WEBVTT

1
00:00:00.000 --> 00:00:02.240
How can you make money from OpenClaw?

2
00:00:02.240 --> 00:00:05.520
Like how can you spin up these OpenClaw instances,

3
00:00:05.520 --> 00:00:08.480
these sub agents, these digital employees

4
00:00:08.480 --> 00:00:11.360
that can go out and make you money while you sleep?

5
00:00:11.360 --> 00:00:12.520
Is it even possible?

6
00:00:12.520 --> 00:00:15.000
Well, in today's episode, I brought on Nick

7
00:00:15.000 --> 00:00:17.320
and he shows a tactical tutorial

8
00:00:17.320 --> 00:00:20.400
for how to spin up multiple OpenClaw machines

9
00:00:20.400 --> 00:00:22.080
in a virtual instance,

10
00:00:22.080 --> 00:00:25.680
how you can be automating tasks on Upwork

11
00:00:25.680 --> 00:00:28.000
and these boring business automations

12
00:00:28.080 --> 00:00:31.440
and how you can actually make money from OpenClaw.

13
00:00:31.440 --> 00:00:33.920
If this doesn't get your creative juices flowing

14
00:00:33.920 --> 00:00:35.160
for the future of SAS,

15
00:00:35.160 --> 00:00:36.800
how people are gonna make money

16
00:00:36.800 --> 00:00:39.440
and how to actually use OpenClaw

17
00:00:39.440 --> 00:00:42.400
from not just acute little use cases,

18
00:00:42.400 --> 00:00:46.520
but actually money making opportunities

19
00:00:46.520 --> 00:00:47.960
then I don't know what will.

20
00:00:47.960 --> 00:00:49.680
I had such a good chat with Nick.

21
00:00:49.680 --> 00:00:51.800
It got my creative juices flowing.

22
00:00:51.800 --> 00:00:53.520
They think it will yours too.

23
00:00:53.520 --> 00:00:56.240
And this is, I think, one of Nick's first podcasts.

24
00:00:56.240 --> 00:00:59.640
So give him a like and comment to juice him up

25
00:00:59.640 --> 00:01:01.560
because he shared that sauce.

26
00:01:01.560 --> 00:01:04.560
I couldn't be more excited to have Nick on the pod.

27
00:01:13.120 --> 00:01:15.320
He's one of my go-to people

28
00:01:15.320 --> 00:01:17.680
when I have questions about OpenClaw.

29
00:01:17.680 --> 00:01:20.200
Nick, by the end of this episode,

30
00:01:20.200 --> 00:01:22.480
what are people gonna learn?

31
00:01:22.480 --> 00:01:25.280
Yeah, people are gonna learn that OpenClaw

32
00:01:25.280 --> 00:01:27.840
is more than just a personal assistant.

33
00:01:27.840 --> 00:01:30.240
You can actually deploy this into businesses.

34
00:01:30.240 --> 00:01:32.600
You could drive the actual business outcomes,

35
00:01:32.600 --> 00:01:35.960
generate revenue off of OpenClaw as an opportunity.

36
00:01:36.920 --> 00:01:38.800
And yeah, we're seeing it on X

37
00:01:38.800 --> 00:01:41.000
like people who are deploying OpenClaw

38
00:01:41.000 --> 00:01:46.000
for kind of executives or individuals who are super busy.

39
00:01:46.720 --> 00:01:48.760
They're making thousands of dollars,

40
00:01:48.760 --> 00:01:50.080
setting OpenClaw up,

41
00:01:50.080 --> 00:01:51.680
getting it up and running for these people

42
00:01:51.680 --> 00:01:52.640
and managing it for them.

43
00:01:52.640 --> 00:01:54.760
So I think there's a huge opportunity here.

44
00:01:54.760 --> 00:01:57.640
And yeah, just excited to jump in.

45
00:01:57.640 --> 00:01:59.560
Cool. And before we get going,

46
00:01:59.560 --> 00:02:01.560
I need you to make a commitment to me

47
00:02:01.560 --> 00:02:03.200
and to the person listening or watching,

48
00:02:03.200 --> 00:02:06.800
which is I need you not to hold back any sauce.

49
00:02:06.800 --> 00:02:08.760
I don't wanna know just about the opportunity.

50
00:02:08.760 --> 00:02:11.240
I wanna know how are people doing it tactically

51
00:02:11.240 --> 00:02:14.280
and by the end of this episode, what I want is

52
00:02:14.280 --> 00:02:15.680
for people to take away,

53
00:02:15.680 --> 00:02:18.200
like I want people to know how they can actually

54
00:02:18.200 --> 00:02:19.400
make a dollar from this.

55
00:02:19.400 --> 00:02:21.480
And is that a commitment, Nick,

56
00:02:21.480 --> 00:02:24.360
that you are willing to make to us?

57
00:02:24.360 --> 00:02:26.600
Absolutely, absolutely.

58
00:02:26.600 --> 00:02:28.600
I'm not gonna hold back anything.

59
00:02:28.600 --> 00:02:33.600
I've in fact, I think OpenClaw is a tool

60
00:02:33.600 --> 00:02:36.120
that allows us to be able to do the things

61
00:02:36.120 --> 00:02:38.200
that we have always been able to make money

62
00:02:38.200 --> 00:02:40.120
from like automation with AI,

63
00:02:40.120 --> 00:02:41.360
but do it even better.

64
00:02:41.360 --> 00:02:43.080
And I'm gonna show you how to get it all set up

65
00:02:43.080 --> 00:02:45.480
so you can do that and the wedge to get going.

66
00:02:45.480 --> 00:02:47.520
So let's do it.

67
00:02:47.520 --> 00:02:48.320
Yeah.

68
00:02:48.320 --> 00:02:50.560
So I guess jumping right in,

69
00:02:50.560 --> 00:02:52.880
as far as like getting set up with OpenClaw,

70
00:02:52.880 --> 00:02:57.040
you can see here, this is Orgo, this is our startup.

71
00:02:57.040 --> 00:02:59.520
You don't have to use Orgo to get started with OpenClaw.

72
00:02:59.520 --> 00:03:01.800
There's full disclaimer, this is what I'm using.

73
00:03:01.800 --> 00:03:05.000
And what I'm gonna do is I have a project here.

74
00:03:05.000 --> 00:03:07.560
You can see I have a couple projects

75
00:03:07.560 --> 00:03:09.440
and I have Greg, I set you up a project.

76
00:03:09.440 --> 00:03:13.840
I hope you enjoy your five computers.

77
00:03:15.120 --> 00:03:17.200
And so you can imagine, Greg,

78
00:03:17.200 --> 00:03:19.440
let's say you're a business owner

79
00:03:19.440 --> 00:03:22.200
and you have a busy life, you know?

80
00:03:22.200 --> 00:03:23.880
You've got all the podcasts going on,

81
00:03:23.880 --> 00:03:25.480
you have all these businesses you're running,

82
00:03:25.480 --> 00:03:28.640
the agency, the idea of browser, all this stuff.

83
00:03:28.640 --> 00:03:31.600
And you need help automating some stuff.

84
00:03:31.600 --> 00:03:33.360
So what I'm gonna do is I'm gonna come in,

85
00:03:33.360 --> 00:03:35.400
I'm scrappy Nick, I'm gonna come in

86
00:03:35.400 --> 00:03:37.560
and I'm gonna help automate some things

87
00:03:37.560 --> 00:03:40.080
in your business, in your life.

88
00:03:41.080 --> 00:03:42.800
You know, as a busy executive.

89
00:03:42.800 --> 00:03:44.480
And I'm gonna get you set up with OpenClaw.

90
00:03:44.480 --> 00:03:47.320
So when you open up a computer here,

91
00:03:47.320 --> 00:03:48.680
you can see I have it open.

92
00:03:48.680 --> 00:03:52.120
This is the CloudBot1 computer I made for you.

93
00:03:52.200 --> 00:03:55.880
If I actually just type in OpenClaw to UI,

94
00:03:55.880 --> 00:03:58.760
this will open up OpenClaw in the terminal.

95
00:03:58.760 --> 00:04:01.000
And you can see I actually already started like,

96
00:04:01.000 --> 00:04:05.480
hey, I'm Greg Eisenberg and it's all ready to get set up.

97
00:04:05.480 --> 00:04:08.360
And so actually what I could do is,

98
00:04:08.360 --> 00:04:09.960
I could invite you to this project

99
00:04:09.960 --> 00:04:11.440
and then you'd be able to do this as well.

100
00:04:11.440 --> 00:04:13.600
Like in your terminal, you'd be able to spin this up

101
00:04:13.600 --> 00:04:15.320
and now you're talking to OpenClaw.

102
00:04:15.320 --> 00:04:17.600
So super easy to get set up.

103
00:04:17.600 --> 00:04:18.960
Once again, you don't have to use Orgo,

104
00:04:18.960 --> 00:04:19.920
you can use whatever you want.

105
00:04:19.920 --> 00:04:24.800
You could use I know Manis just dropped their own version

106
00:04:24.800 --> 00:04:26.840
of like one click deployment OpenClaw.

107
00:04:27.840 --> 00:04:30.080
Also Kimi launched their version as well.

108
00:04:30.080 --> 00:04:31.200
X is down right now.

109
00:04:31.200 --> 00:04:33.320
So we can't actually pull it up on Twitter or anything

110
00:04:33.320 --> 00:04:35.480
but Kimi launched their version.

111
00:04:35.480 --> 00:04:38.400
And so there's all these options as far as getting started.

112
00:04:38.400 --> 00:04:40.120
You could use a Mac Mini, whatever.

113
00:04:40.120 --> 00:04:43.200
So the key here Greg with OpenClaw

114
00:04:43.200 --> 00:04:45.160
and actually creating money from it

115
00:04:45.160 --> 00:04:47.520
is to have the wedge to know

116
00:04:47.560 --> 00:04:51.080
what is the specific use case in a person's business

117
00:04:51.080 --> 00:04:52.880
that we're going to automate like first.

118
00:04:52.880 --> 00:04:57.280
Because when you see OpenClaw on Twitter,

119
00:04:57.280 --> 00:04:59.520
it's very much a personal assistant.

120
00:04:59.520 --> 00:05:00.440
It's exciting.

121
00:05:00.440 --> 00:05:01.680
It's fun.

122
00:05:01.680 --> 00:05:05.240
But all the demos that go viral, including me, I get it.

123
00:05:05.240 --> 00:05:07.280
I'm guilty of this too.

124
00:05:07.280 --> 00:05:12.040
All the demos that go viral are a little bit kind of toyish.

125
00:05:12.040 --> 00:05:13.680
They're a little flashy.

126
00:05:13.680 --> 00:05:17.400
But the real power is in finding the thing

127
00:05:17.400 --> 00:05:20.880
that actually drives business outcomes, saves time

128
00:05:20.880 --> 00:05:23.200
for a business, finding that,

129
00:05:23.200 --> 00:05:25.080
and building the automation around that.

130
00:05:25.080 --> 00:05:27.840
So I have something running here.

131
00:05:27.840 --> 00:05:31.760
This is my OpenClaw doing looking up products

132
00:05:31.760 --> 00:05:36.040
for a business that I deployed for,

133
00:05:36.040 --> 00:05:39.000
this is a promotional distributorship.

134
00:05:39.000 --> 00:05:43.120
And what this is doing is it's looking up products

135
00:05:43.120 --> 00:05:46.240
and actually downloading all the product information.

136
00:05:46.240 --> 00:05:48.000
And then parsing all that information,

137
00:05:48.000 --> 00:05:50.320
there's all these reports that needs to download.

138
00:05:50.320 --> 00:05:54.480
And then uploading that into a Zoho CRM,

139
00:05:55.480 --> 00:05:58.440
so it can essentially create a central source

140
00:05:58.440 --> 00:06:00.120
of truth for this client.

141
00:06:00.120 --> 00:06:04.440
So this is a perfect example here of actually creating

142
00:06:04.440 --> 00:06:08.720
an agent that OpenClaw deploys to be able

143
00:06:08.720 --> 00:06:10.400
to automate something end-to-end.

144
00:06:11.400 --> 00:06:15.760
So just to recap, are we all good so far?

145
00:06:16.680 --> 00:06:19.600
Yeah, so a few things I just want to talk about.

146
00:06:19.600 --> 00:06:23.400
So one thing is when you showed that Orgo screen,

147
00:06:23.400 --> 00:06:25.480
you had like five machines running.

148
00:06:25.480 --> 00:06:28.640
So what's interesting is,

149
00:06:28.640 --> 00:06:32.440
I've got my Mac mini going, I've got one instance.

150
00:06:32.440 --> 00:06:36.240
So using something like this is cool

151
00:06:36.240 --> 00:06:39.080
because you can have multiple instances going, right?

152
00:06:39.080 --> 00:06:41.760
And you can see them all in one screen, in one view.

153
00:06:41.760 --> 00:06:42.600
So that's really cool.

154
00:06:42.600 --> 00:06:45.080
That was sort of in a hot moment for me.

155
00:06:45.080 --> 00:06:46.920
Yeah, absolutely, yeah.

156
00:06:46.920 --> 00:06:49.680
It's like everyone, you know,

157
00:06:49.680 --> 00:06:52.160
as far as where you deploy your main OpenClaw,

158
00:06:52.160 --> 00:06:55.480
you can see here, I like, I starred the main one.

159
00:06:56.520 --> 00:06:58.920
And so that's like, you could have that wherever,

160
00:06:58.920 --> 00:07:00.720
but what people don't realize is,

161
00:07:00.720 --> 00:07:02.960
OpenClaw can spawn sub agents.

162
00:07:02.960 --> 00:07:06.880
And I think this is gonna be huge for people who start,

163
00:07:06.880 --> 00:07:10.600
like right now we're at the phase of having one OpenClaw.

164
00:07:10.600 --> 00:07:11.680
It's gonna happen quickly.

165
00:07:11.680 --> 00:07:13.360
You're already seeing on it on Twitter,

166
00:07:13.360 --> 00:07:16.680
memes about like, oh, what if you have a 10 Mac minis

167
00:07:16.680 --> 00:07:19.000
or all the Mac studios are being bought out?

168
00:07:19.000 --> 00:07:20.400
So this is happening faster.

169
00:07:20.400 --> 00:07:23.520
You're gonna want 10 OpenClaws, you know, 100 OpenClaws.

170
00:07:23.520 --> 00:07:25.120
Right now you can have one OpenClaw

171
00:07:25.120 --> 00:07:28.280
and just have it spawn up to I think eight sub agents.

172
00:07:28.280 --> 00:07:31.400
And each sub agent could have its own computer.

173
00:07:31.400 --> 00:07:32.920
And you can do this like, this is why,

174
00:07:32.920 --> 00:07:35.040
this is like kind of where Orgo shines

175
00:07:35.120 --> 00:07:36.880
as you can spin up multiple computers

176
00:07:36.880 --> 00:07:39.680
for each individual sub agent of your OpenClaw.

177
00:07:39.680 --> 00:07:42.880
And in this case here, I had it looking on Upwork

178
00:07:42.880 --> 00:07:45.720
for actual things that we could automate with OpenClaws.

179
00:07:45.720 --> 00:07:47.920
It's like a little hack here as far as like,

180
00:07:47.920 --> 00:07:49.600
well, I want to make money with OpenClaw.

181
00:07:49.600 --> 00:07:52.320
Well, oh, I don't know any business

182
00:07:52.320 --> 00:07:54.520
that I can reach out to that could automate stuff for.

183
00:07:54.520 --> 00:07:56.960
A great place to start is Upwork

184
00:07:56.960 --> 00:07:59.640
because there's jobs on Upwork that are literally posted.

185
00:07:59.640 --> 00:08:01.320
They're asking you, they're like,

186
00:08:01.400 --> 00:08:05.880
I want to pay $500,000, $1,500, $3,000, $20,000

187
00:08:05.880 --> 00:08:07.400
for this AI workload.

188
00:08:07.400 --> 00:08:10.880
And you can, I spawned a sub agent, it went viral into it.

189
00:08:10.880 --> 00:08:13.400
I spawned sub agents to go find all these jobs

190
00:08:13.400 --> 00:08:16.200
and then build out little demos for each of them.

191
00:08:16.200 --> 00:08:17.600
And then we picked the best one.

192
00:08:17.600 --> 00:08:20.160
And okay, let's apply for that proposal with that.

193
00:08:20.160 --> 00:08:24.360
So that's a little tidbit there on Upwork and all of that.

194
00:08:24.360 --> 00:08:26.480
So, which is just kind of hilarious

195
00:08:26.480 --> 00:08:29.840
because I mean, Upwork is designed

196
00:08:29.840 --> 00:08:31.600
for human beings to complete work, right?

197
00:08:31.600 --> 00:08:33.600
It's not designed for machines,

198
00:08:33.600 --> 00:08:36.760
let alone multiple machines to complete work.

199
00:08:38.160 --> 00:08:41.000
But I mean, as long as the quality of work is good,

200
00:08:42.800 --> 00:08:44.840
customers can be happy, right?

201
00:08:44.840 --> 00:08:49.840
Yeah, and I think it's good to treat it as a starting point.

202
00:08:51.080 --> 00:08:53.440
If you can save time preparing a proposal

203
00:08:53.440 --> 00:08:57.640
for a job on Upwork, it's like, what is that worth?

204
00:08:57.640 --> 00:09:01.160
And if you could do just a 100x volume, what is that worth?

205
00:09:01.160 --> 00:09:03.600
So there's a couple of things on that of like

206
00:09:03.600 --> 00:09:06.800
the parallelization of work with OpenClaw.

207
00:09:06.800 --> 00:09:10.440
So there's like, could you have 10 OpenClaw's working

208
00:09:10.440 --> 00:09:13.000
on a given task and it breaks up that task

209
00:09:13.000 --> 00:09:14.440
into 10 sub tasks.

210
00:09:14.440 --> 00:09:17.000
And so each OpenClaw does one of those sub tasks.

211
00:09:17.000 --> 00:09:19.400
That's one way of having parallelization.

212
00:09:19.400 --> 00:09:23.640
But another way is to have 10 OpenClaw's working

213
00:09:23.640 --> 00:09:27.040
on the same task, just 10 different instances of it.

214
00:09:27.040 --> 00:09:28.600
And that was kind of like what I was doing here.

215
00:09:28.600 --> 00:09:33.080
It's four different instances of the one OpenClaw,

216
00:09:33.080 --> 00:09:35.360
or four different OpenClaws doing the same thing

217
00:09:35.360 --> 00:09:37.720
of looking up different jobs on Upwork

218
00:09:37.720 --> 00:09:38.680
that they can apply to.

219
00:09:38.680 --> 00:09:40.680
So that's kind of an interesting topic there.

220
00:09:42.560 --> 00:09:44.960
But yeah, so like as far as OpenClaw goes,

221
00:09:44.960 --> 00:09:47.160
it's a huge opportunity.

222
00:09:47.160 --> 00:09:49.960
I just want to like throw this in here.

223
00:09:49.960 --> 00:09:52.960
And recent Horowitz talks about computer use agents.

224
00:09:52.960 --> 00:09:55.080
I view OpenClaw as a computer use agent.

225
00:09:55.080 --> 00:09:58.400
You're giving an agent a computer and it's able to do,

226
00:09:58.400 --> 00:10:00.000
it's able to use that computer.

227
00:10:00.000 --> 00:10:01.560
It's like, it's a computer use agent.

228
00:10:01.560 --> 00:10:02.560
But that's half the story.

229
00:10:02.560 --> 00:10:05.200
The other half of the story is for it to be able to like,

230
00:10:05.200 --> 00:10:09.800
click around, actually operate a graphical user interface

231
00:10:09.800 --> 00:10:12.520
on like legacy softwares and systems.

232
00:10:12.520 --> 00:10:16.320
And so you can imagine like, you know,

233
00:10:16.320 --> 00:10:18.760
in here this automation I built,

234
00:10:18.760 --> 00:10:21.600
this is navigating a legacy platform for this client

235
00:10:21.600 --> 00:10:23.160
that doesn't have any clean APIs

236
00:10:23.160 --> 00:10:26.200
and it's able to click in, download reports,

237
00:10:26.200 --> 00:10:29.360
and actually, you know, be the universal API

238
00:10:29.360 --> 00:10:30.400
to be able to solve problems

239
00:10:30.400 --> 00:10:32.520
that you couldn't previously solve without computer use agents.

240
00:10:32.520 --> 00:10:34.600
So I think there's a huge opportunity here

241
00:10:34.600 --> 00:10:36.320
and in recent Horowitz, they talk about it

242
00:10:36.320 --> 00:10:39.680
and they say, we believe that the properly,

243
00:10:39.680 --> 00:10:42.040
to properly verticalize computer use agents

244
00:10:42.040 --> 00:10:44.120
and assist companies and adopting it

245
00:10:44.120 --> 00:10:47.040
will be a major area of exploration for startups.

246
00:10:47.040 --> 00:10:49.640
And this is like, to me, this screen's OpenClaw.

247
00:10:49.640 --> 00:10:52.840
You know, can you create a vertical use case for OpenClaw

248
00:10:52.840 --> 00:10:56.000
for a business and actually assist that company

249
00:10:56.000 --> 00:10:57.280
and adopting it?

250
00:10:57.280 --> 00:10:59.600
I think that's the huge opportunity here.

251
00:11:01.680 --> 00:11:05.880
So going back to this like workspace we have set up for you,

252
00:11:06.680 --> 00:11:09.520
you know, with all the people setting up OpenClaw,

253
00:11:09.520 --> 00:11:12.000
you could set it up as easy as, you know,

254
00:11:12.000 --> 00:11:15.360
I invite Greg to this workspace.

255
00:11:15.360 --> 00:11:18.080
I create him a new computer, we could just do it now.

256
00:11:18.080 --> 00:11:19.760
I hope OpenClaw too.

257
00:11:20.760 --> 00:11:25.920
I select how much RAM let's do, let's do eight gigs,

258
00:11:25.920 --> 00:11:30.720
launch that, open that up.

259
00:11:30.720 --> 00:11:34.840
And then all we need now to get you set up

260
00:11:34.840 --> 00:11:39.280
is to get the, let me get the, the curl command

261
00:11:39.280 --> 00:11:42.200
for OpenClaw, I copy that from their website.

262
00:11:42.200 --> 00:11:44.880
And then as this computer loads,

263
00:11:44.880 --> 00:11:47.480
we'll be able to then literally just paste it

264
00:11:47.480 --> 00:11:48.640
into the terminal.

265
00:11:49.640 --> 00:11:53.040
And once I see the interface pop up, boom, okay.

266
00:11:53.040 --> 00:11:54.400
So now I'm gonna hit Enter.

267
00:11:55.680 --> 00:11:57.720
And now we're off to the races installing OpenClaw.

268
00:11:57.720 --> 00:12:00.040
So like, it's as easy as that.

269
00:12:00.040 --> 00:12:02.840
And I think there's, this in and of itself

270
00:12:02.840 --> 00:12:05.760
is like a workspace where you can invite people

271
00:12:05.760 --> 00:12:08.640
and get them set up with OpenClaw or Cloud Code.

272
00:12:08.640 --> 00:12:12.640
I think it was like a huge opportunity as well

273
00:12:12.640 --> 00:12:15.040
of like just, there are executives right now

274
00:12:15.800 --> 00:12:20.040
like reaching out to me, like law firms, insurance companies.

275
00:12:20.040 --> 00:12:21.840
They're like, can you, can I just like pay you

276
00:12:21.840 --> 00:12:23.720
to teach me how to use this stuff?

277
00:12:23.720 --> 00:12:25.760
So that's it's own whole thing as well.

278
00:12:25.760 --> 00:12:27.520
As far as like, if you're savvy enough

279
00:12:27.520 --> 00:12:29.160
to even know how to install OpenClaw

280
00:12:29.160 --> 00:12:31.760
and get it set up in the first place,

281
00:12:31.760 --> 00:12:33.440
I just think there's a huge opportunity around like,

282
00:12:33.440 --> 00:12:36.680
just helping executives, businesses adopt it.

283
00:12:38.160 --> 00:12:40.760
So yeah, you can see it's as easy as this to get set up.

284
00:12:41.760 --> 00:12:46.360
And then as far as like, what specific things

285
00:12:46.360 --> 00:12:50.320
can we, can we automate with OpenClaw?

286
00:12:50.320 --> 00:12:52.200
There's a couple, it takes a little bit

287
00:12:52.200 --> 00:12:54.040
of a design thinking approach.

288
00:12:54.040 --> 00:12:56.400
So when you go into a business,

289
00:12:56.400 --> 00:12:58.400
let's say you find a project on Upwork

290
00:12:58.400 --> 00:13:00.440
and you wanna automate that with OpenClaw

291
00:13:00.440 --> 00:13:02.160
or let's say you go into a business

292
00:13:02.160 --> 00:13:05.360
and you're talking to the executive, the decision maker

293
00:13:05.360 --> 00:13:06.600
and it's clear that they have things

294
00:13:06.600 --> 00:13:07.520
that need to be automated.

295
00:13:07.520 --> 00:13:10.600
Well, as far as like, design thinking goes,

296
00:13:10.600 --> 00:13:13.640
like you need to have a clear way of like,

297
00:13:13.640 --> 00:13:16.360
first mapping like all the different possibilities.

298
00:13:16.360 --> 00:13:18.360
You can see here, this is something I did in the past

299
00:13:18.360 --> 00:13:20.200
of like, there's all these different things

300
00:13:20.200 --> 00:13:22.240
that you can automate and you wanna map them

301
00:13:22.240 --> 00:13:25.080
off two very simple metrics.

302
00:13:25.080 --> 00:13:26.960
What is the value that we can create

303
00:13:26.960 --> 00:13:28.400
by automating this thing?

304
00:13:28.400 --> 00:13:31.320
And what's the relative effort cost and time?

305
00:13:31.320 --> 00:13:33.160
And so we ultimately wanna start with,

306
00:13:33.160 --> 00:13:35.280
okay, we wanna automate things with OpenClaw

307
00:13:35.280 --> 00:13:38.320
that are high value and low effort cost and time.

308
00:13:38.320 --> 00:13:40.840
And that's like your low hanging fruit.

309
00:13:42.280 --> 00:13:43.920
And so you start there.

310
00:13:43.920 --> 00:13:47.200
And so like for this client, this was like,

311
00:13:47.200 --> 00:13:48.880
this was that, this was like, okay,

312
00:13:48.880 --> 00:13:50.760
we're looking at products on this website,

313
00:13:50.760 --> 00:13:53.880
we're downloading them, we're parsing all the information,

314
00:13:53.880 --> 00:13:55.480
that's the low hanging fruit.

315
00:13:55.480 --> 00:13:58.240
So start with the design thinking approach

316
00:13:58.240 --> 00:14:02.760
of like, okay, simplest, fastest to deploy.

317
00:14:03.040 --> 00:14:07.400
And then you need to map out like the systems design

318
00:14:07.400 --> 00:14:10.640
around how is this thing going to be automated, right?

319
00:14:10.640 --> 00:14:13.520
So for this client, she's like, okay,

320
00:14:13.520 --> 00:14:16.880
I send an email to a client of hers, right?

321
00:14:16.880 --> 00:14:18.440
She sends an email to a client

322
00:14:18.440 --> 00:14:21.520
and she has a presentation link with all these products.

323
00:14:21.520 --> 00:14:23.640
Okay, and all those products,

324
00:14:23.640 --> 00:14:25.200
she needs to look at all of them up

325
00:14:25.200 --> 00:14:27.160
and get all the information on them

326
00:14:27.160 --> 00:14:29.160
and then upload them into Zoho.

327
00:14:29.160 --> 00:14:30.520
So then your next step

328
00:14:30.520 --> 00:14:32.800
after identifying the opportunities

329
00:14:32.800 --> 00:14:35.720
to literally like map this out, I use Figma,

330
00:14:35.720 --> 00:14:36.840
you can use whatever.

331
00:14:36.840 --> 00:14:39.760
But map out the actual workflow process of like,

332
00:14:39.760 --> 00:14:41.400
okay, step one, step two, step three.

333
00:14:41.400 --> 00:14:43.480
What is this automation going to look like,

334
00:14:43.480 --> 00:14:46.240
tip to tail so that we can do the whole thing?

335
00:14:47.200 --> 00:14:48.920
Because with OpenClaw and computer use,

336
00:14:48.920 --> 00:14:50.840
now you can do that.

337
00:14:50.840 --> 00:14:52.920
You can do things from tip to tail.

338
00:14:52.920 --> 00:14:55.400
It's not like, you know, it used to be

339
00:14:55.400 --> 00:14:57.680
where you'd have to like go into a website

340
00:14:57.680 --> 00:14:59.040
and click a button,

341
00:14:59.040 --> 00:15:01.720
and then you'd be able to do some 50% of the whole thing,

342
00:15:01.720 --> 00:15:02.960
but then you'd have to copy that

343
00:15:02.960 --> 00:15:04.680
and paste that somewhere else and do it on your own.

344
00:15:04.680 --> 00:15:06.280
Like we can do it tip to tail.

345
00:15:08.320 --> 00:15:09.680
So quick recap,

346
00:15:10.800 --> 00:15:14.440
install OpenClaw into a computer, identify the next,

347
00:15:14.440 --> 00:15:18.000
identify the low-hanging fruit opportunities,

348
00:15:18.000 --> 00:15:20.000
the highest value opportunities,

349
00:15:20.000 --> 00:15:22.240
and then begin to map out what that even looks like

350
00:15:22.240 --> 00:15:23.800
to begin with.

351
00:15:23.800 --> 00:15:26.520
Couldn't you, you know, sort of, this is meta,

352
00:15:26.520 --> 00:15:28.560
but couldn't you use OpenClaw

353
00:15:28.560 --> 00:15:32.680
to actually do some of the prioritization

354
00:15:32.680 --> 00:15:34.920
on the automation and actually,

355
00:15:34.920 --> 00:15:37.720
I mean, you as a human being did the Figma,

356
00:15:37.720 --> 00:15:39.760
but couldn't you actually just use the OpenClaw

357
00:15:39.760 --> 00:15:43.520
or Cloud Code or something like that to help you with that?

358
00:15:43.520 --> 00:15:47.880
So for example, like if you go back to the Figma,

359
00:15:48.320 --> 00:15:54.680
like you could walk into a business and basically say,

360
00:15:54.680 --> 00:15:58.240
hey, I want to figure out what we can automate here

361
00:15:58.240 --> 00:16:02.520
and you do customer interviews with different people

362
00:16:02.520 --> 00:16:03.360
on the team.

363
00:16:03.360 --> 00:16:05.040
You record those customer interviews,

364
00:16:05.040 --> 00:16:08.120
you get the transcripts, you upload the transcripts,

365
00:16:08.120 --> 00:16:10.240
and then you say, hey, based on that,

366
00:16:10.240 --> 00:16:12.560
then you're like, you can actually say, you know,

367
00:16:12.560 --> 00:16:15.080
you give this as a reference image.

368
00:16:15.080 --> 00:16:16.920
Basically, say like, hey, I want to figure out

369
00:16:16.920 --> 00:16:20.480
which automation opportunities have the highest amount

370
00:16:20.480 --> 00:16:23.600
of value, low amount of effort, cost and time,

371
00:16:23.600 --> 00:16:27.680
give me the top three and then create, you know, Figma.

372
00:16:27.680 --> 00:16:29.920
And I think there's like a Figma, MCP,

373
00:16:29.920 --> 00:16:33.720
even that you can use and you can say like,

374
00:16:33.720 --> 00:16:35.240
hey, like, can you map this thing out

375
00:16:35.240 --> 00:16:36.800
based on these customer transcripts?

376
00:16:36.800 --> 00:16:38.880
Does that make sense or am I?

377
00:16:38.880 --> 00:16:40.400
Oh, absolutely.

378
00:16:40.400 --> 00:16:41.480
Yeah.

379
00:16:41.480 --> 00:16:43.080
No, that's the way to do it.

380
00:16:43.080 --> 00:16:46.360
Like, whenever I do any kind of call with a client

381
00:16:46.360 --> 00:16:48.360
or cut like a potential customer,

382
00:16:48.360 --> 00:16:53.240
oh my gosh, Gemini for Google Meet is amazing.

383
00:16:53.240 --> 00:16:55.240
You just have it take all the notes.

384
00:16:55.240 --> 00:16:57.720
And then actually, that's how I even got,

385
00:16:57.720 --> 00:16:59.080
because I don't know about, I don't know about you Greg,

386
00:16:59.080 --> 00:17:00.960
but sometimes when you're in these calls,

387
00:17:00.960 --> 00:17:03.920
you kind of, you know, this industry, you know,

388
00:17:03.920 --> 00:17:06.640
you're in a new industry, you're helping this customer,

389
00:17:06.640 --> 00:17:10.080
you don't understand their domain expertise, the lingo.

390
00:17:10.080 --> 00:17:12.320
And so you got to go back and like, okay,

391
00:17:12.480 --> 00:17:14.120
what was it that they said?

392
00:17:14.120 --> 00:17:17.600
And so half the granola or Gemini notes or whatever,

393
00:17:17.600 --> 00:17:19.400
and then literally ask it to like, okay,

394
00:17:19.400 --> 00:17:22.600
what's the step-by-step workflow and then map it out?

395
00:17:22.600 --> 00:17:25.280
It just helps me to map it out visually.

396
00:17:25.280 --> 00:17:27.800
But you could literally ask it, yeah, like you said,

397
00:17:27.800 --> 00:17:29.080
based on this transcript,

398
00:17:29.080 --> 00:17:32.200
what is the automation workflow look like?

399
00:17:32.200 --> 00:17:33.200
You know, stuff like that.

400
00:17:33.200 --> 00:17:34.920
And if you don't want to use Figma,

401
00:17:34.920 --> 00:17:38.360
you can also even say, you know,

402
00:17:38.360 --> 00:17:42.880
do output in Mermaid code,

403
00:17:42.880 --> 00:17:44.600
and then you can use the Mermaid code

404
00:17:44.600 --> 00:17:48.040
and insert that into an Excala draw or a TL draw

405
00:17:48.040 --> 00:17:48.960
or something like that.

406
00:17:48.960 --> 00:17:51.720
So a little pro tip there.

407
00:17:51.720 --> 00:17:55.440
Nice, nice, yeah, and the Figma MCP is pretty cool too.

408
00:17:55.440 --> 00:17:56.880
Yeah, definitely check that out.

409
00:17:56.880 --> 00:18:00.080
So yeah, so then once you figure out

410
00:18:00.080 --> 00:18:02.840
what the workflow is, right?

411
00:18:02.840 --> 00:18:05.480
This is where like, you have to actually

412
00:18:06.120 --> 00:18:11.160
be able to know, okay, how much can I really ask,

413
00:18:11.160 --> 00:18:14.920
like, how much can I just say like to OpenClaw right now,

414
00:18:14.920 --> 00:18:18.000
hey, like build this, like, hey, build this thing,

415
00:18:18.000 --> 00:18:22.400
and you just describe the workflow versus genuinely

416
00:18:22.400 --> 00:18:25.280
using something like Cloud Code to build out

417
00:18:25.280 --> 00:18:27.960
like what that workflow would look like with, you know,

418
00:18:27.960 --> 00:18:32.480
Python, APIs, a genuine automation pipeline and process

419
00:18:32.480 --> 00:18:35.960
that your OpenClaw can actually just trigger

420
00:18:35.960 --> 00:18:39.720
upon whenever it's like contextual relevant.

421
00:18:39.720 --> 00:18:42.680
So for instance, this whole pipeline here

422
00:18:42.680 --> 00:18:44.200
of like going to these websites,

423
00:18:44.200 --> 00:18:46.120
looking at this product information,

424
00:18:46.120 --> 00:18:47.760
downloading the information, parsing it,

425
00:18:47.760 --> 00:18:50.960
uploading it to Zoho, the trigger of all of that

426
00:18:50.960 --> 00:18:56.160
is, you know, the OpenClaw being CCed in an email,

427
00:18:56.160 --> 00:18:58.320
and it's seeing that email, it has a link

428
00:18:58.320 --> 00:19:01.400
that is relevant for this type of workflow to be triggered.

429
00:19:01.400 --> 00:19:04.640
So that is like the thing, that's like the listening event

430
00:19:04.640 --> 00:19:07.520
that OpenClaw can do with like a cron job that it sets up

431
00:19:07.520 --> 00:19:10.200
to just like, okay, listen for this trigger.

432
00:19:10.200 --> 00:19:12.240
And then once that trigger starts,

433
00:19:12.240 --> 00:19:16.040
it can then activate the whole Python script,

434
00:19:16.040 --> 00:19:17.880
workflow automation, everything that you would need

435
00:19:17.880 --> 00:19:19.000
downstream of that.

436
00:19:19.000 --> 00:19:23.800
So you're not relying too much on OpenClaw's like abilities

437
00:19:23.800 --> 00:19:24.760
in and of itself.

438
00:19:24.760 --> 00:19:27.720
You're more of so creating specialized AI workers

439
00:19:27.720 --> 00:19:30.360
under each, underneath the OpenClaw

440
00:19:30.360 --> 00:19:33.000
that it can call individually, if that makes sense.

441
00:19:34.040 --> 00:19:36.240
You know, you talked earlier about sub agents.

442
00:19:36.240 --> 00:19:39.480
I think a lot of people are confused about

443
00:19:39.480 --> 00:19:42.640
what is a sub agent versus a task and stuff like that.

444
00:19:42.640 --> 00:19:46.160
Can you just clearly explain that?

445
00:19:46.160 --> 00:19:47.480
Yeah.

446
00:19:47.480 --> 00:19:52.480
So sub agents are, there's a couple of ways to view them, right?

447
00:19:53.280 --> 00:19:56.640
So like in the context of, because in the reason I say

448
00:19:56.640 --> 00:19:57.480
there's a couple of ways to view them,

449
00:19:57.480 --> 00:19:59.640
it's because there's a couple of ways of using them.

450
00:19:59.640 --> 00:20:02.000
So in the context of like OpenClaw,

451
00:20:02.000 --> 00:20:05.160
you can ask it to spin up five research sub agents

452
00:20:05.160 --> 00:20:08.000
that I'll go and research some given task.

453
00:20:08.000 --> 00:20:11.800
And like I said earlier, you can have it parallelized

454
00:20:11.800 --> 00:20:15.760
that task across, splitting it up across each sub agent

455
00:20:15.760 --> 00:20:18.920
or having each sub agent actually go and do the same task

456
00:20:18.920 --> 00:20:20.920
across five different instances.

457
00:20:20.920 --> 00:20:23.320
But the next thing around sub agents

458
00:20:23.400 --> 00:20:25.840
is that you can actually, like you said,

459
00:20:27.000 --> 00:20:29.160
maybe think of them in terms of skills.

460
00:20:29.160 --> 00:20:31.280
So if you're familiar with anthropic skills,

461
00:20:32.840 --> 00:20:36.560
you can have like these specialized instructions and rules

462
00:20:36.560 --> 00:20:40.000
along with actual code that you can provide to your agent

463
00:20:40.000 --> 00:20:43.280
for it to be able to go and do a given task.

464
00:20:43.280 --> 00:20:47.880
And this is really nice because it gives you

465
00:20:47.880 --> 00:20:52.160
a very more powerful general purpose agent

466
00:20:52.160 --> 00:20:55.880
that I can do many of your specific nuance tasks

467
00:20:55.880 --> 00:20:57.720
across various domains.

468
00:20:57.720 --> 00:21:01.560
But the thing is it's like I want my general agent

469
00:21:01.560 --> 00:21:06.720
to be freed up and to more so just be the orchestrator.

470
00:21:06.720 --> 00:21:09.960
And what if the general agent, this one, right?

471
00:21:09.960 --> 00:21:13.720
The one I have start here can just call a sub agent,

472
00:21:13.720 --> 00:21:16.280
like worker number four here,

473
00:21:16.280 --> 00:21:19.560
to do a given skill that you have created.

474
00:21:19.560 --> 00:21:22.040
So if your skill is that it goes on Twitter

475
00:21:22.040 --> 00:21:26.120
and finds the most viral ideas and it bookmarks them,

476
00:21:26.120 --> 00:21:27.880
rather than having your main agent do that,

477
00:21:27.880 --> 00:21:29.480
and now you can't talk to your main agent

478
00:21:29.480 --> 00:21:31.720
for the next 20 minutes because it's working on that,

479
00:21:31.720 --> 00:21:34.480
can it call that skill into a sub agent

480
00:21:34.480 --> 00:21:36.000
and have the sub agent do that?

481
00:21:36.000 --> 00:21:39.560
That I think is where things get really interesting

482
00:21:39.560 --> 00:21:43.040
and in terms of the context of like deploying open clause

483
00:21:43.040 --> 00:21:46.760
for businesses, I would think of everything that you have

484
00:21:46.760 --> 00:21:48.840
in terms of an AI automation opportunity

485
00:21:48.840 --> 00:21:52.640
around workflows, skills, tasks, et cetera.

486
00:21:52.640 --> 00:21:57.480
I would actually just create that as its own specific sub agent

487
00:21:57.480 --> 00:22:01.000
with its own skill that your open clock can then call.

488
00:22:01.000 --> 00:22:03.080
If that does that make sense?

489
00:22:03.080 --> 00:22:04.760
It does, it does.

490
00:22:04.760 --> 00:22:05.840
Nice.

491
00:22:05.840 --> 00:22:09.600
It's, you know, I think the basic idea is like,

492
00:22:10.600 --> 00:22:15.600
you know, in layman terms, it's as soon as you have

493
00:22:15.600 --> 00:22:20.840
your open clause instance, you know, do something,

494
00:22:20.840 --> 00:22:25.680
they're busy, you know, it's like they've got a mug of hot coffee.

495
00:22:25.680 --> 00:22:31.480
And so your job is you want to leverage this as much as possible.

496
00:22:31.480 --> 00:22:35.720
So you don't want your agent to hold a hot coffee.

497
00:22:35.720 --> 00:22:40.520
So if you ask it to do, to move this desk into this area,

498
00:22:40.520 --> 00:22:43.560
you know, it says, no, I'm holding a cup of hot coffee,

499
00:22:43.560 --> 00:22:44.640
I can't do that.

500
00:22:44.720 --> 00:22:49.560
So what sub agents do is it basically creates leverage

501
00:22:49.560 --> 00:22:51.120
for your open clause.

502
00:22:51.120 --> 00:22:52.880
And it basically says like, okay,

503
00:22:52.880 --> 00:22:54.640
you're going to create a set of sub agents

504
00:22:54.640 --> 00:22:57.600
who are going to be good at x, y, z thing.

505
00:22:57.600 --> 00:23:02.600
And that way it frees up your main agent to, you know,

506
00:23:04.600 --> 00:23:08.120
as you say, orchestrate to basically be the manager

507
00:23:08.120 --> 00:23:09.760
of the sub agents.

508
00:23:10.600 --> 00:23:15.480
And you know, what that can mean is like looking at quality

509
00:23:15.480 --> 00:23:18.840
of work, it can mean checking for certain things

510
00:23:18.840 --> 00:23:20.520
and stuff like that.

511
00:23:20.520 --> 00:23:22.360
Exactly, exactly.

512
00:23:22.360 --> 00:23:24.320
I think that's going to be huge, you know,

513
00:23:24.320 --> 00:23:27.880
when you start working with these businesses and customers

514
00:23:27.880 --> 00:23:32.480
who want things to be automated, once you show them what's possible,

515
00:23:32.480 --> 00:23:35.840
their eyes light up, they get all these ideas themselves.

516
00:23:35.840 --> 00:23:37.320
These are high agency people, you know,

517
00:23:37.480 --> 00:23:40.480
they come up with creative ideas that they want to start implementing.

518
00:23:40.480 --> 00:23:43.800
And then what you realize is there's just a huge,

519
00:23:43.800 --> 00:23:45.600
a huge list of things that can be automated

520
00:23:45.600 --> 00:23:47.280
and they're excited about that.

521
00:23:47.280 --> 00:23:51.800
And so actually like the ability to, okay,

522
00:23:51.800 --> 00:23:55.680
first solve a vertical specific workflow for a customer.

523
00:23:55.680 --> 00:23:57.040
And then that opening up their mind

524
00:23:57.040 --> 00:23:59.560
and then then being like, oh, I wonder if I could,

525
00:23:59.560 --> 00:24:02.120
could I text this thing and it does this?

526
00:24:02.120 --> 00:24:05.600
This is kind of where like the whole open clause moment

527
00:24:05.600 --> 00:24:06.720
is really powerful.

528
00:24:06.720 --> 00:24:10.960
It's like, it's the assistant like capability.

529
00:24:10.960 --> 00:24:13.000
It's the, you know, I have it, I have it here.

530
00:24:13.000 --> 00:24:15.960
It's like, a lot of people might get confused about,

531
00:24:15.960 --> 00:24:17.920
why is it that open clause so special?

532
00:24:17.920 --> 00:24:21.360
It's the ability that has its own computer,

533
00:24:21.360 --> 00:24:24.880
it's running 24.7, you can text it

534
00:24:24.880 --> 00:24:26.880
and you can schedule tasks.

535
00:24:26.880 --> 00:24:30.840
And really, if we just removed open clause from this card here

536
00:24:30.840 --> 00:24:33.640
and you just called this a really good employee,

537
00:24:33.640 --> 00:24:35.000
it would just make sense.

538
00:24:35.000 --> 00:24:38.840
It'd be like, oh, works 24.7, can code, can schedule tasks,

539
00:24:38.840 --> 00:24:41.280
I can text it and they have their own computer.

540
00:24:42.960 --> 00:24:45.680
So I think that's like kind of why open clause

541
00:24:45.680 --> 00:24:46.520
exciting for a lot of people.

542
00:24:46.520 --> 00:24:50.040
And if it's not, you know, if some people think it's overhyped,

543
00:24:50.040 --> 00:24:52.280
you have to kind of look at the whole picture, I think.

544
00:24:52.280 --> 00:24:54.560
And then you're able to really gauge it.

545
00:24:54.560 --> 00:24:58.920
So I'm certainly bought in on this idea that,

546
00:24:58.920 --> 00:25:02.280
you know, it could be a really good employee.

547
00:25:02.280 --> 00:25:04.640
I think there's also cases where people

548
00:25:04.640 --> 00:25:06.120
aren't setting up their open clause

549
00:25:06.120 --> 00:25:08.800
in the right way where it ends up being a bad employee.

550
00:25:08.800 --> 00:25:13.800
And I think, you know, that's sort of like the issue

551
00:25:15.640 --> 00:25:18.120
with that is, you know, sometimes you have a bad employee

552
00:25:18.120 --> 00:25:20.120
because the manager, the coach essentially,

553
00:25:20.120 --> 00:25:23.320
is not doing a good job at giving the right context

554
00:25:23.320 --> 00:25:24.160
at the right time.

555
00:25:24.160 --> 00:25:27.720
So I think, you know, do you have any tips and tricks

556
00:25:27.720 --> 00:25:30.840
around besides spinning up sub agents?

557
00:25:30.840 --> 00:25:33.600
Like how could people listening to this,

558
00:25:33.600 --> 00:25:38.040
if they want to go after this opportunity of essentially

559
00:25:38.040 --> 00:25:42.320
verticalize, you know, open clause and automating

560
00:25:42.320 --> 00:25:47.280
some of these flows, how could people actually,

561
00:25:47.280 --> 00:25:49.840
you know, take their open claw from a bad employee

562
00:25:49.840 --> 00:25:51.760
to a good employee?

563
00:25:51.760 --> 00:25:54.800
Yeah, I think it comes down to, let's,

564
00:25:54.800 --> 00:25:56.400
maybe we should walk through.

565
00:25:56.400 --> 00:25:59.360
So I know you have, I was actually looking at this.

566
00:25:59.360 --> 00:26:00.440
So idea browser.

567
00:26:00.440 --> 00:26:02.680
So for everyone, if you don't know,

568
00:26:02.680 --> 00:26:05.440
Greg has this amazing product, idea browser.

569
00:26:05.440 --> 00:26:08.800
And I love this trend, this idea today,

570
00:26:08.800 --> 00:26:11.360
take talk trend tool that catches viral waves

571
00:26:11.360 --> 00:26:12.840
before they peak.

572
00:26:12.840 --> 00:26:15.680
So I was actually looking at this this morning,

573
00:26:15.680 --> 00:26:17.160
I was like, wow, this is actually something

574
00:26:17.160 --> 00:26:21.040
that you can maybe turn into a skill for an open clause

575
00:26:21.040 --> 00:26:23.680
to create a specialized skill around this.

576
00:26:23.680 --> 00:26:26.560
So let's just do a live, let's see.

577
00:26:26.560 --> 00:26:29.200
I generally don't know how far can we get,

578
00:26:29.200 --> 00:26:30.600
can we build this out now?

579
00:26:30.600 --> 00:26:33.680
Let's see, if I copy all of that,

580
00:26:33.680 --> 00:26:36.840
and now I'm here in the open claw that we set up for you,

581
00:26:36.840 --> 00:26:38.240
and you can see it just got set up,

582
00:26:38.240 --> 00:26:40.720
it's, I just told it, hey, I'm Greg Eisenberg.

583
00:26:40.720 --> 00:26:42.560
So we're getting started, that's it.

584
00:26:42.560 --> 00:26:47.560
And let's just say I want to build a specialized skill

585
00:26:50.160 --> 00:26:53.000
to be able to do the following.

586
00:26:53.000 --> 00:26:58.000
And I'm just gonna paste that entire idea browser idea.

587
00:26:59.000 --> 00:27:02.720
And I'm gonna ask it as far as creating these automations,

588
00:27:02.720 --> 00:27:05.480
creating workflows, or doing anything with open claw,

589
00:27:05.480 --> 00:27:09.120
my number one tip is always ask it to ask you questions.

590
00:27:09.120 --> 00:27:14.120
So what do you need from me to be able to build this out?

591
00:27:16.760 --> 00:27:17.960
Let's create a plan.

592
00:27:17.960 --> 00:27:21.720
And so a lot of people need to remember open clause

593
00:27:21.720 --> 00:27:25.200
like a, almost a little bit of a wrapper around like

594
00:27:25.200 --> 00:27:26.440
a clawed code in a way.

595
00:27:27.960 --> 00:27:30.280
So let's see, okay, cool.

596
00:27:30.280 --> 00:27:32.720
This is a big vision, Greg, I like it.

597
00:27:32.720 --> 00:27:37.240
Let's break down a realistic build as an open claw skill.

598
00:27:38.520 --> 00:27:42.040
So now it's saying, okay, I need data access.

599
00:27:42.040 --> 00:27:45.240
I need the scope, I need the niche focus.

600
00:27:46.080 --> 00:27:49.440
That'll shape what we build a lean skill.

601
00:27:49.440 --> 00:27:52.480
So the first thing that we can maybe do is,

602
00:27:53.400 --> 00:27:56.600
here in like Orgo, we have this playground mode.

603
00:27:56.600 --> 00:27:59.960
And you can ask the playground agent here

604
00:27:59.960 --> 00:28:04.000
to do things in this computer, in this computer environment.

605
00:28:04.000 --> 00:28:06.280
So one thing I might test first is like,

606
00:28:06.280 --> 00:28:11.080
can we get this agent to even just spin up TikTok

607
00:28:11.080 --> 00:28:14.920
and just scroll TikTok and identify

608
00:28:14.920 --> 00:28:16.640
what is on the for you page TikTok?

609
00:28:16.640 --> 00:28:17.840
So let's like maybe start there.

610
00:28:17.840 --> 00:28:18.880
Does that sound good?

611
00:28:18.880 --> 00:28:20.680
Yeah, absolutely.

612
00:28:20.680 --> 00:28:21.600
Let's do that.

613
00:28:21.600 --> 00:28:26.600
Open Firefox, go to TikTok, scroll.

614
00:28:27.320 --> 00:28:31.400
I'm gonna say scroll TikTok looking for

615
00:28:31.400 --> 00:28:36.400
what the most common videos are on the for you page.

616
00:28:37.360 --> 00:28:39.160
Give me a summary.

617
00:28:39.160 --> 00:28:41.520
So let's see how it's able to do this.

618
00:28:43.440 --> 00:28:46.200
So this is using our playground mode.

619
00:28:47.200 --> 00:28:51.200
And boom, opens up Firefox.

620
00:28:54.040 --> 00:28:56.600
It's gonna go to TikTok.

621
00:28:56.600 --> 00:28:59.760
And for those listening, I'm just gonna talk

622
00:28:59.760 --> 00:29:01.760
through a little bit about what this agent

623
00:29:01.760 --> 00:29:03.360
in our playground is doing.

624
00:29:03.360 --> 00:29:05.240
It's visually interacting with the screen.

625
00:29:05.240 --> 00:29:08.520
It's clicking into Firefox, it's opening up the browser.

626
00:29:08.520 --> 00:29:13.040
Now it's typing in TikTok.com, it's going there.

627
00:29:13.920 --> 00:29:17.760
And let's see.

628
00:29:20.200 --> 00:29:22.200
This is always such a magical experience

629
00:29:22.200 --> 00:29:26.480
just watching a computer navigate the web

630
00:29:26.480 --> 00:29:28.360
like a human being.

631
00:29:28.360 --> 00:29:29.840
It's amazing.

632
00:29:29.840 --> 00:29:32.280
You know, this is, yeah, there it goes.

633
00:29:33.600 --> 00:29:35.560
It's on the homepage.

634
00:29:36.600 --> 00:29:40.120
And now it's gonna actually scroll TikTok.

635
00:29:40.120 --> 00:29:42.200
It's gonna probably take a screenshot of this,

636
00:29:42.240 --> 00:29:43.560
get the context of,

637
00:29:43.560 --> 00:29:46.400
based off that screenshot of what the video is about.

638
00:29:46.400 --> 00:29:48.120
You could even see it, it has hashtags,

639
00:29:48.120 --> 00:29:49.440
movie hashtag for you page

640
00:29:49.440 --> 00:29:51.560
that's gonna be able to infer a lot of things.

641
00:29:51.560 --> 00:29:54.960
Boom, it's scrolls, it's scrolls, it's gets a pop up,

642
00:29:54.960 --> 00:29:56.480
it's gonna close out the pop up.

643
00:29:58.280 --> 00:30:01.120
But on your point, Greg, Darryl Amade,

644
00:30:01.120 --> 00:30:06.120
the CEO of Anthropic, he just had a podcast with Dora Keshe

645
00:30:06.200 --> 00:30:09.000
and it came out a couple of days ago.

646
00:30:09.000 --> 00:30:10.280
And you know what's really interesting,

647
00:30:10.360 --> 00:30:13.120
what he said in that podcast.

648
00:30:13.120 --> 00:30:18.120
He said, he said this idea of his around this data center

649
00:30:18.240 --> 00:30:21.800
full of brilliance, you know, scientists

650
00:30:21.800 --> 00:30:25.160
and Nobel Prize winners, essentially his concept

651
00:30:25.160 --> 00:30:27.120
of what AGI will be like.

652
00:30:27.120 --> 00:30:29.800
He says the constraint to getting there

653
00:30:29.800 --> 00:30:31.480
is computer use agents.

654
00:30:31.480 --> 00:30:34.680
The ability to have an AI that can operate a computer

655
00:30:34.680 --> 00:30:36.280
like you and I can, but better.

656
00:30:36.280 --> 00:30:39.000
You know, interact with the visual interface,

657
00:30:39.000 --> 00:30:41.080
also be able to do things under the hood,

658
00:30:41.080 --> 00:30:42.320
kind of like clawed code.

659
00:30:43.480 --> 00:30:45.360
This is the constraint, he said.

660
00:30:45.360 --> 00:30:48.720
And I mean, it makes perfect sense.

661
00:30:48.720 --> 00:30:50.880
If it can do anything that you and I can do on a computer,

662
00:30:50.880 --> 00:30:53.200
that seems like it can go pretty far.

663
00:30:54.480 --> 00:30:57.800
OpenClaw is like a chat GBTmo when I think

664
00:30:57.800 --> 00:31:00.960
for this kind of idea of computer use.

665
00:31:00.960 --> 00:31:04.560
And as far as like building these

666
00:31:04.560 --> 00:31:07.680
computer use agents out, you can see this is clearly working

667
00:31:07.680 --> 00:31:09.200
so we know this is possible.

668
00:31:10.560 --> 00:31:12.960
You can build, a lot of people might get confused

669
00:31:12.960 --> 00:31:14.680
with Orgo when they come to our site,

670
00:31:14.680 --> 00:31:16.400
they see computers for agents.

671
00:31:16.400 --> 00:31:17.400
And like, what does that mean?

672
00:31:17.400 --> 00:31:18.720
Well, I think they get it now.

673
00:31:18.720 --> 00:31:20.200
It's like, okay, you want your codbot

674
00:31:20.200 --> 00:31:21.600
to have its own computer.

675
00:31:21.600 --> 00:31:25.080
But also we provide, and I'll show this,

676
00:31:25.080 --> 00:31:27.800
we provide the doc in our docs.

677
00:31:27.800 --> 00:31:30.960
Like we're actually provide the programmatic APIs

678
00:31:30.960 --> 00:31:33.800
so that you can create custom computer use agents

679
00:31:33.800 --> 00:31:36.800
that do a given task very well.

680
00:31:36.800 --> 00:31:38.440
So you can bring any model, you can get

681
00:31:38.440 --> 00:31:42.480
Kimi 2.5 is like the super cheap Chinese model,

682
00:31:42.480 --> 00:31:44.440
it's very good at computer use.

683
00:31:44.440 --> 00:31:47.320
And you can give it the ability to click, drag, scroll,

684
00:31:47.320 --> 00:31:49.640
type in the keyboard, spin up a computer.

685
00:31:50.960 --> 00:31:53.440
And you could create specialized, really fast,

686
00:31:53.440 --> 00:31:56.600
performance, low cost computer use agents,

687
00:31:56.600 --> 00:31:57.800
using our docs.

688
00:31:57.800 --> 00:32:00.160
And I just say that to say like as far as,

689
00:32:00.160 --> 00:32:02.080
this thing that we're doing here,

690
00:32:02.080 --> 00:32:04.840
creating a skill around scrolling TikTok,

691
00:32:05.840 --> 00:32:08.440
I think that's how we can actually give it a shot.

692
00:32:08.440 --> 00:32:11.640
It's, let's just, we see this is working now.

693
00:32:11.640 --> 00:32:14.640
Let me just grab the or go docs

694
00:32:14.640 --> 00:32:16.440
and let's start building the skill out.

695
00:32:16.440 --> 00:32:18.480
If you didn't see here, I'm at the or go docs.

696
00:32:18.480 --> 00:32:22.800
I click this thing called LLMSFull.txt.

697
00:32:22.800 --> 00:32:25.320
This is like all the instructions for the LLM

698
00:32:25.320 --> 00:32:27.040
to be able to build on top of or go.

699
00:32:28.000 --> 00:32:31.680
And what I'm going to do is, I'm going to wait for,

700
00:32:32.680 --> 00:32:34.920
I have to probably, I'm going to wait for this to finish,

701
00:32:34.920 --> 00:32:37.080
this it's task and then I'll be able to tell it.

702
00:32:39.160 --> 00:32:42.400
But while this goes Greg, you have any thoughts?

703
00:32:42.400 --> 00:32:45.600
Well, just sort of a thing I was thinking about is,

704
00:32:45.600 --> 00:32:50.600
it sounds like whenever you're trying to do a new automation,

705
00:32:51.880 --> 00:32:55.360
you start by thinking about what is a lightweight skill

706
00:32:55.360 --> 00:32:57.800
that I should create, is that correct?

707
00:32:57.800 --> 00:33:00.520
Right, exactly, what's the MVP?

708
00:33:00.520 --> 00:33:03.800
Yeah, so you start with a lightweight skill,

709
00:33:05.400 --> 00:33:08.400
you test it and then from there,

710
00:33:08.400 --> 00:33:10.520
you're probably like, okay, here's what went wrong,

711
00:33:10.520 --> 00:33:13.640
here's what could be better, that sort of thing.

712
00:33:13.640 --> 00:33:15.240
Right, exactly.

713
00:33:15.240 --> 00:33:17.360
Just fine-tuning, you know, debug,

714
00:33:17.360 --> 00:33:21.480
like I think the design thinking process

715
00:33:21.480 --> 00:33:25.480
around all of this is super important of like, okay,

716
00:33:25.480 --> 00:33:29.040
if I want to build a car, maybe the first thing I do

717
00:33:29.040 --> 00:33:32.240
isn't to build the frame of the car

718
00:33:32.240 --> 00:33:37.240
or to build the whole body of the car.

719
00:33:37.240 --> 00:33:38.840
That's not the first thing, the first thing I should do

720
00:33:38.840 --> 00:33:41.320
if I want to build a car, well, why do I want to build a car?

721
00:33:41.320 --> 00:33:43.760
Well, I want to build a car to go from point A to point B.

722
00:33:43.760 --> 00:33:46.040
Oh, okay, so really, maybe I should start

723
00:33:46.040 --> 00:33:47.240
by building a skateboard.

724
00:33:49.120 --> 00:33:54.120
How can we accomplish the task to get the dream outcome

725
00:33:55.120 --> 00:33:59.640
and sometimes that means like starting with something

726
00:33:59.640 --> 00:34:02.560
that's completely different than the end state?

727
00:34:03.920 --> 00:34:05.880
So this is done here.

728
00:34:05.880 --> 00:34:09.480
I think it interrupted itself, probably a context thing,

729
00:34:09.480 --> 00:34:12.360
but now we can actually what we can do.

730
00:34:12.360 --> 00:34:15.960
This is all live, so you're seeing this in real time.

731
00:34:15.960 --> 00:34:19.240
You can literally, there's a couple ways to go about this.

732
00:34:19.240 --> 00:34:22.600
You can actually install Cloud Code into this computer

733
00:34:22.600 --> 00:34:26.440
and have it build out the automation in here

734
00:34:26.440 --> 00:34:29.280
or we can just ask this agent to do it for us.

735
00:34:29.280 --> 00:34:33.960
So let's see, I want to build a computer use agent

736
00:34:35.120 --> 00:34:37.080
that does this exact thing,

737
00:34:38.960 --> 00:34:43.960
but more programmatically using the Orgo API docs.

738
00:34:45.760 --> 00:34:50.120
And I paste that here, what do we need to get started?

739
00:34:50.120 --> 00:34:51.480
And we send that off.

740
00:34:51.480 --> 00:34:52.840
All right, here we go.

741
00:34:52.840 --> 00:34:54.320
So we need an Orgo API key.

742
00:34:54.320 --> 00:34:56.480
Here's the architecture of what we'll build.

743
00:34:57.560 --> 00:34:58.800
Okay, this looks good.

744
00:35:00.920 --> 00:35:04.720
Orgo key andthropic key, it has all the code here.

745
00:35:06.320 --> 00:35:07.160
Cool.

746
00:35:08.160 --> 00:35:10.000
You're looking at a TikTok video extract,

747
00:35:10.000 --> 00:35:13.440
the username, video description, the category,

748
00:35:13.440 --> 00:35:16.680
the appropriate like count, boom, boom, boom.

749
00:35:17.520 --> 00:35:18.560
Okay, cool.

750
00:35:18.600 --> 00:35:20.640
So I have all these things already

751
00:35:20.640 --> 00:35:23.600
and don't worry, I'm going to delete these keys.

752
00:35:23.600 --> 00:35:25.800
I'm not worried about leaking or anything.

753
00:35:25.800 --> 00:35:27.480
I'm going to copy my Orgo API key.

754
00:35:27.480 --> 00:35:28.680
I'm going to paste it in there.

755
00:35:29.680 --> 00:35:34.080
And I'm going to say Orgo API key.

756
00:35:34.080 --> 00:35:36.080
I also have my andthropic key.

757
00:35:36.080 --> 00:35:37.520
Let me grab that.

758
00:35:37.520 --> 00:35:39.680
I'm going to paste that here.

759
00:35:39.680 --> 00:35:41.520
Can we build this out?

760
00:35:42.680 --> 00:35:45.760
And so now we're going to have our playground mode

761
00:35:45.760 --> 00:35:47.760
build out this computer use agent

762
00:35:47.760 --> 00:35:50.040
to be able to go do this thing that we just test it out.

763
00:35:50.040 --> 00:35:52.040
We know it works, we know we can do it.

764
00:35:52.040 --> 00:35:54.240
Let's turn it into something that could be more programmatic

765
00:35:54.240 --> 00:35:55.600
kind of like a skill.

766
00:35:55.600 --> 00:35:58.840
And let's give it to Clawbot so it always has access to it.

767
00:35:58.840 --> 00:35:59.840
The dream.

768
00:36:03.200 --> 00:36:07.120
It's literally any idea you have, you can just build it.

769
00:36:10.040 --> 00:36:12.960
And that's sort of the arbitrage opportunity, right?

770
00:36:12.960 --> 00:36:16.280
Especially like in our world, of course,

771
00:36:16.280 --> 00:36:18.600
like we're so used to this now.

772
00:36:18.600 --> 00:36:20.920
Even though it's only been like two months.

773
00:36:22.000 --> 00:36:26.400
But, you know, the opportunity is the vast majority

774
00:36:26.400 --> 00:36:29.320
of people on this planet and businesses

775
00:36:29.320 --> 00:36:31.640
would love to have better automation

776
00:36:31.640 --> 00:36:36.840
and would love to, you know, have computer use agents

777
00:36:36.840 --> 00:36:39.800
working for them, AK, really good employees

778
00:36:39.800 --> 00:36:43.280
working for them, but they don't know how.

779
00:36:43.320 --> 00:36:46.240
And so I think which is cool that you're like,

780
00:36:46.240 --> 00:36:48.640
you're showing us like some of the best practices

781
00:36:48.640 --> 00:36:50.160
and how to do it.

782
00:36:50.160 --> 00:36:51.160
Exactly.

783
00:36:51.160 --> 00:36:54.440
And I think this is also like, I can imagine,

784
00:36:54.440 --> 00:36:57.520
you know, the whole audience of this podcast,

785
00:36:57.520 --> 00:36:59.480
we're all pretty tech savvy.

786
00:36:59.480 --> 00:37:01.680
We know how to do things like vibe code

787
00:37:01.680 --> 00:37:04.760
and play around with Clawbot and be able to do these things.

788
00:37:04.760 --> 00:37:09.760
And we take it for granted in terms of what that value is worth.

789
00:37:10.440 --> 00:37:14.680
As far as like open Claw and the opportunity around that,

790
00:37:14.680 --> 00:37:17.960
I mean, open Claw started going viral on Twitter

791
00:37:17.960 --> 00:37:19.760
around two to three weeks ago.

792
00:37:19.760 --> 00:37:21.960
Only now is it starting and I'm starting to see

793
00:37:21.960 --> 00:37:23.880
it's starting to go viral on TikTok,

794
00:37:23.880 --> 00:37:25.360
a little more mainstream.

795
00:37:25.360 --> 00:37:30.360
So yeah, I just mean, you know, people are catching on

796
00:37:33.320 --> 00:37:35.760
and a lot of people still need help with getting,

797
00:37:35.760 --> 00:37:37.280
you know, up and running on this type of stuff.

798
00:37:37.280 --> 00:37:38.920
And what you might think is,

799
00:37:38.920 --> 00:37:40.840
oh, I mean, I have a basic understanding

800
00:37:40.840 --> 00:37:44.640
of open Claw and Clawbot and Clawed Code.

801
00:37:44.640 --> 00:37:46.720
And you might under underwrite that.

802
00:37:46.720 --> 00:37:48.000
A lot of people find that valuable.

803
00:37:48.000 --> 00:37:50.680
So being able to help businesses adopt it

804
00:37:50.680 --> 00:37:54.000
or just people in general, I think it's super good opportunity.

805
00:37:54.000 --> 00:37:56.720
Yeah, I think my only advice for people

806
00:37:56.720 --> 00:38:01.720
would be to focus, like, don't be everything to everyone.

807
00:38:02.720 --> 00:38:06.720
Like, don't help any every business help real estate agents,

808
00:38:08.920 --> 00:38:12.320
for example, or like pick a vertical that maybe you have

809
00:38:12.320 --> 00:38:15.440
some unfair advantage for some particular reason.

810
00:38:15.440 --> 00:38:17.720
And that unfair advantage doesn't necessarily mean

811
00:38:17.720 --> 00:38:19.080
you have 20 years of experience.

812
00:38:19.080 --> 00:38:21.040
It might mean just that, you know,

813
00:38:21.040 --> 00:38:24.120
you want to build something for real estate agents

814
00:38:24.120 --> 00:38:26.800
because your mom was a real estate agent.

815
00:38:26.800 --> 00:38:29.720
So you know the customer, right?

816
00:38:30.560 --> 00:38:35.560
So I think that's the way to do this.

817
00:38:36.120 --> 00:38:39.440
Absolutely, it's whatever you know,

818
00:38:39.440 --> 00:38:40.480
that's your advantage.

819
00:38:40.480 --> 00:38:43.840
And I mean, I guess as far as like some,

820
00:38:43.840 --> 00:38:46.680
maybe this doesn't apply to everyone.

821
00:38:46.680 --> 00:38:48.080
Obviously, if you're in this industry,

822
00:38:48.080 --> 00:38:48.920
then go for it.

823
00:38:48.920 --> 00:38:51.120
But like, you know, probably things to avoid,

824
00:38:51.120 --> 00:38:53.240
things that I have a lot of like red tape,

825
00:38:53.240 --> 00:38:56.320
like healthcare finance, you know,

826
00:38:56.320 --> 00:38:59.120
I recommend maybe starting something,

827
00:38:59.120 --> 00:39:00.880
yeah, like you said, you know,

828
00:39:00.880 --> 00:39:02.640
you could do even manufacturing,

829
00:39:02.640 --> 00:39:05.720
or there's a lot of distributors ships out there

830
00:39:05.720 --> 00:39:09.800
who, you know, they distribute merchandise,

831
00:39:09.800 --> 00:39:11.640
just like the demo I was showing earlier

832
00:39:11.640 --> 00:39:13.640
of that computer usage agent working.

833
00:39:14.960 --> 00:39:17.840
So yeah, I think as far as, you know,

834
00:39:17.840 --> 00:39:20.840
doing what you know, and starting there,

835
00:39:20.840 --> 00:39:22.240
and then the market will tell you,

836
00:39:22.240 --> 00:39:25.200
you know, the market will pull you into specific verticals,

837
00:39:25.200 --> 00:39:28.560
you'll start seeing ways that as you build these specific,

838
00:39:28.640 --> 00:39:29.760
this is a great point.

839
00:39:29.760 --> 00:39:31.960
As you build out these specific workflows

840
00:39:31.960 --> 00:39:34.240
around these verticals of like,

841
00:39:34.240 --> 00:39:35.240
let's say you do, you know,

842
00:39:35.240 --> 00:39:36.920
let's say you do the manufacturing thing,

843
00:39:36.920 --> 00:39:38.960
and you do manufacturing for doors.

844
00:39:40.160 --> 00:39:42.040
Over time, you're gonna have a workflow

845
00:39:42.040 --> 00:39:45.080
for almost like every kind of thing you could imagine

846
00:39:45.080 --> 00:39:46.920
in that industry.

847
00:39:46.920 --> 00:39:49.240
And if you have the agents all built out,

848
00:39:49.240 --> 00:39:51.760
can you imagine you have a workspace,

849
00:39:51.760 --> 00:39:55.280
and you invite, you know, some new company

850
00:39:55.280 --> 00:39:57.600
into this workspace that for,

851
00:39:57.600 --> 00:40:00.080
for automations, for manufacturing, for doors,

852
00:40:00.080 --> 00:40:02.320
like luxury doors, and you invite them,

853
00:40:02.320 --> 00:40:04.360
and they have all these AI employees

854
00:40:04.360 --> 00:40:07.040
in the workspace and Orgo set up ready to go.

855
00:40:07.040 --> 00:40:09.800
You just see them all here, and they're all ready to go.

856
00:40:09.800 --> 00:40:13.200
It's like, you feel like you just hired not a person,

857
00:40:13.200 --> 00:40:14.200
but a team.

858
00:40:15.760 --> 00:40:18.320
I think that's a very near future.

859
00:40:18.320 --> 00:40:20.200
In fact, I don't think there's anything stopping us

860
00:40:20.200 --> 00:40:21.800
from having that right now.

861
00:40:21.800 --> 00:40:24.720
It's all about just, you know, who's gonna go out there,

862
00:40:24.720 --> 00:40:26.680
and put in the work to actually do that.

863
00:40:26.680 --> 00:40:30.480
And if you do, I think it's pretty clear.

864
00:40:30.480 --> 00:40:33.720
Yeah, I mean, and that's why I truly believe

865
00:40:33.720 --> 00:40:36.640
that, you know, agents are the new SaaS.

866
00:40:36.640 --> 00:40:37.840
Like, you know what I mean?

867
00:40:37.840 --> 00:40:41.560
So, yeah, I agree with the vision you painted.

868
00:40:41.560 --> 00:40:44.520
I think that, you know, in the past,

869
00:40:44.520 --> 00:40:47.480
you know, we created software that we would sell

870
00:40:47.480 --> 00:40:50.880
to these businesses, and then they would have people

871
00:40:50.880 --> 00:40:54.240
actually, you know, press the buttons,

872
00:40:54.240 --> 00:40:57.080
touch the knobs to make it useful.

873
00:40:57.080 --> 00:40:59.680
Now, you're not gonna create software,

874
00:40:59.680 --> 00:41:01.880
and invite them to the software.

875
00:41:01.880 --> 00:41:03.440
You're going to create agents,

876
00:41:03.440 --> 00:41:06.040
and you're gonna invite them to the agents,

877
00:41:06.040 --> 00:41:08.840
and then the agents are going to do work

878
00:41:08.840 --> 00:41:11.240
that creates value for these companies.

879
00:41:11.240 --> 00:41:12.800
So that's the mindset shift.

880
00:41:12.800 --> 00:41:17.120
And, you know, it's only recently actually

881
00:41:17.120 --> 00:41:20.640
that people have been, you know,

882
00:41:20.640 --> 00:41:22.040
I think over the last like two weeks,

883
00:41:22.040 --> 00:41:23.240
I would say two, three weeks,

884
00:41:23.240 --> 00:41:27.440
that people have been like, on acts talking about this,

885
00:41:29.520 --> 00:41:32.080
how agents are the new SaaS.

886
00:41:32.080 --> 00:41:35.440
But I do think that, like over,

887
00:41:35.440 --> 00:41:37.400
you're gonna see over the next two to three months,

888
00:41:37.400 --> 00:41:39.360
like some really big winners,

889
00:41:40.400 --> 00:41:41.560
and you're gonna start to see it work.

890
00:41:41.560 --> 00:41:43.320
So I'm excited for people listening

891
00:41:43.320 --> 00:41:47.240
because I think that this is the type of audience

892
00:41:47.240 --> 00:41:50.880
that will act on some of this stuff,

893
00:41:50.880 --> 00:41:54.200
and it'll be interesting to you what happens.

894
00:41:54.200 --> 00:41:57.840
Yeah, yeah, I think it was a Sam Altman

895
00:41:57.840 --> 00:41:59.360
that just said, you know,

896
00:41:59.360 --> 00:42:02.320
every company is turning into an API company.

897
00:42:02.320 --> 00:42:03.320
Yeah.

898
00:42:03.320 --> 00:42:05.440
And that's interesting because interfaces

899
00:42:05.440 --> 00:42:08.680
are, in a sense, dying in that way of like,

900
00:42:08.680 --> 00:42:09.920
you know, you won't interact.

901
00:42:09.920 --> 00:42:12.080
Like the ultimate interface for whatever reason

902
00:42:12.080 --> 00:42:15.160
seems to be chat and text message.

903
00:42:16.360 --> 00:42:19.200
And it's happened twice now of like,

904
00:42:19.200 --> 00:42:21.760
okay, the chat GVT moment was a chat box.

905
00:42:21.760 --> 00:42:23.200
And now it's the open-cloth moment,

906
00:42:23.200 --> 00:42:25.640
which is like a text message or telegram.

907
00:42:25.640 --> 00:42:27.920
So it's like, okay, chat has happened twice.

908
00:42:29.560 --> 00:42:31.120
And that just means, okay,

909
00:42:31.120 --> 00:42:32.840
so we just want to weigh for our agents

910
00:42:32.840 --> 00:42:34.760
to be able to use all the tools that we use.

911
00:42:34.760 --> 00:42:36.960
And we don't really care about how it does it.

912
00:42:36.960 --> 00:42:38.200
It just needs to be able to do it,

913
00:42:38.200 --> 00:42:42.320
and it runs in the background and does it under the hood.

914
00:42:44.560 --> 00:42:46.640
So here, okay, I started,

915
00:42:46.920 --> 00:42:49.760
I built, you could see that this agent in the playground

916
00:42:49.760 --> 00:42:53.880
built out the TikTok agents.py inside the computer.

917
00:42:53.880 --> 00:42:56.240
So now I asked open-cloth,

918
00:42:56.240 --> 00:43:00.360
hey, I built the TikTok agent.py in your desktop.

919
00:43:00.360 --> 00:43:01.760
Can you take a look?

920
00:43:01.760 --> 00:43:04.000
And it's, it says, okay, I've read it.

921
00:43:04.000 --> 00:43:05.640
Here's what I see.

922
00:43:05.640 --> 00:43:08.480
There's a skeleton for, you know, using Orgo

923
00:43:09.360 --> 00:43:11.440
plus Anthropics API.

924
00:43:13.600 --> 00:43:14.680
Go ahead Greg.

925
00:43:14.680 --> 00:43:16.320
Now you keep going.

926
00:43:16.320 --> 00:43:18.320
And it was, and it was, you know,

927
00:43:18.320 --> 00:43:22.160
all the actual TikTok logic, trend detection,

928
00:43:22.160 --> 00:43:24.720
all this stuff, it's like, okay,

929
00:43:24.720 --> 00:43:26.680
maybe we should build that out.

930
00:43:26.680 --> 00:43:29.640
And it's like, oh, you got the API keys hard coded, et cetera, et cetera.

931
00:43:29.640 --> 00:43:34.240
So let's just say, let's use this script,

932
00:43:34.240 --> 00:43:38.160
build on top of it, or whatever you need to do.

933
00:43:39.160 --> 00:43:46.160
And let's spin up a Orgo VM inside of the Greg Eisenberg workspace.

934
00:43:48.160 --> 00:43:53.160
So accomplish this automation with TikTok.

935
00:43:54.160 --> 00:43:57.160
Let's demo it just working.

936
00:43:57.160 --> 00:44:04.160
We can spawn a sub agent and VM in this workspace.

937
00:44:04.160 --> 00:44:10.160
And let me just give it API key just in case it needs that.

938
00:44:12.160 --> 00:44:15.160
What really blows my mind about this whole thing is that,

939
00:44:18.160 --> 00:44:21.160
you know, I was going to say we were building a business

940
00:44:21.160 --> 00:44:22.160
in a very short amount of time,

941
00:44:22.160 --> 00:44:24.160
but it's really like we're building an asset.

942
00:44:24.160 --> 00:44:29.160
Like the amount of assets that people are going to be building

943
00:44:30.160 --> 00:44:35.160
using tools like this is going to be crazy.

944
00:44:35.160 --> 00:44:36.160
Yeah.

945
00:44:36.160 --> 00:44:38.160
It's insane.

946
00:44:38.160 --> 00:44:41.160
Honestly, it comes down to like, honestly,

947
00:44:41.160 --> 00:44:43.160
like it comes down to taste now.

948
00:44:43.160 --> 00:44:44.160
Good ideas.

949
00:44:47.160 --> 00:44:50.160
If you have a good idea, you could just build it.

950
00:44:50.160 --> 00:44:52.160
And yeah, there's going to be like, oh my goodness,

951
00:44:52.160 --> 00:44:55.160
what's going to happen with all of these assets,

952
00:44:55.160 --> 00:44:57.160
like you said, that people are just going to build and build and build.

953
00:44:57.160 --> 00:44:59.160
There's going to be so many assets.

954
00:44:59.160 --> 00:45:05.160
I think is this what we mean when we talk about the abundance

955
00:45:05.160 --> 00:45:09.160
that AI will bring, you know, of solving all these problems?

956
00:45:09.160 --> 00:45:10.160
Yeah.

957
00:45:10.160 --> 00:45:14.160
Well, I think what ends up happening is, you know,

958
00:45:14.160 --> 00:45:18.160
unfortunately, there are going to be more and more layoffs

959
00:45:18.160 --> 00:45:21.160
as AI helps with productivity.

960
00:45:21.160 --> 00:45:26.160
At the same time, I think there's going to be a renaissance,

961
00:45:26.160 --> 00:45:30.160
the golden age of entrepreneurship and people creating assets,

962
00:45:30.160 --> 00:45:33.160
products like this, one person businesses,

963
00:45:33.160 --> 00:45:38.160
and that's how I see it playing out.

964
00:45:38.160 --> 00:45:39.160
Yeah.

965
00:45:39.160 --> 00:45:43.160
Even maybe we don't know officially,

966
00:45:43.160 --> 00:45:46.160
but maybe it's already happened with Peter Steinberger,

967
00:45:46.160 --> 00:45:47.160
the creator of OpenClaw.

968
00:45:47.160 --> 00:45:52.160
You know, he just officially announced he's joining OpenAI.

969
00:45:53.160 --> 00:45:56.160
Like, I think it was just him who built OpenClaw.

970
00:45:56.160 --> 00:45:58.160
How much did he get, you know, aquahired for?

971
00:45:58.160 --> 00:46:01.160
I think it was a lot.

972
00:46:01.160 --> 00:46:04.160
So that's, I mean, that's cool.

973
00:46:04.160 --> 00:46:07.160
I think this is the best time to be a builder,

974
00:46:07.160 --> 00:46:10.160
slash tinkerer, to get creative.

975
00:46:10.160 --> 00:46:13.160
I'm excited to see what people build with OpenClaw

976
00:46:13.160 --> 00:46:16.160
and computer use agents in general of like,

977
00:46:16.160 --> 00:46:18.160
is there's so many things, you know,

978
00:46:18.160 --> 00:46:22.160
whether it's a super fast chess computer use agent

979
00:46:22.160 --> 00:46:24.160
or if it's something that's genuinely driving

980
00:46:24.160 --> 00:46:26.160
and outcoming your business?

981
00:46:26.160 --> 00:46:30.160
Like, I just think there's so many things that can be built.

982
00:46:30.160 --> 00:46:32.160
Playing around with these tools, getting familiar,

983
00:46:32.160 --> 00:46:34.160
learning how to, how to leverage them.

984
00:46:34.160 --> 00:46:38.160
Yes, AI is going to be replacing a lot of jobs,

985
00:46:38.160 --> 00:46:41.160
but also it's going to enable a lot of people to do things,

986
00:46:41.160 --> 00:46:43.160
you know, that they've never been able to build before

987
00:46:43.160 --> 00:46:45.160
and now they can do it.

988
00:46:46.160 --> 00:46:48.160
Yeah, I'm just, oh, here it is.

989
00:46:48.160 --> 00:46:50.160
Okay, so you can see,

990
00:46:50.160 --> 00:46:52.160
it spun up this computer,

991
00:46:52.160 --> 00:46:54.160
TikTok trend hunter.

992
00:46:54.160 --> 00:46:56.160
And my screen's just refreshing.

993
00:46:56.160 --> 00:46:59.160
Let me just click into it.

994
00:46:59.160 --> 00:47:03.160
And, and I could just tab back and forth a little bit

995
00:47:03.160 --> 00:47:06.160
to see, okay, the M is up, it's opening Firefox.

996
00:47:06.160 --> 00:47:08.160
Let me wait for the agent loop to start.

997
00:47:08.160 --> 00:47:09.160
Here we go.

998
00:47:09.160 --> 00:47:11.160
So it spun up its own computer.

999
00:47:11.160 --> 00:47:12.160
This is insane.

1000
00:47:12.160 --> 00:47:16.160
It's spun up its own computer, open-clotted.

1001
00:47:16.160 --> 00:47:19.160
And now it's, it's using its own Python script

1002
00:47:19.160 --> 00:47:20.160
that it just made just now.

1003
00:47:20.160 --> 00:47:22.160
We took it from Idea Browser.

1004
00:47:22.160 --> 00:47:28.160
And now it's going to go do this thing that we just built out.

1005
00:47:28.160 --> 00:47:30.160
I don't know how long this took last like 10 minutes.

1006
00:47:30.160 --> 00:47:32.160
Kind of just, you know, in between we're talking

1007
00:47:32.160 --> 00:47:34.160
and having our coffee.

1008
00:47:34.160 --> 00:47:36.160
It's like, this is insane.

1009
00:47:36.160 --> 00:47:39.160
I don't know, it kind of, it makes me, it gets me giddy.

1010
00:47:39.160 --> 00:47:43.160
It's like, and it's going to figure this out.

1011
00:47:43.160 --> 00:47:45.160
It's like, you know, it's going to debug.

1012
00:47:45.160 --> 00:47:46.160
Okay.

1013
00:47:46.160 --> 00:47:47.160
What's going on?

1014
00:47:47.160 --> 00:47:49.160
Why am I on the ads.tiktok.com?

1015
00:47:49.160 --> 00:47:50.160
Let me reroute myself.

1016
00:47:50.160 --> 00:47:52.160
I'm sure it's going to figure all this out.

1017
00:47:52.160 --> 00:47:54.160
But it's just cooking, you know?

1018
00:47:54.160 --> 00:47:56.160
Crazy dude.

1019
00:47:56.160 --> 00:47:58.160
Crazy.

1020
00:47:58.160 --> 00:48:03.160
Anything else you want to cover before we head out?

1021
00:48:03.160 --> 00:48:08.160
Yeah, I think as far as other things to cover,

1022
00:48:08.160 --> 00:48:14.160
I mean, I just want people to start thinking about these tools.

1023
00:48:14.160 --> 00:48:18.160
You know, yes, they're, once again, they're great personal assistants.

1024
00:48:18.160 --> 00:48:22.160
But if you start thinking of OpenClaw as an in a den,

1025
00:48:22.160 --> 00:48:27.160
or you start thinking of it as, you know, like a Lindy AI of like,

1026
00:48:27.160 --> 00:48:30.160
you know, people, there are real businesses right now.

1027
00:48:30.160 --> 00:48:35.160
Like, we can go to Upwork right now and find jobs that are being posted

1028
00:48:36.160 --> 00:48:39.160
around, you know, things like this is what you do.

1029
00:48:39.160 --> 00:48:40.160
You just go to Upwork.

1030
00:48:40.160 --> 00:48:43.160
Upwork's great because you get to see what the market's asking for.

1031
00:48:43.160 --> 00:48:46.160
And you just type in a robotic process automation.

1032
00:48:46.160 --> 00:48:53.160
You know, this old outdated way of programmatically automating tasks

1033
00:48:53.160 --> 00:48:58.160
that it's like clunky and it breaks and it's not intelligent.

1034
00:48:58.160 --> 00:49:03.160
If the button isn't in the exact UI space that you just delegated it to,

1035
00:49:03.160 --> 00:49:04.160
it won't work.

1036
00:49:04.160 --> 00:49:08.160
And you go here, like, posted yesterday, Android RPA automation,

1037
00:49:08.160 --> 00:49:12.160
posted yesterday, automation pipeline for client upload.

1038
00:49:12.160 --> 00:49:17.160
You go here, you could just do, let's find $500, $1,000, $5,000.

1039
00:49:17.160 --> 00:49:19.160
Let's look at all these projects.

1040
00:49:19.160 --> 00:49:23.160
Okay, this one, boom, right here.

1041
00:49:23.160 --> 00:49:25.160
$1,000 budget.

1042
00:49:25.160 --> 00:49:29.160
I'm looking for experienced automation engineer to build desktop automation

1043
00:49:29.160 --> 00:49:32.160
for my software business.

1044
00:49:32.160 --> 00:49:34.160
We sell a specialized dynamic PDF.

1045
00:49:34.160 --> 00:49:38.160
You take all this context, give it to OpenClaw, give it to CloudCode.

1046
00:49:38.160 --> 00:49:43.160
How much of it can you build out as a demo based off of this context alone?

1047
00:49:43.160 --> 00:49:44.160
Send a proposal.

1048
00:49:44.160 --> 00:49:48.160
You have your first customer right here, $1,000.

1049
00:49:48.160 --> 00:49:52.160
Get some case studies, leverage that.

1050
00:49:52.160 --> 00:49:55.160
Maybe go deeper into this industry that this person's in.

1051
00:49:55.160 --> 00:49:58.160
Start building out specialized workplaces in Cloudbot OpenClaw.

1052
00:49:59.160 --> 00:50:03.160
For all the different vertical use cases in that, create a workspace of it.

1053
00:50:03.160 --> 00:50:06.160
I think that's where we're at right now, and I'm excited by this.

1054
00:50:06.160 --> 00:50:10.160
So, yeah, I guess we'll just see where it goes.

1055
00:50:10.160 --> 00:50:13.160
Yeah, okay, this needs a little debugging.

1056
00:50:13.160 --> 00:50:15.160
As to be expected, we spent 10 minutes on it.

1057
00:50:15.160 --> 00:50:18.160
But I think you get the gist.

1058
00:50:18.160 --> 00:50:21.160
And yeah, I'm excited to see what everyone builds.

1059
00:50:21.160 --> 00:50:24.160
From your lips to God's ears, baby.

1060
00:50:25.160 --> 00:50:30.160
I think your approach makes complete sense.

1061
00:50:30.160 --> 00:50:34.160
It's the exact approach I would use, totally recommended.

1062
00:50:34.160 --> 00:50:36.160
People get your hands dirty, get tinkering.

1063
00:50:36.160 --> 00:50:39.160
I'm excited for what you build.

1064
00:50:39.160 --> 00:50:42.160
Nick doesn't do a lot of podcasts.

1065
00:50:42.160 --> 00:50:46.160
I think I was only able to find him to do one livestream before.

1066
00:50:46.160 --> 00:50:50.160
So, show him some love in the comment section.

1067
00:50:50.160 --> 00:50:53.160
Like the video, show him some love.

1068
00:50:53.160 --> 00:50:59.160
And Nick, I hope you come back on, share more use cases.

1069
00:50:59.160 --> 00:51:05.160
I'm going to be sharing more use cases that I haven't done too many publicly.

1070
00:51:05.160 --> 00:51:09.160
I'm going to be sharing more of my use cases, both on virtual machines,

1071
00:51:09.160 --> 00:51:13.160
and on my own Mac Mini for my OpenClaw stuff.

1072
00:51:13.160 --> 00:51:16.160
So, get ready for that, folks.

1073
00:51:16.160 --> 00:51:24.160
And Nick, wait, is there anything else before we go that you want to share?

1074
00:51:24.160 --> 00:51:26.160
I just want to share with you, Greg.

1075
00:51:26.160 --> 00:51:30.160
You don't know this, but I've been a long time follower.

1076
00:51:30.160 --> 00:51:36.160
This is my YouTube rewind 2025 top 0.5%.

1077
00:51:36.160 --> 00:51:40.160
So, if you're watching this and you love Greg's podcast,

1078
00:51:40.160 --> 00:51:43.160
you love his videos, you get building on top of it,

1079
00:51:43.160 --> 00:51:46.160
you put cool stuff out there.

1080
00:51:46.160 --> 00:51:51.160
You join your top YouTube channel on your rewinds of the year.

1081
00:51:51.160 --> 00:51:52.160
I love it.

1082
00:51:52.160 --> 00:51:53.160
I love it, Nick.

1083
00:51:53.160 --> 00:51:54.160
You're a legend.

1084
00:51:54.160 --> 00:51:55.160
You got to come back on.

1085
00:51:55.160 --> 00:51:56.160
You're one of us.

1086
00:51:56.160 --> 00:51:57.160
You're one of us.

1087
00:51:57.160 --> 00:52:00.160
So, appreciate that.

1088
00:52:00.160 --> 00:52:01.160
Thank you, Greg.

1089
00:52:01.160 --> 00:52:03.160
Thank you for having me.