Are we digital tenants? Tech aspect nobody talks about!

Speaker: 00:00:00

they have read the whole internet.

2

: 00:00:02

Have they lived?

3

: 00:00:04

no. Feel the rush.

4

: 00:00:05

It's a gravity high.

5

: 00:00:08

Talking tech is the planets fly.

6

: 00:00:11

The upgrades on the show begins.

7

: 00:00:14

Sit right back and let the future's.

8

: 00:00:21

All right then.

9

: 00:00:22

Welcome, Fatima.

10

: 00:00:23

You are number one guest

that we're doing a remote recording with.

11

: 00:00:28

So that's

why I'm really excited for this one.

12

: 00:00:32

So for anybody who doesn't know you, let's

start with a quick sort of introduction.

13

: 00:00:37

Who you are, what you do. And,

you know, we'll take it from there.

14

: 00:00:39

Okay. Okay.

15

: 00:00:41

Thank you.

16

: 00:00:42

It's an honor to be here.

17

: 00:00:44

Hello to anyone and everyone listening.

18

: 00:00:47

My name is Fatima

I’ve been a software developer.

19

: 00:00:52

And maybe you can call me

entrepreneur for.

20

: 00:00:56

For most of my life.

21

: 00:00:58

Yeah. I started so very many years ago.

22

: 00:01:02

First with PCs,

then with the IBM I systems.

23

: 00:01:08

Then I of course featured to the web

when it came and mobile.

24

: 00:01:12

I see.

So you've been through that transition.

25

: 00:01:15

Yeah. Yeah.

26

: 00:01:16

You got to you cannot,

you know, be on the past.

27

: 00:01:20

You get to to to update.

28

: 00:01:23

And I've been doing and blockchain

for like ten years

29

: 00:01:26

and lately last five years

I'd been also been tinkering with AI.

30

: 00:01:33

I see.

31

: 00:01:35

So do you see any specific

32

: 00:01:37

trends where things are

where things are going right now?

33

: 00:01:40

What would you say is the technology

of of, you know, of the future?

34

: 00:01:44

Let's say, you know,

35

: 00:01:45

is it all going to be really around AI

or are there like other sort of,

36

: 00:01:48

you know, technologies

that people don't really talk about,

37

: 00:01:50

but they are going to be really like

critical here?

38

: 00:01:53

Well, we are seeing them

some things happening.

39

: 00:01:56

You know, first it's around here.

40

: 00:02:00

It's like a new layer,

you know, so far is constructed in layers.

41

: 00:02:04

At first. Of course.

42

: 00:02:06

People were toning on and off switches

to make ones and zeros.

43

: 00:02:10

Yeah.

44

: 00:02:11

Until then, the punched cards came.

45

: 00:02:14

Then you got a keyboard and a printer,

and then you got a screen,

46

: 00:02:19

and then you

you didn't use Cirrus and once anymore.

47

: 00:02:22

But there was a language assembly language

to program the processor that then

48

: 00:02:28

in this maybe in the 60s between the 60s

49

: 00:02:32

and 70s,

the temporal programing languages.

50

: 00:02:36

Yeah, starting with the old ones,

which most are not around

51

: 00:02:41

Algol, Pascal, which is still around

COBOL, which is still around.

52

: 00:02:46

And so yes,

then programing languages were able been

53

: 00:02:51

and we got today

more sophisticated languages.

54

: 00:02:54

There came Jabba Jabba with the break 90s

55

: 00:02:58

and for all

modern languages can Python, PHP,

56

: 00:03:03

Ruby, most mostly everything.

57

: 00:03:07

She was great.

58

: 00:03:08

Whatever is used today

and people were coding.

59

: 00:03:12

Yeah.

60

: 00:03:12

In those languages you code you wrote,

you write a program and the computer

61

: 00:03:18

translates it to assembly language

62

: 00:03:21

and once and zeros in the end.

63

: 00:03:24

But now there's a new layer.

64

: 00:03:27

Yeah. Now that you get something came up.

65

: 00:03:29

There was a was a concept brick

at the beginning of the century, you know,

66

: 00:03:34

because our intelligence pretty old

is from the 80s,

67

: 00:03:38

but it didn't work it much.

68

: 00:03:42

You know.

69

: 00:03:42

What was the obstacle back then?

70

: 00:03:44

Can you, can you sort of talk

a little bit about that?

71

: 00:03:46

Was it just the computing power back then?

72

: 00:03:49

Well, that was the limitation, you know.

73

: 00:03:52

Nobody knew.

74

: 00:03:53

Nobody knew. It was kind of a mystery.

75

: 00:03:57

Okay.

76

: 00:03:58

At some

point in the at the beginning of the nice,

77

: 00:04:01

some people got you know, how it started.

78

: 00:04:04

Is very interesting to understand

how it works, how it started.

79

: 00:04:08

This was this guy in IBM at IBM

80

: 00:04:11

that was trying to plan games,

you know, like TikTok too.

81

: 00:04:15

Yeah, that was very easy.

82

: 00:04:17

Yes, but then he tried to protect

checkers.

83

: 00:04:20

Yeah. And chickens is a bit more complex.

84

: 00:04:23

You cannot just do if, if, if

because there's a lot of pieces involved.

85

: 00:04:28

So what it is, it was very interesting.

86

: 00:04:32

He started making the program record

the matches.

87

: 00:04:36

Yeah.

88

: 00:04:37

And so and so he started

to have a database of place,

89

: 00:04:42

you know, and he started

making the computer react to that.

90

: 00:04:47

So if you drew that move, that move,

the computer had examples

91

: 00:04:52

and go say no, go this way, this way. No.

92

: 00:04:55

Because we did it on Willows.

93

: 00:04:57

Yeah. So I already machine learning.

94

: 00:04:59

Yeah.

95

: 00:05:00

That's the beginning of machine learning.

96

: 00:05:02

And that. Became. The 90s.

97

: 00:05:04

They were able to read hand

written numbers from checks.

98

: 00:05:09

That was sort of a breakthrough,

but nothing else.

99

: 00:05:12

Workers.

100

: 00:05:13

And then came the where I know

the bunny went to the web, you know,

101

: 00:05:17

and they're keeping it what was called

the second AI winter.

102

: 00:05:22

But then.

103

: 00:05:23

I see the web changed everything. Why?

104

: 00:05:26

Well, it's obvious

because now everyone is connected

105

: 00:05:31

to everyone,

which was not possible before.

106

: 00:05:34

Yeah, but it brought a second effect.

107

: 00:05:37

That was the the size of applications

changed it completely

108

: 00:05:43

when we were doing applications in this,

you know, beginning of the 90s,

109

: 00:05:49

if you had like 500 users,

it was a very big app.

110

: 00:05:53

In a corporation, you will have 500 users,

111

: 00:05:58

you may have 3000 users.

112

: 00:06:01

Yeah. And it was wow, very big.

113

: 00:06:04

But then the question yes, broke a scale

that was never seen before.

114

: 00:06:10

Users in the web is nothing.

115

: 00:06:12

Some websites appeared

that had millions of users.

116

: 00:06:17

Yes. And that

117

: 00:06:19

means that you have enormous databases

and that is called big data.

118

: 00:06:25

Big data,

because you have these enormous databases.

119

: 00:06:29

You got a million users,

you got a billion records of what

120

: 00:06:33

they are doing, what they are buying,

what they are writing.

121

: 00:06:36

And then so many different data points

as well.

122

: 00:06:39

Yeah, yeah.

123

: 00:06:40

The size changed and then came Fei-Fei Li,

you know, Fei-Fei Li.

124

: 00:06:46

I think she's Chinese.

125

: 00:06:47

She's from from the east clearly.

126

: 00:06:51

And as she had this, this insight.

127

: 00:06:54

Is it that the AI algorithms

128

: 00:06:58

do not work or is it that they work?

129

: 00:07:01

But we are feeding them too little data?

130

: 00:07:04

It's a brilliant question.

131

: 00:07:06

You know. And what she did. Yes.

132

: 00:07:08

She created an online database of pictures

133

: 00:07:12

and asked people

to upload pictures of cats.

134

: 00:07:16

Okay. In different positions is possible.

135

: 00:07:20

Yeah. Face, whole body.

136

: 00:07:22

The cat doing doing their thing

that, you know,

137

: 00:07:25

in their strange positions

and so on and so on.

138

: 00:07:29

So she got this

enormous database of cat pictures,

139

: 00:07:34

and then she ran the same algorithm

that didn't work before.

140

: 00:07:39

And it worked.

141

: 00:07:40

So it was the volume of data

that was basically the obstacle.

142

: 00:07:44

It wasn't enough for that

to work in the past.

143

: 00:07:47

I see.

144

: 00:07:48

Oh, wow.

They're going somewhere starving.

145

: 00:07:51

Somehow they didn't have enough data

when she fed the algorithm

146

: 00:07:57

with a lot of data, he started working.

147

: 00:08:00

Yeah. Wow.

148

: 00:08:02

And so the new I started,

and everybody started doing it.

149

: 00:08:07

Google? Yeah.

150

: 00:08:09

Google are ready

to have interesting things.

151

: 00:08:12

They have a database for handling

152

: 00:08:15

the their users, but the users search for

153

: 00:08:18

they put it in a database

that is called Bigtable.

154

: 00:08:22

Yeah. Not by chance.

155

: 00:08:24

And so they started running algorithms

on these big, big, big data collections.

156

: 00:08:29

And they started working.

157

: 00:08:31

And so Google was into it.

158

: 00:08:33

After sometime Facebook was into it,

Apple was into it.

159

: 00:08:37

But then something happened somewhat.

160

: 00:08:40

That cuckoo thought that it would be

may be useful

161

: 00:08:45

if when you are writing an email.

162

: 00:08:47

Yeah, but the email gave you an option

to continue your phrases.

163

: 00:08:52

You saw that thing that you start typing

and he appears,

164

: 00:08:57

it appears a gray adult,

a possible continuation.

165

: 00:09:01

Yeah. For the phrase. Yeah, Yeah.

166

: 00:09:04

That's the prediction. Yeah.

167

: 00:09:05

The algorithm is trying to guess

what you need to write

168

: 00:09:10

after what you have already touched. Yeah.

169

: 00:09:13

And so they started modeling language.

170

: 00:09:17

Yeah. Human language.

171

: 00:09:18

Okay. Initially English of course.

172

: 00:09:21

Here they started modeling English to see

173

: 00:09:25

because they had already this enormous

that the database

174

: 00:09:28

that was Google scanning the whole web,

all the texts in, in the whole.

175

: 00:09:33

What in the world here?

176

: 00:09:35

Yes. What the users are searching.

177

: 00:09:38

Okay.

178

: 00:09:39

The way they write, the way they consume.

179

: 00:09:42

Yeah. Of course.

180

: 00:09:43

And they they started modeling language.

181

: 00:09:46

Yeah. And when they.

182

: 00:09:47

And then they started modeling language,

and then the borders became

183

: 00:09:53

bigger and bigger and bigger

because you have so much text to index.

184

: 00:09:58

Yeah.

185

: 00:09:59

And that's

what is now called large language models.

186

: 00:10:04

Yeah.

187

: 00:10:05

I see that big, big models here

that are learning.

188

: 00:10:10

Same as the check curse program in the 60s

or 70s.

189

: 00:10:15

Yeah, that was recording place

matches here and see what happens.

190

: 00:10:20

But this, this this new algorithms,

they map language.

191

: 00:10:26

Yeah. You are me talking, for example.

192

: 00:10:29

Yeah, yeah.

193

: 00:10:30

And they start to learn

what combinations of words are.

194

: 00:10:34

Sentences are valid.

195

: 00:10:38

I see.

196

: 00:10:40

And so and so it came.

197

: 00:10:42

And then they started trying

this in the lamps this language models.

198

: 00:10:48

And you can chat with them.

199

: 00:10:52

Yes. They can predict what come next.

200

: 00:10:54

So you say hello to them

and they say hello.

201

: 00:10:57

How are you.

202

: 00:10:58

Because they had seen

it had salient times in this big database.

203

: 00:11:02

Exactly.

204

: 00:11:04

In also in the context.

205

: 00:11:06

So that's that's

the important part as well.

206

: 00:11:09

Right. Yeah. The context.

207

: 00:11:11

Yeah.

208

: 00:11:11

You know, you you start writing an email

and when you start,

209

: 00:11:15

what they say is not so good.

210

: 00:11:18

But after you wrote 20 lines,

they have seen a sealion.

211

: 00:11:22

Similar emails and what they offer.

212

: 00:11:27

Because much better.

213

: 00:11:29

Yeah. I see.

214

: 00:11:32

And that's where we are today. Yeah.

215

: 00:11:35

And so.

216

: 00:11:36

Companies appeared that are

217

: 00:11:38

dedicated to training this

218

: 00:11:43

language models

219

: 00:11:44

OpenAI Anthropic, some Chinese companies.

220

: 00:11:48

Yeah.

221

: 00:11:49

That's the point where we are today.

222

: 00:11:53

And just very quickly because,

223

: 00:11:55

you know, there is this,

this concept of hallucinations obviously.

224

: 00:11:58

So it's trying to respond,

not having the input or

225

: 00:12:01

specific information that's backed up

by that, say, actual data.

226

: 00:12:06

So do you have any, any sort of insights

on the origin of that problem?

227

: 00:12:11

Well.

228

: 00:12:12

You know,

let's say that you have this rent.

229

: 00:12:15

Yeah.

230

: 00:12:16

That lives in the library.

231

: 00:12:19

Okay. Library.

232

: 00:12:21

Right. Yeah.

233

: 00:12:22

You got this from that from school?

234

: 00:12:25

Yeah. That never goes out of the library.

235

: 00:12:28

He likes very much to read.

236

: 00:12:30

And he got a shop in the library,

and he said all day, all night.

237

: 00:12:35

He never goes out.

238

: 00:12:37

Yeah, he's the reason.

239

: 00:12:38

Okay. Does he know a lot?

240

: 00:12:40

Well, yes. Yes. Kind of.

241

: 00:12:43

Yeah. I know he's really theory.

242

: 00:12:45

Yeah.

243

: 00:12:45

He's reading all the books

from billion, from brilliant people.

244

: 00:12:51

So he knows a lot.

245

: 00:12:52

Does he had real life experience?

246

: 00:12:56

Nobody lives.

247

: 00:12:58

Oh, I see where you're going with this.

248

: 00:13:01

You say, hey, why don't we go this Sunday

to see this F1 race?

249

: 00:13:06

I see.

250

: 00:13:06

So yeah, in 1973,

Fittipaldi in one and then and the car

251

: 00:13:13

and then the in the 90s and the serials

and now and he knows a lot

252

: 00:13:18

but he never drove a car

and he has never been in the races.

253

: 00:13:22

So he knows a lot.

254

: 00:13:23

But he tends to fantasize

255

: 00:13:27

a bit

because he lacks real life experience.

256

: 00:13:30

This is what happens to language models.

257

: 00:13:34

They are kind of blind.

258

: 00:13:36

Yeah, they have read the whole internet.

259

: 00:13:39

Have they lived? no.

260

: 00:13:41

So they tend to hallucinate -

it’s natural.

261

: 00:13:44

If you think about it and you do,

a real life comparison is natural.

262

: 00:13:49

I know, so

now there's a new thing coming Fei-Fei

263

: 00:13:53

Li I like her book a lot.

264

: 00:13:56

Of course.

265

: 00:13:57

Fei-Fei

Li is now working on a spatial model.

266

: 00:14:03

Okay.

267

: 00:14:04

A model of the physical world.

268

: 00:14:06

Because the limbs

only know about language.

269

: 00:14:10

But when dealing with the physical world,

they are not so good.

270

: 00:14:14

So she's she's working.

271

: 00:14:17

She has a new company.

272

: 00:14:18

He got, she got of course he masters.

273

: 00:14:21

And so.

274

: 00:14:22

And she's trying to build her

275

: 00:14:26

model of physical reality here.

276

: 00:14:29

I see because I see.

277

: 00:14:31

Because the next frontier

is probably robotics.

278

: 00:14:35

We are starting to see, for example,

279

: 00:14:39

self-driving cars.

280

: 00:14:42

Yes, of course.

281

: 00:14:43

Which are not that good,

but they are starting.

282

: 00:14:46

She has started to be as good as bad

drivers.

283

: 00:14:50

Yeah.

284

: 00:14:51

Yes. Yes, exactly. Yeah. Okay.

285

: 00:14:54

They are not as good as their experience

driver.

286

: 00:14:58

Maybe, but they are not worse than

287

: 00:15:02

but rather.

288

: 00:15:03

Yeah, they can go out

and they don't kill people.

289

: 00:15:07

That's it. Yeah of course.

290

: 00:15:10

And but.

291

: 00:15:12

Of course.

292

: 00:15:12

And I think we've all seen

a video of Tesla avoiding an accident.

293

: 00:15:16

So that also happens on occasion.

294

: 00:15:18

Yeah yeah. Yeah yeah.

295

: 00:15:20

Also doing

people are creating self-driving drones.

296

: 00:15:25

Yeah.

297

: 00:15:26

Because most runs at first

you will control them by radio.

298

: 00:15:30

But for example in Ukraine

both sides are jamming

299

: 00:15:35

all the frequencies

so you can control your arms.

300

: 00:15:38

So they put a lesser cable.

301

: 00:15:42

Yeah. A fiber optic. Yes. Yeah.

302

: 00:15:45

That maybe the that the drones goes away.

303

: 00:15:50

Yeah. And they did.

304

: 00:15:51

And the fiber optic will extend you,

you have a roll

305

: 00:15:55

and you give it more and more and more.

306

: 00:15:58

Cable like kilometers.

307

: 00:15:59

Long, five kilometers long.

308

: 00:16:01

It's a mess.

309

: 00:16:02

Yeah.

310

: 00:16:02

I have seen a picture,

and it was pretty simple.

311

: 00:16:05

But for the optics. Yeah.

312

: 00:16:07

But now they are trying to do

how tournament runs, for example.

313

: 00:16:12

So probably the next frontier is there

is there here that.

314

: 00:16:17

I see my shins that can drive themselves,

that can do things.

315

: 00:16:22

With the drones.

316

: 00:16:23

I think there was a lot of work done

by Palantir and these kind of companies,

317

: 00:16:26

but it's all for military use, obviously,

because that's where money goes first.

318

: 00:16:30

Yeah.

319

: 00:16:30

So I've seen some presentations

of of these things.

320

: 00:16:33

But yeah, it's so fascinating.

321

: 00:16:35

Yeah, that that's our current

technological environment.

322

: 00:16:40

We could say.

323

: 00:16:41

Yes I see.

324

: 00:16:42

So would you say that

the race that that we're sort of,

325

: 00:16:45

you know, part of right now,

now it's going to be more about

326

: 00:16:49

the efficiency of the eyes because

obviously they draw a lot of power,

327

: 00:16:52

all the data centers, etc..

328

: 00:16:54

There is this massive push in the AK,

for example, in the US, you know,

329

: 00:16:59

they all want to have the grid

that's capable of powering

330

: 00:17:01

a lot of data centers

to to cope with the gathering of the data.

331

: 00:17:05

Do you see any insights on on that front?

332

: 00:17:08

Well, you know,

the Chinese models are much smaller.

333

: 00:17:14

You can run it. Yes.

334

: 00:17:15

Not on a normal PC.

335

: 00:17:17

You know, I get this PC, we're talking

that there is my auxiliary here,

336

: 00:17:22

that I have a normal operating system.

337

: 00:17:25

I have one here.

338

: 00:17:27

But let's say that normal operating

339

: 00:17:30

systems today are windows, Mac OS,

and maybe it won't on

340

: 00:17:35

a couple of Linux Clipper.

341

: 00:17:39

Then in the upper.

342

: 00:17:39

This is a small normal machine

with eight bytes of memory.

343

: 00:17:44

In the other one I run a specific

OS for security and blah blah blah.

344

: 00:17:50

It has 64 bytes of memory.

345

: 00:17:54

It's a bit bigger,

346

: 00:17:56

but even

that is a bit small for running AI

347

: 00:18:00

because you need a graphics card.

348

: 00:18:04

Cards are very good at doing math

349

: 00:18:07

because all the graphics in the PC

are calculations.

350

: 00:18:10

Yeah. So yes.

351

: 00:18:14

Do you have a guy in a big

PC with 1 or 2 graphics cards?

352

: 00:18:20

You already can run AI today,

or if you rent a server,

353

: 00:18:25

the internet with a graphics card

you can run a.

354

: 00:18:29

The Chinese model are pretty small,

and if you don't want to keep your data

355

: 00:18:33

to anthropic or to open a,

you can read a private model and it works.

356

: 00:18:39

I've been doing it and it works.

357

: 00:18:41

And I'm not sure if the big companies

are that interested in optimizing.

358

: 00:18:46

They got all the money than me.

359

: 00:18:49

So sure. For them.

360

: 00:18:51

I've seen in the last few days

I've seen a guy presenting a new language

361

: 00:18:57

coding language,

which allows the LMS to compress

362

: 00:19:01

it a lot more than the traditional

languages used now.

363

: 00:19:04

So it becomes like way more efficient

because it can snapshot

364

: 00:19:07

what the code is supposed to be doing in a

in a bigger scale, basically.

365

: 00:19:11

So you can kind of zoom out and, you know,

366

: 00:19:13

you can see the classes,

the parameters, etc.

367

: 00:19:16

everything

that's involved in a much smaller

368

: 00:19:19

sort of zoomed out sort of scale in a way.

369

: 00:19:21

So, so you're like

370

: 00:19:22

keeping tabs on all individuals,

sort of parts of your application.

371

: 00:19:25

So that was interesting, you know, to say,

because it looks like that's going

372

: 00:19:29

to be part of the the direction as well

where everybody was going to go.

373

: 00:19:33

I think so,

because I think that many people,

374

: 00:19:37

including government,

are not really aware that

375

: 00:19:43

when you when you use the cloud,

376

: 00:19:46

the cloud is just someone else computer.

377

: 00:19:50

Exactly, exactly.

378

: 00:19:51

Yeah. People

don't realize that this is literally it.

379

: 00:19:54

It's you're using somebody else's

hard drive.

380

: 00:19:57

You're

just renting a computer on the internet.

381

: 00:20:00

Maybe it's a big computer.

382

: 00:20:02

Yeah, it's a rack or a number of racks or.

383

: 00:20:07

Yes. No block full of computers

384

: 00:20:11

that is such that when. You.

385

: 00:20:14

Use the cloud and when you use the

the language models from them,

386

: 00:20:18

you use it for an open air,

you use CloudFront dropping,

387

: 00:20:23

you're giving all your data to them

388

: 00:20:26

and you're giving some control to them.

389

: 00:20:30

Yes. Because one day they can say,

390

: 00:20:31

well, no, today you cannot use a computer.

391

: 00:20:35

Yeah, that could be complicated

in certain situations.

392

: 00:20:39

Yeah, I think that in a few years

393

: 00:20:43

maybe people and governments

and companies are going to become,

394

: 00:20:51

you know, Abuja or this of this.

395

: 00:20:54

Okay.

396

: 00:20:55

So okay.

397

: 00:20:58

You can speak today.

398

: 00:21:00

There are some conversations already

about digital sovereignty Okay.

399

: 00:21:06

And AI sovereignty.

400

: 00:21:08

Easy. Yes.

401

: 00:21:09

Are you data yours?

402

: 00:21:11

Really?

403

: 00:21:11

Are your computers yours? Really.

404

: 00:21:14

Are your AI yours?

405

: 00:21:16

Your yours renting from somebody else?

406

: 00:21:20

Exactly. Do you? Exactly.

407

: 00:21:21

Do you want to rent or you know,

408

: 00:21:26

the guns for the police from someone else.

409

: 00:21:29

And one day they can press

a button and say, no, you can't shoot.

410

: 00:21:33

You know, nothing. Works now.

411

: 00:21:35

Well, maybe you don't like it that much.

412

: 00:21:38

Yeah.

413

: 00:21:39

If you're renting the sheeps

and they say no,

414

: 00:21:42

they press a button

and they say, no, the ships don't.

415

: 00:21:45

Don't say it today. It's complicated.

416

: 00:21:48

Yeah, yeah.

417

: 00:21:48

So I think that this is

418

: 00:21:52

starting to, to become to become an issue.

419

: 00:21:55

And as much as people

is becoming aware of this,

420

: 00:21:59

they will say, no, wait,

we need our own data center.

421

: 00:22:03

We need to run our own AI.

422

: 00:22:06

We need to to keep our own data.

423

: 00:22:09

Yeah,

you look is starting to think about it.

424

: 00:22:12

I, I read about it.

425

: 00:22:15

They are a bit worried

that they are under US control

426

: 00:22:20

about the centers AI, the UK also.

427

: 00:22:25

Yeah.

428

: 00:22:25

I think that is this is going to become

a commonplace worry for many.

429

: 00:22:31

I'm not surprised

because it can become a very big issue,

430

: 00:22:34

like you say, because there is one

central point controlling everything.

431

: 00:22:38

And many economies

rely on this at this point.

432

: 00:22:41

Yeah, yeah.

433

: 00:22:42

The UK government's prediction from 2024,

434

: 00:22:46

if I remember correctly, was that 63

billion with a B is going to be lost

435

: 00:22:51

to the economy, to in the economy,

to the lack of digital skills.

436

: 00:22:55

So not just, you know,

like really like deep diving

437

: 00:22:59

into like coding and similar niches,

but in general digital skills.

438

: 00:23:02

And many entrepreneurs, business founders,

business owners don't feel comfortable

439

: 00:23:07

making decisions around

buying a specific piece of software.

440

: 00:23:10

And that's one of the reasons

why I've started this show to actually,

441

: 00:23:13

you know, bring some clarity around, okay,

what are the tools that are worth using,

442

: 00:23:18

what can you leverage

in your business, etc.?

443

: 00:23:20

So maybe if you could share a little bit

about

444

: 00:23:23

what's really useful from what you see,

obviously you create custom things.

445

: 00:23:26

So are there any specific trends

that you would see, you know, are,

446

: 00:23:30

you know, let's say,

do you see that many clients

447

: 00:23:33

ask for the same or similar thing

these days?

448

: 00:23:35

Is it, you know, along the lines of Zapier

and make

449

: 00:23:38

so you link different sort of applications

for them through AI layer?

450

: 00:23:42

You know, just just,

451

: 00:23:43

you know, give us some examples of

of what you would see these days.

452

: 00:23:46

I think that's something that many people

453

: 00:23:49

don't know how to ask or what

to ask first.

454

: 00:23:52

Okay.

455

: 00:23:53

Another thing I have been for many years,

how can I call myself

456

: 00:23:58

a supporter of what we call free software?

457

: 00:24:02

Yeah. Okay.

458

: 00:24:03

Yeah.

459

: 00:24:03

Free software doesn't mean.

460

: 00:24:06

Doesn't mean free as in free beer.

461

: 00:24:09

But it's more that free as in freedom.

462

: 00:24:11

Yeah. Because.

463

: 00:24:12

Got it.

464

: 00:24:13

Because, you know, to create. So for you.

465

: 00:24:16

You write code? Yeah. It doesn't.

466

: 00:24:20

You don't need to really technically

understand how it happens,

467

: 00:24:24

but you need to understand that

someone writes something in some language

468

: 00:24:29

which is not a human language,

but a language for computers.

469

: 00:24:33

I think that we are in the 21st century.

470

: 00:24:36

Many people can can get this

pretty much intuitively, you know.

471

: 00:24:42

Yes, programing language is okay.

472

: 00:24:44

Something that you write that they think

that we that we brought amorous.

473

: 00:24:50

And so for developers that we see

the screen and we understand everything.

474

: 00:24:54

No it's infinite.

475

: 00:24:56

Yeah I can read maybe five

programing languages.

476

: 00:25:00

Yeah. Maybe ten. Yeah.

477

: 00:25:02

But there are 200.

478

: 00:25:04

I need to call more

people if I need to. Yes.

479

: 00:25:09

Yeah.

480

: 00:25:09

But it's something that is written

for the computer

481

: 00:25:12

to convert into ones

and zeros and execute.

482

: 00:25:16

That's it.

483

: 00:25:16

That's a concept I think.

484

: 00:25:18

I think people are failing

in digital skills

485

: 00:25:22

because they are failing to have the

the concepts.

486

: 00:25:26

Okay.

487

: 00:25:27

They don't get the principles. Yeah.

488

: 00:25:30

There are like ten, 20 concepts

that you have to understand.

489

: 00:25:35

They they're not so complex, but

you need that someone explain them to you.

490

: 00:25:39

I didn't understand it by magic.

491

: 00:25:42

You know, I had to read or someone's

playing it to me and say, look, yes,

492

: 00:25:46

this is how it works.

493

: 00:25:47

This is a concept

and that concepts translate to software.

494

: 00:25:52

And then one concept is, yeah,

someone writes this code

495

: 00:25:58

and it's the computer translates

to one and zero.

496

: 00:26:02

And two things can happen.

497

: 00:26:04

Yeah.

498

: 00:26:05

They give you the code

499

: 00:26:09

audacious keep you the translated version

500

: 00:26:12

into one and zero into ones and zeros,

501

: 00:26:15

which is the executive

that we call the executable.

502

: 00:26:19

You're trapped. Yes. Okay.

503

: 00:26:21

If you have to execute

a woman, you're trapped.

504

: 00:26:26

Yeah.

505

: 00:26:26

They will.

506

: 00:26:27

You can't do anything.

507

: 00:26:28

You can't do.

508

: 00:26:29

You can run it. Yeah.

509

: 00:26:31

Can you run you your business. Yeah.

510

: 00:26:33

But you can modify. It here today.

511

: 00:26:35

You can write your business.

512

: 00:26:37

What happens tomorrow?

513

: 00:26:38

The market changes, the government

regulation

514

: 00:26:41

changes, your business changes.

515

: 00:26:44

You say, hey, let's do it this way.

516

: 00:26:47

Because we may

we will make a bit more money.

517

: 00:26:50

You need to change the software.

518

: 00:26:52

If you don't have the code,

you are tied up,

519

: 00:26:55

your hands are changed

and you can do nothing.

520

: 00:26:59

That's the key of digital dependency.

521

: 00:27:04

Yeah. Yeah.

522

: 00:27:06

So by being a big defender

523

: 00:27:09

of free software free software

means it comes with the code.

524

: 00:27:14

And so you can modify it yourself.

525

: 00:27:17

Now if you're not a developer. No.

526

: 00:27:19

But you can you can contract any developer

to modify it.

527

: 00:27:24

You can do a lot of things. Okay.

528

: 00:27:28

I think

529

: 00:27:29

that most people don't understand this.

530

: 00:27:32

And they so they they fall

531

: 00:27:35

into the trap of to resolve for.

532

: 00:27:38

Yeah.

533

: 00:27:39

Where the, where the creator keeps

the code and just sells

534

: 00:27:44

you or rents you.

535

: 00:27:48

The rights to use it.

536

: 00:27:49

Yeah, they're a permission to use it,

not the right.

537

: 00:27:52

It's just a permission

to use it for a while.

538

: 00:27:55

You have no rights.

539

: 00:27:57

As a matter of fact, you only have rights

when you have the code.

540

: 00:28:00

Yeah.

541

: 00:28:01

So I will say for many, many businesses

it will do them a lot of good.

542

: 00:28:07

If they book, try to get closer

543

: 00:28:11

to the free software community.

544

: 00:28:15

They see.

545

: 00:28:16

If they can run their business

on free software,

546

: 00:28:20

which is possible for mostly any business.

547

: 00:28:24

Yeah, it's easier than it looks.

548

: 00:28:26

Of course you need a

you need a bit of help. Yes.

549

: 00:28:29

You always need a bit of help.

550

: 00:28:31

Okay.

551

: 00:28:32

We need doctors, dentists.

552

: 00:28:34

We even need people to tailor our clothes.

553

: 00:28:37

Living a specialized society.

554

: 00:28:40

But the thing here

is, you're not trapped for life.

555

: 00:28:43

You can't change if the vendor doesn't

give you a good deal,

556

: 00:28:47

you can switch to the easily

to another band.

557

: 00:28:51

Yeah, and the same goes with the guide.

558

: 00:28:53

As long as you use cloud or you pity,

you are mostly trapped in

559

: 00:28:59

when you start using free

software or independent.

560

: 00:29:03

Yeah, I think this is something

that it's not people doesn't understand.

561

: 00:29:08

People don't understand it mostly.

562

: 00:29:10

So I won't.

563

: 00:29:11

Say people who create software

obviously will know about it

564

: 00:29:14

because everybody will say,

oh, I need to have my own devs.

565

: 00:29:17

I don't I cannot outsource

creating the app from the start

566

: 00:29:20

because obviously they would own the code.

567

: 00:29:22

So that's that's where yeah,

also that's where it becomes expensive.

568

: 00:29:25

Right?

569

: 00:29:25

Because when you have your own team,

you control what happens, how many people

570

: 00:29:29

you assign to it

where you just, you know, send an order

571

: 00:29:33

to create me an app

which does this, this, this and that.

572

: 00:29:36

Obviously, you're not involved

in the creation process.

573

: 00:29:39

Literally, right? Yeah.

574

: 00:29:40

Yeah.

575

: 00:29:41

So I think that it is hard

to pick problems

576

: 00:29:44

that people have today

that make so far so complex.

577

: 00:29:48

One, they don't have the basic concepts

which are not technical mostly.

578

: 00:29:53

Yeah, there's a solution thing,

you know, okay.

579

: 00:29:56

It's a solution can sing mostly anybody.

580

: 00:29:59

I work many times, I teach classes

and so of course,

581

: 00:30:03

and mostly anybody with any background

doctors, business

582

: 00:30:08

people, accountants,

mostly anybody can understand it.

583

: 00:30:12

You write this code,

584

: 00:30:13

the computer translated to one

and zeros is not science fiction anymore.

585

: 00:30:17

You know, it's almost 100 years old

586

: 00:30:21

today

that we are talking about all the future.

587

: 00:30:24

No, it's not the future anymore.

588

: 00:30:26

It has been the present for many decades.

589

: 00:30:29

But when people had has these concepts,

then they can reason about it.

590

: 00:30:35

This little thing. Yeah.

591

: 00:30:36

And when they, they become used to

to using free software,

592

: 00:30:41

they say, oh, this is mine

or is in the public domain.

593

: 00:30:44

It's like you, you know,

if you, you're making cars.

594

: 00:30:48

Yeah. Let's say you're making cars. Okay.

595

: 00:30:51

You say fatty, my god this semester

and we are going to do cars in the UK.

596

: 00:30:56

Yeah. Okay.

597

: 00:30:57

You need drawings. Yeah. But today.

598

: 00:31:01

Of. Course,

we don't do drawings by hand anymore.

599

: 00:31:04

We use cat computer aided design,

which when you see the drawing

600

: 00:31:08

in the screen that you can rotate it

and you can say,

601

: 00:31:12

make it bigger, make it smaller,

take this and and that and that.

602

: 00:31:16

It's obvious

that you will want to have the cat files.

603

: 00:31:21

Yeah.

604

: 00:31:21

Of course, of course. Yes.

605

: 00:31:23

Because you can find them. Next year.

606

: 00:31:25

You can take the take the cat file

and modify it and make a new car.

607

: 00:31:31

If the cat files are owned

by a third party, your chain again.

608

: 00:31:37

Okay. Yes, yes.

609

: 00:31:39

So this thing of digital ownership

610

: 00:31:42

and this little subreddit

is what many people don't understand.

611

: 00:31:48

It's very interesting that you say that.

612

: 00:31:50

It's very interesting

from what I've seen, 100%.

613

: 00:31:53

I agree,

because I've seen a single business who's

614

: 00:31:56

using Linux as their main system

for the actual users within the business,

615

: 00:32:01

everybody else is windows,

the typical stock exchange

616

: 00:32:05

for for actual emails

and communications and stuff like this.

617

: 00:32:08

A single business out

of, let's say 200, 300.

618

: 00:32:12

You know, however many I've seen so far

619

: 00:32:14

was using Linux as their system,

you know, for the servers as well.

620

: 00:32:18

It was. Yeah. Yeah.

621

: 00:32:20

Really it's not as, as common knowledge.

622

: 00:32:23

Right?

623

: 00:32:23

It's not common.

624

: 00:32:24

But today is the key to being to be

the owner of your life.

625

: 00:32:30

That's what I'm favor.

626

: 00:32:32

Let's say I like the team people

627

: 00:32:36

to be the owners of the digital life.

628

: 00:32:39

And if you're a country,

you can be sobering.

629

: 00:32:42

Yeah. Okay.

630

: 00:32:44

Things that that's what people

631

: 00:32:47

doesn't understand these

because the so far industry

632

: 00:32:51

don't tell them

because they don't want to be independent.

633

: 00:32:55

Not in their interest because.

634

: 00:32:57

It's a, it's a it's an interest here.

635

: 00:32:59

They don't want us to be independent.

636

: 00:33:02

They want to just to be our masters.

637

: 00:33:05

This is really true.

638

: 00:33:06

Okay.

639

: 00:33:07

They will say, oh,

you're a conspiracy theorist.

640

: 00:33:10

No, no no no.

641

: 00:33:12

Look, I don't I don't need no theory

642

: 00:33:15

and I don't need to to,

643

: 00:33:18

you know, create any story.

644

: 00:33:21

I've been inside the sofa industry

645

: 00:33:24

for 43 years now.

646

: 00:33:28

Okay?

647

: 00:33:28

I started when I was 17 years old.

648

: 00:33:32

Yeah.

649

: 00:33:33

And I'm 60 now, so

650

: 00:33:37

I'm an insider.

651

: 00:33:38

You could say I'm an insider.

652

: 00:33:40

I work at in all kind of places,

a multinational corporation,

653

: 00:33:45

software companies, small companies.

654

: 00:33:48

I had my own company for many years.

655

: 00:33:51

I'm an insider.

656

: 00:33:52

So what I tell you?

657

: 00:33:54

Yes, I would say that model less.

658

: 00:33:56

Believe me, I know what I'm talking about.

659

: 00:34:00

Yeah, but I talk about this.

660

: 00:34:01

How you can be free in a digital world

and how can you be?

661

: 00:34:05

Yeah.

662

: 00:34:06

You know, anyone who has lived

for a number of years

663

: 00:34:10

knows that if you own your house,

664

: 00:34:13

you're in a much better position,

that if you rent first, codes first,

665

: 00:34:17

and if you own your car,

you're in a much better position

666

: 00:34:20

that if you rent a car or take a taxi,

you're the whole being the whole.

667

: 00:34:25

In every society in history,

free people were the ones

668

: 00:34:29

who owned land

owned the house owned animals.

669

: 00:34:34

Owned machines.

670

: 00:34:35

Yeah, yeah, it's the way,

it's the way it is.

671

: 00:34:39

You know.

672

: 00:34:40

The world didn't really change that much.

673

: 00:34:42

No, in that sense, no.

674

: 00:34:44

But what is not is not intuitive.

675

: 00:34:48

It's.

676

: 00:34:48

I would say

this is one of my favorite words.

677

: 00:34:52

It's counterintuitive to understand

678

: 00:34:55

that in the digital world,

you can also be an owner or a tenant.

679

: 00:35:00

Yeah.

680

: 00:35:01

Yes. It's not obvious because we live now.

681

: 00:35:05

I tend to explain when I teach courses.

682

: 00:35:08

And so that we have three parallel worlds

today.

683

: 00:35:11

We got the physical

world. Yeah. Of course.

684

: 00:35:15

Yeah.

685

: 00:35:15

But since maybe 200

686

: 00:35:18

years ago, we have a financial world.

687

: 00:35:22

Okay. Which is kind of superimpose.

688

: 00:35:25

Yeah. To the physical world.

689

: 00:35:27

But in the last hundred years

we have a third world, the physical,

690

: 00:35:32

the financial and the digital world

and the interplay.

691

: 00:35:36

Interplay. Okay. Yeah.

692

: 00:35:39

And in these three parallel worlds,

you can be an owner or a ten.

693

: 00:35:44

The three of them, the physical,

the financial and the digital.

694

: 00:35:48

The owner has all of the advantages here.

695

: 00:35:51

That's very true.

696

: 00:35:52

That's my philosophy. I think it's facts.

697

: 00:35:56

It's literally just, you know,

this is like the truth, isn't it?

698

: 00:35:59

You know, we see it,

you know, me coming from the corporate,

699

: 00:36:03

you know, background as well.

700

: 00:36:04

You know, I've seen it,

you know, from, from the inside.

701

: 00:36:07

Right.

702

: 00:36:07

If you own the application,

you just basically

703

: 00:36:11

sell, you know, imaginary licenses, right?

704

: 00:36:14

It doesn't cost you anything

because you're not generating any physical

705

: 00:36:17

objects.

706

: 00:36:18

It's just a piece of code saying yes,

707

: 00:36:20

somebody X, Y, and Z can use it

for a year, and then a year later

708

: 00:36:24

they have to pay again because otherwise

we will switch it off for them.

709

: 00:36:27

It's like money.

710

: 00:36:30

Yeah.

711

: 00:36:30

Yes, exactly.

712

: 00:36:31

That's how Microsoft became so powerful.

713

: 00:36:35

Yeah.

714

: 00:36:35

Printing windows

licenses was like 20 money here.

715

: 00:36:40

And so they they were the

they had the first net worth.

716

: 00:36:45

They were bigger

bigger bigger bigger bigger bigger.

717

: 00:36:48

Yeah. Yeah, yeah.

718

: 00:36:50

Because they were the owners of windows.

719

: 00:36:53

And so they became so rich.

720

: 00:36:55

Yeah yeah. Yeah yeah.

721

: 00:36:58

Okay.

722

: 00:36:58

And you as a,

723

: 00:37:00

you know, as a, as a contractor,

you know, creating custom applications.

724

: 00:37:03

What are the useful tools for you

these days.

725

: 00:37:06

What would you say

726

: 00:37:07

is the core of what's helping

you deliver the services that you do?

727

: 00:37:11

And from what I remember, you know,

728

: 00:37:12

you obviously have a team

that delivers on the work.

729

: 00:37:15

But you know what?

730

: 00:37:16

What is the latest thing

that sort of helps you the most, you know,

731

: 00:37:19

to run this kind of business?

732

: 00:37:20

Well, these last two years

we have been switching

733

: 00:37:24

to using a lot of AI,

but we know how to use it and we.

734

: 00:37:28

Whatever we do, we keep the code

735

: 00:37:31

so we don't get tied up to an AI

or the other.

736

: 00:37:36

Yeah, yeah.

737

: 00:37:37

That's been a big make a change.

738

: 00:37:41

Yeah, yeah.

739

: 00:37:42

What about some like basic things that are

that say off the shelf communications?

740

: 00:37:46

Because I'm assuming you wouldn't have

a custom thing for such a simple thing

741

: 00:37:50

like comms.

742

: 00:37:51

Or do you actually have your own thing

for communications between teams?

743

: 00:37:54

That depends on how much privacy you need.

744

: 00:37:57

Yeah, okay.

745

: 00:37:59

If there are like three labels,

746

: 00:38:02

normal privacy won't harm you.

747

: 00:38:05

If it like if it's like him.

748

: 00:38:07

Well, you can use anything.

749

: 00:38:11

Yeah.

750

: 00:38:11

If you teams zoom all these slack.

751

: 00:38:15

But yeah yeah. Yeah yeah. No.

752

: 00:38:18

If you need to handle commercial secrets,

753

: 00:38:22

industrial secrets, whatever,

then you need a second level of privacy.

754

: 00:38:27

You need to use free tools

that are audited because, for example,

755

: 00:38:32

WhatsApp tells you

communications are encrypted.

756

: 00:38:37

Yeah.

757

: 00:38:37

But there. Yeah, you can get to it

from what I.

758

: 00:38:39

Yeah. But yeah.

759

: 00:38:41

What's the key?

760

: 00:38:42

Yeah. Close.

761

: 00:38:45

Yeah. Okay.

762

: 00:38:46

Now tell me who has a key.

763

: 00:38:49

It's not enough that that is always closed

okay.

764

: 00:38:52

It's lock.

765

: 00:38:53

The door is locked.

766

: 00:38:54

But now give me a license. Who has a key?

767

: 00:38:58

Because if there's a hundred keys around,

well, there is safe.

768

: 00:39:03

It isn't very safe. Yes.

769

: 00:39:05

So if you need more privacy,

you need a different set of tools,

770

: 00:39:10

usually open source tools

that are out by a lot of people.

771

: 00:39:15

And, you know, nobody's

reading your communications.

772

: 00:39:19

If you need high level privacy,

then you need

773

: 00:39:23

the free tools with a specialized setup.

774

: 00:39:26

I think those are the three levels.

775

: 00:39:28

It's it's first you need to

we need to think about

776

: 00:39:34

what is critical for your business,

for your organization, what is critical.

777

: 00:39:39

Okay.

778

: 00:39:39

There's not a recipe that you can use

for everybody.

779

: 00:39:43

But of course having concepts,

having understanding some concepts,

780

: 00:39:48

you can start to discuss those things

where.

781

: 00:39:51

It's a bit

like being diagnosed by a doctor.

782

: 00:39:53

You have to have a list of symptoms.

783

: 00:39:55

You know, the requirements, the problems.

784

: 00:39:58

Yeah, I see. What you do. Yeah.

785

: 00:40:01

Because, you know, are you

are you sitting many hours.

786

: 00:40:05

Do you walk? Yeah. Walking. Yeah.

787

: 00:40:09

Do you agree situation is different. Okay.

788

: 00:40:12

Yes. Yes.

789

: 00:40:14

Every situation is different.

790

: 00:40:16

So yeah diagnosis is the first thing.

791

: 00:40:19

Speaking clients about what they do,

what is critical for them.

792

: 00:40:24

What will be critical

not to help here okay.

793

: 00:40:29

And then we start giving recommendations

794

: 00:40:32

and maybe building things.

795

: 00:40:35

Yeah.

796

: 00:40:36

So does it ever happen that, you know, you

you know,

797

: 00:40:39

let's say the client approaches

you and they, you know,

798

: 00:40:42

they think

that they will need something custom.

799

: 00:40:44

But it turns out,

800

: 00:40:45

no, you just buy this thing because it's

going to do everything that you need.

801

: 00:40:48

Do you ever say that

or it actually ends up being a

802

: 00:40:51

I don't want to rip off

people of their money.

803

: 00:40:56

We try to keep value. Yeah.

804

: 00:40:58

So if you can spend less, spend

less this year okay.

805

: 00:41:03

Next year, maybe I will have to tell you.

806

: 00:41:05

Look, this is pensive.

807

: 00:41:07

Sorry.

808

: 00:41:07

I can do nothing.

809

: 00:41:08

Yeah, but if it can be cheaper

to do it. Cheaper?

810

: 00:41:11

Of course.

811

: 00:41:11

That's what a good brand.

812

: 00:41:13

Or does it feel you go.

813

: 00:41:16

So you build. Trust me.

814

: 00:41:18

Go to buy a car.

815

: 00:41:20

And they are all always trying to

to sell you the most expensive car.

816

: 00:41:25

You're going to start having the help

817

: 00:41:29

of course.

818

: 00:41:29

Say they only want to rip me off.

819

: 00:41:32

I don't need her. Yes.

820

: 00:41:33

Don't need a Ferrari to go

pick up the children.

821

: 00:41:37

Yes, yes.

822

: 00:41:38

Many times in the conversations

that I have on the show,

823

: 00:41:41

it basically becomes very clear.

Sometimes you just need to.

824

: 00:41:44

Just to start the business.

825

: 00:41:45

You need a fancy spreadsheet, right?

826

: 00:41:47

Just so you have a piece of data

somewhere,

827

: 00:41:50

just so you keep a record of something.

828

: 00:41:51

And then over time

you switch to something more elaborate,

829

: 00:41:54

something more,

you know, you know, complex, etc.

830

: 00:41:58

you just build on top of that,

that simple base.

831

: 00:42:01

You don't need to spend 100,000 from day

one on a fancy software,

832

: 00:42:05

because you're not going to use

even the half of it.

833

: 00:42:08

Suddenly you need to adapt to your scale

and to on your level of sophistication.

834

: 00:42:13

If you say, no,

I'm going to be an atomic center.

835

: 00:42:18

Well, okay.

836

: 00:42:19

Yeah, yeah, you will have to spend.

837

: 00:42:21

Yeah, that's

that's the difference. Yes. Of course.

838

: 00:42:23

Everything will be a better bit expensive.

839

: 00:42:26

Yeah, but if you're running

a simple operation okay.

840

: 00:42:30

Well it's simple software.

841

: 00:42:32

Yeah, it should be true.

842

: 00:42:33

Yeah, yeah,

it should adapt to your scale and your.

843

: 00:42:36

And this sophistication of your operation.

844

: 00:42:39

Probably. And yes.

845

: 00:42:42

Probably

846

: 00:42:44

both.

847

: 00:42:44

Most organizations have a 90

bad cut of 1%,

848

: 00:42:50

maybe 2% in some cases.

849

: 00:42:53

Yes. If you're going up,

you got to justify it okay.

850

: 00:42:57

Yeah. First of course.

851

: 00:42:58

Yeah I've seen that up close.

852

: 00:43:00

At some point in the previous life

I was in automotive business.

853

: 00:43:04

It was a global company.

854

: 00:43:05

And you know, we were sort of making sure

that the purchasing applications,

855

: 00:43:09

they were all custom made from scratch.

856

: 00:43:11

It was for SAP environment.

857

: 00:43:13

So we were using react as a layer

858

: 00:43:16

to communicate with the back end in SAP.

859

: 00:43:20

So it was functioning okay

and in terms of the performance etc..

860

: 00:43:25

But you some of the apps

that the guys had to look at the code

861

: 00:43:28

was, you know, older than the developers

looking at the code.

862

: 00:43:32

So it was

863

: 00:43:32

really interesting thing to sort of,

you know, you, you know, actually see.

864

: 00:43:36

Yeah,

yeah yeah, yeah. So far is an investment.

865

: 00:43:39

And look, I remember doing an application

866

: 00:43:43

for a client in the nest

at the beginning of this.

867

: 00:43:46

As a matter of fact,

there is still running it.

868

: 00:43:49

Yeah. Wow.

869

: 00:43:52

Yeah.

870

: 00:43:52

Sergey, you know, I said, well,

still running.

871

: 00:43:55

I know this guy still works for them.

872

: 00:43:58

From time to time.

873

: 00:43:59

I speak with him.

874

: 00:44:00

He said, oh yeah right.

875

: 00:44:03

Yeah yeah, yeah.

876

: 00:44:04

But you know, there

that only tells me that you've done

877

: 00:44:07

a great job back then

because it didn't need change for all.

878

: 00:44:11

They,

they give him the time to do it. Yeah.

879

: 00:44:13

I say look, if you want this

well done, let me walk on it.

880

: 00:44:18

And I was

I bought it on it for four years.

881

: 00:44:21

It was a lot.

But turns out to be worth it.

882

: 00:44:24

30 years later they are still using it.

883

: 00:44:27

Still using it.

884

: 00:44:28

Yeah. So wow.

885

: 00:44:30

You got to have some perspective.

886

: 00:44:32

You know, what you're doing is important

long term for your business.

887

: 00:44:37

Or this is just something

that is for the time being.

888

: 00:44:41

Yeah. Those are things to consider. Yeah.

889

: 00:44:45

In some. Things I.

890

: 00:44:46

See you want

you do want to spend some real money.

891

: 00:44:50

Yeah.

892

: 00:44:50

Because his core business

and probably won't change, you know.

893

: 00:44:54

Yeah, yeah.

894

: 00:44:55

Okay.

895

: 00:44:56

I've got a couple of questions

and I've got an impression

896

: 00:44:59

that I will know what you're going to say,

because it ties

897

: 00:45:02

with the ownership of the digital assets,

let's say.

898

: 00:45:05

But I've got a small community,

and I always give them an opportunity

899

: 00:45:09

to ask a question to the guest.

900

: 00:45:11

So in your case,

because you are so closely tied with

901

: 00:45:15

with AI and relevant work, you know,

what would you say about the process of,

902

: 00:45:20

okay, somebody has a concept of an app

and they go and let's say build

903

: 00:45:24

something on it, or base 44, one of those

low code or no code platforms, right.

904

: 00:45:29

Obviously

they don't own the source code, right.

905

: 00:45:31

So that's one obstacle I can already

foresee that you're going to point out.

906

: 00:45:35

But then what would you advise these guys

to do if they want to eventually own it?

907

: 00:45:40

What would you sort of

advise them to to do?

908

: 00:45:42

Because if they really want it

to be their thing and then sell it

909

: 00:45:45

and make it into a business,

what would you advise them to do?

910

: 00:45:48

I would say go a status,

which is my favorite recipe going station.

911

: 00:45:54

Okay, you got the idea.

912

: 00:45:56

Well of course, go and got it. Go.

913

: 00:45:58

No gold. Whatever.

914

: 00:46:00

Do it.

915

: 00:46:00

Because, yeah, you know,

we'll have a first version and it's much

916

: 00:46:06

better to discuss over a first version,

no matter how imperfect is.

917

: 00:46:11

But we are seeing it

and we can discuss it.

918

: 00:46:15

Yeah. So yeah.

919

: 00:46:16

Yeah. Please go and do that first version.

920

: 00:46:20

You know, when people come to me and say,

I got this idea for

921

: 00:46:24

I said, put it on paper

so we can discuss it better if you do it

922

: 00:46:28

on Replit on whatever,

no code platform, please do it.

923

: 00:46:32

But you got to know that it will serve you

for a while when your business scales.

924

: 00:46:38

If you are successful and you start

having more clients, more clients,

925

: 00:46:42

more clients, obviously there will come

a point where it's not big enough.

926

: 00:46:48

Yeah.

927

: 00:46:50

Yes. Then

928

: 00:46:51

then then

when you're seeing that point approaching,

929

: 00:46:55

try not to wait until it's bars.

930

: 00:47:00

Like in everything in life,

931

: 00:47:01

try not to wait until it bars.

932

: 00:47:05

Yeah.

933

: 00:47:05

When you sit that it's

starting to feel a bit uncomfortable,

934

: 00:47:10

bring in someone who knows what he or

she is doing.

935

: 00:47:14

Well, I will, but that can take the arc

and say, well, that is okay.

936

: 00:47:21

Now let's discuss

how it is serving your business.

937

: 00:47:25

What?

938

: 00:47:26

What do you see

a hit in the next two years?

939

: 00:47:29

What will you want to add?

940

: 00:47:31

What will you want to change?

941

: 00:47:33

And then the one you were shown

professionally? Yes.

942

: 00:47:37

And in this new version, you can use more

943

: 00:47:40

sophisticated tools and keep the code.

944

: 00:47:43

Yeah.

945

: 00:47:44

And you start in there.

946

: 00:47:46

You got a real business application.

947

: 00:47:48

Yeah, but having a first version is

always fantastic.

948

: 00:47:52

You know,

949

: 00:47:54

best based applications

950

: 00:47:57

usually have a first version

that maybe is not so good.

951

: 00:48:02

And you just improve it over time.

952

: 00:48:04

Yeah.

953

: 00:48:04

Because then you can discuss and

and improve it.

954

: 00:48:07

You know something that I do,

I like for example, I like an artist

955

: 00:48:11

know a group, a singer, something.

956

: 00:48:14

I go to Wikipedia.

957

: 00:48:15

Yeah. And I see the list of their albums.

958

: 00:48:18

Yeah.

959

: 00:48:19

And I stopped listening to them on YouTube

starting from the first album.

960

: 00:48:25

Yeah.

961

: 00:48:25

And going ahead

second, third, fourth, fifth.

962

: 00:48:28

And sometimes it's incredibly surprising

963

: 00:48:31

how bad the first album

is, is not very good.

964

: 00:48:35

I wasn't sure where

you're going with this, but now I get it.

965

: 00:48:38

Okay. It's.

966

: 00:48:40

Yeah. You're saying.

967

: 00:48:41

Yeah. It's interesting.

968

: 00:48:42

Yeah, that's a good song.

969

: 00:48:44

But I mean, big artists.

970

: 00:48:46

No, no. Like I did it with

971

: 00:48:49

let's say

972

: 00:48:51

first album has a good song.

973

: 00:48:53

The rest.

974

: 00:48:54

Well it's interesting this

like they are searching for something,

975

: 00:48:59

but clearly they don't find it.

Yes and yes.

976

: 00:49:02

Listen, at the second one

you say you're catching it.

977

: 00:49:06

And then the third album is wow, it's

very good.

978

: 00:49:09

And the fourth, wow, that's a big album

979

: 00:49:12

that is historical,

but applications are the same.

980

: 00:49:16

You do.

981

: 00:49:16

The first version is not very good,

but it it starts

982

: 00:49:21

rolling the discussion.

983

: 00:49:24

Yes. Same as probably the musicians

984

: 00:49:26

had a

lot of heavy discussions in the room.

985

: 00:49:31

Say no, no, don't play that.

986

: 00:49:32

Don't you see that you're

breaking everything, guy to change it?

987

: 00:49:37

No, no.

988

: 00:49:37

But that happens in real life

so far is the same.

989

: 00:49:41

You do the first version of the app

and then you say, look,

990

: 00:49:45

these two screens are horrible.

991

: 00:49:47

Users are complaining these two work well.

992

: 00:49:50

We want to more.

993

: 00:49:52

We want.

994

: 00:49:52

And then you do a second version

and is much better

995

: 00:49:56

that system

I told you that they did in the 90s.

996

: 00:50:00

First version was designed by another guy.

997

: 00:50:04

Yeah. Okay.

998

: 00:50:05

He gave us the design and say,

write this application with it.

999

: 00:50:09

It was horrible.

: 1000

00:50:10,480 --> 00:50:14,560

When the user got it, said, no,

I worked the other way around.

: 1001

00:50:15,960 --> 00:50:17,320

Everybody thinks differently.

: 1002

00:50:17,320 --> 00:50:19,280

That's another problem, right?

: 1003

00:50:19,280 --> 00:50:22,200

But then they call me and say,

can you fix this?

: 1004

00:50:22,200 --> 00:50:23,200

And you know what?

: 1005

00:50:23,200 --> 00:50:28,760

Having that first person

that was a new sabo was very good

: 1006

00:50:28,760 --> 00:50:33,640

because the user saw it on the screen

and say, no, I do it the other way around.

: 1007

00:50:33,680 --> 00:50:38,000

I cannot start here because I don't have

the data I need to start here.

: 1008

00:50:38,240 --> 00:50:41,320

Then you do this calculation and

: 1009

00:50:41,320 --> 00:50:45,200

and tell me this and this and this,

and I will be able to do this.

: 1010

00:50:45,240 --> 00:50:48,080

And that was for me. Fantastic.

: 1011

00:50:48,080 --> 00:50:51,080

I wrote it, I didn't tell them

: 1012

00:50:51,920 --> 00:50:55,080

because they won't say yes anyway.

: 1013

00:50:55,120 --> 00:51:01,960

But I told them that I was like going

to a route 50% because that was my idea.

: 1014

00:51:02,000 --> 00:51:03,800

But then when I sat down,

: 1015

00:51:03,800 --> 00:51:07,880

I said, no, I need to start from zero

and I started from zero.

: 1016

00:51:07,880 --> 00:51:09,440

But I knew what to do.

: 1017

00:51:09,440 --> 00:51:13,160

And the second version was pretty good,

you know,

: 1018

00:51:13,200 --> 00:51:16,480

six months later I had a second version

: 1019

00:51:16,600 --> 00:51:19,480

was so good, call him by hand.

: 1020

00:51:19,480 --> 00:51:20,240

At that time.

: 1021

00:51:20,240 --> 00:51:22,160

It took me like for six months

: 1022

00:51:22,160 --> 00:51:26,680

to have a usable Belarusian,

but when I heard it, it was pretty good.

: 1023

00:51:26,720 --> 00:51:31,800

They sat down and tried it and say, this

saves me for the save me a lot of time.

: 1024

00:51:32,480 --> 00:51:34,280

And that's it. Yeah.

: 1025

00:51:34,280 --> 00:51:39,040

And we kept on building versions

and after three years

: 1026

00:51:39,040 --> 00:51:43,000

it was all of the team

which was like 12th person

: 1027

00:51:43,040 --> 00:51:46,240

were working together,

sharing the same database.

: 1028

00:51:46,280 --> 00:51:50,320

They will touch something on one screen,

it will show, on the other,

: 1029

00:51:50,360 --> 00:51:51,760

it will show on the other.

: 1030

00:51:51,760 --> 00:51:54,840

They were it changed completely.

: 1031

00:51:54,880 --> 00:51:56,040

But they work.

: 1032

00:51:56,040 --> 00:51:57,400

They were working. Yeah.

: 1033

00:51:57,400 --> 00:52:01,080

So I would say that going is always yeah,

: 1034

00:52:01,160 --> 00:52:05,920

don't try to to

to jump to do it in one shot. Yes.

: 1035

00:52:06,880 --> 00:52:07,320

Okay.

: 1036

00:52:07,320 --> 00:52:08,760

So the end goal

: 1037

00:52:08,760 --> 00:52:12,520

should be to recreate the concept

that you've built on that third party

: 1038

00:52:12,520 --> 00:52:16,320

platform with your own developers,

with the team that you would hire to

: 1039

00:52:16,360 --> 00:52:17,640

actually build it from zero.

: 1040

00:52:17,640 --> 00:52:20,040

But following that draft. Yeah.

: 1041

00:52:20,040 --> 00:52:21,560

For the win or. Not, very good.

: 1042

00:52:21,560 --> 00:52:23,880

But you have something to discuss.

: 1043

00:52:23,880 --> 00:52:25,000

That's fantastic.

: 1044

00:52:25,000 --> 00:52:28,280

Of course,

because it's concrete, it's real.

: 1045

00:52:28,280 --> 00:52:29,800

And you can say no, no, no.

: 1046

00:52:29,800 --> 00:52:30,080

Yeah.

: 1047

00:52:30,080 --> 00:52:33,520

No, don't put that in that screen

because I don't have it.

: 1048

00:52:33,560 --> 00:52:34,800

Yeah. When I'm talking.

: 1049

00:52:34,800 --> 00:52:36,960

So you have that feedback work on. Yeah.

: 1050

00:52:36,960 --> 00:52:40,160

Because so far show with be

: 1051

00:52:40,160 --> 00:52:43,160

a reflection of what you do in your work.

: 1052

00:52:43,600 --> 00:52:47,040

And if you put some

something in the screen when you're

: 1053

00:52:47,040 --> 00:52:50,040

talking on the phone with the client

and you don't have it

: 1054

00:52:50,040 --> 00:52:54,680

and you can cannot pass

to the next screen, it's a nightmare.

: 1055

00:52:54,760 --> 00:52:57,320

Yeah, you cannot work. But

: 1056

00:52:58,520 --> 00:52:59,080

once you

: 1057

00:52:59,080 --> 00:53:02,120

have a person, you can say, no, no,

take that out.

: 1058

00:53:02,160 --> 00:53:03,160

That goes.

: 1059

00:53:03,160 --> 00:53:06,360

Today's later, after

I spoke with the client.

: 1060

00:53:06,360 --> 00:53:08,360

After we did some work.

: 1061

00:53:08,360 --> 00:53:11,400

After then I have such data.

: 1062

00:53:11,440 --> 00:53:12,120

Yeah.

: 1063

00:53:12,120 --> 00:53:14,280

That's that's

kind of what I discuss when I do

: 1064

00:53:14,280 --> 00:53:16,320

my tech stack audits with clients, right?

: 1065

00:53:16,320 --> 00:53:18,280

You know, we exactly do that. So. Okay.

: 1066

00:53:18,280 --> 00:53:19,760

How do you physically do things?

: 1067

00:53:19,760 --> 00:53:21,360

Where does the paper go?

: 1068

00:53:21,360 --> 00:53:24,440

Like what's the paper trail,

you know, is is the root card, you know,

: 1069

00:53:24,480 --> 00:53:27,080

being generated with this data

or with that data.

: 1070

00:53:27,080 --> 00:53:28,360

Where does it come from.

: 1071

00:53:28,360 --> 00:53:30,720

So we have those exact

sort of conversations.

: 1072

00:53:30,720 --> 00:53:31,920

So 100%.

: 1073

00:53:31,920 --> 00:53:32,600

Yeah I agree.

: 1074

00:53:32,600 --> 00:53:35,920

Thank you for saying that because

yeah it's great to hear it from an expert.

: 1075

00:53:35,960 --> 00:53:38,600

You know, with such

you know, such a massive experience.

: 1076

00:53:38,600 --> 00:53:39,040

You know.

: 1077

00:53:39,040 --> 00:53:44,960

When they get used to

to using the up and go become a child.

: 1078

00:53:45,120 --> 00:53:50,080

Because if you see the app they will come

saying we can do this better, you know.

: 1079

00:53:50,880 --> 00:53:52,360

Yeah, we could.

: 1080

00:53:52,360 --> 00:53:54,480

Switch the order of the steps.

: 1081

00:53:54,480 --> 00:53:56,880

We could do this way, that way.

: 1082

00:53:56,880 --> 00:54:00,880

Because they in the process,

we all became spells

: 1083

00:54:00,880 --> 00:54:03,880

in the in the thing that we were doing,

you know.

: 1084

00:54:04,080 --> 00:54:05,080

Yes. Yeah.

: 1085

00:54:05,080 --> 00:54:05,720

Yeah yeah.

: 1086

00:54:05,720 --> 00:54:09,280

That's a that's a process when it works.

: 1087

00:54:09,280 --> 00:54:09,880

Yeah.

: 1088

00:54:09,880 --> 00:54:13,840

You, you do iterations

and we all became experts.

: 1089

00:54:13,880 --> 00:54:14,560

Yeah.

: 1090

00:54:14,560 --> 00:54:18,680

I started becoming an expert in the work

a bit.

: 1091

00:54:18,720 --> 00:54:21,680

Not as much as them, but quite a bit.

: 1092

00:54:21,680 --> 00:54:25,720

I start understanding what they do

and they start understanding

: 1093

00:54:25,720 --> 00:54:29,240

what they want on this screen to be able

: 1094

00:54:29,240 --> 00:54:31,920

to work less, work faster.

: 1095

00:54:31,920 --> 00:54:34,880

Yes. Yeah. Do more. Yeah, yeah.

: 1096

00:54:34,880 --> 00:54:37,920

In the in the previous slide,

that automotive business I mentioned,

: 1097

00:54:37,960 --> 00:54:40,640

you know, there was this presentation

that we've seen.

: 1098

00:54:40,640 --> 00:54:44,880

So SAP

was switching to S for Hana Environment.

: 1099

00:54:44,880 --> 00:54:48,680

So it was the cloud version of

of the base of the old version.

: 1100

00:54:48,680 --> 00:54:52,200

But it was able to cut down

a number of clicks

: 1101

00:54:52,200 --> 00:54:55,200

that you do

to get to finalize certain actions.

: 1102

00:54:55,200 --> 00:54:57,480

There was less

clicking on the screen happening.

: 1103

00:54:57,480 --> 00:55:01,240

So there was this big presentation

that they would put on on YouTube even,

: 1104

00:55:01,240 --> 00:55:04,600

and all these like platforms,

and it was literally just calculating

: 1105

00:55:04,600 --> 00:55:07,920

the time spent clicking on the screen,

filling out the data,

: 1106

00:55:07,920 --> 00:55:11,200

whether it was the purchasing thing

or something around the process.

: 1107

00:55:11,240 --> 00:55:11,600

Right.

: 1108

00:55:11,600 --> 00:55:15,560

And you know, the let's say

15 seconds saved.

: 1109

00:55:15,600 --> 00:55:19,600

But if you multiply it

by 10,000 people using it

: 1110

00:55:19,600 --> 00:55:23,200

globally or 100,000 people,

then obviously it's a massive saver.

: 1111

00:55:23,680 --> 00:55:26,280

And that's how businesses

don't usually look at it.

: 1112

00:55:26,280 --> 00:55:28,560

Can you sort of advise anything

on, you know,

: 1113

00:55:28,560 --> 00:55:32,400

so if somebody wants to calculate

the return on investment, right.

: 1114

00:55:32,440 --> 00:55:35,440

When it comes to applications

making some changes, you know,

: 1115

00:55:35,440 --> 00:55:37,760

is there any advice

that you can give us on that front?

: 1116

00:55:37,760 --> 00:55:40,080

That was also one of the questions

from the community.

: 1117

00:55:40,080 --> 00:55:44,520

You got to consider

how much it costs you to do it manually.

: 1118

00:55:44,560 --> 00:55:46,560

How much does this go?

: 1119

00:55:46,560 --> 00:55:49,560

Does it cost to you to do it with an app?

: 1120

00:55:49,560 --> 00:55:50,040

Yeah.

: 1121

00:55:50,040 --> 00:55:51,360

Usually it will be much.

: 1122

00:55:51,360 --> 00:55:53,520

Cheaper in terms of the time spent. Right.

: 1123

00:55:53,520 --> 00:55:56,480

The time spent on actually

like clicking and.

: 1124

00:55:56,480 --> 00:55:58,480

Yeah. Yeah okay. Yeah.

: 1125

00:55:58,480 --> 00:56:01,640

And today we got a new thing to consider

: 1126

00:56:01,640 --> 00:56:04,800

which is can we ought to make things okay.

: 1127

00:56:04,960 --> 00:56:08,960

We start

we are starting having a good enough

: 1128

00:56:09,000 --> 00:56:11,560

to automate a lot of things, you know.

: 1129

00:56:11,560 --> 00:56:15,560

And that will also save

you time and money which is.

: 1130

00:56:16,520 --> 00:56:18,600

Tokens are cheaper than labor basically.

: 1131

00:56:18,600 --> 00:56:20,040

Yeah, yeah, yeah.

: 1132

00:56:20,040 --> 00:56:23,000

Computing power

is becoming cheaper than labor.

: 1133

00:56:23,000 --> 00:56:27,280

And maybe you want, you know, fire people,

but you will have

: 1134

00:56:27,280 --> 00:56:30,960

your people free to do valuable things.

: 1135

00:56:32,080 --> 00:56:34,760

Yes. Speak with clients to think

: 1136

00:56:34,760 --> 00:56:39,480

how to solve things better to

yeah yeah yeah yeah yeah yeah.

: 1137

00:56:39,840 --> 00:56:40,880

Okay.

: 1138

00:56:40,880 --> 00:56:43,680

All right then

I think we've, we've covered everything

: 1139

00:56:43,680 --> 00:56:47,560

sort of relevant to, to what we,

we planned to do.

: 1140

00:56:47,560 --> 00:56:51,640

So you know, the couple

of selfish questions that I've put there

: 1141

00:56:51,680 --> 00:56:54,920

that was also for, for my benefit

to hear it from you as an expert.

: 1142

00:56:54,920 --> 00:56:57,160

So thank you for

for covering that as well.

: 1143

00:56:57,160 --> 00:57:00,200

And it was really great

to have you on the show.

: 1144

00:57:00,200 --> 00:57:03,480

And I really appreciate the time that

you've sacrificed, you know, to be here.

: 1145

00:57:03,480 --> 00:57:05,760

So I really appreciate that.

: 1146

00:57:05,760 --> 00:57:07,240

For you in some ways.

: 1147

00:57:07,240 --> 00:57:12,400

Nice to to to be able to speak to people

and explain things is fantastic.

: 1148

00:57:12,920 --> 00:57:14,280

Very good, very good.

: 1149

00:57:14,280 --> 00:57:16,960

Just as a last thing,

you know, if you can share, you know

: 1150

00:57:16,960 --> 00:57:21,120

who is your ideal client, you know,

are you operating within Europe, UK

: 1151

00:57:21,640 --> 00:57:22,560

just just for the guys.

: 1152

00:57:22,560 --> 00:57:25,840

Also, if anybody's interested

in collaborating with with you,

: 1153

00:57:25,880 --> 00:57:26,600

with your business.

: 1154

00:57:26,600 --> 00:57:33,360

You surely work with startup and medium

sized companies for UK and Europe.

: 1155

00:57:33,400 --> 00:57:35,560

That's how you use your target.

: 1156

00:57:35,560 --> 00:57:37,560

So how do we understand them?

: 1157

00:57:37,560 --> 00:57:39,480

Yeah, we can do things.

: 1158

00:57:39,480 --> 00:57:42,680

Yeah yeah okay. Fantastic. All right.

: 1159

00:57:42,720 --> 00:57:46,040

And obviously, you know, we'll

put the link down to the description

: 1160

00:57:46,040 --> 00:57:51,040

so people can find your business

once we have the episode published.

: 1161

00:57:51,040 --> 00:57:53,360

But yeah for now that's that's everything.

: 1162

00:57:53,360 --> 00:57:55,200

Thank you very, very much.

: 1163

00:57:55,200 --> 00:57:57,480

I really appreciate

the time you spent with us.

: 1164

00:57:59,040 --> 00:57:59,840

Thank you.