198: You're not ready for the next phase of data analytics

Speaker: 00:00:00

Avery Smith-1: You're not ready for the

next phase of data analytics because

2

: 00:00:03

there is a lot going on with AI right

now and it is impossible to keep up.

3

: 00:00:08

And I'm guessing that most of you

guys who are listening are not ready

4

: 00:00:11

for what's coming and I don't even

know if I'm ready for what's coming.

5

: 00:00:14

But in this episode, I will try to

explain what I see coming in the

6

: 00:00:17

near future with data analytics and

becoming a data analyst as well as.

7

: 00:00:21

Tell you how you can prepare yourself

for that future to best succeed, give

8

: 00:00:24

yourself the best chance of landing a

data job, getting promoted, and ultimately

9

: 00:00:28

succeeding in the data analytics field.

10

: 00:00:30

But if you're new here,

my name is Avery Smith.

11

: 00:00:32

I help people land their first data job.

12

: 00:00:34

I've worked with companies like

ExxonMobil, Harley-Davidson, hp, and a

13

: 00:00:38

lot of other companies help analyze data,

and now I make contents teaching people

14

: 00:00:41

about how to land their first data job.

15

: 00:00:43

Now, lemme tell you what's going

on with AI and why I think the

16

: 00:00:46

future, we're not prepared for it.

17

: 00:00:47

So AI is getting better every single

day at a lot of different tasks.

18

: 00:00:51

And I think the most recent groundbreaking

moments where I've been reading

19

: 00:00:55

a lot online, specifically in the

software development space on Twitter,

20

: 00:00:58

some people are calling it like a.

21

: 00:00:59

Gutenberg Grass Moment, it's Claude Code.

22

: 00:01:02

If you've never heard of Claude Code,

it's from a company called Anthropic.

23

: 00:01:05

They make a very similar product

to Chatt called Claude, but they

24

: 00:01:08

also have a programming version

that's called Claude Code.

25

: 00:01:11

And Claude Code is just like really good.

26

: 00:01:13

It's basically like an AI

programmers way you can think of it.

27

: 00:01:15

And they just recently released

what's called Claude Cowork, which is

28

: 00:01:18

supposed to be code for non-coding.

29

: 00:01:21

Task.

30

: 00:01:21

I've played around with it.

31

: 00:01:22

I haven't been super

blown away or shocked yet.

32

: 00:01:24

In fact, a lot of times it hasn't worked.

33

: 00:01:26

But a lot of developers are

pretty impressed with clot code.

34

: 00:01:29

It's probably the number one AI product

that's being talked about right now, and

35

: 00:01:32

people are using it to build all sorts of

different software, uh, a lot faster, a

36

: 00:01:35

lot quicker, a lot cheaper than you know.

37

: 00:01:37

Development has happened in the past,

and I think that data is a little

38

: 00:01:41

bit behind software in terms of the

adoption of ai, but I think that's

39

: 00:01:44

where we're going to in the future.

40

: 00:01:46

So down the road, maybe it's

Claude Cowork, I don't know.

41

: 00:01:48

I don't think it is, but there's

gonna be some sort of a tool that

42

: 00:01:51

can basically replace a data analyst.

43

: 00:01:53

Now when I say replace a data analyst, I

don't actually mean take a data analyst.

44

: 00:01:57

Job.

45

: 00:01:58

I see AI only as a tool that people

are going to use to do their jobs

46

: 00:02:02

better, and I'll explain why.

47

: 00:02:03

I think that's the case.

48

: 00:02:04

I'll make my argument and how really

AI just shifts how we work instead

49

: 00:02:07

of, I guess, how much we work.

50

: 00:02:09

Going back to this cloud code

thing, the biggest thing that I

51

: 00:02:11

think has happened is this is the

number one AI product on the marker.

52

: 00:02:14

Right now.

53

: 00:02:14

Everyone loves cloud code and recently

at the developer, the main guy for

54

: 00:02:18

Claude Code has revealed that all the

updates to Claude Code were actually.

55

: 00:02:22

Built by Claude Coate.

56

: 00:02:23

Now that's really meta, but basically

this AI tool is building itself.

57

: 00:02:27

Now, that's not to say that, that

there's not like a whole team behind it.

58

: 00:02:30

There definitely is, and humans are still

needed, but the idea that this number

59

: 00:02:33

one AI tool is actually built by the

number one AI tool is pretty impressive.

60

: 00:02:37

So I think this is a moment where we all

need to sit back as data analysts and

61

: 00:02:41

be like, what does the future look like?

62

: 00:02:42

And first off, I wanna say, I don't think

much is gonna change in the near future.

63

: 00:02:45

Companies are really slow to adopt ai,

like terribly slow to adopt anything

64

: 00:02:49

new, and it's gonna take a long

time to get inside of corporations

65

: 00:02:52

and actually get things to work.

66

: 00:02:53

So that's the first thing.

67

: 00:02:54

In the near future, I don't see a

whole lot changing necessarily, but

68

: 00:02:56

let's say five years down the road,

what does it actually look like?

69

: 00:02:59

And I don't think AI

is gonna take your job.

70

: 00:03:01

I don't think if you're trying to break

in the data analytics that you should,

71

: 00:03:04

you know, go somewhere else, try something

else, because AI is gonna take over.

72

: 00:03:07

I don't think that's the case.

73

: 00:03:08

I see it more of a, as like a

hammer, like a tool, and I think

74

: 00:03:11

it's going to change how we work.

75

: 00:03:12

Now, this has actually happened many

times before and unfortunately I'm not old

76

: 00:03:17

enough to remember a lot of them, right?

77

: 00:03:19

But like, obviously I'm shooting

this right now on my iPhone.

78

: 00:03:22

This episode, I'm recording

it on these wireless mics.

79

: 00:03:24

These didn't exist.

80

: 00:03:25

20 years ago, and now it completely

changes the way that we do video, that

81

: 00:03:28

we do content, those types of things.

82

: 00:03:29

Technology ends up just changing how

our job looks, not necessarily the

83

: 00:03:34

problems that we're actually solved.

84

: 00:03:35

Another example, I don't know if you

guys have seen the movie Hidden Figures.

85

: 00:03:38

I know there's a book, but basically it's

about these three African American women.

86

: 00:03:41

In the United side of the United

States that work for nasa, and

87

: 00:03:43

they're basically math computers.

88

: 00:03:45

They're hand doing math

calculations for space shuttle

89

: 00:03:48

landings and stuff like that.

90

: 00:03:49

Now, I haven't admittedly worked

for nasa, although one of my

91

: 00:03:52

students, uh, who graduated from

my bootcamp, landed a job at nasa.

92

: 00:03:56

So maybe we can ask him.

93

: 00:03:57

Evan, if you're listening, um.

94

: 00:03:59

I don't think they're doing like a lot of

hand calculations like at NASA right now.

95

: 00:04:03

Maybe they are.

96

: 00:04:04

Maybe they are.

97

: 00:04:04

Maybe.

98

: 00:04:04

I don't know how it is, but my guess

is they're using a lot of computers

99

: 00:04:07

and it's like these mathematicians,

let's just say that when computers

100

: 00:04:10

came out, did they lose their job?

101

: 00:04:11

No.

102

: 00:04:11

Their job just transferred from doing

the math calculations by hand to doing

103

: 00:04:15

the math calculations on a computer.

104

: 00:04:16

And that's honestly how I see the

future of data analytics going is that

105

: 00:04:20

data analysts might not be doing their

analysis in Excel or SQL or Python in

106

: 00:04:24

the future, but they'll be doing their

analysis in some sort of AI tool, some

107

: 00:04:27

sort of cloud code tool, some sort of

whatever AI tool you wanna, you know,

108

: 00:04:30

chat GBT interface to analyze their data.

109

: 00:04:32

And I don't think that those

tools are going to be able to

110

: 00:04:35

do things without the humans.

111

: 00:04:36

Now is cloud code programming itself?

112

: 00:04:38

Yes.

113

: 00:04:38

But there's supervision and that's

the big thing I wanna talk to you

114

: 00:04:41

is about the future of maybe every

job is less about doing the job.

115

: 00:04:46

And more about becoming

a little supervisor.

116

: 00:04:48

And I've heard the CEO of multiple

companies talk about this.

117

: 00:04:51

I'm forgetting the one where I

specifically heard this in some interview,

118

: 00:04:54

but basically like he sees individual

contributors now becoming like managers

119

: 00:04:58

to many AI services down the road.

120

: 00:05:01

And so instead of being individual

contributor, we're all becoming managers,

121

: 00:05:04

managing like little AI employees.

122

: 00:05:06

Is that going to happen?

123

: 00:05:07

I don't know.

124

: 00:05:07

But I definitely think that we are all

going to be doing less hands-on tasks.

125

: 00:05:12

We're going to be getting

AI a lot more of the tasks.

126

: 00:05:14

So our job becomes less of an instrument

player, more of a conductor, less

127

: 00:05:19

of a writer, more of an editor,

you know, more of a manager role

128

: 00:05:22

where we're actually like, we're

setting things up at the beginning.

129

: 00:05:25

Um, and it's really interesting because,

you know, five years ago when I quit

130

: 00:05:28

my, my data scientist job at ExxonMobil.

131

: 00:05:30

I was just an individual

contributor at ExxonMobil.

132

: 00:05:32

I was working on different AI

projects and it was a lot of fun.

133

: 00:05:35

I had a lot of fun.

134

: 00:05:36

I wasn't a manager at all.

135

: 00:05:37

I quit my job.

136

: 00:05:38

I started my own business, and over the

last five years we've grown quite a bit

137

: 00:05:40

to the point now where I have like a small

team of, let's just say five to 10 people.

138

: 00:05:44

All of a sudden, I'm a manager now and

I don't know what the heck I'm doing,

139

: 00:05:46

but it's really interesting because the

way I manage employees is also the way

140

: 00:05:50

I've realized that you need to manage

AI as well, and that's number one.

141

: 00:05:53

You need to set the right expectations.

142

: 00:05:54

You need to give them all the

resources upfront so that way they can

143

: 00:05:57

actually know what they need to do.

144

: 00:05:58

It's just really been an interesting

process where it's like at the beginning

145

: 00:06:01

you have to do a lot of work to set up

everything correctly, and at the end

146

: 00:06:04

you have to do a lot of work to make

sure that your employees did everything

147

: 00:06:06

correctly to your liking that they,

you know, didn't mess anything up.

148

: 00:06:08

And so it's like a lot of work at the

beginning to set things up, a lot of work

149

: 00:06:11

at the end to make sure everything went

well and some back and forth in between

150

: 00:06:14

to make sure that it stays on task right.

151

: 00:06:16

And I'm, I'm not trying

to liken employees, ai.

152

: 00:06:19

My point here is we're all

gonna have the mindset of being

153

: 00:06:21

conductors have the bigger vision.

154

: 00:06:23

And what that means for you specifically,

especially for those of you who are trying

155

: 00:06:26

to land your first data job, is the what

or rather, the how of doing data analytics

156

: 00:06:31

that we've been so focused on as like

a culture and a society for the last 10

157

: 00:06:34

years is gonna matter a lot less like the

tutorials of how to do things in Excel.

158

: 00:06:38

The tutorials in Power BI or

sql, they're gonna matter less.

159

: 00:06:41

I still think they're gonna be important.

160

: 00:06:42

I still think there's gonna

be a lot of data analysts.

161

: 00:06:44

In fact, basically my job at Exxon, this

is before AI even really existed, right?

162

: 00:06:50

My job at Exxon was to basically use

mathematics and machine learning to do

163

: 00:06:54

someone else's job, to do a trader's job.

164

: 00:06:56

So I worked on buying oil from

all around the world, right?

165

: 00:06:59

And in the past, historically, there

was just kind of a buyer, well, their

166

: 00:07:01

gut feeling and maybe some like stock,

like, oh, this stock's up so we're

167

: 00:07:05

gonna buy this oil, or whatever, right?

168

: 00:07:06

My job was to create math to make the

right decision on what oil to buy.

169

: 00:07:10

And then also another project

I worked on was where should we

170

: 00:07:12

send gasoline to around the world?

171

: 00:07:14

Like wherever you're living at

right now, your local ExxonMobil gas

172

: 00:07:17

station, how much gasoline is there

right now in like their storage?

173

: 00:07:20

That was my job.

174

: 00:07:21

And before, once again, it was like

a trader who would do that basically.

175

: 00:07:24

And my job was to use math

to replace those people.

176

: 00:07:27

It wasn't actually to replace those

people, it was to supplement those people.

177

: 00:07:30

Those people, their job

wasn't in jeopardy at all.

178

: 00:07:32

I was helping them create tools to

do their job faster and more accurate

179

: 00:07:35

and with more confidence, and that's

how I kind of see it being with AI as

180

: 00:07:39

well, is it's really just something

that's not gonna replace us, it's

181

: 00:07:42

just going to supplement our work.

182

: 00:07:44

What that means for you specifically is

like, it might not be as important to

183

: 00:07:47

know the difference between Index match

and Excel and a an X lookup like that

184

: 00:07:52

might not be as important down the road.

185

: 00:07:54

I think is really important and the

thing that I'm not prepared for, the

186

: 00:07:57

thing that you're probably not prepared

for and something that I really hope

187

: 00:08:00

to be doing more on this channel,

on this podcast and in my newsletter

188

: 00:08:04

is talk more about the why are we

doing this or the, what are we doing?

189

: 00:08:08

So not necessarily how to do

something, but the why and the what.

190

: 00:08:11

That is what I think is going to

be the most important thing down

191

: 00:08:14

the road, is knowing what to do

when not necessarily how to do it.

192

: 00:08:17

'cause I think AI is gonna know how to

do it, and I think we're gonna use AI

193

: 00:08:20

most of the time to know how to do it.

194

: 00:08:21

I still think it's really important

to learn the how to make sure

195

: 00:08:23

that AI is doing it correctly.

196

: 00:08:25

But I think the what and the

when is what really matters.

197

: 00:08:28

And so what I'm actually doing

is I run a bootcamp, it's called

198

: 00:08:30

Data Analytics Accelerator.

199

: 00:08:31

We'll have a link to the show notes

down below if you wanna, if you're

200

: 00:08:33

curious, you wanna check it out.

201

: 00:08:34

I think I need to go through the

entire thing again and really

202

: 00:08:37

focus on the what and the when.

203

: 00:08:39

'cause the how I've been, I've

nailed the, how we have had so many

204

: 00:08:41

students go through this program.

205

: 00:08:42

They've really enjoyed it.

206

: 00:08:43

They become great data

analysts at the end of it.

207

: 00:08:44

But I think the most important

thing is going through and going

208

: 00:08:47

through, okay, why are we doing this?

209

: 00:08:49

When would you do this again?

210

: 00:08:50

You know, how did I know to do this?

211

: 00:08:51

How did, how should you know to do this?

212

: 00:08:53

When you get a data set in the future,

what are some different things that you

213

: 00:08:55

can do with it and when would you do it?

214

: 00:08:56

When is it appropriate?

215

: 00:08:57

That is what's going to be.

216

: 00:08:58

That's what's gonna make you a Golden

data analyst in this new era of ai, and

217

: 00:09:02

I really hope that I will be part of

your journey in learning how to do that.

218

: 00:09:05

So that's why it's really important

that no matter what you're listening,

219

: 00:09:08

you hit subscribe and you stay tuned

because over the next six to 12 months,

220

: 00:09:12

I'm gonna be hitting this really hard

and I don't want you guys to miss out.

221

: 00:09:14

So thanks for listening, and

I'll catch you in the next one.