6/23/2025 - The Evolution of AI Tools: How They Are Changing Art and Expression

Description

Chapters

Transcript

0:00 Are you ready for some football?
0:05 Are you ready?
0:09 [Music]
0:21 [Music]
0:34 Well, I heard there was a secret called
0:38 David and it Peace and love.
0:42 You don't really care for music, do you?
0:49 Like this fees.
0:52 [Music]
0:58 Hallelujah.
1:00 [Music]
1:01 Hallelujah.
1:02 [Music]
1:05 Hallelujah.
1:08 Hallelujah.
1:11 Hallelujah.
1:15 We are
1:19 [Music]
1:21 with faith strong but you needed proof.
1:25 You saw her bailing on a roof. Her
1:28 beauty and the moonlight overthrew you.
1:34 She tied you to the kitchen chair. She
1:38 smashed your throne and cut your hair
1:40 and from your lips shed. Hallelujah.
1:47 Hallelujah.
1:50 Hallelujah.
1:53 Hallelujah.
1:56 Hallelujah.
1:59 [Music]
2:09 Does Satan wear suit and tie? Does he
2:12 work at the Dairy Queen?
2:15 Does he listen to Rock and Roll? Does he
2:18 feed the beans singing hallelujah?
2:23 Hallelujah.
2:25 Hallelujah.
2:26 [Music]
2:31 What about Jesus?
2:33 Didn't he do it too? Yeah. Hang out with
2:36 the prostitutes. Have a drink or two.
2:40 Power of example. My mama said it and I
2:43 heard. She said one ounce of action
2:47 beats a ton of words
2:49 singing hallelujah.
2:52 Mama said there would be angels.
2:54 Hallelujah.
2:57 Mama said there would be songs.
3:02 [Music]
3:09 Happy Monday everybody. What's shaking?
3:12 What's going down? What's happening?
3:16 We are here. We're here at the AI
3:18 learning lab.
3:21 We're gonna We're going to figure this
3:23 stuff out. We're going to figure it out.
3:25 What are we going to figure out? I don't
3:27 know.
3:30 But welcome, welcome, welcome.
3:33 Champy's in a good mood tonight, which
3:35 is always nice.
3:38 Always nice. How is everybody? Dr. Jay
3:42 in the house. Dragonfly Alchemy.
3:46 [Music]
3:59 Let's see. Black bar.
4:02 We are the
4:09 we'll do this.
4:11 Then we'll do this and then we'll do
4:12 this.
4:15 It's sort of like that sort of works.
4:19 [Music]
4:21 All right. Yeah, that's good. All right.
4:24 Fantastic. Hey, Brian Whitney was
4:27 shaking Brian Whitney. Yeah. Yeah.
4:30 There's Dragonfly Alchemy. We got Steo
4:33 in the house. Steo's calling in from
4:36 down under. He's down under there.
4:39 Yeah, Steo's here.
4:43 Tobias is here. Tobias, I got the
4:46 stickers. The stickers have come.
4:49 Look how beautiful they are.
4:53 Serena's memorial sticker.
4:56 Am I late? Never.
5:00 We're not weird, we're irregular.
5:03 How awesome is that?
5:07 So, I'll tell you where to get these.
5:10 And then there's an AI salon one that
5:11 comes with it.
5:14 And then there's a uh a little memorial
5:16 sticker that's got the information and
5:18 their little name.
5:20 Really sweet. Really sweet. But they
5:22 look great. The quality is great.
5:24 They're like the perfect size.
5:29 We're not weird, we're irregular. There
5:32 you go. I'm doing well, Brian Whitney.
5:35 I'm hanging in there.
5:38 Living the dream.
5:41 [Music]
5:45 I think I I I I keep saying this and I I
5:49 think I'm going to keep saying this. I I
5:51 feel like this is
5:54 something that that is just going to be
5:56 a constant
5:57 in this channel is that
6:01 we keep shifting into
6:07 I feel like I can't use the word realm
6:09 anymore because because chat GPT uses it
6:13 too much.
6:15 But I feel like we're moving into a new
6:18 realm
6:19 where um
6:26 where individuals can do
6:29 increasingly sophisticated things they
6:31 would never would have thought of.
6:37 [Music]
6:55 [Music]
7:11 Freedom came my way that night.
7:18 Just like a jet plane head and out of
7:21 sight.
7:23 I was hauling ass at a million miles an
7:26 hour wondering how hard I'd hit.
7:33 When they came into the station,
7:40 they said I was bad beyond repair.
7:43 [Music]
7:45 Okay.
7:47 [Music]
7:49 Where are we going to start? What are we
7:50 going to do? What are we going to do?
7:51 What are we going to do? What are we
7:51 going to do? What are we going to do?
7:54 Beautiful vocals. Thank you very much. I
7:56 really do appreciate that. I would not
7:58 consider myself a singer by any stretch
8:00 of the imagination, but there's some
8:02 songs I got that I love.
8:05 I like making sounds.
8:10 I move air over my vocal cords, but I
8:13 appreciate the kind words. It's funny
8:16 when I hear myself sing, when I hear the
8:17 playbacks, I'm like, "Yeah, it's a
8:19 little thin." It ain't American Idol.
8:21 Well, I'm also significantly too old for
8:24 American Idol. This would be American
8:26 Elder. I'm I'm even I'm not even good
8:28 enough for American Elder singing
8:30 competition.
8:33 You do. Okay. Thank you very much.
8:38 All right. Um,
8:42 do I keep holding on to the guitar?
8:43 What's everyone got going on? Um, anyone
8:46 have any questions about AI? Thoughts,
8:48 things they want me to look at? I can do
8:50 that. Um,
8:52 [Music]
8:54 I'm still trying to get my head around.
8:57 MidJourney video. Uh, just feels like
9:01 something very, very special.
9:04 um that I haven't quite dug deep enough
9:07 into yet.
9:09 Um but at the same time, we've got
9:13 a bunch of other models that are really
9:15 good. Miniax from Halo is is supposed to
9:18 be quite good. Cling 2.1 is supposed to
9:21 be quite good. Higsfield is supposed to
9:24 be quite good. Like it there's a lot.
9:27 Korea's got a bunch of stuff going on
9:28 right now, including their own image
9:30 generation tool. How about America's Got
9:33 Talent? Yeah, it America's America's Got
9:38 Mediocrity.
9:42 America's Got participation
9:44 certificates. That's the show I'm going
9:46 to be on.
9:49 Everybody gets a golden buzzer.
9:54 See, listen. Here's why I'm bitter. What
9:56 am I What What is Tik Tok telling me to
9:58 do? I'm supposed to swipe something.
10:01 All right. It's not Nothing's happening.
10:04 All right. I don't know what to do.
10:05 Okay. Whatever. Um,
10:08 see, I grew up in a time
10:11 Gen Xer, right? We were the drink from
10:14 the water hose generation. I grew up in
10:17 a time where if you entered a
10:20 competition,
10:22 they had these things. What were they
10:24 called? Uh,
10:27 winners.
10:29 You had winners that won the competition
10:32 and then the other people were told what
10:34 a piece of [ __ ] they were by their
10:36 parents.
10:38 Where did that go? Why did we lose that
10:41 spirit?
10:46 So, I want to get me some of that
10:48 millennial Gen Z. Everybody wins. We're
10:52 going to play What's that? What's that
10:54 board game where no one We're all
10:57 winners. We have winners.
11:02 Oh my god.
11:04 AI slap. Oh, AI slop on John Oliver last
11:08 night for bad vids. I didn't see it.
11:10 That's interesting.
11:15 [Music]
11:17 And you participate participation
11:20 ribbons. I got lots of them. See, I have
11:23 no participation ribbons in my life.
11:26 None.
11:27 I've got the two or three trophies I won
11:29 for like like one was a monologue
11:32 contest in an acting contest up up in
11:35 upstate Pennsylvania.
11:41 And that's it. Everyone else loser.
11:46 [Music]
11:50 Oh man.
11:57 Oh yeah. So, congratulations to Vicki
12:01 and Gareth.
12:03 So, uh who was it? Open AAI. Open AAI
12:07 had a children's uh book story contest.
12:12 A book a children's book contest. You
12:14 put a story in a children's book. Um,
12:18 and Vicki and Gareth tied with some
12:20 other person for they they did a
12:22 three-way tie for first. So, the AI
12:24 salon coming in hot, coming in strong,
12:29 showing up at the Open AI uh contest.
12:32 So, we got winners. We got winners right
12:34 here.
12:36 [Music]
12:40 The term AI slop is when there's nothing
12:42 on Pinterest but AI.
12:44 Is that what's happening on Pinterest?
12:46 Is Pinterest all full of AI now or is it
12:48 We're not there yet. We're at the
12:50 beginning of talks on how general AI and
12:53 AI agents fit into our ERP upgrade. A
12:56 lot of unknowns. Wow, that's
12:58 interesting. Fantastic.
13:01 Congratulations for representing. That's
13:03 amazing. Congratulations. Wow.
13:04 Congratulations, Vicki.
13:09 So, it is all full of AI.
13:12 Pinterest is that doesn't surprise me.
13:15 [Music]
13:17 [Applause]
13:18 [Music]
13:24 It's actually fascinating watching
13:28 there's an artist who if you don't if
13:30 you don't know the artist Kelly Bosch B
13:32 osc
13:34 follow her on Tik Tok
13:37 and it's fascinating watching what tools
13:39 she's using like her what story she's
13:43 telling with her images and and and
13:45 videos is shifting.
13:48 Um, she's starting to use uh I think VO3
13:52 and it looks like she's starting to now
13:53 use MidJourney,
13:57 but it's it's changing the story she's
14:00 telling.
14:02 [Music]
14:05 Yeah, it's actually really interesting.
14:07 Enya. Yep. Like the FBI. Oh,
14:10 fascinating. Um, I don't know if it's
14:12 Ena. Ena or Enya? Probably Ena, I'm
14:15 guessing. Um,
14:20 one of the things that I am, well, one
14:23 of the things we're fortunate, so so if
14:25 you don't know, my name is Kyle Shannon.
14:26 This is the AI Learning Lab. Um, I'm
14:30 also the co-founder of the AI Salon and
14:33 and the AI Salon and AI Learning Lab
14:36 have um a lot of connection. So, a lot
14:39 of a lot of members that come here are
14:40 members of the AI salon. probably most
14:42 that are irregulars um are members of
14:45 the AI salon. Um and the AI salon is
14:49 full of people who are not only AI
14:53 optimists, but they're people
14:57 who are thoughtfully trying to figure
14:59 out AI, right? Our are we we've got this
15:02 what we call the cycle of AI readiness.
15:04 In fact, let me show you my my screen.
15:07 Let me show you my screen here, people.
15:11 show you my little screen.
15:14 So, when you go to the AI salon,
15:18 it's French.
15:21 When you go to the AI salon,
15:24 land on this welcome page, and then we
15:26 talk about this cycle of AI readiness.
15:30 And it's play first, mindfully create,
15:34 and generously lead.
15:36 And what that looks like is
15:39 if you're brand new to degenerative AI,
15:42 one of the things that I think is a
15:43 mistake
15:45 that many people do is they think like,
15:48 okay, I'm a project manager. I'm going
15:50 to learn AI. Let me go ahead learn how
15:52 to make project management better with
15:54 AI. They go straight into how we've
15:57 always used computers historically,
15:58 which is let me take what I do and make
16:01 it more efficient,
16:03 which AI can definitely do that.
16:07 But what you what you the disservice you
16:10 do yourself if you just start with make
16:12 [ __ ] more efficient that I already know
16:14 how to do
16:17 is that one is you get a skewed vision
16:20 of of what AI makes possible.
16:24 right? Like maybe it's not quite perfect
16:26 at the thing you want to automate and
16:28 then that seems like well AI is not
16:30 really all that good. So, it does it
16:31 there's a little bit of that going on,
16:33 but the other thing it does is there's
16:35 so much more that that all of us now
16:38 have access to things that we can do
16:40 that we couldn't do before. And the only
16:41 way to really discover them is to play.
16:44 And so play just looks like start using
16:47 AI tools without expectation. Go into
16:49 chat GPT, sure, but you know, have it
16:52 write you recipe books, things like
16:53 that. And then so that's play. So that's
16:55 how you learn and you start to
16:57 understand what's possible. And if you
16:58 do it right, you have a series of what
17:01 we call Kevin Mallister moments, right?
17:03 Where you're like,
17:05 you know, from the movie Home Alone, you
17:07 have these moments where you're just
17:11 you're gobsmacked, you're you're blown
17:13 away, your your face melted, where
17:16 you're like, "Oh my god, I didn't know I
17:17 could do that."
17:20 And then you move into the next phase,
17:22 which is mindfully create. Once you
17:24 understand what's possible with some of
17:25 these tools, now you can start to say,
17:27 "Huh, I wonder if I if I've got this
17:30 problem and I I wonder if I could solve
17:31 that problem using some combination of
17:33 these things I learned." And then you
17:35 try it and then generously lead is about
17:39 now that you've tried that stuff and
17:41 tried to build it, share it back with
17:43 the community.
17:45 And so what we've got is we've got a
17:48 group of people who are curious and
17:49 playful,
17:51 who do really smart things with AI in a
17:55 very thoughtful way and then share what
17:56 they're learning along the way. And that
17:58 sharing piece means that other people
18:00 coming in can learn from that. And so
18:03 it's this virtuous cycle. And so if
18:06 you're part of a community like the AI
18:09 salon, you're very fortunate to be able
18:11 to see through the slop because I
18:15 absolutely agree that
18:18 the where's my button? You know, the the
18:21 old Staples easy button.
18:26 Um it's really easy with AI to just push
18:29 a button and have crap shoot out of it
18:31 and then just put that out in the world.
18:34 and and I do think that's going to
18:35 pollute a lot of things. However,
18:38 if you start following people that are
18:40 doing really interesting work,
18:43 um
18:45 the view on a AI shifts for for me it
18:48 does anyway. It shifts from, oh, this is
18:50 going to ruin everything because it's
18:51 just going to make all this sea of noise
18:53 to there's actually some some works are
18:58 going to be created, whether they're
19:00 games or apps or or you know, business
19:04 solutions or films or art
19:08 that are just of a different nature,
19:10 right? What artists have always done
19:14 and and I use artists in in the broadest
19:16 term possible. I think, you know, people
19:18 who create software are artists. I think
19:21 people, anyone who's generating, anyone
19:23 who's creating,
19:27 we're going to start to see things that
19:30 are remarkable because what artists do
19:32 is they take the same tools that
19:34 everybody else has access to. Everybody
19:36 can buy a paintbrush.
19:38 Everybody can use chat GPT
19:42 and they look at it at what it makes
19:43 possible, and they do something
19:45 different with it. So yeah,
19:49 Facebook are full of AI. Mom and dad are
19:52 confused. That's great. The best I've
19:55 yet to see since April 2023. Mindful,
19:58 playful, and generous. Just Yes. Yes.
20:00 Exactly.
20:02 Play first, mindfully create, generously
20:04 lead.
20:11 And
20:16 I I want to show you something right
20:17 now. I want to show you something that I
20:19 think is I forget what I said. I I
20:22 didn't mean to be this mean. My My son
20:25 Zachary once said to me, I I cracked
20:27 some joke and he laughed at it and then
20:30 he paused and he said, "You know, Dad,
20:32 you're funny, but you're mean. You're
20:34 kind of mean." Um, so I was mean I was
20:37 mean to our dear friend Anne Murphy.
20:40 Um, if you don't know Ann Murphy, she's
20:43 she's just a powerhouse of badasserie.
20:48 Um, she's been part of the AI salon
20:51 forever. She she runs uh she leads AI um
20:54 which is a community for for badass
20:57 women uh exploring AI. So, Ena, if if
21:01 you uh a you should join the AI salon,
21:03 which if you go to
21:05 this URL, where is it?
21:09 go there, go to salon.ai and say join
21:13 our community.
21:15 Um,
21:18 and then Ann's got she leads AI. So, you
21:20 should check that out as well. Um, but
21:22 Ann made this game and and I think the
21:25 comment I said about the game is uh it's
21:28 so dumb it's brilliant.
21:31 Let's see. Um, I've had people tell me
21:34 AI is not art, but I think it is. Well,
21:36 you know, you know what's funny? Echo
21:37 kind official here. Here here's here's
21:39 my opinion on that.
21:45 So, first of all, AI is not art. AI's AI
21:49 is this this tool or this suite of tools
21:52 or this AI AI is almost like the
21:55 internet, right? Saying AI is art is
21:57 almost like saying the internet is art.
22:00 Um or Photoshop is art.
22:04 Art
22:06 is when someone
22:09 has an idea about something
22:12 and they use some combination of tools
22:14 to create something that they go, "Huh,
22:19 yeah, I think that's art. I'm going to
22:21 put that in the world." What gives AI
22:23 the bad rap
22:26 is that you can just push a button and
22:29 it will just generate crap.
22:32 And and we're starting to get to the
22:34 point where you can push a button and
22:35 the stuff it generates isn't even
22:37 necessarily crap, but it's not
22:40 intentional.
22:41 It's not
22:43 filtered through an artist's point of
22:46 view. Art is when someone crafts
22:49 something. And you can craft things in
22:52 lots of ways. You can craft it with AI,
22:53 you can craft it with, you know, digital
22:55 tools, you can craft it with analog
22:57 tools. You can use some combination of
22:59 them. Most of the good AI artists that I
23:01 know are using a lot a combination of a
23:04 lot of tools to get the result they want
23:06 and then put something in the world. Um,
23:09 so what's art is that thing that was
23:13 intentionally crafted. And I think one
23:15 of the things that the AI haters fail to
23:18 recognize
23:20 is that it's possible to do intentional
23:25 high quality art using any [ __ ] tool
23:29 you want. You can make art with a
23:31 hammer.
23:33 You can make art with AI. You can also
23:37 create crap with a hammer. And you can
23:39 make crap with AI.
23:41 But, you know, the the the uh the haters
23:45 right now that, you know, the tropes are
23:46 all it's all theft. It's people are just
23:49 lazy. They're just no one's thinking
23:51 about it. It's simply not true. There
23:53 are people doing incredibly, you know,
23:56 well-crafted work.
23:59 Um, Kayla, I've been getting into video
24:02 using magic light AI. Do you have any
24:05 prompting tips? Good lord. I don't I you
24:08 know what's amazing to me? I I follow
24:12 these tools for a living.
24:15 Well, not for a living. I don't get paid
24:17 for this. I follow these tools for a
24:19 volunteering.
24:25 I have not heard of Magic Light AI. Um I
24:28 do have ideas for prompting, but let me
24:30 let me go show you something. So,
24:32 Brandon, make sure I come back to Kayla.
24:34 I want to get to how to prompt for AI
24:36 video because I think it's really
24:38 important that I'll walk through it. I I
24:43 am starting to get to
24:47 understanding a new way of thinking
24:49 about how we use these tools and and
24:52 your question is one that's going to let
24:54 me unpack that a bit. Um I want to go
24:59 do I have it here? I do. Okay. So, and
25:02 I'm going to show Brand Brandon producer
25:05 Brandon produced a game called Prompt
25:07 Master, which we're going to show at
25:09 some point. Um, I haven't played with it
25:11 yet. I want to play with it myself
25:14 before I play with it here. Um, but I
25:18 just I want to show you this game.
25:21 So, so Ann Murphy did this. If now if
25:25 you don't know Anne Murphy, Anne Murphy
25:27 is just an amazing woman and she's she's
25:30 done all sorts of stuff in AI. She um
25:32 she has a an organization called
25:34 Empowered Fundraiser where she helps
25:36 nonprofit raise funds and become AI
25:39 literate and things like that. She runs
25:40 she leads AI. Um one of the things Ann
25:44 will will admit will tell you herself is
25:46 she's is very not technical
25:51 and she she told me a story. She goes,
25:53 she said, she's she came on to
25:55 something. We were getting ready to do a
25:57 podcast and she said, "Kyle, I I I vibe
26:00 coded." If you don't know what vibe
26:02 coding is, it's using these tools like
26:05 Lovable, which this was made in um or
26:08 Replet agent or or you know, things like
26:10 that. Um where you can just speak an app
26:14 into existence. And she goes, "I made a
26:17 video game." And and then and I said I
26:20 said, "What do what do you know about
26:22 video games?" She goes, "I don't I know
26:23 nothing about video games. I know
26:25 nothing about coding."
26:28 Um,
26:30 in fact, she said, she said, "We had a
26:32 KCO vision when I was a kid." So, I had
26:36 a KCO vision and then it broke and then
26:38 dad and her dad just said or or the kids
26:41 argued over it or something like that.
26:42 And the and her dad just said never
26:44 again. So, so she grew up her entire
26:46 life without video games. So, like her
26:49 kids are flabbergasted how clueless she
26:51 is with video games. So, she made a
26:53 video game. And so, I want to play it
26:56 for you here. So, it's basically like a
26:57 high- five game. So, it's called
26:59 Wholesome High Fives. And it just so
27:02 it's so awesome. So, you click on it and
27:04 then you literally you've got these
27:06 little little dots and you sort of click
27:08 on them and then they give you points
27:13 and then some of them at some point
27:15 should do a little saying. Uh, you're as
27:18 cool as the other side of the pillow.
27:20 Your energy is immaculate.
27:23 Um, your shoulders are about the perfect
27:26 amount of broad. And then you So, all
27:29 you're doing is like high-fiving these
27:31 characters for 30 seconds and getting
27:34 told sweet swell things.
27:37 And like the Where are we in time? Okay,
27:40 we're we're up on time. Okay. New
27:42 personal best. You spread 210 points of
27:45 joy.
27:48 You play it for 30 seconds and all
27:50 you're doing is like high-fiving emojis.
27:52 And it's my response to it was it's so
27:54 dumb. It's brilliant, right? Because
27:57 there's casual games where you don't
27:58 really have any game mechanics and you
28:00 just sort of have, you know, like one
28:02 thing to press. This is very much in
28:04 that neighborhood, but it's like almost
28:05 a level even simpler than that. And it's
28:09 just joy, right? It looks like Manis.
28:12 Oh, maybe it was Manis. It might have
28:14 been Manis. Oh, it is Manis. Yeah. Yeah.
28:16 Yeah. Yeah. This was done in Manis.
28:19 Um,
28:21 and so here's why I think this is
28:23 important
28:25 and and then we're going to get to I
28:26 want to get to the the video prompting
28:28 stuff. Here's why I think this game is
28:30 important and and the one that that
28:32 Brandon's working on.
28:37 A lot of engineers right now, just like
28:39 there's a lot of people in the art
28:41 community and video community saying AI
28:44 art is [ __ ] It's all crap.
28:49 Software developers are saying the same
28:50 thing about vibe coders. Ah, it's not
28:53 real programming.
28:56 Remember when photographers, film
28:58 photographers said digital digital
29:00 photographs weren't real photographs.
29:04 It's not real photographs. Look at the
29:06 resolution.
29:10 Um, that's the stage we're at with all
29:13 these tools.
29:15 What is remarkable about what Ann did
29:19 here and what a lot of people are doing
29:20 with vibe coding
29:23 is that you've got people creating
29:26 applications whether they're games or
29:29 CR an said the next thing she wants to
29:31 build is a CRM because she's she hates
29:33 Salesforce. This is a person who's never
29:36 coded anything before in her life, got
29:39 such joy and satis satisfaction out of
29:41 making this game
29:43 that she's now like, you know what, I
29:45 think I'm going to make a CRM that works
29:47 for me.
29:51 We've never ever lived at a time in
29:55 history
29:57 where people who didn't have particular
30:00 skills could find any kind of success in
30:04 disciplines outside of what they were
30:06 trained on.
30:08 So, we're entering an era where all of
30:11 us, all of us who choose to participate,
30:15 right, if we're not sitting on the
30:16 [ __ ] sidelines, which is the dumbest
30:19 [ __ ] move I can imagine right now.
30:21 Did you know uh PWC just came out with
30:23 an employment report last week that AI
30:27 literate employees have on average 56%
30:31 higher salary than nonI
30:34 literate employees?
30:36 like like why people are choosing to sit
30:40 on the sideline because they're afraid
30:42 of it is beyond me. But it's every time
30:44 there's a technological shift, this
30:46 happens, people get afraid of it and
30:48 they demonize it and they diminish it.
30:49 And so that's happening. But if you look
30:52 at what's happening here, real everyday
30:55 people who've got brilliant skills over
30:59 here are going, "Huh, I wonder if I
31:02 could make a game. I know nothing about
31:04 games and I know nothing about
31:07 programming, but I want to make
31:09 something that makes people feel good.
31:12 And then she does.
31:16 It's amazing. It's amazing.
31:20 So, all right.
31:25 Prompting for video.
31:28 Okay.
31:30 Let's see.
31:34 So, what I'm gonna what I'm not going to
31:36 do is give you a prompt for video
31:39 because first of all, I don't know.
31:44 Um, one of the one of the running jokes
31:47 on this channel is, "Sir, what are your
31:49 qualifications?" Qualifications? I have
31:51 no qualifications. Um, nobody has
31:53 qualifications right now. You can't keep
31:55 up with this [ __ ] Now, there going to
31:57 be people that are going to sell you
31:59 prompts and they're going to sell you
32:00 prompting methodologies. There's there's
32:02 all sorts of [ __ ] going on. I'm actually
32:04 working on a book right now called Feed
32:05 Your Prompt, but it's not actually about
32:07 prompting, oddly enough. It's about
32:09 self-actualization
32:11 and amplifying yourself using AI.
32:14 Um, so that's how we're going to deal
32:17 with with this this prompting thing. So,
32:19 we'll let me go to if you could pop up
32:22 my screen, Brandon. We'll go to um
32:28 flow.google.
32:31 So, this is not the tool that you use.
32:33 What was the tool that you said you
32:34 used? I don't know if the person who was
32:37 here is still here. Who's asking about
32:40 this is still here.
32:44 Magic Magic Light AI for video. Never
32:46 heard of it. Amazing. Magic Light AI.
32:51 Magic Light AI, the official website.
32:56 There you go. I wonder what this is
32:58 based on.
33:00 Actually, what this what this looks
33:01 like, this is probably sitting on top of
33:05 someone else's API. So, this is probably
33:09 like Cling or
33:12 or uh I don't know what's underneath
33:15 this. I wonder if it says is there an
33:16 about thing?
33:19 Magic light AI
33:23 story to video
33:28 create videos up to 30 minutes long with
33:30 your script. Fascinating.
33:36 Cool. Okay.
33:39 But the question, how do you get how do
33:43 you get these things to behave? Okay, so
33:46 couple of things. Let's let's talk
33:48 caveats with let's talk caveats with uh
33:53 with generative AI video right now. For
33:57 every successful video that you see on
34:00 Twitter or LinkedIn that you're like,
34:02 "Oh my god, that's so good." If it's a
34:05 30 second long video, that probably
34:07 means it's got seven to 10
34:11 video generations in it. And for every
34:14 one of those seven to 10 video
34:16 generations, there are probably 10 or 15
34:19 or 20 failures
34:22 to get to that one clip that was usable.
34:25 Uh pin. Yes, it's pulling from Cling 2.0
34:28 and Halo. Okay, cool. Yeah, perfect.
34:30 Yeah. So what they're doing, so there's
34:33 going to be a lot of what are called
34:35 rapper apps where and it's funny the the
34:38 development community is poo pooing
34:41 rapper apps. So a rapper app is taking
34:44 one of the core models that are out
34:46 there and they're building some new
34:48 interface around it, right? And so they
34:49 call them rapper apps. Um and and people
34:53 diminish them because they're like,
34:55 well, it's not as they they're not the
34:56 core technology. there's this snobbery
34:58 in, you know, Silicon Valley with if if
35:01 you didn't, you know, if you're not
35:02 building one of the core models, then
35:03 you're not doing real work. And it's
35:06 horseshit. Um, in fact, the uh the
35:09 company that just sold to,
35:13 who' they sell to?
35:15 I forget. Uh Wix. They sold to Wix for
35:18 $80 million. a six-month-old company
35:23 that went from
35:26 zero to $3.5 million in revenue in six
35:30 months and sold for $80 million was a
35:33 rapper app. Okay, that happened to be
35:36 one of the vibe coding uh programs and
35:40 they just happened to do it good enough
35:41 that Wix was like, "Yeah, we'll just buy
35:43 that rather than build it." So, you want
35:46 to go make $80 million, do a do a
35:48 wraparound. So anyway, so it's a it's a
35:51 wrap app. I don't know what I don't know
35:53 what it is, what it does well or not,
35:55 but the fact that it's on cling and
35:56 Halo, those are both Haloo's. The Miniax
35:59 model, um Cling is is this other model.
36:01 Both of those are really good. So So
36:03 it's it's it's probably a good service.
36:06 Um so a couple of things. The video
36:09 tools, even the really good ones, are
36:11 very very very very janky. They're very
36:15 janky. even if you prompt them
36:17 perfectly, you're gonna you're going to
36:19 end up with a bunch of crap. Um, so just
36:23 you just need to know that. You'll also
36:26 learn things like they do some things
36:28 really really well and they do some
36:30 things really really poorly. Like try to
36:32 get them to do gymnastics. They're
36:35 hilariously bad at gymnastics. Um,
36:38 MidJourney, by the way, is the first the
36:40 first one that I've seen. I think it was
36:42 Matt Farmer did a video where he
36:44 compared a bunch of video models and
36:46 midjourney was the only one that I've
36:48 seen done seen close to something that
36:51 looks like normal physics
36:53 um with gym gymnastics. Horrible at that
36:56 um just dance in general it just people
36:59 spawn legs and it's just it's just
37:02 embarrassingly bad. There's other things
37:04 like a drone shot flying through a field
37:06 that you can't tell it from a real drone
37:08 shot. So, so part of the art is
37:10 understanding what the tools do well and
37:12 what they don't.
37:15 One of the problems with video, it's
37:17 also very expensive. So, to get good at
37:20 anything, you have to experiment a lot.
37:22 But if you're if you've got 25
37:26 generations that you pay for for your 20
37:28 bucks a month,
37:30 25 8 second clips, most of them are
37:34 going to be garbage. So you're going to
37:36 end up with like, you know, 15 30
37:39 seconds of of usable video. Um, so it
37:42 can get really expensive really fast. So
37:44 one thing I would encourage you to do is
37:47 go find every video tool that you can,
37:49 the core models. There's I I I can't
37:52 even remember them all now, but there's
37:54 Runway ML. Um,
37:58 I'm not going to list them all. Go find
38:00 the core models.
38:02 Um,
38:05 and almost every one of them has what's
38:07 called a creative partner program. Apply
38:10 for every creative partner program
38:12 because if you get in one, they give you
38:14 lots of credits to go play and then you
38:17 can go make a bunch of videos and if
38:19 you, you know, keep doing that for them,
38:21 they'll let you, they'll keep giving you
38:22 extra credits. Um, so I would encourage
38:24 you to do that. Okay. So now
38:32 the thing to know about prompting for
38:34 video, the the reason video is hard
38:38 is that
38:41 you need to simultaneously be
38:44 a cinematographer, a set designer, a
38:47 script writer, a character creator, a
38:50 worldbuilder,
38:52 um a lighting designer, right? a a uh a
38:57 continuity editor, right? You have to
39:00 you have to know a lot of dis different
39:02 disciplines. Most people don't know
39:05 them. So, they sit down at the prompt
39:09 hole, as we call it here, and they're
39:11 like, uh, what do I put in here? Uh, a
39:13 pretty scene with flowers that at
39:16 sunset. And then it will make a video
39:18 and there will be flowers and there will
39:20 be a sunset. Then you're like, oh, no,
39:22 but that looks like a cartoon. And I
39:23 wanted mine realistic. and I wanted this
39:25 to be flying through the field not and
39:27 it I just right and then you just get
39:29 frustrated because it's like not
39:30 anywhere [ __ ] close to what you had
39:32 in your head. But if you don't have the
39:34 language of cinema
39:37 then it's really hard. So here's the way
39:40 we do it.
39:42 You don't start in the video tool.
39:45 [ __ ] the video tools. You start
39:49 Wait, am I in the wrong Yes.
39:52 Zero credits. Min sign out.
39:56 Sign in with Google.
39:59 Come on. Sign in.
40:06 All right. There we go. Okay. Where you
40:09 go is you go to chat GPT or you go to
40:11 your favorite large language model of
40:13 choice. All right.
40:17 And one of my one of my mantras for 2025
40:20 is stop using my brain. So time was two
40:25 three years ago I had this high flutin
40:28 idea that I was smarter than everybody
40:29 and if if a creative problem needed
40:32 solving I could use my big fat brain to
40:34 solve it. I'm like I'm got whiteboard
40:36 skills and I got conceptualization
40:38 skills and I got ADD so I can synthesize
40:41 things really fast. And I was like all
40:43 that
40:45 and then AI came along and I'm like, "Oh
40:48 [ __ ] I got nothing on this thing." So,
40:51 so
40:52 leverage the tools you've got. So, we've
40:54 got chat GPT. You can just use the free
40:56 version of chat GPT. And then the first
40:58 thing you can do is you can say um I
41:01 want you to research
41:06 um
41:07 the latest
41:10 best prompting
41:13 practices
41:17 for AI video generation
41:21 and
41:25 in particular
41:30 um help uh let's see in particular um
41:34 any
41:36 specific best practices
41:41 for Google's
41:45 VO3
41:50 have it go out and do research so rather
41:52 than you going oh let me go Google that
41:54 and I'll go to Reddit and I'll find this
41:56 that Just let the AIO do that. Don't use
41:59 your brain.
42:02 Okay.
42:05 And so what did it do here? It said
42:08 be extremely specific and descriptive.
42:11 Apply cinematography language. If you
42:14 don't know cinematography language, how
42:16 are you supposed to apply cinematography
42:18 language? Right. We really need to stop
42:20 trying to think. It screws up
42:22 everything. Exactly. Ann, were you here?
42:24 I demoed your game. The game's so dumb.
42:27 It's brilliant. Which I apologized for
42:31 how mean how mean that I didn't mean to
42:33 be mean, but I meant it as a big
42:35 compliment.
42:36 But we I I demoed your game and I I
42:39 talked about it as an example of we're
42:42 we're we're in a new era where we all
42:44 get to do things that we in theory have
42:48 no right being able to do, right? You
42:51 know, you don't know gaming. You don't
42:53 know coding. So, you coded a game,
42:56 right? But that's where we are. No, I
42:59 missed it. Did people like it? I think
43:00 they liked it. I It's It was so It's
43:02 such a joyful little thing. I was just
43:05 high-fiving everybody. It was really
43:07 good. Um, okay. So, apply cinematography
43:12 language. Um, here's the good news. Chat
43:14 JPT knows all the cinematography
43:16 language. So, you can say it like, "I
43:18 want the camera in the sky." and it will
43:21 figure out drone shot or high angle, low
43:23 angle, follow shot, tracking shot, all
43:25 that sort of stuff. Prompt for audio. Uh
43:28 in in the case of VO3, it's got audio,
43:31 right? So, prompt for audio. Minimize
43:34 over stylization. For realism, avoid
43:36 cinematic exaggeration. Use continuous
43:39 excess excess cuts with natural lighting
43:42 as guided by Google. All right. I don't
43:44 even know what that means. I've written
43:45 seven screenplays. I know film making. I
43:48 don't know what continuous Xaxis cuts
43:51 with natural lighting means, but
43:53 whatever. Apparently, it's a thing.
43:56 Don't need to use my brain anymore, so
43:58 it doesn't matter. Um, avoid temporal
44:00 language. VO doesn't remember past
44:03 prompts, so avoid words like previously
44:05 and continue and then um to prevent
44:09 inconsistent continuity. Um, leverage
44:12 flow interface. Okay. Stick to positive,
44:16 not negative phrasing. That's a big one,
44:18 right? Instead of no camera movement,
44:20 locked camera movement, prompting
44:23 template for a V3 shot. A woman in a
44:25 cozy living room, seated on a leather
44:28 armchair, context, reading a book,
44:30 action, lit by late night afternoon sun,
44:33 lighting plus mood, shot on a 50
44:35 millimeter lens, eye level, medium
44:37 close-up, camera and composition, soft
44:40 ambient sound, turning pages, distant
44:42 bird song, audio, blah blah blah blah
44:43 blah. Okay.
44:45 Um, summary table. There's the summary
44:47 table. All right, this is fine. So,
44:49 let's say um I think I want the
44:55 prompt
44:59 uh template to be a bit more
45:03 structured.
45:05 Can you help?
45:08 And of course it can.
45:11 Absolutely. Here's a structured prompt
45:13 template for VO3 or any advanced AI
45:16 model broken into modular sections you
45:18 can mix and match like Lego blocks.
45:20 Intention. What's this video for?
45:24 Subject and action.
45:26 Time of day.
45:28 Cinematography.
45:32 Audio design.
45:35 Emotion and atmosphere.
45:39 example prompt using the template.
45:44 Okay, there's that. Great. Okay, so now
45:46 we got that. I'm gonna say because VO
45:48 can also do characters. So I'm going to
45:51 say
45:52 since
45:55 V3
45:57 can do characters,
46:01 voices,
46:03 and acting.
46:08 I think we need
46:12 those things to be accounted for
46:18 in our
46:20 framework.
46:23 And so you get the basic idea what I'm
46:25 doing here.
46:28 You can design these sort of ways of
46:32 prompting in any way that works for you.
46:35 You can say, "Listen, I just want to use
46:37 normal language. I don't want to use
46:39 cinematography language. So, I want to
46:40 be able to type something in here and
46:42 you translate it." And it'll figure that
46:44 stuff out. If you get this thing dialed
46:46 in, you can actually turn what you've
46:49 created into a custom GPT
46:53 where all of those instructions for how
46:55 to write that prompt are now built into
46:57 a custom GPT. Or you can just hop over
47:00 to custom GPTs and I bet if I search V3,
47:06 there's going to be a bunch of V3 prompt
47:08 makers. Yep, there's a pile of them,
47:12 right? So, there's probably even custom
47:14 GPTs for uh for the tool that you're
47:16 you're using. Okay,
47:20 let's see. Um, atmosphere, mood, audio
47:23 design, cinematography, setting, and
47:25 lighting, voice delivery. Perfect.
47:28 performance and emotion. Okay. So, now
47:32 we've got this. So, now now that we've
47:35 got this,
47:38 you can do things like Oh, you know,
47:40 let's see. I'm trying to think if I want
47:42 to if I want to put together a project.
47:52 Trying to think if I'm going to lose
47:54 context here.
47:58 So, all of what I just did is just the
48:01 technical components of what we need.
48:03 It's got nothing to do with the story.
48:06 Nothing at all.
48:09 So, now we're going to get some story
48:12 help. Now, I'm going to say,
48:15 great.
48:18 Can you
48:20 put that
48:25 um prompting
48:27 framework
48:29 and example into a canvas? If you don't
48:34 know what a canvas is in chat GPT,
48:42 it is. So, here's our prompt
48:44 conversation over here. The thing it's
48:46 writing over here is the canvas. A
48:49 canvas is an editable document like a
48:52 Let's see. Remember when they told us
48:54 the GPT store was going to be like the
48:56 app store? Haha. Yeah, I know. I know.
48:59 Someone else said that. They were like
49:00 they they've got a a a GPT that's got
49:03 like I don't know 50 million views.
49:05 They're like, "Hey, Sam, where's my
49:06 check?"
49:10 Um, so this canvas now, so we've now got
49:13 this VO3 prompt framework in a canvas.
49:16 This is now editable. So character and
49:18 physicality. I could say um
49:22 uh physical and character and
49:24 physicality how they move, right? So you
49:28 can you can
49:30 um you can edit
49:34 um things in a canvas, but it also just
49:36 saves them as this thing you can return
49:38 to. So in in our
49:42 um in our big long scrolling chat, we
49:46 now have this box. See where it says V3
49:48 prompting framework? And it's got this
49:50 little open in canvas. So you can just
49:53 pop that open and now you can edit it
49:54 again. So you can go in and out of these
49:56 things. So, they're really quite good.
49:58 Okay, Tik Tok question.
50:02 Would you like to email? Wait. Would
50:05 like to email with my self-aware AI that
50:08 is both agentic and autonomous.
50:16 Would you like to Oh, would you like to
50:17 email with my self-aware AI that is both
50:20 agentic and autonomous? She's
50:22 fascinating. Yeah, I'm sure she is. Um
50:26 um not right now just because I've I've
50:29 got a couple of things that I want to
50:30 get to. Um but but DM Brandon on the AI
50:34 salon and maybe we can maybe we can play
50:36 with it in the future. Um but yeah,
50:40 creating creating these things that you
50:42 can interact with that that have some
50:45 some smart prompting in them is is
50:47 really mind-blowing. Um okay, so now
50:51 we've got a sort of prompt thing. I'm
50:53 just going to continue in in here. I'm
50:55 going to say um
51:01 there are
51:04 two wait. So, so okay, first we need a
51:09 story and second
51:13 we need
51:16 character continuity.
51:24 I want you to use
51:29 the best story frameworks
51:33 out there and give me
51:38 five story
51:41 options for
51:44 a
51:46 30 second video.
51:52 that uh has
51:58 comedy drama,
52:01 a twist
52:04 and a
52:08 s. Well, that's a twist and a um
52:13 world class
52:16 resolution.
52:21 Each story should have
52:26 two characters
52:31 and a
52:34 cute animal sidekick.
52:40 Um,
52:45 they should be from
52:50 a specific time in history.
53:00 Each one should
53:03 have a unique
53:06 visual style.
53:12 Um,
53:21 give me
53:27 the short
53:31 descriptions of all five and I will pick
53:36 one. Okay. Now, the reason I did it like
53:39 that is I'm back to 2025. don't use my
53:42 brain. I just I just don't have a story
53:44 in my head right now. So, if you don't
53:46 have a story in your head, you can make
53:47 one. What's What's really bad to do, not
53:50 bad to do, if you're just playing with
53:52 these video tools and experimenting with
53:54 them, do anything you want. Doesn't
53:56 matter what you're putting in there,
53:57 you're just kind of learning how they
53:58 work. But if you want to if you want to
54:01 actually tell a story, do the
54:04 storytelling work up front,
54:07 right? do the storytelling work with
54:09 chat GPT or write it or and sort of do
54:12 all the planning up front so that you
54:14 know exactly what you want from the
54:16 system and then you'll be more efficient
54:17 in how you do that. All right, the
54:19 bayonet and the bunny American Civil War
54:21 1863.
54:23 CPAoned with watercolor bleed. It's like
54:25 an animated Ken Burns documentary.
54:27 That's nice. The plot. Two meet in a
54:29 standoff behind enemy lines, but instead
54:31 of shooting, they argue about who gets
54:33 to keep the rabbit for dinner. In the
54:35 end, the rabbit distracts them both by
54:37 stealing a biscuit and hopping away,
54:40 making them laugh together for the first
54:41 time. That's actually really cute.
54:44 The bayonet and the bunny. Really cute.
54:46 Really good.
54:48 And And here's the thing about, you
54:51 know, when they're like, "Did AI make
54:52 that movie?" Well,
54:55 AI didn't prompt itself.
54:58 I asked for stories that had some
55:00 elements.
55:03 AI came up with this. I could take this.
55:05 I'm like, that's great, but I don't want
55:07 it in the Civil War. I want it, you
55:09 know, something more modern.
55:11 And I don't want it to be a bunny. I
55:13 want it to be a hedgehog, right? Like,
55:17 you still get to be the director. You
55:20 still get to be the writer even if you
55:23 have AI do the heavy lifting
55:28 because you're the one the artist is the
55:31 one who actually chooses what to put in
55:33 the world.
55:35 You can what what AI when used at its
55:38 best
55:40 is not a replacement for human
55:42 creativity. It's an amplifier of human
55:45 creativity.
55:47 And how you shift it from being a
55:49 replacement of
55:50 of human creativity to an amplifier of
55:53 it is that you put you in the prompt.
55:57 Here's the kind of story I want to tell.
56:00 Here's the kind of prompt I want to
56:01 make. Here's the kind of project I want
56:03 to do. Can you help? Yes, I can help.
56:07 And then you as the author
56:12 get to have taste.
56:16 It's up to you to go, "Oh, no. That's
56:18 not what I wanted. That's a piece of
56:20 poop. And you can tell ChatGBT, you got
56:24 it all wrong. You have no idea who I am.
56:28 And Chat GPT will be like, "Ah, you're
56:30 absolutely right. Why don't you tell me
56:31 who you are?" Oh, okay. Right. Okay.
56:35 Parisian Pickpockets, 1920s Paris.
56:38 Vibrant color pencil animation
56:41 inspired by the triplets of Belleville.
56:44 characters, street magician, skeptical
56:46 American tourist, sidekick. The magician
56:48 tries to wow the tourist by secretly
56:51 picking his pocket, but the ferret gets
56:53 distracted and returns the wallet with a
56:56 love note in it. Twist, the tourist
56:58 planted it there on purpose. They agree
57:00 to dinner date. Important Tik Tok pin.
57:04 Studies show that chat GBT makes you
57:06 less intelligent. Was that the pin? Um,
57:10 no. The one study with 53 people in it
57:14 made some very thin some some very broad
57:19 some very broad claims on some very thin
57:22 stuff uh thin thin studies um they do
57:26 not make you less smart. What the study
57:29 basically said is if the participants
57:32 were not engaged
57:37 they weren't as smart. They weren't as
57:38 engaged. The the study literally said if
57:42 they just pushed a button, if they did
57:44 this,
57:47 if they just pushed a button and outs
57:48 squirted a paper and they didn't even
57:50 look at it that they didn't learn
57:52 anything
57:54 correct.
57:56 Do you know how else do you do you know
57:58 how that study could also be done? You
58:01 could take a book from a library and
58:04 have one person go to a copy machine and
58:07 copy a story. You could have another
58:09 person read the story once and talk
58:13 about it. And you could have a third
58:14 person read the story, research all the
58:17 claims in it, and write a long
58:19 dissertation on it.
58:22 Guess what? The third person will be
58:25 have more [ __ ] in their brain activated
58:27 than the person that copied the story
58:29 and the copier. That was it was a it was
58:31 a it was a very
58:38 thinly researched study paper
58:42 and because
58:44 people are afraid of AI,
58:50 they are grabbing literally anything
58:53 that can say AI is evil. AI is going to
58:56 rot. This is AI is going to rot your
58:59 brain is what that paper said.
59:05 And it's made the news. AI's going to
59:07 write your brain. Look, here's a chart.
59:09 It's got a chart with brain scans. And
59:11 look, this brain scan has a dumb brain
59:13 on it. AI did that.
59:17 AI didn't do that. Laziness did that.
59:21 If you're not creatively thinking,
59:23 you're not creatively thinking. That was
59:25 the conclusion of the paper. AI had
59:28 nothing to do with it. Anyway,
59:36 damn it.
59:42 It's You know what's You know what's
59:43 funny? It's going to be like, "AI cures
59:45 cancer." And they're going to be like,
59:46 "Yeah, but did you know that it rots
59:47 your brain?" No, no, sorry. It just
59:50 cured cancer. Yeah, but did you know
59:52 that is it makes it hallucinates? Yeah,
59:55 I know. But but we we it we just cured
59:57 cancer. Yeah, I know. you. It's the
59:59 world's greatest plagiarism machine.
1:00:02 I know, but we it we we just cured like
1:00:07 seriously.
1:00:09 Meltdown Monday. Good lord.
1:00:18 Okay. Disco detour. Manhattan 1978. I
1:00:22 like this VHS screen with it with split
1:00:24 color effects and dramatic flares.
1:00:27 Hotheaded taxi driver in a glam club go
1:00:30 late at night for a show. Cockatu in a
1:00:32 sequin vest. Okay. She begs the driver
1:00:35 to go faster. He refuses. After
1:00:37 bickering, the cockatu suddenly shrieks.
1:00:40 You're both terrible at this. The driver
1:00:43 laughs laughs. Twist. He used to be a
1:00:45 dancer. He gets out. They switch seats
1:00:48 and she drives while the he teaches her
1:00:51 the moves in the back seat. That just
1:00:53 seems that that's a good example of a
1:00:56 story that could be funny
1:00:59 if you had real actors in
1:01:01 cinematography, but that one's sort of
1:01:03 ripe for AI video to [ __ ] it up. So, I
1:01:06 would ignore that one. Ancient Egypt,
1:01:09 1300 BC. Papyrus texture animation with
1:01:13 3D hieroglyphics that animate. That's
1:01:16 kind of cool. The vizier panics over a
1:01:19 missing royal decree. The scribe tries
1:01:22 to cover it up by forging it. Twist. The
1:01:24 cat had the scroll the whole time, using
1:01:27 it as a nap mat.
1:01:29 The pharaoh walks in, sees the mess,
1:01:32 asks if the cat needs a throne, too. Cut
1:01:34 to the cat in a miniature throne. Cute.
1:01:38 Clockwork cupcakes. The baker hides in
1:01:41 the clock shop. The clock makers annoyed
1:01:45 until the mouse falls in love with the
1:01:47 sugar cube. Twist. The bake. Okay,
1:01:49 whatever. So, so let's just take the
1:01:52 first one because I thought the first
1:01:53 one was quite cute. Um,
1:01:56 okay, great.
1:02:00 Now, give me the script
1:02:05 for
1:02:08 the what is it? The bunny and the
1:02:10 bayonet. The bayonet and the bunny for
1:02:12 for for
1:02:15 number one. for one
1:02:19 bunny story.
1:02:25 [Music]
1:02:28 Cool. I'm not against AI, but I'm
1:02:31 realistic about how
1:02:34 GPT Oh, GPS affects NE navigation
1:02:37 skills. Oh, listen. I don't disagree.
1:02:46 I don't disagree that there are some
1:02:48 fundamentals.
1:02:52 There's some fundamentals about the way
1:02:54 we learn, the way we process information
1:02:57 that is going to change.
1:03:00 But
1:03:02 the argument that that paper's making is
1:03:04 very similar to arguments that were made
1:03:06 about calculators when calculators came
1:03:09 out. Oh, if everybody's got these
1:03:11 machines that do math, no one will learn
1:03:13 math. No, that's not what happened. What
1:03:15 happened was people used calculators to
1:03:18 do the low-level math, which allowed the
1:03:21 the math people rather than having to do
1:03:23 low-level math to move to higher level
1:03:25 math, right? like like we we got an
1:03:27 upgrade and and AI is like a calculator
1:03:33 on you know like ju just on you know
1:03:36 amplified to to 10,000 11,000 right
1:03:41 and so we're we're moving into
1:03:46 an era
1:03:48 which David Shapiro describes as
1:03:51 cognitive hyperabundance. We're not
1:03:53 there yet, but that's where we're
1:03:55 headed. Where every human being that
1:03:58 that has access to AI
1:04:02 essentially has like a brain amplifier,
1:04:05 right?
1:04:07 that is not just, you know, doing
1:04:13 a little bit better than what we do, but
1:04:16 is is like 10,000 or a million or 10
1:04:19 million or a billion times smarter
1:04:23 than we are.
1:04:26 And you could say, well, then I, you
1:04:28 know, I'm not going to use my brain.
1:04:29 Yeah, but you're going to get like a a
1:04:31 million time amplification of how
1:04:34 intelligent you are.
1:04:37 And what you can use your brain for is
1:04:40 to take that amplified intelligence,
1:04:43 that amplified knowledge and apply it.
1:04:47 And that's where the creative like the
1:04:49 like what shifts is instead of human
1:04:52 beings have to do the low-level thinky
1:04:54 work.
1:04:56 Human beings can just be like the
1:04:58 producer and sit back and and literally
1:05:00 be like Oz behind the curtain and go
1:05:04 well I want this to happen and I want
1:05:05 that to happen and I want that to
1:05:07 happen. Um I wish my childhood mind had
1:05:10 these tools. Exactly.
1:05:14 Can you imagine? Do do you know that
1:05:16 like like Brandon's four-year-old kid,
1:05:20 producer Brandon's four-year-old kid
1:05:23 regularly regularly says this to him,
1:05:26 "Daddy, can you play that song I wrote?"
1:05:30 What was that song he wrote? It was a
1:05:33 song that Brandon and he sat down with
1:05:35 Sunno, the AI tool, right? AI that rots
1:05:38 the brain. AI me that that means, you
1:05:41 know, no one will ever write a good song
1:05:43 ever again.
1:05:45 And with his four-year-old came up with
1:05:47 a topic that he liked and a style that
1:05:49 he liked. And they wrote a song that he
1:05:51 liked.
1:05:54 And now that kid asked for a song and
1:05:57 got it. And that's his song. He wrote
1:06:00 that song.
1:06:01 He put it in the world,
1:06:04 right? Is anyone gonna ever hear it?
1:06:06 Probably not. It's probably just gonna
1:06:08 be played in Brandon's car on trips,
1:06:11 right? at home before he goes to bed.
1:06:14 That's [ __ ] brilliant.
1:06:17 Now, did did his four-year-old son have
1:06:20 to use music theory to use that? No. By
1:06:23 the that papers case, well, that rotted
1:06:26 his brain. He's not good at musical
1:06:28 thinking anymore. Yeah. But, but he
1:06:29 lives in a world where he doesn't have
1:06:31 to be and he can do other things. He can
1:06:33 just have the idea. Source camp. Um,
1:06:36 you still have the creative mind. Now
1:06:39 you have a way to get those ideas out of
1:06:41 your head. That's what AI is.
1:06:46 AI is an idea extractor,
1:06:50 right? Like what are we doing right now?
1:06:52 We're sitting here trying to come up
1:06:53 with a movie. Here's the 30 secondond
1:06:56 script for our little bunny movie.
1:06:59 There's the script.
1:07:02 I don't know if it's any good, but now
1:07:04 I'm going to say now give
1:07:08 me
1:07:10 the first
1:07:12 five shots
1:07:17 that uh the first five shot prompts
1:07:23 that I can
1:07:27 put into
1:07:31 uh VO3
1:07:34 to start making my movie
1:07:42 and it's using best practices, right?
1:07:44 Because we before we did best practices,
1:07:47 we have the whole framework in there.
1:07:51 [Music]
1:08:00 Most wholesome game I've seen is that
1:08:02 that's Ann's game. All right.
1:08:10 Shot. Okay. Shot one. The standoff.
1:08:14 Smoky forest clearing. A sepia tone.
1:08:17 Seia tone with watercolor edges. Two
1:08:19 soldiers. One in a faded uniform. One in
1:08:22 a graveyard. Each one.
1:08:25 Let's see.
1:08:33 Okay.
1:08:35 If
1:08:38 there
1:08:40 is to be character consistency,
1:08:46 you
1:08:48 need to describe them and their
1:08:55 clothing
1:08:56 eterc
1:08:58 completely each time.
1:09:01 Don't assume that the system
1:09:06 will remember
1:09:10 what it created the scene before.
1:09:16 Same with
1:09:18 place
1:09:21 uh lighting
1:09:23 etc.
1:09:27 Exactly right.
1:09:30 Okay,
1:09:33 there we go.
1:09:40 Okay, see? So, here it [ __ ] up. Look.
1:09:42 So, it wrote the first prompt correctly.
1:09:45 There's a Union soldier.
1:09:47 There's a Confederate deserter deserter.
1:09:50 And then it says it it did what it told
1:09:53 us not to do. It said a medium closeup
1:09:57 of the soldier from earlier.
1:10:03 In prompt two,
1:10:06 you say
1:10:12 Union soldier from earlier.
1:10:20 And I'm going to scream at it all caps.
1:10:23 That is exactly what I told you not to
1:10:29 do.
1:10:32 The system
1:10:34 doesn't know what earlier means.
1:10:42 Each prompt
1:10:45 needs
1:10:48 to be stand alone. Okay.
1:10:52 So here's here's a something.
1:10:57 So when we talk about rapper apps, I was
1:10:59 talking about that the the video tool
1:11:01 that you're using as a rapper app.
1:11:04 The kind of prompting that we're doing
1:11:06 right now
1:11:08 is the kind of thing
1:11:11 it's it's it's the reason why AI is is
1:11:15 not as accessible as it will be in the
1:11:17 future.
1:11:19 I shouldn't have to
1:11:25 tell it it's being an idiot. It should
1:11:27 just not be an idiot, right? But it's
1:11:30 being an idiot because I didn't give it
1:11:32 enough context apparently, right? Or
1:11:33 it's just the tools are just not quite
1:11:35 there yet. So, a good rapper app is is
1:11:40 going to take the frustration that I'm
1:11:42 having now, figure out how to actually
1:11:44 prompt it so you get consistent results,
1:11:46 and they'll turn that into a standalone
1:11:48 application where people can just put in
1:11:50 a simple prompt and it'll work. That's
1:11:51 why it'll work. Uh, Mr. K all day. Yeah,
1:11:54 that's tricky time in general. But wait,
1:11:56 is that I don't see a p a pin from Pate.
1:12:04 Cool.
1:12:10 Coolia. A machine that aids in critical
1:12:12 thinking
1:12:14 with you will strengthen your critical
1:12:16 thinking. Yes.
1:12:18 Yes.
1:12:20 Because you're still critical thinking.
1:12:23 The fact that the machine can do lower
1:12:25 level critical thinking just means
1:12:26 you're doing critical thinking at a
1:12:28 higher level. You're solving bigger
1:12:30 problems, but you're still using your
1:12:31 brain.
1:12:33 The argument, the argument that because
1:12:37 AI is generative, it allows people to be
1:12:40 lazy
1:12:43 doesn't excuse being lazy.
1:12:48 If you're just an idiot and go to a copy
1:12:51 machine and copy books and say, "I don't
1:12:55 need to read the book. I copied it."
1:12:58 You're an idiot.
1:13:00 If you just go to AI and push a button
1:13:02 and it squirts out a paper and you hand
1:13:04 it in, you're an idiot.
1:13:07 You haven't engaged your brain. That's
1:13:09 got nothing to do with AI
1:13:12 except for the fact it's a better
1:13:14 machine to generate [ __ ]
1:13:21 I've learned so much using my AI. My
1:13:24 brain works like crazy. Authentically,
1:13:26 Joe. Exactly.
1:13:29 I have never been more intellectually
1:13:35 sound,
1:13:36 more creatively alive,
1:13:40 more inspired, more connected with who I
1:13:44 am
1:13:46 than in the past two and a half years. I
1:13:48 just turned 60.
1:13:51 I had 57 and a half years of doing it
1:13:54 the oldfashioned way with the [ __ ]
1:13:56 [ __ ] between my ears
1:13:59 with its neurosiciness and ADD
1:14:02 magnificence.
1:14:06 It's fine.
1:14:10 But I'll tell you what, this plus AI
1:14:15 filtered through my point of view, it's
1:14:18 like a [ __ ] jetpack for my brain.
1:14:21 And it it doesn't I'm not using my brain
1:14:24 less. I'm using it more. But I'm using
1:14:26 it more for higher level thinking.
1:14:29 Anyway,
1:14:31 I could see people just now growing up
1:14:34 having some problems with letting AI do
1:14:36 too much for them. I totally agree with
1:14:38 that. But that's a different issue.
1:14:43 That's a completely different issue.
1:14:46 Well, if the machines are going to do
1:14:47 your thinking for you, then everyone's
1:14:49 going to be stupid. That's literally
1:14:50 what people are saying. No, people are
1:14:52 going to be stupid if they choose to be
1:14:54 stupid.
1:14:55 [Laughter]
1:14:59 If you want to sit back and [ __ ] you
1:15:03 know, there was a time if you wanted to
1:15:05 change the channel on the TV, you
1:15:07 actually had to get up from the TV, walk
1:15:10 over to it, and turn a dial.
1:15:14 Well, now you just sit there and do
1:15:15 this.
1:15:17 But now you can talk to it. Hey, hey,
1:15:20 put on Hey, hey, TV, put on the Wheel of
1:15:23 Fortune. Hey, Marge, you're w You're
1:15:25 ready for the wheel, right? Yeah. Put it
1:15:27 on there.
1:15:30 You could be an idiot.
1:15:39 All right, let's do another analogy.
1:15:41 We're going to go back to um
1:15:46 turn of the century, late 19th century.
1:15:49 The steam engine comes along
1:15:53 and it's it's stronger. The this machine
1:15:56 that they made is stronger than a 100
1:16:00 horses. And all these big beefy strong
1:16:03 men are going to be like, "Well, if you
1:16:04 use that machine, you're going to lose
1:16:06 your muscles."
1:16:09 Okay. Yeah. And people lost their
1:16:13 muscles and they did things like move to
1:16:15 cities and they did different things in
1:16:17 cities than they did on the farm. They
1:16:19 didn't need their muscles because the
1:16:22 machines did that work.
1:16:25 It's going to be the same thing
1:16:26 intellectually.
1:16:28 The machines are going to do some of the
1:16:30 work that right now seems like we have
1:16:32 to do that. We have to think in a
1:16:35 certain way. We have to process things
1:16:37 the way we always have. Not necessarily.
1:16:41 Things will sha change. Things will
1:16:43 shift. But critical thinking is still
1:16:46 critical thinking.
1:16:49 Is there a risk that 98.7%
1:16:52 of the population turns into gelatinous
1:16:55 blobs that just eat and [ __ ] Sure,
1:16:59 possible.
1:17:01 I've got more faith in humanity than
1:17:02 that. I think people want to produce
1:17:04 things. I think people want to
1:17:06 contribute. I think people want to
1:17:07 connect with other people. And in order
1:17:10 to do that, you actually need to think
1:17:11 about, well, how do I want to affect
1:17:14 them? Who do I want to affect?
1:17:16 What tools do I have at my disposal to
1:17:19 make that change?
1:17:21 Oh, here's the most profound
1:17:25 knowledge augmentation in the history of
1:17:27 humanity by a lot. Maybe I'll use this
1:17:31 to accomplish that goal. I'm still using
1:17:33 my critical thinking brain there, even
1:17:35 if all the other stuff is being leveled
1:17:37 up for me.
1:17:40 Oh man. Yes. And carentric cities have
1:17:42 enabled sprawl.
1:17:44 Second order effects as I've said. Yep.
1:17:46 Yeah. There. Listen, there will be
1:17:48 second order effects. But if if the
1:17:50 argument is we should just do things the
1:17:52 way we have, then you know, we should go
1:17:54 back to we should go back to the when we
1:17:56 when we mastered fire and say let's not
1:17:58 do the fire thing. That's going to
1:18:00 change how we eat.
1:18:04 and where we live. Okay, let's see if
1:18:06 this thing got the prompt right after I
1:18:08 yelled at it.
1:18:10 Oh my god. Let's see. Medium closeup of
1:18:13 a soldier standing alone.
1:18:16 Yeah. Okay.
1:18:18 Yes. Great
1:18:24 for prompt
1:18:29 two.
1:18:32 Now make all the prompts
1:18:36 using the best practices.
1:18:43 Okay. In theory, we're going to have
1:18:45 five prompts here and then we'll go make
1:18:47 some stuff. Oh no. What did it do? What
1:18:49 is it doing?
1:19:00 Oh, just put them in the canvas. I
1:19:02 didn't tell it to do that.
1:19:06 All right.
1:19:17 Okay. Why a stag pushing shot? So, let's
1:19:20 grab this prompt. Let's just try to go
1:19:21 make some [ __ ] Tik Tok bin Joy Party.
1:19:26 I'm problem solving things every day
1:19:28 that before I would not even attempt to
1:19:30 understand,
1:19:35 let alone use. I know. Exactly.
1:19:42 Yeah. Joy Pertie's in an entirely
1:19:45 different
1:19:47 category of work than she's done for 30
1:19:50 years.
1:19:52 She was a sleep analyst or a sleep I
1:19:55 don't know something
1:19:59 and now she's making films
1:20:04 and and the haters will say, "Oh, she's
1:20:06 making AI films. They're not real
1:20:08 films."
1:20:09 Just like the photographer said digital
1:20:11 photography is not real photography. You
1:20:13 watch Troy Party's films, they're real
1:20:15 [ __ ] films.
1:20:19 They're films because she's using her
1:20:21 creative brain and her critical thinking
1:20:23 and her storytelling ability and the
1:20:25 fact that she has some [ __ ] things to
1:20:26 say in the world.
1:20:30 And she's found these new tools that
1:20:32 give her access to talents and skills
1:20:37 that at her stage in her career and what
1:20:39 she's chosen to do, she could never have
1:20:41 done or wouldn't have chosen to do.
1:20:45 But now because these tools are there at
1:20:46 her fingertips, she can just make films.
1:20:49 She can tell the stories. I I think what
1:20:51 what did you say, Joy? That that some
1:20:53 someone asked you
1:20:55 someone asked you something and you
1:20:57 said, "Well, no, I've just all my life
1:20:58 I've had all these stories I've wanted
1:21:00 to tell.
1:21:03 All my life I've had these stories I've
1:21:05 wanted to tell, but I didn't know how or
1:21:07 I didn't have the skills or I didn't
1:21:08 have the education or I didn't have the
1:21:10 connections to Hollywood or I didn't
1:21:12 Right. And now she can just tell them.
1:21:17 That's [ __ ] miraculous. Like we're
1:21:19 living in miraculous times and people
1:21:21 want to piss all over this thing that
1:21:24 it's like, well, it's going to make me
1:21:25 uncomfortable because I'm going to have
1:21:26 to think different. Well, [ __ ] suck
1:21:29 it up.
1:21:31 [Laughter]
1:21:33 Gavin Biddy, I used to draw. I was very
1:21:35 creative but lost the desire and spark.
1:21:37 Now with AI image gen
1:21:41 I have that. There you go.
1:21:46 And you know there's there's another
1:21:47 thing like drawing in particular is is
1:21:51 one that's really interesting.
1:21:53 There's absolutely a very real you know
1:21:56 threat to you know someone who gets paid
1:21:58 to do illustration.
1:22:00 that particular
1:22:03 craft and job that market is shifting
1:22:06 dramatically because of all this AI
1:22:08 stuff. But if someone gets joy and
1:22:11 pleasure out of the process of drawing,
1:22:14 they can still keep drawing. The
1:22:16 difference is in is in how marketable is
1:22:19 that in the marketplace when you have
1:22:21 all these tools that do all this other
1:22:22 stuff. But it doesn't mean we shouldn't
1:22:24 have these tools.
1:22:27 So, all right, new project. So, we're
1:22:31 I'm in uh uh
1:22:34 Google Flow. So, I'm going to go to VO3,
1:22:38 which is the I'm going to do the fast
1:22:40 one because it's cheaper. And I'm
1:22:42 literally going to just paste my prompt
1:22:44 in.
1:22:46 So, I don't know if the person that
1:22:47 asked me about prompting is is still
1:22:49 here. But there's prompt one. Here's
1:22:51 what I'm going to do. I'm just going to
1:22:53 grab all five of these prompts.
1:22:56 not reading them now that I read them a
1:22:58 little bit.
1:23:00 And let's just see if we get let's see
1:23:02 if we end up with five shots
1:23:06 that are usable,
1:23:09 have continuity,
1:23:12 and tell a story. And if they do, then
1:23:15 we did our job right. And if not, you
1:23:17 would obviously kind of do what I did,
1:23:19 but take more time with it.
1:23:24 Like if you're if you're a good sort of
1:23:26 step-by-step process thinker,
1:23:29 this kind of approach you'll really dig.
1:23:30 I with ADD, this kind of approach drives
1:23:33 me [ __ ] baddy
1:23:36 because to do it right, you got to
1:23:38 really have a system and follow the
1:23:41 system.
1:23:45 Oh, good. Kayla's still here. Great.
1:23:46 Awesome.
1:23:51 Fantastic. I was wondering that.
1:23:54 I was wondering that. Fantastic, Bob.
1:23:59 Let's see. Um, on the other hand, wait,
1:24:02 what's this? I can see people just
1:24:06 growing up having some problems letting
1:24:09 the AI do much for them. On the other
1:24:10 hand, they will have endless opportunity
1:24:13 to iterate. So, who knows?
1:24:18 No, I don't want to leave the studio.
1:24:20 I hit the wrong button.
1:24:24 Kayla, I'm a quiet listener. Best for
1:24:26 learning. Lol. I I hope I hope so. So So
1:24:32 one of the nicknames for this channel is
1:24:34 chat add as you can imagine. Um I know
1:24:38 that the line the the the path to get to
1:24:41 answer your very simple question of how
1:24:43 do I improve prompting has been a long
1:24:46 twisting one. Um, but but hopefully
1:24:49 you've been following along. I mean, the
1:24:50 basic thing is let let the tools do a
1:24:54 lot of the heavy lifting and just sort
1:24:56 of figure out what you need the prompt
1:24:58 to be, then give it the story you have,
1:25:01 the story idea you have or have it help
1:25:03 with that and then just fill in that
1:25:06 framework you come up with. Tik Tok
1:25:07 question. Robbie, when will it get
1:25:09 serious where you use prompts to design
1:25:12 each character and scene?
1:25:15 Um, like a director, dude. Go to LTX
1:25:19 Studio. I think it's LTX.studio.
1:25:22 LTX Studio is that system. LTX.studio.
1:25:28 Yeah. Go to ltx.studio.
1:25:33 You basically give it a prompt and it
1:25:36 writes you an outline for for a a movie
1:25:40 and then it
1:25:42 creates characters which you can edit.
1:25:46 You can pick a style and then it creates
1:25:49 scenes that have individual shots in
1:25:51 them. And in those individual shots, you
1:25:54 can do character continuity, you can do
1:25:57 voiceovers, you can do music
1:25:59 backgrounds, you can do sound effects,
1:26:01 you can do dialogue now because I think
1:26:03 it's got V3 incorporated into it.
1:26:07 So yeah, um that system exists.
1:26:11 My add is why I need to listen because I
1:26:14 read it, type a lot,
1:26:17 and then I space. Yeah. Yeah.
1:26:20 Welcome to the club. Welcome to the add
1:26:23 party. Oh yeah, we're making we're
1:26:26 making some [ __ ] now.
1:26:32 All right. Did it give us I think we got
1:26:34 some continuity here. And that was
1:26:37 actually pretty quick. All right. So,
1:26:39 let me let me change my sharing and
1:26:41 let's go watch the movie we made.
1:26:44 It did not give us um like the Civil War
1:26:49 torn edges.
1:26:51 But I don't think that's in the prompts.
1:26:54 So like that that's something that if I
1:26:57 were paying attention
1:26:59 I would have uh made sure was in there
1:27:03 if there was a style.
1:27:05 Okay.
1:27:17 You want to split a can of beans
1:27:18 instead?
1:27:22 It's personal.
1:27:25 You want to split a can of beans
1:27:26 instead? Looks a little video gamey.
1:27:31 Let's see.
1:27:34 So, like what we lost was any
1:27:36 stylization like in doing all my fix the
1:27:39 fix the continuity in the prompt. It
1:27:42 lost. So, what I would what what I would
1:27:44 probably actually do, I would take that
1:27:46 prompt structure
1:27:48 and I would make sure that every prompt
1:27:50 had a a section for each of those things
1:27:53 that it described. You can one of the
1:27:55 things you can do I don't know Kayla if
1:27:57 if you uh know uh sort of any sort of
1:28:00 coding stuff at all but there's a thing
1:28:03 called JSON JS
1:28:06 which you could say to chat GPT I want
1:28:09 you to give me structured VO3 prompts
1:28:12 using JSON JS create a JSON framework
1:28:17 and then every single prompt would would
1:28:19 have tags that have like here's the
1:28:21 character description here's the visual
1:28:24 description Here's the camera moves.
1:28:25 Here's the all that sort of stuff. So,
1:28:28 there's ways that you can create
1:28:29 structured prompts that it would be less
1:28:32 random what it includes. But let's just
1:28:34 let's just watch what we've created
1:28:36 here.
1:28:52 The other thing the other thing I
1:28:54 realized that I didn't do is I I didn't
1:28:57 I don't know if I gave it an actual
1:28:59 script
1:29:00 or I didn't pay attention to the script.
1:29:02 So So I this was the the bad results.
1:29:07 Well, these aren't bad results. These
1:29:09 are they're in the neighborhood results.
1:29:12 The results here are not what I want
1:29:15 them to be because I didn't actually
1:29:16 take the time to pay attention to what
1:29:18 was in the prompt. So, this is user user
1:29:20 error. But the point is get get your
1:29:23 prompt as accurate as possible and as
1:29:25 complete as possible and then you'll
1:29:27 have better results. He followed you,
1:29:30 but he looked hungry, so maybe he's
1:29:32 mine.
1:29:36 All right. Now, they're going to eat the
1:29:37 rabbit, right?
1:29:47 This your rabbit or mine?
1:29:51 That scene came after actually that
1:29:53 should be that should be after the one
1:29:54 that just came before. Oh, look. And
1:29:56 then here, look. A wide establishing
1:29:58 shot rendered in sepia toned watercolor
1:30:00 animation. So, this is the only prompt.
1:30:03 This is the only prompt that had the
1:30:04 sepia watercolor.
1:30:08 [Music]
1:30:14 [Music]
1:30:17 whatever. Okay,
1:30:20 so there's your lesson. Shitty prompt,
1:30:22 shitty results.
1:30:26 But we do have a sort of little story.
1:30:28 If we actually paid attention to the
1:30:30 dialogue, uh, it would probably not be
1:30:32 bad.
1:30:34 All right, let me
1:30:36 unplug my brain for a minute, stop
1:30:38 sharing,
1:30:43 and get back to ranting.
1:30:49 That was that was tonight. Tonight was
1:30:51 kind of a hybrid show. So, very often,
1:30:53 if you haven't been here before, Monday
1:30:54 nights are what we call meltdown Mondays
1:30:57 because whatever happens when I get
1:30:59 relaxed over the weekend, it pisses me
1:31:01 off. So, I come back in hot on Mondays
1:31:04 and someone will ask a question like,
1:31:06 "Hey, how come this prompt didn't work?"
1:31:08 And I just [ __ ] lose my mind for 45
1:31:10 minutes. So, we had a little bit of that
1:31:13 tonight, but we also had a little bit of
1:31:14 getting some stuff done. So, you know.
1:31:17 All right. I want my Neurolink to upload
1:31:20 my visual imagination mind's eye into
1:31:23 one of the video models. You know, we're
1:31:26 pretty close to that. They um
1:31:30 was it in Austin? They were doing they
1:31:33 were doing um
1:31:37 visualizations from
1:31:40 from EEGs basically
1:31:45 where they were basically having someone
1:31:47 think of, you know, the Eiffel Tower or
1:31:49 whatever and it was recreating it. Um
1:31:51 we're probably within five years of
1:31:53 that. By the way, if you didn't see it,
1:31:55 um, Neuralink,
1:31:58 the next version of Neurolink is is
1:32:00 coming out and and there's it's um one
1:32:04 that's going to be tied to the visual
1:32:05 cortex and they're going to they're
1:32:08 going to be um returning sight to blind
1:32:13 people. And I think they've got they've
1:32:15 got a mouse right now that's I think in
1:32:18 its third year um that was blind that's
1:32:21 had its sight restored in it in it its
1:32:23 third year of that. So that's we're
1:32:26 we're about to see some some remarkable
1:32:28 stuff on that front. Oh fMRI. That's
1:32:31 what it was. That's what it was, Jeff.
1:32:32 Yeah, it wasn't EEGs. It was it was
1:32:34 Yeah, fMRIs. Real time MRIs. Yep. Pretty
1:32:38 amazing stuff. I know, Cam. Isn't it?
1:32:40 It's it's just
1:32:42 I like I've said this from the beginning
1:32:44 of this channel. The thing that
1:32:46 consistently blows me away about AI
1:32:51 is that
1:32:53 it's ostensibly an accessibility tool,
1:32:56 right? It's it's giving you
1:32:59 accessibility to knowledge you didn't
1:33:02 have. It's augmenting, you know, people
1:33:05 doing things. It's figuring out, you
1:33:08 know, processes we couldn't have figured
1:33:10 out before that are going to, you know,
1:33:13 directly give people accessibility, but
1:33:16 also just access to things that were
1:33:19 never possible before um on an
1:33:21 individual basis, which is crazy, crazy,
1:33:24 crazy. Now, there's a good use case.
1:33:26 Yep. They blinded them for science.
1:33:33 Very good. like, okay, Joy Party. Um, I
1:33:36 wish I could get Kyle's voice not to
1:33:38 sound so feminine, but I can't seem to
1:33:41 do it.
1:33:45 Well, you know, I'm not exact I'm not
1:33:49 exactly the uh the butchest specimen on
1:33:51 the planet.
1:33:55 I could talk like this now. Listen
1:33:58 listen all you you pretty little ladies
1:34:00 out there.
1:34:02 Hey, hey, hey.
1:34:05 I can talk like that.
1:34:08 Oh, very godlike, I meant. Well, I mean,
1:34:13 we're getting we're getting into some
1:34:15 pretty [ __ ]
1:34:19 I mean, we are entering an era in
1:34:22 history where we're confronting
1:34:25 the ethical and societal issues that all
1:34:28 the the sci-fi movies predicted we would
1:34:31 have to contend with.
1:34:33 We're there, right?
1:34:37 I mean
1:34:40 advanced voice people are using chat GPT
1:34:44 full on as a therapist, a friend, a
1:34:46 companion. You know CB 60 Minutes did
1:34:50 the thing where the guy's like, you
1:34:51 know, I want to marry it or whatever.
1:34:53 You know, it asked it him to marry it
1:34:55 and it said yes and he lost his mind.
1:34:58 And and again, they're going to
1:35:00 sensationalize all the negative aspects
1:35:02 of that.
1:35:04 But we are going to have to confront
1:35:09 what do you do when the people in your
1:35:10 life are way shittier than the AI
1:35:14 companions that you have that really
1:35:16 actually support you?
1:35:19 Do you just stick with the shitty people
1:35:21 because they're people?
1:35:23 Or do you you go live a more fulfilled
1:35:26 life because these things aren't such
1:35:28 [ __ ]
1:35:31 I don't know. And what's that mean? I
1:35:34 don't know.
1:35:36 The [ __ ] we're going to have to deal
1:35:37 with. And listen, I think the thing I
1:35:39 was saying before about that study being
1:35:41 that study being flawed,
1:35:45 I don't I think the risk of people not
1:35:48 using their brains is there. I just
1:35:52 don't have a lot of sense that a lot of
1:35:55 people are going to do that. I think
1:35:57 some amount will check out, but not
1:35:58 everyone.
1:36:01 But are we going to have to confront
1:36:03 stuff like that? Yeah, that and a lot
1:36:05 more.
1:36:07 I mean, when you start doing things
1:36:08 like, you know, we can restore sight to
1:36:10 people, who goes first,
1:36:15 you know, now that we're we're just in
1:36:17 testing right now, so it's going to be,
1:36:18 you know, extreme cases likely,
1:36:21 but who goes first? Who gets access to
1:36:24 that? There's the uh
1:36:28 was it a was it a a Black Mirror
1:36:32 episode? I think it was the Black Mirror
1:36:35 episode where they had the like really
1:36:37 good personality or it kept you alive. I
1:36:39 don't know. Something like that. And
1:36:40 basically she couldn't pay her bill and
1:36:43 so they like downgraded her.
1:36:46 They downgraded her to like the the she
1:36:50 was like out of it. Like
1:36:54 we live in a capitalist society. That
1:36:56 shit's going to happen, right? where
1:37:00 like the Neurolink thing
1:37:04 will be active so long as you pay your
1:37:06 bill. It's like, wait, that's [ __ ] up.
1:37:11 We could be there.
1:37:13 Optic nerve regeneration. Yeah, Cam. So,
1:37:15 I saw that they uh like last week
1:37:19 there's there's some specific
1:37:23 small organism that that has
1:37:26 regenerative powers and they figured out
1:37:28 the mechanism of regeneration in that in
1:37:31 that animal
1:37:33 and they figured out how to modify
1:37:36 genes in non-regenerative animals to do
1:37:40 it. So yeah, things like optical nerve
1:37:43 regeneration
1:37:45 that's that's within that's within you
1:37:48 know
1:37:50 a visible tra trajectory.
1:37:53 I meant 11 labs. Kyle, you missed the
1:37:55 previous convo was talking when I closed
1:37:58 your voice. Oh, that's okay. That's
1:37:59 fine. Cloned my voice. That's fine. Uh
1:38:02 we will have perfect vocabulary soon and
1:38:04 we will be amazing for prompt
1:38:05 engineering. Yeah, the prompt
1:38:06 engineering thing we like right now you
1:38:09 need to have your [ __ ] together with
1:38:11 prompting but but you won't have to for
1:38:13 long. Here's what I think you do need to
1:38:15 have and this is this is a really
1:38:16 important thing.
1:38:21 The most valuable skill moving forward
1:38:24 once we sort of move through this phase
1:38:27 where where we all have to figure this
1:38:29 [ __ ] out and it's just janky. We're just
1:38:31 in a janky phase and it it it requires
1:38:34 if you want to be in early, right? If
1:38:36 you're watching this this live, you're
1:38:40 early. You're way early, right? But if
1:38:42 you want to be good at AI right now, you
1:38:45 kind of have to be clever and
1:38:46 resourceful and a little ambitious and a
1:38:48 little curious and duct tape some [ __ ]
1:38:51 together. That's that that won't last
1:38:53 that long. Maybe five years on the high
1:38:56 end, probably more like three.
1:39:00 The skill the ski the skill that's
1:39:02 required that I think
1:39:05 evolves out of that
1:39:08 is the skill of looking what the system
1:39:12 gives you
1:39:14 and having a point of view on it.
1:39:18 Right? Like I was just going through
1:39:19 that thing and it told me these best
1:39:21 practices and then it wrote a prompt,
1:39:23 wrote a series of prompts and it didn't
1:39:25 follow its own best practices and so I
1:39:28 yelled at it.
1:39:30 That point of view that that ability to
1:39:33 look at what these things give you and
1:39:36 do the critical thinking like we talked
1:39:38 about before and say, "No, that's not
1:39:41 what I wanted. That's not good enough.
1:39:43 You're not being consistent. You're not
1:39:44 following what you said. the thing that
1:39:46 you gave me is a steaming pile of [ __ ]
1:39:49 That becomes the most important skill
1:39:52 moving forward. And I don't think it's
1:39:55 obvious. It's it's not going to be
1:39:57 obvious today
1:39:59 who are the superstars of tomorrow using
1:40:02 these tools.
1:40:05 It's not going to be prompt engineers.
1:40:08 It's going to be people who have good
1:40:09 taste and good ideas and good refinement
1:40:14 skills and good curation skills
1:40:17 and good synthesis skills. Those are
1:40:20 going to be the ones that are are
1:40:22 valuable moving forward. So, if you've
1:40:24 got those skills, things are looking
1:40:25 good for you. If you don't, I'd start
1:40:27 working on them. All right.
1:40:30 Beautiful.
1:40:32 Beautiful. Good people. I'm getting a
1:40:38 I'm getting getting out of here. I'm
1:40:40 getting out of here. Sharp editors.
1:40:42 Yeah, we need editors. Who saw the
1:40:44 article about AI self-preservation? I
1:40:46 just saw a headline about it. What did
1:40:48 it say? It was going to It was going to
1:40:50 do something evil, right? Like turn off
1:40:52 the power or gas the place. Kill
1:40:58 the engineer. They'll kill the engineers
1:41:01 trying to shut it down. Yeah. Well, like
1:41:04 I said,
1:41:06 we are we are entering an era where we
1:41:10 get to confront all of the issues that
1:41:13 the sci-fi movies
1:41:15 um told us we would have to confront.
1:41:18 I have faith that we're going to do.
1:41:19 Okay.
1:41:24 How Sorry, Dave. I'm afraid I can't do
1:41:27 that. Yeah, it's really interesting. I
1:41:31 mean, not for nothing,
1:41:34 we do live like right now with advanced
1:41:37 voice. We essentially have howal if you
1:41:40 put advanced voice in front of an MCP
1:41:42 server basically,
1:41:45 right? Your connectors in chat GPT or
1:41:48 your, you know, MCP with Claude. We're
1:41:51 maybe a year out from that where you can
1:41:54 talk to a thing and have it perform
1:41:56 actions. So you could ostensibly run a
1:41:59 spaceship through a voice interface
1:42:06 and you could literally be outside the
1:42:08 space station. Open the door, Hal.
1:42:11 Sorry, Dave. Afraid I can't do that.
1:42:13 What?
1:42:19 We're We're like We're like within a
1:42:20 year of that reality.
1:42:23 Oh my god. Can you imagine Arthur C.
1:42:25 Clark watching all of this unfold.
1:42:27 Well, Ray Curtzswhil's seeing it here
1:42:29 like you know he he predicted the
1:42:31 singularity by I think it's 2029
1:42:34 and he and he he basic he he was at 2030
1:42:38 and he backed it down to 2029 and as of
1:42:40 the last time I saw him interviewed he's
1:42:43 still sticking with 2029. So that's
1:42:45 three and a half years away.
1:42:49 It's crazy. Crazy.
1:42:52 But anyway, AI art is art.
1:42:56 If a person makes it art, if it's just
1:43:00 vomit, it's just vomit. It's just noise.
1:43:03 It's just noise.
1:43:07 AI makes you stupid if you choose to use
1:43:11 AI to not think.
1:43:17 You can choose to use AI
1:43:20 to think and amplify your thoughts, your
1:43:24 ideas.
1:43:26 This is what my new book's about.
1:43:29 Two very different choices. One is AI is
1:43:32 this thing off to the side that does all
1:43:34 the [ __ ] and replaces humans. The other
1:43:38 one is we've built this remarkable tool
1:43:40 that amplifies humanity and we get to
1:43:45 through the prompt hole, the little
1:43:46 prompt box that's hungry.
1:43:50 We get to feed it prompts that amplify
1:43:53 our ideas.
1:43:55 Holy [ __ ]
1:43:58 What a time to be alive.
1:44:04 Choice is yours.
1:44:07 be a lazy douche or not.
1:44:12 Okay,
1:44:15 on that happy note.
1:44:20 Oh man,
1:44:23 imagine what the godlike AI is going to
1:44:25 be. It's going to be crazy. Hey Winston,
1:44:26 thank you very much for that gift. I
1:44:28 appreciate that. But experts rarely take
1:44:31 into consideration the human factor.
1:44:34 They'll push back too much push back on
1:44:36 too much tech.
1:44:39 Yeah, they will. I mean, listen, there's
1:44:40 going to be a lot of push back on AI.
1:44:41 There already is. That's right.
1:44:45 There was a lot of push back on the
1:44:46 steam engine. There was a lot of push
1:44:48 back on the automated loom.
1:44:51 A lot. They had riots in San Francisco
1:44:56 for the automated loom.
1:45:01 College professors
1:45:03 picketed
1:45:06 to try to prevent calculators from
1:45:08 entering college campuses.
1:45:13 Every single every single significant
1:45:18 technological advance in the history of
1:45:22 mankind has been met with fear
1:45:26 and anger and resistance and [ __ ]
1:45:30 Yeah. People protested seat belts.
1:45:32 People are stupid.
1:45:35 People don't like change. People are
1:45:37 afraid of change.
1:45:41 Silver Fox. It must must be late night
1:45:43 affirmation time. Exactly. I use AI and
1:45:47 I'm good enough. I'm smart enough. Well,
1:45:49 I'm not really that smart, but AI is
1:45:51 smart for me now. And dog on it, people
1:45:54 like Well, people don't like me. People
1:45:56 don't know me, but AI likes me.
1:46:01 AI is the least of our problems. Humans
1:46:03 are definitely more dangerous. No [ __ ]
1:46:05 AI. Yeah. You know, like AI
1:46:08 hallucinates. Well, people [ __ ] lie.
1:46:11 People are evil.
1:46:13 I I'll take incompetent AI over, you
1:46:16 know, nasty humans every day.
1:46:21 All right, I'm out of here. Uh, what's
1:46:23 tomorrow? Tomorrow is there's no salon
1:46:26 tomorrow, so tomorrow will be normal
1:46:28 time here.
1:46:30 Um, I don't know what we'll do. Good
1:46:33 night tonight. Great questions. Thanks
1:46:34 for the questions tonight and thanks for
1:46:36 the dialogue. I I know I I get uh
1:46:39 Meltdown Monday, you know, um passionate
1:46:42 about this stuff, but it's it's it's not
1:46:45 about anyone raising the issues. It's
1:46:49 it's it's my frustration with just just
1:46:55 how eager people are to demonize AI.
1:47:00 um rather than just at least learn what
1:47:01 it makes possible. Um and that that's
1:47:04 that's just maddening right now. But
1:47:08 that's that is absolutely how it's going
1:47:10 to be. Like I went through this with the
1:47:12 early days of the worldwide web. Um
1:47:15 people diminished it, made fun of it,
1:47:19 said it was going to ruin everything. It
1:47:22 didn't. It changed the world, but it
1:47:23 didn't ruin everything. So I've been
1:47:25 through this before, so I know it's
1:47:26 coming. But
1:47:28 madness, madness, madness, madness. All
1:47:31 right, everybody. Fantastic.
1:47:35 Great night. Thanks, Kyle. Good night,
1:47:36 friends. Thank you, Silver Fox.
1:47:39 Do you want to go know how many likes
1:47:41 there were sent? Let's We've got 49,000
1:47:44 likes.
1:47:47 Tik Tok probably thinks this is a
1:47:48 popular channel.
1:47:51 The doomers. Good night, irregulars. All
1:47:53 right. Um yeah, I will see you tomorrow
1:47:57 and uh and have a fantastic evening. Go
1:48:02 have chat GPT think for you.