03/16/2026 - Exploring OpenClaw Agents and the Reality of Building a Personal AI Assistant

Description

Chapters

Transcript

0:06 people.
0:29 down.
0:37 What do you think, champ?
1:08 Well, you don't love me and you bet I
1:11 felt it. Tried to beat you, but you so
1:13 hot that I melted. fill right through
1:15 the cracks. Now I'm trying to get back
1:20 before the cool down and I'm giving it
1:22 my best. Nothing going to stop me but
1:24 divine an event. Making it again my turn
1:28 to win some old
1:31 but I won't hesitate
1:35 no more. No more. It cannot wait. I'm
1:41 young.
1:48 Yeah.
1:50 Yeah. Yeah.
1:59 What's going down good people? Look at
2:02 that [ __ ] hair. That is crazy hair.
2:05 That is crazy hair. He's got crazy hair.
2:07 Happy Monday. Crazy hair. Welcome to AI
2:10 Learning Lab
2:12 in ASMR Village. Hi, welcome to the AI
2:16 learning lab. How are y'all doing today?
2:20 Let me go get some cake.
2:23 We'll have some delicious cake. I'll
2:25 chew some cake on AI Learning Lab.
2:29 You'll hear me chew and your misophonia
2:32 will make you so angry you'll throw your
2:34 computer out the window. And you won't
2:36 ever need to worry about AI again. It'll
2:39 be beautiful.
2:41 Here's Johnny.
3:20 I heard on the Tik Tok yesterday
3:24 that if you have misophonia, which I
3:26 definitely have,
3:28 someone chewing, smacking their lips,
3:31 It's just like
3:36 it drives me [ __ ] crazy. Apparently,
3:38 it's a sign of intelligence.
3:44 You know what that You know, you know
3:46 who came up with that? If you've got
3:47 misophonia, it's a sign of intelligence.
3:50 It was a dumb guy with anger issues.
5:04 Oh my god, I must be very dumb because I
5:07 watch those people eating videos on
5:09 repeat.
5:16 I I cannot
5:18 I get one of these.
5:21 Oh jeez. I even myself I can't do it.
6:01 Wait.
6:15 Here
6:29 we are.
7:04 I don't even call you out of trouble.
7:10 I didn't mean to cause you any pain.
7:19 Only one time to see you laughing.
7:28 Black bars. Yeah, the hair's there's
7:30 nothing we can do about the hair. Hair
7:31 the hair is what it is.
7:34 as as bald guys like to tell me. At
7:36 least you got some
7:45 resemble that remark.
7:47 You got some
7:50 um What am I trying to do? Oh, show
7:52 always show bookmark bar. I didn't Oh,
7:54 no. I'm fine. Okay.
7:57 I was missing what was happening in the
7:59 black bar. I didn't put it there.
8:04 Someone just invited me to co-host with
8:07 them somewhere.
8:09 All right, there's that.
8:12 There's me looking douchy
8:15 like like a douchy 80s rocker
8:19 in 2026. Yeah, man.
8:29 Your coffee is in the shot.
8:33 Oh yeah,
8:37 that's cool. That's kind of cool
8:39 actually. That was cool in the shot.
9:14 My locator, Los Angeles,
9:17 I am profoundly bald and think you have
9:19 pretty great hair. Thank you. It's just,
9:22 listen, I I like my hair, don't get me
9:24 wrong. Not not bitching about the hair.
9:26 It's just it's a little unpredictable.
9:28 Sometimes I get here at night and it's
9:30 like flopped down in my head or and then
9:33 sometimes I get here and I think it's
9:35 flopped down in my head and I look like
9:38 Brian Settzer from the Stray Cats. The
9:40 Stray Cats, if you're younger than 72,
9:43 we're a band
9:45 for old people.
9:52 Oh my god.
9:57 Waiting for the song You Got a Friend. I
9:59 used to sing that.
10:03 Hey, dude.
10:09 Um
10:22 I see the rabbit there. Did you hear
10:25 people are using that to access their
10:27 open claw? That's kind of a cool idea.
10:34 I could I could use I could put my
10:36 rabbit to use at this point with my with
10:39 my Adam. I'm I'm I'm I am frustrated
10:42 with Adam. Here's one of the things I'm
10:46 learning about openclaw agents is much
10:49 like everything in life.
10:53 They are a perfect reflector of you.
10:58 Max headroom. Yeah, I got a little max
11:00 headroom going on. Although Max was a
11:03 little slimmer on the sides. I've got
11:04 the This is the This is When I go in for
11:07 the uh for the haircut, I ask for this.
11:10 I just I say I want the the Sears Robuck
11:13 carpet salesman. They know what that
11:15 means. That's what they That's what they
11:16 give me. These
11:18 this sort of curved Hi, welcome Sears.
11:21 I'm Kyle.
11:23 Can I interest you in a washer dryer
11:25 set? We got combos.
11:44 I mentioned Hollywood Squares to a 24
11:47 year old this week. They didn't know
11:48 what it was. Yeah. Yeah. There's there's
11:51 there there's those cultural markers in
11:53 life where you realize, oh, I'm no
11:56 longer relevant.
11:57 Mine happened in 1997.
12:01 I made a Soilent green joke to a room
12:03 full of 24 year olds. There were blank
12:06 stairs. They didn't know what Soil and
12:08 Green was.
12:26 There's been something, baby, I've been
12:28 trying to say.
12:32 What age? And it seems I don't know how.
12:35 Wow.
12:38 Past and a future now surrounding me.
12:44 Surrender to whatever true thrill can be
12:46 found.
12:49 There's been little trouble
12:53 since you came to my rescue.
13:00 And if you were like all of the rest, I
13:03 would have quit long ago. But I couldn't
13:06 do that.
13:12 Oh, tell me now when I never went too
13:16 well.
13:20 Crazy make him cold as hell.
13:25 I know one night you wish me well.
13:30 But in spotty trying,
13:32 still going to have to find my own way
13:36 through
14:12 Oh, good lordy good people. What's
14:15 happening? What? What? What? The dog
14:18 can't handle the blues. The dog just
14:20 loves anything music. A minor. Right.
14:22 Right. Champy.
14:24 Huh?
14:29 No, you're not going to sing to a minor
14:30 tonight.
14:54 Um
14:56 um I had an interesting meeting today. I
14:58 was I I uh I consult with a company
15:03 um here in Denver that's doing
15:06 like really interesting stuff with
15:08 health
15:10 liability for doctors and they're
15:13 building really really fascinating
15:16 technology to sort of take 30 years
15:19 worth of this business's data and like
15:21 modernize it and AI it
15:25 meteller
15:28 and the avatar designer basically um to
15:32 to think about ways we can take the
15:34 learnings out of this data and turn it
15:36 into stories and education and content
15:39 things like that.
15:41 And
15:44 so I get to see the the architecture,
15:46 how they're building it, and
15:49 kind of the
15:53 watching a technologist who who like
15:56 incredibly incredibly competent,
15:59 sophisticated
16:01 lead engineer, watching him explain the
16:04 platform and the
16:07 how everything talks to everything and
16:10 which LL LMS they can use because
16:12 they're approved for healthcare and
16:15 which ones they can't and how they're
16:17 segregating that data and you know
16:18 because of HIPPA and things like that.
16:21 Um I love the idea of having an agent
16:24 but I have no bandwidth to set one up
16:25 right now. You do need bandwidth to set
16:27 one up. Um and I also didn't bring in a
16:30 water so let me go get that in a second.
16:31 But um but I shared with them so they
16:34 wanted they wanted me to talk to them
16:36 about storytelling and about avatars and
16:39 I talked about that. But I led the
16:40 conversation off
16:43 with a little bit of talk about
16:50 who does it make sense to retain right
16:52 now. If you're a company and you've got
16:54 to lay people off, who do you keep?
16:57 Who do you keep? Who's who's valuable
16:59 today might not be who's valuable three
17:01 years from now or even a year from now?
17:04 So, how do you identify those people? So
17:06 I was talking a little bit about that
17:07 about you want to find people that are
17:09 curious and adaptable and
17:14 can look at the whole of a problem not
17:16 just the peace parts of it. Um and
17:19 they've got judgment and they've got
17:21 taste
17:27 and none of those things are skills.
17:29 Those are what would technically be
17:31 called soft skills which in today's
17:33 world are completely devalued. Like if
17:36 you have a soft skills program in your
17:38 company and your company hits a rough
17:40 patch financially, first thing that's
17:42 going is the soft skills training
17:46 because hard skills skills
17:49 expertise is valued. Soft skills are
17:51 not. That iceberg's about to flip upside
17:55 down. Hydrate. Yeah, I know.
17:58 Look at that producer Brandon taking
18:00 care of me. Thank you.
18:04 Corporate corporate won't touch an
18:06 entrepreneur with a 10 foot pole. No,
18:08 they don't know what to listen.
18:09 Corporations do not know what to do with
18:11 with entrepreneurs. They don't know what
18:12 to do with creative people. They don't
18:14 know what to do with generalists. I'll
18:15 be back. Pull. Please entertain
18:17 yourselves.
18:32 What
18:34 idiot?
19:30 So, I'm back. I got my water. I'm
19:32 hydrating.
19:35 So, anyway, so this Kai Health, so I was
19:38 talking about that. I got a good song.
19:42 Sing You Don't Have to by Marilyn Maku.
19:45 But it's sung in a deep man's country
19:47 voice. Be the first.
19:50 Sing You Don't Have to Be a Star by
19:52 Marilyn Maku. All right, I'll find that.
19:55 Here's my problem. My Locator is uh I
19:59 haven't learned a new song in ages,
20:03 and I know I know like 17 of them. So,
20:05 if you come here enough, you'll hear
20:07 them all.
20:09 You think you're sick of the ones I play
20:10 now? Wait till you've heard them all 17
20:12 times. Um, I'm fighting with myself at
20:15 the expense
20:17 of the new Mac. Totally can't afford it,
20:20 but new toy. Congrats.
20:24 Well, I Yeah, I a new Mac is somewhere
20:27 in my future here. Am I that chatty
20:30 tonight or is everyone else that quiet?
20:32 Well, it's it's a there's not a ton of
20:33 folks in here, so if we can get some
20:35 more folks in here, it'll get a little
20:37 more
20:40 peppy, chatty. Um, back to Kai Health.
20:48 Do they know this is an AI channel? Do
20:50 we? Oh, that was cold. That was cold.
20:53 Well, I'm going to talk about AI. Like,
20:55 what I'm going to actually talk about is
20:58 I showed Kai Health this slide.
21:01 Kyle is the name of the company
21:06 and
21:14 this here slide
21:17 the seven economies.
21:20 So what they asked me to do is come in
21:22 and talk about avatars. And what I did
21:24 was I came in and I talked about what I
21:26 think's about to happen to our to our
21:29 workforce. What am I? Chopped liver.
21:32 Chef Kelly, what's happening?
21:36 Um,
21:37 give me a B for Brandon.
21:42 My brain. Oh, black bar. B for black
21:44 bar. Damn it. I knew it was something.
21:46 Okay, there you go.
21:49 Um,
21:51 so if you haven't seen this this
21:53 graphic, this infographic that I made,
21:56 um,
21:57 this was a thing that came up with on
22:00 the live, I don't know, a week or two
22:01 ago, and
22:04 it's called the seven economies, and
22:05 this is the answer to the question that
22:08 Kelly Camp posed because she's she's
22:10 overly chatty over there in the TikTok.
22:13 She's just chatting away.
22:15 But she said one night, "Hey, you keep
22:18 saying it's going to get weird. What's
22:19 weird going to look like?" And so th
22:21 this was my answer. And and I just sort
22:24 of thought of it on the fly. And and as
22:26 I thought about it, it seemed really
22:27 clear. And now that I've visualized it,
22:29 it seems even more clear. And so I I
22:33 shared this today
22:35 with the with the folks at Kai Health.
22:37 And what was cool about it, and what it
22:38 is is basically from left to right.
22:42 Do you know that? Do you think Kelly has
22:44 a recipe for turtle soup?
22:50 Sorry for the extra e and Kelly. No,
22:52 that's right. It's it's it's ke ly e.
22:56 You got it. You got it almost right
22:58 there. Um,
23:03 but this is basically on the left hand
23:04 side of this chart is the analog
23:08 economy, right? No computers or just,
23:11 you know, go farming. Um,
23:14 the second economy is we're just going
23:16 to do things the way we always have. The
23:18 third economy is we'll use AI a little
23:20 bit to make things more efficient, but
23:22 we don't really give a [ __ ] about it.
23:23 The transition economy is number four.
23:25 That's like, oo, AI is really cool.
23:27 We're going to keep doing our business,
23:28 but we're going to try to innovate
23:29 within it. And then the the last three
23:32 economies, five is the AI aggressor
23:34 economy where you've got a seauite
23:36 that's like, [ __ ] it, let's blow up
23:38 everything and and like really live into
23:41 the future. The six is the AI native
23:44 economy. So that's people starting
23:46 companies from scratch, you know, highly
23:49 AI powered. And then
23:52 uh economy seven is the multi- aent
23:54 operator economy, right? So open claw
23:57 running agents and and having businesses
24:00 that are, you know, 20 autonomous agents
24:03 off doing your thing.
24:05 Um, so I presented this and what was
24:08 amazing was how quickly everyone
24:11 self-identified into segments and they
24:13 they identified not only themselves but
24:15 their customers. This might be cool for
24:17 you, Kelly. Um they were saying, "Oh,
24:20 well so so Brandt the CEO
24:24 thinks thinks they're an economy for
24:26 kind of company where they're doing
24:28 innovative stuff with some existing
24:30 tech, you know, existing business
24:32 models." And then the the lead engineer
24:35 was like, "No, no, we're like a five,
24:37 six or seven company." Because he was
24:39 talking about how they're going to have
24:41 swarms of agents doing things and then,
24:44 you know, no one's going to be able to
24:45 keep up.
24:47 So, it's a really interesting
24:48 conversation, but they also work in
24:50 healthcare and a lot of health care
24:51 companies are going to live down in
24:53 legacy economy or efficiency economy.
24:57 They're going to be very very slow and
24:58 they're going to they're going to be
24:59 slow to adapt and slow to, you know,
25:02 just do basic basic stuff. Um, ah, and
25:08 so I thought that was a really cool
25:10 conversation like you know, understand
25:12 where you are, understand where you want
25:14 to be.
25:15 Um, you know, if the CEO thinks you're
25:18 in economy 4 and the the uh the CTO
25:23 wants to take it to economy 7, well then
25:25 there's there's some talk to do. There's
25:27 some conversations to be had. Um, I was
25:31 very impressed with Gemini's pros
25:32 writing lately.
25:36 Just since last month, it's dramatically
25:37 improved or it got better or I got
25:39 better at prompting. I noticed the the
25:42 last time I used Gemini for creative
25:44 writing was probably six months ago,
25:46 eight months ago maybe. And I thought it
25:48 it surprised me how good it was. Um
25:50 Gemini's Gemini is quite a strong model.
25:53 It's quite good.
26:00 Um
26:03 let's see what else. Um, Manis got into
26:05 the game with um
26:10 with a downloadable an app that you
26:17 you can run Manis agents locally on your
26:20 machine. So, similar to Claude Co-work
26:23 um similar to OpenClaw, similar like you
26:27 know these local control your local
26:29 computer kind of things. So, everyone's
26:31 going to start getting into the game. Um
26:33 Jensen Hang from uh Nvidia today said
26:37 that he thinks that uh OpenClaw is the
26:40 most significant technology in ages. And
26:45 um
26:47 Open Claw in terms of open-source
26:50 adoption
26:51 got to a level of adoption that took
26:54 Linux 30 years to get to and did it in a
26:58 few weeks.
27:00 the the the open claw adoption curve is
27:02 basically a straight line. It's a
27:04 straight line up. Um let's see. Black
27:08 bar.
27:23 I haven't had a Mac in 20 years, but I'm
27:24 so over Microsoft. I'm telling you, one
27:27 of the reasons I'm hating um I'm hating
27:31 Adam right now is because I I did it on
27:33 a Windows machine. I need to get a Mac.
27:35 I just need to get a Mac. Uh
27:41 yeah,
27:42 yeah, yeah, yeah. Well, Linux sucks.
27:45 Yeah, exactly. Linux does suck. Chef
27:49 Kelly
28:07 I use Gemini a lot. It seems like the
28:10 quiet smart kid. That's interesting.
28:16 So
28:18 So Silver Fox, did you buy one? Did you
28:20 buy a Mac Mini?
28:23 Did you Did you get sucked into the into
28:26 the
28:28 You know, I'm guessing that that uh
28:31 Peter Steinberger is just like, you
28:33 know, Steve Jobs's cousin or something
28:37 like that. Like it's just a marketing
28:39 ploy to sell mech minis.
29:03 Linux is the foundational layer of
29:05 everything. No, my PC hard drive is
29:08 done. So, I'm at a precipice.
29:10 I want the new M5. I know. The new M5
29:12 looks really good. That's what I
29:14 recommended. my my co-founder at
29:16 Storymine wants a machine
29:19 and she's like, "ChachiPT said I should
29:21 get a Mac Mini." And I said, "No, get
29:23 the new M5 Power Book MacBook."
29:26 And uh
29:28 and then Chachi PT goes, "Oh, your
29:30 founder is very smart. He's getting you
29:32 the latest latest and greatest.
29:36 What's going on, folks?" Nothing LDG.
29:38 So, thank you to LDG. LDG is helping us
29:43 make Adam more productive.
29:46 So LDG wrote a cool script, a cool
29:49 application that pulls down um what you
29:53 call it transcripts from the YouTube and
29:56 puts them in a database and then lets
29:59 our open claw Adam access them. I've
30:03 somehow screwed up. Adam is still
30:05 screwed up LD LDG, I think. But we'll
30:08 see after tonight. So, we'll see
30:09 tomorrow if this episode generates a new
30:13 transcript and and the thing works.
30:15 We'll we'll see.
30:17 But right now, I'm just like, h I don't
30:19 know if Adam works. And I didn't work on
30:22 him today. Um because I had I had to be
30:26 in a in a in that session with Kai
30:28 Health today. So, but thank you for your
30:30 work. Oh, no. I know. Exactly.
30:34 He might be okay. Tell LG how much I
30:36 appreciate him indexing YouTube videos.
30:38 I use them often. Oh, that's very cool.
30:41 Uh, Kelly Camp uses the stuff you're
30:44 building, LDG.
30:47 It was so funny. We're there working
30:49 working away on a Sunday and uh LDG G's
30:53 like, I'm I'm lazy. I I just sit around
30:56 not doing any work. I'm like, you're
30:57 working your ass off. He goes, oh, this
30:59 is fun.
31:01 You're amazing, dude.
31:04 I thought that was quite sweet. Um,
31:10 let's see. What do we want to do?
31:15 It's Monday.
31:17 I don't feel particularly passionate
31:19 about anything right now. Although,
31:20 someone could team me up could tee me up
31:23 on a topic. I could probably get wound
31:25 up.
31:32 M
31:34 St. Patrick's Day tomorrow.
31:38 LDG have been using actually, you know,
31:43 that could be fun.
31:49 Oh,
31:55 what
31:57 is the original
32:00 St. Patrick's Day. Like,
32:09 why not give OpenClaw a notion account?
32:11 I could give OpenClaw a notion account.
32:13 Right now, my Claudebot has a his own
32:17 email. He's got a an AI salon email.
32:21 He's got
32:23 um access to to Google to Google Drive
32:26 and Gmail and [ __ ] like that.
32:29 Um, he's got some some database stuff
32:33 that he's doing. He's got a a vector
32:35 database for memories. So, he's got some
32:38 he's got some decent stuff. I haven't
32:40 added things like self-improvement yet.
32:42 You mentioned Manis. What are they doing
32:44 these days? They just today, Kelly, they
32:46 launched a a Mac and a Windows
32:49 application that you can run just like
32:51 uh Claude Co-work. You can run on your
32:54 local machine and do [ __ ] Um, so that's
32:56 what's going on. That's what's going on
32:58 with Manis.
33:11 Um, let's see. When did it start?
33:25 Oh, St. Patrick days was in the early
33:28 1600s.
33:29 The Catholic Church placed it on the
33:31 lurggical calendar in 1631.
33:36 All right,
33:38 let's see. So, let's see. Make me a
33:42 painting
33:45 that you feel
33:50 would
33:52 be a contempt or
33:56 very temp or
33:59 style
34:04 and subject matter
34:09 from 1631.
34:14 All right,
34:16 we're going to have chatbt make us
34:18 historical St. Patrick's Day paintings.
34:23 Play along, kids. You can play along at
34:25 home.
34:30 Oh, I get I think we're getting some
34:31 clergy. We're getting some Wait, that
34:34 was fast.
34:36 Holy [ __ ] that was fast.
34:44 So, two two things that I noticed today
34:46 about chatbt. Now, one of those one of
34:49 them is that the image that image
34:51 generation was quite fast. I'm not used
34:52 to that. And then GPT 5.4 Pro launched
34:57 today for team accounts.
35:01 All right. So, so that was 1631. There's
35:03 there's St. Patrick
35:06 St. St. Patrick blessing in a church.
35:09 I'm going to say okay. Uh, what was a
35:12 good
35:15 next date in
35:19 or let's see good next milestone
35:24 in the history of St. Patrick's.
35:31 In fact, I'm going to turn on deep
35:33 research.
35:36 Do I want deep research? No, I don't
35:37 give a [ __ ] In fact, I'll just say
35:39 research
35:41 research.
35:43 Um
35:48 major
35:54 historical markers
35:58 that would be relevant
36:03 to St.
36:05 Patty's Day
36:08 from then until now.
36:14 That's chat. Yeah, that was chat GBT.
36:15 Did you see how fast that was?
36:18 Mad Libs. Cheers. Cheers, brethren. Also
36:21 events. I'm pulling together a real
36:23 timeline from the formal feast day
36:25 forward. It's thinking
36:28 bratannica localvhistory.com
36:31 local.gov of Britannica.
36:35 I've got early religious. Now I'm
36:37 filling in the milestones that explain
36:40 how it became the green global thing
36:42 people recognize today
36:47 cuz St. Patrick's Day is really [ __ ]
36:49 annoying, you know? It's just just drunk
36:53 people being idiots. So let's let's
36:56 figure out how we got there. Let's let's
36:58 do some learning. We're doing some
37:00 educating.
37:08 All right.
37:10 A strong next milestone after 1631 is
37:13 1762
37:15 where Irish soldiers serving in the
37:16 British army marched in New York City.
37:19 Oh, that's good. Let's see a 16 or no
37:24 1762
37:27 image in
37:29 the style
37:32 of painting for that year.
37:37 Make it uh landscape.
37:48 Oh, it's still writing. I see
37:51 orientation.
37:57 1903 official public holiday in Ireland.
38:00 1931
38:02 first state sponsored Dublin parade.
38:04 1962 Chicago River goes green. So it was
38:08 Chicago's fault. 9596 Ireland launches
38:12 the modern St. Patrick's Day festival.
38:15 All right, we're doing 1762 now.
38:20 It's a real trip in New York City. Drunk
38:22 people everywhere. Yeah. New York City
38:23 around St. Patrick's Day. Especially
38:25 downtown, the Wall Street area.
38:29 Jesus.
38:34 Wait, look how fast this is going.
38:38 Bang.
38:40 Holy [ __ ] That's
38:43 10 seconds.
38:45 Chat GBT used to be really slow. And is
38:48 that 16 by9?
38:50 I think it's 16 by9.
38:55 Copy image. Do I have Photoshop open?
38:57 Oh, I know what I can do. And go to
38:59 Keynote.
39:01 Paste it in there. Hang on. Copy image.
39:05 Paste. Oh, because paste.
39:09 Oh, it's not it's not 16 by9. Okay.
39:15 I I still don't have any faith in them.
39:18 Yeah. See See the white bars on the left
39:20 and right. [ __ ] All right.
39:25 All right. What was the next one? 1903.
39:29 Um let's see.
39:31 There has to be something between
39:37 1762 and 1903.
39:41 Come on.
39:43 Give me
39:48 something.
39:56 St. Patty's Day gathers on Greenville
39:58 Avenue in Dallas. It's nuts.
40:02 I played with both sides. As a drinker,
40:04 I loved it. As a bartender, it made me
40:06 feel bad things about people.
40:12 If you've bartended, you have a
40:14 relationship with St. Patty's Day. I
40:16 bartended for two years in New York
40:18 City.
40:20 Oh my god.
40:23 Yeah, bartending. Bartending's wild.
40:26 It's wild.
40:30 Uh uh
40:32 late 1700s green shamrock
40:37 famine immigration makes huge diaspora
40:40 identity event in in the Americas. All
40:43 right,
40:44 let's do the
40:52 18
40:55 40s stuff.
41:01 abusive relationship. It is St.
41:03 Patrick's Day drinkers and bartenders.
41:07 That's That's an abusive relationship.
41:10 It That is a It is. It's It's bad.
41:13 Luckily, I I bartended at a bar that
41:16 just had
41:18 um low-level non-officer cops in it down
41:21 near one police plaza in Manhattan in
41:23 Chinatown and uh or right next to
41:26 Chinatown.
41:27 And so my bar, it had Irish people in
41:31 it, but like most of the Irish officers
41:35 were at the bar up the street. So So
41:38 they got all the idiots. I was with the
41:40 with the guys in the police station that
41:43 would beat people up just because they
41:44 thought it was fun. Those were my
41:46 customers. So if some college idiot came
41:50 in all dressed in green, they didn't
41:51 last long at my bar. Not because of me,
41:54 I would have served them. But
41:58 my customers, they didn't take kindly to
42:00 that. Um, what happened here?
42:05 Why did we stop and make the image?
42:29 Oh, that's cool. All right. So, wait.
42:30 So, we've got
42:32 You know what we're going to do? No,
42:34 Kyle.
42:36 We're going to go
42:54 St.
42:56 Patrick's
42:58 Day.
43:03 A photograph
43:06 graphic
43:10 history.
43:15 Oh, by the way, by the way,
43:19 important.
43:23 If you have not seen the movie Hamnet,
43:28 you got to see that movie.
43:32 Listen, it's slow. It's depressing.
43:38 They color graded it as if people
43:42 back in those days all had glaucoma.
43:46 You can barely see it. It's beautiful.
43:49 But the performance by Jesse Buckley,
43:52 the the wife of William Shakespeare,
43:56 is is like no other performance I've
43:58 ever seen. I I was like I watched the
44:00 Oscars last night just to see if like
44:03 she had to win the Oscar.
44:06 She had to win the Oscar and she won the
44:08 Oscar. It was really good. So if you
44:10 haven't seen it, do yourself a favor.
44:14 Do yourself a favor.
44:18 Okay, what are we going to do here?
44:19 We're going to go
44:23 do that one.
44:26 Bang.
44:28 Let's see. How do we want to do this?
44:31 We're gonna go We're gonna go.
44:48 All right. Beautiful. Beautiful.
44:51 Fantastic. So, this is going to be 1631.
44:56 Um,
45:01 an old dude
45:05 blesses a peasant.
45:13 It's going to be fun. I think I'm going
45:15 to like this. Our little history here.
45:17 It's a little history lesson. You're
45:19 welcome.
45:27 An old dude blesses a peasant. This is
45:31 good. It's good.
45:34 That and we're come over here and we're
45:36 go down and we're going to get
45:39 All right. Wait. What's this? This is
45:41 1762.
45:43 Parade tradition takes off in New York.
45:47 First parade in New York City. Here we
45:49 go. 1762. Copy image.
46:11 1762.
46:13 By the way, you could be making St.
46:15 Patty's Day images and putting them in
46:18 irregulars if you want.
46:20 Um,
46:22 first parade
46:25 in New York City.
46:29 All right. Beautiful. Fantastic. That's
46:33 really good. Garb.
46:36 Yeah. Thanks, Siri. Siri, you've been
46:38 excused.
46:41 Nobody needs another Siri in their
46:43 house, do they?
46:48 Hey, young man. Want to see me make this
46:51 snake disappear?
46:57 That just reminded me of a joke? I It'll
47:00 get me banned if I say it. Um, but the
47:04 punchline is tada.
47:14 Uh, wait, that's 1840s. Let's see what
47:17 was 1840s.
47:19 Famine integration makes
47:23 um,
47:26 we did a parade for 1762.
47:31 You told me
47:34 um, 1840s
47:41 to 50s
47:44 was about
47:46 famine. Uh, make an image about that,
47:52 dumb dumb.
47:57 This feels like a job for Notebook LM.
48:01 Nah, this is at this right now. All this
48:04 is is an add
48:07 um moment that you're that you're
48:09 unfortunately having to go along for the
48:12 ride on.
48:14 But we did
48:17 we did learn that the uh image
48:20 generation
48:21 is uh is quite fast for chat GPT when it
48:25 actually decides to make an image. What
48:27 is it doing now?
48:46 Did you guys see that? There was a big a
48:48 bunch of engineers from Anthropic,
48:51 OpenAI,
48:53 maybe Grock.
48:57 They all got together and basically
48:59 said, um, AI is coming really fast and
49:02 we're not prepared for it. It's like,
49:05 yeah,
49:07 I know what what's going on, chat GBT.
49:11 Um, let's see. What are you
49:15 doing?
49:22 I made the wrong image.
49:27 Yeah, no [ __ ] Make the right image.
49:31 Damn it.
49:36 This is what it's like working with
49:37 Adam. This is what it's This is what
49:39 it's like. I'm like, "Hey, could you do
49:41 that thing for me?" Sure thing. I'm off
49:43 to do that thing for you. Nothing.
49:45 Nothing. Nothing.
49:47 Did you make that thing for me? Oh, no.
49:49 I had an error.
49:51 Uhhuh.
49:53 When were you going to tell me? Yeah,
49:54 that's You're right. You're right, boss.
49:57 That's that's that's my bad.
50:00 But boy, thank you for uh thank you for
50:03 for helping me out with that.
50:09 Oh my god, that's depressing. Okay,
50:11 that's good though.
50:14 St. Patrick's Day and immigrant
50:15 community. Very sad. Very sad. Okay, so
50:19 back here. Now we're going to duplicate
50:21 this.
50:33 gank.
50:36 All right. So, we're going to do we'll
50:37 call this 1847
50:40 and then we'll go um
50:44 immigrants
50:46 and St.
50:50 Patty's St.
50:53 patch
50:54 day.
50:58 All right,
51:01 beautiful. Let's go back to the chat.
51:04 JPai.
51:05 All right, so now we need I think 1903
51:08 was the next one.
51:11 1903.
51:13 Here we go.
51:17 Make this image.
51:31 Come on.
51:33 Make it
51:36 make the image.
51:39 Um, what what
51:43 is your problem?
51:58 ding
52:01 ding ding ding ding ding ding
52:10 rut row.
52:12 Oh, here we go. All right. Once you get
52:15 the diffusion looking thing, it looks
52:17 like it does like the top quarter inch
52:19 of the image.
52:24 Bang. It just did it.
52:28 So, wait, what was 1903
52:34 official holiday in Ireland? Okay.
52:40 Copy image.
52:59 Okay. Fantastic. That looks really good.
53:04 Hey, here Kyle. Yeah. Yeah. Listen. So,
53:08 this is 1903.
53:11 Um,
53:16 St. Patrick's Day
53:22 is an official
53:26 holiday
53:28 in Ireland.
53:31 All right. Beautiful.
53:34 All right. Going back here. Look, we're
53:37 getting an education. We're look we're
53:39 learning how to make stuff with the
53:41 things.
53:43 We got prompts.
53:50 1931 first state sponsored Dublin
53:53 parade.
53:56 All right. 1931. Okay. So, let's do this
53:58 one. We'll go do this one. Uh, make this
54:03 image as a faded photograph.
54:09 from the era.
54:19 Do
54:41 Jensen Huang said, "Open Claw is the
54:44 operating system for personal AI." Yeah,
54:46 Todd. The uh we should probably go watch
54:48 some of that um some of that stuff.
54:52 Copy image. So, this is what 1931. Yeah.
54:57 What is this? The first state sponsored
54:59 parade. Okay.
55:01 So on the hat we go right.
55:06 Wait, that's the same image.
55:10 You dumb dumb? Oh my god. Chat GPT.
55:14 Really?
55:15 Hey
55:17 dum dum.
55:21 You took the image from 1903
55:26 and just made it black and white.
55:31 for the 1931 image. Um, that is
55:38 ridiculous
55:40 and you should
55:44 be ashamed.
55:48 Um,
55:50 learn
55:52 what
55:54 image uh would be iconic
55:59 and make me a 1931
56:03 photo or foe.
56:09 Color is extra. I asked it to do a
56:11 photograph of the era, so it made it
56:13 black and white. That was appropriate.
56:15 But oh my god.
56:19 Wait. When did
56:24 I'm just laughing at myself. Uh, when
56:28 did color
56:33 photography
56:34 make the scene? Was it the 40s?
56:45 Oh, the late 19th century practical,
56:48 widely accessible color photography
56:50 arrived with the autochrome Lumiere
56:52 process in 1907,
56:55 which used dyed potato starch to create
56:57 color images.
56:59 Huh,
57:01 that's cool.
57:08 1935 the launch of Kodakchrome
57:12 revolutionized the field.
57:23 1960s color photography became widely
57:26 common with 1967
57:29 considered a tipping point for popular
57:31 usage. I assume that was about Vietnam.
57:37 Um,
57:39 let's see. Why
57:41 was 1967 the tipping point
57:48 for color photography?
57:54 The affordability, cultural push,
57:57 snapshots,
57:58 improved technology and cost,
58:01 cultural shift,
58:04 artistic adoption. Joel Marowitz,
58:08 and then transition from slides to
58:10 prints. Ah, interesting, huh? Wild. All
58:14 right, where were we?
58:17 Thought for a couple of seconds. Look
58:18 how [ __ ] up chat GBT is.
58:21 It's unbelievable.
58:23 Um, make the image.
58:37 What happened to being generative? I
58:39 don't know what's going on with chat GBT
58:41 tonight. It's being weird.
58:45 Image and irregulars. Okay,
58:48 while that's doing that, let's go look
58:50 at irregulars.
58:57 Thinks Paul Simon created Kodakchrome.
58:59 Time to listen to Zombie by the
59:01 Cranberries again.
59:08 Little Champy Champy made the scene.
59:11 It's been weird for at least three
59:13 weeks. Catchy BT's been weird for at
59:14 least three weeks. at least three weeks.
59:16 Greg chat GBT has been weird for like
59:18 three months. Like it's just not
59:22 I don't know what they did.
59:26 I you know you know when they lost that
59:28 whole rash of senior engineers a while
59:31 back. Now is about the time we would see
59:34 the the impact of that. So I have a
59:37 feeling that that OpenAI was kicking ass
59:40 because they had the right 17 people
59:42 there and now they're down to three of
59:44 them.
59:46 All right, there we go.
59:49 This is the official. Copy the image.
59:52 Now we're talking, you big dumb dums.
59:55 Okay,
59:59 I'll do this.
1:00:06 So this is 1931
1:00:12 parade
1:00:16 in Dublin.
1:00:40 Let's do this. 1962 and we'll do this in
1:00:43 Kodakchrome.
1:00:45 Let's
1:00:48 make this image tall aspect ratio
1:00:56 photograph.
1:00:59 Good lord.
1:01:01 Photograph
1:01:03 on Kodak Chrome.
1:01:22 Maybe OpenAI is bad because they're
1:01:23 working on releasing their own open
1:01:24 claw. They will they will release their
1:01:26 own open claw. Um, I don't think that's
1:01:29 why they're bad. Something Something's
1:01:30 up in there. Something's up that they're
1:01:32 they're not I don't know what's going on
1:01:34 at OpenAI. I It It is possible that they
1:01:38 are their servers can't handle the load
1:01:41 and so they're throttling back inference
1:01:43 time or something like that, inference
1:01:45 cycles. I don't know. Um, Chat GPT just
1:01:49 decided not to make another image. Um um
1:01:52 hey chat GPT
1:01:57 grp hey chat gtp gpt
1:02:02 I'd like to formally invite you
1:02:08 to do your [ __ ] job.
1:02:14 Make the image
1:02:18 I asked for.
1:02:22 You are sucking.
1:02:27 Military is using it and we're getting
1:02:29 throttled. Yeah, something like that.
1:02:32 Although the military is using Claude
1:02:33 right now. They're they're trying to
1:02:35 unwind it because of politics, I
1:02:38 suppose.
1:02:41 But
1:02:47 it looked like they were investigating
1:02:48 whether or not um the targeting of that
1:02:52 school was a a Claude error that
1:02:55 basically they just let Claude do the
1:02:58 targeting for the Iran war.
1:03:01 And uh yeah,
1:03:03 I don't know if it's true or not. All
1:03:05 right. Here's the 1962
1:03:08 river being green. Beautiful.
1:03:13 So, we're going to go duplicate. Then
1:03:16 we're going to go.
1:03:20 We're going to do 19. What was it? 196.
1:03:23 What?
1:03:29 62.
1:03:31 1962.
1:03:37 Chicago
1:03:39 River
1:03:41 dyed green.
1:03:45 Is that how you spell dyed? No, it's
1:03:47 not.
1:03:50 I'm looking at it like, well, it is how
1:03:51 you spell dyed. It's not this particular
1:03:53 one. Dyed green. Um,
1:03:57 this is when
1:04:00 the
1:04:02 idiocy
1:04:04 began,
1:04:07 right? This is when it Oh, no. What's
1:04:09 going on? Hang on.
1:04:12 All right. This is when the idiocy
1:04:14 began.
1:04:21 Is it true people are leaving chat GPT
1:04:23 in in droves? Well, it's true that um I
1:04:27 don't know if in droves is is it chat
1:04:30 GPT has 900 million weekly active users
1:04:33 which is a lot almost a billion people.
1:04:37 Um very very few of them pay for it.
1:04:42 So a lot of people use it not a lot of
1:04:44 people pay for it. I think
1:04:47 a decent chunk of the people that pay
1:04:49 for it started paying for Claude because
1:04:51 chat GBT's been sucking so bad. Um and
1:04:54 and then Claude also with Opus 4.6 that
1:04:58 got so much press because of engineers
1:05:01 being able to vibe code well.
1:05:04 Um
1:05:06 along with some of the stuff that
1:05:07 they're they're launching um like Claude
1:05:11 Co-work was a really big uh boon for
1:05:14 them and like their revenues in the past
1:05:17 two months have just been crazy.
1:05:20 Um, okay. There's 1962.
1:05:24 And then let's do
1:05:27 we've got to finish the little the
1:05:29 little presentation.
1:05:35 Let's do um
1:05:40 let's go for
1:05:44 let's do
1:05:49 1996
1:05:51 New York City Times Square
1:06:00 magazine
1:06:04 photo of
1:06:07 drunk
1:06:09 idiots
1:06:11 in New York City. Um,
1:06:15 horizontal
1:06:17 photo.
1:06:20 Please, please, maybe please will work.
1:06:31 Thought for a few seconds and did
1:06:33 nothing. Make the [ __ ] image.
1:06:39 I think we're in a loop here where every
1:06:40 time I'm gonna ask for an image, it's
1:06:41 going to fail and then I'll cuss at it
1:06:43 and then I'll make it. Then I think this
1:06:45 is just how it's going to be tonight.
1:06:49 I only want this image and one more.
1:06:58 Could you just have chat GPT generated
1:07:00 all as a presentation? probably could,
1:07:04 but uh I haven't found chat GPT to be
1:07:07 reliable at that at all.
1:07:10 Um but maybe if I did it as as the agent
1:07:13 rather than just doing it in chat here.
1:07:27 I think that's going to be our last
1:07:28 image.
1:07:35 That's really good. That is really,
1:07:38 really good.
1:07:40 I like it.
1:07:49 Happy
1:07:53 Saint Patrick's
1:07:57 Day.
1:08:06 Happy Saint Patrick's Day,
1:08:10 you
1:08:12 big
1:08:15 dumb
1:08:17 ninny.
1:08:33 Okay, this is really good. This is
1:08:37 what we're learning here in the AI
1:08:38 learning lab tonight
1:08:41 is um
1:08:45 you too
1:08:48 can be an idiot
1:08:51 just like me.
1:09:00 All right, there we go. Beautiful.
1:09:03 All right, you want to see our little
1:09:04 presentation?
1:09:13 Get rid of that.
1:09:18 All right, kids.
1:09:21 Shouldn't he be holding a potato in the
1:09:23 other hand?
1:09:28 Y'all are funny. All right.
1:09:32 1631, an old dude blesses a peasant.
1:09:39 1762, first parade in New York City.
1:09:43 You go, Irish immigrants.
1:09:46 1847, immigrants in St. Patrick's Day.
1:09:49 It's a sad state of affairs.
1:09:52 1903, St. Patrick's Day is an official
1:09:54 holiday in Ireland. 1931,
1:09:57 the the official first parade in Dublin.
1:10:00 Look at that.
1:10:02 Gorgeous.
1:10:04 1962, Chicago River died green. This is
1:10:07 when the idiocy begins.
1:10:11 and today.
1:10:19 All right.
1:10:21 If you have not yet checked out the AI
1:10:25 salon, please go to
1:10:26 community.thesalon.ai
1:10:29 and go to the irregulars channel,
1:10:33 which is
1:10:35 the channel for this year, this year
1:10:37 live.
1:10:39 and uh either in
1:10:42 um community feed at the top or the
1:10:46 irregulars channel. Go make some stupid
1:10:49 St. Patty's Day stuff. Or if you don't
1:10:51 celebrate that, don't
1:10:58 if you feel about St. Patrick's Day like
1:11:00 I do. You know, that was fun. Those are
1:11:03 some amazing images. That one's perfect.
1:11:07 Look at it. Got the bud. It got
1:11:08 Budweiser right. It got the logo right.
1:11:10 And the Coca-Cola looks correct. Those
1:11:12 logos look right from 1995
1:11:16 or six. It's pretty cool.
1:11:20 Pretty cool.
1:11:22 All right.
1:11:29 Oh my god. Oh my god. It's so good. It's
1:11:33 so good. All right. Um,
1:11:38 I'm gonna get out of here. I don't know.
1:11:39 I don't have much to say tonight. Um, so
1:11:43 let me check on out. I will be back
1:11:46 tomorrow night. Do we have anything
1:11:47 tomorrow? No, tomorrow's the 16th. Um,
1:11:51 just be prepared. Next week is probably
1:11:53 going to be a fair amount of chaos. I
1:11:55 may not be here. Today is the 17th.
1:11:57 Wait. Oh, today's the 16th. Yeah,
1:12:00 tomorrow's Tomorrow's St. Patty's Day.
1:12:03 Yeah.
1:12:04 Yeah, that's good. I'm just getting a
1:12:07 jump on things. Just getting a jump on
1:12:09 things. So, so tomorrow is St. Patty's
1:12:12 Day and uh we've already we've already
1:12:14 already done our obligatory um
1:12:17 historical keynote presentation
1:12:23 actually. Yeah, that would be
1:12:25 interesting to have
1:12:30 Gen Spark or one of those things make us
1:12:32 a slideshow.
1:12:34 I bet it would be better better than
1:12:36 this.
1:12:40 You can do it as an agent for multiples.
1:12:43 Groovy. All right, cool. I will see you
1:12:45 guys here tomorrow night. Have a
1:12:47 fantastic evening. And uh I'm going to
1:12:50 go drink some more water. I'm a I'm a
1:12:52 little dry. The past week or so,
1:12:53 something's up with my throat. It's very
1:12:55 scratchy. I don't know what's going on.
1:12:57 Might just be spring pollen. I don't
1:13:00 know.
1:13:02 But it's weird. All right, peace out
1:13:04 y'all. Have a good night. All right,
1:13:06 bye.