184: What I’d Learn Instead of Data Science in 2026

Speaker: 00:00:00

If you're starting from absolute

scratch, I don't think you

2

: 00:00:02

should be a data scientist.

3

: 00:00:04

At least not yet.

4

: 00:00:05

And let me explain why not,

and what I would do instead.

5

: 00:00:09

Now I wanna make it very clear.

6

: 00:00:10

I don't think data science is dead at all.

7

: 00:00:13

Like you might see a

lot of YouTubers saying.

8

: 00:00:15

I don't think it's dead in the least.

9

: 00:00:17

I freaking love data science and I

think it's gonna continue to thrive,

10

: 00:00:21

but at the current moment, I do think

there's a better path for you to

11

: 00:00:25

take if you ultimately wanna become

a data scientist down the road.

12

: 00:00:28

So let's first get into the reasons why

not to be a data scientist right now.

13

: 00:00:32

Reason number one is it takes a

long time to learn data science and

14

: 00:00:36

ultimately become a data scientist.

15

: 00:00:38

And basically, in order to be a data

scientist, you have to do two things.

16

: 00:00:43

Number one, you have to know some math.

17

: 00:00:45

And number two, you have to be able

to do that math with programming.

18

: 00:00:48

And unfortunately.

19

: 00:00:49

Just the way it is.

20

: 00:00:51

Math is pretty hard to learn.

21

: 00:00:52

It is not easy to learn.

22

: 00:00:54

And knowing things like calculus,

linear algebra obviously are important

23

: 00:00:59

to do things like machine learning.

24

: 00:01:00

But even if you, let's just say you ignore

linear algebra and calculus because it

25

: 00:01:05

makes your job as data scientist better.

26

: 00:01:06

You're a better data scientist

if you know those things.

27

: 00:01:09

But it's not a necessity.

28

: 00:01:10

Like everyone makes it out to

seem , it's helpful, but it's not

29

: 00:01:13

like you need a hundred percent to

understand those and know those.

30

: 00:01:16

A million percent.

31

: 00:01:17

Even just ignoring all of that,

the algorithms, the logic behind

32

: 00:01:21

the algorithms is pretty tricky.

33

: 00:01:22

And it takes like, uh, a lot of

patience and understanding and a

34

: 00:01:27

lot of like math logic to get these

machine learning algorithms down.

35

: 00:01:31

And as a data scientist, that's like

your number one job is to be, you know,

36

: 00:01:34

using and creating these machine learning

algorithms to do data science stuff,

37

: 00:01:38

to predict stuff, to classify stuff.

38

: 00:01:41

So you're gonna be needing to know

mathematics, which just takes a

39

: 00:01:44

while to, to learn, to be honest.

40

: 00:01:45

And once you even figure out the math

and you can use the machine learning

41

: 00:01:48

algorithms, you have to be able to

use them pretty much via programming.

42

: 00:01:53

There are some data scientists

jobs out there that probably use.

43

: 00:01:56

Less programming than you might think.

44

: 00:01:59

They use tools that kind of do

the programming for them, but I

45

: 00:02:02

think that's few and far between.

46

: 00:02:05

I think that's where knowing like R or

Python comes in, those are pretty much

47

: 00:02:09

the two programming languages that you're

going to be using as a data scientist.

48

: 00:02:13

To do machine learning.

49

: 00:02:15

And unfortunately,

programming is also hard.

50

: 00:02:17

It takes a long time to learn.

51

: 00:02:19

If you've never programmed at all

or you've only done a little bit of

52

: 00:02:21

programming, it's a lot of effort to

learn programming from absolute scratch.

53

: 00:02:25

You're gonna have to learn about

variables, you're gonna have to

54

: 00:02:27

learn about functions, you're

gonna have to learn about loops.

55

: 00:02:30

You're gonna have to learn about like.

56

: 00:02:31

Classes and, uh, I'll have parameters

and arguments and a bunch of other things

57

: 00:02:36

that I'm probably forgetting right now.

58

: 00:02:37

The point is there's actually a lot to

learn when it comes to programming and it

59

: 00:02:40

doesn't really come naturally to everyone.

60

: 00:02:43

Uh, it is definitely a learned skill

that takes patients in years to master.

61

: 00:02:48

Um, so if you're starting from absolute

scratch and you're like not a math

62

: 00:02:50

whiz, and you're not like a programming

expert, going from zero to data scientist

63

: 00:02:55

is gonna take a long time because

before you're gonna be qualified.

64

: 00:02:58

To land data scientist roles,

you're gonna have to get Okay.

65

: 00:03:01

At the math and okay.

66

: 00:03:02

At programming.

67

: 00:03:03

And don't get me wrong, I love math.

68

: 00:03:05

I love programming.

69

: 00:03:06

If you're loving it, then maybe

it is something to pursue.

70

: 00:03:09

, But I think I still think there's

a better path to get there.

71

: 00:03:11

We'll talk about that here in a second.

72

: 00:03:13

Um, but just know that like it's really

hard to lend a data science job to even

73

: 00:03:16

be qualified to land a data science job

right now because the amount of math and

74

: 00:03:20

the amount of programming is quite high.

75

: 00:03:22

The number two reason that you

maybe shouldn't be a data scientist

76

: 00:03:25

right now is it honestly requires

a little bit of like bons.

77

: 00:03:29

I don't even know if I'm using that

word correctly, but like you need

78

: 00:03:32

some, not certifications, but honestly

it's seeming like when you're applying

79

: 00:03:37

data scientist jobs, most of them

are saying master's in data science.

80

: 00:03:41

Now I haven't actually analyzed that

statistically, , but that's just kinda

81

: 00:03:45

what I've been seeing anecdotally.

82

: 00:03:46

Um, and once again.

83

: 00:03:47

Masters.

84

: 00:03:48

It's gonna take a long time to get.

85

: 00:03:50

Uh, the good news is you probably will

learn some of the programming, some of

86

: 00:03:54

the math that I talked about earlier, but

it's just gonna take a freaking long time.

87

: 00:03:57

We're talking like probably two years

to get a master's in data science.

88

: 00:04:01

I guess maybe if you're doing it full

time, maybe it would only take one year.

89

: 00:04:04

but one, it's just taking

a long time, right?

90

: 00:04:06

We're talking years, not months here.

91

: 00:04:08

And two, it's also gonna

be expensive, right?

92

: 00:04:10

Because masters are not cheap.

93

: 00:04:13

I think the cheapest that you

can get, like a master's in data

94

: 00:04:16

science is like probably, yeah.

95

: 00:04:20

$13,000, if I'm gonna be honest.

96

: 00:04:22

I got my master's in data

analytics from Georgia Tech.

97

: 00:04:26

I have a review on it if you

want to check it out sometime.

98

: 00:04:28

It costs me about $13,000

and I think that's about the

99

: 00:04:31

cheapest that you could go.

100

: 00:04:32

So you.

101

: 00:04:33

Kind of need a master's.

102

: 00:04:35

If you're gonna try to be a data

scientist at, at least you have

103

: 00:04:37

a lot better shot because it is

kind of listed as a requirement.

104

: 00:04:40

I don't think it is a

requirement necessarily.

105

: 00:04:42

I think you could land a data scientist

job without a master's degree, but I

106

: 00:04:46

think it's honestly gonna be pretty hard.

107

: 00:04:48

So unless you're wanting to spend like

two years, unless just say like, on

108

: 00:04:51

average $20,000 in student loans maybe

you shouldn't be a data scientist.

109

: 00:04:55

Reason number three that maybe you

should avoid the data scientist role

110

: 00:04:59

right now is there's actually a lot

more openings in different data roles.

111

: 00:05:04

And let me explain.

112

: 00:05:05

So if you wanna do this experiment,

you can, you know, uh, I did

113

: 00:05:08

this experiment and I'll tell

you the results here in a second.

114

: 00:05:11

But there's actually more data

engineering jobs open right now

115

: 00:05:14

than there are data scientist jobs.

116

: 00:05:16

Um, not by a lot about.

117

: 00:05:17

10% more.

118

: 00:05:19

And there's actually double data

analyst jobs open than there are

119

: 00:05:22

data scientist roles open right now.

120

: 00:05:24

And that's in the US

and I, I used LinkedIn.

121

: 00:05:26

You can go to LinkedIn and go

to the search bar and you can

122

: 00:05:28

type in data analyst or data

scientist and I did United States.

123

: 00:05:32

So it might be different if

you're in a different country.

124

: 00:05:34

Um, and it shows you the number of

results and I think the results for.

125

: 00:05:38

Data scientist was 8,000,

data engineer was 9,000 and

126

: 00:05:43

data analyst was like 17,000.

127

: 00:05:46

So data scientist has like the

lowest amount of openings right

128

: 00:05:50

now, and I'll talk about why.

129

: 00:05:51

I think that's the case here in a second.

130

: 00:05:53

Data engineering has a little bit more.

131

: 00:05:55

But there's also, a lot less

barrier to entry for data engineer

132

: 00:05:59

and for a data analyst as well.

133

: 00:06:02

Data engineering, it requires a lot of

programming and a lot of logic, a little

134

: 00:06:06

bit less math, and you're not doing as

much like machine learning necessarily,

135

: 00:06:09

but honestly, probably more programming.

136

: 00:06:11

So if you are kind of a programmer,

maybe that's the route you wanna go.

137

: 00:06:14

Because there is more job openings right

now and there's not master's degrees.

138

: 00:06:18

Like there's not really a ton of data

engineering master's degrees out.

139

: 00:06:21

This degree doesn't even really exist yet.

140

: 00:06:23

So that's.

141

: 00:06:23

That's nice, right?

142

: 00:06:24

'cause when there's a degree that

doesn't exist, you don't have

143

: 00:06:26

to have it to land the roles.

144

: 00:06:28

And uh, I think data engineering is

kind of exploding with AI recently

145

: 00:06:32

because AI really, at the core of

it, at, I guess at the beginning

146

: 00:06:37

is a data engineering problem.

147

: 00:06:39

'cause it's lot of data.

148

: 00:06:40

You wanna basically feed these

models and it's a lot of unstructured

149

: 00:06:44

data, so it's like, how do we best

structure the unstructured data?

150

: 00:06:48

Um, so data engineering roles

are, are getting more popular.

151

: 00:06:51

They don't require like a master's

degree, but there is a lot of programming.

152

: 00:06:54

I don't wanna make it seem like

it is a lot, it is easy to land a

153

: 00:06:58

data engineering job because it's.

154

: 00:07:00

But I do think it is easier to

land a data engineering job right

155

: 00:07:02

now than it is a data scientist.

156

: 00:07:04

Easier than both of those, I

think is a data analyst role.

157

: 00:07:07

One, there's a lot more

roles open right now, right?

158

: 00:07:09

About double, , and two, like the

barrier to entry is so much lower.

159

: 00:07:13

You don't have to know nearly as

much programming and you don't

160

: 00:07:15

have to know nearly as much math.

161

: 00:07:17

So you're able to land a

role a lot more quickly.

162

: 00:07:20

And that's like huge.

163

: 00:07:21

So spoiler alert, that is my

whole pitch to you is like, Hey.

164

: 00:07:25

Put becoming a data scientist on the

shelf for just a little bit, become a data

165

: 00:07:29

analyst first, and then pivot into that.

166

: 00:07:31

We'll talk about that here in a second.

167

: 00:07:33

the fourth reason you should maybe

consider not being a data scientist

168

: 00:07:35

right now is I do think the data

scientist jobs are at least stagnant.

169

: 00:07:40

I don't think they're down necessarily.

170

: 00:07:42

But data science, It takes

longer to make a business impact.

171

: 00:07:46

And what I mean by that is if you're

a business and you're looking to

172

: 00:07:50

make money right now, you're looking

for profits today data science is

173

: 00:07:53

much more of an investment than data

analytics and data engineering, data

174

: 00:07:57

engineering is a really good investment

for companies right now because.

175

: 00:08:01

You can't really do data analytics or data

science without good data engineering.

176

: 00:08:05

You can't really do much with

data if you don't have good data

177

: 00:08:09

engineering, because it's like how do

the data scientists access the data?

178

: 00:08:12

How do they know that it's clean?

179

: 00:08:14

So businesses investing in data

engineering, it makes a lot of sense

180

: 00:08:16

right now because you can't even really

get much return on investment spent

181

: 00:08:21

in data until you have data, right?

182

: 00:08:23

And data engineering's all about

having data and storing it.

183

: 00:08:27

Properly and effectively.

184

: 00:08:28

Once you have that data stored right,

and, and everything's all set up,

185

: 00:08:31

that's when you can start doing

data science or data analytics.

186

: 00:08:34

I differentiate those between data

science is Looking more towards the

187

: 00:08:39

future, like predicting stuff, right?

188

: 00:08:41

And data analytics, looking more towards

the past and saying what happened.

189

: 00:08:45

Um, so data analytics is like

more reporting, like this is what

190

: 00:08:48

happened in the past type of thing.

191

: 00:08:50

Data science being this is

what will happen in the future

192

: 00:08:52

or, or predicting some sort of

behavior or something like that.

193

: 00:08:55

It's just a lot easier

to do data analytics.

194

: 00:08:58

It's like so much easier to do, uh, data

analytics than it is data science and

195

: 00:09:02

you can get those results a lot quicker.

196

: 00:09:04

Like I can make a report on what

happened in the past and probably what

197

: 00:09:08

a fourth of the time is that's gonna

gonna take me to predict the future.

198

: 00:09:12

And as someone who's been a data analyst.

199

: 00:09:14

And a data scientist, I just know, doing

the data science work takes longer,

200

: 00:09:18

'cause once again, it's more complex.

201

: 00:09:19

You're doing more programming

you're, you're doing more math.

202

: 00:09:22

It's a harder problem to solve.

203

: 00:09:24

And so it just takes longer

for the businesses to actually

204

: 00:09:27

see the fruits of their labor.

205

: 00:09:29

Versus a data analyst, you

can almost see the results.

206

: 00:09:32

Kind of immediately.

207

: 00:09:33

So I, I think you'll have more

business impact as a data analyst

208

: 00:09:37

'cause the results are very clear.

209

: 00:09:39

And I think as a data scientist, I

think some of their, these long-term

210

: 00:09:42

projects, a lot of these long-term

projects in data science fail too.

211

: 00:09:45

Like, I worked as a data

scientist for ExxonMobil.

212

: 00:09:48

Uh, I worked on let's see, I don't

know, like four big initiatives.

213

: 00:09:52

And I would say half of them

probably failed, if I'm being honest.

214

: 00:09:55

I don't know that for a fact.

215

: 00:09:56

I kind of left before some of

those, Products were finished.

216

: 00:10:00

One of them that was even

considered a success.

217

: 00:10:03

Like it wasn't even really being

implemented or used I basically

218

: 00:10:06

built that project in what,

: 2018

219

: 00:10:12

It was like everyone really liked it.

220

: 00:10:14

They're like, this is awesome.

221

: 00:10:15

But I don't really think we had

very many users of the tool.

222

: 00:10:18

So I don't know if you even count that

as success or not, but my point here

223

: 00:10:22

is like looking at that, that was two

and a half years to even get to the

224

: 00:10:25

point where, oh, we think this is gonna

be a success, but it hasn't been yet.

225

: 00:10:29

So I hope that gives you a little

bit of an idea of how long it might

226

: 00:10:32

take to actually impact the business.

227

: 00:10:34

And I think when you're

impacting business one.

228

: 00:10:37

You get promoted more often, you

get raises, those types of things,

229

: 00:10:40

you get more clout, I guess.

230

: 00:10:42

Uh, but two more roles of

those types of roles open up.

231

: 00:10:45

And I just think that the, the return

on investment for data scientists

232

: 00:10:48

right now might be a little bit fuzzy.

233

: 00:10:51

Now, that's not true for every company.

234

: 00:10:52

Like a lot of companies make their

money with data science work and

235

: 00:10:55

they have already have a really good

data engineering infrastructure.

236

: 00:10:59

That might be some of the bigger

tech companies that like are.

237

: 00:11:01

Bajillionaire, right?

238

: 00:11:02

Like those types of companies, they

still probably make their, their

239

: 00:11:06

money with, with data science,

I think it's very valuable.

240

: 00:11:08

, Like social media apps like Instagram,

Facebook, those types of things.

241

: 00:11:11

But I think a lot of smaller operations,

they might be getting a little bit

242

: 00:11:15

tight on data scientists because

the return on investment, it's high

243

: 00:11:18

risk, high reward, I guess, uh, data

engineering and data analysts, it's,

244

: 00:11:22

it's a lot more sure of an investment.

245

: 00:11:24

So I think that might be one

of the reasons why you should

246

: 00:11:27

actually consider these roles.

247

: 00:11:28

I think you should become a data

analyst because like I said earlier.

248

: 00:11:31

It's easy, it's easy to learn.

249

: 00:11:34

There's a lower barrier to entry

and there's a lot more roles open.

250

: 00:11:37

And my whole philosophy is I think you

should become a data scientist someday

251

: 00:11:41

if you want to be a hundred percent.

252

: 00:11:43

But the cool thing is this data

analyst role is kind of like a

253

: 00:11:46

gateway role where the, the, the

fence is not hard to get over.

254

: 00:11:50

The barrier is not hard.

255

: 00:11:51

You can get in this data analyst role.

256

: 00:11:53

Pretty easily.

257

: 00:11:54

I have a whole roadmap on how to

actually become a data analyst.

258

: 00:11:57

Um, you can watch that on YouTube, uh, up

here in the card, or if you're listening

259

: 00:12:01

to the podcast via audio, I'll have it

a link in the show notes down below.

260

: 00:12:05

That'll kind of explain everything

that you need to do step by step.

261

: 00:12:08

But my whole point is like

you can become a data analyst.

262

: 00:12:11

And then you can get paid to become

a data scientist down the road.

263

: 00:12:15

Because that's, that's also true.

264

: 00:12:17

Companies will pay you to learn.

265

: 00:12:19

I actually have another video

about like my old philosophy of

266

: 00:12:22

like how to get paid to learn.

267

: 00:12:23

I'll have it right here on a YouTube

card or in the show notes down below.

268

: 00:12:27

Um, but just like the

short of it is this, that.

269

: 00:12:30

Once you're in a company, they're

going to invest in you to learn things.

270

: 00:12:35

You're gonna have access to

like free LinkedIn learning.

271

: 00:12:37

You're gonna have access

to like go to conferences.

272

: 00:12:39

Uh, a lot of these companies will

even pay for a master's degree.

273

: 00:12:42

I know I have some students in

my accelerator bootcamp who.

274

: 00:12:45

I've worked with

275

: 00:12:46

I actually interviewed them both,

so I'll pop them up in, in cards

276

: 00:12:49

up here and have their, their

links in the show notes down below.

277

: 00:12:52

Um, but one was a math teacher

and one was in quality assurance,

278

: 00:12:56

and I helped them both pivot into

like more data analyst roles.

279

: 00:13:00

And now their companies are

paying for them to go get a

280

: 00:13:03

data science master's degree.

281

: 00:13:05

And I think that's awesome

because now instead of.

282

: 00:13:07

You know, going into debt $20,000

to become a data scientist,

283

: 00:13:10

you already have a data job.

284

: 00:13:11

You have income coming in, data income,

you know, data analysts get paid well.

285

: 00:13:15

It's not like it's a, a crappy,

uh, salary and they can get paid to

286

: 00:13:20

learn on the job and have school.

287

: 00:13:22

Paid for by the company.

288

: 00:13:24

So I think that is like a win, win win.

289

: 00:13:26

I think that is the route

that you should take.

290

: 00:13:28

So I'm not saying data science is dead.

291

: 00:13:30

I'm not saying don't

become a data scientist.

292

: 00:13:33

I'm just saying if you want to become

a data scientist, I think you should

293

: 00:13:36

become a data analyst first, then learn

to become a data scientist on the job.

294

: 00:13:40

You can learn Python on the job, you

can learn machine learning on the job.

295

: 00:13:42

And that way you're

getting income coming in.

296

: 00:13:44

While you're learning, because I can't

tell you how many people have come to me.

297

: 00:13:47

Hey, Avery, I have a

master's in data science.

298

: 00:13:48

I can't land a job.

299

: 00:13:49

It's so much easier once you

already have some sort of a data

300

: 00:13:52

job and a current data job too.

301

: 00:13:55

You can watch this video if you're

watching on YouTube next, that

302

: 00:13:57

will help you learn to get started.

303

: 00:13:59

Check out the show notes if

you're listening on the podcast.

304

: 00:14:01

Thank you guys for listening, and

uh, I'll see you in the next episode.