WEBVTT

1
00:00:00.240 --> 00:00:09.320
Blandam the stitution that needs no miss. You're behind being innovation as new business

2
00:00:09.359 --> 00:00:14.960
models, we invent every industry industry. Inside Analysis is your source of information

3
00:00:15.080 --> 00:00:19.320
and insight about how to make the
most of this exciting new era. Learn

4
00:00:19.399 --> 00:00:24.760
more at inside analysis dot Comside Analysis
dot com. And now here's your host,

5
00:00:24.839 --> 00:00:35.920
you, Eric Kavanaugh. Oh yes, ladies and gentlemen, Welcome is

6
00:00:35.960 --> 00:00:40.399
what the future is here? Already
itched yet? That's our line from future

7
00:00:40.439 --> 00:00:43.479
Proof Your host here, Eric Kavanaugh, the only coast to coast radio show

8
00:00:43.520 --> 00:00:48.240
all about the information economy, and
boy am I excited for today, folks.

9
00:00:48.320 --> 00:00:51.679
Last week we had some great guests, including Levine Rao, the CEO

10
00:00:51.759 --> 00:00:55.719
of Mosaic mL, which just got
acquired by Data Bricks for one point three

11
00:00:56.200 --> 00:01:00.719
billion dollars and one of the fastest
deals ever made in our industry. Quite

12
00:01:00.719 --> 00:01:03.920
frankly, really impressive stuff, and
Data Bricks is just on a tear.

13
00:01:03.480 --> 00:01:07.680
They also rolled out their own vector
database. We'll talk about those on the

14
00:01:07.719 --> 00:01:11.760
show today. They're doing amazing things
with their data warehouse and with data orchestration,

15
00:01:11.799 --> 00:01:15.719
and even this whole thing about Iceberg
and Delta Lake and Hootie and the

16
00:01:15.760 --> 00:01:21.680
different possibilities. They solved it by
just looking at the parquet files. So

17
00:01:21.840 --> 00:01:25.640
look all that stuff up. It's
a really impressive array of innovations coming out

18
00:01:26.000 --> 00:01:29.079
of Silicon Value these days, and
of course all around the world. But

19
00:01:29.200 --> 00:01:32.480
our guests today knows a lot about
all those things and also knows a lot

20
00:01:32.560 --> 00:01:36.799
about LMS, So it's going to
joke anyone remember data science. We used

21
00:01:36.799 --> 00:01:38.480
to talk about data science all the
time. Then we started talking about data

22
00:01:38.519 --> 00:01:42.879
match and of course data products,
and then it's large language models and AI

23
00:01:42.959 --> 00:01:48.000
has just taken the world by storm. Well, data science is still around,

24
00:01:48.280 --> 00:01:51.159
and we're going to talk about today
is really automating the hard part of

25
00:01:51.239 --> 00:01:55.840
data science to get to the value
quickly, so cost to value, time

26
00:01:55.920 --> 00:02:00.560
to value, get those really into
a nice happy place. That's going to

27
00:02:00.640 --> 00:02:04.319
make CEO is happy, it's going
to make customers happy. Or talk all

28
00:02:04.319 --> 00:02:08.319
about these large language models and ETL
for LMS. So with that, let

29
00:02:08.360 --> 00:02:13.840
me bring in Brian Raymond from Unstructured
dot Io. Brian, you folks are

30
00:02:13.879 --> 00:02:15.879
doing some really interesting things out there, and I'd like to help our audience

31
00:02:15.960 --> 00:02:20.879
kind of wrap their heads around the
magnitude of what's happening and I think you

32
00:02:20.919 --> 00:02:23.919
had mentioned to me that there's been
something like thirty billion dollars of investment in

33
00:02:24.080 --> 00:02:28.560
lms in the last six to eight
months or so, which is a pretty

34
00:02:28.719 --> 00:02:32.599
staggering number. That shows you that
the VC community sees where the rabbit is

35
00:02:32.680 --> 00:02:37.039
running now and that's where they're going, right. Yeah, it's an incredible

36
00:02:37.039 --> 00:02:39.479
shift, especially from where we were
twelve months ago, everyone expecting a recession,

37
00:02:39.639 --> 00:02:45.240
massive tech layoffs, contraction on innovation
budgets in the commercial sector, to

38
00:02:45.719 --> 00:02:49.360
the run that we've seen since the
beginning of the year on the heels of

39
00:02:49.719 --> 00:02:53.680
diffusion models first last summer and then
large language models, which is really sparked

40
00:02:53.919 --> 00:02:59.000
by the introduction of chat GPT last
November, but now in dozens and dozens

41
00:02:59.039 --> 00:03:04.199
of lms that are emerging every month. Yeah. And so our audience,

42
00:03:04.280 --> 00:03:08.639
our core audience who understand data warehousing
and analytics and business intelligence, well they

43
00:03:08.680 --> 00:03:14.840
all know ETL extract transform load.
That's been the mainstay of data warehousing for

44
00:03:15.000 --> 00:03:19.520
arguably almost forty years or so,
since the earliest days when we figured out

45
00:03:19.560 --> 00:03:23.159
that you could not run analytics on
enterprise systems like ERPs, for example.

46
00:03:23.159 --> 00:03:27.400
They're not designed to do that.
They're designed to move stuff around and get

47
00:03:27.439 --> 00:03:30.439
the job done. So we invented
this whole data warehouse concept ETLT. You

48
00:03:30.479 --> 00:03:36.520
move things, but you're doing a
very special kind of ETL and of unstructured

49
00:03:36.599 --> 00:03:39.039
data. So we talk all the
time about eighty percent of the data the

50
00:03:39.120 --> 00:03:45.520
enterprise is unstructured. It's in word
documents, emails, PDFs, PowerPoint presentations,

51
00:03:45.520 --> 00:03:49.919
etc. The structured data, well
that's the numbers of how much product

52
00:03:49.919 --> 00:03:54.639
we've sold, etc. What the
cost was. The context for the structured

53
00:03:54.680 --> 00:04:00.520
data comes from the unstructured data,
but historically it's been just of the analysts

54
00:04:00.639 --> 00:04:04.000
or the user to kind of piece
that together themselves. But now with these

55
00:04:04.039 --> 00:04:09.560
large language models, in particularly with
what unstructured dot io is doing, you

56
00:04:09.599 --> 00:04:14.599
can actually pull out this corpus of
context and you also kind of clean it

57
00:04:14.639 --> 00:04:18.240
and refine it and get it to
the textual value that's going to be used.

58
00:04:18.360 --> 00:04:21.519
Right, tell us about that and
how that works. Yeah, I

59
00:04:21.519 --> 00:04:27.399
mean we use the phrase ETL for
ELBMS. That's that's a metaphor that breakly

60
00:04:27.439 --> 00:04:33.639
it's probably something closer to connect transformed
stage and what that means in practice interests

61
00:04:33.639 --> 00:04:39.079
that we're building enterprise grade connectors kind
of like what five trend has done an

62
00:04:39.079 --> 00:04:43.920
airbyte, but that are built from
the ground up for files containing natural language

63
00:04:43.959 --> 00:04:48.199
dayler major repositories where those live.
So think amazon As three as your blob

64
00:04:49.120 --> 00:04:55.319
SharePoint. Wherever these these files contain
natural language, they are are generated and

65
00:04:55.360 --> 00:05:00.639
then save down so we're able to
grab those on the transform stage. Instead

66
00:05:00.639 --> 00:05:04.959
of mapping relational databases to one another, what we're doing is we're doing a

67
00:05:04.959 --> 00:05:11.319
couple of things. We're doing file
transformation first, so those pds, pptxs,

68
00:05:11.439 --> 00:05:15.920
xls, whichever the file extension might
be getting, extracting the natural enguage

69
00:05:16.000 --> 00:05:20.839
data out of it and rendering it
into a common JSON format typically and then

70
00:05:20.879 --> 00:05:25.920
on the staging side, getting it
ready for the analytics downstream. And so

71
00:05:25.959 --> 00:05:32.839
that might be tokenization, vectorization,
chunking scheme of mapping to the particular Jason

72
00:05:32.959 --> 00:05:38.680
scheme that's required given the task downstream. Either way, what's it has in

73
00:05:38.800 --> 00:05:42.040
common with ETL is that this is
data engineering, and this is a data

74
00:05:42.040 --> 00:05:47.240
engineering in service of data science of
machine learning and making this data available to

75
00:05:47.319 --> 00:05:51.560
models. And then we'll talk more
about how it's being made available through training

76
00:05:51.639 --> 00:05:56.920
or through vector databases. But that
critical step of how do you connect to

77
00:05:56.920 --> 00:06:00.480
that knowledge where it's being stored,
transform into a common format, and then

78
00:06:00.519 --> 00:06:05.560
get it ready is a very slow
and manual process still today in twenty twenty

79
00:06:05.560 --> 00:06:10.439
three. Yeah, and you had
mentioned something that I think is quite compelling

80
00:06:10.519 --> 00:06:14.800
that when you're extracting, let's say, from a word document or from a

81
00:06:14.839 --> 00:06:23.480
PowerPoint presentation, your technology has the
capacity to identify headers versus subheads versus body

82
00:06:23.519 --> 00:06:26.560
text, etc. And folks had
probably seen this now. And there are

83
00:06:26.600 --> 00:06:30.360
some great algorithms in the engines that
we use, like on Google and YouTube

84
00:06:30.360 --> 00:06:35.600
and other places where it'll automatically put
tables of contents together, or you look

85
00:06:35.639 --> 00:06:41.160
at something like otter ai taking notes
on your meetings and it'll say who said

86
00:06:41.199 --> 00:06:45.519
what when, and even distill that
for you and do some analysis dynamically for

87
00:06:45.600 --> 00:06:48.279
you, which is incredibly powerful.
I mean, you think about all the

88
00:06:48.319 --> 00:06:50.959
ways we used to take notes and
just try to keep track of what we

89
00:06:50.959 --> 00:06:54.720
had heard and go back to your
notes and explain, well, now it's

90
00:06:54.759 --> 00:06:59.079
all recorded and it's all translated,
and it's organized for you. And it

91
00:06:59.120 --> 00:07:01.439
sounds like that's kind of what you're
doing with the Sunstructured data is you're able

92
00:07:01.480 --> 00:07:06.079
to pull out the value the signal, which is once you want anyway,

93
00:07:06.279 --> 00:07:11.519
and then plug that signal into the
model to stage it somewhere to be able

94
00:07:11.560 --> 00:07:15.360
to leverage it for some analytical purpose. Rights. That's that's exactly right,

95
00:07:15.399 --> 00:07:20.560
and in practice we think about it
as both cleaning the data and so getting

96
00:07:20.600 --> 00:07:27.000
rid of unwanted artifacts off the OCR
something use a operable character recognition or undergo

97
00:07:27.040 --> 00:07:30.720
that file transformation process. There's almost
always just nasty artifacts that data engineers and

98
00:07:30.759 --> 00:07:33.920
data scientsts who are doing this data
engineering, you have to confine white space,

99
00:07:34.240 --> 00:07:38.160
unicode characters, sentence fragments, these
sorts of things, so that you

100
00:07:38.240 --> 00:07:41.639
got to clean it up. But
then you need to curate your data.

101
00:07:41.759 --> 00:07:45.199
And this is something that folks have
been thinking about a long time. When

102
00:07:45.199 --> 00:07:49.480
you have columns and rows but with
respect and natural anage processing. Suppose you

103
00:07:49.519 --> 00:07:53.360
have an analyst research report. Everyone's
kind of seen one of these states on

104
00:07:53.360 --> 00:07:58.639
Ford Motor Company over the last quarter. That analyst research report might GM.

105
00:07:58.680 --> 00:08:01.879
It may have a section, have
another section talking about supply chain issues,

106
00:08:01.920 --> 00:08:07.639
but just like a meandering body paragraph, that's actually or the security and that's

107
00:08:07.680 --> 00:08:11.120
what you want to push through your
machine learning model, your pipeline, or

108
00:08:11.160 --> 00:08:13.800
make available to your large language model. And so in that example, how

109
00:08:13.839 --> 00:08:18.160
do you take thousands of these that
all have different lived and you know,

110
00:08:18.240 --> 00:08:22.360
a dozen different file formats and curate
that natural English data and so you can

111
00:08:22.360 --> 00:08:26.279
park get somewhere to make it available
right now? The answer is, and

112
00:08:26.399 --> 00:08:31.240
historically the answer has been custom regular
expressions, Python scripts for every single document

113
00:08:31.320 --> 00:08:35.960
layout. So it's a slow,
painful process, and that's where we're Yeah,

114
00:08:37.080 --> 00:08:41.000
that's really remarkable. And you know, when I think about topics like

115
00:08:41.080 --> 00:08:45.240
information life cycle management, right like
information comes in, it gets used,

116
00:08:45.240 --> 00:08:48.879
it gets stored. I mean just
even the basics of storage, like when

117
00:08:48.919 --> 00:08:50.960
I win. Over the weekend,
I've been thinking about our conversation last week

118
00:08:52.000 --> 00:08:54.639
and thinking to myself, this is
big in so many ways. I mean

119
00:08:54.679 --> 00:09:00.960
just d duping for example, just
the number of duplicate documents that people have

120
00:09:01.080 --> 00:09:05.279
in large organizations. What's the average
like seven they say per documents, I

121
00:09:05.320 --> 00:09:07.159
mean, depending upon which study you
look at. So that's a whole lot

122
00:09:07.159 --> 00:09:11.799
of wasted space and wasted time and
effort, etc. And it's also just

123
00:09:11.919 --> 00:09:16.639
chaos there. I mean, basically
unstructured data is chaos right now in most

124
00:09:16.759 --> 00:09:22.200
organizations, and you are providing a
mechanism to extract a meaning from those environments

125
00:09:22.360 --> 00:09:28.320
and then feed them into these foundational
models or large language models or whatever.

126
00:09:28.159 --> 00:09:31.159
That is a huge deal. I
mean, it's almost the kind of thing

127
00:09:31.279 --> 00:09:35.840
and I joked about this when I
was writing an article about knowledge graphs that

128
00:09:35.360 --> 00:09:39.399
you know, so many organizations have
had to choose do we actually make sense

129
00:09:39.440 --> 00:09:41.840
of all the stuff that we have, or do we just focus on the

130
00:09:41.879 --> 00:09:45.919
now and do what's needed at the
moment. And of course they all do

131
00:09:45.960 --> 00:09:48.440
the latter and just sweep the rest
under the SharePoint, as I joked about

132
00:09:48.440 --> 00:09:52.200
it, but now you'll be able
to extract that value, and then you

133
00:09:52.200 --> 00:09:56.279
could do even interesting things about the
storage. Right once you have it out,

134
00:09:56.519 --> 00:09:58.919
if you're not required by law to
keep these things, you can drop

135
00:10:00.120 --> 00:10:03.480
all that's or at least put it
in super cold storage somewhere where it's not

136
00:10:03.519 --> 00:10:09.480
costing a lot of money. Point
being this is it's a revolution of information

137
00:10:09.559 --> 00:10:13.519
management. What do you think.
I think we've seen kind of three stages

138
00:10:13.559 --> 00:10:18.639
fold since last November and so,
especially with respect to large language models.

139
00:10:18.679 --> 00:10:20.320
The first was, hey, let's
just use what's in memory let's just ask

140
00:10:20.440 --> 00:10:24.799
chat GPT questions and see see how
far we get. And that was incredible.

141
00:10:26.360 --> 00:10:31.519
Um, and then around the December
January February timeframe, folks started saying,

142
00:10:31.519 --> 00:10:35.080
hey, what if we connected these
with external data right and made it

143
00:10:35.120 --> 00:10:39.200
available either by retraining models like bloom
which was one of the first open source

144
00:10:39.240 --> 00:10:45.159
elms, or putting in a vector
database. And so they grabbed data mostly

145
00:10:45.200 --> 00:10:48.960
off the Internet that was already in
a fairly clean format that you can get

146
00:10:48.960 --> 00:10:52.360
over API and made it available and
built some really cool applications on top of

147
00:10:52.360 --> 00:10:56.200
that. And then as we progress
throughout the spring, folks were like,

148
00:10:56.399 --> 00:10:58.720
you know, what if we're actually
going to deploy these things. If we're

149
00:10:58.720 --> 00:11:03.000
gonna you know, the my T
study, Goldman Sack studies showing the productivity

150
00:11:03.120 --> 00:11:05.639
promise of these, the productivity that, you know, the promise, the

151
00:11:05.639 --> 00:11:09.480
productivity gains with these, we're going
to have to take all of our private

152
00:11:09.559 --> 00:11:13.720
data and utilize the conjunction of these
models. Then it said, okay,

153
00:11:13.799 --> 00:11:16.320
we can't. It's not just what's
in memory. It's not just the data

154
00:11:16.360 --> 00:11:20.080
that's easy that's already out there.
It's the hard data that we're continuing to

155
00:11:20.120 --> 00:11:24.200
produce tens of thousands of files every
day. How do we do that?

156
00:11:24.279 --> 00:11:28.320
And that's the reality that the industry
as a whole is confronting kind of head

157
00:11:28.360 --> 00:11:31.399
on right now as we cool prototypes
and start to transition them into production.

158
00:11:31.840 --> 00:11:35.519
Right that's such a big deal.
And you had told me before the show

159
00:11:35.600 --> 00:11:39.759
too that you know, really there
are two paths that organizations can take,

160
00:11:39.799 --> 00:11:43.879
and one is to embed your data
that you chose to use or put it

161
00:11:43.879 --> 00:11:48.000
into these vector databases, and then
you've got a layer of abstraction. Basically,

162
00:11:48.039 --> 00:11:50.000
you've got the model and then it
reaches out to the vector database as

163
00:11:50.000 --> 00:11:54.279
it needs to. That strikes me
as the way to go. Frankly,

164
00:11:54.320 --> 00:11:58.120
it's sort of loosely coupled, if
you will, as opposed to tightly coupled.

165
00:11:58.399 --> 00:12:01.279
Tell us about that and how your
part to the left of either option.

166
00:12:01.039 --> 00:12:05.159
Sure, So, as you mentioned
kind of past, Number one is

167
00:12:05.200 --> 00:12:09.039
to either pre train or fine tune
a model with your data. And so

168
00:12:09.080 --> 00:12:15.759
this is where mosaic mL comes in. They're fantastic where customers of mosaic mL

169
00:12:15.799 --> 00:12:20.600
we've pre trained our own model using
them. And the objective there is how

170
00:12:20.639 --> 00:12:24.000
do you take your data and then
encode it into that into the memory of

171
00:12:24.000 --> 00:12:28.919
that model, and so it knows
it's seen and is familiar with and answer

172
00:12:28.960 --> 00:12:33.600
questions about your data. There's two
problems with that, or shortcomanies I should

173
00:12:33.639 --> 00:12:37.639
say why that's not sufficient in and
of itself. The first is recency latency

174
00:12:37.679 --> 00:12:41.440
of the data and so it's frozen
in time. And so what that requires

175
00:12:41.519 --> 00:12:46.600
is for you to periodically curate large
batches of new data, label it and

176
00:12:46.639 --> 00:12:50.120
then re encode it in. So
you're paying for label and class. You're

177
00:12:50.120 --> 00:12:54.799
paying for the compute costs and a
time and energy to run those projects that

178
00:12:54.840 --> 00:13:01.000
we've been in call it over the
last ten years. And it's still it's

179
00:13:01.039 --> 00:13:05.399
still required for a lot of the
cap like you know, the large lement

180
00:13:05.399 --> 00:13:09.320
based projects that we're doing. However, a new kind of path has emerged,

181
00:13:09.440 --> 00:13:15.279
and it's on the back of a
technique called retrieval augmented generation RAG or

182
00:13:15.960 --> 00:13:20.279
databases. And here what you do
is you park a bunch of your data

183
00:13:20.480 --> 00:13:24.200
in a vector database with vectorized embedding
representation. So think of it like a

184
00:13:24.240 --> 00:13:30.279
seventy digit string of numbers. It's
a multidimensional representation of that text. And

185
00:13:30.840 --> 00:13:33.919
as you prompt a model. What
it does. It looks at what it

186
00:13:33.919 --> 00:13:37.320
has in memory, but then also
reaches over and grabs relevant data from the

187
00:13:37.399 --> 00:13:43.399
vector database, considers it all together, and and produce a response that solves

188
00:13:43.399 --> 00:13:48.679
the second problem, which is hallucined
hallucinations. It's not a silver bullet.

189
00:13:48.720 --> 00:13:50.919
But what you can do is you
can anchor it. You can make it

190
00:13:50.000 --> 00:13:54.840
show homework and so you have chain
of thought reasoning. A buch great way

191
00:13:54.879 --> 00:13:58.120
to describe it. Yeah, I'm
right in this town. You can anchor

192
00:13:58.120 --> 00:14:03.000
it. That's exactly right. So
you're you're optimizing your chances of getting a

193
00:14:03.200 --> 00:14:07.440
real, governed, trusted answer out
of this thing, as opposed to hallucination.

194
00:14:07.480 --> 00:14:09.000
And you know, I'd mentioned to
you that when I was covering the

195
00:14:09.039 --> 00:14:13.960
Data Breaks conference to go to market, guy Adam something from chat GPT was

196
00:14:15.120 --> 00:14:18.840
engaging with me on Twitter, which
of course is now called xs. Elon

197
00:14:18.960 --> 00:14:22.320
took a page out of Mark Zuckerberg's
book to Arenas, but he was pointing

198
00:14:22.320 --> 00:14:26.080
out that hallucinations are a feature.
Note with these large language models, they

199
00:14:26.080 --> 00:14:30.759
are predictive engines, and their job
is to predict what text they think you

200
00:14:30.840 --> 00:14:33.720
want to hear based upon your prompts. Right, So they're really fusing vectors

201
00:14:33.720 --> 00:14:39.960
of information together dynamically in pros formats
to create content, either marketing content or

202
00:14:41.000 --> 00:14:45.559
business proposals or whatever. So it's
not going to go away completely, but

203
00:14:45.600 --> 00:14:48.519
to your point, to anchor it
like that's to give it tremendous weight and

204
00:14:48.600 --> 00:14:52.320
to basically gravitated toward the truth,
right, Yeah, and for it to

205
00:14:52.320 --> 00:14:56.559
be able to show that somewhere.
So say it produces h you know,

206
00:14:56.600 --> 00:15:01.759
a TPS memo, a page line
smml. Right, you should be able

207
00:15:01.840 --> 00:15:07.000
to drill down and on this paragraph, Um, you know what documents underlie

208
00:15:07.399 --> 00:15:13.039
the inference. You know that that
that generated that, and that's achievable using

209
00:15:13.080 --> 00:15:18.240
techniques like RAG and you know some
of the iterations of that. Yeah,

210
00:15:18.279 --> 00:15:20.639
I mean that's very very powerful.
And so you you do wind up getting

211
00:15:20.679 --> 00:15:26.279
into a conversation with this engine and
you walk through it and you help it

212
00:15:26.360 --> 00:15:28.960
understand things, and it helps you
understand things. It's a very bidirectional thing

213
00:15:30.000 --> 00:15:33.720
in that sense, right, because
you're trying to get to the bottom of

214
00:15:33.759 --> 00:15:37.039
something and you're trying to generate.
That's what they do. These are generative

215
00:15:37.080 --> 00:15:41.919
models. Generative AI generate stuff.
It creates new things, right, It

216
00:15:41.960 --> 00:15:43.919
absolutely does, and so, like
the it is a feature, not a

217
00:15:43.919 --> 00:15:48.600
bug. But at the same time, if you're depending on this for business

218
00:15:48.600 --> 00:15:52.399
purposes, you don't want a five
year old that's completely fabricating information. You

219
00:15:52.480 --> 00:15:58.600
want something something more like Encyclopedia Britannica
where it's written, but that you can

220
00:15:58.639 --> 00:16:03.000
trill down on sources inside and so. And there's a wide spectrum there in

221
00:16:03.000 --> 00:16:07.320
between what and what you can produce
using these models. Yeah, there is

222
00:16:07.440 --> 00:16:11.080
an absolutely wide spectrum, and it's
a learning process. But that's the other

223
00:16:11.080 --> 00:16:15.720
thing that gets me really excited about
this is that, especially when you train

224
00:16:15.799 --> 00:16:22.440
it using your own data, you
now have this very robust window pane into

225
00:16:22.679 --> 00:16:30.000
your organizations text and imagery and language
and messaging, so you can look for

226
00:16:30.039 --> 00:16:33.080
these things, and then that's a
whole discovery process, right And to me,

227
00:16:33.559 --> 00:16:37.279
discovery has always been one of the
most important parts of any sort of

228
00:16:37.279 --> 00:16:41.480
analytical process because you know, first
of course, you gather data, you

229
00:16:41.679 --> 00:16:45.840
organize data. In the old world, you had to come up with your

230
00:16:45.919 --> 00:16:49.600
data model early, your schema,
basically your start your snowflake or whatever you're

231
00:16:49.600 --> 00:16:55.000
gonna do Kimble versus in mind,
that kind of stuff, But you had

232
00:16:55.039 --> 00:17:00.399
to decide ahead of time that plumbing
was going to look like. Well,

233
00:17:00.440 --> 00:17:07.400
that's very fragile, and that's very
difficult to use in a fast changing world.

234
00:17:07.680 --> 00:17:12.000
And so what we have now is
a much more robust way to understand

235
00:17:12.039 --> 00:17:17.160
what the data is telling us.
Tell me what the model should look like,

236
00:17:17.240 --> 00:17:21.000
tell me what the processes should look
like. Still vet it, still

237
00:17:21.000 --> 00:17:23.599
make your own decisions, but it's
a much different and much more dynamic and

238
00:17:23.640 --> 00:17:26.079
flexible world. Folks, don't touch
that. I'll be right back. If

239
00:17:26.079 --> 00:17:33.000
you're listening to inside amountmals. What
if you could own a piece of the

240
00:17:33.039 --> 00:17:37.119
future. What if you could build
your next castle not on sand, but

241
00:17:37.240 --> 00:17:41.000
on the bedrock of a modern blockchain
ecosystem. The first Internet gold Rush made

242
00:17:41.000 --> 00:17:45.960
millionaires, the second wave is minting
billionaires, but the third wave is at

243
00:17:47.039 --> 00:17:51.039
anyone can get in on the Adding
to crowdpointtech dot com to learn how you

244
00:17:51.039 --> 00:17:56.240
can secure a foothold in the blockchain
revolution. Whatever your passion, wherever you

245
00:17:56.279 --> 00:18:00.599
want to go in life waiting you
right now. Go to crowdpoint tech dot

246
00:18:00.599 --> 00:18:04.119
com to learn how the blockchain who
will fuel the next generation of innovation in

247
00:18:04.160 --> 00:18:11.759
this globally connected world. That's crowdpointtech
dot com your trusted agent in an untrusted

248
00:18:11.799 --> 00:18:15.119
world. What's the longest running radio
show in the world focused on data?

249
00:18:15.359 --> 00:18:19.680
DM Radio? Want to be a
guest sometime? Send an email to info

250
00:18:19.720 --> 00:18:25.960
at dmradio dot biz. That's info
at DM radio dot biz. Can your

251
00:18:26.039 --> 00:18:30.039
ire financial crisis that our top economist
are saying is that are allocating a percentage

252
00:18:30.039 --> 00:18:34.480
of your IRA into physical gold and
silver. With a tax free rollover,

253
00:18:34.720 --> 00:18:40.079
you can diversify in, safeguard your
holdings from trigulent markets and echo up putting

254
00:18:40.079 --> 00:18:44.759
your IRA back on the gold standard. Find out how to safeguard your STS

255
00:18:44.920 --> 00:18:48.960
with a tax free rollover with a
Genesis Gold IRA, the only IRA that

256
00:18:49.000 --> 00:18:53.720
can hold physical precious metals. Call
now for your free gold and silver report.

257
00:18:53.920 --> 00:18:59.839
Protect your IRA today with one simple
phone call and learn how to qualify

258
00:19:00.000 --> 00:19:04.279
if we're up to ten thousand dollars
in free silver called Genesis Gold Group Empowering

259
00:19:04.400 --> 00:19:10.960
Faith Driven Stewardship. Eight hundred six
four four eight six one one eight hundred

260
00:19:11.119 --> 00:19:17.079
six four foot eight hundred six four
four that's eight hundred six four four eighty

261
00:19:17.119 --> 00:19:23.319
six eleven. When a player's sudden
cardiac event brought a national football game to

262
00:19:23.400 --> 00:19:27.880
a halt. It's shown a spotlight
on the importance of CPR readiness. Now,

263
00:19:29.079 --> 00:19:33.720
with youth sports in full swing,
the American Heart Association is rallying parents

264
00:19:33.759 --> 00:19:37.279
and coaches to be ready in an
emergency. To be ready learn hands only

265
00:19:37.279 --> 00:19:41.319
CPR. It's He'll am and learn
in minutes. Just visit Heart dot org

266
00:19:41.359 --> 00:19:48.039
slash hands only CPR. Hands Only
CPR is nationally supported by Elephant's Health Foundation.

267
00:19:48.839 --> 00:19:52.720
Each year, three hundred and fifty
thousand Americans died from a cardiac arrest.

268
00:19:52.839 --> 00:19:56.799
When seconds matter most, CPR can
be the difference in whether a friend

269
00:19:56.880 --> 00:20:03.680
or family member survives. That's why
the American Heart Association is challenging every household

270
00:20:03.720 --> 00:20:07.880
to elect at least one person to
learn a CPR. If you have ninety

271
00:20:07.920 --> 00:20:11.480
seconds, you can be your family
CPR hero. Just watch the American Heart

272
00:20:11.480 --> 00:20:17.839
Astiation tell at heart dot org and
become a hero. Do you need to

273
00:20:17.880 --> 00:20:21.920
get your hands on some extra money
right now? Maybe twenty five thousand or

274
00:20:22.000 --> 00:20:25.359
more. If you're a homeowner,
Now is a perfect time to get cash

275
00:20:25.400 --> 00:20:29.319
out. While homes in many neighborhoods
like yours have gone up in value.

276
00:20:29.519 --> 00:20:32.519
You can use the money for anything. It's yours. You can buy an

277
00:20:32.519 --> 00:20:36.960
investment property, payoff higher interest debt, or make home improvements if you need

278
00:20:37.000 --> 00:20:41.759
twenty five thousand, fifty thousand or
more. Now is the time home values

279
00:20:41.799 --> 00:20:44.680
are up, and so is your
equity. We offer you a way to

280
00:20:44.799 --> 00:20:48.359
use it. No need to use
your savings called New American Funding. Now

281
00:20:48.400 --> 00:20:51.920
and see how much cash out you
can get. Call eight hundred seven one

282
00:20:52.039 --> 00:20:56.160
h three seven three nine, eight
hundred seven one h three seven three nine

283
00:20:56.319 --> 00:21:02.519
one O three seven three nine.
That's eight seven one h thirty seven thirty

284
00:21:02.599 --> 00:21:07.119
nine NMLS sixty six h six www
dot MLS Consumer exit commitment to end subject

285
00:21:07.160 --> 00:21:11.160
to borrow. We're improperty qualifications,
not all of borrow, whereas we'll qualify

286
00:21:11.319 --> 00:21:19.319
terms. Welcome back to Inside Analysis. Here's your host, Eric Tavanaugh.

287
00:21:22.759 --> 00:21:26.200
Oh yes, folks, take us
to the future. Indeed, that's exactly

288
00:21:26.200 --> 00:21:30.039
what large language models are doing.
They're taking us to the future right now.

289
00:21:30.440 --> 00:21:33.160
We're going there right now, folks. It's very exciting stuff. I've

290
00:21:33.160 --> 00:21:37.799
played around with these. They are
very powerful and they're good to lots of

291
00:21:37.839 --> 00:21:41.039
different things. Like one of the
things they're very good at is summarizing,

292
00:21:41.119 --> 00:21:44.039
like you can give it one hundred
page document and tell it give me a

293
00:21:44.119 --> 00:21:47.400
two page summary of this, and
then dive into this, diving of that

294
00:21:47.799 --> 00:21:52.400
very very powerful research tool and content
creation tool as well. And of course,

295
00:21:52.480 --> 00:21:56.720
underpinning all this is an LP natural
language processing, which has been around

296
00:21:56.720 --> 00:22:00.039
for a very long time, decades, quite frankly, and it really didn't

297
00:22:00.079 --> 00:22:03.480
get anywhere for a long time,
and it was kind of stumbling. And

298
00:22:03.480 --> 00:22:07.519
then of course Siria came out and
Bixby and all these others, and we

299
00:22:07.640 --> 00:22:11.680
got better at that stuff. But
one of the things I've learned talking with

300
00:22:11.839 --> 00:22:15.880
unstructured here is that the dirty secret
of natural language processing these days is that

301
00:22:17.000 --> 00:22:22.440
many data science teams are still building
one off artisanal models and it takes a

302
00:22:22.440 --> 00:22:26.559
tremendous amount of time. It takes
a tremendous amount of effort, and by

303
00:22:26.559 --> 00:22:29.640
the time you're done, the world
has probably changed a little bit. So

304
00:22:29.680 --> 00:22:33.799
that's what's really changing, and that
affects the cost, that affects the viability,

305
00:22:33.799 --> 00:22:37.519
that affects what people try. Because
thinking about it in large organizations,

306
00:22:37.559 --> 00:22:40.960
man, you better have a good
feeling about the bet that you're going to

307
00:22:41.039 --> 00:22:45.200
put a million bucks on that it's
going to come to pass and generate value,

308
00:22:45.599 --> 00:22:47.880
and that's been a very difficult thing. So tell us a bit about

309
00:22:47.920 --> 00:22:52.640
that, Brian. It's good thing. So I would say that with any

310
00:22:52.799 --> 00:22:59.079
NLP project industry wide, you're looking
at about a twenty percent success rate,

311
00:22:59.160 --> 00:23:04.599
historically with enterprise about eighty percent failure
rate. And there's a number of challenges

312
00:23:04.599 --> 00:23:07.480
here. This isn't because the folks
were working on it didn't try hard enough,

313
00:23:07.880 --> 00:23:15.200
smart enough. You're challenged in two
different areas on the NLP side itself

314
00:23:15.359 --> 00:23:17.880
right before it, right at the
at the point of inference, and one

315
00:23:17.920 --> 00:23:22.880
is on the pre processing side.
So starting with like the NLP side itself,

316
00:23:22.200 --> 00:23:29.279
it's huge lead forward with the introduction
to transform based models over their predecessors

317
00:23:29.599 --> 00:23:30.960
in terms of the power of the
models. However, let me give you

318
00:23:30.960 --> 00:23:34.960
an example. He wanted to train
a model to extract people, like a

319
00:23:36.079 --> 00:23:40.400
named nity recognition model. You can
do that. Then if you wanted to

320
00:23:40.880 --> 00:23:45.519
have it identify folkses you know where
they work or where they live, or

321
00:23:45.599 --> 00:23:49.839
you know where they went to college, you need some relation extraction model for

322
00:23:49.920 --> 00:23:56.839
each one of those relations person and
college or person and towns and maintenance.

323
00:23:56.880 --> 00:24:03.799
So you end up with these Byzantine
model pipelines in order to produce the structured

324
00:24:03.920 --> 00:24:07.680
data that they labor intensive, and
so in terms of economics, it had

325
00:24:07.720 --> 00:24:11.240
to drive a tremendous amount of value
to make sense. However, if you

326
00:24:11.319 --> 00:24:18.920
introduced any complexity in terms of the
data feeding into that pipeline, compounding the

327
00:24:18.960 --> 00:24:23.039
economic challenges, and this is the
data engineering challenge that never received any BC

328
00:24:23.240 --> 00:24:26.920
money. It didn't receive from any
attention. This is where we're really focused.

329
00:24:26.920 --> 00:24:30.680
Which was all right, if you're
grabbing Twitter's API and you're feeding that

330
00:24:30.759 --> 00:24:34.200
in, Okay, you understand the
schema, you can map that schema,

331
00:24:34.279 --> 00:24:37.039
you can clean it, curate it, and send it in. It's a

332
00:24:37.119 --> 00:24:40.359
one time investment. That's good to
go. However, if you're wanting to

333
00:24:40.400 --> 00:24:44.160
do that on your data and taking
your emails and your power points and your

334
00:24:44.200 --> 00:24:49.599
memos and your call transcripts or any
other type of natural ers data and whatever

335
00:24:49.640 --> 00:24:56.119
formatity, any time the document layout
changed, anytime the file extension change,

336
00:24:56.440 --> 00:25:00.920
you needed to go and rebuild the
preprocessing pipeline completely costom every time. And

337
00:25:02.000 --> 00:25:06.400
so this meant that the amount the
cost of value and the time to value

338
00:25:06.680 --> 00:25:11.400
to get these up and running was
extremely expensive and also meant if anything changed

339
00:25:11.440 --> 00:25:15.559
at all, you have this in
this huge professional services bill that came your

340
00:25:15.599 --> 00:25:19.319
way in order to keep it running. Right, So LMS changed that first

341
00:25:19.359 --> 00:25:26.359
problem where you don't need these byzantine
model pipelines, but you still face that

342
00:25:26.440 --> 00:25:30.720
economic challenge on the preprocessing side,
and that's why we're focused there. That

343
00:25:30.799 --> 00:25:34.160
makes a lot of sense. And
you know, these transformer models. One

344
00:25:34.200 --> 00:25:38.640
of the things they allow you to
do is leverage this attention component, right,

345
00:25:38.680 --> 00:25:45.279
which is basically how much attention the
computer spends examining this versus examining that.

346
00:25:45.880 --> 00:25:48.079
And it doesn't take too much thought
to realize how important that is,

347
00:25:48.240 --> 00:25:53.759
right, because your brain has this
you you're constantly modifying what gets more attention.

348
00:25:53.920 --> 00:25:56.640
Someone just barked outside or barked dog, just bark, Oh, what's

349
00:25:56.680 --> 00:26:00.519
that? You can shift your focus
very very quickly. And these models,

350
00:26:00.559 --> 00:26:03.359
I mean, it used to be
that you just said go and it had

351
00:26:03.359 --> 00:26:07.680
to go through and do the whole
thing, no matter what. Now it's

352
00:26:07.720 --> 00:26:10.640
more like go, but I want
you to find this, and it starts

353
00:26:10.640 --> 00:26:14.119
down one path and goes I'm not
finding this, and it switches. So

354
00:26:14.160 --> 00:26:18.680
it has the capacity to change its
focus and attention. That's a really really

355
00:26:18.720 --> 00:26:21.839
big deal. Can you talk about
how that happens and what's actually happening on

356
00:26:22.000 --> 00:26:25.880
the cover. Yeah, So where
we were five years ago up to about

357
00:26:25.920 --> 00:26:29.880
a year ago is it would look
to the left and I look to the

358
00:26:29.960 --> 00:26:33.599
right, about two hundred to two
hundred and fifty tokens typically like a five

359
00:26:33.640 --> 00:26:37.559
to twelve token window token about two
thirds of a word. You can think

360
00:26:37.559 --> 00:26:41.559
about it that way, right,
and so it like it had blinders on,

361
00:26:41.799 --> 00:26:47.000
Think a horse with blinders on,
and it's just looking at these kind

362
00:26:47.000 --> 00:26:49.039
of chunks of text, and it's
looking left and looking right and hunting for

363
00:26:49.079 --> 00:26:56.559
answers and looking and evaluating that multidimensional
text space within there. With the introduction

364
00:26:56.599 --> 00:27:02.799
of large language models, what's been
really exciting is that that five twelve token

365
00:27:02.839 --> 00:27:04.799
window went to one thousand, went
to two thousand, and went to four

366
00:27:04.880 --> 00:27:11.519
thousand, and recently has gone to
one hundred thousand and beyond. And so

367
00:27:11.559 --> 00:27:14.680
what that means is that you can, as you mentioned earlier, you can

368
00:27:14.680 --> 00:27:17.599
give it one hundred pages attacks and
say please summarize this, because it can

369
00:27:17.640 --> 00:27:22.200
look at one hundred pages from thirty
thousand feet read everything at once and then

370
00:27:22.240 --> 00:27:26.880
take it all in together. Wow. One of the challenges that's emerged over

371
00:27:26.920 --> 00:27:30.839
the last three months, and some
really important research that's been done by folks

372
00:27:30.880 --> 00:27:34.319
at Stanford and elsewhere has been looking
at okay, we have these large attention

373
00:27:34.359 --> 00:27:41.400
windows, but is the first quarter
of a document or of that corpus considered

374
00:27:41.400 --> 00:27:42.680
the same as the second quarter as
a third quarter. And what it was

375
00:27:42.720 --> 00:27:45.759
finding was that there's this big donut
hole right in the middle that are looking

376
00:27:45.799 --> 00:27:49.799
at the beginning, the first ten
thousand or the last ten thousand tokens and

377
00:27:49.839 --> 00:27:55.119
would ignore most of the meddal thing. And so this is where that relationship

378
00:27:55.160 --> 00:27:59.200
with these vector databases comes in,
which is like, okay, let's not

379
00:27:59.359 --> 00:28:03.319
try and feed one hundred thousand tokens
at a time. Let's give it a

380
00:28:03.359 --> 00:28:06.319
more reasonable amount, but then give
it access to all of this data that's

381
00:28:06.559 --> 00:28:11.799
a vector database. It's cheaper,
so the inference cost is cheaper, and

382
00:28:11.839 --> 00:28:15.480
you're going to get better performance.
That's interesting too, and I'm guessing that

383
00:28:15.559 --> 00:28:18.720
you know the way the data is
persisted on the vector databases, there is

384
00:28:18.759 --> 00:28:22.440
some element of metadata. In other
words, this data is about the color

385
00:28:22.519 --> 00:28:26.279
of cars, This data is about
the speed of cars. This stata is

386
00:28:26.279 --> 00:28:30.839
about engines or etc. I mean, I even think about on parquet files.

387
00:28:30.920 --> 00:28:33.720
Right on a parque file, that
first bit of the parque file gives

388
00:28:33.720 --> 00:28:37.680
you like the men and the max
and some basic information. There was a

389
00:28:37.720 --> 00:28:41.519
company infobright that at a database which
did that kind of thing, and I

390
00:28:41.519 --> 00:28:45.599
think they unfortunately went away. But
you know, the parque file format was

391
00:28:45.680 --> 00:28:51.920
designed to facilitate analytics, and you
know data bricks solve that Iceberg hooty Delta

392
00:28:52.000 --> 00:28:56.039
Lake thing by just going to the
parquet files. So I'm just guessing there

393
00:28:56.079 --> 00:29:00.039
that in the mechanism that that to
persist the data in the vector database and

394
00:29:00.039 --> 00:29:03.920
then of course thee leves accessing it, there is some component where it can

395
00:29:03.960 --> 00:29:07.359
go, Oh he wants to know
more about colors of cars, I'll grab

396
00:29:07.400 --> 00:29:11.920
it from here. Is that about
right? Well, it's right in a

397
00:29:11.000 --> 00:29:14.599
roundabout sort of way, Eric,
and it's about it's a back to the

398
00:29:14.599 --> 00:29:18.160
future story. And Verry over at
Lama Index and Harrison Chase and others and

399
00:29:18.680 --> 00:29:25.200
of Metal are doing some really cool
work here. So call it this past

400
00:29:25.279 --> 00:29:30.839
February and before you were doing something
called nearest neighbor search so co signed similarity,

401
00:29:30.359 --> 00:29:36.000
So have a vectorized representation of your
prompt, Hey tell me how to

402
00:29:36.039 --> 00:29:38.839
make spaghetti right, and it would
victorize that, and it would look for

403
00:29:38.960 --> 00:29:45.720
similar vectorizing bettings in that vector database
and bring it back together. Where the

404
00:29:45.759 --> 00:29:51.039
tech is going now, especially in
the last quarter to two quarters, has

405
00:29:51.079 --> 00:29:56.400
been Hey, there's this whole world
over here of of data engineering in these

406
00:29:56.440 --> 00:29:59.799
parquet files and all of this meta
data that could be valuable. How do

407
00:29:59.839 --> 00:30:03.359
we do cosign similarity and all this
super you know, complicated math around nearest

408
00:30:03.440 --> 00:30:08.839
DA research and then also leverage that
metadata to produce even better inference. And

409
00:30:08.880 --> 00:30:14.599
so those those worlds are just now
coming together, you know, the spring

410
00:30:14.640 --> 00:30:19.400
and end of the summer. That's
crazy. Yeah, it's really it's really

411
00:30:19.440 --> 00:30:25.880
fascinating too to see how quickly things
are changing. So we had another really

412
00:30:26.079 --> 00:30:29.920
really cool two three months snowflake for
a while to Wild is now doing something

413
00:30:29.920 --> 00:30:33.680
with a relational AI, I think, And he was even telling me,

414
00:30:33.720 --> 00:30:37.759
he said, three months ago,
I would have told you a very different

415
00:30:37.759 --> 00:30:41.000
thing that I'm going to tell you
now, because that's how much advancement we've

416
00:30:41.000 --> 00:30:45.960
made in that short period of time. And that's what really is exciting,

417
00:30:45.960 --> 00:30:48.559
and there are lots of factors to
that, right. The open source movement

418
00:30:48.640 --> 00:30:52.319
is a big factor, big contributing
factor. Just in the mindset of committing

419
00:30:52.359 --> 00:30:56.359
code to public and then leveraging it
and being able to share so that we

420
00:30:56.440 --> 00:31:03.079
all collectively build these foundations and different
vendors can focus on their part speaks to

421
00:31:03.319 --> 00:31:07.000
tremendous success in the near future with
these technologies. What do you think,

422
00:31:07.640 --> 00:31:11.079
Well, it's funny you mentioned him. His most recent book, DA Data

423
00:31:11.240 --> 00:31:17.079
Preneurs just bought for everyone in the
company and gave it a copy to everyone

424
00:31:17.519 --> 00:31:21.680
at our offsite last week. Nice
folks. This is kind of a forty

425
00:31:21.720 --> 00:31:26.440
year view on the world that we're
at today, the importance of data for

426
00:31:26.599 --> 00:31:33.400
driving business outcomes, and that it's
the architecture that's evolving. Yeah, and

427
00:31:33.480 --> 00:31:36.839
the architecture is evolving. Mean,
that's kind of where I was going earlier

428
00:31:36.839 --> 00:31:41.880
in the show about how just having
an information strategy. So my European partner

429
00:31:42.000 --> 00:31:45.240
Eve Mulkers and I were going to
be working on this, and I've had

430
00:31:45.240 --> 00:31:48.079
this theory for a long time now
and we're going to try to crystallize this

431
00:31:48.160 --> 00:31:52.720
a bit. But every organization must
have and that can be focused on any

432
00:31:52.799 --> 00:31:56.119
number of different thing be one area
of focus understanding what are us? What

433
00:31:56.160 --> 00:31:59.880
kind of a team do we need
to put together to be able to let

434
00:32:00.359 --> 00:32:04.640
these things? And that's the job
of the Information Strategy Group, and you

435
00:32:04.680 --> 00:32:07.599
hash out the use cases. You
know, when we had Navine round the

436
00:32:07.599 --> 00:32:10.920
show last week, he said,
yes, I mean marketing, content creation,

437
00:32:12.400 --> 00:32:15.960
business proposals, the sales and marketing
in general is a big area of

438
00:32:16.039 --> 00:32:19.480
focus for us. It's going to
be the next couple of years, he

439
00:32:19.559 --> 00:32:22.880
thinks. And it's great because it
turns every writer into a ten X writer,

440
00:32:23.319 --> 00:32:27.359
just in the same way that rules
will turn a good engineer into a

441
00:32:27.359 --> 00:32:30.079
ten X engineer. And who doesn't
want a bunch of ten X writers and

442
00:32:30.160 --> 00:32:35.119
creators out there? It helps you
by giving you eighty percent of what you

443
00:32:35.200 --> 00:32:37.480
need, and then if you get
good at the prompt you can get to

444
00:32:37.519 --> 00:32:39.759
eighty five to ninety percent, and
you've got to fine tune it on top.

445
00:32:39.799 --> 00:32:45.680
But still, if every marketer walked
in every morning and had eighty percent

446
00:32:45.680 --> 00:32:50.279
of their traditional job done for them, that's a pretty good change in workflow

447
00:32:50.359 --> 00:32:55.640
and in productivity increase. Right,
It's not only that it's triggered a foot

448
00:32:55.720 --> 00:33:02.079
race, and it's a foot race
in terms of product within this economy that

449
00:33:02.079 --> 00:33:07.359
we're all operating again. And so
what you're seeing now is shareholder pressure and

450
00:33:07.400 --> 00:33:13.640
competitive pressure to adopt and integrate this
technology as quickly as possible because it's producing

451
00:33:14.119 --> 00:33:19.200
true quality and improvements and the work
that's done, but also the efficiency with

452
00:33:19.240 --> 00:33:24.200
which it's accomplished. And so I
think one of the most surprising things over

453
00:33:24.240 --> 00:33:28.839
the last year has been that a
lot of the you know, in the

454
00:33:28.839 --> 00:33:31.799
past decade that low skilled jobs we're
going to be the first impacted by AI,

455
00:33:32.200 --> 00:33:36.839
right, But we're seeing in practice
is that's actually high school that are

456
00:33:36.880 --> 00:33:42.519
impacted, but that's on everything sideways. And so goldman sacts are both extremely

457
00:33:42.559 --> 00:33:47.079
bullish, both extremely bullish on you
know, the effect that these models are

458
00:33:47.079 --> 00:33:51.599
going to have on growing the economy, but it's going to reshuffle it at

459
00:33:51.599 --> 00:33:53.640
the same time. Yeah, well, and we're also going to see some

460
00:33:53.759 --> 00:33:59.359
organizational changes too. I was talking
about this last week on the show in

461
00:34:00.319 --> 00:34:06.039
hierarchies in teams and how you build
teams and who's responsible for what in certain

462
00:34:06.079 --> 00:34:08.280
teams, you know, I mean, I think back to my old days

463
00:34:08.320 --> 00:34:14.960
in the print press world where you
had this whole linear process of laying out

464
00:34:15.000 --> 00:34:17.239
the paper and getting all the articles
that you're not going to be updating that

465
00:34:17.320 --> 00:34:22.280
document unless you're going to print another
eighty thousand and so as a very different

466
00:34:22.280 --> 00:34:24.599
world. And that's kind of like
what we are in this world today where

467
00:34:24.639 --> 00:34:28.559
the Web comes along. Now I
can publish something that I can change it

468
00:34:28.599 --> 00:34:30.400
instantly. There's a mistake, boom
changed, we got it out of there,

469
00:34:30.480 --> 00:34:34.039
right. Even if you think about
databases, how it used to be

470
00:34:34.119 --> 00:34:38.280
overwriting records and databases. Now we're
appending records so all. And I think

471
00:34:38.320 --> 00:34:43.599
we're going to see some real changes
in organizational structures that reflect that. Meaning

472
00:34:43.639 --> 00:34:47.239
you're gonna you know, people who
were just copywriting now they're doing more than

473
00:34:47.280 --> 00:34:51.360
copywriting. They're actually engaging more with
the salespeople to learn from them, or

474
00:34:51.360 --> 00:34:55.880
whatever the case may be. You're
going to see the traditional flow of work

475
00:34:57.000 --> 00:35:00.400
managing that I'm seeing Companies like Drift
for examp Ample are out what used to

476
00:35:00.400 --> 00:35:05.639
take three or four different people using
what used to take three or four different

477
00:35:05.639 --> 00:35:08.000
people. They're all doing it in
real time by scoring someone who comes to

478
00:35:08.000 --> 00:35:12.239
your website, creating some copy to
send them and all this stuff. It's

479
00:35:12.239 --> 00:35:15.559
like, wow, that's how that's
how day to day life changes. Right.

480
00:35:17.760 --> 00:35:21.840
One of the most forward leading organizations
I know folks probably won't believe me

481
00:35:21.880 --> 00:35:24.960
on this is actually the Pentagon and
the Intelligence organizations. And you know I

482
00:35:25.320 --> 00:35:29.519
have a background and intelligence and we've
been doing a lot of work with them

483
00:35:29.519 --> 00:35:34.920
from the very beginning. Analytic engines
in the form of lms that don't require

484
00:35:34.960 --> 00:35:39.880
as much retraining or fine tuning.
If you have your data available and provision

485
00:35:40.039 --> 00:35:45.159
like what we're helping to do and
what can you do, you can quickly

486
00:35:45.199 --> 00:35:50.079
build apps on top of that scaffolding. They're actually calling that AI scaffolding.

487
00:35:50.440 --> 00:35:54.039
Craig market a chief day to day
office, and they envisioned a future where

488
00:35:54.079 --> 00:35:59.079
you have these cheap, fast apps
that are quickly built on top of the

489
00:35:59.239 --> 00:36:02.800
data, and they analytic engines that
you're not having to spend years building this.

490
00:36:04.079 --> 00:36:07.239
You can have internal teams that are
doing this. And maybe do you

491
00:36:07.280 --> 00:36:12.679
own an annuity either fixed rate,
indexed or variable? Are you paying high

492
00:36:12.719 --> 00:36:15.800
fees and getting low returns? If
so, Annuity General would like you to

493
00:36:15.840 --> 00:36:20.679
have this free book to learn the
pitfalls and mistakes of buying an annuity.

494
00:36:20.760 --> 00:36:24.280
The Annuity Dues and Don'ts for Baby
Boomers contains the little known truths about annuities,

495
00:36:24.599 --> 00:36:29.199
like how to help reduce your fees
and increase retirement income. And it's

496
00:36:29.239 --> 00:36:32.280
free, that's right free. As
a bonus, we'll also throw in a

497
00:36:32.360 --> 00:36:37.320
free annuity rate report just for calling. We researched over one thousand annuities and

498
00:36:37.400 --> 00:36:43.679
summarized rates and benefits from financially strong
insurers. You get Annuity Dues and Don'ts

499
00:36:43.679 --> 00:36:47.599
for Baby Boomers and the Annuity Rate
Report, both absolutely free for calling Annuity

500
00:36:47.639 --> 00:36:52.719
General Today. Hurry supplies are limited. Call now eight hundred two four or

501
00:36:52.800 --> 00:36:59.119
five one six nine seven, eight
hundred two four, five one six nine

502
00:36:59.280 --> 00:37:04.199
seven, eight hundred two four or
five one six nine seven. That's eight

503
00:37:04.320 --> 00:37:08.599
hundred two four or five sixteen ninety
seven. Do you own a timeshare?

504
00:37:08.880 --> 00:37:13.639
We'll face the facts. You made
a mistake, you made a bad purchase.

505
00:37:13.920 --> 00:37:17.760
A timeshare is not an investment.
It's a money pit that continues forever.

506
00:37:19.000 --> 00:37:22.599
If you use your time share,
that's great. But if you don't

507
00:37:22.639 --> 00:37:27.440
and you want illegally get out of
your contract, call my friends right now

508
00:37:27.480 --> 00:37:31.320
at the timeshare Exit Hotline. They're
an experienced team of lawyers who help good

509
00:37:31.320 --> 00:37:36.760
people like you get out of a
timeshare contract that they just don't want.

510
00:37:37.079 --> 00:37:39.920
Don't throw away your money on maintenance
fees. Use it for things you really

511
00:37:39.960 --> 00:37:45.199
want. We can help you end
your time share contract and stop the money

512
00:37:45.280 --> 00:37:49.360
drain immediately. If you are ready
to move on with your time share,

513
00:37:49.639 --> 00:37:53.159
call our team right now. Cancel
your time share now with a free call

514
00:37:53.360 --> 00:37:59.800
six seven O five eight hundred two
ninezero six eight hundred two nine zero six.

515
00:38:00.039 --> 00:38:05.000
That's eight hundred two nine zero sixty
seven. Do you own a timeshare,

516
00:38:05.280 --> 00:38:09.280
We'll face it. Do you own
a timeshare, We'll face the facts.

517
00:38:09.480 --> 00:38:14.199
You made a mistake. You made
a bad purchase. A timeshare is

518
00:38:14.280 --> 00:38:17.800
not an investment. It's a money
pit that continues forever. If you use

519
00:38:17.840 --> 00:38:22.000
your time share, that's great.
But if you don't and you want illegally

520
00:38:22.239 --> 00:38:27.880
get out of your contract, call
my friends right now at the Timeshare Exit

521
00:38:27.960 --> 00:38:31.639
Hotline. They're an experienced team of
lawyers who help good people like you get

522
00:38:31.679 --> 00:38:37.119
out of a timeshare contract that they
just don't want. Don't throw away your

523
00:38:37.119 --> 00:38:39.800
money on maintenance fees. Use it
for things you really want. We can

524
00:38:39.880 --> 00:38:45.960
help you end your time share contract
and stop the money drain immediately. If

525
00:38:45.000 --> 00:38:50.000
you are ready to move on with
your time share, call our team right

526
00:38:50.000 --> 00:38:53.320
now. Cancel your time share now
with a free call eight hundred two eight

527
00:38:53.440 --> 00:39:00.239
nine O four one three eight hundred
two eight nine O four one three eight

528
00:39:00.320 --> 00:39:04.880
hundred two eight nine oh four one
three. That's eight hundred two eight nine

529
00:39:05.079 --> 00:39:08.480
zero four thirteen. If you served
in the Marine Corps, by now you

530
00:39:08.559 --> 00:39:13.079
know about You probably have a lot
of questions. We have some answers.

531
00:39:13.280 --> 00:39:16.320
You could be entitled to compensation.
Billions of dollars are being allocated to pay

532
00:39:16.320 --> 00:39:21.880
for damages to anyone stationed at Camp
Lejune during that time. Unfortunately, it

533
00:39:21.880 --> 00:39:25.400
appears that officials may have known the
contaminated water problem existed and did little to

534
00:39:25.440 --> 00:39:30.440
protect their men. The Semper five
code was not honored. If you or

535
00:39:30.559 --> 00:39:34.719
someone in your family has developed a
serious illness, including various forms of cancer,

536
00:39:35.079 --> 00:39:38.119
called this Camp Lejune Legal Support line
right now. You can't turn back

537
00:39:38.119 --> 00:39:42.840
the clock and change what happened,
but you can certainly call right now and

538
00:39:42.960 --> 00:39:45.760
learn your rights as a marine.
Here's the number. Call eight hundred two

539
00:39:45.920 --> 00:39:52.519
five four three two one eight eight
hundred two five four three two one eight.

540
00:39:52.840 --> 00:40:01.960
That's eight hundred two five four thirty
two eighteen. Welcome back to Inside

541
00:40:01.960 --> 00:40:10.800
Analysis. Here's your host Eric Kabinat
all right, folks, back here on

542
00:40:10.960 --> 00:40:15.480
Inside Analysis. Automating the hard part
of data science, and folks, that

543
00:40:15.599 --> 00:40:19.719
is what AI does. It automates
a lot of the hard stuff, the

544
00:40:19.840 --> 00:40:24.239
difficult things, but also the tedious
tasks, for example, gathering information,

545
00:40:24.639 --> 00:40:30.119
optimizing pipelines. Just been writing about
observability. We have our top twenty list

546
00:40:30.159 --> 00:40:34.440
of observability vendors out there. That's
all testiful because Google had a great vision

547
00:40:34.840 --> 00:40:38.599
about container orchestration about really the next
generation of the foundation of enterprise software is

548
00:40:38.639 --> 00:40:42.480
what we're talking about. But right
there at the end of the last segment,

549
00:40:42.800 --> 00:40:50.000
Brian, this next generation of really
fast and fast developed apps. That

550
00:40:50.599 --> 00:40:53.159
which is very exciting stuff because now, and it's interesting, I'm seeing this

551
00:40:53.199 --> 00:40:59.440
as a trend in the marketplace of
AI infused apps or analytics infused apps,

552
00:40:59.719 --> 00:41:05.119
or you're not just hitting my sequel
database or you know, a Maria DP

553
00:41:05.360 --> 00:41:10.559
or whatever. Instead you're hitting some
of these models and thus can can leverage

554
00:41:10.599 --> 00:41:15.840
the power of analysis, the power
of AI very very quickly. That's like

555
00:41:15.880 --> 00:41:20.440
a Cambrian explosion right now, right
It's one of the biggest stories of the

556
00:41:20.480 --> 00:41:23.760
last six months has been the fact
that this is not just a data science

557
00:41:23.800 --> 00:41:28.679
and machine learning engineer world anymore.
Directly on top of these LEM architectures very

558
00:41:28.719 --> 00:41:32.519
easily, and we've first started in
March with a huge demand for markdown and

559
00:41:32.599 --> 00:41:38.000
other file types in the length chain
ecosystem, one of the key kind of

560
00:41:38.079 --> 00:41:47.039
orchestration tools in the LM space,
and since then it's unlocked thousands and thousands

561
00:41:47.239 --> 00:41:53.239
of new users space for thousands thousands
of new users, entities that hadn't talked

562
00:41:53.280 --> 00:41:58.400
to each other. Um. You
know, you had the data scientists that

563
00:41:58.400 --> 00:42:00.440
we're working on everything under the hood, and they'd hand it often. Then

564
00:42:00.440 --> 00:42:04.000
you'd have some React developers build on
top it. And now you're having build

565
00:42:04.000 --> 00:42:06.840
directly on these element engines, and
it's really exciting to see what they're coming

566
00:42:06.880 --> 00:42:09.599
up with. That's very cool too. And we were amusing in the break

567
00:42:09.840 --> 00:42:15.480
about my observation over the last five
to seven years as data science has really

568
00:42:15.480 --> 00:42:19.719
taken route and gotten a lot of
investment, a lot of activity. I

569
00:42:19.840 --> 00:42:22.559
told you that my experiences those teams
never talk to the data management teams,

570
00:42:23.000 --> 00:42:28.000
which is like, what, like, why would you have your data warehousing

571
00:42:28.039 --> 00:42:30.960
people not talking to your data science
teams? And that's the answer. The

572
00:42:30.960 --> 00:42:35.440
short answer as well, it's different
projects that they're working on, but from

573
00:42:35.440 --> 00:42:40.480
an organizational cohesion perspective, you want
these people talking to each other and working

574
00:42:40.519 --> 00:42:45.199
on the same things because you could
be That's the beauty of open source,

575
00:42:45.280 --> 00:42:49.519
right is you don't have to always
rebuild and reinvent wheels because that's what companies

576
00:42:49.519 --> 00:42:52.920
were doing is reinventing wheels every day. And I think the excitement and the

577
00:42:53.000 --> 00:42:59.679
power of these models is really going
to force new conversations to get people to

578
00:42:59.679 --> 00:43:02.920
talk to each other and get a
much more cohesive approach to what we're doing,

579
00:43:02.960 --> 00:43:07.440
which is going to save money,
save time, and do my favorite

580
00:43:07.440 --> 00:43:09.960
thing, which is improved morale of
your workforce. What do you think,

581
00:43:12.039 --> 00:43:15.039
Well, a lot of the you
know, aiml in issues with thein large

582
00:43:15.119 --> 00:43:20.119
organizations have been computer versions and looking
into syboy lines. You're looking at crops

583
00:43:20.400 --> 00:43:23.880
IoT data right and doing large ternal
analysis on that time serious analysis on that

584
00:43:24.920 --> 00:43:28.960
or kind of you know, really
in depth and LP projects and if you

585
00:43:29.000 --> 00:43:30.480
introduce new data to any of those. What you have to do you have

586
00:43:30.559 --> 00:43:34.639
to go get new label data to
retrain and rebuild a stack. You have

587
00:43:34.719 --> 00:43:37.440
to monitor it. So what you
have is you had these I don't want

588
00:43:37.440 --> 00:43:40.360
to call them stovepipe, but you
almost had these kind of experiments in an

589
00:43:40.360 --> 00:43:45.119
incubator, right, or these capabilities
that could just look at this little window

590
00:43:45.159 --> 00:43:50.519
of data and now what the model
is. Becoming as powerful as they are,

591
00:43:50.639 --> 00:43:52.400
you're able to bring to bear a
lot more data in different types of

592
00:43:52.480 --> 00:43:59.960
data to this analytic foundation, and
meaning that us as people right within organization,

593
00:44:00.559 --> 00:44:02.599
we're talking to folks that we haven't
talked to before. And so to

594
00:44:02.760 --> 00:44:07.320
your point, exact theory. And
this gets very interesting too because I think

595
00:44:07.320 --> 00:44:12.519
about the amount of heavy lifting that
has to be done to get certain projects

596
00:44:12.760 --> 00:44:16.079
afloat and get the analysis that you
want, get the insights that you need

597
00:44:16.119 --> 00:44:20.480
to make your decisions. And I
think what we're going to see here is

598
00:44:20.519 --> 00:44:23.800
a sparking out of all these different
ideas from different sources. We're gonna be

599
00:44:23.800 --> 00:44:29.039
able to get signal from places we
never thought we could get signal because people

600
00:44:29.039 --> 00:44:32.039
are going to be able to see
it in the models and the results and

601
00:44:32.079 --> 00:44:36.599
what they get, you know,
seeing is believing. One of my favorite

602
00:44:36.599 --> 00:44:40.880
philosophers, Wittgenstein, he wrote this
book, Tractatus Logico Philosophicus, which was

603
00:44:40.960 --> 00:44:45.639
really a big joke almost on himself
and on the industry, because you get

604
00:44:45.679 --> 00:44:46.400
to the end of the book and
he said, yeah, everything I said,

605
00:44:46.400 --> 00:44:50.960
forget about it. He says.
In fact, he said, use

606
00:44:51.000 --> 00:44:52.679
it as a ladder to climb up
into the clouds, and then kick the

607
00:44:52.719 --> 00:44:58.920
ladder away, and then you have
reached enlightenment. Now you understand things.

608
00:44:58.920 --> 00:45:02.039
And one of his big points was
some things cannot be said, they must

609
00:45:02.119 --> 00:45:07.519
be shown right and it's like,
wow, that's pretty interesting. And that's

610
00:45:07.559 --> 00:45:12.079
what is happening when people use these
models and start seeing what they can see

611
00:45:12.159 --> 00:45:16.599
into the details of things. My
buddy Steve Lucas from Booming was joking they

612
00:45:16.800 --> 00:45:20.719
connected an ALM to some of their
data and he said, show me all

613
00:45:20.760 --> 00:45:22.679
the different versions of order to cash
that we have right now, and it

614
00:45:22.760 --> 00:45:28.000
did and he was just like,
how did you do that? So this

615
00:45:28.119 --> 00:45:30.880
learning experience is going to be I
think very inspiring, yes, daunting,

616
00:45:31.280 --> 00:45:35.000
yes, a bit challenging or you
know, causing a little bit of fear.

617
00:45:35.360 --> 00:45:37.480
But I think on balance, people
are going to be like, oh

618
00:45:37.519 --> 00:45:39.960
my goodness, this is really powerful. Let's get the ball rolling. What

619
00:45:39.960 --> 00:45:44.159
do you think, Oh, ten
years ago we were talking a bit about

620
00:45:44.159 --> 00:45:46.079
big data. What we didn't have
to go as big data was big analytics.

621
00:45:46.440 --> 00:45:50.960
And you had the data, but
you didn't have a means for actually

622
00:45:51.159 --> 00:45:54.599
harness leveraging that data to answer the
questions that you really wanted answers to.

623
00:45:55.199 --> 00:45:59.079
Rather, you're just curious on that
you wanted to noodle on and pursue your

624
00:45:59.079 --> 00:46:02.480
curiosity. Do the things that make
you human, right, that are uniquely

625
00:46:02.519 --> 00:46:07.079
human. And now that we have
big analytics, like with LMS, you're

626
00:46:07.079 --> 00:46:09.519
able to do that. And to
see what folks were able to uncover is

627
00:46:09.679 --> 00:46:16.840
incredibly exciting and is unlocking entirely new
job job descriptions like prompt engineers who don't

628
00:46:16.840 --> 00:46:20.679
need to have any coding ability and
are getting paid three hundred thousand dollars at

629
00:46:20.679 --> 00:46:22.960
some places right out of the gate, right Well, and you know,

630
00:46:22.039 --> 00:46:25.079
just think about because I tell you, when I first started to playing with

631
00:46:25.119 --> 00:46:28.360
these things, I'm like, hmm, if you can write in French and

632
00:46:28.360 --> 00:46:30.199
German and English and different languages,
I bet I can write in coalball in

633
00:46:30.280 --> 00:46:35.800
JavaScript, and the short answer is
yes, yes it can. And now

634
00:46:35.840 --> 00:46:39.719
I think is in data bricks coming
out with their encoder or something or their

635
00:46:39.800 --> 00:46:44.440
decodes kind of fire. That's calling
it something that basically you can pump code

636
00:46:44.440 --> 00:46:47.760
into it and say what does this
code do? Right? Which was the

637
00:46:47.880 --> 00:46:52.880
big miss, the big issue with
coball developers because a lot of these old

638
00:46:52.079 --> 00:46:57.000
systems were written in coball and if
you don't have someone who knows Coball hitherto,

639
00:46:57.639 --> 00:46:59.920
you were just at the end of
the line folks, and you'd have

640
00:47:00.159 --> 00:47:02.880
to just work around it. But
now you can actually go in And this

641
00:47:04.159 --> 00:47:08.440
is what really excites me is somewhere
down the future being able to examine entire

642
00:47:08.599 --> 00:47:15.599
information architectures and understand the flow of
data and systems and what's doing what in

643
00:47:15.639 --> 00:47:21.719
which case and optimize that because that's
great for sustainability, for efficiency, for

644
00:47:21.920 --> 00:47:25.119
data quality, for process quality,
all these things. We're going to get

645
00:47:25.280 --> 00:47:30.239
really good at not wasting time.
Final thoughts from you. I think that

646
00:47:30.719 --> 00:47:36.800
what we're finding is that language really
is the native what you just described as

647
00:47:37.159 --> 00:47:38.920
code to text. We have image
attacks, we have speech to text,

648
00:47:39.440 --> 00:47:44.360
and we have tax to text,
right and I think the trick is going

649
00:47:44.400 --> 00:47:47.280
to be how do we actually make
this Like the last ten percent is going

650
00:47:47.320 --> 00:47:51.960
to be the hardest, and so
how do you make this production bread where

651
00:47:52.000 --> 00:47:53.760
you can trust it enough to hand
over the work that you don't want to

652
00:47:53.800 --> 00:47:58.960
do so that you can work on
the things that really are most rewarding for

653
00:47:59.039 --> 00:48:02.199
you and they are the most impactful
to the organization. We're at that moment

654
00:48:02.280 --> 00:48:07.159
right now where enterprises are just beginning
to engage engage with this, and that's

655
00:48:07.239 --> 00:48:12.280
on us as an industry to help
solve that problem. That's interesting, Yeah,

656
00:48:12.280 --> 00:48:15.519
and I think that you know,
companies do need to be practical and

657
00:48:15.639 --> 00:48:20.559
understand what are these tools supposed to
be used for. Hammers are used to

658
00:48:20.639 --> 00:48:23.679
hammer nails, screw drivers were used
to screw things in. You have different

659
00:48:23.719 --> 00:48:28.599
tools that have different purposes, and
it's going to be important for people to

660
00:48:28.639 --> 00:48:32.519
really understand what lms should do right
now and what they should not do.

661
00:48:34.000 --> 00:48:37.519
Like there's the fun case of the
guy who asked it to give him case

662
00:48:37.599 --> 00:48:40.239
law that it just made up and
it was all hallucinations. Well, don't

663
00:48:40.239 --> 00:48:45.400
do that. I think that is
not a proper use of the technology.

664
00:48:45.400 --> 00:48:47.679
It will get you in trouble.
But to your point, you've got to

665
00:48:47.719 --> 00:48:51.239
get someone look at it this stuff, folks, I promise you know,

666
00:48:51.639 --> 00:48:55.199
marketing people, you're curiosity people,
right that, the people who are naturally

667
00:48:55.320 --> 00:49:00.360
curious and want to know things.
That's who you want playing around with this

668
00:49:00.400 --> 00:49:04.679
stuff. That's who you want playing
around with these engines. To figure out

669
00:49:04.719 --> 00:49:07.960
the use cases and then get your
use cases and just you know, be

670
00:49:07.079 --> 00:49:12.039
methodical about it, but don't not
go down the road. Right, this

671
00:49:12.119 --> 00:49:15.880
is the kind of thing where everyone
is going to be leveraging these technologies in

672
00:49:15.920 --> 00:49:20.400
some way, shape or form,
and they're they're advancing so quickly. I'll

673
00:49:20.400 --> 00:49:22.719
give you one minute left. That's
one of the other really interesting things is

674
00:49:22.760 --> 00:49:25.840
how fast they're evolving and improving.
So the problem that was a big issue

675
00:49:25.960 --> 00:49:30.440
last month may not be an issue
anymore. Right, Yeah, that's exactly

676
00:49:30.559 --> 00:49:34.320
right. And so what we've seen
is, you know, folks were really

677
00:49:34.400 --> 00:49:37.320
worried about this becoming an oligopoly over
the winter months and just a couple of

678
00:49:37.320 --> 00:49:40.920
companies owning everything. Right, Well, you've seen you know, the teams

679
00:49:40.920 --> 00:49:45.880
and the capabilities that Navine and others
have built a mosaic and the good work

680
00:49:45.880 --> 00:49:50.320
on the architecture side have pulled the
cost down to building these dramatically to several

681
00:49:50.440 --> 00:49:52.760
hundred thousand dollars or several million dollars, and so from a capital investment standpoint,

682
00:49:52.880 --> 00:49:57.599
is more tractable. From an inference
side, that's that's an area where

683
00:49:57.639 --> 00:50:00.559
it's still really expensive and we need
to work on. But but it's it's

684
00:50:00.599 --> 00:50:06.360
far more accessible than it was twelve
months ago, and infinitely more accessible than

685
00:50:06.360 --> 00:50:08.800
it was when GPT three was first
introduced in June at twenty twenty. Yeah,

686
00:50:08.840 --> 00:50:14.039
that's a pretty fast innovation curve.
I would call that logarithmic. Just

687
00:50:14.159 --> 00:50:16.719
go walk, it's going like that. Will folks look these guys up online?

688
00:50:16.760 --> 00:50:22.960
Brian Raymond from Unstructured dot io Just
like it sounds unstructured dot io ETL

689
00:50:23.079 --> 00:50:25.599
for LMS. You're gonna be hearing
a lot more from these folks. You're

690
00:50:25.639 --> 00:50:30.480
gonna be hearing a lot more about
LM's folks sending email info and Inside analysis

691
00:50:30.559 --> 00:50:31.519
dot com. I want to know
what you want to know. We'll talk

692
00:50:31.559 --> 00:50:37.360
to you next time you've been listening
to Inside Analysis. Did you know that

693
00:50:37.480 --> 00:50:42.559
every year millions of men, women, and children are trafficked worldwide, including

694
00:50:42.639 --> 00:50:54.400
right here in the United States.
You know, pure pouty arc supert comes

695
00:50:54.400 --> 00:50:58.599
from the only tree in the world
that fungus does not grow on. As

696
00:50:58.639 --> 00:51:01.400
a result, it naturally has anti
fungal, anti infection, anti viral,

697
00:51:01.440 --> 00:51:07.440
antibacterial, anti inflammation, and anti
parasite properties. So the T is great

698
00:51:07.440 --> 00:51:10.079
for healthy people because it helps build
the immune system, and it can truly

699
00:51:10.079 --> 00:51:15.639
be miraculous for someone fighting a potentially
life threatening disease due to an infection,

700
00:51:15.719 --> 00:51:20.599
diabetes, or cancer. The T
is also organic and naturally caffeine free.

701
00:51:20.639 --> 00:51:23.400
A one pound package if T is
forty nine ninety five which includes shipping.

702
00:51:23.599 --> 00:51:28.679
To order, please visit to eebotclub
dot com. T Hibo is spelled T

703
00:51:29.079 --> 00:51:32.599
like tom, A H, E, B like boy oh. Then continue

704
00:51:32.639 --> 00:51:37.239
with the word T and then the
word club. The complete website is to

705
00:51:37.360 --> 00:51:40.800
hubot club dot com or call us
at eight one eight six one zero eight

706
00:51:40.920 --> 00:51:45.760
zero eight eight Monday through Saturday nine
am to five pm California time. That's

707
00:51:45.800 --> 00:51:51.519
eight one eight six one zero eight
zero eight eight toebot club dot com.

708
00:51:51.960 --> 00:51:54.719
It's time to make the Tri City
Center in Redlands a regular part of your

709
00:51:54.760 --> 00:52:00.079
weekly shopping experience. Tri City is
home to a wide assortment of quality businesses,

710
00:52:00.280 --> 00:52:06.719
including the all new Ocean Aquatics.
Check out their variety of exotic tropical

711
00:52:06.760 --> 00:52:10.599
fish along with fish food accessories and
tanks of all shapes and sizes. The

712
00:52:10.719 --> 00:52:15.400
Tri City Center is located just off
of Alabama and the Tennessee exits in Redlands.

713
00:52:15.760 --> 00:52:20.360
Visit the Tri City Center today and
find out why it's called the mall

714
00:52:20.480 --> 00:52:24.840
with a heart. When you're alone
and life is making you lonely, you

715
00:52:24.920 --> 00:52:32.719
can't always go Downtown. Nestled in
the heart of downtown San Bernardino is living

716
00:52:32.800 --> 00:52:37.679
history and the place you want to
be on the Internet. It's three twenty

717
00:52:37.679 --> 00:52:42.559
downtown dot com. That's three twenty
downtown dot com. The Enterprise Building with

718
00:52:42.559 --> 00:52:45.880
its rich interiors, it's a place
so special you just have to see it.

719
00:52:45.880 --> 00:52:52.079
It's a three twenty Northeast street in
downtown San Bernardino. The Enterprise Building

720
00:52:52.320 --> 00:52:55.480
is the heartbeat and entertainment life of
downtown San Bernardino, as well as a

721
00:52:55.559 --> 00:53:00.559
distinguished space for your new officer.
Building you can grow with its newly renovated

722
00:53:00.599 --> 00:53:05.400
banquet area, meeting rooms, three
twenty bar top deck, terrace, and

723
00:53:05.559 --> 00:53:10.639
plenty of parking space with over eighty
nine square feet of reasonable and available opportunity

724
00:53:10.760 --> 00:53:15.440
Today. It's family owned and operated
by Alicia Allen and their son Ryan.

725
00:53:15.679 --> 00:53:21.239
They've rolled out the red carpet and
crafted a geracious space, keeping the historical

726
00:53:21.280 --> 00:53:24.719
feel of the building while providing the
opportunity to create the future memories of your

727
00:53:24.800 --> 00:53:30.679
upcoming wedding or celebration. It's three
twenty downtown dot Com. That's three twenty

728
00:53:30.679 --> 00:53:45.079
downtown dot com the Enterprise Building.
It's Downtown waiting for you. This segment

729
00:53:45.159 --> 00:53:51.119
sponsored by the all new Rams Market, Rams Express car Wash and the Rams

730
00:53:51.239 --> 00:53:55.079
Ultimate One Stop Shop at twenty ninety
mentone Boulevard on the corner of Crafton Hills

731
00:53:55.119 --> 00:54:00.760
Avenue, just across from Jacento Farms. Rams is now open, all new,

732
00:54:00.960 --> 00:54:04.920
well stocked, bringing fresh and sharp
convenience to Mentone. With their clean,

733
00:54:05.000 --> 00:54:07.800
safe chemicals, they'll leave your car
all shiny and protected. When you

734
00:54:07.840 --> 00:54:12.639
buy gas inside the one stop shop, you'll get savings on your car wash.

735
00:54:12.679 --> 00:54:15.079
Make life easy and so we eat
it all with Rams. You can

736
00:54:15.159 --> 00:54:21.480
save with the Dinosaur and your top
tier Sinclair Fuel. Rams has great car

737
00:54:21.519 --> 00:54:25.400
wash membership deal and you can save
on gas when you wash your car satisfy

738
00:54:25.480 --> 00:54:30.559
your appetite when you drop in the
one stop shop convenience store. Smell the

739
00:54:30.679 --> 00:54:36.119
chicken delectable, mouthwatering, crispy,
crunchy fried chicken that is. RAMS is

740
00:54:36.159 --> 00:54:40.199
now open at twenty ninety Menton and
boulevarded Mentone. Thank you to RAMS for

741
00:54:40.239 --> 00:54:45.679
their community spirit and sponsoring this radio
station. Rams and Mentone is waiting for

742
00:54:45.760 --> 00:54:54.440
you. I always hear from our
clients who hired another firm that they wish

743
00:54:54.480 --> 00:54:59.000
they'd hire DNA Financial first. Don't
have regrets about your I R S tax

744
00:54:59.039 --> 00:55:00.880
case. Just hire the best in
the first place. One owed one hundred

745
00:55:00.880 --> 00:55:05.360
fifty thousand to the IRS and it's
spent thousands on another firm. We stopped

746
00:55:05.360 --> 00:55:07.719
the levies and negotiated a payment plan
and had their penalties forgiven. And while

747
00:55:07.719 --> 00:55:12.280
every case is different, we guarantee
that we'll find your perfect resolution and get

748
00:55:12.320 --> 00:55:15.199
it done right. For a free
consultation, call us at eight six six

749
00:55:15.280 --> 00:55:17.880
two zero one zero one five six. That's eight six six two zero one

750
00:55:19.079 --> 00:55:22.679
zero one five six. Then you
can say DNA dude right by me.

751
00:55:22.079 --> 00:55:28.599
You're listening to Cadre, where we
play all the best classic water wasting hits.

752
00:55:28.599 --> 00:55:32.119
Like watering your lawn during the hottest
time of the day and leaky fossils.

753
00:55:34.000 --> 00:55:37.119
There are tunes from another time because
now that we face a hotter,

754
00:55:37.199 --> 00:55:42.039
dryer future, every drop counts despite
an extremely wet winter. Keep using less

755
00:55:42.039 --> 00:55:45.159
water, fixing leaks, and reusing
indoor water for your garden. Visit save

756
00:55:45.239 --> 00:55:50.320
our water dot com for more ways
to conserve Up next, brushing your teeth

757
00:55:50.360 --> 00:55:55.880
with the water on through. It's
a bird, it's a plan. No,

758
00:55:57.280 --> 00:56:01.800
it's superraw. Okay, a gimmicky
opening for a commercial about super Roth

759
00:56:01.960 --> 00:56:07.320
universal life insurance, but I'm sure
it got your attention. Now what is

760
00:56:07.440 --> 00:56:12.079
a super roth, you ask.
It's a permanent indexed universal life insurance that's

761
00:56:12.239 --> 00:56:16.400
totally liquid and easily accessible once it
matures. Can be used to supplement retirement

762
00:56:16.400 --> 00:56:21.800
savings or a death benefit, or
both. Has no income or contribution limit,

763
00:56:22.039 --> 00:56:25.480
has no five year rule like roth
iras, has no ten percent penalty

764
00:56:25.519 --> 00:56:30.079
for accessing the funds before age fifty
nine and a half. Oh, and

765
00:56:30.159 --> 00:56:35.119
the average historical returns are five to
seven percent annually. Tax free. Super

766
00:56:35.199 --> 00:56:37.880
roths also lock in gains, which
means you don't lose your money when the

767
00:56:37.920 --> 00:56:44.360
market is down. Sounds incredible right
sounds. Super super Roths are the way

768
00:56:44.360 --> 00:56:47.480
of the future, specifically your future. To see if you qualify for a

769
00:56:47.519 --> 00:56:53.559
super Roth, go online to the
super Roth dot com. This segment is

770
00:56:53.559 --> 00:56:59.440
sponsored by press Print, Southern California's
best full service union printer and mailhouse.

771
00:56:59.559 --> 00:57:04.400
The offer the lowest prices around with
unmatched service and reliability and free delivery throughout

772
00:57:04.440 --> 00:57:07.960
So cal. Press Print can print
anything from letterhead, business cards, and

773
00:57:07.039 --> 00:57:12.280
campaign literature to mailers of any size, lawn signs, banners, doorhangers,

774
00:57:12.360 --> 00:57:15.719
or just about anything you might want. Press Print promises to save money for

775
00:57:15.840 --> 00:57:17.760
you, your business, or your
campaign. If you'd like to learn more,

776
00:57:17.960 --> 00:57:22.679
contact Mike Croombrin at press Print seven
one, four, three, nine

777
00:57:22.800 --> 00:57:28.800
nine eight, seven zero eight.
Get the Union Bug. You've eaten lots

778
00:57:28.840 --> 00:57:32.039
of great food and lots of great
food at restaurants. Cowboy Burgers in Fontana

779
00:57:32.320 --> 00:57:37.800
and now on Arlington and Riverside will
fast become one of your favorites with their

780
00:57:37.840 --> 00:57:43.719
delicious, mouth watering burgers and breakfast
burritos. Cowboy Burgers in Barbecue also serves

781
00:57:43.719 --> 00:57:47.719
fantastic smoke barbecue, baby back ribs, dry tip chicken pull Park sandwiches as

782
00:57:47.719 --> 00:57:52.320
well as lunch and dinner plates.
Everything is made from scratch, including their

783
00:57:52.320 --> 00:57:55.360
delicious side dishes like cole slap,
potato salad, barbecue beans, and much

784
00:57:55.440 --> 00:58:00.360
much more. Check out their rich, decadent chocolate brownie him food critic Gallenborgan

785
00:58:00.679 --> 00:58:04.679
and you can dine in, take
food out, or have them cater your

786
00:58:04.719 --> 00:58:09.159
next special event. I highly recommend
Cowboy Burgers and Barbecue at their new location

787
00:58:09.239 --> 00:58:14.840
at five five seven three Arlington Avenue
in Riverside. Just looked them up on

788
00:58:14.880 --> 00:58:19.679
the Internet. That's Cowboy Burgers and
Barbecue, happy eating and perfect for the

789
00:58:19.679 --> 00:58:24.800
holidays. Cowboy Burgers in Barbecue is
also available for catering. That's Cowboy Burgers

790
00:58:24.800 --> 00:58:32.719
and Barbecue in Fontana and now in
Riverside on Arlington. The Tri City Shopping

791
00:58:32.719 --> 00:58:37.400
Center in Redlands is serving up some
really cool ice cream at Lamicho Acana.

792
00:58:37.519 --> 00:58:42.719
Then get your chocolates and other delights
from Seas Candies. Moms and future moms

793
00:58:42.760 --> 00:58:45.280
who visit the mall can cool off
and relax while they get treated like royalty

794
00:58:45.320 --> 00:58:50.920
at Shiny Nails or Francis Nails and
then pampered at Texture Hair. The Tri

795
00:58:51.039 --> 00:58:53.519
City Center is filled with retailers who
care about you. Shop at the Tri

796
00:58:53.639 --> 00:58:58.400
City Center in Redlands and see why
they call it the Mall with a Heart.

797
00:59:00.920 --> 00:59:07.239
Now here's a new concept. Digital
network advertising for businesses. Display your

798
00:59:07.360 --> 00:59:12.639
ad inside their building. If a
picture is worth a thousand words, your

799
00:59:12.639 --> 00:59:17.239
company is going to thrive with digital
network advertising. Choose your marketing sites or

800
00:59:17.400 --> 00:59:22.920
jump on the DNA system and advertise
with all participants. Your business ad or

801
00:59:23.000 --> 00:59:30.119
logo is rotated multiple times an hour
inside local businesses where people will discover your

802
00:59:30.239 --> 00:59:37.480
company. Digital network advertising DNA a
novel way to be seen and remembered.

803
00:59:37.840 --> 00:59:43.159
Digital network advertising with networks in Redlands
and KAIPA call in the nine on nine

804
00:59:43.199 --> 00:59:47.599
area two two two nine two nine
three for introductory pricing. That's nine O

805
00:59:47.760 --> 00:59:54.440
nine two two two nine two nine
three for digital network advertising. One last

806
00:59:54.480 --> 01:00:00.280
time Digital network advertising nine O nine
two two two nine two nine three.

807
01:00:01.159 --> 01:00:08.360
Now here's a new concept. Digital
door on board casey AA's inlandok Express casey a