AI Learning Lab

4/2/2025 - From Kids' Drawings to AI Art: Transforming Simple Sketches

aB4zTh_hgWQ
Live Stream2025-04-031:42:3970 views

Description

This lively discussion explores the rapid advancements and persistent quirks of AI image and music generation tools. Lyle Shannon showcases experiments with various platforms, highlighting the impressive capabilities of ChatGPT, Suno, and Luma Labs in transforming simple drawings and prompts into sophisticated visuals, music, and even animations. The conversation also touches on the challenges of navigating copyright restrictions and the surprising limitations of these powerful tools, emphasizing the importance of playful experimentation over outcome-driven approaches. Kyle shares anecdotes about generating magazine covers, video game concepts, and even a Star Wars-inspired "Bar Wars" poster, revealing the potential for creative expression and humor within the evolving landscape of AI. Kyle also introduces Lindy, an AI-powered task automation tool, and discusses its potential applications in sales and research. They further delve into the ethical implications of AI in the animation industry and express excitement about future possibilities, including the development of "emotional architecture" and world-building tools. The conversation concludes with a reflection on the current "janky" state of AI technology and a prediction that these early challenges will be viewed with nostalgia as the field continues to advance. The speaker encourages viewers to join the AI Salon community for further exploration and collaboration. Learn more about AI on TikTok: https://tiktok.com/@aiLearningLab. #AI #ArtificialIntelligence #ImageGeneration #MusicGeneration #AICreativity #AIArt #AITools #AISalon Chapters: 00:00:00 Intro Music and Song 00:03:03 Comments Section Issues 00:04:00 Podcast Recording and X Posts 00:04:51 TikTok's Uncertain Future 00:08:25 Recording Audio and Song Creation - "Walty Groove" 00:10:11 Writing Lyrics About a Singing Dog 00:11:45 Managing Expectations and Technical Difficulties 00:13:14 Refining the Song Description and Genre 00:15:58 Instrumental vs. Cover Song Discussion 00:16:41 Exploring Lyric Platforms and Refusion 00:17:03 Greeting Viewers and Technical Issues 00:17:25 Writing Lyrics with ChatGPT - "Bruises That Shine" 00:19:34 Generating More Lyrics and Using Refusion 00:22:00 Refusion Website Issues and Tab Management 00:24:00 Listening to "Bruises That Shine" - First Version 00:26:28 Refining the Song and Generating More Versions 00:29:32 Covering a Generated Song 00:34:14 Vibe Coding and Experimentation 00:35:33 Troubleshooting Lyrics and Cover Song Creation 00:38:53 Showing Image Generation Experiments 00:41:31 Showcase of AI-Generated Images: Book Covers, Posters, Restaurant 00:46:51 Exploring Twitter and AI Developments 00:48:24 Get Lindy and Agent Swarms 00:52:32 Exploring Lindy.ai Pricing 00:57:16 OpenAI Academy and Image Generation Updates 00:59:20 Exploring Image Generation with Kids' Drawings 01:01:00 Generating Images Based on a Child's Drawing 01:02:24 More Examples of AI-Generated Images from Drawings 01:04:21 Generating a Story and Video from a Child's Drawing 01:06:07 Discussion on Stock Photography and AI 01:06:47 Generating a Video from a Child's Drawing in Luma Labs 01:08:31 Showcasing More AI-Generated Images from Irregulars 01:13:04 Discussion on the Animation Industry and AI 01:15:01 Generating a Cartoon Video in Sora 01:19:04 Discussion on the Animation Union and AI 01:20:02 Reimagining a Photo as a Rocket Ship in ChatGPT 01:22:01 Creating an Interior Shot of the Rocket 01:24:01 Using Luma Labs for Video Generation 01:26:13 Adding a Driver to the Luma Labs Video 01:28:24 More AI-Generated Images and Discussion 01:30:30 Replicating a Storyboard Style in ChatGPT 01:33:21 Discussion on TikTok's Future and Alternatives 01:35:22 Generating a Five-Step Plan to Save TikTok 01:36:06 Finalizing the Storyboard Image 01:37:39 Reflecting on the Week's Theme and Janky Technology 01:39:00 Emotional Architecture and World Building 01:41:11 Outro and Future Plans

Chapters

0:00Intro Music and Song3:03Comments Section Issues4:00Podcast Recording and X Posts4:51TikTok's Uncertain Future8:25Recording Audio and Song Creation - "Walty Groove"10:11Writing Lyrics About a Singing Dog11:45Managing Expectations and Technical Difficulties13:14Refining the Song Description and Genre15:58Instrumental vs. Cover Song Discussion16:41Exploring Lyric Platforms and Refusion17:03Greeting Viewers and Technical Issues17:25Writing Lyrics with ChatGPT - "Bruises That Shine"19:34Generating More Lyrics and Using Refusion22:00Refusion Website Issues and Tab Management24:00Listening to "Bruises That Shine" - First Version26:28Refining the Song and Generating More Versions29:32Covering a Generated Song34:14Vibe Coding and Experimentation35:33Troubleshooting Lyrics and Cover Song Creation38:53Showing Image Generation Experiments41:31Showcase of AI-Generated Images: Book Covers, Posters, Restaurant46:51Exploring Twitter and AI Developments48:24Get Lindy and Agent Swarms52:32Exploring Lindy.ai Pricing57:16OpenAI Academy and Image Generation Updates59:20Exploring Image Generation with Kids' Drawings1:01:00Generating Images Based on a Child's Drawing1:02:24More Examples of AI-Generated Images from Drawings1:04:21Generating a Story and Video from a Child's Drawing1:06:07Discussion on Stock Photography and AI1:06:47Generating a Video from a Child's Drawing in Luma Labs1:08:31Showcasing More AI-Generated Images from Irregulars1:13:04Discussion on the Animation Industry and AI1:15:01Generating a Cartoon Video in Sora1:19:04Discussion on the Animation Union and AI1:20:02Reimagining a Photo as a Rocket Ship in ChatGPT1:22:01Creating an Interior Shot of the Rocket1:24:01Using Luma Labs for Video Generation1:26:13Adding a Driver to the Luma Labs Video1:28:24More AI-Generated Images and Discussion1:30:30Replicating a Storyboard Style in ChatGPT1:33:21Discussion on TikTok's Future and Alternatives1:35:22Generating a Five-Step Plan to Save TikTok1:36:06Finalizing the Storyboard Image1:37:39Reflecting on the Week's Theme and Janky Technology1:39:00Emotional Architecture and World Building1:41:11Outro and Future Plans

Transcript

0:02 Get that look in
0:04 [Music]
0:05 mine. Every time I see your
0:09 mouth, I hear that
0:12 [Music]
0:13 smile in the early misty morning that I
0:18 heard the engine
0:21 turning the old for
0:25 [Music]
0:31 You leaving
0:33 me again
0:37 today. You convince
0:40 me again
0:44 today. Leaving this
0:46 home looking for someone else's golden
0:51 ring.
0:55 should
0:57 have so long
1:00 [Music]
1:06 [Applause]
1:07 [Music]
1:18 Suzana. Don't you cry for me.
1:23 [Music]
1:29 sharing jeans and cigarettes and keeping
1:32 them out on the
1:34 road. Chasing down a lifestyle out on
1:38 Highway
1:41 24. New York State was a rolling breeze
1:44 in the sunshine with a blue sky falling.
1:49 Chill of all September
1:56 creep. You were leaving
1:59 me again
2:03 today. You convince
2:06 me again
2:09 today. Leaving this home looking for
2:14 someone else's golden ring.
2:18 [Music]
2:22 So
2:24 [Music]
2:28 long now. Don't you
2:34 cry. So
2:36 [Music]
2:41 long. Don't you cry for me.
2:47 [Music]
2:55 [Music]
3:00 boom bang bang bang bang
3:03 boom. All right, let's see. Did
3:06 something similar yesterday. What is
3:09 going on over here? Something is messed
3:11 up in my comment section. There we go. I
3:15 mean, it got all
3:17 collapsed. I needed to write some very
3:19 personal notes. Chat GPT came up with
3:22 exactly the right words when I couldn't.
3:23 I did that today, too. I had to come up
3:25 with the words with something that was
3:27 kind of heartfelt, but I didn't want it
3:31 to be
3:32 schmaltzy. And so, I had chat GPT write
3:35 me something and it was schmaltzy and
3:36 just wrong. And then I said, "Fix it
3:38 like this." And it fixed it. And then I
3:40 still had to write it.
3:43 Schmaltzy word of the
3:45 day. You You can't Oh, love your ex,
3:50 Kyle. Yeah, I'll show you some of the I
3:52 went on a bit of a tear. So,
3:55 um So, I had So, Ann Murphy and I are um
4:00 our podcast was today, but we recorded
4:03 it Sunday, but I streamed our recording
4:06 live, so I could like comment, right?
4:08 you know, make comments like it was it
4:11 was a live thing. So, while the podcast
4:14 was rolling and I was commenting in
4:16 between that, I'm I'm in chat GBT making
4:20 funny things and posting them on X. So,
4:24 I'll show I'll show you the stupid [ __ ]
4:26 I
4:31 did. Oh, man.
4:34 [Music]
4:42 Um
4:45 [Music]
4:52 um I don't know what's going on with Tik
4:53 Tok. Oh, let me flip my cameras here. I
4:56 don't know what's going on with Tik Tok.
4:58 Um, Saturday it's supposed to go away,
5:03 but I heard that there was a lastm
5:05 minute offer by Amazon today for it, but
5:09 it was a Tik Tok video, so I don't know
5:10 if it was just [ __ ] or not. Um, and
5:13 then with uh Happy Liberation Day, by
5:16 the way, uh, we've been liberated thanks
5:18 to thanks to all the tariffs, I guess.
5:23 All I know is the the uh after hours
5:26 trading is not very happy about it. Um
5:29 but uh I have a feeling because of that
5:33 Trump's going to be
5:34 distracted and uh so Tik Tok might go
5:37 away Saturday. So just so you know,
5:41 nothing I can do about it. That's I mean
5:43 part of the reason we're doing the whole
5:44 dual broadcasting thing is just because
5:46 I don't know what the [ __ ] going on
5:48 with Tik Tok. This is not ideal.
5:51 Um, Mr. Shannon, we we try to avoid
5:54 talking about politics in the AI
5:55 learning lab. Thank you. Yeah.
5:58 Yes. Exactly. If you want the full
6:00 stream, you can get on to YouTube. Don't
6:02 forget to like and subscribe. Yeah.
6:04 Exactly. Exactly.
6:08 [Music]
6:17 I mean, it apparently it's going to be
6:20 saved, but you know, there's been no
6:23 announcement of that, so we should just
6:25 assume it's going to go
6:27 away. Um, but yeah, subscribe to YouTube
6:30 if you're not because that's where I'll
6:32 be. Liked and subscribed. Thank you,
6:33 Jeff Flanigan. What? What? Tik Tok is
6:36 supposed to go away again. Yep. Janet
6:37 Jackson, sorry. Sorry to break it to
6:40 you. the last reprieve was was only a
6:43 90-day reprieve
6:45 uh or a 70-day reprieve, something like
6:48 that. And Saturday is the the end of it.
6:54 So, I do not know. I do not know. I'm
6:58 but a lowly tick
7:00 tocker. I am a I am a I am a semi
7:04 influencer. The only thing that makes me
7:07 an influencer is the fact that I call
7:08 myself an influencer.
7:12 [Music]
7:36 Actually, that's kind of cool. Let's go
7:37 try something. Let me let me change up
7:40 my tab
7:42 tabios. Am I going to be able to
7:43 remember that?
7:45 [Music]
7:52 Um there's just something cool about
7:54 that. Let's Let's uh And there I'm going
7:58 to stop this and start this and do
8:01 this. Going to click that. You're going
8:03 to watch this. You're going to love it.
8:04 It's going to be so good. It's going to
8:06 be so good. Oh, now I got to figure out
8:08 where this button is
8:11 again to record. It was in
8:16 simple.
8:18 Uh, where was it? Song styles. More
8:23 options. Add
8:25 audio. Record audio.
8:28 [Music]
8:37 Okay.
8:40 [Music]
8:48 [Music]
9:09 Nice. We get the doggy in
9:10 [Laughter]
9:13 there. Save without
9:17 trimming. All right. I don't know what
9:19 that song is, but it's
9:21 something. Nice groove. Yeah, that's a
9:23 cool groove, wasn't it? There's just
9:25 something about
9:26 the 1 2 3 1 2 3 Oh, it's a it's a
9:30 [Music]
9:36 waltz. 1 2 3 1 2 3 1 2 3 1 2 3 That's
9:42 cool. Um,
9:45 okay. Confirm. Oh, we're going to title
9:48 this thing. So, we're going to go um
9:53 uh
9:55 Walty
9:58 Walty Groove. Walt
10:01 Groove Walt Walt doesn't have a T in it.
10:04 Has a T in it, right? Walty Groove.
10:07 Confirm and
10:09 save. All right. So, we got the Waltsy
10:12 Groove. What are we going to make? Let's
10:14 say
10:20 um a
10:26 groovy
10:28 waltz. It's a double time waltz, right?
10:32 Double time walts, I think.
10:37 Waltz
10:40 about a
10:42 singing dog
10:44 named
10:46 Champy. But let's see. A groovy double
10:50 time folk rock
10:55 waltz. All right, here we go. Bang. Uh,
10:59 by the way, if you're if you're uh new
11:02 here, you don't know what the hell I'm
11:03 doing. I just I just had a brain fart.
11:06 It's Welcome. They sometimes call the
11:09 channel
11:10 ChatD. My name is Kyle Shannon. I'm the
11:13 add part of this
11:15 [Laughter]
11:21 operation. And then wait, why did it
11:23 only make 9 seconds songs?
11:30 Folks,
11:32 [Music]
11:35 oh. Oh, named Champy. I wrote lyrics. I
11:38 didn't write it in the musical style.
11:45 Just hang on, people. Just calm down.
11:48 Calm down. I understand that that you're
11:52 all
11:53 um your expectations are that this going
11:56 to be some sort of educational
11:59 experience. Check that at the [ __ ]
12:01 door. All
12:04 right. At least I'm at least I'm bitter
12:07 for the third night in a row.
12:09 At least I have my tab shared,
12:14 right? Wait, did I create something?
12:17 Yes, it's creating.
12:19 [Music]
12:21 Okay, let's create more than two because
12:25 you
12:26 can, right?
12:33 [Music]
12:39 That didn't capture the green one. Oh,
12:41 wait. Did I do a cover or did I do a I
12:45 did an extend. So, it should do an
12:47 extend
12:49 [Music]
13:14 Okay, we we got to get rid of got to
13:18 take the word waltz out of the song
13:22 [Music]
13:25 description. No, no, it's all wrong.
13:28 It's all wrong, people. All right. a
13:30 groovy
13:34 uh we'll just say folk rock
13:40 uh high
13:47 harmony
13:48 song about a singing dog named Champy.
13:53 Create create.
13:57 All
13:59 right. Bitter. You betcha. The San
14:03 Antonio stroll. Yeah, that was
14:05 definitely that was definitely not what
14:08 I had in mind.
14:11 [Music]
14:28 He's got the
14:29 [Music]
15:02 That's pretty cool.
15:05 [Music]
15:21 Do it
15:22 down. I do it down. Do it
15:30 down. I do it down. Do it down. I do it
15:34 down.
15:35 I do it down.
15:38 [Music]
15:39 I do it down. Do it down. Do it
15:43 [Music]
15:44 [Laughter]
15:46 [Music]
15:50 [Applause]
15:59 down. Did I hit instrumental? I didn't.
16:02 All right, let's do a cover instead of
16:05 a whatever. We'll do
16:09 um that's exclude style. So, we'll
16:13 do warm
16:17 harmonies. We'll
16:23 do Afrobe. Let's see what that does.
16:28 YouTube question. Yes.
16:30 What platform is good for someone to
16:33 create
16:34 lyrics but doesn't have Wait, what
16:37 platform is good for someone who creates
16:39 lyrics but doesn't have instrumentals?
16:41 I'm on it. This is This is it. Or
16:43 Refusion. RFU Sio N. So,
16:49 um, yeah, you can just come in here.
16:52 Here, we can do it here. I'll pop over
16:55 to chat GPT. Oh, you're not going to be
16:57 able to see it. Hang on a sec.
17:03 Hey Dan Driver, what's happening? We're
17:05 shaking. What's going down there, good
17:07 buddy? What's happening, good buddy?
17:10 Yeah, fantastic. Bob, tell him what he's
17:12 won. Uh he he he hasn't won anything.
17:16 It's This is not a game
17:22 show. Why do I work
17:26 here? All right. So, let's say we're
17:28 over in chat
17:30 jeep, you know, like you
17:32 do. Wait, why is the screen all off the
17:37 screen? Oh, cuz
17:38 my cuz my window ain't
17:42 right. You got to get your window right,
17:45 boy. Wait, it's still not right.
17:49 What the hell is going
17:52 on? Oh, no. Do I have
17:56 a I have a chrome problem, don't
17:58 [Laughter]
18:03 I? All right. Well, it doesn't matter.
18:05 Uh, let's see. Uh,
18:09 write me some
18:13 lyrics that is real
18:19 good, not
18:21 crappy and all.
18:24 But real
18:27 good. Like
18:29 I got
18:33 away with
18:36 words, you know. All
18:43 right. Too many tabs. Shut
18:46 up. Okay. I don't chase the gold. I mine
18:50 it slow. putting verses out the dark
18:53 where the wild things go. Every line a
18:56 lantern in a hallway of doubt. I speak
18:59 in thunder cuz I'm tired of shouting
19:02 out. Yeah, this is
19:06 good. Wait, let's ask it
19:10 something.
19:12 Wait, what's it called?
19:18 We'll do this in refusion and in here.
19:21 Weigh with words or where the silence
19:23 breaks or bruises that shine. Okay,
19:26 we're going to go with bruises that
19:27 shine. So, we're going to go copy our
19:30 lyrics. Uh, it didn't give us very many.
19:34 Uh, make more of
19:38 it. By the way, I'll be teaching uh
19:41 prompt mastery classes uh this weekend.
19:53 Wow. Chat GPT is slower than
19:56 [ __ ] Look at that. You can actually
19:58 watch the words typing out. This is This
20:01 actually reminds me of like 300 baud
20:03 modems back in the early internet days.
20:06 You know why it's this slow, right?
20:07 because all you idiots are out there
20:10 making magazine covers, which by the
20:12 way, in a second, I'll show you all the
20:13 magazine covers I made. All right, so we
20:16 got all these lyrics like like I was
20:19 good with words and all.
20:23 Right. And then what what did we call
20:25 it? Bruises. Bruises that shine. So,
20:28 we're going to go back to uh
20:31 Suno. We're going to paste our lyrics
20:33 in. Um, if you put things like the
20:36 bridge and verse in square brackets, it
20:39 just tells it that that's what those
20:41 are. It won't say those out
20:44 loud. And
20:46 then let's
20:49 see. Let's get rid of the dog named
20:52 Champy. And let's see, do we have a
20:54 title? We have a title.
20:58 And the title is going to be called
21:02 Bruises That
21:04 Shine. It's a horrible name. It's a
21:07 horrible
21:12 title. A groovy rock, high harmony song,
21:16 warm harmonies, afro beat, uplifting
21:19 melody,
21:22 uh, reflective tone. So, you can just
21:25 click on buttons if you want. Oops. And
21:28 then you can just create and then
21:31 create. Oh, and it's
21:34 using. It's using my uh my groove that I
21:38 did, which that's probably kind of
21:40 unfortunate for it, but that's all
21:45 right. Let me I'll I'll get rid of this
21:47 and we'll do I'll do two or three
21:51 generations without
21:54 it. And then let's go over we'll go over
21:58 to
22:00 refusion.
22:02 RF
22:11 refusion. Wait, why is that
22:13 not loud refusion? You not so
22:18 working. Um, r i fus
22:23 io n must not
22:26 be.com.
22:28 No, must be aai. R
22:31 ffus n.ai. A
22:44 [Music]
22:58 I thought that's what I typed. Maybe I
23:00 just spelled it wrong. R if
23:03 FFU sio
23:09 N.com. It ain't going there. Did Do they
23:13 not have their domain set up right? Do I
23:15 have to have
23:21 www? It didn't go in there. All right.
23:24 FF
23:29 uh the browser's screwed.
23:31 Uh I think you're right. All right.
23:34 Well, let me let me change my tabs.
23:36 We'll just we'll just play with Sunno
23:38 then. But anyway, you can go to
23:39 refusion.com. If you're if unlike me,
23:43 you've only got six tabs open and you're
23:46 good at tab management and you restart
23:48 your browser
23:51 occasionally, you can use other app. You
23:54 can use other sites. I can't. Oh man.
24:01 Okay. Bruises that shine. So, here's the
24:04 here's the lyrics we wrote in chat GPT
24:06 where I asked it for good words.
24:08 [Music]
24:11 [Applause]
24:16 [Music]
24:38 Unbelievable. Why no words? Why no
24:40 words? What? What's happening?
24:43 [Music]
24:51 I don't chase the gold.
24:53 I
24:55 slowing out the wild
25:00 things. Every line a lantern in a
25:03 hallway of
25:04 doubt. I speak in thunder cuz I'm tired
25:08 of shouting
25:11 out. Truth ain't clean. It's got dirt on
25:15 it. It's whiskey.
25:18 [Music]
25:27 I
25:28 don't mind it slow. Pulling verses out
25:31 the dark where the wild things go. Every
25:34 line a lantern in a hallway of doubt. I
25:38 speak in thunder cuz I'm tired of
25:40 shouting out. Truth ain't clean. It's
25:43 got dirt on its knees. It's whiskey on
25:45 the breath and a midnight breeze. I
25:48 don't write for applause or a fleeting
25:50 cheer. I write cuz the silence got too
25:54 loud in here. I got away with words like
25:58 rivers got away with time. Every rhymes
26:02 a bruise that learned how to
26:05 shine. I don't need a map. I just follow
26:08 the ache. Turn heartbreak into hymns
26:11 only real ones make.
26:16 I've been burned by the light, kissed by
26:19 the fall. Learn to love the echo when no
26:22 one
26:24 [Music]
26:26 calls. I wear them like armor. There's
26:29 something actually really sweet about
26:31 this. It's like the lyrics are like I
26:34 write lyrics real good in a song and I
26:37 make them make a song song like a good
26:41 song.
26:46 Don't quite fit. Hope ain't soft. It's a
26:50 barbed wire thread tied to every
26:54 [Music]
27:01 That's funny.
27:35 Weird. Weird. A lot of no words
27:39 [Music]
28:00 tonight. That was kind of cool.
28:03 [Music]
28:22 I lost my lyrics. Let's go back to chat
28:24 GPT and grab them. Hang on a sec.
28:35 D. Those last ones were cool. I kind of
28:38 like the grooves on those, but I didn't
28:40 have lyrics in there. I don't know what
28:42 happened to my lyrics. They They went
28:44 away. All right, so we're going to put
28:45 more lyrics back
28:48 in. And then I'm just going to
28:56 go Wait. Okay. Groovy Folk Rock.
29:04 Oh, these were the ones. Wait, were the
29:07 Were any of the covers the cool ones?
29:09 Oh, they weren't. So, these were the
29:11 ones where I got rid of the the audio
29:15 input. That's That's why they sound
29:20 [Music]
29:22 better. That one's good. Yeah, this one
29:24 This one's really good.
29:27 [Music]
29:33 So, you know what we can probably do? We
29:35 can probably cover that one. Let's go
29:36 grab all the lyrics. Copy. Wait, am I
29:40 generating some? Yeah. Let me let me
29:42 generate a few more of
29:44 those. And then let's take this one and
29:47 we'll do create cover
29:52 song. And then we'll do an
29:55 extend. And we'll do it like right
30:02 there. Extend it 43 seconds
30:09 in. We got that. We'll add in our
30:12 lyrics and we'll say
30:24 create.
30:34 Create
30:36 create. All right.
30:39 So, so we'll that one will be there. All
30:42 right.
30:44 Um, that doesn't have lyrics. These
30:46 probably do.
30:56 I don't chase the gold, I slow. No,
31:01 that's cheesy.
31:03 [Music]
31:10 I don't chase the gold. I mine it slow.
31:13 Pulling verses out the dark where the
31:16 wild things go. Every line of lantern in
31:19 a hallway of doubt.
31:21 I speak
31:24 in truth ain't clean.
31:31 Love that.
31:33 [Music]
31:43 Cool.
31:44 [Music]
31:49 This is very Jason
31:52 Morz pulling out the dark where the wild
31:56 things go. Every line of lantern in a
31:58 hallway of
32:01 doubt truth ain't clean. It's got dirt
32:03 on it. It's whiskey on the breath and a
32:06 midnight breeze.
32:08 [Music]
32:09 That's very Jason Morz.
32:16 [Music]
32:22 I don't chase the gold. I mine it slow.
32:26 Pulling verses out the dark where the
32:27 wild things go. Every line of lantern in
32:30 a hallway of doubt. I speak in thunder
32:32 cuz I'm tired of shouting out. Truth
32:34 ain't clean. It's got dirt on his knees.
32:40 [Music]
32:48 It's I don't chase the
32:51 gold, I slow.
32:56 No,
33:02 a lot of these are cool. I like whatever
33:04 this prompt is. This prompt is
33:07 Groovy folk rock, high harmony song.
33:10 High warm harmonies, afro beat,
33:12 uplifting melody, reflective tone. It's
33:15 a good prompt.
33:22 [Music]
33:28 I
33:28 [Music]
33:31 don't mind it slow.
33:35 These are the
33:42 extensions. So, wait, how does that
33:46 work? Save prompt. Clear
33:49 all. We do a
33:54 cover. Put the lyrics back in.
33:58 Keep that. We'll do a cover of this
34:00 song.
34:02 Create. Create. I don't quite know what
34:04 I'm doing as you can
34:09 see. But I'll tell you
34:12 what, the spirit of how I'm doing this,
34:15 you know, they talk about vibe coding.
34:17 This is vibe music making. It's like
34:20 there
34:22 are an absolutely joyful way to play
34:27 with AI is to have zero expectations. Go
34:31 into it not trying to accomplish
34:33 something because when where AI gets
34:37 really frustrating is where you go into
34:39 it trying to do a specific
34:42 thing. But if you just go into it and
34:44 go, "Fuck it. Let's just have some fun."
34:46 You can you can stumble on ways to do
34:49 things you never would have gotten to
34:51 otherwise.
34:53 Um, all
34:55 [Music]
34:57 right. Okay. So, this is a cover
35:02 of
35:03 [Music]
35:09 this. All
35:11 right. So, that's what we're
35:17 [Music]
35:33 Where's my
35:40 words, [ __ ]
35:42 [Music]
35:48 Do I have the words in here? Oh, I
35:51 don't. Why does it keep eating my
35:54 lyrics,
35:57 people? Okay, we're doing a cover of
36:01 this with
36:03 lyrics. We don't have instrumental
36:09 checked. Generate.
36:12 Why it not doing it good?
36:19 [Music]
36:41 [Music]
36:43 in here. I got away with
36:47 [Music]
36:58 words. I don't chase the gold. I mine it
37:02 slow. Pulling verses out the dark where
37:04 the wild things go. Every line a lantern
37:08 in a hallway of doubt. I speak in
37:11 thunder cuz I'm tired of sounds like bad
37:14 musical
37:15 [Music]
37:29 crappy. I don't chase the gold, I
37:34 slow out the bad. I don't chase the
37:39 gold, I
37:41 slowing out the wild things. Every lime
37:46 turn in a hallway of doubt. I speak in
37:49 thunder cuz I'm tired of shining out.
37:52 Truth ain't clean. It's got dirt on it.
37:55 It's whiskey on the breath.
38:00 That was kind of groovy. I like his
38:01 voice. I don't chase the gold, I slow
38:06 out the wild things. Every line of
38:14 lantern truth ain't clean. It's got dirt
38:17 on it. It's whiskey on a breath in a
38:20 midnight. All right, whatever. There you
38:22 go. There's some song making for you. We
38:25 didn't get We didn't get a weird Mary,
38:27 but you know what are you going to do?
38:28 You win some, you lose
38:30 some. It understands the music is good,
38:32 but your words suck. I think it actually
38:38 [Music]
38:54 does. All right, let me let me show you
38:56 some stuff I played with today in the
38:58 old image gen. What was the thing about
39:00 image genen? Jeff Flanigan open AAI says
39:03 image creation just got better. Chat now
39:07 thinks a little longer to make a more
39:10 accurate and detailed images. Learn
39:12 more. It looks like another image
39:16 upgrade. I hadn't seen this one. Oh,
39:20 well this is this is the one this is the
39:22 one that we've been playing with. So,
39:24 this is the one that came out last
39:25 Friday or
39:27 Thursday. Yeah, it's this is a big deal.
39:30 This image gen. Although Sam Alman said
39:33 today, version two of Image Gen he said
39:36 is going to blow your
39:37 mind, which I thought was
39:40 interesting. They can't keep up with the
39:42 demand as it is. And he's teasing the
39:44 next thing.
39:46 Buttthead. Oh, the other thing. You know
39:48 how I I [ __ ] and m whine and moan about
39:53 the webp format and the fact that chat
39:57 GBT outputs WEBP and then it can't
40:00 import it. You have to convert
40:02 everything. Like, are you [ __ ]
40:04 kidding me? Well, they they changed
40:06 that. It's now downloading PGs.
40:09 So,
40:11 um, yeah, the new, uh, the new image
40:14 generation stuff is quite remarkable.
40:16 It's slow and it's also, um, they've
40:20 nerfed it, which sucks. So, when it
40:23 first came out, it was pretty much doing
40:24 anything you wanted it to do. And now,
40:28 like today, I tried to get it to make a
40:29 Star Wars poster, and it would not do
40:32 it. So, I had to change it to bar wars.
40:36 But what was what was interesting was it
40:38 would not touch Star
40:40 Wars. Um, but I said I want it to be a
40:44 Netflix
40:45 show, a Comic Con 2025 poster and Bar
40:50 Wars and it made that poster. So
40:53 basically what it's doing is it will
40:55 generate a piece of artwork and then it
40:57 analyzes it and says, "Does this
40:59 infringe on any copyrights from a big
41:01 ass company that is likely to sue us
41:03 like Disney?" And if the answer's yes,
41:06 it won't give you the
41:08 image, [ __ ] And then I heard
41:11 that Sora was less nerfed than chat GPT,
41:14 but I didn't I couldn't get Sora to do
41:16 anything either.
41:18 So anyway, all right, let's go show you
41:20 what I've been what I was playing with
41:32 today. All right, so
41:35 [Laughter]
41:39 I took my cowboy Kyle Shannon dreams
41:42 picture and I I created a cowboy book
41:44 called My Name is
41:46 Spanx. A Mormon tale of a cowboy more
41:49 shy than
41:54 smart. I don't know why that cracked me
41:56 up so good, but it did.
42:03 Oh yeah. And then so then I took I
42:07 figured well you know that was this book
42:10 was was written in the in the 90s but
42:13 this was based on an Atari video game
42:16 called My Name is
42:18 Spanx. And then here's this is the this
42:22 is a game that's more rare than ET um
42:25 but with worse gameplay. So, there's the
42:27 Atari 2600
42:29 uh gameplay of My Name is Spanx use with
42:32 joysticks joystick
42:37 controllers. Actually, yeah, I could go
42:39 make this in Lovable. Make an Atari
42:43 2600 style game of a of a horse, a
42:47 cowboy on a horse that rescues a damsel
42:50 in
42:53 distress. So, that that I did. And then
42:56 so again, I was I was uh I was
42:58 babysitting a live stream basically. And
43:01 so while I was babysitting it, I was
43:03 just building all this stuff. So then I
43:04 thought, well,
43:06 maybe maybe our our our favorite cowboy
43:09 friend here got on the cover of uh GQ
43:15 West, Embracing Toxic Masculinity, How
43:19 to Raise Your Horse, Right. Six healthy
43:21 campfire
43:29 exercises. Oh my
43:32 god. It Oh, it just added it to the chat
43:37 hole. Yeah. Yeah. Yeah. I think Yeah,
43:40 they just they they they did add
43:41 something to the chat hole that now says
43:44 uh make image. I think that's just for
43:46 new people that that don't understand
43:49 how chat GBT
43:51 works. And then I did uh so there was my
43:55 live
43:56 stream and then I did cool AARP
43:59 magazine. Oh, you can't really see it on
44:02 on Tik Tok, but it says Kyle Shannon
44:06 secrets to senior beard mastery.
44:14 Oh yeah, man. That's
44:17 comedy. Then I did a vampire, a vampire
44:20 book, Just a Nibble, New York Times
44:23 bestseller by Sebastian
44:32 Moistgate. Oh, fantastic,
44:36 Bob. Then I did uh I opened a
44:39 restaurant with uh with robots. You you
44:44 can only use humanoid robots in the
44:46 kitchen other than the main chef, which
44:49 is giant
44:51 me. And
44:53 then where did I
44:58 [Music]
45:00 go? Okay. And then so that was that. And
45:02 then the final thing I did was this was
45:04 the the bar
45:06 [Laughter]
45:07 [Music]
45:11 wars. Uh, you can't really see it. Damn
45:14 it. It says Bar Wars: The
45:18 Sluggish, starring Kyle Shannon as Girth
45:23 Vapor. ComicCon
45:26 2025. Streaming soon on
45:28 [Laughter]
45:31 Netflix. So, you know, you could just do
45:34 dumb [ __ ] now and put it out
45:36 [Laughter]
45:40 there. Here. Here's my philosophy on
45:42 deep fakes. If you do enough of your own
45:45 stupid ass deep fakes, then you're going
45:47 to be less likely to have people come at
45:50 you. He's already good. He's already
45:53 embarrassed
45:57 himself. Oh man. It's also a little sad
46:01 they will trample all other all over
46:04 other IPs that aren't so grumpy. Looks
46:05 like low integrity. Yeah. I mean, low
46:09 integrity or just, you know, what did
46:12 they just raise? $40
46:14 billion architect. I
46:17 mean, when you raise $40 billion, those
46:20 don't come without strings, right? Like
46:23 it like Microsoft, you know, probably
46:26 has some say about how they operate.
46:28 Whoever just gave him the $40 billion
46:30 probably does, too. And so the investors
46:33 are probably like, you know, don't
46:35 infringe on Disney, you know, who
46:38 whoever the big ones are, New York
46:40 Times,
46:42 Disney, the ones that are going to sue
46:44 you. So that's just that's just that's
46:47 just corporate corporate [ __ ] man.
46:50 That's just the way it
46:52 goes. Um, let's see.
46:56 [Music]
46:59 I I I kind of wanted to just roll
47:01 through um Twitter a little
47:06 bit because there like a lot of stuff
47:09 dropped today. But what I couldn't I was
47:12 too busy doing other [ __ ] that I
47:13 couldn't actually go test a bunch of
47:16 things. It was really hard to figure out
47:18 what actually dropped, what were just
47:20 announcements, what were just papers.
47:23 What is
47:25 that? Spine navigation. A breakthrough
47:27 in surgical intelligence that leverages
47:29 AR and AI. Minimal minimally invasive
47:33 spine procedures. Cool. That's
47:36 great. Wind
47:39 surf.
47:44 Um, people seem to be liking Runway ML,
47:47 but I still don't have Oh, we can go
47:50 check and see if I have it. the
47:53 uh reference
47:56 stuff. They said it's rolling out
47:58 slowly. Nope, I don't have
48:05 [Music]
48:11 it. What's this? Bring your logo to
48:14 life. Is that Pika?
48:25 Okay, get Lindy was one of these. So,
48:27 Get
48:29 Lindy. Let's watch this video. Hang on.
48:32 Let me go change my tabs for
48:42 y'all. A link with a list of fun. Hi,
48:46 this is Flo from Lindy and today we're
48:48 introducing agent swarms which let you
48:50 automate entire lists of tasks using AI
48:52 and get them done all at once. No
48:54 matter, you know, if you've got an
48:56 accent that thick, why don't you use hey
48:59 Jen and just speak French and then just
49:03 translate it to English so we can
49:04 actually [ __ ] understand the size of
49:05 the list. So for example, you can use an
49:08 agent swarm to automate a research of a
49:11 list of conference attendees. So you
49:12 would give your agents form a link with
49:14 a list of conference attendees and it's
49:17 going to spin up one agent for each
49:19 attendee in the conference and all these
49:22 agents are going to work in parallel to
49:24 research all these
49:25 attendees and you guys following this.
49:28 So you're going to search up all your
49:30 attendees of a conference and this is
49:32 going to create an agent per attendee
49:34 and go do a bunch of research on each
49:36 person as a swarm. Crazy. Once it's
49:40 done, you're going to receive one nice
49:41 report about all these attendees, all
49:44 this research happening all at once
49:46 thanks to the agent in the swarm. So
49:48 today we're going to talk about three
49:49 use cases for agent swarms. So I'm going
49:52 to start with a swarm that prepares me
49:53 for my meetings. Every morning it wakes
49:55 up, it looks at my calendar for the day
49:57 and then for every meeting it kicks off
49:59 one agent that researches the people in
50:01 this meeting and in the end it sends me
50:03 a digest by email. This is the digest
50:05 that I received yesterday. You can see
50:07 all the meetings I had. the context for
50:09 these meetings and the LinkedIn links of
50:11 the people that I am meeting with. We're
50:13 also using agent swarms for competitive
50:15 analysis. So, I have this spreadsheet
50:16 with a list of companies. Good lord, I
50:19 can't I'm like I'm reading the [ __ ]
50:22 subtitles. I can't understand them. And
50:24 I have this agent's phone that wakes up
50:26 every month. Tell us how you really
50:28 feel. Looks at the list of competitors
50:31 and for each competitor gets employee
50:33 count, traffic estimates, the recent
50:35 news from this company. It sends me a
50:36 report about it. So I'm going to ask it
50:39 to begin right here is pulling the list
50:43 of competitors and kicking off one agent
50:45 for each competitors and it's performing
50:47 all of these results and once it's done
50:49 I can go back to the swarm here and see
50:51 the reports that it brought together
50:52 about these
50:54 companies. Finally we use agent swarms
50:57 for sales outreach. I have a Google
50:59 sheets with a list of unicorn funders
51:01 and I have set up this agent swarm to
51:03 reach out to them for real estate
51:05 services. I just P. It works with the
51:08 most basic of inputs.
51:11 Amazing. So, it's your own personal
51:14 stalker
51:15 app. Y'all are getting
51:17 it. Jury selection breakthrough. How
51:21 dare these people have accents? I just
51:23 can't understand him. Just give it to
51:24 Everyone can have accents. I'm just
51:26 saying we have technology now where, you
51:28 know, you could you could do this so it
51:30 would be easy for everyone to Google
51:32 Sheets. is going to get it and it's
51:33 going to kick off one agent in the swarm
51:36 for each of these people. And then for
51:38 each of them, it's going to go ahead and
51:40 perform research about them online and
51:43 use the information it found to draft a
51:45 personalized
51:48 email. So here in a prepare to be
51:51 assaulted with AI generated personalized
51:54 email. Few seconds I had 20 outreach
51:57 emails drafted for these people. It
51:58 could have been 200. I'll tell you
51:59 what's a a a big business to have is if
52:02 you can create a thing that generates
52:05 automated emails that don't sound like
52:07 automated emails because I I don't
52:10 whoever's doing the prompting of the
52:11 [ __ ] that's coming to my inbox and my
52:13 LinkedIn, it's so bad. Like the the bar
52:17 the bar is pretty low. Like it just has
52:19 to not sound like [ __ ] Like the
52:22 quantity doesn't matter because they all
52:23 happen simultaneously
52:25 anyway. So this is agents forms. super
52:28 excited to announce it today and it's
52:29 available immediately on lindy.ai. Ah,
52:33 lindy.ai available immediately. Let's go
52:35 over there and see
52:37 if we can figure this [ __ ]
52:39 out. I wonder I I would think if it's
52:42 doing simultaneous
52:46 agents, I would think you would run up a
52:48 bill really quickly with this. Let's go
52:50 see what the pricing is for it.
52:53 If it's like $10 a month for a,000
52:56 credits, I'm going to be like,
52:59 "No, you're going to blow through a
53:01 thousand credits on a test. Try for
53:04 free." E that See, that already looks
53:07 automate your CRM data. So, they're
53:10 positioning this as a sales tool. So,
53:13 they're going to charge a lot for it.
53:14 Get started for free. Let's sign up with
53:19 Gmail because why not?
53:22 Uh oh, I'm in the
53:40 wrong. Share your screen, Kyle. I am,
53:43 aren't I? Oh, no. Cuz you're right.
53:46 Fine.
53:47 Fine. Shut
53:49 up.
53:55 Uh what he said
54:00 um verify it's
54:09 you. Let's change this out and go
54:13 to that
54:16 persona and then
54:20 No, I don't want
54:44 that. That doesn't look right.
54:48 Wait, hang on. Let me I got to go deal
54:50 with passwords. Let me not share my
54:52 screen and share all my passwords
54:54 inadvertently
54:56 here. He's done it before.
55:04 [Music]
55:22 That's not
55:23 right. What is going on, man?
55:42 Okay, now I think we're
55:54 good.
55:58 Uh, let's see what the cost is for this
56:01 thing. This is for
56:04 Lindy. M it would never occur to me to
56:08 gather information like this and if I
56:09 did have absolutely no idea what to do
56:12 with
56:19 it. What's your job
56:24 title? What's your main use case? sales
56:27 recruiting, high value, high volume
56:29 customer support, email automation,
56:32 scheduling, workflow
56:36 automation. This is going to be pricey
56:38 as
56:43 [ __ ] How'd you hear about
56:46 us? Twitter
56:49 X, skip the building. Book a call with
56:52 our book a call to receive a custom
56:55 agent workflow. No
56:59 thanks. 50 bucks a
57:04 month. Yeah. Not interested. All right.
57:08 Might be cool. Could be
57:11 cool. Feels expensive. Not going
57:14 [Laughter]
57:17 there. Oh
57:21 man. All right. Back to uh back to X.
57:34 Gen run gen 4 a lot of Gen 4
57:39 videos lenticular
57:42 art openai academy if you didn't see it
57:46 openai launched uh training yesterday if
57:50 you go to academy.openai.com
58:11 openai.com. There was another there was
58:13 another agent thing that came out. I
58:15 thought I would stumble on it by now.
58:20 Figure
58:30 update. Yeah. I don't know. All
58:34 right. What do we want to do? What do we
58:37 want to do? Do we want to go? Let's see.
58:57 Um, yeah, we could do that.
59:03 So here they here's the the new
59:06 interface element where they they added
59:08 the create image button
59:11 here. Create an image of a magazine
59:15 cover. I was also trying to think of
59:17 like what are all the other things that
59:18 you could make.
59:20 Um there's all sorts of crazy [ __ ] you
59:23 can do. Like we can go let's go to
59:25 Google for a second.
59:27 Um Google
59:32 com. Brandon was telling me about this
59:34 earlier. So, we'll do kids
59:40 drawing. We'll do
59:48 images. Let's go find a really janky ass
59:54 drawing. Let's see.
1:00:00 Um, magic
1:00:04 marker. Oh my god. Magic marker.
1:00:09 Handdrawn. Everything's [ __ ]
1:00:11 marketing. Everything's search engine
1:00:13 optimized. Stop
1:00:16 it. I just want to find pictures. Shitty
1:00:20 pictures. You mean you don't just want
1:00:22 to draw something? No.
1:00:29 Let's see more results. Let me get rid
1:00:33 of magic
1:00:37 [Music]
1:00:39 marker. I'm trying to get something that
1:00:42 will be
1:00:45 impressive if it
1:00:47 works. Oh, here's a good one.
1:01:00 copy image. Okay. So, let's go to chat
1:01:08 GPT. And so, we threw in So, it's it's a
1:01:11 just a picture of a kid throwing a ball
1:01:12 to someone else. So, we'll say,
1:01:16 um, make me a photo
1:01:22 of two kids on the
1:01:27 beach playing with a ball at
1:01:35 sunset. So, there's our little picture.
1:01:37 I don't Can you see that on You can't
1:01:40 see it on TikTok.
1:01:43 [Music]
1:01:44 H I shared an example of this on a
1:01:47 regular sky. Okay, cool.
1:01:51 [Music]
1:01:57 Um, this is such a janky [ __ ] setup I
1:02:00 have here for the for the people on Tik
1:02:03 Tok. This is bad.
1:02:08 All right, let's hop over while that
1:02:10 thing while that thing draws. Let's go
1:02:11 over to
1:02:13 um Where did you put it? In
1:02:16 irregulars. All you have to say is make
1:02:18 this image come to life. Oh, that's
1:02:24 cool. Oh, there you go. Okay, that's
1:02:27 really
1:02:32 cool. Is that just a screenshot of those
1:02:34 two?
1:02:36 So, here's her son's drawing to the left
1:02:39 and then what ChatPD did with it. That's
1:02:42 really slick. God, he must have loved
1:02:44 that. You can animate that in Sora. Oh,
1:02:47 yeah. Right. Then you can take that over
1:02:49 to Sora and animate
1:02:50 it. That's pretty
1:02:53 slick. Pretty flipping
1:02:58 slick. Paint modder. Oh, some V2 videos.
1:03:01 Been playing with V2 recently.
1:03:14 A sticky gooey peanut butter
1:03:19 tractor. I think my I think my browser's
1:03:21 broke. Oh no, there it goes. That's
1:03:24 pretty
1:03:27 cool. Come on.
1:03:38 All right, look what I made for
1:03:40 his. How to perform matrix
1:03:48 multiplication. Is the math right in the
1:03:51 matrix multiplication paint? I assume
1:03:55 not. It'd be cool if it
1:03:59 was. All right, look what I
1:04:05 made. Google whisk.
1:04:12 Interesting. So, here's the original
1:04:14 artwork. So, there's the original
1:04:16 artwork that Brandon's daughter
1:04:18 did. Cool. And
1:04:21 then there's what it turned
1:04:25 into. Oh, and you turned it into a
1:04:27 video. Oh, cool. All right. I need
1:04:29 audio. Hang on. Let me switch my
1:04:42 tab. The rainbow windmill. In a peaceful
1:04:45 valley where the tulips dance in the sky
1:04:47 is always swirling with color. There
1:04:48 stood a windmill unlike any other. It
1:04:51 blades shimmered in every shade of the
1:04:53 rainbow. purple, blue,
1:04:57 green, yellow, orange, and even the
1:05:00 tiniest stripe of red and black at the
1:05:02 tips. Locals said the windmill could
1:05:04 catch not just the wind, but dreams.
1:05:06 Each blade was painted by a different
1:05:08 child from the village. Their
1:05:10 imagination soaked into every brush
1:05:12 stroke. The windmill didn't grind grain
1:05:15 or pump water like others. It powered a
1:05:17 magical light that glowed from its
1:05:19 center, especially bright during
1:05:21 festivals. But what made it truly
1:05:23 special was what happened at night. As
1:05:26 the wind turned the rainbow blades, it
1:05:27 made a soft hum one that carried
1:05:30 stories. Children lying in bed would
1:05:32 hear gentle tales floating through their
1:05:35 windows of brave brave. I teach a teach
1:05:38 a master class and tab management. I
1:05:41 could teach you what not to do. Very van
1:05:43 go. Yeah, this is very cool. Very cool.
1:05:45 Brandon, I love that. So that's chatgpt
1:05:49 cartisia which is the audio the voice
1:05:52 over suno for the music and then flex
1:05:55 clip to make the video. That's pretty
1:05:57 cool. Very very cool man. We are living
1:06:01 in radical times. I dig it. I wasn't
1:06:05 able
1:06:07 to Oh, that's funny. I grabbed a sketch
1:06:10 that had a watermark on it and it won't
1:06:13 it won't make a uh it won't make an
1:06:16 image of it. I was wondering if it would
1:06:18 do
1:06:19 that. I actually just want a shitty kids
1:06:22 drawing and I can't find one because
1:06:24 everything
1:06:27 is Here's
1:06:31 one. Everything's copy copyright stock
1:06:35 photography. I mean, holy [ __ ] Can you
1:06:38 imagine having a stock photography
1:06:39 business right now? Cuz who's going to
1:06:41 need stock photography? You just go make
1:06:43 your own stuff. Um, bring this to
1:06:47 life.
1:06:49 Um,
1:06:56 cinematically tabs. Oh, I'm not sharing.
1:06:58 No, I'm not. No. See, Kelly, you spoke
1:07:02 too
1:07:03 soon. I can't teach a class in tab
1:07:06 management.
1:07:08 I can only teach a class in what not to
1:07:10 [Laughter]
1:07:13 do. If you'd like, I can mock it up.
1:07:16 Yes. Yes,
1:07:21 please. Try having Chat GBT makes the
1:07:24 kids drawing first. Oh, yeah. That's a
1:07:26 good
1:07:27 idea. Now I have to go dig out my kids
1:07:29 drawings. That's
1:07:33 cool. Well, was cool. last night on the
1:07:36 um at the AI salon. Um Adnan
1:07:42 uh he he wrote this book and his
1:07:46 daughter's like seven. He wrote this
1:07:48 book with Brian Moahan and his daughter
1:07:51 said, "What's the book about?" And he's
1:07:52 like, "Well, you can read it, but it's
1:07:53 like some of the concepts will be over
1:07:55 your head." So she goes, "Well, tell me
1:07:57 about it." So they ended up writing a
1:08:00 version of the book. He he co-wrote it
1:08:03 with his seven-year-old daughter. So,
1:08:05 he's now got the adult version of the
1:08:06 book and then the kids version of the
1:08:09 book co-written with his daughter, which
1:08:11 that's a pretty cool That's a pretty
1:08:13 slick
1:08:15 little little thing.
1:08:20 [Music]
1:08:32 That's pretty
1:08:38 funny. So from
1:08:43 that, come
1:08:48 on. It's so frustrating. I can't change
1:08:51 the brightness for Tik Tok. Everything's
1:08:54 blown
1:08:55 out to that. That's pretty cool. I love
1:08:59 that it kept the hand shape, but it made
1:09:01 it like a felt like a felt art project
1:09:04 with real chicken
1:09:05 [Laughter]
1:09:10 legs.
1:09:13 Crazy. Um, trying to think what else we
1:09:16 could
1:09:20 do. This is kind of endless, which
1:09:24 is which is a bit intimidating.
1:09:31 [Music]
1:09:33 Well, let me jump over to the salon.
1:09:36 Um, if you're not a member of the salon,
1:09:39 the AI salon, go to the
1:09:41 salon.ai. So, go
1:09:43 there and
1:09:46 um, click on join our community and that
1:09:49 will take you over to this site, which
1:09:51 I'm on right now. And there's a there's
1:09:54 a seven a little sevenst step, you know,
1:09:57 onboarding process. You don't have to do
1:09:58 all seven steps, but if you do do all
1:10:01 seven steps, you can earn your first
1:10:03 badge. Um, and then this is a community
1:10:06 of AI optimists, and there's a little
1:10:09 community area, there's a news area,
1:10:11 there's a showand tell area where you
1:10:13 can do projects and collaborations and
1:10:15 learn what tools to use. Um, and then
1:10:17 there's hubs and clubs and and those are
1:10:20 all little sub communities within within
1:10:22 the community. Um, so so go check out
1:10:26 the salon if you haven't been there
1:10:27 before. Um, tomorrow night is the the
1:10:32 business guild. So that's Kelly Camp's
1:10:35 uh business guild. So if you're trying
1:10:38 to figure out how to use AI and b
1:10:40 business, do that. Saturday, Pate M um
1:10:44 is doing a demystifying another AI class
1:10:47 I think this fifth fifth or sixth I
1:10:49 think it's the fifth one uh that he's
1:10:52 been doing. So if you want to understand
1:10:54 uh how how all this magical seemingly
1:10:57 magical stuff works he can actually tell
1:10:59 you the math behind it but do it in a
1:11:01 really
1:11:03 um accessible way and he even builds uh
1:11:07 custom Google notebooks that you can run
1:11:09 code. He writes little bits of code to
1:11:11 to really bring the concepts to life for
1:11:13 how this stuff works. Pretty pretty
1:11:15 powerful.
1:11:17 Um I don't know what shout outs you're
1:11:19 talking
1:11:20 about. Let's see. In the business club
1:11:22 tomorrow night at 6:30
1:11:25 central, you were going to start
1:11:27 building a virtual staff. Oh, cool. Like
1:11:29 a like
1:11:30 a Oh, the kudos things. Oh, yeah. So,
1:11:33 there's another cool thing in the salon
1:11:35 that we just
1:11:37 added where if you let's say we'll go to
1:11:40 someone's post here, like so here's
1:11:45 Brandon. So, there's now this little
1:11:48 button that says if you click if you
1:11:51 roll over someone's name or click on it,
1:11:55 it pops up a little box and there's a
1:11:57 there's a button there that says
1:11:58 recognize. So, I can click on that to
1:12:01 recognize Brandon. And then I can
1:12:03 recognize him for one of the five values
1:12:06 of the salon. Curiosity, playfulness,
1:12:09 empathetic, generous, and brave. So,
1:12:11 we'll just say we'll go
1:12:13 playful, and generous. Oh, I can only do
1:12:16 one. And then I'll say give more
1:12:18 details. Uh, playful. Uh, you make
1:12:23 um being a producer look like fun.
1:12:30 even though I know it's not because
1:12:32 you're producing me and then so now he
1:12:35 just got that recognition and so you
1:12:37 earn points by participating within the
1:12:39 community and then you can recognize
1:12:41 people for the values of the community
1:12:43 so it's a nice little gamification
1:12:44 feature they added source camp our first
1:12:47 build will be for an assistant secretary
1:12:50 hope everyone will join us if you've not
1:12:52 played with um digital twins or or you
1:12:56 know virtual assistants like a virtual
1:12:59 work crew. This is a really good skill
1:13:02 to start to learn right now. Really
1:13:05 good. Um, let's go look at some stuff in
1:13:07 irregulars. There's there's one of the
1:13:09 clubs is called Irregulars. That's for
1:13:11 this channel for this live. Here's Mr.
1:13:14 IT duty-free designs.
1:13:18 Nice. He put intense man cologne. Very
1:13:24 nice. That's awesome. Gareth. Oh, that's
1:13:27 nice. What is that? Look what I
1:13:30 spitefully made because I was trying to
1:13:31 show capabilities of AI and this guy I
1:13:34 was talking to and he was being kind of
1:13:36 a jerk. So I said
1:13:39 something and he and wait, what did he
1:13:42 say? So I said some So I said something
1:13:44 and I'll show you I can replicate
1:13:53 it. Oh, and then you replicated
1:13:56 it. He also said it could not be done.
1:14:00 Did it. I I assume his his jaw hit the
1:14:07 floor. That's pretty funny. My late
1:14:10 brother playing at the place where his
1:14:12 ashes will be scattered. Oh, that's very
1:14:14 nice, Jeff. Sorry for your
1:14:20 loss. That's really beautiful.
1:14:25 YouTube
1:14:28 spotlight. I was
1:14:31 talking to a friend that's an animator
1:14:34 for
1:14:35 Simpsons and she said that animator
1:14:38 industry is up in arms and trying to put
1:14:41 rules around how much AI can be used in
1:14:43 their industry. Well, the ships [ __ ]
1:14:46 sailed in that case. I mean, the writers
1:14:49 struck what two years ago to to have the
1:14:52 the studios not completely decimate the
1:14:54 writing industry and then the actors
1:14:56 followed suit. I can't believe if the
1:14:59 animators are just starting right now.
1:15:04 Uh
1:15:05 [Laughter]
1:15:07 h it's not good. Let's see. Let's go to
1:15:14 Sora. So, here we are in Sora. So, we'll
1:15:17 just go. So, here's a bunch of images. I
1:15:19 can probably just grab any one of those,
1:15:21 but let me go in here. I'll do I'll do
1:15:23 an image, a 3x two image. We'll do a
1:15:25 four up. So, I'm at
1:15:28 sora.com right
1:15:32 now. And I got to do my black bar for
1:15:35 the for the the Tik Tok folk.
1:15:43 And then I'm going to go um make me a
1:15:49 um 70s
1:15:52 style
1:15:55 Saturday morning
1:16:00 cartoon inspired by Speed
1:16:04 Racer about uh let's see
1:16:09 featuring a super cool
1:16:14 muscle
1:16:15 car racing
1:16:18 down the Pacific Coast Highway, the PCH.
1:16:24 Boom. All right, added to the queue.
1:16:27 Let's go look at our little
1:16:30 [Music]
1:16:39 images. That surprises me that the
1:16:41 animators are just now getting to
1:16:45 this. Is that a chicken made of chicken
1:16:48 nuggets? Probably was. Yes, that is a
1:16:51 chicken made of chicken
1:16:52 [Laughter]
1:16:57 nuggets. Big Bang Theory as vinyl
1:17:03 dolls. Taxi
1:17:07 driver. Very cute.
1:17:18 sleeping bad. That's good. They just
1:17:22 look
1:17:22 [Laughter]
1:17:30 tired. I mean, even stuff like like
1:17:33 that, like, you know, make ice cube out
1:17:35 of ice cubes.
1:17:38 Some people are going to come up with
1:17:39 some creative ways to to do some
1:17:42 things. The AI
1:17:45 journey. Magic equals math plus code
1:17:48 plus curiosity. There you go, Pate. Look
1:17:50 at the top corner of that. Magic equals
1:17:53 math, code, and curiosity. I love it.
1:17:57 The The thing they're missing there,
1:17:58 it's math, code, training data, and
1:18:01 curiosity. Because the training data is
1:18:03 what makes all this AI [ __ ] so weirdly
1:18:06 human.
1:18:13 [Music]
1:18:16 All right, here we go. Look. No, it it
1:18:20 totally brought
1:18:21 in that
1:18:25 dude. How do I I can't Oh, yeah. Here I
1:18:32 can. All right. So, there's that. So,
1:18:35 let's create a video out of
1:18:51 this. All
1:18:54 right, we'll see how she do.
1:19:04 Oh, also found out they're not even
1:19:06 animating. They just draw the playbook
1:19:08 and then send overseas to animate and
1:19:11 get it back in 24 hours.
1:19:14 Unbelievable. It's the animation union
1:19:16 that's been working on it. But I mean,
1:19:18 seriously, like they're just working on
1:19:21 it now. I was
1:19:23 making like crazy good images in 2022.
1:19:32 that that feels a little
1:19:34 late for that union to be they're like
1:19:39 hang on people are prompting you know 30
1:19:42 30 minute cartoons into existence. Hey,
1:19:45 we might want to do something about
1:19:46 that. I'm pretty sure I'm pretty sure
1:19:49 the ship has sailed on that one,
1:19:51 kids. That's funny. I mean, it's not
1:19:54 funny. It's it's like it's surprising.
1:20:02 staggering one video cued. All right,
1:20:05 we'll come back to that. Let's see what
1:20:07 else do we what can we have chat GBT do?
1:20:10 Let's see. What do we
1:20:13 want? Um, let me go Oh, I know what I
1:20:16 can do. Let me go find a photo.
1:20:23 [Music]
1:20:33 So, here's some photos that I've
1:20:39 done. That's a fun one.
1:20:51 Let's get that
1:20:53 one. So, I'm going to I'm going to copy
1:20:55 this
1:21:09 image and I'm going to paste it into
1:21:11 chat GPT. So that it it was just a uh
1:21:14 like a cool metal round building with a
1:21:16 with the sunset setting on it. So, I'm
1:21:19 going to say,
1:21:21 um,
1:21:24 reimagine this photo of a
1:21:28 building as
1:21:31 a giant
1:21:33 rocket
1:21:37 ship getting ready for takeoff at
1:21:41 sunset.
1:21:50 I was kind of blown away she was that
1:21:53 she she didn't realize what was
1:21:55 possible. Honest to God,
1:21:58 I the whole purpose of me doing this
1:22:01 channel is to try to get as many people,
1:22:04 you know, up to speed on this stuff as
1:22:06 possible. There are so many people that
1:22:09 don't have a [ __ ] clue what's
1:22:12 happening. Well, I I you know, it's
1:22:15 funny. Open AAI had half a million or
1:22:20 has half a million customers and they've
1:22:22 been adding, you know, many many more
1:22:25 customers because of this image gen
1:22:28 stuff.
1:22:29 Um I I have a I have a a strong
1:22:33 suspicion now that the image gen stuff
1:22:35 is going to be the is going to be the
1:22:38 catalyst that gets a lot of people aware
1:22:39 of what's happening.
1:22:45 Oh, that's kind of cool what it's
1:22:48 doing. It's looking very Jules
1:22:52 Vern. Oh, that's nice. Just it just
1:22:55 adjusted the
1:23:06 light.
1:23:09 Huh? Oh, I know what we'll do. So, okay.
1:23:11 Okay. So, we'll generate this rocket and
1:23:13 then maybe we can generate an
1:23:21 astronaut. Oh, it's taking off. That's
1:23:25 cool. All right. So, now let's
1:23:28 say create an
1:23:32 interior or
1:23:34 shot of the astronaut.
1:23:39 not
1:23:41 um in the same
1:23:45 style as if there is a camera in
1:23:51 the
1:23:53 cockpit or whatever the [ __ ] it's called
1:23:55 that they go
1:23:59 in. All right, let me go back to Sora
1:24:01 and see if we have an animation now.
1:24:07 Well, we sure
1:24:09 do. But what the
1:24:11 [Laughter]
1:24:23 [ __ ] sore is
1:24:25 weird. Let me take that image and go go
1:24:28 over to Lumalabs. I bet Lumalabs is way
1:24:31 better.
1:24:43 Where is my luma labs? Hello, Luma.
1:24:46 Luma, where are you? Wherefore art thou,
1:24:48 Luma
1:24:50 Labs? I've lost thee. I have lost thee
1:24:53 in a sea of tabs. Luma
1:24:56 Labs.
1:24:57 Luma. Luma Labs.
1:25:01 Join us. Try now in dream
1:25:06 machine. Camera motion is
1:25:09 here.
1:25:12 Paste key
1:25:14 frame rate
1:25:19 two. We'll do it in
1:25:21 720p. 5 seconds is fine. And we'll just
1:25:24 say
1:25:26 um car
1:25:29 races down
1:25:32 highway
1:25:35 as
1:25:38 camera. Oh, here's camera movement
1:25:41 here. Zoom out. Oh, that's cool. as
1:25:44 cameras zoom
1:25:48 out. They just added it as as a visual
1:25:52 visual set of
1:25:54 choices. Oh, let's do aerial
1:25:58 drone. All
1:26:06 right. I don't even know what accent I
1:26:08 did. What accent did I do?
1:26:14 Put your cartoon face in place of his.
1:26:16 Oh, that's a good idea. Oh, yeah. Let's
1:26:19 do that. Can I do it here? Edit
1:26:26 prompt. Yeah, I
1:26:28 can. Oh, no, I can't. Wait. Plus. Yes, I
1:26:34 can. Choose from library. Upload. What
1:26:37 do I have in the library? Do I have
1:26:38 anything? No.
1:26:42 plus upload from
1:26:45 device. We'll look up uh
1:26:59 dreams. Okay, that's good.
1:27:06 Um, make the
1:27:09 driver. Wait, did it upload that that
1:27:13 image? Upload from
1:27:17 device dreams.
1:27:34 Make the driver look like
1:27:37 him.
1:27:39 Remix.
1:27:42 Okay. It's supposed to zoom in and show
1:27:44 a video of it
1:27:46 driving. Oh, yeah. That was that would
1:27:49 be pretty
1:27:51 cool. But let's go back to Lumal Lab.
1:27:54 See what's going on
1:27:56 there. Well, we got a we got a zoom out.
1:27:59 We got an aerial drone of
1:28:03 it. Actually, if you reverse this, it
1:28:06 would be
1:28:22 fine.
1:28:25 Crazy. It's crazy Italia. Crazy
1:28:29 Italia. Crazy crazy crazy
1:28:33 Italia. Your regulars. Go look at the
1:28:36 regulars. Boy. Boy. Go do
1:28:39 it. Go do it. Fierce where nobody knows
1:28:43 your name. The social anxiety
1:28:49 bar. That's pretty good. I like it. Oh,
1:28:53 that one's nice, Vicki. Except why is
1:28:57 the why are the two gnomes have this
1:29:00 one's
1:29:01 head? He doesn't have a
1:29:03 [Laughter]
1:29:08 body. Oh, here's the original. Oh,
1:29:10 that's really
1:29:13 cool. That's amazing. It's actually
1:29:16 amazing.
1:29:18 Honest to God, I feel like you take a
1:29:21 drawing like that from a kid and with
1:29:24 them
1:29:25 present, you know, make
1:29:30 that it feels like it's a
1:29:33 reinforcing kind of
1:29:36 magical
1:29:38 gift, something for Brandon, Danielle,
1:29:40 and Kyle.
1:29:43 While the producer is busy with post-it
1:29:45 notes, Kyle's shock
1:29:50 collar. That's really
1:29:55 good. Creeps very, very slowly up the AI
1:29:59 learning lab with a better
1:30:03 idea. Wow, that's really those are good
1:30:06 and
1:30:07 disturbing. What's the prompt for that?
1:30:10 Did did you do like is that like a like
1:30:13 a storyboard with text on it cuz it it
1:30:16 did it in courier which is good
1:30:19 screenplay kind of typography. That was
1:30:21 kind of
1:30:22 cool. The videos are all backwards.
1:30:25 Yeah, it's
1:30:31 weird. Yeah, these are sawlike images.
1:30:33 They are sawlike images. They're really
1:30:35 good.
1:30:37 I told chatb4 I'm paying you. Why is
1:30:40 there a delay in creating my image? I
1:30:43 don't think that'll help, but good for
1:30:45 you for go for for going for it. I did
1:30:48 that last week. Text was Canva. Oh,
1:30:50 okay. Yeah, that makes sense. Actually,
1:30:53 wait. So, watch. Copy image. Okay, this
1:30:55 will this will be good. Silver Fox,
1:30:57 watch this. Oh, that's
1:31:00 nice. Yeah, that's sort of in the same
1:31:03 same style, isn't it?
1:31:05 Although he doesn't look like he's
1:31:06 taking off, but whatever. Okay, so let's
1:31:09 do a new
1:31:11 chat. We'll put in Silver Fox's
1:31:15 thing. I love this style of storyboard.
1:31:24 Love the
1:31:30 imagery,
1:31:33 typography, and layout.
1:31:38 Give me 10
1:31:41 ideas in
1:31:48 text that would
1:31:51 look good in this
1:31:57 style. A meeting. The meeting that never
1:32:00 ends. A long conference table stretching
1:32:02 into the difference. Figures slumped
1:32:03 over it.
1:32:05 Right panel. A single unplugged
1:32:07 extension cord crawling down the
1:32:10 hallway. Okay. The algorithm has a
1:32:14 dream. Marketing's fifth circle.
1:32:16 Creative brief with a trap. Focus
1:32:20 group. The founders
1:32:24 ghost. Let's just do the first one.
1:32:27 That's pretty good.
1:32:29 Uh, show
1:32:31 me number one.
1:32:35 Let's see how close we can get it to
1:32:38 get that
1:32:42 style cuz that's creepier than
1:32:49 [ __ ] Jesse from Austin. That's damn Tik
1:32:52 Tok channel. Damn it. You're damn right
1:32:54 it is. Um, it might end Saturday, but
1:32:58 that's why we're dual broadcasting on
1:33:01 YouTube
1:33:02 because we knew this day might
1:33:09 come. Let's put our money together and
1:33:12 buy Tik Tok. We would have to put it
1:33:14 together pretty quick. It's Wednesday.
1:33:16 It goes out Saturday.
1:33:21 They laughed at a $20 million 20 billion
1:33:24 dollar offer
1:33:25 [Laughter]
1:33:31 already. Oh, you did overlays and
1:33:34 backgrounds and such. Good
1:33:36 lord. You got five on it. That's good.
1:33:40 All right, we got five bucks. We just
1:33:42 need We just need
1:33:44 uh three billion.
1:33:48 I mean 300
1:34:00 billion.
1:34:03 Okay. You made the image square and it
1:34:09 cut everything off. Make it
1:34:14 wide so it
1:34:21 doesn't
1:34:26 think can't we just make it a lovable
1:34:29 there's a Mark Cuban actually Mark Cuban
1:34:31 put out a call I don't know two months
1:34:35 ago that if anyone wanted to recreate
1:34:39 Tik Tok on the AT protocol which is I
1:34:41 some open source protocol, he would fund
1:34:44 it. And there's a group that has been
1:34:47 talking about that. They put something
1:34:48 together. He actually just stepped in
1:34:50 and funded it um two days ago. So I
1:34:55 think it's what's that called? Is it
1:34:56 called? It's not called Blue Sky. It's
1:34:58 called It's something like Blue Sky. I
1:35:00 don't know what it
1:35:01 is. Skylight. Yeah, it's called
1:35:03 Skylight.
1:35:05 Um, so Skylight could be a thing, but
1:35:09 you know, they're they're they're just
1:35:10 they're trying to build it from the
1:35:12 ground up with just a couple of
1:35:15 people.
1:35:18 [Music]
1:35:23 So, please, Chat GBT, please create me a
1:35:26 five-step plan to save Tik Tok by
1:35:28 Friday.
1:35:36 It's really not worth it without the
1:35:38 algorithm. Although I would think that
1:35:40 you could Well, I don't know. [ __ ]
1:35:44 know. I was just going to say I was just
1:35:47 going to say the algorithm is probably
1:35:49 easy and then PE will jump through the
1:35:52 screen and smack me.
1:36:04 [Music]
1:36:07 But it did copy the the uh
1:36:09 typography. Silver
1:36:15 Fox, come on. Finish it
1:36:18 up. Finish it up. There you go. Look at
1:36:21 that. That's pretty cool.
1:36:30 It's not quite as as edgy as the image
1:36:34 you
1:36:36 did, but you could jank it up. While
1:36:38 strategy aligns its quarterly KPIs, the
1:36:41 power quietly unplugs itself from the
1:36:44 [Laughter]
1:36:50 room. On a side note, I ran into Hulk
1:36:53 Hogan today.
1:36:55 That electric cord is creepy as [ __ ] It
1:36:58 is. It's also a British electric cord.
1:37:00 It's It's those round the round holes
1:37:02 with the two prongs or the two
1:37:07 holes. I got some chicken nuggets to
1:37:09 give to the Tik Tok fund. Yeah, I think
1:37:11 we can pull it off. A group a a group of
1:37:16 43 irregulars has made a bid for Tik
1:37:18 Tok. $5 and a large box of chicken
1:37:22 nuggets in the shape of a
1:37:27 [Music]
1:37:27 [Laughter]
1:37:33 chicken. Oh man. All right, let me get
1:37:37 out of here. Do I have anything else to
1:37:38 check out? Nothing else is still
1:37:39 rendering, right? Oh, no. This this uh
1:37:42 stylish driver
1:37:44 transformation.
1:37:48 [Music]
1:37:50 Well, okay. That's a fail. I said I said
1:37:55 make
1:37:58 make that driver look like
1:38:03 me make make that driver look like me.
1:38:06 And it made
1:38:10 that. It's all in the jacket.
1:38:19 Lest you think these tools are not still
1:38:24 janky. Oh man. All right, I'm getting
1:38:28 out of here. That's a That's a beautiful
1:38:30 fail to end
1:38:35 on. The theme for the week. The theme
1:38:38 for the week is in spite of impressive
1:38:41 advancements in
1:38:43 technology, this shit's still janky. And
1:38:47 then the other theme for the week is
1:38:50 don't try to do anything good. Just have
1:38:52 fun. Because the minute you try to do
1:38:55 something good, it's going to piss you
1:39:01 off. Lord, digital gods, thank you for
1:39:03 the super sticker. I appreciate that. We
1:39:06 should write a comedy book about this.
1:39:08 Oh my god. This week would be a good
1:39:10 one. This is just a tragedy of errors
1:39:12 this
1:39:15 week. Woo. What did I start? Oh, the
1:39:18 thing the thing I'm actually really
1:39:20 excited about I I think I want to find a
1:39:23 way to build the app. And it's not even
1:39:26 an app. I think it's a protocol.
1:39:33 um is the emotional architecture
1:39:36 thing.
1:39:38 Like wouldn't it be
1:39:42 cool? Wouldn't it be cool to be able to
1:39:44 do the equivalent of
1:39:46 like create mood boards in midjourney?
1:39:49 Here's a bunch of song styles, you know,
1:39:51 generated from whatever that you wrote
1:39:53 or you want, you know, you used AI to
1:39:56 generate. Here's a bunch of writing
1:39:58 examples. Here's a bunch
1:40:01 of, I don't know, world examples. Here's
1:40:04 a bunch of
1:40:06 characters. And just be able to like add
1:40:11 all these frameworks into this world
1:40:14 builder or content generator and let
1:40:18 people come in and generate content in
1:40:20 your style and like co-create with them.
1:40:23 There's I I don't know. There's
1:40:24 something there that I'm really excited
1:40:25 about. Kyle the Trucker looking like
1:40:28 Jensen Wong. I I really do need to get
1:40:30 me a leather
1:40:34 jacket. Go with the Go with the uh the
1:40:36 leather jacketed CEO look. I like
1:40:39 [Laughter]
1:40:43 it. Oh
1:40:46 man, saved someone hundreds of dollars
1:40:49 today. They needed a house staged and I
1:40:52 said to do it virtually. Oh, that's
1:40:53 really smart.
1:40:56 That's really smart. That's pretty
1:40:59 good. Once it all works together, we'll
1:41:01 forget about the janky stuff. Yeah, I
1:41:03 know. We will. I the I I still I I know
1:41:08 this to be the case. We will look back
1:41:11 on these these janky times with
1:41:13 nostalgia when this stuff gets really
1:41:15 really good. Um when you had to when you
1:41:18 had to actually figure it out and you
1:41:20 had to be as frustrated as I get
1:41:22 sometimes. Um, it's not going to be like
1:41:25 this for that long. So, we will look
1:41:28 back on these times fondly, I do
1:41:31 believe. All right. Um, tomorrow's
1:41:33 Thursday, normal time
1:41:35 tomorrow. Um, we'll be back. Hopefully,
1:41:38 I'll have some time to to look at some
1:41:40 things and actually put some thoughts
1:41:42 together about what to do tomorrow
1:41:43 night. But if not,
1:41:46 not. We'll just chat. Add it. Um, coffee
1:41:50 is vital. I started uh I started uh
1:41:53 methyline blue methylene blue today. So
1:41:58 didn't do anything other
1:42:00 than change my pee of particular
1:42:06 color. Will we be alive when it gets
1:42:08 that good? Yes, we will. And if we're
1:42:11 alive for another five years, we we
1:42:13 potentially start to hit uh life escape
1:42:19 velocity. Well, that explains a
1:42:27 lot. All right. All right, everyone. I'm
1:42:30 out of here. I will see you tomorrow
1:42:32 night. Peace out. Go to the uh business
1:42:35 guild tomorrow night with Kelly Camp
1:42:37 later.