8/21/2025 - AI and the Future of Creative Work: A Songwriter's Perspective

Description

Chapters

Transcript

0:01 He's so trampy.
0:05 That little champ.
0:07 [Music]
0:22 10,000 words around my head. 10 million
0:26 more in books written beneath my bed.
0:29 [Music]
0:34 I wrote old demo when searching in my
0:39 care how to hold my hands
0:46 and I know you need me
0:49 over. I'm stuck in here all paralyzed.
0:58 4 months I got myself in ruts. Too much
1:01 time spinning mirrors framed in yellow
1:04 balls.
1:07 [Music]
1:09 Ain't it like most people? I'm no
1:12 different. Like to talk on things we
1:14 don't know about.
1:17 [Music]
1:20 Ain't it like most people? I'm no
1:22 different. Life knock on things we don't
1:25 know about.
1:29 [Music]
1:52 It's not simple to say
1:56 that most days
1:58 I don't recognize me that these shoes
2:03 apron that place and its patrons
2:07 have taken more than I gave them.
2:12 >> It's not easy to know
2:15 I'm not anything
2:18 like I used to be. Oh, it's true. I was
2:22 never attention sweet center. I still
2:26 remember that girl.
2:29 She's imperfect,
2:31 but she tried.
2:34 She is good, but she lies.
2:38 She is
2:40 hard on herself.
2:42 She is broken and won't ask for help.
2:47 She is messy, but she's kind.
2:51 She's lonely.
2:53 Most of the time she's all of this mixed
2:57 up and baked a beautiful pie.
3:02 She's gone, but she used to be mine.
3:09 [Music]
3:31 Oh, good people. What is happening? What
3:34 is happening on the internet? What did
3:36 Sam Waltman launch today? What's
3:38 happening? What's going down? was
3:41 shaking. Anybody? Anybody? Anyone?
3:46 [Music]
3:48 Oh, good lord.
3:51 All right. Well, it's an intimate crowd.
3:56 It's an intimate crowd here at the lab
3:59 late on a Thursday night.
4:06 You know what's going on. A lot of
4:08 people have kids going back to school
4:11 and so they're getting drunk. I know
4:16 that's what's that's what's happening.
4:21 They're like they're finally out of the
4:23 house. Let's get drunk.
4:27 [Music]
4:37 There's no audio on Tik Tok.
4:44 People on Tik Tok can't hear me. No. Is
4:48 it just a Vicki problem?
4:55 Confirmed. No audio.
4:58 Right.
5:02 One of the things we're going to be
5:03 redoing is your park.
5:05 [Music]
5:13 Hello. Can you hear me now? Hello.
5:17 Hello. Can you hear me? I was wondering.
5:22 I was wondering. Good chaps. Good fella.
5:24 We can't hear you.
5:26 But it's okay. You're singing. That was
5:28 rude. I love it. That was a good troll.
5:32 Solid troll.
5:35 Oh, let me start over.
5:38 [Music]
5:50 10,000. All right, that's it. All right.
5:53 Fantastic. Fantastic. Fantastic. Good
5:57 people. Yeah. Lip sync is off for me.
6:03 What do you mean
6:06 the lip sync is off?
6:18 [Music]
6:26 My voice is lagging. I wonder if I've
6:27 got bad internet.
6:30 Read the pin. There's no pin.
6:33 [Music]
6:40 I'm good for Tobias. Tobias is the only
6:43 one I care about. So long as my lips
6:45 sync is is on for Tobias. To hell with
6:48 everyone else.
6:50 I think you all understand that, right,
6:53 Vicki? The I was trying to lipre.
6:58 Oh, when I when I was just making fun.
7:00 Fine, I'll log back in. All right. Yeah,
7:03 that's good. Let's put this on Dr. J.
7:05 This is all on Dr. J. Hey, Joy Party.
7:08 What's happening?
7:11 [Music]
7:28 archetypal architect. I can't hear you
7:30 on Tik Tok, but I don't have Tik Tok up.
7:34 That's That's comedy. That's That's
7:36 technology comedy right there. I got to
7:38 tell you, it's solid. That's That's some
7:40 solid solid comedy works right there.
7:43 And I guess I could do my black bar. And
7:45 I guess I could flip my camera. I could
7:47 do my camera flipping. There we go.
7:53 [Music]
7:58 You wonder why it's an intimate crowd.
8:04 [Music]
8:10 I closed the app and came back and now
8:12 the audio is fixed. Yay.
8:18 I'm learning rotoscoping this week. You
8:21 don't need rotoscoping anymore, Joy. I
8:24 mean, sure, if you want to actually
8:25 control things, but doesn't runway now
8:28 have that new thing, the runway al where
8:31 you can just upload a video clip and
8:32 just change the location,
8:35 I'm pretty sure. Or like change the main
8:38 character to be in cartooning. I don't
8:40 think you need to learn rotoscoping.
8:43 Not to crap all over your parade, Joy.
8:45 Listen, you've learned more about video
8:48 editing and production in the past 6
8:50 months than I'll ever learn in my life.
8:52 And I own a video company, but to be
8:55 fair, my video company exists because I
8:58 hated editing so much.
9:06 Yes, but I was losing quality. Ah, oh,
9:08 you wanted quality. Yeah, that's You're
9:11 not going to get that with AI. Not just
9:13 yet. You're gonna have to give that two
9:15 years. I do fear.
9:20 And quite frankly, Joy, as you know,
9:22 learning the craft of how to do it the
9:23 oldfashioned way will pay you dividends.
9:25 It'll get you jobs because you'll be
9:27 able to speak the language.
9:29 [Music]
9:53 Have I heard about Nano Banana? I have.
9:55 And I'm like, honey, that's kind of rude
9:58 to say to your husband.
10:04 It's a micro penis joke.
10:12 Uh yes, I have heard about Nano Banana,
10:14 the the code name for the Google image
10:17 editing tool.
10:19 Sorry, sorry. I just had sometimes you
10:21 just have to go seventh grade. You got
10:23 to go sophomoric humor and if you don't
10:25 do it, it's an opportunity lost. Whis
10:28 whiskey club is good. Yes.
10:38 Have I heard of Nano Banana? That was my
10:40 nickname in college.
10:47 No, it really was.
10:52 [Music]
11:21 What did I just walk into? Literally the
11:24 thing I hear is micro penis.
11:29 We were just talking about the code name
11:32 of Google's new radically powerful image
11:35 editing tool rumored to be Google.
11:39 Uh and you know according to Google size
11:41 doesn't matter.
11:43 [Music]
11:49 [Applause]
11:50 [Music]
12:06 I didn't mean to cause you any sorrow.
12:10 [Music]
12:13 Didn't mean to cause you any pain.
12:18 [Music]
12:20 Only one time to see you laughing.
12:23 [Music]
12:26 Only want to see you laughing in that
12:29 purple rain. Purple rain. Purple rain.
12:34 Purple
12:37 rain. Purple rainbow
12:42 [Music]
12:45 rain.
12:46 [Music]
12:51 Only want to see you laughing in that
12:55 purple ring.
12:58 Um, yeah. So, Nano Banana looks good. I
13:01 don't know when we're going to have
13:02 access to it. And Google didn't confirm
13:04 that it was theirs on their Pixel 10
13:06 event. So, it's still rumor at this
13:08 point, but it looks like that's what it
13:09 is. Princeton acoustic version, too.
13:12 Yeah, that version that I was just
13:13 singing, that's the Martin Ston version.
13:15 The mart the Martin Ston Purple Rain is
13:17 amazing. I stopped singing it right
13:19 there because that's the point where he
13:21 actually starts singing really good and
13:25 I don't.
13:28 So I sing I sing the setup and then I
13:32 stop.
13:36 Oh, good lord. Good people. He's got a
13:39 falsetto. He does. He's got an amazing
13:41 voice. Martin Ston.
13:45 His falsetto. There's something about
13:47 the way his face is designed. The
13:50 structure of his face. It's just like
13:51 the sound comes out of his face like
13:54 trumpets. It's amazing. Absolutely
13:57 amazing. Amazing.
14:02 Harry Nelson. Yeah, I like him too.
14:14 There's a Harry Nielson song that Nico
14:16 Casease covered. She covered it on that
14:19 Elvis Costello show.
14:22 She was on with Jesse Winchester,
14:24 Cheryl Crowe,
14:27 Jesse,
14:30 forget Jesse's last name,
14:32 and Nico Casease. It was a really good
14:34 episode, but it's it's weird. It's not A
14:36 lot of those shows are on YouTube, but
14:38 that one's missing for some reason.
14:43 Has Carl Kyle been gardening tonight? I
14:45 haven't been gardening, but I was
14:46 drinking whiskey. So, you know, I heard
14:50 on the news about the farmer's tragedy.
14:55 I did not hear about that, so I don't
14:57 know what that is. Corey Sandler, I saw
15:00 Martin Ston in the early 90s at the
15:02 Seapport Folk Festival. Wow, that's
15:05 amazing. I saw my my my I've I've seen
15:09 Martin Ston, I don't know, four or five
15:11 times now. And uh but I I saw him in a
15:14 songwriting course. He he did a two-hour
15:17 master class in a songwriting course
15:19 that I was in and oh my god, it was
15:21 amazing.
15:23 [Music]
15:31 Nice. Erica, I was on stage with Prince
15:34 while he played that song on the sound
15:37 stage in the house. Life-changing
15:38 moment. That's amazing. I actually got
15:40 to talk to Prince once. He was um he was
15:44 trying to find someone to build his
15:46 first website
15:48 and someone gave him our name and I was
15:49 the chief creative off officer at the
15:51 agency and so I talked to his manager
15:54 and then he goes, "Okay, I'm going to
15:55 put Prince on now and you know he got on
15:58 the phone." He's like, "Hello, hello,
16:00 hello." I can't even go go as low as it
16:03 was. His voice was so low
16:05 like
16:07 that was crazy.
16:09 [Music]
16:17 What's your whiskey of choice? I am a
16:19 Bllandon's man. A blanton. A blanton. A
16:24 blanton.
16:26 [Music]
16:29 Cha.
16:31 Uh, I am a Bllandon's guy. Um, tonight I
16:34 had an Old Forester
16:37 old fine whiskey 93 proof double cask.
16:42 [Music]
16:44 The theme of the whiskey night tonight
16:45 was campfire and they started a campfire
16:48 when it was 90° out. We all went in the
16:51 air conditioning.
16:55 Basil Hidden probably my favorite. Very
16:58 nice.
17:01 I love that. Oh, he's been the reason
17:02 you started your agency and left TV
17:04 news. Wow, that's so cool. He's a good
17:06 hugger. Yeah, I never got to meet him in
17:08 person, but talking to him on the phone
17:10 was a trip, man. He was so, you know,
17:13 I'm like me, like I'm peppy, and if
17:16 there was a counter energy to who I am,
17:19 it was Prince. It was weird.
17:23 I'd be like,
17:26 be just like silence.
17:27 >> I'd be like,
17:29 >> "Hey, Kyle,
17:30 >> hello.
17:32 Um,
17:34 >> oh yeah, the Tik Tok audio is working
17:36 now. That was just a confirmation. Um,
17:39 but I I I have something for you. Uh, if
17:43 you actually want to talk about AI at
17:44 the AI learning lab, I just
17:46 >> Yeah, just in case.
17:47 >> Wait, AI?
17:50 >> I talked about AI at Whiskey. I don't
17:52 need to talk about it here.
17:54 >> Okay. Well, in case you change your
17:56 mind, I just I just sent you a uh
17:59 Twitter DM from Robbie Stein. Uh he is
18:04 VP of product at Google search and um
18:08 they had some significant updates with
18:11 AI mode and it's a five-part tweet. So,
18:16 >> all right, let's go look at that. All
18:17 right, I like it.
18:19 Let's go share it. Let's go share it.
18:22 All right, we're going rogue. Well, no,
18:25 we're going we're following the
18:27 recommendation of our producer, which if
18:30 you've learned anything in radio, if
18:33 you've learned anything in radio, you do
18:35 the opposite of what the producer
18:37 suggests.
18:39 All right,
18:42 Robbie Stein, today we're announcing new
18:45 agentic and personalization features in
18:48 AI mode that make search even more
18:50 useful and tailored to you.
18:52 And a further nail in the coffin to SEO
18:55 agencies everywhere.
19:00 Agentic web brows web browsing
19:02 capabilities from Project Mariner
19:06 make it easier than ever to find and
19:08 book restaurant reservations. Oh my god.
19:10 I feel like these AI companies, it's
19:13 like their only [ __ ] use case is
19:17 you've got to book restaurant
19:19 reservations.
19:22 Anyway, sorry. Like, can we come up with
19:25 another use case?
19:27 You've got to find the
19:30 uh
19:32 dog kibble manufacturer in your
19:35 neighborhood.
19:38 I live relatively close to the Pyina
19:40 plant here in Denver. It's disgusting.
19:43 Um, tell AI mode exactly what you're
19:45 looking for and search will present a
19:47 curated list of restaurants with
19:49 available tables and time slots for you
19:52 to choose from. Okay, so that's sort of
19:54 like search,
19:57 but okay, I guess.
20:01 Get personalized dining recommendations
20:03 tailored to your unique taste. Okay,
20:04 that's different. Now, when you search
20:07 for dining related topics in AI mode,
20:09 you'll see suggestions that are more
20:10 relevant and personalized based on your
20:13 previous conversations and places you've
20:15 searched for or tapped in
20:18 search and maps
20:21 available for available for people in
20:23 the US who've opted into AI mode in
20:25 experimental. Finally, we've added new
20:27 link sharing capability in AI mode, so
20:30 it's even easier to collaborate with
20:31 friends and family. All right, that's
20:35 kind of cool.
20:39 Check out the new AI mode capabilities.
20:44 Where that's is that just at Google? Do
20:46 I do I have that enabled?
20:49 If I just go to google.com
20:53 AI mode. Yes, I do. Um,
20:57 >> it's not on the Twitter tab though.
21:01 Oh,
21:05 now it is right.
21:07 >> Yes.
21:09 >> Um,
21:14 reservations
21:20 for meat
21:22 in Denver
21:26 tomorrow
21:28 for me. And make sure you hit AI mode uh
21:32 instead of just enter because if you do
21:33 enter, it'll just do regular Google
21:34 search.
21:35 >> Oh, interesting.
21:41 >> All right, kicking off eight searches.
21:43 Steak dinner reservations. Garden Grace,
21:45 Modern Steakhouse. Reservations
21:48 recommended. Call
21:50 or visit their website. Visit their
21:52 website's not a hyperlink.
21:58 Bizarre. All right, whatever. A5
22:01 steakhouse
22:04 urban farmer. Like, that's pretty good.
22:06 That's a pretty good
22:08 search result.
22:16 Yeah, you're not going to get the um
22:18 agentic behavior yet because it's not
22:20 opted in in labs, but this gives you a
22:22 pretty good idea of what they've been
22:24 working on behind the scenes.
22:27 Yeah, pretty cool.
22:29 >> All right,
22:31 gy.
22:34 All right, what are we going to talk
22:35 about tonight, people? What is the plan?
22:38 What is the plan?
22:41 H, wait, we're still talking about micro
22:43 penises. Oh, micro first micro penis and
22:46 and now me.
22:48 I was talking about steak.
22:51 I see a pattern here. The pattern is I'm
22:54 an idiot. It's It's a pretty consistent
22:57 pattern and it's been here for a while.
23:02 Speaking of which,
23:04 I think it's only appropriate. I haven't
23:06 dawned the pink bow in a while.
23:10 All right. And I know ladies, I'm not
23:12 wearing it right. I get told that every,
23:14 you know, that's not how you're supposed
23:15 to wear it. You know, we, you know, pink
23:19 bows are very specific. They got a very
23:21 specific use case for us. I know. I
23:24 know. I know. Pinkiey's out.
23:29 [Music]
23:30 All right.
23:31 That's cutting off the blood to my
23:33 brain. I think that's not a good idea.
23:36 All right. Let's see.
23:40 Um,
23:42 what are we going to do tonight?
23:58 could go finish that song we started
24:01 two nights ago or last night whenever
24:03 that was.
24:07 Oh, I know what I You know what? I want
24:10 to play you a video.
24:20 [Music]
24:26 the honeymoon round.
24:31 Sorry if that was loud. Where's my audio
24:34 here?
24:36 [Music]
24:43 It's this one.
24:45 >> And then it's this one.
24:48 All right.
24:50 Um,
24:53 so this is a video. I did a duet to this
24:56 today. I don't ever do duets, but I did
24:58 one to this today because I think it's I
25:00 think it's an important video. How about
25:03 a connection?
25:05 How about connections in chat GPT and
25:08 Gemini? Connections like the game, like
25:10 the New York Times connections thing. I
25:12 don't know quite how that goes, but I
25:14 can ask it. Anyway, um we we talk a lot
25:18 about extensions. What's that mean?
25:24 What's extensions mean?
25:28 Oh, extensions connect to Google Drive.
25:30 I don't know how to do that, Corey. I
25:32 would do that. No, the new connections
25:34 and customizations
25:36 like make them and see if they work. The
25:38 last time the last time I tried it, it
25:40 didn't work. Like I connected them and
25:42 then I tried to like look for something.
25:44 and he goes, "I don't have access to
25:46 your Gmail." I'm like, "Well, I just
25:47 [ __ ] connected it."
25:50 So, I haven't tried it since. Um,
25:53 anyway, let me play this video and then
25:55 we'll figure out something else to do.
25:56 Um, so I talk a lot in here about Kevin
25:59 Mallister moments and and I also talk
26:02 about the five stages of AI adoption.
26:06 And if you go to the AI salon in the in
26:09 the welcome to the salon page, that's
26:11 got the Oops. Why is that not sharing?
26:14 Because
26:18 Okay, that's just weird.
26:22 I honestly don't know what's going on
26:26 with the sharing on this machine lately.
26:28 Anyway, if you go to the AI salon, um
26:32 the salon.ai or community.thesalon.ai
26:37 on the welcome page, we've got the five
26:38 stages of AI adoption. Stage one is
26:40 dismissal or denial. So that's where a
26:43 lot of people are right now is AI is
26:45 crap. AI steals. I I don't believe in
26:48 AI. I don't think it's all that good. I
26:50 tried it once. All that sort of stuff is
26:52 stage one. Stage two is awe. It's a
26:55 Kevin Mallister moment where you
26:57 realize, holy [ __ ] the world's
26:58 different. I didn't know it could do
27:00 that. And then stage three is wonder and
27:04 fear. So reverence and dread
27:06 simultaneously.
27:07 So, this video that I'm going to show
27:09 you is is stages two and three um kind
27:12 of rolled into one video. Um, and it's
27:15 it's it's pretty amazing video to
27:17 witness.
27:19 Um, okay. So,
27:22 let me move that over there. Let me move
27:24 that over there. All right. Can
27:27 everybody see? Is everybody good? All
27:30 right.
27:30 >> Hi, my name is Fernando Pomo. I am a
27:33 record producer and singer songwriter
27:36 currently living in Los Angeles.
27:37 >> I can barely hear that.
27:39 >> This might be the most important video I
27:41 ever post. Um,
27:44 my job is in jeopardy because of AI. And
27:47 I say that with
27:50 some crazy emotions going right now
27:54 because I'm equally offended as I am
27:57 impressed. But um,
27:59 >> I'm I'm equally offended as I am
28:01 impressed. AI music creation software
28:04 has now reached
28:07 >> and I am honestly blown away by what I'm
28:09 about to show you. Um, so
28:12 >> why can I not make this louder? Hang on.
28:16 Sound settings.
28:21 Output.
28:24 Output volume is maxed.
28:38 So um there is uh this program called
28:42 Sumo and it's a program that a lot of
28:44 people use to create AI music and the
28:47 usual way you do it is you type in the
28:50 lyrics and you put a prompt and it
28:52 creates a song. Well, it has a new
28:55 feature. new feature is hum a tune,
28:59 take a tap out a beat or sing a melody.
29:02 Well, I'm going to now show you uh an a
29:06 bridged version of my song called
29:07 California Moon, which is one of my fan
29:09 favorite songs. And it was okay. So, if
29:12 you didn't hear that, you you'll be able
29:13 to hear the rest of this. If you didn't
29:15 hear that, what he basically says is
29:16 he's a songwriter and a producer, and
29:18 his job is in jeopardy because of what
29:20 he just learned about SNO, the uh the
29:23 songwriting tool. And he said what he
29:25 knew Sunno could do before was you could
29:28 type in text and it would generate a
29:29 song. What he didn't know was that you
29:32 could play the beginning of a song and
29:33 it will just finish it for you. And so
29:35 that's what he's actually about to demo.
29:38 It's
29:38 >> recorded uh it was actually covered by
29:40 Pat Sansone of Wilco and Defense. One of
29:44 the most beautiful voices ever. But
29:46 here's an abridged version of the song.
29:47 We're going to do a verse and chorus.
29:51 [Music]
29:55 Instead of dancing in the dark,
30:00 let's go out tonight.
30:03 The moon is shining bright.
30:08 Let's go out tonight.
30:14 Moon is shining,
30:19 shining on you.
30:23 California moon
30:26 is shining
30:29 shining on you.
30:32 [Music]
30:33 Oh,
30:37 shining on you.
30:41 [Music]
30:45 Stop.
30:46 Now I'm going to paste in. He needs the
30:50 chips.
30:57 >> And I'm just gonna hit create.
31:01 Let's count together.
31:03 >> Um, so Ryan, to your comment on doesn't
31:06 this say more about the songs than AI?
31:08 No, he does he does a follow-up video of
31:11 of this video where he takes his most
31:14 complicated song, his most sophisticated
31:17 cordal chord progression song and puts
31:20 it in sunno. Now, you'll you'll see just
31:22 listen to what what Sunno does with this
31:25 with what he just played. One
31:26 Mississippi 2 Mississippi 3 Mississippi
31:29 4 Mississippi 5 Mississippi 6
31:33 Mississippi 7 Mississippi 8 Mississippi
31:36 9 Mississippi 10 Mississippi 10 seconds
31:40 this is what it created in 10 seconds
31:45 [Music]
31:55 [Music]
31:56 I'm not faking this. This actually just
31:58 happened.
32:01 >> I don't know if you heard that. He goes,
32:02 "I'm not faking this. This actually just
32:03 happened." Yeah. This This is the Jimmy
32:05 Buffet version.
32:09 [Music]
32:14 >> It caught all the chord progress
32:18 to G.
32:21 >> That's nice.
32:24 Now just do the B minor.
32:29 [Music]
32:30 >> There's the B minor.
32:34 >> Now, are you ready for the death punch?
32:37 >> In that 10 seconds, I love that. That's
32:39 very easy. In that same 10 seconds, it
32:41 created two versions.
32:46 >> Version two.
32:50 instead of
32:51 [Music]
32:57 >> Wow.
32:58 >> Yeah. Yeah. Erica, the the the the awe
33:01 and sadness on his face are a little
33:03 heartbreaking. I I I commented to him a
33:06 couple of times today. I I mean
33:11 this
33:14 I think that every single knowledge
33:17 worker and creative worker, anyone who
33:19 works on computers
33:22 is going to confront this moment. I like
33:24 what what I really appreciate him
33:27 sharing this is he's sharing that it
33:31 kind of in the moment where he's still
33:33 aruck by it, right? and terrified and
33:38 impressed and
33:40 but he's not shying away from it. He's
33:42 like he's like I'm gonna I'm gonna I'm
33:45 gonna face this demon, right? And and I
33:49 think I think for me this this is
33:51 actually an important video because I
33:53 think everyone's going to go through
33:54 this if you haven't gone through it
33:56 already. Anyone who's not into AI is
33:58 going to have this moment at some point.
34:02 [Music]
34:04 Why would anybody need a producer?
34:07 >> Why would anyone need a producer? He
34:08 just said
34:19 >> I'm going to let this one play out. I
34:22 don't even know what's happening,
34:24 >> but this is crazy.
34:30 [Music]
34:34 You see my eyes dying.
34:38 >> He just said, "Do you see my eyes
34:39 dying?"
34:46 [Music]
34:52 >> That's really clever.
34:56 [Music]
35:00 Oh my god.
35:03 [Music]
35:07 >> This is crazy.
35:18 I'm equally impressed
35:20 >> that I am like
35:22 >> I'm equally impressed that I am bummed
35:25 offended.
35:27 >> Offended.
35:32 I bet you do a guitar solo. It's going
35:33 to do a guitar solo.
35:40 He just said it's I bet it's going to do
35:42 a guitar solo. He goes, "It's going to
35:43 do a guitar solo." And then it does.
35:47 [Music]
35:49 >> So,
35:51 I don't know what you guys are doing.
35:54 >> Yeah. Let me let me pop over and play
35:57 the other one for a second.
36:00 >> For my final test of sumo tonight, I'm
36:02 going to uh sing into it an unfinished
36:05 song idea.
36:06 And
36:07 >> that wasn't it. Let's see. This one.
36:09 >> Today's test of Sunno's production
36:11 skills is my most complicated
36:13 composition called Feels. This song has
36:15 more chords in the intro than the entire
36:17 Ramon's catalog. So, um, let's see if I
36:20 can stump Sunno. And also, I've always
36:22 wanted to hear this song with a string
36:23 quartet and a vocal. Let's see what
36:26 happens here.
36:30 [Music]
36:44 feels
36:46 when the stars align and you're on my
36:51 mind. You're my favorite finds
36:57 [Music]
36:58 running down my spine. Knowing you are
37:03 mine tonight.
37:07 [Music]
37:14 >> His name is Fernando
37:19 Perdome. It's Fernando Perdome music. P
37:22 E R D O M
37:25 Pommo.
37:27 Fernando Perdomo music is his Tik Tok
37:30 channel.
37:32 But here here he's playing his most
37:34 complicated song.
37:39 [Music]
37:40 Source camp encountered a hostile AI
37:43 hater tonight. Lost his job. 25year-old
37:47 son can't get a job. Yep.
37:50 Yep. Yeah. This this is 2025 has been
37:54 the year that AI has gone from a novelty
37:58 to competent.
38:00 2026 is where it goes from competent to
38:04 like world class and where impact on
38:08 jobs is going to be felt like the the
38:10 first real clear stuff is is going to be
38:14 either late this year or early next
38:16 year.
38:19 mess.
38:23 [Music]
38:33 Take that, Sununo.
38:36 Oh, man. I feel like I'm bullying it.
38:39 All right. All right. Lyrics have been
38:40 pasted in.
38:41 >> So, one of the things one of the things
38:43 that's really interesting here,
38:46 the way the way
38:49 AI embedding works, the way training
38:51 these models work, you can train them on
38:54 anything
38:56 and they don't know the difference
38:58 between what we find complicated and
39:00 what we find simple. So for a large
39:05 language model to understand Chinese is
39:07 not inherently more complicated for it
39:10 to understand English or French or
39:12 Spanish where with us if it's something
39:15 like Chinese where it's way far out of
39:17 our you know out of our comfort zone and
39:20 out of our you know lineage of language
39:23 it's a it's a completely different thing
39:25 but it doesn't experience it that way.
39:27 Same with the music. He's interacting
39:29 with like with this like because he put
39:32 in more complicated chords that Sunno is
39:36 actually analyzing those chords in some
39:38 way. It's not right. It's just it's
39:40 basically creating a probability going
39:43 into what's I I assume the equivalent of
39:46 a diffusion model for images and just
39:48 generating a song out of that. So like
39:51 absolutely fascinating
39:53 >> style
39:56 male vocal
40:00 string quartet.
40:03 >> Oh yeah, he said he wanted to hear this
40:05 as a string quartet. That was that was a
40:07 cool thing.
40:08 >> No guitar.
40:12 No drones.
40:19 All right, here we go.
40:22 One Mississippi 2 Mississippi 3
40:25 Mississippi 4 Mississippi 5 Mississippi
40:29 6 Mississippi 7 Mississippi 8
40:32 Mississippi 9 Mississippi 10 Mississippi
40:35 10 seconds
40:36 >> Tom they're they're not complicated
40:38 chords. They may not be, but that it's
40:40 this is his most complicated song.
40:47 [Music]
40:52 >> Whoa.
40:53 [Music]
41:04 >> It's got a very Beatles kind of boppy
41:06 string section, doesn't it? And you're
41:15 a
41:15 >> little more bouncy than I expected.
41:17 down.
41:20 [Music]
41:23 >> It's touching every chord,
41:26 [Music]
41:31 >> man.
41:42 Love is real.
41:52 [Music]
41:54 So anyway, so I mean here's my thought
41:57 on this th this guy. So there there's a
41:59 couple of things. A bunch of things are
42:01 are blasting through my head right now.
42:02 One is I just I I think Erica your point
42:06 about
42:08 this guy being vulnerable enough to
42:10 share
42:12 like like his real time reaction to this
42:15 rather than just [ __ ] run and hide in
42:17 the closet. Um I think is I think is a
42:21 really incredible thing that he did. Um
42:25 because you know he's not going to be
42:27 the only one that goes through this. And
42:31 if you look at um
42:34 Timberland, if you know the producer
42:36 Timberland, he's got a similar video
42:38 where he he put out the video the first
42:41 time he heard Sunno and his his
42:43 associate producer put in a fragment of
42:45 a song that Timberland had created at
42:48 some point and just put it into finished
42:51 the song and
42:55 you see his reaction and he's he's
42:57 dumbfounded. He's like, "This thing
42:59 can't sing, can it?" And then it starts
43:00 singing and he's like just blown away at
43:03 the chord progressions and at the at
43:05 what it what it produced.
43:07 But Timberland now like works with Sunno
43:10 and he's using Sunno in his, you know,
43:14 in his music writing, in his music
43:16 production workflow. He's he's figured
43:18 out how to roll it in and how to
43:20 collaborate with it and treat it as a
43:22 collaborator.
43:24 Um,
43:26 this guy, I mean, he's clearly fil, fil
43:30 with songwriting, really knows music
43:32 theory, really knows his songs, really
43:33 knows his taste.
43:36 If he embraces this, he's going to be
43:39 better at Sunno than any schlub that
43:43 doesn't know anything about music,
43:45 right? So, a lot of people are just
43:46 going to be pressing the buttons on,
43:49 squirt out a song, squirt out a song,
43:50 squirt out a song. And those are going
43:52 to be generic run-of-the-mill songs. But
43:54 if you take someone with skills and
43:56 someone that's been writing as long as
43:57 this guy has, he's going to get
43:59 something different out of it. Right?
44:01 And this gets back to Rick Rick Rubin.
44:04 There's a there's a video of Rick Rubin
44:06 recently
44:07 where someone asked him about do you
44:10 think AI music can ever be music?
44:14 And he said, he said, "My my the thing I
44:18 find the most interesting about art
44:21 is the point of view of the person
44:23 behind it."
44:26 And that struck me as
44:30 you can use AI mindlessly and you can
44:33 create AI slop just like you can use a
44:36 camera mindlessly and and shoot
44:38 snapshots, right? you just you're just
44:40 shooting tick tick tick tick tick tick
44:42 tick tick tick and and those are
44:44 snapshots of your vacation, right? Or if
44:47 you've got a strong creative point of
44:48 view, you can wait until sunset at the
44:50 beach and make sure that someone's
44:52 standing on the surf and the the light
44:55 is coming through the waves as they're
44:56 crashing and you take the picture at
44:58 just the right moment. You've got a
45:00 point of view and all of a sudden that
45:02 exact same camera and that exact same
45:04 push of the shutter button now creates a
45:07 photograph,
45:08 not a snapshot.
45:11 You can do the same with AI.
45:13 AI can absolutely be a snapshot. Click,
45:15 click, click, click, click, click,
45:16 click, click, click. Right?
45:20 And I think that's how a lot of people
45:21 are going to use it and that's what
45:22 gives it a bad reputation and that's
45:24 what pisses people off about it. But if
45:26 you take a guy like this and he takes
45:28 fragments of songs that he never
45:30 finished and puts them in here and
45:32 explores them and comes up with a song
45:34 and he's like, "Oh yeah, that's really
45:35 good." And then he learns it and then he
45:37 produces it.
45:41 He's going to do this in a different way
45:42 than someone else. It's going to it's
45:44 going to be a new tool that gets worked
45:46 into his workflow.
45:48 or
45:51 he can choose to be the victim of it
45:55 and just lament the fact that we don't
45:57 make songs the way we used to, right?
46:01 Um,
46:02 and I think that's every single
46:06 knowledge worker, every single creative,
46:08 every single person that does anything
46:10 with computers.
46:12 You're going to have this moment, right?
46:15 The minute you try to do the stuff you
46:17 do and you realize, "Oh [ __ ] that's at
46:20 least as good as me, if not better."
46:24 Like, one of my favorite things to do in
46:26 the world is brainstorm. I loves me a
46:28 good whiteboard. Get me on the
46:29 whiteboard. I'll [ __ ] brainstorm up a
46:32 storm. You know how many how much
46:33 whiteboard brainstorming I've done in
46:35 the past year? Maybe twice.
46:39 I used to do it twice a week. I don't
46:42 need to anymore. It's better than me at
46:44 that.
46:45 And then there's two ways to react to
46:47 that. Okay, chat GPT is better at
46:49 brainstorming than I am.
46:53 I can either use it or not, right? Like
46:56 I do have the choice to not use it.
47:01 But if I don't use it, I'm actually
47:03 competing with people that do.
47:08 And so that's going to come back and
47:10 bite me at some point because I'm not
47:12 going to learn how to take my skills and
47:15 amplify them using this really amazing
47:17 idea generating machine, you know. So I
47:20 got to turn the air back on.
47:21 >> It's getting hot, people. It's getting
47:23 hot in here.
47:25 Hey, Chippy.
47:27 Hey, Champ.
47:33 [Music]
47:41 chainsaw. I have hours of oneliners and
47:44 chord progressions and lick licks. Now
47:46 they will become songs. Yeah. Um
47:50 let's see.
47:52 We could do
47:56 [Music]
48:04 Heat. Heat.
48:06 [Music]
49:11 Oh man, I don't know. I was going to do
49:13 something pseudo, but [ __ ] it. I don't
49:15 think so. Um, thoughts. What are your
49:18 thoughts? What What are all your
49:20 thoughts on Mr. Fernando Perdomo music.
49:26 This is absolutely why I'm writing my
49:29 book instead of running. Yep. So sick of
49:31 the fear-mongering. Damn. I know. I
49:34 know. It's like
49:37 the the whole point of this channel
49:41 that you know I started I started this
49:43 channel the week after chat GPT launched
49:47 because the minute I saw it
49:49 and the minute I I saw Microsoft's
49:52 investment go from 1 billion
49:56 to10 billion in like a month and a half
49:58 period. The only thing that I knew was
50:02 one, the technology is profound and two,
50:04 it's not going away.
50:08 Like it is crystal clear to me. It was
50:10 crystal clear to me the day chat GPT
50:12 came out. It ain't going away.
50:15 And so this channel exists
50:19 for the binary choice of what are you
50:21 going to do about that? It's binary.
50:24 It's absolutely binary.
50:26 You either deal with it or you don't.
50:29 Right. And from where I sit, just having
50:32 been on the planet long enough,
50:35 if you don't deal with it, AI is going
50:38 to happen to you. You're going to be the
50:40 victim of it, and it's going to [ __ ]
50:42 suck.
50:44 And if you choose to use it,
50:48 what you're going to what you're going
50:49 to very likely have is the moment that
50:51 Fernando just had,
50:55 which is, "Holy [ __ ] [ __ ] I'm
50:58 terrified."
51:00 And did you notice that he was not only
51:02 terrified, he was also invigorated.
51:05 And he's like, "Let me throw my my
51:07 toughest stuff at it." Oh, yeah. It
51:09 handled that, too. Huh.
51:14 And that's that's this this cycle of AI
51:16 adoption is very much like the cycle of
51:19 accepting death.
51:23 You're in denial. You're angry. You're
51:24 pissed. You have this moment of oh my
51:27 god.
51:29 You have this moment of joy and sadness
51:32 all at the same time.
51:34 And then you kind of move through that
51:36 into this place of like, wait a minute.
51:39 What if I really embrace this [ __ ] and
51:42 try it on? This can take my ideas and
51:44 amplify them. Those songs that were
51:48 produced by Suno are his songs.
51:53 Amplified.
51:56 Now imagine if he took those songs,
51:58 amplified them, got reinspired by them,
52:02 turned them up another notch, put them
52:04 back in suno, turned them up another
52:05 notch,
52:07 and he ends up with six different
52:08 versions of the songs that he can sell
52:10 to different artists.
52:13 Why not?
52:19 What if he's always wanted to make his
52:20 own music videos, but he didn't have
52:23 camera equipment and he didn't have the
52:24 skills, but now he can.
52:28 Right.
52:32 I completely ignore Wait, Tik Tok pin
52:35 adaptation is where it's at. Kake,
52:38 that's
52:40 the word for the next three years is be
52:43 adaptive. Be be adaptable.
52:46 Adaptation.
52:48 And and here's what I gotta say. I I
52:53 chat GBT5 in a weird way kind of kind of
52:56 [ __ ] me up a bit.
52:58 Not cuz I had an AI girlfriend. They
53:00 took away my girlfriend. It wasn't that
53:08 I've been the one that's out here been
53:09 out here preaching adaptability
53:12 and curiousness and play and you got to
53:14 play and you got to be adaptable. I've
53:16 been out here preaching that
53:20 it it it was roughly 18 months from when
53:23 chat GPT4 came out until chat GPT5 came
53:26 out.
53:27 And even though they kept improving GPT4
53:30 over those 18 months, it was still GPT4.
53:34 GPT5 is actually a fundamentally
53:36 different model. It it works
53:38 differently.
53:40 It responds differently. It's got a
53:41 different personality.
53:43 And what struck me was, "Oh [ __ ] I have
53:47 to start over
53:49 with learning how to use chat GPT."
53:52 Now, I don't have to start completely
53:54 over, but a lot of the ways that I
53:57 learned to prompt chat GPT4
54:01 no longer apply.
54:04 And there was a real sadness in that for
54:06 me. Like, there was there were a couple
54:07 of days in there where I was like, "Oh,
54:08 that's a [ __ ] bummer." I think part
54:11 of being adaptable is how quickly you
54:14 can
54:17 you can move through
54:20 these sad moments, right? Because
54:24 they're they're going to be sad, right?
54:25 It's going to be, you know, I took time
54:28 to learn this [ __ ] I went to this
54:29 stupid [ __ ] AI learning lab with this
54:32 old fat dude that just rambles on. He
54:35 never makes a point. And I did that for
54:36 like a year and a half so I could learn
54:38 how to use chat GBT. And now they
54:39 [ __ ] change it. Do I have to spend
54:42 another year with this bozo? He wears
54:45 pink bows and his dog sings.
54:48 Really?
54:50 [Laughter]
54:56 I still seem to tell GPD5 its role. Who
55:00 it is? Yeah.
55:03 How quickly you can b bounce back from a
55:05 beatdown. It Yeah.
55:08 Yep. How quick can you pivot?
55:13 Thank you, Erica.
55:19 Stop yapping and get a job. Exactly.
55:22 Exactly.
55:24 But, you know, I just my hope for
55:27 Fernando,
55:29 like I I have hope for him that that
55:32 he's not gonna
55:35 that he's not going to let this shut him
55:37 down,
55:41 that that he's willing to share his
55:43 experience. says to me that that he
55:46 might be willing to find a path into
55:49 into
55:50 at a minimum at a minimum really
55:54 understanding what Sunno makes possible
55:56 for him.
55:59 Not what what it does as a capability,
56:02 but what it makes possible for him. And
56:04 the only way he can do that, the only
56:07 way he can do that is to go through it,
56:12 right? and it's going to suck and he's
56:15 going to have moments of elation and
56:16 he's going to have moments of just
56:17 absolute devastation.
56:22 And it doesn't mean he's never going to
56:23 write a song on the guitar again. He's
56:25 always going to write songs on the
56:26 guitar.
56:28 But what it might mean is that he
56:29 doesn't get hired as much to do
56:32 producing. But maybe what it means is he
56:34 gets to produce more of his own music.
56:37 There's a possibility
56:42 Tik Tok pin.
56:44 Uh, pizza. I started using AI with
56:47 curiosity, but now I use it out of
56:49 necessity. Yeah. Yeah. Exactly.
56:52 Exactly.
56:54 Yeah. I I used to use AI as a curiosity,
56:58 right? Here's a parlor trick. Look at
57:00 this cute little thing it does.
57:03 And now it's like I couldn't imagine
57:07 bringing any idea to life without it at
57:09 this point.
57:16 All right. Launched last week. Okay.
57:18 Let's go check that out.
57:23 Oh man.
57:28 In my opin, in my humble opinion, AI
57:30 music production so far has no soul.
57:32 Yeah, I I I don't disagree. I mean, if I
57:36 look at every video that's been
57:38 produced, even even the most
57:42 well-crafted AI videos,
57:45 the most well-crafted ones are still
57:48 weird.
57:49 The physics are weird. They lack
57:51 humanity. There's just something weird
57:53 about them. But they're a [ __ ] ton less
57:56 weird than they were a year ago.
57:59 And a year before that, they weren't
58:01 even watchable.
58:03 Right? You remember the original um Will
58:06 Smith spaghetti eating video where it's
58:09 like his face is morphing and the
58:11 spaghetti's all over the place.
58:17 So just because AI music right now
58:19 doesn't have a soul,
58:22 don't don't assume that's going to be
58:23 the case for very long.
58:26 Don't assume video is going to be like
58:27 that. Don't assume the writing is going
58:28 to be like that. Assume that these
58:31 things are going to get better and
58:32 better and better. And then if that's
58:33 the case,
58:35 if if we truly get to the point that AI
58:37 generated stuff is indistinguishable
58:39 from human created stuff,
58:42 then what rises above the noise? And
58:44 that goes back to the Rick Rubin thing.
58:46 What's the point of view of the human
58:48 being behind it?
58:52 So
58:55 anyway,
58:58 people are going to get better at
58:59 prompting, too. Well, I think some
59:02 people are going to get better at
59:03 prompting. I think the other thing
59:05 that's going to happen is
59:08 people will have to prompt less and less
59:11 that we will have agentic systems that
59:13 will be doing the prompting for us. that
59:15 what we will give AI is an idea or an
59:19 imperative,
59:21 a direction.
59:23 I want a movie that's got the excitement
59:25 of Tarantino and the comedy of Ryan
59:30 Gosling. I don't [ __ ] know. And and I
59:33 and I want there to be trolls and
59:35 fantasy and lasers.
59:38 And then it will go figure out what
59:41 storyline makes the most sense and what
59:43 setting makes the most sense and what
59:46 special effects and what characters and
59:47 how they interact.
59:50 Um,
59:52 I think that's where it ultimately goes,
59:57 but I also don't think it prevents any
59:59 of us from handcrafting work or from
1:00:04 using AI tools with chain of craft to do
1:00:07 sophisticated things that we're proud
1:00:09 of. Like I I think it all just becomes a
1:00:11 spectrum. Some things will be fully
1:00:13 automated and hyperpersonalized.
1:00:15 Some things will be created with AI and
1:00:18 you know human AI collaboration.
1:00:21 Some things will be just digitally
1:00:23 produced with humans. Some things will
1:00:24 still be produced with physical cameras,
1:00:28 right? Maybe there's a big push back to
1:00:30 to physical film and hand editing
1:00:33 played on a projector outside on a
1:00:35 street on a sheet stretched against a
1:00:37 wall. Maybe that becomes the new the new
1:00:41 preferred way of watching movies. Um,
1:00:44 let's go look at So, Pizza Later, um, is
1:00:47 is one of our friends on Tik Tok and has
1:00:50 a site called o really oi.com.
1:01:01 Discover AI creators, explore the latest
1:01:04 and AI,
1:01:06 generative AI talent, and find creatives
1:01:09 for your next project. There you go. Is
1:01:11 is Joy Perie in here?
1:01:16 Yo, producer Brandon.
1:01:17 >> If she is, she'd be telling you to
1:01:18 change your tabs.
1:01:20 [Laughter]
1:01:24 >> And producer Brandon, he's good.
1:01:26 >> So, uh, this is this is actually Pizza
1:01:29 Later's friend that started this last
1:01:31 week and it looks super cool.
1:01:34 >> Yeah, it's very cool. I think this is
1:01:36 quite smart. Um because
1:01:40 you're already seeing so so um the last
1:01:44 two salons have been hosted by Liz
1:01:47 Miller Gersfeld who's a she was a an
1:01:50 advertising producer for like 20 years
1:01:52 out of Chicago and in the past two years
1:01:55 she's been getting up to speed on AI and
1:01:57 she's putting been putting together a
1:01:59 portfolio and she's now starting to get
1:02:01 hired by production companies you know
1:02:04 production companies with actual film
1:02:06 directors
1:02:07 like Emmy awardwinning film directors
1:02:10 and they're hiring her as a behind the
1:02:12 scenes prototyping engine.
1:02:17 So they know she's got good aesthetic
1:02:20 capabilities. They now know that she's
1:02:22 good at AI. They now know that they're
1:02:24 not good at AI, but their clients are
1:02:26 asking for it. And so this is a really
1:02:29 smart site. So if you are good at making
1:02:34 stuff, get started. Just as a creator,
1:02:36 join as an agency.
1:02:39 Discover and hire AI creatives. Join as
1:02:42 a creator. Nice.
1:02:48 Joyy's here. There you go, Joy.
1:02:56 Joy Perie, thanks. You're living proof
1:02:58 that artists like you make amazing art
1:03:00 with AI tools. Yeah. Here's the here's
1:03:03 the thing. The biggest trope the biggest
1:03:06 trope with By the way, nice work pizza
1:03:10 later on on that site. That's really
1:03:12 cool. Um I'm continuing to sign up.
1:03:16 The biggest trope, the biggest
1:03:17 misunderstanding,
1:03:20 please check this box. Yes.
1:03:23 The biggest misunderstanding with um
1:03:29 Hang on.
1:03:30 Okay, I got to make a profile. All
1:03:32 right, I'll make a profile. The biggest
1:03:34 misunderstanding with AI art is that
1:03:37 people that don't use AI still assume
1:03:42 Where's my Wait, where's my button?
1:03:45 Oh, I did I drop it the other day? I
1:03:47 dropped it the other day.
1:03:49 >> That's the one.
1:03:53 Most people that don't understand AI,
1:03:55 when you go, "Oh, yeah. I used AI for
1:03:57 that." They think this is what you did.
1:04:02 >> Here's your video.
1:04:06 It's not how it works.
1:04:09 even an image.
1:04:11 If you look at the work that Kelly Bosch
1:04:14 does, the images that she does that she
1:04:15 then turns into these videos,
1:04:19 that's years
1:04:21 of her getting ideas that have been in
1:04:23 her head for decades and figuring out
1:04:26 how to prompt midjourney in such a way
1:04:28 that she gets results that kind of match
1:04:30 this idea in her head. And then she
1:04:32 generates dozens and dozens and dozens
1:04:35 of images for the 20 or 30 that you see
1:04:38 in any one of her videos.
1:04:44 The the AI artists that are in the AI
1:04:47 salon that that come here to the
1:04:49 learning lab
1:04:53 are artists. They're they are they they
1:04:55 are learning a new craft of how to
1:04:59 string together multiple tools
1:05:02 to generate an output.
1:05:06 What's that sound like? Sounds like
1:05:07 movie making to me. Sounds like music
1:05:10 production to me. Sounds like image
1:05:12 making to me. Sounds like project
1:05:14 management to me. Right? Every [ __ ]
1:05:18 job is some collection of tools that we
1:05:21 interact with to generate some kind of
1:05:24 output.
1:05:35 My friend that launched it is trying to
1:05:37 help connect clients with creators that
1:05:39 might not have typical a agency
1:05:41 experience. Yeah. Yeah. Because what's
1:05:43 going to happen? What's going to happen
1:05:45 pizza is
1:05:47 the creative directors at the agencies
1:05:50 that have well they all have egos. The
1:05:54 creative directors at the agencies that
1:05:57 have egos and are insecure
1:06:00 and not willing to have the moment that
1:06:02 Fernando just had where they're just
1:06:04 like, "Ah, I'll deal with AI when it
1:06:06 gets good enough." That's a that's a
1:06:09 that's a little below my talent level.
1:06:13 I talked to a lot of agency douchebags
1:06:16 that are that that's their attitude. I
1:06:19 talked I just met a guy in LA
1:06:22 um runs
1:06:24 an agency
1:06:26 and his attitude was Yeah.
1:06:29 Yeah. Yeah. I mean, some of my people
1:06:30 are using it. Yeah. Whatever. All right.
1:06:35 The guy I was talking to was like he's
1:06:37 about to have his ass handed to him.
1:06:44 The only way through this is through. If
1:06:47 the only way through it is through. I
1:06:49 mean, it's literally So, Erica, for your
1:06:51 book, you know, this is this is you've
1:06:55 got a beach here. You got two choices.
1:06:57 Okay. There's there the the the waves
1:07:00 are getting bigger and and on the beach
1:07:03 is a really cozy tiki hut with pina
1:07:06 coladas in it and a nice fan and and and
1:07:09 and a beautiful person operating the
1:07:12 fan, right? And a and an ice cold pina
1:07:15 colada and and next to the the tiki hut
1:07:20 is a surfboard
1:07:22 and you don't know how to surf.
1:07:25 Those are your two choices.
1:07:27 That's where we are right now. The only
1:07:30 [ __ ] choice, the only [ __ ] choice
1:07:33 is grab the [ __ ] surfboard and get in
1:07:35 the water in my opinion because the
1:07:37 other one is going to be [ __ ] awful,
1:07:42 right? You're going to have temporary
1:07:44 bliss be, you know, while the tsunami
1:07:47 wave, you know, builds and builds and
1:07:49 builds and then just comes [ __ ]
1:07:51 smashing in.
1:07:53 So yeah,
1:07:56 shoot list meaning no film crews, no
1:07:58 location permits, no craft services,
1:08:00 etc. Right. Yeah, in some cases I I mean
1:08:04 Ryan, if you look at look at um the um
1:08:08 Netflix just put out a statement about
1:08:11 their um some film that they released in
1:08:15 I think it was April or May um that had
1:08:18 a big sequence in it, a big special
1:08:20 effects sequence that was done with AI
1:08:23 and and they didn't announce it until
1:08:25 late late July or early August, but they
1:08:28 talk about like how they did it and why
1:08:29 they did it. And they basically said
1:08:31 that movie couldn't have been made
1:08:32 without AI because that sequence they
1:08:35 would have either had to cut that
1:08:36 sequence which I guess would have [ __ ]
1:08:38 up the story or
1:08:41 you know it it would it would just would
1:08:42 have been prohibitive and so they did it
1:08:45 with AI. But couple of things that are
1:08:46 fascinating there. Who was the AI artist
1:08:49 that did that work? How did they get
1:08:51 hired? How did they find them? That's
1:08:54 why something like the uh the the O
1:08:57 really site is is really powerful. Like
1:08:59 if if you're someone that wants to find
1:09:01 someone that's got their [ __ ] together
1:09:03 with making AI video, go find them
1:09:06 there.
1:09:08 So there's going to be a lot of people
1:09:10 hired to do stuff for larger productions
1:09:13 that they just don't have the budget
1:09:15 anymore or just producers being
1:09:17 producers. Wait, instead of spending $5
1:09:21 million for this film, I could spend two
1:09:23 if we use AI. Uh, yeah, [ __ ] use AI
1:09:27 because I'm still going to make the same
1:09:28 money on the other side of this thing
1:09:31 whether we use AI or not. But if I don't
1:09:32 have to spend an extra $3 million and
1:09:35 put that at risk, why don't we do this?
1:09:38 Why don't I'll give you $2 million for
1:09:41 this movie with AI and then I'll give
1:09:43 you another $2 million for the sequel to
1:09:45 it if it's successful
1:09:48 and then I'll put a million dollars into
1:09:50 some punk that's just making AI movies.
1:09:54 So, I don't think you're necessarily
1:09:56 going to see a complete obliteration of
1:09:59 the film business. I just think you're
1:10:01 going to see a massive fragmenting of it
1:10:03 and probably an amplification of it. But
1:10:06 the current power structures are going
1:10:08 to lose their [ __ ] minds.
1:10:11 Erica, I truly think that the show
1:10:13 Adolescence will be one of the last true
1:10:15 oneshot TV shows that are done in such a
1:10:17 grand scale in a beautiful way without
1:10:18 AI. Yeah, I know it's Yeah, there's a
1:10:21 couple of them, but I I don't know that
1:10:23 it'll be the last one, Erica, but I
1:10:24 think what it'll be is
1:10:27 I think there's going to be a hunger for
1:10:30 analog produced work. We We've got to
1:10:34 get there first, right? Like we like
1:10:36 right now we've got all the shows that
1:10:38 are shot on, you know, curved LED
1:10:40 screen, you know, those LED rooms like
1:10:43 the Mandalorian.
1:10:45 So, we're going to work our way through
1:10:46 that
1:10:48 and then that's going to give way to
1:10:49 full AI productions and the AI
1:10:52 productions will get better and better
1:10:53 and better and within two or three years
1:10:54 they'll be indistinguishable from what
1:10:56 we have today. And I think I I think
1:10:58 probably at that point somewhere around
1:11:00 three years out from now someone's gonna
1:11:03 go like like do you remember when
1:11:05 Tarantino did did the Hateful Eight? He
1:11:08 wanted to shoot it on the anamorphic 70
1:11:10 millimeter 1.21 or 2.1 to one aspect
1:11:15 ratio and he had to go to he had to go
1:11:18 to uh Panavision and they they had two
1:11:21 of these lenses in existence. There's
1:11:23 two of them in existence and one of them
1:11:26 worked and one of them didn't. So, one
1:11:28 of them became like the donor body and
1:11:30 they rebuilt this lens
1:11:33 and then it had to be shown in special
1:11:35 theaters that had anamorphic projection
1:11:38 lenses
1:11:41 and so but like some and I I actually
1:11:44 went to see Hateful 8 um in LA in a
1:11:50 theater with you know with the actual
1:11:53 anamorphic film and it was insane. It
1:11:55 was crazy. Like you could see the
1:11:57 flickering of 24 frames a second and
1:12:00 there was just something janky about it.
1:12:02 There was just something right about it
1:12:03 because the subject matter was the wild
1:12:05 west, right? I think we're going to see
1:12:08 a return to a hunger for analog stuff,
1:12:13 in-person, you know, events, in-person
1:12:16 movies, making movies the oldfashioned
1:12:19 way. You know, a resurgence. We've seen
1:12:21 like a resurgence to film with
1:12:23 photography.
1:12:25 Um,
1:12:27 you know, we've seen a resurgence of
1:12:28 polaroids. So, I I I don't think it goes
1:12:31 away completely, but I think I think the
1:12:33 core way you produce things is
1:12:36 absolutely going to transform. It's
1:12:38 going to absolutely transform. And and I
1:12:40 think that the quantity of content goes
1:12:42 up dramatically,
1:12:44 which how the [ __ ] do we keep up with
1:12:46 it? That I don't know. I mean, I can't
1:12:50 I do. You guys Do you guys have TV show
1:12:55 anxiety? I totally have TV show anxiety,
1:12:58 which is this. A show comes on and I'm
1:13:01 like, "Oh, I want to watch that." And
1:13:02 then we start watching it and then it's
1:13:05 and it's one of those annoying shows
1:13:06 where you can't binge the whole season.
1:13:08 You have to watch it week by week. And
1:13:10 then the next week, you know it's on
1:13:12 Tuesday, but you're like, "Oh, [ __ ]
1:13:15 What channel is that?" And you're
1:13:16 literally like, is it is it Paramount
1:13:19 Plus? Is it Tuby? Is it is it Netflix?
1:13:22 Is do you think it's on YouTube? And
1:13:25 it's like 15 minutes and like the time
1:13:27 is coming up to when it's supposed to
1:13:28 start and and it's streaming. So you
1:13:30 have to watch that one live because they
1:13:32 don't show it for 24 hours in the
1:13:34 [ __ ] replay because they're [ __ ]
1:13:36 And you're like, "Where's what channel
1:13:38 is it?"
1:13:42 That's going to get worse.
1:13:47 So many platforms. It's so so many
1:13:50 platforms.
1:13:53 Bit Torrent. Yeah, exactly. Uh I think
1:13:56 there's going to be a big art fusion
1:13:58 era, meaning combining of art with AI.
1:14:02 Yeah, Joy. And why not?
1:14:05 Th this this whole [ __ ] idea that
1:14:09 artists that are on the outside of AI
1:14:11 right now have this all or nothing
1:14:14 attitude
1:14:16 like AI art is evil and everything else
1:14:18 is good.
1:14:20 Well, it wasn't always that way. Digital
1:14:24 cameras
1:14:26 were evil to film photographers.
1:14:31 They didn't consider digital photography
1:14:34 real photographs for years.
1:14:37 And the same things going on with AI
1:14:39 right now. But I think you're right,
1:14:40 Joy. It's like this is just another
1:14:44 mechanism in a series of mechanisms,
1:14:47 mechani,
1:14:49 that make [ __ ]
1:14:52 Some things are cameras, some things are
1:14:54 digital audio workstations, some things
1:14:56 are nonlinear editors. Now we've got AI
1:14:59 video generators. It's just generating
1:15:02 more crap to put in the chain of craft.
1:15:06 So yeah, I think you're right. I think
1:15:08 the experimentation of
1:15:13 like here's an idea.
1:15:20 Imagine a theater troop
1:15:24 Imagine
1:15:32 a combination of a theater troop and a
1:15:36 film production studio
1:15:39 and a bunch of AI prompt engineers.
1:15:43 Okay, go with me here.
1:15:46 and they rent an old olden timey movie
1:15:49 theater
1:15:52 and you go in as the audience
1:15:56 and it's like an improv show
1:16:00 where some actors come out on stage or
1:16:02 some you know ho the hosts come out on
1:16:04 stage or come out in the front of the
1:16:06 audience and like what are we doing
1:16:08 tonight
1:16:10 and then from the audience they're like
1:16:11 we want to see a western someone goes
1:16:13 sci-fi
1:16:15 all right you wanna sci-fi cowboy thing.
1:16:20 Space aliens or or space cowboys?
1:16:24 No, I want to see a race movie. Okay,
1:16:26 you want a race movie? What year?
1:16:30 And then they start prompting it.
1:16:34 And
1:16:36 what goes up on screen is
1:16:38 hyperpersonalized for that audience that
1:16:40 night, brought to life by people that
1:16:43 can that have that craft.
1:16:47 And every 15 minutes you stop the film
1:16:50 and say, "All right, we going left or we
1:16:53 going right? Do we still want to be in
1:16:54 this genre? We want to go costume
1:16:57 drama." Okay, so from here on out, it's
1:16:59 still going to be a race movie, but
1:17:00 we're going costume drama. Okay.
1:17:03 You sure you want that? Here we go.
1:17:06 Another 15 minutes of show. Like that
1:17:09 could be an entirely new genre of
1:17:11 entertainment.
1:17:16 But if the [ __ ] artists
1:17:19 are just sitting on the side like this,
1:17:22 I'm not going to use it. It's
1:17:23 plagiarism.
1:17:27 How do we ever discover that?
1:17:32 It's people like Fernando
1:17:35 who goes, "Holy [ __ ] I had no idea,
1:17:38 but then has the balls
1:17:41 to like make the videos to say, "Hey,
1:17:43 I'm going to go through this with you.
1:17:44 Let me show you what I'm learning here.
1:17:45 Let me tell you how I'm feeling about
1:17:47 it." It's people like Fernando that are
1:17:49 going to go, "Huh, what if we put
1:17:52 together a band
1:17:55 that their specialty their specialty was
1:17:58 they're a cover band,
1:18:01 but they're only covering songs that
1:18:03 were written that night."
1:18:08 And and while the warm-up act is is
1:18:10 playing, the audience is is giving input
1:18:16 and the AI is writing songs and the
1:18:18 cover band is breaking down those songs,
1:18:21 learning the chords, learning the
1:18:23 lyrics, and they go out and they play
1:18:25 original songs every [ __ ] night.
1:18:31 Why not?
1:18:33 Why not? Tik Tok pit joy party. Ask an
1:18:37 LLM to give you art fusion ideas. Some
1:18:40 of the ideas are pretty awesome. That's
1:18:41 a great idea actually. Yeah. Yeah. Yeah.
1:18:44 Yeah. Yeah. The impromptu cover band.
1:18:46 Exactly. Why not?
1:18:49 Like that's a whole different skill of
1:18:51 musician. You take like you'd have to
1:18:53 have some pretty [ __ ] baller chops,
1:18:56 right? Some really tight musicians that
1:18:58 would be able to just take, you know,
1:19:00 they they'd be more like jazz bands,
1:19:02 right? They could just take chord
1:19:03 progressions, a couple of lyrics, and
1:19:05 just [ __ ] nail it.
1:19:10 Every song could start with here's these
1:19:12 version that we created.
1:19:17 They fade that out as the band comes in.
1:19:23 Could be really exciting
1:19:28 or not.
1:19:30 New Tik Tok pin. Corey,
1:19:33 use my fabulous film finder. GPT Kyle
1:19:35 finds any platform
1:19:38 anything is on. Oh, that's cool.
1:19:40 Fabulous film finder. Okay, cool. I'll
1:19:41 check it out.
1:19:43 I would love that cover band. Early AM
1:19:46 meeting night all night, KKE. Thank you.
1:19:57 Yeah, people won't write a book first.
1:19:59 They'll write they'll make the movie
1:20:00 first and then write a book based on the
1:20:02 movie. Yeah, exactly. Why not? All
1:20:05 right. Um it's getting to be that time.
1:20:08 So, tomorrow um I keep threatening to do
1:20:11 this and I'm going to do it at some
1:20:12 point here. I'm going to get graphics
1:20:14 for this. Tomorrow at 11:00 a.m.
1:20:17 Mountain time on LinkedIn, I do AI
1:20:20 office hours and um it's for anyone who
1:20:24 wants to come. So, if you go to my
1:20:25 LinkedIn account, Kyle Shannon on
1:20:28 LinkedIn, click on events, it's been the
1:20:31 same event URL. It's a Google Meet
1:20:33 event. Um, it's 11:00 a.m.
1:20:37 Mountain time tomorrow. Okay? So, just
1:20:41 go find that and come to that. And then
1:20:44 tomorrow night is Friday night date
1:20:46 night. It's just a normal Friday. I
1:20:48 don't think we have any gallivanting to
1:20:50 do in the art district. So, it should be
1:20:52 just regular 8:00 Friday night date
1:20:54 night. So, get your nachos, get your hot
1:20:56 pocket in the dirty microwave. Get it
1:20:58 all ready.
1:20:59 Probably some some orange drink or red
1:21:02 Kool-Aid.
1:21:04 And and preferably not a date. If you
1:21:06 have a date, bring them. If you don't
1:21:09 want to see them again, you bring them.
1:21:12 But otherwise, we'll be each other's
1:21:13 date. Okay.
1:21:21 Ah.
1:21:23 All right. Um,
1:21:26 I don't even know what we talked about
1:21:28 tonight,
1:21:35 but I know I got peppy about it. Um, all
1:21:39 right, cool, Griffy. Hope you all have a
1:21:43 fantastic evening, and I will see you
1:21:44 tomorrow night. Peace.