Let's generate $$$ with Grok Bot (LIVE: Watch me work)
2857 segments
All right, good morning, good afternoon,
and good evening. Welcome back to
another Clear Mud live stream. Today, I
want to focus I need to stop hitting
that button. I want to focus today's
episode around how people are making
money with Grockbot. Uh we are going to
try a few of these tactics ourselves.
And who knows, maybe we come up with
some brilliant idea to make some money
on the side while we're learning, while
we're experimenting. Um, so let me first
this Grockbot system that I set up
actually in a video just a few days ago.
I'm turning into our team bot. Oh, I
need to
I'm turning into our team bot. So
instead, I want to sign out to my other
account and I want to clean up that
other account and use that other account
for experiments for these live streams.
So, let's go ahead and log into that
other account. Give me one second. And
we're going to first start off with
start off with some cleanup here. And
then we'll uh still trying to log me
back in. It's kind I wish there was an
easy way to toggle between accounts.
That's a bummer. So, I got to log out of
that other cursor account and sign back
in.
Let's reopen the link.
Oh, I hit it preemptively. Now continue.
Sign in.
All right. Open Grogbot. So, this is our
original one. I need to finish migrating
the clear mud bots, but we can keep the
chief of staff. We can keep these chiefs
here. As you can see, I started
experimenting with a money pod and a
database factory as well as an
experiment team. So, we're going to keep
those, but we're going to delete
everything else because I've already
migrated the X team. I've already
migrated the housekeeping. Um,
keeper was actually something fun. I was
joined my first dynasty fantasy football
league and uh I thought let me ask some
questions.
And I don't know how I feel about it,
but uh let's let's do some cleanup.
All right, so let's do some cleanup. Go
ahead and delete the X team. That's
successfully migrated. Uh delete
housekeeping.
And we're going to keep the database
factory. We're going to keep the money
pod. We're going to keep the experiment
uh team or section rather. And
and then we're going to keep Codex and
Claude. But everything else,
oh, prep me a handoff markdown file for
the clear mud team. Uh, as we are going
to finish migrating that over. You can
actually delete partnerships that's
already migrated. Uh, but water, mud,
and sky, we still need to hand off that
agent data, that context, those
instructions, how those operate, uh,
into a markdown file. And then once
you've created that, go ahead and
cleanse this system, please.
All right, just starting off with
sending off a big prompt here.
I'm wondering if we should change our
chief of staff for this. Uh, ah, we'll
keep it Elon Musk.
Yo, what's up, Franc? How we doing
today?
There we go. I should probably though
move this video for Grockbot because it
covers feel like an important part of
the app.
So, let's grab that unlock. Move that
over to the right.
And perfect.
Keep that over there.
How's everyone doing today though
otherwise? And what's up with that
background music? Um, but yeah, what
inspired me, if I wasn't clear at the
start of this live stream is I was
watching this one here from 6 days ago
from Greg Eisenberg, and he's had this
guy on the on the show before. This guy
makes money off of creating database
websites. Database websites work like
this. You provide value. You create a
database and you focus on getting as
many visitors as possible. Then you
monetize that traffic, right? you
monetize that traffic in the form of
promoting certain tools,
whether brands pay for you or not.
Essentially, ad space. Um, there's
there's more to it than that, but that's
just a quick TLDDR of what a database
factory or a database website is. Um,
and this sent me down a rabbit hole of
what if we create our own team to do
that. So, today's episode is going to be
all focused on that. It's not going to
be focused on how I'm using it for clear
mud, how I'm using it for the team. Um,
we'll do another video about that. But
now, as you can see, I have two separate
systems. One for clear mud and team, and
then one for experiments and
demonstrations and just random personal
bot use. Um, that's how I'm kind of
separating things now.
Yo, what's up, Mark? Nice. Nice. Are you
just dipping your toes? Have you created
your first uh your first bot yet?
Listen, what one tip I have for
beginners, don't worry about a chief of
staff. When I was first playing with
Grockbot, I was just creating bots. I
was creating bots, having them do
things, and just learning. That's the
best way to start. Um, it's it can get
very
distracting if you try to follow
somebody else's advice too soon before
you find a way to utilize it best for
yourself, your life, your business,
whatever it is you want to uh build out
your Grockbot team or system for. uh
that's unique to you, right? I think
getting ideas from other people is very
important. Um it's amazing. Uh somebody
in our community was talking about this,
but since January, since my OpenClaw
days, right? And by the way, I still use
OpenClaw every day. I never liked having
individual channels to chat with all my
agents. I just wanted one executive
assistant who handles everything else.
But back in those early days, everyone
was creating these mechanisms to
communicate with each individual agent.
And it's just like it it made no sense
to me. Now with Grockbot, it it seems
like everyone's now routing through this
one agent does all. It's it's funny.
It's just funny.
I can't delete seats. All right.
Cannot delete rows either. Interesting.
So, we'll go ahead and delete it
ourselves.
Let's delete,
delete, delete,
delete.
Yo, what's up, Derek? How we doing
today?
All
right, that hopefully
hard to refresh this.
We can delete housekeeping as well.
And I just wanted to make sure add
housekeeping to the list. I can't
remember if we created a handoff doc for
the other system. Um, I think we did,
but just double check before I delete
that one.
I've already deleted X.
And then we can delete partnerships
because we already took care of that.
All right.
Wait, are we not live? Oh, we are. Woo.
I I was previewing my YouTube video on
uh another tab and it just showed the
thumbnail, but it showed 12 watching now
and it made no sense to me. I had to
refresh that
housekeeping. All right, let's download
that. Throw that on our desktop. We'll
hand that over to the other system uh
when I go back into it after stream.
All right, keep that. Keep that.
Yeah, you know, Grockbots's hitting when
uh normie friends of mine are hitting me
up about Grock about Grockbot. Also,
Omari, my goodness, I swear like that
dude is on a terror right now. Everybody
and their mothers are talking about
Omari and Framework Laptops.
I have nothing against them at all. Um,
it's just the power of podcasts. It
started off with, I believe it was the
Senra podcast and that was weeks ago,
but then he recently got on the Lex
Freedman podcast and I think a few
others and he's just been on an absolute
tear to where I now have normie friends
saying I'm buying a framework laptop and
I'm installing Omari. I'm like, you do
understand that's just Arc Linux, right?
Uh, nothing against it, but um
I'm curious to see if that is a fad or
not, right? Um,
very curious. But soon as Elon's done
archiving and preparing these handoff
docs, I'm going to finish cleaning up
this team and then we're going to dig
into what I've done so far with the
database factory. I actually did deploy
uh a super basic version one uh based on
the type of clients I work with. I
figured why not build a database of
resources uh for that industry for free,
right? Um
that come from two primary sources,
research painpoints on social media, but
also what I'm building out for those
clients as well. what types of agents
and bots that we're creating uh for
those clients without of course
revealing client data.
Um so I'm very excited to show you that
one. I'll give you a little teaser
though. If you go to uh AI agents
construction, this is the super version
one of the database that's already live
and uh Grockbot handled all of this for
me. This team handled it for me. It's
really awesome.
All right, let's go make sure all this
handoff doc. So, he's still working.
Let's just give it a moment. Why is that
video or music cutting out? It needs to
playlist.
So my site I'm working on rates, videos,
hosts, guests. Bill Hor got the
interesting interesting.
So what I understand is your site just
rates long form versus
podcast videos. Trying to understand
that a little bit better. I can create
shorts now. Faceless. Amazing. Way
faster with Grockbot. Hey, let's go. Oh,
I'm happy to hear it, man. Can't wait to
see what you do. Share the channel and
the growth as as as you grow. I'd love
to to kind of see
what you build.
So, we're going to delete housekeeping.
I'm wondering if we should create a a
trading bot. What do you think about
this, Derek? You want to help us create
this live? I feel like that would be
insane if we can get that working. I'm
happy to fund it. Uh, I think we should
do a prediction markets trading bot if
anything or trading bot team.
I still wish we could see what he's
doing like real time, line by line, tool
calling, etc.
Think we're going to unpin Sky because
we're getting rid of him.
All right.
Let's go. Looking forward to seeing it,
man.
Now, in the meantime, let's watch this
video because I didn't finish watching
it,
>> which is helpful.
>> Yeah. So, to be less involved, but you
you really have to run you have to be in
the weeds with the team for a couple
days, a couple weeks before you can run
it autonomously. And you do that with
the routines, like we said. Uh, and you
do it with having agents review work
that you you teach them how to review
like you. So a useful thing could be so
say Greg like I gave all this feedback
on these posts that it did. I could say
okay take what we did in this thread and
turn it into a skill uh that another
agent can use when it's reviewing. So it
reviews exactly like Billy, right?
>> That's exactly what you would do. And so
if you want to see do you want to see
how the newsletter actually works? Do
you do you want to talk about like the
business case?
>> Yeah, I think yeah, absolutely.
>> Okay. So
>> I do I do because I think that
there's thousands of potential
newsletter businesses that could be
built and run with Grockbot.
Yeah, thousands for sure.
So this is this is a local newsletter.
Um, but I've run newsletters in tech and
sports for different clients and
basically all you need to get started is
a niche and where to get the best
information to to put in the newsletter,
right? So, if you're really into I I
don't know, board games and you know all
the best board game subreddits and the
best board game review sites and and
message boards, then you can make a
newsletter where you aggregate and
editorialize.
>> So, I think a lot of people are going to
watch listen to this and really like the
whole agent team idea is going to click
in their heads.
>> Yes. Um, I think that there's tons of
different business ideas that you can
build on top of this. You know, one of
them is local newsletters.
What else could people be building?
>> Any type of new any type of newsletter
doesn't need to be local. Uh, probably
would be easier not to get the ball
rolling on one that's not local. That's
a whole different can of worms. Um,
Shopify. I'm I've got a diff a separate
Grockbot account that I paid the $200
for because I'm so crazy about keeping
the contact separate and it's just
Shopify. And so I'm training it to be uh
to use the Shopify developer tools and
to
>> source products. I have a sourcing
agent, a developer agent, and like an
email agent right now. Um, and I'm still
setting those up, but Grockbot does an
outstanding job of researching products,
sourcing products. Uh, and so I'll just
I what I would do is say I have an idea
for this product. Um, I have I really
like this company and its product. I
want to do something that's in in the
same category, uh, similar price point.
and it knew to go look at import Yeti,
which like can just look at people's
Shopifies and tell you like exactly what
they're sourcing.
>> All right, let's uh let's just double
check everything here.
Hey, I wanted to touch base. What's the
purpose of this team, this experiment
pod? What are we doing here?
All
right, I just wanted to chat here in the
team channel. Uh, how's the money pod
looking? It looks like uh we started
building this out, but it's been stale
for 5 days now.
What's our goals?
What's the team built for? Give me a
TLDDR.
All right. So when I go through these
experiments, also keep in mind that this
account has been laying dormant. Uh
we're just reviving it right now. So
database factory.
Hey, I just wanted to get a pulse check
of this team.
Where do we stand? What do you need from
me? Um, I think we should also create
a markdown file showing me our current
status as far as
the original plan we created for this. I
know we're live. I know you're updating
with new roles. Looks like you updated
yesterday, you updated today. So, you're
actually building things out.
But, you know, we need a plan for the
future. We need a plan for revenue.
All right. And yeah, me personally, I
love having these separate team
channels. Um, that way Elon's in the mix
so he gets the context. Uh, and I don't
just have to message Elon up here and
worry that the more I message him, the
more he will, you know. Oh, that's what
you're talking about. I just put it
together. You're talking about Billy
Horl here. I don't know why I thought it
was like some other podcaster
how he did in the video. So if it's an
interview it raises his question it's
push back hype BS and then same for the
interviewer. Nice. Nice
interesting prediction markets like
Palshi my agent ran 10,000 tests and the
best ones were esports and weather
believe it or not. So betting on weather
and esports events or outcomes. Yo,
what's up, Will? Completely new with
zero tech skills. No, you are not. I
don't think you are. I think that your
willingness to learn this stuff is the
advantage, right? Um it's never been
easier to build things, right? All you
have to do is have a vision or or a bit
of curiosity and just start asking
questions. They don't even have to be
the right questions. Just start asking
questions. I would say, you know, I said
this at the beginning of the live stream
from a beginner standpoint. Don't go and
create some crazy team. Create one bot.
Test it, one use case, then expand, then
get the idea of creating this chief of
staff, right? I don't think everybody
needs to start with that type of setup.
I think starting as simply as possible
and just building one thing, one
automation, one checker. You could do
something surface level, some email
related work, automations, etc. But
that's all too surface level for me.
You'll never catch me creating a video
or, you know,
telling you to do that via, you know,
uh,
I want to choose my words carefully. Um,
you'll never see me hyping up the basic
standard workflows uh that everybody
seems to be spitting out and copying
right now. Yo, what's up? How we doing
today? Congrats. That's awesome, Trout.
It's good to hear. Congrats.
Let's go.
All right,
let's go check back with our team here.
Okay.
So, let's go check on it because I want
to give it some feedback here.
We need to do a better job at uh
ready today works with supervision.
Okay.
I see. So that navigates through
jobs, president, owner,
this is publishing, next project
manager,
and then there's specific jobs and use
cases per. Okay.
Feel like we can improve on this quite a
bit though
for collaboration. Yes. Yes. Uh
collaboration with humans. Definitely.
Currently growing my Grotbot team to
automate my residential real estate
business. I have a ton of agents and
created my own system and process over
the last few days and slowly automating.
Right on. Coach Clay, that's good to
hear.
And correct me if I'm wrong. Did you
just start with one thing before you
then start feeding more responsibilities
to your team?
Yo, good afternoon, Anthony. How we
doing today? I'm afraid to touch
Grockpot after the whole Grock build
code theft thing. Yeah,
it's uh
How long ago was that?
How long ago was that? I'm not sure. I
remember that being a a topic of
conversation, but
I mean, look at the look at it this way.
Didn't um Hugging Face get hacked
recently and then they were just
acquired for what is it 12 billion or 17
billion? I can't remember the number. Um
so these types of vulnerabilities aren't
necessarily a bad thing. Sure, they're
bad in the moment. Um but they get
fixed, they get resolved, they get
addressed rather quickly. So,
ah, that's right. That's right. Um, oh,
you're talking about it creating a copy
of private repositories if I remember
correctly. Right. I believe they
answered that. I I can't remember what
it was, but listen, you know, I I
operate based off of one simple rule. U
there's no more privacy if you're using
a smartphone. Like, this is not a
conspiracy. if you're using a
smartphone, if you're on the internet,
um there's just no more privacy. So, to
me, I I'm not as concerned about those
things anymore because, you know,
something that's super private, I would
never hand off to an agent first and
foremost, right? Like, um so I think
it's just understanding that we live in
a world where uh that aspect of things.
I mean,
yeah. Yeah. No, and I totally
understand, right? I totally understand.
and they were using it as training data,
but it wasn't every private repository.
It was specific workflows. It was
specific um harnesses that would do
this. Um but it wasn't Grock as a whole
that was doing this. I think it was a
specific like um pipeline of using Grock
that was causing that to happen.
I mean, a number of different ways. Um
a number of different ways. I can sit
here and tell you that Grockbot has
closed some sponsorship deals. Uh I can
sit here and tell you that Grockbot has
gone in and uh cancelled subscriptions I
didn't know I had. Right? These are a
lot of the common use cases in videos
you're seeing right now. Uh but what
this live stream is all about is let's
build a team that's actually going to
make us money. And it may not make us
money today, but it's going to build out
those automations and those bots for us
to then make us money in the long term.
Um, and this one here, this database
factory team is this is version one, but
I'm building a database factory
essentially, uh, for AI agents in the
construction industry.
I have two chief bots, but they are all
next to bots that work in parallel.
They're on the left side of your menu.
Do you think there's a way to organize
them as if only in a circle in a bubble
and when you click there the bots of
each of the main bots can be broken
down? I don't know if I'm going to be
able to do that. Um trying to So here's
what I personally like to do. Let me
maybe this will help you. Um
I have pinned agents here. Obviously I
removed Sky. Sky was another one. But
this is the chief of staff who's
essentially routed into every team
channel. you see here. So, when I start
talking with the team here,
Elon's routed in, has context. Uh, I can
go in here, message him, and he'll know
exactly what happened in the team
channel for the database factory. Same
goes for here. Same goes for here,
right? Uh, if you're thinking you want
like a team of chiefs, then what I would
probably do is either
still have one master orchestrator uh,
and then create a Chiefs. I need to
rename this. These are actually
harnesses.
Um
maybe re, you know, create a new section
with all of your chief of staffs. Uh and
then create a team channel with them so
that you can communicate with all of
those head of departments in one
channel. Uh that might be the move if
I'm understanding your build correctly.
you. Yeah, you don't need to have an X
account to use Grock. Um,
yeah, you don't need to have an X
account to use Grock. You can sign up a
number of different ways.
Nice. Nice. There we go. Hey, don't be
ashamed. You're testing. You're
experimenting, right? Like we learn what
works for us and what doesn't work for
us. Um, I paid for Replet early on, well
over a year ago, right? I was paying for
Replet, uh, dabbling in cloud code and
then as soon as it clicked, I think I
believe I made a video about this, but I
calculated how much I was spending per
month on Replet versus how much I would
spend equivalently if I just took that
money into Cloud Code and paid for a
bigger subscription. and I was getting
more usage out of it. The only negative
per se between the two comparing was
Replet made it really easy to deploy
right and host. I have a network
engineer background. The DevOps side of
things is no problem for me. The hosting
side of things is no problem for me. So,
um, Cloud Code was a no-brainer. But
now, I mean, all you got to do is ask an
agent or ask a bot and they'll they'll
guide you through the steps if they
don't do it for you. Uh, making it so
easy. Like, you don't need to have the
background that I have to host things
and deploy things yourselves anymore.
Uh, that is just that's just gone. You
just need to know how to ask the
question. That's it.
Yeah. Bots as needed when the skill or
process calls for it. Yep. Interesting.
Derek, should we build one? You want to
help us build one right now?
Main difference between Hermes agent and
Grockbot? Well, Grockbot,
there's a couple of differences. One is
open source and one can be completely
hosted on your own hardware, whereas one
is not open source and one cannot be
hosted on your own hardware. So, you're
at the mercy of Space X data centers.
I doubt they would ever go offline,
right? But that's like one of the big
differences right there. Another one is
with Grockbot, you're limited to using
Grock 4.6 or just Grock in general. Uh
whereas with Hermes, you could connect
your Grock, you can connect your OpenAI,
your Claude, whatever harness you want.
Now, I'm going to put an asterisk on
that comparison because there are
workarounds to connect your codecs and
Claude subscriptions within Grockbot.
Now, this bot itself is still powered by
Grock, but the underlying work it does
routes through the CLI. Whether you set
up Peekaboo on your Mac, assuming you're
with a Mac, um it'll just invoke codeex
or Claude in the CLI, do the work,
report back, and then the bot itself is
still powered by Grock. But fun little
workarounds to use uh other models
within Grockbot. I have this working
better on my other system that I created
on the demo which has turned it into my
primary. But nevertheless, fun little
workarounds.
Uh VPS to Grockbot, you don't need to.
It's not self-hosted. It's hosted by
SpaceX. You don't VPS is out of the
question. If you're talking about
Hermes, yes. Yes.
I got to run out for a bit. Daughter's
kindergarten. Oh man. Busy busy. Hey,
right on. Nice. We can always do
something uh on a live stream next week,
too.
So, I hope you have a a great rest of
the day. I hope it's an easy day. I hope
the teachers don't give you a hard time
for whatever reason.
I don't know what it's like with the
whole back to school stuff, but
it's funny you say kindergarten open
house. I've never heard it's like, are
they becoming real estate agents? What
is going on here?
Sorry, I had to mute there.
Oh, that's right. I created tags in our
mailbox and created aliases.
Yeah. So, what happens daily right now?
What happens weekly and what happens
monthly? What can we do to improve?
Because the front-end UI leaves much to
be desired for. So, we got to we got to
put in some work there. But, I'm just
trying to get a pulse checker on,
you know, what's happening daily,
weekly, and monthly at the moment.
And what's our goal? I'm pretty sure we
defined it, but I want you to tell me
what's our goal for this project.
Uh, this pod sells one thing.
Interesting.
Interesting. Really, I could have sworn
that the money pod was something else.
So the money pot it looks like we are
I see.
So what's currently configured? What's
our goal? And what's configured daily,
weekly, and monthly right now.
This is what happens when you build too
much.
We do a lot of experimenting here on the
channel though, so it's all a part of
the process. The one that I know is
going to stick is the database factory.
We're still experimenting here. We might
delete these teams and then just start
from scratch.
Wait a second.
To find one clear mud product a month,
not to ship every morning. I hunt what's
actually moving on X last 30 days pulls
what people are saying architect turns
it the pain into a pack GTM or go to
market bot says whether anyone would pay
water and mud only jump in if we pursue
ah
okay
okay so I I think we need to shift this
because we're moving clear mud related
responsibilities to another system so I
wonder if we just don't simplify this
into a
listen for pain points,
create a prototype,
create a go to market strategy, and then
present me with that output one new
product every single day. And then I
choose whether or not we want to move
things down the pipeline, right? Also,
what do we have configured as far as
routines are concerned within this team
when it comes to daily, weekly or
monthly routines?
Just know that Grockbot is essentially
Elon's digital optimus.
Feel like he's learning from all the
Grockpot usage to eventually release
agents that actually do Yeah. Yeah.
Yeah, for sure. Yo, what's up, Jordan?
Hey, no problem for me.
I mean, one account got close to 100%,
but I was very, very mindful of it. Uh,
and then we got that that reset two days
ago, which is awesome. Which is awesome.
Well, I think we're going to see more of
those personally. Um, but I'm I'm on
average to spend about
10%
maybe 12% a day, which leaves me with
some extra percentage. Uh, and that's on
my Clear Mud system. On this system,
it's completely separate. So, I have two
different Grockbot systems right now
that I'm testing with.
One I actually built out in that video
from a few days ago, which is now turned
into my primary system for ClearBud. So,
I'm not going to demo that one too much
more moving forward. Uh, I'm primarily
going to demo this account here. That
other one's running off of a Super Gro
Heavy and this one's running off of
Cursor Ultra.
So, we actually experimented with this
on Wednesday's live stream and uh we
were following a certain list of hacks
that were recommended on X. Um,
yes, it could do it, but it's not
natively built in. You have to do you
have to take a lot of steps to to pull
that off
and you have to set up a a phone number
based off of how we did it at least. Can
it send voice notes in chat? Let's ask.
Hey, if I send you voice notes, can you
send me voice notes?
Hey, if I send you voice notes, can you
send me voice notes?
What's going on here? I'm noticing a
weird little bug is when I try to record
something here.
Hey, I'm curious. The microphone button
on this,
you know, field where I message you
doesn't work. Uh, I've tried both ways.
Also, can you send me voice notes? Can
we send voice notes back and forth? I'm
very curious.
Okay. Right on. Yeah. Hope hope it's
helpful.
I already answered this. I think so. I
think so. I think it's worth an
experiment for sure if you have the
budget, right? If it's not going to hurt
you, if it's no skin off your back, then
yeah, why not?
Cursor Ultra is uh $200 a month.
Yeah, exactly. Exactly. The reason I
justified what I'm spending is that
super Grock heavy account in itself,
it's already paying for itself. It's
already paying for itself with what it's
doing for me. Uh it's freeing up my
time, which is awesome. This other one
is more of a business expense for Clear
Mud to experiment, to demo. This is for
you guys more than it's for me, right?
Um, so if you need something built
specifically, if you want to see how to
build something, just ask. I'm happy to
build something live showing you the
steps that I would take.
Yeah, I think they are right. Maybe I
can't remember. Let's go take a look.
I want to say if we go to bot
believe they shared yeah all super grock
plus cursor pro plus and cursor team
subscribers now have access.
So, Super Grock Plus. Uh, how do we
easily get to
go to grock.com?
Looks like I'm actually signed in with a
different one here.
Where are the plans?
Grock plans. Super Grock.
I'm pretty sure it's the $30 a month
plan if I'm not mistaken.
Oh, maybe it is 100. Okay.
H interesting.
Ah, very nice. Very nice.
You can do it from a laptop. They don't
have an Android app for it yet, which I
am bummed about.
I am very bummed about that because I'm
an Android first mobile user, but then I
do have this Mac as a
computer here as my daily driver for
creative.
Interesting how our music keeps pausing
in the background.
Go back to flow music.
Do I have to go in here?
No. Let's go back. I think we should be
able to play the playlist as a whole.
And we just hit
There we go. I think what it might be
doing is it might be shuffling down to
the end and then ends right here. So,
let's turn off shuffle and just do
automatically re-re replay.
I cannot send voice notes back. No audio
tool on the seat. I send text. Okay.
The other org has outbone voice to you.
That's a different machine. Yeah.
Got it. Got it.
You know, I'd love to will I really
wanted to have um because I'm not a
trader first and foremost.
I really wanted Derek to help to to
build it with us live because he has
that experience.
Um, I'm a, you know, I really stand
behind making sure you have the
experience to build something. Uh,
otherwise you're just going to build a
super surface level trading bot that's
not going to really make you any money.
I think from a trader's perspecting ha
perspective having that foundational
knowledge of how to trade of
if then statements of certain criteria
right to track you're going to be far
more successful in creating a trading
bot with that foundational knowledge and
configuration but just telling it to say
hey I just gave you $50 on this MetaMask
account I want you to go to this
prediction market start making us some
money that's not enough. It's most like
nine times out of 10 it's going to lose
money, right? Um but
if you give it specific tasks, specific
routines to do certain research, a
specific niche to target, uh I think
that's when you become more profitable.
I think it looks like we're going to
build that next week because Derek had
to take off. Dererick's a busy man. He's
got like what is it 17 kids? No, I'm
joking. He's got seven though, which is
a lot. Um, so he had to take off. But to
answer your will, I think next week I'm
going to DM Derek to see if we can't uh
get him on a live to to essentially just
build a trading bot live. That would be
a fun one.
I want to go back and focus on this
database factory. Let's see.
Interesting. It set a goal for 100k a
month by year end from that. What does
that mean? By year end from that year
end from that. From what?
Where is it getting that from? Hey, I'm
with it. I'm not I'm not mad at that
goal right there, you know.
All right.
messaging claude for contractor
blueprint spec and codeex to ship it. So
it looks like they are working right
now.
Um
passing that to Claude. Okay. So I mean
based off of this original prompt, I
didn't even tell them to do anything. I
just said the front end UI leaves much
to be desired for. So, we have to put in
some work there. And they got started
off of
prompting our uh our bots that handle
the codeex and cloud harnesses. Awesome.
Awesome.
So, let's go check in on Money Pod.
Interesting. Really, I could have sworn
we had money pod with something else.
Goal one paid to install bid desk.
That's me followup.
Got it.
Oh, interesting. I kind of want to start
from scratch with this team here.
Yeah, I think we're going to trash this
experiment and start from scratch.
Enjoy your last few moments.
Oh, I cracked myself up. I'll let them
enjoy their last few moments. Let's see
what they say as bots.
Yes. New loop. Listen for pain. Ship a
prototype. Write the go to market
strategy.
Put it on the walk-in tab. One new
product every day. You decide what
continues. Got it.
All right. So, I'm also curious as to
why we're using water and mud. Those are
our old Clear Mud team members. Uh we
primarily use the team we've built here
plus our Codeex and Claude bots um that
route work through each respective model
whether you're choosing Opus or Fable or
GPT Soul or GPT Terra Luna etc etc. All
that work gets routed through them.
See what did they say?
Enjoy your last few moments. Come on,
say something.
I guess Elon's the only one answering
right now. Um,
yeah, we're just going to delete. Going
to delete. Fun experiment. Let's clean
things up in this system.
All right, I went ahead and deleted the
money pod team.
That was an experiment. And uh
we want to go back to the drawing board
there. So, also I'm noticing that
sometimes mud and sky and water are kind
of spliced into certain team channels.
That shouldn't be the case. They're a
clear mud team. We're handing that team
off to um to our other system, right?
So, I'm getting ready to delete those
agents shortly.
You know what? We might keep Mud because
didn't we build a Opus clone here? We
did. We did.
But I've already added that repo.
Did we add this mud clip? Feel like we
do. Let's go check.
Mud clip.
Is it mud dash clip?
Give me a link to this uh repo, please.
Yeah. If anything, I'll keep mud and
I'll move him into
actually rename this one engineering.
Move mud down there
because he was more of our engineer than
anything.
I'm going to remove that tag for now.
And then let's go ahead and create a
channel for Elon. Elon Mud, Claude, and
Codex
team chat or team channel.
Move that up here. That way they all
have their own way to communicate. And
then this is one we need to archive and
delete shortly.
Yeah. So, what I'm ending up doing here
is I moved mud into engineering. Um, so
we're going to delete and archive. Well,
first we'll archive, then we'll delete
water and sky shortly because those
responsibilities we're handing off to
the other system.
Dang, Mr. NY,
what angle did they pick? Who
is it? Nice.
I have not uh I have not experimented
with uh GLM 5.3 yet,
but are you still using Hermes and
OpenClaw? Yes, I'm not using them with
Grockbot. I'm using them independently.
So, I'm going to be honest. Hermes I
deploy Hermes agent systems for clients
um
because of how stable it has been since
day one and reliable. That's why I chose
Hermes Agent. I have my own Hermes
agent, but it doesn't do much aside from
I don't give it too many
responsibilities for for clear mode as a
whole. My openclaw setup on the other
hand that I originally built out, you
know, end of December, beginning of
January, I still use daily. I still rely
on daily. Um, yeah, I still, long story
short, yes, I still use Open Claw. I
still use Hermes. I still use Buzz. Um,
but Buzz is more for collabs now because
I want to use the collaboration feature
of bringing in other humans and agents
and those are its responsibilities right
now. Open Claw's day-to-day clear mud
ops. Grockbot now is I'm giving it some
responsibilities, but it's not going to
outright replace OpenClaw or Hermes for
me personally. Um, but I am, you know,
it it it's going to be doing a lot for
me in the future. But is it gonna
outright replace OpenClaw? I don't think
so. I'm too much of a a techie and an
experimentter to to just trash something
like that, trash something that's fully
in your control on your own hardware,
right? Sure, you run the risk of that
hardware crashing, etc., etc., but
that's the type of person I am, right?
So, I'm never going to trash my open
class setup, personally. Um,
and never say never. You never know what
happens in 10 years, 15 years, but for
the foreseeable future, like I don't I
may make content around all these new
solutions that drop. Um, but I'm not
like these other guys who were saying I
now migrated everything to these other
systems. Uh, if that were true, then
they never really utilized those systems
in the first place. uh because m keep
continuously migrating from system to
system to system shows how little you've
actually done in those systems uh in my
opinion right because
if you if you're truly relying on that
tech uh the amount of time it's going to
take to migrate from one system to the
next I is insane is insane like um I
don't know I'll stop talking there I'll
stop talking there
yeah this right here This is this is a
great use case, great little workaround.
Um, but you're still at the end of the
day using Grockbot, right? Hermes does
have its own bot feature. If you guys
know, Hermes blatantly copied Grockbot
within hours. I feel like within their
desktop app, it was hilarious to watch.
Nice. Nice.
Yeah. So, this was my original thing is
Grockbot is primarily going to focus
around X growth for clear mud and being
more consistent across socials. Uh,
primarily X, right? I'm giving it more
social first responsibilities than
anything. I'm not giving it YouTube
responsibilities because I have certain
automations in place with my open cloth
setup that are just tried and true. They
work. I don't even have to think about
it. All these live streams, they've been
automated uh when it comes to
post-prouction. So, you'll notice within
24 to 48 hours after all of our live
streams since the beginning of the year,
right since I set this up with OpenClaw,
uh it'll go in, download the transcript,
create timestamps, create an SEO
optimized description based off of
what's discussed in the live stream, add
tags, and then update YouTube. I don't
have to touch it. Sure, the first few
weeks I was monitoring how it was doing.
I was tweaking it. It did break once, I
want to say, in April. Um, but I just
set it, I forget about it. And that's
one of many automations I have in place
with OpenClaw. I have no intention of
migrating that to Grock or anything
else, right? There's just no point. It's
a waste of time. A waste of time.
Yeah. So, here, Grizz, this is actually
a great I love this comment. It's like
Hermes with training reels. So, listen.
Grockbot, I don't like
it's not revolutionary. It's not
gamechanging in the sense of it's
inventing anything new. What it's done
is reduced that friction that people had
with trying to set up an OpenClaw or
Hermes or Buzz system for that matter.
It's so easy and seamless to get
started. And that's what's gamechanging
about Grockbot. It's not the foundation
of it. It's just how easy it is to use.
Like that's what we needed. And I think
that's why it's making waves and and
it's essentially has gone viral, right?
Is it? Yeah, I haven't played with it
yet. So, one of my clients, we are
setting up individual desktop a access
for the employees. Uh, I did start
testing from an employees perspective
with the bots. It's super fun to use.
Um, not the same as Grockbot, but still
super fun to use. That staff's going to
have a lot of fun uh once they get
access to that next week. I love
Grockpot. It just blows through my usage
like crazy. Such a low pool with Super
Gro Heavy. Yeah, I hope they fix this.
Um,
the here's the reason I have two subs,
Adam, is because one cursor ultra sub
for me giving it responsibilities for
clear mud, but then also demoing and
experimenting with Grockbot from a
YouTube demo perspective. I was max I
maxed out way too quick, right? So, I
decided to just bite the bullet, so to
speak. I have a super grog heavy for
clear mud now. And then this cursor
ultra which is going to serve as a
YouTube demonstrated experiment lane of
just testing and having fun. And I
haven't hit 100% yet. Um
I it's my hope that they fix this
though. It's my hope that they fix this.
I think they're going to improve uh and
and and also become more transparent as
well. Like I'd like to see a
side-by-side comparison into how much
usage can I use via Cursor Ultra versus
how much usage can I use via Super Grock
Heavy. If it's the same and the only
reason I have these subscriptions is for
Grockbot, then it makes sense to only
pay for Cursor Ultra. But if you get
slightly more out of that Super Gro
heavy account than you do out of that
Cursor Ultra account when when I I think
we're going to get more clarity in the
coming weeks. Um, we just got to be a
little bit patient.
It's a good gateway for newbies.
Then they can graduate to her. Yeah,
they can graduate to whatever they want
to set up, right? It doesn't have to
just be Hermes. Listen, I,
as you've learned, I'm deploying Hermes
agent systems, but I'm still team
OpenClaw. If I had to choose one over
the two, it's OpenClaw.
That's just my preference.
acting like they're going insane, left
running. Yeah. So, I think a lot of this
is
lack of context around what you want
them to accomplish, which is sometimes a
bad thing, sometimes a good thing
depending on what you want to
accomplish, right? Um
yeah, I think a lot of this is lack of
lack of uh context of like what's its
purpose, right? Uh if you clearly define
its purpose and the certain loops you
want to achieve with your bots, uh when
you want them to stop doing work, that
won't happen, right? But if you leave it
open-ended, I could I could see a world
where that that could be a problem.
We've had some comments on some of the
videos we posted of people just uh
sending a task and the bot not stopping
because they weren't specific enough
with, you know, stop at this output,
right? Um so I I think um
yeah, I think that's case by case there.
Newb integration they enabled for X
after releasing Grockbot makes this
concept interesting. The bot can now
have its own presence on the platform
associated with your account. Yeah. So,
one thing I did notice though, if you
give full access and automation to your
Grockbot to post on your behalf, every
post says made with AI. But if you say
stage this draft or stage this reply and
then you go on mobile or desktop and you
go into your drafts and you make one
little change and hit post, that made
with AI tag is gone, right? So, if you
automate the posting and replying, yes,
you're just going to look like a bot,
though. Uh, I don't want to look like a
bot. I want it to help me, but
ultimately I'm the last person to
essentially send that prompt or let me
rephrase that. Send that post.
No. All good. I think it's on topic. So,
I hope that answers it.
Does low friction reduce usability
compared to other harnesses? Yeah. So,
I don't think so. I don't think so. I
think it's, you know, Grockadot is built
for a very specific type of person,
right? Um,
and we all have a touch of that type of
person. Like, a lot of us are techies, a
lot of us aren't, right? For some people
in installing Open Claw is intimidating,
for some it's not, right? It's like uh
and I I think as Grock gets better, I
I've said it before in the past and I've
gotten a lot of uh flak for it, but I
think Grock, Space XX, let me let me
rephrase this. SpaceX AI is going to be
one of the top frontier models in the
coming years.
Uh Claude is already I their downfall
needs to be studied.
Like, yo, what's up, David? How we doing
today?
They confirmed it's the same.
Interesting.
Yeah. Yeah. Very cool. I think we saw
that when we briefly went to go look.
Share as template.
Nice.
Now, does it pull
I'm curious if it pulls out any like
sensitive data in those instructions
automatically. I would imagine it does.
See where are we at? 11%. I'm not sure
where we started this session is. So I
wish I I kept track of that.
Yeah, I'm not going to delete them yet.
I want to go into my other system and
finish migrating these two.
Okay,
let's go back to
our database factory.
So, deal looks like he has some
responsibilities.
Subscribe form is live. Cool.
I'm curious to see like what uh what the
new UI is going to look like.
Hm. Okay.
Now, let's go back down to our
experiment pod.
Going to rename this experiment pod.
All right, cool. Let's um let's force
run one right now. Build me something.
and list out a table of all the routines
associated with this experiment pod as
well. So I get a
just a visual proof of what happens each
and every day.
Oh, you know what I'm realizing? We need
to add uh
how do we add
should be an easier way to add members,
right?
Huh? It's weird how Elon's not in here.
Is it settings? No.
You would think there'd be an easier way
to just add a member. No.
Do I have to create a new channel for
that? I'm shocked he's not in here.
Also,
we need to add Elon, our chief of staff,
into uh
Huh. into this uh team channel.
Hey, I noticed you're not in the team
channel for the experiment pod. Can you
add yourself? Can't find an easy way to
add you.
Interesting.
I mean, I could just create manually
another channel, but since they're
already doing work,
let's wait here.
Interesting. Okay.
Yeah, very interesting.
Okay. Looks like he's added in there.
Awesome.
And they adjusted the team, right? Check
out the engineers cuz they don't need to
be a part of each team pod.
Great.
Looks like Fable is building something
called same model. Don't know what that
is. Let's see how it looks.
Let's say model. Paste a stealth slug
plus date window show identifiers.
job people actually did name the
hangover.
This is very brief. Interesting.
All right. I'm very curious about
Grockbot as a whole. Do you operate
based off of certain skill MDs or
certain
a certain file system that we can then
organize certain markdown files so that
you follow them or each individual agent
follows them? I want to get a better
understanding of how we can improve
certain automations namely for database
factory for the experiment pod and any
other future teams we create.
My buzz partner installed Grockbot and
wanted me to collaborate on his
projects. It seemed impossible.
Interesting.
Yeah, I think right collaborating I from
to my understanding you can't bring
another human into your Grockbot system.
What you can do on the other hand is
either have one of your bots communicate
with another person through the remote
computer. Um,
give your Grockbot bot and underlying
bots an agent mail identity so that they
can email people at will.
Um, or if you're building a web app or
whatever, create a private GitHub
repository with your bots, have it sh,
you know, commit to Git that way and
then share that repository with somebody
else to collaborate.
That's that would probably be the most
efficient route there, AV.
But as I learn more, I'll uh I'll
definitely share. Matos
to the bane too. How are you?
Been AFK last couple weeks, but I've
watched your streams. Testing Crockpot
stickked more than Buzz. Nice. Nice. I
love it. Hey, observing is what it's all
about. Listen, I still have a Buzz boot
camp coming out. Actually, I'm still
going to record a firsttime setup use
for Buzz. Uh because I think Buzz is
still very powerful for its use case.
Um, I'm also going to be teasing
something to where we're launching a new
series on the channel, um, in
collaboration with a digital marketing
agency. Uh, the CEO is going to come on.
It's not going to be live. We're going
to record these videos. Uh, but we're
going to be creating agents and bots for
her business, which are then going to
serve as kind of like a dual purpose,
right? We're building out her agent team
for her, but we're also doing it live on
video so that we can then postedit and
create series on how to create an an XYZ
bot for XYZ. And I I say XYZ because
that's like fill in the blank, right? Uh
I think it's going to be super valuable.
Um, that is
at least two weeks out until we start
posting those videos. Uh, possibly
three. But it's it's on the calendar to
record. It's it's on the calendar. Um,
forgive me. I need to put this on
silent. I hate these little They just
They distract me too much. Um, it's in
motion. Let's just put it that way. It's
in motion. So, I'm super excited to
record those videos and uh share more
about um this project as soon as
possible.
Already had an iMessage gateway running
on Mac Mini for Hermes to use.
Nice. Lovely workound here. Awesome,
dude. Awesome.
Each seat has a job description in
memory. Yes, that's standing intent. A
routine is a clock. Yep. Plus a prompt.
Yeah, we know. A skill is a shared
recipe any seat can run. Okay. I don't
think he understood my question. Um to
make a new team tighter.
No, what I meant is is there a directory
somewhere like let me go check on my
route.
Um, command shiftG I think to search.
I guess not. Maybe I need to be in
finder.
There it is.
No. Yeah, it would be on the system
itself. I don't think it would be local.
Yeah.
Skills love on Grock bots computer. Got
it. Got it.
Show me a file tree
of how it looks and where to access that
on uh the remote computer.
really wish this was rich.
Could I get those responses in like rich
text? Is that possible here in Grockbot
chat?
because that table is uh while I can
read it, it's not uh it's not very uh
user friendly.
All right, we are back.
There we go.
So, daily products
7 days a week. It hunts.
It's live. Last 30 days, none. Dump when
this room asks. Okay. So, it's on
demand. Product architect
802. So, it's an hour later. Pick latest
live. Go to market. That's 30 minutes
roughly 30 minutes later. Every single
day pulls from latest. Okay.
Okay.
Okay.
I have an idea.
Let me go find this repository. One
second.
Home box scratchwork is
forward/workspace skills you care about.
Okay.
Hey, this is good to know here.
Ha. They call them agents.
They don't call them bots.
Awesome
man. This in itself can be a video right
here.
Interesting.
Wow. So, this is very valuable
information. Um, I'm kind of thinking
about turning this into a YouTube video
where it's about, you know, how to
properly configure your bots via
Grockbot.
Um,
any thoughts?
Oh, Nikolai was already on top of it.
I want you to build a um build an
Excalibur canvas.
I think we should have a screenshot of
this file tree, but also you know where
things get placed, what happens when you
create a new bot, how you should create
a new bot, um what you should be telling
your bot on how to store certain
information. uh I think would be super
valuable and I can walk through this uh
and we can walk through it in a YouTube
video recording and and you know a show
off existing systems but then b set up
new bot teams based off this.
This is taking a while.
Hey, where do we stand with this?
Yeah, they're already working there.
Looks like they've already built a
prototype here.
Super basic. So, I need to uh
Oh, I am. I am for sure.
So, what I do with that canvas tool is I
first create an Excal drawing and then I
hand it over to it and then it it
creates its version.
But I haven't recorded a video like that
since last week cuz the last video I
recorded, we really shot from the hip on
Tuesday to where I thought it would be
fun to when I set up that second account
to just record what I was doing and it
turned into a very great video, super
valuable. So, I'm going to start to to
record more videos like that to where I
don't do any prep work. I just have an
idea and I'm recording my screen and I'm
talking through it, right, real time.
Um, and then of course adding those
those types of elements if we need to.
But yeah, it's a two-part thing. I don't
build the canvas from scratch. I use an
Excal drawing to help me come up with
our canvas on clear mud canvas, if that
makes sense.
Oh,
yes.
Yeah, I just noticed I didn't see any
notification saying I needed to approve
a command within press. I think um you
know a rule of thumb moving forward I
want to be as handsoff as possible here.
Um so there should be no blockers, no
allow once, allow all. Uh whatever we
can do to override that within this
grouping would be great.
Sweet.
Oh, Grockbot Opus clone is what it was
called.
Yeah. So, we s saved Mud clip. So, I
could in theory get rid of that one,
but I think we're going to keep mud for
now.
Interesting.
We're about to find out what it does.
And let's see. Last time I checked,
we're at 11%. So, we burned through 3%
already. Interesting.
And how many days left do we have? Reset
in seven days. Change limit. Oh, it
takes us to cursor. Okay.
It'd be curious. I want to monitor this,
but I'd be curious to see, you know,
with these automations between the
experiment pod and the database factory.
Um,
no. Are you talking about my team? So, I
have my team specifically set to use uh
these harnesses here with these two
bots.
Where are you getting that from? Are you
getting that from my screen? Because I I
you know I set this up to to work a
certain way so that it could utilize
Opus or Fable through my Claude sub and
then also for GBT Soul um through my
codec bot.
So, the current plan that we're demoing
today is the Cursor Ultra $200 month
sub.
No, all good. All good.
Chris McConey.
I mean, he posts a lot, but I'm not sure
where you're seeing this from.
Ah, okay. Okay.
Okay.
Um, feel like it just made it worse,
right? What just happened?
I'm gonna wait for them to finish.
Um
Trying to find something. I could have
sworn I bookmarked it. I guess I didn't.
Maybe it's in here. Let me go check.
off screen.
I had this repo that I wanted to start
using. I was just trying to wait for
a good time
to use it. Hold on. I can't seem to find
it though. It's not here. Shoot.
That's a bummer.
Oh, what was I think it was
Okay, I did find a UI UX skill that I do
want to actually feed to this team.
All right. I'm not sure if you guys are
done with the work yet, but I didn't
think it would have been possible, but
you know, a the website looks exactly
the same and now the nav header is
essentially misaligned. Um, I'll send
you a screenshot, but
I'm wondering if I need to do a better
job at giving you guys a skill to follow
moving forward. Um, because you're just
not doing the work that I expect you to
do. So,
let me know how I can prompt you to do a
better job here.
Welcome back. Welcome back.
Oh, interesting.
I wonder if this is Yeah, interesting.
Very interesting.
I'll have to look into that.
Oh, I see what you mean. I see what you
mean. So, it was inspired off of this
Greg Eisenberg video, right? Um, that's
why I used like the same dollar sign,
dollar sign, dollar sign uh in our
title. And I've already created the B
database factory here. So this is not a
short-term plan, right? This is a
long-term plan, but it's not something
random, right? I built this uh I
squatted on this domain a while back
because I started building AI agents for
the construction industry, primarily
GC's uh and then construction companies
focused around demo and abatement. Um
but I thought what if we build our own
database factory? So, this is kind of
like the version one of it and its
purpose is to generate traffic. That's
our first goal is how do we get our
monthly visitors up to 10,000, then
20,000, then 50,000. You see my point?
Um,
and then eventually monetize it in a
number of different ways. But we can't
do that all in one day, right? Like this
is a slow burning project. So
that's, you know, route number one.
Route number two is this experiment pod
to where we build products every day,
see which one we want to ship live right
at the end of the week. Uh, and then I
want to build a trading bot, which I
think is more of a short-term
benefit to the viewer. Um, and I'd love
to collaborate with somebody experienced
in trading rather than me building a
super surface level trading bot. So
yeah, today we're just talking about the
idea of making money with Crockbot,
showing the foundational kind of bricks
I'm laying to make money down the road
off of this experimental account. If
that makes sense. Hope that answers
answers it, Matios.
Yeah, I do think uh let what tech stack
is this site built off of? Let's start
there. Um because I really love the idea
of using shadec components. So, if you
go to shadescenblocks.com,
you'll find a massive library of cards,
forms, like everything we need to build
an awesome website. Let's just put it
put it at that. And I can get a little
bit more specific, but maybe our tech
stack doesn't support those elements.
I'd be shocked to hear this, but please
let me know.
shade CN blocks.
It misspelled it, didn't it? Let me
double check that that's the right
domain.
Yeah, there we go.
Hey, let's go.
I'd love to be the judge of that right
there.
Astro 7 static custom CSS no react no
tailwind. Ah
well let's uh let's make these changes
then. Can we still keep everything
hosted on Cloudflare?
You also mentioned a faceless YouTube
channel in some past live stream. Um,
yeah, I think I know what you're talking
about, but the generative AI at that
moment in time, it just it didn't make
sense to release more slop in such a
sloppy
ecosystem. Um,
might be a good time to revisit that if
I'm being honest with you.
The question is, do we focus around
short form or long form? Feel like long
form is the move. No.
So, we started when I started looking at
it, we were at 11%. We might have
started at 9%. So, so far in this
session,
it's creeping up there. Interesting.
But I did do a lot of maintenance. We
did do a lot of prompting, a lot of
handoff documentation, a lot of deleting
uh after the handoff docs were ready.
So, I wonder how much percentage um that
was responsible for. You know what I
mean?
go sheets CN blocks. I want to give it a
specific um
nav header here.
Yeah, I like the idea of using this
header.
because it has a really awesome search
feature built into it. Um, let me know
if you have any questions.
And maybe I changed that. Maybe I
changed that. Let's see. Let's go back
to nav on.
For some reason, that is not working.
Why is that not working?
And then I think all obviously we should
have breadcrumb crumbs incorporated uh
into the nav at the top.
Yeah, it's crazy. I don't want to
contribute towards that, right? So, you
give me some ideas, though. You give me
some ideas.
Let me go check something. Maybe we
build out a team right now for this.
You just go pull up an old project.
Where did I put that? That's the
question though.
Was it in YouTube ideiation?
Or did I
Now, I could have sworn I had a certain
project somewhere.
Must have been a different account.
I can't seem to find it.
I honestly kind of thought it was a a
GPT, like a custom GBT that I built out.
Ah, maybe I just found it.
Where is all the context related to
this?
Uh, review GPT. No, that's gonna review
GPT.
Edit.
Welcome back. Configure.
Oh, wow. Yeah.
Oh, yeah. Remember reserving that
channel tag, huh?
Huh? Let me uh go full screen.
Okay, maybe we uh we build out that team
right now. So, we're going to call this
uh
oh, we create a new section from the
agent. You can't just create a whole new
section.
All right, I want to build out a new
team.
And this team is going to manage
a new Faceless YouTube channel, a tag
that I reserved called ABC Vault.
All right, I'm going to give you a copy
paste of some context about the idea
behind this channel. Keep in mind some
of this is outdated, so we'll need to
make some improvements, but I want to
build a team that complements this.
All right. And then I'm going to paste
this in there completely.
Oh man, time flies, huh?
All right, they're still building out
the new version of that.
We are burning through usage with this
experiment here.
Yeah, I'd imagine we want to create a
faceless team. So
ideiation
ideiation then leads to copyright
slashscripting
which then leads to voice over which
then leads to
video output which then leads to an
editor
or a composer so to speak that brings
everything together. the audio, the
video, the thumbnails. We need to add
that into the pipeline as well. Um, but
I want this working end to end. Of
course, we're going to take some time to
build out these specific bots that are
going to handle this pipeline, but let
me know if you have any questions. The
goal here is to automate a faceless
YouTube channel entirely through our
Grockbot
team
to where one weekly routine, sets in
motion
this pipeline
which ideates scripts, does the voice
over, creates the video outputs, splices
them all together, and then leaves me
with an MP4 that I can upload. Ideally,
we would want um a bot handling the
upload process as well, but the only
part of the blocker I want to be is me
reviewing the video and letting you know
thumbs up or thumbs down.
How funny is this quote from XAI talking
about why Grockbot is different? Reminds
me of our last 20 website from forever
we talked about. Oh yeah, I remember
what you were talking about. Last 20.
There's a huge difference between 90%
done and 100% done. Most AI gets you
there. Grockbot can finish the swing
because the work lands where a human
would put it in the actual tool. Yeah,
trying to remember. I I remember seeing
this post specifically.
Um, interesting. Did he create a new
team that I just don't see yet? Oh,
unassigned.
Yep. He's straight up creating an
entirely new team.
So, what we're going to do is add to
section.
Faceless
YouTube
pod.
Let's move this down here. I'm just
going to add Can I select more than one
agent at a time? Yes. Nice.
What? Oh, but then when you drag them,
it doesn't drag them all together.
That's rude.
Why are you going to allow us to select
multiple agents at once but not drag
them all? It's just rude.
That's just absolutely rude.
Yes, you can for sure. Um, there's a
little caveat there. There's a little
caveat. So, the bot itself is still
powered by Grock, but what's happening
is I installed Claude in the CLI on the
VM in this particular system. So,
anytime I want work done, I specify
which model or sometimes I don't specify
which model. Uh, but they essentially
prompt in that team chat what they want
from Claude. Claude then reports back
with its findings and then Grock the bot
itself, uh, whichever bot you're
powering essentially receives that
information and then handles it. So,
there's still that there's there's a
little asterisk there. On my other
system, I set up Peekaboo on my Mac to
where instead of having to install um
the CLI within the VMs.
It just prompts my system, right? And I
didn't have to authenticate anything
from that route. It was actually super
easy and convenient to do. Um but then
again, you're giving Grockbot access via
Peekaboo to your Mac OS, assuming you're
using Mac OS. But yeah, little asterisk
there, but you can get it to work.
All
right, we're just going to wait here. I
do need to
do something really quickly. So, BRB in
like two or 3 minutes.
Maybe we'll do some movie time in the
meantime.
And I was about to get up. Let me answer
this. Yeah, exactly. Exactly. Exactly.
Apparently though, we need to look into
usage through our cursor account. But
there might be another route to doing
it. Um, I just need to do some research
before sharing my findings. Shout out
Anthony aka Pokemon here in chat who
shared this account. Kind of brought
this to our attention from Chris
Mcconey. Apparently, he's finding that
his cursor account is using different
models. Now, the question is how's this
tied to Grockbot, right? U so I need to
look into this and like why this is
happening, how it's happening. Um but
nevertheless, yeah, but that's one
workaround I found. I did both routes as
I mentioned. Uh great question though.
Hope it's helpful.
Now, we're going to set up some movie
movie time while we watch these agents
work. And uh I need to go AFK for a sec
and like what their margins are. It did
that on its own. I I was kind of blown
away and it made these
reports on each product idea and was
like here's my recommendation. Always
ask for recommendations by the way here.
Like my bulletproof system for testing
business ideas with Crockbot is I want
to do this, this, or this. So, I've got
three ideas. Maybe they're similar,
maybe they're different. Go research
each of them. Figure out how I can price
it, what the cost is going to be, and
then give me your top three
recommendations on how to do the damn
thing. And it just works. Uh you it your
task should not be open-ended when
you're doing research or you're just
going to end up doing research for two
weeks.
>> That's fair. Okay. So, you think I mean
you think Shopify that there's still
opportunity in in finding products,
finding audiences for those products and
then having agent teams manage that
whole process. That's,
you know, that's interesting because
like a, you know, on on this podcast we
talk a lot about like digital products,
SAS, vertical, AI,
um, not so much on the Shopify side. So,
it it is it's just interesting. Yeah, I
I'm not like an ecom expert. This is an
experiment uh I'm working on. But I
think it's nice to work on something
physical because the cost of creating
digital things is going to zero. There's
no friction there.
>> Uh actually making something physical uh
a brand that you can just like build
over time. I think that I think that's a
smart thing to do.
>> And over here I see you have site
builder stack. What do you mean by that?
Yeah. So I I just have an agent uh on my
other account that has this simple stack
of Cloudflare um the CLI tools. Uh pork
bun is where I buy domains. So it can
log into that on its computer and buy a
domain. Um and then I use this thing
called Astro for for it's a framework
for creating static sites and it's
connected to my GitHub. And so basically
like I needed I needed a link for people
to book meetings with me. Uh and I
didn't want to pay for Calendarly. So I
said I got it set up on this stack and I
said go embed my three different Google
calendars on this and I was literally
sitting on the phone on the beach and it
would send me okay here's a mockup of
this and I would tell it okay make
another branch and do a mockup in this
style and it would come back and through
Cloudflare it would have separate
developer links that I could review the
different styles in and then when I said
okay I like that one it connected it to
the domain that that I owned. Uh, and it
went live all from the beach and I set
that up in two or three hours that that
that stack. Uh, and I'll probably use it
for a long time just for simple I need
something to exist on the internet. It
doesn't need to be perfect or pretty.
>> So, is you know is that just like
internal tools building internal tools
or is there an opportunity to build some
of these tools and sell it to other
businesses?
Yeah. So, what I would use this for is
landing pages. So, yeah, it's kind of an
internal tool. If I want to sell, say, a
local business on being a sponsor, I
could just say, "Hey, um, you know, make
a landing page that speaks directly to
this customer or directories." Uh, so
Astro specifically is really good at
directories. So, if you if you're
interested in the SEO game,
just drew through that. This is the
stack for testing and iterating.
And I would just have a research agent
that adds to the directory. So if you
want to make a directory of pickle ball
equipment, I would set up a coding agent
that uses this stack and a research
agent. And every day the research agent
investigates a new brand and writes a
great page on that brand as markdown and
then it hands it to the engineer who
posts it. Um, then you would, you know,
set up an agent that collects referral
links and does outbound to get to get
deals once you have the traffic. Uh, and
you could rinse and repeat that if
you're interested in doing kind of the
SEO directory game.
>> Yeah. So, I've done a few episodes on
directories in the past. I'll include
the latest one uh with my friend Frey
who does a good job of kind of
explaining why directories matter, how
it's a good like 10k, 25k, 100k a month
business. Um you know what is a
directory? It it's basically just
curated data on a website. So it's
you're picking a topic and you're like
here's the best xyz in this topic. And
in an AI world, we actually need
directories more than ever because um
there's just too much information. So
when it's curated,
uh it's helpful. Also, the whole like
SEO game is changing right now. Like
what's being, you know, like people are
going to chat GBT and they're going to
Google AI overviews
um to actually get, you know, con to
consume content and they need platforms
like that need directories. So I think
using
>> all right we are back and you know
what's funny right as I sat down
>> and in an AI world u no setup
>> I've done what is a directory it it's
basically just curated data on a website
so it's your just curated data on a
website
I don't know if I've defined that
can you guys define what a database
website is
>> and Frey who does a good job of kind of
explaining why directories matter, how
it's a good like 10K, 25K, 100K a month
business.
>> Let me rephrase that. A directory
website.
>> Um, you know what is a directory? It's
basically just curated data on a
website. So, it's you're picking a topic
and you're like,
All right. to simplify it, right? A
directory website is just basically a
website with curated data about a
specific topic or niche. In our
particular case for this project,
especially AI agents.construction,
it says it in the name of the domain,
right? I know we have a very specific
target audience, but I'm wondering if we
we don't broaden this to essentially be
the go-to directory for humans and
agents to learn how to build agents or
bots for
every job role in whether it's a GC
environment or construction companies of
all sizes, small, medium, large,
enterprise,
right? The goal is to make this
directory website, this domain,
uh, so recognizable and like so insanely
used to where then we can start to
monetize off of it.
>> Here's the best XYZ in this.
>> Hey, we'll see you in a year if this
works. I'm going to keep this running
cuz, you know, I'm doing work with the
construction industry, with a number of
different companies, uh, building agents
and managing agents, uh, to where I'm
not going anywhere out of this industry.
So, that's why I chose this industry for
this directory website. Just a heads up,
I didn't just randomly select an
industry
topic. And in an AI world, we actually
need directories more than ever because
um, there's just too much information.
So when it's curated,
uh, it's helpful. Also, the whole like
SEO game is changing right now. Like
what's being, you know, like people are
going to chat GBT and they're going to
Google AI overviews.
>> What is Pork Bun, though? I've never
heard of Pork Bun. It sounds funny. Pork
Bun.
Oh, it's a domain site.
Out of all the domains, Pork Bun.
Really?
there's just too much information. So
when it's curated,
uh
this right here, I built a directory
site. My site is backline local. Love
idea of it for restaurant repair. Right
up your alley, right? Like so like Yeah.
Oh, is it really? Is it like the new
namecheep that's super that's still
super niche and small?
It's helpful. Also, the whole like SEO
game is changing right now. Like what's
being, you know, like people are going
to chat GBT and they're going to Google
AI overviews
um to actually get, you know, con to
consume content and they need platforms
like that need directories. So, I think
using Grockbot as the basis for your
directory is absolutely brilliant. And
the be beautiful thing about a directory
is it's a once you build a strong
directory, it's a foundation. And yeah,
yeah,
yeah,
yeah. I know. I agree. I agree. It's uh
I'm reminded of it as well when I have
friends and family calling me for advice
and and questions. Um,
while it's easy,
you still have to work your way through
understanding how to communicate.
Um,
yeah, but you got to also understand
like listen, got a lot of respect for
Greg. Um, but his videos are a specific
are targeting a specific type of person
who would rather consume the content and
be wowed by the idea but not actually
take action on the idea. Sorry, Greg,
but that's like, you know, a lot of the
guests that you have on, they talk about
this idea and they get super hyped, but
they don't really give specifics.
They're super broad, right? They're
super like imaginative and like wow
factor, but wow factor does not
translate into execution, right? So like
having a guest on who shows the
execution of a database site, who shows
the proof that they're making a h
100,000 a month off of a database
website
or a directory, let me rephrase that.
Directory website, that's a whole
different realm, right? But that's not
who he's targeting, right? And you got
to understand that as well. And
>> you're getting thousands of visits a
month, let's say. And then you could
say, you know what, I want to build a
Shopify store. Actually, you know what?
I want to build a newsletter. Oh, you
know what? I actually want to build a
SAS. Or, you know,
>> I heard Pork Bun, but I've been doing it
through Versa. What?
>> But I want to build a mobile app. So,
it's actually a really good place for
people to start.
>> Yeah. And the cool thing about a
directory is
>> big fan of Greg. This was the first
video that really got me. Even this
ended up being just a bagel shop emailed
him and his Grockbot saw email and said,
"Hey, they want to be
Yeah. Yeah. Yeah. Yeah.
>> It ties in with this. Not a local
newsletter, but niche newsletter." So,
say we're doing that board game idea.
You could have it build out like board
games that are good for for parties
could be your niche directory. And you
could have a newsletter that talks about
that too. Uh you could have this on a
sub subdomain or a different domain than
your newsletter and uh they tie into
each other. So you can monetize uh by
linking people to the directory and have
a referral link there. Uh you could do
those warm leads from those referral
links could become sponsors in your
actual newsletter. Uh the I think the
the lowest barrier to entry with
Crockbot to start a business is probably
newsletter and then directory. I think
that stack goes together really nice.
>> Beautiful. All right. I think uh I think
we covered what we wanted to cover. Is
there anything anything else you want to
leave people with?
>> Yeah. Um I would do this try this for a
month. set up Grockbot for one project
and uh run it like a real team like we
went over and
let us like let me know like tweet at me
like did you make progress doing this? I
I like this is your sign to pick
something that you've been thinking
about and pursue it aggressively versus
trying a new tool and researching ideas
for four weeks. Uh so do this. Build a
Spotify or newsletter or a directory. uh
with Rockbot and I think you'll see what
I mean. This I haven't been excited
about a tool like this um
maybe like all year. Maybe this is my
favorite this year. Um
yeah, so tweet at me.
>> What's so interesting about this tool is
we've seen
>> Vendor Deso.
Yeah. Yeah. like the dream uh the sales
the seller of dreams, right?
The seller of dreams. Yeah, I could see
that. Yeah. Yeah.
>> A glimpse of this and other tools.
>> And don't get me wrong, like great show.
I get a lot of ideas from it, but you
also have to understand,
you know, it's a certain format for a
certain type of individual and it only
gives you c a certain amount of
information. Um,
it does not mean that that is tried and
true. It's an idea.
There's generally never any proof shown.
Rarely any proof shown. Um, depending on
the guest, I think it's it's all about
how honest the guest is and how much
they're willing to share. Um,
yeah, I'm sure Greg's regretted having
certain people on the show, you know.
Um, but that that's not his concern,
right? He he's he's making videos for a
specific target audience. So, got to
understand that.
One positive I will say about this video
is the ad. You make the platform calls,
not the agent.
What ad? I didn't see the ad. I must
have skipped through it.
Now where is is it unassigned? No, we
have everybody we need here. Question is
all right. I want to visually see this
pipeline that you guys have created. I
also want to visually see a rich table
here just showing me what's running,
whether it's daily, weekly, what day of
the week, uh each step of the pipeline,
who's handling what, and when I get
pinged to either green light it or red
light it.
Listen, I'll run an experiment for a
month. I'll find an API key. I hate it.
I hate the idea of it, but let's do it.
Also, I think, correct me if I'm wrong,
but would it be helpful if I send you a
few YouTube channels I think we should
draw inspiration from as far as quality
of output, narration style, education
style. Um,
because we're going to want to fine-tune
our process first. We're just seeing
what you come up with based off of the
minimal context I've given you. Um, but
I do want to fine-tune our process. I do
want to fine-tune the pipeline here.
Look at this. Great. So, Tuesday 11:00
a.m. Vault case voice script that sends
two ways. Frame pack, tapewave, cut mp4,
YouTube thumbnails,
Elon upload. Interesting. Interesting
how Elon handles the uploading. I feel
like I'm we're going to want an
uploader,
right? I don't know if I want to give
you uh Elon that responsibility.
So Elon fires it.
Okay. Okay. Block, no voice, block, no
stitcher. Okay, we got to fix those. So,
wait, are you blocked by me?
All right, let me uh go sign up for an
11 Labs account real quick.
11 Labs
developers
API keys.
What account am I in?
Oh, I have some credits in here somehow.
What is going What? Since when do I have
credits?
Do I use 11 creative, 11 agents, or 11
API? I think I know the answer, but
please let me know.
Um,
let's see. Settings.
Interesting.
I guess I'll keep that account live for
now.
Does it? H can I see this? Uh
these two visuals, the pipeline above
and the table imagined
or generated via Grock imagine please.
11 API. Got it. Create
key.
So
baseless
YouTube
experiment expire after call it 30-day
experiment. Create key.
Copy to clipboard.
It should detect it. So, I'm just going
to let it do its thing.
All right.
Yeah. So, we started off I I drew
inspiration from this Greg Eisenberg
video. So, we did a little bit of movie
time. Uh, but this video talks about
creating a directory website. Um so I
have built a directory team. I don't
know why I called it database. Let's
called it a directory
factory where so we primarily service
the construction industry right now. Um
so I have retainers in place, contracts
in place that
um primarily create, manage um and
maintain AI agent systems for small to
midsize construction based companies and
GC's for that matter. Um so I chose our
target audience.
I have a certain uh domain that I squat
on squatted on a while ago called AI
agents.construction.
So, we've built out this team to
automate the process of building out
that directory for us based off of that
video. So,
how is YouTube guidelines reacting to AI
videos or shorts? You know, they're
supportive of it, but they're also
making changes for, you know, how they
manage AI slop in general. I'm one thing
I don't want to do is contribute to AI
slop. Let's be very, very clear. Um,
but yeah, that's I mean that's ever
evolving, right? They allow it, but I'm
not going to just blindly upload videos,
right? I want to be the blocker. I want
to be that um that final handoff before
I say upload, right? Uh but it's another
idea that we have. So, we've built out a
team to complement that, a faceless
YouTube pod. And then we built out
another team, an experiment pod that
automatically builds us a product based
off of uh certain pain points they
discover across socials using the last
30 days skill. So yeah, as far as the
the title of this live stream, I was
very clear if you watched the very
beginning about it uh beginning of the
live stream. This is not a promise that
I'm going to teach you how to make x
amount of dollars. We're just
experimenting from the perspective of
making money, right? So, it's a longer
term gonger term game with this. And
then I think next week we're going to
build a trading bot live. I think that's
going to be the move for next week. Uh
because I want to bring in some expert
traders to do that. Uh because for me to
build a trading bot, not having
foundational knowledge of if then
statements, what to do when, right? Like
how to trade based off of certain
variables and metrics. um I would not be
as effective making a trading bot as
somebody in the industry, right? So stay
tuned. Next week I'm going to plan to to
build a trading bot live. I'm not sure
which day of the week yet, but we'll do
it next week.
So for fun, I'm going to create an agent
to do tracker of Greg's videos. If his
guest wears a hat, he does.
If they don't, he doesn't. I want to see
percentage. Oh, really? That's funny. I
never noticed that. That is funny.
Looks like they're not done with the
work yet here.
All right. Where What's the status on
the rebuild?
I'm just getting impatient here.
Oh, wow. Vault just came up with a ABC
vault case. Oh, I love that structure of
how it's doing it.
Solid.
Key is scoped. Not because it's wrong.
Key is set, but it's locked down.
Edit that key and turn on text to speech
and voices. Text to speech and voices.
Let's see.
Edit text to speech
and voices
read.
Same key is fine.
Done.
Okay, great.
Awesome.
Now, when did I reserve this channel?
I'm very curious. Let's go.
Create view channel.
Did I really do this in December of
2024?
Huh?
Also, Elon, shouldn't we have an
uploader
or do you really want to take on the
responsibilities? Is that really your
role to upload as well?
Here's our YouTube URL that we reserved
this domain a while back.
Okay.
So, since I started looking
when I first toggled this, this was 11%
at the start of the stream. So, we've
we've burned through 10% already.
Now, how much of it? I'm going to
monitor this. So, chances are I'm going
to record a video a little bit later
today. Still not sure the topic yet. Um,
and then I'm just going to let this run
day by day and I'm going to check in
every 24 hours to see what that
percent's at.
I guess instead
Oh, yeah. Look at that. It did a pretty
good job, Derek. Good idea.
Good idea.
All right. So,
what I'd like to also do is set a daily
checker for usage. Can you do that? Can
you see exactly how much of our weekly
usage we're using every 24 hours so that
I can start to manage and and just keep
an eye keep tabs on how much
usage we're burning through based off of
all of the different pods and teams
we've created and routines that we've
created.
Let's see what he says here.
I appreciate you. Thank you. Thank you.
Thank you for the uh the donation there
as well.
I'm running a manufacturer
I'm running a manufacturing theme in
India building been building systems
with AI on Google sheet
ordering monitoring stuff but I've
learned the system works only if the
people who are using it do it properly.
Okay. Okay. So,
are you looking into using bots to help
automate the process or maybe creating a
different solution uh to make sure that
it's easier for the people to use the
system properly? Is that what I'm
understanding?
Kayleie burn check. I'll find the
reason.
Oh, nice.
Very nice.
Am I going to be disappointed? Okay,
looks better.
But it's just the header. Like what
happened down here?
Okay,
there needs to be like
feel like we need to get even more clear
here.
Oh, you already know that instantly
happened. You must be delayed for a
manufacturing theme. Where do you feel
Crockpot can come in handy? ERP is
business central. So MCP I can give
Grockbot Eagles view on all ops. Yeah,
you know, I like only speaking on things
I truly know and have experience in. So
as far as how it could come in handy in
your industry,
let me think on that right because I
don't have a manufacturing background.
So what what is in your brain is
different than is what in my brain like
as far as context related to it. Um
now what size facilities are you more
talking about from the
um
how do I how do I word this? So like
from the perspective of the staffing
that's running this manufacturing
facility whether it's small, medium or
or rightme is um
forgive me I don't know that
abbreviation off the top of my head but
I'd imagine it means small to mediumsiz
enterprises
just knowing what I know about SMB
meaning small to midsize companies
businesses.
So your ERP is business central. So
Sophia MCP I can go with GRA.
So you answered it right there. Forgive
me. Um
yeah, I would honestly, this is going to
be super surface level, but start with
one responsibility. Start with
using Grockbot a bot or number of bots
from the perspective of what's a tedious
repeatable task that you do that you
think is a waste of your time. automate
that with a bot or two or three and then
rinse and repeat. You know what I mean?
I think it's very easy to get to fall
into the trap of
wanting to do too much. Visualizing this
big grandiose vision of how it can
automate your life and your company and
and and your business or the company you
work for, but not taking the steps
needed to like build up to that. I'm I'm
a big advocate for start simple, start
small, start with one thing. And the
best thing to start with is is something
you know well, right? We all know what
we waste our time with. We all know
tedious parts of our dayto-day, whether
it's personal or professional or working
for another company of man, I really
wish I didn't have to do this task every
single day or every single week or every
single month. automate that thing first
and then build from there. And then
you'll start to get little unlocks of
more ideas of how to utilize bots or
agents for that matter um for your
specific industry. And and um yeah, hope
that's helpful. Hope that's helpful.
All right, so they're still working.
This is great. We're burning through
usage. We'll see how this ma maintains.
Uh pulling the episode now.
Oh, interesting. Claude is doing some
work.
Okay. So,
I asked you to update the text stack of
this directory website AI
agents.construction
and it looks like you half did the job.
The nav header is fixed. Great. But the
website itself
was not touched, right? All the contents
also that nav header should be sticky.
So as you scroll, it should still be
visible. I think the only scenario where
it's not visible is an article. an
article for a specific job or a specific
task to where as you scroll then it
shows those breadcrumbs of sections one,
two, three, and four that stay sticky
for a user to easily navigate. On second
thought, keep the nav header and
breadcrumb portion of the page sticky,
but then as a user scrolls, there should
be a
I don't know on the left or right column
1/4 column um essentially a site map or
a glossery for that specific article for
that specific knowledge resource.
allowing a user to easily click between
job and this is just an example on one
of the pages between 1 2 3 4 5 6 right
um
I think that's a great starting All
right.
Yo, what's up, Lego
geeks? Nice. Nice. Yes, but I don't It's
not fully automated, right? But yes, a
lot of what I do is automated um to
where I can just focus on live streaming
and making videos. I still record the
videos. I still edit the videos cuz I
enjoy that process. But a lot of those
tedious parts of my dayto-day are
automated primarily with OpenClaw and
now a touch of Grockbot. Uh Buzz I use
from a collaborative perspective. I
don't really automate too much through
it. Um, and then Hermes, I build out
Hermes agent systems for uh companies.
So
I I don't only build them, I maintain
them as well. I manage them. Um, there's
a lot of different options from that
perspective.
Fix map. You know, I don't have any
input on that.
I wish I had more insight into that. I
can look into it, but I I have nothing
for you at this moment.
Wish I could be more helpful.
Hey, I'm happy to help. I'm happy to
help. Do you have any agents or bots
running through X or other channels
getting you news or updates for examples
on AI? Yes. Yes. So, my OpenClaw agent,
Muddy, my OG, I get a morning brief
every morning. There's an industry pulse
checker. I also created something
similar but specifically targeting X,
what's trending on X. And then based off
of what's trending on X, I have the last
30 days skill run uh based off of that
topic or, you know, one of the topics it
chooses from. And then it cross
references with Google Trends, YouTube
search data, and then it tries to find a
nice middle ground to then feed me video
ideas based off of what are people
talking about X, is there a similar like
crossover on YouTube and then what can
we find in the middle to create content
around, right? So, I have a lot of
different agents from that perspective,
but I really only rely on two for, you
know, reading morning news. And then, of
course, I still scroll X on on certain
time blocks of the day just to see what
people are talking about, just for my
own eyes, my own, you know, cuz I don't
like automating the posting of content.
I'm still that final human that presses
the button always. Uh, except for
updating certain older videos. I have
agents doing all that heavy lifting for
me. Uh for the first few weeks though, I
did monitor their outputs to say, you
know, save it as a draft first so that I
could see how you how you write and then
I make my edits. And then at the end of
each day, there was an automation to
build a voice profile based off of the
differences of what they wrote versus
the differences of what I edited before
I hit post. That way, the model or the
agent itself will train itself with
time, right? Based off of you just being
a little patient and not letting it post
on your behalf blindly. Uh cuz I'm
really against contributing to this
world of AI slop personally. So, I'm
trying to to make sure that everything I
put out there has my touch, right, and
not just blindly get sent.
How bad are usage limits? I mean,
they're bad if you just work non-stop
out of it. They're pretty bad. Um, I
hope we get some more clarity on that
front
as far as usage is concerned. And I
think, you know, they just handed out a
reset what, two days ago, if I'm not
mistaken.
So, but it's not that bad. Not, you
know, it's bad, but it's also not that
bad. It's it's all a you know from the
perspective of
if you know what you want to accomplish
with Grockbot and your bots and
automations then I think it's actually
pretty efficient if you just set things
up on a reoccurring basis to do the
work. Uh but if you're experimenting
building teams trashing teams setting
automations realizing that that's not
the move and trashing those that's when
the usage burns up. Also, if you have it
scanned thousands of emails, that's when
usage burns up. I've personally
experimented with all this. So, but if
you come into a Grockbot setup knowing
what you want to configure, knowing that
the task is simple and you're just
setting up one at a time, usage isn't
that bad. Uh, but keep in mind, I am on
the maxed out accounts, so I'm not sure
how it is on the 20 or $60 a month
accounts.
Yeah. Yeah. Yeah. Exactly. And they
still do. They still do. They still do.
I still read that brief every single
morning. It's awesome. It's tried and
true.
Yeah, reach out. I'm not on LinkedIn
often, but uh when I am, I'll make sure
to uh touch base on there.
Yeah, my first focus is
getting more consistent on X and then
we're going to get a little bit more
consistent on um LinkedIn. All in due
time. I'm not rushing, right? I'm not
losing over sleep into, you know, being
worried that I'm not being active on
everything all the time. You know, my
main focus is YouTube and I'm slowly
expanding out, right?
Hey, let's go, Maestro.
Okay.
Yeah, I think we need to have, you know,
maybe one or two different template
styles of
the posts themselves cuz I don't get me
wrong, I love the way that it's just
super informative here. I feel like this
is while it's very basic from a UX
perspective or a UI perspective, it is a
great user experience depending on the
target audience, but then it's also a
great agent experience, which I believe
is being dubbed AX. Um, I feel like it's
very easy for an agent to dissect all
this information and create whether
an agent, a bot, a team, um, an
automation based off of,
you know, each one of these roles. Um,
so I great job on that front. I'm just
wondering if we can't just slightly
improve uh, based off of this new tech
stack.
I'm very curious to see what they're
doing. Uh,
never ask me. I should never be a
blocker. Just prioritize based off of
what came in first.
Okay.
Damn, it's already 11:30. I may have to
uh
we're going to have to switch gears here
pretty soon in our day.
My coalition is working better than
ever. Got Fable to tweak their pipeline.
Nice. Nice. I can do my best. I can do
my best.
Check the uh YouTube uh the YouTube link
takes you to our website which has our
our LinkedIn uh links on there.
Everything you need is through that
link. I'm launching a bamboo socks
company. What do you feel about the
current AI proudness? It's worth
automation, for example, driving demand,
marketing, etc. H
I'm thinking right like like there's a
lot of different routes you can take,
but I'm going to go back to my original
comment. Start simple, right? I would
first maybe if you're launching a bamboo
socks company, I'm sure there's some fun
facts about bamboo that the average
person doesn't know about, right? So,
getting and this is going to trigger
other events, but getting uh a deep dive
into bamboo
every single day, right? Maybe new
things you didn't know about bamboo uh
and then creating some type of
educational style newsletter about
bamboo. Obviously, the goal is to
promote your socks company, but make it
focused around the the material itself
first. Build an audience through that.
And
I mean, then you get a like endless
content ideas from creating the
automation, doing research to anything
bamboo related, whether it's bamboo in
the environment or bamboo in this
country, bamboo in that country, right?
creating a directory, so to speak, of
all the Bamboo alpha worldwide based off
continent, uh, based off of fun use
cases, based off of DIY hacks, right?
Um, just creating a content engine
around bamboo, uh, and then just start
automating little bits and pieces of
that that pipeline.
That's probably what I would do. Bamboo
socks does sound wild. Does sound wild.
Yeah. Welcome back, Maestro.
Welcome back, Maestro.
Yeah, if you just go to our website and
scroll down to the footer, you'll find
it. It's either in the header or the
footer. It's there for sure.
Drop the bamboo alpha. I know, right? I
know.
Hey, but I hate to uh abruptly kind of
end things. I do have to switch gears
here in about 20 minutes and um I want
to prep for this call. So,
so yeah, I think we're going to end it
here. I think it's been a fun one. Today
was just all about experimenting with
Grockbot. Uh I'm also going to have to
record a video a little bit later as I'm
not sure I'm going to use two of the
videos, two of the other videos I
recorded this week. So stay tuned. I'm
either going to drop it today or
tomorrow.
Uh, but I just want to thank everybody.
Thank everybody for hanging out with me
on your Friday. Uh, wherever you are,
wishing you a good morning, good
afternoon, and good evening. I just want
to take time to say thank you.
No, you have to go to our website. So,
go to clearma.ai.
The difference is quality in normal
cotton polyester bamboo is the
difference. Okay. Okay. Yeah, I'll have
to try it out myself, too. But hey,
thank you, Maestro. Thank you, Lego.
Thank you, Sar Sarvesh. I I hope I'm
pronouncing that correctly. Thank you,
Derek. Thank you, Alexa
or Alex A. I can't tell. Uh, thank you
for the kind comment. Really appreciate
the donation.
Abishek. Abishek. I hope I'm I'm hope
I'm not butchering this. Um, thank you,
Matos.
Thank you, Nikolai.
Thank you, Pokemon.
And I can't keep scrolling, so I can't
uh can't thank anybody else. But yeah,
just want to say thank you. Hope
everybody has a great weekend. Don't
forget to go touch some grass.
Disconnect. Uh, in a world of AI and
tech, it's nice to get away from the
computer screen every now and again. You
know what I mean? Um,
and yeah, we'll be back at it Monday,
same time. Until then, happy building.
Cheers.
Hey, happy to help. Thanks for being
here.
Ask follow-up questions or revisit key timestamps.
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.
Videos recently processed by our community