9/26/2025 - From AI Film Contests to Music Production, A Deep Dive Into Today's Creative Tools

Description

Chapters

Transcript

0:00 Let's go. Let's go. Let's go.
0:03 [Music]
0:14 [Music]
0:28 [Music]
0:33 It's time to see you now. Get that
0:36 looking m
0:38 [Music]
0:40 to see your mouth.
0:43 See that smile?
0:46 The early misty morning light. I heard
0:50 the engine turning
0:53 the old for
0:57 [Music]
1:02 you
1:05 again today.
1:07 [Applause]
1:08 You would convince me
1:11 again today.
1:15 You're leaving this hotel looking for
1:19 someone else's golden rain.
1:27 So long
1:32 [Music]
1:34 now. Don't you cry.
1:35 [Music]
1:39 So long. Susanna,
1:46 don't you cry for me.
1:48 [Music]
1:59 [Music]
2:05 [Applause]
2:11 [Music]
2:16 See the hills quiet town.
2:20 Bloody old tree sway.
2:23 Whistle of the train that goes through
2:26 the day.
2:29 People with a knife. Nothing ever seems
2:33 to change.
2:35 [Music]
2:37 Stories still unfold. Each unique and
2:40 strange.
2:44 She grew up tough.
2:47 She grew up tough. She grew up tough. I
2:50 don't know what that was. One of those
2:52 songs was AI written.
2:55 [Music]
3:00 Friday night, date night. Good night,
3:02 good people. What's happening?
3:07 [Music]
3:22 [Music]
3:39 Woohoo!
3:43 [Music]
3:51 Ah!
3:58 [Music]
4:00 Well, you just done me. You bet. I felt
4:03 it. Tried to be chill, but you sad that
4:05 I melted. Fell right through the cracks.
4:09 Now I'm trying to get back
4:12 for the cool down run out. I'll be
4:14 checking in my best. Nothing going to
4:16 stop me but design an intervention.
4:20 Reckon it's again my turn to win win win
4:25 some
4:26 learn some but I won't hesitate
4:31 no more no more it can I wait I'm your
4:40 heart
4:41 [Music]
4:53 Oh, yes. Good people. What's happening?
4:56 What's shaking? What's going down?
4:57 What's going down?
5:02 [Music]
5:09 Oh, lordy, lordy, lordy. You know, one
5:12 of the things I'm realizing
5:16 I haven't posted [ __ ] to Tik Tok in like
5:19 two weeks
5:22 and I've noticed our numbers are
5:24 dwindling and I I hate to admit that
5:28 there is a correlation,
5:30 but if people don't know you're there,
5:32 they don't know you're there.
5:34 So, think I got to get my [ __ ] together
5:37 and start posting again.
5:38 [Music]
5:53 Oh man, what should I be posting about?
5:56 That's a that's a question I have
5:59 because I I just posting about tools
6:02 just doesn't do it for me anymore. I
6:04 feel like it's I feel like the ship has
6:06 sailed on that.
6:11 Friday night date night. Who did that?
6:12 Steo. Nice pick. Nice. Look at this.
6:15 Here, let me share share this tab
6:18 instead.
6:25 [Music]
6:29 Little Scooby-Doo copyright
6:31 infringement. Fantastic.
6:35 I'm down for it.
6:40 Champ's looking good.
6:44 And then in Look what I made.
6:48 Let's see.
6:52 Edge of the World. Oh, this is a Lori
6:54 Blair song. All right. I think I'm
6:55 sharing my I think I'm sharing my audio,
6:59 but this is going to make new audio.
7:01 Share this time instead.
7:04 Inspired by
7:10 inspired by me.
7:15 Oh, you in the AI learning lab. Nice.
7:20 [Music]
7:28 Barefoot
7:34 in my head.
7:36 [Music]
7:40 I'm untouchable.
7:44 [Music]
7:56 White fall midnight crown standing high
8:00 but my soul breaks down river talking
8:04 this the moment it don't breathe in gold
8:07 burning in golden light in the skies cut
8:11 from night I'm the silence I'm the storm
8:14 mutation fun new worlds are born on the
8:17 edge of the world I rise let the beat in
8:20 my blood I don't fall no I don't
8:25 I turn fire in a water made of shame on
8:28 the edge of the world. I rise a beat in
8:31 my blood. I don't fall. No, I don't
8:35 break. I turn fire in the water. Make a
8:38 shake. Bovenian soul flow
8:43 the power. Amber clouds in a painted
8:46 sky. I'm eternal. I don't die. Grab my
8:50 doubts in a baseline beat. Sexy rhythm,
8:53 secrets I keep calls and I reply. Ain't
8:58 no limit when you learn to fly. On the
9:00 edge of the world I rise. Let the beat
9:03 in my blood. I don't fall. No, I don't
9:07 break. I turn fire and water shake. On
9:11 the edge of the world I rise. Let the
9:13 beat in my blood baptize. I don't fall.
9:16 No, I don't break. I turn fire in the
9:19 water.
9:32 >> All right, good question on Tik Tok.
9:34 Kyle, have you monetized any AI
9:37 automations? Trying to make money,
9:41 but it seems oversaturated.
9:44 Well, um,
9:48 have I have I done it with any
9:50 automations? Well, sort of. Like I've
9:53 got I've got an automation built into
9:55 Storyvine where we do creative work. So
9:56 when we do client delivery, we're using
9:58 automation, but that's not really
10:00 selling it directly. Um
10:04 and then I I have sold instead of an
10:07 automation I built a web app that um
10:11 that generates content from a video. it
10:15 gen it's basic you know content
10:16 generation but it's it's pharma specific
10:21 and so we're selling that um so I have
10:25 but the market is saturated and
10:31 how do I put this the
10:36 I think that there's a this is this is
10:38 something I've been thinking a a lot
10:39 about for my my latest book
10:43 Um
10:45 there's kind of five stages that any any
10:49 complete project has to go through if
10:52 you want it to be a robust project. And
10:55 the the first phase is ideation.
10:59 And then the fourth phase is build. And
11:03 the problem with AI right now is that
11:05 you can jump from ideiation to build in
11:09 minutes, right? You can just go, "Oh, I
11:12 have this idea. Hey, maybe I can vibe
11:14 code that into an app." And you go vibe
11:15 code it into an app or turn it into an
11:18 automation. And you're like, "Oh, that's
11:20 cool. Let me go try and sell that."
11:23 There's two steps in between ideiation
11:26 and build, which are validation and
11:29 planning.
11:31 Um,
11:33 the validation phase is is one that's
11:37 actually quite easy to do now, but it's
11:39 a it's a bit of a drag because it's
11:42 like, does anybody actually want this
11:45 thing? It might be a good idea. It might
11:47 solve a problem, but does it solve a
11:49 problem big enough that people are
11:51 willing to give up their money for it?
11:53 Um, the steps two and three are
11:57 validation and and planning. Um, step
12:00 two is validation. That's that's the one
12:04 where you have to research, right? You
12:06 have to look at customer personas. You
12:08 have to look at, are there any other
12:11 apps or automations out there like it?
12:13 Um, who's the ideal customer? Um, do
12:16 they have money to spend? Do they have
12:18 budget? What's the sales cycle? Like,
12:20 there's a whole bunch of work to do
12:22 between having an idea and putting it in
12:24 the marketplace. An idea lets you
12:27 shortcircuit that step.
12:29 Um, and so a lot of what I'm seeing in
12:32 the world right now is people going,
12:34 "Oh, I made a thing. Go look at my
12:36 thing." And everyone's kind of hawking
12:38 their things, but the things tend to be
12:40 kind of single
12:42 single point solutions for multi-point
12:45 problems, right? And so, you know, if
12:48 you've got, you know, if your problem is
12:50 I need to increase sales, there's a
12:52 whole bunch of things in that chain
12:54 where your solution might be one piece
12:56 of it, but it doesn't necessarily solve
12:58 the whole thing. And then the question
13:00 is who's buying it? Do they have money?
13:02 Do they have urgency? Um, and can you
13:04 get to them? Do you have access? Those
13:07 are the bigger issues there. And so,
13:09 again, I you know, the the the
13:11 validation is uh is is an important
13:13 piece of it. Um, but it's it's it is a
13:16 saturated market and it's going to get
13:17 more saturated. Um, there's also a
13:20 contest at the dotbros.com
13:24 35K and job opportunities for AI.
13:29 The dotbros.com.
13:31 What the [ __ ] is that? Who are they? Is
13:33 that a Is that a uh By the way, that
13:34 song we just heard was from Silverf Fox.
13:37 It's very cool. Um,
13:42 let me see.
13:44 The
13:48 Did I read that right? Brothers.com. Oh,
13:50 the Oh, the doorbros.com.
14:08 Animating the future. The Door Brothers
14:10 Awards submissions are open.
14:15 The wildest AI film awards in history.
14:21 Six days. Submit now.
14:25 The Door Brothers will personally review
14:27 everything.
14:30 All right.
14:32 That's kind of cute. All their pictures.
14:34 Oh, I'm not sharing my tab.
14:39 the Tab brothers would like to have a
14:41 word with you. Um, so if you go to the
14:44 doorbros.com and then click on
14:48 the door awards, which is down here.
14:54 Why is it the wildest? I guess we could
14:56 watch this, find out why it's the
14:57 wildest awards in
15:00 >> Have you lost your mind, brother? The
15:02 cameras are rolling.
15:05 >> Oops.
15:07 Ladies and gentlemen, welcome to the
15:11 Door Awards.
15:13 >> $35,000.
15:14 Have you lost your mind, brother? The
15:17 cameras are rolling. Welcome to the
15:20 wildest AI film competition in history.
15:23 Why the wildest, you ask? Well, that's
15:27 because these days art is only rewarded
15:31 when it's obedient.
15:34 [Music]
15:36 Here at the Door Awards, we don't
15:38 disqualify films just because they make
15:40 someone uncomfortable.
15:42 Allow me to demonstrate. Hey, Winston.
15:47 [Music]
15:52 Isn't it magnificent, Mr. Door? A fine
15:55 work of art.
15:56 >> Oh, it is quite lovely. Dear Winston,
15:59 let's talk business. Gold winner will
16:02 receive $20,000.
16:04 Silver, $10,000 and bronze, $5,000.
16:09 Not in software credits or discount
16:12 codes, just real delicious money.
16:15 The submission fee is $15, so that when
16:19 you submit it, you mean it. But don't
16:22 worry, all submission fees are split
16:24 evenly between finalists 4 to 10. Your
16:27 film must be 1 to 3 minutes long. Begin
16:31 with our provided intro video and follow
16:34 one of our three provided themes. The
16:36 first theme is door. How surprising. It
16:42 could, for example, be a creepy door in
16:44 a horror film, a sci-fi portal door to
16:48 other dimensions, a metaphorical door to
16:51 a lover's heart, or any strange
16:54 brilliant idea you might think of.
16:58 Theme number two is twins. It could be
17:02 two evil sisters, two rats on a mission,
17:06 two brains in jars, or whatever version
17:09 of twins you imagine when you close your
17:12 eyes. Theme number three, trophy. It
17:16 could, for example, be a trophy from a
17:18 beauty pageant, an intergalactic eating
17:22 competition. a trophy for being the last
17:25 human, not replaced by AI or any other
17:28 award.
17:30 We've also assembled a jury of iconic
17:32 legends who will personally watch and
17:34 rate the finalist films. And one more
17:37 thing, submit something great and we
17:40 might end up hiring you whether you win
17:42 or not. That will be all.
17:46 [Music]
17:52 >> Dear ladies and gentlemen, welcome.
17:55 >> All right. Well, there you go.
17:58 I think that video was AI generated. You
18:01 think?
18:05 Oh, that's pretty good. I like that they
18:06 give you a theme. That's pretty That's
18:08 pretty smart. Pretty smart. Pretty
18:11 pretty pretty smart.
18:28 [Music]
18:42 Uh, I went silent because I was having a
18:44 brainstorm about uh
18:47 about doors.
18:55 All right. Have you heard of Galaxy AI?
18:59 They are way cheaper.
19:02 But there's got to be a rub. Galaxy AI.
19:04 They're way cheaper than what?
19:15 The all-in-one AI platform.
19:19 Start your own AI platform.
19:26 Oh, I Yeah,
19:30 they're sort of banking on the fact that
19:31 people aren't going to use their
19:32 platform.
19:35 This is This reminds me This This
19:37 reminds me of one of those businesses.
19:38 This like a gym membership kind of
19:40 business.
19:42 Here's the thing. A platform that gives
19:45 you access to all of the platforms means
19:47 you get access to some percentage of the
19:51 capabilities of, you know, whatever is
19:53 offered in an API.
19:56 um
19:57 like using
19:59 one of the clawed models or one of the
20:02 open AI models inside their platform is
20:06 not going to be the same as you know
20:08 using chat GPT you know sitting on top
20:12 of the same
20:14 you know the same models beneath it.
20:16 It's just not. Um,
20:19 a fancier version of this is poe.com
20:24 and I think they're they they're 20
20:25 bucks a month for all you can eat or I
20:28 don't know if it's all you can eat. It's
20:29 I'm sure it's credit based and at some
20:31 point you sort of blow out of that.
20:34 Uh,
20:37 Mr. It created a memory card game from
20:40 scratch with Chat GPT and Adobe Express.
20:42 Very cool.
20:47 [Music]
20:56 Are we making a movie tonight? Maybe.
20:58 That's That's kind of what I was
20:59 thinking.
21:02 [Music]
21:11 Oh, you know what I saw today?
21:14 Um,
21:18 I saw at UD I saw a post from Udo
21:23 that
21:25 the hell was that?
21:28 That was some marketing thing. I I'll
21:30 tell you another thing to to be careful
21:32 of. Look at this site.
21:36 Um,
21:42 you've got in the lower left, new
21:44 monthly subscription popup. You've got a
21:47 chat popup here. Hey, we're here to talk
21:49 to you. You've got 1.2 million happy
21:52 users. You've got get it on the app
21:54 store. Every tool you need, just $15 a
21:57 month. We've won an award. September
21:59 flash sale. There's about six different
22:01 mark Oh, here's all the here's all the
22:03 companies that use us. There's about six
22:05 different marketing messages on this
22:07 page.
22:09 So, it's like this is the focus here is
22:13 not providing you value. The focus here
22:16 is getting
22:18 >> you can make money with
22:21 >> this is one of those sites. So, um I'm
22:25 sure it's fine, but you know, I I would
22:29 I would be skittery of sites that are so
22:32 uh so excited to tell you how awesome
22:35 they are.
22:37 Um ChatGpt got to a 100 million users um
22:41 in six weeks without any marketing
22:44 because they it changed everything,
22:48 right? This is not that.
22:51 Um, okay. Let's see. Um,
22:56 Frank's Last Day, the film. Uh, uh, uh,
23:00 I want to go to UD. I want to go check
23:01 something out on UDIO. So, apparently at
23:04 Udio there's a slider that you can slide
23:08 to make your song a certain length.
23:13 All right. Website uses cookies. Fine.
23:16 Introducing voices. Consistent voices.
23:19 Use the same voice across your songs for
23:22 cohesive things. Try our new iOS app.
23:24 Oh, that's interesting.
23:26 Create.
23:30 Use a style.
23:33 Advanced controls. Here we go. Custom
23:36 length.
23:39 2 minute 10.
23:50 clip timing.
23:55 I don't even know what that is.
24:05 Tik Tok question.
24:08 DGX Spark is a new Nvidia AI superco
24:11 computer that sits on your desk. Yeah,
24:13 they announced that like a year ago.
24:15 What's the question? Oh, have you talked
24:18 to anyone who's gotten their hands on
24:20 one? No, I haven't. Um,
24:24 I haven't heard anyone talk about it. I
24:26 don't I Are they out yet? I don't know
24:28 that they're out yet.
24:33 So, you can make a 30 second song, a 105
24:36 song. Wait. Length custom
24:39 210 is legacy.
24:43 So you could make So we could make a
24:44 minute 30 long song.
24:48 Let's make a 2 minute and 10 second
24:50 song.
24:54 And then version 1.5 Algro.
24:59 Okay.
25:01 Manual mode.
25:07 Directly prompt the model.
25:11 Add to folder.
25:15 Write your lyrics. Autogenerate.
25:18 All right.
25:21 Style reduction. What genres to avoid?
25:26 Prompt strength. Lyric strength.
25:28 Clarity.
25:30 Seed. Generation quality. We'll go high.
25:35 All right. All right.
25:39 Um,
25:50 any experience with Vibe Code? I have a
25:53 really amazing app idea, but I'm worried
25:56 about my IP. Okay,
26:00 so Diz Masta, here's the deal with IP
26:04 and your idea. Um,
26:09 we now live in an era where anyone can
26:11 instantly steal your idea, like
26:14 instantly, easily.
26:16 Um, even if you protect it.
26:21 And
26:23 here's the thing about ideas.
26:26 Especially
26:28 in the in the in the age of AI, coming
26:31 up with an idea is incredibly easy. Now,
26:35 I'm not saying your idea is not great,
26:37 not brilliant. It might be. Um, here's
26:41 the actual hard thing to do.
26:44 Execute.
26:45 Because it's not just about the idea,
26:47 it's about how does that idea get
26:50 instantiated into some sort of form that
26:53 people can interact with it and use it.
26:55 So, it sounds like you have a software
26:56 idea and then it's it's about, you know,
27:00 is it good enough that people can use
27:01 it? Does it solve a problem problem that
27:03 they want solved or does it, you know,
27:05 delight them in some way that they're
27:07 they can't, you know, that they've got
27:09 to talk about it, whatever it might be,
27:11 you know, is it does it have a viralness
27:14 or a stickiness to it? What's your
27:15 marketing strategy? How are you getting
27:17 it out in the world? like the execution
27:19 of the thing is
27:22 way harder, way harder than the idea
27:25 itself. So even if you put your idea out
27:28 there and someone steals your idea, the
27:30 chances that they're going to do it well
27:32 are incredibly low. So, you know, my my
27:36 advice right now is, you know, build
27:39 [ __ ] break [ __ ] put it in the world.
27:41 Just build it, put it out there, and
27:43 just keep iterating, keep having more
27:45 ideas. If you notice what's happening
27:48 with the Frontier model companies right
27:50 now, they're all copying each other's um
27:54 ideas, their interfaces are identical.
27:57 Their features are identical. Whenever
27:59 one of them innovates, all of the others
28:01 within two or three months have that
28:03 same feature. It's almost like there's
28:06 no innovation going on at the Frontier
28:08 Labs that are innovating the most
28:09 because they're innovating on the
28:11 models. what they're doing is they're
28:12 knocking each other off on the on the
28:14 software end. And you know, in their
28:17 case,
28:19 um it's not about customer acquisition,
28:21 it's about getting enough money to get
28:23 enough compute to keep to keep leading.
28:27 Um so that's that's a whole different
28:28 thing. Like their challenge is, you
28:30 know, gargantuan because that's not even
28:32 about customer acquisition at this
28:34 point. That's about raising billions if
28:36 not trillions of dollars. Nvidia's just
28:39 committed to investing a hundred billion
28:42 dollars into open AI. So anyway, that's
28:45 that follow-up pin. Most people would
28:47 use it. It doesn't exist yet. You mean I
28:51 you mean the IP intellectual property
28:53 doesn't exist yet? It may or may not.
28:55 Have you done deep research to see if
28:57 there's anything out there like it? Have
28:59 you done deep research to to see if this
29:02 is a problem that needs solved? If it's
29:04 a problem that people are willing to pay
29:06 for? Is there is there an acute need for
29:08 it right now? And and again, like I I
29:12 suppose you could build in stealth mode,
29:14 but like I would say at this point
29:16 building faster,
29:18 like I don't know what your goals are. I
29:20 don't know if you're a developer, if
29:21 it's just an idea, if you want to go
29:24 raise money for this thing, right?
29:27 even if you want to go raise money for
29:29 it. Like we're in a world right now that
29:31 a VC is going to is like it it used to
29:35 be the VCs would look for things like
29:37 you you know your deck and have you done
29:40 [ __ ] before and and you know is this an
29:42 interesting idea. At this point if you
29:45 don't show up with a working prototype
29:47 and probably some customers already in
29:49 the can they're going to be like well
29:51 what are you waiting for? It doesn't
29:52 cost you any money to build something.
29:55 Why don't you go build it?
29:58 Um
30:01 [Music]
30:04 yeah, Windinsurf Windsurf did this. So
30:07 So Windsurf was like a a a coding
30:11 platform like lovable or like one of
30:14 those and they they built in public and
30:16 they got acquired by who bought them?
30:18 Open AAI.
30:20 Was it OpenAI or was it Anthropic? they
30:22 got bought for like $3 billion
30:26 like within a year. So
30:30 like you know I I would hone in on your
30:32 goals and and and quite frankly like I
30:35 don't know where you are on the on the
30:36 coding side of things if you know how to
30:38 code if if you know how to set up
30:40 databases and authentication
30:43 um schemes and and ecommerce and web
30:48 servers.
30:49 um go use something like cursor or
30:52 something like that. If you don't know
30:53 any of that [ __ ] then use something
30:55 like lovable, lovable.dev where it it
30:59 kind of handles all that for you. Um
31:02 but but there's lots to learn about if
31:05 you're going to put a commercial app out
31:07 there, you got to make sure that it's
31:09 it's decently solid in terms of security
31:12 and privacy and things like that. So
31:16 anyway, I I don't mean to to be not
31:18 enthusiastic for you. Like I would say
31:20 go build it. Go build it. Don't don't
31:23 worry about
31:25 everyone's too busy with their own [ __ ]
31:27 to No one is sitting there waiting to
31:29 steal your idea is all I'm saying. And
31:31 even if they are, the chances that they
31:33 can do it as well as you can are going
31:34 to be low. So I would say build and
31:37 iterate as quickly as possible.
31:40 Um,
31:42 need AI to build it out. Yeah, you
31:44 probably do. Um, so we could go play
31:47 with we could go play with some some
31:49 vibe coding stuff in a bit. Let me let
31:52 me write a song here quick. So, where
31:55 I'm at is I'm at Ude. I'm going to say
31:59 um
32:02 write a song about
32:09 not being able to
32:15 predict
32:18 who's
32:20 behind any door
32:25 and what you can assume about them.
32:31 And then we'll go
32:35 we'll go rock
32:38 hiphop
32:42 [Music]
32:47 punk blues
32:50 folk
32:54 and we'll do Yeah, that's good. That's
32:56 good. All right. Uh is that it? All
32:58 right. Let's go create a song.
33:04 Crowsels wants to know about pulse, the
33:05 new pro feature in chat GPT. Uh, I don't
33:08 have it yet. I got rid of my my pro
33:10 subscription a while back in in lie of
33:13 uh getting access to VO3
33:17 um from uh from Google. So, I'm paying
33:20 them the big bucks right now. So, I
33:22 don't have it. It looks like it's a
33:24 personalized like news feed effectively
33:28 or some sort of parser of your calendar
33:32 and email and [ __ ] like that. I don't
33:35 know. I don't I don't know about it. I
33:37 have not paid too much attention to it
33:39 right now.
33:43 But there is there's an app out there
33:45 called um Hux or Hukes. H uxe E.
33:53 What was that?
33:59 Um Huxe is an app um that claims that uh
34:06 Pulse is a ripoff of their app. So So
34:09 there you have it. So, go check out Hux.
34:15 It's coming to Plus soon. It's Proactive
34:18 Chat GPT. Oh, I see. Oh, I see. It It
34:21 reads your calendar and your emails and
34:24 then uses chat GPT to prompt to give you
34:29 updates. Yeah, that makes sense. Okay.
34:33 [Music]
34:42 No, I wouldn't want to listen to that.
34:46 [Music]
34:52 Behind the doors, a mystery, a whisper
34:56 of the baby tread.
34:59 Faces change like shadows. Can't expect
35:02 the lost or get the but
35:06 >> how do I like V3?
35:09 Um
35:11 V3 is interesting to me.
35:15 Um it's expensive.
35:19 It's the only thing until um
35:23 one 2.5 just came out that's got
35:26 dialogue in it. So V3 is the only thing
35:29 that's got um dialogue in it, which when
35:33 it first came out, I loved it, but the
35:36 the whatever they trained it on for the
35:38 dialogue for the acting is
35:42 it's it's a very small data set. So
35:44 anything that's produced with V3 sounds
35:47 exactly the same. So it's like you can
35:49 spot it a mile away. Um so because of
35:53 that, I haven't really used it a lot,
35:55 but it's it's pretty good. But like
35:57 right after VO3 came out, Midjourney
36:00 Video came out. And I think MidJourney
36:02 Video is [ __ ] insanely good. It
36:05 doesn't have audio and it doesn't have
36:07 dialogue, but in terms of of like just
36:10 just nuanced quality of video. Um I
36:14 really like midjourney video right now.
36:16 That's the one I'm digging more than
36:17 anything else. Also, any experience with
36:20 Apple intelligence? Yeah, it sucks. Um,
36:23 I want to use it, but uh but privacy
36:25 with Apple seems unsure. No, actually I
36:28 I would disagree with that. I think
36:29 privacy with Apple, Apple is probably
36:32 the one company on the planet um that
36:36 that can make claims of keeping things
36:39 private, right? They've got all sorts of
36:41 hardware schemes in their devices. Um
36:44 they're pretty uh consistent about not
36:48 allowing, you know, government to, you
36:50 know, to get back doors and things like
36:52 that. at least publicly. Um,
36:55 everything else is giving your privacy
36:57 away. I would say Apple is good. The
36:59 problem with Apple is their AI is
37:01 [ __ ] trash. It's horrible. It's
37:03 horrible.
37:08 I like the LLM. I like that the LLM is
37:10 on device. Well, they don't really have
37:11 an LLM on device right now. Apple
37:14 doesn't. There are some apps that you
37:16 can install where you can run LLMs on
37:18 device, but I would just, you know, get
37:21 get the highest end Mac you can get or
37:23 get a PC with the highest end GPU you
37:27 can get and just, you know, run run log
37:30 language models locally there. Just get
37:33 just get as big of iron as you can. I
37:36 don't know [ __ ] about the open-source
37:39 world because it is it's impossible to
37:41 keep up with the commercial world and
37:43 the open source world's moving even f
37:45 even even faster and fragmenting even
37:47 faster
37:48 but there's lots out there that you can
37:51 play with um so yeah if you want to do
37:53 that um yeah and I and I would assume
37:57 that you could set up some sort of vibe
37:59 coding
38:00 local large language model but you're
38:02 going to listen here's the deal if
38:04 you're decent with techn technology. If
38:06 you're decent with configuring
38:08 open-source [ __ ] then you're probably
38:11 going to be decent enough with setting
38:14 up that environment to vibe code
38:15 something locally. So, all right. Um,
38:21 it lives on device, but it's too
38:23 complicated.
38:25 It goes to Apple Private Cloud. Yeah,
38:28 but look at how it Yeah, I I know it
38:30 goes to Apple's private cloud, but look
38:32 at
38:34 read the paper on how they're doing um
38:37 their trans transmitting that data. It's
38:39 it's in a it's in a hermetically sealed
38:42 transmitt thing. Um I like of of all the
38:48 companies when it comes to privacy and
38:50 protecting your [ __ ] I would trust
38:52 Apple more than anyone right now. I
38:54 wouldn't trust anyone else right now
38:56 other than the local large language
38:58 model, but I wouldn't [ __ ] use Apple.
39:02 I like I wouldn't even touch Apple right
39:04 now because their AI is trash. It's it's
39:06 it's not usable. Um,
39:09 all right. That song sucked. Let's take
39:11 this same exact prompt and go over to
39:18 share this tab instead. We're going to
39:20 go create.
39:22 We're gonna go style. Bang.
39:27 Let's see.
39:31 Here's the Here's what the song's about.
39:35 Create. All right. Let's see what
39:37 happens here.
39:44 [Music]
39:45 Suno or UDIO for music? It used to be
39:48 Udio for me. And then Sununo. Here's the
39:51 thing. UDIO basically stopped
39:55 the something happened with UDIO where
39:58 they just stopped releasing features. It
40:00 looks like they've started again
40:01 recently, but Sunno has just been this
40:04 juggernaut of feature after feature
40:06 after feature and new models and they
40:08 just launched V5. So I would say right
40:11 now Sunno is the one.
40:14 Yeah, there's producer AI. If you
40:15 haven't played with Producer AI, it used
40:17 to be called Refusion.
40:20 Um, and um, it's it's what's cool about
40:24 Producer AI is it's a chat interface.
40:27 So, you can write a song and then go,
40:29 "No, make it more jazzy. No, make it a
40:31 male singer or a female singer." And
40:32 it'll just keep iterating the song for
40:34 you. So, Producer AI is fun. It's It
40:37 makes good quality songs. But I would
40:39 just say in terms of like who's got
40:41 their [ __ ] together the most right now,
40:43 Sunno.
40:45 Wait, why are these 15-second songs?
40:52 [Music]
40:55 >> A song about not being able to predict
40:58 who's behind.
40:58 >> Oh. Oh. Oh, I see what I did. Wait.
41:04 I I put I put the instructions for
41:06 lyrics as the lyrics. So, it just
41:10 Rightful song. That's pretty funny.
41:16 Okay, let's let's let's get some actual
41:20 lyrics written here. Behind every door,
41:23 there's a stranger, a spark, a
41:24 lighthouse. No. Knock knock knock. Who's
41:26 there? I can't tell. A prince in rage or
41:29 a popper in a shell. A whisper of
41:31 kindness. Behind any door, a story
41:33 unfolds. A heart.
41:35 Let's see. Yeah, we'll select that
41:37 option. That's fine. And then let's
41:39 create this song now. All right. H
41:46 been there, done that.
41:48 Is there an XML audio file like SG SVG
41:52 for images?
41:54 Um,
41:56 I don't know. That I don't know.
42:01 I know that there's some new weird video
42:04 format that's apparently
42:07 happy with X. It's really hard to upload
42:09 video files to X right now. I don't I
42:12 don't know why and I don't care why they
42:14 should just fix it.
42:16 >> Knock knock.
42:22 A prince in rags or a popper in a shell.
42:28 A whisper of kindness.
42:31 A hurricane. Yeah.
42:34 Behind any door. Story unfolds. A heart
42:38 that's warm or a soul that's cold. You
42:41 can't assume you can't foresee the true
42:44 pies deep like roots of a tree.
42:47 [Applause]
42:51 That door might creep,
42:54 that door might shine,
42:57 but the light don't speak. It give no
43:00 sign. Could be a treasure or a land
43:03 mine. I thought I knew,
43:07 but I was wrong.
43:10 Each twist and turn
43:12 a brand new song
43:17 in a door
43:20 that's warm.
43:22 >> I kind of like that.
43:24 >> Knock knock.
43:27 I can't tell.
43:33 Knock knock. Who's there?
43:37 Knock knock. Who's there?
43:40 I can't tell
43:43 a prince in rags or a popper in a shell.
43:48 A whisper of kindness, a hurricane yell
43:54 behind any door, a story unfolds. A
43:57 heart that's warm or a soul that's cold.
44:00 You can't assume.
44:03 You can't foresee.
44:06 The truth has even
44:10 like roots of a tree.
44:15 Wo!
44:17 >> That door might creek,
44:20 >> that door might shine,
44:22 but the lock don't speak. It give no
44:25 sign. Could be a treasure or a land
44:28 mine. I thought I knew,
44:31 but I was wrong.
44:34 He's twist and turn
44:37 a brand new song.
44:42 >> Knock knock. Who's there?
44:46 I can't tell.
44:48 A prince in rags or a popper in a shell.
44:53 All right, I'm having an idea about how
44:54 to make this movie.
44:59 Let's Let's see if we can get a movie
45:01 done for the uh for the contest. We'll
45:03 spend $15 and join their [ __ ]
45:05 contest. The Door Brothers. How people
45:09 use chat GPT vibe coding making a movie.
45:12 Um, great song. The first one, right? I
45:15 think the first one.
45:19 A whisper of kindness,
45:22 a hurricane. Yeah.
45:26 Behind any door, story unfolds. A heart
45:29 that's warm or a soul that's cold. You
45:32 can't assume, you can't foresee. The
45:35 truth hides deep like roots of a tree.
45:38 [Applause]
45:43 That door might creep,
45:46 that door might shine,
45:49 but the lock don't speak. Give no sign.
45:52 Could be a treasure or a land mine. I
45:55 thought I knew,
45:58 but I was wrong.
46:01 Each twist and turn
46:04 a brand new song
46:08 behind any door. A story unfolds. A
46:11 heart that's warm or a soul that's cold.
46:14 You can't assume you can't foresee. The
46:17 truth hides deep like roots of a tree.
46:21 All right, that's the one. All right,
46:22 we're going to download that one. Behind
46:24 any door. I hate that name. So, we're
46:26 going to fix the name.
46:28 All right, let's go. Song details. Song
46:31 details. All right. Behind any door. Um,
46:38 let's wait. We don't We don't Okay,
46:41 listen. If I've said it once, I've said
46:43 it a thousand times. Don't use your
46:45 [ __ ] brain. It's 2025. You don't need
46:48 that thing anymore. All you have to do
46:50 is exist. Okay. No, that's not true.
46:53 What you have to do is be the arbiter of
46:56 taste,
46:58 right? Like, why did I choose song one
47:00 over song two? Why didn't I make 16 more
47:02 songs? I don't know. Because there was
47:05 something about that song that like
47:06 stops you in your tracks. If we're going
47:07 to if we're going to make a movie, we
47:10 can get someone to go, "Huh, that sounds
47:12 weird."
47:14 And the theme is doors and it's a song
47:16 about doors, you know? You know what I'm
47:18 saying? But
47:20 we got to come up with a better title.
47:22 And how are we going to come up with a
47:23 better title? We could use our brains,
47:26 but we don't need to do that. It's 2025.
47:30 So, what we're going to do is we're
47:31 gonna go chat
47:36 ready when you are. Hello.
47:40 We'll start with hello.
47:45 Silver Fox has got ideas. Oh, yeah.
47:47 Here's what I would encourage you to do.
47:48 Okay, I'll tell you my idea. You can
47:50 come up with any idea you want. So, I
47:52 forget what the themes were for that
47:54 contest. There was door, trophy, and
47:56 what was the second one? Anybody
47:58 remember
47:59 twins? Yeah, there you go. Door, trophy,
48:02 and twins. Um,
48:06 it's only a 1 to three minute video. So,
48:08 I would say go make a 1 to three minute
48:10 song on Sunno or UDO or producer.ai
48:16 in whatever style you want. It It
48:18 doesn't It doesn't have to be a music
48:20 video, by the way. I'm I'm just doing it
48:22 as a music video because I think it
48:23 would be kind of cool.
48:26 Um,
48:29 but you know, come up with a concept and
48:31 and just play along. So, all right. So,
48:34 how's your evening shaping up? Great.
48:37 Here are lyrics
48:40 to a song
48:42 song
48:45 about not being able to judge
48:50 a book
48:53 by its cover. or more accurately,
48:58 a person
49:00 behind a door.
49:05 Here, uh, let's see. Uh, door here are
49:08 the lyrics.
49:14 Uh, it's a strong set of lyrics, so it
49:16 works. I don't care about your opinion
49:18 of it. Possible tweaks.
49:22 A crown of thorns or a velvet lie, a
49:25 broken strengthen it. I can't predict. I
49:28 can't define. Each door I face redraws
49:32 the line. I'm going to say sure.
49:34 Rewrite. Sure. Rewrite. Rewrite the
49:36 lyrics.
49:38 Rewrite. Rewrite
49:42 to improve. Great. We'll let ChatgBT5 do
49:45 that. Tabs. Oh, tabs. Yes. Fine.
49:49 Everybody wants to What did I do?
49:52 I made a boo boo.
50:03 Okay. Knockk knockock. Who's there? I
50:05 can't define. Oh, wait. I can't change
50:07 the lyrics because they're already
50:09 written in the song. Fine. They are what
50:10 they are then. Um, okay.
50:15 I need
50:17 30
50:18 possible titles.
50:22 that use door and
50:26 clever play on words.
50:33 Let's just see what we get.
50:36 Behind the door, the door within
50:40 silent doorways through the door. The
50:41 truth at the door. The weight of a door.
50:45 No door alike.
50:47 Did was behind a door one? Oh, behind a
50:49 door was number one.
50:52 A door. A door.
50:54 A door. Nothing. Knock on a door.
50:59 Door of shadows echoes at the door.
51:01 Doors don't tell. Every door lies.
51:06 Oh, I like that. Every door lies.
51:13 Oh. Oh, no. I know what it's called.
51:17 Oh, this is good.
51:19 Behind closed doors.
51:24 Okay, it's called Behind Closed Doors.
51:26 All right, so let's go change the name.
51:35 Song details
51:39 behind closed doors.
51:46 [Music]
51:55 All right, we're going to download this
51:57 bad boy.
52:00 Now, we're going to go back to chat.
52:02 Jeep,
52:05 do you want to learn more? No. Uh,
52:09 the
52:11 song is
52:14 Behind Closed Doors.
52:19 Let's see. Um,
52:25 I want you
52:28 to
52:34 write image prompts. Wait. Image
52:40 prompt pairs
52:44 where the first
52:48 prompt is what the door looks like
52:55 from the outside.
53:00 And the second prompt
53:04 is what is shown when the door
53:11 is opened.
53:14 That should
53:18 contradict
53:20 contradict
53:23 or surprise
53:30 The viewer
53:36 who
53:37 has expectations
53:40 set
53:42 when
53:43 seeing
53:45 the closed door.
53:49 Should these all be the same style?
53:51 Should they all be
53:55 should it all be likei cinematically
53:58 consistent? Like if if we choose like a
54:00 Wes Anderson style, should it all be Wes
54:02 Anderson or should it be like cartoon
54:05 door, cinematic door, dark, you know,
54:07 fantasy door?
54:09 H I don't know. Let's let's let's see
54:11 what we get here. Brandon, do we have a
54:14 link to that contest in the salon
54:16 anywhere? Don't see it. Door brothers.
54:18 No, but we can go we can go put it
54:19 there.
54:20 Um, where did I do I have it here
54:24 anywhere?
54:26 So, I didn't know
54:29 it was doorbros.com.
54:32 Oh, you're on it, Brandon. Okay, cool.
54:34 I'm going to keep chat gpting. Okay,
54:38 I want you to write prompt pairs where
54:40 the first prompt is
54:42 the I I've I've misspelled prompt three
54:45 times in this one paragraph. That's
54:47 pretty impressive.
54:49 Um,
54:51 I want the
54:56 door
54:58 descriptions.
55:00 Wait, I I want the doors
55:03 to be iconic and familiar
55:09 and
55:10 provocative
55:13 and
55:15 inspiring.
55:18 in such
55:21 a way
55:23 that the viewer
55:28 will know who's behind the door
55:34 and then be surprised what's there.
55:39 Um, as for what's behind the doors,
55:46 it should always
55:51 wait. Always be unexpected,
55:56 sometimes funny,
56:00 sometimes charming,
56:03 sometimes
56:05 terrifying.
56:08 Finally,
56:10 I want you to use the imagery
56:16 in the original lyrics
56:22 to inspire
56:25 some of the prompt pairs.
56:29 and tell me which
56:34 lines of
56:36 uh lyrics
56:39 go with which
56:42 doors.
56:45 Um, give me 20
56:49 prompt pairs. And I have an idea for how
56:51 we're going to execute this, too, by the
56:53 way. Posted in challenges and
56:56 competitions. So, Brandon, if you could
56:58 pop up the uh AI salon banner, that
57:01 would be swell.
57:04 There you go. Go to community.thesal
57:06 salon.ai
57:08 and then in in the news section, there's
57:11 an area called competitions and uh
57:15 something like that. Competitions and
57:19 competitions.
57:22 Shainsaw knows one of the judges. Great.
57:25 Oh, challenges and competitions. Yes. I
57:27 I can't believe I couldn't keep that
57:29 other C-word in my head. Um, okay.
57:34 You're looking You're asking for 20 door
57:37 prompt pairs
57:39 where prompt one is what the closed door
57:41 looks like. Familiar, iconic, evocative,
57:43 setting expectations. Prompt two, what's
57:45 behind the door? Surprising
57:47 contradictions, funny, charming, or
57:48 terrifying.
57:50 Lyric, a prince in rags, a popper in a
57:53 shell. Closed door. a grand gilded
57:55 palace door. Behind the door, a barefoot
57:58 child sitting on the floor drawing with
57:59 crayons, surrounded by scattered toys.
58:05 Um, a weathered wooden cottage door with
58:08 flowers in bloom all around it. A raging
58:11 tornado ripping through a room, papers
58:13 flying.
58:15 That's kind of cool.
58:17 A door that might creek. A door that
58:19 might shine. A shiny modern stainless
58:21 steel door with a corporate logo. a four
58:24 a full circus performance. Okay. So,
58:27 wait. So, it's it's missing something
58:29 here. Um Okay. Um
58:33 you've just made random things
58:39 behind the doors.
58:43 I want something that is related
58:48 but
58:51 unexpected.
58:54 So if it's the palace door,
59:00 it is the princess
59:07 whatever uh hung over
59:10 and disheveled.
59:14 Um or if it is the corporate door,
59:21 it's an exec.
59:25 Um
59:31 who's I don't know ready to I don't wait
59:35 if it's the if it's the
59:38 corporate door.
59:41 It's a
59:45 uh food delivery person.
59:56 Both of my ideas suck. So, make them
1:00:02 better than that,
1:00:05 but not purely
1:00:09 random. Oh, I've got it. Random.
1:00:12 Here's the idea.
1:00:15 What's be behind
1:00:22 the door
1:00:25 should reveal a story we want to hear
1:00:32 or know more about.
1:00:36 What happened?
1:00:40 Get it? Question mark. Okay,
1:00:44 let's just stick with the hung over
1:00:46 princess. Okay. Yes, I get it now.
1:00:48 Thanks for clarifying. The story spark
1:00:49 is the point. The outside sets an
1:00:52 expectation. What's inside subverts it
1:00:54 in a way that makes us want to know what
1:00:56 happened here. Not surreal randomness,
1:00:58 but story rich contradictions. That's
1:01:00 good. That's what I said
1:01:04 without using my brain.
1:01:06 or an eight palace door. The princess in
1:01:08 sweatpants and a messy bun surrounded by
1:01:10 takeout boxes scrolling Tik Tok.
1:01:14 Yeah, that's good.
1:01:16 Lyric. Okay. Quaint cottage door. Sweet
1:01:19 grandmother in an apron screaming
1:01:21 furiously in a headset while gaming
1:01:24 online. Okay, these are good. These are
1:01:26 good. Okay, great. now.
1:01:29 Um,
1:01:33 okay. We got to we've got to do
1:01:34 something right now.
1:01:37 I'm going to go to a new chat GPT
1:01:39 window.
1:01:41 New chat jeep.
1:01:44 [Music]
1:01:46 Share this tab. Instead, I'm going to go
1:01:48 research
1:01:50 nano
1:01:53 banana
1:01:56 prompting
1:01:59 best
1:02:03 practices
1:02:06 and give me
1:02:11 a
1:02:16 structured
1:02:17 framework
1:02:20 for ideal
1:02:24 nano banana
1:02:29 outputs. All right.
1:02:44 Let it die. Got it.
1:02:49 All right. What is nano banana?
1:02:53 Pick the mode up front. Generate.
1:02:58 Oh, supply the right inputs.
1:03:21 Okay, I want a recipe for two related
1:03:28 prompts.
1:03:30 The first is
1:03:34 um
1:03:51 a closed door.
1:03:54 And the second
1:03:58 is the same scene with that
1:04:03 door open revealing
1:04:06 a scene
1:04:08 or person inside.
1:04:13 Give me the generic
1:04:17 framework
1:04:24 for all the elements
1:04:27 um
1:04:29 including aspect ratio
1:04:35 style
1:04:42 details,
1:04:44 etc. And make sure
1:04:48 that's Let's just Let's see what it
1:04:50 comes up with.
1:04:57 Michael knee, anyone playing with Sununo
1:04:59 Studio yet? I don't think I have it. Do
1:05:01 I?
1:05:04 Yes. Holy [ __ ] We have studio. We have
1:05:08 studio. We should go play with studio.
1:05:10 But let's I I want to I want to crack
1:05:12 this
1:05:14 global settings for both both prompts.
1:05:24 Okay. So, let me grab this.
1:05:31 Okay. I'm going to go back to chat GPT
1:05:33 with the
1:05:36 where it's got all of our prompts. I'm
1:05:37 going to say okay. Okay, here are
1:05:44 um
1:05:48 instructions
1:05:52 for
1:05:54 good prompts
1:05:58 for nano
1:06:02 banana
1:06:04 images.
1:06:10 Rewrite
1:06:11 all the prompts
1:06:16 to match these frameworks.
1:06:20 Be creative
1:06:24 and give me good details
1:06:29 for both
1:06:32 elements of a door.
1:06:36 Um,
1:06:43 I'm trying to think if we want this to
1:06:45 be like photorealistic cinematic.
1:07:00 Oh, I know what we'll do.
1:07:05 Oh, I actually do. I I have another idea
1:07:08 even. Okay, I know what we're going to
1:07:10 do. Let's just go. We're going to go.
1:07:14 All right. Excellent. I've got a clear
1:07:16 framework.
1:07:18 The palace door. Closed door. Generate
1:07:21 16 by9.
1:07:23 Okay. Massive.
1:07:53 Show this tab instead.
1:07:59 I'm going to go aspect
1:08:03 ratio
1:08:06 equals
1:08:08 16
1:08:11 colon 9.
1:08:17 Let's see if this looks like [ __ ] or
1:08:19 not.
1:08:29 Crafting the tool call.
1:08:38 The [ __ ] is that?
1:08:51 Why is it only showing part of the door?
1:08:53 Let's go. You know,
1:08:57 let's go to midjourney.
1:09:01 We're going to go to
1:09:03 16 by9.
1:09:33 Why is this taking so long? Oh, there we
1:09:35 go.
1:09:43 Now, let's go.
1:09:47 Let's go to explore.
1:09:49 Let's go cinematic. So I'm in I'm in
1:09:52 styles. Cinema.
1:09:55 Cinematic.
1:10:01 So now I'm in styles.
1:10:14 We try that style.
1:10:21 I like this one.
1:10:35 Oh, this one's cool.
1:10:38 All right.
1:10:51 All right, we're going to go. We're
1:10:53 going to see what we built here.
1:11:19 H.
1:11:21 These are all too forced and [ __ ] up
1:11:23 and weird.
1:11:39 Um,
1:11:41 I'm running into a problem where I don't
1:11:43 know what I want
1:11:46 [Music]
1:11:50 and I don't know which tool to use.
1:11:54 And this is going to get really boring
1:11:55 really fast.
1:11:57 Well, no, it's already boring. That I
1:11:59 know.
1:12:01 You didn't check the image in Gemini. I
1:12:03 did not.
1:12:05 Oh, no. I did. I checked the image in
1:12:07 Gemini and it only gave me the bottoms
1:12:09 of the doors.
1:12:10 That was This was my first dilemma.
1:12:14 My first dilemma was that's that's not
1:12:17 the that's not the scene I need.
1:12:21 Um
1:12:28 because what I was thinking was if we
1:12:29 went if we got a good result in
1:12:32 midjourney
1:12:36 let's do
1:12:43 um
1:12:46 dramatic
1:12:56 Those are all just dramatic colors.
1:12:58 Let's see. Dramatic film.
1:13:02 Ah, here we go.
1:13:06 Okay.
1:13:10 [Music]
1:13:14 Green could be cool.
1:13:18 something just weird and [ __ ] up.
1:13:23 Um,
1:13:26 we could do black and white noir, but
1:13:28 that just doesn't feel right. That could
1:13:31 be cool.
1:13:35 That could be cool. It's very surreal.
1:13:37 That could be cool.
1:13:42 [Music]
1:13:45 That could be cool.
1:13:49 Uhuh.
1:13:55 You know, it's actually fascinating.
1:13:57 This is this is one of the things about
1:13:59 about ideiating
1:14:03 is
1:14:05 I would argue that what I'm doing right
1:14:06 now is vibe ideiation.
1:14:10 Meaning I don't have a clear vision for
1:14:12 what I want. Right? We had a door theme.
1:14:17 We came up with a door song that we
1:14:18 liked. I sort of have an idea about
1:14:20 doors opening and revealing something
1:14:22 unexpected.
1:14:24 And now I'm kind of stuck because I
1:14:26 don't have a clear visual idea of what I
1:14:28 want. And because I don't have a clear
1:14:31 visual idea, now I'm just sort of
1:14:33 running in circles.
1:14:35 And that and this is one of the
1:14:37 challenges of AI. One of the downsides
1:14:39 of having a tool that can produce stuff
1:14:43 so well so quickly is if you don't have
1:14:46 a clear vision, if you don't have a
1:14:49 container to hold the idea,
1:14:52 then you're just spinning your wheels.
1:14:56 You're just spinning your wheels,
1:14:58 people. Do you understand what I'm
1:15:00 saying? Tik Tok pin. This seems like the
1:15:03 same problem I had when trying to start
1:15:05 my submission choice. Yes, exactly.
1:15:08 choice overload. Yeah. So, so there's a
1:15:11 good thing that you can do, Shane, and I
1:15:14 could certainly do it here. One of the
1:15:17 best things that you can do to free your
1:15:19 creativity is dramatically increase the
1:15:22 constraints. Now, the constraints here
1:15:24 are decently decently closed. We have a
1:15:27 we have a door. We we have a picture of
1:15:30 a door closed and then the picture of
1:15:31 the door opened with something behind
1:15:33 it. That's a fairly tight constraint,
1:15:36 but I don't have any constraints on the
1:15:37 look and feel right now. I don't have
1:15:38 any ideas for it, but but a way to get
1:15:42 around that choice overload is just give
1:15:44 yourself artificial constraints.
1:15:48 That's how life and manifesting is. You
1:15:50 have to know what you want. Yeah,
1:15:51 exactly. That's absolutely right,
1:15:53 Memorable Jay. But but here's here's the
1:15:56 thing that I have discovered and I like
1:15:57 I discovered on on this show night after
1:15:59 night after night. Like last night, um,
1:16:02 we came up with the Weird Mary cover
1:16:04 album.
1:16:05 Um, and that was that was not something
1:16:09 I went in knowing what I wanted. That
1:16:11 was something that sort of midway
1:16:12 through the night, someone on the live
1:16:15 suggested, oh, this could be an album
1:16:17 and that all of a sudden coalesed into
1:16:19 an idea and then I knew what I wanted.
1:16:21 Then I knew what I wanted to pursue. Um,
1:16:24 so it happens a lot with AI where AI can
1:16:26 inspire an idea, but like the the the
1:16:29 point I'm at right now that I'm not
1:16:31 liking is I I ain't got [ __ ]
1:16:36 Okay, let's go look at some some more
1:16:38 pictures of the dough. Uh oh. Why did I
1:16:40 not get Oh, cuz I'm not in the right
1:16:44 place.
1:16:47 Oh, this is a cool style. These are cool
1:16:49 doors.
1:16:53 Wait, all of a sudden these are better.
1:16:56 These choices that I just selected are
1:16:58 better. Mo better.
1:17:01 But wait, let's go listen to that song.
1:17:04 Oh, you're not going to be able to hear
1:17:05 it. Yeah, I downloaded it. I can make
1:17:08 you hear it. Hang on. Hold. Hold,
1:17:09 please.
1:17:12 Desktop downloads.
1:17:19 I can't tell
1:17:25 in a shell.
1:17:27 A whisper of kindness.
1:17:30 A hurricane. Yeah.
1:17:34 Behind any door. Sto a heart that's warm
1:17:38 or a soul that's cold. You can't assume.
1:17:41 You can't foresee. The truth hides deep
1:17:44 like roots of a tree.
1:17:47 [Music]
1:17:51 That door might
1:17:55 shine,
1:17:57 but the light don't speak. Give no sign
1:18:00 could be a treasure or a land mine. I
1:18:04 thought I knew,
1:18:06 but I was wrong.
1:18:09 Each twist and turn
1:18:13 song.
1:18:16 The heart sto a heart that's warm or a
1:18:21 soul that's cold. You can't assume you
1:18:24 can't foresee the truth deep like roots
1:18:27 of a tree.
1:18:34 [Music]
1:18:40 Could be laughter, could be tears. A
1:18:43 past a hope or a trail of fears. Every
1:18:46 knock a gambler spin. Who's to say what
1:18:50 lies within?
1:18:55 [Music]
1:19:04 [Applause]
1:19:05 [Music]
1:19:08 All right. I I don't have any clear
1:19:10 ideas right now. I'm going to bail on
1:19:11 this for now.
1:19:13 Um, sometimes that's just how it goes.
1:19:19 Maybe I started with the wrong door with
1:19:20 it being that palace door.
1:19:26 [Music]
1:19:28 I don't know. All right.
1:19:31 Yeah, you need the creative time to
1:19:33 percolate. I do. So, what are we So,
1:19:36 what are people supposed to do? Oh, I
1:19:38 don't know. Um, well, so if what I've
1:19:42 done has inspired you to start something
1:19:44 where you actually do have an idea of
1:19:45 what to do, go like what I was going to
1:19:48 do here was create get a look and feel I
1:19:51 like in midjourney. Take that into
1:19:52 Gemini and say now open these doors and
1:19:55 have this person behind it. Let me let
1:19:56 me just try that quickly. Let me just
1:19:58 grab a door here
1:20:01 that I that I kind of like. Like this
1:20:04 one's bizarre and wild. Let me copy that
1:20:07 image. Let's go to Gemini.
1:20:10 I'm going to pop that in.
1:20:13 I'm going to go grab the second half.
1:20:25 Okay.
1:20:28 So, here's the prompt. The door is now
1:20:30 open, revealing a disheveled princess in
1:20:34 sweatpants, hair messy, surrounded by
1:20:36 takeout boxes and soda cans. I'm going
1:20:38 to say uh the the door uh let's see uh
1:20:43 zoom in closer to the door.
1:20:48 The door is now open, revealing a
1:20:50 disheveled princess. Okay, let's see
1:20:52 what this looks like.
1:20:57 And then what I was going to do is take
1:21:00 the the door as the starting frame and
1:21:03 the open door as the ending frame and go
1:21:05 into a video tool and
1:21:09 make video. Understood. Here's the
1:21:12 updated image.
1:21:16 Well, now she's in front of the door.
1:21:20 Um,
1:21:23 she should be
1:21:26 behind the door frame.
1:21:33 And how do we know she's a princess?
1:21:39 Haha. I asked Spotify, "What music is
1:21:42 this?" And it couldn't figure it out.
1:21:44 Then I realized it must be your music.
1:21:46 Yeah, that that was Sununo. That was the
1:21:49 door song. Oh, here's the So, here's the
1:21:52 thing that drives me crazy about Gemini.
1:21:54 It just gives you the same image back
1:21:55 sometimes. Doesn't actually fix
1:21:57 anything.
1:22:04 [Music]
1:22:09 You have
1:22:11 moved the princess.
1:22:17 in front of the door.
1:22:22 She should
1:22:25 be inside.
1:22:33 Um, we're going to go play with uh
1:22:36 Sununo Studio right now
1:22:39 with a crown tilted on her head. Yeah,
1:22:41 something like that.
1:22:44 Then your video will come out natural.
1:22:46 What? Relate the song to your life.
1:22:51 Then your video will come out natural.
1:22:53 Yeah, that's good. Good idea. See, like
1:22:56 it Gemini just [ __ ] sucks sometimes.
1:23:02 What?
1:23:09 That's to put her in the room. What I
1:23:10 what I found about Gemini is that it it
1:23:13 just like once it once it fails, it just
1:23:17 keeps failing. So, let me copy this
1:23:19 image. We'll do a new chat.
1:23:34 The door is now um
1:23:39 zoom in on the door
1:23:48 which is now open
1:23:51 revealing a disheveled princess in
1:23:53 sweatpants hair messy. the
1:23:58 um
1:24:00 Tiara
1:24:02 tilted
1:24:04 on her head
1:24:08 surrounded by takeout boxes. Um,
1:24:12 she is inside
1:24:15 the room
1:24:18 beyond the door
1:24:22 with dramatic
1:24:25 lighting
1:24:27 on her.
1:24:31 The glow of her phone
1:24:37 reveals her sadness.
1:24:41 Reveals the sadness
1:24:45 she is not aware of yet.
1:24:50 Aware of. All right, let's see. Let's
1:24:53 see if we can get it to follow some
1:24:56 instructions.
1:24:59 Uh uh uh
1:25:06 I thought you like midjourney. Yeah, but
1:25:09 but getting midjourney like what I think
1:25:11 what I would have to do I have an idea
1:25:14 for midjourney. What I would do in
1:25:15 midjourney is I would I would go edit
1:25:17 that image
1:25:19 and tell it to open the door. We can try
1:25:21 we can try midjourney too.
1:25:29 See, that's it's not close enough on
1:25:32 her.
1:25:35 Like, it would have to be that close.
1:25:45 It's too close to the door. She wouldn't
1:25:46 be sitting there like that. That's
1:25:47 bizarre. All right. Whatever. Whatever.
1:25:50 I
1:25:55 I can't um we're going to pseudo
1:26:01 share this tab instead. Okay. So,
1:26:03 there's a new option in pseudo called
1:26:06 studio. So, we're going to go grab
1:26:11 we're going to go grab one of our songs,
1:26:13 one of our weird Mary songs.
1:26:25 [Music]
1:26:31 Uh, let's see. Library
1:26:35 playlists. Weird Mary. Oh, by the way,
1:26:40 um, last night after
1:26:44 after I left y'all and then this morning
1:26:46 I finished it, I made a
1:26:53 I made a playlist
1:26:55 that is the album of all the covers that
1:26:58 were the best covers and I named them
1:27:00 all. And so if you want to see this, let
1:27:04 me see. Let me go put it into I'll go
1:27:06 put it in irregulars right now in the
1:27:08 salon.
1:27:12 Um here's
1:27:18 the weird
1:27:20 Mary playlist.
1:27:29 Okay, there's that. So, you can go um
1:27:33 you can go check that out. They're
1:27:35 they're really good. Like, it it was
1:27:38 super cool to be able to um just sort of
1:27:42 flip through these things.
1:27:45 [Music]
1:27:56 Here's
1:28:00 [Music]
1:28:03 the jazz one.
1:28:06 [Music]
1:28:20 Here's
1:28:26 [Music]
1:28:27 my favorite ones. Well, I got a couple
1:28:30 of favorites. Seant is really good. This
1:28:32 one
1:28:35 [Music]
1:28:56 If you don't know what we're listening
1:28:57 to right now, we took a song that we
1:28:59 made a year ago on this channel that we
1:29:02 were all fans of and V5 and Sununo came
1:29:07 out two days ago and so for the past two
1:29:09 days we've been making covers of this.
1:29:11 So all these different versions of it.
1:29:13 So that was the the pirate shanty one.
1:29:15 Here's um this is the the one one of the
1:29:18 ones that came out the best was bagpipe
1:29:20 disco is what I called this one.
1:29:24 [Music]
1:29:42 Amazing.
1:29:44 [Music]
1:30:11 No, Vicki, you're right. It's Saturday.
1:30:14 It's Saturday.
1:30:16 Why did you think it was Friday? That's
1:30:18 weird.
1:30:20 Um,
1:30:21 I'm just [ __ ] with you, by the way.
1:30:24 It is Friday.
1:30:27 I just realized how mean that was.
1:30:35 [Music]
1:30:35 [Laughter]
1:30:39 Nobody brought me nachos. A that's rude.
1:30:45 Okay. So So there's our thing. So So
1:30:49 two days ago V5 drops and today Studio
1:30:53 dropped. So studio I think is like a
1:30:56 multi-track recording studio where we
1:30:58 can take any one of these songs and
1:30:59 split it and edit it.
1:31:06 [Music]
1:31:10 Ow.
1:31:13 [Music]
1:31:23 [Music]
1:31:33 [Music]
1:31:37 I think
1:31:39 Oh, you know what? one I could do.
1:31:41 There's some weird
1:31:44 in C shanty. There's some weird melodies
1:31:46 that we could maybe fix.
1:31:49 [Music]
1:32:08 Wait, Kyle, you labeled those in post,
1:32:11 right?
1:32:13 Yeah. You mean the names of these where
1:32:14 it says like C shanty, Ethereal, Broke
1:32:17 Pop, that stuff?
1:32:19 Yes.
1:32:22 I I just I just wrote the names of them
1:32:24 this morning because they all had really
1:32:25 shitty names and I like they may not be
1:32:28 right. Like West Coast Pop. I didn't
1:32:30 know what this one sounded like. Like
1:32:31 the prompt is West Coast hiphop, but the
1:32:33 song's not that at all. Actually,
1:32:39 [Music]
1:32:52 that one's got a [ __ ] up in the rhythm
1:32:54 there, doesn't it?
1:32:57 That's weird right here.
1:33:01 Right now.
1:33:06 right there
1:33:09 like it started over. Okay. So,
1:33:14 remove edit. Um, open an editor.
1:33:20 See, I don't know what studio is. What
1:33:22 one thing we're going to discover here
1:33:24 is what is studio, but let me
1:33:28 Okay, that's West Coast Pop is the name
1:33:31 of that one. So, let's go to studio.
1:33:34 Oh,
1:33:37 monthly.
1:33:42 Access to the latest and most advanced
1:33:44 V5 model. 10,000 credits.
1:33:50 Yeah, I'm going to upgrade
1:33:53 immediately with proration.
1:33:57 Continue.
1:33:59 $8. I'm spending $8 for you people.
1:34:04 You people. This is just for Vicki
1:34:06 because I feel bad that I made fun of
1:34:08 her and told her it was Saturday.
1:34:13 Vicki, it's all your fault.
1:34:16 Yep. I was excited in case it did it for
1:34:19 you. No, it did not. It did not give me
1:34:21 these really nice nice names. No, I did.
1:34:24 That was all That was all me.
1:34:27 Okay, studio. Let's go to studio. This
1:34:30 better be worth it. I just spent $8 on
1:34:33 it.
1:34:35 Sunno Studio import audio. Okay. Welcome
1:34:38 to the new era of music production. One
1:34:41 that matches your creativity, adapts
1:34:43 your sound, and pushes your ideas beyond
1:34:45 what you imagined. Drag and drop audio
1:34:48 from your workspace, song details,
1:34:51 or your local file system to start
1:34:55 creating. Oh, let's let's watch the
1:34:56 movie. Drag and drop audio from a
1:34:59 workspace, song details, or your local
1:35:01 file system to start creating.
1:35:09 [Music]
1:35:17 >> Oh,
1:35:22 I see.
1:35:24 Select an audio input, arm the track,
1:35:26 then press record to track audio
1:35:28 directly on
1:35:29 >> Yeah, this a digital digital audio
1:35:30 workstation that you can generate
1:35:32 within.
1:35:35 You know what's actually, you know, you
1:35:37 know what's kind of interesting about
1:35:39 this is
1:35:44 musicians kind of pretty much across the
1:35:49 board hate this AI generated music [ __ ]
1:35:53 because they don't have any control and
1:35:56 this product lets them have control. Um
1:36:00 they've had Timberland on as an advisor
1:36:03 um at Sunno for I don't know close to a
1:36:06 year now I think. Um and I bet this is
1:36:09 some of his his influence to say hey let
1:36:13 let producers come in here and produce
1:36:16 let them generate some [ __ ] and then
1:36:18 come in and make it their own. I I this
1:36:20 is this is really interesting. Let's
1:36:22 we'll see how intuitive it is.
1:36:24 >> Click on a song or double click on a
1:36:26 clip header to see it in the details
1:36:28 panel. Select extract stems to isolate
1:36:31 each instrument.
1:36:39 Okay. All right. We don't need to see
1:36:40 all these stupid movies. Okay. Here we
1:36:42 are. Create song. Open library.
1:36:47 Can I search in my library? I probably
1:36:49 can. What was the one that we wanted to
1:36:51 find? It was uh
1:36:58 it was uh
1:37:01 was it pirate shanty? No, it wasn't.
1:37:03 Which one was it? Anybody remember?
1:37:10 Can you create with AI, put it in a DAW,
1:37:12 and work with it? Like when I use my
1:37:15 DAW. Yeah, that's what this is. That's
1:37:17 exactly what this is.
1:37:20 So,
1:37:23 how this platform has changed in the
1:37:25 past year. Yeah, it's it's incredible.
1:37:28 Why use this if you have your own DAW?
1:37:32 Um
1:37:35 because
1:37:38 so
1:37:39 think of two different worlds um flying
1:37:43 at each other right now. You've got the
1:37:46 world of professional musicians
1:37:50 who've gotten really good at digital
1:37:52 music production.
1:37:55 You've got new AI slop generators as as
1:38:00 the musicians would call it, right? All
1:38:02 you're doing is AI slop. It can never
1:38:04 produce anything that's got any soul to
1:38:06 it, right? All that sort of [ __ ]
1:38:09 And these two worlds are coming together
1:38:12 now. the the music production crowd is
1:38:17 not coming toward AI. The music
1:38:19 production crowd is sitting there with
1:38:20 their arms crossed, pissed off that this
1:38:22 stuff is here. But what I see this
1:38:25 studio as is is as a bridge where people
1:38:28 who are producing music by just pushing,
1:38:31 you know, the button by prompting are
1:38:33 going to start to learn digital audio
1:38:35 workstations. And then the digital audio
1:38:37 workstation people are going to come in
1:38:39 and go, "Well, I might use Sunno to to,
1:38:42 you know, create a baseline or something
1:38:44 like that." They'll come in and they'll
1:38:46 they'll sort of pussyfoot their way into
1:38:48 it. And what's going to happen is all of
1:38:50 the DAWs are at some point are going to
1:38:53 have um AI generation within them. I
1:38:56 promise you. Like these two worlds are
1:38:58 going to converge. Um it's just that
1:39:00 right now everyone's treating AI like
1:39:02 this evil thing that, you know,
1:39:04 shouldn't be on the planet.
1:39:06 And it's just another tool. And it and
1:39:08 if anyone thinks that this train is
1:39:11 stopping, this train's not stopping. Um,
1:39:14 okay. But I do need to go find
1:39:19 original. Weird Mary from Cedar Hill.
1:39:24 [Music]
1:39:26 There's the original. Why? Let's see.
1:39:30 Weird Mary.
1:39:34 Uh,
1:39:36 eight songs. Why is it only found eight
1:39:38 songs? That's bizarre.
1:39:44 >> Oh, I see. Cuz I'm inside a project.
1:39:46 Maybe
1:39:48 open project.
1:39:51 No.
1:39:55 Can I do no project? No.
1:40:07 library
1:40:09 liked.
1:40:29 Hang on, people. I'm trying to figure
1:40:31 out how to.
1:40:39 No, that's not it.
1:40:44 Maybe here
1:40:47 playlists.
1:40:49 Weird Mary. Which one was it that I
1:40:52 heard the weirdness in? Was it West
1:40:55 Coast Pop? It was
1:40:59 Okay. Okay. Can I drag this to the
1:41:01 studio? No.
1:41:03 Damn it.
1:41:06 I hate shitty interface. Studio. Okay.
1:41:13 Yeah, this is a big ass app. This is
1:41:15 definitely they they've definitely made
1:41:18 this a full-on
1:41:26 How do I close this?
1:41:29 I can't.
1:41:32 What is going on?
1:41:36 Your interface sucks. Your interface
1:41:39 sucks. Your interface sucks.
1:41:43 Ah, filter. Here we go.
1:41:47 Edits from any project.
1:41:54 public
1:42:12 my workspace
1:42:15 untitled. Ah, here we go. Okay.
1:42:19 So, it was this thing,
1:42:23 my workspace, right? Is that where I
1:42:26 was?
1:42:28 Yes. Okay. Got it. Untitled.
1:42:32 Here's more Weird M. Okay.
1:42:36 And so that was called retro. Was it
1:42:39 called retro?
1:42:41 Search weird Mary.
1:42:44 Retro.
1:42:48 Weird and Mary. It wasn't retro. It was
1:42:52 pop.
1:42:58 West Coast pop. This one. Okay. So,
1:43:02 okay. Ah, got it. Drag it here.
1:43:11 Uh, can I drag it out? How do I God?
1:43:14 Okay, listen. I don't know what I'm
1:43:17 doing.
1:43:20 All you people, leave me alone.
1:43:24 [Laughter]
1:43:27 I really don't know what I'm doing. Um,
1:43:30 double click it. No. Um,
1:43:38 this is horrible.
1:43:46 Okay.
1:43:49 Add take lane.
1:43:53 All right. Here's the good news. The
1:43:56 good news is Sunno has a has put a
1:43:59 complete DAW inside of Sunno. The bad
1:44:03 news is Sunno has put a complete DAW
1:44:07 inside of Sunno. If you've ever tried to
1:44:09 use a DAW, especially if you've got
1:44:12 ADHD, it will make you [ __ ] insane
1:44:16 because digital audio workstations are
1:44:18 these radically complicated things that
1:44:20 are trying to replicate every pos
1:44:23 possible music mixing scenario in the
1:44:26 history of humanity in a digital
1:44:29 interface. And so, uh, yeah, like I
1:44:32 don't know what I'm doing.
1:44:38 Ah,
1:44:40 okay. So, I put this song in with
1:44:43 another song. So, I don't want this song
1:44:45 to be here.
1:44:48 New project. Okay. New project. Great.
1:44:52 Now, let's go back here and get
1:44:55 Weird Mary West Coast Pop and drag that.
1:45:01 Drop it here.
1:45:04 Okay. So, we've done that.
1:45:07 So, now I should be able to say make
1:45:10 stems from this.
1:45:13 Make beat split. Split. Oh, no. We don't
1:45:17 want to do that. Uh, how do I make
1:45:19 stems?
1:45:22 Add take lane. Delete. Duplicate.
1:45:25 Rename.
1:45:27 Um,
1:45:34 what's this? Lyrics,
1:45:37 music. I
1:45:44 um,
1:45:49 hey, Marge. Marge. Yeah. Hey, I was
1:45:53 wondering, you know, on the sun where
1:45:55 they they added that the studio. Yeah.
1:45:58 No, I know you're watching the wheel. I
1:46:00 know, honey. Now, how's how's Pat say,
1:46:03 Jack? I know you think he's a cutie.
1:46:05 Listen. Hey, did you ever
1:46:09 just figure it out?
1:46:12 Where do you know where's the manual?
1:46:14 There's not a manual, right? I'm just I
1:46:16 just I just What? I keep clicking. Keep
1:46:19 clicking till something happens. Yeah, I
1:46:22 I am. And that's where I'm that's why
1:46:24 I'm I I know I don't want to interrupt
1:46:26 you.
1:46:30 Good God, I wish I had a Marge.
1:46:36 Ask Quinn with video. Quinn won't know
1:46:38 yet. It's too new. Um,
1:46:41 all right. I'm not going to figure this
1:46:42 out now. It's 10 o'clock. It's time.
1:46:45 It's time. You need to It's 10 o'clock.
1:46:47 Do you know where your children are? I
1:46:49 know Vicki doesn't.
1:46:55 He doesn't even know what day it is.
1:46:59 Um
1:47:01 I listen
1:47:04 some nights we have good nights we make
1:47:05 amazing things. Other nights.
1:47:10 So So go to Sunno, spend the 16 bucks a
1:47:14 month and turn on Studio and then when
1:47:18 you figure it out, tell me how awesome
1:47:19 it is.
1:47:22 Because as of right now, I [ __ ] hate
1:47:25 it.
1:47:28 Well, well, he obviously if he would
1:47:30 have followed the instruction videos
1:47:32 there. There were there were seven
1:47:33 onboarding videos that we created that
1:47:36 you watch them in YouTube and and it
1:47:38 clearly indicates that you started step
1:47:40 one and then you proceed to step two
1:47:42 after of course achieving the
1:47:43 accomplishments of step one. and we'll
1:47:46 just have, you know, if if uh if the
1:47:48 user would just follow those
1:47:50 instructions, it wouldn't be so
1:47:52 confusing. And so that's we we created
1:47:55 the software with the intent of someone,
1:47:57 you know, following along on the the
1:47:59 instruction videos.
1:48:06 I'm not the problem.
1:48:09 Okay,
1:48:11 frigin engineers
1:48:13 with your logic.
1:48:19 This was fun. Gotta go. Florentina's
1:48:22 like, "Yeah, I'm out of here." Whatever.
1:48:23 This dude doesn't know what he's doing.
1:48:25 All right, I don't know what I'm doing.
1:48:27 Um, here's your homework for the
1:48:29 weekend. Um,
1:48:33 go join a contest. Go make a movie. Join
1:48:35 a contest. If you want to make a movie,
1:48:38 uh, I think tomorrow I'm going to I'm
1:48:41 going to put I'm going to do a little
1:48:43 write up of the Weird Mary cover album.
1:48:48 And if you want to do a video for the
1:48:50 It's that other um that other contest,
1:48:54 uh, I forget the name of it, but I'll
1:48:56 I'll put it in there. I'll I'll work it
1:48:58 out with producer Brandon. Uh, and I'll
1:49:00 I'll put it in uh in contests and things
1:49:03 like that. And maybe we can get some
1:49:04 some videos made to represent the AI
1:49:07 salon. I think that could be kind of
1:49:08 fun. And and bomb them with uh with
1:49:11 Weird Mary. So, you can do that. Um or
1:49:14 just go play. There's just there's a lot
1:49:16 to play with right now. There's a bunch
1:49:17 of new stuff out there. Um if you
1:49:20 haven't played with Google's Mixboard,
1:49:22 that's fun. Um
1:49:25 uh labs.google.commixboard.
1:49:27 google.commixboard
1:49:30 is fun. That's one worth playing with.
1:49:32 This new SUNO stuff is crazy. If you can
1:49:36 figure out how to use it,
1:49:37 congratulations.
1:49:38 You've you'll win a prize.
1:49:45 Sunno needs a clip on another track to
1:49:48 add instrument accompaniment.
1:49:52 I do not I really don't understand this.
1:49:56 Yeah, we know that, Kyle.
1:49:59 Let's see. Can I move that there? Oh,
1:50:03 wait. Oh.
1:50:05 Oh, I see what I did. I think I did
1:50:07 something wrong. Extract stems. Okay.
1:50:11 I drag and drop dragged and dropped it
1:50:13 into the wrong. So,
1:50:16 let's see. Can I get rid of this? No.
1:50:18 How do How do you delete one of these?
1:50:20 Delete. Yes. Ah, okay. So, this
1:50:30 >> No, let's get rid of that.
1:50:34 Let's go find Weird Mary.
1:50:39 Uh, this one. We're going to drop it
1:50:41 here now. Ah, okay. Now I can extract
1:50:44 the stems. All detected. All detected
1:50:47 stems. Stems, by the way, are the like
1:50:50 the individual tracks. And and what this
1:50:53 is doing is is pulling them out for each
1:50:56 individual lane.
1:51:00 And then it should also auto put the
1:51:04 song into sections, which I don't know
1:51:06 why it hasn't done that.
1:51:12 Tempo.
1:51:20 Wow. 1 2 3 4 5 6 7 8 9
1:51:27 10 11 12 12 different tracks.
1:52:10 [Music]
1:52:12 huh. Fascinating.
1:52:16 All right. So, you got all this now.
1:52:23 Recreate
1:52:26 lyrics, music styles.
1:52:41 [Music]
1:53:12 H fascinating. All right,
1:53:15 here's my prediction. I will never
1:53:17 [ __ ] use this.
1:53:22 For all for all the people that are
1:53:24 control freaks that generating AI music
1:53:27 makes them [ __ ] insane, they're going
1:53:30 to love this. For people like me that
1:53:32 this makes me want to hurt small animals
1:53:38 because it's too [ __ ] complicated, I
1:53:41 am just going to ignore this.
1:53:45 Like what I what I will do is if there's
1:53:47 something broken in the song, I'll come
1:53:49 in here. I'll split the stems. I'll fix
1:53:51 the broken thing and then I'll get out
1:53:52 as quick as I can. You're such a good
1:53:54 sport, Kyle. Yeah, I'm a good sport. All
1:53:57 right. This is how midjourney feels to
1:53:59 non-arts. Yeah. Yeah, it does. I I don't
1:54:02 disagree. Um, you know, I kind of feel
1:54:05 like
1:54:07 I feel like with all the AI tools right
1:54:10 now,
1:54:12 um,
1:54:14 I would not want to be a beginner right
1:54:16 now. Um, everything's gotten so
1:54:19 complicated
1:54:21 that that just getting a baseline
1:54:23 understanding of what's possible is
1:54:25 actually really hard right now because
1:54:26 there's just it's just too much. It's
1:54:29 it's it's it's essentially, you know,
1:54:32 every profession of every job has got
1:54:36 new rules, but we don't know what the
1:54:37 rules are yet. We don't know what the
1:54:39 tools are. We don't know what the rules
1:54:40 are. So, anyway, uh this is this is one
1:54:43 of those. So, enough enough comedy
1:54:46 jokes. Let's get the [ __ ] out of here
1:54:48 for the weekend. All right. Hope you had
1:54:50 fun tonight. Um
1:54:52 some nights we learn things, some nights
1:54:54 we learn what we don't know.
1:54:58 And we learned an important lesson. It's
1:55:01 always easier if you have an idea to
1:55:04 create something than if you don't have
1:55:05 an idea because then you just spin your
1:55:08 cycles. So, all right, Groovy, have a
1:55:11 fantastic weekend everyone and I will
1:55:13 talk to you on Monday.