HomeVideos

Let's generate $$$ with Grok Bot (LIVE: Watch me work)

Now Playing

Let's generate $$$ with Grok Bot (LIVE: Watch me work)

Transcript

2857 segments

0:02

All right, good morning, good afternoon,

0:04

and good evening. Welcome back to

0:05

another Clear Mud live stream. Today, I

0:08

want to focus I need to stop hitting

0:11

that button. I want to focus today's

0:12

episode around how people are making

0:15

money with Grockbot. Uh we are going to

0:18

try a few of these tactics ourselves.

0:21

And who knows, maybe we come up with

0:22

some brilliant idea to make some money

0:24

on the side while we're learning, while

0:27

we're experimenting. Um, so let me first

0:33

this Grockbot system that I set up

0:35

actually in a video just a few days ago.

0:38

I'm turning into our team bot. Oh, I

0:42

need to

0:46

I'm turning into our team bot. So

0:48

instead, I want to sign out to my other

0:50

account and I want to clean up that

0:52

other account and use that other account

0:54

for experiments for these live streams.

0:58

So, let's go ahead and log into that

1:00

other account. Give me one second. And

1:03

we're going to first start off with

1:05

start off with some cleanup here. And

1:07

then we'll uh still trying to log me

1:11

back in. It's kind I wish there was an

1:13

easy way to toggle between accounts.

1:16

That's a bummer. So, I got to log out of

1:19

that other cursor account and sign back

1:21

in.

1:26

Let's reopen the link.

1:33

Oh, I hit it preemptively. Now continue.

1:38

Sign in.

1:40

All right. Open Grogbot. So, this is our

1:43

original one. I need to finish migrating

1:46

the clear mud bots, but we can keep the

1:49

chief of staff. We can keep these chiefs

1:51

here. As you can see, I started

1:53

experimenting with a money pod and a

1:55

database factory as well as an

1:58

experiment team. So, we're going to keep

2:00

those, but we're going to delete

2:01

everything else because I've already

2:03

migrated the X team. I've already

2:05

migrated the housekeeping. Um,

2:09

keeper was actually something fun. I was

2:12

joined my first dynasty fantasy football

2:15

league and uh I thought let me ask some

2:18

questions.

2:19

And I don't know how I feel about it,

2:21

but uh let's let's do some cleanup.

2:26

All right, so let's do some cleanup. Go

2:28

ahead and delete the X team. That's

2:30

successfully migrated. Uh delete

2:33

housekeeping.

2:35

And we're going to keep the database

2:37

factory. We're going to keep the money

2:39

pod. We're going to keep the experiment

2:41

uh team or section rather. And

2:45

and then we're going to keep Codex and

2:47

Claude. But everything else,

2:49

oh, prep me a handoff markdown file for

2:53

the clear mud team. Uh, as we are going

2:55

to finish migrating that over. You can

2:58

actually delete partnerships that's

3:00

already migrated. Uh, but water, mud,

3:03

and sky, we still need to hand off that

3:08

agent data, that context, those

3:10

instructions, how those operate, uh,

3:12

into a markdown file. And then once

3:14

you've created that, go ahead and

3:16

cleanse this system, please.

3:20

All right, just starting off with

3:22

sending off a big prompt here.

3:26

I'm wondering if we should change our

3:27

chief of staff for this. Uh, ah, we'll

3:30

keep it Elon Musk.

3:32

Yo, what's up, Franc? How we doing

3:35

today?

3:40

There we go. I should probably though

3:44

move this video for Grockbot because it

3:46

covers feel like an important part of

3:49

the app.

3:51

So, let's grab that unlock. Move that

3:54

over to the right.

4:00

And perfect.

4:02

Keep that over there.

4:06

How's everyone doing today though

4:07

otherwise? And what's up with that

4:09

background music? Um, but yeah, what

4:12

inspired me, if I wasn't clear at the

4:14

start of this live stream is I was

4:16

watching this one here from 6 days ago

4:18

from Greg Eisenberg, and he's had this

4:19

guy on the on the show before. This guy

4:21

makes money off of creating database

4:23

websites. Database websites work like

4:25

this. You provide value. You create a

4:27

database and you focus on getting as

4:30

many visitors as possible. Then you

4:34

monetize that traffic, right? you

4:36

monetize that traffic in the form of

4:38

promoting certain tools,

4:40

whether brands pay for you or not.

4:42

Essentially, ad space. Um, there's

4:44

there's more to it than that, but that's

4:46

just a quick TLDDR of what a database

4:48

factory or a database website is. Um,

4:52

and this sent me down a rabbit hole of

4:54

what if we create our own team to do

4:55

that. So, today's episode is going to be

4:57

all focused on that. It's not going to

4:59

be focused on how I'm using it for clear

5:01

mud, how I'm using it for the team. Um,

5:03

we'll do another video about that. But

5:06

now, as you can see, I have two separate

5:08

systems. One for clear mud and team, and

5:10

then one for experiments and

5:11

demonstrations and just random personal

5:13

bot use. Um, that's how I'm kind of

5:16

separating things now.

5:20

Yo, what's up, Mark? Nice. Nice. Are you

5:23

just dipping your toes? Have you created

5:24

your first uh your first bot yet?

5:27

Listen, what one tip I have for

5:29

beginners, don't worry about a chief of

5:31

staff. When I was first playing with

5:33

Grockbot, I was just creating bots. I

5:34

was creating bots, having them do

5:36

things, and just learning. That's the

5:38

best way to start. Um, it's it can get

5:42

very

5:45

distracting if you try to follow

5:47

somebody else's advice too soon before

5:49

you find a way to utilize it best for

5:52

yourself, your life, your business,

5:54

whatever it is you want to uh build out

5:58

your Grockbot team or system for. uh

6:01

that's unique to you, right? I think

6:03

getting ideas from other people is very

6:05

important. Um it's amazing. Uh somebody

6:08

in our community was talking about this,

6:09

but since January, since my OpenClaw

6:11

days, right? And by the way, I still use

6:13

OpenClaw every day. I never liked having

6:17

individual channels to chat with all my

6:19

agents. I just wanted one executive

6:21

assistant who handles everything else.

6:24

But back in those early days, everyone

6:26

was creating these mechanisms to

6:28

communicate with each individual agent.

6:31

And it's just like it it made no sense

6:33

to me. Now with Grockbot, it it seems

6:35

like everyone's now routing through this

6:37

one agent does all. It's it's funny.

6:42

It's just funny.

6:44

I can't delete seats. All right.

6:51

Cannot delete rows either. Interesting.

6:54

So, we'll go ahead and delete it

6:55

ourselves.

6:58

Let's delete,

7:00

delete, delete,

7:03

delete.

7:05

Yo, what's up, Derek? How we doing

7:06

today?

7:10

All

7:13

right, that hopefully

7:16

hard to refresh this.

7:27

We can delete housekeeping as well.

7:31

And I just wanted to make sure add

7:32

housekeeping to the list. I can't

7:34

remember if we created a handoff doc for

7:37

the other system. Um, I think we did,

7:42

but just double check before I delete

7:43

that one.

7:45

I've already deleted X.

7:52

And then we can delete partnerships

7:54

because we already took care of that.

7:59

All right.

8:05

Wait, are we not live? Oh, we are. Woo.

8:08

I I was previewing my YouTube video on

8:11

uh another tab and it just showed the

8:13

thumbnail, but it showed 12 watching now

8:16

and it made no sense to me. I had to

8:17

refresh that

8:21

housekeeping. All right, let's download

8:24

that. Throw that on our desktop. We'll

8:28

hand that over to the other system uh

8:31

when I go back into it after stream.

9:05

All right, keep that. Keep that.

9:13

Yeah, you know, Grockbots's hitting when

9:15

uh normie friends of mine are hitting me

9:17

up about Grock about Grockbot. Also,

9:20

Omari, my goodness, I swear like that

9:24

dude is on a terror right now. Everybody

9:27

and their mothers are talking about

9:29

Omari and Framework Laptops.

9:32

I have nothing against them at all. Um,

9:35

it's just the power of podcasts. It

9:39

started off with, I believe it was the

9:40

Senra podcast and that was weeks ago,

9:43

but then he recently got on the Lex

9:45

Freedman podcast and I think a few

9:46

others and he's just been on an absolute

9:48

tear to where I now have normie friends

9:50

saying I'm buying a framework laptop and

9:52

I'm installing Omari. I'm like, you do

9:54

understand that's just Arc Linux, right?

9:57

Uh, nothing against it, but um

10:02

I'm curious to see if that is a fad or

10:06

not, right? Um,

10:09

very curious. But soon as Elon's done

10:12

archiving and preparing these handoff

10:14

docs, I'm going to finish cleaning up

10:15

this team and then we're going to dig

10:17

into what I've done so far with the

10:19

database factory. I actually did deploy

10:22

uh a super basic version one uh based on

10:26

the type of clients I work with. I

10:28

figured why not build a database of

10:32

resources uh for that industry for free,

10:35

right? Um

10:38

that come from two primary sources,

10:40

research painpoints on social media, but

10:43

also what I'm building out for those

10:45

clients as well. what types of agents

10:47

and bots that we're creating uh for

10:49

those clients without of course

10:51

revealing client data.

10:53

Um so I'm very excited to show you that

10:55

one. I'll give you a little teaser

10:57

though. If you go to uh AI agents

11:00

construction, this is the super version

11:03

one of the database that's already live

11:05

and uh Grockbot handled all of this for

11:08

me. This team handled it for me. It's

11:12

really awesome.

11:16

All right, let's go make sure all this

11:17

handoff doc. So, he's still working.

11:20

Let's just give it a moment. Why is that

11:22

video or music cutting out? It needs to

11:28

playlist.

11:41

So my site I'm working on rates, videos,

11:44

hosts, guests. Bill Hor got the

11:48

interesting interesting.

11:53

So what I understand is your site just

11:57

rates long form versus

12:00

podcast videos. Trying to understand

12:03

that a little bit better. I can create

12:05

shorts now. Faceless. Amazing. Way

12:07

faster with Grockbot. Hey, let's go. Oh,

12:10

I'm happy to hear it, man. Can't wait to

12:12

see what you do. Share the channel and

12:13

the growth as as as you grow. I'd love

12:16

to to kind of see

12:18

what you build.

12:28

So, we're going to delete housekeeping.

12:36

I'm wondering if we should create a a

12:38

trading bot. What do you think about

12:40

this, Derek? You want to help us create

12:41

this live? I feel like that would be

12:44

insane if we can get that working. I'm

12:46

happy to fund it. Uh, I think we should

12:48

do a prediction markets trading bot if

12:52

anything or trading bot team.

13:08

I still wish we could see what he's

13:09

doing like real time, line by line, tool

13:12

calling, etc.

13:28

Think we're going to unpin Sky because

13:29

we're getting rid of him.

13:55

All right.

14:11

Let's go. Looking forward to seeing it,

14:13

man.

14:24

Now, in the meantime, let's watch this

14:26

video because I didn't finish watching

14:28

it,

14:28

>> which is helpful.

14:30

>> Yeah. So, to be less involved, but you

14:32

you really have to run you have to be in

14:34

the weeds with the team for a couple

14:36

days, a couple weeks before you can run

14:38

it autonomously. And you do that with

14:40

the routines, like we said. Uh, and you

14:44

do it with having agents review work

14:47

that you you teach them how to review

14:49

like you. So a useful thing could be so

14:52

say Greg like I gave all this feedback

14:55

on these posts that it did. I could say

14:58

okay take what we did in this thread and

15:01

turn it into a skill uh that another

15:04

agent can use when it's reviewing. So it

15:05

reviews exactly like Billy, right?

15:07

>> That's exactly what you would do. And so

15:09

if you want to see do you want to see

15:11

how the newsletter actually works? Do

15:13

you do you want to talk about like the

15:14

business case?

15:15

>> Yeah, I think yeah, absolutely.

15:17

>> Okay. So

15:18

>> I do I do because I think that

15:21

there's thousands of potential

15:23

newsletter businesses that could be

15:25

built and run with Grockbot.

15:28

Yeah, thousands for sure.

15:32

So this is this is a local newsletter.

15:34

Um, but I've run newsletters in tech and

15:40

sports for different clients and

15:42

basically all you need to get started is

15:45

a niche and where to get the best

15:47

information to to put in the newsletter,

15:50

right? So, if you're really into I I

15:53

don't know, board games and you know all

15:55

the best board game subreddits and the

15:57

best board game review sites and and

16:00

message boards, then you can make a

16:02

newsletter where you aggregate and

16:05

editorialize.

16:08

>> So, I think a lot of people are going to

16:11

watch listen to this and really like the

16:13

whole agent team idea is going to click

16:15

in their heads.

16:16

>> Yes. Um, I think that there's tons of

16:20

different business ideas that you can

16:21

build on top of this. You know, one of

16:23

them is local newsletters.

16:27

What else could people be building?

16:29

>> Any type of new any type of newsletter

16:31

doesn't need to be local. Uh, probably

16:34

would be easier not to get the ball

16:36

rolling on one that's not local. That's

16:38

a whole different can of worms. Um,

16:42

Shopify. I'm I've got a diff a separate

16:44

Grockbot account that I paid the $200

16:47

for because I'm so crazy about keeping

16:50

the contact separate and it's just

16:52

Shopify. And so I'm training it to be uh

16:55

to use the Shopify developer tools and

16:57

to

16:58

>> source products. I have a sourcing

17:00

agent, a developer agent, and like an

17:03

email agent right now. Um, and I'm still

17:07

setting those up, but Grockbot does an

17:09

outstanding job of researching products,

17:13

sourcing products. Uh, and so I'll just

17:16

I what I would do is say I have an idea

17:19

for this product. Um, I have I really

17:23

like this company and its product. I

17:25

want to do something that's in in the

17:27

same category, uh, similar price point.

17:30

and it knew to go look at import Yeti,

17:33

which like can just look at people's

17:35

Shopifies and tell you like exactly what

17:39

they're sourcing.

17:40

>> All right, let's uh let's just double

17:43

check everything here.

17:48

Hey, I wanted to touch base. What's the

17:50

purpose of this team, this experiment

17:53

pod? What are we doing here?

17:58

All

18:00

right, I just wanted to chat here in the

18:02

team channel. Uh, how's the money pod

18:04

looking? It looks like uh we started

18:06

building this out, but it's been stale

18:08

for 5 days now.

18:12

What's our goals?

18:14

What's the team built for? Give me a

18:17

TLDDR.

18:22

All right. So when I go through these

18:24

experiments, also keep in mind that this

18:27

account has been laying dormant. Uh

18:29

we're just reviving it right now. So

18:31

database factory.

18:34

Hey, I just wanted to get a pulse check

18:36

of this team.

18:39

Where do we stand? What do you need from

18:41

me? Um, I think we should also create

18:47

a markdown file showing me our current

18:50

status as far as

18:55

the original plan we created for this. I

18:57

know we're live. I know you're updating

19:00

with new roles. Looks like you updated

19:02

yesterday, you updated today. So, you're

19:04

actually building things out.

19:07

But, you know, we need a plan for the

19:09

future. We need a plan for revenue.

19:16

All right. And yeah, me personally, I

19:18

love having these separate team

19:19

channels. Um, that way Elon's in the mix

19:22

so he gets the context. Uh, and I don't

19:25

just have to message Elon up here and

19:28

worry that the more I message him, the

19:31

more he will, you know. Oh, that's what

19:35

you're talking about. I just put it

19:37

together. You're talking about Billy

19:38

Horl here. I don't know why I thought it

19:40

was like some other podcaster

19:44

how he did in the video. So if it's an

19:46

interview it raises his question it's

19:47

push back hype BS and then same for the

19:49

interviewer. Nice. Nice

19:55

interesting prediction markets like

19:57

Palshi my agent ran 10,000 tests and the

19:59

best ones were esports and weather

20:01

believe it or not. So betting on weather

20:04

and esports events or outcomes. Yo,

20:07

what's up, Will? Completely new with

20:09

zero tech skills. No, you are not. I

20:14

don't think you are. I think that your

20:16

willingness to learn this stuff is the

20:18

advantage, right? Um it's never been

20:21

easier to build things, right? All you

20:23

have to do is have a vision or or a bit

20:25

of curiosity and just start asking

20:28

questions. They don't even have to be

20:29

the right questions. Just start asking

20:32

questions. I would say, you know, I said

20:34

this at the beginning of the live stream

20:38

from a beginner standpoint. Don't go and

20:40

create some crazy team. Create one bot.

20:43

Test it, one use case, then expand, then

20:47

get the idea of creating this chief of

20:49

staff, right? I don't think everybody

20:50

needs to start with that type of setup.

20:53

I think starting as simply as possible

20:55

and just building one thing, one

20:57

automation, one checker. You could do

21:00

something surface level, some email

21:02

related work, automations, etc. But

21:04

that's all too surface level for me.

21:06

You'll never catch me creating a video

21:08

or, you know,

21:11

telling you to do that via, you know,

21:14

uh,

21:16

I want to choose my words carefully. Um,

21:20

you'll never see me hyping up the basic

21:22

standard workflows uh that everybody

21:25

seems to be spitting out and copying

21:27

right now. Yo, what's up? How we doing

21:30

today? Congrats. That's awesome, Trout.

21:34

It's good to hear. Congrats.

21:37

Let's go.

21:39

All right,

21:41

let's go check back with our team here.

21:54

Okay.

21:59

So, let's go check on it because I want

22:01

to give it some feedback here.

22:10

We need to do a better job at uh

22:14

ready today works with supervision.

22:22

Okay.

22:27

I see. So that navigates through

22:34

jobs, president, owner,

22:37

this is publishing, next project

22:40

manager,

22:41

and then there's specific jobs and use

22:43

cases per. Okay.

22:46

Feel like we can improve on this quite a

22:48

bit though

22:54

for collaboration. Yes. Yes. Uh

22:57

collaboration with humans. Definitely.

23:02

Currently growing my Grotbot team to

23:05

automate my residential real estate

23:07

business. I have a ton of agents and

23:09

created my own system and process over

23:11

the last few days and slowly automating.

23:12

Right on. Coach Clay, that's good to

23:14

hear.

23:16

And correct me if I'm wrong. Did you

23:18

just start with one thing before you

23:20

then start feeding more responsibilities

23:22

to your team?

23:24

Yo, good afternoon, Anthony. How we

23:26

doing today? I'm afraid to touch

23:29

Grockpot after the whole Grock build

23:30

code theft thing. Yeah,

23:33

it's uh

23:36

How long ago was that?

23:38

How long ago was that? I'm not sure. I

23:40

remember that being a a topic of

23:42

conversation, but

23:47

I mean, look at the look at it this way.

23:49

Didn't um Hugging Face get hacked

23:52

recently and then they were just

23:53

acquired for what is it 12 billion or 17

23:56

billion? I can't remember the number. Um

23:58

so these types of vulnerabilities aren't

24:00

necessarily a bad thing. Sure, they're

24:03

bad in the moment. Um but they get

24:06

fixed, they get resolved, they get

24:08

addressed rather quickly. So,

24:12

ah, that's right. That's right. Um, oh,

24:15

you're talking about it creating a copy

24:19

of private repositories if I remember

24:21

correctly. Right. I believe they

24:23

answered that. I I can't remember what

24:25

it was, but listen, you know, I I

24:27

operate based off of one simple rule. U

24:30

there's no more privacy if you're using

24:32

a smartphone. Like, this is not a

24:34

conspiracy. if you're using a

24:35

smartphone, if you're on the internet,

24:37

um there's just no more privacy. So, to

24:40

me, I I'm not as concerned about those

24:43

things anymore because, you know,

24:44

something that's super private, I would

24:46

never hand off to an agent first and

24:48

foremost, right? Like, um so I think

24:50

it's just understanding that we live in

24:52

a world where uh that aspect of things.

24:55

I mean,

24:57

yeah. Yeah. No, and I totally

24:59

understand, right? I totally understand.

25:00

and they were using it as training data,

25:02

but it wasn't every private repository.

25:04

It was specific workflows. It was

25:07

specific um harnesses that would do

25:09

this. Um but it wasn't Grock as a whole

25:13

that was doing this. I think it was a

25:15

specific like um pipeline of using Grock

25:20

that was causing that to happen.

25:23

I mean, a number of different ways. Um

25:26

a number of different ways. I can sit

25:29

here and tell you that Grockbot has

25:31

closed some sponsorship deals. Uh I can

25:34

sit here and tell you that Grockbot has

25:36

gone in and uh cancelled subscriptions I

25:39

didn't know I had. Right? These are a

25:41

lot of the common use cases in videos

25:43

you're seeing right now. Uh but what

25:45

this live stream is all about is let's

25:47

build a team that's actually going to

25:48

make us money. And it may not make us

25:50

money today, but it's going to build out

25:53

those automations and those bots for us

25:55

to then make us money in the long term.

25:57

Um, and this one here, this database

25:59

factory team is this is version one, but

26:01

I'm building a database factory

26:03

essentially, uh, for AI agents in the

26:05

construction industry.

26:11

I have two chief bots, but they are all

26:14

next to bots that work in parallel.

26:16

They're on the left side of your menu.

26:17

Do you think there's a way to organize

26:19

them as if only in a circle in a bubble

26:22

and when you click there the bots of

26:25

each of the main bots can be broken

26:27

down? I don't know if I'm going to be

26:28

able to do that. Um trying to So here's

26:33

what I personally like to do. Let me

26:34

maybe this will help you. Um

26:38

I have pinned agents here. Obviously I

26:40

removed Sky. Sky was another one. But

26:42

this is the chief of staff who's

26:44

essentially routed into every team

26:46

channel. you see here. So, when I start

26:48

talking with the team here,

26:52

Elon's routed in, has context. Uh, I can

26:55

go in here, message him, and he'll know

26:57

exactly what happened in the team

26:59

channel for the database factory. Same

27:01

goes for here. Same goes for here,

27:03

right? Uh, if you're thinking you want

27:05

like a team of chiefs, then what I would

27:08

probably do is either

27:11

still have one master orchestrator uh,

27:13

and then create a Chiefs. I need to

27:15

rename this. These are actually

27:17

harnesses.

27:19

Um

27:21

maybe re, you know, create a new section

27:24

with all of your chief of staffs. Uh and

27:27

then create a team channel with them so

27:29

that you can communicate with all of

27:31

those head of departments in one

27:33

channel. Uh that might be the move if

27:36

I'm understanding your build correctly.

27:50

you. Yeah, you don't need to have an X

27:53

account to use Grock. Um,

27:58

yeah, you don't need to have an X

27:59

account to use Grock. You can sign up a

28:01

number of different ways.

28:10

Nice. Nice. There we go. Hey, don't be

28:12

ashamed. You're testing. You're

28:14

experimenting, right? Like we learn what

28:16

works for us and what doesn't work for

28:17

us. Um, I paid for Replet early on, well

28:21

over a year ago, right? I was paying for

28:23

Replet, uh, dabbling in cloud code and

28:26

then as soon as it clicked, I think I

28:28

believe I made a video about this, but I

28:30

calculated how much I was spending per

28:32

month on Replet versus how much I would

28:34

spend equivalently if I just took that

28:36

money into Cloud Code and paid for a

28:39

bigger subscription. and I was getting

28:41

more usage out of it. The only negative

28:43

per se between the two comparing was

28:46

Replet made it really easy to deploy

28:49

right and host. I have a network

28:52

engineer background. The DevOps side of

28:55

things is no problem for me. The hosting

28:58

side of things is no problem for me. So,

29:01

um, Cloud Code was a no-brainer. But

29:03

now, I mean, all you got to do is ask an

29:06

agent or ask a bot and they'll they'll

29:08

guide you through the steps if they

29:10

don't do it for you. Uh, making it so

29:13

easy. Like, you don't need to have the

29:15

background that I have to host things

29:16

and deploy things yourselves anymore.

29:19

Uh, that is just that's just gone. You

29:21

just need to know how to ask the

29:22

question. That's it.

29:27

Yeah. Bots as needed when the skill or

29:29

process calls for it. Yep. Interesting.

29:32

Derek, should we build one? You want to

29:34

help us build one right now?

29:40

Main difference between Hermes agent and

29:42

Grockbot? Well, Grockbot,

29:47

there's a couple of differences. One is

29:49

open source and one can be completely

29:51

hosted on your own hardware, whereas one

29:54

is not open source and one cannot be

29:56

hosted on your own hardware. So, you're

29:58

at the mercy of Space X data centers.

30:01

I doubt they would ever go offline,

30:03

right? But that's like one of the big

30:06

differences right there. Another one is

30:08

with Grockbot, you're limited to using

30:11

Grock 4.6 or just Grock in general. Uh

30:14

whereas with Hermes, you could connect

30:17

your Grock, you can connect your OpenAI,

30:19

your Claude, whatever harness you want.

30:21

Now, I'm going to put an asterisk on

30:23

that comparison because there are

30:26

workarounds to connect your codecs and

30:28

Claude subscriptions within Grockbot.

30:30

Now, this bot itself is still powered by

30:33

Grock, but the underlying work it does

30:38

routes through the CLI. Whether you set

30:40

up Peekaboo on your Mac, assuming you're

30:41

with a Mac, um it'll just invoke codeex

30:44

or Claude in the CLI, do the work,

30:46

report back, and then the bot itself is

30:49

still powered by Grock. But fun little

30:51

workarounds to use uh other models

30:53

within Grockbot. I have this working

30:56

better on my other system that I created

30:58

on the demo which has turned it into my

31:00

primary. But nevertheless, fun little

31:03

workarounds.

31:05

Uh VPS to Grockbot, you don't need to.

31:09

It's not self-hosted. It's hosted by

31:11

SpaceX. You don't VPS is out of the

31:14

question. If you're talking about

31:15

Hermes, yes. Yes.

31:19

I got to run out for a bit. Daughter's

31:21

kindergarten. Oh man. Busy busy. Hey,

31:23

right on. Nice. We can always do

31:26

something uh on a live stream next week,

31:28

too.

31:30

So, I hope you have a a great rest of

31:32

the day. I hope it's an easy day. I hope

31:35

the teachers don't give you a hard time

31:36

for whatever reason.

31:39

I don't know what it's like with the

31:40

whole back to school stuff, but

31:43

it's funny you say kindergarten open

31:45

house. I've never heard it's like, are

31:48

they becoming real estate agents? What

31:50

is going on here?

31:56

Sorry, I had to mute there.

32:10

Oh, that's right. I created tags in our

32:15

mailbox and created aliases.

32:28

Yeah. So, what happens daily right now?

32:30

What happens weekly and what happens

32:32

monthly? What can we do to improve?

32:34

Because the front-end UI leaves much to

32:38

be desired for. So, we got to we got to

32:40

put in some work there. But, I'm just

32:42

trying to get a pulse checker on,

32:46

you know, what's happening daily,

32:48

weekly, and monthly at the moment.

32:50

And what's our goal? I'm pretty sure we

32:52

defined it, but I want you to tell me

32:54

what's our goal for this project.

33:01

Uh, this pod sells one thing.

33:04

Interesting.

33:10

Interesting. Really, I could have sworn

33:12

that the money pod was something else.

33:17

So the money pot it looks like we are

33:21

I see.

33:24

So what's currently configured? What's

33:26

our goal? And what's configured daily,

33:29

weekly, and monthly right now.

33:43

This is what happens when you build too

33:44

much.

33:47

We do a lot of experimenting here on the

33:49

channel though, so it's all a part of

33:52

the process. The one that I know is

33:54

going to stick is the database factory.

33:56

We're still experimenting here. We might

33:58

delete these teams and then just start

34:00

from scratch.

34:06

Wait a second.

34:09

To find one clear mud product a month,

34:11

not to ship every morning. I hunt what's

34:14

actually moving on X last 30 days pulls

34:16

what people are saying architect turns

34:18

it the pain into a pack GTM or go to

34:22

market bot says whether anyone would pay

34:26

water and mud only jump in if we pursue

34:28

ah

34:31

okay

34:34

okay so I I think we need to shift this

34:36

because we're moving clear mud related

34:38

responsibilities to another system so I

34:41

wonder if we just don't simplify this

34:43

into a

34:48

listen for pain points,

34:58

create a prototype,

35:00

create a go to market strategy, and then

35:02

present me with that output one new

35:06

product every single day. And then I

35:08

choose whether or not we want to move

35:10

things down the pipeline, right? Also,

35:13

what do we have configured as far as

35:15

routines are concerned within this team

35:18

when it comes to daily, weekly or

35:20

monthly routines?

35:30

Just know that Grockbot is essentially

35:32

Elon's digital optimus.

35:36

Feel like he's learning from all the

35:37

Grockpot usage to eventually release

35:39

agents that actually do Yeah. Yeah.

35:42

Yeah, for sure. Yo, what's up, Jordan?

35:46

Hey, no problem for me.

35:49

I mean, one account got close to 100%,

35:52

but I was very, very mindful of it. Uh,

35:55

and then we got that that reset two days

35:57

ago, which is awesome. Which is awesome.

36:02

Well, I think we're going to see more of

36:03

those personally. Um, but I'm I'm on

36:07

average to spend about

36:10

10%

36:12

maybe 12% a day, which leaves me with

36:14

some extra percentage. Uh, and that's on

36:17

my Clear Mud system. On this system,

36:18

it's completely separate. So, I have two

36:20

different Grockbot systems right now

36:22

that I'm testing with.

36:24

One I actually built out in that video

36:27

from a few days ago, which is now turned

36:28

into my primary system for ClearBud. So,

36:31

I'm not going to demo that one too much

36:33

more moving forward. Uh, I'm primarily

36:35

going to demo this account here. That

36:37

other one's running off of a Super Gro

36:39

Heavy and this one's running off of

36:41

Cursor Ultra.

36:44

So, we actually experimented with this

36:47

on Wednesday's live stream and uh we

36:51

were following a certain list of hacks

36:53

that were recommended on X. Um,

36:59

yes, it could do it, but it's not

37:01

natively built in. You have to do you

37:03

have to take a lot of steps to to pull

37:05

that off

37:09

and you have to set up a a phone number

37:11

based off of how we did it at least. Can

37:13

it send voice notes in chat? Let's ask.

37:19

Hey, if I send you voice notes, can you

37:22

send me voice notes?

37:27

Hey, if I send you voice notes, can you

37:29

send me voice notes?

37:32

What's going on here? I'm noticing a

37:34

weird little bug is when I try to record

37:36

something here.

37:38

Hey, I'm curious. The microphone button

37:42

on this,

37:44

you know, field where I message you

37:46

doesn't work. Uh, I've tried both ways.

37:49

Also, can you send me voice notes? Can

37:51

we send voice notes back and forth? I'm

37:53

very curious.

37:59

Okay. Right on. Yeah. Hope hope it's

38:01

helpful.

38:06

I already answered this. I think so. I

38:10

think so. I think it's worth an

38:12

experiment for sure if you have the

38:14

budget, right? If it's not going to hurt

38:16

you, if it's no skin off your back, then

38:17

yeah, why not?

38:22

Cursor Ultra is uh $200 a month.

38:30

Yeah, exactly. Exactly. The reason I

38:33

justified what I'm spending is that

38:35

super Grock heavy account in itself,

38:38

it's already paying for itself. It's

38:40

already paying for itself with what it's

38:42

doing for me. Uh it's freeing up my

38:44

time, which is awesome. This other one

38:46

is more of a business expense for Clear

38:48

Mud to experiment, to demo. This is for

38:51

you guys more than it's for me, right?

38:53

Um, so if you need something built

38:56

specifically, if you want to see how to

38:58

build something, just ask. I'm happy to

39:00

build something live showing you the

39:02

steps that I would take.

39:06

Yeah, I think they are right. Maybe I

39:10

can't remember. Let's go take a look.

39:14

I want to say if we go to bot

39:23

believe they shared yeah all super grock

39:25

plus cursor pro plus and cursor team

39:28

subscribers now have access.

39:33

So, Super Grock Plus. Uh, how do we

39:37

easily get to

39:41

go to grock.com?

39:44

Looks like I'm actually signed in with a

39:45

different one here.

39:49

Where are the plans?

39:51

Grock plans. Super Grock.

40:01

I'm pretty sure it's the $30 a month

40:02

plan if I'm not mistaken.

40:06

Oh, maybe it is 100. Okay.

40:13

H interesting.

40:19

Ah, very nice. Very nice.

40:24

You can do it from a laptop. They don't

40:26

have an Android app for it yet, which I

40:28

am bummed about.

40:30

I am very bummed about that because I'm

40:33

an Android first mobile user, but then I

40:36

do have this Mac as a

40:39

computer here as my daily driver for

40:42

creative.

40:56

Interesting how our music keeps pausing

40:58

in the background.

41:05

Go back to flow music.

41:15

Do I have to go in here?

41:18

No. Let's go back. I think we should be

41:21

able to play the playlist as a whole.

41:24

And we just hit

41:28

There we go. I think what it might be

41:30

doing is it might be shuffling down to

41:32

the end and then ends right here. So,

41:36

let's turn off shuffle and just do

41:38

automatically re-re replay.

41:44

I cannot send voice notes back. No audio

41:46

tool on the seat. I send text. Okay.

41:50

The other org has outbone voice to you.

41:52

That's a different machine. Yeah.

41:58

Got it. Got it.

42:06

You know, I'd love to will I really

42:08

wanted to have um because I'm not a

42:10

trader first and foremost.

42:13

I really wanted Derek to help to to

42:15

build it with us live because he has

42:17

that experience.

42:19

Um, I'm a, you know, I really stand

42:22

behind making sure you have the

42:24

experience to build something. Uh,

42:26

otherwise you're just going to build a

42:27

super surface level trading bot that's

42:29

not going to really make you any money.

42:32

I think from a trader's perspecting ha

42:35

perspective having that foundational

42:37

knowledge of how to trade of

42:42

if then statements of certain criteria

42:46

right to track you're going to be far

42:48

more successful in creating a trading

42:50

bot with that foundational knowledge and

42:53

configuration but just telling it to say

42:55

hey I just gave you $50 on this MetaMask

42:58

account I want you to go to this

42:59

prediction market start making us some

43:01

money that's not enough. It's most like

43:04

nine times out of 10 it's going to lose

43:06

money, right? Um but

43:09

if you give it specific tasks, specific

43:13

routines to do certain research, a

43:15

specific niche to target, uh I think

43:18

that's when you become more profitable.

43:20

I think it looks like we're going to

43:21

build that next week because Derek had

43:23

to take off. Dererick's a busy man. He's

43:26

got like what is it 17 kids? No, I'm

43:29

joking. He's got seven though, which is

43:31

a lot. Um, so he had to take off. But to

43:35

answer your will, I think next week I'm

43:37

going to DM Derek to see if we can't uh

43:40

get him on a live to to essentially just

43:43

build a trading bot live. That would be

43:45

a fun one.

43:48

I want to go back and focus on this

43:50

database factory. Let's see.

44:01

Interesting. It set a goal for 100k a

44:03

month by year end from that. What does

44:07

that mean? By year end from that year

44:09

end from that. From what?

44:14

Where is it getting that from? Hey, I'm

44:16

with it. I'm not I'm not mad at that

44:19

goal right there, you know.

44:23

All right.

44:32

messaging claude for contractor

44:33

blueprint spec and codeex to ship it. So

44:36

it looks like they are working right

44:38

now.

44:41

Um

44:44

passing that to Claude. Okay. So I mean

44:47

based off of this original prompt, I

44:50

didn't even tell them to do anything. I

44:51

just said the front end UI leaves much

44:53

to be desired for. So, we have to put in

44:55

some work there. And they got started

44:58

off of

45:00

prompting our uh our bots that handle

45:02

the codeex and cloud harnesses. Awesome.

45:05

Awesome.

45:06

So, let's go check in on Money Pod.

45:08

Interesting. Really, I could have sworn

45:10

we had money pod with something else.

45:11

Goal one paid to install bid desk.

45:13

That's me followup.

45:22

Got it.

45:24

Oh, interesting. I kind of want to start

45:26

from scratch with this team here.

45:31

Yeah, I think we're going to trash this

45:33

experiment and start from scratch.

45:37

Enjoy your last few moments.

45:47

Oh, I cracked myself up. I'll let them

45:50

enjoy their last few moments. Let's see

45:51

what they say as bots.

45:56

Yes. New loop. Listen for pain. Ship a

45:58

prototype. Write the go to market

45:59

strategy.

46:01

Put it on the walk-in tab. One new

46:03

product every day. You decide what

46:04

continues. Got it.

46:07

All right. So, I'm also curious as to

46:09

why we're using water and mud. Those are

46:12

our old Clear Mud team members. Uh we

46:14

primarily use the team we've built here

46:17

plus our Codeex and Claude bots um that

46:20

route work through each respective model

46:23

whether you're choosing Opus or Fable or

46:26

GPT Soul or GPT Terra Luna etc etc. All

46:30

that work gets routed through them.

46:36

See what did they say?

46:39

Enjoy your last few moments. Come on,

46:40

say something.

46:42

I guess Elon's the only one answering

46:44

right now. Um,

46:50

yeah, we're just going to delete. Going

46:53

to delete. Fun experiment. Let's clean

46:55

things up in this system.

47:07

All right, I went ahead and deleted the

47:09

money pod team.

47:12

That was an experiment. And uh

47:15

we want to go back to the drawing board

47:17

there. So, also I'm noticing that

47:20

sometimes mud and sky and water are kind

47:23

of spliced into certain team channels.

47:26

That shouldn't be the case. They're a

47:27

clear mud team. We're handing that team

47:29

off to um to our other system, right?

47:32

So, I'm getting ready to delete those

47:34

agents shortly.

47:39

You know what? We might keep Mud because

47:42

didn't we build a Opus clone here? We

47:44

did. We did.

47:47

But I've already added that repo.

47:51

Did we add this mud clip? Feel like we

47:55

do. Let's go check.

48:00

Mud clip.

48:04

Is it mud dash clip?

48:19

Give me a link to this uh repo, please.

48:34

Yeah. If anything, I'll keep mud and

48:35

I'll move him into

48:38

actually rename this one engineering.

48:42

Move mud down there

48:45

because he was more of our engineer than

48:47

anything.

48:50

I'm going to remove that tag for now.

49:01

And then let's go ahead and create a

49:03

channel for Elon. Elon Mud, Claude, and

49:09

Codex

49:12

team chat or team channel.

49:18

Move that up here. That way they all

49:21

have their own way to communicate. And

49:23

then this is one we need to archive and

49:24

delete shortly.

49:30

Yeah. So, what I'm ending up doing here

49:32

is I moved mud into engineering. Um, so

49:35

we're going to delete and archive. Well,

49:37

first we'll archive, then we'll delete

49:39

water and sky shortly because those

49:41

responsibilities we're handing off to

49:43

the other system.

49:49

Dang, Mr. NY,

49:52

what angle did they pick? Who

49:58

is it? Nice.

50:01

I have not uh I have not experimented

50:04

with uh GLM 5.3 yet,

50:09

but are you still using Hermes and

50:10

OpenClaw? Yes, I'm not using them with

50:13

Grockbot. I'm using them independently.

50:15

So, I'm going to be honest. Hermes I

50:17

deploy Hermes agent systems for clients

50:20

um

50:22

because of how stable it has been since

50:26

day one and reliable. That's why I chose

50:28

Hermes Agent. I have my own Hermes

50:30

agent, but it doesn't do much aside from

50:33

I don't give it too many

50:35

responsibilities for for clear mode as a

50:37

whole. My openclaw setup on the other

50:39

hand that I originally built out, you

50:41

know, end of December, beginning of

50:42

January, I still use daily. I still rely

50:45

on daily. Um, yeah, I still, long story

50:49

short, yes, I still use Open Claw. I

50:51

still use Hermes. I still use Buzz. Um,

50:54

but Buzz is more for collabs now because

50:57

I want to use the collaboration feature

51:00

of bringing in other humans and agents

51:03

and those are its responsibilities right

51:05

now. Open Claw's day-to-day clear mud

51:09

ops. Grockbot now is I'm giving it some

51:14

responsibilities, but it's not going to

51:16

outright replace OpenClaw or Hermes for

51:18

me personally. Um, but I am, you know,

51:22

it it it's going to be doing a lot for

51:24

me in the future. But is it gonna

51:26

outright replace OpenClaw? I don't think

51:28

so. I'm too much of a a techie and an

51:31

experimentter to to just trash something

51:33

like that, trash something that's fully

51:35

in your control on your own hardware,

51:37

right? Sure, you run the risk of that

51:39

hardware crashing, etc., etc., but

51:42

that's the type of person I am, right?

51:45

So, I'm never going to trash my open

51:47

class setup, personally. Um,

51:51

and never say never. You never know what

51:53

happens in 10 years, 15 years, but for

51:55

the foreseeable future, like I don't I

51:58

may make content around all these new

52:00

solutions that drop. Um, but I'm not

52:03

like these other guys who were saying I

52:04

now migrated everything to these other

52:06

systems. Uh, if that were true, then

52:09

they never really utilized those systems

52:11

in the first place. uh because m keep

52:14

continuously migrating from system to

52:16

system to system shows how little you've

52:19

actually done in those systems uh in my

52:21

opinion right because

52:24

if you if you're truly relying on that

52:26

tech uh the amount of time it's going to

52:28

take to migrate from one system to the

52:30

next I is insane is insane like um I

52:35

don't know I'll stop talking there I'll

52:37

stop talking there

52:40

yeah this right here This is this is a

52:42

great use case, great little workaround.

52:45

Um, but you're still at the end of the

52:47

day using Grockbot, right? Hermes does

52:49

have its own bot feature. If you guys

52:51

know, Hermes blatantly copied Grockbot

52:54

within hours. I feel like within their

52:57

desktop app, it was hilarious to watch.

53:05

Nice. Nice.

53:08

Yeah. So, this was my original thing is

53:11

Grockbot is primarily going to focus

53:13

around X growth for clear mud and being

53:16

more consistent across socials. Uh,

53:18

primarily X, right? I'm giving it more

53:20

social first responsibilities than

53:23

anything. I'm not giving it YouTube

53:25

responsibilities because I have certain

53:27

automations in place with my open cloth

53:29

setup that are just tried and true. They

53:31

work. I don't even have to think about

53:32

it. All these live streams, they've been

53:34

automated uh when it comes to

53:36

post-prouction. So, you'll notice within

53:40

24 to 48 hours after all of our live

53:42

streams since the beginning of the year,

53:44

right since I set this up with OpenClaw,

53:46

uh it'll go in, download the transcript,

53:48

create timestamps, create an SEO

53:51

optimized description based off of

53:52

what's discussed in the live stream, add

53:55

tags, and then update YouTube. I don't

53:58

have to touch it. Sure, the first few

54:00

weeks I was monitoring how it was doing.

54:02

I was tweaking it. It did break once, I

54:05

want to say, in April. Um, but I just

54:09

set it, I forget about it. And that's

54:10

one of many automations I have in place

54:12

with OpenClaw. I have no intention of

54:14

migrating that to Grock or anything

54:16

else, right? There's just no point. It's

54:17

a waste of time. A waste of time.

54:20

Yeah. So, here, Grizz, this is actually

54:23

a great I love this comment. It's like

54:25

Hermes with training reels. So, listen.

54:27

Grockbot, I don't like

54:31

it's not revolutionary. It's not

54:33

gamechanging in the sense of it's

54:35

inventing anything new. What it's done

54:38

is reduced that friction that people had

54:41

with trying to set up an OpenClaw or

54:44

Hermes or Buzz system for that matter.

54:46

It's so easy and seamless to get

54:49

started. And that's what's gamechanging

54:51

about Grockbot. It's not the foundation

54:53

of it. It's just how easy it is to use.

54:57

Like that's what we needed. And I think

54:59

that's why it's making waves and and

55:01

it's essentially has gone viral, right?

55:06

Is it? Yeah, I haven't played with it

55:07

yet. So, one of my clients, we are

55:10

setting up individual desktop a access

55:12

for the employees. Uh, I did start

55:15

testing from an employees perspective

55:16

with the bots. It's super fun to use.

55:19

Um, not the same as Grockbot, but still

55:22

super fun to use. That staff's going to

55:25

have a lot of fun uh once they get

55:26

access to that next week. I love

55:29

Grockpot. It just blows through my usage

55:31

like crazy. Such a low pool with Super

55:33

Gro Heavy. Yeah, I hope they fix this.

55:36

Um,

55:38

the here's the reason I have two subs,

55:40

Adam, is because one cursor ultra sub

55:43

for me giving it responsibilities for

55:46

clear mud, but then also demoing and

55:48

experimenting with Grockbot from a

55:51

YouTube demo perspective. I was max I

55:54

maxed out way too quick, right? So, I

55:56

decided to just bite the bullet, so to

55:59

speak. I have a super grog heavy for

56:01

clear mud now. And then this cursor

56:03

ultra which is going to serve as a

56:05

YouTube demonstrated experiment lane of

56:09

just testing and having fun. And I

56:11

haven't hit 100% yet. Um

56:15

I it's my hope that they fix this

56:17

though. It's my hope that they fix this.

56:19

I think they're going to improve uh and

56:22

and and also become more transparent as

56:26

well. Like I'd like to see a

56:28

side-by-side comparison into how much

56:31

usage can I use via Cursor Ultra versus

56:34

how much usage can I use via Super Grock

56:36

Heavy. If it's the same and the only

56:39

reason I have these subscriptions is for

56:41

Grockbot, then it makes sense to only

56:43

pay for Cursor Ultra. But if you get

56:46

slightly more out of that Super Gro

56:48

heavy account than you do out of that

56:49

Cursor Ultra account when when I I think

56:53

we're going to get more clarity in the

56:54

coming weeks. Um, we just got to be a

56:56

little bit patient.

56:58

It's a good gateway for newbies.

57:01

Then they can graduate to her. Yeah,

57:03

they can graduate to whatever they want

57:04

to set up, right? It doesn't have to

57:05

just be Hermes. Listen, I,

57:11

as you've learned, I'm deploying Hermes

57:13

agent systems, but I'm still team

57:14

OpenClaw. If I had to choose one over

57:16

the two, it's OpenClaw.

57:19

That's just my preference.

57:25

acting like they're going insane, left

57:27

running. Yeah. So, I think a lot of this

57:30

is

57:34

lack of context around what you want

57:37

them to accomplish, which is sometimes a

57:40

bad thing, sometimes a good thing

57:41

depending on what you want to

57:42

accomplish, right? Um

57:47

yeah, I think a lot of this is lack of

57:49

lack of uh context of like what's its

57:52

purpose, right? Uh if you clearly define

57:55

its purpose and the certain loops you

57:57

want to achieve with your bots, uh when

58:01

you want them to stop doing work, that

58:03

won't happen, right? But if you leave it

58:06

open-ended, I could I could see a world

58:07

where that that could be a problem.

58:09

We've had some comments on some of the

58:10

videos we posted of people just uh

58:13

sending a task and the bot not stopping

58:16

because they weren't specific enough

58:18

with, you know, stop at this output,

58:21

right? Um so I I think um

58:26

yeah, I think that's case by case there.

58:30

Newb integration they enabled for X

58:32

after releasing Grockbot makes this

58:33

concept interesting. The bot can now

58:35

have its own presence on the platform

58:37

associated with your account. Yeah. So,

58:39

one thing I did notice though, if you

58:41

give full access and automation to your

58:43

Grockbot to post on your behalf, every

58:45

post says made with AI. But if you say

58:49

stage this draft or stage this reply and

58:52

then you go on mobile or desktop and you

58:55

go into your drafts and you make one

58:57

little change and hit post, that made

59:01

with AI tag is gone, right? So, if you

59:04

automate the posting and replying, yes,

59:06

you're just going to look like a bot,

59:07

though. Uh, I don't want to look like a

59:09

bot. I want it to help me, but

59:12

ultimately I'm the last person to

59:14

essentially send that prompt or let me

59:16

rephrase that. Send that post.

59:20

No. All good. I think it's on topic. So,

59:23

I hope that answers it.

59:29

Does low friction reduce usability

59:31

compared to other harnesses? Yeah. So,

59:35

I don't think so. I don't think so. I

59:38

think it's, you know, Grockadot is built

59:40

for a very specific type of person,

59:42

right? Um,

59:45

and we all have a touch of that type of

59:47

person. Like, a lot of us are techies, a

59:49

lot of us aren't, right? For some people

59:52

in installing Open Claw is intimidating,

59:54

for some it's not, right? It's like uh

59:57

and I I think as Grock gets better, I

60:00

I've said it before in the past and I've

60:02

gotten a lot of uh flak for it, but I

60:04

think Grock, Space XX, let me let me

60:07

rephrase this. SpaceX AI is going to be

60:09

one of the top frontier models in the

60:11

coming years.

60:13

Uh Claude is already I their downfall

60:16

needs to be studied.

60:18

Like, yo, what's up, David? How we doing

60:21

today?

60:23

They confirmed it's the same.

60:24

Interesting.

60:27

Yeah. Yeah. Very cool. I think we saw

60:29

that when we briefly went to go look.

60:38

Share as template.

60:41

Nice.

60:43

Now, does it pull

60:46

I'm curious if it pulls out any like

60:48

sensitive data in those instructions

60:50

automatically. I would imagine it does.

60:56

See where are we at? 11%. I'm not sure

60:58

where we started this session is. So I

61:00

wish I I kept track of that.

61:10

Yeah, I'm not going to delete them yet.

61:11

I want to go into my other system and

61:13

finish migrating these two.

61:17

Okay,

61:20

let's go back to

61:22

our database factory.

61:34

So, deal looks like he has some

61:36

responsibilities.

61:41

Subscribe form is live. Cool.

61:47

I'm curious to see like what uh what the

61:50

new UI is going to look like.

62:12

Hm. Okay.

62:14

Now, let's go back down to our

62:15

experiment pod.

62:19

Going to rename this experiment pod.

62:32

All right, cool. Let's um let's force

62:35

run one right now. Build me something.

62:42

and list out a table of all the routines

62:46

associated with this experiment pod as

62:48

well. So I get a

62:52

just a visual proof of what happens each

62:55

and every day.

62:58

Oh, you know what I'm realizing? We need

63:00

to add uh

63:03

how do we add

63:06

should be an easier way to add members,

63:07

right?

63:13

Huh? It's weird how Elon's not in here.

63:20

Is it settings? No.

63:26

You would think there'd be an easier way

63:28

to just add a member. No.

63:31

Do I have to create a new channel for

63:33

that? I'm shocked he's not in here.

63:36

Also,

63:38

we need to add Elon, our chief of staff,

63:41

into uh

63:46

Huh. into this uh team channel.

63:55

Hey, I noticed you're not in the team

63:57

channel for the experiment pod. Can you

63:59

add yourself? Can't find an easy way to

64:01

add you.

64:03

Interesting.

64:05

I mean, I could just create manually

64:07

another channel, but since they're

64:09

already doing work,

64:12

let's wait here.

64:36

Interesting. Okay.

64:40

Yeah, very interesting.

64:43

Okay. Looks like he's added in there.

64:45

Awesome.

64:52

And they adjusted the team, right? Check

64:55

out the engineers cuz they don't need to

64:56

be a part of each team pod.

65:00

Great.

65:04

Looks like Fable is building something

65:07

called same model. Don't know what that

65:09

is. Let's see how it looks.

65:15

Let's say model. Paste a stealth slug

65:18

plus date window show identifiers.

65:21

job people actually did name the

65:23

hangover.

65:30

This is very brief. Interesting.

65:39

All right. I'm very curious about

65:44

Grockbot as a whole. Do you operate

65:47

based off of certain skill MDs or

65:50

certain

65:53

a certain file system that we can then

65:56

organize certain markdown files so that

66:00

you follow them or each individual agent

66:02

follows them? I want to get a better

66:04

understanding of how we can improve

66:08

certain automations namely for database

66:10

factory for the experiment pod and any

66:13

other future teams we create.

66:20

My buzz partner installed Grockbot and

66:22

wanted me to collaborate on his

66:24

projects. It seemed impossible.

66:25

Interesting.

66:27

Yeah, I think right collaborating I from

66:30

to my understanding you can't bring

66:32

another human into your Grockbot system.

66:35

What you can do on the other hand is

66:38

either have one of your bots communicate

66:41

with another person through the remote

66:43

computer. Um,

66:47

give your Grockbot bot and underlying

66:50

bots an agent mail identity so that they

66:53

can email people at will.

67:00

Um, or if you're building a web app or

67:04

whatever, create a private GitHub

67:06

repository with your bots, have it sh,

67:09

you know, commit to Git that way and

67:11

then share that repository with somebody

67:12

else to collaborate.

67:16

That's that would probably be the most

67:18

efficient route there, AV.

67:21

But as I learn more, I'll uh I'll

67:23

definitely share. Matos

67:27

to the bane too. How are you?

67:31

Been AFK last couple weeks, but I've

67:33

watched your streams. Testing Crockpot

67:35

stickked more than Buzz. Nice. Nice. I

67:38

love it. Hey, observing is what it's all

67:40

about. Listen, I still have a Buzz boot

67:42

camp coming out. Actually, I'm still

67:44

going to record a firsttime setup use

67:47

for Buzz. Uh because I think Buzz is

67:49

still very powerful for its use case.

67:52

Um, I'm also going to be teasing

67:54

something to where we're launching a new

67:57

series on the channel, um, in

67:59

collaboration with a digital marketing

68:01

agency. Uh, the CEO is going to come on.

68:04

It's not going to be live. We're going

68:05

to record these videos. Uh, but we're

68:08

going to be creating agents and bots for

68:10

her business, which are then going to

68:13

serve as kind of like a dual purpose,

68:15

right? We're building out her agent team

68:16

for her, but we're also doing it live on

68:22

video so that we can then postedit and

68:24

create series on how to create an an XYZ

68:28

bot for XYZ. And I I say XYZ because

68:31

that's like fill in the blank, right? Uh

68:34

I think it's going to be super valuable.

68:36

Um, that is

68:41

at least two weeks out until we start

68:44

posting those videos. Uh, possibly

68:46

three. But it's it's on the calendar to

68:49

record. It's it's on the calendar. Um,

68:54

forgive me. I need to put this on

68:55

silent. I hate these little They just

68:59

They distract me too much. Um, it's in

69:02

motion. Let's just put it that way. It's

69:03

in motion. So, I'm super excited to

69:05

record those videos and uh share more

69:08

about um this project as soon as

69:11

possible.

69:14

Already had an iMessage gateway running

69:16

on Mac Mini for Hermes to use.

69:20

Nice. Lovely workound here. Awesome,

69:26

dude. Awesome.

69:39

Each seat has a job description in

69:41

memory. Yes, that's standing intent. A

69:44

routine is a clock. Yep. Plus a prompt.

69:46

Yeah, we know. A skill is a shared

69:48

recipe any seat can run. Okay. I don't

69:50

think he understood my question. Um to

69:53

make a new team tighter.

69:58

No, what I meant is is there a directory

70:01

somewhere like let me go check on my

70:04

route.

70:10

Um, command shiftG I think to search.

70:15

I guess not. Maybe I need to be in

70:17

finder.

70:19

There it is.

70:28

No. Yeah, it would be on the system

70:30

itself. I don't think it would be local.

70:35

Yeah.

70:37

Skills love on Grock bots computer. Got

70:39

it. Got it.

70:44

Show me a file tree

70:47

of how it looks and where to access that

70:49

on uh the remote computer.

71:12

really wish this was rich.

71:15

Could I get those responses in like rich

71:17

text? Is that possible here in Grockbot

71:19

chat?

71:22

because that table is uh while I can

71:24

read it, it's not uh it's not very uh

71:28

user friendly.

71:49

All right, we are back.

71:54

There we go.

71:59

So, daily products

72:02

7 days a week. It hunts.

72:05

It's live. Last 30 days, none. Dump when

72:08

this room asks. Okay. So, it's on

72:10

demand. Product architect

72:12

802. So, it's an hour later. Pick latest

72:17

live. Go to market. That's 30 minutes

72:20

roughly 30 minutes later. Every single

72:21

day pulls from latest. Okay.

72:27

Okay.

72:45

Okay.

72:47

I have an idea.

72:49

Let me go find this repository. One

72:51

second.

74:44

Home box scratchwork is

74:46

forward/workspace skills you care about.

74:51

Okay.

74:56

Hey, this is good to know here.

75:07

Ha. They call them agents.

75:10

They don't call them bots.

75:13

Awesome

75:23

man. This in itself can be a video right

75:25

here.

75:32

Interesting.

75:35

Wow. So, this is very valuable

75:38

information. Um, I'm kind of thinking

75:41

about turning this into a YouTube video

75:45

where it's about, you know, how to

75:47

properly configure your bots via

75:50

Grockbot.

75:52

Um,

75:55

any thoughts?

76:01

Oh, Nikolai was already on top of it.

76:17

I want you to build a um build an

76:20

Excalibur canvas.

76:24

I think we should have a screenshot of

76:26

this file tree, but also you know where

76:29

things get placed, what happens when you

76:31

create a new bot, how you should create

76:33

a new bot, um what you should be telling

76:37

your bot on how to store certain

76:39

information. uh I think would be super

76:42

valuable and I can walk through this uh

76:44

and we can walk through it in a YouTube

76:46

video recording and and you know a show

76:49

off existing systems but then b set up

76:52

new bot teams based off this.

77:03

This is taking a while.

77:07

Hey, where do we stand with this?

77:16

Yeah, they're already working there.

77:21

Looks like they've already built a

77:22

prototype here.

77:25

Super basic. So, I need to uh

77:31

Oh, I am. I am for sure.

77:35

So, what I do with that canvas tool is I

77:37

first create an Excal drawing and then I

77:40

hand it over to it and then it it

77:42

creates its version.

77:44

But I haven't recorded a video like that

77:46

since last week cuz the last video I

77:49

recorded, we really shot from the hip on

77:51

Tuesday to where I thought it would be

77:53

fun to when I set up that second account

77:55

to just record what I was doing and it

77:57

turned into a very great video, super

78:00

valuable. So, I'm going to start to to

78:02

record more videos like that to where I

78:04

don't do any prep work. I just have an

78:05

idea and I'm recording my screen and I'm

78:07

talking through it, right, real time.

78:10

Um, and then of course adding those

78:12

those types of elements if we need to.

78:16

But yeah, it's a two-part thing. I don't

78:17

build the canvas from scratch. I use an

78:21

Excal drawing to help me come up with

78:24

our canvas on clear mud canvas, if that

78:27

makes sense.

78:54

Oh,

78:55

yes.

79:02

Yeah, I just noticed I didn't see any

79:05

notification saying I needed to approve

79:07

a command within press. I think um you

79:10

know a rule of thumb moving forward I

79:12

want to be as handsoff as possible here.

79:14

Um so there should be no blockers, no

79:17

allow once, allow all. Uh whatever we

79:20

can do to override that within this

79:22

grouping would be great.

79:28

Sweet.

79:32

Oh, Grockbot Opus clone is what it was

79:34

called.

79:37

Yeah. So, we s saved Mud clip. So, I

79:39

could in theory get rid of that one,

79:45

but I think we're going to keep mud for

79:46

now.

79:54

Interesting.

80:01

We're about to find out what it does.

80:03

And let's see. Last time I checked,

80:04

we're at 11%. So, we burned through 3%

80:08

already. Interesting.

80:10

And how many days left do we have? Reset

80:12

in seven days. Change limit. Oh, it

80:16

takes us to cursor. Okay.

80:22

It'd be curious. I want to monitor this,

80:24

but I'd be curious to see, you know,

80:26

with these automations between the

80:28

experiment pod and the database factory.

80:31

Um,

80:37

no. Are you talking about my team? So, I

80:40

have my team specifically set to use uh

80:43

these harnesses here with these two

80:45

bots.

80:47

Where are you getting that from? Are you

80:48

getting that from my screen? Because I I

80:50

you know I set this up to to work a

80:52

certain way so that it could utilize

80:54

Opus or Fable through my Claude sub and

80:57

then also for GBT Soul um through my

81:01

codec bot.

81:07

So, the current plan that we're demoing

81:09

today is the Cursor Ultra $200 month

81:12

sub.

81:31

No, all good. All good.

81:36

Chris McConey.

82:18

I mean, he posts a lot, but I'm not sure

82:20

where you're seeing this from.

82:23

Ah, okay. Okay.

82:50

Okay.

83:10

Um, feel like it just made it worse,

83:15

right? What just happened?

83:28

I'm gonna wait for them to finish.

83:32

Um

84:14

Trying to find something. I could have

84:16

sworn I bookmarked it. I guess I didn't.

84:18

Maybe it's in here. Let me go check.

84:21

off screen.

84:24

I had this repo that I wanted to start

84:27

using. I was just trying to wait for

84:31

a good time

84:33

to use it. Hold on. I can't seem to find

84:36

it though. It's not here. Shoot.

84:46

That's a bummer.

85:15

Oh, what was I think it was

85:55

Okay, I did find a UI UX skill that I do

85:59

want to actually feed to this team.

86:07

All right. I'm not sure if you guys are

86:08

done with the work yet, but I didn't

86:10

think it would have been possible, but

86:14

you know, a the website looks exactly

86:17

the same and now the nav header is

86:23

essentially misaligned. Um, I'll send

86:26

you a screenshot, but

86:29

I'm wondering if I need to do a better

86:30

job at giving you guys a skill to follow

86:33

moving forward. Um, because you're just

86:36

not doing the work that I expect you to

86:39

do. So,

86:46

let me know how I can prompt you to do a

86:48

better job here.

86:57

Welcome back. Welcome back.

87:16

Oh, interesting.

87:21

I wonder if this is Yeah, interesting.

87:24

Very interesting.

87:26

I'll have to look into that.

87:33

Oh, I see what you mean. I see what you

87:35

mean. So, it was inspired off of this

87:38

Greg Eisenberg video, right? Um, that's

87:42

why I used like the same dollar sign,

87:43

dollar sign, dollar sign uh in our

87:45

title. And I've already created the B

87:48

database factory here. So this is not a

87:50

short-term plan, right? This is a

87:52

long-term plan, but it's not something

87:54

random, right? I built this uh I

87:57

squatted on this domain a while back

87:59

because I started building AI agents for

88:01

the construction industry, primarily

88:03

GC's uh and then construction companies

88:06

focused around demo and abatement. Um

88:09

but I thought what if we build our own

88:12

database factory? So, this is kind of

88:14

like the version one of it and its

88:18

purpose is to generate traffic. That's

88:20

our first goal is how do we get our

88:22

monthly visitors up to 10,000, then

88:25

20,000, then 50,000. You see my point?

88:28

Um,

88:31

and then eventually monetize it in a

88:34

number of different ways. But we can't

88:36

do that all in one day, right? Like this

88:37

is a slow burning project. So

88:41

that's, you know, route number one.

88:43

Route number two is this experiment pod

88:46

to where we build products every day,

88:49

see which one we want to ship live right

88:51

at the end of the week. Uh, and then I

88:53

want to build a trading bot, which I

88:55

think is more of a short-term

88:57

benefit to the viewer. Um, and I'd love

89:01

to collaborate with somebody experienced

89:03

in trading rather than me building a

89:06

super surface level trading bot. So

89:08

yeah, today we're just talking about the

89:11

idea of making money with Crockbot,

89:12

showing the foundational kind of bricks

89:15

I'm laying to make money down the road

89:18

off of this experimental account. If

89:21

that makes sense. Hope that answers

89:23

answers it, Matios.

89:55

Yeah, I do think uh let what tech stack

89:58

is this site built off of? Let's start

90:00

there. Um because I really love the idea

90:02

of using shadec components. So, if you

90:05

go to shadescenblocks.com,

90:08

you'll find a massive library of cards,

90:13

forms, like everything we need to build

90:18

an awesome website. Let's just put it

90:21

put it at that. And I can get a little

90:22

bit more specific, but maybe our tech

90:26

stack doesn't support those elements.

90:28

I'd be shocked to hear this, but please

90:31

let me know.

90:38

shade CN blocks.

90:41

It misspelled it, didn't it? Let me

90:42

double check that that's the right

90:44

domain.

90:47

Yeah, there we go.

91:29

Hey, let's go.

91:32

I'd love to be the judge of that right

91:34

there.

91:49

Astro 7 static custom CSS no react no

91:52

tailwind. Ah

92:02

well let's uh let's make these changes

92:05

then. Can we still keep everything

92:06

hosted on Cloudflare?

92:20

You also mentioned a faceless YouTube

92:23

channel in some past live stream. Um,

92:26

yeah, I think I know what you're talking

92:28

about, but the generative AI at that

92:30

moment in time, it just it didn't make

92:32

sense to release more slop in such a

92:35

sloppy

92:37

ecosystem. Um,

92:41

might be a good time to revisit that if

92:43

I'm being honest with you.

92:46

The question is, do we focus around

92:48

short form or long form? Feel like long

92:51

form is the move. No.

92:58

So, we started when I started looking at

93:01

it, we were at 11%. We might have

93:04

started at 9%. So, so far in this

93:06

session,

93:09

it's creeping up there. Interesting.

93:13

But I did do a lot of maintenance. We

93:15

did do a lot of prompting, a lot of

93:16

handoff documentation, a lot of deleting

93:19

uh after the handoff docs were ready.

93:21

So, I wonder how much percentage um that

93:25

was responsible for. You know what I

93:27

mean?

93:45

go sheets CN blocks. I want to give it a

93:48

specific um

93:51

nav header here.

94:49

Yeah, I like the idea of using this

94:51

header.

94:52

because it has a really awesome search

94:54

feature built into it. Um, let me know

94:57

if you have any questions.

95:05

And maybe I changed that. Maybe I

95:08

changed that. Let's see. Let's go back

95:10

to nav on.

95:17

For some reason, that is not working.

95:29

Why is that not working?

95:39

And then I think all obviously we should

95:41

have breadcrumb crumbs incorporated uh

95:44

into the nav at the top.

95:56

Yeah, it's crazy. I don't want to

95:57

contribute towards that, right? So, you

96:01

give me some ideas, though. You give me

96:02

some ideas.

96:09

Let me go check something. Maybe we

96:10

build out a team right now for this.

96:15

You just go pull up an old project.

96:28

Where did I put that? That's the

96:30

question though.

96:38

Was it in YouTube ideiation?

96:55

Or did I

97:34

Now, I could have sworn I had a certain

97:36

project somewhere.

97:39

Must have been a different account.

97:42

I can't seem to find it.

97:55

I honestly kind of thought it was a a

97:58

GPT, like a custom GBT that I built out.

98:05

Ah, maybe I just found it.

98:09

Where is all the context related to

98:11

this?

98:13

Uh, review GPT. No, that's gonna review

98:17

GPT.

98:19

Edit.

98:22

Welcome back. Configure.

98:25

Oh, wow. Yeah.

98:32

Oh, yeah. Remember reserving that

98:33

channel tag, huh?

98:40

Huh? Let me uh go full screen.

98:57

Okay, maybe we uh we build out that team

99:00

right now. So, we're going to call this

99:03

uh

99:11

oh, we create a new section from the

99:13

agent. You can't just create a whole new

99:15

section.

99:17

All right, I want to build out a new

99:19

team.

99:21

And this team is going to manage

99:26

a new Faceless YouTube channel, a tag

99:29

that I reserved called ABC Vault.

99:54

All right, I'm going to give you a copy

99:56

paste of some context about the idea

100:00

behind this channel. Keep in mind some

100:02

of this is outdated, so we'll need to

100:04

make some improvements, but I want to

100:05

build a team that complements this.

100:12

All right. And then I'm going to paste

100:14

this in there completely.

100:39

Oh man, time flies, huh?

101:15

All right, they're still building out

101:16

the new version of that.

101:21

We are burning through usage with this

101:23

experiment here.

101:32

Yeah, I'd imagine we want to create a

101:34

faceless team. So

101:37

ideiation

101:42

ideiation then leads to copyright

101:46

slashscripting

101:48

which then leads to voice over which

101:52

then leads to

101:55

video output which then leads to an

101:57

editor

101:59

or a composer so to speak that brings

102:03

everything together. the audio, the

102:06

video, the thumbnails. We need to add

102:08

that into the pipeline as well. Um, but

102:11

I want this working end to end. Of

102:14

course, we're going to take some time to

102:16

build out these specific bots that are

102:19

going to handle this pipeline, but let

102:20

me know if you have any questions. The

102:23

goal here is to automate a faceless

102:25

YouTube channel entirely through our

102:27

Grockbot

102:29

team

102:32

to where one weekly routine, sets in

102:35

motion

102:38

this pipeline

102:40

which ideates scripts, does the voice

102:43

over, creates the video outputs, splices

102:47

them all together, and then leaves me

102:49

with an MP4 that I can upload. Ideally,

102:52

we would want um a bot handling the

102:56

upload process as well, but the only

102:58

part of the blocker I want to be is me

103:01

reviewing the video and letting you know

103:03

thumbs up or thumbs down.

103:20

How funny is this quote from XAI talking

103:22

about why Grockbot is different? Reminds

103:24

me of our last 20 website from forever

103:28

we talked about. Oh yeah, I remember

103:29

what you were talking about. Last 20.

103:34

There's a huge difference between 90%

103:36

done and 100% done. Most AI gets you

103:37

there. Grockbot can finish the swing

103:40

because the work lands where a human

103:42

would put it in the actual tool. Yeah,

103:45

trying to remember. I I remember seeing

103:47

this post specifically.

104:30

Um, interesting. Did he create a new

104:32

team that I just don't see yet? Oh,

104:34

unassigned.

104:36

Yep. He's straight up creating an

104:38

entirely new team.

104:41

So, what we're going to do is add to

104:42

section.

104:45

Faceless

104:46

YouTube

104:50

pod.

104:53

Let's move this down here. I'm just

104:56

going to add Can I select more than one

104:58

agent at a time? Yes. Nice.

105:03

What? Oh, but then when you drag them,

105:06

it doesn't drag them all together.

105:08

That's rude.

105:11

Why are you going to allow us to select

105:13

multiple agents at once but not drag

105:15

them all? It's just rude.

105:18

That's just absolutely rude.

105:38

Yes, you can for sure. Um, there's a

105:41

little caveat there. There's a little

105:43

caveat. So, the bot itself is still

105:46

powered by Grock, but what's happening

105:48

is I installed Claude in the CLI on the

105:51

VM in this particular system. So,

105:54

anytime I want work done, I specify

105:56

which model or sometimes I don't specify

105:58

which model. Uh, but they essentially

106:00

prompt in that team chat what they want

106:03

from Claude. Claude then reports back

106:06

with its findings and then Grock the bot

106:08

itself, uh, whichever bot you're

106:11

powering essentially receives that

106:13

information and then handles it. So,

106:16

there's still that there's there's a

106:18

little asterisk there. On my other

106:20

system, I set up Peekaboo on my Mac to

106:22

where instead of having to install um

106:25

the CLI within the VMs.

106:28

It just prompts my system, right? And I

106:30

didn't have to authenticate anything

106:31

from that route. It was actually super

106:33

easy and convenient to do. Um but then

106:36

again, you're giving Grockbot access via

106:39

Peekaboo to your Mac OS, assuming you're

106:41

using Mac OS. But yeah, little asterisk

106:43

there, but you can get it to work.

106:47

All

106:52

right, we're just going to wait here. I

106:54

do need to

106:56

do something really quickly. So, BRB in

106:58

like two or 3 minutes.

107:03

Maybe we'll do some movie time in the

107:04

meantime.

107:15

And I was about to get up. Let me answer

107:17

this. Yeah, exactly. Exactly. Exactly.

107:20

Apparently though, we need to look into

107:22

usage through our cursor account. But

107:25

there might be another route to doing

107:28

it. Um, I just need to do some research

107:30

before sharing my findings. Shout out

107:32

Anthony aka Pokemon here in chat who

107:36

shared this account. Kind of brought

107:39

this to our attention from Chris

107:41

Mcconey. Apparently, he's finding that

107:43

his cursor account is using different

107:46

models. Now, the question is how's this

107:48

tied to Grockbot, right? U so I need to

107:51

look into this and like why this is

107:53

happening, how it's happening. Um but

107:56

nevertheless, yeah, but that's one

107:59

workaround I found. I did both routes as

108:01

I mentioned. Uh great question though.

108:03

Hope it's helpful.

108:06

Now, we're going to set up some movie

108:07

movie time while we watch these agents

108:09

work. And uh I need to go AFK for a sec

108:17

and like what their margins are. It did

108:19

that on its own. I I was kind of blown

108:22

away and it made these

108:26

reports on each product idea and was

108:28

like here's my recommendation. Always

108:30

ask for recommendations by the way here.

108:32

Like my bulletproof system for testing

108:34

business ideas with Crockbot is I want

108:36

to do this, this, or this. So, I've got

108:39

three ideas. Maybe they're similar,

108:41

maybe they're different. Go research

108:43

each of them. Figure out how I can price

108:46

it, what the cost is going to be, and

108:49

then give me your top three

108:50

recommendations on how to do the damn

108:52

thing. And it just works. Uh you it your

108:56

task should not be open-ended when

108:58

you're doing research or you're just

108:59

going to end up doing research for two

109:01

weeks.

109:02

>> That's fair. Okay. So, you think I mean

109:05

you think Shopify that there's still

109:07

opportunity in in finding products,

109:10

finding audiences for those products and

109:12

then having agent teams manage that

109:15

whole process. That's,

109:18

you know, that's interesting because

109:19

like a, you know, on on this podcast we

109:22

talk a lot about like digital products,

109:24

SAS, vertical, AI,

109:27

um, not so much on the Shopify side. So,

109:30

it it is it's just interesting. Yeah, I

109:34

I'm not like an ecom expert. This is an

109:37

experiment uh I'm working on. But I

109:40

think it's nice to work on something

109:41

physical because the cost of creating

109:44

digital things is going to zero. There's

109:46

no friction there.

109:47

>> Uh actually making something physical uh

109:50

a brand that you can just like build

109:53

over time. I think that I think that's a

109:55

smart thing to do.

109:56

>> And over here I see you have site

109:58

builder stack. What do you mean by that?

110:00

Yeah. So I I just have an agent uh on my

110:04

other account that has this simple stack

110:07

of Cloudflare um the CLI tools. Uh pork

110:11

bun is where I buy domains. So it can

110:12

log into that on its computer and buy a

110:14

domain. Um and then I use this thing

110:17

called Astro for for it's a framework

110:19

for creating static sites and it's

110:22

connected to my GitHub. And so basically

110:27

like I needed I needed a link for people

110:30

to book meetings with me. Uh and I

110:32

didn't want to pay for Calendarly. So I

110:34

said I got it set up on this stack and I

110:37

said go embed my three different Google

110:39

calendars on this and I was literally

110:42

sitting on the phone on the beach and it

110:45

would send me okay here's a mockup of

110:48

this and I would tell it okay make

110:50

another branch and do a mockup in this

110:52

style and it would come back and through

110:54

Cloudflare it would have separate

110:56

developer links that I could review the

110:58

different styles in and then when I said

111:00

okay I like that one it connected it to

111:03

the domain that that I owned. Uh, and it

111:05

went live all from the beach and I set

111:08

that up in two or three hours that that

111:11

that stack. Uh, and I'll probably use it

111:14

for a long time just for simple I need

111:17

something to exist on the internet. It

111:19

doesn't need to be perfect or pretty.

111:20

>> So, is you know is that just like

111:23

internal tools building internal tools

111:26

or is there an opportunity to build some

111:28

of these tools and sell it to other

111:31

businesses?

111:33

Yeah. So, what I would use this for is

111:35

landing pages. So, yeah, it's kind of an

111:38

internal tool. If I want to sell, say, a

111:40

local business on being a sponsor, I

111:42

could just say, "Hey, um, you know, make

111:44

a landing page that speaks directly to

111:46

this customer or directories." Uh, so

111:49

Astro specifically is really good at

111:51

directories. So, if you if you're

111:53

interested in the SEO game,

111:56

just drew through that. This is the

111:58

stack for testing and iterating.

112:01

And I would just have a research agent

112:03

that adds to the directory. So if you

112:05

want to make a directory of pickle ball

112:09

equipment, I would set up a coding agent

112:12

that uses this stack and a research

112:14

agent. And every day the research agent

112:18

investigates a new brand and writes a

112:20

great page on that brand as markdown and

112:24

then it hands it to the engineer who

112:26

posts it. Um, then you would, you know,

112:28

set up an agent that collects referral

112:31

links and does outbound to get to get

112:34

deals once you have the traffic. Uh, and

112:36

you could rinse and repeat that if

112:38

you're interested in doing kind of the

112:40

SEO directory game.

112:42

>> Yeah. So, I've done a few episodes on

112:45

directories in the past. I'll include

112:47

the latest one uh with my friend Frey

112:51

who does a good job of kind of

112:52

explaining why directories matter, how

112:54

it's a good like 10k, 25k, 100k a month

112:57

business. Um you know what is a

112:59

directory? It it's basically just

113:02

curated data on a website. So it's

113:05

you're picking a topic and you're like

113:08

here's the best xyz in this topic. And

113:12

in an AI world, we actually need

113:14

directories more than ever because um

113:18

there's just too much information. So

113:20

when it's curated,

113:22

uh it's helpful. Also, the whole like

113:26

SEO game is changing right now. Like

113:28

what's being, you know, like people are

113:30

going to chat GBT and they're going to

113:32

Google AI overviews

113:34

um to actually get, you know, con to

113:37

consume content and they need platforms

113:39

like that need directories. So I think

113:42

using

113:43

>> all right we are back and you know

113:44

what's funny right as I sat down

113:48

>> and in an AI world u no setup

113:52

>> I've done what is a directory it it's

113:55

basically just curated data on a website

113:59

so it's your just curated data on a

114:02

website

114:04

I don't know if I've defined that

114:08

can you guys define what a database

114:10

website is

114:14

>> and Frey who does a good job of kind of

114:16

explaining why directories matter, how

114:18

it's a good like 10K, 25K, 100K a month

114:21

business.

114:23

>> Let me rephrase that. A directory

114:25

website.

114:28

>> Um, you know what is a directory? It's

114:31

basically just curated data on a

114:34

website. So, it's you're picking a topic

114:37

and you're like,

115:03

All right. to simplify it, right? A

115:06

directory website is just basically a

115:09

website with curated data about a

115:11

specific topic or niche. In our

115:13

particular case for this project,

115:15

especially AI agents.construction,

115:18

it says it in the name of the domain,

115:22

right? I know we have a very specific

115:25

target audience, but I'm wondering if we

115:27

we don't broaden this to essentially be

115:30

the go-to directory for humans and

115:33

agents to learn how to build agents or

115:36

bots for

115:39

every job role in whether it's a GC

115:42

environment or construction companies of

115:44

all sizes, small, medium, large,

115:46

enterprise,

115:49

right? The goal is to make this

115:50

directory website, this domain,

115:53

uh, so recognizable and like so insanely

115:58

used to where then we can start to

116:00

monetize off of it.

116:05

>> Here's the best XYZ in this.

116:07

>> Hey, we'll see you in a year if this

116:09

works. I'm going to keep this running

116:10

cuz, you know, I'm doing work with the

116:13

construction industry, with a number of

116:15

different companies, uh, building agents

116:17

and managing agents, uh, to where I'm

116:20

not going anywhere out of this industry.

116:21

So, that's why I chose this industry for

116:23

this directory website. Just a heads up,

116:25

I didn't just randomly select an

116:27

industry

116:28

topic. And in an AI world, we actually

116:32

need directories more than ever because

116:35

um, there's just too much information.

116:38

So when it's curated,

116:40

uh, it's helpful. Also, the whole like

116:44

SEO game is changing right now. Like

116:46

what's being, you know, like people are

116:48

going to chat GBT and they're going to

116:49

Google AI overviews.

116:51

>> What is Pork Bun, though? I've never

116:53

heard of Pork Bun. It sounds funny. Pork

116:56

Bun.

116:58

Oh, it's a domain site.

117:01

Out of all the domains, Pork Bun.

117:04

Really?

117:07

there's just too much information. So

117:09

when it's curated,

117:11

uh

117:13

this right here, I built a directory

117:14

site. My site is backline local. Love

117:18

idea of it for restaurant repair. Right

117:20

up your alley, right? Like so like Yeah.

117:25

Oh, is it really? Is it like the new

117:29

namecheep that's super that's still

117:31

super niche and small?

117:35

It's helpful. Also, the whole like SEO

117:38

game is changing right now. Like what's

117:41

being, you know, like people are going

117:42

to chat GBT and they're going to Google

117:44

AI overviews

117:46

um to actually get, you know, con to

117:49

consume content and they need platforms

117:51

like that need directories. So, I think

117:54

using Grockbot as the basis for your

117:58

directory is absolutely brilliant. And

118:00

the be beautiful thing about a directory

118:02

is it's a once you build a strong

118:05

directory, it's a foundation. And yeah,

118:08

yeah,

118:13

yeah,

118:14

yeah. I know. I agree. I agree. It's uh

118:19

I'm reminded of it as well when I have

118:21

friends and family calling me for advice

118:23

and and questions. Um,

118:27

while it's easy,

118:29

you still have to work your way through

118:32

understanding how to communicate.

118:34

Um,

118:38

yeah, but you got to also understand

118:40

like listen, got a lot of respect for

118:42

Greg. Um, but his videos are a specific

118:46

are targeting a specific type of person

118:49

who would rather consume the content and

118:52

be wowed by the idea but not actually

118:55

take action on the idea. Sorry, Greg,

118:57

but that's like, you know, a lot of the

118:59

guests that you have on, they talk about

119:03

this idea and they get super hyped, but

119:06

they don't really give specifics.

119:09

They're super broad, right? They're

119:11

super like imaginative and like wow

119:14

factor, but wow factor does not

119:17

translate into execution, right? So like

119:20

having a guest on who shows the

119:22

execution of a database site, who shows

119:25

the proof that they're making a h

119:27

100,000 a month off of a database

119:29

website

119:31

or a directory, let me rephrase that.

119:32

Directory website, that's a whole

119:34

different realm, right? But that's not

119:37

who he's targeting, right? And you got

119:39

to understand that as well. And

119:41

>> you're getting thousands of visits a

119:44

month, let's say. And then you could

119:46

say, you know what, I want to build a

119:48

Shopify store. Actually, you know what?

119:49

I want to build a newsletter. Oh, you

119:50

know what? I actually want to build a

119:51

SAS. Or, you know,

119:53

>> I heard Pork Bun, but I've been doing it

119:56

through Versa. What?

120:02

>> But I want to build a mobile app. So,

120:03

it's actually a really good place for

120:05

people to start.

120:07

>> Yeah. And the cool thing about a

120:09

directory is

120:10

>> big fan of Greg. This was the first

120:12

video that really got me. Even this

120:14

ended up being just a bagel shop emailed

120:16

him and his Grockbot saw email and said,

120:19

"Hey, they want to be

120:21

Yeah. Yeah. Yeah. Yeah.

120:22

>> It ties in with this. Not a local

120:25

newsletter, but niche newsletter." So,

120:27

say we're doing that board game idea.

120:29

You could have it build out like board

120:32

games that are good for for parties

120:33

could be your niche directory. And you

120:35

could have a newsletter that talks about

120:37

that too. Uh you could have this on a

120:39

sub subdomain or a different domain than

120:41

your newsletter and uh they tie into

120:44

each other. So you can monetize uh by

120:46

linking people to the directory and have

120:49

a referral link there. Uh you could do

120:52

those warm leads from those referral

120:54

links could become sponsors in your

120:56

actual newsletter. Uh the I think the

121:00

the lowest barrier to entry with

121:03

Crockbot to start a business is probably

121:06

newsletter and then directory. I think

121:08

that stack goes together really nice.

121:10

>> Beautiful. All right. I think uh I think

121:13

we covered what we wanted to cover. Is

121:15

there anything anything else you want to

121:17

leave people with?

121:19

>> Yeah. Um I would do this try this for a

121:22

month. set up Grockbot for one project

121:26

and uh run it like a real team like we

121:29

went over and

121:32

let us like let me know like tweet at me

121:34

like did you make progress doing this? I

121:36

I like this is your sign to pick

121:38

something that you've been thinking

121:40

about and pursue it aggressively versus

121:43

trying a new tool and researching ideas

121:46

for four weeks. Uh so do this. Build a

121:49

Spotify or newsletter or a directory. uh

121:53

with Rockbot and I think you'll see what

121:56

I mean. This I haven't been excited

121:57

about a tool like this um

122:00

maybe like all year. Maybe this is my

122:02

favorite this year. Um

122:05

yeah, so tweet at me.

122:09

>> What's so interesting about this tool is

122:11

we've seen

122:15

>> Vendor Deso.

122:18

Yeah. Yeah. like the dream uh the sales

122:21

the seller of dreams, right?

122:25

The seller of dreams. Yeah, I could see

122:28

that. Yeah. Yeah.

122:31

>> A glimpse of this and other tools.

122:33

>> And don't get me wrong, like great show.

122:36

I get a lot of ideas from it, but you

122:38

also have to understand,

122:40

you know, it's a certain format for a

122:43

certain type of individual and it only

122:45

gives you c a certain amount of

122:47

information. Um,

122:50

it does not mean that that is tried and

122:53

true. It's an idea.

122:56

There's generally never any proof shown.

122:58

Rarely any proof shown. Um, depending on

123:01

the guest, I think it's it's all about

123:04

how honest the guest is and how much

123:05

they're willing to share. Um,

123:08

yeah, I'm sure Greg's regretted having

123:11

certain people on the show, you know.

123:13

Um, but that that's not his concern,

123:16

right? He he's he's making videos for a

123:19

specific target audience. So, got to

123:22

understand that.

123:30

One positive I will say about this video

123:32

is the ad. You make the platform calls,

123:35

not the agent.

123:38

What ad? I didn't see the ad. I must

123:40

have skipped through it.

123:50

Now where is is it unassigned? No, we

123:53

have everybody we need here. Question is

124:01

all right. I want to visually see this

124:02

pipeline that you guys have created. I

124:04

also want to visually see a rich table

124:07

here just showing me what's running,

124:10

whether it's daily, weekly, what day of

124:12

the week, uh each step of the pipeline,

124:15

who's handling what, and when I get

124:16

pinged to either green light it or red

124:19

light it.

124:29

Listen, I'll run an experiment for a

124:31

month. I'll find an API key. I hate it.

124:33

I hate the idea of it, but let's do it.

124:52

Also, I think, correct me if I'm wrong,

124:54

but would it be helpful if I send you a

124:56

few YouTube channels I think we should

124:59

draw inspiration from as far as quality

125:02

of output, narration style, education

125:05

style. Um,

125:08

because we're going to want to fine-tune

125:10

our process first. We're just seeing

125:12

what you come up with based off of the

125:14

minimal context I've given you. Um, but

125:17

I do want to fine-tune our process. I do

125:19

want to fine-tune the pipeline here.

125:26

Look at this. Great. So, Tuesday 11:00

125:29

a.m. Vault case voice script that sends

125:32

two ways. Frame pack, tapewave, cut mp4,

125:37

YouTube thumbnails,

125:39

Elon upload. Interesting. Interesting

125:41

how Elon handles the uploading. I feel

125:44

like I'm we're going to want an

125:45

uploader,

125:47

right? I don't know if I want to give

125:49

you uh Elon that responsibility.

125:54

So Elon fires it.

125:59

Okay. Okay. Block, no voice, block, no

126:02

stitcher. Okay, we got to fix those. So,

126:08

wait, are you blocked by me?

126:15

All right, let me uh go sign up for an

126:18

11 Labs account real quick.

126:21

11 Labs

126:23

developers

126:26

API keys.

126:31

What account am I in?

126:34

Oh, I have some credits in here somehow.

126:37

What is going What? Since when do I have

126:42

credits?

126:46

Do I use 11 creative, 11 agents, or 11

126:49

API? I think I know the answer, but

126:51

please let me know.

127:02

Um,

127:05

let's see. Settings.

127:09

Interesting.

127:12

I guess I'll keep that account live for

127:13

now.

127:38

Does it? H can I see this? Uh

127:43

these two visuals, the pipeline above

127:46

and the table imagined

127:48

or generated via Grock imagine please.

127:59

11 API. Got it. Create

128:04

key.

128:06

So

128:08

baseless

128:09

YouTube

128:11

experiment expire after call it 30-day

128:15

experiment. Create key.

128:20

Copy to clipboard.

128:28

It should detect it. So, I'm just going

128:30

to let it do its thing.

128:33

All right.

128:47

Yeah. So, we started off I I drew

128:50

inspiration from this Greg Eisenberg

128:52

video. So, we did a little bit of movie

128:54

time. Uh, but this video talks about

128:56

creating a directory website. Um so I

129:00

have built a directory team. I don't

129:03

know why I called it database. Let's

129:04

called it a directory

129:06

factory where so we primarily service

129:11

the construction industry right now. Um

129:15

so I have retainers in place, contracts

129:16

in place that

129:18

um primarily create, manage um and

129:23

maintain AI agent systems for small to

129:26

midsize construction based companies and

129:28

GC's for that matter. Um so I chose our

129:31

target audience.

129:34

I have a certain uh domain that I squat

129:36

on squatted on a while ago called AI

129:39

agents.construction.

129:41

So, we've built out this team to

129:43

automate the process of building out

129:45

that directory for us based off of that

129:47

video. So,

129:50

how is YouTube guidelines reacting to AI

129:53

videos or shorts? You know, they're

129:55

supportive of it, but they're also

129:57

making changes for, you know, how they

130:01

manage AI slop in general. I'm one thing

130:04

I don't want to do is contribute to AI

130:06

slop. Let's be very, very clear. Um,

130:11

but yeah, that's I mean that's ever

130:12

evolving, right? They allow it, but I'm

130:16

not going to just blindly upload videos,

130:18

right? I want to be the blocker. I want

130:20

to be that um that final handoff before

130:24

I say upload, right? Uh but it's another

130:26

idea that we have. So, we've built out a

130:28

team to complement that, a faceless

130:30

YouTube pod. And then we built out

130:32

another team, an experiment pod that

130:34

automatically builds us a product based

130:36

off of uh certain pain points they

130:39

discover across socials using the last

130:41

30 days skill. So yeah, as far as the

130:44

the title of this live stream, I was

130:46

very clear if you watched the very

130:48

beginning about it uh beginning of the

130:50

live stream. This is not a promise that

130:52

I'm going to teach you how to make x

130:54

amount of dollars. We're just

130:56

experimenting from the perspective of

130:58

making money, right? So, it's a longer

131:00

term gonger term game with this. And

131:04

then I think next week we're going to

131:06

build a trading bot live. I think that's

131:08

going to be the move for next week. Uh

131:10

because I want to bring in some expert

131:12

traders to do that. Uh because for me to

131:15

build a trading bot, not having

131:18

foundational knowledge of if then

131:20

statements, what to do when, right? Like

131:24

how to trade based off of certain

131:26

variables and metrics. um I would not be

131:30

as effective making a trading bot as

131:32

somebody in the industry, right? So stay

131:35

tuned. Next week I'm going to plan to to

131:37

build a trading bot live. I'm not sure

131:39

which day of the week yet, but we'll do

131:41

it next week.

131:43

So for fun, I'm going to create an agent

131:45

to do tracker of Greg's videos. If his

131:47

guest wears a hat, he does.

131:50

If they don't, he doesn't. I want to see

131:52

percentage. Oh, really? That's funny. I

131:58

never noticed that. That is funny.

132:03

Looks like they're not done with the

132:05

work yet here.

132:14

All right. Where What's the status on

132:15

the rebuild?

132:20

I'm just getting impatient here.

132:33

Oh, wow. Vault just came up with a ABC

132:37

vault case. Oh, I love that structure of

132:41

how it's doing it.

132:44

Solid.

132:50

Key is scoped. Not because it's wrong.

132:53

Key is set, but it's locked down.

132:57

Edit that key and turn on text to speech

133:00

and voices. Text to speech and voices.

133:02

Let's see.

133:05

Edit text to speech

133:09

and voices

133:12

read.

133:14

Same key is fine.

133:18

Done.

133:23

Okay, great.

133:26

Awesome.

133:45

Now, when did I reserve this channel?

133:46

I'm very curious. Let's go.

133:50

Create view channel.

133:54

Did I really do this in December of

133:56

2024?

133:59

Huh?

134:10

Also, Elon, shouldn't we have an

134:12

uploader

134:15

or do you really want to take on the

134:17

responsibilities? Is that really your

134:18

role to upload as well?

134:22

Here's our YouTube URL that we reserved

134:24

this domain a while back.

134:37

Okay.

134:44

So, since I started looking

134:48

when I first toggled this, this was 11%

134:50

at the start of the stream. So, we've

134:51

we've burned through 10% already.

134:57

Now, how much of it? I'm going to

134:59

monitor this. So, chances are I'm going

135:02

to record a video a little bit later

135:04

today. Still not sure the topic yet. Um,

135:08

and then I'm just going to let this run

135:09

day by day and I'm going to check in

135:11

every 24 hours to see what that

135:12

percent's at.

135:15

I guess instead

135:22

Oh, yeah. Look at that. It did a pretty

135:25

good job, Derek. Good idea.

135:29

Good idea.

135:32

All right. So,

135:36

what I'd like to also do is set a daily

135:41

checker for usage. Can you do that? Can

135:44

you see exactly how much of our weekly

135:46

usage we're using every 24 hours so that

135:50

I can start to manage and and just keep

135:52

an eye keep tabs on how much

135:56

usage we're burning through based off of

136:00

all of the different pods and teams

136:02

we've created and routines that we've

136:04

created.

136:12

Let's see what he says here.

136:22

I appreciate you. Thank you. Thank you.

136:26

Thank you for the uh the donation there

136:28

as well.

136:31

I'm running a manufacturer

136:36

I'm running a manufacturing theme in

136:38

India building been building systems

136:41

with AI on Google sheet

136:44

ordering monitoring stuff but I've

136:45

learned the system works only if the

136:46

people who are using it do it properly.

136:48

Okay. Okay. So,

136:52

are you looking into using bots to help

136:54

automate the process or maybe creating a

136:56

different solution uh to make sure that

136:59

it's easier for the people to use the

137:02

system properly? Is that what I'm

137:04

understanding?

137:20

Kayleie burn check. I'll find the

137:22

reason.

137:30

Oh, nice.

137:38

Very nice.

137:46

Am I going to be disappointed? Okay,

137:50

looks better.

137:54

But it's just the header. Like what

137:56

happened down here?

138:03

Okay,

138:05

there needs to be like

138:08

feel like we need to get even more clear

138:10

here.

138:29

Oh, you already know that instantly

138:30

happened. You must be delayed for a

138:33

manufacturing theme. Where do you feel

138:36

Crockpot can come in handy? ERP is

138:39

business central. So MCP I can give

138:40

Grockbot Eagles view on all ops. Yeah,

138:44

you know, I like only speaking on things

138:47

I truly know and have experience in. So

138:50

as far as how it could come in handy in

138:54

your industry,

138:56

let me think on that right because I

139:00

don't have a manufacturing background.

139:02

So what what is in your brain is

139:06

different than is what in my brain like

139:07

as far as context related to it. Um

139:14

now what size facilities are you more

139:17

talking about from the

139:20

um

139:21

how do I how do I word this? So like

139:23

from the perspective of the staffing

139:25

that's running this manufacturing

139:28

facility whether it's small, medium or

139:30

or rightme is um

139:34

forgive me I don't know that

139:35

abbreviation off the top of my head but

139:36

I'd imagine it means small to mediumsiz

139:38

enterprises

139:42

just knowing what I know about SMB

139:44

meaning small to midsize companies

139:46

businesses.

139:50

So your ERP is business central. So

139:53

Sophia MCP I can go with GRA.

139:56

So you answered it right there. Forgive

139:58

me. Um

140:02

yeah, I would honestly, this is going to

140:04

be super surface level, but start with

140:05

one responsibility. Start with

140:10

using Grockbot a bot or number of bots

140:13

from the perspective of what's a tedious

140:15

repeatable task that you do that you

140:17

think is a waste of your time. automate

140:20

that with a bot or two or three and then

140:23

rinse and repeat. You know what I mean?

140:25

I think it's very easy to get to fall

140:29

into the trap of

140:34

wanting to do too much. Visualizing this

140:37

big grandiose vision of how it can

140:39

automate your life and your company and

140:41

and and your business or the company you

140:43

work for, but not taking the steps

140:46

needed to like build up to that. I'm I'm

140:48

a big advocate for start simple, start

140:51

small, start with one thing. And the

140:54

best thing to start with is is something

140:55

you know well, right? We all know what

140:58

we waste our time with. We all know

141:00

tedious parts of our dayto-day, whether

141:02

it's personal or professional or working

141:05

for another company of man, I really

141:07

wish I didn't have to do this task every

141:08

single day or every single week or every

141:10

single month. automate that thing first

141:13

and then build from there. And then

141:14

you'll start to get little unlocks of

141:16

more ideas of how to utilize bots or

141:19

agents for that matter um for your

141:22

specific industry. And and um yeah, hope

141:25

that's helpful. Hope that's helpful.

141:30

All right, so they're still working.

141:33

This is great. We're burning through

141:35

usage. We'll see how this ma maintains.

141:38

Uh pulling the episode now.

141:42

Oh, interesting. Claude is doing some

141:44

work.

141:51

Okay. So,

141:54

I asked you to update the text stack of

141:56

this directory website AI

141:58

agents.construction

142:00

and it looks like you half did the job.

142:02

The nav header is fixed. Great. But the

142:09

website itself

142:12

was not touched, right? All the contents

142:14

also that nav header should be sticky.

142:16

So as you scroll, it should still be

142:19

visible. I think the only scenario where

142:22

it's not visible is an article. an

142:25

article for a specific job or a specific

142:27

task to where as you scroll then it

142:31

shows those breadcrumbs of sections one,

142:33

two, three, and four that stay sticky

142:36

for a user to easily navigate. On second

142:39

thought, keep the nav header and

142:41

breadcrumb portion of the page sticky,

142:45

but then as a user scrolls, there should

142:48

be a

142:53

I don't know on the left or right column

142:55

1/4 column um essentially a site map or

143:00

a glossery for that specific article for

143:05

that specific knowledge resource.

143:08

allowing a user to easily click between

143:11

job and this is just an example on one

143:13

of the pages between 1 2 3 4 5 6 right

143:17

um

143:22

I think that's a great starting All

143:23

right.

143:42

Yo, what's up, Lego

143:45

geeks? Nice. Nice. Yes, but I don't It's

143:49

not fully automated, right? But yes, a

143:51

lot of what I do is automated um to

143:55

where I can just focus on live streaming

143:57

and making videos. I still record the

143:59

videos. I still edit the videos cuz I

144:01

enjoy that process. But a lot of those

144:03

tedious parts of my dayto-day are

144:05

automated primarily with OpenClaw and

144:08

now a touch of Grockbot. Uh Buzz I use

144:11

from a collaborative perspective. I

144:12

don't really automate too much through

144:14

it. Um, and then Hermes, I build out

144:17

Hermes agent systems for uh companies.

144:20

So

144:22

I I don't only build them, I maintain

144:24

them as well. I manage them. Um, there's

144:27

a lot of different options from that

144:28

perspective.

144:30

Fix map. You know, I don't have any

144:33

input on that.

144:35

I wish I had more insight into that. I

144:38

can look into it, but I I have nothing

144:40

for you at this moment.

144:45

Wish I could be more helpful.

144:51

Hey, I'm happy to help. I'm happy to

144:53

help. Do you have any agents or bots

144:55

running through X or other channels

144:57

getting you news or updates for examples

144:59

on AI? Yes. Yes. So, my OpenClaw agent,

145:03

Muddy, my OG, I get a morning brief

145:06

every morning. There's an industry pulse

145:08

checker. I also created something

145:10

similar but specifically targeting X,

145:13

what's trending on X. And then based off

145:15

of what's trending on X, I have the last

145:18

30 days skill run uh based off of that

145:21

topic or, you know, one of the topics it

145:24

chooses from. And then it cross

145:26

references with Google Trends, YouTube

145:29

search data, and then it tries to find a

145:32

nice middle ground to then feed me video

145:35

ideas based off of what are people

145:37

talking about X, is there a similar like

145:41

crossover on YouTube and then what can

145:43

we find in the middle to create content

145:45

around, right? So, I have a lot of

145:48

different agents from that perspective,

145:49

but I really only rely on two for, you

145:52

know, reading morning news. And then, of

145:54

course, I still scroll X on on certain

145:58

time blocks of the day just to see what

145:59

people are talking about, just for my

146:01

own eyes, my own, you know, cuz I don't

146:02

like automating the posting of content.

146:05

I'm still that final human that presses

146:08

the button always. Uh, except for

146:11

updating certain older videos. I have

146:14

agents doing all that heavy lifting for

146:16

me. Uh for the first few weeks though, I

146:19

did monitor their outputs to say, you

146:22

know, save it as a draft first so that I

146:24

could see how you how you write and then

146:26

I make my edits. And then at the end of

146:28

each day, there was an automation to

146:31

build a voice profile based off of the

146:35

differences of what they wrote versus

146:36

the differences of what I edited before

146:39

I hit post. That way, the model or the

146:42

agent itself will train itself with

146:44

time, right? Based off of you just being

146:47

a little patient and not letting it post

146:49

on your behalf blindly. Uh cuz I'm

146:52

really against contributing to this

146:53

world of AI slop personally. So, I'm

146:56

trying to to make sure that everything I

146:58

put out there has my touch, right, and

147:01

not just blindly get sent.

147:06

How bad are usage limits? I mean,

147:08

they're bad if you just work non-stop

147:10

out of it. They're pretty bad. Um, I

147:13

hope we get some more clarity on that

147:16

front

147:19

as far as usage is concerned. And I

147:20

think, you know, they just handed out a

147:22

reset what, two days ago, if I'm not

147:25

mistaken.

147:28

So, but it's not that bad. Not, you

147:31

know, it's bad, but it's also not that

147:32

bad. It's it's all a you know from the

147:35

perspective of

147:38

if you know what you want to accomplish

147:40

with Grockbot and your bots and

147:42

automations then I think it's actually

147:44

pretty efficient if you just set things

147:46

up on a reoccurring basis to do the

147:49

work. Uh but if you're experimenting

147:52

building teams trashing teams setting

147:54

automations realizing that that's not

147:56

the move and trashing those that's when

147:59

the usage burns up. Also, if you have it

148:02

scanned thousands of emails, that's when

148:05

usage burns up. I've personally

148:07

experimented with all this. So, but if

148:10

you come into a Grockbot setup knowing

148:13

what you want to configure, knowing that

148:15

the task is simple and you're just

148:16

setting up one at a time, usage isn't

148:18

that bad. Uh, but keep in mind, I am on

148:21

the maxed out accounts, so I'm not sure

148:23

how it is on the 20 or $60 a month

148:25

accounts.

148:28

Yeah. Yeah. Yeah. Exactly. And they

148:29

still do. They still do. They still do.

148:32

I still read that brief every single

148:34

morning. It's awesome. It's tried and

148:37

true.

148:39

Yeah, reach out. I'm not on LinkedIn

148:40

often, but uh when I am, I'll make sure

148:42

to uh touch base on there.

148:48

Yeah, my first focus is

148:52

getting more consistent on X and then

148:54

we're going to get a little bit more

148:55

consistent on um LinkedIn. All in due

148:57

time. I'm not rushing, right? I'm not

148:59

losing over sleep into, you know, being

149:03

worried that I'm not being active on

149:04

everything all the time. You know, my

149:06

main focus is YouTube and I'm slowly

149:08

expanding out, right?

149:11

Hey, let's go, Maestro.

149:25

Okay.

149:29

Yeah, I think we need to have, you know,

149:31

maybe one or two different template

149:32

styles of

149:38

the posts themselves cuz I don't get me

149:40

wrong, I love the way that it's just

149:42

super informative here. I feel like this

149:44

is while it's very basic from a UX

149:48

perspective or a UI perspective, it is a

149:51

great user experience depending on the

149:52

target audience, but then it's also a

149:54

great agent experience, which I believe

149:56

is being dubbed AX. Um, I feel like it's

150:00

very easy for an agent to dissect all

150:02

this information and create whether

150:05

an agent, a bot, a team, um, an

150:08

automation based off of,

150:12

you know, each one of these roles. Um,

150:16

so I great job on that front. I'm just

150:18

wondering if we can't just slightly

150:20

improve uh, based off of this new tech

150:22

stack.

150:30

I'm very curious to see what they're

150:32

doing. Uh,

150:36

never ask me. I should never be a

150:39

blocker. Just prioritize based off of

150:41

what came in first.

151:21

Okay.

151:59

Damn, it's already 11:30. I may have to

152:01

uh

152:04

we're going to have to switch gears here

152:06

pretty soon in our day.

152:08

My coalition is working better than

152:10

ever. Got Fable to tweak their pipeline.

152:13

Nice. Nice. I can do my best. I can do

152:16

my best.

152:20

Check the uh YouTube uh the YouTube link

152:24

takes you to our website which has our

152:25

our LinkedIn uh links on there.

152:27

Everything you need is through that

152:28

link. I'm launching a bamboo socks

152:31

company. What do you feel about the

152:32

current AI proudness? It's worth

152:34

automation, for example, driving demand,

152:36

marketing, etc. H

152:51

I'm thinking right like like there's a

152:54

lot of different routes you can take,

152:56

but I'm going to go back to my original

152:57

comment. Start simple, right? I would

153:00

first maybe if you're launching a bamboo

153:04

socks company, I'm sure there's some fun

153:06

facts about bamboo that the average

153:07

person doesn't know about, right? So,

153:09

getting and this is going to trigger

153:12

other events, but getting uh a deep dive

153:16

into bamboo

153:18

every single day, right? Maybe new

153:20

things you didn't know about bamboo uh

153:22

and then creating some type of

153:24

educational style newsletter about

153:26

bamboo. Obviously, the goal is to

153:29

promote your socks company, but make it

153:32

focused around the the material itself

153:35

first. Build an audience through that.

153:39

And

153:41

I mean, then you get a like endless

153:44

content ideas from creating the

153:46

automation, doing research to anything

153:49

bamboo related, whether it's bamboo in

153:51

the environment or bamboo in this

153:52

country, bamboo in that country, right?

153:55

creating a directory, so to speak, of

153:58

all the Bamboo alpha worldwide based off

154:00

continent, uh, based off of fun use

154:03

cases, based off of DIY hacks, right?

154:07

Um, just creating a content engine

154:09

around bamboo, uh, and then just start

154:12

automating little bits and pieces of

154:14

that that pipeline.

154:17

That's probably what I would do. Bamboo

154:19

socks does sound wild. Does sound wild.

154:30

Yeah. Welcome back, Maestro.

154:34

Welcome back, Maestro.

154:36

Yeah, if you just go to our website and

154:38

scroll down to the footer, you'll find

154:40

it. It's either in the header or the

154:42

footer. It's there for sure.

154:46

Drop the bamboo alpha. I know, right? I

154:48

know.

154:50

Hey, but I hate to uh abruptly kind of

154:53

end things. I do have to switch gears

154:55

here in about 20 minutes and um I want

154:58

to prep for this call. So,

155:01

so yeah, I think we're going to end it

155:03

here. I think it's been a fun one. Today

155:04

was just all about experimenting with

155:06

Grockbot. Uh I'm also going to have to

155:08

record a video a little bit later as I'm

155:11

not sure I'm going to use two of the

155:13

videos, two of the other videos I

155:14

recorded this week. So stay tuned. I'm

155:17

either going to drop it today or

155:19

tomorrow.

155:20

Uh, but I just want to thank everybody.

155:22

Thank everybody for hanging out with me

155:24

on your Friday. Uh, wherever you are,

155:27

wishing you a good morning, good

155:28

afternoon, and good evening. I just want

155:29

to take time to say thank you.

155:35

No, you have to go to our website. So,

155:37

go to clearma.ai.

155:44

The difference is quality in normal

155:46

cotton polyester bamboo is the

155:48

difference. Okay. Okay. Yeah, I'll have

155:51

to try it out myself, too. But hey,

155:54

thank you, Maestro. Thank you, Lego.

155:56

Thank you, Sar Sarvesh. I I hope I'm

156:00

pronouncing that correctly. Thank you,

156:03

Derek. Thank you, Alexa

156:07

or Alex A. I can't tell. Uh, thank you

156:12

for the kind comment. Really appreciate

156:14

the donation.

156:16

Abishek. Abishek. I hope I'm I'm hope

156:19

I'm not butchering this. Um, thank you,

156:22

Matos.

156:24

Thank you, Nikolai.

156:26

Thank you, Pokemon.

156:28

And I can't keep scrolling, so I can't

156:30

uh can't thank anybody else. But yeah,

156:32

just want to say thank you. Hope

156:34

everybody has a great weekend. Don't

156:35

forget to go touch some grass.

156:37

Disconnect. Uh, in a world of AI and

156:40

tech, it's nice to get away from the

156:42

computer screen every now and again. You

156:44

know what I mean? Um,

156:47

and yeah, we'll be back at it Monday,

156:49

same time. Until then, happy building.

156:51

Cheers.

156:54

Hey, happy to help. Thanks for being

156:56

here.

Interactive Summary

In this live stream, the host explores practical strategies for monetizing AI-powered workflows using Grockbot, moving away from simple automations to building business-focused systems. He demonstrates the creation of a 'database factory' for the construction industry and discusses setting up pipelines for faceless YouTube channels. Throughout the session, he emphasizes the importance of starting with simple, specific tasks rather than overly complex setups, while also managing AI usage costs and maintaining a human-in-the-loop approach to content production.

Suggested questions

3 ready-made prompts