Speaker: 00:00:00

Exactly.

00:00:00,861 --> 00:00:02,612

No, you're right.

00:00:02,633 --> 00:00:08,197

And I think, you know, we had a bit of a space race when it came to AI.

00:00:08,197 --> 00:00:20,065

So you had these hospitals last year, they had like 80, you know, I read an article that

said one academic uh medical center had 80 AI widgets, software solutions, whatever you

00:00:20,065 --> 00:00:21,176

want to call it, last year.

00:00:21,176 --> 00:00:23,307

This year, they now have 200.

00:00:23,307 --> 00:00:25,989

And we're only in September.

00:00:26,031 --> 00:00:28,981

So they're probably going to be up to 250 by the end of the year.

00:00:41,306 --> 00:00:45,809

Hello and welcome to Haverin About, our brand new podcast where we

00:00:45,809 --> 00:00:52,763

mix stories, some health information technology strategy and a fair bit of Scottish style

blethering.

00:00:52,883 --> 00:00:54,284

I'm Stuart Miller.

00:00:54,699 --> 00:00:57,270

And I am Bethany, his daughter.

00:00:57,686 --> 00:00:59,057

excellent stuff.

00:00:59,057 --> 00:01:09,564

Well, before we get started with today's topic, I was kind of curious this week in terms

of like, you know, what in the news caught your attention or in your life as you were sort

00:01:09,564 --> 00:01:11,897

of like navigating through the past week.

00:01:11,897 --> 00:01:15,920

I have been slowly binging The Pitt which I'm late to it.

00:01:15,920 --> 00:01:17,381

I know I'm late to it.

00:01:17,381 --> 00:01:17,872

I know.

00:01:17,872 --> 00:01:20,645

Well, I have a almost five-year-old.

00:01:21,206 --> 00:01:24,328

She's not interested in watching The Pitt, sadly.

00:01:24,589 --> 00:01:28,201

But it's convinced me that I low-key maybe should have become a doctor.

00:01:28,201 --> 00:01:32,294

Because there's been like three cases where I'm like, clearly it's this.

00:01:32,815 --> 00:01:34,896

And it's turned out like right upper quadrant pain.

00:01:34,896 --> 00:01:37,126

was like, gallstones, gallstones.

00:01:37,126 --> 00:01:39,387

And I was like, I think that was too easy of a diagnosis.

00:01:39,387 --> 00:01:41,467

I bet you something else is going on.

00:01:41,867 --> 00:01:43,378

He did heart attack, died.

00:01:43,378 --> 00:01:47,679

And then the baby came in and they were like, no, they don't know how to fix this baby.

00:01:47,679 --> 00:01:50,369

We're doing all these tests and meningitis.

00:01:50,630 --> 00:01:56,551

The minute I saw socks on that baby, I said that baby has a hair tourniquet on her toe.

00:01:57,632 --> 00:02:06,864

And look, and it turned out, you know, yeah, and you know how you get rid of it with, with

Nair, with, with that cream to melt, melt the hair away.

00:02:07,718 --> 00:02:10,634

So I'm now like, should I go back to school and become a doctor?

00:02:10,634 --> 00:02:13,780

Like, I'm sure it's just as easy as that, right?

00:02:13,780 --> 00:02:20,102

should explain to our viewers, this is a decades long Miller family tradition.

00:02:20,783 --> 00:02:30,567

Because when I first met Sarah, Bethany's mom, we, uh my wife, still my wife, I should

lead with that.

00:02:30,767 --> 00:02:33,608

But when I first met her, we were both in nursing.

00:02:33,608 --> 00:02:39,389

And there was a show in the UK called Casualty that had just launched on the BBC.

00:02:39,457 --> 00:02:49,002

And it was your archetypal emergency department show, know, three cases of the week and

then drama within the team.

00:02:49,002 --> 00:02:54,625

So we've been playing this game of, I wonder what they're going to do with this one.

00:02:55,126 --> 00:03:00,869

For, I can't even remember how old Casualty is now, about 40 years?

00:03:00,869 --> 00:03:02,150

Probably something like that.

00:03:02,150 --> 00:03:05,221

It's probably something like that because it predated you and Jamie.

00:03:05,221 --> 00:03:07,512

So it's got to be something like that.

00:03:07,512 --> 00:03:08,545

00:03:08,545 --> 00:03:15,287

remember, and I mean, I was watching the show at like seven, eight, it's not like, like I

probably shouldn't have.

00:03:15,287 --> 00:03:19,268

It wasn't that gory, because it was on BBC, so it's not super gory.

00:03:19,268 --> 00:03:20,248

But there's definite blood.

00:03:20,248 --> 00:03:27,240

And I remember there was one where they had to like take a pen and like pull it apart to

like the tubey bit.

00:03:27,240 --> 00:03:30,031

And they shoved the tube in to let air out.

00:03:30,031 --> 00:03:33,932

And that was another one that was on The Pitt They were like, no, there's air collecting.

00:03:33,932 --> 00:03:36,774

I was like, just get a biro pen and shove it in.

00:03:36,774 --> 00:03:38,636

And they didn't use a biro pen.

00:03:38,636 --> 00:03:45,921

use like a scalpel, like actual medical devices.

00:03:45,921 --> 00:03:49,431

So yeah, so I just, maybe I should be a doctor.

00:03:49,431 --> 00:04:01,448

Well, and I will say in kudos to the Pitt think that the drama, number one, the medical

accuracy of a lot of the cases is so good.

00:04:01,448 --> 00:04:09,752

I mean, we've got a lot of experience of being in the clinical area, not only as a

clinician, but also as a patient.

00:04:09,752 --> 00:04:14,015

And so therefore, and I've had some pretty high acuity stuff done.

00:04:14,015 --> 00:04:18,745

And in our family, we've had other family members who have been pretty sick.

00:04:18,745 --> 00:04:21,805

And so therefore we've got quite a lot of experience with some of that stuff.

00:04:21,805 --> 00:04:32,445

And I have to say, we watch some medical stuff and it is like, oh, you shudderingly bad

when it comes to medical accuracy.

00:04:32,445 --> 00:04:33,745

The Pitt is not that.

00:04:33,745 --> 00:04:37,145

Whoever their advisors are, are doing a sterling job.

00:04:37,145 --> 00:04:39,345

And I don't mean just the medicine.

00:04:39,345 --> 00:04:43,825

I mean the nursing and even some of the kind of hospital administration drama.

00:04:43,825 --> 00:04:46,819

But I got to say, Noah Wyle number one,

00:04:46,819 --> 00:04:52,251

kudos dude for leaning back in on being an emergency doctor later in your career.

00:04:52,251 --> 00:05:05,935

And I will say it was marvelous to see him do his shout out to FIGS, the company that

makes scrubs and is very well known in the healthcare provider industry, by having them

00:05:05,935 --> 00:05:08,186

make that tuxedo for him.

00:05:08,186 --> 00:05:13,057

I thought that was an excellent shout out to all of our healthcare warriors.

00:05:13,449 --> 00:05:14,169

Yeah.

00:05:14,169 --> 00:05:17,571

And to be clear, I could not be a doctor.

00:05:18,951 --> 00:05:19,822

absolutely not.

00:05:19,822 --> 00:05:22,964

Could I diagnose a hair tourniqet on a toe?

00:05:22,964 --> 00:05:23,435

Yes.

00:05:23,435 --> 00:05:25,196

That is the limit of my...

00:05:25,196 --> 00:05:31,619

We joke, we're medically literate, but we are not clinicians.

00:05:31,619 --> 00:05:36,421

lot of respect for the people that have actually studied this and know what they're doing.

00:05:36,693 --> 00:05:38,034

yeah, 100%, yeah.

00:05:38,575 --> 00:05:49,106

Well, on that happy topic and AI replacing doctors and Stuart and Bethany not replacing

doctors, maybe we should go on to talk about today's topic.

00:05:49,499 --> 00:05:55,491

Yes, today's topic is cutting through the AI FUD in healthcare.

00:05:55,892 --> 00:05:59,893

AI is everywhere in the world, in life.

00:05:59,913 --> 00:06:06,476

I used chat GPT this morning to make uh coloring pages for my daughter's birthday party

this weekend.

00:06:06,476 --> 00:06:07,697

They're very cute.

00:06:07,697 --> 00:06:10,858

I'll see if maybe on YouTube we can put an image of them.

00:06:10,858 --> 00:06:12,318

They're very cute.

00:06:13,019 --> 00:06:14,899

So AI really is everywhere, right?

00:06:14,899 --> 00:06:19,067

And if you go to any of the conferences, I went to one in Europe this summer.

00:06:19,067 --> 00:06:25,539

I went to another just this week in Atlanta, all the booths, artificial intelligence, just

everywhere.

00:06:25,539 --> 00:06:26,116

AI, AI.

00:06:26,116 --> 00:06:33,319

Yeah, we are definitely in the full throng of an AI hype cycle.

00:06:33,319 --> 00:06:49,128

um And it is, I think it's very confusing uh for a lot of the people who are interested in

how the technology can work in health, in healthcare and how it can have a meaningful

00:06:49,128 --> 00:06:54,310

impact, not just on clinical care, but on hospital operations in lots of different ways.

00:06:54,411 --> 00:07:05,720

But also for the vendors to try and work out, how do I position my solution, my AI widget

against this landscape?

00:07:05,720 --> 00:07:07,251

And how do I differentiate?

00:07:07,251 --> 00:07:14,717

How do I deal with the fears, uncertainties and doubts, the objections that are going to

come up from the other side?

00:07:14,717 --> 00:07:19,412

How do I navigate that space to be able to be successful with it?

00:07:19,412 --> 00:07:26,498

Well and it's difficult, because I think you've got people on two, like all things in

life, two very extreme points of view.

00:07:26,498 --> 00:07:32,513

You've got the people on one side who are saying, AI is going to replace doctors.

00:07:32,513 --> 00:07:37,007

We're not going to need any nurses in 20 years because everything will be done by a robot.

100

00:07:37,007 --> 00:07:46,515

And we will just trust computers because they'll be able to diagnose because they have

access to all of this information that, you know, a regular human could not contain in

101

00:07:46,515 --> 00:07:47,636

their brain.

102

00:07:47,698 --> 00:07:56,144

And then you have people on the other side going, well, no, no, no, I want to do

everything manually because then it's me touching it and I know what's happening and I

103

00:07:56,144 --> 00:08:00,176

don't trust AI and I don't want AI and it's absolutely not.

104

00:08:00,457 --> 00:08:10,724

And as with all things in life, there are the extreme points of view are fine, but the

truth and where things, where you're really going to find success is somewhere in the

105

00:08:10,724 --> 00:08:11,404

middle.

106

00:08:11,404 --> 00:08:17,390

We're not going to replace doctors anytime soon, but we're also not going to sit on the

sidelines and not

107

00:08:17,390 --> 00:08:21,690

let advances in technology really help us in our life and in our work.

108

00:08:21,690 --> 00:08:26,611

If AI can help to save someone's life, we need to be using it.

109

00:08:26,611 --> 00:08:28,323

yeah.

110

00:08:28,323 --> 00:08:40,704

And I think that it is kind of interesting because it's one of the things that we might

touch on a little bit more in depth in a later episode is there's an analogy that's used

111

00:08:40,704 --> 00:08:54,013

and it's actually been, it's existed for some time and it's been used in the science

fiction context as well, which is the centaur cyborg balance whereby, know, fully,

112

00:08:54,013 --> 00:08:59,955

handing over to AI is equated to being like a cyborg.

113

00:08:59,996 --> 00:09:07,319

The AI is so integrated, it's taking over elements of your brain function, not just

helping empower you.

114

00:09:07,319 --> 00:09:18,234

Whereas the Centaur model is that analogy of the two components working together, but the

human part of it still being very much in charge.

115

00:09:18,234 --> 00:09:21,385

So that being the torso and head and brain.

116

00:09:21,546 --> 00:09:26,415

but taking advantage of the power of the lower part being the horse part.

117

00:09:26,415 --> 00:09:33,380

So that's a very interesting analogy that we might explore a little bit more deeply in

another episode.

118

00:09:33,380 --> 00:09:35,883

So we just need horse bodies now.

119

00:09:36,608 --> 00:09:49,609

Well, if you look at some of this stuff in terms of the mechanized robots that are now

coming, I mean, you know, how far are we away from having mechs like, you know, Sigourney

120

00:09:49,609 --> 00:09:53,609

Weaver in Alien or the many other science fiction films that have been there?

121

00:09:53,609 --> 00:09:55,769

The answer is not far.

122

00:09:55,949 --> 00:10:05,371

I do woodworking and there are there's a company, German company called Festool, who

already have automated powered

123

00:10:05,371 --> 00:10:07,952

like arm limbs to assist.

124

00:10:07,952 --> 00:10:13,515

And they're designed for people who do a lot of construction work with their arms up like

this.

125

00:10:13,515 --> 00:10:25,802

You're up on high places like ceilings and doing light fittings and doing wiring runs And

the whole idea is that instead of taking their hands all the way down and shaking to get

126

00:10:25,802 --> 00:10:34,271

blood back into the muscles, et etcetera, they're able to just relax them so the muscles

aren't so tense that the blood can flow into them.

127

00:10:34,674 --> 00:10:36,394

But that technology is available.

128

00:10:36,394 --> 00:10:44,415

You buy that for about two grand to be able to sort of like do it and it's powered by

plug-in batteries just like your Dewalt drill.

129

00:10:44,415 --> 00:10:46,775

And so that kind of technology is already there.

130

00:10:46,775 --> 00:10:49,675

So the Centaur model is not that far away.

131

00:10:49,874 --> 00:10:57,735

The only thing is I don't think there's AI in there that's deciding which direction your

arm should move and what you should do next.

132

00:10:57,735 --> 00:11:00,277

Because that truly would start to become like.

133

00:11:00,277 --> 00:11:12,092

cyborg arms, you know, that are totally robotic, etcetera And then the AI takes over and

then, yeah, and then Skynet occurs and then the Terminator.

134

00:11:12,092 --> 00:11:13,552

Yes, I know.

135

00:11:13,613 --> 00:11:20,130

the, end of humanity, oh

136

00:11:20,130 --> 00:11:20,510

right?

137

00:11:20,510 --> 00:11:25,901

AI is definitely, as you said, in the kind of upper bit of the hype cycle.

138

00:11:25,901 --> 00:11:28,332

I don't think we're at the very tippy top yet.

139

00:11:28,332 --> 00:11:30,232

I think there is more to come.

140

00:11:30,473 --> 00:11:38,955

But I think suddenly everyone is kind of looking at uh artificial intelligence and saying,

okay, what do we do with this?

141

00:11:38,955 --> 00:11:40,316

Okay, we know that it's useful.

142

00:11:40,316 --> 00:11:43,637

We know that there are some incredible use cases in healthcare.

143

00:11:43,637 --> 00:11:45,297

I mean, if you walk around,

144

00:11:45,297 --> 00:11:53,931

any of the big conferences like HLTH and Vive and HIMSS And you go and talk to all of

these little vendors who are doing amazing things, whether it's, you know, they're

145

00:11:53,931 --> 00:12:00,964

automating and using AI on one little piece of the workflow, one little piece of clinical

decision support.

146

00:12:01,044 --> 00:12:02,504

It's incredible.

147

00:12:02,565 --> 00:12:04,115

And if we could just...

148

00:12:05,705 --> 00:12:13,681

bundle them all up and pop them in a hospital for a good amount of money, it would

completely revolutionize how...

149

00:12:13,681 --> 00:12:15,522

how medicine is delivered.

150

00:12:16,082 --> 00:12:17,765

But we can't do that.

151

00:12:17,765 --> 00:12:24,809

It's not realistic from a financial perspective, from a change management perspective,

from so many perspectives.

152

00:12:24,809 --> 00:12:26,956

So I think today...

153

00:12:26,956 --> 00:12:44,176

know, one of the big issues is a lot of hospitals are risk averse because, you know, and

people always equate to the aviation industry where if a mistake happens, know, 300 people

154

00:12:44,176 --> 00:12:46,416

will die, et etcetera, et etcetera, et etcetera.

155

00:12:46,436 --> 00:12:49,128

So there are lots of controls there.

156

00:12:49,324 --> 00:12:57,312

The challenge with medicine is it's infinitely more complex from a decision-making and

data points perspective as to what you're doing.

157

00:12:57,312 --> 00:13:07,482

And so actually, the potential for an error having a fatal outcome um are infinitely

larger.

158

00:13:07,743 --> 00:13:09,584

And they may only affect that one patient.

159

00:13:09,584 --> 00:13:13,788

Or actually, they may affect a whole cadre of patients because

160

00:13:13,808 --> 00:13:18,612

you've programmed something wrong and it isn't spotted until it's affected several hundred

patients.

161

00:13:18,612 --> 00:13:24,117

So I think that that's that's that health care is very, very risk averse.

162

00:13:24,117 --> 00:13:37,571

And so therefore, when you are looking at how do you address that, their tolerance for

potential error is significantly lower than it might be in other places.

163

00:13:37,571 --> 00:13:40,319

And so I think that's that's one of the things that

164

00:13:40,319 --> 00:13:52,275

that as you're thinking about how you position, particularly if you're trying to bring a

technology that has worked very well in another industry and apply it to healthcare, um

165

00:13:52,275 --> 00:13:53,365

it's not that simple.

166

00:13:53,365 --> 00:13:57,937

And we'll probably cover that a little bit more as we go through this conversation.

167

00:13:58,417 --> 00:13:59,679

Yeah, well, let's dive into that.

168

00:13:59,679 --> 00:14:03,553

Let's look first at the hospital buyer's lens.

169

00:14:03,654 --> 00:14:08,580

And we know that that is not buyer apostrophe S, that is buyer S apostrophe.

170

00:14:08,580 --> 00:14:10,082

There is not one buyer.

171

00:14:10,082 --> 00:14:13,685

There is a whole army of buyers.

172

00:14:13,685 --> 00:14:19,539

methodologies that says, oh, you only need to go and talk to the guy who signs the check

or signs the approval.

173

00:14:19,999 --> 00:14:24,482

That's not how hospital and health systems work.

174

00:14:24,503 --> 00:14:31,979

They are huge branches of committees that actually you have to navigate your way through.

175

00:14:52,769 --> 00:14:53,399

Absolutely.

176

00:14:53,399 --> 00:15:01,424

And you actually posted a great sub stack about this in the last couple of weeks about the

Knights who say no.

177

00:15:01,644 --> 00:15:04,476

And we'll put a link in the description here.

178

00:15:04,476 --> 00:15:08,990

Everyone loves Monty Python, but it really was.

179

00:15:08,990 --> 00:15:14,826

It's, and it's, you know, saying no sounds so defeatist and definite, but

180

00:15:15,040 --> 00:15:18,261

there's a lot of reasons why somebody might say no, right?

181

00:15:18,261 --> 00:15:27,394

So there is the financial aspect, there's clinical, there's compliance, there's privacy,

there's security, there's workflow, there's change management, there's, I just don't want

182

00:15:27,394 --> 00:15:30,424

to do it now because we have too many other things going on.

183

00:15:30,424 --> 00:15:33,425

There are a lot of reasons why people say no.

184

00:15:34,225 --> 00:15:34,906

Exactly.

185

00:15:34,906 --> 00:15:43,694

So there are so many reasons, but I think, dad, what do you think a hospital leader, when

they are being pitched an AI,

186

00:15:43,894 --> 00:15:50,613

solution from either a new company or an existing partner, what are they thinking about?

187

00:15:50,613 --> 00:15:53,726

What do you think is top of mind for that hospital buyer?

188

00:15:54,334 --> 00:16:02,516

I think it is that kind of risk reward analysis that says, well, okay, so this sounds very

cool.

189

00:16:02,516 --> 00:16:04,377

know, what you're pitching me sounds great.

190

00:16:04,377 --> 00:16:05,797

It sounds awesome.

191

00:16:05,857 --> 00:16:12,559

But, you know, one, what's the risk that I'm introducing with this tool?

192

00:16:12,559 --> 00:16:17,080

You know, how is it going to change things?

193

00:16:17,841 --> 00:16:23,332

Secondly, you're telling me it will have this impact.

194

00:16:23,559 --> 00:16:27,939

But have you got any proof that says that'll definitely happen?

195

00:16:28,319 --> 00:16:44,719

There's an article I wrote, which was a kind of post-mortem of Olive Health, a company

that was a super rock star, incredibly well-funded unicorn that in the space of about

196

00:16:44,719 --> 00:16:51,238

three years rattled through a billion dollars

197

00:16:51,238 --> 00:16:52,222

promising stuff

198

00:16:52,323 --> 00:16:54,014

and then not delivering it.

199

00:16:54,034 --> 00:17:01,697

when you look at that as a case study oh of the kind of promise that was made, it was very

frothy.

200

00:17:01,697 --> 00:17:06,559

A lot of people were very interested in what Olive Health were pitching.

201

00:17:06,820 --> 00:17:13,202

But the challenge was there was nothing on the other side that actually resembled what

they were pitching.

202

00:17:13,502 --> 00:17:20,445

They had acquired some other legacy technology along the way, some robotic process

automation type stuff.

203

00:17:20,869 --> 00:17:33,696

and they were selling that, but then they were trying to pitch it as if it was, you know,

some very cool state of the art, you know, AI driven workflow stuff.

204

00:17:33,977 --> 00:17:42,501

So when you look at that as a context, there's lots and lots of experience of these

hospital leaders whereby they've been made promises before.

205

00:17:42,501 --> 00:17:49,145

So they're pretty skeptical when it comes to stuff that has gone promising, a big ROI.

206

00:17:49,439 --> 00:17:52,471

that is in cash dollar terms.

207

00:17:52,471 --> 00:18:00,535

Even more skeptical when it's something that actually is delivering intangible benefits,

i.e.

208

00:18:00,535 --> 00:18:04,487

well, this will actually save our doctors three hours a week.

209

00:18:06,476 --> 00:18:08,868

Okay, so each doctor is gonna save three hours a week.

210

00:18:08,868 --> 00:18:10,128

What are they gonna do with that time?

211

00:18:10,128 --> 00:18:18,964

Because believe me, I have enough trouble hiring doctors, right, and filling the vacancies

I have to lay off doctors.

212

00:18:18,964 --> 00:18:20,896

That's not even on the table.

213

00:18:20,896 --> 00:18:29,402

So you saving me three hours a week, that's great for them, but it doesn't have an impact

directly on the way the hospital works.

214

00:18:29,402 --> 00:18:34,751

And if I'm in a non-US marketplace, actually,

215

00:18:34,751 --> 00:18:43,458

the demand driven side, if you're state funded by the government, the government will just

expect you to, well, okay, treat some more people that are sat on the waiting list,

216

00:18:43,458 --> 00:18:44,629

shuffle them through.

217

00:18:44,629 --> 00:18:46,751

So I think that that's on that dynamic.

218

00:18:46,751 --> 00:19:00,072

oh On the other side, they'll also be looking at what is gonna be the impact on this and

is this gonna create, let's say it's very successful and well adopted, is that then gonna

219

00:19:00,072 --> 00:19:03,361

drive up the budget for the use of this?

220

00:19:03,361 --> 00:19:16,301

And that's one of the problems you're seeing with a lot of the generative AI stuff is that

suddenly the unit cost of this is making the cloud budget go up because suddenly you're

221

00:19:16,301 --> 00:19:23,621

accessing a lot more data, you're storing a lot more data, you're having to do all of

these other things, you're having to apply other PCs to try and make it safer.

222

00:19:23,621 --> 00:19:29,537

So as you introduce more technology, suddenly your infrastructure cost goes up and

you're...

223

00:19:29,537 --> 00:19:33,997

Utilization cost goes up because the doctors are suddenly going, oh, I like I quite like

this.

224

00:19:33,997 --> 00:19:34,997

This is quite good.

225

00:19:34,997 --> 00:19:48,077

And so I think that there's going to end up being some assessment by a lot of these

hospital leaders of well, who do we allow to use this and how much more because it's got a

226

00:19:48,077 --> 00:19:58,933

it's got a downside whereby even if it's got an ROI, is that ROI to get there going to

actually cost me more because

227

00:19:58,933 --> 00:20:03,367

to get there, we're going to have to consume more in terms of infrastructure to get there.

228

00:20:03,367 --> 00:20:05,893

So total cost of ownership.

229

00:20:05,997 --> 00:20:13,472

that's the big secret, the secret that I think everybody knows is all the big cloud

platforms that exist today.

230

00:20:13,472 --> 00:20:17,083

And I've sold in partnership with most of them.

231

00:20:17,184 --> 00:20:19,315

Definitely AWS, Azure and Google.

232

00:20:19,315 --> 00:20:25,088

I've sold in partnership with them, you know, side by side, buy my thing.

233

00:20:26,309 --> 00:20:29,391

All they want is consumption, cloud consumption.

234

00:20:29,478 --> 00:20:31,568

Is this going to drive cloud usage?

235

00:20:31,568 --> 00:20:35,563

Are they going to use more AWS servers or Google servers or Azure?

236

00:20:35,563 --> 00:20:37,466

That's all that they care about.

237

00:20:38,230 --> 00:20:39,151

Uh-huh.

238

00:20:39,424 --> 00:20:41,867

% in one day?

239

00:20:41,887 --> 00:20:54,252

It's because they came out with some predictions, predictions, not results, predictions of

what the growth of their cloud consumption was going to be driven by AI adoption.

240

00:20:54,392 --> 00:21:01,575

And that was enough to make investors go, oh, man, this is awesome.

241

00:21:02,015 --> 00:21:03,676

So that

242

00:21:03,868 --> 00:21:15,556

increased consumption is going to have to be paid by somebody and hospitals don't have the

ability to just go and create an inelastic revenue.

243

00:21:15,556 --> 00:21:18,097

They don't have the ability to grow that.

244

00:21:18,098 --> 00:21:25,323

They typically are kind of relatively stuck in terms of how much revenue they can have.

245

00:21:25,323 --> 00:21:30,726

And the government and the private insurers are all trying to push that down.

246

00:21:30,726 --> 00:21:33,844

So therefore you've got a challenge whereby

247

00:21:33,844 --> 00:21:42,842

To be able to address more of this and consume more of this means that you're dealing with

an inelastic budget on the income side.

248

00:21:42,842 --> 00:21:44,193

that's the other.

249

00:21:44,554 --> 00:21:55,664

Again, we're back to talking macroeconomics here, but we're looking at this from the lens

of a hospital buyer and the hospital buyer as a CFO is going to be looking that way.

250

00:21:57,370 --> 00:22:00,602

That can't keep growing exponentially.

251

00:22:00,602 --> 00:22:14,301

So I think if you, and this is kind of veering more into our next, well, it's actually,

it's a very nice natural transition into the seller's lens of if you're already in a

252

00:22:14,301 --> 00:22:23,868

hospital, in a health system, you are going to be scrutinized in the next 12 months

because suddenly someone up there, one of those Knights or maybe one of those Knights

253

00:22:23,868 --> 00:22:25,939

bosses is going to go hold up.

254

00:22:27,247 --> 00:22:28,127

What does this do?

255

00:22:28,127 --> 00:22:29,108

What are the results?

256

00:22:29,108 --> 00:22:30,539

What is actually happening here?

257

00:22:30,539 --> 00:22:33,930

Why has our cloud consumption cost gone up?

258

00:22:35,091 --> 00:22:36,342

What did we get out of that?

259

00:22:36,342 --> 00:22:44,557

So if you're existing, you're going to have to really prove your worth in the next few

months as your renewals are coming up.

260

00:22:44,557 --> 00:22:56,741

On the flip side, if you are going in, it is going to become significantly harder because

people have suddenly realized like, oh, this is, know, patient data.

261

00:22:56,741 --> 00:22:58,452

This is financial data.

262

00:22:58,452 --> 00:23:09,259

Maybe we shouldn't just be trusting all these little companies that have only just popped

up in the last few months in some cases.

263

00:23:10,821 --> 00:23:12,022

Should we really trust them?

264

00:23:12,022 --> 00:23:19,887

So I do think there's gonna be very much a pushback and it's gonna start with the big

health system.

265

00:23:19,887 --> 00:23:20,988

It always does, right?

266

00:23:20,988 --> 00:23:25,340

Let alone the poor rural hospitals barely even have one AI agent.

267

00:23:25,340 --> 00:23:26,661

exactly, yeah.

268

00:23:26,661 --> 00:23:29,692

uh And I think that that's the thing.

269

00:23:29,692 --> 00:23:31,523

There is a big disparity as well.

270

00:23:31,523 --> 00:23:47,869

mean, in the article I talk, and by the way, the article that I wrote, the Knights who say

no, we, you know, again, Monty Python The reason that I wrote that was I was inspired by

271

00:23:47,869 --> 00:23:54,547

another article, LinkedIn post, by a CEO of a startup actually in Canada.

272

00:23:54,547 --> 00:24:09,657

who was addressing exactly this issue of how do you create the link between value that

you're delivering and the solution that you're bringing and being able to argue that case

273

00:24:09,657 --> 00:24:12,049

with hospital leadership and on from there.

274

00:24:12,049 --> 00:24:19,033

And I wrote the article because he gave a couple of examples of the obstacles that are

there.

275

00:24:19,033 --> 00:24:23,708

And I said, having worked, because I think, you know,

276

00:24:23,708 --> 00:24:31,133

those of you who know me, I worked for a long time on the vendor side of things and I now

work for a large health system.

277

00:24:31,754 --> 00:24:38,779

And now that I worked inside the beast, I now understand so much more.

278

00:24:38,779 --> 00:24:42,321

I always knew, I was always quite a strategic seller.

279

00:24:42,321 --> 00:24:44,913

And so I knew that there were these committees and things.

280

00:24:44,913 --> 00:24:52,109

I had no idea how Byzantine and how bureaucratic ah some of this was.

281

00:24:52,109 --> 00:25:02,907

If you look at it and you've got a doctor who's very excited about your product, let's

just take a scenario, and they say, I'm gonna go to procurement and tell them I really

282

00:25:02,907 --> 00:25:04,318

need to have this.

283

00:25:05,059 --> 00:25:06,699

Here's what will happen.

284

00:25:07,120 --> 00:25:16,066

Procurement will have a bunch of their stuff that they want you to jump through and you'll

go, that's not too bad, there's a list of things, et etcetera.

285

00:25:16,207 --> 00:25:19,779

Then procurement will defer.

286

00:25:20,229 --> 00:25:30,329

they'll send it off to IT to the enterprise architects and say, hey, can you check some

boxes and make sure this fits with our general architectural direction?

287

00:25:30,709 --> 00:25:42,109

And then the enterprise architects will get into it and they'll go, oh, well, and the

enterprise architects will say, has this passed through CISO, which is the chief

288

00:25:42,109 --> 00:25:45,969

information security officer in whichever set of committees they've got?

289

00:25:45,969 --> 00:25:48,934

And then suddenly it needs to go there and then they will go.

290

00:25:48,934 --> 00:25:54,994

Well, CISO will go and assess it for information security and say, has this been through

compliance?

291

00:25:55,354 --> 00:26:01,754

And then the compliance group, the legal compliance group will get involved in terms of

looking at what's the impact of this and so on and so forth.

292

00:26:01,754 --> 00:26:04,294

And they will have a whole set of checks and balances.

293

00:26:04,294 --> 00:26:10,314

And then they will defer and they'll say, well, has the clinical body approved this?

294

00:26:10,314 --> 00:26:17,123

And they will send it off to something that has been around for decades now, which is

called the

295

00:26:17,123 --> 00:26:26,988

the Pharmaceuticals and Therapeutics Committee, the P &T committee is normally what it's

shortened down to, which is a group of doctors who also sit there and go, well, you we

296

00:26:26,988 --> 00:26:34,182

have to consider keeping a balanced uh viewpoint of what is the standard of care here in

this organization.

297

00:26:34,182 --> 00:26:43,517

And they will act as a kind of regulator to say, well, we're going to balance this out and

say, well, this is an approved practice and this is an approved medication and this is an

298

00:26:43,517 --> 00:26:44,908

approved implant.

299

00:26:44,908 --> 00:26:47,019

And so they will come up with their set.

300

00:26:47,139 --> 00:27:02,353

And then if it's something novel that they haven't come across, like AI, then they're

gonna go, well, we need to do a uh study and actually assess what the potential clinical

301

00:27:02,353 --> 00:27:04,534

benefit versus risk is here.

302

00:27:04,554 --> 00:27:10,516

So we're gonna send it to institutional review board and suddenly you've got a new

committee involved.

303

00:27:10,516 --> 00:27:12,876

And IRB are then gonna go through.

304

00:27:12,996 --> 00:27:15,937

The important thing about those last, you may have thought,

305

00:27:15,937 --> 00:27:24,780

everybody up until that point was fairly bureaucratic and you know, sort of more than my

job is worth to let something get purchased.

306

00:27:24,780 --> 00:27:41,144

The thing that is fascinating about the politics inside P &T committees and the uh IRB

sort of like oversight board is they typically are staffed by some of the longest serving

307

00:27:41,144 --> 00:27:43,758

and therefore most conservative

308

00:27:43,758 --> 00:27:50,781

and then most Luddite, if you want to take a point of view on it, of the clinicians that

are in the hospital.

309

00:27:50,781 --> 00:28:00,425

So you have this great champion and he, you know, this doctor is up against the machine.

310

00:28:00,425 --> 00:28:05,847

It's not just you, it's him as well saying, here's a technology that I want to move

forward.

311

00:28:05,847 --> 00:28:07,328

And he could be quite senior.

312

00:28:07,328 --> 00:28:10,309

He could be, I always remember,

313

00:28:10,345 --> 00:28:17,588

being told you should look for who is the clinician that generates the most revenue for

the hospital.

314

00:28:18,048 --> 00:28:29,839

And it was always like, you know, was all the cardiovascular surgeons or the transplant

surgeons or it was the people that were making, generating big money and that depended on

315

00:28:29,839 --> 00:28:31,313

that type of procedures.

316

00:28:31,354 --> 00:28:36,536

Yeah, yeah, you're struggling if they are employed inside a health system.

317

00:28:36,596 --> 00:28:39,657

So I think that my point is

318

00:28:39,993 --> 00:28:54,105

that when you are looking, and this particularly applies to early stage and startup

organizations, you've got to not be naive about what that political mine field is that

319

00:28:54,105 --> 00:28:56,066

you're stumbling into.

320

00:28:56,206 --> 00:29:06,273

And just because you think you figured out at Health System A, it could be very different,

not just because of the personalities involved.

321

00:29:06,273 --> 00:29:08,043

but they could be structured differently.

322

00:29:08,043 --> 00:29:12,515

They could have some little side committees that are actually involved as well.

323

00:29:12,515 --> 00:29:23,418

So mapping that out early and understanding it is very important because you can't expect

your champion doctor because their day job is caring for patients.

324

00:29:23,598 --> 00:29:33,400

And even if they've been in an organization for 10 years, if this is the first product

that they've sponsored through saying, I'd really like us to adopt this, they may be naive

325

00:29:33,400 --> 00:29:35,572

about all of this stuff because it's not

326

00:29:35,572 --> 00:29:39,872

typically transparent to people who are working on the frontline.

327

00:29:39,872 --> 00:29:41,097

328

00:29:41,097 --> 00:29:50,741

When I think as a seller, what that teaches you or what that should guide you to do is you

have to not only find the right champion and map out all of those relationship maps and

329

00:29:50,741 --> 00:29:53,362

the decision trees and things that we've been doing for years, right?

330

00:29:53,362 --> 00:29:55,593

I've done that my whole career.

331

00:29:55,934 --> 00:30:00,797

But what's different now is you really have to be an educator.

332

00:30:00,797 --> 00:30:02,168

So you got to sell, right?

333

00:30:02,168 --> 00:30:05,839

And turn up and do the song and dance and show off your nice shiny demo.

334

00:30:06,341 --> 00:30:11,453

but you've got to explain it in such a way that your champion can go and explain it.

335

00:30:11,453 --> 00:30:25,579

And if you get that little five minute coffee meeting with someone who's on one of these

boards or you run into them in the hallway, you've got to know, I have 30 seconds to say

336

00:30:25,579 --> 00:30:27,580

my most important things.

337

00:30:27,660 --> 00:30:34,483

And if they're approving everything from supply chain workflow to new implants to...

338

00:30:34,947 --> 00:30:42,831

I don't know, the lawn care company that the hospital uses, they are not experts in any of

those fields necessarily.

339

00:30:42,892 --> 00:30:45,113

So they may not be an expert in AI.

340

00:30:45,113 --> 00:30:48,977

So it is your job to educate what's really going on.

341

00:30:48,977 --> 00:30:53,860

And I always have been very transparent in my career about the risks, right?

342

00:30:53,880 --> 00:30:58,903

You have to explain not what can go wrong with your software, that sounds...

343

00:30:59,667 --> 00:31:04,431

It sounds like you don't have good software, in which case you should probably find a new

job and a new company.

344

00:31:04,431 --> 00:31:09,154

But if your software works, but you do need to explain limitations, right?

345

00:31:09,154 --> 00:31:17,820

There are, I have been in so many sales meetings where, where sellers will just sort of

bend the truth and be like, yeah, I could do that.

346

00:31:17,820 --> 00:31:24,725

No, you have to be just so transparent about what your product does, what its impact is.

347

00:31:24,725 --> 00:31:27,947

You need to know the technical details.

348

00:31:28,299 --> 00:31:33,226

It is not okay to just rely on your solutions architect or your solutions consultant.

349

00:31:33,226 --> 00:31:43,475

You really have to be an expert enough that you can confidently explain your solution and

the value that it brings.

350

00:32:06,400 --> 00:32:09,621

I provide coaching to product managers now.

351

00:32:09,881 --> 00:32:12,161

It's a part of my new life.

352

00:32:12,161 --> 00:32:16,943

know, as I used to be, I managed and cracked the Whip over sales teams.

353

00:32:16,943 --> 00:32:27,207

But actually, know, know, I've uh studied with Pragmatic Institute, a well-known training

and consulting company in the product management world.

354

00:32:27,388 --> 00:32:28,948

But one of the things that...

355

00:32:28,948 --> 00:32:39,093

that they often talk about, and I know salespeople consume these, even though product

management might come up with them, is kind of ideal user profiles, right?

356

00:32:39,093 --> 00:32:46,996

Whereby you basically have a description of, well, this is a typical buyer of this type.

357

00:32:46,996 --> 00:32:54,959

And so therefore, as part of your sales enablement training that you go through for a new

product launch or something, they'll give you this archetype.

358

00:32:55,139 --> 00:32:56,721

I'm here to tell you.

359

00:32:56,721 --> 00:33:03,566

It's as useful as a two tone PowerPoint template.

360

00:33:03,566 --> 00:33:14,213

It is not going to help you in terms of actually addressing, yeah, it can give some clues

and it's useful from a training perspective to try and get there.

361

00:33:14,273 --> 00:33:25,811

But I think one of the things that you have to do, and this is an opportunity to use AI to

your advantage, is there are ways and techniques that you can actually build

362

00:33:25,811 --> 00:33:32,426

much more specific ideal user profiles for the type of buyers you'll encounter.

363

00:33:32,426 --> 00:33:34,588

So think of all those committees I described.

364

00:33:34,588 --> 00:33:39,711

They're all staffed by archetypes who are actually real people.

365

00:33:39,892 --> 00:33:52,241

So as you are researching an organization, if the campaign is valuable enough to you to

put in this effort, you should actually go and look and pull their

366

00:33:52,636 --> 00:34:05,851

LinkedIn profiles, their resume if you can find it, if they're a clinician you typically

can find clinicians CVs out there and pull all that information in and then use the

367

00:34:05,851 --> 00:34:11,693

instructions that might be used and maybe this is something that you talk to your product

management leaders about.

368

00:34:11,874 --> 00:34:20,137

Use the information you've got from your archetypes and then apply it against those real

people in the organization and say

369

00:34:20,137 --> 00:34:27,161

Hey chat GPT slash Claude slash Comet slash, which you know, name your GPT of choice.

370

00:34:27,422 --> 00:34:32,315

Here's a prompt that actually says, you know, I'm trying to create this and this is what

I'm trying to do.

371

00:34:32,315 --> 00:34:33,866

This is who I'm selling to.

372

00:34:33,866 --> 00:34:50,131

This is their profile, upload document and actually say, help me tune my messaging about

my product that I just told you about to be able to talk to and address

373

00:34:50,131 --> 00:34:52,883

potential objections this person may come up with.

374

00:34:52,883 --> 00:35:06,273

So have those talking points ready before you ever go and set foot in the organization so

that you can uh be much more accurate in terms of what you're trying to do rather than

375

00:35:06,273 --> 00:35:12,318

just relying on, well, I got taught these 15 uh objection responses.

376

00:35:12,318 --> 00:35:19,911

Charlie says that he's gonna be worried about this and analytics Anna, Anna is going to be

concerned about this.

377

00:35:19,911 --> 00:35:21,913

Yeah.

378

00:35:21,913 --> 00:35:33,121

think that's something that people, where you can actually use AI in your work life to try

and be much more successful and powerful as soon as you are ready to have that interaction

379

00:35:33,121 --> 00:35:33,712

with that person.

380

00:35:33,712 --> 00:35:40,837

So dad, in the next 10 years, obviously we don't know which, we can't predict which

companies are actually gonna blow up, right?

381

00:35:40,837 --> 00:35:44,570

Because we would buy their stock and we would not be doing this podcast.

382

00:35:44,570 --> 00:35:47,742

We would be on a island sipping some drinks.

383

00:35:47,742 --> 00:35:53,626

um But we can talk about what we think the overarching trends are gonna be.

384

00:35:53,932 --> 00:36:00,673

And I mean, off the bat, I've got to imagine that in the next, I'm thinking even six

months as the...

385

00:36:00,673 --> 00:36:09,529

as the new kind of administration gets into place and really starts to dive down into

health, that they've already said a lot of stuff about interoperability and all these

386

00:36:09,529 --> 00:36:10,299

different pieces.

387

00:36:10,299 --> 00:36:18,024

I've got to imagine that an AI regulatory crackdown is going to come at some point.

388

00:36:18,024 --> 00:36:24,619

And you can even look, the NHS posted an incredible guideline about ambient scribing.

389

00:36:24,619 --> 00:36:28,061

So very, very particular AI use case.

390

00:36:28,270 --> 00:36:37,246

And if, but if you read it, it can be generalized for all things artificial intelligence,

because it's looking at things like, where is that data coming from?

391

00:36:37,246 --> 00:36:38,447

Where is it being stored?

392

00:36:38,447 --> 00:36:40,517

What is happening to that data?

393

00:36:40,557 --> 00:36:44,541

All the things that AI does, cause it's got to whoop up to the cloud and whoop back down.

394

00:36:44,541 --> 00:36:46,722

Look at that NHS guideline.

395

00:36:46,722 --> 00:36:52,815

If you are either in a hospital and you're trying to figure out what's going to happen, or

if you are a seller.

396

00:36:52,955 --> 00:36:57,507

Those are the kind of questions that you're going to be getting asked very, very soon.

397

00:36:57,528 --> 00:37:03,608

It's a wonderful, wonderful place to start if you are a AI skeptic or an AI lover.

398

00:37:03,608 --> 00:37:06,270

It's just a really great guideline that they pull together.

399

00:37:06,270 --> 00:37:07,030

Yeah.

400

00:37:07,030 --> 00:37:17,516

And I think there's going to be more of those, you for all of the kind of reasons I

touched on earlier on um that obviously at a federal level, there's various types of

401

00:37:17,516 --> 00:37:20,278

regulations slash enablement coming in.

402

00:37:20,278 --> 00:37:24,620

Even sort of like changes that Dr.

403

00:37:24,620 --> 00:37:31,265

Oz and the administration, you know, are being talking about in terms of health and human

services.

404

00:37:31,265 --> 00:37:33,015

They'd be talking about

405

00:37:33,121 --> 00:37:40,808

trying to mandate automating and making smoother, quicker, easier prior authorization.

406

00:37:40,808 --> 00:37:42,189

That's gonna happen at that level.

407

00:37:42,189 --> 00:37:50,486

Now I have a whole bunch of opinions about that, Bethany and I were chatting about earlier

on and we have two very different viewpoints on it, which is exactly, and both of them are

408

00:37:50,486 --> 00:37:54,979

exactly right, but we'll maybe save that for another podcast.

409

00:37:55,400 --> 00:37:57,802

But I think that you're gonna have that level of regulation.

410

00:37:57,802 --> 00:37:59,693

But then you're gonna have,

411

00:38:00,045 --> 00:38:02,090

the state level regulation.

412

00:38:02,285 --> 00:38:03,641

California already does.

413

00:38:03,641 --> 00:38:04,411

Yep.

414

00:38:04,772 --> 00:38:12,619

And then you're going to have the health system level, the knights who say no, you know,

that level of compliance management.

415

00:38:12,619 --> 00:38:18,243

And I think that um I'm already working with our enterprise architecture team.

416

00:38:18,244 --> 00:38:26,190

And one of the things that we're looking at is, you know, how do we police against our

architecture and so on and so forth?

417

00:38:26,271 --> 00:38:31,539

It's it is chaos even for an organization as large as the one I work for.

418

00:38:31,539 --> 00:38:34,841

and that has a lot of resources available to it.

419

00:38:34,861 --> 00:38:49,833

They have got uh a real challenge because if you take the example Bethany referenced

earlier on, from 80 AI widgets to 200 in the space of seven months, how do you validate

420

00:38:49,833 --> 00:38:51,554

where all that data is flowing?

421

00:38:51,554 --> 00:38:59,629

How do you validate that the models that are underpinning some of these, because

understand a lot of these AI widgets are built on

422

00:38:59,829 --> 00:39:09,807

another company's model and widget and then had some secret sauce applied to them to make

it relevant to that workflow.

423

00:39:10,248 --> 00:39:23,538

And so when you look at that, you're not even dealing with the end part of the machine

that is doing a lot of that calculation and that thought process and machine learning on

424

00:39:23,538 --> 00:39:26,041

your data to be able to apply it.

425

00:39:26,041 --> 00:39:29,168

So what do you do from

426

00:39:29,168 --> 00:39:40,471

a governance and compliance perspective to try and track and look at all of these models

that are suddenly in your ecosystem and how do you manage them?

427

00:39:40,471 --> 00:39:47,422

How are you gonna deal with all those data flows that are suddenly seeping out of your

organization?

428

00:39:47,422 --> 00:39:58,715

And I gotta tell you that the organization I'm part of is very, very concerned about that

because they are very proud.

429

00:39:58,951 --> 00:40:08,314

of their data, they know that the history of their data is worth an enormous amount to

them from a research perspective, from an operating perspective, etc.

430

00:40:08,314 --> 00:40:22,898

So they're very concerned about data leaking out through some of these side doors to be

able to go just because it's being consumed by one of these AI widgets.

431

00:40:22,898 --> 00:40:27,695

And I think this is the landscape that is going to dominate

432

00:40:27,695 --> 00:40:31,336

the conversation over the next 12 months.

433

00:40:31,336 --> 00:40:37,981

I used to sell or I worked for a company that sold governance risk and compliance

solutions.

434

00:40:38,642 --> 00:40:43,588

And I think when we're talking about AI, everyone's thinking about the sexy AI, right?

435

00:40:43,588 --> 00:40:56,639

Like the sexy AI where you can speak and like the words appear magical or you can get a uh

imaging done and it can detect within seconds if you're having a stroke.

436

00:40:56,639 --> 00:40:57,814

Amazing.

437

00:40:57,814 --> 00:40:58,685

super sexy.

438

00:40:58,685 --> 00:41:06,344

If it can save my life versus waiting for someone to look at it and you can give you that

medicine and you save my life with AI, incredible.

439

00:41:06,344 --> 00:41:07,937

And those are great solutions.

440

00:41:07,937 --> 00:41:18,897

But until I was selling GRC solutions, I didn't realize the Unsexy software is very, very

powerful in a health system.

441

00:41:18,897 --> 00:41:22,980

And they spend a lot of money on the Unsexy software.

442

00:41:22,980 --> 00:41:25,122

You know, I was in the interoperability space.

443

00:41:25,122 --> 00:41:27,084

Nobody thinks about integration engines.

444

00:41:27,084 --> 00:41:29,125

It's like the pipes in your wall.

445

00:41:29,246 --> 00:41:30,156

My pipes work.

446

00:41:30,156 --> 00:41:31,748

I don't know who made my pipes.

447

00:41:31,748 --> 00:41:32,909

They bring me water.

448

00:41:32,909 --> 00:41:35,616

I know that I have a nice faucet and I can like...

449

00:41:35,616 --> 00:41:38,618

wave my hand and water comes out and that's magical.

450

00:41:38,618 --> 00:41:53,605

But I don't care about the pipes unless my pipes break or suddenly your sewer line

explodes and like, oh no, my water, my data is everywhere outside.

451

00:41:53,605 --> 00:41:54,786

That's not good.

452

00:41:54,786 --> 00:41:57,027

I don't want that outside of my house.

453

00:41:57,187 --> 00:42:05,171

And so those unsexy pieces of software are very, very important in a health system.

454

00:42:05,171 --> 00:42:06,813

in a hospital and a clinic.

455

00:42:06,813 --> 00:42:16,475

And so I've got to imagine as AI becomes, as these solutions just become so enmeshed and

every doctor is screaming, I need this, I need that, I need that.

456

00:42:16,475 --> 00:42:22,293

And not just the doctors, but the supply chain folks and legal, everyone is using AI.

457

00:42:22,293 --> 00:42:24,030

It's not just in the clinical sphere.

458

00:42:24,030 --> 00:42:29,350

obsessed about PHI, quite rightly, in terms of, from a confidentiality perspective.

459

00:42:29,350 --> 00:42:34,664

And that's another topic for another day in terms of who owns the patient's data.

460

00:42:34,725 --> 00:42:35,445

Interesting.

461

00:42:35,445 --> 00:42:38,686

uh Lots of viewpoints about that.

462

00:42:38,686 --> 00:42:41,307

And also, that's very regional.

463

00:42:41,307 --> 00:42:49,825

But confidential business information, CBI, is just as important because that is

oftentimes

464

00:42:49,825 --> 00:42:59,782

the health system's secret sauce of how it operates and how it's able to be successful or

otherwise in terms of its endeavors.

465

00:42:59,782 --> 00:43:04,486

So that CBI leaking out suddenly becomes highly relevant.

466

00:43:04,486 --> 00:43:10,449

Think of all of kind of personal information that might be there that is about your

employees.

467

00:43:10,449 --> 00:43:16,803

Think about it being some of the um intellectual property that's being created.

468

00:43:16,985 --> 00:43:18,726

by some of your employees.

469

00:43:18,726 --> 00:43:26,790

It may not have patient information in it, but it is very unique and specialized and truly

intellectual property.

470

00:43:26,831 --> 00:43:37,217

And yet that can leak out because some researcher decides they're going to put it in chat

GPT and suddenly it's often as part of training the model.

471

00:43:37,217 --> 00:43:42,900

um And so, yeah, there's a lot of concern in that area.

472

00:43:42,900 --> 00:43:44,261

And I think...

473

00:43:44,596 --> 00:43:53,633

If you wanted to start a business or you wanted to find the next big thing to sell, I

think, start a podcast.

474

00:43:53,633 --> 00:43:54,463

Yeah, I wish.

475

00:43:54,463 --> 00:43:55,924

I don't know where the money's coming from.

476

00:43:55,924 --> 00:43:57,716

Insert advert here.

477

00:43:57,716 --> 00:44:05,021

um But I think that um that's going to be an area that is going to explode.

478

00:44:05,021 --> 00:44:09,846

People looking for governance sort of solutions.

479

00:44:10,186 --> 00:44:16,329

that are actually managing that a bit like some of the compliance solutions that exist

already for some of my other areas.

480

00:44:16,329 --> 00:44:28,644

But being able to track that and then being able to embed that and I know ServiceNow and

some of those companies who do manage some of the internal, you know, of uh IT assets.

481

00:44:28,644 --> 00:44:35,396

There's another company that I really like called ARDOQ who are doing some amazing stuff,

Swedish company.

482

00:44:35,396 --> 00:44:36,897

I know you like Swedish companies.

483

00:44:36,897 --> 00:44:39,738

484

00:44:40,210 --> 00:44:52,636

And so from that perspective, and I don't mean IKEA, by the way, folks, ah but I do think

that those kind of technologies are going to become increasingly important because as

485

00:44:52,636 --> 00:44:55,377

Bethany says, yes, the plumbing, but guess what?

486

00:44:55,377 --> 00:44:59,819

When the plumbing is broken, it messes up everybody's life.

487

00:44:59,819 --> 00:45:05,682

So I think that that's where there's going to be a lot of focus on how do we try and make

sure that

488

00:45:05,742 --> 00:45:18,444

we've got control over all of the AI technologies we go forward and if I was looking for a

job, selling job or a product management job, I would be looking in that area.

489

00:45:18,444 --> 00:45:27,802

And if I was uh an innovator that had an inkling of how to work around those workflows,

that's who I'd be trying to convince.

490

00:45:27,802 --> 00:45:32,838

So VCs be looking at those platforms, they might not be sexy.

491

00:45:32,838 --> 00:45:42,200

the might not get you on CBS News, but my goodness, I think everybody's gonna be looking

for that over the next probably couple of years.

492

00:45:42,395 --> 00:45:44,247

Yeah, I completely agree.

493

00:45:44,247 --> 00:45:50,582

And if, you know, we just need some of them to hit so that we can make a lot of money.

494

00:45:51,074 --> 00:45:51,854

So.

495

00:45:52,192 --> 00:45:57,437

Exactly, All right, so do you want to wrap us up here, Bethany?

496

00:45:57,573 --> 00:45:59,355

Yes, absolutely.

497

00:45:59,355 --> 00:46:01,246

So thank you so much for listening.

498

00:46:01,246 --> 00:46:09,413

If you enjoyed this, please make sure that you share it with all your friends, your peers,

your colleagues, put it in that Teams chat, in your Slack channels.

499

00:46:09,413 --> 00:46:15,567

There's these two Scottish weirdos talking about healthcare technology.

500

00:46:15,868 --> 00:46:21,193

But our next episode, we are gonna drive a little bit more into our personal work

experience.

501

00:46:21,193 --> 00:46:26,941

Dad has been selling healthcare technology for 30, 40 years.

502

00:46:26,941 --> 00:46:28,621

Let's call it 30.

503

00:46:28,793 --> 00:46:30,853

30, we'll go 30.

504

00:46:30,993 --> 00:46:40,533

I have for coming up on 12 years now, about half the time, but we're going to talk a

little bit more about our own experience and around negotiations.

505

00:46:40,693 --> 00:46:43,053

How have we approached deals in the past?

506

00:46:43,053 --> 00:46:43,993

What has worked?

507

00:46:43,993 --> 00:46:44,853

What hasn't worked?

508

00:46:44,853 --> 00:46:46,533

What are some of our tricks?

509

00:46:46,613 --> 00:46:55,269

What do we think, you know, as we were just saying in today's episode in 2025, it's

different than even last year.

510

00:46:55,269 --> 00:47:03,855

So what are the techniques that we think are really gonna stick around and we're really

gonna lean into and see more of in the future?

511

00:47:03,855 --> 00:47:05,176

It's gonna be fun.

512

00:47:05,176 --> 00:47:06,739

We'll share all our secrets.

513

00:47:06,739 --> 00:47:12,465

that for all of you who have joined us, I really appreciate you coming and joining.

514

00:47:12,465 --> 00:47:17,921

As I say, if you would like, comment, give us some feedback about what we've covered here

today.

515

00:47:17,921 --> 00:47:20,393

If you've got a different point of view, tell us.

516

00:47:20,393 --> 00:47:21,995

We'd love to hear from you.

517

00:47:21,995 --> 00:47:28,641

So go ahead and comment down below, but like and subscribe help us in terms of growing the

channel.

518

00:47:28,690 --> 00:47:36,497

But the other thing is feel free to reach out to us either through the comments or through

any other messaging format.

519

00:47:36,497 --> 00:47:40,161

And we look forward to seeing you on the next episode.

520

00:47:40,161 --> 00:47:42,605

But until then, keep haverin'.