WEBVTT

1
00:00:03.000 --> 00:00:06.679
All right, ladies and gentlemen,
Hello and welcome back once again to the

2
00:00:06.679 --> 00:00:11.039
only coast to coast radio show all
about the information economy. It's time for

3
00:00:11.119 --> 00:00:14.839
inside analysis. You're sure that the
Eric Cavanaugh is here with the good buddy

4
00:00:14.839 --> 00:00:18.960
of mine, one of my besties
from the Pittsburgh region, Andy Hanna,

5
00:00:19.199 --> 00:00:21.519
who has quite a pedigree. He's
done a lot of really cool things.

6
00:00:21.559 --> 00:00:26.440
We're going to talk about the origin
story for Blue Street Data. I came

7
00:00:26.480 --> 00:00:32.640
across Andy on LinkedIn, which is
obviously the ideal networking platform in social media

8
00:00:32.679 --> 00:00:36.039
for business. Those folks have done
an excellent job. They've also spun out

9
00:00:36.079 --> 00:00:40.520
some great technology. In fact,
a Pachi Kofka which is now everywhere under

10
00:00:40.520 --> 00:00:44.479
the stewardship of a company called Confluence, that came out of LinkedIn. That's

11
00:00:44.560 --> 00:00:49.679
the engine used to run LinkedIn was
called Apachi Kofka. It's now a huge

12
00:00:49.719 --> 00:00:53.200
source and in fact it's a source
of data for one of my other clients,

13
00:00:53.200 --> 00:00:57.960
a company called Oceans. They just
rolled out an announcement today the landed

14
00:00:58.000 --> 00:01:02.079
another forty nine point four million dollars
of investment and extension of their Series B

15
00:01:02.840 --> 00:01:06.000
and Chris glad when their CEO was
telling me the conflut is one of the

16
00:01:06.000 --> 00:01:10.879
big sources of data for their platform, so LinkedIn good stuff. That's how

17
00:01:10.879 --> 00:01:15.319
I met Andy through one of his
master's students at fourteen eighty six Labs.

18
00:01:15.359 --> 00:01:19.159
But he's got a longer origin story
than that, and we'll kind of dive

19
00:01:19.200 --> 00:01:23.920
in. It'll be fun to remember
where we've come from and how we got

20
00:01:23.959 --> 00:01:26.760
here. It's kind of hard to
really wrap your head around the changes.

21
00:01:26.799 --> 00:01:32.760
They've been so significant and so compelling
over the last two decades. But here

22
00:01:32.799 --> 00:01:36.879
we are, so with that Andy
Hannah from fourteen eighty six Labs and the

23
00:01:36.959 --> 00:01:40.439
University of Pittsburgh and oh thought,
well, welcome to Inside Analysis. How's

24
00:01:40.439 --> 00:01:42.400
it going, Eric, Thank you
so much for inviting me. This is

25
00:01:42.519 --> 00:01:46.719
so much fun. I'm so excited
to be on the show this afternoon and

26
00:01:46.840 --> 00:01:49.400
can't wait to chat with you about, you know, the past really thirty

27
00:01:49.480 --> 00:01:57.000
years and what's changed and how we've
evolved over that time. Yeah, And

28
00:01:57.079 --> 00:02:00.480
I remember a friend of mine came
over to my apartment in nw wor Orleans

29
00:02:00.480 --> 00:02:02.640
and was all purpose driven and was
like, you are going to get on

30
00:02:02.719 --> 00:02:05.480
this thing. I'm like, what
are you talking about? It? And

31
00:02:05.560 --> 00:02:07.240
it was the Internet, and I
was like, all right, I had

32
00:02:07.280 --> 00:02:10.560
heard about it through Prodigy. In
fact, when I was a newspaper editor

33
00:02:10.599 --> 00:02:15.000
in Lamont, one of my assistants
was all into Prodigy and it was like,

34
00:02:15.039 --> 00:02:19.039
oh, that's that is pretty interesting. And of course it took off

35
00:02:19.199 --> 00:02:22.879
very very quickly. But you'd have
to think for a second, how did

36
00:02:22.919 --> 00:02:29.360
we get information before the Internet?
And the answer is on floppy drives and

37
00:02:29.639 --> 00:02:32.039
you know others hard medium that you
would carry around. And I was in

38
00:02:32.080 --> 00:02:37.199
the printed uh business, so you
couldn't use floppy drives. You had to

39
00:02:37.240 --> 00:02:39.080
use these big, like portable discs. And I thought it was so cool

40
00:02:39.120 --> 00:02:44.759
to have my portable discs I'd carry
around because that was cool. But you

41
00:02:44.800 --> 00:02:46.520
were there too, so you remember
the early days. How did we get

42
00:02:46.560 --> 00:02:51.520
information around back then? Sure,
I mean you're exactly right, floppy disc

43
00:02:51.560 --> 00:02:53.719
but I remember, you know,
if you were in financial services, you

44
00:02:53.759 --> 00:02:58.840
had multiple terminals on your desk,
right, you had a Bloomberg terminal,

45
00:02:59.680 --> 00:03:04.879
Tom and Reuter's terminal, you had
a terminal from Moody. So it's like

46
00:03:04.919 --> 00:03:09.840
you couldn't even combine the sources or
the you know, the systems in order

47
00:03:09.840 --> 00:03:14.599
to have one screen. You had
to have multiple screens. It was a

48
00:03:14.680 --> 00:03:19.560
really inefficient time when it came into
retrieving data. I mean, I'm sure

49
00:03:19.560 --> 00:03:23.759
you remember if you wanted to give
somebody a file, you had to stick

50
00:03:23.800 --> 00:03:25.759
it somewhere out there on the internet. Literally, you couldn't attach it to

51
00:03:25.800 --> 00:03:30.599
an email. Remember you had an
FTP it, right, And then I

52
00:03:30.680 --> 00:03:34.560
have to say, hey, Eric, here's the secret passageway to where the

53
00:03:34.599 --> 00:03:38.520
information is on the Internet. Go
out there, grab it and download it,

54
00:03:39.000 --> 00:03:43.479
right. I mean, that's what
the early nineties were like, whenever

55
00:03:44.400 --> 00:03:49.599
you were trying to share information.
Yeah, and it's changed a lot.

56
00:03:49.599 --> 00:03:53.680
We're watching Young Sheldon now on Netflix
and they were just joking about that,

57
00:03:53.800 --> 00:03:59.960
how Young Sheldon was on an FTP
transferring files. That's what you would do,

58
00:04:00.159 --> 00:04:03.960
and it moved pretty darn quickly,
you know. Once I remember the

59
00:04:04.000 --> 00:04:08.879
early days. We call them asps, right, Application service providers. And

60
00:04:09.039 --> 00:04:10.960
I once had a smart guy in
my show. I said, what's the

61
00:04:10.960 --> 00:04:14.280
difference between aspns and suffer as a
service and they go SaaS works. I'm

62
00:04:14.280 --> 00:04:20.720
like, oh, I get it, smarty pants. But architectures changed,

63
00:04:20.759 --> 00:04:25.199
and you know, there's tremendous innovation
that came out of companies like Yahoo in

64
00:04:25.279 --> 00:04:29.199
the early days, like had Dupe
that whatever to Google and really started this

65
00:04:30.000 --> 00:04:34.319
or I should say kick start at
the open source movement into very serious data

66
00:04:34.439 --> 00:04:39.639
territory, right because Linux was the
foundation of open source. And then ten

67
00:04:39.680 --> 00:04:44.120
odd years later you had this whole
revolution in data management with had Dupe and

68
00:04:44.160 --> 00:04:48.160
a whole ecosystem around that. Of
course, Apache Kafka is open source and

69
00:04:48.680 --> 00:04:54.759
all the so the channels from moving
information have been rapidly advancing. We've got

70
00:04:54.759 --> 00:04:58.319
Snowflake, now we've got data bricks. We all have all these amazing companies

71
00:04:58.759 --> 00:05:01.240
that are allowing the use of analytics. But you and I have been there

72
00:05:01.279 --> 00:05:04.079
from the word go, so we
remember what it was like to try to

73
00:05:04.079 --> 00:05:09.879
cobble this stuff together. And really
you have to constantly change your mindset about

74
00:05:09.959 --> 00:05:14.000
things and recognize that the times have
changed, and recognize you have to be

75
00:05:14.199 --> 00:05:17.439
using the new platforms as time goes
by. And I guess that's an epiphany

76
00:05:17.439 --> 00:05:21.519
that you had as well, Right, Yeah, THO was in the mid

77
00:05:21.600 --> 00:05:26.120
nineteen nineties. I would say,
right around ninety five. There's a really

78
00:05:26.279 --> 00:05:30.160
visionary guy's named Gary Muller, wonderful
guy who was doing research in Poland.

79
00:05:30.959 --> 00:05:35.639
And this is remember not too long
after the fall of the Soviet Union,

80
00:05:36.279 --> 00:05:41.839
and the former Soviet countries all wanted
to privatize their assets. So what Gary

81
00:05:41.959 --> 00:05:47.000
was doing was doing research about a
particular industry, and what he was finding

82
00:05:47.079 --> 00:05:50.920
is that there was a dearth of
information. And so a bunch of us

83
00:05:50.959 --> 00:05:57.839
got together and said, Okay,
let's build this information infrastructure about the former

84
00:05:57.920 --> 00:06:02.720
Soviet Union. So countries like Russia, pull in Ukraine, you know,

85
00:06:03.000 --> 00:06:08.519
the Czech Republic, et cetera.
And what we did is we sent little

86
00:06:08.600 --> 00:06:14.680
teams into each of those countries and
they mine data. Literally we either we

87
00:06:14.759 --> 00:06:19.279
got it somehow digitally or we went
to copy machines and actually copied it.

88
00:06:20.279 --> 00:06:25.199
And I'm serious that because it wasn't
in existence, we had to go to

89
00:06:25.279 --> 00:06:30.560
banks, companies, analysts, and
then we would consolidate it and all in

90
00:06:30.639 --> 00:06:33.199
digital form. And we needed a
way to distribute this information. We weren't

91
00:06:33.199 --> 00:06:36.839
going to do it over Bloomberg or
Reuters, and so we said, hey,

92
00:06:36.839 --> 00:06:41.639
there's this really cool thing called the
internet. You know, Yahoo seems

93
00:06:41.639 --> 00:06:46.639
to be making it work, and
so we literally put you know, we

94
00:06:46.759 --> 00:06:49.279
probably created one of the first SaaS
based applications out there. We put it

95
00:06:49.319 --> 00:06:55.600
on the one and we got banks
and large companies wanted to investor do business

96
00:06:55.600 --> 00:07:00.519
in the former Soviet Union. They
connected into these these database and they paid

97
00:07:00.600 --> 00:07:05.480
per drink essentially, and so meaning
that every time that they went when in

98
00:07:05.600 --> 00:07:09.680
pulled information off, they would they
would pay. It was a it was

99
00:07:09.720 --> 00:07:13.800
a total gas. We had to
put the first Internet Service Provider I s

100
00:07:13.920 --> 00:07:16.160
P for all those who are kind
of from a different generation, but we

101
00:07:16.199 --> 00:07:20.720
had to put the ISP first I
P and poland out there. We actually

102
00:07:20.720 --> 00:07:26.120
owned the first ISPN poll wow,
and so that was what it was.

103
00:07:26.199 --> 00:07:30.439
It was crazy, it was a
wild time and it was it was pre

104
00:07:30.600 --> 00:07:34.240
infrastructure, and so we developed our
own accounting systems, we developed our own

105
00:07:34.279 --> 00:07:39.519
distribution systems. It's kind of crazy. It's a lot of fun. Yeah.

106
00:07:39.560 --> 00:07:43.519
Well, and you were at the
very forefront of what is now called

107
00:07:43.800 --> 00:07:46.279
alternative data. And you and I
have talked about this. I have done

108
00:07:46.279 --> 00:07:48.959
a lot of research into it,
and you know, like you, I

109
00:07:49.000 --> 00:07:53.000
think down the road, like hmm, how is this going to pan out?

110
00:07:53.360 --> 00:07:57.720
And clearly the amount of alternative data
being bought and sold today is just

111
00:07:58.199 --> 00:08:01.800
vast. It's stunning and a lot
of people don't realize that credit card companies

112
00:08:01.839 --> 00:08:05.519
sell what they call exhaust data,
which is then bought by a whole variety

113
00:08:05.560 --> 00:08:11.319
of firms, most notably investment bankers. And so I wrote about this a

114
00:08:11.360 --> 00:08:15.839
number of years ago, saying that
it's a pretty unfair playing field if these

115
00:08:15.879 --> 00:08:20.800
investment bankers can have all that data
at their disposal, because now they can

116
00:08:20.879 --> 00:08:24.399
see, for example, who's going
to meet or beat market estimates on revenue

117
00:08:24.399 --> 00:08:28.240
because they have the actual raw data
and they can compare it to a baseline.

118
00:08:28.439 --> 00:08:33.720
So why don't we have some publicly
facing consumer data lake with my suggestion,

119
00:08:33.159 --> 00:08:35.559
so you could level the playing field, and so the rest of us

120
00:08:35.559 --> 00:08:39.159
could all benefit from that. And
you're kind of seeing the start of that

121
00:08:39.279 --> 00:08:43.759
in sort of in different places.
And of course you also have a history

122
00:08:43.879 --> 00:08:48.320
in the machine learning plus data space
through this company called O thought right tell

123
00:08:48.399 --> 00:08:52.120
us about that? Yeah, sure, I think one stop before then I'll

124
00:08:52.159 --> 00:08:58.440
talk a little bit about sort of
that transformation from data is powerful to data

125
00:08:58.440 --> 00:09:03.159
plus analytics or AI is more powerful. And that was the time that I

126
00:09:03.200 --> 00:09:07.039
spent A good friend of mine started
a company called the International Institute for Analytics.

127
00:09:07.120 --> 00:09:09.720
His name is Jack Phillips along with
Tom Davenport, who we all know

128
00:09:09.960 --> 00:09:16.159
is sort of the guru and business
analytics. And you know, about fifteen

129
00:09:16.240 --> 00:09:18.919
years ago they came together and said, you know what, we need to

130
00:09:18.960 --> 00:09:22.960
be the leading research firm about how
firms are using analytics, looking at the

131
00:09:24.000 --> 00:09:30.480
maturity of companies in terms of low
versus high, and then whether or not

132
00:09:30.519 --> 00:09:35.799
they outperform companies that actually are less
analytic. And you know, the research

133
00:09:35.840 --> 00:09:39.720
holds true, the more analytic you
are, the better you use data the

134
00:09:41.440 --> 00:09:45.600
top of the heap you'll be.
So good evidence of that is if we

135
00:09:45.639 --> 00:09:48.480
look at the top ten companies of
the s and P five hundred. You

136
00:09:48.480 --> 00:09:54.120
know, these are the Metas,
the Googles, the Microsoft's, the Tesla's,

137
00:09:54.240 --> 00:09:58.240
you know, they're using data in
a very different way Amazon, and

138
00:09:58.639 --> 00:10:03.440
they represent four percent of the total
value of the SMP five hundred. While

139
00:10:03.759 --> 00:10:09.480
so they learned they're the early adopters. They understood power data power of analytics

140
00:10:09.519 --> 00:10:16.080
together and I feel very fortunate to
spend time with Jack and Tom and do

141
00:10:16.240 --> 00:10:20.720
some consulting some very large companies and
understand this transformation that was going on.

142
00:10:22.600 --> 00:10:26.399
And then me and a guy named
John Evattico, guy named Jeremy Garvey started

143
00:10:26.480 --> 00:10:31.559
said hey, we're going to do
this. So twenty fourteen we started this

144
00:10:31.639 --> 00:10:35.080
company called Othought and the idea of
out o thought was, can we at

145
00:10:35.159 --> 00:10:43.799
scale take data about students, high
school students and help universities colleges enroll the

146
00:10:43.840 --> 00:10:48.600
best fit students who are going to
persist and graduate, and how can we

147
00:10:48.639 --> 00:10:56.440
help those institutions get do better,
get more graduated on time, get better

148
00:10:56.519 --> 00:11:01.440
jobs, et cetera. So that's
where we took machine learning coupled with all

149
00:11:01.480 --> 00:11:09.159
the rich data that high school students
provide to universities, combine those two together

150
00:11:09.480 --> 00:11:13.679
and we build a thought. And
that was a seven or eight year journey.

151
00:11:13.159 --> 00:11:18.399
Wow. Well, and just to
explain to our audience here, what

152
00:11:18.480 --> 00:11:22.840
you're able to do with data at
scale and analytics and especially machine learning is,

153
00:11:22.919 --> 00:11:26.960
of course, identify patterns. But
then the real magic is once you've

154
00:11:26.960 --> 00:11:31.360
identified a pattern, let's say a
pattern of a successful student. Once you've

155
00:11:31.360 --> 00:11:35.840
identified that pattern, then you can
sort of distill it, understand it as

156
00:11:35.840 --> 00:11:39.360
almost like a recipe, and then
apply it to the rest of the data

157
00:11:39.679 --> 00:11:45.039
and see Aha, So these forty
out of one hundred students actually fit that

158
00:11:45.159 --> 00:11:48.639
pattern of behavior, in terms of
interests, in terms of writing style.

159
00:11:48.679 --> 00:11:52.200
There are all sorts of different bits
and pieces you can cobble together. And

160
00:11:52.279 --> 00:11:56.720
that's really the power of this analytical
technology, right, is that you can

161
00:11:56.840 --> 00:12:01.080
understand what works and what doesn't work. You can model both and then analyze

162
00:12:01.120 --> 00:12:05.679
current data to say, Okay,
according to our model that we've built,

163
00:12:05.039 --> 00:12:09.080
these forty students are going to do
very well. These sixty probably not so

164
00:12:09.159 --> 00:12:11.039
much. And you'll never get it
completely right. I mean, there's I

165
00:12:11.039 --> 00:12:15.360
think a bit of a misconception that, oh, with all the best analytics,

166
00:12:15.399 --> 00:12:16.360
you will always be right. No, you won't always be right,

167
00:12:16.559 --> 00:12:20.759
but you will greatly increase your chances
of being correct, and you'll create what's

168
00:12:20.799 --> 00:12:26.720
what's called lift right one hundred percent. So that's so let's let's even maybe

169
00:12:26.759 --> 00:12:30.240
easier for the audience to understand,
is if we're looking at the probability of

170
00:12:30.320 --> 00:12:33.519
somebody to enroll, right, So
what let's say that Eric, we want

171
00:12:33.559 --> 00:12:37.320
you to enroll at the University of
Pittsburgh. We see that you're twenty percent

172
00:12:37.480 --> 00:12:41.759
likely to enroll because that's what the
machine learning has told us, and that's

173
00:12:41.799 --> 00:12:46.279
because of who you are and how
you behave, so we know all the

174
00:12:46.360 --> 00:12:48.519
data about where you went to high
school, your grades, you know,

175
00:12:48.559 --> 00:12:52.840
the activities, et cetera. And
then we see your behavior. Right,

176
00:12:52.879 --> 00:12:56.120
so you're hitting website, what you're
reading, you know, the visits that

177
00:12:56.200 --> 00:13:00.720
you go to. Are you an
engaged individual? So you land this twenty

178
00:13:00.759 --> 00:13:03.960
percent? Well, we want to
do if this you know, we want

179
00:13:03.120 --> 00:13:07.559
Eric, we want to be eighty
percent. So we look at the prescriptions.

180
00:13:07.159 --> 00:13:11.320
What can we do that's going to
increase his probability? So for example,

181
00:13:11.320 --> 00:13:16.399
we might send you a particular marketing
campaign, give you a particular scholarship,

182
00:13:16.480 --> 00:13:18.159
visit you in a high school,
and all of a sudden, because

183
00:13:18.159 --> 00:13:22.000
we are studying the data in your
patterns, just like you said, we

184
00:13:22.039 --> 00:13:28.360
see that that those activities raise your
probability from twenty to eighty percent. This

185
00:13:28.519 --> 00:13:31.440
and this applies to every industry.
This happens to be higher education that we're

186
00:13:31.480 --> 00:13:35.080
talking about. But this is how
Amazon gets you to buy all the products

187
00:13:35.120 --> 00:13:39.480
that it gets you to buy to
that recommendation engine. Yeah, and it's

188
00:13:39.720 --> 00:13:43.679
what's cool is that you can learn
about the students you know. So it's

189
00:13:43.720 --> 00:13:48.639
great for improving efficacy, but it's
also great for educating the people who are

190
00:13:48.639 --> 00:13:52.399
involved in the process, right,
because when you play with these models,

191
00:13:52.399 --> 00:13:56.440
when you play with the data,
that's when you start to understand how it

192
00:13:56.480 --> 00:14:00.360
all fits together, how the tumblers
align. And I think at a very

193
00:14:00.399 --> 00:14:05.320
exciting time in our business world and
the education world as well, because we've

194
00:14:05.360 --> 00:14:07.919
reached a bit of a critical mass
in that regard, meaning we have the

195
00:14:07.919 --> 00:14:11.399
compute power now, we have tons
and tons of data, we have smart

196
00:14:11.399 --> 00:14:16.240
people who have methodologies, who have
published these methodologies that we can follow and

197
00:14:16.320 --> 00:14:20.360
understand. So it's really all coming
together in terms of being able to leverage

198
00:14:20.559 --> 00:14:24.960
this stuff. And then you look
at artificial intelligence CHAT, GBT large language

199
00:14:26.000 --> 00:14:28.879
models, which are of course are
very very powerful. They are going to

200
00:14:28.879 --> 00:14:35.159
require some efforts to govern and to
manage responsibly. But I think we're up

201
00:14:35.159 --> 00:14:37.480
to the task now, right because
of all this experience, because what we

202
00:14:37.559 --> 00:14:43.240
have is so valuable, and I
think that transparency and ethics are really going

203
00:14:43.279 --> 00:14:46.039
to come into play here in the
near future. What do you think hugely,

204
00:14:46.240 --> 00:14:48.759
hugely important. I mean, one
of the things that you said that

205
00:14:48.799 --> 00:14:52.440
people don't often get is it's not
just about predicting the future. These models

206
00:14:52.440 --> 00:14:56.519
that you build allow you to diagnose
where you might have issues, so you

207
00:14:56.559 --> 00:15:01.480
can detect bias through these. So
if you're if there's bias in the data

208
00:15:01.559 --> 00:15:05.039
or biased and the interpretation of the
results, you can look at it and

209
00:15:05.080 --> 00:15:09.440
you can do some reverse engineering and
say, hey, what if I change

210
00:15:09.480 --> 00:15:13.320
the gender? What if I change
does the results? Do the results change?

211
00:15:13.559 --> 00:15:18.960
Oh? Wow, this particular sub
segment is disadvantage because of that,

212
00:15:20.120 --> 00:15:22.320
because it just just by changing the
gender, which means that we have a

213
00:15:22.360 --> 00:15:26.559
problem. Right, So then we've
diagnosed it. Now we can fix it.

214
00:15:26.120 --> 00:15:31.360
So and I think that that's an
area that we're really just starting to

215
00:15:31.440 --> 00:15:35.320
dive into so that we can understand
how to fix the problems that exist.

216
00:15:35.440 --> 00:15:39.799
And by the way, we'll never
fix all the problems. We just got

217
00:15:39.840 --> 00:15:43.639
to put on the big ones.
Yeah. Well, and it's a journey,

218
00:15:43.720 --> 00:15:46.080
right, it's a process. And
of course, when you fix one

219
00:15:46.120 --> 00:15:48.120
thing, other things break. It's
like painting your house. You paint one

220
00:15:48.120 --> 00:15:52.200
wall, you got to paint them
all, right, I cannot unless you

221
00:15:52.279 --> 00:15:54.679
do that. But you're hitting on
something very important, which I think is

222
00:15:54.720 --> 00:16:02.919
probably the most valuable proposition about a
machine learning is that it helps you get

223
00:16:03.000 --> 00:16:08.039
to understanding the problems. That discovery
side is very challenging because it could be

224
00:16:08.080 --> 00:16:12.000
anywhere. I often use the example
of you lose your wallet, You're looking

225
00:16:12.000 --> 00:16:15.039
for it all around the house.
Is it even in the house. It's

226
00:16:15.120 --> 00:16:17.840
very frustrating because you don't know exactly
where it is, don't know where to

227
00:16:17.840 --> 00:16:19.320
look. That's kind of the way
it is with large sets of data too.

228
00:16:19.399 --> 00:16:25.039
You have to start filtering it and
taking it a hard look at things.

229
00:16:25.080 --> 00:16:27.600
Like you said, let's flip out
the gender see if things change.

230
00:16:27.679 --> 00:16:32.960
Oh wow, they change dramatically.
That's part of the process of working with

231
00:16:33.080 --> 00:16:36.759
the data, of trying to understand
the data. And it takes effort.

232
00:16:36.799 --> 00:16:38.960
I mean it's we're getting better and
better with the technologies, but still it

233
00:16:40.000 --> 00:16:44.120
takes human beings in the loop it
analyzing the data. I'll give you a

234
00:16:44.159 --> 00:16:47.360
fun observation before a first break,
come ef for in a minute. What

235
00:16:47.519 --> 00:16:49.480
good buddy of mine, who also
runs an analytics institute, He goes,

236
00:16:49.519 --> 00:16:53.639
Yeah, machines don't have the ability
to go huh, that's kind of weird,

237
00:16:55.639 --> 00:16:59.399
but human beings do, right,
Yeah, I mean that's a good

238
00:16:59.440 --> 00:17:03.600
point. You know, I thought
our chief data scientist, Mark Fordman,

239
00:17:03.200 --> 00:17:07.359
he's actually developed what we call the
automated data scientists. So does the automated

240
00:17:07.400 --> 00:17:11.720
data scientist does exactly what you're talking
about. It scans results, it scans

241
00:17:11.880 --> 00:17:17.440
data, it scans changes, you
know, model drift, and it says

242
00:17:17.480 --> 00:17:21.240
aha, it tells our people,
aha, we got a problem over here.

243
00:17:21.279 --> 00:17:23.200
So it's like a member of the
team, which is how I think

244
00:17:23.200 --> 00:17:26.480
that we need to start thinking about
AI as a member of the team and

245
00:17:26.519 --> 00:17:30.160
not an adversary. That's a really, really good point. And we're up

246
00:17:30.200 --> 00:17:33.920
to our first break here, but
don't touch up now, folks. We're

247
00:17:33.960 --> 00:17:37.039
talking with Andy Hannah all about machine
learning and data and the origin story of

248
00:17:37.079 --> 00:17:44.960
blue street data. Will be right
back. You're listening to Inside Analysis.

249
00:17:47.599 --> 00:17:55.559
Welcome back to Inside Analysis. Here's
your host, Eric Tavanaugh. Okay,

250
00:17:55.559 --> 00:17:59.839
folks, back here on Inside Analysis
with the one and only Andy Hannah for

251
00:18:00.039 --> 00:18:03.000
teen eighty six Labs or get to
that in just a minute. And of

252
00:18:03.039 --> 00:18:06.160
course, oh thought, and in
the break there, we're just chatting about

253
00:18:06.200 --> 00:18:10.279
the fact that data is changing,
the nature of data is changing, and

254
00:18:10.319 --> 00:18:14.640
what we do with it is changing. You have these cycles. Basically,

255
00:18:14.920 --> 00:18:18.440
I'll just throw one fun observation out
there that I've seen in the recent past,

256
00:18:18.799 --> 00:18:23.079
which is observability. This whole space
blew up called observability, and basically

257
00:18:23.079 --> 00:18:29.200
it's just windows into feeds of machine
data that's cruising around of things that are

258
00:18:29.200 --> 00:18:33.680
happening, connections to data sets,
for example. And within like a year

259
00:18:33.720 --> 00:18:37.359
to two years of observability becoming a
thing, there's already the company that sits

260
00:18:37.400 --> 00:18:41.720
on top of your observability data and
allows you to manage it. So there's

261
00:18:41.720 --> 00:18:45.119
already a meta solution. We had
them on the show last year that allows

262
00:18:45.160 --> 00:18:52.160
you to It's called Mesmo, and
they allow you to transform and process and

263
00:18:52.200 --> 00:18:56.920
correlate data streams from observability. Right, So these are new ways of dealing

264
00:18:56.920 --> 00:19:00.279
with data. It's not all unstructured
data like human texts and images and things

265
00:19:00.319 --> 00:19:03.640
that they think. A lot of
it is machine data, so as they

266
00:19:03.640 --> 00:19:07.599
talk to each other, it's capturing
that information. But the point is this

267
00:19:07.640 --> 00:19:11.000
is all grist for the mill,
as they say, and what data analysts

268
00:19:11.000 --> 00:19:15.039
and business people in general need to
appreciate these days is that it is changing

269
00:19:15.079 --> 00:19:19.079
and it continues to change. So
you have to kind of change the pace

270
00:19:19.119 --> 00:19:23.720
of your own processes and workflows and
understandings to be able to leverage itself,

271
00:19:23.720 --> 00:19:26.839
and you'll never be completely up to
date, you know what I mean.

272
00:19:26.880 --> 00:19:30.000
You're never going to be riding the
crush of the way for any significant time

273
00:19:30.319 --> 00:19:33.000
because everything is changing all around us. So you kind of have to be

274
00:19:33.440 --> 00:19:37.079
very agile and very open minded and
by the way, put some governance into

275
00:19:37.079 --> 00:19:41.640
place. Right. Governance, I
think is going to be a hugely popular

276
00:19:41.680 --> 00:19:47.319
topic for many years to come because
we're in such a turbulent time and because

277
00:19:47.519 --> 00:19:52.200
the outcomes are still relatively unknown.
Well, if you don't have governance and

278
00:19:52.240 --> 00:19:55.920
you flip the switch of aon and
your aion and your organization, you're going

279
00:19:56.000 --> 00:19:57.799
to get some crazy stuff flying on
the other end. It's not going to

280
00:19:57.839 --> 00:20:02.880
work well for you. But give
me your thoughts on that, the changing

281
00:20:03.039 --> 00:20:06.799
nature of the data landscape and how
you have to change with it. Yeah,

282
00:20:06.839 --> 00:20:11.319
So we're I think we're really lucky
that we've seen the evolution of this

283
00:20:11.640 --> 00:20:15.920
technology and the source of the power
of that technology, which is the data

284
00:20:17.440 --> 00:20:19.920
over to pass forty years, right, So if you go back forty years,

285
00:20:19.960 --> 00:20:23.200
we know, you know, Yahoo
had just been going public, you

286
00:20:23.200 --> 00:20:27.839
know, or just formed. So
we've come a long way since those first

287
00:20:27.960 --> 00:20:33.279
days and even if it's only two
decades since Yelp and Twitter have been around,

288
00:20:33.319 --> 00:20:37.720
and really the growth of that and
the use of that data is over

289
00:20:37.759 --> 00:20:41.200
the past decade. So the ability
and the examples that you give are wonderful

290
00:20:41.240 --> 00:20:47.400
are but the ability to think about
the use of that data that didn't exist,

291
00:20:47.960 --> 00:20:52.079
That user level data did not exist
until a decade ago. And so

292
00:20:52.119 --> 00:20:56.839
the ability for and we see that
as we mentioned, of the top companies

293
00:20:56.279 --> 00:21:02.960
in the country using that data to
understand behavior and to change that behavior to

294
00:21:03.039 --> 00:21:07.440
increase probability of buying those types of
things. How do we you know,

295
00:21:07.559 --> 00:21:10.440
how does the rest of the world. Maybe this is a good question for

296
00:21:10.519 --> 00:21:14.559
you. We know, it's like
the rich getting richer and how does the

297
00:21:14.599 --> 00:21:18.079
poor move up? Right? Because
these large companies are the ones who are

298
00:21:18.079 --> 00:21:23.839
pushing the edges of technology and the
use of data, when the eighty percent

299
00:21:23.880 --> 00:21:29.960
of the companies that are out there
are still trying to understand digital transformation and

300
00:21:30.000 --> 00:21:33.599
what can I do with that data? They're learning about the use cases and

301
00:21:33.640 --> 00:21:37.440
the applications. You know, are
we going to continue to have this huge

302
00:21:37.519 --> 00:21:41.480
divide so that the you know,
the true value that's driven by AI and

303
00:21:41.720 --> 00:21:47.519
data is really housed in a small
number of companies. What do you I

304
00:21:47.559 --> 00:21:49.440
mean, what do you think,
Derek Ver, I mean, that's that's

305
00:21:49.480 --> 00:21:52.960
a very very big question. I
think transparency is going to be the key.

306
00:21:53.440 --> 00:21:56.759
I think this idea through at the
top of the hour of a consumer

307
00:21:56.799 --> 00:22:02.640
facing data lake basically would be because
it's a place where we can all dive

308
00:22:02.680 --> 00:22:06.720
in and start of better understand what's
happening with this data and ownership. You

309
00:22:06.720 --> 00:22:10.960
know, data ownership is a big
issue these days. You're starting to see

310
00:22:11.039 --> 00:22:15.079
some of the bigger companies at least
give lip service to this, but allowing

311
00:22:15.119 --> 00:22:18.240
you to own your data, giving
you control of when your data can be

312
00:22:18.359 --> 00:22:22.039
used to train algorithms for example.
There are lots of people coming up with

313
00:22:22.119 --> 00:22:26.559
business models to allow you to own
your data and control that. But it's

314
00:22:26.759 --> 00:22:32.039
it's hard. It's going to be
hard to unseat the Amazons and the Microsofts

315
00:22:32.039 --> 00:22:33.839
and the Googles of the world.
But I will tell you that the big

316
00:22:33.880 --> 00:22:37.799
guys are nervous too. There is
tremendous tension at Google right now. They

317
00:22:37.799 --> 00:22:42.240
are very nervous about what just happened
with chat GPT in particular, and then

318
00:22:42.279 --> 00:22:47.720
Microsoft stepping in to buy it basically
or invest heavily in it, I should

319
00:22:47.720 --> 00:22:51.319
say. And then what happened.
The board for open Ai kicks out Sam

320
00:22:51.359 --> 00:22:53.960
Altman, and over the course of
the weekend, Sata Nadela offers to hire

321
00:22:55.000 --> 00:22:59.279
him, and you're just like,
Holy Christmas, what is going on?

322
00:23:00.160 --> 00:23:06.000
Yeah, we do need transparency,
and I think that while I fully respect

323
00:23:06.079 --> 00:23:11.000
the concept of intellectual property and having
to protect these sorts of things, you

324
00:23:11.079 --> 00:23:15.519
get right down to it, these
models are so incredibly complex that you could

325
00:23:15.680 --> 00:23:21.839
offer complete transparency and I don't think
you're jeopardizing your business model really much at

326
00:23:21.880 --> 00:23:25.839
all. And in fact, I
just heard today that apparently Elon Musk is

327
00:23:25.839 --> 00:23:30.680
going to open source Grock. Of
course Mark Zuckerberg open source Lama or Lama

328
00:23:30.720 --> 00:23:33.559
too, So you know, again
you kind of see these these movements,

329
00:23:33.640 --> 00:23:38.519
and I just read today that who
is it? I think Google is following

330
00:23:38.559 --> 00:23:44.839
suits and allowing you to exit their
data centers right because they're all these exit

331
00:23:44.880 --> 00:23:48.880
fees. Basically that we're just killing
people. I mean, that was the

332
00:23:48.480 --> 00:23:51.839
big joke is you can put all
your data in the cloud for cheap,

333
00:23:51.839 --> 00:23:53.039
but if you try to take it
out those egress fees, we're going to

334
00:23:53.119 --> 00:23:57.559
hammer you. I do think there's
pressure on the big guys, but I

335
00:23:57.559 --> 00:24:02.000
think you're right that we need to
be very careful. And you know there

336
00:24:02.079 --> 00:24:07.160
was concern what if this kind of
LLLM capability is restricted to a handful of

337
00:24:07.200 --> 00:24:10.799
people, Well, that would be
a pretty bad situation. I think we

338
00:24:10.839 --> 00:24:14.519
want we all want access to these
technologies, and I think it's good that

339
00:24:14.559 --> 00:24:18.519
we have competition at the highest levels
that's pushing towards that transparency. But I

340
00:24:18.559 --> 00:24:25.519
think transparency and education are the two
massive keys to success to give power to

341
00:24:25.559 --> 00:24:27.799
the little people out there. What
do you think, Yeah, I agree

342
00:24:27.799 --> 00:24:33.000
with that. I think that that's
a tall order. I think you know

343
00:24:33.240 --> 00:24:37.880
that we were talking earlier about data
warehousing, and you know there's still there's

344
00:24:37.880 --> 00:24:42.039
still lots of companies trying to figure
out what's my single source of truth as

345
00:24:42.079 --> 00:24:48.039
opposed to thinking about what's the latest
in architecture that I can use both external

346
00:24:48.119 --> 00:24:52.319
data and internal data to outcompete those
who are in my space. Right,

347
00:24:52.359 --> 00:24:59.039
So it's I just I look across
the spectrum of companies and I am just

348
00:24:59.319 --> 00:25:07.079
for by how the distribution of capabilities, the distribution of knowledge, and there

349
00:25:07.119 --> 00:25:11.000
there are so many companies that are
just at the beginning of this journey,

350
00:25:11.279 --> 00:25:14.279
and you and I, you know, understand this the end of the journey

351
00:25:14.279 --> 00:25:17.119
because we work with the larger companies
are on the edge. You know,

352
00:25:17.160 --> 00:25:22.000
we understand retail media networks, and
we understand data clean rooms and data fabric

353
00:25:22.079 --> 00:25:26.720
and all this. But you know, most most companies are going to Gartner's

354
00:25:26.720 --> 00:25:30.039
annual conference to learn about what the
heck these things are, right, let

355
00:25:30.079 --> 00:25:34.440
alone being able to gain value out
of out of this technology. So I

356
00:25:34.519 --> 00:25:38.079
completely agree with you, you know, the transparency and the education. I

357
00:25:38.200 --> 00:25:44.079
just don't know how it's going to
happen without mechanisms to say, Okay,

358
00:25:44.519 --> 00:25:48.319
here's how you can get value,
you know, and then let's help.

359
00:25:48.440 --> 00:25:52.119
Let's help each other, Like on
fraud, as an example, we can

360
00:25:52.200 --> 00:25:56.640
share knowledge about fraud detection, right, It's it's do we want to compete

361
00:25:56.680 --> 00:25:59.920
on fraud detection? I hope not? Right, that's something that we want

362
00:26:00.119 --> 00:26:03.920
to share. So why don't we
have a common way of looking at fraud

363
00:26:03.319 --> 00:26:10.279
and helping each other in every industry
understand where there's fraudulent customers and or customers

364
00:26:10.319 --> 00:26:12.799
that don't even exist. And so
I think we have to figure out a

365
00:26:12.839 --> 00:26:18.319
way to get people to that base
camp of use cases and understand how they

366
00:26:18.400 --> 00:26:22.680
work. I like the concept the
base camp of use cases. You're absolutely

367
00:26:22.759 --> 00:26:26.279
right about fraud. I've talked about
that on this show even fifteen years ago,

368
00:26:26.359 --> 00:26:30.200
with the same exact thought. It's
like, let's share threat detection data,

369
00:26:30.559 --> 00:26:34.599
at least within industries and the financial
services, in healthcare and insurance and

370
00:26:34.640 --> 00:26:41.519
different industries where they're the same business
model. Let's share information about threats because

371
00:26:41.559 --> 00:26:45.000
they can spread very quickly and they
can bring companies down. I mean,

372
00:26:45.000 --> 00:26:48.640
you were hearing all these stories about
ransomware and just terrible things that are happening

373
00:26:48.640 --> 00:26:52.519
with people stealing your data, stealing
access to your data and causing all kinds

374
00:26:52.559 --> 00:26:56.759
of trouble. But I will say
the good news is that the young people

375
00:26:56.799 --> 00:27:02.160
today, they're all on their iPhone, they're on their Samsungs, they're on

376
00:27:02.200 --> 00:27:07.519
their different devices, and they understand
technology as as a first class citizen,

377
00:27:07.960 --> 00:27:10.880
whereas a lot of us, you
know, middle aged folks, let's say,

378
00:27:11.200 --> 00:27:12.960
you know, we've had the whole
journey of knowing what it was like

379
00:27:14.079 --> 00:27:18.559
before even computers for crying out loud, before that was mainstream, where like

380
00:27:18.599 --> 00:27:22.480
reading books, and you know,
you do have to. I think every

381
00:27:22.519 --> 00:27:26.599
so often gut check yourself and mind
check yourself and say, all right,

382
00:27:26.680 --> 00:27:30.160
do am I still lingering in these
old habits and answers? You probably are,

383
00:27:30.839 --> 00:27:33.759
So how do you shake out of
that? And I think you do

384
00:27:33.799 --> 00:27:38.240
it by listening to people and just
by going to events and going to webinars

385
00:27:38.279 --> 00:27:42.319
and things, and just taking a
moment to hear what other people are talking

386
00:27:42.359 --> 00:27:45.680
about and to see and to watch
how they interact with data. You know.

387
00:27:45.720 --> 00:27:51.359
That's why I'm so excited about transparency
of data sets. And it's really

388
00:27:51.359 --> 00:27:53.039
come a long way in the last
i'd say ten years. And you guys,

389
00:27:53.039 --> 00:27:56.680
of course, are part of this, but transparency data because it's only

390
00:27:56.759 --> 00:28:02.559
once you really start playing around they
say, munging the data that you start

391
00:28:02.559 --> 00:28:06.279
to get a better understanding for what
that process is like and how it really

392
00:28:06.319 --> 00:28:08.920
does take time. Even if you
have the best technologies, it takes time

393
00:28:08.960 --> 00:28:12.519
to play around flat things, move
things around, see it in motion,

394
00:28:14.160 --> 00:28:15.960
you know. But that's where you
start to get it and you're like,

395
00:28:17.039 --> 00:28:19.599
oh, wait a minute, And
that's the beginning of a career, right,

396
00:28:21.640 --> 00:28:25.200
yeah, it's you know, it's
interesting. We're you know, a

397
00:28:25.240 --> 00:28:29.599
lot of people don't know this.
We now have a major, an analytics

398
00:28:29.640 --> 00:28:33.519
major at the University of Pittsburgh for
undergraduate business students. See, these students

399
00:28:33.559 --> 00:28:38.400
are coming in in their freshman year
and they're learning to program in Python.

400
00:28:38.440 --> 00:28:42.880
They're learning how to mind data,
they're learning how to use that data and

401
00:28:44.599 --> 00:28:49.039
you know, different models to make
decisions. I mean, so they're they're

402
00:28:49.279 --> 00:28:56.440
graduating with with a language that most
people at the organizations that they're going into

403
00:28:56.640 --> 00:29:00.880
only casually understand. I find that
as a very interesting sort of You have

404
00:29:00.960 --> 00:29:08.799
this group of really fluent in technology
new hires and then up to levels people

405
00:29:10.079 --> 00:29:14.119
are their their their level knowledge is
what they've read, you know, from

406
00:29:14.599 --> 00:29:19.160
McKenzie or from you know, something
from you know, towards data science or

407
00:29:19.200 --> 00:29:22.200
something like that, that they are
getting in the nuance of it, but

408
00:29:22.279 --> 00:29:25.880
they haven't lived yet. So how
do how do we deal with that in

409
00:29:26.440 --> 00:29:30.240
these companies that divide between sort of
three level I'm not talking about Eppermansion,

410
00:29:30.319 --> 00:29:34.240
I'm even talking about two or three
levels up in those new hires, and

411
00:29:34.279 --> 00:29:40.279
that that that ability to translate what
one set knows versus what the other doesn't.

412
00:29:40.839 --> 00:29:44.759
Yeah, well, I tell you. One of the lessons of life

413
00:29:44.759 --> 00:29:48.720
that I've taken away is you've got
to realize that you do have to reset

414
00:29:48.039 --> 00:29:52.799
and recognize there are a lot of
things that you don't know, and and

415
00:29:52.039 --> 00:29:56.359
embrace the change. I suppose you
know. I try to do it on

416
00:29:56.400 --> 00:29:59.319
these shows by talking to smart people. Of course, you are in the

417
00:29:59.400 --> 00:30:03.759
universities base, so you're talking to
the next generation of business executives. I

418
00:30:03.799 --> 00:30:07.599
mean, that's who these folks are. If you take a master's in analytics,

419
00:30:07.799 --> 00:30:10.440
you know you're going to go somewhere
in business. I'm quite sure you're

420
00:30:10.440 --> 00:30:12.960
going to have a pretty good career
path ahead of you, and it's going

421
00:30:14.000 --> 00:30:18.640
to be constantly changing. So you
know, understanding the data understanding. As

422
00:30:18.640 --> 00:30:22.400
you've said, the use case is
a base camp of use cases, and

423
00:30:22.400 --> 00:30:26.079
that's of course what you're providing with
fourteen eighty six labs, right, Yeah,

424
00:30:26.160 --> 00:30:29.920
so that's the idea. You know, we're very fortunate we have a

425
00:30:29.920 --> 00:30:37.400
company named Liaison based in Boston,
an ed tech company really really focused on

426
00:30:37.519 --> 00:30:44.559
helping the student journey and become successful. That solved the value in O thoughts

427
00:30:44.599 --> 00:30:47.440
and they purchased A thought about two
and a half years ago. I still

428
00:30:47.480 --> 00:30:51.559
remain as a president of the AI
division, really looking at the next generation

429
00:30:51.640 --> 00:30:56.319
of products, and so you know, with that, I also have a

430
00:30:56.359 --> 00:31:02.000
lot of flexibility to be creative.
And so what I wanted to do in

431
00:31:02.079 --> 00:31:07.599
combining sort of what Liaison does in
creativity around the data and analytics side and

432
00:31:07.359 --> 00:31:15.480
the resources, especially the young sort
of very energetic and very knowledge thirsty students,

433
00:31:15.960 --> 00:31:21.240
you know, put together an entrepreneurial
lab where we will develop companies at

434
00:31:21.240 --> 00:31:26.759
the intersection of data and analytics.
And that's you know, fourteen eighty six

435
00:31:26.839 --> 00:31:32.359
Labs is all about that. So
we have twenty one interns right now and

436
00:31:32.519 --> 00:31:36.920
they're building our first company, which
is called Blue Street Data. And Blue

437
00:31:36.920 --> 00:31:41.640
Street Data is aimed directly at how
do we make it easier for the buyers

438
00:31:41.640 --> 00:31:45.640
of data to understand what they're buying
and what they need to buy in order

439
00:31:45.680 --> 00:31:49.920
to make these use cases work?
What powers those use cases? How do

440
00:31:49.960 --> 00:31:55.079
we understand it and how do we
get the right best, highest quality,

441
00:31:55.240 --> 00:32:00.000
lowest price data to power those those
use cases. Yeah, it's really important.

442
00:32:00.079 --> 00:32:05.079
I can tell you that in my
industry, I've tried a couple times

443
00:32:05.119 --> 00:32:07.480
over the years, and I was
nervous each time. To buy data sets

444
00:32:07.519 --> 00:32:13.279
of contact information. It's a very
dicey thing to do. You typically get

445
00:32:13.359 --> 00:32:15.359
very low quality data and it's just
hard to know. I mean, you

446
00:32:15.400 --> 00:32:17.720
get all these different companies offering it, it's like, well, how did

447
00:32:17.759 --> 00:32:21.559
you get that data? Where did
you find it? These days, there

448
00:32:21.599 --> 00:32:25.680
are technologies like seamless AI that we
use somewhat extensively, which are quite compelling.

449
00:32:25.720 --> 00:32:30.160
It's almost like an engine that spins
up a real time Yellow Pages for

450
00:32:30.319 --> 00:32:34.720
you to find contact information. And
it's real time. It's not just a

451
00:32:34.759 --> 00:32:37.759
repository like a big data set.
It's a set of algorithms that in the

452
00:32:37.839 --> 00:32:42.880
moment will reach out and capture information
and they do persist a lot of it.

453
00:32:43.519 --> 00:32:45.000
But you have to ask yourself the
question and like how good is this

454
00:32:45.119 --> 00:32:49.920
data? And that's the service you're
providing, is being able to act as

455
00:32:49.960 --> 00:32:53.279
a liaison or a shepherd or a
schirper or something to help people understand what

456
00:32:53.279 --> 00:32:57.200
are the use cases? When do
you use this stuff? For? What

457
00:32:57.279 --> 00:33:00.319
purpose do you use this stuff?
How do you use it? You have

458
00:33:00.359 --> 00:33:01.440
to be careful about how you use
these things. But I can tell you

459
00:33:01.480 --> 00:33:07.519
buying email database is buying email lists
very bad idea like, you cannot just

460
00:33:07.559 --> 00:33:09.359
throw that stuff into production. I
mean, if anything, you can use

461
00:33:09.400 --> 00:33:14.160
it as a starting point to start
looking around and finding people. But there's

462
00:33:14.200 --> 00:33:15.480
a lot of work to be done, folks. There's a lot of time

463
00:33:15.480 --> 00:33:17.920
and effort that goes into this,
and you have to be careful about how

464
00:33:19.000 --> 00:33:20.960
you spend your time. But don't
touch that. That will be right back.

465
00:33:20.960 --> 00:33:31.200
You're listening to Inside Analysis. Welcome
back to Inside Analysis. Here's your

466
00:33:31.279 --> 00:33:37.880
host, Eric Tabanac. All right, folks, back here on Inside Analysis

467
00:33:37.880 --> 00:33:43.079
with Andy Hannah of fourteen eighty six
Labs and oh Thought and the University of

468
00:33:43.079 --> 00:33:45.759
Pittsburgh and of course Blue Street Data
and Andy you were just staying on the

469
00:33:45.759 --> 00:33:51.680
break there that you are now looking
at data quality through some new lenses.

470
00:33:52.039 --> 00:33:53.640
That's one thing that machine learning is
very good at, by the way,

471
00:33:53.799 --> 00:34:00.200
is finding mistakes. You know,
people are looking for use cases for Lllmsing

472
00:34:00.279 --> 00:34:02.079
mistakes in your code is really good
stuff. You can just throw a bunch

473
00:34:02.079 --> 00:34:05.920
of goats, say where's the mistake, It'll find it. So there are

474
00:34:05.960 --> 00:34:08.800
lots of things these engines do that
are not just text or image generative in

475
00:34:08.960 --> 00:34:12.760
nature. That's for fider for another
show, we'll do a show in a

476
00:34:12.800 --> 00:34:16.480
couple of weeks on that, but
let's drive really dive into the quality conversation

477
00:34:16.599 --> 00:34:22.000
because quality is so important with data. Think about when you get an outreach

478
00:34:22.039 --> 00:34:25.159
and they misspell your name or something
that's not a good customer experience. What

479
00:34:25.199 --> 00:34:29.960
are you doing to improve quality and
to vet the quality of data sets.

480
00:34:30.239 --> 00:34:34.360
Yeah, so I think this is
a really interesting topic because when we first

481
00:34:34.360 --> 00:34:37.280
started to attack this angle, we
went to all the typical sort of metrics

482
00:34:37.280 --> 00:34:43.199
around quality, the consistency, the
comparability, the time and liss et cetera,

483
00:34:43.239 --> 00:34:46.000
everything that everybody thinks about. And
as we were exploring it, we

484
00:34:46.119 --> 00:34:51.679
came to realize that, hey,
so much of the quality of data is

485
00:34:51.800 --> 00:34:57.639
dependent on the particular use case and
how important that use case is to the

486
00:34:57.719 --> 00:35:01.320
business. So it becomes very individualized. It becomes very difficult. Like you

487
00:35:01.360 --> 00:35:06.599
said, Eric, it's there's great
technologies out there to do anomally detection,

488
00:35:06.880 --> 00:35:10.039
right, or to define missing values
or whatever it may be, but you

489
00:35:10.360 --> 00:35:15.320
often can't use those techniques unless you
get the full data set. You can

490
00:35:15.320 --> 00:35:19.880
do it on a sample, but
as we know, sample rarely represents the

491
00:35:19.920 --> 00:35:23.679
full data set. So you know, one of our advisors is Malcolm Hawker,

492
00:35:23.719 --> 00:35:28.039
who you know well, and we
started thinking about how do we need

493
00:35:28.079 --> 00:35:31.679
to approach quality And this takes me
back to sort of some of my days

494
00:35:31.719 --> 00:35:37.000
in a company called Plextronics, where
you look at the supplier, right,

495
00:35:37.079 --> 00:35:44.519
and you start to evaluate what is
the process that the supplier has in order

496
00:35:44.559 --> 00:35:50.840
to ensure that it's delivering a top
notch product, right, So it is

497
00:35:50.920 --> 00:35:54.079
the processes behind that. How many
sources do you use before you say,

498
00:35:54.400 --> 00:35:59.840
hey, I have a valid sleeve
of data related to a particular attribute.

499
00:36:00.280 --> 00:36:04.719
So it is that as a transparent
do they tell us how they do it,

500
00:36:05.159 --> 00:36:07.960
what's the source of the data,
is an ethically sourced data, what's

501
00:36:08.000 --> 00:36:14.800
the quality of the company itself?
And then of course we look at the

502
00:36:14.840 --> 00:36:19.800
expert opinion about the database or of
the data set or the data product,

503
00:36:19.840 --> 00:36:24.039
depending of people who've actually used it
before, the pros and the cons.

504
00:36:24.480 --> 00:36:30.039
So when you put that sort of
how we make the product together with expert

505
00:36:30.159 --> 00:36:35.760
opinion, you have a much different
view of quality than we typically look at

506
00:36:36.000 --> 00:36:42.199
when we're trying to find these anomalies
or problems with the data sets. So

507
00:36:42.239 --> 00:36:45.960
it's like, to a certain extent, it's an expert, crowd sourced view

508
00:36:46.119 --> 00:36:52.360
of quality because it's not just random
let's say Yelp reviewers, and many of

509
00:36:52.400 --> 00:36:55.719
those can be fake. No,
you have trusted sources that you rely upon

510
00:36:55.920 --> 00:37:02.079
to give you sophisticated analysis of data
sets. Right that coupled with are they

511
00:37:02.119 --> 00:37:07.159
transparent in the process that they use
to make the product? If you know,

512
00:37:07.199 --> 00:37:10.320
if you think about the material science
world, you know, you have

513
00:37:10.559 --> 00:37:15.320
your the spec of the product has
to be on and you go in and

514
00:37:15.519 --> 00:37:21.519
you you you explore the manufacturing process
of the manufacturer. That's what I say

515
00:37:21.559 --> 00:37:25.760
is all about. We need that
ISO type of concept here with data production.

516
00:37:27.320 --> 00:37:32.360
It's just like thinking about a product
that any company might produce. Well,

517
00:37:32.360 --> 00:37:38.440
that's interesting because you're you're kind of
hinting at or alluding to observability rights.

518
00:37:39.000 --> 00:37:43.280
When we talk about observability in the
data world, what you're really talking

519
00:37:43.280 --> 00:37:47.639
about is getting windows into streams of
information and then being able to correlate those

520
00:37:47.679 --> 00:37:52.760
to understand what happened. So this
is what I love about open source.

521
00:37:52.800 --> 00:37:55.440
That's why I always tout open source
and I'm a huge fan for lots of

522
00:37:55.440 --> 00:38:00.400
different reasons. One is because many
eyes make few the errors or as the

523
00:38:00.440 --> 00:38:05.079
old expression goes bad, code goes
away. Right. That's one of the

524
00:38:05.079 --> 00:38:07.280
benefits of open source. Now there
are downsides, which is it's hard to

525
00:38:07.320 --> 00:38:10.519
make money. Like if you're open
source in the code, how do you

526
00:38:10.639 --> 00:38:15.000
ensure that you get the money.
We were joking a couple of years ago

527
00:38:15.119 --> 00:38:20.199
that Amazon was strip mining the open
source community and not doing a lot for

528
00:38:20.280 --> 00:38:23.199
it. They have since changed their
tune somewhat and now they kind of understand,

529
00:38:23.199 --> 00:38:28.000
which gets us back to ethics,
right and too being ethical and how

530
00:38:28.039 --> 00:38:30.159
you run your business, how you
run your operations, and you know,

531
00:38:30.400 --> 00:38:36.639
there are these sort of competing virtuous
and vicious cycles, it seems to me,

532
00:38:37.039 --> 00:38:42.000
and we want to put the balance
of power in the virtuous circles where

533
00:38:42.679 --> 00:38:45.480
good virtue breeds good virtue, which
breeds good virtue and it keeps going up.

534
00:38:45.639 --> 00:38:50.039
But there is always this downward pressure
of you know, the vicious cycles.

535
00:38:50.079 --> 00:38:52.559
And how do you solve that.
I think you solve it through good

536
00:38:52.679 --> 00:38:59.239
ethical constructs and transparency and just ongoing
work. Basically it's never going to be

537
00:38:59.559 --> 00:39:01.880
solved. To keep hammering away at
it, right, Yeah, and the

538
00:39:02.320 --> 00:39:07.920
transparency, as we've talked about,
I think every single segment here is critically

539
00:39:07.920 --> 00:39:10.639
imparent. It's transparency and what the
data is, how you're using it,

540
00:39:10.719 --> 00:39:15.880
how it's produced. But you said
something that's kind of interesting that I think

541
00:39:15.880 --> 00:39:19.239
a lot about, which is you
said use the words sleeves of data.

542
00:39:20.280 --> 00:39:22.880
And the more that I talk,
you know, I probably talked to a

543
00:39:22.920 --> 00:39:28.639
dozen people every week in this in
the industry about external data, and one

544
00:39:28.679 --> 00:39:32.119
of the most common things are that
I have to typically as a buyer,

545
00:39:32.360 --> 00:39:37.639
buy bundles of data. I want
streams of data. I want particular data

546
00:39:37.679 --> 00:39:40.480
elements. Yeah, I have.
I'll have to find a way to make

547
00:39:40.480 --> 00:39:45.320
sure my matching algorithms are able to
pen my existing database. I get that,

548
00:39:45.880 --> 00:39:50.840
But why do I have to buy
the whole anchilada when I just want

549
00:39:51.000 --> 00:39:54.079
you know, the you know,
the uh, you know, the guacamole

550
00:39:54.239 --> 00:39:57.719
or whatever. It might be.
What a bad analogy, but you get

551
00:39:57.719 --> 00:40:00.880
what I mean is it's like,
so you want that, you want the

552
00:40:00.880 --> 00:40:07.039
ability to slice down into the individually, into the granular level, and we

553
00:40:07.079 --> 00:40:10.679
really don't have that. That's not
the way that data and information has been

554
00:40:10.719 --> 00:40:15.519
sold in the pass, and so
we have to move to that to that

555
00:40:15.599 --> 00:40:21.199
type of granularity, to the individual
level data element. And that's going to

556
00:40:21.280 --> 00:40:23.119
be hard. That's going to be
hard for a lot of the existing companies

557
00:40:23.159 --> 00:40:28.760
who are doing it the old way
to transform. But if we do it

558
00:40:28.840 --> 00:40:30.480
in the right way, we're going
to be able to say, oh,

559
00:40:30.599 --> 00:40:35.360
for this use case, here's a
particular element you need to buy. You

560
00:40:35.360 --> 00:40:39.239
can buy the best data and it's
a streame off of this company that's never

561
00:40:39.320 --> 00:40:45.119
sold data in the market before.
Because so we give that opportunity for those

562
00:40:45.159 --> 00:40:49.519
types of companies to come to market
with their data with the particular use that's

563
00:40:49.559 --> 00:40:54.440
better than anything else that's out there. That's the way to think about restructuring

564
00:40:54.480 --> 00:40:59.559
of the data markets, which we
desperately need if we're going to move to

565
00:40:59.639 --> 00:41:05.199
next level of value from data.
That's very interesting, and I was mentioning

566
00:41:05.239 --> 00:41:09.960
before you're seeing early indicators for this
kind of technique or of standard, if

567
00:41:10.000 --> 00:41:14.440
you will, and one of which
is like Google with your timeline, and

568
00:41:14.480 --> 00:41:16.440
how Google will now give you your
timeline, right. I mean, I

569
00:41:16.440 --> 00:41:21.239
think Twitter allows you to download your
tweets. Other platforms allow you to kind

570
00:41:21.239 --> 00:41:23.880
of download things, and there is
value to that, but being able to

571
00:41:24.000 --> 00:41:28.360
reflect back to these Amazon is actually
pretty good at this at being able to

572
00:41:28.360 --> 00:41:30.960
capture all of your transactions very quickly, show you where they are so you

573
00:41:31.000 --> 00:41:34.239
can go find things. So you
think about all this stuff used to have

574
00:41:34.239 --> 00:41:36.840
to keep track of yourself, and
now you realize, oh, okay,

575
00:41:36.840 --> 00:41:38.840
well they're going to keep track of
that for me, like a web portal

576
00:41:38.920 --> 00:41:42.960
for the hospital, for example,
I don't have to write down everything for

577
00:41:43.119 --> 00:41:45.440
my wife's appointments. I know it's
up in the cloud, so I can

578
00:41:45.480 --> 00:41:50.079
go there. And I think the
more we learn to trust those sources and

579
00:41:50.119 --> 00:41:52.679
rely on those sources, the more
we'll be able to focus on what we're

580
00:41:52.719 --> 00:41:55.639
trying to do and not worry about
trying to do all the other stuff that

581
00:41:55.679 --> 00:42:00.119
other people are doing for us.
Right, Yeah, and it gets down

582
00:42:00.559 --> 00:42:07.559
you're right. So the way that
we capture information has dramatically improved over the

583
00:42:07.559 --> 00:42:09.679
past decade. The way that we
store it so that we can track it

584
00:42:09.760 --> 00:42:14.800
down to any individual is what gives
us the power to be able to do

585
00:42:14.840 --> 00:42:17.480
this. It's going to open up
a whole bunch of different revenue models.

586
00:42:17.519 --> 00:42:21.760
It's going to open up a whole
bunch of different companies. I mean,

587
00:42:21.800 --> 00:42:27.159
we're really flooded with Jenai companies right
now. But I think the underlying sort

588
00:42:27.199 --> 00:42:32.039
of hidden gold is what companies are
doing very creative things with the data to

589
00:42:32.079 --> 00:42:37.960
make it much more usable. And
when you can combine that value of that

590
00:42:37.119 --> 00:42:43.519
usable data with these technologies, just
like we've seen over the past forty years,

591
00:42:44.400 --> 00:42:49.840
you're going to companies will merge that
will disrupt their industries, will disrupt

592
00:42:49.880 --> 00:42:53.159
the other companies that they compete against. Yeah, no, that's right,

593
00:42:53.239 --> 00:42:59.440
and I think it's going to be
amazing. I think that all the tumblers

594
00:42:59.440 --> 00:43:04.519
are align right now and it's kind
of wild, you know. I think

595
00:43:04.599 --> 00:43:09.159
the more people get access to useful
data sets, like a data sleeve,

596
00:43:09.199 --> 00:43:13.079
as you're talking about, just a
little piece to fill in a gap,

597
00:43:13.760 --> 00:43:15.880
the more success you have with that, the better. The problem is that

598
00:43:15.920 --> 00:43:20.519
people do get burned. I mean, one of my partners does content syndication,

599
00:43:21.119 --> 00:43:23.039
and I trust this guy as much
as I trust any human being in

600
00:43:23.079 --> 00:43:25.840
the world. He's incredibly professional,
he's been around a long long time,

601
00:43:27.280 --> 00:43:30.239
and the projects they've done with this
have been spectacular. I mean, the

602
00:43:30.320 --> 00:43:32.039
quality of the leads is amazing,
but it's still hard to sell. Why

603
00:43:32.119 --> 00:43:36.679
because people have been burned in the
past and once you get burned, you

604
00:43:36.679 --> 00:43:38.719
know, once bitten, twice shy. Basically, it's hard to kind of

605
00:43:38.760 --> 00:43:43.480
move past that. I mean with
chatbots, you're kind of seeing this right

606
00:43:43.519 --> 00:43:47.480
now. Chatbots are coming back with
some vigor. They first came out,

607
00:43:47.519 --> 00:43:51.159
I mean it first came out like
fifteen years ago. In the earliest days,

608
00:43:51.159 --> 00:43:53.400
they were just dreadful, right,
it was just horrible. It's like,

609
00:43:53.440 --> 00:43:57.119
what is going on here? These
people are making me crazy. So

610
00:43:57.679 --> 00:44:00.639
it's like there, you know,
well, it's like Pittsburgh still to this

611
00:44:00.760 --> 00:44:01.760
day to people who have never been
here, like, oh, it's not

612
00:44:01.840 --> 00:44:05.679
a dirty city. I'm like,
oh god, no, it's not a

613
00:44:05.719 --> 00:44:07.960
dirty city. It hasn't been thirty
for It was from the steel days and

614
00:44:08.000 --> 00:44:10.920
that ended, you know, in
terms of the toxins and all that stuff.

615
00:44:12.000 --> 00:44:15.079
That's this distant past now, but
it takes a long time for these

616
00:44:15.079 --> 00:44:17.360
things to die. Well. Podcast
bonus segments coming up next. Bot.

617
00:44:17.400 --> 00:44:24.159
Don't shut that out. You're listening
to Inside Analysis, all right, folks,

618
00:44:24.159 --> 00:44:28.800
back your time with the podcast bonus
segment here with Andy Hannah of fourteen

619
00:44:28.840 --> 00:44:31.000
eighty six Labs and oh Thought and
the University of Pittsburgh, and we're talking

620
00:44:31.000 --> 00:44:37.840
all about AI and data and ethics
and speaking of everyone's talking about responsible AI

621
00:44:37.960 --> 00:44:42.239
and ethical AI. And you know
it's good because we need to have these

622
00:44:42.239 --> 00:44:45.639
conversations. And you, my friend, were recently given the chair of what

623
00:44:45.760 --> 00:44:50.039
is it the University of Pittsburgh's Responsible
Data science sport? Is that right?

624
00:44:50.079 --> 00:44:54.320
Tell us about that? That is
right? So we're really fortunate at the

625
00:44:54.400 --> 00:45:00.199
University of Pittsburgh where we have a
lot of individuals that are focused on data

626
00:45:00.239 --> 00:45:07.800
science and artificial intelligence. What we're
focused on in a responsible data science community

627
00:45:08.480 --> 00:45:14.280
is three things. How do we
make sure that the infrastructure behind all these

628
00:45:14.360 --> 00:45:20.320
models, etc. Are responsible From
a curriculum perspective, from a research perspective,

629
00:45:20.400 --> 00:45:24.960
from a workforce development perspective. So
we want to lead the nation on

630
00:45:25.119 --> 00:45:30.199
how we think about the deployment of
all these technologies, regardless of whether you're

631
00:45:30.199 --> 00:45:37.360
in academia or you're in business.
My responsibility as chair of the advisory board

632
00:45:37.639 --> 00:45:44.519
is to bring industry into the conversation. So that very very thoughtful approach that

633
00:45:44.599 --> 00:45:47.119
says that we can't do this in
a vacuum. So if we're working on

634
00:45:47.239 --> 00:45:51.960
and we literally use the words we're
working on use cases and data sets,

635
00:45:52.039 --> 00:45:58.000
to show how you can responsibly use
the techniques that we've been talking about and

636
00:45:58.639 --> 00:46:02.360
solve real problems. And so in
those real problems come from industry. So

637
00:46:02.400 --> 00:46:13.199
we're focused on retail, finance,
and healthcare and what i'll call civic applications

638
00:46:13.960 --> 00:46:20.639
of this technology. That's pretty cool, you know when I think about policies,

639
00:46:20.840 --> 00:46:23.039
and I'm glad that you're saying you're
bringing the business world into this,

640
00:46:23.159 --> 00:46:27.480
because that's really what it takes.
Because you know, you and I can

641
00:46:27.519 --> 00:46:30.880
have all sorts of theories about how
things operate in a particular industry, but

642
00:46:30.960 --> 00:46:34.639
until you sit down with people from
that business, you're probably not going to

643
00:46:34.679 --> 00:46:37.840
be able to figure out some of
the really critical components of processes and workflows.

644
00:46:37.880 --> 00:46:40.639
And they will tell you. I
mean, I remember, just as

645
00:46:40.679 --> 00:46:44.920
an example, many years ago,
I was on as at a meeting.

646
00:46:44.960 --> 00:46:47.400
I used to represent the Downtown Development
District of New Orleans, and we're in

647
00:46:47.400 --> 00:46:52.960
a meeting with the new executive director
and a whole bunch of other people's stakeholders

648
00:46:52.000 --> 00:46:55.280
are there, including a lieutenant from
the police force and a bunch of other

649
00:46:55.320 --> 00:46:59.360
people, and the new executive director
was like, yeah, we decided we're

650
00:46:59.360 --> 00:47:01.239
going to come up with ourn police
force and they're going to handle these stuff

651
00:47:01.239 --> 00:47:06.039
downtown. And you can imagine the
lieutenant of the police force in New Orleans

652
00:47:06.079 --> 00:47:08.480
said, excuse me, what are
you going to do? That's our responsibility.

653
00:47:08.480 --> 00:47:12.960
Why what are you talking about here? They've never even broached the subject

654
00:47:12.960 --> 00:47:15.159
with them, and it's like,
that's kind of a problem. You know.

655
00:47:15.199 --> 00:47:19.480
You if you sit down with people
in private first and have conversations or

656
00:47:19.519 --> 00:47:23.280
even have hearings, for example,
to vet these issues, that's the way

657
00:47:23.400 --> 00:47:28.239
to get there and make sure you
have the stakeholders at the table, because

658
00:47:28.239 --> 00:47:30.360
they're the ones who are going to
recognize those red flags. It sounded like

659
00:47:30.400 --> 00:47:34.480
a great idea in the other boardroom, but when you share it in the

660
00:47:34.800 --> 00:47:37.360
bigger context, you realize, oh, I guess there are some problems with

661
00:47:37.400 --> 00:47:40.840
that. That's the point of having
these meetings, right, yeah, one

662
00:47:40.880 --> 00:47:45.280
hundred percent. And you know,
we have to look at it from every

663
00:47:45.400 --> 00:47:51.679
angle in terms of responsibility, and
so the lens when I think about that

664
00:47:51.840 --> 00:47:55.719
concept are responsible. I really think
about it from three different lenses. The

665
00:47:57.119 --> 00:48:00.960
first one is purpose. You know, why are we using this technology?

666
00:48:01.079 --> 00:48:05.480
What is the big reason why we're
using this technology? Is it to solve

667
00:48:05.559 --> 00:48:12.400
organizational societal issues or is it to
cause some type of catastrophic harm? Right,

668
00:48:12.480 --> 00:48:15.199
I mean, that's the easiest way
to think about it. So purpose

669
00:48:15.320 --> 00:48:20.800
is number one, So how that
that that perspective. The second one is

670
00:48:20.920 --> 00:48:27.480
all about ethics, So it's what
is what is the intent of the individuals

671
00:48:27.519 --> 00:48:35.639
involved in the project? Are they
looking to benefit the society or benefit the

672
00:48:35.639 --> 00:48:37.960
company, or benefit or everything that
they're doing? Is it? Is it

673
00:48:38.079 --> 00:48:43.639
leaning towards the good as opposed to
the bad? And then the last piece

674
00:48:44.000 --> 00:48:51.559
is about you avoiding bias or having
anyone treated unfairly. And that's a really

675
00:48:51.679 --> 00:48:54.639
interesting piece of it because that's where
we go back all the way to the

676
00:48:54.639 --> 00:49:00.119
beginning of our conversation on a diagnosis. Where are people being this advantage because

677
00:49:00.159 --> 00:49:06.280
of this technology? And so if
we can combine that purpose and ethics and

678
00:49:06.760 --> 00:49:13.119
fairness together, we're going to have
a very powerful and reliable and trustworthy set

679
00:49:13.159 --> 00:49:16.840
of technologies and data that we can
then build the next generations of business on.

680
00:49:17.800 --> 00:49:22.639
Yeah, you just reminded me of
emanual cons categorical imperative. You may

681
00:49:22.760 --> 00:49:27.880
you may recall act only on that
maxim which you can at the same time

682
00:49:27.960 --> 00:49:32.159
will as universal law, which is
a nuanced approach of do as you would

683
00:49:32.199 --> 00:49:36.559
be done by. Basically, it's
like, if you were going to engage

684
00:49:36.559 --> 00:49:39.719
in a policy, ask yourself,
would it be reasonable for every other company

685
00:49:39.760 --> 00:49:43.360
to engage in the same policy?
And if the answer is yes, then

686
00:49:43.920 --> 00:49:46.840
ethically speaking, you're on pretty solid
ground, right. And first, you

687
00:49:46.880 --> 00:49:52.199
know, I like the physician's oath, right, So first we do no

688
00:49:52.320 --> 00:49:55.760
harm, right, So if you
literally have that in your mindset, and

689
00:49:55.840 --> 00:50:01.119
so I think that's you know,
the diversity of people around the problem is

690
00:50:01.159 --> 00:50:05.639
a really critical one. So if
you have a bunch of people that are

691
00:50:06.039 --> 00:50:08.760
the same gender, same ethnicity,
you know, the same level of wealth,

692
00:50:08.840 --> 00:50:12.960
whatever it is, trying to solve
a problem, you're not going to

693
00:50:12.960 --> 00:50:16.400
get a diverse view of the problem. So the people around the problem and

694
00:50:16.519 --> 00:50:22.440
the techniques that we use to determine
whether or not somebody could be treated unfairly,

695
00:50:22.920 --> 00:50:27.960
the combination of those two things very
powerful m HM. And you have

696
00:50:28.039 --> 00:50:31.440
to kind of maintain an ongoing balancing
act, right, because you can also

697
00:50:31.599 --> 00:50:36.639
overreact to things. And that's something
we see a lot in our world is

698
00:50:36.679 --> 00:50:38.039
pendulum swings one way and then it
swings the other. Way, and it

699
00:50:38.079 --> 00:50:40.719
swings one way and it swings the
other way. You know, for every

700
00:50:40.760 --> 00:50:45.480
action there's an equal and opposite reaction, So you kind of have to take

701
00:50:45.519 --> 00:50:49.960
all that into consideration and just be
reasonable. Right. There's something to be

702
00:50:50.039 --> 00:50:53.519
said for being a reasonable person,
for understanding there are differences, and you

703
00:50:53.519 --> 00:50:57.039
know, not everyone's out to get
you and all these kinds of things.

704
00:50:57.079 --> 00:50:59.559
You do have to keep that all
in mind. But I think the real

705
00:50:59.639 --> 00:51:04.239
key is, as you suggest,
bring an ethically oriented mindset, what are

706
00:51:04.280 --> 00:51:07.360
we trying to do, how are
we trying to do it, and then

707
00:51:07.519 --> 00:51:10.239
honestly gauging the success. Is it
working. It is nothing like the law

708
00:51:10.239 --> 00:51:15.039
of unattended consequences to give you a
Rohm Dinger, or do you think you're

709
00:51:15.079 --> 00:51:16.760
doing the right thing and it all
falls apart. I mean, have you

710
00:51:16.760 --> 00:51:20.599
ever seen the video of Yellowstone when
they let the wolves back in. Have

711
00:51:20.639 --> 00:51:22.920
you seen that video that all of
them? Look it up, man,

712
00:51:23.000 --> 00:51:28.760
it'll absolutely blow your mind. They
did a study and they brought wolves back

713
00:51:28.800 --> 00:51:34.480
to Yellowstone National Park and they said
the results were absolutely mind blowing. You

714
00:51:34.480 --> 00:51:37.039
have to get into it to understand, but basically it reset the balance of

715
00:51:37.119 --> 00:51:40.679
power amongst all the different animals that
were there, the creeks came back,

716
00:51:40.719 --> 00:51:45.159
All sorts of things happened that absolutely
blew people's minds, which shows you it's

717
00:51:45.199 --> 00:51:47.679
kind of like the butterfly's wings effect. Right, one little change can have

718
00:51:47.760 --> 00:51:52.159
a very big impact, But you
have to be open to the change,

719
00:51:52.159 --> 00:51:54.679
and you have to be open to
tracking the change and just being realistic.

720
00:51:54.840 --> 00:51:58.360
And that's what business is all about. Right in the business, you have

721
00:51:58.440 --> 00:52:02.000
to be pragmatic. If you go
too far in one direction, you're not

722
00:52:02.039 --> 00:52:06.159
going to be in business anymore.
But final thoughts, how do people find

723
00:52:06.199 --> 00:52:09.280
out more about Blue Street Data and
fourteen eighty six Labs. Yeah, so

724
00:52:09.679 --> 00:52:15.159
the easy go to www dot Blue
Street Data dot com or www dot fourteen

725
00:52:15.159 --> 00:52:19.920
eighty six Labs dot com. You
can learn about us from fourteen eighty six,

726
00:52:20.000 --> 00:52:23.440
how we're building next generation companies with
really great students, and of course

727
00:52:23.440 --> 00:52:30.280
Blue Street Data, how we facilitate
the purchase of high quality, right price

728
00:52:30.960 --> 00:52:36.199
data that can drive analytic engines.
I love it, folks talking next time

729
00:52:36.199 --> 00:52:40.760
you've been listening to Inside Analysis.
Former President Trump is defending his comments that

730
00:52:40.800 --> 00:52:45.880
he would not defend NATO members that
don't meet defense spending targets. When I

731
00:52:45.960 --> 00:52:49.320
said that, and when I made
that statement, I want to energize them

732
00:52:49.400 --> 00:52:52.239
to pay. Daring an interview with
Fox News, Trump said NATO countries pay

733
00:52:52.280 --> 00:52:55.239
far less than what the US pay
is into the alliance, and some take

734
00:52:55.280 --> 00:53:00.639
advantage from faced backlash after saying he
would tell Russia to quote do whatever the

735
00:53:00.679 --> 00:53:05.000
hell they want to countries who don't
meet spending goals. The NATO Secretary General

736
00:53:05.079 --> 00:53:08.840
recently criticized the re marks, arguing
that any suggestion that allies will not defend

737
00:53:08.880 --> 00:53:15.519
each other undermines security for all allies. National Security Council spokesman John Kirby says

738
00:53:15.519 --> 00:53:19.480
it's up to the Israeli people to
decide if they'll have new elections and leadership.

739
00:53:19.840 --> 00:53:22.559
He is the Prime Minister of Visual
We respect the sovereignty of the Israeli

740
00:53:22.639 --> 00:53:28.159
people. Speaking on Fox News Sunday, Kirby was referring to Israeli Prime Minister

741
00:53:28.199 --> 00:53:32.079
Benjamin Nett Yahoo following Senate majority of
Leader Chuck Schumer's speech calling for new elections

742
00:53:32.079 --> 00:53:37.559
in Israel. Schumer also criticized net
Yahou's approach to the war in Gaza.

743
00:53:37.800 --> 00:53:42.480
Herby said President Biden recognized as Schumer
was speaking for many Americans concerned about how

744
00:53:42.519 --> 00:53:45.320
the war is going. He added
that Biden and net Yah, who have

745
00:53:45.400 --> 00:53:47.480
known each other for forty years,
and while they don't always agree, they

746
00:53:47.519 --> 00:53:52.280
do talk. Five states will hold
their presidential primaries on Tuesday. Scott Carr

747
00:53:52.360 --> 00:53:58.079
has more candidates. Donald Trump and
President Joe Biden both already have enough delegates

748
00:53:58.119 --> 00:54:01.079
to clinch their party's respective nomine nations, but will be looking to increase their

749
00:54:01.159 --> 00:54:07.039
numbers of delegates in Kansas, Arizona, and Florida. Ohio and Illinois will

750
00:54:07.079 --> 00:54:10.920
also be holding their state and congressional
primaries. Reports say President Biden may also

751
00:54:10.960 --> 00:54:16.800
be competing with some uncommitted voters who
may cast protest votes against US military aid

752
00:54:16.880 --> 00:54:22.440
to Israel. I'm Scott Carr in
Washington. Today is Selection Sunday for the

753
00:54:22.519 --> 00:54:27.039
NCAA Men's Basketball Tournament. The field
of sixty eight will be revealed by the

754
00:54:27.039 --> 00:54:30.440
Selection Committee at six pm Eastern on
CBS. Thirty two of the spots will

755
00:54:30.480 --> 00:54:36.320
be filled by teams and automatically qualified
by winning their conference tournaments. I'm Tammy

756
00:54:36.320 --> 00:54:40.599
trheo when you shop with local merchants, more of your money stakes Closest to

757
00:54:40.639 --> 00:54:45.039
home by helping support local businesses.
We all benefit this message from my tran

758
00:54:45.159 --> 00:54:49.639
e Shop, family owned and operated
for over thirty years. Save and bring

759
00:54:49.679 --> 00:54:52.840
your foreign or domestic vehicle to my
tran e Shop located at two tenty eight

760
00:54:52.960 --> 00:54:57.760
nine to one Alessandro Boulevard and Marino
Valley for called nine five to one,

761
00:54:57.840 --> 00:55:02.159
two four three seven six seventy five. Visit online at my transhop dot com.

762
00:55:02.199 --> 00:55:08.559
That's my tra and keshop dot com
that's my tran e Shop, proud

763
00:55:08.559 --> 00:55:16.119
supporters of our local businesses. The
Tri City Shopping Center in Redlands is serving

764
00:55:16.199 --> 00:55:21.440
up some really cool ice cream at
lemcho Akana. Then get your chocolates and

765
00:55:21.480 --> 00:55:24.519
other delights from Seased Candies. Moms
and future moms who visit the mall can

766
00:55:24.519 --> 00:55:29.639
cool off and relax while they get
treated like royalty at Shiny Nails or Francis

767
00:55:29.719 --> 00:55:34.039
Nails and then pampered at Texture Hair. The Tri City Center is filled with

768
00:55:34.079 --> 00:55:37.920
retailers who care about you. Shop
at the Tri City Center in Redlands and

769
00:55:37.960 --> 00:55:44.599
see why they call it the Mall
with a heart. Redloods. Aviation of

770
00:55:44.639 --> 00:55:49.760
Redloods reminds is the too often our
veterans and their spouses have trouble finding jobs.

771
00:55:50.199 --> 00:55:53.760
If you're an employer, joined the
Alliance to help support our transitioning military

772
00:55:53.800 --> 00:55:59.599
service. Minted women and their families
bring elite skills, agility, admission,

773
00:55:59.599 --> 00:56:06.280
dedicate your organization higher smart ed,
higher vets that patriotic reminders courtesy of Redlands

774
00:56:06.320 --> 00:56:09.320
Aviation of Redlands and Acreage where they
are on the air, supporting those who

775
00:56:09.360 --> 00:56:14.840
serve and the families who remain at
home. Hi, I'm food critic Allen

776
00:56:14.880 --> 00:56:19.119
Morgan, and I'm excited to tell
you about a small local company called Sooring

777
00:56:19.159 --> 00:56:24.559
Swine Acres that creates and sells over
three hundred delicious tasting and unique jams,

778
00:56:24.800 --> 00:56:30.119
jellies, honeys, sauces, pickled
products, and their new Voodo Queen holistic

779
00:56:30.239 --> 00:56:36.039
teas that you can't find anywhere else. You can visit Sooring swine Acres store

780
00:56:36.320 --> 00:56:42.079
and u Kaipa, or you can
purchase their products on their website Sooringswineacres dot

781
00:56:42.079 --> 00:56:45.360
com. That's s O A r
I n G s w i n E

782
00:56:45.480 --> 00:56:51.960
acres ac r s dot com.
Savor such delicious items like rosemary pear jelly,

783
00:56:52.440 --> 00:56:57.039
ginger orange jam, hmm, cherry
chili jam, smoked apple jelly,

784
00:56:57.280 --> 00:57:01.360
plus so many more sensational products.
They make great gifts for the holidays as

785
00:57:01.400 --> 00:57:06.280
well as for birthdays or other special
occasions. Or you can just pay for

786
00:57:06.320 --> 00:57:12.119
yourself and your family's taste buds.
Visit them at sowingswineacres dot com. That's

787
00:57:12.119 --> 00:57:16.599
soaringswineacres dot com. Take it from
me, food critic Almborgan These food products

788
00:57:16.639 --> 00:57:34.320
are delicious. That's soaringswineacres dot com
Happy eating. Was your car involved in

789
00:57:34.360 --> 00:57:37.639
an accident or just need help with
dents? All Magic Paint and Body Collision

790
00:57:37.719 --> 00:57:44.039
Centers in business for over thirty years. They're highly trained staff and certified technicians

791
00:57:44.119 --> 00:57:47.039
and friendly staff are the best in
the business and treat each car as if

792
00:57:47.159 --> 00:57:51.880
it was their own. All Magic
Paint and Body Collision Centers are family owned

793
00:57:51.960 --> 00:57:55.159
and offer state of the art equipment
and tools to ensure optimum results. They

794
00:57:55.239 --> 00:58:00.400
use the latest technology in computerized color
matching and specialize in frame repairs with their

795
00:58:00.440 --> 00:58:06.719
modern laser measuring systems. They're OEM
certified and they have four locations to serve

796
00:58:06.800 --> 00:58:10.039
you. All Magic Paint and Body
Collision Centers offer rental car assistance with free

797
00:58:10.119 --> 00:58:15.599
dropoff and pickup services too, and
their work has a lifetime guarantee. All

798
00:58:15.639 --> 00:58:19.760
Magic Paint and Body Collision Centers are
in Narco, East Vale, Marino Valley

799
00:58:19.840 --> 00:58:22.960
and in Fontana. Call them at
one eight hundred sixty one Magic. That's

800
00:58:22.960 --> 00:58:28.599
one eight hundred sixty one Magic All
Magic Paint and Body Collision Centers, one

801
00:58:28.639 --> 00:58:32.559
eight hundred sixty one Magic All Magic
Paint and aut of Bodies says drive careful.

802
00:58:37.599 --> 00:58:39.599
Hey guys, this is Andrew Caravella. Just wanted to let you know

803
00:58:39.679 --> 00:58:44.360
that CACAA has a new listener line
for all of you that have a phone.

804
00:58:44.519 --> 00:58:46.880
Now you can listen to KCAA in
the grocery store, in the doctor's

805
00:58:46.920 --> 00:58:50.480
office, or when you want to
pretend you're on the phone with someone,

806
00:58:50.519 --> 00:58:53.239
just so that creeper will leave you
alone. Call seven two oh eight three

807
00:58:53.320 --> 00:58:59.639
five three oh ninety nine today and
listen to your favorite KCAA radio shows that

808
00:58:59.719 --> 00:59:05.519
number again seven to two oh eight
three five three oh nine nine KCAA Radio.

809
00:59:05.639 --> 00:59:12.320
The station ley's no caller or listener
behind. Do you have a debt

810
00:59:12.400 --> 00:59:15.400
problem? Are you being sued for
an unpaid debt? Is your paycheck being

811
00:59:15.440 --> 00:59:20.159
garnished or your bank account emptied out. Do you feel like you're running out

812
00:59:20.199 --> 00:59:23.719
of options? The Fullman Firm is
your friendly local law firm next door and

813
00:59:23.960 --> 00:59:28.679
has helped thousands of people just like
you. Give us a call at eight

814
00:59:28.719 --> 00:59:31.400
three to three Fullman and see if
we can help you too. It's a

815
00:59:31.440 --> 00:59:37.559
free confidential consultation. You have nothing
to lose for debts above seven thousand dollars.

816
00:59:37.599 --> 00:59:39.559
Give us a call at eight three
to three Fullmen, or check out

817
00:59:39.679 --> 00:59:45.840
Fullmanfirm dot com. That's eight three
three f U L M A N the

818
00:59:45.880 --> 00:59:50.199
Fullman Firm. Let our lawyers get
you on the path to financial freedom.

819
00:59:50.519 --> 00:59:53.000
The preceding is not intended as a
guarantee or estimate of the outcome of your

820
00:59:53.039 --> 00:59:57.360
case. Every case is unique and
past. Yourselves are not indicators of the

821
00:59:57.360 --> 01:00:00.880
success of your case. In control
of your financial future and callie three three

822
01:00:00.960 --> 01:00:05.199
Foreman or visitforman firm dot com now
h