Advice for founders on how to sell into healthcare: Amit Mehta - BuildersVC

1
00:00:00,000 --> 00:00:03,000
Hi everyone, thanks so much for joining me today.

2
00:00:03,000 --> 00:00:05,840
Today I'm talking to Amit Mehta.

3
00:00:05,840 --> 00:00:10,280
He's an interventional radiologist and a general partner at Builders VC.

4
00:00:10,280 --> 00:00:15,180
Builders VC is a $450 million venture capital firm.

5
00:00:15,180 --> 00:00:18,880
They invest across all verticals, including healthcare.

6
00:00:18,880 --> 00:00:23,200
I hope you guys enjoyed the conversation with Amit as much as I did.

7
00:00:23,200 --> 00:00:26,440
Amit, thanks so much for coming on the podcast.

8
00:00:26,440 --> 00:00:28,960
Really looking forward to this call.

9
00:00:28,960 --> 00:00:34,760
To start, if you want to give us a brief introduction and then we'll get right into it.

10
00:00:34,760 --> 00:00:35,760
Sure.

11
00:00:35,760 --> 00:00:38,560
So, I grew up in Canada.

12
00:00:38,560 --> 00:00:45,400
I spent my childhood in Toronto, went to school in Canada and then saw some need for some

13
00:00:45,400 --> 00:00:51,160
technology as I was moving through medical school and figured that I couldn't do it there

14
00:00:51,160 --> 00:00:52,160
in Canada.

15
00:00:52,160 --> 00:00:58,960
So, for residency, decided to come to the US and work specifically on some specific

16
00:00:58,960 --> 00:00:59,960
problems.

17
00:00:59,960 --> 00:01:05,840
So, I trained through the Harvard system at Mass General, did essentially all my training

18
00:01:05,840 --> 00:01:12,320
there, fellowship and then ran a lab, research lab there for a few years while I was doing

19
00:01:12,320 --> 00:01:15,240
residency and did a few other things that were of interest.

20
00:01:15,240 --> 00:01:20,480
But during that time, got interested in investing and startup and got involved in that whole

21
00:01:20,480 --> 00:01:22,600
ecosystem there.

22
00:01:22,600 --> 00:01:31,600
Once I finished training, I had a thesis to run that lab there but then decided, had spent

23
00:01:31,600 --> 00:01:36,760
an infinitesimally small amount of money on my medical school education compared to my

24
00:01:36,760 --> 00:01:38,400
cohorts in the US and felt bad.

25
00:01:38,400 --> 00:01:41,960
So, I said, I'm going to come back to Canada and not contribute to the brain drain.

26
00:01:41,960 --> 00:01:46,760
So, moved back to Canada and worked in Canada for a couple of years, again, trying to move

27
00:01:46,760 --> 00:01:51,040
forward some of these ideas that I was working on in telemedicine and teleradiology and all

28
00:01:51,040 --> 00:01:53,040
of this kind of stuff.

29
00:01:53,040 --> 00:01:58,320
In that process, my wife who was from LA hated the cold and said, we got to move somewhere

30
00:01:58,320 --> 00:02:00,040
warm.

31
00:02:00,040 --> 00:02:05,320
And so, in that journey, moved back to the US and kept going on all the things that I

32
00:02:05,320 --> 00:02:06,320
was doing.

33
00:02:06,320 --> 00:02:13,040
Tell me a bit more about what tech did you foresee that was not feasible in Canada and

34
00:02:13,040 --> 00:02:16,280
you hinted towards teleradiology and telemedicine.

35
00:02:16,280 --> 00:02:22,560
And was it not feasible because of reimbursement issues, regulatory issues or something else?

36
00:02:22,560 --> 00:02:24,040
Yes.

37
00:02:24,040 --> 00:02:28,080
Basically, all of the above.

38
00:02:28,080 --> 00:02:32,680
So conceptually, what I was working on when I was doing residency and fellowship was a

39
00:02:32,680 --> 00:02:37,180
combination of telemedicine but also applying telemedicine and teleradiology into the clinical

40
00:02:37,180 --> 00:02:39,240
trial sphere.

41
00:02:39,240 --> 00:02:41,800
And that didn't work in Canada for a couple of reasons.

42
00:02:41,800 --> 00:02:46,240
One, I mean, there was no concept of telemedicine from both a technologic infrastructure standpoint

43
00:02:46,240 --> 00:02:49,680
nor from a reimbursement standpoint.

44
00:02:49,680 --> 00:02:55,160
There was no drive to even innovate in that space, venture and startup and all of which

45
00:02:55,160 --> 00:03:00,320
now is very robust in Canada through Mars and through BDC and a lot of the entities

46
00:03:00,320 --> 00:03:07,240
that as a VC now, those entities are actually limited partners in our funds.

47
00:03:07,240 --> 00:03:09,000
There was none of that.

48
00:03:09,000 --> 00:03:11,120
This was 15 years ago.

49
00:03:11,120 --> 00:03:17,320
And so I tried to push this agenda of what at the time I thought was sort of revolutionary

50
00:03:17,320 --> 00:03:21,400
in terms of bringing care to underserviced areas, which was a big problem in Canada,

51
00:03:21,400 --> 00:03:22,400
right?

52
00:03:22,400 --> 00:03:24,960
Northern Ontario and lots of other places.

53
00:03:24,960 --> 00:03:27,080
There was just no appetite for it.

54
00:03:27,080 --> 00:03:31,280
And the extension of that was what I initially had dreamt of was working on a clinical trial

55
00:03:31,280 --> 00:03:37,800
company where we had some technology and some ideas around making clinical trials more efficient

56
00:03:37,800 --> 00:03:43,160
and subsequently, several years later, did start that company, ran that and then we sold

57
00:03:43,160 --> 00:03:45,440
that company last year to a public.

58
00:03:45,440 --> 00:03:50,760
So in the end, it turned out and there was fruition to all the goals, but in Canada,

59
00:03:50,760 --> 00:03:53,160
I just didn't feel like I could get it done.

60
00:03:53,160 --> 00:03:57,880
Have you seen, and I've had similar experiences with my startup in Canada and talking to the

61
00:03:57,880 --> 00:04:04,080
Ministry of Health here as their primary goal is being reelected and nothing else.

62
00:04:04,080 --> 00:04:11,160
Have you seen any changes now and if you were a resident or in med school now, would you

63
00:04:11,160 --> 00:04:14,680
have stayed in Canada or do you think things are more or less the same still?

64
00:04:14,680 --> 00:04:18,200
No, I think it's done a complete 180 and it's completely changed.

65
00:04:18,200 --> 00:04:24,420
I mean, we're sourcing a lot of our deals on the venture side from Canada.

66
00:04:24,420 --> 00:04:29,720
We have a mandate because these Canadian entities, Alberta Enterprise and a few others are LPs

67
00:04:29,720 --> 00:04:30,720
in our fund.

68
00:04:30,720 --> 00:04:33,960
We have a Canadian mandate that we need to spend a certain amount of money in Canada

69
00:04:33,960 --> 00:04:37,160
and we don't have trouble finding deals.

70
00:04:37,160 --> 00:04:42,000
So I think that entire innovation ecosystem has completely blossomed from the time that

71
00:04:42,000 --> 00:04:47,320
I was there and the difficulty what I had in terms of just the whole life cycle of a

72
00:04:47,320 --> 00:04:52,000
startup from fundraising to infrastructure to human capital.

73
00:04:52,000 --> 00:04:53,600
I just don't see that happening now.

74
00:04:53,600 --> 00:04:58,240
I mean, it's not, I don't think it's Silicon Valley, but it's also not as difficult it

75
00:04:58,240 --> 00:05:00,840
was when it was when I was doing it.

76
00:05:00,840 --> 00:05:02,960
Okay, yeah, that's good to hear.

77
00:05:02,960 --> 00:05:05,400
Let's go back to your childhood, Ahmed.

78
00:05:05,400 --> 00:05:10,960
Let's go back to when you were about 10, 15, maybe in high school.

79
00:05:10,960 --> 00:05:13,440
What did you want to be then?

80
00:05:13,440 --> 00:05:17,320
And once you were in med school, did you, most people in med school think they're going

81
00:05:17,320 --> 00:05:21,280
to be a doctor forever, as did I.

82
00:05:21,280 --> 00:05:23,120
When did that change for you?

83
00:05:23,120 --> 00:05:28,840
And did you say, okay, this isn't enough for me and I want to do something else or something

84
00:05:28,840 --> 00:05:29,840
more?

85
00:05:29,840 --> 00:05:31,680
I always wanted to be a doctor.

86
00:05:31,680 --> 00:05:35,960
I think I grew up in a family with physicians and all of the rest of, you know, typical

87
00:05:35,960 --> 00:05:41,720
sort of immigrant story, but I had a very, I think an interest in technology and a technological

88
00:05:41,720 --> 00:05:48,120
bent in terms of tinkering and, you know, ran illegal bulletin boards out of my attic

89
00:05:48,120 --> 00:05:52,920
and got in trouble with my parents and all of that kind of stuff when I was a kid.

90
00:05:52,920 --> 00:05:55,400
And so I was at an interest in technology.

91
00:05:55,400 --> 00:06:01,840
And so I was driven towards a technologic field and it was more finding the right time

92
00:06:01,840 --> 00:06:07,360
to start working on those ideas as opposed to trying to generate the idea or is it something

93
00:06:07,360 --> 00:06:08,360
I want to work on?

94
00:06:08,360 --> 00:06:10,320
I had determined I wanted to work on it very early.

95
00:06:10,320 --> 00:06:16,640
It was just, you know, back then it wasn't cool to be 17 and start a company as it is

96
00:06:16,640 --> 00:06:19,440
now, if you're 30, it's not cool.

97
00:06:19,440 --> 00:06:21,440
You need to be 15 to start a company.

98
00:06:21,440 --> 00:06:23,240
So that dynamic was very different.

99
00:06:23,240 --> 00:06:28,920
It was hard to get taken seriously as a 17, 18 year old, you know, and I mean, you guys

100
00:06:28,920 --> 00:06:32,240
are in Canada, you're in Canada, the back of the, I mean, I did the two year university

101
00:06:32,240 --> 00:06:33,520
four year med school thing.

102
00:06:33,520 --> 00:06:38,440
And so on top of that, I was going into being, you know, medical school really young, like

103
00:06:38,440 --> 00:06:39,440
16, 17.

104
00:06:39,440 --> 00:06:44,000
So it was hard to get, it was hard to be taken seriously.

105
00:06:44,000 --> 00:06:47,400
But I had those ideas and I was always working on those ideas as I kind of progressed through

106
00:06:47,400 --> 00:06:48,400
the system.

107
00:06:48,400 --> 00:06:53,280
So medicine was sort of, it was the day job and work on going through medical school and

108
00:06:53,280 --> 00:06:57,320
the academic component of it, but was working on those things at my own time.

109
00:06:57,320 --> 00:06:59,040
Okay, perfect.

110
00:06:59,040 --> 00:07:02,120
And why did you decide to pick intervention radiology?

111
00:07:02,120 --> 00:07:07,000
Primarily, again, like I'm at a very tech bent.

112
00:07:07,000 --> 00:07:10,280
It was very, I think it was a combination of a couple of things.

113
00:07:10,280 --> 00:07:13,240
One is, you know, you say, there's so many experiences shape who you are.

114
00:07:13,240 --> 00:07:15,440
So like I was telling you, like I went to medical school very early.

115
00:07:15,440 --> 00:07:20,720
Like I remember taking a primary care rotation and just patients just were like, you're too

116
00:07:20,720 --> 00:07:21,720
young to be my doc.

117
00:07:21,720 --> 00:07:24,160
Like I just was young and so people didn't take me seriously.

118
00:07:24,160 --> 00:07:29,800
So that was sort of steered me away from patient facing things almost off the bat because of

119
00:07:29,800 --> 00:07:37,080
this sort of this discrimination on age to a certain extent.

120
00:07:37,080 --> 00:07:38,720
But I did have a very tech bent.

121
00:07:38,720 --> 00:07:41,280
And so I was looking for things that were very heavy in that.

122
00:07:41,280 --> 00:07:45,840
So it almost fed me towards surgical specialties off the bat.

123
00:07:45,840 --> 00:07:50,880
And as I got further, I realized that maybe I'm a little bit ADHD and that, you know,

124
00:07:50,880 --> 00:07:53,400
I like to move very quickly and efficiently.

125
00:07:53,400 --> 00:07:56,400
And so radiology was something that was like that one.

126
00:07:56,400 --> 00:08:00,560
It wasn't as patient facing as I didn't want completely not patient facing, which, you

127
00:08:00,560 --> 00:08:02,840
know, interventional reality still is patient interaction.

128
00:08:02,840 --> 00:08:07,040
It's not all interpretation, but it was less patient facing.

129
00:08:07,040 --> 00:08:08,320
It was tech oriented.

130
00:08:08,320 --> 00:08:12,240
I could innovate in that space and it was still it was a high volume of cases moving

131
00:08:12,240 --> 00:08:15,680
very quickly, which was very attuned to my personality.

132
00:08:15,680 --> 00:08:18,760
So I think I was kind of lucky to find that early on that I found something that fits

133
00:08:18,760 --> 00:08:19,760
my personality.

134
00:08:19,760 --> 00:08:23,360
I never went through that concentration of is this the right thing for me or is this

135
00:08:23,360 --> 00:08:24,360
the right specialty?

136
00:08:24,360 --> 00:08:26,960
Like I loved it from the beginning and I've always loved I mean, to this day, I still

137
00:08:26,960 --> 00:08:27,960
enjoy it.

138
00:08:27,960 --> 00:08:28,960
Okay, perfect.

139
00:08:28,960 --> 00:08:33,840
You've had a lot of different experiences in the past 10 years.

140
00:08:33,840 --> 00:08:38,560
I'm and you know, someone would look at the resume and say, I'm a doesn't stick with things

141
00:08:38,560 --> 00:08:46,480
to an extent is tell me a bit more about joining your most favorite experience and then your

142
00:08:46,480 --> 00:08:49,720
least favorite if you'd like to.

143
00:08:49,720 --> 00:08:54,040
And what made the decision to were you like, I'm here for this job.

144
00:08:54,040 --> 00:08:56,920
I'm going to take this company from point A to point B when it's a point B, I'm going

145
00:08:56,920 --> 00:09:00,160
to leave or what was the thought process behind that?

146
00:09:00,160 --> 00:09:01,360
Yeah, it's funny.

147
00:09:01,360 --> 00:09:04,680
I think my detractors would say the opposite, I stick with things too long.

148
00:09:04,680 --> 00:09:09,240
So a lot of these Yeah, interestingly, I probably doing too many things and doing all of them

149
00:09:09,240 --> 00:09:12,720
at the same time as opposed to doing one and letting it go.

150
00:09:12,720 --> 00:09:16,600
But you know, the most interesting thing I think is definitely all for me as always been

151
00:09:16,600 --> 00:09:21,360
is been venture and being a GP and a fund and investing and all of the cool ideas and

152
00:09:21,360 --> 00:09:24,920
things you think about and figuring out a process of how to be a good investor and how

153
00:09:24,920 --> 00:09:25,920
to work on that.

154
00:09:25,920 --> 00:09:28,600
I mean, absolutely, that's been it's been fun.

155
00:09:28,600 --> 00:09:31,160
And it's been the exposure is incredible.

156
00:09:31,160 --> 00:09:34,960
You get to see things and touch things and do things that just you don't get any other

157
00:09:34,960 --> 00:09:36,480
way.

158
00:09:36,480 --> 00:09:39,560
Interesting way I would tell you that the flip side of that probably and I wouldn't

159
00:09:39,560 --> 00:09:43,320
say this is the worst experience, but the most difficult has been starting and selling

160
00:09:43,320 --> 00:09:49,000
this company like it was eight years of, you know, it was 100 hour weeks just working,

161
00:09:49,000 --> 00:09:50,840
it was stressful.

162
00:09:50,840 --> 00:09:55,280
And you know, it was enjoyable to a certain extent for certain things.

163
00:09:55,280 --> 00:09:59,440
But at the same time, it was very stressful to running a company running a you know, being

164
00:09:59,440 --> 00:10:03,720
a physician doing venture like it was just probably a lot of things at the same time.

165
00:10:03,720 --> 00:10:09,280
But at the same rate, it was instrumental, I think, in in forming a foundation for me

166
00:10:09,280 --> 00:10:13,940
to be a better investor, you know, coming from an operator perspective, to becoming

167
00:10:13,940 --> 00:10:19,080
an investor is a very different mode than just trying to be a primary investor.

168
00:10:19,080 --> 00:10:23,920
You know, I mean, this morning, I spent, you know, three hours on calls going through with

169
00:10:23,920 --> 00:10:29,600
entrepreneurs looking at 2023 budgets and pro forma is an understanding how to work

170
00:10:29,600 --> 00:10:34,520
out, you know, how to run a pro forma to make sure that your cause doesn't exceed your revenue

171
00:10:34,520 --> 00:10:38,400
and that your human capital is being, you know, all of that only comes from operational

172
00:10:38,400 --> 00:10:40,160
experience.

173
00:10:40,160 --> 00:10:45,040
It's hard to get that unless you've seen hundreds and hundreds of things or you've spent time

174
00:10:45,040 --> 00:10:49,480
very focused and understanding that specific problem.

175
00:10:49,480 --> 00:10:53,800
And so it was an education by fire to a certain extent that I think has helped me as an investor

176
00:10:53,800 --> 00:10:56,160
Okay, perfect.

177
00:10:56,160 --> 00:11:03,320
A lot of people are looking at the rising interest rates, recession and kind of buckling

178
00:11:03,320 --> 00:11:08,400
down, keeping more money in cash, not deploying as much capital.

179
00:11:08,400 --> 00:11:11,320
You know, I look at public and private markets as two different entities.

180
00:11:11,320 --> 00:11:14,040
I don't even try and compare the two.

181
00:11:14,040 --> 00:11:17,520
What is your strategy moving over the next six or 12 months?

182
00:11:17,520 --> 00:11:19,840
Are you changing the way you look at startups?

183
00:11:19,840 --> 00:11:25,640
Are you looking for startups that are almost maybe revenue neutral and whereas before you

184
00:11:25,640 --> 00:11:27,880
would be have more laxity there?

185
00:11:27,880 --> 00:11:33,240
Is your investment thesis changing at all and are you deploying less or more capital

186
00:11:33,240 --> 00:11:36,320
over the next six or 12 months?

187
00:11:36,320 --> 00:11:40,280
So I mean, there's a personal component and then there's an institutional component, right?

188
00:11:40,280 --> 00:11:46,000
And I would say on the institutional side, we are a series A fund and a seed fund.

189
00:11:46,000 --> 00:11:52,560
And so what's just evolved in the venture landscape is sort of valuations more than anything else.

190
00:11:52,560 --> 00:11:57,360
I mean, we're still with the companies we were investing in already were in a situation

191
00:11:57,360 --> 00:12:00,520
where they were often were not that revenue positive.

192
00:12:00,520 --> 00:12:05,400
They were beyond proof of concept, beyond seed stage companies, but they had not hit

193
00:12:05,400 --> 00:12:08,200
that hockey stick of growth yet.

194
00:12:08,200 --> 00:12:09,200
So that hasn't changed.

195
00:12:09,200 --> 00:12:12,440
I mean, we're still investing at the same level and series.

196
00:12:12,440 --> 00:12:18,520
It's just some of the deal metrics have changed in terms of valuations, in terms of ownership

197
00:12:18,520 --> 00:12:20,680
or what we're doing for that company.

198
00:12:20,680 --> 00:12:25,720
Our stage size of check and involvement in the company has stayed the same as it was

199
00:12:25,720 --> 00:12:28,280
a year or two or five ago.

200
00:12:28,280 --> 00:12:33,040
So we're not, you know, there's just these, I think, you know, part of the benefit of

201
00:12:33,040 --> 00:12:35,720
now being older is that this is cyclical, right?

202
00:12:35,720 --> 00:12:39,880
And I've been through a couple of cycles of this now and understand that all we did really

203
00:12:39,880 --> 00:12:44,120
honestly be truthful with you is we told our companies that, you know, you should stockpile

204
00:12:44,120 --> 00:12:50,120
30 months, you know, going out into the fundraising environment in the next 12 to 18 months is

205
00:12:50,120 --> 00:12:52,720
going to be really difficult unless you are one of the winning companies.

206
00:12:52,720 --> 00:12:58,280
I mean, if you're a winning company, there is money in the system, venture, whatever

207
00:12:58,280 --> 00:13:00,440
you want to say that has to be deployed, right?

208
00:13:00,440 --> 00:13:04,080
By definition, funds work by deploying capital.

209
00:13:04,080 --> 00:13:07,480
So that capital is going to be deployed to the winners.

210
00:13:07,480 --> 00:13:10,920
And so if you're a winner, it should be okay, but if you're on the precipice of being a

211
00:13:10,920 --> 00:13:14,000
winner or you're not a winner, then you need to hoard cash now.

212
00:13:14,000 --> 00:13:15,840
It's not that we think you're a bad company.

213
00:13:15,840 --> 00:13:19,800
It's just your life cycle over the next 18 months is going to be very different than

214
00:13:19,800 --> 00:13:22,800
a year ago when everyone was falling all over themselves.

215
00:13:22,800 --> 00:13:25,840
We had deals that were valued at 100 X ARR.

216
00:13:25,840 --> 00:13:29,840
I mean, I haven't seen a deal like that in nine months.

217
00:13:29,840 --> 00:13:34,560
Like, I mean, if you try to sell a deal now at 100 X ARR, people will just laugh at you

218
00:13:34,560 --> 00:13:36,840
and be like, you know, good luck.

219
00:13:36,840 --> 00:13:43,280
So I think that's the biggest shift is be conservative and, you know, we all hope our

220
00:13:43,280 --> 00:13:44,280
companies are winners.

221
00:13:44,280 --> 00:13:45,520
So if they're a winner, it's great.

222
00:13:45,520 --> 00:13:47,080
We can help them raise money.

223
00:13:47,080 --> 00:13:50,200
But if they're not or their inflection point is going to be a little more difficult.

224
00:13:50,200 --> 00:13:52,160
Okay, perfect.

225
00:13:52,160 --> 00:13:56,040
Talk to me about your first LP.

226
00:13:56,040 --> 00:13:59,520
How long did it take to get that investment?

227
00:13:59,520 --> 00:14:03,240
And does it get much easier after the first LP?

228
00:14:03,240 --> 00:14:07,880
There's a few GPs with smaller funds who listen to this, and I think they'd be very interested

229
00:14:07,880 --> 00:14:11,880
in the answer.

230
00:14:11,880 --> 00:14:16,760
Contrary to what everyone else says, I find fundraising the most difficult thing in the

231
00:14:16,760 --> 00:14:17,760
world.

232
00:14:17,760 --> 00:14:21,840
I mean, every single meeting is a different minefield.

233
00:14:21,840 --> 00:14:27,200
So in essence, right, in venture, I mean, as well as personal probably trackers, if

234
00:14:27,200 --> 00:14:30,600
you've had a track record of great investments, it speaks for itself.

235
00:14:30,600 --> 00:14:31,920
And this is like when you run a company.

236
00:14:31,920 --> 00:14:34,320
You know, when there's sales, everyone's happy.

237
00:14:34,320 --> 00:14:36,800
When there's no sales, then all of the problems start to come up.

238
00:14:36,800 --> 00:14:39,820
And, you know, this person is not doing this and that person is not doing this or this

239
00:14:39,820 --> 00:14:41,500
infrastructure is broken.

240
00:14:41,500 --> 00:14:45,780
So in venture, if your funds are doing well and you have returned that follow up meeting

241
00:14:45,780 --> 00:14:50,080
with an LP to ask for a check for the next fund is very straightforward because so look,

242
00:14:50,080 --> 00:14:53,040
we returned you 5x on our last fund.

243
00:14:53,040 --> 00:14:54,200
It's almost a fait accompli.

244
00:14:54,200 --> 00:14:57,280
Like, why wouldn't you put money into this next fund?

245
00:14:57,280 --> 00:15:01,680
The trouble is when your funds are not doing well, and you're asking for re-ups on checks.

246
00:15:01,680 --> 00:15:03,280
That's when it's difficult.

247
00:15:03,280 --> 00:15:07,240
But my philosophy has always been, you know, and this is somewhat medical, you know, driven

248
00:15:07,240 --> 00:15:11,640
in medicine into that if you're not doing something well, then maybe you need to look

249
00:15:11,640 --> 00:15:15,200
at some other, you know, don't kill people, like maybe pick another surgical specialist

250
00:15:15,200 --> 00:15:18,120
or something like you need to you need to pivot.

251
00:15:18,120 --> 00:15:21,280
And so luckily, our funds have done very well from the inception.

252
00:15:21,280 --> 00:15:26,860
And so we, you know, going into these LPs is more of a relationship driven fundraise

253
00:15:26,860 --> 00:15:32,560
rather than trying to prove our economics, our DPI and our, you know, everything has

254
00:15:32,560 --> 00:15:33,560
been good.

255
00:15:33,560 --> 00:15:35,480
So we haven't had that issue.

256
00:15:35,480 --> 00:15:36,960
Okay, perfect.

257
00:15:36,960 --> 00:15:42,280
When I think about my past investments, when I look at companies, I'm looking to invest

258
00:15:42,280 --> 00:15:49,080
right now, I usually look at the level of conviction I have and the potential reward.

259
00:15:49,080 --> 00:15:54,800
And usually sometimes I may have a lower level of conviction, but I'm getting in at a cheap

260
00:15:54,800 --> 00:15:59,000
valuation and there's a high level of reward.

261
00:15:59,000 --> 00:16:04,200
Talk to me about, I'll ask you about your best or your favorite investment first and

262
00:16:04,200 --> 00:16:07,520
we'll talk about your least favorite next.

263
00:16:07,520 --> 00:16:08,800
What was your level of conviction?

264
00:16:08,800 --> 00:16:14,280
What was the balance between your level of conviction and a potential reward in your

265
00:16:14,280 --> 00:16:16,400
favorite investment and then your least favorite?

266
00:16:16,400 --> 00:16:24,200
Yeah, I mean, I think that's a pretty difficult question to answer from a VC perspective.

267
00:16:24,200 --> 00:16:26,000
The decision is multifactorial, right?

268
00:16:26,000 --> 00:16:29,600
There's a myriad of things that go into deciding whether you make an investment, whether it's,

269
00:16:29,600 --> 00:16:33,680
you know, of which the entrepreneur is very important, the total market is important,

270
00:16:33,680 --> 00:16:37,200
the product is important, fit, all of those things go into that decision.

271
00:16:37,200 --> 00:16:42,920
I would say I've never made an investment that I didn't have a hundred percent full

272
00:16:42,920 --> 00:16:44,640
conviction on.

273
00:16:44,640 --> 00:16:49,360
Whatever got me to that level of conviction was different probably every time.

274
00:16:49,360 --> 00:16:50,360
And there's ebbs and flows.

275
00:16:50,360 --> 00:16:53,240
And like I think, again, like I said, as you understand, there's cycles in this stuff,

276
00:16:53,240 --> 00:16:56,640
there's different things at different times that are applicable, especially in healthcare,

277
00:16:56,640 --> 00:16:57,640
right?

278
00:16:57,640 --> 00:17:01,840
Healthcare investing is probably to me is more difficult than some of the other disciplines

279
00:17:01,840 --> 00:17:05,000
just because it's so heterogeneous.

280
00:17:05,000 --> 00:17:08,600
There's multiple verticals, there's payers, there's providers, there's patients, there's

281
00:17:08,600 --> 00:17:13,760
a FinTech component, there's a tech component, there's, you know, applicability in the market,

282
00:17:13,760 --> 00:17:17,400
there's hospitals, ACOs, clinics, I mean, it goes on and on and on, right?

283
00:17:17,400 --> 00:17:22,280
And so it's always at the end of the day, it's always been a hundred percent conviction,

284
00:17:22,280 --> 00:17:24,240
but the conviction has been around different things.

285
00:17:24,240 --> 00:17:29,120
And that's dependent on the cycle we're in, you know, decreasing reimbursement rates,

286
00:17:29,120 --> 00:17:34,000
is there opportunity of FinTech opportunities in healthcare right now, as opposed to, I

287
00:17:34,000 --> 00:17:38,600
think, during COVID and the pandemic, there was a lot of opportunity for remote care and

288
00:17:38,600 --> 00:17:44,080
moving tertiary care outside of the hospital to the home, remote monitoring, IOT.

289
00:17:44,080 --> 00:17:48,960
So there's a different, the conviction comes around different elements of the opportunity

290
00:17:48,960 --> 00:17:51,040
at the time in the cycle we're at.

291
00:17:51,040 --> 00:17:52,440
Okay, perfect.

292
00:17:52,440 --> 00:17:59,400
Amit, would you rather have a life of many successes, but no major success?

293
00:17:59,400 --> 00:18:03,280
Or would you rather have a life similar to Vincent van Gogh's?

294
00:18:03,280 --> 00:18:09,640
He had massive, what we would call failures, was not recognized in his own life, you know,

295
00:18:09,640 --> 00:18:11,920
infamously famous now.

296
00:18:11,920 --> 00:18:15,960
Would you rather have Vincent van Gogh's life, a hard life, lots of failures, no success

297
00:18:15,960 --> 00:18:23,320
while you were living, or a life of many successes, but you're not known, and you don't change

298
00:18:23,320 --> 00:18:26,960
the world in any substantial capacity?

299
00:18:26,960 --> 00:18:34,120
Yeah, I mean, I don't think the human brain is probably not wired to enjoy failure.

300
00:18:34,120 --> 00:18:38,000
So I don't, you know, I think it'd probably be the former.

301
00:18:38,000 --> 00:18:40,680
Yes, obviously, success feels good.

302
00:18:40,680 --> 00:18:46,320
But success and failure, I think, again, are also probably very, it's a very individual

303
00:18:46,320 --> 00:18:47,320
term.

304
00:18:47,320 --> 00:18:53,640
And what one may, you know, if you have an exit out of your company for $50,000, is that

305
00:18:53,640 --> 00:18:57,200
successful versus someone who has a $50 million or a $500 million exit?

306
00:18:57,200 --> 00:19:01,520
Once you've had a $500 million exit, then your next exit needs to be $5 billion, or

307
00:19:01,520 --> 00:19:02,520
you weren't successful.

308
00:19:02,520 --> 00:19:06,880
You know, your next exit at $500K is not successful, right?

309
00:19:06,880 --> 00:19:10,680
So I think some of that is just shaped on your experience.

310
00:19:10,680 --> 00:19:14,400
And what I've done, you know, I've tried to build a career around three facets, right?

311
00:19:14,400 --> 00:19:19,440
So there's a medical career, an actual physical operating career, the investment career, and

312
00:19:19,440 --> 00:19:26,480
then the innovative or the operator career of starting a company, running a company,

313
00:19:26,480 --> 00:19:31,000
and tried to bring all three of those things together to optimize what I can do as an investor,

314
00:19:31,000 --> 00:19:33,840
as a physician, and as an operator for each of those.

315
00:19:33,840 --> 00:19:38,880
So I think, I feel like success has been different in each of those things, and failure has been

316
00:19:38,880 --> 00:19:39,880
in each of those things.

317
00:19:39,880 --> 00:19:43,640
Like, you know, our failure on the operating side is very different than failure on the

318
00:19:43,640 --> 00:19:46,880
investing side, which means we lost $15 million in investment.

319
00:19:46,880 --> 00:19:50,800
Failure on the operating side is that we couldn't motivate our salespeople to make enough sales.

320
00:19:50,800 --> 00:19:55,440
You know, the magnitude of importance of those things are very different as for me as an

321
00:19:55,440 --> 00:20:00,520
individual, but as in my responsibility in each of those entities, is probably of equal

322
00:20:00,520 --> 00:20:01,520
importance.

323
00:20:01,520 --> 00:20:07,240
So it's just very dependent on what lens you're looking at it from, but I would say I don't

324
00:20:07,240 --> 00:20:08,800
like, I mean, I don't like failure.

325
00:20:08,800 --> 00:20:11,800
So I'm going to go with success more than failure.

326
00:20:11,800 --> 00:20:13,560
Okay, perfect.

327
00:20:13,560 --> 00:20:19,000
What advice do you have for founders trying to sell to health systems right now?

328
00:20:19,000 --> 00:20:24,120
How much should they, a lot of founders that come to me, they focus too much on clinical

329
00:20:24,120 --> 00:20:32,000
ROI and not enough on immediate financial ROI, which I feel a lot of health systems,

330
00:20:32,000 --> 00:20:40,000
ACOs, family health teams are more focused on and decisions are made more on financial

331
00:20:40,000 --> 00:20:41,280
ROI.

332
00:20:41,280 --> 00:20:42,760
You have a lot more experience than I do.

333
00:20:42,760 --> 00:20:46,600
It's more of a selfish question for me and what advice I should give to my founders right

334
00:20:46,600 --> 00:20:47,600
now.

335
00:20:47,600 --> 00:20:51,360
What advice would you give to founders trying to sell into health systems right now?

336
00:20:51,360 --> 00:20:55,000
You know, I said the thing that I think is the hardest thing to do in the world is to

337
00:20:55,000 --> 00:20:56,000
raise money from LPs.

338
00:20:56,000 --> 00:21:00,960
Let me give you a very, very close second is to sell into a hospital system.

339
00:21:00,960 --> 00:21:07,820
So, you know, selling to a hospital system, the problem is the incentives are probably

340
00:21:07,820 --> 00:21:16,680
perverse and it's very difficult to understand what those are in that the driver for a hospital

341
00:21:16,680 --> 00:21:18,080
system may not be what you...

342
00:21:18,080 --> 00:21:27,320
If you are not a provider or been on the medical staff or a CMO or a CMIO, whatever it may

343
00:21:27,320 --> 00:21:33,960
be, you may not understand what the driver is for that hospital system, hospital administrator

344
00:21:33,960 --> 00:21:39,120
who, you know, not a popular thing to say, but as we all know, they seem to rotate every

345
00:21:39,120 --> 00:21:40,120
two months.

346
00:21:40,120 --> 00:21:41,360
There's a new hospital administrator, right?

347
00:21:41,360 --> 00:21:46,280
The velocity of change in the hospital administrator is incredible.

348
00:21:46,280 --> 00:21:52,680
So I would say try to focus on products that probably don't sell to the hospital first,

349
00:21:52,680 --> 00:21:57,440
but saw more of a large clinical need where the hospital is where you end up as your larger

350
00:21:57,440 --> 00:22:02,000
clients once you've established a base with whatever else it may be.

351
00:22:02,000 --> 00:22:03,960
So telemedicine is a good example.

352
00:22:03,960 --> 00:22:09,880
Telemedicine is very applicable to hospital interactions, but maybe start in another setting,

353
00:22:09,880 --> 00:22:13,520
both prove out your infrastructure, prove out your business model, prove out your finances,

354
00:22:13,520 --> 00:22:16,440
and then that's the bigger fish to go after.

355
00:22:16,440 --> 00:22:17,800
Okay, perfect.

356
00:22:17,800 --> 00:22:24,120
Yeah, I mean, A16Z recently released their report about the pay-vider model and their

357
00:22:24,120 --> 00:22:28,460
banking on BTC consumer health.

358
00:22:28,460 --> 00:22:32,320
Let's talk a bit about the reimbursement structures in healthcare.

359
00:22:32,320 --> 00:22:37,960
And then particularly, do you think healthcare should be profitable?

360
00:22:37,960 --> 00:22:41,060
If so, what parts of healthcare should be profitable?

361
00:22:41,060 --> 00:22:42,400
Should it be primary care?

362
00:22:42,400 --> 00:22:44,640
Should it be more acute care, hospital care?

363
00:22:44,640 --> 00:22:53,240
As it stands now, the most profitable parts of healthcare are cancer therapy, surgeries,

364
00:22:53,240 --> 00:22:56,880
which I think should be perhaps at least profitable.

365
00:22:56,880 --> 00:23:02,920
And I also think innovation in healthcare should be decoupled from delivery of healthcare.

366
00:23:02,920 --> 00:23:07,840
So I'd like to get your thoughts on innovation and delivery in healthcare.

367
00:23:07,840 --> 00:23:12,320
Should they be different financial systems and incentives for both and profitability

368
00:23:12,320 --> 00:23:13,320
in healthcare?

369
00:23:13,320 --> 00:23:16,960
Yeah, that's a very interesting question.

370
00:23:16,960 --> 00:23:21,720
So I've worked in the US system and the Canadian system, and we've done a lot of work in the

371
00:23:21,720 --> 00:23:22,840
UK system.

372
00:23:22,840 --> 00:23:28,840
So I've seen three very different models of healthcare, worked in two of them and worked

373
00:23:28,840 --> 00:23:30,880
very closely in the third.

374
00:23:30,880 --> 00:23:35,240
And so I will tell you, I think my personal opinion is there's probably no optimal system.

375
00:23:35,240 --> 00:23:39,440
But the answer to your first question is I do think healthcare should be profitable.

376
00:23:39,440 --> 00:23:45,960
I think there's a governor on how profitable and some of that is ensuring that it needs

377
00:23:45,960 --> 00:23:50,360
... I think part of the problem right now is, and CMS and some of the other systems

378
00:23:50,360 --> 00:23:54,160
have made an effort towards this, but there is an essential decoupling of quality versus

379
00:23:54,160 --> 00:23:55,640
reimbursement.

380
00:23:55,640 --> 00:24:00,360
And we're moving towards trying to fix that, but in a lot of ways that's almost not fixable

381
00:24:00,360 --> 00:24:01,440
or has not been fixed.

382
00:24:01,440 --> 00:24:05,600
But I think that it is reasonable to make it a profitable system where you're based

383
00:24:05,600 --> 00:24:07,800
on quality and pay for quality.

384
00:24:07,800 --> 00:24:11,920
And the reason I say that is, I know you're saying a decouple innovation, but I think

385
00:24:11,920 --> 00:24:16,720
innovation is driven by that.

386
00:24:16,720 --> 00:24:22,440
People do go into medicine because financially it is satisfying.

387
00:24:22,440 --> 00:24:24,600
You do make good money as a doctor.

388
00:24:24,600 --> 00:24:26,000
It's not a secret.

389
00:24:26,000 --> 00:24:31,920
If you start eliminating the financial benefit, if you're building a system from de novo

390
00:24:31,920 --> 00:24:33,640
and that there's that opportunity, that's different.

391
00:24:33,640 --> 00:24:37,120
But in the current system, if you said, well, doctors are going to start making a 10th of

392
00:24:37,120 --> 00:24:40,600
what they make now, you're going to lose some of the best and brightest from going into

393
00:24:40,600 --> 00:24:43,600
medicine and then quality will drop and et cetera.

394
00:24:43,600 --> 00:24:44,600
There's a fallout from that.

395
00:24:44,600 --> 00:24:49,240
So I think in the confines of the current system, it does need to be profitable, but

396
00:24:49,240 --> 00:24:53,960
I do believe insurance companies have taken this to a different level and the system has

397
00:24:53,960 --> 00:24:58,560
been perverted in order to have oversized returns for United.

398
00:24:58,560 --> 00:25:02,560
If you look at just look at their filing, the amount of money insurance is making is

399
00:25:02,560 --> 00:25:07,360
out of proportion to what we see on the street and what happens in terms of patients being

400
00:25:07,360 --> 00:25:12,920
either not getting a pre-op or being steered towards certain providers or whatever is happening

401
00:25:12,920 --> 00:25:14,560
for the people in the trenches.

402
00:25:14,560 --> 00:25:16,320
Okay, perfect.

403
00:25:16,320 --> 00:25:18,800
Let's talk about tailwinds.

404
00:25:18,800 --> 00:25:23,760
One tailwind I'm banking on as a hybrid home care model for acute care.

405
00:25:23,760 --> 00:25:25,440
Take me back 10 years.

406
00:25:25,440 --> 00:25:29,440
What are some tailwinds you were identifying then that you were right about and what are

407
00:25:29,440 --> 00:25:33,360
some you were wrong about?

408
00:25:33,360 --> 00:25:39,540
So I sort of built at Builders our healthcare thesis around sort of four pillars of which

409
00:25:39,540 --> 00:25:41,360
I think they've all sort of come true.

410
00:25:41,360 --> 00:25:48,880
So the first and the paramount was the concept of moving care from hospital, surgery center,

411
00:25:48,880 --> 00:25:54,480
tertiary care center to the home as much as possible, cheaper, faster, better, more convenient

412
00:25:54,480 --> 00:25:57,120
and using technology in order to do that.

413
00:25:57,120 --> 00:26:01,360
The first iteration or generation of that tech was sort of remote monitoring, IOT, host

414
00:26:01,360 --> 00:26:07,040
of techniques such as that companies like Lavongo and even Fitbits and that kind of

415
00:26:07,040 --> 00:26:08,200
thing have used.

416
00:26:08,200 --> 00:26:14,600
So a concept of moving, it's not necessary to treat a copper strep in a tertiary care

417
00:26:14,600 --> 00:26:16,480
hospital at three grand a visit.

418
00:26:16,480 --> 00:26:20,040
You can do that at home, telemedicine for 15 bucks.

419
00:26:20,040 --> 00:26:26,840
So that is a generic concept of moving care away from the sort of very expensive to cheaper

420
00:26:26,840 --> 00:26:27,840
care centers.

421
00:26:27,840 --> 00:26:31,760
That was a big thesis of which we made some investments around that happening into a company

422
00:26:31,760 --> 00:26:36,460
called Carbon Health and a couple of others.

423
00:26:36,460 --> 00:26:42,840
Second was a true understanding that I think the omics space is going to revolutionize

424
00:26:42,840 --> 00:26:47,760
a lot of what we do whether it's genomics, protanomics, metabolomics, radioma, all of

425
00:26:47,760 --> 00:26:49,640
these omic kind of things.

426
00:26:49,640 --> 00:26:55,200
Basically driven towards a better understanding of each individual rather than at a population

427
00:26:55,200 --> 00:26:56,200
level.

428
00:26:56,200 --> 00:27:00,240
And so I can tell you personally, we see this all the time, people that come into the emergency

429
00:27:00,240 --> 00:27:06,000
room, they have abdominal pain, trigger a CT, which is a thousand dollar test.

430
00:27:06,000 --> 00:27:12,000
They had some bad chicken and they go home because it's a normal CT.

431
00:27:12,000 --> 00:27:18,280
I think we had a test at home that we knew that you had a, you know, whatever it may

432
00:27:18,280 --> 00:27:22,840
be, a tryptophan allergy and you probably shouldn't drink this kind of red wine.

433
00:27:22,840 --> 00:27:25,560
And it's because you had that red wine with your chicken and that's why you have belly

434
00:27:25,560 --> 00:27:26,560
pain.

435
00:27:26,560 --> 00:27:30,140
And we just saved the system a thousand dollars, an ER visit and a CT scan.

436
00:27:30,140 --> 00:27:33,720
So that personalized medicine component of it, I think is coming.

437
00:27:33,720 --> 00:27:38,560
And we're seeing that with companies like Everly Well and a lot of the home testing,

438
00:27:38,560 --> 00:27:42,760
you know, home testing has become much more robust and a lot of companies doing so just,

439
00:27:42,760 --> 00:27:45,280
you know, either whether it's blood, urine.

440
00:27:45,280 --> 00:27:51,520
I was on the board of a company that has a breast cancer detection algorithm out of tears.

441
00:27:51,520 --> 00:27:56,980
So lots of that kind of stuff, saliva, tears, sweat, blood, urine, stool, that having these

442
00:27:56,980 --> 00:28:01,880
tests that will give us a lot more information about an individual that can give us more

443
00:28:01,880 --> 00:28:04,960
preventative healthcare.

444
00:28:04,960 --> 00:28:09,440
Number three was some novel techniques around the finance of healthcare.

445
00:28:09,440 --> 00:28:11,120
So we were an early investor in Oscar.

446
00:28:11,120 --> 00:28:14,800
This is a healthcare insurance company in New York that took advantage of a lot of the

447
00:28:14,800 --> 00:28:15,800
Obamacare.

448
00:28:15,800 --> 00:28:17,480
They were doing some very interesting things with tech.

449
00:28:17,480 --> 00:28:21,040
You know, they were able to, for example, they were one of the first ones who would

450
00:28:21,040 --> 00:28:22,920
give you a fitness tracker.

451
00:28:22,920 --> 00:28:26,840
And if you were using the fitness tracker and walking a certain amount every day, you

452
00:28:26,840 --> 00:28:29,120
would get a reduction on your rates.

453
00:28:29,120 --> 00:28:35,840
They understood, you know, they would look at clusters of claims and they, you know,

454
00:28:35,840 --> 00:28:39,920
I remember a very good example where they had a cluster of asthmatic kids in one area

455
00:28:39,920 --> 00:28:44,280
in New York and they realized what was happening was that there was an infestation of bed bugs

456
00:28:44,280 --> 00:28:47,880
and these bed bugs were eliciting asthmatic reactions in these kids.

457
00:28:47,880 --> 00:28:52,520
So the insurance company paid for exterminators to go exterminate these bed bugs because the

458
00:28:52,520 --> 00:28:57,280
Senate exterminated it out for 59 bucks was cheaper than three hospital visits for the

459
00:28:57,280 --> 00:28:58,280
asthmatic kid.

460
00:28:58,280 --> 00:29:02,560
So, you know, all the understandings of sort of payment mechanisms, both whether it's a

461
00:29:02,560 --> 00:29:07,640
pay-by-their, but whether it's third-party administrators, whether it's ACOs or whatever

462
00:29:07,640 --> 00:29:12,280
it may be, the novel mechanisms of payment for healthcare that can make the system more

463
00:29:12,280 --> 00:29:18,440
efficient and not necessarily do sort of just the conventional insurance model.

464
00:29:18,440 --> 00:29:22,200
And then the last is just AI as a generic catch-all for sort of things.

465
00:29:22,200 --> 00:29:27,200
And, you know, generative AI and diagnostic AI, I hesitate to use that term because every

466
00:29:27,200 --> 00:29:32,320
company we look at now in 2022 has some AI component and it doesn't really mean anything.

467
00:29:32,320 --> 00:29:39,000
But I think there is a rule, especially as a radiologist, we see a ton of AI and my clinical

468
00:29:39,000 --> 00:29:43,200
trial company that we had worked with that did almost all the FDA approvals for all the

469
00:29:43,200 --> 00:29:45,200
AI algorithms and radiology.

470
00:29:45,200 --> 00:29:49,480
So we had a lot of exposure and a lot of work in that space.

471
00:29:49,480 --> 00:29:53,520
But I think that's something that is going to keep on growing.

472
00:29:53,520 --> 00:29:59,120
Okay, I think I have a trillion questions.

473
00:29:59,120 --> 00:30:05,840
The one thing I'll ask first, I ask you the next question, the fitness tracker for reduction

474
00:30:05,840 --> 00:30:07,920
on rates, did that work?

475
00:30:07,920 --> 00:30:08,920
No.

476
00:30:08,920 --> 00:30:11,040
So I think there was two problems.

477
00:30:11,040 --> 00:30:15,800
One is they were using Misfit, which was an old school tracker, which, you know, there's

478
00:30:15,800 --> 00:30:17,440
a limited amount of information.

479
00:30:17,440 --> 00:30:21,720
And we've now invested in a company that is actually the reiteration of Jawbone.

480
00:30:21,720 --> 00:30:25,360
I don't know if you remember that Jawbone had a headset and a boombox.

481
00:30:25,360 --> 00:30:29,160
Well, they've like come back and they actually have a fitness tracker that they went and

482
00:30:29,160 --> 00:30:33,420
bought a bunch of technology for a bunch of sensors, green light sensors from Phillips.

483
00:30:33,420 --> 00:30:38,680
But they're building an ecosystem around a full fitness tracker for humans.

484
00:30:38,680 --> 00:30:42,420
And so you think about it sort of like in your car where you get a red light for when

485
00:30:42,420 --> 00:30:44,160
something's going wrong with the engine.

486
00:30:44,160 --> 00:30:49,160
It's a tracker for it's tracking everything 24 seven in real time, but then informing

487
00:30:49,160 --> 00:30:50,580
you when there's when there's an issue.

488
00:30:50,580 --> 00:30:53,800
So I think tracker technology has progressed a significant amount.

489
00:30:53,800 --> 00:30:54,800
That was number one.

490
00:30:54,800 --> 00:30:57,680
And number two, just adoption by people.

491
00:30:57,680 --> 00:31:02,080
You know, at the end of the day, the problem with that technology was people weren't walking.

492
00:31:02,080 --> 00:31:06,480
Like it wasn't enough of an incentive for them to say, if I get three dollars off my

493
00:31:06,480 --> 00:31:10,160
insurance premium, but I have to walk seven miles a day, I'm not going to do it.

494
00:31:10,160 --> 00:31:11,700
We have more data now.

495
00:31:11,700 --> 00:31:16,960
What is beneficial both from a human hack perspective, from both a diet, from sleep,

496
00:31:16,960 --> 00:31:18,480
from all of these things.

497
00:31:18,480 --> 00:31:20,760
Again, we're talking about personalized medicine, right?

498
00:31:20,760 --> 00:31:23,920
Ketosis, prognomics, intermittent fasting, walking, all of these things.

499
00:31:23,920 --> 00:31:28,280
We just have a better understanding so we can be more meaningful now about what we're

500
00:31:28,280 --> 00:31:32,840
what an insurance company would recommend if you're trying to get a true reduction in

501
00:31:32,840 --> 00:31:33,840
rates.

502
00:31:33,840 --> 00:31:35,960
OK, perfect.

503
00:31:35,960 --> 00:31:39,200
Let's talk about how we make decisions.

504
00:31:39,200 --> 00:31:42,640
And the basis of this is health care is a need.

505
00:31:42,640 --> 00:31:43,980
It's not a want.

506
00:31:43,980 --> 00:31:46,400
We pay more for wants than our needs.

507
00:31:46,400 --> 00:31:51,680
If we look at Robert Cialdini's principles of marketing, and I'm a big fan of his reciprocity,

508
00:31:51,680 --> 00:31:57,680
commitment, consistency, social proof, liking and scarcity, needing something is not a principle.

509
00:31:57,680 --> 00:32:02,760
How do we gamify health care so prevention is a want, not a need?

510
00:32:02,760 --> 00:32:05,680
Yeah, I mean, it's interesting.

511
00:32:05,680 --> 00:32:11,920
I would say, if I put my investor hat on, I don't think I'm not prepared right now to

512
00:32:11,920 --> 00:32:16,720
make bets in gamifying health care just because human behavior, the way the system is set

513
00:32:16,720 --> 00:32:19,120
up probably doesn't lend itself to being successful.

514
00:32:19,120 --> 00:32:22,920
I think the things that we've the companies we've seen have tried to gamify things have

515
00:32:22,920 --> 00:32:25,760
not traditionally been overly successful.

516
00:32:25,760 --> 00:32:30,400
Not that it can't be done in specific things for children, for you know, it works.

517
00:32:30,400 --> 00:32:35,240
But I think as an overall health care system, I think that's difficult to do.

518
00:32:35,240 --> 00:32:41,480
So I think, you know, for me personally, some of that the motivation to do that again is

519
00:32:41,480 --> 00:32:44,840
I strongly believe, unfortunately, or fortunately, is driven by money.

520
00:32:44,840 --> 00:32:49,400
You know, if your insurance rates are lower, we've seen and shown that people are motivated

521
00:32:49,400 --> 00:32:50,480
to do things.

522
00:32:50,480 --> 00:32:53,920
And then there's an upper and a lower band of what is considered reasonable, right?

523
00:32:53,920 --> 00:32:58,600
Like, you know, if we said to you, we'll reduce your insurance by, like I said, $5, but you

524
00:32:58,600 --> 00:33:00,720
have to walk 10 miles a day, you're not going to do it.

525
00:33:00,720 --> 00:33:05,000
But if we reduce your insurance by $50, and you have to walk three miles, maybe you'll

526
00:33:05,000 --> 00:33:06,000
do it.

527
00:33:06,000 --> 00:33:10,120
So there's some upper and lower band of human dynamics that you just need to understand.

528
00:33:10,120 --> 00:33:14,120
And we're looking and working with insurance providers who are trying to figure that out.

529
00:33:14,120 --> 00:33:18,000
You know, in a they're trying to figure out in a smaller cohort, we're working with a

530
00:33:18,000 --> 00:33:21,160
company that works only with small businesses, and they have a lot of healthy population.

531
00:33:21,160 --> 00:33:24,280
And so they're able to do that because they are there.

532
00:33:24,280 --> 00:33:27,280
There's not a lot of variation between their sickest and their healthiest person.

533
00:33:27,280 --> 00:33:29,920
So they have an internal control that they can look at.

534
00:33:29,920 --> 00:33:36,680
So I think it's it's understanding human behavior and human dynamic, and then trying to layer

535
00:33:36,680 --> 00:33:44,160
on realistic things that will allow those things to be to be valuable.

536
00:33:44,160 --> 00:33:47,160
And I remember, you know, there was a time when I was looking at, you know, trying to

537
00:33:47,160 --> 00:33:48,160
health hack myself.

538
00:33:48,160 --> 00:33:51,600
And I was watching these movies, Game Changers and Fat, Sick and Nearly Dead.

539
00:33:51,600 --> 00:33:54,520
And this guy was like juicing himself across the country.

540
00:33:54,520 --> 00:33:58,520
And I remember I live in Texas, but I remember this very vividly sat down with these these

541
00:33:58,520 --> 00:34:01,080
people in a in some barbecue place in Texas.

542
00:34:01,080 --> 00:34:03,760
And I said, you know, I'm juicing my way across the country.

543
00:34:03,760 --> 00:34:07,000
And the guy said, I'd rather die than not eat barbecue.

544
00:34:07,000 --> 00:34:10,240
And I think that that sentiment is probably there are people who are like, it doesn't

545
00:34:10,240 --> 00:34:11,240
matter.

546
00:34:11,240 --> 00:34:16,760
I'd rather smoke, drink, eat this, that, whatever, and just die happy than, you know, eat lettuce

547
00:34:16,760 --> 00:34:19,280
every day for 30 years and live an extra 20.

548
00:34:19,280 --> 00:34:22,600
I mean, there's just there's a human dynamic component that you have to understand.

549
00:34:22,600 --> 00:34:25,760
And it's very population driven in the population you're serving.

550
00:34:25,760 --> 00:34:29,920
And so it's a local and regional thing that you have to look at.

551
00:34:29,920 --> 00:34:33,720
And I think the companies that we're looking at are successful, have different models for

552
00:34:33,720 --> 00:34:34,720
different regions.

553
00:34:34,720 --> 00:34:35,720
Perfect.

554
00:34:35,720 --> 00:34:36,720
I'll say something a little bit inappropriate.

555
00:34:36,720 --> 00:34:37,720
I don't know if I should.

556
00:34:37,720 --> 00:34:38,720
I looked into juicing.

557
00:34:38,720 --> 00:34:45,720
And the reason I didn't do it is because there was a lot of mention of enemas.

558
00:34:45,720 --> 00:34:46,720
Yeah.

559
00:34:46,720 --> 00:34:49,720
Let's talk about each other.

560
00:34:49,720 --> 00:34:50,720
Yeah, exactly.

561
00:34:50,720 --> 00:34:54,960
But not for me, just to clarify.

562
00:34:54,960 --> 00:35:00,420
Let's talk about how to best incentivize people for healthy behavior.

563
00:35:00,420 --> 00:35:03,680
And I'll just give you two options to make it simple.

564
00:35:03,680 --> 00:35:08,360
Is it better to tug at loss prevention, which is essentially reducing rates or asking people

565
00:35:08,360 --> 00:35:13,320
for money and giving them back some if they meet the targets?

566
00:35:13,320 --> 00:35:19,960
Or is it better to tug at positive reinforcement or give them money on top of whatever the

567
00:35:19,960 --> 00:35:20,960
rates are?

568
00:35:20,960 --> 00:35:27,960
And it's more of a way of how you phrase or deliver this reward that is essentially the

569
00:35:27,960 --> 00:35:29,000
game here.

570
00:35:29,000 --> 00:35:31,480
Which one do you think is better?

571
00:35:31,480 --> 00:35:37,280
I would say all of the education and reading that I've done would illustrate that positive

572
00:35:37,280 --> 00:35:38,880
reinforcement is better than negative.

573
00:35:38,880 --> 00:35:42,000
Carrots are better than sticks.

574
00:35:42,000 --> 00:35:46,140
I would say that my real world experience of the same thing and me personally has been

575
00:35:46,140 --> 00:35:51,600
a reduction in rate or whatever is more motivating than the other side.

576
00:35:51,600 --> 00:35:54,120
So I don't know the answer to that question.

577
00:35:54,120 --> 00:35:56,720
We've looked at companies that have done both.

578
00:35:56,720 --> 00:36:02,240
So we've made investments in the secondary healthcare companies that run incentive programs.

579
00:36:02,240 --> 00:36:04,600
If you join the gym, you get a discount.

580
00:36:04,600 --> 00:36:07,760
If you do something else, you get a free Apple Watch.

581
00:36:07,760 --> 00:36:10,820
So that's the two things you're talking about.

582
00:36:10,820 --> 00:36:13,560
One is a reduction because you do some of the other is you get a free Apple Watch if

583
00:36:13,560 --> 00:36:15,560
you do this.

584
00:36:15,560 --> 00:36:20,060
And I think when they've studied that, it's been very, again, very heterogeneous depending

585
00:36:20,060 --> 00:36:22,560
on the population they're serving.

586
00:36:22,560 --> 00:36:24,880
And it's not one size fits all.

587
00:36:24,880 --> 00:36:28,960
A 75-year-old probably doesn't care about getting an Apple Watch.

588
00:36:28,960 --> 00:36:31,560
A 21-year-old wants an Apple Watch.

589
00:36:31,560 --> 00:36:36,320
The 21-year-old doesn't understand a reduction in rate will give you money that you can compound

590
00:36:36,320 --> 00:36:39,560
interest for the rest of your life and become a millionaire because you are healthy.

591
00:36:39,560 --> 00:36:44,880
The 75-year-old, so it's very different and you have to, it has to be applicable to the

592
00:36:44,880 --> 00:36:47,000
population that you're trying to serve.

593
00:36:47,000 --> 00:36:48,680
And that I think has been part of the problem.

594
00:36:48,680 --> 00:36:52,080
It's a one size fits all and it doesn't fit all.

595
00:36:52,080 --> 00:36:56,520
And so some of these innovative payments, like we have a family of four, we're fairly

596
00:36:56,520 --> 00:36:57,520
healthy.

597
00:36:57,520 --> 00:36:59,080
We use almost no healthcare.

598
00:36:59,080 --> 00:37:04,360
I pay a huge deductible, a huge premium every month, but we don't use healthcare because

599
00:37:04,360 --> 00:37:06,640
we're at a phase in our life where we're healthy.

600
00:37:06,640 --> 00:37:07,640
We'll need that later.

601
00:37:07,640 --> 00:37:11,680
Well, if you gave me an incentive to save some money and do things that would set up

602
00:37:11,680 --> 00:37:14,960
the insurance company and you could amortize that over some period of years, that would

603
00:37:14,960 --> 00:37:16,120
be a novel way of doing it.

604
00:37:16,120 --> 00:37:19,560
So we're looking at sort of those kinds of models where insurance companies are saying,

605
00:37:19,560 --> 00:37:21,200
let's take a small cohort.

606
00:37:21,200 --> 00:37:25,080
We have to analyze you as a cohort and see whether this fits for you.

607
00:37:25,080 --> 00:37:28,280
Traditionally, that's been too much manpower to do that.

608
00:37:28,280 --> 00:37:33,900
But now with algorithms, AI, whatever it may be, you can get that information on a much

609
00:37:33,900 --> 00:37:36,920
more efficient basis and make those decisions.

610
00:37:36,920 --> 00:37:41,860
So I mean, we looked at a company that on your phone, it can read all your vitals.

611
00:37:41,860 --> 00:37:47,040
So just by looking at the screen, it projects a red light, projects it off your skin, gets

612
00:37:47,040 --> 00:37:51,640
heart rate, temperature, O2 saturation, blood pressure, all from that.

613
00:37:51,640 --> 00:37:54,360
They're starting a pilot with an insurance company in Europe.

614
00:37:54,360 --> 00:37:56,520
So if you not even, you don't even have to check in.

615
00:37:56,520 --> 00:38:01,200
If you put the app on your phone, when you do FaceTime or whatever, or unlock your phone,

616
00:38:01,200 --> 00:38:03,200
it records that data and sends it to the insurance company.

617
00:38:03,200 --> 00:38:07,600
So now insurance company has a longitudinal view of your blood pressure, your heart rate,

618
00:38:07,600 --> 00:38:11,040
and now they can do your premiums without you having to do anything.

619
00:38:11,040 --> 00:38:16,280
That kind of novel stuff is I think where the future of this is where you can now, like

620
00:38:16,280 --> 00:38:21,480
you said, cancer and surgery, you can still pay for quality up there.

621
00:38:21,480 --> 00:38:25,400
Because the number of people using that is less out of the general population, but it's

622
00:38:25,400 --> 00:38:28,520
not the general population subsidizing that anymore.

623
00:38:28,520 --> 00:38:33,200
You can actually divert dollars where they need to go and save money where it doesn't

624
00:38:33,200 --> 00:38:37,940
need to go and have a very novel sort of payment and an insurance mechanism as opposed to this

625
00:38:37,940 --> 00:38:39,240
one size fits all.

626
00:38:39,240 --> 00:38:41,560
Yeah, I think that makes sense.

627
00:38:41,560 --> 00:38:48,960
Risk pooling is needed, but distribution of that risk can be triaged and personalized.

628
00:38:48,960 --> 00:38:51,200
What's the end game for you, Amit?

629
00:38:51,200 --> 00:38:57,640
If we define retirement by when you reach a point where every day is complete in itself,

630
00:38:57,640 --> 00:39:02,960
you're looking forward to things, but you're perfectly happy with the day as it was.

631
00:39:02,960 --> 00:39:05,420
Would you consider self-retire right now?

632
00:39:05,420 --> 00:39:10,900
And then the second part of this question is, I'll take you 50 years down the line.

633
00:39:10,900 --> 00:39:17,000
What do the last five years of your life look like in an ideal world?

634
00:39:17,000 --> 00:39:21,400
I think my wife probably has a very different answer to this question than I do.

635
00:39:21,400 --> 00:39:24,000
So for me, I don't think there is retirement.

636
00:39:24,000 --> 00:39:28,620
Part of what I like about investing and what I like is I pick a topic every year and go

637
00:39:28,620 --> 00:39:34,200
deep and try to understand, read, experience, whatever it may be around that topic.

638
00:39:34,200 --> 00:39:35,800
And there's no end of topics.

639
00:39:35,800 --> 00:39:37,920
And I enjoy doing that.

640
00:39:37,920 --> 00:39:39,840
So I'll continue to do that.

641
00:39:39,840 --> 00:39:44,480
And as long as I can be in it, obviously, there's a potential finite life as an investor.

642
00:39:44,480 --> 00:39:47,360
I mean, if your funds aren't returning, then there is no next fund.

643
00:39:47,360 --> 00:39:50,080
And it may be self-fulfilling that it's over for me.

644
00:39:50,080 --> 00:39:57,600
I'll continue to do the intellectual and experiential investigation that I do now on a yearly basis.

645
00:39:57,600 --> 00:40:02,920
But I think there's a physical component to when do you tire out?

646
00:40:02,920 --> 00:40:10,680
I stand in a lab 13 hours a day wearing lead, heavy lead for when I'm doing cases.

647
00:40:10,680 --> 00:40:14,880
At some point, physically, you wear out from being able to do that.

648
00:40:14,880 --> 00:40:15,880
But I'm not at that point.

649
00:40:15,880 --> 00:40:17,480
So I think I enjoy that.

650
00:40:17,480 --> 00:40:21,520
I'm now that I'll keep doing that until physically I'm not capable.

651
00:40:21,520 --> 00:40:24,280
But one thing that I've learned to do is multitask.

652
00:40:24,280 --> 00:40:28,880
And so while I can pursue these intellectual things at my level of my satisfaction, I'll

653
00:40:28,880 --> 00:40:30,980
just keep doing that till I can.

654
00:40:30,980 --> 00:40:35,360
So I don't know what the last five years, that last five is.

655
00:40:35,360 --> 00:40:40,240
I think there's a sentence that, tell me how I'm going to die so I can avoid it.

656
00:40:40,240 --> 00:40:47,600
So when that's coming, I hope I have some personalized medicine, genomic, proteomics

657
00:40:47,600 --> 00:40:53,240
markers that will tell me that you have 447 days left.

658
00:40:53,240 --> 00:40:54,520
Would you want to know?

659
00:40:54,520 --> 00:40:56,280
Yeah, I think I would.

660
00:40:56,280 --> 00:41:00,160
I mean, my personality is I'd rather know than not know.

661
00:41:00,160 --> 00:41:01,900
Both our kids, I'm a radiologist.

662
00:41:01,900 --> 00:41:04,640
We ultrasounded them in the first time.

663
00:41:04,640 --> 00:41:07,800
We knew what sex of kids we were going to have before we had them.

664
00:41:07,800 --> 00:41:08,800
There was no surprises.

665
00:41:08,800 --> 00:41:10,880
We're not doing, there was no gender reveal stuff.

666
00:41:10,880 --> 00:41:15,680
So yeah, I've always been a personality that I'd rather know and deal with what it is and

667
00:41:15,680 --> 00:41:17,880
have the knowledge rather than the surprise.

668
00:41:17,880 --> 00:41:19,760
The surprise doesn't engage me.

669
00:41:19,760 --> 00:41:20,760
Interesting.

670
00:41:20,760 --> 00:41:26,160
What's the perfect birthday for you?

671
00:41:26,160 --> 00:41:31,160
You know, I mean, I like to, you know, this sort of, I like to work.

672
00:41:31,160 --> 00:41:32,480
Like I like to do things.

673
00:41:32,480 --> 00:41:37,000
And so it doesn't necessarily have to be medicine or investing or these other things that I

674
00:41:37,000 --> 00:41:38,840
do, but active things.

675
00:41:38,840 --> 00:41:41,320
So obviously spend time.

676
00:41:41,320 --> 00:41:42,560
Family is obviously very important.

677
00:41:42,560 --> 00:41:49,160
You know, we have young kids, so probably that's the majority of it is that carve out

678
00:41:49,160 --> 00:41:55,040
time specifically to be with family during what is usually a very hectic week.

679
00:41:55,040 --> 00:42:01,720
Do you work on your birthday?

680
00:42:01,720 --> 00:42:04,200
I work if I have to.

681
00:42:04,200 --> 00:42:11,440
If I, if I can maneuver out of it, I would, but if I have to, I would.

682
00:42:11,440 --> 00:42:17,480
Would you classify yourself as an operator, someone excellent at execution or a visionary?

683
00:42:17,480 --> 00:42:23,840
And when you look at a founding team, do you look for both and which one is easier to replace

684
00:42:23,840 --> 00:42:27,600
or find source externally from your network?

685
00:42:27,600 --> 00:42:31,800
It's an interesting question.

686
00:42:31,800 --> 00:42:35,320
It's hard to say you're a vision, in my opinion, to say you're a visionary without being an

687
00:42:35,320 --> 00:42:36,760
egotistical visionary.

688
00:42:36,760 --> 00:42:39,640
So I'm going to, I'm going to dance away from that.

689
00:42:39,640 --> 00:42:45,240
But I think, yeah, my, I think my skillset has been more around strategy and vision than

690
00:42:45,240 --> 00:42:46,240
operation.

691
00:42:46,240 --> 00:42:51,120
You know, when we accepted this company, my company, I mean, I got on the ground and plugged

692
00:42:51,120 --> 00:42:57,560
in computers and wrote SOPs and did whatever it took as an operator, more as a skillset,

693
00:42:57,560 --> 00:43:02,480
learning how to do it so that, you know, when someone failed at it, I was the backup, I

694
00:43:02,480 --> 00:43:03,600
could do it.

695
00:43:03,600 --> 00:43:08,040
I wouldn't say I enjoyed the operational component as much as the strategy, vision, guidance

696
00:43:08,040 --> 00:43:09,040
part of it.

697
00:43:09,040 --> 00:43:12,240
And so I think also, you know, as you look at things that you do in your life and you

698
00:43:12,240 --> 00:43:16,080
converge into things that are more your personality, you know, as it takes time to figure that

699
00:43:16,080 --> 00:43:20,360
out, you know, that's why I prefer to be on the board of a company and help with direction

700
00:43:20,360 --> 00:43:26,200
and strategy using the experience from being an operator to help them rather than, than

701
00:43:26,200 --> 00:43:27,200
being the operator.

702
00:43:27,200 --> 00:43:29,720
Having said that, we've been toying with being an operator again.

703
00:43:29,720 --> 00:43:32,920
Your friend of mine has come to me and said, like, let's start another company.

704
00:43:32,920 --> 00:43:39,680
And so maybe that does come again, but it has to be the right idea and the right execution.

705
00:43:39,680 --> 00:43:45,400
But I prefer the, you know, he's an, he's an incredible operator and has operated at

706
00:43:45,400 --> 00:43:47,160
a very high level.

707
00:43:47,160 --> 00:43:50,440
So we compliment each other really well and that I'm more strategy vision.

708
00:43:50,440 --> 00:43:53,520
He's very operate, he's less strategy vision.

709
00:43:53,520 --> 00:43:56,280
And you know, so that compliment works well.

710
00:43:56,280 --> 00:44:01,240
And do you look for both in a founding team and which one is easier to source from your

711
00:44:01,240 --> 00:44:02,240
network?

712
00:44:02,240 --> 00:44:03,240
I mean, absolutely.

713
00:44:03,240 --> 00:44:06,200
I look for it in a founding team.

714
00:44:06,200 --> 00:44:08,640
I mean, it's always in the back of your mind.

715
00:44:08,640 --> 00:44:12,320
The issue is more at the stage of when you're making an investment.

716
00:44:12,320 --> 00:44:17,440
So if it's a series CD, whatever, I mean, they figured it out, right?

717
00:44:17,440 --> 00:44:20,040
I mean, you're not getting to a hundred million in revenue.

718
00:44:20,040 --> 00:44:26,360
If the CEO is a tech guy, was no tech person who has no operational experience and can't

719
00:44:26,360 --> 00:44:27,360
motivate people.

720
00:44:27,360 --> 00:44:28,800
Like it's just not going to happen.

721
00:44:28,800 --> 00:44:29,800
Right.

722
00:44:29,800 --> 00:44:35,080
Or it's very rare that that happens at the stage where we invest, which is seed and a,

723
00:44:35,080 --> 00:44:39,920
all of that is very important, but hasn't necessarily always manifested.

724
00:44:39,920 --> 00:44:44,640
So we see lots of companies where the CEO is great for the seed or for the early a,

725
00:44:44,640 --> 00:44:49,880
but once you get to a B or a C, they really need to step aside and bring in adult supervision,

726
00:44:49,880 --> 00:44:52,000
so to speak, because it's not there.

727
00:44:52,000 --> 00:44:56,320
They're a great founding CEO and they're the ones, you know, two guys in a garage, you

728
00:44:56,320 --> 00:45:01,280
know, all of that sort of stuff, but they're not the ones that sit with an investment bank

729
00:45:01,280 --> 00:45:03,760
to negotiate, you know, book running an IPO.

730
00:45:03,760 --> 00:45:05,600
I mean, that's not who they are, right?

731
00:45:05,600 --> 00:45:08,840
They don't, the attention to detail, the experience, whatever it may be.

732
00:45:08,840 --> 00:45:12,400
And so that, I think it's very much dependent on what stage you're investing at to figure

733
00:45:12,400 --> 00:45:14,560
out who the right founders are.

734
00:45:14,560 --> 00:45:15,560
Okay.

735
00:45:15,560 --> 00:45:22,440
You know, something I'm doing right now is looking at companies that have made it and

736
00:45:22,440 --> 00:45:25,240
going back to see if I would have invested in them.

737
00:45:25,240 --> 00:45:29,560
And oftentimes the answer I'm getting is no.

738
00:45:29,560 --> 00:45:33,880
The companies I'm talking about, I'll just say some names as Romans and Four Hymns, you

739
00:45:33,880 --> 00:45:39,640
know, they've done really well and they're pivoting to more being almost at home hybrid

740
00:45:39,640 --> 00:45:43,760
health system, which is great, but that's not what they were when they started.

741
00:45:43,760 --> 00:45:45,200
Do you do this exercise?

742
00:45:45,200 --> 00:45:49,400
Do you go back and look at companies you did not invest in and maybe you didn't come across

743
00:45:49,400 --> 00:45:52,040
or you did and you passed on them?

744
00:45:52,040 --> 00:45:56,720
And then do you change your investment thesis to any extent and say if you get one of these

745
00:45:56,720 --> 00:46:01,320
or say if you get 10 of these companies?

746
00:46:01,320 --> 00:46:05,240
I would tell you, I don't do the exercise, but my partners definitely do the exercise

747
00:46:05,240 --> 00:46:07,240
for me because they send me the deals.

748
00:46:07,240 --> 00:46:08,500
You said no on this.

749
00:46:08,500 --> 00:46:10,040
Look at this billion dollar exit.

750
00:46:10,040 --> 00:46:11,680
So no, I'm kidding.

751
00:46:11,680 --> 00:46:12,680
They don't.

752
00:46:12,680 --> 00:46:14,120
Thankfully we have a great partnership.

753
00:46:14,120 --> 00:46:15,120
We don't do that.

754
00:46:15,120 --> 00:46:23,960
But I would say I do, I look at that from a, from a just an experiential intellectual

755
00:46:23,960 --> 00:46:24,960
perspective.

756
00:46:24,960 --> 00:46:27,920
I don't beat myself up about not making the investment.

757
00:46:27,920 --> 00:46:31,200
I mean, some of these ones that you're talking about, we had the opportunity to invest.

758
00:46:31,200 --> 00:46:35,160
I think if you stay fundamental to a thesis, some things just don't fit.

759
00:46:35,160 --> 00:46:37,880
And those, the ones that you hear about are the outliers.

760
00:46:37,880 --> 00:46:42,320
One thing I've learned when I got into investing, I mean, we look at 70, 80, a hundred deals

761
00:46:42,320 --> 00:46:43,320
a week, right?

762
00:46:43,320 --> 00:46:46,080
But you can't invest in all of them.

763
00:46:46,080 --> 00:46:50,720
And the ones that make the headlines are the ones that break away like that.

764
00:46:50,720 --> 00:46:56,040
But categorically, if it doesn't fit your thesis, then you'll make a mistake 99 times

765
00:46:56,040 --> 00:46:57,960
and have success once.

766
00:46:57,960 --> 00:47:03,040
And I think that's okay in venture if you've right sized your venture investments probably,

767
00:47:03,040 --> 00:47:07,160
you know, that one has a hundred billion exits, so to speak.

768
00:47:07,160 --> 00:47:11,760
It's fine that the 99 went to zero in venture math, but you better be pretty sure that you

769
00:47:11,760 --> 00:47:16,120
can do that on a repeated basis or there is no fun two or fun three or fun four.

770
00:47:16,120 --> 00:47:22,360
So I don't beat myself about the misses, but I do, I do look at them and I try to understand,

771
00:47:22,360 --> 00:47:26,800
you know, did I actually miss something or is this off thesis or this is a one off?

772
00:47:26,800 --> 00:47:30,400
Yeah, I'm really happy you said that.

773
00:47:30,400 --> 00:47:31,400
And I agree with it.

774
00:47:31,400 --> 00:47:34,640
You should not define your thesis chasing after Google.

775
00:47:34,640 --> 00:47:36,560
It was the 18th search engine.

776
00:47:36,560 --> 00:47:40,860
If you decided this wasn't for you after the 16th one, you would have missed out.

777
00:47:40,860 --> 00:47:47,040
So it's something you just have to have an infinite amount of money to pull off, I feel.

778
00:47:47,040 --> 00:47:48,040
Yeah.

779
00:47:48,040 --> 00:47:49,400
I mean, a lot of these businesses, you know, it's interesting.

780
00:47:49,400 --> 00:47:54,880
We've made an investment in a company called Stampede, which is a novel movie studio.

781
00:47:54,880 --> 00:48:02,800
This guy named Greg Silverman, who was the producer manager at Warner Brothers for 20

782
00:48:02,800 --> 00:48:06,160
years, super successful, highest grossing producer of all time.

783
00:48:06,160 --> 00:48:09,920
He did Harry Potter series, Batman, Hangover.

784
00:48:09,920 --> 00:48:12,040
This guy, I mean, this is the guy.

785
00:48:12,040 --> 00:48:19,040
We ran an exercise at our annual meeting last month where he had a grid of seven by seven

786
00:48:19,040 --> 00:48:23,000
movies, which were script titles that were pitched to him.

787
00:48:23,000 --> 00:48:29,760
And we were supposed to pick based on, you know, it was a British kid who is a wizard

788
00:48:29,760 --> 00:48:36,600
who lives below a stairwell and has plays at this school that's, you know, whatever.

789
00:48:36,600 --> 00:48:38,720
And everyone's like, absolutely would never do that.

790
00:48:38,720 --> 00:48:39,720
Right.

791
00:48:39,720 --> 00:48:40,720
It was Harry Potter.

792
00:48:40,720 --> 00:48:44,280
And so it was interesting when you go through what they, you know, his trade craft and his

793
00:48:44,280 --> 00:48:45,840
discipline of picking these movies.

794
00:48:45,840 --> 00:48:49,240
I mean, he can obviously pick winners.

795
00:48:49,240 --> 00:48:52,640
The skill set to pick, I mean, he even said it in our meeting.

796
00:48:52,640 --> 00:48:54,040
We get lucky sometimes, right?

797
00:48:54,040 --> 00:48:56,400
Like there are outliers to the Harry Potter.

798
00:48:56,400 --> 00:49:01,360
He picked he was everyone had rejected that every studio had rejected the books before,

799
00:49:01,360 --> 00:49:05,580
you know, and now look at the one of the great biggest grossing franchises of all time.

800
00:49:05,580 --> 00:49:10,800
So I think you could learn from it, but I don't like you said, I don't we don't I try

801
00:49:10,800 --> 00:49:13,800
not to define by what I consider mistakes.

802
00:49:13,800 --> 00:49:19,040
Do you think you learn more from your successes than your failures?

803
00:49:19,040 --> 00:49:20,320
And I will give my answer.

804
00:49:20,320 --> 00:49:21,320
Yes.

805
00:49:21,320 --> 00:49:23,840
I have learned more from my successes.

806
00:49:23,840 --> 00:49:25,960
And I know this is not the popular answer.

807
00:49:25,960 --> 00:49:32,000
I have learned humility and introspection and reflection from my failures.

808
00:49:32,000 --> 00:49:36,600
But in terms of decision making, I feel like I've learned more from successes.

809
00:49:36,600 --> 00:49:38,800
Yeah, I would agree with that.

810
00:49:38,800 --> 00:49:40,120
I mean, I think I've learned more from the success.

811
00:49:40,120 --> 00:49:45,000
But I think what I've learned from failures is how to be more successful, not necessarily

812
00:49:45,000 --> 00:49:47,880
learn a task or a skill or an idea.

813
00:49:47,880 --> 00:49:54,760
It was, you know, we had we invested in a company where the founder had, you know, we

814
00:49:54,760 --> 00:49:59,960
should have done better due diligence on the founder and his background, etc, etc.

815
00:49:59,960 --> 00:50:03,040
So now, you know, we run a background check on everybody before we make it like those

816
00:50:03,040 --> 00:50:04,040
kinds of things.

817
00:50:04,040 --> 00:50:07,760
They're operational infrastructure things that I think I've learned over the years.

818
00:50:07,760 --> 00:50:12,480
Investor, you know, I've never done a reverse merger IPO with an Israeli shell company.

819
00:50:12,480 --> 00:50:16,560
Well, now we've sort of done it and we understand how to like those kinds of things from those

820
00:50:16,560 --> 00:50:20,760
kinds of failures where we ended up in a situation that we had to navigate out of.

821
00:50:20,760 --> 00:50:25,080
We've learned something that now for other companies that we have, we have a better toolbox

822
00:50:25,080 --> 00:50:27,960
of saying, hey, this is one of the possibilities now.

823
00:50:27,960 --> 00:50:31,640
So I don't for me, it's not been so cut and dry of the conventional, you know, is your

824
00:50:31,640 --> 00:50:34,440
success versus failure learning.

825
00:50:34,440 --> 00:50:37,840
And you know, you teach this to your kids or they tell you to teach it to your kids

826
00:50:37,840 --> 00:50:41,680
that your kids fail so that they can understand success kind of thing.

827
00:50:41,680 --> 00:50:45,220
It's so different, I think, in investing and understanding how this kind of stuff works,

828
00:50:45,220 --> 00:50:51,560
because there's a mechanistic component to just investing, like how to read a term sheet

829
00:50:51,560 --> 00:50:56,760
and how to do a deal and IPO and this and caps and all of these things that come into

830
00:50:56,760 --> 00:51:01,040
liquidity preference, how it all works, that plays such an important role in success or

831
00:51:01,040 --> 00:51:02,040
failure.

832
00:51:02,040 --> 00:51:03,560
You can have something that's very successful.

833
00:51:03,560 --> 00:51:07,440
We're looking at a company right now that they're recapping the company.

834
00:51:07,440 --> 00:51:09,920
We looked a few years ago, it was too rich for our blood.

835
00:51:09,920 --> 00:51:14,880
Now they're recapping the company where the founders have ended up with 1% of the company

836
00:51:14,880 --> 00:51:19,120
after two rounds and that everyone's realized that there is no company unless we recap the

837
00:51:19,120 --> 00:51:21,280
company and give the founders back 30% equity.

838
00:51:21,280 --> 00:51:24,120
You know, that kind of thing is just an experiential thing.

839
00:51:24,120 --> 00:51:26,040
It's got nothing to do with no one fail.

840
00:51:26,040 --> 00:51:29,560
It's just the fundraising went in a way that there were so many people involved that the

841
00:51:29,560 --> 00:51:33,440
cap table got completely squeezed down that these guys sold all their equity out and now

842
00:51:33,440 --> 00:51:34,760
there's no motivation to grow a company.

843
00:51:34,760 --> 00:51:36,160
But the company is good.

844
00:51:36,160 --> 00:51:40,080
So there's investors who are interested in recapping the company.

845
00:51:40,080 --> 00:51:47,000
Okay, let's talk about hiring employees, co-founders and finding a partner.

846
00:51:47,000 --> 00:51:53,040
The lens I want you to answer this question through is Danny Kahneman's decision-making

847
00:51:53,040 --> 00:51:56,040
framework.

848
00:51:56,040 --> 00:52:01,320
I used to hire based on intuition and that was the wrong thing for me.

849
00:52:01,320 --> 00:52:08,580
I've made some terrible mistakes because how do you approach hiring employees and looking

850
00:52:08,580 --> 00:52:14,680
for a co-founder, do you have a structured approach and how do you factor in intuition?

851
00:52:14,680 --> 00:52:20,080
Yeah, I mean, I think co-founder is a very different whole thing, right?

852
00:52:20,080 --> 00:52:24,120
I mean, co-founder is there's a personal component, a professional component, an intellectual

853
00:52:24,120 --> 00:52:25,680
component and a financial component.

854
00:52:25,680 --> 00:52:31,240
I mean, it's very different and that for me and all of the things that I've done where

855
00:52:31,240 --> 00:52:35,960
I've had a co-founder or a partner in a business has been we were friends first and then we

856
00:52:35,960 --> 00:52:38,120
moved into a business relationship.

857
00:52:38,120 --> 00:52:44,520
And we've but very upfront built a business relationship with tax and blocks and tackles

858
00:52:44,520 --> 00:52:49,520
in the documents that shotgun clauses, whatever it was that we understand upfront that if

859
00:52:49,520 --> 00:52:53,200
we get into a fight, which everyone tells you don't do business with your friends, which

860
00:52:53,200 --> 00:52:57,900
I have done business with all my friends is we just build a way that there's a structure

861
00:52:57,900 --> 00:53:00,480
that no one gets into a fight and everyone understands how it's going to be.

862
00:53:00,480 --> 00:53:03,660
So I think as a co-founder, it's a very different exercise.

863
00:53:03,660 --> 00:53:09,880
On the employee side, both my company, our fund, all of these other things, we've tried

864
00:53:09,880 --> 00:53:13,800
360 evaluations and neograms.

865
00:53:13,800 --> 00:53:19,240
There's a host of tools that you fill in these questions, all of this stuff.

866
00:53:19,240 --> 00:53:24,640
I personally have never had that much success with these orchestrated tools.

867
00:53:24,640 --> 00:53:29,300
I think they give you, if you have a limited time of assessment with somebody, they do

868
00:53:29,300 --> 00:53:34,840
give you insight that's very valuable in terms of are they a high performer?

869
00:53:34,840 --> 00:53:36,760
Are they obsessive compulsive?

870
00:53:36,760 --> 00:53:40,480
Whatever these factors that you're looking for.

871
00:53:40,480 --> 00:53:45,160
But I think if you spend enough time with somebody that all of those things as an employer,

872
00:53:45,160 --> 00:53:47,780
as a partner, we're able to garner.

873
00:53:47,780 --> 00:53:49,640
And so that's what we try to focus on.

874
00:53:49,640 --> 00:53:53,840
We try to spend a good amount of time with someone upfront and then to a certain extent

875
00:53:53,840 --> 00:53:56,360
hope for the best.

876
00:53:56,360 --> 00:54:00,720
I've been proven wrong on lots of the 360 evaluations that we've done where the thing

877
00:54:00,720 --> 00:54:04,920
says this person is awesome and they turn out to be awful and vice versa.

878
00:54:04,920 --> 00:54:08,320
We've had people who we thought there's no way this person could succeed and then they're

879
00:54:08,320 --> 00:54:11,680
the most successful person on whatever entity it was.

880
00:54:11,680 --> 00:54:12,680
Okay.

881
00:54:12,680 --> 00:54:13,960
Do you have time for one more question?

882
00:54:13,960 --> 00:54:14,960
If not, we can...

883
00:54:14,960 --> 00:54:15,960
I do.

884
00:54:15,960 --> 00:54:16,960
Yeah.

885
00:54:16,960 --> 00:54:17,960
Let's talk about...

886
00:54:17,960 --> 00:54:23,920
And I had questions for each one of the four pillars and I think maybe in the next round

887
00:54:23,920 --> 00:54:26,040
we'll get to it.

888
00:54:26,040 --> 00:54:28,520
I'm already marking down the next round by the way.

889
00:54:28,520 --> 00:54:33,680
Let's talk about AI and explainability.

890
00:54:33,680 --> 00:54:36,240
I talked to this about Amit from Tao Ventures.

891
00:54:36,240 --> 00:54:37,240
I don't know if you know him.

892
00:54:37,240 --> 00:54:38,240
I do.

893
00:54:38,240 --> 00:54:40,160
We're invested in a company together actually.

894
00:54:40,160 --> 00:54:41,160
Okay.

895
00:54:41,160 --> 00:54:42,160
Yeah.

896
00:54:42,160 --> 00:54:44,560
And he gave an amazing answer.

897
00:54:44,560 --> 00:54:46,360
So the pressure is on Amit.

898
00:54:46,360 --> 00:54:47,360
All right.

899
00:54:47,360 --> 00:54:53,840
AI and explainability in the sense of right now the regulatory framework in healthcare

900
00:54:53,840 --> 00:55:00,600
requires us to an extent to understand what the AI is doing, to understand the networks

901
00:55:00,600 --> 00:55:04,320
or the processes behind the outcomes.

902
00:55:04,320 --> 00:55:07,440
When do we let go of that need for explainability?

903
00:55:07,440 --> 00:55:13,840
If the AI is providing amazing outcomes, maybe outcomes better than humans, when do we say,

904
00:55:13,840 --> 00:55:20,360
okay, let us let this algorithm do its job, provide us with care, with decisions on when

905
00:55:20,360 --> 00:55:25,240
to operate, when not to operate, what drugs to use even, what diagnoses, and maybe we

906
00:55:25,240 --> 00:55:27,640
don't even answer the diagnoses.

907
00:55:27,640 --> 00:55:30,760
At what point do we let go of that need for explainability?

908
00:55:30,760 --> 00:55:34,840
Because by not doing so, we're limiting the scope of AI, I feel.

909
00:55:34,840 --> 00:55:35,840
Yeah.

910
00:55:35,840 --> 00:55:40,320
So interestingly, like radiology has been on the forefront of this, right?

911
00:55:40,320 --> 00:55:44,480
And so through our clinical trial company, we did the approvals for a lot of these AI

912
00:55:44,480 --> 00:55:45,480
products.

913
00:55:45,480 --> 00:55:51,160
The FDA as a regulatory body has already given blessing on not knowing how the AI works.

914
00:55:51,160 --> 00:55:55,520
So in a compositional neural network, in a pure CNN, it truly is a black box, right?

915
00:55:55,520 --> 00:55:57,640
So it's input in, input out, back test.

916
00:55:57,640 --> 00:56:00,960
So no one understands how, I mean, there is no understanding of how the black box works

917
00:56:00,960 --> 00:56:03,880
per se, but those algorithms are approved.

918
00:56:03,880 --> 00:56:09,720
So we have FDA approval for several algorithms that nobody knows how they work, quote unquote,

919
00:56:09,720 --> 00:56:10,720
right?

920
00:56:10,720 --> 00:56:11,720
Oh, perfect.

921
00:56:11,720 --> 00:56:17,400
So AI exists, and the FDA is moving through this in terms of defining, there's a whole

922
00:56:17,400 --> 00:56:18,600
host of secondary.

923
00:56:18,600 --> 00:56:24,720
So AI is going through this evolution of sort of what I call AI 1.0, which has been traditionally

924
00:56:24,720 --> 00:56:28,280
sort of detection technology, and it's not controversial, right?

925
00:56:28,280 --> 00:56:30,660
Is there a mass on this picture?

926
00:56:30,660 --> 00:56:33,200
Is there an abnormality in the CBC?

927
00:56:33,200 --> 00:56:35,160
Is there whatever it may be?

928
00:56:35,160 --> 00:56:36,480
It's not controversial in understanding.

929
00:56:36,480 --> 00:56:40,100
It's a detection technology, and there's been detection technology for a long time.

930
00:56:40,100 --> 00:56:41,560
No one has an issue with that.

931
00:56:41,560 --> 00:56:45,880
The detection is getting better, and the detection is getting to a level where we're asymptotic

932
00:56:45,880 --> 00:56:50,480
and where you cannot, as computing power goes to, the cost of computing goes to zero, this

933
00:56:50,480 --> 00:56:54,160
gets better and you can detect better and be more holistic about it.

934
00:56:54,160 --> 00:56:58,840
This next generation of AI is, which I don't think we're at, which, and I'm going to get

935
00:56:58,840 --> 00:57:02,920
to your point, is sort of where there's an interpretation of what is detected.

936
00:57:02,920 --> 00:57:04,920
So you're making an evaluation.

937
00:57:04,920 --> 00:57:08,120
And so for example, you find a mass on a mammogram.

938
00:57:08,120 --> 00:57:14,440
There's a company that has a heat map that it takes and ingests your BRCA, it ingests

939
00:57:14,440 --> 00:57:18,320
your family history, it ingests your diet, a whole bunch of other risk factors to give

940
00:57:18,320 --> 00:57:22,480
an assessment of whether or not that mass is going to turn into a cancer.

941
00:57:22,480 --> 00:57:25,280
1.0 tells you, is it a cancer or not?

942
00:57:25,280 --> 00:57:30,240
And if it's not, 2.0 tells you, based on all these other risk factors, it's going to become

943
00:57:30,240 --> 00:57:31,380
a cancer.

944
00:57:31,380 --> 00:57:35,600
So that's sort of 2.0, and I think 3.0 is what you're talking about, where it's a true

945
00:57:35,600 --> 00:57:36,600
decision maker.

946
00:57:36,600 --> 00:57:42,160
I think as a society, we probably don't want computers to make that decision, open AI or

947
00:57:42,160 --> 00:57:43,720
all the money, fear and all that.

948
00:57:43,720 --> 00:57:51,360
But I think it's inevitable that a component of what we do will become, and AI will adjunct

949
00:57:51,360 --> 00:57:52,360
what we do.

950
00:57:52,360 --> 00:57:54,880
And imaging is a great example, right?

951
00:57:54,880 --> 00:58:01,120
I think in 2022, as long as I am a licensed physician who is trained, who's willing to

952
00:58:01,120 --> 00:58:07,240
sign off on something, I can use the AI to adjunct and augment what I do.

953
00:58:07,240 --> 00:58:10,800
There's an algorithm that can find a rib fracture on a CT scan.

954
00:58:10,800 --> 00:58:13,880
Does anyone in the world have a problem with me using that algorithm?

955
00:58:13,880 --> 00:58:18,600
Absolutely not, as long as the algorithm detects and I say, I agree, and that's a rib fracture.

956
00:58:18,600 --> 00:58:23,080
So that, I think we're there, and that's where we are today.

957
00:58:23,080 --> 00:58:28,680
So the question becomes is, as we evolve, are we okay with AI making the ultimate decision?

958
00:58:28,680 --> 00:58:33,120
And then the constituents there are going to be, again, the provider, the payer, and

959
00:58:33,120 --> 00:58:34,720
the patient.

960
00:58:34,720 --> 00:58:42,680
I think the patient reflects on society, FDA, everybody else to make sure, just like the

961
00:58:42,680 --> 00:58:48,560
Consumer Protection Bureau protects consumers from ovens going on fire, right?

962
00:58:48,560 --> 00:58:52,200
We as a society expect these regulatory bodies to protect us.

963
00:58:52,200 --> 00:58:53,800
So I don't think the patients make a decision.

964
00:58:53,800 --> 00:58:58,240
The patients say, if the FDA says it's okay, it's okay, as we've done with almost everything

965
00:58:58,240 --> 00:58:59,680
else.

966
00:58:59,680 --> 00:59:04,280
The providers, I think as long as they're getting an economic benefit or do not lose

967
00:59:04,280 --> 00:59:07,400
an economic benefit, are absolutely fine using it.

968
00:59:07,400 --> 00:59:12,440
So in my rib fracture example, if insurance is willing to pay for me to use the rib fracture

969
00:59:12,440 --> 00:59:14,800
algorithm, I have no problem using it.

970
00:59:14,800 --> 00:59:16,920
Makes me a better doctor, better diagnostician.

971
00:59:16,920 --> 00:59:20,040
I miss less fractures all day long.

972
00:59:20,040 --> 00:59:23,640
So it comes down to the payer, right?

973
00:59:23,640 --> 00:59:26,360
And it comes down to money, which is what all of this comes down to.

974
00:59:26,360 --> 00:59:28,920
So the payer says, why am I willing to pay for this?

975
00:59:28,920 --> 00:59:32,760
The only reason I'm willing to pay for this, honestly, right, if we're true to ourselves,

976
00:59:32,760 --> 00:59:35,600
is that it reduces my cost of providing care.

977
00:59:35,600 --> 00:59:39,160
So if we go back to my rib fracture example, and I use this specifically, if you have a

978
00:59:39,160 --> 00:59:43,720
rib fracture on a CT scan that's not significantly displaced, you know what we tell you to do?

979
00:59:43,720 --> 00:59:45,960
Take two Advil and we'll call you in the morning, right?

980
00:59:45,960 --> 00:59:47,680
There's no treatment for that per se.

981
00:59:47,680 --> 00:59:49,880
I'm not talking about a flail chest or a pneumothorax.

982
00:59:49,880 --> 00:59:52,500
I'm just talking about a non-displaced rib fracture.

983
00:59:52,500 --> 00:59:56,200
So at the end of the day for the payer, did it really matter if I found that rib fracture

984
00:59:56,200 --> 00:59:58,280
probably not, right?

985
00:59:58,280 --> 01:00:02,160
Academically, for us, it matters that we don't miss a finding.

986
01:00:02,160 --> 01:00:03,160
For the patient, it matters.

987
01:00:03,160 --> 01:00:06,440
Now they have an explanation for their right side of chest pain.

988
01:00:06,440 --> 01:00:11,320
But the payer, and this at some point was supposition, now it's playing out because

989
01:00:11,320 --> 01:00:16,720
I've seen many, many, many AI companies that have closed shop because they went to the

990
01:00:16,720 --> 01:00:18,640
payer and the payer said no.

991
01:00:18,640 --> 01:00:20,820
They went to the patient and the patient said no.

992
01:00:20,820 --> 01:00:24,500
They went to the hospital, the hospital said no, we're not paying for this.

993
01:00:24,500 --> 01:00:26,800
And they have to close shop because they couldn't find a payer.

994
01:00:26,800 --> 01:00:30,920
So I think that's what ends up driving some of this stuff at the end of the day is, is

995
01:00:30,920 --> 01:00:37,480
there a economic benefit to that algorithm independent of, lots of time there's a clinical

996
01:00:37,480 --> 01:00:41,020
benefit and there's no question there's a clinical benefit.

997
01:00:41,020 --> 01:00:43,520
But that's not what the driver of the technology is.

998
01:00:43,520 --> 01:00:45,740
The driver is who's going to pay for it.

999
01:00:45,740 --> 01:00:51,120
So that's experientially what I've learned as an investor is like when we have these

1000
01:00:51,120 --> 01:00:53,760
companies that come, we say, who's going to pay for this?

1001
01:00:53,760 --> 01:00:56,120
Oh, well, we'll figure that out later.

1002
01:00:56,120 --> 01:00:57,120
Not investing in your company.

1003
01:00:57,120 --> 01:01:02,640
Or I'm going to get a CPT code and I'm going to get reimbursement for this.

1004
01:01:02,640 --> 01:01:03,640
Not investing in that.

1005
01:01:03,640 --> 01:01:07,880
You have not been through the system because the payer is not going to pay for that unless

1006
01:01:07,880 --> 01:01:09,760
you can prove benefit.

1007
01:01:09,760 --> 01:01:16,440
How far are we from the AI being the provider, at least for certain verticals like, you know,

1008
01:01:16,440 --> 01:01:18,960
UTI and birth control refills?

1009
01:01:18,960 --> 01:01:19,960
Yeah.

1010
01:01:19,960 --> 01:01:23,680
I mean, I don't know if you've used chat GPT yet, but I haven't.

1011
01:01:23,680 --> 01:01:25,200
Yeah, I mean, just type in.

1012
01:01:25,200 --> 01:01:26,200
I got a cold one.

1013
01:01:26,200 --> 01:01:27,560
I mean, chat GPT is amazing.

1014
01:01:27,560 --> 01:01:33,120
So AI inner sort of generational AI, I think, is at that level now.

1015
01:01:33,120 --> 01:01:34,120
So I think we're there.

1016
01:01:34,120 --> 01:01:40,960
I mean, if it's algorithmic algorithmic treatment, which is in essence what you know, there's

1017
01:01:40,960 --> 01:01:44,600
a whole host of companies doing asynchronous care, right?

1018
01:01:44,600 --> 01:01:50,240
You have an app, you log on, you say, I've got a cough, I've got a fever, I've got green

1019
01:01:50,240 --> 01:01:52,840
sputum, I've got shortness of breath.

1020
01:01:52,840 --> 01:01:55,120
Pretty high chance you got a pneumonia.

1021
01:01:55,120 --> 01:01:58,360
And if it's hurting on the right side, it's in the right lower lobe, right?

1022
01:01:58,360 --> 01:02:03,260
So that asynchronous data capture, they then send you for a chest x-ray and some labs,

1023
01:02:03,260 --> 01:02:05,120
and then they get that result.

1024
01:02:05,120 --> 01:02:07,400
And then they, you know, you get antibiotics.

1025
01:02:07,400 --> 01:02:10,800
There isn't a need for a level five or an ND visit.

1026
01:02:10,800 --> 01:02:13,640
I mean, you can go see a PA or maybe an NP if you need to.

1027
01:02:13,640 --> 01:02:19,360
But a lot, you know, for the most part, that visit doesn't add AI and even chat GPT, which

1028
01:02:19,360 --> 01:02:24,520
sounds crazy, but conversational AI on asynchronous data capture and treatment, I think is there

1029
01:02:24,520 --> 01:02:25,520
today.

1030
01:02:25,520 --> 01:02:29,160
Now it goes back to the same problem of, you know, again, in the money, talking about the

1031
01:02:29,160 --> 01:02:33,480
money again, there's obviously a, you know, the litiginous part of it.

1032
01:02:33,480 --> 01:02:35,480
There's, you know, we worry about all that.

1033
01:02:35,480 --> 01:02:37,000
What happens when the AI makes a mistake?

1034
01:02:37,000 --> 01:02:38,000
Who do you sue?

1035
01:02:38,000 --> 01:02:39,000
All that kind of stuff.

1036
01:02:39,000 --> 01:02:45,480
But I think as it augments to the position in 2022, it's worth, I mean, there is capability

1037
01:02:45,480 --> 01:02:48,280
there and radiology has been one of those places where it's been first.

1038
01:02:48,280 --> 01:02:53,400
I mean, we augment what we do in mammography with cat with AI, right?

1039
01:02:53,400 --> 01:02:57,560
We bring up the mammogram, we look at it, we hit a button, the algorithm will highlight

1040
01:02:57,560 --> 01:02:59,240
a mass recalcitration.

1041
01:02:59,240 --> 01:03:02,960
We use that and use that to make a better interpretation.

1042
01:03:02,960 --> 01:03:07,280
So I think for the foreseeable future, it's AI augmenting the physician.

1043
01:03:07,280 --> 01:03:10,920
There is a day where I think it crosses over, but it has to get to a level of sensitivity

1044
01:03:10,920 --> 01:03:18,280
and specificity where it ensures providers feel comfortable that the system isn't going

1045
01:03:18,280 --> 01:03:19,280
to screw you.

1046
01:03:19,280 --> 01:03:20,280
Okay, perfect.

1047
01:03:20,280 --> 01:03:23,080
Amit, thanks so much for taking the time to be here.

1048
01:03:23,080 --> 01:03:24,080
Yeah, absolutely.

1049
01:03:24,080 --> 01:03:26,960
I had an amazing time and we'll have to do it again.

1050
01:03:26,960 --> 01:03:30,920
Yeah, I appreciate the time and I'm excited about what you guys are doing and love to help

1051
01:03:30,920 --> 01:03:35,760
out with the forum and as people sort of make their way through investments, you know, feel

1052
01:03:35,760 --> 01:03:36,760
free to reach out.

1053
01:03:36,760 --> 01:03:37,760
I'm happy to help in any way.

1054
01:03:37,760 --> 01:03:38,760
Thanks.

1055
01:03:38,760 --> 01:03:41,760
Amit.