March 1, 2023

Data Is The New Oil: What’s Next In Banking

Apple Podcasts podcast player iconSpotify podcast player icon
Apple Podcasts podcast player iconSpotify podcast player icon

The release of Chat GPT has spurred all sorts of conversations and speculation on where the technology is taking humanity- and if this is a cause for alarm or celebration. In the banking industry leveraging the new advances in AI can increase lending accessibility and customer experience.


Tim Shangle, Innovation and Data Analytics at ChoiceOne Bank and a member of the American Bankers Association's Emerging Leaders Council, is all about the use of AI to develop breakthrough tactics to make lending more efficient and personal.

Join us as Jeff and Tim discuss:

  • Practical applications of Chat GPT
  • Innovation to scale and innovation to time
  • Data as the new oil
  • Remembering the people behind the data
Want to learn more about how Upstart partners with credit unions? Check out this case study mentioned in the episode.
WEBVTT

1
00:00:02.919 --> 00:00:07.839
You're listening to Leaders and Lending from
Upstart, a podcast dedicated to helping consumer

2
00:00:07.919 --> 00:00:13.039
lenders grow their programs and improve their
product offerings. Each week, here,

3
00:00:13.080 --> 00:00:17.640
decision makers in the finance industry offer
insights into the future of the lending industry,

4
00:00:18.079 --> 00:00:22.839
best practices around digital transformation, and
more. Let's get into the show.

5
00:00:26.679 --> 00:00:30.399
Welcome to Leaders in Lending. I'm
your host, Jeff Keltner. This

6
00:00:30.440 --> 00:00:35.240
week's episode features my conversation with Tim
Schangel. He works on data and analytics

7
00:00:35.320 --> 00:00:40.200
at Choice one Bank and as part
of the American Bankers Associations Merging Leaders Council.

8
00:00:40.320 --> 00:00:43.079
And we really you know, Tim
has been in the tech world for

9
00:00:43.119 --> 00:00:46.640
a while, has a CS degree, and we really dove into the technology

10
00:00:46.640 --> 00:00:50.640
a little deep, talking about chat, gpt AI, some of the new

11
00:00:50.719 --> 00:00:54.399
things and where they might fit into
the banking environment and also where they don't

12
00:00:54.439 --> 00:00:59.359
fit, how you combine AI and
human resources and human touches. I think

13
00:00:59.399 --> 00:01:04.599
importantly Tim's perspective that sometimes we overestimate
what to expect from a technology in the

14
00:01:04.640 --> 00:01:07.640
short term, but underestimate in that
medium term what maybe a ten year time

15
00:01:07.680 --> 00:01:11.599
horizon. While things may not change
dramatically in the next two years or the

16
00:01:11.599 --> 00:01:15.200
next ten the banking environment might be
unrecognizable and so really interesting conversation. And

17
00:01:15.239 --> 00:01:21.079
the last bit we delve into his
background going to seminary studying theology and how

18
00:01:21.120 --> 00:01:23.280
that impacts the way he thinks about
banking. It's a really interesting conversation that

19
00:01:23.359 --> 00:01:26.120
ranges a lot of topics, but
deep on the tech this week. Please

20
00:01:26.200 --> 00:01:30.719
enjoy my conversation with Tim. Tim, Welcome to the podcast. Thanks for

21
00:01:30.760 --> 00:01:34.519
making the time to join us today. Yeah, thank you, thank you

22
00:01:34.560 --> 00:01:38.799
for having me. Excited to talk
here. A little bit of a body

23
00:01:38.840 --> 00:01:42.079
innovation. Just right before this call, I was chatting with chat gpt,

24
00:01:42.319 --> 00:01:46.959
so definitely into it. What did
what did chat GPT tell you? Buddy

25
00:01:46.959 --> 00:01:49.439
of mine called me. It's like
I had a business idea for you financial

26
00:01:49.439 --> 00:01:52.439
services chat GPT go. I was
like, all right, man, I

27
00:01:52.439 --> 00:01:55.719
feel like everybody's trying to get the
chat GPT of X. Yeah. No,

28
00:01:55.840 --> 00:01:59.400
I was just kind of testing it
and I said, I said,

29
00:01:59.519 --> 00:02:04.079
rate letter for me to give to
my employer that I changed bank accounts.

30
00:02:04.480 --> 00:02:06.959
And that's all I said to it
because I wanted to see what kind of

31
00:02:07.000 --> 00:02:09.159
what it did. And it gave
a perfect form letter and said like insert

32
00:02:09.159 --> 00:02:13.680
your name here, insert your new
bank's name here, your account number,

33
00:02:13.719 --> 00:02:15.960
your routing number. And so then
of course I had to follow up with

34
00:02:16.000 --> 00:02:19.919
it, and I asked it,
well, where else should I change my

35
00:02:19.919 --> 00:02:23.080
bank account information? And that rattled
off the next like three or four things

36
00:02:23.120 --> 00:02:29.599
and gave me template letters for all
those, and so it's pretty It was

37
00:02:29.599 --> 00:02:32.439
funny. I saw this Twitter post
or something somewhere somebody said, you know,

38
00:02:32.479 --> 00:02:35.759
chat GPT is great and all,
but I'll never be used in medicine.

39
00:02:35.960 --> 00:02:38.879
And the next thing I saw were
two doctors, and one I said,

40
00:02:38.039 --> 00:02:40.960
well, whenever I get denied insurance
coverage for a treatment, I go

41
00:02:42.000 --> 00:02:45.159
to chat GPT and tell it,
you know, write me a letter to

42
00:02:45.280 --> 00:02:49.400
United Healthcare justifying. You know.
It spits out a perfect letter. And

43
00:02:49.439 --> 00:02:51.159
then I talked to a buddy mine. It's like, oh man, I

44
00:02:51.240 --> 00:02:53.960
use it all the time to write
reference letters, because I wouldn't write these

45
00:02:54.039 --> 00:02:58.240
nice reference letters. But I can
go and say, write a reference letter

46
00:02:58.400 --> 00:03:00.800
from the head whatever position, at
this position, and for a nurse with

47
00:03:00.960 --> 00:03:06.599
this name and these you know,
she's cheerful and on time and effect whatever

48
00:03:06.680 --> 00:03:08.240
professional. Give a couple of adjectives, and that it pops out this perfect

49
00:03:08.280 --> 00:03:10.919
layers, Like what would have been
hours with work is now five minutes of

50
00:03:12.039 --> 00:03:14.639
putting in and it's still small.
Cousin. They edited it a little bit,

51
00:03:14.759 --> 00:03:15.800
but I was like, yeah,
right there, I can't never work

52
00:03:15.800 --> 00:03:19.560
in medicine. I went, well, I mean it's not doing diagnosies.

53
00:03:19.599 --> 00:03:23.080
I great, but like that's doing
a pretty good job exactly, you're actually

54
00:03:23.120 --> 00:03:24.280
changing your bank or was that just
the test game. No, No,

55
00:03:24.360 --> 00:03:30.000
that was just a complete test just
to see how how close they could get

56
00:03:30.000 --> 00:03:31.400
it. So, I mean,
but even like writing emails like okay,

57
00:03:31.439 --> 00:03:35.560
I want to announce this new feature
coming out, I throw it in chat

58
00:03:35.560 --> 00:03:39.400
GPT first and five drafts later,
I have one that's perfect that to get

59
00:03:39.439 --> 00:03:43.639
that initial draft that sometimes the hardest
thing to do. Yeah. I actually

60
00:03:43.680 --> 00:03:46.280
saw somebody who's built a I forget
the name of it, but it's like

61
00:03:46.319 --> 00:03:51.199
a Chrome plug in that uses Chat
JPT. It kind of embedds it in

62
00:03:51.199 --> 00:03:54.560
the Gmail composed experience, so you
can be in Gmail. Well you can

63
00:03:54.560 --> 00:03:58.159
plug like a grammarly plug in will
come in and make you sound that our

64
00:03:58.199 --> 00:04:00.599
our it T T team will let
us use that, and we'll let us

65
00:04:00.680 --> 00:04:02.039
use this. But you could give
it a few prompts in it will pop

66
00:04:02.039 --> 00:04:03.919
a right in the Gmail so you
don't have to have two witness open.

67
00:04:03.919 --> 00:04:08.639
It just kind of like uses it
to kind of throw stuff right right into

68
00:04:08.680 --> 00:04:11.639
the email body for you. It
sounds pretty fun. Oh that's awesome.

69
00:04:12.120 --> 00:04:14.639
Yeah, I mean, I gotta
find that one. So when we talked

70
00:04:14.680 --> 00:04:16.240
to earlier time, you use this
really interesting phase that I wanted you to

71
00:04:16.279 --> 00:04:19.480
explain it, which was innovation to
scale and to time, And I thought

72
00:04:19.519 --> 00:04:25.000
that was an interesting description of innovation
or ways you think about innovation and access

73
00:04:25.000 --> 00:04:27.399
along way. So talk to a
little bit about what you mean by innovation

74
00:04:27.439 --> 00:04:30.839
to scale and innovation to time.
Yeah, So innovation to scale, like

75
00:04:30.920 --> 00:04:36.000
especially when we think about like small
community banks like Choice one. I mean

76
00:04:36.000 --> 00:04:39.639
we have thirty four locations. There's
a lot of banks smaller than us,

77
00:04:39.639 --> 00:04:42.639
and there's a whole lot better,
a whole lot bigger than us. One

78
00:04:42.680 --> 00:04:47.720
big thing is they with all of
like cat GPT and artificial intelligence and all

79
00:04:47.759 --> 00:04:51.040
the information we have. One of
the big kind of pushes as well,

80
00:04:51.079 --> 00:04:56.480
you have to have a customized experience
for all of your customers, and there's

81
00:04:56.480 --> 00:04:59.519
a lot of different ways to do
that. But sometimes they call it like

82
00:04:59.639 --> 00:05:03.480
niche banking. But when we think
about like our geographic area with niche banking,

83
00:05:03.560 --> 00:05:08.279
and like, let's say we wanted
to do something for dentists, Well,

84
00:05:08.279 --> 00:05:13.079
we only have a limited geography and
there's only so many dentists that can

85
00:05:14.079 --> 00:05:16.720
bank in that geographical area. So
then you have to kind of weigh it,

86
00:05:16.839 --> 00:05:20.000
and it's like, Okay, do
you really want to create a special

87
00:05:20.040 --> 00:05:28.240
product and special user interface for a
niche market when the maximum number of people

88
00:05:28.240 --> 00:05:31.480
you can get is one hundred people? And I'm just throwing those numbers out

89
00:05:31.519 --> 00:05:35.959
that probably isn't correct for dentists in
our area. But it's still one of

90
00:05:35.959 --> 00:05:41.000
those where it's like, Okay,
with scale, it's like if you're launching

91
00:05:41.000 --> 00:05:45.279
stuffing nationally, it's very different than
if you're a community bank that's trying to

92
00:05:45.319 --> 00:05:48.959
reach a geographic area. Interesting,
that's the innovation to scale. How about

93
00:05:49.120 --> 00:05:53.680
time? How do you think about
that question? Yeah? And then the

94
00:05:54.000 --> 00:06:00.319
time it's interesting because you innovation happens
in the next five years, there's not

95
00:06:00.519 --> 00:06:03.600
going to be really that much that
changes. I mean, there'll be new

96
00:06:03.639 --> 00:06:08.040
products on the market, you know, there'll be early adopters that are doing

97
00:06:08.040 --> 00:06:12.160
really exciting things, and some of
them will fail. There'll be other people

98
00:06:12.160 --> 00:06:15.199
that have literally not changed at all
in the next five years and it will

99
00:06:15.279 --> 00:06:19.839
still be acceptable. But then if
we go ten years out and we look

100
00:06:19.879 --> 00:06:23.639
backwards, it'll be like, oh, my gosh, everything has changed in

101
00:06:23.639 --> 00:06:27.800
the last ten years. And I
think about my own career in banking.

102
00:06:27.839 --> 00:06:30.959
I started about ten years ago,
and you very much saw that. I

103
00:06:30.959 --> 00:06:33.560
mean, when I started ten years
ago, there were still the number of

104
00:06:33.560 --> 00:06:40.079
banks that didn't have mobile banking yet, it didn't have mobile deposit. Mobile

105
00:06:40.079 --> 00:06:46.399
deposit was still questionable whether or not
the everything was laid out in the best

106
00:06:46.399 --> 00:06:51.199
way possible to prevent fraud. There
was no type of mobile wallet. And

107
00:06:51.240 --> 00:06:56.160
that's only ten years ago. And
I think with innovation, we have to

108
00:06:56.240 --> 00:06:58.680
kind of keep in mind that there's
gonna be a whole lot that changes.

109
00:06:59.439 --> 00:07:01.279
We just don't feel it changing.
So you have to kind of look at

110
00:07:01.279 --> 00:07:04.480
that ten years out and then remember
that, like five years from now,

111
00:07:04.519 --> 00:07:09.079
it's okay, You're gonna have plenty
of time to help on. Yeah,

112
00:07:09.160 --> 00:07:13.040
it's interesting. I think it's a
quote by Roy Amara who said something along

113
00:07:13.079 --> 00:07:15.639
lines if we tend to overestimate the
impact of technology and the short term and

114
00:07:15.759 --> 00:07:18.319
underestimate it in the long term.
I think it's kind of about the five

115
00:07:18.360 --> 00:07:21.600
the ten years we see chatting team. We expect next week, I'm gonna

116
00:07:21.600 --> 00:07:25.680
be chatting on my banking website with
an AI thing and it's gonna be great,

117
00:07:25.680 --> 00:07:27.079
and then it doesn't. Kind of
go, oh, this thing never

118
00:07:27.079 --> 00:07:29.639
worked out. It's like, well, in ten years, you may not

119
00:07:29.680 --> 00:07:31.360
recognize anything about the experience, but
in you know, in two weeks and

120
00:07:31.560 --> 00:07:34.920
in a year that those things take
a little time to percolate, so we

121
00:07:34.959 --> 00:07:39.839
tend to we tend to not notice
it or expect it to be instantaneous,

122
00:07:39.879 --> 00:07:42.600
and it's actually faster than you might
have thought in the long term, but

123
00:07:42.639 --> 00:07:44.800
in the short term it just never
quite lives up to expectations. I think

124
00:07:44.800 --> 00:07:47.160
there's a lot of that going around
exactly, and quite often there's this kind

125
00:07:47.160 --> 00:07:50.199
of fear that oh, I'm going
to miss out, Like this is moving

126
00:07:50.199 --> 00:07:54.639
so fast, I'm gonna miss out
on it. But just the same way

127
00:07:54.680 --> 00:07:59.639
that customers are so slow to move, there will be some customers that are

128
00:07:59.639 --> 00:08:03.519
earlier ops. But overall, it's
a lot of work to change a bank

129
00:08:03.519 --> 00:08:07.519
account. It's a lot of work
just to convince them to change their morning

130
00:08:07.600 --> 00:08:11.800
routine and to hop on a peloton
instead of happening on a treadmill. And

131
00:08:11.959 --> 00:08:16.319
eventually they do, but they have
to kind of ramp up and scale just

132
00:08:16.360 --> 00:08:20.839
the same way that we do.
Yeah. The other thing that occurs to

133
00:08:20.879 --> 00:08:22.120
me is the kind of the Gartner
version of this, the hype cycle.

134
00:08:22.199 --> 00:08:24.560
Yeah, I'm sure you're familiar with
the hypes you can kind of. I

135
00:08:24.639 --> 00:08:28.000
think it's it's related to this kind
of you know, short term overestimation long

136
00:08:28.079 --> 00:08:33.440
term because the short term overestimation is
that hype peak whatever, they call it,

137
00:08:33.519 --> 00:08:37.200
the peak of whatever, and then
they've got the trough of disillusionment,

138
00:08:37.879 --> 00:08:41.840
which is where the real work is
being done to build the long term lasting

139
00:08:41.840 --> 00:08:45.399
that kind of longer term vision.
One of the things you how do you

140
00:08:45.440 --> 00:08:48.519
think about sustaining because there's this kind
of like the AH as a flash in

141
00:08:48.559 --> 00:08:50.919
a pan and it's not going to
be it just didn't do what I wanted

142
00:08:50.919 --> 00:08:54.559
in six months, and yet it's
really the people investing in building it in

143
00:08:54.600 --> 00:08:56.840
the right ways over the next six
seven years that are going to reap the

144
00:08:56.919 --> 00:09:01.679
rewards on that timescale. How do
you think about the like what you do

145
00:09:01.759 --> 00:09:05.639
to optimize your ability to really see
those things and then continue in, you

146
00:09:05.679 --> 00:09:09.559
know, iterating on and investing in
the important areas of technology over time so

147
00:09:09.679 --> 00:09:11.639
you can you can actually be the
ones developing the new thing over the ten

148
00:09:11.720 --> 00:09:16.679
year horizon that maybe doesn't come into
Yeah. I think the big part with

149
00:09:16.840 --> 00:09:20.919
that is to kind of keep your
eyes focused on that ten year or to

150
00:09:20.320 --> 00:09:24.519
kind of know what's happening ten years
from now and kind of know, Okay,

151
00:09:24.519 --> 00:09:28.720
where the trends going, what should
we have our eyes on. Like

152
00:09:28.799 --> 00:09:33.600
I think of ten years ago,
like sending you around a table and talking

153
00:09:33.600 --> 00:09:37.679
about mobile design, like anything that
we're going to invest in, we're going

154
00:09:37.720 --> 00:09:39.519
to make sure it has mobile design, is going to be compatible with phones.

155
00:09:41.080 --> 00:09:45.080
And ten years ago, like we
had phones, but there was a

156
00:09:45.159 --> 00:09:50.240
question of should should a website even
be mobile responsive or should we just go

157
00:09:50.360 --> 00:09:52.840
straight to the app? And we
kind of made an executive decision like no,

158
00:09:52.919 --> 00:09:56.120
everything going forward is going to be
mobile responsive. You're going to have

159
00:09:56.159 --> 00:10:01.080
to get zone on the phone.
And I think making those type of decisions

160
00:10:01.919 --> 00:10:05.240
so they they're kind of the smaller
the decisions that have a big impact on

161
00:10:05.320 --> 00:10:11.639
the kind of overall direction then along
with you know, looking to see when

162
00:10:11.679 --> 00:10:16.840
those kind of smaller changes kind of
reach fruition. You know. I think

163
00:10:16.840 --> 00:10:22.919
of like the Microsoft Power Apps platform
with power BI and a lot of the

164
00:10:22.080 --> 00:10:26.480
kind of no code automated, you
know, type of things that have been

165
00:10:26.480 --> 00:10:30.360
promised for the last twenty years.
I think has finally reached a point of

166
00:10:30.360 --> 00:10:35.279
maturity where you can start using them
in production. And I think that's the

167
00:10:35.360 --> 00:10:39.519
kind of an area of excitement.
You said two things to me there that

168
00:10:39.799 --> 00:10:43.360
I want to dig into. One
is kind of like what these tenure trends

169
00:10:43.360 --> 00:10:46.240
are You talk about power tools being
on the lipic if I want to take

170
00:10:46.240 --> 00:10:48.840
a step back. You talked about
these small decisions that end up having a

171
00:10:48.879 --> 00:10:52.720
big impact, and that was really
interesting to me. Jeff bezos Wants described

172
00:10:52.759 --> 00:10:56.799
them as like thinking about you know, one way doors versus two way doors,

173
00:10:56.879 --> 00:10:58.799
right, and like move quickly through
the two way door and slowly through

174
00:10:58.840 --> 00:11:03.960
a decision that can't be undone or
reverse it. And I think in tech

175
00:11:03.000 --> 00:11:07.519
often it's things like data architecture where
you can build code quickly and replace it.

176
00:11:07.519 --> 00:11:09.200
But if you architect the data wrong, well God help you. But

177
00:11:09.360 --> 00:11:13.960
replacing that fixing that adding the feature
that wasn't that the data infrastructure didn't support

178
00:11:15.159 --> 00:11:18.200
becomes nigh impossible two years down the
road because stuff gets built on top.

179
00:11:18.679 --> 00:11:22.480
And I guess my question is,
how do you think about what the small

180
00:11:22.519 --> 00:11:26.600
decisions that have a big impact are
though you want to be careful about or

181
00:11:26.639 --> 00:11:30.320
get right or thoughtful about, versus
the small decisions that are like easy to

182
00:11:30.399 --> 00:11:33.600
change and not a big deal and
you should just make them with little information

183
00:11:33.600 --> 00:11:35.759
and move quickly as opposed to trying
to get them right. Because it seems

184
00:11:35.759 --> 00:11:41.000
really important to understand the difference between
the big, the big decisions that make

185
00:11:41.080 --> 00:11:43.399
you know, maybe small in the
moment but make a big difference on the

186
00:11:43.440 --> 00:11:45.399
road, and the small on the
moment that are like easy to fix and

187
00:11:45.440 --> 00:11:46.919
don't worry about this, just go
and try a bunch of stuff. They

188
00:11:46.960 --> 00:11:50.519
feel like very different. How do
you how do you think about discerning what

189
00:11:50.559 --> 00:11:54.320
you're looking at? Yeah, so
I mean like big decisions. I mean

190
00:11:54.320 --> 00:11:58.480
you're always going to look at like
okay, cost availability and struggle with the

191
00:11:58.480 --> 00:12:03.960
contract to whole bunch and get tied
up with lawyers, and so I mean

192
00:12:03.000 --> 00:12:05.960
like, like big decisions are always
going to be big decisions. But then

193
00:12:07.000 --> 00:12:11.480
I think it's there's other kind of
small, varying factors they can kind of

194
00:12:11.519 --> 00:12:15.919
plug into there. So whether it's
like, Okay, does this new application

195
00:12:16.000 --> 00:12:18.159
have an API? Like we have
no plans to use an API, but

196
00:12:18.360 --> 00:12:22.519
can I get my data out of
it? And it's those type of decisions

197
00:12:22.519 --> 00:12:26.799
where you have A, B and
C products that you're evaluating, and if

198
00:12:26.799 --> 00:12:30.840
all of them are kind of equal, but this one has a really good

199
00:12:30.840 --> 00:12:33.440
API that you have no intention of
using, but maybe someday you will,

200
00:12:35.039 --> 00:12:39.320
it's an easier decision to go with
that product. And I think it's like

201
00:12:39.039 --> 00:12:43.879
artificial intelligence will be another one where
it's like, Okay, does this new

202
00:12:43.919 --> 00:12:48.559
product can it integrate with have an
API that will allow it to integrate with

203
00:12:48.639 --> 00:12:54.360
chat GPT or something similar to it. We haven't kind of went down that

204
00:12:54.440 --> 00:12:58.519
road yet, but I think very
quickly will get there if it, especially

205
00:12:58.519 --> 00:13:01.240
if chat GPT continues the way it
is going. Yeah, well, yeah,

206
00:13:01.399 --> 00:13:03.600
well we'll talk about chat TB too
many. I think your answer is

207
00:13:03.600 --> 00:13:09.919
so interesting because it highlights this like
import of the ability to do things I

208
00:13:09.919 --> 00:13:13.039
don't yet need or want to do, but kind of the optionality of having

209
00:13:13.120 --> 00:13:16.480
that capability if in what I would
want to because so often like it's the

210
00:13:16.519 --> 00:13:20.240
thing you didn't need two years ago, that you didn't optimize for and then

211
00:13:20.279 --> 00:13:24.360
and two years ago, and oh
the lack of that thing is killing my

212
00:13:24.399 --> 00:13:26.320
ability to do the next thing I
want to do. And what if I

213
00:13:26.440 --> 00:13:30.720
put myself in this corner and as
fascinating to think about as things you're not,

214
00:13:31.000 --> 00:13:35.159
I'm making a decision based on the
future. I don't even use you

215
00:13:35.200 --> 00:13:39.480
do that. You're kind of future
proofing if you will your your thought process,

216
00:13:39.480 --> 00:13:43.039
which I think is a really interesting
mental construct. Right well, and

217
00:13:43.759 --> 00:13:46.840
like people do it all the time. I mean, I mean they go

218
00:13:46.919 --> 00:13:50.679
to buy a car, and till
newlywed's going to buy a car, Well,

219
00:13:50.120 --> 00:13:52.799
maybe I shouldn't buy the sports car
because we just got married and we're

220
00:13:52.799 --> 00:13:56.440
gonna have kids soon. I don't
know when we're gonna have kids. Maybe

221
00:13:56.440 --> 00:13:58.000
we'll have kids, maybe we won't
have kids. But let's go with a

222
00:13:58.080 --> 00:14:01.960
Sudan this car I got sports car, and then upgrade. That's what I

223
00:14:03.000 --> 00:14:03.759
would be my advice. You get
to take it, take it while you

224
00:14:03.759 --> 00:14:05.679
can have. You is a long
time do you go back the other way?

225
00:14:05.720 --> 00:14:09.360
So just trader in when it's time
and enjoy it while you can.

226
00:14:09.279 --> 00:14:11.759
It's my regret. I think I
went with a Preuss instead of the sports

227
00:14:11.759 --> 00:14:16.240
car. That was definitely not the
same kind of enjoyment for the first few

228
00:14:16.279 --> 00:14:20.840
years before we had the kids.
Small stakes, they are small mistakes.

229
00:14:20.879 --> 00:14:22.759
What it is? What do you
see as that you know you talked,

230
00:14:22.759 --> 00:14:26.759
We talked about chat, GPT,
talked about kind of the Microsoft power Tools

231
00:14:26.799 --> 00:14:28.919
power BI What are the big if
you think the big waves are the big

232
00:14:28.960 --> 00:14:31.080
trends, the thing you say,
hey, we've got to be this is

233
00:14:31.120 --> 00:14:33.639
going to be a thing over this
time horizon. The question is how do

234
00:14:33.679 --> 00:14:35.679
we take advantage of it, What
do we do about it, how do

235
00:14:35.759 --> 00:14:39.440
we work with it? What are
the ones you've got your eye on?

236
00:14:39.519 --> 00:14:41.480
What are the things you're saying,
Hey, this is the stuff that's really

237
00:14:41.519 --> 00:14:46.360
interesting. Obviously you've got AI in
particularly generative AI and in large language models

238
00:14:46.399 --> 00:14:48.440
with chat, GPT and GPT three
on the list. Anything else that goes

239
00:14:48.440 --> 00:14:52.919
into that kind of bucket. Yeah, I would say all the kind of

240
00:14:52.960 --> 00:14:58.360
no code low code solutions. So
whether it's power apps or like Jecutary's workflow

241
00:14:58.480 --> 00:15:01.879
or I mean, there's so many
on the market right now. That allows

242
00:15:01.879 --> 00:15:07.440
someone to kind of somewhat quickly develop
kind of their own custom program with with

243
00:15:07.639 --> 00:15:13.080
fairly little programming experience. I mean, it still takes a lot of work,

244
00:15:13.120 --> 00:15:15.519
and it will get better in the
future, you know. But I

245
00:15:15.519 --> 00:15:18.440
think that's one area that you'll see
a lot of kind of investment in.

246
00:15:18.639 --> 00:15:24.519
And finally, the automation of the
manual Excel spreadsheets that a lot of people

247
00:15:24.559 --> 00:15:28.200
are using. Yeah. I once
heard somebody describe, as you know,

248
00:15:28.279 --> 00:15:31.360
in the in the sort of committee, like why do I get a great

249
00:15:31.360 --> 00:15:33.279
business idea? They go, if
you want to be to b SaaS business

250
00:15:33.320 --> 00:15:37.879
idea, just find anything that's done
in a spreadsheet and then build an app

251
00:15:37.919 --> 00:15:41.200
that does it better. Yea.
And that's like going to build, build,

252
00:15:41.200 --> 00:15:43.039
build a billion dollar business on that. And there are a lot of

253
00:15:43.039 --> 00:15:46.399
businesses that are essentially really complicated Excel
sheets turned into an app. But the

254
00:15:46.480 --> 00:15:48.879
no code loco and I've not heard
that before. That's I mean, I'm

255
00:15:48.879 --> 00:15:52.320
familiar with it. Are you seeing
a lot of development and interest in that

256
00:15:52.360 --> 00:15:54.360
space? It's a really I think
it's a fascinating space to open up the

257
00:15:54.480 --> 00:15:58.159
creative power. If you think back
to the origins of the Internet and hackers,

258
00:15:58.159 --> 00:16:00.960
they were really builders who wanted to
build stuff, and it was like

259
00:16:02.200 --> 00:16:04.559
it was behind this technical wall of
stuff you had to learn to be able

260
00:16:04.559 --> 00:16:07.240
to build in that environment. And
the no code is kind of going,

261
00:16:07.240 --> 00:16:11.960
Hey, if you're a builder but
don't don't know the sophisticated languages and the

262
00:16:11.080 --> 00:16:15.000
kind of technical background, you can
still build interesting stuff, which I think

263
00:16:15.080 --> 00:16:19.200
is a really fascinating unlocking of the
power of compute for a user who never

264
00:16:19.279 --> 00:16:22.919
learned the secret language of you know, c Sharp or Swift or Java or

265
00:16:22.919 --> 00:16:26.679
whatever you pick your language of choice. Python. What are you seeing done

266
00:16:26.720 --> 00:16:30.080
with that that's interesting to you?
Yeah? Really, I see the exact

267
00:16:30.159 --> 00:16:33.559
same things that I was doing twenty
five years ago. And micros opt to

268
00:16:33.559 --> 00:16:37.480
access Yeah, I mean, just
to be completely frank and honest, but

269
00:16:37.639 --> 00:16:41.879
it's one of those where micros opt
to Access was. I mean, there

270
00:16:41.919 --> 00:16:45.200
are ways to roll it out across
the network, but really power apps is

271
00:16:45.279 --> 00:16:48.240
kind of the embodiment of that of
like, Okay, you have all this

272
00:16:48.399 --> 00:16:52.120
data, We're going to put a
pretty interface on top of it and allow

273
00:16:52.159 --> 00:16:53.440
you to work with it. Then
of course, I mean, you can

274
00:16:53.440 --> 00:16:56.480
plug in with different APIs and you
know, they made it a whole lot

275
00:16:56.519 --> 00:17:00.279
better, But at the end of
the day, it's really take those kind

276
00:17:00.279 --> 00:17:04.279
of like database oriented data, you
know, crunching type of tools and making

277
00:17:04.519 --> 00:17:08.720
a beautiful way to view it and
process that and work with that data.

278
00:17:11.279 --> 00:17:12.640
Yeah, you've said one of the
I think one of the through lines of

279
00:17:14.240 --> 00:17:19.200
my conversations has been the increasing import
of data. And you know, I've

280
00:17:19.200 --> 00:17:22.079
had this ongoing debate like if data
is the new oil, and I actually

281
00:17:22.160 --> 00:17:26.559
I'm curious to take I'll give you
my perspective, and I'm kind of curious

282
00:17:26.559 --> 00:17:27.880
your reaction, which is like,
I hear this data is the new oil,

283
00:17:27.880 --> 00:17:32.359
and I kind of think it's a
totally wrong analogy because oil is something

284
00:17:32.400 --> 00:17:34.720
that's like rare, only a couple
of people have it, and it's mostly

285
00:17:34.759 --> 00:17:37.039
just having it that's valuable, and
I kind of feel like data is the

286
00:17:37.039 --> 00:17:41.119
opposite. Data is like everywhere,
everybody's got tons of it, and it's

287
00:17:41.160 --> 00:17:45.200
not the having of it that's important. It's the being able to make use

288
00:17:45.240 --> 00:17:48.599
of it in some important way that
is the key differentiator. So I kind

289
00:17:48.599 --> 00:17:52.680
of think it's like well and someway, it's kind of more like sand it's

290
00:17:52.720 --> 00:17:55.319
everywhere. But it's like the turning
it into silicon and chips. It's like,

291
00:17:55.359 --> 00:17:57.200
that's the part that's really magical.
It's not the like I've got the

292
00:17:57.240 --> 00:18:00.599
sand, like everyone's got sand.
It's not a big deal oils. More

293
00:18:00.599 --> 00:18:03.000
like you got it. It's good
and like wow, So I know what

294
00:18:03.079 --> 00:18:06.200
you're taking is but I feel like
there's this unlocking the power of data to

295
00:18:06.319 --> 00:18:08.319
do things in a better way that
is a part of whatever the tools are,

296
00:18:08.359 --> 00:18:11.079
there's a part of the way of
the future that is really important for

297
00:18:11.079 --> 00:18:15.240
people to be investing in and getting
right. Yeah, I really like that

298
00:18:15.240 --> 00:18:21.400
analogy. Yeah that you gave it. I think I definitely agree with it

299
00:18:21.440 --> 00:18:27.559
and resonate with it. The whole
idea of data being the new oil,

300
00:18:27.799 --> 00:18:32.359
I definitely don't think I agree with
it. I think it does take a

301
00:18:32.400 --> 00:18:36.480
lot of it's really the processing of
it, and that's really I mean,

302
00:18:36.519 --> 00:18:41.200
what you know chat GPT is doing
is it's processing these huge amounts of data.

303
00:18:41.400 --> 00:18:44.000
I mean I will read different articles
and stuff and they talk about like,

304
00:18:44.079 --> 00:18:47.079
oh, yeah, you need a
mine your data to get the best

305
00:18:47.119 --> 00:18:52.720
financial decision for your customer. It's
like, well, probably at least a

306
00:18:52.799 --> 00:18:56.079
third of my customers the best financial
decision they can do is to balance their

307
00:18:56.119 --> 00:19:00.920
checkbooks. It's like, I don't
really need a big day to tell me

308
00:19:00.960 --> 00:19:04.079
that. And I think of Adam
Greenland, who's our CFO and spend a

309
00:19:04.079 --> 00:19:10.079
bunch of time processing all these reports
and identifying different customers. I gave him

310
00:19:10.119 --> 00:19:12.640
like a top ten list and because
I bet it's these ten, and sure

311
00:19:12.720 --> 00:19:15.680
enough, that's exactly who it was. Because he was so familiar with the

312
00:19:15.759 --> 00:19:21.160
data. I think sometimes we lose
sight of just the human that's able to

313
00:19:21.240 --> 00:19:25.839
process the data on their own,
and then we need to figure out ways

314
00:19:25.880 --> 00:19:30.079
to process the data in a way
that the human can still be involved in

315
00:19:30.079 --> 00:19:33.160
it. At the end of the
day, it's a very human centric problem

316
00:19:33.240 --> 00:19:37.519
as opposed to a problem we're just
going to push over to the machine and

317
00:19:37.640 --> 00:19:44.759
have them solve it. Hey,
Sarah buch Here, VP of Consumer Lending

318
00:19:44.759 --> 00:19:48.680
at First Federal Bank of Kansas City. A few years back, our executive

319
00:19:48.680 --> 00:19:52.119
team knew we needed to grow our
unsecured personal loan portfolio, so we turned

320
00:19:52.119 --> 00:19:56.640
to Upstart. Three years later,
I am proud to share that we've scaled

321
00:19:56.640 --> 00:20:00.720
our loan volume target from five hundred
thousand a month to twelve million a month

322
00:20:00.839 --> 00:20:03.480
and acquired over a thousand new customers. If you want to hear more about

323
00:20:03.519 --> 00:20:07.960
our partnership, check out the full
case study on upstart dot Com forward slash

324
00:20:08.079 --> 00:20:15.039
Lenders. Once again, that's Upstart
dot Com Forward slash Lenders. All right,

325
00:20:15.559 --> 00:20:18.519
let's get back to the show.
We've talked a lot about chat GPT.

326
00:20:18.640 --> 00:20:21.759
Are you are you guys deploying any
kind of generative AI be it you

327
00:20:21.799 --> 00:20:25.640
have any specific use cases for where
you see that's kind of the large language

328
00:20:25.680 --> 00:20:29.359
model version that's text based, and
of course the other big generative AI space

329
00:20:29.440 --> 00:20:33.079
right now is its image base.
Are you guys looking at any use of

330
00:20:33.200 --> 00:20:37.240
utilization of any of those tools in
any specific ways or more just kind of

331
00:20:37.279 --> 00:20:41.559
tinkering and playing at this point,
not in a very formal way. A

332
00:20:41.559 --> 00:20:45.400
lot of the reasons, you know, we don't want to put a bunch

333
00:20:45.440 --> 00:20:49.000
of data in there and then have
it be I don't know if there's quite

334
00:20:49.000 --> 00:20:55.119
a business structure around it. That's
all. We'll plus all the audit inspections,

335
00:20:55.160 --> 00:20:57.839
but we're definitely using it for the
things that we would google, you

336
00:20:57.920 --> 00:21:03.319
know, so we would be like
you might Google, like a good email,

337
00:21:03.359 --> 00:21:07.359
descend to so and so for X, Y and Z. Instead put

338
00:21:07.359 --> 00:21:11.759
that in a chat GPT and it
gives you the full letter without any advertising

339
00:21:11.920 --> 00:21:15.920
or pop ups or fear that you're
going to click on the wrong link.

340
00:21:15.440 --> 00:21:18.759
How do you think about what the
line of that is? Because the thing

341
00:21:18.759 --> 00:21:21.599
I worry about for chat GPT,
and I've told my kids is because now

342
00:21:21.960 --> 00:21:23.039
I don't think my kids have done
it, but I don't know, but

343
00:21:23.119 --> 00:21:27.160
I know their friends have where there's
like they're they're putting homework prompts into chat

344
00:21:27.200 --> 00:21:30.240
CHPT and getting out and they're like, oh, we ddit it before,

345
00:21:30.119 --> 00:21:33.440
Yeah, but didn't right most.
But the interesting thing is it's it's like

346
00:21:33.480 --> 00:21:37.720
a very good writer, yeah,
but it's not always a good fact checker,

347
00:21:37.759 --> 00:21:41.839
which is to say, sometimes it
writes stuff confidently and well, but

348
00:21:41.960 --> 00:21:45.920
it's wrong and you're right. That's
something like write me a letter to a

349
00:21:45.920 --> 00:21:51.039
customer who X that that tells them
about why you can change the factual and

350
00:21:51.039 --> 00:21:53.319
accuracies if you ask it, like, you know, at a six year

351
00:21:53.400 --> 00:21:56.680
at a six percent annual rate of
return, how long does it take to

352
00:21:56.720 --> 00:22:00.640
double my money? I don't know
that I trust it to like get that

353
00:22:00.720 --> 00:22:03.839
answer. I got factual stuff is
not there, So I'm curious how you

354
00:22:03.839 --> 00:22:06.640
think about I think that's one of
the reason whole it's never gonna touch finance

355
00:22:06.680 --> 00:22:08.200
because it's wrong about stuff. But
that's very different than like, write me

356
00:22:08.200 --> 00:22:11.799
a letter to a customer in a
you know, in a disaster area that

357
00:22:11.839 --> 00:22:15.119
needs forbearance on their loan to you
know, and that you can imagine that

358
00:22:15.400 --> 00:22:18.799
actually writing a really good thing.
Can you do you have any heuristics for

359
00:22:18.880 --> 00:22:22.519
how you think about the kinds of
problems you would think this is a good

360
00:22:22.519 --> 00:22:25.119
fit to help you solve and those
that you go, h, I'm not.

361
00:22:25.160 --> 00:22:27.200
I don't want to trust a language
model to do the math for me

362
00:22:27.240 --> 00:22:30.240
because it doesn't do very good maths
sometimes. Yeah. Well, I mean

363
00:22:30.920 --> 00:22:34.160
the bank already has, you know, things that can calculate all of those

364
00:22:34.880 --> 00:22:38.039
items. Yeah. Yeah, So, I mean we really don't need another

365
00:22:38.079 --> 00:22:42.720
tool to calculate the financial information for
us. We already have tools that can

366
00:22:42.720 --> 00:22:48.359
do that. Where we struggle is
is is taking those those answers, those

367
00:22:48.400 --> 00:22:53.119
financial and actually get them in a
way that our customers are willing to view

368
00:22:53.160 --> 00:22:56.559
them and see them and listen to
us. Yeah. So I think that's

369
00:22:56.599 --> 00:23:02.119
where chat GPT comes in is here's
my bullet points, here's the financial data.

370
00:23:02.440 --> 00:23:06.440
Write a good message you know that
appeals to a gen Z audience.

371
00:23:07.640 --> 00:23:12.160
Facebook post, Yeah, yeah,
using the prompt to change like we talk

372
00:23:12.240 --> 00:23:15.839
about like the personalization of the message
and the content and emojis in it to

373
00:23:17.000 --> 00:23:21.000
can you know what emoticate? I
don't speak that way, but you know

374
00:23:21.039 --> 00:23:22.880
the young kids do, and so
if you want to reach them on a

375
00:23:22.920 --> 00:23:26.240
level that's really interesting. Can I
tell you my favorite chat GPT instance here

376
00:23:26.240 --> 00:23:30.079
that the funny story about math?
And so there was an instance of chat

377
00:23:30.160 --> 00:23:33.279
GPT and somebody said what is five
plus two and said five plus two a

378
00:23:33.359 --> 00:23:37.359
seven? Says my wife says five
plus two is eight and says your wife

379
00:23:37.400 --> 00:23:40.519
is incorrect five plus two a seven. And then the person says, my

380
00:23:40.599 --> 00:23:44.720
wife is never wrong, and chat
with your wife is correct plus to his

381
00:23:44.799 --> 00:23:48.079
eight. A very smart tool.
It's smarter than I gave it credit for.

382
00:23:48.279 --> 00:23:52.880
But you know, um, you
got to be careful when my wife

383
00:23:52.880 --> 00:23:56.200
actually has taken a couple of college
classes and she has to write papers and

384
00:23:56.559 --> 00:24:00.599
she's using chat GPT in a number
of different ways. So show right out

385
00:24:00.640 --> 00:24:03.640
of prompt and it's way too long, and have chat JPC helper summarize it

386
00:24:03.720 --> 00:24:08.480
and other times helper kind of stretch
out bullet points into a well written essays.

387
00:24:08.519 --> 00:24:12.279
So well, yeah, I think
that's the you know, people think

388
00:24:12.680 --> 00:24:17.559
chat GPT will replace lots of jobs. It will replace some these kinds of

389
00:24:17.960 --> 00:24:21.559
chat particular but large language models.
But I do think the interesting thing is

390
00:24:21.599 --> 00:24:23.039
like how do you get good at
using it? As people said the same

391
00:24:23.039 --> 00:24:26.960
thing about computers and every technology that's
long and calculators, and they did replace

392
00:24:27.000 --> 00:24:30.960
certain jobs, but it was a
people who were effective at using the tool,

393
00:24:30.960 --> 00:24:33.200
who could use Excel well, that
were able to take the most And

394
00:24:33.240 --> 00:24:37.200
so the skill of like how do
I get good at writing a prompt or

395
00:24:37.240 --> 00:24:40.599
figuring out? Hey, chat GPT
can help me generate a bunch of ideas,

396
00:24:40.599 --> 00:24:42.119
but then I've got to take them
and do X or I want to

397
00:24:42.200 --> 00:24:45.920
use it to expand on this point
that I'm kind of stuck on my writing

398
00:24:45.119 --> 00:24:48.680
or structure a thing. But it's
it's not replacing me, it's letting me.

399
00:24:49.039 --> 00:24:52.000
You know, what's the old job
quote A bicycle for the mind.

400
00:24:52.039 --> 00:24:55.279
It's it's helping me accelerate my productivity, and I've gotten good at using it

401
00:24:55.319 --> 00:24:57.000
as a tool. I think that's
going to be an amazing skill to develop

402
00:24:57.000 --> 00:25:00.200
over an imp. And I don't
think it's going to happen overnight. I

403
00:25:00.200 --> 00:25:02.960
mean, like we said before,
I mean this technology takes a long time.

404
00:25:03.079 --> 00:25:06.880
So even if it was ten times
as good as what it is now,

405
00:25:07.799 --> 00:25:11.400
writers are still going to write for
probably at least for our generation,

406
00:25:11.440 --> 00:25:15.480
if not many more to come.
You know, it's it's not gonna be

407
00:25:15.519 --> 00:25:18.759
a like, Okay, we're gonna
fire everyone and replace you a chat GPT.

408
00:25:18.960 --> 00:25:22.200
It's going to be more kind of
bi attrition and you're gonna increase your

409
00:25:22.200 --> 00:25:26.240
workflow. You're gonna be able to
crank out twice as many papers as what

410
00:25:26.240 --> 00:25:27.839
you could before. I did hear
it right, it's a pretty good Seinfeld

411
00:25:27.839 --> 00:25:30.240
episode if you ask for that.
That's like it's a lot of content.

412
00:25:32.440 --> 00:25:37.119
Yeah, George walks into a restaurant. I could figure out what to do.

413
00:25:37.640 --> 00:25:41.839
I wanted to end. We normally
I start this podcast by asking people

414
00:25:41.880 --> 00:25:45.119
about their path into banking, and
we kind of jumped right into we're too

415
00:25:45.160 --> 00:25:48.920
excited about chat GPT whatever. We
would jump right in. But you've had

416
00:25:48.960 --> 00:25:53.960
a particularly interesting path I think into
the into the banking space in one that

417
00:25:55.000 --> 00:25:57.240
I think really highlights the human element
that you're you're talking about with these technologies.

418
00:25:57.240 --> 00:26:00.240
Tell me a little bit about about
how you got your excited. It's

419
00:26:00.480 --> 00:26:06.400
it's a fascinating story. Yeah.
So I I have a undergraduate from a

420
00:26:06.519 --> 00:26:11.039
Quintas College in Grand Rapids in computer
science. You know, so I have

421
00:26:11.119 --> 00:26:15.839
a tech background and worked in tech
for many many years. And then from

422
00:26:15.880 --> 00:26:18.839
there I went on I got my
Master of Arts and Theological Study from Asbury

423
00:26:18.920 --> 00:26:23.240
Theological Seminary. So all I was
down there is working on all their websites

424
00:26:23.240 --> 00:26:29.240
and varying into the tech space,
but reading in depth philosophy, fee and

425
00:26:29.400 --> 00:26:36.680
theology and really enjoyed it. And
then kind of came back home to Michigan

426
00:26:36.720 --> 00:26:41.480
and ended up at Choice One.
So Kelly Pots who I is from the

427
00:26:41.559 --> 00:26:45.720
same small town and grew up with
my parents and I kind of new Choice

428
00:26:45.720 --> 00:26:51.519
one is the small hometown bank in
the small town of Sparta, and end

429
00:26:51.640 --> 00:26:56.319
up working in their customer service center
about ten years ago. So let me

430
00:26:56.359 --> 00:26:59.880
just ask one question about that,
which is, what do you take for

431
00:27:00.200 --> 00:27:03.400
your you know, theological training from
seminary that's applicable to the banking space.

432
00:27:03.519 --> 00:27:07.839
It's such a different area of study
of being in the technology side of banking.

433
00:27:07.880 --> 00:27:11.519
I can imagine what the CS degree
brings, but you know what perspective

434
00:27:11.640 --> 00:27:15.000
or different point of view do you
bring from that theological studies into the work

435
00:27:15.000 --> 00:27:18.359
you do today. I mean,
there's so much. I mean, I

436
00:27:18.400 --> 00:27:21.079
could talk for hours on it,
but I think that if I had to

437
00:27:21.200 --> 00:27:23.920
focus on one thing, it would
be that, I mean, we're all

438
00:27:23.920 --> 00:27:29.480
people, I mean, and people
are interesting creatures. I mean, we

439
00:27:29.559 --> 00:27:33.279
have we make all kinds of mistakes, but then we also do all kinds

440
00:27:33.319 --> 00:27:37.359
of good things. And I think
it's, you know, one of those

441
00:27:37.400 --> 00:27:40.519
that when we look at the numbers, we sometimes you have to remember like,

442
00:27:40.559 --> 00:27:44.759
Okay, these are people behind this
data. And once we do that,

443
00:27:44.799 --> 00:27:49.519
we can approach them better and we
can turn around financial lives and it

444
00:27:49.559 --> 00:27:53.079
makes everything just a little bit easier. I love that. Do you have

445
00:27:53.079 --> 00:27:57.759
any tips on how you help the
organization to I mean, it's it's one

446
00:27:57.759 --> 00:28:00.839
of the things I worry about when
I ended up coming from the technology space

447
00:28:00.839 --> 00:28:06.160
into financial services is with how numbers
oriented. It is, mean, loans

448
00:28:06.200 --> 00:28:11.200
deposits, ratios, right interest rates, deltas, spread your nettedge whatever you

449
00:28:11.240 --> 00:28:15.839
can get in this world where you
lose sight of the human beings behind the

450
00:28:15.880 --> 00:28:18.720
numbers and you optimize the business.
You've got knowledge. And there's a great

451
00:28:18.799 --> 00:28:22.400
article about someone at the credit card
Visit, a large credit card credit card

452
00:28:22.440 --> 00:28:26.160
bank, who felt, I got
a couple of years in it was like

453
00:28:26.160 --> 00:28:29.039
a McKenzie project for them, and
they were dialing the numbers to optimize and

454
00:28:29.039 --> 00:28:32.599
they were really getting people in debt
that they couldn't repay and like, and

455
00:28:32.720 --> 00:28:36.880
you lose sight of the impact of
those decisions on that people, particularly as

456
00:28:36.880 --> 00:28:40.839
you get removed from being a frontline
employee, right. And I love that

457
00:28:40.880 --> 00:28:44.200
perspective, but I'm curious if you
have any advice on how you help the

458
00:28:44.279 --> 00:28:48.920
organization or the people who have less
customer facing experience kind of maintain front and

459
00:28:48.920 --> 00:28:52.799
center that perspective of the human nature
of the products and the people that we're

460
00:28:52.799 --> 00:28:56.079
serving and the import of these decisions
on my livesty. I feel like for

461
00:28:56.119 --> 00:28:59.640
me, I've always worried about losing
sight of that and optimizing something that made

462
00:28:59.640 --> 00:29:02.079
the business look great and then you
woke up on more Oh my god.

463
00:29:02.119 --> 00:29:04.200
I've been like I've been hurting people
about making this decision in a way that

464
00:29:04.200 --> 00:29:07.640
I shouldn't have and I didn't even
realize it. Yeah, I think back

465
00:29:07.680 --> 00:29:12.680
to it was right before one of
the first panel discussions it was ever on,

466
00:29:12.839 --> 00:29:15.400
and I was talking with my uncle, you know, Keith Brophy,

467
00:29:15.559 --> 00:29:18.440
and he was I kind of asking
him like, hey, do you have

468
00:29:18.480 --> 00:29:22.200
any advice you know, like for
when I sat on this panel discussion.

469
00:29:22.559 --> 00:29:25.960
He's like, Tim, like,
you can memorize as many facts and figures

470
00:29:26.000 --> 00:29:27.839
as you want. You can talk
about all the statistics, but what people

471
00:29:27.880 --> 00:29:30.240
are going to remember at the end
of the day is the stories you share.

472
00:29:30.839 --> 00:29:34.599
So he goes the more that you
can tie whatever big point you want

473
00:29:34.599 --> 00:29:40.640
to get across and tie that to
a story that's gonna help stick in people's

474
00:29:40.680 --> 00:29:45.039
minds. And I think we can
do the same thing at the banks that

475
00:29:45.079 --> 00:29:48.119
we work at, where it's like, Okay, if we want to try

476
00:29:48.160 --> 00:29:51.880
to do this and improve financial lives
this way, well, let's share ways

477
00:29:51.960 --> 00:29:56.880
where we actually improve financial lives.
Let's share kind of our wins and successes

478
00:29:56.880 --> 00:30:02.599
on a very kind of human personal
level as opposed to just we increased X

479
00:30:02.640 --> 00:30:06.480
by ten percent. Interesting, Yeah, the story is thing really resonates with

480
00:30:06.519 --> 00:30:10.799
me because I do think we are
at the core storytelling creature. We remember

481
00:30:10.839 --> 00:30:15.200
things that way, and so I
often love in um in my slides,

482
00:30:15.400 --> 00:30:18.519
like having single visuals that tie to
stories as opposed to fifteen bullet points,

483
00:30:18.519 --> 00:30:22.200
because if you'll people remember one thing, like tell them the story with a

484
00:30:22.279 --> 00:30:26.119
visual, they remember and they're more
likely to like take that message home and

485
00:30:26.200 --> 00:30:29.680
resonate and remember it than like this
was a slide with six bullet points each

486
00:30:29.680 --> 00:30:33.440
with twenty words. I have no
deal with that slide where I've forgotten,

487
00:30:33.440 --> 00:30:37.599
but like a good story and a
visual to stick in the brain and they

488
00:30:37.599 --> 00:30:41.519
can change your perspective and change your
behavior. All right, Tim, I

489
00:30:41.559 --> 00:30:44.480
got three questions I ask everybody at
the end of this podcast, and unfortunately

490
00:30:44.519 --> 00:30:47.240
my clocks is about thirty minutes are
gone, so it's it's time for us

491
00:30:47.279 --> 00:30:51.279
to drive through them. Um,
and those are first. What's the best

492
00:30:51.279 --> 00:30:56.039
piece of career advice you've ever gotten? Probably the best piece of career advice?

493
00:30:56.759 --> 00:31:00.400
Yeah, really, just try your
best And I got that from both

494
00:31:00.400 --> 00:31:06.279
of my parents of just give forth
your best effort and uh and try your

495
00:31:06.279 --> 00:31:11.440
hardest. How did they impart as
a as a father of children teenagers in

496
00:31:11.480 --> 00:31:15.000
the less, how did they convince
you of the truth of the wisdom of

497
00:31:15.000 --> 00:31:19.119
that perspective at a younger age.
Yeah, really, it's just through practice,

498
00:31:19.200 --> 00:31:22.880
you know. So my mom is
a township treasurer, so really into

499
00:31:22.920 --> 00:31:29.319
the numbers and everything, but very
much into uh, serving her community in

500
00:31:29.319 --> 00:31:33.319
that sense of civic duty. My
dad's a number of different jobs and uh,

501
00:31:33.359 --> 00:31:40.240
you know, from everything from a
plant manager at a large international shoe

502
00:31:40.279 --> 00:31:45.599
factory to h Now he's currently a
living statue. Yea, So he's internationally

503
00:31:45.640 --> 00:31:52.279
renowned living statue, goes around travels
the world in living statue competitions. But

504
00:31:53.359 --> 00:31:56.759
I didn't know that we're living statue. You can be a competitive living statue.

505
00:31:57.039 --> 00:32:00.480
Oh yeah, yeah, yeah,
he does it. It's really big

506
00:32:00.480 --> 00:32:05.799
in Europe, but I mean he
does it and Mackinaw City and kind of

507
00:32:05.799 --> 00:32:08.519
all over. But he uh,
just seeing the passion that he has for

508
00:32:08.599 --> 00:32:13.519
that and kind of any job that
he's ever had, he has passion forward

509
00:32:13.559 --> 00:32:15.480
and just drives his best at it. So put your best foot forward,

510
00:32:15.920 --> 00:32:20.480
love it. Second question, what's
the best piece of advice you've gotten about

511
00:32:20.480 --> 00:32:23.640
the consumer banking or consumer lending industry
as a whole. Yeah, I think

512
00:32:23.680 --> 00:32:28.640
going back to what my you know, uncle said about telling a story.

513
00:32:28.559 --> 00:32:31.400
You know, that's really helped me. So trying to find other people's stories,

514
00:32:31.400 --> 00:32:35.920
trying to put a story to whatever, you know, trying to do,

515
00:32:36.960 --> 00:32:38.640
put a story behind whatever you're trying
to do. All right? And

516
00:32:39.240 --> 00:32:42.880
final question, what's one bold prediction? If I feel like we've cared a

517
00:32:42.880 --> 00:32:45.559
lot of ground for bold predictions,
but I always like one bold prediction and

518
00:32:45.599 --> 00:32:47.160
we can bring you back and hold
your feet to the fire. So give

519
00:32:47.200 --> 00:32:51.839
me one bold prediction for the future. Yeah, if I had to do

520
00:32:51.960 --> 00:32:55.960
one bold one, I would say
within ten years. So it's a longer

521
00:32:57.400 --> 00:33:01.480
timeline, but within ten years there'll
be much more virtual reality, especially in

522
00:33:01.519 --> 00:33:07.240
a business environment. You're a virtual
reality okay, ten years virtual? Have

523
00:33:07.240 --> 00:33:10.839
you have you played with the meta
whatever, the quest whatever they call it,

524
00:33:10.880 --> 00:33:14.640
pro and all the different sides.
Have you have you used them and

525
00:33:14.640 --> 00:33:16.759
you found what are the most compelling
use cases you found for how they're being

526
00:33:16.799 --> 00:33:20.400
utilized to day or what do you
see that being used for. Yeah?

527
00:33:20.519 --> 00:33:22.799
I don't actually use my Oculus two
nearly as much as I used to because

528
00:33:22.880 --> 00:33:29.279
my son, he was eleven soon
to be twelve next Wednesday, is constantly

529
00:33:29.279 --> 00:33:32.160
on Guerrilla Tag, so he's swinging
his arms all around him in a virtual

530
00:33:32.200 --> 00:33:37.279
world, enjoying it. But for
me, the biggest use case I saw

531
00:33:37.319 --> 00:33:40.160
where I was in the light bulb
went off is I put on the I

532
00:33:40.200 --> 00:33:44.039
had my laptop in front of me, and I put on the headset,

533
00:33:44.680 --> 00:33:51.200
and all of a sudden, I
had a huge, like forty foot computer

534
00:33:51.279 --> 00:33:53.440
screen. And at that point I
kind of realized, like, Okay,

535
00:33:53.519 --> 00:33:57.119
this is this is how it will
change, is once you get into the

536
00:33:57.200 --> 00:34:01.279
business world and you have like an
augmented reality environment where instead of having a

537
00:34:01.319 --> 00:34:06.960
computer monitor, you have five different
computer monitors whatever size you want, and

538
00:34:07.160 --> 00:34:09.800
all the visualizations around you. I've
heard that use case. I've heard that

539
00:34:09.840 --> 00:34:15.039
the virtual reality meetings for people who
aren't there are actually very compelling. The

540
00:34:15.079 --> 00:34:19.440
integration. I think Microsoft Teams has
some integrations with the latest Oculus release,

541
00:34:19.519 --> 00:34:22.920
and they're supposed to be pretty compelling. I've not tried them myself, but

542
00:34:22.960 --> 00:34:27.000
I do hear that. Yeah,
all right, so bold prediction on VR

543
00:34:27.000 --> 00:34:28.920
in the future, But you gave
me a ten year time horizon, so

544
00:34:28.920 --> 00:34:30.159
I don't know if this podcast will
be I don't know if podcasts will be

545
00:34:30.239 --> 00:34:35.000
round. This podcast will be round
ten years. Maybe you'll find an interim

546
00:34:35.039 --> 00:34:37.440
period before ten years to check in
on how that one's doing. But Tim,

547
00:34:37.480 --> 00:34:40.000
thanks so much for joining me today. I appreciate all your perspectives.

548
00:34:40.039 --> 00:34:45.559
This is a lot of fun.
Yeah, thank you. Up Start partners

549
00:34:45.559 --> 00:34:49.960
with banks and credit unions to help
grow their consumer loan portfolios and deliver a

550
00:34:50.000 --> 00:34:54.599
modern all digital lending experiments. As
the average consumer becomes more digitally savvy,

551
00:34:54.880 --> 00:35:00.320
it only makes sense that their bank
does too. Up Starts AI lending form

552
00:35:00.480 --> 00:35:06.719
uses sophisticated machine learning models to more
accurately identify risk and approve more applicants than

553
00:35:06.760 --> 00:35:12.800
traditional credit models. With fraud rates
near zero, Upstarts all digital experience reduces

554
00:35:12.880 --> 00:35:17.719
manual processing for banks and offers a
simple and convenient experience for consumers. Whether

555
00:35:17.760 --> 00:35:22.480
you're looking to grow and enhance your
existing personal and auto lending programs or you're

556
00:35:22.480 --> 00:35:28.480
just getting started, Upstart can help. Upstart offers an end to end solution

557
00:35:28.559 --> 00:35:31.519
that can help you find more credit
worthy borrowers within your risk profile. With

558
00:35:31.639 --> 00:35:37.159
all digital underwriting, onboarding, loan
closing, and servicing, it's all possible

559
00:35:37.159 --> 00:35:42.840
with Upstart in your corner. Learn
more about finding new borrowers, enhancing your

560
00:35:42.840 --> 00:35:47.119
credit decisioning process, and growing your
business by visiting upstart dot com slash four

561
00:35:47.280 --> 00:35:53.039
dash banks. That's upstart dot com
slash four dash banks. You've been listening

562
00:35:53.039 --> 00:35:57.920
to leaders and lending from Upstart.
Make sure you never miss an episode.

563
00:35:58.000 --> 00:36:01.440
Subscribe to Leaders in Lending in your
face podcast player using Apple podcasts. Leave

564
00:36:01.480 --> 00:36:05.840
us a quick rating by tapping the
number of stars you think the show deserves.

565
00:36:05.840 --> 00:36:07.760
Thanks for listening until next time.