every week.
35:10
Speaker 2: So now you're going to have four hundred terror flops
35:12
Speaker 2: of data every single week on the progression of your body.
35:16
Speaker 2: And what are you going to care about when you're
35:18
Speaker 2: you know, you're young and healthy. You know you're not
35:21
Speaker 2: going to be like, oh, I'm just going to go
35:22
Speaker 2: get an MRI. But if you have this as a
35:25
Speaker 2: part of your your health journey and as a part
35:27
Speaker 2: of when you go to a gym, you're on a
35:29
Speaker 2: health journey, yeah, because you're trying to build muscle, you're
35:32
Speaker 2: trying to improve your cardiovascu your health, you're trying to
35:34
Speaker 2: reduce weight. And so if you have this ability to
35:37
Speaker 2: see how your body is responding to a different workout,
35:41
Speaker 2: you're starting to see fat, you know, your BMI reduce,
35:44
Speaker 2: You're starting to see your cardioc your cardiac health increase.
35:48
Speaker 2: These are all and then if you change, have major
35:50
Speaker 2: changes to your diet. You're going to see how it
35:52
Speaker 2: changes your body. And if every week you're getting a
35:54
Speaker 2: scan and that scan is costing you know, just a
35:57
Speaker 2: gym membership, that's going to change your life. Yeah, it's
36:01
Speaker 2: gonna give you greater awareness. It's going to teach you
36:03
Speaker 2: that those behaviors that doctor Oz wanted you to not
36:07
Speaker 2: have and just hit a wall, you know, he would
36:10
Speaker 2: want you to be healthier sooner. And that's really a
36:13
Speaker 2: key component that the next wave of personal health is
36:17
Speaker 2: not just the devices that you wear. It's now the
36:20
Speaker 2: ability to not only get that image, but have all
36:23
Speaker 2: that data for AI to look at, Hey, you know
36:27
Speaker 2: what's happening in your body and if they see something earlier.
36:31
Speaker 2: The biggest way to reduce health care costs in mass
36:36
Speaker 2: is to first of all, teach people to be healthier,
36:38
Speaker 2: but second of all to find something early because if
36:40
Speaker 2: you find it early, you can intervene at very low costs.
36:44
Speaker 2: Who knows you could just change behaviors to get to
36:47
Speaker 2: you know, sometimes as if behaviors progress, then you have
36:50
Speaker 2: to go into surgery. But if you are aware of
36:53
Speaker 2: some behavior that's having a negative outcome, you can change
36:56
Speaker 2: your behavior and then not even enter the health that
36:59
Speaker 2: the secure system all together.
37:01
Speaker 1: Yeah. No, I've seen that firsthand because covering the diabetes companies,
37:05
Speaker 1: I've tried both the stello and the lingo, right, you know,
37:08
Speaker 1: consumer ICGM.
37:10
Speaker 2: That lingo is cool.
37:11
Speaker 1: It is, I mean, and it's but it's interesting because
37:13
Speaker 1: I did notice subtly my eating habits changed as I
37:18
Speaker 1: was wearing it, because it gamified keeping you in that
37:21
Speaker 1: range like they always kept you you. I always kept
37:23
Speaker 1: watching them like and I would stress test that. I'm like, okay,
37:26
Speaker 1: I have another little fun sized candy bars. I getta good.
37:29
Speaker 1: That's the rest I get up to like one Ford like,
37:31
Speaker 1: oh okay, no more, no more, no more. And but
37:32
Speaker 1: I saw that, and I with the smokers, it sounds
37:35
Speaker 1: like if someone is starting to smoke, they do this
37:37
Speaker 1: full body scan, they see their lungs, they see the
37:40
Speaker 1: damage being done to their lungs. They're like, okay, let's
37:43
Speaker 1: maybe let's not stress test my lungs as much. Maybe, right,
37:47
Speaker 1: And even if they cut from one pack a day
37:49
Speaker 1: to a couple of cigarettes a day, that's still healthier
37:52
Speaker 1: for them in the long run. That all makes Yeah,
37:55
Speaker 1: that makes sense there and just you know, being able
37:57
Speaker 1: to use utilize that technology to be able to do
37:59
Speaker 1: that in the future.
38:01
Speaker 2: That's exactly right, because once you see the image of
38:03
Speaker 2: your lung that has an impact. Those things will change
38:07
Speaker 2: your They will change your behavior. So instead of being
38:10
Speaker 2: shocked by having to have your chest cut open and
38:12
Speaker 2: having a surgery thirty years later, you can be shocked
38:15
Speaker 2: a lot earlier and hopefully in that behavioral change. That
38:18
Speaker 2: is probably avoidance of any type of these interventions into
38:23
Speaker 2: our healthcare system is the greatest impact that can occur.
38:28
Speaker 2: That we can actually really have a healthcare system that
38:31
Speaker 2: was backstopped with the ability because look, I'm not looking
38:35
Speaker 2: down my nose at our healthcare system with the greatest
38:37
Speaker 2: healthcare system at the world. You have a disease, you
38:41
Speaker 2: have an injury, You're going to be fixed faster and
38:44
Speaker 2: well and all that is extraordinary. But if you can
38:48
Speaker 2: factor out the avoidable based upon behavior you're going to
38:52
Speaker 2: you're and then you put the money there so people
38:54
Speaker 2: can live longer, healthier lives. I think Mid Journey, what
38:58
Speaker 2: they're doing with Mid Journey Medical and their skin is
39:01
Speaker 2: just an extraordinary leap forward for humanity, especially because of
39:06
Speaker 2: their view on cost and access.
39:09
Speaker 1: Yeah, no, exactly, And yeah, normally I close out these
39:12
Speaker 1: episodes with talking about how you see this market in
39:15
Speaker 1: five years, but I think we really don't fip into that.
39:17
Speaker 1: Let me jump into just the Apollo one. That's your
39:19
Speaker 1: fifth generation semiconductor, how I mean. And we talked about
39:23
Speaker 1: Moore's laws, so I'm assuming some level of Moore's Laws
39:26
Speaker 1: being applied there. But how much is you know, how
39:30
Speaker 1: much improvement are you expecting to see in the new design?
39:33
Speaker 1: And then what other features are there that you are
39:35
Speaker 1: kind of interesting for you.
39:37
Speaker 2: So what we've seen is from our IQ plus with
39:42
Speaker 2: which is our second generation probe, to IQ three, which
39:45
Speaker 2: is our third generation probe. We doubled the processing power
39:49
Speaker 2: and we saw that we were able to use less battery,
39:54
Speaker 2: we were able to get better imaging faster, and it
39:57
Speaker 2: was an extraordinary leap. Apollo doesn't double imaging power, it
40:03
Speaker 2: actually takes it takes our overall power up twenty times.
40:07
Speaker 1: Wow.
40:08
Speaker 2: So we now are going to be able to process
40:11
Speaker 2: twenty times the amount of data than we could today.
40:16
Speaker 2: So it's going to allow for us to not only
40:19
Speaker 2: get a better image, but also be able to stream
40:22
Speaker 2: that data into AI to be able to run models
40:26
Speaker 2: to make to help doctors make clinical decisions. A Butterfly
40:31
Speaker 2: is now going to be leading the way in physical AI,
40:35
Speaker 2: you know where you see so much of what AI.
40:39
Speaker 2: Our experience with AI today is we ask a question
40:42
Speaker 2: and then it goes into the cloud, right, and then
40:45
Speaker 2: the cloud is where the AI models are processed because
40:48
Speaker 2: they're so big, and then the answer comes back to
40:51
Speaker 2: us from the cloud and that's awesome. But so much
40:54
Speaker 2: of the AI processing between all the different companies is
40:58
Speaker 2: over time going to become more more generic. And what's
41:01
Speaker 2: going to be unique is how that AI is actually
41:04
Speaker 2: put into a device. You've heard, you know, Jonny ives
41:09
Speaker 2: has you know, joined open Ai AND's developing this personal
41:13
Speaker 2: AI device that will be coming out in a year.
41:15
Speaker 2: That it's all these different things where how AI at
41:19
Speaker 2: a local level can impact and be inside a device
41:23
Speaker 2: to give us a substantial amount of intelligence where you
41:26
Speaker 2: need it. That's what Apollo is going to do. And
41:29
Speaker 2: Apollo with AI and the image quality, I think it's
41:33
Speaker 2: going to start start changing how ultrasound images are acquired.
41:38
Speaker 2: You know, when you go into a CT suite or
41:40
Speaker 2: an MR suite, the technicians have to worry about a
41:43
Speaker 2: lot of things, but what they don't have to worry
41:45
Speaker 2: about is how the image is going to be acquired
41:47
Speaker 2: and how they're going to interpret that image. It's an
41:50
Speaker 2: automatic protocol, they say, line the table, don't hit your
41:52
Speaker 2: head when you go in. But then the machine automatically
41:55
Speaker 2: acquires the image and then a trained person listens to
41:58
Speaker 2: that image. With ultra sound, it's hard to acquire and
42:03
Speaker 2: it's hard to interpret. The AI tools are going to
42:05
Speaker 2: make it a lot easier and are going to make
42:08
Speaker 2: this low cost, low consequence imaging modality very very accessible.
42:14
Speaker 1: Yeah, it's gonna be exciting to see this play out.
42:17
Speaker 1: Just lastly, is there any timeline for that Apollo yet
42:20
Speaker 1: or is that still under development? And you know, more
42:23
Speaker 1: to be heard from a later date.
42:25
Speaker 2: Apollo's probably end of twenty eight or early twenty nine.
42:28
Speaker 2: So well, I.
42:29
Speaker 1: Hope to have you back a lot sooner than that.
42:31
Speaker 2: Yeah, I will come back.
42:33
Speaker 1: But yeah, what a fascinating episode. Joe. Thank you so
42:35
Speaker 1: much for joining us today.
42:36
Speaker 2: Appreciate it, Matt, and thank you to.
42:38
Speaker 1: Our producers Adidias Somani and Mariam Thawaile for all the
42:42
Speaker 1: behind the scenes work. And thank you to our listeners.
42:45
Speaker 1: For tuning in today. We hope you join us for
42:47
Speaker 1: future episodes, and if you like to stay up to date,
42:49
Speaker 1: you can click the subscribe button on Spotify or your
42:52
Speaker 1: favorite streaming platform. Take care.
43:16
Speaker 2: Uses asses chases, cosble
43:28
Speaker 1: Bases,