8/19/2025 - Generative AI: A Collaboration with Ghosts

Description

Chapters

Transcript

0:00 [Music]
0:11 Girl looking
0:15 find a night
0:18 [Music]
0:19 fellas got you in his sight.
0:26 What you doing with a clown like me?
0:32 You're surely one of life's little
0:36 [Music]
0:38 mystery.
0:40 So down at last
0:46 [Music]
0:49 I ever
0:51 [Music]
0:52 love.
0:55 What did I do?
0:58 What did I say
1:01 [Music]
1:04 to turn your angel as my way?
1:11 [Music]
1:37 Woohoo!
1:44 [Music]
1:46 Hey! Hey!
1:51 [Music]
1:56 Look at her.
2:03 [Music]
2:15 [Music]
2:38 Good evening, good people.
2:41 [Music]
2:42 Come
3:12 here.
3:14 [Music]
3:23 Freedom came my way that night.
3:27 [Music]
3:30 Just like a jet plane in and out.
3:36 I was hauling ass at a million miles an
3:40 hour.
3:42 wondering how hard I'd hit
3:49 when they came into the station.
3:54 Yeah. Um, what's happening good people?
3:58 It is Tuesday night. I've got the hair a
4:00 little under control. It was wild
4:02 before. Producer Brandon was like, "Get
4:05 to get you in hair and makeup, boy."
4:07 We took care of it. All's good now. Uh,
4:10 no one's on TikTok. We got a few people
4:12 over on the YouTube. Good day everybody.
4:14 How is it going?
4:20 [Music]
4:31 [Music]
4:38 [Music]
4:48 Now
4:52 [Music]
5:06 Oh man. All right. What are we going to
5:08 talk about tonight? What do you want to
5:09 talk about? What are we going to talk
5:12 about? People, good people. Good people
5:15 of the internet. Hey, good people.
5:23 All right. Fantastic. Fantastic. We got
5:26 six people on Tik Tok. Seven.
5:29 Do I have seven? Seven people on Tik
5:31 Tok.
5:35 Enjoy party steo.
5:37 [Music]
5:50 I could do that. I'll tell you what, if
5:52 we don't get more people on TikTok, I'm
5:54 going to twerk. I'll do it. I'll
5:56 flipping do it. You know, I'll do it.
5:59 Did you see me in my little Catholic
6:00 schoolgirl outfit? I thought I looked
6:02 cute.
6:04 [Music]
6:33 Talk about pika.
6:35 Irregulars on
6:38 what? What does that say?
6:42 I cannot read that for the life of me.
6:44 Oh, Irregulars on Netflix 101.
6:47 [Music]
7:21 in a westerly
7:23 direction.
7:28 This car is my train.
7:33 [Music]
7:34 Been driving. I've been wondering
7:38 what it is I want from again.
7:43 Feel like an 8-year-old man
7:47 holding on a 29.
7:52 head on that horizon.
7:55 [Music]
7:57 This California line
7:59 [Music]
8:06 up ahead of trucks carrying a wide load.
8:12 Preab house cut in half.
8:16 Cute little front door and two windows.
8:18 My love
8:20 ain't sure whether a cry should laugh.
8:25 You see, I broke a home up myself once.
8:29 As I stumbled to that door,
8:34 I read a note by the dawn light.
8:38 It said, "Don't you come around here
8:43 anymore."
8:46 Well, I've had enough.
8:51 How this freedom of
8:55 never was good with decision.
8:59 At least that's what I've been told.
9:02 [Music]
9:11 Um, I just had an idea something we
9:14 could do tonight.
9:15 Um,
9:19 thank you, Silver Fox.
9:22 Thank you. Thank you. Thank you.
9:26 If you missed it tonight, we had a
9:28 really good meet and greet. We had a
9:29 special edition meet and greet where we
9:32 uh we officially launched the uh AI
9:35 readiness training program
9:38 and uh
9:41 we brought back speakers from AI
9:43 Festivus who caught us up on where they
9:45 were. It was really good. It's really
9:46 fun. My multitasking this evening is
9:49 creating contracts.
9:52 Do I know what I'm doing?
9:56 No, but chat GPT does. There you go.
10:00 So, what I thought might be kind of fun.
10:03 [Music]
10:19 All
10:47 right. I have an idea. I have an idea.
10:50 All right, for anyone who didn't come
10:52 tonight, they're losers. What are we up
10:55 to? 16. All right,
10:58 here's what we're going to do.
11:03 I just remembered there was a thing.
11:05 Sorry, I don't I don't know why I'm hair
11:09 obsessed tonight, but you know,
11:13 feel like I've got a tick.
11:17 um
11:19 like like a a physical tick, not the
11:22 insect tick.
11:25 Two different ticks. Not important right
11:27 now. Here's what I thought might be fun.
11:30 So, do you remember back in the olden
11:32 timey days, like two years ago,
11:36 there was a thing called
11:48 It was it was when Dolly
11:52 when the Dolly image generator first
11:54 came out. It was this second separate
11:57 experience. You'd made made images in a
11:59 different place than you made. No, it
12:01 was in the same place. Still in chat
12:03 GPT.
12:05 But chat GPT roundtpping was you would
12:09 generate an image,
12:12 then you would have chat GPT describe
12:14 the image,
12:16 then you would use that description to
12:18 create another image, and then you would
12:19 do that in cycles.
12:25 So, what if we did a sophisticated
12:27 version of that tonight?
12:29 We'll start in chat GPT. Maybe we'll
12:32 start somewhere else.
12:37 wherever. We'll start somewhere. We'll
12:39 get some ideas for visuals. Then we'll
12:42 go to midjourney. We'll make a bunch of
12:44 visuals using some some random sref
12:48 style references.
12:52 Then we'll come back to chat GPT.
12:57 Tell us the backstory of the image.
13:11 Then once we have the backstory of the
13:13 image, we'll have us have it write us
13:15 lyrics for a song. Then we'll go to
13:18 producer.ai,
13:20 make the song,
13:28 then maybe we'll start to make a video.
13:30 I don't know how much of that will get
13:31 done, but that'll be fun.
13:35 Oh my god. A show called the regulars.
13:41 Get it? Yeah, exactly.
13:50 [Laughter]
13:55 All right. All right. That's a good
13:57 point. But we can do this fast. Crazy
14:03 little thing called love.
14:07 [Music]
14:12 All right, let's see. Where are we going
14:15 to start? Start in chat. GPT. Yeah.
14:19 Yeah, we'll start in the chat. Japa.
14:22 All right, let me I've got to share my
14:24 screen apparently. Apparently.
14:27 Apparently.
14:28 Apparently I gotta share my screen.
14:31 Apparently
14:34 because apparently
14:36 I got to share my screen.
14:39 Okay, here we are.
14:56 I want you to act
15:01 like you're channeling
15:06 people from times past.
15:09 And I want you to describe
15:12 10 different unforgettable characters
15:14 that didn't really
15:17 exist.
15:19 And I want them to all exist
15:22 in some really unique town or really
15:24 unique place. And I want them to have a
15:27 nickname. And I want them to have a
15:28 backstory. And I want you to describe
15:31 them physically in excruciating detail.
15:37 Period.
15:40 Oh, no.
15:45 I want you to act like you're channeling
15:49 entities from another time. And I want
15:53 the people that you're channeling to be
15:57 unforgettable and unique and black sheep
16:01 within their communities.
16:04 And I want their
16:07 locations to be as interesting as they
16:10 are.
16:12 And I want you to explain them in
16:15 excruciating detail
16:17 and the place in excruciating detail and
16:20 how old they are and anything
16:23 interesting about representing them
16:27 visually. And I want 10 of them. Go.
16:34 Oh, come on. Seriously.
16:46 Okay. I'm hating life.
16:49 I'm hating life. We got to reload. I
16:52 might have to restart.
16:55 Talk at you. Can I talk at you? Will you
16:58 write it down?
17:01 This looks better. Okay.
17:06 All right. I hope this works this time.
17:09 I want you to channel I want you to act
17:12 as someone who can channel people from
17:14 other times. I want you to find
17:18 interesting times in the 20th century.
17:21 And I want you to channel 10 different
17:24 people, all from interesting times, all
17:26 from interesting places. Each of them is
17:29 a black sheep of their community.
17:32 has a fascinating nickname that implies
17:34 a deep and traumatic story.
17:37 And I want you to describe them at a
17:41 specific location that is as memorable
17:43 as they are.
17:45 And I want you to make 10 of them. And I
17:47 want you to make them now.
17:53 No,
17:56 it did. It Ite ate it again.
18:11 just type. You know what? Here's the
18:13 thing, producer Brandon. Every time I
18:15 type, someone's like, "Why don't you
18:16 just use the microphone, you big dumb
18:18 dumb?" So now I start out using the
18:20 microphone because I don't want to
18:21 torture the people, and now we're
18:23 torturing the people. Do you understand?
18:26 Do you understand? Let me try another
18:28 way.
18:31 [Music]
18:36 Oh, this is broken now.
18:41 Can you
18:44 Oh, that's not working. All right. So,
18:46 my speech.
18:50 All right. That's what's going on. All
18:51 right. So, we're going to type it. So,
18:54 one of the first lessons of AI is never
18:59 start a Tik Tok channel or a YouTube
19:02 live stream
19:04 demoing AI live. Never do that. That's
19:09 that's the first lesson in AI is don't
19:12 start a streaming service where you
19:14 position yourself as the expert and
19:16 can't even get letters into the chat
19:18 hole. Okay.
19:21 Can you fix your mic? What's wrong with
19:23 my mic?
19:25 My ears.
19:34 Is that better? Hello. Hello.
19:38 Microphone is way better. It should
19:40 work. Yes, it should work. But I'd have
19:42 to restart my machine because it's not
19:45 working.
19:47 But both ways of inputting uh things are
19:49 not working. Archetypal Architect. Okay.
19:52 I tried to use Sora today instead of V3.
19:54 Holy crap is it stupid. Sora is really
19:57 bad. Absolutely refused to come close to
20:00 following my instructions. ignored my
20:02 starting frame for the most part. Sora
20:05 is really bad. I Whatever. Whatever.
20:12 Open AI is a mess. Okay.
20:16 I can't believe I've got to type this.
20:18 Okay. I want you to act
20:23 as a channeler
20:27 of
20:30 memorable
20:33 fictional people from
20:38 the 20th century.
20:42 They should have
20:46 nicknames that imply
20:49 some traumatic past
20:53 or
20:55 comedic
20:57 event
20:59 that stuck.
21:02 They should live in
21:07 equally interesting
21:10 places.
21:12 and times.
21:16 I want you to
21:19 describe them in excruciating
21:24 detail.
21:25 Oh, you know what we're going to try
21:26 here also?
21:28 Now, Kyle, what
21:31 we're going to do? We're going to do the
21:32 same prompt with three different models.
21:36 Uh, I want you to describe them in
21:37 excruciating detail.
21:41 as well as an interesting
21:45 location they
21:48 are in. And give each
21:53 one a
21:56 surprising
22:01 uh
22:03 item
22:06 they possess.
22:10 All right.
22:12 So,
22:15 Knuckles McGary,
22:18 Chicago 1928 at the height of
22:20 Prohibition
22:38 Wait, did it only give me four? I asked
22:41 for 10.
22:44 Maybe I didn't. Fine. 10 it is. Oh, by
22:47 the way, I've got So, this is
22:49 fascinating. If you haven't done this
22:50 yet, um
22:53 I I changed my personality of my chat
22:56 GBT to a cynic
22:59 and
23:02 about 50% of the time
23:06 I'm just expecting chat GPT to be like
23:08 chat GPT. And it's like, oh, so you
23:11 really want to do that? Are you really
23:12 sure you want that? It's like it's like
23:15 a smartass.
23:19 It's as if chat GPT is me
23:23 and all I want is an answer and I get a
23:25 smartass answer and it's a it's it's a
23:27 pain in the ass. But because I have ADD,
23:30 I don't change it back to the default.
23:32 So,
23:34 okay. So, there's those. So, now I need
23:37 now I need a detailed image prompt
23:45 for each one.
23:48 with a uh
23:52 unique
23:58 unique
24:01 visual style
24:04 that matches
24:07 the character.
24:10 All right,
24:13 here's 10 walking calamities. A hulking
24:16 Irish American bootleger on a foggy
24:19 Chicago river docket dock at night. His
24:22 knuckles bandaged and bruised holding a
24:24 violin case like it's contraband. This
24:26 is great style black and white film
24:29 noir. All right, so here's what we're
24:31 going to do. We're just going to copy
24:32 these. We're going to go to uh
24:36 we're going to go to
24:39 that's producer. We'll come back to
24:40 that. Hang on. Just calm down people.
24:42 Calm down. Midjourney. I know I got to
24:44 switch my tabs, Brandon. I know. I know
24:47 I have to switch my tabs. I'm very well
24:49 aware of that. Okay, so let's get our
24:54 let's get our stylization at 200.
24:58 Let's turn weirdness up to,400 because
25:01 we want we want art.
25:04 We'll go 16 by9.
25:12 All right.
25:14 Boom. Okay. So, you're not going to see
25:17 me copy and paste, but you'll So,
25:20 spilled milk. Rosenblat, a young woman
25:22 in pastel 1950s kitchen. All right.
25:26 Beautiful.
25:29 Okay.
25:32 Next one is burnt toast duca.
25:42 So, you see what I'm doing here, people?
25:44 You may or may not see what I'm doing
25:46 here, but what I'm doing here
25:48 is
25:50 doing something creative without using
25:52 my brain,
25:54 which which is a kind of there's a kind
25:58 of intentionality to what I'm doing,
26:00 which is I want chat GPT to inspire me,
26:05 right? I don't want to have to be the
26:07 inspiration here.
26:09 So, I'm getting chat GPT
26:13 to give me crazy ideas. Now, I gave it
26:16 like the the prompt here was pretty
26:19 good, I thought.
26:24 So, we're going to go get all these
26:25 going. We're going to get 10 of these
26:26 going.
26:28 And then I'm going to take that same
26:29 prompt and I'm going to do it with a
26:31 different model, the the thinking mini
26:34 model, which is really good at writing,
26:35 and see if those are any better, any
26:38 more better.
26:40 Copy.
26:42 Paste.
26:44 Go.
26:48 Copy.
26:53 And you might be asking, "But Kyle, is
26:55 there a method to your madness? It just
26:57 seems like madness.
27:01 I don't know. If it comes out good, then
27:05 there was a method to it. If it comes
27:07 out bad, then there wasn't.
27:11 Okay. Going to take this prompt.
27:15 Oh, and then I had a followup about
27:17 making it into images, didn't I?
27:26 Yeah. Okay. So, we're going to go to
27:31 notes.
27:38 Come on.
27:42 No notes. There's notes. Notes.
27:47 Copy.
27:49 Paste. Okay. Copy. All right. Right. So,
27:52 we're going to go new chat.
27:54 Um,
27:57 we're going to change this to thinking
27:59 mini. I'm going to put in that same
28:01 prompt. Let's see if these are more
28:03 interesting or
28:06 thinking
28:12 formatting characters in towns. Fine.
28:15 You wanted memorable fictional 20th
28:17 century people. Oneeyed Monty.
28:21 Montigue Monty Hargreaves.
28:26 I gotta tell you, I just love the
28:28 writing
28:30 of thinking mini. Monty is the size of a
28:34 small island gale. Barrel chest, a patch
28:37 of white hair like a lighthouse on his
28:40 left temple.
28:42 That's gorgeous. All right.
28:46 Love these.
28:48 So, let's go get our image prompt
28:54 and
28:56 do that.
29:09 Okay.
29:11 Picture prompts. Prompt. Whimsical
29:14 nautical watercolor plus inkwash. This
29:16 is oneeyed Monty.
29:23 Look at these things. They're crazy.
29:34 Okay. Copy
29:36 black bar. I knew that. Shook.
29:44 Look how look at look how much bigger.
29:46 Oh, you don't you can't see that. Pin
29:49 pin pen. Just found you guys. I hope I
29:54 see this again because I'm just on my
29:57 way out for work stuff. So this thing,
30:01 so you can find this on YouTube. If you
30:05 go to YouTube, the account is learning
30:08 lab-ai.
30:10 On Tik Tok, it's AI learning lab. on
30:12 YouTube, it's learning lab-ai
30:15 and this live will be streamed there.
30:18 So, keep coming back. And we're live uh
30:21 five nights a week starting at 8:00 PM
30:23 Mountain time except tonight we were
30:25 late because I had an AI AI salon. Okay,
30:28 but thank you. Welcome.
30:31 We're going to do 10 more of these now.
30:33 [Music]
30:39 Shake it down.
31:00 [Music]
31:02 I'm sharing these, right? Yeah. Okay. Uh
31:05 oh, producer Brandon's about to write me
31:07 a nasty gram. I don't know what's
31:09 happening. I don't know what I did.
31:13 YouTube comment. Sup, Kyle? I finished
31:16 my first standalone
31:20 app with open. It worked great. There
31:22 you go. This is the thing, you know.
31:25 Wait, why is this writing these as
31:27 stories?
31:32 Is that what it prompted?
31:36 I guess I should read the prompts that
31:37 it wrote.
31:49 These are annoying. I don't like them at
31:51 all
31:55 on the table
31:57 style. Whimsical watercolor with bright
31:59 saturated tropical texture fabric weave
32:02 tablecloth camera.
32:09 I think these are overly complicated.
32:11 All right, so let's go. Um, these
32:14 prompts are overly
32:18 complicated.
32:20 Um, make them less structured
32:26 and more
32:29 interesting visually.
32:39 So Mary Mary on Tik Tok wanted to know
32:42 what we were doing and I'm failing to
32:45 figure out how to encapsulate what it
32:47 exactly is that we're doing in the
32:50 comments space allowed by Tik Tok.
32:52 >> Yes, that's a good that's a good
32:54 question. So Mary Mary, here's what
32:55 we're doing.
32:58 Back in the back in the olden timey
33:00 days, we did that there was this fun
33:02 parlor trick called chat GPT image
33:05 roundtpping where you would have chat
33:07 GPT write a prompt. You would then take
33:10 that prompt, put it in chat GPT's image
33:13 maker, Dolly Dolly 2.
33:16 It would generate the image. You would
33:18 then upload that image to ChatGpt, have
33:20 it describe it, and then take that
33:22 prompt, put it back in the image
33:24 generator, and so so you do this thing.
33:26 So, I thought, what if we did a a
33:30 more sophisticated version of that where
33:33 we're going to start in chat GPT, have
33:35 it write a bunch of ideas, go to
33:37 midjourney, generate a bunch of styles
33:40 until we find something that's really
33:42 cool that we really like, take the image
33:44 back to ChatGpt, have ChatGpt tell us
33:47 the story of the image,
33:50 then create a set of lyrics, then go to
33:53 producer AI, write a song about it, then
33:55 come back and based on the song, maybe
33:59 do a story board. I don't know. Maybe
34:02 like I don't know how far we're going to
34:03 get because already I'm taking too
34:06 [ __ ] long. Although these pictures
34:08 are really cool right here. So that's
34:10 that's the plan. That's the plan.
34:14 That's the plan.
34:16 You all right? You understand what I'm
34:18 saying? You understand what I'm saying
34:21 here? Huh? Yeah. It's fantastic. Yeah.
34:25 Mhm. Yeah. That's that's super cool.
34:29 Yeah. Yeah. You know, he was on an AI
34:31 learning lab and he was he was he uh he
34:36 he he told people that he knew what he
34:38 was doing and then he just kept pasting
34:40 things into in the midjourney and it it
34:44 didn't seem like anyone was learning
34:46 anything. But they were all still there.
34:48 They were watching him. I I don't know
34:51 why they were watching him. I don't know
34:52 what they were why they were
34:55 That channel's very confusing to me. He
34:59 He's kind of a pleasant fellow, but
35:02 doesn't doesn't seem super bright. He
35:05 doesn't seem super bright.
35:07 And
35:10 talented. I I wouldn't use the word
35:12 talented. Probably not the right word to
35:14 use.
35:16 Uh I I don't know if he's got any
35:18 qualifications whatsoever. Uh I need
35:23 eight
35:26 to 10. Come on.
35:32 Oh my god. 10 new characters. Stop.
35:36 Stop. Stop. All right. Let's go back to
35:39 Let's go back here now. Okay. So, now
35:42 we've got
35:44 Let's go see if we've got
35:47 a songw worthy image.
35:50 These black and white ones are pretty
35:52 cool, aren't they? Look at that one.
35:53 That's got some like an old photographic
35:57 watermark on it. That's kind of cool.
36:01 Okay, those are nos. No.
36:06 Interesting.
36:08 Oh, really interesting.
36:14 Is that blood on her dress?
36:24 What? What the hell? There's There's
36:27 pancake batter splattering all over the
36:29 phone.
36:35 Oh, that looks like Hemingway or is
36:38 Yeah, Hemingway.
36:42 Oh, these are fascinating. Oh, that
36:43 one's really cool.
36:53 All right. I like these styles.
36:59 All right. So, th this is one we're
37:02 going to come back to. A wiry man
37:05 hunched over a realtore tape.
37:12 Cute.
37:17 That's interesting. That's interesting.
37:34 80s. That one's kind of cool.
37:40 Here's
37:48 these weird ones.
37:51 Don't like any of those.
37:54 That's cool. That's really cool.
38:04 Let's turn this one into a
38:09 Let's turn this one into a low motion
38:11 animation.
38:19 Love her.
38:22 She lived a hard life.
38:32 Which by the way, if you're not taking
38:34 in
38:37 what I think is kind of a remarkable
38:39 thing about this approach is there are
38:43 images here that are absolutely
38:45 inspiring.
38:47 Like you you could take some of these
38:50 images and really run down some rabbit
38:52 holes.
38:58 Like I like that one. All right, so
39:00 here's what we're going to do.
39:04 Too many choices. This is one of the
39:06 problems with AI. Too many choices.
39:08 These are really cool. Like I love her.
39:15 All right, so here's what we're going to
39:16 do. We're going to just download the
39:18 ones that are interesting and then we're
39:20 gonna
39:23 That one's interesting.
39:30 That one's interesting.
39:38 That one's interesting.
39:44 She's got a story to tell.
39:48 So does he. We already made the video of
39:51 that one.
40:02 I think it's that that dude. Oops. What
40:04 did I do?
40:09 That dude.
40:22 That dude.
40:26 That dude.
40:33 That dude.
40:36 It's all dudes.
40:38 There's a There's a lady.
40:40 All right. These guys.
40:44 I think this one.
40:49 You're like, Kyle, what are you going to
40:51 do? You'll see.
40:55 All right. We didn't like those.
40:58 That one's fascinating. Oh, damn it. The
41:02 hell is happening?
41:09 And him. Okay. So now now we got all our
41:12 we got all our starter images.
41:15 So let's go look at Oh, the video is not
41:18 quite done yet. I want to see the video,
41:37 huh? Midjourney slow tonight. Slow,
41:40 slow, slow. All right, we're going to go
41:42 back to chat GBT here in a second.
41:45 I want to see this video. I want to see
41:47 if the videos are any good. Come on,
41:49 chat GBT or Mid Journey. Stop being so
41:51 slow.
41:53 This is making me whiny. Is anyone still
41:56 here?
41:58 A few people.
42:04 Some nights I actually explain what I'm
42:05 doing. This isn't one of those nights
42:07 apparently.
42:27 All right, whatever. Okay, back to chat
42:29 GBT.
42:32 Okay, now
42:38 I'm going to take I'm going to take one
42:40 image
42:44 and we're going to do the same simple
42:46 prompt with four different models
42:49 because part of part of what I'm
42:51 exploring right now is
42:58 I I guess just the nuances of writing.
43:00 Like I kind of knew what chat GBPT40
43:04 would do in in most of the cases that I
43:06 used. I don't have a clue about chat
43:09 GBT5 right now. But if you didn't know
43:12 this, one of the things that AI is
43:15 really good at is analyzing other [ __ ]
43:18 So rather than generating stuff from
43:20 scratch, analyzing other stuff. So
43:23 which one do we want to start with?
43:25 Let's start with
43:30 that one's kind of cool.
43:33 That one's kind of cool. I think that
43:35 one
43:37 or her, huh?
43:49 Oh, all right. That's there's some story
43:52 here. Okay. So, this is the one we're
43:54 gonna upload.
43:57 Which one is it?
44:00 Slouched.
44:15 Okay.
44:19 What's
44:22 the backstory here
44:30 if you've never seen chatbt do this?
44:33 It's remarkable.
44:35 The backseat of the apocalypse. In the
44:37 waning golden light of a dying Los
44:39 Angeles afternoon, Eddie Greasefire
44:42 Nolan slumps into the backseat of his n
44:44 his battered 1963 Lincoln Continental.
44:48 His office, his confession booth, his
44:50 cage. Once a celebrated stunt driver in
44:53 Bm movie circles, Eddie's name
44:57 used to flash in grindhouse credits
44:59 right after the words all driving
45:01 performed by. But this isn't about glory
45:04 days. Eddie's got blood on his shirt,
45:06 not his. From the guy who tried to
45:08 collect what Eddie owed to the wrong
45:10 people in his lap, a weathered leather
45:13 satchel. So, so just good good story
45:16 there, right?
45:18 Okay. So, now
45:22 I wonder how I do this without
45:29 it having memory. So, I think how I'll
45:31 do it is this. I'll do a temporary chat
45:35 where we'll upload the same image
45:42 and we're going to go to GPT5 fast.
45:47 What's
45:51 the backstory here?
45:59 This looks like a scene pulled straight
46:01 out of some sweaty neon tinged pulp
46:04 novel. Like notice the difference. The
46:06 the the
46:08 GPT4 one told us a story. This one it's
46:13 a little more literal. The backstory
46:14 probably this. He was once a hot shot
46:17 stunt driver. It got that out of the
46:19 file name by the way because stunt
46:21 driver is in the file name.
46:24 All right. That's crappy. writing. All
46:26 right. I don't like that. So, we like
46:28 GPT40 better than that. Now, let's do um
46:33 new chat. Temporary chat.
46:37 GPT5 thinking mini.
46:43 There he is. What's
46:45 the backstory here?
46:53 Thinking. Thanking.
46:58 Nice picture. Here are three tidy little
47:01 origin stories for this greasy-haired
47:03 man. The stunt man on layaway. He used
47:07 to eat screeching tires for breakfast,
47:09 movie sets, exploding cars, applause.
47:12 Now the calls stopped. The stunt
47:14 coordinators moved on to younger spines.
47:16 And he's been selling a few of his
47:18 keepsakes to pay the hospital bills. The
47:21 fixer who messed that's this is
47:23 interesting. It gave us three. Pick one
47:26 mentally and I'll expand it into a Titan
47:28 scene.
47:30 Huh. All right. Fascinating. And now
47:34 let's do GPT5 thinking.
48:00 All right. What's the backstory here?
48:06 [Music]
48:11 Yeah. By the way, you should all be
48:12 playing along. Just be just be playing.
48:15 Probably from that woman in the kitchen.
48:17 Yeah, exactly.
48:21 Oh, actually that's really cool.
48:24 That gives me a really good idea. Silver
48:27 Fox.
48:29 We're gonna figure out which of these we
48:30 like better. I I think I think right now
48:32 I'm liking 40 the best. The old model
48:36 response. He's a stunt driver. Not
48:39 glamorous. Not the glamorous kind with
48:41 jawline sponsorship. The other kind that
48:44 gets paid in cash. Arnica and don't ask.
48:47 What's arnica?
48:51 Name he gives is Colt Reyes because Los
48:54 Angeles rewards aliases more than
48:56 honesty.
48:57 The turquoise sofa on wheels is 63
49:00 Impala the studio keeps off books for
49:03 Bunit crashes.
49:07 A week ago he rigged a rig complication.
49:10 The patch on the pouch is a boy scout
49:13 emblem. This is pretty good.
49:17 So, it could be either one of these.
49:21 All right. So, what we're going to do is
49:22 we're going to do we're going to do two
49:24 of these. So, I'm going to do new chat.
49:27 I'm going to do one in thinking mini
49:30 and then we're going to upload two
49:32 different images. We're going to upload
49:34 stunt driver
49:36 and we're going to upload
49:42 kitchen lady.
49:46 that one.
49:55 Two locations,
50:05 two times.
50:11 Two people
50:16 How are they connected?
50:21 All right. So, that's our prompt. So,
50:22 we're going to copy this prompt.
50:25 Boom. We'll let thinking think on that
50:28 for a minute. And then we're going to go
50:31 to chat GPT again.
50:37 And I'm going to go to chat GPT40.
50:42 And we're going to upload those same two
50:43 images.
50:56 Same prompt.
51:20 Eh, didn't like that one.
51:23 Net network connection lost. Sweet.
51:27 All right, reload.
51:30 New chat thinking. Damn it.
51:34 I do not like technology. Technology is
51:37 not your friend.
51:40 class.
51:44 [Music]
51:54 Uh,
51:56 I think we're going to do a different
51:57 prompt even. Let me go back to
52:01 Oh, wait. You're not seeing what I'm
52:02 seeing. Let me go back to this one. Um,
52:08 new prompt.
52:12 These
52:15 images
52:18 were taken at the exact same moment. No,
52:24 let's see. Wait, new prompt. Um, let's
52:27 say, um,
52:30 tell me their story.
52:33 You know what? We're going to do a new
52:35 new chat
52:38 because it's it's it's not going to
52:40 forget the old story.
52:49 Tell me their story
52:54 and what's
52:56 happening now.
52:59 All right.
53:00 I kind of like that prompt.
53:40 H interesting
53:44 thought for 48 seconds. Oh, wait. Let me
53:47 share this one. Share this tab instead.
53:49 Okay.
53:52 Their family split by decades and coast.
53:55 Queens 1950s.
53:58 The woman's the kitchen switchboard.
54:00 Taking calls and skimming notes for a
54:02 small time racket. Stashing cash and
54:04 papers in the beatup canvas pouch. Los
54:07 Angeles 1970s. The guy's her son. A
54:10 stunt driver moonlighting as a bagman.
54:13 Clutching the very same pouch on a job.
54:17 Does she have a pouch there? No, she
54:20 doesn't. I just kind of made that up.
54:21 That's cool.
54:30 Right.
54:34 This is a great
54:36 American
54:40 story to be told. I want
54:46 I want
54:48 to tell it
54:50 in song.
54:54 Write
54:56 the style of the song in detail
55:01 and
55:03 write the lyrics.
55:08 That's kind of cool. I like this one.
55:11 Erica, what's happening? We're shaking.
55:14 We're shaking. Was shaking.
55:18 All right. Writing song lyrics there.
55:20 Let's go back to this one. Share this
55:23 tab instead.
55:24 Static on the line. Um,
55:31 what if she is his mother
55:36 and
55:38 these
55:40 images are when they were
55:45 the same age? What
55:50 is their story?
56:07 In the world where time bleeds, echoes
56:09 loop, and memories are magnetic, their
56:12 timelines have crossed in the frayed
56:14 wires of a broken switchboard.
56:19 She's a brilliant but emotionally
56:20 unstable telephone operator turned DIY
56:23 cryptographer. She believes she
56:25 discovered a pattern in the static coded
56:28 pulses buried in long-distance hum of
56:31 trans transcontinental calls. She became
56:33 obsessed. She becomes obsessed. Records
56:35 everything. This is good. She begins
56:39 deciphering strange coordinates names,
56:41 one of which is her own son's birthday
56:43 30 years before his birth.
56:47 She's laughed out of every institution.
56:50 Nick Von, a tormented stunt driver. What
56:52 he doesn't know is
56:55 the current moment. They're the same age
56:57 now visually and spiritually. Time is
56:59 folded.
57:01 He's sitting in a car that no longer
57:02 exists. She's leaning over a counter
57:04 that was condemned in 68.
57:07 But the signal is strong today.
57:12 Huh?
57:15 Can you turn that into a song
57:22 lyrics
57:25 that is haunting,
57:29 memorable,
57:32 and
57:35 easy to understand.
57:45 All right, let's go back to this one.
57:47 Share this instead. Song in detail.
57:49 Genre heart noir Americana. Think
57:52 Springsteen grit meets the Nighthawks
57:54 Diner at 2:00 a.m. with a touch of surf
57:57 tremolo.
58:02 Chord progressions.
58:05 Huh.
58:11 Instrumentation, structure, lyrics,
58:16 pink wall kitchen, sun on for micica,
58:18 rotary, rotary ringing like a church
58:21 bell lie.
58:24 Recipe cards turned ledgers and
58:26 nicknames. Tomato stains where the
58:29 numbers dry. She writes fast shoulder on
58:32 the handset. Voice like steam from a
58:35 midnight rag. This is actually good.
58:37 Baby hush. I'm only taking messages.
58:41 Tucks the world in a canvas bag. She
58:44 says, "Count your blessings. Never count
58:45 the cash. Keep your zipper facing
58:49 in when the sirens flash.
58:53 Two coast, two clocks, same stitches
58:57 in the same old bag from from the Line
59:01 rattle to Lerea drag.
59:05 If the past tried shotgun, let the
59:08 future.
59:09 Okay, this is this is bonkers. Let's
59:12 let's copy all of this
59:16 and we're going to go over to
59:21 producer.
59:28 All right, here we go. We're going to go
59:31 to producer now.
59:34 So, this is producer.ai.
59:37 >> Hey, Kyle. Um, speaking of audio, you've
59:40 got a error message on the Tik Tok
59:43 version of the browser view.
59:47 >> Wait, the browser
59:50 has lost connection to screen share
59:52 audio.
59:57 Oh.
1:00:07 All right, we should be good now. All
1:00:08 right.
1:00:13 All right. Here's a crazy
1:00:16 Do I have to log in? Wait.
1:00:22 Wait. Stop sharing. Stop sharing.
1:00:25 Hang on a sec. Something's Something is
1:00:27 definitely weird here.
1:00:32 All right.
1:00:36 Okay. I'm logged in.
1:00:43 All right. We've got all of our stuff.
1:00:46 We've got the lyrics.
1:00:48 We've got all the in detailed
1:00:50 instructions.
1:00:51 I have no idea if producer can handle
1:00:54 this.
1:00:58 This is a killer setup. Total late night
1:01:00 Americana ephemera mood. The kind of
1:01:03 track that lives in the cracks between
1:01:05 neon and memory.
1:01:07 Producer is really good.
1:01:11 Let me break down how I'd push this to
1:01:13 record ready and then we'll set up a
1:01:15 session for tracking
1:01:19 creative approach
1:01:21 guitar motifs.
1:01:32 um
1:01:34 mix of male and female.
1:01:38 This should be a killer duet.
1:01:51 I like lowfi. idea
1:01:56 should feel
1:02:00 more like
1:02:02 memory surfing
1:02:05 than time travel.
1:02:12 Perfect. Going with lowfi memory haunted
1:02:15 duet.
1:02:25 Use the lyrics
1:02:29 I provided
1:02:31 and
1:02:34 make the full
1:02:36 duet.
1:02:43 [Laughter]
1:03:17 [Music]
1:03:20 The low trim line
1:03:22 that feels like telephone wire as
1:03:26 [Music]
1:03:35 pink wall kitchen
1:03:39 like a church bell. Mississippi cars
1:03:41 turn ledges and nicknames tomato stains
1:03:44 where the numbers dry.
1:03:46 >> She rides fast shoulder to the handset.
1:03:48 Voice like steam from a midnight rag.
1:03:51 Baby hush. I'm only taking messages.
1:03:53 Tucks the world in a canvas bag. Says
1:03:57 count your blessing.
1:03:59 Never count the cash. Keep the zipper
1:04:02 facing in when sirens flash.
1:04:06 Two coats, two cuts, same stitches in
1:04:08 the same old man. From
1:04:12 the aine ratt
1:04:17 shotgun,
1:04:20 [Music]
1:04:22 two clock stitches in the same bag.
1:04:26 >> Huh? There's something here
1:04:29 that's [ __ ] remarkable, right? Like
1:04:33 you go you look at these two images.
1:04:37 Oops, that's broken.
1:04:42 Um,
1:04:50 take that image.
1:04:56 Which one is it?
1:05:04 Oh, am I Oh, I'm not sure. Oh, yeah. I'm
1:05:07 not sharing the same thing. Well, those
1:05:09 two images, I mean, they they sound like
1:05:10 that song. Oh, wait. Show this tab
1:05:12 instead. These these two images sound
1:05:15 like that song.
1:05:17 Like, it's not great. It's it's it's
1:05:19 messing up the uh pronunciation on a
1:05:21 bunch of words.
1:05:25 >> Backseat like palm tree shadows on a
1:05:28 busted dash. Driving knuckles tapping
1:05:31 movie time. Waiting for the handoff.
1:05:33 Take me cash ring box like thumbs a
1:05:36 zipper smells like clothes in a
1:05:38 laundromat flat. He hears a hum in the
1:05:41 radiator.
1:05:43 Every mile is mother's dresses
1:05:46 hands are too.
1:05:48 Don't look.
1:05:51 Keep your name off paper and your heart
1:05:53 off track.
1:05:55 Two cos
1:06:00 >> that line. Two coast, two clocks, same
1:06:02 stitches in the same old bag. That's a
1:06:04 good line.
1:06:06 Two coast, two clocks, same stitches in
1:06:08 the same old bag. So good. From queens
1:06:13 confessions to a sunset shame
1:06:16 [Music]
1:06:24 stitches in the same old
1:06:29 deal crooked by the tower records
1:06:32 laughing bills like birds take flights
1:06:35 that Chevy through a citrus seat choose
1:06:37 a break instead of buys a mail near A
1:06:41 donut counter seals a letter with a
1:06:43 ready tag. I'm done being someone's
1:06:48 sending home.
1:06:50 I sold those scenes from your father's
1:06:53 jacket. Said a prayer in every
1:06:57 way it keeps back. Put down the weight
1:07:02 that never
1:07:04 running fing.
1:07:10 Let the rest our own free.
1:07:14 Two cos, same stitches in the same old
1:07:17 bag
1:07:19 from
1:07:21 >> huh. It's kind of a mess, but it's kind
1:07:23 of there's something sweet about it. Um,
1:07:28 let's see.
1:07:32 I'm gonna say it's kind of
1:07:36 a mess.
1:07:38 There are a lot of lines in the lyrics
1:07:44 that don't belong there.
1:07:47 I love the chorus
1:07:54 lyrics
1:07:57 uh and the vibe
1:08:00 of the song.
1:08:03 Clean up the lyrics.
1:08:09 for me.
1:08:11 [Music]
1:08:18 I like the bright tonal quality
1:08:21 they gave her voice. Kind of Allison
1:08:23 Krauss. Yeah. Very Allison Krauss. Like
1:08:25 this is this is very much in the in the
1:08:30 in the neighborhood of the images. It's
1:08:34 it's really good.
1:08:42 I gotta I gotta tell you like there
1:08:43 there's something
1:08:51 there's something really fulfilling
1:08:53 about
1:09:02 it's a couple of things. There's
1:09:04 something fulfilling about following and
1:09:07 leading,
1:09:09 right? So like I started out, I gave it
1:09:11 sort of an idea, then it came up with 10
1:09:14 ideas, then I picked one, right? Like
1:09:17 like I'm kind of letting it come up with
1:09:20 ideas, then I'm curating those ideas,
1:09:22 then I'm giving it more direction, then
1:09:24 and then I'm switching tools. Then I'm
1:09:27 coming back and let the images I made
1:09:30 over there inform the words over here.
1:09:32 Then we're going to the song.
1:09:34 There's something about the creative
1:09:43 um improvisation.
1:09:45 That's what it is. This is That's what
1:09:47 this is. This is creative improv.
1:09:50 This is cool.
1:09:55 Okay.
1:09:56 Create cleared clean lyric demo.
1:10:01 All right. So, so they cleaned up the
1:10:03 lyrics.
1:10:05 They kept our chorus as is. We still
1:10:07 have our two two cars, two clocks, same
1:10:11 stitches in the same old bag or whatever
1:10:13 that is.
1:10:16 It's a total discourse. It is creative
1:10:19 flow. It's It's creative improv. It It
1:10:22 really is creative improv. This is like
1:10:24 This is like Yes. Anding
1:10:28 like I'm like here's an idea. chat. GBT
1:10:30 is like, "Here's 10." I'm like, "I like
1:10:31 this one." Right.
1:10:34 [Music]
1:10:43 Pink wall kitchen,
1:10:46 sun on Farmica,
1:10:49 rotary ringing with a promise in a side.
1:10:55 Recipe cards turn ledgers. tomato stains
1:11:00 dry. She tucks a world in a canvas bag
1:11:04 and tries. She says, "Count your
1:11:08 blessings,
1:11:10 not the cash.
1:11:11 [Music]
1:11:13 Keep the zipper in when the sirens
1:11:17 flash.
1:11:21 Two cos, two clocks, same stitches in
1:11:23 the same old bag.
1:11:26 From the airline rattle
1:11:29 to the labyr.
1:11:37 [Music]
1:11:45 >> Um,
1:11:46 okay. So,
1:11:50 let's see. Go back here.
1:11:52 We lost the duet
1:11:55 and
1:11:59 and the haunting
1:12:03 quality
1:12:05 of
1:12:08 the two of them
1:12:15 at the same stage in their life.
1:12:20 decades
1:12:22 apart.
1:12:24 I don't know how much further I'm going
1:12:25 to go in this,
1:12:30 but like I could imagine doing this for
1:12:32 hours. Like there's something about this
1:12:34 that is
1:12:43 you know it's funny the
1:12:49 what generative AI has been trained on
1:12:52 is the output of like all the people
1:12:56 that have come before us. you know, like
1:12:58 not all of them, but like all of them
1:13:00 that got their crap on the internet
1:13:03 basically, right?
1:13:06 And so in a very real sense, what I'm
1:13:08 doing right here
1:13:11 is I'm collaborating with like ghosts of
1:13:19 people and creatives and
1:13:26 but they're but they're alive again in
1:13:28 this weird kind of way. It's just
1:13:30 bizarre. Just bizarre.
1:13:35 Um,
1:13:38 give me the full production.
1:13:43 I want to win a Grammy
1:13:47 with this one.
1:13:54 Love that energy. Let's build this for
1:13:55 Grammy level impact. Focusing on a true
1:13:58 duet, emotional resonance, and cinematic
1:14:01 pallet. Here's the full production.
1:14:03 production vision blah blah blah blah
1:14:04 blah. Now it's gonna ask me if I really
1:14:06 want to do it.
1:14:11 Generate
1:14:15 make the song.
1:14:19 All right.
1:14:27 They changed the melody. It's more
1:14:29 twangy now.
1:14:31 Yeah.
1:14:33 This is a good lesson video. Yeah. I
1:14:36 don't I mean I I hope what what you all
1:14:40 understand is I have no idea what I'm
1:14:43 doing here, but that that
1:14:46 I kind of I kind of had an idea that it
1:14:50 might be fun to navigate like this.
1:14:54 That's it.
1:14:55 [Applause]
1:15:01 Pink wall kitchen sun on form of corro
1:15:05 ring with promise and sigh.
1:15:10 Recipe cards are ledges worn edges
1:15:13 tomato stains with a number dry
1:15:18 the world in a canvas bag.
1:15:26 >> All right, you [ __ ] God damn it.
1:15:29 Okay. Um,
1:15:31 your dumb song is
1:15:36 singing all the stage directions.
1:15:42 Remove the stage
1:15:45 directions
1:15:47 from the lyrics.
1:15:54 Um,
1:15:57 also
1:16:02 spell
1:16:04 for Micah.
1:16:07 Um,
1:16:09 what's it called when you spell
1:16:10 something phonetically?
1:16:13 Phonetically,
1:16:16 since you can't pronounce it.
1:16:20 Okay.
1:16:26 Good catch. Stage directions shouldn't
1:16:28 be in the song lyrics.
1:16:33 Here's your cleaned lyrics. For Micah.
1:16:36 Okay. For Micah is now spelled as it
1:16:40 should be. Okay. Now,
1:16:45 yes. Final song
1:16:49 with new lyrics.
1:16:55 haunting
1:16:57 duet.
1:17:02 Make it sing
1:17:05 like the past,
1:17:09 present, and future
1:17:12 all at the same time. Let's see what
1:17:14 that does. Maybe we'll get in some weird
1:17:17 Cynthy kind of things.
1:17:29 This is fascinating. I've got a lot more
1:17:31 people on YouTube and and X or whatever
1:17:34 and almost no one like Tik Tok are
1:17:36 dropping those flies which it doesn't
1:17:38 surprise me because I'm just I'm not
1:17:40 really talking that much.
1:17:47 Pink wall kitchen sun on from my rotary
1:17:51 ringing with a promise in a sigh recipe
1:17:54 cards for ledges worn edges tomato
1:17:57 stains where the numbers dry tucks the
1:18:00 world in a canvas bag. She whispers
1:18:05 count your blessings not your cash
1:18:10 zipper facing in one siren flesh.
1:18:16 Two coast two clock same stitches in the
1:18:19 same old bag
1:18:22 from the line rattle to the bread drag.
1:18:29 If the past right shot gone, let the
1:18:31 future take the wheel.
1:18:35 We had
1:18:38 two coast, two clock, same stitches and
1:18:41 same old
1:18:47 >> teal vinyl backseat. Heat like lemons.
1:18:50 Palm tree shadows on a crack dash ring
1:18:54 by slide thumb on the zipper. her song
1:18:57 in the steam and radiators crash. Every
1:19:00 mile is a mother's crash.
1:19:03 >> He mutters hands at 10 and two. Don't
1:19:06 look in the sad.
1:19:10 Name off paper, heart off track.
1:19:16 Two coast, two clock, same stitches in
1:19:19 the same old bed.
1:19:22 From queens confessions to a sunset shed
1:19:29 flux of pay phone drop a quarter for the
1:19:32 life
1:19:35 you
1:19:38 two coach two clock same stitches in the
1:19:41 same old bag.
1:19:44 Sold those seams from your father's
1:19:46 jack. said a prayer and in every attack
1:19:51 from how you packed it.
1:19:53 >> Yeah.
1:19:54 >> But the road keeps winding back.
1:19:57 >> Yeah. I mean, yeah. Producer Brandon was
1:20:00 saying it's like like the the the
1:20:02 difference between good and bad with
1:20:05 this AI stuff is crazy.
1:20:09 Um, let's listen to the very first
1:20:11 version that we had because that was
1:20:12 pretty good.
1:20:19 trim line. It feels like telephone
1:20:23 wires. I'm over asphalt.
1:20:25 [Music]
1:20:32 [Music]
1:20:35 Pink wall kitchen
1:20:38 like a church bell. Recipe cards turn
1:20:41 ledges and nicknames. Tomato stains
1:20:43 where the numbers dry.
1:20:45 >> She rides fast. Show it to the handset.
1:20:48 Voice like steam from a midnight rag.
1:20:50 Baby hush. I'm only taking messages.
1:20:53 Tucks the world in the canvas bag. Says
1:20:56 count your blessing.
1:20:58 Never count the cash. Keep the zipper
1:21:01 facing in. When the sirens flash.
1:21:06 Two co. Same stitches in the same old
1:21:09 man
1:21:11 from the airline rattled to my bra drag.
1:21:15 If the past right shot let the future
1:21:18 take the wheel we had
1:21:21 two co same like birds take flights that
1:21:25 Chevy through a citrus seat
1:21:28 two co same stitches in the same old
1:21:33 queen confessions to a sunset ship
1:21:37 [Music]
1:21:39 phone drop a quarter for the life you
1:21:43 two co same stitches in the same old.
1:21:47 >> All right, I got an idea. We're going to
1:21:48 try something here. Okay, so let's
1:21:51 download this puppy.
1:21:53 And I know it's it's time is getting of
1:21:56 the essence. We're going to be quick.
1:22:00 How do I download this? Is it upload?
1:22:03 Copy to clipboard. Use vibe remix.
1:22:06 Here we go. Download.
1:22:09 Famous last words.
1:22:11 MP3. Okay. So, here's what we're going
1:22:14 to do.
1:22:16 Let me share different.
1:22:35 So, we're going to open this in
1:22:37 QuickTime Player.
1:22:48 [Music]
1:22:52 Keep your name off paper and your heart
1:22:54 off
1:22:55 [Music]
1:23:05 Too close to clock. Same stitches.
1:23:08 [Music]
1:23:13 Teal vinyl backseat. Heat like lemons.
1:23:16 Palm tree shadows on a busted dash.
1:23:19 Don't drive knuckles tapping movie time.
1:23:21 Waiting for the handoff. Make it cash
1:23:24 ring but slipper
1:23:26 smells like clothes.
1:23:28 [Music]
1:23:32 Every mile is mother's dresses.
1:23:37 >> Okay.
1:23:40 thumbs
1:23:43 in a laundry flag
1:23:45 in the rad.
1:23:54 [Music]
1:24:00 >> Okay. So, now I'm going to split the
1:24:01 clip
1:24:03 there.
1:24:05 We're going to get rid of the beginning
1:24:06 of that.
1:24:10 Hands are too.
1:24:12 Don't look in the sand.
1:24:14 Keep your name off paper and your heart
1:24:16 off track.
1:24:19 Two cos stitches in the same old bed.
1:24:24 From queens confessions to a sunset
1:24:27 ship.
1:24:30 Drop a quarter for the
1:24:34 two.
1:24:39 >> Okay, I'm going split it there.
1:24:43 Then I'm going to get rid of end. So now
1:24:46 we just have the c the pre chorus and
1:24:48 the chorus. So now we're going to save
1:24:51 that
1:24:53 as two coasts
1:24:58 chorus.
1:25:04 All right, save it to the desktop. Now,
1:25:10 let's go let's let's leave producer
1:25:14 and let's go to Sunno.
1:25:20 We're going to go create
1:25:29 inspo. Select a playlist. No.
1:25:34 How do I
1:25:36 Anybody know how to upload stuff here?
1:25:41 Oh, plus audio. Describe your song.
1:25:43 Haunting.
1:25:48 Haunting duet
1:25:51 about a mother and a son
1:25:56 grifting
1:25:59 decades apart
1:26:01 at the same age.
1:26:04 And then we're going to go plus audio
1:26:07 upload.
1:26:11 [Music]
1:26:15 That's the full version. And we want the
1:26:17 little snippet we made.
1:26:19 Oh, and it saved it as a movie. So now
1:26:21 we got to go back over here. And we got
1:26:24 to export audio only
1:26:28 because file formats apparently [ __ ]
1:26:30 matter. We have AI now. We shouldn't
1:26:32 have to worry about file format, should
1:26:34 we? Shouldn't it just figure this [ __ ]
1:26:36 out? The answer is yes, by the way. The
1:26:40 answer is yes. It should figure this
1:26:42 [ __ ] out. Okay.
1:26:51 Okay.
1:26:54 So, we're going to upload that. Can't
1:26:56 fetch the uploaded audio. Please upload
1:26:58 a different audio file. Oh my god. Black
1:27:01 bar. Okay, fine. Um, upload audio. We'll
1:27:04 try it one more time. And then we're
1:27:06 gonna have to go to the internet to
1:27:09 change it to MP3 because M4A isn't
1:27:13 supported [ __ ] anywhere.
1:27:16 Yeah. Okay. So now watch this.
1:27:21 Unbelievable.
1:27:23 M4
1:27:25 A to MP3 web free
1:27:30 and then Oh, am I not sharing?
1:27:33 Oh, I am. Yeah. Yeah. Okay.
1:27:36 Um. Oh, not descript. No, no, no. God
1:27:39 damn it.
1:27:42 Free convert. I think this one works.
1:27:46 Choose files. M4A.
1:27:50 Convert to MP3. Convert.
1:27:54 Okay.
1:27:58 [Music]
1:28:00 Converting.
1:28:02 Download.
1:28:05 Save to device.
1:28:11 Did it save
1:28:28 and it crashed my machine?
1:28:32 Really?
1:28:52 We're having trouble streaming on
1:28:54 LinkedIn.
1:29:06 This might be a sign that it's time to
1:29:08 reboot your computer.
1:29:11 >> Yeah, this is a reboot the computer
1:29:12 thing.
1:29:16 Did that download anything?
1:29:23 Oh, it did. Okay. All right. So, let me
1:29:27 close that.
1:29:29 We go back to Sunno audio.
1:29:33 You can still see stuff, right? I'm
1:29:34 still streaming sort of.
1:29:39 Okay. All right. Upload audio.
1:29:43 Upload audio.
1:29:59 All right, and we're done for the night.
1:30:01 Fantastic. Fantastic. I hope you enjoyed
1:30:04 that.
1:30:12 Well, anyway, what I was going to do was
1:30:14 take that little snippet, use that as
1:30:17 the source to create a completely
1:30:19 different song in Suno,
1:30:22 but we're not going to do that now
1:30:23 because apparently I have to restart my
1:30:26 machine and my hair is
1:30:31 special.
1:30:33 [Laughter]
1:30:38 Unbelievable. Um, okay. So, here's the
1:30:41 lesson, important lesson. Um,
1:30:45 use all the AI you want, so long as you
1:30:48 never use a computer. If you use AI with
1:30:52 a pen and a pencil or like a Sharpie,
1:30:54 fully endorse that. Anything where a
1:30:57 computer's involved, I would avoid it. I
1:30:59 just wouldn't do it. All right. Hope
1:31:02 that helps.
1:31:04 Hasn't this been fantastic tonight? Huh?
1:31:08 Huh?
1:31:09 You're welcome. What are your
1:31:11 qualifications? Just watch this video.
1:31:14 You'll see my qualifications. They come
1:31:16 out hard. Huh?
1:31:24 My apologies for lack of a
1:31:28 coherent sentence tonight. Uh, hope you
1:31:32 had fun. Anyway, okay. Tomorrow,
1:31:36 uh, Ann Murphy and I, the AI Readiness
1:31:39 Project podcast,
1:31:42 400 PM Mountain time, 3 PM Pacific,
1:31:46 6 PM Eastern, Chicago, you're on your
1:31:49 own.
1:31:51 All right.
1:31:53 If you go to uh aire readiness
1:31:56 project.com,
1:31:58 boom.
1:32:00 Tomorrow at 10:00 am Pacific time, 11:00
1:32:02 a.m. Mountain, Google is doing some
1:32:06 event. They've got Who's Who's hosting
1:32:09 it again? It's someone big, right?
1:32:15 Jimmy Fallon.
1:32:17 So, Google Google coughed up what are we
1:32:20 guessing? 50 grand, 75, 100 grand.
1:32:24 Google coughed up a hundred grand to get
1:32:26 Jimmy Fallon to go talk about their cool
1:32:28 new phone.
1:32:30 Uh producer Brandon thinks it's going to
1:32:33 be an ondevice AI
1:32:38 uh like local AI. I'm not confident, but
1:32:41 there could be cool things announced.
1:32:45 Life hacks is also tomorrow. the life
1:32:47 hacks um
1:32:51 club in the uh AI salon mastermind.
1:32:56 What time is that?
1:32:58 5:30
1:33:00 Mountain time. So 5:30 Mountain time
1:33:03 tomorrow. You you you pick that up at
1:33:06 the uh AI salon, the information about
1:33:08 it. And then finally, the the big launch
1:33:12 tonight at the AI salon was the AI
1:33:14 readiness training program. So, that is
1:33:16 now live at are readyforai.com.
1:33:21 Here, let me show you the URL because
1:33:23 it's not a
1:33:25 it's not an easy one.
1:33:29 I recommended that one.
1:33:34 All right. Um, I hope you had fun
1:33:35 tonight, even though there were tech
1:33:37 snafoos at the end there. I promise I'll
1:33:39 restart my machine tomorrow. I I only
1:33:41 restart my machine about once every
1:33:43 three weeks just to piss paid off.
1:33:48 You know, it's not logical. You You
1:33:50 should really be clearing your memory
1:33:51 caches on a regular basis. Uh yeah,
1:33:54 that's that's my problem. That's it's
1:33:56 definitely a memory cache issue that I
1:33:59 don't clear it on a regular basis. I
1:34:00 know that. I know that's one of my
1:34:02 shortcomings as a human.
1:34:07 But anyway, all right. Hope you had fun
1:34:08 tonight. I'll see you tomorrow. No love
1:34:11 for central time. We love you. Chicago
1:34:14 Ill. You Illinoisers.
1:34:16 Illinois,
1:34:18 huh?
1:34:20 Vicki, you're Illinois.
1:34:31 All right. Um, I hope you have fun
1:34:34 tonight. Bye, everybody. I'll see you
1:34:36 tomorrow.
1:34:37 [Music]