WEBVTT

1
00:00:00.120 --> 00:00:04.120
Stration that the House can't even vote
on his resolution that condemns Hamas and supports

2
00:00:04.160 --> 00:00:08.640
Israel. The Foreign Affairs Committee chair
added that the House's inability to govern empowers

3
00:00:08.640 --> 00:00:13.439
and Embolden's adversaries to claim that democracy
doesn't work. Nine House Republicans are now

4
00:00:13.439 --> 00:00:18.239
considered likely candidates for Speaker after the
party dropped Jim Jordan is their nominee.

5
00:00:18.239 --> 00:00:24.160
On Friday, I'm Chris Karragio NBC
News Radio. You're on board kcaa's Inland

6
00:00:24.440 --> 00:00:29.480
Express KCAA Home A Linda ten fifty
Am, the station that needs no mess.

7
00:00:29.519 --> 00:00:40.119
Year behind, the information economy has
a ride. The world is teeming

8
00:00:40.200 --> 00:00:46.159
with innovation as new business models reinvent
every industry industry. Inside Analysis is your

9
00:00:46.280 --> 00:00:51.119
source of information and insight about how
to make the most of this exciting new

10
00:00:51.159 --> 00:00:55.880
era. Learn more at inside analysis
dot com, Insideanalysis dot com. And

11
00:00:55.960 --> 00:01:03.840
now here's your host, Eric Kavanaugh. Yes, oh yes, folks,

12
00:01:03.879 --> 00:01:07.680
welcome to the future. Indeed,
your host Eric Cavanaugh here on the only

13
00:01:07.799 --> 00:01:11.599
coast to coast radio show in the
US of Ada. It's all about the

14
00:01:11.640 --> 00:01:15.519
information economy Inside Analysis, and folks, I'm very excited today to have a

15
00:01:15.640 --> 00:01:21.159
very special guest in industry visionary,
someone who has been working in the space

16
00:01:21.280 --> 00:01:23.959
for a number of years and has
always seemed to have his finger on the

17
00:01:23.959 --> 00:01:29.560
pulse of where things are and where
things are going, and that's especially difficult

18
00:01:29.599 --> 00:01:34.519
to do in these days with this
new remarkable power of generative AI. We'll

19
00:01:34.560 --> 00:01:38.079
be talking to Andrew Turner. He's
the general manager of a company called Proxy

20
00:01:38.159 --> 00:01:41.920
Data. They're doing some very interesting
things in that space. And before we

21
00:01:42.000 --> 00:01:45.480
dive in, I wanted to throw
out a bit of my past here as

22
00:01:45.519 --> 00:01:49.519
I think about a metaphor to describe
what we're seeing in the world and what

23
00:01:49.560 --> 00:01:53.640
this GENAI stuff is really doing.
And I'm reminded of the philosopher about whom

24
00:01:53.680 --> 00:01:59.000
I wrote my thesis in college in
philosophy, Jacques dere Da. He wrote

25
00:01:59.000 --> 00:02:02.159
this very dense piece called Structure,
sign and Play in the Discourse of the

26
00:02:02.239 --> 00:02:07.800
Human Sciences, and it was deep
and he starts off with this whole concept.

27
00:02:07.840 --> 00:02:12.120
He says, let us assume that
something has happened like an event,

28
00:02:12.479 --> 00:02:15.919
if that word weren't so loaded,
and he refers to as a sort of

29
00:02:16.759 --> 00:02:22.960
rupture or redoubling. And what he's
kind of talking about is is a movement

30
00:02:23.000 --> 00:02:28.120
to another level of self awareness,
almost like cultural self awareness, and how

31
00:02:28.159 --> 00:02:32.360
that cultural self awareness is now shaping
and sort of reshaping how we interact with

32
00:02:32.400 --> 00:02:38.759
each other, how cultures are shaped
around the world and change. And if

33
00:02:38.759 --> 00:02:44.159
you think about this generative AI stuff, it's really what these guys have created.

34
00:02:44.159 --> 00:02:47.680
What these folks have created is a
reflection of the text of our world.

35
00:02:49.199 --> 00:02:52.599
That's really what it boils down to. These engines like open AI and

36
00:02:52.639 --> 00:02:57.319
Barred and LAMA and Lama too,
and all these huge foundational models as they're

37
00:02:57.360 --> 00:03:00.639
called. They have been trained on
the corpus of text on the internetch maybe

38
00:03:00.639 --> 00:03:05.840
it's Wikipedia or any other Twitter for
example, is a very good source of

39
00:03:05.840 --> 00:03:09.479
information excess, it's now called,
and they are now reflecting back to us

40
00:03:09.960 --> 00:03:15.439
this corpus of text, of language
of communication. And Andrew will talk a

41
00:03:15.479 --> 00:03:20.840
bit about the transformer models and some
other things that we're enabling. En schmid

42
00:03:20.960 --> 00:03:23.319
Huber, who I've interacted with,
very smart guy. He wrote papers on

43
00:03:23.360 --> 00:03:29.319
these transformer models and they do really
interesting things. So in the earlier days

44
00:03:29.360 --> 00:03:34.000
of these large language models, you
could only see like a token or two

45
00:03:34.319 --> 00:03:37.159
to the left or right, in
other words, backwards or forwards. You

46
00:03:37.159 --> 00:03:39.800
can think of a token as like
half a word basically, and what these

47
00:03:39.800 --> 00:03:45.960
engines are doing is just predicting text
that the engine thinks you want to see

48
00:03:46.000 --> 00:03:49.840
based upon your prompt. Well,
there's this whole issue around embeddings you have

49
00:03:49.919 --> 00:03:54.360
to use, which are like memories
to a human basically, But your embeddings,

50
00:03:54.400 --> 00:03:59.080
combined with the generative power of these
models, this predictive capability, is

51
00:03:59.080 --> 00:04:04.120
what has created these things. The
transformers allowed the engine to have multiple agents,

52
00:04:04.159 --> 00:04:08.680
if you will, so now you
can see a number of tokens forward,

53
00:04:08.719 --> 00:04:12.560
a number of tokens back and get
a much clearer picture. Just think

54
00:04:12.560 --> 00:04:15.639
of it as extending your view as
you're driving down the highway. Before it's

55
00:04:15.680 --> 00:04:19.079
like you're driving in dense fog.
Now the fog has gone and you can

56
00:04:19.120 --> 00:04:25.639
see pretty far forward as you're driving
along. So's it helps to create more

57
00:04:25.720 --> 00:04:30.399
accurate depictions of what the prompt is
telling the engine to create. But it

58
00:04:30.480 --> 00:04:33.480
still is an engine just predicting text
basically. But the other nice thing about

59
00:04:33.560 --> 00:04:39.120
these transformers is that they allow for
there to be multiple agents I refer to

60
00:04:39.120 --> 00:04:45.319
as almost like a peanut gallery of
participants who are suggesting what the next word

61
00:04:45.360 --> 00:04:47.360
should be, and then at a
certain point in time, the engine goes,

62
00:04:47.399 --> 00:04:49.439
Okay, I'm going to go with
this word next, I'm going to

63
00:04:49.439 --> 00:04:54.160
go with that word next, etc. Well that's what they're doing. But

64
00:04:54.240 --> 00:04:58.079
we're just at the beginning of this
I mean interactive AI I think is going

65
00:04:58.120 --> 00:05:00.279
to come next, and that's going
to be even more interesting. But let's

66
00:05:00.319 --> 00:05:03.800
just stay focused on the JENAI and
what do we mean my dark data?

67
00:05:04.000 --> 00:05:08.759
So the title of today's episode includes
dark data. That's all the data you

68
00:05:08.839 --> 00:05:12.800
have that you don't really know about. That you have a catalog that's sitting

69
00:05:12.879 --> 00:05:16.920
on servers somewhere or sitting in a
file system somewhere, and that's probably more

70
00:05:17.000 --> 00:05:21.240
than eighty percent of the data that
any mid size or large organization has.

71
00:05:21.560 --> 00:05:26.800
So you'd better be very careful about
pointing this AI gun at a vast trove

72
00:05:26.879 --> 00:05:30.519
of dark data because you don't know
what's coming back. And this stuff is

73
00:05:30.560 --> 00:05:33.480
so complex now that it's difficult to
figure out exactly how the engine got the

74
00:05:33.519 --> 00:05:38.240
answer that it got. And this
is partly why you have hallucinations. As

75
00:05:38.240 --> 00:05:42.000
a friend of mine, Eugene Burke
says, these large language models don't have

76
00:05:42.040 --> 00:05:46.759
an epistemological boundary, meaning they don't
know what they don't know, and that

77
00:05:46.800 --> 00:05:49.079
could be a very dangerous thing because
that's when they start making stuff up.

78
00:05:49.240 --> 00:05:53.600
Like chat GPT told me I wrote
three books. I was like, did

79
00:05:53.600 --> 00:05:56.560
I are you seeing in the future
or something? Because not overcall writing any

80
00:05:56.600 --> 00:05:59.879
books, but maybe. But anyway, someone who has focused on this for

81
00:06:00.079 --> 00:06:02.360
a while now and it's working with
a company that is trying to harness and

82
00:06:02.399 --> 00:06:06.040
if you will team this dark data, shine a light on this dark data

83
00:06:06.160 --> 00:06:10.879
is Andrew Turner of Practice Data.
Andrew, welcome to Inside Analysis. Tell

84
00:06:10.920 --> 00:06:14.920
us a bit about what you're working
on today and how you're helping companies make

85
00:06:14.959 --> 00:06:19.279
the most of this technology while mitigating
the risks. Well, thank yes,

86
00:06:19.319 --> 00:06:23.920
thank thanks. Firstly, thank you
for letting me join you from a very

87
00:06:23.959 --> 00:06:29.120
cold London tonight, so slightly different
times on to where you are. Yeah,

88
00:06:29.120 --> 00:06:31.920
I mean, I think, as
we all know, all these technologies

89
00:06:31.959 --> 00:06:38.000
go through a certain cycle, and
you know, we're now in this kind

90
00:06:38.000 --> 00:06:43.439
of I want to say hi,
but we're in this kind of roller coaster

91
00:06:43.560 --> 00:06:49.079
ride of you know, recovering from
opening eyes movement of releasing chat GPT.

92
00:06:49.279 --> 00:06:53.160
You know, well just just about
twelve months ago. Really now is it?

93
00:06:53.240 --> 00:06:56.519
So? Now? Now? I
suppose what I would say is the

94
00:06:56.920 --> 00:07:02.519
kind of the the door of actually, okay, let's let's stop talking about

95
00:07:03.959 --> 00:07:06.360
all the things we could do with
it. Let's talk about something we can

96
00:07:06.439 --> 00:07:09.399
do with it. Right, So
I think, you know, I think

97
00:07:09.399 --> 00:07:13.279
this is where the pragmatism kicks in, and you know, the marketing's got

98
00:07:13.319 --> 00:07:16.319
to stop and the actual delivery of
value is going to happen. So yeah,

99
00:07:16.360 --> 00:07:19.240
that whole thing, not saying that
the marketing won't stop, but yeah,

100
00:07:19.279 --> 00:07:23.680
I mean I think, you know, it was a really good intro.

101
00:07:23.759 --> 00:07:26.000
Thanks for doing Thanks for doing the
intro. I think the interesting thing

102
00:07:26.000 --> 00:07:30.079
that we find is that people are
still struggling with what to do with this

103
00:07:30.120 --> 00:07:33.600
stuff. So they you know,
they've been working on data related programs for

104
00:07:34.120 --> 00:07:40.720
multiple years and then this this kind
of tsunami of genreity AI happens. I

105
00:07:40.800 --> 00:07:45.720
was talking to the chief their officer
a very famous media company in London,

106
00:07:46.439 --> 00:07:51.240
and she the positive thing was that
she said that actually the CEO literally contacted

107
00:07:51.279 --> 00:07:56.920
her when this whole chat GPT thing
hit the hit the business and said,

108
00:07:57.000 --> 00:07:59.680
what do we do from a policy
point of view? Do we use it

109
00:08:00.079 --> 00:08:03.639
race it that we actually you know, put you know, kind of moratorium

110
00:08:03.680 --> 00:08:07.160
on it so none you can use
it. So I think the positive sheets

111
00:08:07.199 --> 00:08:11.319
awt of that is actually the CEO
contacted her to say, how do we

112
00:08:11.360 --> 00:08:13.800
govern this type of technology? You
know, let's not shut it down,

113
00:08:15.199 --> 00:08:18.439
let's work out how we can actually
you know, coexist with it. So

114
00:08:18.600 --> 00:08:22.000
but it's probably happening in lots of
boardrooms at the moment. Is what do

115
00:08:22.040 --> 00:08:24.680
you do with it? You know, what is the business use case?

116
00:08:24.120 --> 00:08:28.639
How do you prove the how do
you prove the value and make sure that

117
00:08:28.680 --> 00:08:31.799
you can get something out of it
quite quickly to build. It's the classic

118
00:08:31.840 --> 00:08:37.399
thing about scoping something that's of you
know, a reasonable scale, not too

119
00:08:37.440 --> 00:08:39.639
big, you know, so you
don't create this Bermuda triangle where you disappear

120
00:08:39.639 --> 00:08:46.200
into Beida triangle or you never could
get out. But you know how find

121
00:08:46.279 --> 00:08:48.039
you know, scope something that's actually
your value and then and actually go and

122
00:08:48.080 --> 00:08:50.840
deliver something and then you can scale
it from there. So it's back to

123
00:08:50.919 --> 00:08:54.679
it's back to those kind of principles, I think, whether you're embracing generative

124
00:08:54.679 --> 00:08:58.039
AI or other things. But yeah, one of the things that we did

125
00:08:58.080 --> 00:09:01.679
at practice data is trying to look
at how do we how do we hit

126
00:09:01.720 --> 00:09:07.080
this conundrum of managing data at scale? And to the point you did in

127
00:09:07.120 --> 00:09:11.440
the intro, is this whole issue
of there's lots of data just set around

128
00:09:11.519 --> 00:09:16.039
and being moved around your organization and
being you know, stored. Yeah,

129
00:09:16.440 --> 00:09:20.960
but do you actually have your arms
around it, so you know, putting

130
00:09:20.960 --> 00:09:26.759
a shining light and putting a real
proper enterprise level discovery across that data set,

131
00:09:26.120 --> 00:09:30.919
whether that's you know, unstructured data, semi structured or structured data,

132
00:09:30.960 --> 00:09:33.879
which is traditionally where people have looked
at just purely structured data. Obviously,

133
00:09:33.919 --> 00:09:37.279
when you get into large long as
one is, you're looking at unstructured and

134
00:09:37.360 --> 00:09:41.799
semi structured. So how do you
get that kind of over you on that

135
00:09:41.960 --> 00:09:46.440
data and then actually identify where are
you risks and opportunities and actually what all

136
00:09:46.440 --> 00:09:52.120
of the things we've done recently and
this can become more important is build effectively

137
00:09:52.519 --> 00:09:58.919
business use cases and industry models to
help these organizations really accelerate because you know,

138
00:10:00.240 --> 00:10:01.039
lots there's lots of tools out there, but a lot of them are

139
00:10:01.120 --> 00:10:05.159
very generic. So you know you're
in a you know you're in a property

140
00:10:05.200 --> 00:10:09.039
and casualty insurer. They don't want
to know, They don't want to have

141
00:10:09.039 --> 00:10:11.960
a generate tool. They want to
look at metadata and particular data models that

142
00:10:13.000 --> 00:10:18.600
actually they can relate to from a
business point of view to make decisions.

143
00:10:18.840 --> 00:10:22.639
That's the work that we've been doing
at Practice Data to be different. Because

144
00:10:22.320 --> 00:10:30.440
I was speaking to Mike Ferguson you
probably know at Big Data a few weeks

145
00:10:30.519 --> 00:10:35.639
ago. He said there's something like
thirty eight different data discovery, data,

146
00:10:35.840 --> 00:10:39.519
data catalog types of solution out there. So if I'm a buyer, if

147
00:10:39.559 --> 00:10:41.279
I go back into where I used
to work at Tesco or ge or whatever,

148
00:10:41.759 --> 00:10:45.679
it's a buyer's nightmare, you know, where how do you decide what

149
00:10:45.919 --> 00:10:50.000
you're going to buy? So what
we do is we do something very very

150
00:10:50.279 --> 00:10:54.759
narrow, but very very deep.
So that's what we're focused on, is

151
00:10:54.159 --> 00:10:58.279
how do we add value but not
try and take, not try and say

152
00:10:58.320 --> 00:11:01.919
we can do everything, but effect
we do automation, do curation as a

153
00:11:03.000 --> 00:11:05.080
service. So you know, we're
looking at how do we use our industry

154
00:11:05.159 --> 00:11:11.720
models, domain expertise and give that
value add into the actual overall analytics value

155
00:11:11.799 --> 00:11:13.840
check. That's what we're focused on
and you know what, you just gave

156
00:11:13.879 --> 00:11:18.320
me a good opportunity to bring ted
Lassow in to the conversation, which of

157
00:11:18.360 --> 00:11:22.799
course is taken America by storm on
Netflix or maybe it's Apple TV. I

158
00:11:22.840 --> 00:11:26.240
don't even know which streaming services.
I don't think it's actually Apple. But

159
00:11:26.879 --> 00:11:30.000
the joke there is, do you
take this thing as like a college football

160
00:11:30.240 --> 00:11:33.399
coach or something, and they bring
them over to England to, you know,

161
00:11:33.559 --> 00:11:37.279
to coach a football actual football We
call it soccer team, which is

162
00:11:37.279 --> 00:11:39.559
an outrageous proposition because it's like,
what the heck does this guy know about

163
00:11:39.679 --> 00:11:43.799
soccer? Or you know, Premier
League football, which is a serious deal.

164
00:11:43.879 --> 00:11:48.519
It's a very very different game.
And the point being here you typically

165
00:11:48.600 --> 00:11:52.919
want someone with domain expertise because you
don't want to hire, you know,

166
00:11:54.120 --> 00:11:56.840
a medical doctor to fix your car, for example. I mean, maybe

167
00:11:56.960 --> 00:12:01.679
he or she knows how to fix
cars just serendipitously, but that's certainly not

168
00:12:01.759 --> 00:12:05.720
their core area of expertise. And
if you're talking about trying to wrangle with

169
00:12:05.919 --> 00:12:11.399
these models that have some of them
one point nine three point two billion vertices

170
00:12:11.519 --> 00:12:16.360
or something, this is incredibly complex
stuff. You can't you can't boil the

171
00:12:16.440 --> 00:12:20.440
ocean. It's an age old mantra
in the IT world, don't boil the

172
00:12:20.519 --> 00:12:24.600
ocean. That's kind of what folks
you're starting to do. But you're not

173
00:12:24.759 --> 00:12:26.679
taking that direction. You're saying,
no, no, hold on, let's

174
00:12:26.720 --> 00:12:33.120
focus on these specific ontologies for certain
industries and businesses because that way you can

175
00:12:33.360 --> 00:12:37.799
pull this off. Because if you
try to use and this was my concern

176
00:12:37.960 --> 00:12:39.840
very early in the game when I
was listening to people talk about this,

177
00:12:39.919 --> 00:12:43.080
I'm like, Okay, I get
where you're going with this, but my

178
00:12:43.240 --> 00:12:46.720
goodness, the training will take forever
in a day, and by the time

179
00:12:46.759 --> 00:12:50.440
you've done training it, you're done
training it. Like, you know,

180
00:12:50.600 --> 00:12:52.919
are you even in business anymore?
I mean it's like, yeah, you

181
00:12:52.960 --> 00:12:56.440
can get value, but geez,
you got to be focused right. Well.

182
00:12:56.480 --> 00:13:00.600
Yeah, the issue as well is
that, you know, the I

183
00:13:00.679 --> 00:13:05.360
mean, I was talking to a
bank in the Naudic region of the day

184
00:13:05.480 --> 00:13:11.360
and there was the issue still is
around where because the technology takes so long

185
00:13:11.440 --> 00:13:16.600
to deliver. Yeah, and then
what you've got to do to is work

186
00:13:16.639 --> 00:13:18.639
out a sustainability model. You have
to actually, you've got to win the

187
00:13:18.720 --> 00:13:22.480
hearts and minds of the people who
are going to manage the day to going

188
00:13:22.559 --> 00:13:26.360
forward. If it takes you know, multiple years to get to that point

189
00:13:26.399 --> 00:13:31.600
where they can actually start adopting it, then you know that the hard stuff

190
00:13:31.679 --> 00:13:35.679
is the soft stuff, actually the
change management to actually engage, reward,

191
00:13:37.360 --> 00:13:43.200
incentivize, train, educate the people
who actually have got that deep domain expertise.

192
00:13:43.600 --> 00:13:46.480
When they see something that's actually they
can relate to very very quickly,

193
00:13:46.480 --> 00:13:48.840
you're going to improve that engagement.
You can improve that. You're going to

194
00:13:48.159 --> 00:13:52.159
change management a lot easier as well, because you talk in their language.

195
00:13:52.759 --> 00:13:56.000
If you've got a generic tool and
you start with a blank canvas, then

196
00:13:56.039 --> 00:13:58.320
you've got a lot of work to
do before you get even to that.

197
00:14:00.039 --> 00:14:01.679
You know, you could say base
camp, you know from an everest or

198
00:14:01.720 --> 00:14:05.559
climbing a mountain kind of analogy,
right, and that that can that you

199
00:14:05.639 --> 00:14:09.279
know, that's a frustrating. If
you're the leader of that business unit or

200
00:14:09.360 --> 00:14:11.440
that particular function, you've got a
lot on your hands to sort out.

201
00:14:11.720 --> 00:14:16.559
So we saw this a few years
ago when we were working with Andrew and

202
00:14:16.679 --> 00:14:20.559
and myself was the CEO and founder
was We're working in another company and we

203
00:14:20.639 --> 00:14:24.799
were doing some herculean work, piece
of work in very large enterprises, but

204
00:14:24.879 --> 00:14:28.039
you're just taking too long to get
to that base camp. So we said,

205
00:14:28.279 --> 00:14:31.279
how can we accelerate that put the
effort into building the models so you

206
00:14:31.320 --> 00:14:35.679
know, you can obviously get you
know, free open source models from hugging

207
00:14:35.759 --> 00:14:39.039
face. There's a lot that's there's
been some really great work they've done,

208
00:14:39.519 --> 00:14:43.360
but we believe that there's some real
added value in those industry use cases and

209
00:14:43.480 --> 00:14:46.879
industry models, and that's what we've
been focused on. As I can have

210
00:14:46.960 --> 00:14:54.919
USP well. And so the ontologies
or the industry models that you start with,

211
00:14:54.519 --> 00:15:00.320
they're really like roadmaps. They're like
a legend if you will, this

212
00:15:00.480 --> 00:15:03.200
discovery process. So at least you
have hallways to go down, rooms to

213
00:15:03.279 --> 00:15:07.840
look into instead of trying to you
know, find something in the entirety of

214
00:15:07.879 --> 00:15:11.320
the United States, which is just
too big to digest, right, And

215
00:15:11.440 --> 00:15:15.600
that's the challenge. If you're trying
to digest too much, you just get

216
00:15:15.639 --> 00:15:20.159
over overloaded and the system, the
brain just crashes and you just move on

217
00:15:20.279 --> 00:15:24.960
to something else. You just can't
do it. So you provided pathways essentially

218
00:15:24.080 --> 00:15:30.639
of discovery, and then you use
the technology to look around and to curate

219
00:15:30.679 --> 00:15:33.480
and you know, I have this
new concept I keep talking about, I

220
00:15:33.600 --> 00:15:35.840
need to write something and stick a
pin, so to speak. But I

221
00:15:35.919 --> 00:15:41.240
kind of view these foundational models as
a second chance for data. And what

222
00:15:41.320 --> 00:15:45.360
I mean by that is we spend
all these years organizing, analyzing, doing

223
00:15:45.480 --> 00:15:48.480
data warehousing and then data lake architectures
and all this stuff, but it was

224
00:15:48.519 --> 00:15:52.840
all sort of the same concept that
you persist data somewhere and then you go

225
00:15:52.960 --> 00:15:54.919
around and you try to analyze it
and number crunch it and so forth,

226
00:15:56.399 --> 00:15:58.279
and you know, we just we
have a mess out there. I mean,

227
00:15:58.320 --> 00:16:03.279
if you think of the the corpus
of information that any mid size to

228
00:16:03.440 --> 00:16:07.559
large organization has, it's just it's
monstrous. It's all over the map,

229
00:16:07.600 --> 00:16:11.759
it's in all sorts of formats,
and there's really no way you're going to

230
00:16:11.919 --> 00:16:15.639
tame that. But with these foundational
models, it's like we got a second

231
00:16:15.759 --> 00:16:19.879
chance to sit down and carefully curate
what we want to put into our embeddings,

232
00:16:21.279 --> 00:16:22.759
into our memories if you will,
for the AI to learn, and

233
00:16:22.840 --> 00:16:26.600
if you do that right, that
I think is going to be really compelling.

234
00:16:26.679 --> 00:16:30.799
What do you think we got two
minutes before the break? Yeah,

235
00:16:30.799 --> 00:16:32.960
I mean I think it's that you
know, a phrase we use messy,

236
00:16:33.080 --> 00:16:34.919
murky doubt data. You know,
how do you grab your arms around it?

237
00:16:36.039 --> 00:16:38.399
How do you make a real impact
on that? And one of the

238
00:16:38.399 --> 00:16:42.120
things we did, we've architected our
solution is do it a cloud scale because

239
00:16:42.159 --> 00:16:45.200
you know, that's no longer a
constrained So we're going to be making it

240
00:16:45.240 --> 00:16:49.519
available on you know, the cloud
provider marketplaces, so you can basically try

241
00:16:49.559 --> 00:16:52.519
before you buy. You can get
grab your arms around this stuff. You

242
00:16:52.600 --> 00:16:56.799
can apply it into your own data
set, you know, get get,

243
00:16:56.080 --> 00:17:00.120
get the momentum, get people on
board, and then scale it from that.

244
00:17:00.559 --> 00:17:03.559
Because you know that thing that no
infrastructure is no longer a constraint anymore.

245
00:17:03.960 --> 00:17:07.559
You can get on with that.
You can you know, you can

246
00:17:07.640 --> 00:17:11.400
select it from the marketplace, you
know, do it by tomorrow. It's

247
00:17:11.480 --> 00:17:14.759
you know, provisioning that type of
instructure is like an overnight thing. It's

248
00:17:14.839 --> 00:17:17.079
all. It takes you a couple
of hours. It's not something you have

249
00:17:17.160 --> 00:17:21.440
to work weeks and weeks and months
for. And I have an army of

250
00:17:21.480 --> 00:17:23.440
people to do that. You can
do that from you know, the established

251
00:17:23.480 --> 00:17:26.119
players. So that's that's also what
we're going to be releasing as well.

252
00:17:26.480 --> 00:17:30.319
To make that make it accessible and
make it, you know, adoptable,

253
00:17:30.519 --> 00:17:33.279
and actually then you're going to win
people over and then people going to adopt

254
00:17:33.319 --> 00:17:37.799
it well, right, because you
think about the patience of the average person

255
00:17:37.880 --> 00:17:41.640
these days, and there's no doubt
that change is always difficult. Change is

256
00:17:41.640 --> 00:17:45.279
always hard. Some people are better
at it than other people, but everyone

257
00:17:45.559 --> 00:17:48.839
has to deal with change. And
you know, probably the older you get,

258
00:17:48.880 --> 00:17:52.079
the less willing you are to change. It's kind of your old habits

259
00:17:52.119 --> 00:17:55.720
die hard, as they say.
But to your point, if someone can

260
00:17:55.799 --> 00:17:59.920
make progress in a fairly short period
of time or at least see where the

261
00:18:00.240 --> 00:18:03.359
going, that's the key. That's
the way you kind of win the hearts

262
00:18:03.400 --> 00:18:04.759
and minds. And you know,
it sort of reminds me of my own

263
00:18:06.079 --> 00:18:07.960
writing process when I have to write
something new. At first, I'm like,

264
00:18:08.039 --> 00:18:11.200
oh goodness, what am I gonna
do. I have to find a

265
00:18:11.359 --> 00:18:15.880
starting point. I have to find
where is that bass camp? Okay,

266
00:18:15.960 --> 00:18:18.920
the base camp is up there,
And once I find that, I can

267
00:18:18.000 --> 00:18:22.680
climb my way to it and then
take a few deep breaths and then prepare

268
00:18:22.759 --> 00:18:25.119
for the next part of the journey. And that's kind of where we are

269
00:18:25.240 --> 00:18:26.480
right now. I think everyone's trying
to find their base camp. But folks,

270
00:18:26.519 --> 00:18:30.079
don't touch on that. I'll be
right back. You're listening to Inside

271
00:18:30.079 --> 00:18:41.119
Analysis. Expected, Welcome back to
Inside Analysis. Here's your host, Eric

272
00:18:41.319 --> 00:18:47.039
Tabanaugh. All right, folks,
take us to the future. Indeed,

273
00:18:47.880 --> 00:18:52.000
I love that song that's by Black
Bananas is the band. You gotta love

274
00:18:52.039 --> 00:18:55.799
it. I love that song and
it reminds me of William Gibson, the

275
00:18:56.000 --> 00:18:57.480
futurists. I've always wanted to meet
a futur rist, to be like,

276
00:18:57.559 --> 00:19:00.319
are you even here right now?
Man? But anyway, he's a smart

277
00:19:00.359 --> 00:19:03.240
guy, and he said, the
future is here already, it's just not

278
00:19:03.480 --> 00:19:07.799
evenly distributed. It's like, oh, that's a good one. And all,

279
00:19:07.880 --> 00:19:11.559
of course, it also means the
past is still here, just in

280
00:19:11.680 --> 00:19:15.200
pockets all around us. So it's
an interesting dynamic. But we're talking today

281
00:19:15.200 --> 00:19:18.839
with Andrew Turner, general manager of
Praxy Data. That's PRXI Data, just

282
00:19:18.960 --> 00:19:22.920
like it sounds. And we're talking
about jen Ai and making sense of dark

283
00:19:23.000 --> 00:19:26.640
data. And we wanted to talk
about the different roles, things like prompt

284
00:19:26.720 --> 00:19:30.559
engineering, for example, which is
a real task now, it's a real

285
00:19:30.640 --> 00:19:33.680
skill set and as many people have
said and I think this is true.

286
00:19:34.160 --> 00:19:37.359
It's not so much that AI is
going to take jobs away, per se

287
00:19:37.559 --> 00:19:41.359
as it is that people who learn
to use AI effectively are going to have

288
00:19:41.440 --> 00:19:45.359
a much better time getting jobs and
doing their jobs than people who don't.

289
00:19:45.400 --> 00:19:49.519
It's like any other tool. And
you don't see too many yaks out there

290
00:19:49.599 --> 00:19:53.440
in the fields tilling the land.
We're using giant machines to do that stuff

291
00:19:53.440 --> 00:19:57.119
these days because they're much more efficient. And I've always believed that machines can

292
00:19:57.160 --> 00:20:00.039
do the job better than people.
Let the machine do the job and let

293
00:20:00.079 --> 00:20:03.599
the people do the more interesting things
on top of what the machines do.

294
00:20:04.279 --> 00:20:07.759
But Andrew, we are going to
have to deal with some pretty heavy duty

295
00:20:08.000 --> 00:20:11.799
change management. Right in our deep
dive earlier, you talked about this McKinsey

296
00:20:11.880 --> 00:20:15.079
quote where they said that we now
believe, or they now believe, we'll

297
00:20:15.119 --> 00:20:21.519
be able to automate many of these
processes ten years sooner than we had previously

298
00:20:21.680 --> 00:20:25.480
thought. That's a pretty big deal. I mean, that's a step change.

299
00:20:25.559 --> 00:20:29.319
You're like skipping a decade basically,
and it's going to require us all

300
00:20:29.440 --> 00:20:32.200
to get our sea legs. Right. What do you think? Yeah,

301
00:20:32.279 --> 00:20:36.720
I mean I think it's absolutely I
mean that quat quote came out of the

302
00:20:36.759 --> 00:20:41.400
fourth of October, so it's only
a few weeks ago, and it's absolutely

303
00:20:41.720 --> 00:20:45.960
you know when you think about the
ramifications of it's absolutely staggering. You know,

304
00:20:45.039 --> 00:20:48.400
you talk about, you know,
I remember the days of business process

305
00:20:48.440 --> 00:20:53.240
reengineering, but that's like, to
your point, it's like BPR on steroids,

306
00:20:53.359 --> 00:20:57.960
you know, times ten. So
you know, it's it's kind of

307
00:20:59.039 --> 00:21:02.960
like it's saying, you know,
you it's a real opportunity to redesign and

308
00:21:03.039 --> 00:21:08.200
really step change and reorganize the organization. And any CEO that's listening to this

309
00:21:08.319 --> 00:21:11.720
session that kind of says, oh, you know, we can put it

310
00:21:11.759 --> 00:21:14.720
on the back burner. We could
ignore the impact of generaty of AI.

311
00:21:15.000 --> 00:21:18.880
You know, it's literally it's it's
a it's that would be a mistake.

312
00:21:18.599 --> 00:21:22.480
The question. But the question I
think is then about, you know,

313
00:21:22.559 --> 00:21:25.039
how do you then sit down with
your people team, you know, what

314
00:21:25.240 --> 00:21:27.920
is the organization of the future,
because I think it is that point you

315
00:21:29.039 --> 00:21:33.119
made a bit earlier. It's not
just replacement, it's it's coexistence and actually

316
00:21:33.200 --> 00:21:38.200
how do you shift the work into
that new type of role and then it

317
00:21:38.319 --> 00:21:44.119
is a piece around upskilling and you
know, how do you upskill the right

318
00:21:44.319 --> 00:21:48.480
roles in the future. And you
know, one of the things that I

319
00:21:48.599 --> 00:21:52.079
was sat in the audience at this
recent session where you know, the couple

320
00:21:52.119 --> 00:21:56.680
of CTOs presenting their take or you
know, from a technical perspective about their

321
00:21:56.799 --> 00:22:00.400
view on generative r and they said, you know, it's pretty much the

322
00:22:00.480 --> 00:22:04.839
whole organization structure is going to be
impacted at different levels. And people that

323
00:22:04.960 --> 00:22:08.640
we you know, we traditionally call
you know, developers that the rock stars

324
00:22:08.680 --> 00:22:12.200
of the future or you know,
the the rock stars of the of the

325
00:22:12.759 --> 00:22:15.960
today and the stars of the future. But then you know, I think

326
00:22:17.000 --> 00:22:19.880
it's going to change that that role
because people are spending a lot more money

327
00:22:19.920 --> 00:22:23.599
on people in this kind of area, and actually, how do you get

328
00:22:23.640 --> 00:22:26.400
the productivity out of those you know
that that you know, everybody wants to

329
00:22:26.440 --> 00:22:30.359
be a data scientist and then you
know, you pay, you pay,

330
00:22:30.559 --> 00:22:33.680
you pay market rate to attract the
best, but then actually, how do

331
00:22:33.720 --> 00:22:37.599
you put the tools around them to
actually become productive and impactful in that organization.

332
00:22:38.160 --> 00:22:41.680
So in generality, it puts that
the next layer of complexity on that

333
00:22:42.079 --> 00:22:47.599
I think there's an opportunity for organizations
to help other, you know, enterprises,

334
00:22:47.680 --> 00:22:49.440
to actually educate them on prompt engineering. I think it's a key,

335
00:22:49.720 --> 00:22:53.799
it's a core competency going to be
going forward. And if you're not done

336
00:22:55.319 --> 00:23:00.200
prompt engineering or interacted with bar GPT, et cetera, then you know,

337
00:23:00.519 --> 00:23:03.359
it seems like talking, you know, talking interacting with a real person.

338
00:23:03.759 --> 00:23:07.279
It's quite it's quite remarkable. Yeah, it really is. And you know,

339
00:23:07.640 --> 00:23:11.079
I had Matt McClarty, the CTO
of Boomy, on the show a

340
00:23:11.160 --> 00:23:15.160
few weeks ago, and he had
this great line where he referred to what

341
00:23:15.799 --> 00:23:18.839
era we're in in a certain sense, and we talked about machine learning and

342
00:23:18.880 --> 00:23:22.079
AI and what's really happening now,
and he said, it's a time of

343
00:23:22.599 --> 00:23:27.640
learning, meaning we humans are now
learning what this stuff can do and how

344
00:23:27.680 --> 00:23:32.799
it works, and we're learning about
how to make it hum we're learning about

345
00:23:32.799 --> 00:23:36.279
what the challenges are. And it
kind of reminds me of something I worked

346
00:23:36.319 --> 00:23:37.960
with a couple of Russians years ago, and it was a good Russian and

347
00:23:38.000 --> 00:23:41.440
the bad Russian. The good Russian
Niko and Minchugol. He was a very

348
00:23:41.480 --> 00:23:45.960
smart guy in Nikolai Minchugo. He
ran the first ever advertising agency in Moscow

349
00:23:47.559 --> 00:23:49.519
because when Pedestoika came out, all
of a sudden, you had a need

350
00:23:49.599 --> 00:23:52.559
for ad agencies. There was no
need before in the communist system because why

351
00:23:52.599 --> 00:23:55.720
would you have an ad agency?
Right, So he was the first to

352
00:23:55.799 --> 00:23:59.279
do that. He's a designer but
a very good coder too, and he

353
00:23:59.359 --> 00:24:02.319
built out this system. But he
would say funny things to me, like

354
00:24:02.400 --> 00:24:06.480
one day said, the funny thing
about the web is a no one knows

355
00:24:06.559 --> 00:24:11.519
how it works. All we know
is it works, right, And that's

356
00:24:11.599 --> 00:24:12.480
kind of where we are at this
lom stuff. I mean, we do

357
00:24:12.720 --> 00:24:15.480
know more about how it works.
You can get into the weeds, and

358
00:24:15.599 --> 00:24:21.960
the embeddings are crucial, right because
the models have been trained on this corpus

359
00:24:22.079 --> 00:24:25.960
of text that is vast. It's
very very large. So there's good stuff,

360
00:24:26.000 --> 00:24:29.480
there's bad stuff, there's truthful stuff, there's not truthful stuff, and

361
00:24:29.559 --> 00:24:33.359
so it has all this material.
But the AI engine itself is kind of

362
00:24:33.400 --> 00:24:38.119
like a teenager with a tremendous vocabulary, but it still doesn't know the sort

363
00:24:38.160 --> 00:24:41.880
of more rays and the rules and
what it should do versus what it shouldn't

364
00:24:41.920 --> 00:24:45.200
do, And these are all important
things. To know and so to me,

365
00:24:45.440 --> 00:24:48.440
those embeddings are crucial. And I
get back to this sort of second

366
00:24:48.519 --> 00:24:55.519
chance for data concept because once your
organization understands what these things are and you

367
00:24:55.640 --> 00:24:57.720
kind of wrap your head around some
of the use cases, like customer support,

368
00:24:59.480 --> 00:25:03.119
for sure, it's a big one. News gathering, news dissemination,

369
00:25:03.440 --> 00:25:07.519
news generation I think is going to
be revolutionized by this stuff. Once you

370
00:25:07.640 --> 00:25:10.880
wrap your head around it, then
you can sort of start the process and

371
00:25:10.920 --> 00:25:12.640
figure out, well, what do
we want. How do we want to

372
00:25:12.720 --> 00:25:15.839
teach our teenage child. Do we
want to send it to Catholic school?

373
00:25:15.880 --> 00:25:18.279
Do we want to send it to
public schools? We want a home school

374
00:25:18.359 --> 00:25:22.880
basically, And you really have to
know like what your appetite is, and

375
00:25:22.920 --> 00:25:26.240
that gets into the training side,
right, So you don't want to be

376
00:25:26.319 --> 00:25:29.319
trying to boil the ocean. Don't
try to, you know, do two

377
00:25:29.440 --> 00:25:33.400
full time jobs and educate your five
kids by yourself. You're probably going to

378
00:25:33.400 --> 00:25:36.480
burn yourself out if you do that. So you have to kind of understand

379
00:25:36.519 --> 00:25:38.400
the scope and the appetite and where
you're going. And if you do that

380
00:25:40.039 --> 00:25:41.920
and you use tools I think like
praxy data for example, I think you're

381
00:25:41.960 --> 00:25:45.319
going to have a really good chance
of success. But what do you think,

382
00:25:45.359 --> 00:25:48.359
Andrew, Well, I mean,
I think what you touched on there

383
00:25:48.480 --> 00:25:52.119
was that what i'll call data toxicity. You know, so that thing about

384
00:25:52.319 --> 00:25:55.440
you know, because of you know, there was a scenario that will slightly

385
00:25:55.519 --> 00:25:59.559
the week where if you get it
wrong the way you connect these things together,

386
00:26:00.240 --> 00:26:03.279
which you can actually generate data and
then it actually publishes that data and

387
00:26:03.359 --> 00:26:07.920
then it feasts on itself, like
it becomes like a carnival for its own

388
00:26:08.000 --> 00:26:14.039
data sets. So the danger there
is that you you know, you go

389
00:26:14.160 --> 00:26:18.160
back to origin, what's the origin, what's the syndicated origin, and what's

390
00:26:18.240 --> 00:26:22.160
true and what's not true? Yeah, right, there's a piece around that,

391
00:26:22.319 --> 00:26:27.039
around you know, lineage and and
the kind of the inheritance of that

392
00:26:27.240 --> 00:26:33.759
data, which becomes back to some
kind of fundamental how you manage data at

393
00:26:33.799 --> 00:26:37.680
scale principles you need to get you
need to get right. So you know,

394
00:26:37.920 --> 00:26:38.960
that's the point you said a bit
earlier about this. You know,

395
00:26:40.279 --> 00:26:42.640
what we see is this messy,
murky data. You've got all this kind

396
00:26:42.680 --> 00:26:47.079
of patchwork quilt. You know,
people have gone out and bought lots of

397
00:26:47.160 --> 00:26:51.279
tools and they've evolved over time.
But actually what we see has being this

398
00:26:51.440 --> 00:26:53.920
kind of like I suppose you know
this, I think we talked about it

399
00:26:55.079 --> 00:26:56.920
that you know, it's backplane,
you know, like a backplane concept.

400
00:26:57.240 --> 00:27:00.519
You know, how do you have
a backplane across your enterprise that you can

401
00:27:00.559 --> 00:27:06.599
actually rely on, that you you
look after properly, that's going to take

402
00:27:06.680 --> 00:27:08.920
in these different types of dat source
So you have that I don't want to

403
00:27:10.400 --> 00:27:15.039
it's an age your phrase about that
golden record definition and people still struggling with

404
00:27:15.160 --> 00:27:21.000
that. So you know, all
we're trying to do is is not not

405
00:27:21.160 --> 00:27:23.160
say you're going to throw away all
the thity investments you've already made, but

406
00:27:23.319 --> 00:27:27.759
how do you actually you know,
tap into this this model, these models,

407
00:27:27.880 --> 00:27:33.240
this this industry insight to to add
value to your you know, getting

408
00:27:33.319 --> 00:27:36.200
those people on board and then dealing
with the change management we talked about a

409
00:27:36.200 --> 00:27:40.559
bit earlier, because that's that is
really really the sustainability model. That's the

410
00:27:40.559 --> 00:27:42.480
biggest issue we can see out there. You know, people can keep buying

411
00:27:42.519 --> 00:27:47.759
these tools, but until you actually
get work out a path to get people

412
00:27:47.839 --> 00:27:52.319
engaged, that's that's the key to
it. So you know, we can't

413
00:27:52.480 --> 00:27:55.519
you know, I don't know what
we have situations in the future where we

414
00:27:55.599 --> 00:28:00.160
have robots managing the data in our
organizations. There's that tacit knowledge that's been

415
00:28:00.240 --> 00:28:03.440
hearing the brain, right, that
people have got experience about that is difficult,

416
00:28:03.640 --> 00:28:07.920
difficult to codify. I'm sure work
out through bring computer into face a

417
00:28:07.960 --> 00:28:11.799
way of codifying at some point in
the future. But for the for the

418
00:28:11.880 --> 00:28:14.759
you know, for the next few
years, I think that's where we're trying

419
00:28:14.799 --> 00:28:18.640
to add some value in this,
in this kind of this ocean of dates

420
00:28:18.680 --> 00:28:21.720
that we're dealing with. Yeah,
and I think, you know, for

421
00:28:21.960 --> 00:28:26.559
use cases, certainly chatbots are going
to get a lot better. You know,

422
00:28:26.559 --> 00:28:29.319
a lot of times in enterprise technology, stuff comes out a bit too

423
00:28:29.440 --> 00:28:33.160
soon. Right when the chatbots first
came out, I think they just they

424
00:28:33.200 --> 00:28:37.200
weren't quite ready for prime time,
and so they didn't do as well as

425
00:28:37.359 --> 00:28:40.720
as we needed them to do.
And you know as well as I do,

426
00:28:40.880 --> 00:28:45.039
that people have a very short attention
span these days and very little patience.

427
00:28:45.319 --> 00:28:47.559
So when something doesn't work out,
like oh, that didn't work,

428
00:28:48.160 --> 00:28:52.480
and then these impressions last, right, these impressions linger for periods of time,

429
00:28:52.519 --> 00:28:55.880
and it's hard to sort of shake
off that old mindset and realize,

430
00:28:55.960 --> 00:28:57.519
no, this is a new tool, this is a new way of doing

431
00:28:57.599 --> 00:29:03.039
things. But I think that the
key is to help people understand that used

432
00:29:03.160 --> 00:29:07.920
properly, this new suite of technologies
is going to make your job a heck

433
00:29:07.960 --> 00:29:12.720
of a lot easier, and it's
going to it's going to make your job

434
00:29:12.799 --> 00:29:15.839
more interesting and more fun, is
I guess a way I'd put it,

435
00:29:15.920 --> 00:29:22.680
because in general, machine learning algorithms
are very good at tackling very tedious tasks

436
00:29:22.240 --> 00:29:27.519
like scanning fifty thousand purchase orders and
finding the twenty seven that have some error

437
00:29:27.599 --> 00:29:32.359
in them. Well, a human
being a record after record after record,

438
00:29:32.440 --> 00:29:34.599
forget about it. Man, it's
a nightmare and it's terrible for morale.

439
00:29:36.319 --> 00:29:40.640
So as long as people understand,
and this this is my beef with the

440
00:29:40.759 --> 00:29:44.400
media writ large, is that I
have this expression I throw out there that

441
00:29:44.519 --> 00:29:48.759
the narrative is always wrong because the
narrative is a story, and the story

442
00:29:48.839 --> 00:29:52.920
is not reality. The narrative is
some story I've spun up around components of

443
00:29:53.039 --> 00:29:57.759
reality. But the reality with AI
is that it is making our jobs easier,

444
00:29:59.200 --> 00:30:02.599
or making certain easier. It's not
necessarily taking jobs away. And it's

445
00:30:02.640 --> 00:30:04.279
not the red eyed robot that's going
to take over the world and all that

446
00:30:04.359 --> 00:30:07.559
stuff. But the media loves to
focus on that. They love to focus

447
00:30:07.640 --> 00:30:11.319
on negative because that sells, It
gets eyeballs and so forth. And I

448
00:30:11.400 --> 00:30:17.240
think it's important that people recognize this
as a whole new tool set. You

449
00:30:17.279 --> 00:30:21.440
know, I mentioned dere Dodd is
sort of redoubling. It's a reflection on

450
00:30:21.680 --> 00:30:26.640
everything that we've seen already. So
now it's almost more like riding a wave

451
00:30:26.720 --> 00:30:29.640
instead of trying to create the waves. We have the waves now now we

452
00:30:29.720 --> 00:30:32.599
have to figure out how to ride
them. What do you think, Yeah,

453
00:30:32.640 --> 00:30:33.799
yeah, I mean I think that
what you were touching on a bit

454
00:30:33.839 --> 00:30:38.720
earlier about that the chatbot situation was
that you know, you obviously the the

455
00:30:38.880 --> 00:30:42.359
data that feeds the chat bot,
right, So you touching on this thing

456
00:30:42.440 --> 00:30:47.440
about there's been some scary situations happened
recently with you know, what you do

457
00:30:47.559 --> 00:30:49.799
privately and what you do publicly.
Yeah, so you know, how do

458
00:30:49.839 --> 00:30:56.720
you deal with that, the risk
around data privacy, preserving privacy around that,

459
00:30:56.240 --> 00:31:00.319
you know, building your own private
you know LLLM models, your own

460
00:31:00.400 --> 00:31:10.720
data models, so putting your domain
expertise with recognize other external data sources,

461
00:31:11.319 --> 00:31:15.799
and that's where you're going to have
that differentiator in service because if you create

462
00:31:15.880 --> 00:31:19.359
that kind of that data source that
then is an input into your chatbot,

463
00:31:19.680 --> 00:31:22.599
you're then going to have to stand
a much better chance of having a very

464
00:31:23.119 --> 00:31:29.480
a delighter as a customer experience.
That's going to drive loyalty. Yes,

465
00:31:29.599 --> 00:31:32.200
So you know, one of the
use cases we've been looking at recently is

466
00:31:32.240 --> 00:31:38.920
with an organization that's dealing with insurance
and roadside recovery and that kind of thing

467
00:31:40.000 --> 00:31:44.440
where you kind of combining lots of
you know, are you getting complaint letters

468
00:31:44.480 --> 00:31:47.799
in from into a call center,
how is the call center agent dealing with

469
00:31:48.039 --> 00:31:51.799
and linking and saying who is that
particular customers that's affected by that? So

470
00:31:51.839 --> 00:31:56.279
you're getting different types of data input, but you've got to connect these dots

471
00:31:56.000 --> 00:31:59.839
and how do you put an instruction
in place to allow you to do that.

472
00:32:00.240 --> 00:32:01.160
You've got to then, you know, you've got to work. It's

473
00:32:01.160 --> 00:32:08.000
a complex object. It's a complex
structure of how you sat in a CRM

474
00:32:08.079 --> 00:32:10.640
system or something like that. But
then actually you've got a different type of

475
00:32:10.680 --> 00:32:16.799
interaction of data coming in through through
letters and call center records, how do

476
00:32:16.880 --> 00:32:20.599
you make sure you connect those dots
to make the right decision to look after

477
00:32:20.640 --> 00:32:23.279
that customer or deal or deal with
that service intervention. So those are the

478
00:32:23.319 --> 00:32:30.720
types of thing where we can see
that having this kind of foundational capability is

479
00:32:30.079 --> 00:32:37.359
very important where you're actually connecting the
different types of data that before were siloed'.

480
00:32:37.759 --> 00:32:42.319
That's a really good point. Yeah, that's very interesting because you're right

481
00:32:42.359 --> 00:32:45.160
when you think about the old way
of connecting systems, and I'm old enough

482
00:32:45.200 --> 00:32:52.160
to remember EAI Enterprise Application Integration right
where we would connect one application to another.

483
00:32:52.720 --> 00:32:54.200
Well, in the old days,
you would look at things like primary

484
00:32:54.319 --> 00:33:04.079
keys and foreign keys and stuff like
this as a a very what not what's

485
00:33:04.160 --> 00:33:07.799
its declarative and what's the other one? I can't even remember the name.

486
00:33:07.839 --> 00:33:12.480
Different kinds of programming. It's very
specific instruction go here, do this,

487
00:33:12.640 --> 00:33:15.400
go here, do that. But
it's it's rigid. It's a very rigid

488
00:33:15.480 --> 00:33:19.680
connection. And what you want is
more of a fluid connection that can accept

489
00:33:19.759 --> 00:33:22.440
inputs from multiple sources, like you
say, from the call center, from

490
00:33:22.880 --> 00:33:28.720
text messages, from mobile, from
your laptops or whatever. They are all

491
00:33:28.759 --> 00:33:31.160
these different sources which to manually connect. Yeah, you can do it,

492
00:33:31.640 --> 00:33:37.559
but it's cluegy and it's also not
very performance and then you wind up spending

493
00:33:37.599 --> 00:33:39.480
all this time and effort trying to
figure out how to optimize the performance.

494
00:33:39.519 --> 00:33:43.920
So let's add some hardware over here, let's add some caching over there.

495
00:33:44.319 --> 00:33:49.400
These are very specific tactical things that
you can do tactical, but it's hard

496
00:33:49.440 --> 00:33:52.920
to pull off. Whereas with these
foundational models, that's what they do for

497
00:33:52.039 --> 00:33:57.039
a living. That's what they're designed
to do, is to take sources of

498
00:33:57.200 --> 00:34:00.640
information and fuse them and then generate
some output. And if you can train

499
00:34:00.759 --> 00:34:06.839
that to do something using only your
trusted data, the second chance for data

500
00:34:06.880 --> 00:34:09.480
that we're talking about, then you're
really going to get somewhere because those things

501
00:34:09.519 --> 00:34:13.760
can do it really really really fast, and they just kind of figure stuff

502
00:34:13.800 --> 00:34:16.239
out about how to do that dynamically, instead of having to give such specific

503
00:34:16.360 --> 00:34:21.880
instructions to your systems people about how
to get this connected to that and not

504
00:34:21.960 --> 00:34:23.480
connected to this. They're kind of
solving that out of the box. But

505
00:34:23.519 --> 00:34:27.480
folks don't touch that that. I'll
be right back. You're listening to inside

506
00:34:27.480 --> 00:34:38.159
Analysis. Welcome back to Inside Analysis. Here's your host, Eric Tabanac.

507
00:34:40.800 --> 00:34:45.480
All right, folks, welcome back
to our show today, the Dark Data

508
00:34:45.760 --> 00:34:50.360
Challenge. Solving the Dark Data Challenge
with Practical jen Ai. You're truly Eric

509
00:34:50.400 --> 00:34:54.400
Havanat talking to Andrew Turner of Praxy
Data Praxy Practical. I don't know if

510
00:34:54.440 --> 00:34:58.480
that was the logic behind it,
but you know. And now let's talk

511
00:34:58.480 --> 00:35:01.239
a little bit about public verse is
private. Right, So when you're on

512
00:35:01.360 --> 00:35:06.440
Google, you're just when you're using
chat GPT, let's say the original one.

513
00:35:06.480 --> 00:35:09.679
That's the public version that's trained on
all this data that's out there,

514
00:35:09.800 --> 00:35:13.280
but it hallucinates, it does all
kinds of different things. What you really

515
00:35:13.320 --> 00:35:19.960
want is a private instance that you
then very carefully train by curating only the

516
00:35:20.079 --> 00:35:22.719
data that you want, only the
trusted data. It's like teaching your kid

517
00:35:22.840 --> 00:35:27.679
how to grow up. You want
to be careful about what you expose them

518
00:35:27.719 --> 00:35:30.039
to. Not too much YouTube,
not too much TikTok. Maybe we screwed

519
00:35:30.119 --> 00:35:32.679
up on that one a little bit
with our kid, but anyway, we

520
00:35:32.800 --> 00:35:37.280
try our best. But tell us
about your thoughts. I think every company

521
00:35:37.360 --> 00:35:39.760
is going to want to have their
own private instance of one of these models

522
00:35:40.320 --> 00:35:46.280
and just over time build it out
and let it support things like customer support,

523
00:35:46.440 --> 00:35:50.719
let it support things like even supply
chain other stuff. But what are

524
00:35:50.719 --> 00:35:54.440
your thoughts, Andrew, Yeah,
I mean I think it's I think people

525
00:35:55.360 --> 00:35:59.960
originally thought there was only one model, right, They only thought there was

526
00:36:00.079 --> 00:36:01.840
only a public one. And you've
probably heard about that. You know,

527
00:36:01.880 --> 00:36:07.920
there's been various examples where you know, the high tech company based in South

528
00:36:07.000 --> 00:36:13.440
Korea, you know, uploaded their
source code into chat, GPT, et

529
00:36:13.480 --> 00:36:16.320
cetera, et cetera. So you
know, but the I think that you

530
00:36:16.400 --> 00:36:19.639
know, this is where the CISOW
comes down into. This is where they

531
00:36:19.719 --> 00:36:23.320
kind of you know, the IT
security compliance piece and that the chief data

532
00:36:23.400 --> 00:36:29.719
officer around you know, data privacy
is you've got to govern your assets as

533
00:36:29.760 --> 00:36:32.039
you will govern you know, any
other assets. So you've got to it's

534
00:36:32.199 --> 00:36:36.960
just a different type of asset that
you're having to manage here. So yeah,

535
00:36:37.400 --> 00:36:39.480
there's there's a raging debate. But
the net net is you've got to

536
00:36:39.480 --> 00:36:44.280
build your own private you know,
generative AI asset asset library, you know,

537
00:36:44.320 --> 00:36:50.920
your set up models to address that
problem. Because you know, we

538
00:36:51.360 --> 00:36:54.039
know there's some there's some class actions
going on with some litigation going on with

539
00:36:54.119 --> 00:37:00.239
certain tech companies at the moment around
you know that they're the things that they've

540
00:37:00.360 --> 00:37:05.920
done prior to coming out to market, and you know it's obviously causing a

541
00:37:05.960 --> 00:37:08.280
lot of concern for a lot of
organizations. So I think getting the right

542
00:37:08.360 --> 00:37:14.920
advice around how you architecture what you're
going to be put in place. But

543
00:37:15.079 --> 00:37:17.159
yeah, you've got to have something
that's protected, it's it's it's like a

544
00:37:17.800 --> 00:37:21.039
it's kind of like sleep at night
factor that you've got to think about,

545
00:37:21.320 --> 00:37:23.239
and you've got to get the right
people around the table. And I think

546
00:37:23.320 --> 00:37:28.320
that you know, the key thing
I think is also if you're a CEO

547
00:37:28.880 --> 00:37:32.280
or a CFO, is looking at
the the leadership team you've got working for

548
00:37:32.480 --> 00:37:37.360
you to give you briefing so you
understand the ramifications of just putting this piece

549
00:37:37.400 --> 00:37:40.440
thing, this this thing together in
the right way. I think then then

550
00:37:40.480 --> 00:37:45.159
you can really exploit it. So
that's the kind of that's the thing that

551
00:37:45.239 --> 00:37:49.239
we're seeing. There's there's people are
definitely speeding up now. They you know,

552
00:37:49.320 --> 00:37:52.199
they've they've been watching and waiting,
they've been dealing with this kind of

553
00:37:52.280 --> 00:37:55.679
this massive tsunami and all the all
the kind of you know, as you

554
00:37:55.760 --> 00:38:00.079
mentioned the kind of media and PR
around this, but people, now what

555
00:38:00.119 --> 00:38:02.400
I wanted to go into, get
into brass tax gett into real use cases,

556
00:38:04.280 --> 00:38:06.760
you know, work out you can
actually make some real big impact in

557
00:38:06.800 --> 00:38:10.480
their organization, reshaping their organizations we
talked about earlier with their organization structures.

558
00:38:12.039 --> 00:38:16.000
And then you know, I think
it's raising the importance of that chief Data

559
00:38:16.000 --> 00:38:21.559
Officer role, the chief Digital office, so the CIO, the CTO to

560
00:38:21.679 --> 00:38:23.199
be at the top table again,
you know, and actually to really help

561
00:38:23.239 --> 00:38:29.119
the CEO and CFO and top team
to make those really really strategic decisions.

562
00:38:29.840 --> 00:38:32.159
Yeah. Well, you think about
digital transformation, which we've been talking about

563
00:38:32.199 --> 00:38:37.440
for twenty years or so, and
we've been doing digital transformation, but this

564
00:38:37.599 --> 00:38:42.599
is almost like wholesale organizational transformation.
Is what can happen if you do this

565
00:38:42.639 --> 00:38:45.679
stuff? Right? What do you
think? Yeah? No, definitely,

566
00:38:45.719 --> 00:38:50.360
I mean I think I mean it
was I was literally two hours away from

567
00:38:50.400 --> 00:38:53.360
presenting something and that that McKinsey quote
out came out. I mean it's absolutely

568
00:38:54.280 --> 00:38:59.599
mind blowing, right, you know, taking ten years off an organization's evolution

569
00:38:59.800 --> 00:39:05.360
with in one stroke. So you
know, obviously you know, is is

570
00:39:05.440 --> 00:39:08.840
that a precursor for you know,
oh please please bring up a very very

571
00:39:08.920 --> 00:39:14.599
large strategic consulting firm to actually help
you with that transformation. But you need

572
00:39:14.679 --> 00:39:15.639
some help. I think, you
know, you need to get the right

573
00:39:15.679 --> 00:39:19.719
people around the table, and you
need external help. There is a and

574
00:39:19.800 --> 00:39:23.079
that's what's driving up the cost of
resources because there is a finite resource of

575
00:39:23.159 --> 00:39:28.480
people who know what what the pass
should be. So you know, that's

576
00:39:28.559 --> 00:39:30.880
that's the key thing now is actually
how do you get the right team on

577
00:39:30.000 --> 00:39:34.760
board? How do you accelerate with
this stuff? Don't don't wait, You've

578
00:39:34.800 --> 00:39:37.000
got to get your arms around it. So yeah, at least the planning

579
00:39:37.400 --> 00:39:42.159
side of things right, You've got
to get the c suite and key directors

580
00:39:42.840 --> 00:39:47.079
on board to recognize that this is
a direction for almost everyone for the future.

581
00:39:47.119 --> 00:39:51.280
In me, to not use this
stuff is a bit crazy. I

582
00:39:51.360 --> 00:39:53.920
think you're just going to get steamrolled
if you don't. It's like the dot

583
00:39:54.000 --> 00:39:58.480
com age when the dot com came
out and everyone realized, oh wow,

584
00:39:58.679 --> 00:40:00.480
so now I can sell everything online. I don't have to have all these

585
00:40:00.519 --> 00:40:05.199
brick and mortar stories. Well that
was a pretty big change in terms of

586
00:40:05.280 --> 00:40:08.440
how we view the internet, and
now you have this new, incredibly powerful

587
00:40:09.079 --> 00:40:13.760
set of technologies. And to your
point, you can use different models.

588
00:40:13.840 --> 00:40:17.159
There's the chat GBT, there's barred
you know, GitHub as co pilots,

589
00:40:19.079 --> 00:40:22.719
you know, Microsoft, they all
have their own version, if you will.

590
00:40:22.400 --> 00:40:27.599
So you have to start somewhere,
so at least start the planning process,

591
00:40:27.679 --> 00:40:30.920
get people to sit down and understand
what's possible. And the best way

592
00:40:30.920 --> 00:40:34.559
to do that on an individual basis
is just to play with this stuff,

593
00:40:34.639 --> 00:40:37.000
right, Just get in there and
start playing it, giving it prompts.

594
00:40:37.239 --> 00:40:39.639
Ye see it to create your lab, create, you know, create,

595
00:40:39.880 --> 00:40:45.199
create your you know, your your
data lab. But it's it's it's reinforcing

596
00:40:45.280 --> 00:40:51.719
that type of innovation lab. And
you know what we're quite impressed with actually

597
00:40:51.760 --> 00:40:53.280
what Meta are doing with Lama too, you know, that's what we've been

598
00:40:53.320 --> 00:40:59.559
pretty impressed with that as a as
a foundational capabilities because they've been you know,

599
00:40:59.599 --> 00:41:01.800
we see been through some quite interesting
you know, pr the last few

600
00:41:01.840 --> 00:41:07.119
years about data privacy and preserving that
kind of dimension. But you know,

601
00:41:07.199 --> 00:41:09.079
we think that we think they've done
a really good job around the stuff they've

602
00:41:09.079 --> 00:41:13.320
been doing around armors. So I
think that's something to think about as you

603
00:41:13.440 --> 00:41:15.880
you know, you put that reference
architecture in place, and you get these

604
00:41:15.960 --> 00:41:17.320
right people in the room, and
you get on with a sprint, you

605
00:41:17.360 --> 00:41:22.480
know, get some value done in
and for a few weeks get that into

606
00:41:22.519 --> 00:41:27.039
the back into the board and show
show what's possible with this stuff, and

607
00:41:27.159 --> 00:41:31.360
that that's the way you're going to
transform your industry rather than being transformed or

608
00:41:31.760 --> 00:41:37.000
put in a difficult position where your
competitors are actually doing something. You know,

609
00:41:37.079 --> 00:41:39.440
we were we were looking as you
know we were looking at recently,

610
00:41:39.679 --> 00:41:45.079
was you can use these types of
tool to do competitive analysis and identifying you

611
00:41:45.159 --> 00:41:50.280
know, what the what the strengths
and weaknesses of your cells and competitors that's

612
00:41:50.320 --> 00:41:53.239
been indexed by these types of model. That's right. If you're not leveraging

613
00:41:53.280 --> 00:41:57.960
that kind of stuff, you're missing
out. Well. And I heard and

614
00:41:58.039 --> 00:42:00.039
I haven't verified this, but I've
heard it from a sources now and it

615
00:42:00.079 --> 00:42:07.280
makes complete sense that Google has been
quietly building a vectorization layer underneath g Drive

616
00:42:07.440 --> 00:42:13.280
and Gmail such that in your prompt
at BARB you can do at Gmail at

617
00:42:13.840 --> 00:42:21.159
Drive and it will then point to
your vectorized information and use that to generate

618
00:42:21.239 --> 00:42:24.320
its answer. And that is pretty
darn cool, I mean to already have

619
00:42:24.480 --> 00:42:28.519
it there and just to explain to
folks what these vectors are, what these

620
00:42:28.559 --> 00:42:30.880
embeddings mean. And embedding can be
any piece of information. It can be

621
00:42:30.960 --> 00:42:36.400
a press release, it could be
any number of things that you're using to

622
00:42:36.519 --> 00:42:40.559
build out the foundation for your personalized
engine, if you will. And then

623
00:42:40.599 --> 00:42:45.679
you turn it into mathematical or numeric
values basically, and that way you can

624
00:42:45.800 --> 00:42:52.119
number crunch and do comparisons and do
nearest neighbor and k meines distant differences basically,

625
00:42:52.199 --> 00:42:54.320
And that's how these things run.
And they're just doing all these calculations.

626
00:42:54.440 --> 00:42:59.719
These just tremendous amounts of calculations.
But you think just in terms of

627
00:43:00.000 --> 00:43:01.559
and them use cases. You know, I sit there and think to myself,

628
00:43:02.199 --> 00:43:07.760
MRIs. The MRI, the actual
file of someone's MRI is a massive

629
00:43:07.880 --> 00:43:09.760
file. It's huge. I mean, it barely fits onto a cd RAM.

630
00:43:10.559 --> 00:43:15.440
But if you vectorize it, you
probably collapse it by ninety nine percent

631
00:43:15.559 --> 00:43:19.400
in terms of size. And now
you can also analyze it more effectively.

632
00:43:20.000 --> 00:43:22.800
So this can solve all kinds of
problems, all the data storage issues that

633
00:43:22.880 --> 00:43:27.320
we have. We can now vectorize
this stuff and then put those big fat

634
00:43:27.440 --> 00:43:31.800
files in just ice cold storage.
That's very very cheap. Its depends what

635
00:43:31.960 --> 00:43:35.599
kind of industry you're in them.
I mean, we we're getting involved in

636
00:43:35.639 --> 00:43:39.039
a lot of healthcare stuff and synthetic
data. You know, actually you know,

637
00:43:39.159 --> 00:43:44.480
creating synthetic data sets which you've got
the characteristics of the actual data.

638
00:43:44.599 --> 00:43:46.039
That's where you've got to You've got
to go down that route as well.

639
00:43:46.119 --> 00:43:50.159
That's another dimension that people need to
think about. It's not just about privacy,

640
00:43:50.639 --> 00:43:52.679
right, it's actually about synthetic data
as well. Is an important you

641
00:43:52.719 --> 00:43:58.000
know principle you've got to work out
as well in certain industries because there's obviously

642
00:43:58.079 --> 00:44:00.320
a you know, a patient record, there's a there's a lot of you

643
00:44:00.360 --> 00:44:02.880
know, biometric data, et cetera, et cetera. You've got to work

644
00:44:02.960 --> 00:44:07.599
out how to manage that data at
scale as well. So that's that's another

645
00:44:07.679 --> 00:44:13.239
thing that you know, it's it's
a lifelong kind of thing to sort out.

646
00:44:13.320 --> 00:44:15.440
But I think you know, generally
of AI is giving the industry a

647
00:44:15.519 --> 00:44:20.079
leg up. It's how it's but
it's a you know, it's a ten

648
00:44:20.159 --> 00:44:22.599
headed hydra. How do you how
do you manage that in a smart way?

649
00:44:24.039 --> 00:44:27.519
A lot of maturity, but also
a lot of risk. So it's

650
00:44:27.639 --> 00:44:30.920
it's it's back to basics. It's
back to basics, but it's also back

651
00:44:30.000 --> 00:44:36.239
to you know, getting executive sponsorship
on this kind of stuff and actually getting

652
00:44:36.320 --> 00:44:38.840
the right people around the table and
then getting you know, try trial and

653
00:44:38.960 --> 00:44:43.599
error. You know sometimes in enterprises
that we see today is there's still a

654
00:44:43.719 --> 00:44:46.039
bit scared of failing. But you
know, you've got to do sprints.

655
00:44:46.079 --> 00:44:49.920
You've got to apply agile principles.
You've got to get to the point where

656
00:44:50.039 --> 00:44:53.000
you can you know, you're going
to learn something out of the back of

657
00:44:53.119 --> 00:44:57.559
something, doing something quickly, and
then you're you're going to iterate on that.

658
00:44:57.840 --> 00:45:01.599
And it sounds like an obvious way
of approaching it, but we still

659
00:45:01.639 --> 00:45:05.400
see some food factor around them.
Yeah, that's exactly right. Well,

660
00:45:05.400 --> 00:45:07.920
folks, look these guys up online
praxy Data Andrew Turner. You've been listening

661
00:45:07.960 --> 00:45:15.440
to Inside Analysis. All right,
folks, time for the podcast bonus segment

662
00:45:15.480 --> 00:45:19.679
here on Inside Analysis, talking today
with Andrew Turner of praxy Data, all

663
00:45:19.719 --> 00:45:24.639
about jen Ai and solving the dark
data challenge. So Andy take it away.

664
00:45:27.159 --> 00:45:30.519
Yeah, so we were talking about
the the situation with a lot of

665
00:45:30.599 --> 00:45:36.440
these models, and you know,
there's obviously a huge advantage you can get

666
00:45:36.440 --> 00:45:38.920
if you get this right, but
also sometimes they get it wrong. So

667
00:45:39.119 --> 00:45:42.880
you know, we've got this situation
where you've probably heard about this thing called

668
00:45:42.920 --> 00:45:45.719
hallucinations. You know, you've heard
the thing where we get an incorrect data.

669
00:45:45.920 --> 00:45:51.079
We're getting bias put into this data
set, and the thing is then

670
00:45:51.159 --> 00:45:54.440
and actually responses from the from the
from the platform. So you know,

671
00:45:54.519 --> 00:45:58.760
how do we keep learning this data, how do we keep improving this data

672
00:45:58.840 --> 00:46:01.599
accuracy and ensure that we get a
good output, a good quality output can

673
00:46:01.639 --> 00:46:05.400
rely on. The Other thing that
we've see in as well, is that

674
00:46:05.679 --> 00:46:12.719
there's this whole debate, raging debate
about business data with public products, public

675
00:46:12.840 --> 00:46:15.920
generator AI products. So you know, how do you make sure you're compliant?

676
00:46:16.079 --> 00:46:19.480
You know, if you're in a
regulated industry, you know, how

677
00:46:19.519 --> 00:46:22.599
do you deal with data leakage,
data theft? You know, how do

678
00:46:22.639 --> 00:46:24.840
you how do you understand that the
third party dat you may be integrating is

679
00:46:24.880 --> 00:46:30.719
actually syndicated and actually good to integrate
with your model? And how do you

680
00:46:30.800 --> 00:46:36.079
do with the usual things around you
know, risks around hacking. So you

681
00:46:36.159 --> 00:46:38.719
know, we've got a strong position
here that we believe that you've really got

682
00:46:38.800 --> 00:46:44.519
to build, you know, your
own private models, and that's where we

683
00:46:44.679 --> 00:46:46.960
see we can augment that with our
industry models to help you get a leg

684
00:46:47.039 --> 00:46:52.119
up even further. One of the
things that we see as have been a

685
00:46:52.199 --> 00:46:58.679
core competence of organizations and individuals and
teams within another organizations is this thing of

686
00:46:58.960 --> 00:47:04.119
what we're saying prompting so or prompt
engineering as people starting to call it,

687
00:47:04.599 --> 00:47:07.400
where you know, you develop this
kind of style to interact with the models

688
00:47:07.920 --> 00:47:10.800
and to get your desired output,
and you have to you know, you

689
00:47:10.840 --> 00:47:15.440
can actually then start to request a
certain way of interacting with it, even

690
00:47:15.559 --> 00:47:19.559
a certain character of how you interact
with it as well. You've got to

691
00:47:19.599 --> 00:47:24.559
provide the context. You've got to
understand that the temperature on that particular prompt

692
00:47:24.599 --> 00:47:29.800
that you're actually integrating as well,
and also make sure that you actually start

693
00:47:29.880 --> 00:47:31.880
to build a library up so you
build the value with library that you can

694
00:47:31.960 --> 00:47:36.239
use as a reference model a bit
like we're done with industry models from our

695
00:47:36.320 --> 00:47:40.920
data point of view, to help
you your organization evolve and become more competent

696
00:47:42.000 --> 00:47:45.400
over time. So you know,
looking at this kind of the magic would

697
00:47:45.400 --> 00:47:50.599
say of generative AI and how you
can combine that with you know, corporate

698
00:47:50.679 --> 00:47:53.000
data. You've got to look at
how do you integrate that so you know,

699
00:47:53.039 --> 00:47:55.800
you can use it individually, you
can go into that consumer version that

700
00:47:55.880 --> 00:48:00.800
chat, GBT, bar et cetera, and music like that. But obviously

701
00:48:00.880 --> 00:48:05.920
ultimately you're going to move from an
individual use case into an integrated business scenario.

702
00:48:06.840 --> 00:48:08.480
You know, this example of you
know, the kind of power you

703
00:48:08.559 --> 00:48:12.880
can do today. You know you
can actually start to say, okay,

704
00:48:13.000 --> 00:48:15.440
you know, asking these questions.
You know, how could I create a

705
00:48:15.519 --> 00:48:20.880
direct competitor to a particular company,
either your own company or a company that

706
00:48:20.960 --> 00:48:23.559
you've seen a preer on your radar. I'd yet you know, use the

707
00:48:23.639 --> 00:48:29.000
models to find out what their strengths
of weakness is doing interactive SWAT analysis,

708
00:48:29.480 --> 00:48:32.960
you know, dynamically with prompt engineering, what IP or trade secrets do they

709
00:48:34.039 --> 00:48:37.719
have or they have published that you
can actually then catalog understand, create a

710
00:48:40.000 --> 00:48:44.519
you know, a way of actually
dealing with that and creating a battle card.

711
00:48:44.679 --> 00:48:46.760
You know, how do you deal
how do you sell against that particular

712
00:48:47.760 --> 00:48:51.480
position they have in the market.
This is the kind of things where you

713
00:48:51.480 --> 00:48:54.840
can actually really use it from a
marketing and strategic point of view that really

714
00:48:54.920 --> 00:49:00.599
make a difference to your day job. You know, to talk about public

715
00:49:00.639 --> 00:49:01.800
and private, you know, so
that what that means is that you've got

716
00:49:01.880 --> 00:49:06.920
to build this competency internally. You've
got to look at you know, we

717
00:49:07.000 --> 00:49:10.559
talked about prompt engineering, we talked
about upskilling, we talked about helping not

718
00:49:10.639 --> 00:49:15.039
only train the models, but train
your organization evolve those roles that new organization

719
00:49:15.119 --> 00:49:17.880
structure of the future. But also
you've got to look at think about how

720
00:49:17.920 --> 00:49:23.000
do you manage that data at scale
and apply some kind of the privacy concepts

721
00:49:23.039 --> 00:49:27.960
of you know, compliance, security
and control. Because whether you're in a

722
00:49:28.039 --> 00:49:30.800
regulated industry a non regulated industry,
you've got to think about building your own

723
00:49:30.880 --> 00:49:35.519
model. So how do you put
these these little pieces of the recipe together

724
00:49:35.960 --> 00:49:38.320
in a way that's going to help
you get to that next level of capability

725
00:49:40.639 --> 00:49:44.519
In terms of the future of generative
A you know, what we see as

726
00:49:44.559 --> 00:49:47.119
a number of things here happening is
you know, with the point when this

727
00:49:47.400 --> 00:49:51.159
is already starting to happen where you
can actually interact with it through voice,

728
00:49:51.719 --> 00:49:55.039
you know, where you can actually
start to evolve this into a what we

729
00:49:55.159 --> 00:50:00.599
say in multimodeal capabilities where you can
combine you know, video with images,

730
00:50:00.840 --> 00:50:07.039
with texts, with audio, with
actual code, so you actually create this

731
00:50:07.159 --> 00:50:10.639
kind of multimodal type of output and
the evolved in that ultimately into some kind

732
00:50:10.679 --> 00:50:15.639
of platform that you can use internally
and that's you know, access to corporate

733
00:50:15.719 --> 00:50:20.559
and personal data, obviously respecting the
kind of you know, GDPR, CCPA

734
00:50:21.000 --> 00:50:23.159
and all different regulations you got across
the world. And the whole point is

735
00:50:23.199 --> 00:50:28.440
about boosting the performance of your organization. That you've got to think about think

736
00:50:28.480 --> 00:50:31.039
this through about not only from a
people point of view, but also how

737
00:50:31.079 --> 00:50:35.039
do you put the process in place
and how do you build the products and

738
00:50:35.199 --> 00:50:38.199
data platforms that you need to do
for the future, So, you know,

739
00:50:38.280 --> 00:50:42.519
finishing off of what we've do practice
data. As I mentioned, we're

740
00:50:42.519 --> 00:50:45.119
trying to make sense of this kind
of messy, murky dark data out there.

741
00:50:45.760 --> 00:50:49.079
There's been a lot of investment in
lots of different tools, lots of

742
00:50:49.079 --> 00:50:52.679
different platforms, but we're still seeing
a real huge problem. So our kind

743
00:50:52.719 --> 00:50:58.639
of area that we're focusing on is
really the domain knowledge and the industry libraries

744
00:50:58.639 --> 00:51:01.719
that we've built to allow you to
have discover data, you know, curate

745
00:51:01.880 --> 00:51:07.000
data at scale, you know,
have a curation as a service, and

746
00:51:07.079 --> 00:51:10.159
also have conversations with your data.
So that's we see as an important thing,

747
00:51:10.199 --> 00:51:15.679
whether that's structured, semi structured,
or unstructured data. And we're saying

748
00:51:15.719 --> 00:51:20.239
that we're effectively delivering that, you
know, a narrow type of capability,

749
00:51:20.280 --> 00:51:22.760
but being very deep, So not
trying to be a suite of products,

750
00:51:22.840 --> 00:51:25.599
not trying to be a Microsoft,
trying to be a Google, trying to

751
00:51:25.639 --> 00:51:30.320
be specifically focused on those industries we're
looking at looking at financial services, banking,

752
00:51:30.400 --> 00:51:32.840
insurance, looking at healthcare, look
at the government use cases, and

753
00:51:32.920 --> 00:51:37.960
ultimately one moving into telco and retail
going downstreaming from there. But we're trying

754
00:51:37.000 --> 00:51:42.920
to be part of that analytics and
insight value stream and also not only delivering

755
00:51:42.920 --> 00:51:46.400
industry libraries, but also delivering business
rules out of the box that you can

756
00:51:46.440 --> 00:51:53.000
apply to identify risks, identify opportunities. That's fantastic, both folks. Yeah,

757
00:51:53.079 --> 00:51:55.719
go ahead, one, let's lie
go ahead, so you know,

758
00:51:55.760 --> 00:52:01.039
it's very fast. I suppose a
big watershed moment is that on the sixth

759
00:52:01.119 --> 00:52:05.480
of November, there's going to be
the first Opening Eye Developer conference over in

760
00:52:05.559 --> 00:52:07.920
San Francisco, So it sounds like
something that people should consider going to.

761
00:52:08.760 --> 00:52:13.639
Obviously, it's going to be an
evolution of the GitHub co pilot, which

762
00:52:13.679 --> 00:52:17.079
has been you know, raging,
raging capability that people can coexist with developers.

763
00:52:17.719 --> 00:52:21.760
And I suppose the key thing is
about, as I said a bit

764
00:52:21.800 --> 00:52:24.400
earlier, skills, compensation and dealing
with that, you know, the food

765
00:52:24.440 --> 00:52:28.159
factor, the fear, uncertainty,
in doubt. I'm going to grab this

766
00:52:28.360 --> 00:52:34.480
initiative by both both with both hands
and see how you can embrace it and

767
00:52:34.840 --> 00:52:37.880
apply it to your organization. A
few books for you to read and think

768
00:52:37.880 --> 00:52:42.039
about as you go forward. There's
a particular one that the one on the

769
00:52:42.119 --> 00:52:46.000
right hand side AI twenty forty one
that was written by the ex CEO of

770
00:52:46.159 --> 00:52:52.760
Google in China, Mustafa Suliman is
the original founder of deep Mind, which

771
00:52:52.760 --> 00:52:55.519
is now part of Google, and
a couple of other interesting things. If

772
00:52:55.519 --> 00:53:00.960
you've got some bedtime reading for you
to finish off. If you want to

773
00:53:00.239 --> 00:53:05.159
know more about praxy data, my
email address is at at practi data dot

774
00:53:05.239 --> 00:53:07.159
com and we'd love to hear from
you to speak to you about our use

775
00:53:07.199 --> 00:53:10.800
case and our capability. Thanks for
your time. Good stuff. Another deep

776
00:53:10.880 --> 00:53:15.719
dive on Inside Analysis. Look at
these guys up online at at praxidata dot

777
00:53:15.800 --> 00:53:17.880
com. We'll talk to you next
time you've been listening to Inside Analysis.

778
00:53:20.880 --> 00:53:30.639
Southern California's Inland Talk Express is kcaa
open for takeout and delivery. Hell Tapiac

779
00:53:30.800 --> 00:53:35.480
Mexican food restaurant in the Tri City
Center of Redlands is back. Their entire

780
00:53:35.559 --> 00:53:38.400
family is on hand to serve up
their delicious burritos, but chaka charizo,

781
00:53:38.679 --> 00:53:44.960
Wavosmanchero's steak and eggs just part of
their mouth watering great food. Since fifteen

782
00:53:45.119 --> 00:53:47.320
thirty one, people have marveled at
the miracle of El Tapiac, and now

783
00:53:47.400 --> 00:53:51.840
you can marvel at the great food
the logo family has been serving up for

784
00:53:51.920 --> 00:53:54.840
over two decades, nestled quietly in
the corner of the Tri City Center Shopping

785
00:53:54.880 --> 00:53:59.679
mall next to Burlington Code Factory.
Support them. They can't wait to serve

786
00:53:59.719 --> 00:54:02.800
you some of their delectable, authentic
South of the Border Mexican fair at great

787
00:54:02.880 --> 00:54:07.039
prices, served up with love.
Support the area's best loved Mexican food restaurant

788
00:54:07.199 --> 00:54:10.800
in these tough times. Order up
a tasting meal on the phone for delivery

789
00:54:10.920 --> 00:54:15.119
or take out for breakfast, lunch, or dinner from ten am to six

790
00:54:15.239 --> 00:54:19.079
pm called nine to nine three oh
seven zero zero one seven. That's nine

791
00:54:19.119 --> 00:54:24.239
to nine three oh seven zero zero
seventeen. Or google Altapiac Redlands and treat

792
00:54:24.320 --> 00:54:30.320
yourself to Hebelta Club's original pure pouty
Arco super tea helps build red corpuscles in

793
00:54:30.360 --> 00:54:35.199
the blood, which carry oxygen to
our organs and cells. Our organs and

794
00:54:35.280 --> 00:54:39.159
cells need oxygen to regenerate themselves.
The immune system needs oxygen to develop,

795
00:54:39.199 --> 00:54:44.360
and cancer dies in oxygen. So
the tea is great for healthy people because

796
00:54:44.400 --> 00:54:47.199
it helps build the immune system,
and it can truly be miraculous for someone

797
00:54:47.280 --> 00:54:52.039
fighting a potentially life threatening disease due
to an infection, diabetes, or cancer.

798
00:54:52.320 --> 00:54:57.280
The tea is also organic and naturally
caffeine free. A one pound package

799
00:54:57.280 --> 00:55:00.519
of tea is forty nine to ninety
five, which includes shipping. To order,

800
00:55:00.599 --> 00:55:04.599
please visit to ebot club dot com. To hebo is spelled T like

801
00:55:04.719 --> 00:55:08.079
tom, a, h ee b
like boy oh. Then continue with the

802
00:55:08.119 --> 00:55:13.320
word T and then the word club. The complete website is to hbot club

803
00:55:13.440 --> 00:55:16.519
dot com or call us at eight
one eight sixty one zero eight zero eight

804
00:55:16.599 --> 00:55:21.920
eight Monday through Saturday, nine am
to five pm California time. That's eight

805
00:55:22.000 --> 00:55:27.400
one eight sixty one zero eight zero
eight eight to ebot club dot com with

806
00:55:27.639 --> 00:55:34.800
sixty years of fascinating facts. This
is the man from Yesterday and back in

807
00:55:34.880 --> 00:55:39.119
time to this time in nineteen sixty
six. Following a Charlie Brown Christmas which

808
00:55:39.280 --> 00:55:45.119
debuted last December, Charles Schultz end
CBS will debut a second Charlie Brown Special.

809
00:55:45.360 --> 00:55:50.280
It's called the Great Pumpkin Charlie Brown. Hey, Charlie Brown, I've

810
00:55:50.360 --> 00:55:53.960
got a football. How about practicing
a few place kicks. I'll hold the

811
00:55:54.039 --> 00:55:59.000
ball and you come running and kick
it. Oh brother, I don't mind

812
00:55:59.039 --> 00:56:01.800
your dishonesty. You must think I'm
stupid. And back in time to this

813
00:56:02.000 --> 00:56:07.679
time in nineteen eighty eight after John
Housman passes. Although he had a long

814
00:56:07.760 --> 00:56:12.320
career starting with radio, John Housman
is best known as the feisty, krusty

815
00:56:12.400 --> 00:56:16.000
law professor in the paper Chase.
He was also known for this commercial Smith

816
00:56:16.239 --> 00:56:21.639
Bonnie. They make money the old
fashioned way. They earned it and from

817
00:56:21.639 --> 00:56:27.119
this time in nineteen fifty nine,
debuting on CBSTV A forgotten favorite, Mister

818
00:56:27.239 --> 00:56:30.880
Lucky. He's swab, he's sophisticated, He's a demon with Lady Luck and

819
00:56:30.480 --> 00:56:44.800
lovely ladies. The theme to Mister
Lucky is by Henry Mancini with more at

820
00:56:44.920 --> 00:56:52.119
man from yesterday dot Com. It's
that time of year again, No,

821
00:56:52.480 --> 00:56:57.119
not the holidays. Medicare open enrollment
and if you have questions about Medicare,

822
00:56:57.280 --> 00:57:00.199
you should talk to the local experts. Paul Barrett and Associates. All of

823
00:57:00.280 --> 00:57:05.320
his agents are certified with plans that
are accepted by most of the medical groups

824
00:57:05.360 --> 00:57:08.880
in our area. Call nine oh
nine seven nine three oh three eighty five.

825
00:57:09.039 --> 00:57:13.639
Their service is free and after forty
two years of the business, their

826
00:57:13.679 --> 00:57:16.519
agents are trained to help you pick
the plan that's right for you. Del

827
00:57:16.559 --> 00:57:20.199
Walmsley here, the first thing you're
going to have to learn is that until

828
00:57:20.239 --> 00:57:23.400
you stop expecting our politicians or anyone
else to change your life, your life

829
00:57:23.480 --> 00:57:27.519
isn't going to change. The only
person who can change your life is you.

830
00:57:27.960 --> 00:57:30.599
But you need to know how.
Listen to my show, The Del

831
00:57:30.639 --> 00:57:34.039
Walmsley Radio Show, where the Hype
ends and the help begins right here on

832
00:57:34.159 --> 00:57:38.320
CACAA, now broadcasting on ten fifty
AM and one oh six point five FM,

833
00:57:38.480 --> 00:57:50.880
the stations that leave no listener behind. Was your car involved in an

834
00:57:50.920 --> 00:57:54.519
accident or just need help with dents? All Magic Paint and Body Collision Centers

835
00:57:54.719 --> 00:58:00.559
in business for over thirty years.
They're highly trained staff and certified visions and

836
00:58:00.719 --> 00:58:04.599
friendly staff are the best in the
business and treat each car as if it

837
00:58:04.800 --> 00:58:07.400
was their own. All Magic Paint
and Body Collision Centers are family owned and

838
00:58:07.559 --> 00:58:12.800
offer state of the art equipment and
tools to ensure optimum results. They use

839
00:58:12.840 --> 00:58:17.159
the latest technology in computerized color matching
and specialize in frame repairs with their modern

840
00:58:17.239 --> 00:58:22.199
laser measuring systems. They're OEM certified
and they have four locations to serve you.

841
00:58:22.519 --> 00:58:27.719
All Magic Paint and Body Collision Centers
offer rental car assistance with free drop

842
00:58:27.800 --> 00:58:31.000
off and pickup services too, and
their work has a lifetime guarantee. All

843
00:58:31.039 --> 00:58:36.159
Magic Paint and Body Collision Centers are
in Narco, East Vale, Marino Valley

844
00:58:36.280 --> 00:58:38.920
and in Fontana. Call them at
one eight hundred and sixty one Magic.

845
00:58:39.159 --> 00:58:44.679
That's one eight hundred sixty one Magic. All Magic Paint and Body Collision Centers

846
00:58:44.880 --> 00:58:47.920
one eight hundred sixty one Magic,
All Magic paint and auto bodies, says

847
00:58:49.159 --> 00:58:58.639
drive carefully, Tune into The Faran
Dozier Show, US Marks Place in Time,

848
00:58:58.880 --> 00:59:04.400
the soundtrack to Life. Sunday nights
at eight pm are KCAA Radio playing

849
00:59:04.440 --> 00:59:08.440
the hottest hits and the coolest conversations. Sunday Nights at APM on The Ferrando

850
00:59:08.559 --> 00:59:14.440
Zier Show with in the array of
music, talk, sports, community outreach,

851
00:59:14.800 --> 00:59:19.800
and veteran resources, his hits from
the sixties, seventies, eighties,

852
00:59:20.360 --> 00:59:25.440
nineties, and today's hits. The
Ferrando Zier Show on KCAA Radio on all

853
00:59:25.599 --> 00:59:31.760
available streaming platforms and on six point
five FM and ten fifty AM. The

854
00:59:31.840 --> 00:59:49.840
Ferrando Zia Show on KCAA Radio,
Killstina KCAA, lovel Linda at one O,

855
00:59:50.000 --> 00:59:52.360
six point five FM, K two
ninety three c f Brito Valley,

856
00:59:54.679 --> 00:59:59.559
NBC News Radio. I'm Chris Kragio. The US is stepping up its military

857
00:59:59.599 --> 01:00:02.320
presence in the Middle East, including
air defense systems. Secretary of State Anthony

858
01:00:02.360 --> 01:00:07.559
Blincoln said that the US is taking
steps to avoid a broader regional conflict.

859
01:00:07.800 --> 01:00:08.480
Blincoln noted that includes making sure that