Episode 14

Jonathan Green's Guide to AI Tools for Entrepreneurs

Welcome to another steaming-hot episode of "Digital Coffee: Marketing Brew." I'm your host, Brett Deister, and today we're going to sift through the grounds of artificial intelligence in the marketing world with our special guest, AI-whisperer and best-selling author Jonathan Green. We'll stir in Jonathan's expert insights on why learning AI is like getting the hang of Facebook's myriad features and why it's essential to scald away the fear that AI will replace jobs.

Jonathan brews up some advice on rolling out AI tools such as Bard, Claude, Perplexity, and ChatGPT into your marketing recipe, and he'll sweeten the deal with a free master prompt for ChatGPT newbies. Despite his preference for fruit tea over a cup of joe, Jonathan is here to energize your entrepreneurial spirit with practical AI applications that promise to be a time-saving espresso shot for your business.

Dive into the discussion as we explore using AI for tasks like research, content generation, and social media management. Jonathan will spill the beans on the importance of human oversight when using AI, the evolving toolkit of AI applications, and the price point sweet spot for OpenAI's services.

Grab your favorite mug and settle in as we debate whether AI can streamline PR roles or if it's just another decaf promise. Plus, Jonathan will share cautionary tales reminding us that like any powerful brew, AI must be handled with care.

All this, while I share my preferred podcast editing tools and Jonathan showcases how AI can imitate top-tier marketing minds. Let’s demystify AI—together, we’ll discover why embracing this revolutionary tool might just be the most important skill to filter into your career. So, charge your earbuds and prepare for an episode that's guaranteed to perk up your marketing strategies right here on "Digital Coffee: Marketing Brew."

3 Fun Facts

1. Jonathan Green prefers fruit tea over coffee because coffee makes him sick.

2. Brett Deister favors using SquadCast over Riverside for podcasting due to better scheduling options.

3. Jonathan Green can switch ChatGPT into Q&A mode for beginners with a special master prompt, which is available for free.

Timestamps:

00:00 Use ChatGPT for practical business improvement options.

05:49 Assistance in generating and creating content ideas.

09:16 AI streamlines video editing, saves significant time.

10:32 Constantly improving, new tools, great time to buy.

16:06 Tools help evaluate writer's style and frequency.

17:34 AI reads books for efficient research analysis.

22:51 Identifying top marketers and their unique styles.

23:30 Analyze and capture language of big brands.

27:02 Addressing hesitations around AI in 7 words: Assuaging concerns about AI replacing human work.

31:20 Some believe AI skills are crucial for employment.

34:14 ChatTBT: Low learning curve, 8 hours suffices.

38:06 Twitter data may help improve results.

💬 Want to get involved? Leave us a comment, give us a 'like,' and follow us for more insights. Join our Locals for lively discussions, and if you've got questions, email us at bdeister@digitalcafe.media!

👕 Check out our mech: www.digitalcafe.store

🌟 Review the Podcast if you loved this episode and share it with fellow marketers who could benefit from a treasure trove of podcasting wisdom. Tune in to "Digital Coffee: Marketing Brew" and let's brew up some success together!

Transcript
Brett Deister:

And welcome to a new episode of Digital Coffee

2

:

Marketing Brew, and I'm your host.

3

:

Brett er, but this week we're gonna

be talking about ai, the thing that

4

:

all marketers are worried about,

trying to figure it out, trying

5

:

to do their best to use it in the

best possible way for this stuff.

6

:

But I have Jonathan

Green with me, and he is.

7

:

Expert using artificial intelligence

for your online business.

8

:

And he's a bestselling author of a

chat Chat, GPT Profits, and he has

9

:

a mailing list of over a hundred

thousand subscribers and he hosts

10

:

a podcast with 250 plus episodes.

11

:

So welcome to show, John.

12

:

Jonathan Green: Thank you for having me.

13

:

I'm excited to be here.

14

:

Love talking about my

favorite subject, ai.

15

:

Brett Deister: Yeah, and the

first question I saw my guest is,

16

:

are you a coffee or tea drinker?

17

:

Tea.

18

:

I'm drinking.

19

:

Any specific,

20

:

Jonathan Green: I drink fruit tea.

21

:

I find tea unbearable, like regular tea.

22

:

So I drink like four fruit

tea, fruit tea or mixed fruit.

23

:

It's the only way I can do it,

but yeah, I can't drink coffee.

24

:

It makes me sick, so I have no choice.

25

:

Brett Deister: I, yeah, that's fair.

26

:

I mean, if it makes you sick, yeah.

27

:

I wouldn't want you, I wouldn't

want anybody to drink coffee.

28

:

It makes 'em feel sick.

29

:

But I gave a brief explanation of your

expertise, but can you summarize it

30

:

a little bit more for our listeners?

31

:

Jonathan Green: Sure.

32

:

What I do is help people to understand

how you can use chat GBT to improve

33

:

your processes, to speed up your

business, to be a little bit better at

34

:

your job, to save time or money, and

really focus on the implementation.

35

:

It's very easy to get distracted

by how technical it can get.

36

:

There's every day there's a new

white paper about how AI works.

37

:

It's not really useful.

38

:

It's also easy distracted by all

this new flashy features, which are

39

:

interesting things, but not useful to you.

40

:

So I try to kind of walk that narrow

band of is this practical use for

41

:

entrepreneurs, small business owners

and people that are trying to grow their

42

:

business or kind of grow their career.

43

:

And in that case, there's a lot

that AI can do for you that can

44

:

really help you to be faster at

your job, to save more time to.

45

:

Cut down on your costs and really increase

your efficiency because I can tell you

46

:

from my business, it's a lot easier to

save a dollar than it's to make a dollar.

47

:

That's actually

48

:

Brett Deister: pretty true.

49

:

I mean, making a dollar is, especially

now in this inflation rate, it's making

50

:

people go, should I actually spend that

money for the AI or whatever else that

51

:

you actually need them to actually buy?

52

:

So I agree with that, but.

53

:

Since we're talking about AI,

is your favorite tool, jet P.

54

:

Jet GPT, or do you like Bard

or whatever else is out there

55

:

that I don't even know about?

56

:

Jonathan Green: I just got really good at

chat GBT first, so I became a power user

57

:

and once you get really good at one, it's

hard to justify jumping into another one.

58

:

I also have Claude, also perplexity.

59

:

I was gonna switch to Claude

because it has a longer memory.

60

:

And then chat.

61

:

GBT released an update last week.

62

:

Then I'm a big fan of

perplexity for research.

63

:

Perplexity is up to date,

so it does research.

64

:

That's all the way through today.

65

:

'cause current internet research,

it's really good for that.

66

:

I think chats that.

67

:

And Chad, GBD did just

change their time window.

68

:

It used to have a cutoff of

September,:

69

:

moved it up to spring of 2023.

70

:

So they added in 18 months

of more data, which is great.

71

:

It's still out of date, but one

of the stories I used to tell, and

72

:

actually I told, I was talking to

an employee at OpenAI last week.

73

:

I was like, you know, if you asked

Chad GPT for travel advice, it might

74

:

give you some places that were great

in:

75

:

So tears out of date, kind

of a big deal for research.

76

:

So.

77

:

I use a lot of tools, but my

core tool really is chat, GPT.

78

:

It's the one that I'm the best at.

79

:

I use some other specialized tools.

80

:

I use another tool for video editing

and some other things, but they

81

:

constantly pull those features into chat.

82

:

GPT, like.

83

:

I was about to start using it

for show notes, but it just

84

:

does great podcast show notes.

85

:

It does all these other things now it can

listen and watch video and see images.

86

:

So they're adding more and more features.

87

:

So it really is the powerhouse.

88

:

That's the main one that I use.

89

:

And that's just part of it's

'cause once you go to the system,

90

:

you kind of wanna stick with it.

91

:

Got

92

:

Brett Deister: you.

93

:

And so how can PR, pro PR and

Marketing Pro start to use like.

94

:

AI with their strategy and everything

else, how can they start to use that?

95

:

Because it's all great.

96

:

We're like, everybody should

use it, but where do they start?

97

:

Jonathan Green: Yeah,

that's a great place.

98

:

There's two areas I kind

of take this approach with.

99

:

The first area is research and the second

area is kind of repetitive processes.

100

:

So anything you do that you do

over and over again every week, you

101

:

spend a lot of time doing, whether

it's responding emails, maybe if

102

:

you're in pr, you send out a lot

of emails about your new clients.

103

:

I get a lot of emails for people

that wanna be guests on my podcast.

104

:

I'm sure you get the same thing.

105

:

I get so many cold emails

that are obviously they've.

106

:

Use an automation process

that has mistakes in it.

107

:

'cause they'll point out the wrong thing

or have the wrong link for my website.

108

:

Like someone sent email, oh, I saw

your podcast at, and then it was a

109

:

link to my blog, not my podcast page.

110

:

And I was like, you didn't even try.

111

:

Like you didn't, you know, and

these processes that we were

112

:

automating in a poor way before,

you can do much better Now.

113

:

You can speed that up, you

can speed up your responses.

114

:

That's one of the first places you

can do is how to help you with that.

115

:

If you are dealing with a lot of

inbound where people message you

116

:

and they want you to represent them,

then you can also use it to sort that

117

:

data, start to develop a criteria.

118

:

Those are some of the

first places to use it.

119

:

The next place for most people is

content generation, whether that's

120

:

writing emails, writing blog posts,

creating social media content.

121

:

And it can either help you with

coming up with ideas if that's one of

122

:

your weaknesses, or it can help you

with actually creating the content.

123

:

So for example, I have Chachi BT write

tons of tweets for me, and then I'll just

124

:

pick, I'll say, gimme a list of 10 tweets

on this topic, and I'll pick two or three

125

:

that I like and add those to my queue.

126

:

So those are some of

the ways that I use it.

127

:

But it really is about looking at

what are things that I do that are

128

:

repetitive or what are things that

I'm not very good at that I have to

129

:

do to support the things I am good at?

130

:

Yeah,

131

:

Brett Deister: and so should

people actually like check it?

132

:

Because I know it's easy

to be like, oh, I did it.

133

:

All I have to do is send it out.

134

:

But like you said before, even with

the automation there, there can be

135

:

issues where it's like, that wasn't

even correct at all, or it's out of

136

:

Jonathan Green: date.

137

:

I would always check it because

you can send something out.

138

:

Like if you send out an email

that is offensive, you can't

139

:

just go, oh, chat GBT wrote it.

140

:

They're still gonna go, yeah, but

you sent it like it doesn't give

141

:

you kind of this magical protection.

142

:

So one of the things to understand is that

AI is not able to do anything on its own.

143

:

If you leave chat GBT to its own devices.

144

:

Like if you're not paying attention

to the responses, it goes insane

145

:

within about 13 to 15 responses

within about five to 10 minutes.

146

:

Responses are fully insane.

147

:

So you cannot take your eyes off it.

148

:

That's not how it works.

149

:

It really is like a car.

150

:

You wouldn't let a car drive itself.

151

:

I mean, I know they talk about

that, but every time you see

152

:

that, it doesn't work out.

153

:

It takes the person plus the

machine to get the best results.

154

:

So you want, your job really shifts

from doing the task to oversight.

155

:

So I always, my, one of my

biggest methods, I always

156

:

ask for three or five or 10.

157

:

Say, write me three emails and

I'll pick the one that's the best.

158

:

Write me three short descriptions.

159

:

I'm always choosing and

this keeps me in that mode.

160

:

Instead of going, I like this or I

don't like it, I need to edit it.

161

:

I would rather just have it write me three

emails and I pick much easier process.

162

:

Yeah,

163

:

Brett Deister: and so there,

there are many different tools,

164

:

like what's the best use case?

165

:

For example, I used cast Magic

for show notes, and it also

166

:

writes additional things.

167

:

I think it uses either open AI or jet.

168

:

GPT, which I think is the same thing, but

the uses that, so is there like specific

169

:

tools that will use Chat GPT that they can

actually use without actually going to it?

170

:

Like how's that work for marketers?

171

:

Because like I said, there's a

ton of tools that we'll use open

172

:

AI and everybody's like, I don't

know which ones which, what?

173

:

What should I use?

174

:

Jonathan Green: Most tools, probably 90

to 95% of them are using the Open ai API.

175

:

A few use the Anthropic

Clauds API, but not very many.

176

:

Every time you see an AI writer, a tool

that does blog posts writing or any

177

:

type of writing there, it's chat, GBT.

178

:

It's just a Reskin 90.

179

:

I'm a hundred percent.

180

:

I've never encountered one that isn't.

181

:

So it's possible the one that exists.

182

:

They'll say they're an ai,

they often charge more.

183

:

Open AI at max is $20 a month.

184

:

So any tool that does writing

that's more than $20 a month, you

185

:

know, they're overcharging you.

186

:

So that's one area.

187

:

There are other types of tools that

can be really used, but there's a

188

:

bunch of tools that are part of my

tech stack that do use chat TPT in

189

:

different ways and they're just faster.

190

:

So I have a tool that I

use for running my Twitter.

191

:

I have a tool I use for

running my LinkedIn.

192

:

There's.

193

:

Purpose built tools.

194

:

I use a software.

195

:

I'm sure you're familiar with these.

196

:

That takes your, my podcast or

my video, long form videos and

197

:

cuts up into social media clips.

198

:

I, of course have to watch those

clips to make sure that they are good.

199

:

That's another type of tool that

I use that speeds up my process.

200

:

I use another AI that after I record

a podcast like this, it edits the

201

:

whole thing for me and switches back

and forth to whoever's talking and

202

:

creates a transcript of the show

and does that heavy lifting for me.

203

:

So I no longer have a

video editor for my show.

204

:

Saves a huge amount of time, takes like

10 minutes for it to render, whereas it

205

:

used to take my editor three to five days.

206

:

Just waiting for it to, for the

turnaround time is so much faster.

207

:

So there are some other tools that are

useful and can be helpful for someone.

208

:

But really most tools are relying

on the brain or the AI engine, which

209

:

usually comes from one of the big

companies and they're all like.

210

:

Claude is partially owned by Google.

211

:

Chad, GBT is 49% owned by Microsoft.

212

:

So each company has one.

213

:

You know, Twitter just came out with one.

214

:

Facebook owns Llama, so they're, each

big company kind of has one, and that's

215

:

their next iteration of social media.

216

:

But the real magic is looking at

your process and going, this is where

217

:

there's an opportunity for me, or this

is an area where I have a problem.

218

:

Then you look for that specific tool.

219

:

So I was, I've gone through

so many video editors.

220

:

Constantly improve my process.

221

:

Things that were, I was using a

purpose-built tool or an Adobe Premiere

222

:

add-on to do the episode editing,

flipping back and forth between

223

:

the speakers, and then my recording

software just added as a feature.

224

:

I was like, oh my gosh, now I don't need

this other software that's $30 a month.

225

:

So things are happening so fast that

there's a new tool or there's a specific

226

:

tool coming out every single week.

227

:

So.

228

:

It's really a great time to be a

buyer, especially because OpenAI with

229

:

their pricing of $20 a month and no

limitations on how many questions you

230

:

can ask or how many responses you get,

has put this massive downward pressure

231

:

on what companies are charging for ai.

232

:

So tools that are coming out this

year are cheaper than tools that

233

:

came out last year, and I think it's

a really, it's like a great time to

234

:

be a consumer in that way, because.

235

:

All these softwares everyone was

looking at, like pricing some

236

:

software like was $150 last year

and now it's $39 a month this year.

237

:

It's like that's a huge difference.

238

:

So it doesn't have to be expensive.

239

:

That's one of the important things is

you can be very lean as far as expense,

240

:

and get really, really good results.

241

:

Brett Deister: Gotcha.

242

:

Yeah, you're right.

243

:

You use, opus clip for the

scheduling out and doing that.

244

:

It does a pretty good job.

245

:

I've used another one, but I think

I like this one a little bit better.

246

:

And yeah, the one for Adobe Premier,

unfortunately it's not for DaVinci

247

:

Resolve, but I know the one you're

talking about that will cut it up for you.

248

:

So are you saying that Premier just added.

249

:

Like a feature for that as well.

250

:

In 2024,

251

:

Jonathan Green: Riverside did, so I use

Riverside, which is very comparable to

252

:

squad cast, but squad cast is better

because it has less lag I've noticed.

253

:

'cause I've used both, I

used for different projects.

254

:

This one it's easier,

I have a conversation.

255

:

Whereas Riverside, there's a slight lag,

but Riverside added that feature and I was

256

:

like, oh my gosh, this is a crazy add-on.

257

:

So.

258

:

Every tool is adding more

features like Descrip.

259

:

Originally I was using that as my

transcription tool, but now it's a

260

:

video editor and also they've added

features like it will make the social

261

:

media clips and it does the thing where

it'll make your eye look at the camera.

262

:

If you look away from the camera,

which is like such a crazy feature,

263

:

it doesn't work perfectly yet.

264

:

When it does it, for my eyes, it looks

weird, but they're all, every tool is

265

:

adding more and more features to kind of.

266

:

Keep you on board.

267

:

So there kind of is a race to add

the most features, the best quality,

268

:

and it's a really good thing.

269

:

But yeah that specific plugin,

I forget what it's called.

270

:

We, we were using it and then Riverside

just added the feature and I was

271

:

like, whoa, this is a huge add on.

272

:

'cause it saves you so much time now.

273

:

It used to be so hard to get

transcription done a couple of years ago.

274

:

I started dealing with transcription

around 20 16, 20 17 'cause I have

275

:

vision problems and I used to

have two full-time employees that

276

:

just did transcription for me.

277

:

And I've tried every tool from drag and

dictate, which used to give you one outta

278

:

20 words was wrong and no punctuation.

279

:

So that's a nightmare.

280

:

Just a giant wall of text.

281

:

You have to re-edit all the way through

human employees, and now everything.

282

:

I ended up with three or four transcripts

of everything I do because then

283

:

Riverside transcriptionist, then that

YouTube creates own transcript, and

284

:

then my descriptive created transcript.

285

:

So it's like I've got transcripts

coming out, my ears and everyone,

286

:

like everyone's making 'em.

287

:

So.

288

:

What used to be hard has become easy.

289

:

And then you just look at, well,

what's just the most accurate?

290

:

Just like you're using Opus,

there's probably 20 competitors.

291

:

I use video, but there's not a

really major difference between them.

292

:

Right?

293

:

It's just personal taste.

294

:

That's the, and that's

like where we're at now.

295

:

And that is like Riverside's

adding it to script is adding,

296

:

everyone's adding those tools.

297

:

So they're kind of having to

compete and create new features.

298

:

I know Opus just released

a bunch of B-roll features,

299

:

which are very interesting.

300

:

I was like, oh, that's cool.

301

:

I wanna check that out.

302

:

And I look at, now every tool

is a social media schedule.

303

:

Do you remember?

304

:

There was no such thing as social

media schedule 10 years ago, and

305

:

now it's like every tool has it and

it's almost overwhelming that you're

306

:

scheduling for every single tool.

307

:

So tools that I used last year I

don't use anymore because there's

308

:

all these really cool things coming

out now that make it so much easier.

309

:

Brett Deister: Yeah, I've still

have a Riverside one, but I, I

310

:

think I've done squad cast because

their scheduling is better.

311

:

'cause Riverside doesn't really

have a scheduling, which annoys me a

312

:

little bit that that's the one part

where I'm like, just get a scheduler.

313

:

So yeah, I can use you

more effectively, but no.

314

:

Yeah, I, I've used video video as well.

315

:

It's actually pretty good.

316

:

But moving back to more like

pr, will this save them time?

317

:

Because I mean, PR people are always

looking and the main thing for PR

318

:

is finding the journalist and making

sure the journalist still works

319

:

there and making sure that they

know what they've written so they

320

:

actually pitch them the right story.

321

:

Will this save them time?

322

:

Because I mean, we're all human.

323

:

We'll have a finite number of hours in

a day, and if you're sending a bunch of

324

:

emails to a bunch of different places.

325

:

You're not really gonna remember all that.

326

:

That's a little data overload or breathing

327

:

Jonathan Green: overload.

328

:

That's a great question

because it's very specific.

329

:

I love it.

330

:

So one of the best areas

is for targeted research.

331

:

So what you can do is feed an AI

the person's name, and especially

332

:

if you have a link to them on their

website, say, Hey, I want you to

333

:

find everything by this person.

334

:

What are the type of

articles they're writing?

335

:

How frequently are they writing?

336

:

Are they writing for any other places?

337

:

So there's something perplexity is better

at than a chat GBT Perplexity AI is free.

338

:

They do have a $20 paid upgrade.

339

:

I've never needed to upgrade because the

free tool is so crazy good for research,

340

:

and you can use these other tools to find

the other articles they've written and

341

:

say, Hey, what type of style do they do?

342

:

How often does this person.

343

:

Right, like negative articles, right?

344

:

Because you know, some people,

they, that's their game.

345

:

They'll write a couple of positive

ones and a bunch of negative ones.

346

:

You don't wanna accidentally put

your client in that situation where

347

:

they might get a hit piece right?

348

:

On accident.

349

:

Like last thing you wanna

do is bring that to someone.

350

:

So it can do things like that and say,

oh, how long has this person been writing?

351

:

What type of stuff do they write?

352

:

And then you go, oh, you know what?

353

:

This person.

354

:

Only does actors, where you notice

oh eight outta 10 articles are for

355

:

actresses, or six outta 10 are singers.

356

:

So you can use it in exactly that

way to kind of collate the research

357

:

so that you don't have to read each

article and look for the similarities.

358

:

Which sometimes for us is like,

that's a boring process, right?

359

:

It's very mechanical.

360

:

But to say, oh, this person has

written an article six months.

361

:

That's good to know.

362

:

That immediately saves you

time, go, oh, they're probably

363

:

not writing there anymore.

364

:

They may not be writing.

365

:

Maybe they've gone off

to write their own book.

366

:

You never know.

367

:

So it can really do that.

368

:

Where it does that actual research,

one of the things you can have it do

369

:

is actually read the person's book.

370

:

So every, my experience is almost

every journalist wants to be an author.

371

:

Every author's trying to get on the news.

372

:

It's like, where.

373

:

So you can have it read their book to

get a sense of what they wrote about.

374

:

And then when you message them, you

can actually specifically say, oh,

375

:

it's interesting that your book talks

about this and this is why my client's

376

:

a good fit for what you talk about.

377

:

Now you've kind of hit them in a new

angle because normally yeah, you'll

378

:

normally, when you're doing a little

research, write, read like their most

379

:

three recent blog posts, but no one reads

someone's book as part of the research.

380

:

But the AI can do that.

381

:

There's a bunch of free

ais that will do that.

382

:

Chatt will do it.

383

:

Claude will do it.

384

:

Perplexity will do it 'cause they

have a large enough memory banks.

385

:

So you can do that really quickly.

386

:

It's one of the things that I do

is I'll have it read a book for

387

:

someone, screen my podcast and say,

oh, come up with questions based on

388

:

this book that I might wanna ask.

389

:

And then I'll choose which ones

from that that I find interesting.

390

:

But they're really, really good

at research, really good at

391

:

spreadsheet stuff, really good at

data analysis, kind of figuring out

392

:

what do these things have in common.

393

:

And from there you can jump off and

say, Hey, who are similar riders?

394

:

Who might be a good fit for

like me to pitch this person to?

395

:

Brett Deister: Yeah, I, I mean, the

other thing is that PR people sometimes

396

:

have the right crisis plans, and are

we gonna have to put a little section

397

:

in for AI if it goes off the rails?

398

:

'cause I mean, you can say that,

oh, AI did this, but I mean it.

399

:

Ultimately is gonna be the

business's fault for using the ai.

400

:

So are we gonna have to have a new

like avenue for AI for crisis and make

401

:

like we let it do whatever it wants?

402

:

'cause I'm pretty sure eventually

it's gonna get to that point.

403

:

We're gonna be like, oh.

404

:

Now we have to like crisis that part out.

405

:

Are we gonna have start

to see that eventually?

406

:

Jonathan Green: So there's a couple of

specific incidents that have happened.

407

:

The first is that every time someone

releases a Twitter ai, people see

408

:

how fast they can turn to racist.

409

:

It's happened like six or seven times.

410

:

Like, and people are always gonna do that

because to a number, to enough people.

411

:

That's funny.

412

:

Whether you think it's funny or

not to enough people, it is to see

413

:

if you can make a robot turn bad.

414

:

Just like how fast were

people trying to turn chat.

415

:

GBT badge.

416

:

So remember Microsoft released a bot

a couple of years ago and I think

417

:

they got it to say some horrible stuff

within, like, it wasn't a full day,

418

:

they had to take it down within a day.

419

:

'cause it was saying like

shocking stuff and it's like.

420

:

It shows you that an AI can be tricked.

421

:

That's really the lesson in that.

422

:

And then a couple of months ago, a lawyer

went to court with a bunch of cases that

423

:

Chad PD gave them, but they weren't real.

424

:

He goes, but I asked Chad to you real.

425

:

I was like, yeah, if someone lies to

you and then you go, are you lying?

426

:

They're not gonna go, yes, I'm lying.

427

:

Right?

428

:

So it goes, yeah, these are real.

429

:

And he brings the 'cause he didn't

understand the kind of the underpinning.

430

:

So he got in a lot of trouble.

431

:

He got censored.

432

:

He almost lost his ability to be a

lawyer, like got in a lot of trouble.

433

:

Like the judge was not pleased.

434

:

It was a really big deal.

435

:

You can't just go, oh,

the computer tricked me.

436

:

So you will have to maintain oversight.

437

:

I don't think you can blame the ai.

438

:

Open.

439

:

AI has said they'll cover you if you get

accused of plagiarism, anything in that

440

:

area if you get sued for plagiarism.

441

:

But yeah, for if it says something

like messed up or something that,

442

:

that's why you do have to read it

because things will slip through.

443

:

You can use it for small things like.

444

:

Keeping track of YouTube comments

say, oh, let me know if a YouTube

445

:

comment needs to be deleted.

446

:

'cause it'll read them

all and look for keywords.

447

:

That's something it can do, but as

far as like actually letting an AI

448

:

write an entire press release and

no human reads, I wouldn't do that.

449

:

I think that the idea that an AI

can do something without human

450

:

involvement, that's a big mistake.

451

:

And you can see that like as soon

as after this episode, when you run

452

:

the thing through Opus Clips, at

least two of them will be terrible.

453

:

Probably more than that, right?

454

:

Two of them will go, this is horrible.

455

:

Whenever I do one, it always grabs

the opening half of the opening music.

456

:

I'm like, that's not a

clip anyone wants to see.

457

:

It's half of the introduction

doesn't make any sense.

458

:

So.

459

:

AI still are, are very error

prone and it's why the human

460

:

part will not disappear.

461

:

So definitely a big mistake you can

make is the a hundred percent AI thing.

462

:

I mean, people still do horrible stuff

and then say they got hacked all the time.

463

:

I don't know if everyone's getting

hacked so frequently in the first

464

:

thing people do is post something

messed up to Twitter maybe.

465

:

But I think that we're still using

these kind of suspicious excuses.

466

:

We don't need to switch

to the ai Did it yet?

467

:

And you have to use it

like any other tool.

468

:

You still have to use it responsibly.

469

:

It can't replace your intelligence.

470

:

It can accelerate your intelligence, but

you have to still do error correction

471

:

because it will make mistakes, and

that's why we still need our jobs.

472

:

Brett Deister: And so can AI help create

great campaigns for marketers for PR pros?

473

:

Because I mean, the analytical side is,

it seems to be like the best spot for now.

474

:

I mean, eventually it'll probably

get better, but the analytical

475

:

side, like you said, could be great.

476

:

So could it help like drive the ideation

of like a great digital marketing campaign

477

:

or a great awareness campaign for pr?

478

:

Jonathan Green: I use it

a ton for copywriting.

479

:

I use it a lot for marketing campaigns.

480

:

It's really good at ideas and

there's a couple of ways to use

481

:

it that are very, very creative

and I'll share a few with you.

482

:

So if you say write in the

style of, and then you provide

483

:

a variable, you'll get a comp.

484

:

It will change everything.

485

:

So you can say, write in the style of

love, and then you can just use it the

486

:

name of an actual marketer and it'll say,

Hey, what would so and so famous marketer?

487

:

Do for this campaign and it will

come up with something and you

488

:

can change the name of the market.

489

:

If you don't know famous marketers,

usually you say, who are the

490

:

top 10 marketers of all time?

491

:

What it will tell you is the 10

marketers has the most data on.

492

:

So then you just go through all

10 until you find one you like.

493

:

I do this with copywriters

also, those brands.

494

:

So I'll write, I say write this sales

letter in the style of Harley Davidson.

495

:

And immediately you're gonna

see the word open road a lot,

496

:

sunset, a lot freedom a lot.

497

:

And then sometimes you have to

do a negative and say, but don't

498

:

use the name of any specific

motorcycles or motorcycle parts.

499

:

'cause it'll start talking about like a

dovetail this and a feather, like specific

500

:

motorcycles, which is not what you want.

501

:

'cause you're talking about motorcycles.

502

:

But now it has an Americana

way of talking and it doesn't

503

:

sound like a specific person.

504

:

If you say, talk like apple.

505

:

Apple's put out a lot of ads in the

last 40 years since they started

506

:

in the early eighties, right?

507

:

So you have a language that's

completely different and you can

508

:

go through really big brands and

capture their language in that way.

509

:

So you could pull in a lot of

different ideas, whether they're

510

:

talking about specific, using specific

marketers, specific copywriters.

511

:

Or specific brands.

512

:

Those are some of the ways that I

quickly create a really different answer.

513

:

And then again, that's

a faster way to do it.

514

:

I'll say, I'll go through 10 and

then I'll just pick my favorite, and

515

:

that does 90% of the work for me.

516

:

So the secret is knowing that chat,

GBT has too much data, not too little.

517

:

So what I'm actually trying to do is

eliminate everyone else except Harley,

518

:

Davids, or everyone else except Apple,

or everyone else except McDonald's.

519

:

Use just this one data set.

520

:

Same thing when I narrow it down

to one author or one copywriter.

521

:

I'm saying, don't use other

copywriters because here's the thing,

522

:

if you say design a commercial,

most commercials are bad, right?

523

:

Most commercials fail because the majority

of commercials are local commercials.

524

:

Just like most websites are bad.

525

:

99% of websites, no one's ever seen them.

526

:

They were made in like the 1990s.

527

:

Those Geo Cities websites, those angel

websites, forget what they're called, or

528

:

the a OL websites, they're all bad, right?

529

:

They used to play music.

530

:

As soon as you visited the website,

it was flashing colors make dizzy.

531

:

But if you narrow out and

say, don't include those.

532

:

Just include this successful

brand or this successful website.

533

:

As a successful, now it's using a

dataset that all has good data and

534

:

you're getting a much better response.

535

:

Yeah, you can merge voices and

create your own custom voice.

536

:

So you can say, oh,

here's the things I like.

537

:

Here's the things I don't like.

538

:

Especially now with the release of

GPTs, which is where you can create

539

:

your own custom character or cus they

call it a GPT, which is kind of annoying

540

:

'cause it's GPT inside of A GPT.

541

:

Like don't give it a name.

542

:

But basically this is what I do a

lot of is I create really specific

543

:

characters that are experts at certain

things, have a certain way of talking.

544

:

So I call them cyber staffers.

545

:

The name doesn't really matter.

546

:

It's just what you call it

is, I have one that's like a

547

:

younger girl who does Twitter.

548

:

I mean, does linked TikTok.

549

:

She's annoying because

she talks very fast.

550

:

She talks in 32nd bits.

551

:

She talks in seven second bits,

but she's really good at TikTok.

552

:

The lady who does my LinkedIn, she's mid

thirties, she's a little more of business.

553

:

She has a different way of talking, and

in between I have my Twitter expert who's

554

:

not that annoying, but she's always happy.

555

:

Sometimes you don't want, but

that's how you wanna post on Twitter

556

:

to resist urge to post negative.

557

:

So there's separate personalities

that have separate expertises

558

:

that speak in a different voice.

559

:

So you can create a skill chain,

a personality, using the big

560

:

five personality type, which

is the O-C-E-A-N personality.

561

:

You can give it a specific personality and

you can also create a custom voice which

562

:

says, oh, you talk about these things,

but you don't talk about those things.

563

:

You kind of are, and you can feed in.

564

:

Those different brands that you like and

the brands that you don't like, and that

565

:

will create a really customized voice.

566

:

And when you do that, that's when you

get something really magical, something

567

:

really amazing that no one else has.

568

:

It's completely unique.

569

:

Brett Deister: Gotcha.

570

:

And.

571

:

What would you say for like

those that are hesitant about ai?

572

:

Because that, because this is still

pretty new for a lot of people.

573

:

I mean, a lot of people are used to

doing it all by hand or by search.

574

:

Google search was the first,

I guess ai, but very basic.

575

:

So what would you say for

those that are hesitant?

576

:

Like, I don't know about this.

577

:

I don't wanna make it feel like

that I'm not really working because

578

:

people can actually probably suit or.

579

:

Surmise that people may not be

working, you're just giving it to

580

:

ai, you're not really doing much.

581

:

So what would you say those people

are hesitant about using it right now?

582

:

Jonathan Green: So AI

is not a replacement.

583

:

It's an accelerator.

584

:

So I can write software.

585

:

I've written several pieces of

software using ai, but I will never

586

:

write software as good as a developer

who's using the same ai, right?

587

:

I can create images with an AI

image generator, but an artist

588

:

will create better images.

589

:

I'm a really good writer.

590

:

I write books.

591

:

I've written a lot of books.

592

:

I've written several hundred bestsellers.

593

:

So with AI I can write amazing books,

but it's 'cause it's accelerating

594

:

what I'm already really good at.

595

:

So whatever you're good at, it

makes you faster and more efficient.

596

:

What it, and it can take, if you're

a low skill person, it can bring

597

:

you up to like an okay, but it

can't take you from zero to 10.

598

:

Like it can't make me

a 10 at a 10 PR person.

599

:

If you're a 10 outta 10 PR person,

it can make you 40% faster.

600

:

I understand the hesitancy because

we think of it, oh, this is gonna

601

:

replace people, but not really.

602

:

What it's actually doing is making

everyone a little better at their jobs.

603

:

What's really happening in this

shift right now is the people who

604

:

adopt earlier are getting faster,

and we go through these phases.

605

:

The phases we go through is first it's

like, oh my gosh, look at this new tech.

606

:

Then we go, it's cheating.

607

:

Then we go, okay, it's

allowed, and then we go.

608

:

It's mandatory.

609

:

If we look at calculators, when

calculators came out and became affordable

610

:

in the early seventies, there were people

who said, how can a kid learn math if

611

:

they're allowed to use a calculator?

612

:

Then what happened?

613

:

They go, oh, key for you later go,

okay, you're allowed to use calculator

614

:

in class, but not during the test.

615

:

Then it was, okay, you can use

calculator during the test.

616

:

Now, when I went to high school in

the nineties, calculator required.

617

:

You had to have a graphing calculator and

they, they gave you a list of choices of

618

:

which one to have, and you had to have

it because you had to do the things.

619

:

I never learned how to draw like

a co-sign graph by hand because I

620

:

learned how to do it on the graphing

calculator, and that was good enough.

621

:

In fact, we would show on the

graphing calculator if it had the

622

:

right picture, you'd get the points.

623

:

Same thing happened with Google, right?

624

:

It was like, oh, you're not

allowed to use Google in school.

625

:

Then it was like, okay, you're,

you're supposed to, you're expected

626

:

to use it, expected to use your cell

phone, and the same thing with ai.

627

:

We're gonna go through,

it's not allowed at school.

628

:

It's cheating to, okay, you're allowed

to do it, but not during the test too.

629

:

Now it's mandatory because what we

wanna do is have people that are

630

:

good at using AI to do their jobs.

631

:

Nobody would hire an accountant

who says, I don't use calculators.

632

:

Right.

633

:

Would you trust a bookkeeper?

634

:

Goes, I don't do spreadsheets.

635

:

I do it by hand.

636

:

You'd be like, what?

637

:

'cause that's really your choice.

638

:

You can have a PR person who can do

10 things a day, or one who can do 50.

639

:

Of course you're gonna

hire the second person.

640

:

Now, there is the fear of a dro, a drop

in quality, which I completely understand.

641

:

That's why you have to oversight.

642

:

So you can say, oh,

it's written 50 emails.

643

:

I'm gonna read all 50 before they go out.

644

:

That saves you the time.

645

:

'cause you can read faster, you

can write, that's really where the

646

:

magic happens, where you don't give

up the thing you're really good at.

647

:

You don't wanna give that up.

648

:

But then you go, I'm

gonna do more of that.

649

:

I don't need to write every email.

650

:

I can go through and just tweak each

one and make it a little better.

651

:

I don't need to do all the research.

652

:

I could just double check

this to make sure it's right.

653

:

That's where you get really fast.

654

:

Like I had an AI edit my last book.

655

:

Why?

656

:

Because.

657

:

That's a huge amount of time saving.

658

:

I don't need to spell every

word right and worry about that.

659

:

'cause chay, BT always

spells every word, right?

660

:

Doesn't make grammatical mistakes.

661

:

So it solves that problem.

662

:

I still read through it.

663

:

'cause sometimes it may change

and I go, no, no, you're drifting.

664

:

I still read every single word to

make sure it didn't go off track.

665

:

So you're still involved.

666

:

That's really where the magic comes.

667

:

It's all about seeing it as the

working together on its own.

668

:

AI can't really do anything and on your

own, you're limited by your ability and by

669

:

time working together, you can accelerate.

670

:

Hmm.

671

:

Brett Deister: And where

do you see the impact of AI

672

:

coming in the next five years?

673

:

Jonathan Green: Sure.

674

:

I think there's gonna be the

people who think it's a fad.

675

:

Most of them are gonna be

unemployed in three to five years.

676

:

They're just gonna be replaced

by people who learn to use AI.

677

:

Because why would you hire someone

who can't use a calculator?

678

:

Right.

679

:

And that's one of the worries I have

for people that don't take it seriously.

680

:

'cause this is, right now

we're in the optional phase.

681

:

People go, oh, it's cool if you learn it.

682

:

There are already people who are getting

raises based on their knowledge of ai.

683

:

People are already talking about, oh, I

got a 20% raise 'cause I'm good at ai.

684

:

I got a 40% raise because I'm good at ai.

685

:

That's a lot of money.

686

:

20%, 40% single raise.

687

:

This is over the last six months.

688

:

That's a huge difference.

689

:

So it's already happening.

690

:

And then there's gonna be companies

that the same thing happens.

691

:

They go, oh, we're not, we're old school.

692

:

We're not gonna get involved in that.

693

:

And what's gonna happen is

that it's the same thing.

694

:

Like all the companies that 20

years ago or 30 years ago said,

695

:

oh, we don't need a website.

696

:

What happened to them?

697

:

They're gone.

698

:

They went outta business.

699

:

Right?

700

:

Then what happened to all the companies

said, oh, we don't do social media.

701

:

They all do it now, right?

702

:

Everyone's advertising on Facebook

and Twitter and whatever, like

703

:

they're all over every platform.

704

:

Why?

705

:

'cause it's not really optional.

706

:

So we are in the phase right now

where it's a big opportunity.

707

:

It's a wide open market

for people to learn.

708

:

You can be the first person in your

company, or the first person in your

709

:

market, your first person in your

area to get good at these tools.

710

:

It's a massive advantage over the

competition, but eventually they're

711

:

gonna shift from optional to.

712

:

You are gonna see this when

you see like job postings.

713

:

It's gonna start going

like, oh, AI skills a plus.

714

:

Then it's gonna switch

to AI skills mandatory.

715

:

That's what's gonna happen.

716

:

That's where you're gonna see it first,

and that's where it's gonna be really

717

:

scary for people who've lost their job

and now they're seeing every job in

718

:

their market has AI skills mandatory.

719

:

Now they gotta go back and try

and catch up on three years

720

:

of everyone else learning.

721

:

The cool thing right now is no one

has more than a year of AI skills.

722

:

Anyone who tells me they have three years

of experience of ai, I go, no, you don't.

723

:

Nothing about AI from

:

724

:

Nothing from last year matters.

725

:

Anything.

726

:

Pre-chat, GB 3.5 is irrelevant

because it's been replaced by

727

:

everything happening this year.

728

:

The cool thing is no one

has a massive advantage.

729

:

I only have like an eight, eight months

more experience than anyone else.

730

:

That's not a huge insurmountable

advantage, and that's why it's.

731

:

It's the people who think it's

a fad who are gonna kind of

732

:

suffer the most, unfortunately.

733

:

Brett Deister: Mm-Hmm.

734

:

And so let's say for those already

into their career and having to pivot,

735

:

how, how should they start to pivot?

736

:

Because I feel like the younger people

are gonna be a little bit easier to be

737

:

like, oh yeah, I'll just try this out.

738

:

But it's mostly the older ones,

the seasoned professionals are

739

:

like, Nope, I've done this way.

740

:

I will not change whatever

people say I can do it better.

741

:

So how can they pivot to

that with all of the, all the

742

:

knowledge that they have as well?

743

:

I,

744

:

Jonathan Green: the cool thing about chat

t BT is it's a really low learning curve.

745

:

The bad thing with all these ais

is it seems really hard 'cause

746

:

they have no onboarding process.

747

:

So I tell most people who go through my

programmer that I deal with, if you just

748

:

read my book or watch a bunch of YouTube

videos, just say, I'm gonna spend one day.

749

:

I'm gonna give this eight hours.

750

:

I.

751

:

And I'm gonna put up with this

and I'm gonna go I, and I don't

752

:

care if the AI thinks I'm stupid.

753

:

If you let go of that fear of the AI

judging you, which I used to feel that

754

:

way too, like I don't wanna ask a stupid

question or write a stupid prompt.

755

:

If you let go of that, you can learn it

in eight hours in one day, one business

756

:

day, you can become a skilled user.

757

:

That's what's really cool.

758

:

It seems really hard 'cause they

don't offer any onboarding and I know

759

:

that and I know why they do that.

760

:

But I also know that you can get

past all of that very quickly.

761

:

So it takes one day to

become a pretty good user.

762

:

Just playing around with it,

just trying prompts, just

763

:

watching some YouTube videos.

764

:

You can grab my book, whatever.

765

:

You don't have to, but you can, and

that's all you need to get started.

766

:

That's how I started.

767

:

I just watched a ton of YouTube videos.

768

:

I go, oh, I think I can do that.

769

:

And then I said, oh, you know what?

770

:

I think I can do that better.

771

:

And that's really how I kind of pushed

myself up is that I kind of have this, I.

772

:

Ideas, and then I test them.

773

:

I go, I could do this better.

774

:

I can do that better.

775

:

I can tweak this.

776

:

I can tweak that.

777

:

That's where all the magic happens.

778

:

So once you have that approach,

that's the real game changer.

779

:

It doesn't take a long time to learn it,

and you can just try and go, oh, it's good

780

:

at this and not, it's not good at that.

781

:

So I'm constantly discovering

new things that it can do.

782

:

'cause they're updating it all the time.

783

:

They don't tell you.

784

:

So when I first tested it, I said, how

many programming languages do you know?

785

:

And chat told me five

programming languages.

786

:

Then I asked it two months

later, it listed 14.

787

:

I was like, you haven't told anyone

that you learned nine new languages.

788

:

You only know if you ask.

789

:

And now I'm sure it's even

more probably dozens now.

790

:

So they're always adding new

features, always adding new skills.

791

:

And really it's just about

learning a new way to use a tool.

792

:

I know it seems daunting,

but compared to learning.

793

:

Some of the software that's out

there now that's so much harder,

794

:

like trying to explain to someone how

to use Facebook from the ground up.

795

:

It's unbelievable, right?

796

:

Facebook has so many features and groups

and pages like trying to explain the

797

:

differences between a group and a page.

798

:

To me, I still don't get it.

799

:

Like, right?

800

:

It's like what is happening here?

801

:

There's so many complex things.

802

:

There's a Facebook event.

803

:

This is a different type of

event, and all these integrations.

804

:

Here's where you upload a picture.

805

:

Here's where you upload

a business picture.

806

:

This is a reel.

807

:

This is a.

808

:

Like all of this stuff, it's really

complicated compared to that, the

809

:

idea that it's basically like a small

person that can help you with tasks.

810

:

If you just think of it as like an

assistant, then it's not so daunting.

811

:

It just, it seems scarier

than it is because most of the

812

:

content online is like, look at

all these crazy things you can.

813

:

Robot that goes out and does all the

work, like you don't need all that.

814

:

What you just need is something

that goes, oh, you know what?

815

:

This does something a

little faster for me.

816

:

Instead of having to read all the

articles, this, read the articles for

817

:

me and gimme the highlights, stuff like

that is where you can start with just

818

:

summarizing and doing those faster, and

you start there and you look at what

819

:

you do every week and go, Hey, what

do I do that's repetitive and that.

820

:

I can replace.

821

:

That's really how I approach it,

is replacing either things I'm not

822

:

very good at that are, that are kind

of next to my area of excellence

823

:

and I look at things that are

repetitive that I just can save time.

824

:

That's where I would start.

825

:

Brett Deister: Gotcha.

826

:

And should they test

out Bard and Chat GPT?

827

:

Just so they know like 'cause I

mean, I'm pretty sure some people

828

:

may like borrowed over Jet CBT.

829

:

Not saying anyone's better or

not, but should they try out all

830

:

of them and figure out which one

works for their better workflow?

831

:

Jonathan Green: Absolutely.

832

:

You should try Bard.

833

:

You should try Claude.

834

:

You should try perplexity

and you should try chat GBT.

835

:

And maybe I hate the name.

836

:

I don't know why they call it grok

from Twitter like that, or from X, like

837

:

they're changing the name of everything.

838

:

So that one might turn out to be good.

839

:

I don't know.

840

:

Because that one is gonna be fed all

the data of every Tweet anyone's ever

841

:

written, which I don't know if that's

a good thing or a bad thing, probably

842

:

a bad thing, but not the best data set.

843

:

But if you wanna be good at Twitter, I

guess that's a great place to use it.

844

:

But I would try them all and see if

there's one that's intuitive to you.

845

:

I just got lucky that chat

GT's very intuitive to me.

846

:

It's not gonna be that way for

most people, but I know people that

847

:

really love Claude and hate JGBT.

848

:

They each work a slight different way.

849

:

I really love perplexed.

850

:

I find it really easy to use.

851

:

I think it's actually more user

friendly than the other ones.

852

:

It's really simple and what it does,

it just does research, but it does it

853

:

very well and it gives you all of its

notes in a really professional way.

854

:

It says, oh, this came from here.

855

:

Here's the highlight, this

from here, here's highlight.

856

:

So I would try 'em all out, see

which one you feel good about, and

857

:

then spend one day learning it.

858

:

You just spend one day, you can master it.

859

:

It's really all it takes.

860

:

And then 'cause it's conversational,

once you get comfortable talking to it,

861

:

you can get better and better and better

and just become a master over time.

862

:

Brett Deister: Gotcha.

863

:

And so where can people find you online?

864

:

Jonathan Green: Sure you can

find out everything about me by

865

:

just Googling Serve No Master.

866

:

So that's my website, that's my

everything, my first book, and

867

:

that will give you everything.

868

:

And if you go to serve master.com/master,

I'll give you for free my master

869

:

prompt, which really switches chat

GBT into question answer mode, so

870

:

that it will do the heavy lifting,

kind of shows my formula for that to

871

:

make it really easy for beginners.

872

:

Brett Deister: All right,

any final thoughts for

873

:

Jonathan Green: listeners?

874

:

I am excited that you're here.

875

:

This is a great topic and showing that

AI is really gonna kind of affect every

876

:

different industry and that if you see

it as something, oh, this is a tool

877

:

that can help me do things a little bit

faster, save me a little bit of time,

878

:

that's really the right way to see it.

879

:

Not something that's gonna take

your job way 'cause it's not

880

:

gonna be replacing very many jobs.

881

:

All right.

882

:

Thanks John

883

:

Brett Deister: for joining

Digital Coffee Marketing Brew,

884

:

and sharing your knowledge on ai.

885

:

Jonathan Green: Thank you for having me.

886

:

And thank you for joining.

887

:

Brett Deister: As always, please

subscribe to this podcast of

888

:

all your favorite podcasting

889

:

Jonathan Green: apps and join us

next month and we'll talk about

890

:

Brett Deister: what's going on

in the PR industry and talking

891

:

to great values at the same time.

892

:

Alright guys, stay safe.

893

:

Get to understanding how AI

can better help your workflow

894

:

and see you next month later.

Next Episode All Episodes Previous Episode

About the Podcast

Show artwork for Digital Coffee: Marketing Brew
Digital Coffee: Marketing Brew
Get your does of marketing with your favorite coffee brew

Listen for free

About your host

Profile picture for Brett Deister

Brett Deister