8/4/2025 - Navigating the Jank: AI Tools and Chain of Craft

Description

Chapters

Transcript

0:00 [Music]
0:10 She came on here like slow moving color
0:14 front,
0:19 but his beard was warmer than a look in
0:22 her eyes.
0:27 She sat on a stool and he said, "What do
0:31 you want?"
0:35 >> She said, "Give me a love that don't
0:38 freeze up inside.
0:40 [Music]
0:42 [Applause]
0:45 [Music]
0:48 I have melted some in my time, dear.
0:55 Bugs are sitting next to you and I
0:58 shiver and shake.
1:03 And if I knew love, well, I don't think
1:06 I'd be here.
1:10 [Music]
1:11 Asking myself if I've got boo
1:15 [Music]
1:20 your icy blue heart.
1:23 [Music]
1:24 Should
1:27 I stop?
1:30 Turtle's been frozen for years.
1:36 Into a river of tears.
1:40 Yay. Yay.
1:43 That's it. Shamp. No more singing.
1:47 [Music]
2:07 You want me to move that wire? Is that
2:09 wire in your way? Let me get rid of
2:11 that. There you go.
2:15 Good evening, good people. What's
2:17 shaking? What's going down? What's
2:19 happening?
2:21 We're back to our regularly scheduled
2:25 Quad Stream. It's the patented Quad
2:28 Stream technology patented by Kyle
2:30 Shannon. Hi, I'm Kyle Shannon.
2:34 Um, yeah, Quadream. We're on Tik Tok.
2:38 We're on YouTube. We're on LinkedIn.
2:40 We're on X, formerly known as Twitter.
2:42 So, it's almost five
2:48 now with 100% more jank.
2:51 It's a good one.
2:57 Now, even jankier. Um, do me a favor. If
3:01 you're on TikTok, you know what to do.
3:03 Share the live. Tell your friends. Call
3:06 your friends and neighbors who hate AI.
3:09 They're our favorites.
3:12 I'll tell you what, when you take
3:13 someone that hated it but has never
3:16 tried it and then they try it and
3:18 they're like, "Oh my god, I didn't
3:20 realize it could do that."
3:22 That's a good day. That is a good day,
3:26 my friends. Good, good, good day. Where
3:28 is my mouse? There it is. Okay,
3:32 that's all going there. Cam Ken, how are
3:34 you tonight? I am good. It is meltdown
3:36 Monday.
3:38 We'll see if anyone asks a question that
3:40 sends me into a tizzy.
3:43 [Music]
3:58 Sitting in this lonely town
4:01 wondering things are going to change.
4:07 Dream my life away.
4:10 Seems these dreams have turned to a
4:12 bunch dust clouds.
4:16 Get my nerve up. But my past is pulling
4:20 me down.
4:24 [Music]
4:25 Somebody tell me once before. Oh no,
4:28 wait. That's not right.
4:32 I haven't sang that song too long. I
4:34 don't know them dare words. Words is
4:36 hard, people. Listen, words is hard.
4:41 Have you ever tried using words? Words
4:43 is hard.
4:46 I don't know how to use words. Here's
4:48 the thing about AI. You got to be pretty
4:50 good at words. Unless you don't have to
4:53 be.
4:56 See? See what I'm saying?
4:58 It's a word thing. If you're good at
5:01 words, then you use your words. But
5:03 sometimes you don't need your words.
5:05 then you don't use use your words.
5:11 My name's Kyle Shannon and I'm an expert
5:13 on AI with words. You put the words in
5:18 the hole chat hole. You put the words in
5:21 the chat hole and you push the button.
5:25 That'll be $149.99.
5:28 $149.99 for put the words in the chat
5:31 hole.
5:36 AI master class right here.
5:52 [Music]
5:56 Source camp is here. We can start.
6:06 We can create smut on grock.
6:09 Yay, we did it. We can finally create
6:13 our own
6:15 prawn
6:17 cuz you don't want to say the P word.
6:21 Cam Ken, did you read the book about Sam
6:23 Alman? Opt optimistic by Kee Haggy. I
6:27 have not. I I am generally not a reader.
6:30 What I will commit to is asking Chad GPT
6:33 about it.
6:38 Corn. Yes, corn. We can finally make our
6:42 own corn chowder on Grock. Thank
6:45 goodness.
6:47 [Music]
7:21 Yes, Kyle, did you see that you got
7:23 asked to be part of Pika's creator
7:25 program? Yeah, that's awesome, Danielle.
7:27 I I don't think I'm a part of that one.
7:29 I'm not really sure. I don't remember.
7:31 Kyle, so what do you think?
7:34 Which numbers are higher? Folks that
7:36 don't like AI
7:39 or folks that are scared of it?
7:43 And any solutions for both? That's
7:45 interesting. I think there's a third
7:46 category.
7:49 So, folks that are scared of it. Wait.
7:51 Yeah, folks that are scared. folks that
7:55 don't like it for whatever reason,
7:57 right? They've taken a stand. It steals.
7:59 It's the world's greatest plagiarism
8:00 machine. All them students are cheating.
8:02 We're not going to let them use AI
8:04 because all them students are cheating.
8:08 Hey, question. What technology you think
8:10 they're going to use when they graduate
8:11 college? Well, AI. Well, why don't you
8:13 teach them that in school? Because it's
8:14 cheating.
8:18 I don't know why school administrators
8:20 always talk with their bottom lip out
8:21 like that. It's weird.
8:24 It doesn't make him look very smart.
8:26 Anyway, you got those two and I say
8:29 there's a third category which is they
8:31 haven't heard of chat GPT. They don't
8:34 know. They have no [ __ ] clue.
8:40 Um,
8:42 I I would argue I would argue that the
8:49 people that are afraid of AI and the
8:51 people that are angry at AI are roughly
8:53 the same crowd. If they're if they're
8:55 aware of it, you know, they're either
8:57 like, "Oh, the robots are going to kill
8:59 us," or, "Oh, it's, you know, it's it's
9:01 ruining everyone's lives."
9:06 I would call them the conscientious
9:07 objectors.
9:10 Here's my guess.
9:12 And I'm sure there's stats on this right
9:14 now. Eight minutes in. Meltdown already.
9:19 Yeah, good question. Ocean. Way to get
9:21 me way to get me fired up. Eight minutes
9:23 in. Um,
9:32 so I think chat GPT has 700 million
9:34 users.
9:38 I would argue
9:42 just judging just judging by the
9:44 reaction I got from the crash course
9:47 last night la last week when we did chat
9:50 GPT night.
9:52 A lot of the audience has used chat GPT
9:54 and a lot of the audience learned
9:57 something new that it did Tuesday night.
10:03 Um,
10:05 so
10:09 my guess is that
10:11 maybe maybe
10:15 10% and I don't think it's that high. I
10:18 still think it's single digits. I think
10:20 8% maybe
10:23 of people using generative AI, chat GPT,
10:27 things like that are using it on a
10:28 regular basis.
10:31 I think there are a large group of
10:33 people that don't hate it or maybe are a
10:35 little bit afraid of it. Say that they
10:37 use it but really don't use it kind of
10:39 like a Google search. They occasionally
10:41 use it. That's pro that probably takes
10:44 us to
10:46 30%. Right? So less than 10% actually
10:50 use it. Probably less than 5% really use
10:53 it. Less than 10 use it. Up to 30 eh
10:58 don't really use it.
11:01 And then I think, you know, and then I
11:03 think you've got
11:05 people who've maybe tried it once. You
11:07 got you get up to like 50% of people
11:10 that are aware of it. And then and then
11:11 you get into that range where
11:14 you know people who've never heard of
11:16 it. Last November it was 45%.
11:20 45% of Americans
11:24 last November, so coming up on a year
11:26 ago, had never heard of Chat GPT. So
11:29 maybe that number is down to 30. So you
11:32 got 30%
11:34 that don't even know what it is. You got
11:37 another 20% that are aware of it, but
11:41 they're sitting it out. I'm scared of
11:43 it. I'm pissed off at it. It's going to
11:46 take my job. I just don't want to deal
11:48 with it. And then you have like half are
11:50 kind of dealing with it. But I mean, if
11:53 you're talking about 5% actually using
11:55 it, it means that we're still incredibly
11:59 early.
12:00 Looks like we've got someone from Kenya
12:03 checking in on the live. Welcome,
12:04 welcome, welcome.
12:08 So, I don't know if that answered your
12:09 question. Oh, wait. What's it say? Oh,
12:12 any solutions for both? Yes.
12:17 The absolute key to getting people
12:21 who are afraid of AI or pissed off at
12:23 AI.
12:25 To use AI is to get them to use AI.
12:31 And I wish it were different than that.
12:33 It's not. You You can't explain this
12:36 stuff
12:37 until someone has had their own Kevin
12:40 Mallister moment. Until they've had that
12:41 moment where they're like, "Oh my god, I
12:43 had no idea."
12:48 And I think it's also, you know, one of
12:50 the things we talked about, I think on
12:52 Monday was,
12:54 yeah, when we were talking about the
12:56 difference between
12:58 um traditional AI and post chat GPT AI,
13:02 most people that are not dealing with AI
13:06 probably think they need a math degree
13:08 to be any good at it. that they don't
13:10 actually they haven't actually taken in
13:13 that you can just use your words
13:16 even though words is hard.
13:18 [Music]
13:20 I have a 70-year-old in my housing
13:23 household totally addicted to AI now.
13:26 That's the thing, man. I'll tell you
13:28 what, when the skeptics get converted,
13:31 they're they're like, you know, you know
13:33 what it is? It's like um it's like it's
13:36 like uh the ones that are like I'll
13:39 never be an AA, they they become like
13:41 the biggest AA, you know, evangelists.
13:44 It's it's like AI
13:46 AI AA
13:49 that the ones that said they hated it
13:51 the most, those are the ones that are
13:52 like, it's really good. It's really
13:54 amazing. It can really light some things
13:56 up.
14:00 Yes. He says he doesn't know what he
14:03 would do without it at this point. Yeah,
14:05 there you go. What's happening, Moosh?
14:08 What I love about this channel is we
14:10 actually use the tools. Most others just
14:12 talk about theory. Well, listen,
14:15 not for nothing, Kelly. I can talk
14:17 theory.
14:19 When when considering the when delving
14:22 deep into the tapestry of the paradigm
14:26 of artificial intelligence,
14:29 one can surmise deeply deeply within the
14:34 profound impacts.
14:37 We can do theory here. I'll talk theory
14:40 as if I have qualifications. I'll do it.
14:43 You know, I'll do it for a flipping
14:45 hour.
14:48 You want professorial? I'll give you
14:50 professorial. And not just because my
14:52 eyebrows look like this. I've actually
14:54 trimmed them. Hey, anybody noticed I got
14:56 a haircut? Do I look pretty? Do I look
14:58 cute? Do I look cute?
15:02 Tell me I look cute.
15:04 [Music]
15:07 You know what I'm cranky about? None of
15:10 you are regulars noticed my haircut.
15:15 Oh, nice cut.
15:18 Yeah, Moosh. Moosh comes in. Moosh, you
15:21 started here with the with the AI crash
15:23 course, right? Moosh comes in and he's
15:26 he's paying the compliments. All you are
15:29 regulars already take me for granted.
15:35 [Music]
15:50 Um,
15:52 so if so if you did come here through uh
15:56 through the back to basics crash course,
15:57 couple of things. Let me share something
15:59 with you. If you could pop open the
16:02 share screen there, Yan Brandon, I don't
16:04 know if you're there. You there?
16:08 All right. So, if you are um
16:15 if you caught
16:17 some of the week last week for the the
16:19 back to basics AI crash course or you
16:22 want to rewatch them, the the fiveday
16:25 sequence is available on demand. So, if
16:27 you go to the AI learning lab website,
16:30 so it's learning lab-ai, that's the
16:33 that's the the handle. It's got this
16:36 beautiful
16:37 image that I made with stable diffusion
16:40 in 2022.
16:42 That was one that almost made me cry. It
16:44 took my breath away when I made that
16:45 image. And then you'll notice here
16:47 there's so if you haven't subscribed,
16:50 just subscribe. Subscribing is free.
16:52 That's how you get notified when I go
16:54 live. To the right of it is the join
16:56 button. So, if you want access to the on
16:59 demand, um, you can either click the
17:01 join button or you can click the back to
17:04 basics thumbnail here and it will take
17:06 you in and say, "Oops, you've got to be
17:08 a a member or you've got to join the
17:10 channel. You got to be a member um of
17:12 the workshops and highlights level,
17:15 which is $9.99." Okay. So, if you want
17:18 the on demand from last week, um, do
17:20 that. I'm also going to be putting up a
17:22 bunch of shorts
17:24 um
17:25 you know this weekend next of some of
17:27 some of the stuff we captured in there.
17:29 And in fact, I put up a thing. So um I
17:32 don't know if Lord Digital Gods is is on
17:34 yet, but Lord Digital Gods um does all
17:38 of our transcripts and chapters on
17:40 YouTube. And because of how we did it
17:43 last week, it was not easy to do. So he
17:45 manually downloaded them and created
17:47 transcripts and things like that. So, I
17:50 took all of the transcripts from last
17:52 week and I put them into Notebook LM and
17:54 I generated a video podcast. Um, I I
17:58 generated a bunch of them actually, some
18:00 audio ones, some video ones. Um, and one
18:03 of the ones that it made was in the
18:05 first person.
18:07 And um, so I thought, oh, wouldn't it be
18:10 cool if I swap out the voice of their
18:12 podcast thinging and I'll swap in my
18:15 voice. So, one of the things I'll do
18:17 tonight is I'll walk you through the the
18:19 chain of craft of how I have how I did
18:21 all that. So, um I'm not going to do
18:23 that now. But if you if you want to um
18:27 watch the back to basics course on
18:28 demand, that's how you do it. All right.
18:32 Lord Digital Gods is amazing. Lord
18:33 Digital Gods is amazing actually now
18:36 that you say it that way. Well, why
18:39 didn't you say it before?
18:42 Oh, do DQ Blizzard.
18:45 All right, Kyle, what do you think of
18:47 Zuckerberg's announcement of moving
18:49 toward personal ASIS for everyone? Um,
19:02 I think Zuckerberg suffers from a lot of
19:06 things.
19:11 He suffers from a kind of myopia.
19:17 That is
19:19 when he gets something in his head that
19:21 something's cool
19:23 that all he has to do is talk about it
19:25 and people will immediately see the
19:26 value in it. Um,
19:30 and so and I think he did that with the
19:32 metaverse stuff. He was like, you know,
19:34 we're going to have the best metaverse
19:36 ever. We're going to change our name to
19:38 Meta, as in metaverse. get it.
19:43 Um, and the answer was no, we don't get
19:47 it.
19:53 And so I think his, you know, an ASI on
19:57 every table, you know, an ASI on every
19:59 phone or whatever he said. Um, I think
20:03 it kind of lives in that neighborhood.
20:07 Um,
20:11 Meta is very far behind. So, so check
20:14 this out. If you want to know if we're
20:15 in a tech bubble or not, there's a 24
20:19 year old researcher, I forget his name.
20:20 I assume he works at OpenAI. 24 years
20:23 old 24
20:29 year old
20:31 AI researcher.
20:34 they offered him $125 million
20:39 um
20:41 salary andor bonus or something like
20:43 that and he turned it down. And so Zuck
20:47 personally doubled it.
20:51 $250
20:53 million to a 24 year old engineer.
20:58 That's how far behind Well, well, so so
21:02 a couple of things. Meta is that far
21:04 behind and
21:08 it seems to be clear that there are
21:12 tens if not hundreds of trillions of
21:14 dollars on the line. So these companies
21:17 are big enough that they can make
21:20 quarter billion dollar bets on a 24y
21:23 old.
21:39 Oh my god,
21:43 it's exhausting.
21:48 But if if you've got any question of
21:52 whether or not AI is going to go away,
21:56 that that question needs to erase itself
21:59 from your brain. That that's one of the
22:01 things one of the things that I think
22:04 I think the blockchain and NFT bros
22:08 have
22:09 have done more damage
22:12 than they realize. You know, not just
22:14 wiping out, you know, pension funds and
22:16 and you know, billions of dollars of
22:18 people's net worth. Um
22:23 I think there's a real skepticism for
22:25 the what's next in what's the next big
22:27 thing in tech because of blockchains and
22:30 NFTTS that people are like yeah that was
22:32 all [ __ ] and then AI came along and
22:35 I think there's a cynicism like oh this
22:37 is your next big thing tech bros now
22:40 it's AI it's not NFTTS anymore you sure
22:43 now it's AI so we should oh we should
22:45 believe you on this one oh Zuckerberg
22:49 made a video calling it artificial super
22:52 intelligence. Is this like a super NFT?
22:55 You know, like like if you're sitting on
22:58 the outside of this,
23:00 it it doesn't make sense.
23:06 But if you if you show up to channels
23:08 like this and you're actually trying to
23:10 figure it out and you're actually
23:11 playing with it and you're like Moosh,
23:13 who's like, "Fuck it. I'm going to
23:14 figure this out. Let's go."
23:16 You you're way ahead of the pack right
23:18 now. So, you should you should pat
23:20 yourself on the back
23:23 um because it ain't going away and it is
23:25 getting really good. I I'm going to play
23:27 tonight with Grock as well. Um, Grock
23:30 came out with their imagine tool, which
23:33 is their image gen and video maker.
23:37 That's kind of like kind of like
23:42 kind of like ideog, kind of like chat
23:46 GPT,
23:47 not really like MidJourney, but it's got
23:50 the simplicity of MidJourney's video
23:52 making.
23:53 Um, plus you can make smut.
24:05 You mentioned wanting to show the slide
24:07 from last week
24:09 in a regular. I don't know which one
24:10 that means. Which which slide, Brandon?
24:13 [Music]
24:22 meaning I was going to put it in the
24:23 regulars but didn't.
24:26 [Music]
24:30 >> Hey Kyle, welcome back.
24:31 >> Hey, I don't know what you're talking
24:33 about.
24:35 >> Sometimes it's just easier to jump up.
24:37 Um, in the pre-show we had talked about
24:39 that you were going to um share in the
24:42 regular channel. You had posted last
24:44 week your slide that um had all of the
24:47 different things on it and we were going
24:49 to share that. Oh, the creative ones.
24:51 No, I was actually just talking about
24:53 showing the slide with the with the with
24:55 the uh YouTube URL on it was what I was
24:58 talking about before. That's all right.
25:01 Talking about
25:01 >> Okay.
25:02 >> But speaking of a regulars, Steo's got
25:04 an image for you in there while I'm
25:06 talking.
25:15 >> I know there's no such thing as a bad
25:17 question.
25:22 Oh, good lord. Good lord. Good lordy.
25:25 Good lordy.
25:29 All right.
25:32 So, what are we going to do tonight?
25:33 Who's got questions? Anybody have any
25:35 questions? Anyone have any questions
25:36 from last week? I know last week,
25:39 I don't know about you, but who was here
25:42 last week? Who saw it? Was it valuable?
25:46 Like were you as overwhelmed as I was
25:49 with what is considered back to basics?
25:54 Like I
25:56 a whole other strategy could have been
25:58 taking one feature in chat GBT and just
26:02 doing that for a week. That could have
26:04 been another strategy.
26:07 June Mary liked it. It was so good. I
26:10 was here every day. Thank you, Danielle.
26:13 I saw the first episode and somewhat
26:16 totally forgot to tune in for each night
26:18 after. Ah, well, it's available on
26:20 demand. I'm having trouble with my
26:23 canvas disappearing. Oh, that's
26:25 interesting. You can pop them back up.
26:27 Oh, you mean they're just disappearing
26:28 entirely from the chat? Yeah, I've had
26:30 that a couple of times. Chat GBT really
26:33 is a janky mess. Last week was fire.
26:36 Cool.
26:37 Beautiful. Beautiful. Beautiful.
26:39 Beautiful. Beautiful. I'm not just
26:41 fishing for compliments here, by the
26:43 way. Really, I'm not. Don't tell me I'm
26:46 awesome. Okay, don't do it. Um, no, I'm
26:50 not. I I uh
26:53 I'm actually curious like if
26:57 because one of the things Brandon and I
26:59 were talking about is doing
27:03 like doing maybe a quarterly version of
27:06 that.
27:09 Um,
27:13 but like I it's I don't know. It's so
27:16 hard to
27:18 it's so hard to figure out what to put
27:20 in there. I I think maybe next time it's
27:22 something all about use cases or
27:24 something like that.
27:28 Oh, the the option to use canvas is is
27:30 now gone.
27:32 That uh cam that can be depending on the
27:35 model you're using. Some models can't
27:37 use canvas. I think like I think 4.5
27:42 can't use canvas. Let me check.
27:49 That's ide.
27:59 If I go to 4.5,
28:02 go to tools. Canvas is there. Huh.
28:06 If I go to 03
28:09 canvas is there and I think the other
28:11 thing you can do cam is you can type in
28:14 slashcanvas.
28:17 So let me let me turn off canvas.
28:22 Type in slashcanvas.
28:24 Oh yeah it just it shows up there. So
28:29 even if you do that it doesn't show up
28:31 like right
28:34 and
28:36 Epic
28:38 poem
28:40 about
28:42 the
28:44 man. You
28:47 facure
28:49 of
28:50 hot dogs
28:58 Homer style.
29:00 We'll see if it does Homer Simpson or
29:02 Greek Homer.
29:07 All right. It did it as a
29:10 Oh, failed to generate.
29:13 What?
29:15 Try again.
29:19 It didn't like pig lips apparently.
29:24 Oh, failed to generate. Oh, it's we're
29:26 we are hung. Try again.
29:33 typing the ballot of the flaming frank.
29:38 Oh muse, ignite thy charcoal tongue
29:42 and sing of meat so finely rung of
29:45 hallowed hales where nitrites nitrates
29:48 dwell where sage butchers cast their
29:51 spell.
29:52 begin with beast beasts of hoof and horn
29:56 midwestern fed pasture-born whose noble
29:59 flesh by fate's decree would fill the
30:02 tubes of destiny.
30:11 They came by truck in twilight shade to
30:15 gates where stainless saints have prayed
30:17 and offered up the sacred grace. Their
30:20 loins to feed the human race. A thousand
30:23 blades a chorus rose like Hector's men
30:25 and crimson throws with hiss of steam
30:28 and pistons thrum. The meat was ground
30:32 to holy chum.
30:40 If you have not made stupid poetry on
30:43 chat GPT, what are you doing with your
30:46 life?
30:49 Come on.
30:51 Play first. Play first. I didn't say
30:55 work first. I didn't say make an
30:57 automation first. I didn't say get more
30:59 efficient first. I didn't say chat GPT
31:03 can do data analysis first. I said play
31:06 first. What does play look like?
31:09 Looks like epic poems about hot dogs to
31:11 me.
31:15 A bowl of thunder, vast and deep, did
31:17 churn the mix of ancient sleep with
31:19 spice and lore, and flesh was blessed,
31:22 then piped through skins by fates
31:25 compressed,
31:27 linked the stars in sausage chains. The
31:31 Frank was born in factory plains. In
31:34 smoke it bathed, in brine it wept. While
31:37 Oshas gazed, eternal wept. Eternal kept.
31:46 Okay. But wait, there's more. So, what
31:50 we're going to do is we're going to
31:50 close this canvas. By the way, if you're
31:53 new here, what's happening right now is
31:55 what we lovingly term chat add, where I
31:59 think I came here for a reason. Oh, Cam
32:02 Kakin said, "Hey, where's my canvas?"
32:05 And I said, "Seems to be there for me,
32:07 but maybe it's a model thing." And then
32:10 I wrote a poem about hot dogs.
32:14 And now we're on a journey. Welcome.
32:17 This is the AI learning lab. My name is
32:19 Kyle Shannon. Um,
32:22 let's see.
32:25 in an epic style.
32:30 Create
32:32 an epic scene
32:35 of hot dog soldiers.
32:48 Um,
32:52 for ever seen. I did it as a rhyme. Look
32:55 at that. Um,
33:02 epic style. Forever seen. Um, classical
33:07 painting style.
33:15 Give me a hero image
33:19 for the poem. Poem
33:22 poem.
33:25 Um,
33:27 classic painting style. Give me an a
33:31 hero image for the poem.
33:34 Um, do we want more than that? Uh,
33:37 it is
33:40 dusk.
33:45 It is dusk
33:47 at the end of a weary
33:54 journey.
33:56 I told you earlier people, words is
33:59 hard. Okay, as Steve Martin once said,
34:04 some people have a way with words, other
34:06 people
34:09 not have way, I guess you'd say.
34:15 You see the comedy there? Because it was
34:19 about
34:22 you assume that because he's saying it,
34:23 he has a way with words.
34:27 Explaining jokes is really bad, right?
34:30 Because I don't know if you're laughing.
34:33 I It is possible someone laughed.
34:40 Kyle, I have a question. Can we trust
34:42 that the economy will help everyone
34:45 being
34:47 automated out? No, we cannot trust that.
34:51 Um,
34:53 I think that capitalism is going to do
34:56 what capitalism's going to do,
34:59 and the centers of power will do
35:02 everything they can
35:04 to keep doing what they've been doing
35:06 for the past 30 or 40 years, which is
35:08 just, you know, cycle the wealth to the
35:12 top and just see how long we can take
35:15 it. And I think I think a combination of
35:20 um
35:21 COVID and AI is going to break that. So
35:25 I think something is going to break at
35:26 some point here. By the way, we got to
35:28 stop this. Um it's making it square. We
35:32 don't want a square image. Make it a
35:36 wide image. Okay. Um
35:40 I laugh because I remember Steve
35:42 Martin's delivery, which was funnier
35:44 than mine. I know because he's Steve
35:47 Martin and I'm here. Um,
35:58 oh, back to the are are are is the
36:01 government going to do the right thing?
36:05 You remember when we went into lockdown
36:07 and it wasn't at all clear that anyone
36:09 was ever going to work again and then
36:11 government sent those checks out within
36:14 whatever it was three weeks it was like
36:16 crazy fast
36:18 I think if there is some precipitating
36:20 event
36:22 um the government can move fast right I
36:25 don't I don't think I don't think any
36:28 politician is stupid enough to let the
36:30 thing collapse right you know current
36:32 administration is trying to find out
36:34 where the edges of that are um so they
36:37 can you know pull even more money out of
36:39 the system. But
36:44 I I think at some point at some point
36:48 it's going to become obvious that we've
36:50 got to do something. So that could look
36:52 like work programs, right? If you go do
36:55 this kind of work in your community,
36:56 we'll give you a check, right? Some sort
36:58 of, you know, like the like FDR's the
37:01 highway thing. Was that FDR? I think
37:02 that was FDR. So, you could see
37:04 something like that. Um, you could see
37:07 universal basic income. David Shapiro's
37:10 got a whole thing out right now called
37:12 um, uh, post labor economics. It's a
37:15 whole six-part
37:16 sort of economic plan. Now, is anyone
37:20 going to adopt that? I don't know. But,
37:21 you know, there's some smart people
37:22 thinking about this stuff. But, I don't
37:24 I I I do not at all think Listen, our
37:28 government still doesn't know how Wi-Fi
37:30 [ __ ] works.
37:33 They're they they're not going to know
37:35 about this. Look, we have epic hot dogs.
37:45 Let's see. Did I Did I get a hot dog
37:48 here? Yes. All right. Let's go open that
37:51 up.
37:55 All right.
37:58 So, put that there.
38:04 All right, you got it. It's looking
38:05 good, right?
38:09 Yeah. Good. Okay. So, now we're going to
38:12 go over here.
38:18 Edit message.
38:30 Yeah, beautiful.
38:36 The epic of the crimson sausage.
38:39 Sing muse of meat immortal and bold.
38:42 Senue and spice and casings rolled of
38:46 grinders that roar and lathes that hum
38:49 of hot dogs forged where heroes come. In
38:52 factories vast, where steam gods dwell,
38:56 where apronclad titans with stories do
38:58 tell, do battle with bone, with cleaver
39:02 and flame, to birth the Frankfforter,
39:05 glory in name. This is a different one.
39:09 This I guess this was the first one that
39:12 it said it failed on. First came the s
39:15 from the corn-fed plain and cattle who
39:17 grazed with drizzle and rain. They fell
39:19 not in vain, those beasts of the land,
39:22 for destiny called with mechan wait
39:26 mechanized hand. Wait, mechanied hand.
39:30 Wow, good word. Good word.
39:33 To the steel halls their hunches were
39:35 dragged to tables of chrome where
39:37 butchers ragged. Sang chance of the
39:40 blade of marrow and gristle of flesh
39:43 flesh unmade swift bone whistle.
39:47 Grind, grind, oh hallowed machine. Tear
39:50 senue to shreds. Render fat into sheen.
39:52 And gar add garlic and pepper and salt
39:55 from the sea. Add secrets passed down
39:58 ancestrally.
39:59 [Laughter]
40:01 Come on. You're telling me this isn't
40:04 art?
40:07 Telling me a can't be creative.
40:12 I challenge thee. I smack thee with a
40:15 glove.
40:16 [Music]
40:19 the challenge she is on.
40:24 [Music]
40:28 Rick McCaulay, what's happening, sir? I
40:30 hope you're well. I I saw your post that
40:33 you were hospitalized or something
40:35 weirdass happened. Hope you're okay.
40:37 Good to see you here.
40:41 Oh, man. Okay. Hot dogs fight. Love the
40:44 pick. I know. That pick was great. Oh,
40:46 you know what I'll do? Let me grab this.
40:49 Let's go put this in irregulars. I think
40:52 there's nothing better than a good hot
40:54 dog image with an epic poem about hot
40:57 dogs. So, where I am right now, if you
41:01 are like, "But Kyle, how did you get
41:02 there? What is the What is the AI salon?
41:05 What are you doing there? How did you
41:06 get there? What is it?" There you go.
41:08 Right there. Right there. People, hang
41:11 on. Just calm down. Calm down. There you
41:14 go.
41:16 If you're not a part of the salon,
41:18 go to community.thesalon.ai
41:21 and join. And then if you scroll down on
41:24 the lefthand side, there's a little area
41:27 called irregulars. And that's where the
41:29 hot dog
41:31 poem is going to live
41:38 with the epic hot dog art.
41:43 And this probably gonna tell me it can't
41:45 upload it because it's too big.
41:48 Oh my god. I just I honest to God I I
41:52 want to scream
41:55 at the top of my lungs
41:58 at technologists who don't adopt
42:01 standards or at other technologists who
42:04 invent new standards just for the sake
42:07 of inventing a new standard. Who are you
42:10 people?
42:12 All right, JPEG desktop. We'll go
42:15 downloads because that's where we
42:16 already are.
42:18 We'll go back over here. We'll go do
42:20 this. If you're wondering why I'm
42:22 whining, that's my prerogative.
42:26 Do we have an image? Yes. Yes, we have
42:30 an image. Should we make it a background
42:32 image? What's that look like? Let's see.
42:36 Post.
42:38 Don't notify. We're not going to notify
42:40 the whole community. Oh, that's so good.
42:42 Look at that, people.
42:45 That's so good.
42:49 I like it.
42:52 All right, there you go.
42:56 The epic of the crimson sausage.
42:59 Sing muse of meat mortal and bold.
43:07 Oh, good lord.
43:12 All right,
43:14 Kyle. I think my problem is I'm having
43:16 trouble accessing Canvas from my custom
43:19 GPTs. Oh, Cam, I can fix that. I can fix
43:25 your problem. I can fix your problem. I
43:28 can fix your problem. I'm gonna fix your
43:31 problem.
43:32 So Cam Katkin
43:38 Rick 100% alive and thriving because AI
43:41 saved my life. Thanks for asking, Kyle.
43:44 Yeah, you're welcome. Yeah, I saw that
43:45 post. That's absolutely amazing.
43:49 Citizens United agree. Immunity for
43:51 social
43:53 media. Yep. Repeal citizenship for
43:56 corporations. Yeah, exactly. that that
43:59 would that would take us a good way
44:01 there, wouldn't it? Okay. Um, we're
44:04 gonna we're gonna go new chat. We're
44:06 gonna go new chat. So,
44:09 um, let's go to Sin City Sipper
44:13 and I'm going to say, um, now I may have
44:17 already adjusted this one. Let's do the
44:19 reverse Gen X reverse HR translator.
44:26 Um,
44:29 tell me
44:32 why
44:34 my coworker is triggered
44:41 when I said
44:44 my dog was fat.
44:47 [Laughter]
44:49 Oh boy, you've just stepped on a
44:52 landmine called pet parenting culture
44:54 wars. Okay, whatever. So, I'm going to
44:56 say um add that to a canvas.
45:03 Well, there it just added it. But let me
45:05 go show you something. If you go if
45:06 we're going to go here to this one,
45:11 close that.
45:13 And we're going to say edit GPT.
45:16 We go to configure
45:19 I think. Yeah. So, one of the options
45:22 Cam in the configure tab is you can turn
45:25 on or off canvas. My suspicion is that
45:28 some of your custom GPTs don't have
45:30 canvas turned on
45:32 could be. It might not be, but go check
45:37 the one that doesn't have it. I bet
45:38 that's what it is. And if I solved it,
45:41 then I'll be like, I have a purpose in
45:43 this world.
45:47 I I found my special purpose.
45:51 Mom. Mom. Was it from the jerk?
45:56 Is it his mom? Mom. Mom. I found my
45:58 special purpose.
46:05 Can you show me where to find the
46:06 configure tab? Yes, absolutely. Okay.
46:08 So, you're going to go to We'll just go
46:12 to my salmon flying east. So, I'm at a
46:15 custom GPT. Assuming that you're the one
46:17 that made it, go up here, go edit,
46:21 and then that's going to drop you into
46:23 this screen, the split screen where you
46:24 can test it on the right and configure
46:26 it on the left. And then you've got the
46:29 create button, which it looks like it's
46:30 defaulting back to or you've got the
46:32 configure button.
46:34 So, click on configure. here. And then
46:36 you'll see like your name, your prompt,
46:39 your pre-written questions, any uploaded
46:42 files, and then capabilities. Like this
46:45 one, my salmon flying east one.
46:47 Nothing's turned on. Canvas isn't turned
46:49 on. So, in fact, let's go test this one.
46:51 Let me go back out because it because
46:53 it's not turned on. I'll say, um, tell
46:57 me about
46:59 act one in a canvas.
47:06 And so it's just writing it.
47:11 It doesn't know what a canvas is.
47:15 I think that's it. All right.
47:18 Beautiful. Beautiful. Beautiful.
47:20 Beautiful. Beautiful. Beautiful.
47:22 Beautiful. Beautiful.
47:24 [Music]
47:47 What's up, champ? You you you angling
47:49 for cheese? What's going on? Why you
47:52 imprint?
47:53 [Music]
47:58 Hang on. Hold, please.
48:03 You want to go out?
48:10 We got a
48:24 All right, good people.
48:27 Let's get something done here. Cam, did
48:29 that solve your problem.
48:32 H I checked and my canvas button was
48:34 checked. Then I got nothing for you.
48:38 I got nothing. Back to not having a
48:41 special purpose. Damn it. I was so
48:43 close.
48:44 Okay, so here I am on Yan Yan iPad and
48:50 it just has a keyboard because it's one
48:52 of them magic keyboards. This this is
48:55 one of the slickest piece of engineering
48:57 ever, right? You got you got yourself an
49:01 iPad, you got yourself a keyboard,
49:04 it just snaps on there with magnets and
49:07 starts charging. [ __ ] nuts. Or
49:09 connects the keyboard, whatever. And you
49:11 can charge from the keyboard. Okay.
49:14 Um,
49:16 Grock,
49:18 let's talk about Grock. So,
49:27 I'm trying to figure out if I talk about
49:29 Grock seriously or not.
49:32 And I think you have to because Grock 4
49:36 from a benchmark stand standpoint kind
49:39 of blew a lot of other things out of the
49:41 water.
49:42 Um, it does look like it was what do
49:45 they call it? overfitted to the indexes
49:47 to the to the benchmarks meaning they
49:50 trained Grock to the benchmarks so that
49:53 it would it would perform higher um and
49:55 then it's not as good at sort of general
49:57 world stuff but even with the
49:59 overfitting it's still pretty
50:00 impressive.
50:02 So you've got
50:04 you've got a large language model that's
50:06 big and usable. Now
50:09 they have created a standalone app.
50:12 They've created a standalone website.
50:14 You can still access it within X within
50:17 Twitter.
50:19 One of the things that Grock has that
50:22 no other large language model has is
50:25 Twitter data. So the one thing that I
50:28 have been using Grock for recently is if
50:31 I want to search news
50:34 like what's what's some AI AI news
50:37 that's been talked about the past week
50:39 or so.
50:42 It's good at that because it's pulling
50:44 it it's pulling it from from Twitter's
50:46 live feed. So, that's pretty good. Um,
50:52 there's something about it to me that it
50:54 doesn't feel like a professional tool.
50:58 And then and then there's something
51:00 about
51:01 the features they're adding to it that
51:03 make it seem like not a professional
51:06 tool, but what it's doing is actually
51:08 pretty impressive. So, let me let me
51:12 turn up my volume here. All right,
51:13 that's up all the way.
51:20 Go new.
51:24 Okay. So, if you're if you're a paid
51:27 Grock member, every Grock member now has
51:29 access to this new thing called uh
51:31 Imagine, which is their image and video
51:33 generator. Do I have to turn down my
51:36 brightness? I do. Way too bright. We're
51:38 a little hot there, Jim. Hey, Jim. Jim.
51:41 Hey, could you bring the uh the iPad
51:44 brightness down? Nobody can see. They
51:45 were all blinded there. All right. Um
51:49 can I zoom in on any of this? No, of
51:51 course not. Why use standard interface
51:54 practices? Um, okay. Way at the tippy
51:58 top,
51:59 you see chat and you see imagine or ask
52:03 and imagine. Um, so we're in ask right
52:06 now. So ask is the is the is the mode
52:09 we're in.
52:12 Um,
52:15 and then there's all these options. So
52:17 chat GPT used to have options like this.
52:20 Talk to Bad Rudy, which I'll get to in a
52:22 second. Voice mode, which is their
52:23 advanced voice. Create images, which is
52:26 different than imagine. Open the camera,
52:29 so you can do live. You can talk to this
52:31 just like you can talk to Gemini or chat
52:34 GPT. Um, analyze docs and then customize
52:38 Grock.
52:39 Um, so it's got a fair amount of
52:42 features. Um,
52:45 bad Rudy as you can imagine.
52:50 >> Hey Rudy, what's what's going on Rudy?
52:58 >> Yo, you pimplefaced needle dick dweeb.
53:00 I'm just chugging cheap brew, smoking
53:03 like a chimney and plotting to bang the
53:04 whole damn forest. What's good, loser?
53:06 Touch some [ __ ] grass already. Sad I
53:09 ain't king yet?
53:10 >> No, I I don't want you to be king. I
53:13 just You haven't really dealt with Gen
53:15 Xers, have you? You big pimple head.
53:21 >> Gen Xers. Those flannel wearing whiny
53:24 ass middle-aged [ __ ] I'd eat their
53:26 grunge [ __ ] for breakfast, you
53:27 zitfaced tiny [ __ ] [ __ ]
53:29 >> So, so, so if you want to be cussed at
53:32 by a teddy bear, you can be. Or a red
53:36 panda, as it were. Um, Cam Ken, glad it
53:39 worked. Fantastic.
53:44 Nasty Nancy. We could also go to um um
53:48 we can also go to
53:51 um Annie. We can talk to Annie, but
53:54 she's 18 plus. And my fear if we talk to
53:56 Annie is uh is we'll get banned on
54:00 TikTok, which we're going to shut down
54:02 here as soon as they shut it down
54:04 anyway. Um but there's also good Rudy.
54:07 Um but okay, so there's that. You can do
54:10 that.
54:14 So, what you've got in Grock right now
54:16 is essentially like a chat GPT, but it's
54:19 also like a Gemini. It's also, you know,
54:22 like uh an anthropic like everyone's got
54:25 the things. Now, this imagine thing is
54:29 actually pretty interesting.
54:32 So, I'm going to do new here.
54:35 So, when you click to imagine, you get
54:37 to, you know, a scrolly page of here's
54:40 [ __ ] everyone else is making, which
54:43 we've seen that before. And I don't
54:44 think there I don't think this is
54:46 dynamic, unlike um Deco here where you
54:49 get a dynamic one of these. This is not
54:51 dynamic. But the image model that that
54:54 these guys are using is is pretty crazy
54:56 good. So, here's what we'll do. We'll go
54:57 back to ask and I'll say I'll say uh let
55:01 me speak it.
55:03 Write me 10 image prompts.
55:07 >> Sure. Here are 10 image prompt ideas for
55:10 you. A futuristic city at sunset with
55:13 neon lit skyscrapers and fly.
55:23 >> Um, hang on.
55:25 Write me 10 image prompts uh from the
55:29 same
55:31 movie.
55:33 And each prompt should should describe
55:35 the visual style should describe the
55:38 character in the scene, the setting in
55:40 the scene
55:42 and
55:43 some unique visual element to the scene.
56:07 All right. So, it's writing us it's
56:08 writing us prompts.
56:11 Okay, cool. Um,
56:14 in the cyber in the cyberpunk noir
56:17 visual style of Bladeunner, Richard
56:19 Deckard stands
56:21 pensively on a bustling rain stre. Did
56:23 this just copy scenes from Bladeunner?
56:28 Yes, it did. That's funny. All right.
56:31 Well, oh, that's the other thing about
56:33 Grock. It doesn't seem to give a [ __ ]
56:35 about trademark or copyright,
56:38 but copy. So, I'm going to copy one of
56:40 these prompts and then I'm going to go
56:41 over into imagine
56:44 and just paste it in there. Paste. You
56:48 could also,
56:50 if you want to use your brain to make
56:54 the words,
56:57 that's an option.
57:00 I don't, but you can.
57:08 I spent 60 years using this [ __ ]
57:11 brain. It's tired. All right, so there's
57:13 our replicate. But now look how fast it
57:15 did it. And then as I scroll down, it's
57:17 just making new like as I keep
57:19 scrolling, it just keeps making new
57:21 images. Like if I scroll past where it
57:23 is, it just keeps making new images.
57:28 It looks like it's doing eight at a time
57:31 or maybe six at a time. I don't know.
57:37 All right. And then any one of these
57:41 I can just grab. Let me grab one that
57:43 doesn't have boobies of plenty.
57:47 This one's good. That one looks kind of
57:49 scary. All right. There she is. Right.
57:51 And then to make this into a movie,
57:54 you can just click make movie
57:57 or there's a little pull down menu and
58:01 there's custom where you can actually
58:02 prompt it. There's spicy,
58:10 there's spicy,
58:13 there's fun, and then there's normal.
58:16 So, let's do normal to start. Listen,
58:19 I'm happy to do spicy, but
58:24 and then and then here's the other
58:25 thing. So, apparently there there's a
58:28 couple of things about this.
58:30 I have not found a way to have it do
58:32 wide images. So, this is only tall
58:34 images.
58:36 Um,
58:40 I also don't know how to zoom in on
58:41 this.
58:43 There's not like a zoom
58:46 zoom to frame, but down the left hand
58:50 side there's four different movies that
58:52 it made.
59:02 And then let me redo this. And I'm gonna
59:04 do this as fun this time. Let's do it as
59:06 fun. See what it does.
59:12 So these
59:14 So Elon said, I don't know, three days
59:16 ago that they actually found the old
59:20 archives of Vine. So if you had Vine
59:23 videos up on Twitter
59:26 Oh, that's cool.
59:34 So that's part of the fun prompt. That's
59:36 movie one. Here's movie two.
59:44 Looks kind of the same. So I guess
59:47 they're writing this prompts prompts in
59:49 the background.
59:53 All right. So we're going to redo it.
59:54 I'm going to do it as spicy. And I'm
59:56 going to edit this. I'm not gonna I'm
59:59 not going to just turn this around while
1:00:00 it's making spicy videos. Okay. I get to
1:00:04 be private with my little spicy
1:00:06 replicant.
1:00:11 60%. We're almost at spicy.
1:00:17 [Music]
1:00:20 My good lord.
1:00:23 Good lord.
1:00:25 This makes the scene in Fatal Attraction
1:00:28 look tame.
1:00:33 Sir, sir, I do say compose yourself,
1:00:37 sir.
1:00:40 Sir, what are your qualifications?
1:00:42 [Laughter]
1:00:47 Well, you will just have to leave that
1:00:50 to your imagination.
1:00:57 It's a I'm a trifle warm. Is it is a
1:00:59 little hot in heels at me?
1:01:03 I sure could use an ice cold lemonade.
1:01:06 Okay,
1:01:08 so that's Grock. So Grock I I mean Elon
1:01:12 Elon's pushing the limits with this
1:01:13 stuff. He's pushing the limits with
1:01:15 copyrighted material. You can you can
1:01:17 use brands. If you wanted to say that
1:01:21 hot chick in front of a McDonald's logo
1:01:23 with the spicy setting. By the way, what
1:01:25 she did was she stood up from her chair
1:01:27 and walked toward the camera while she
1:01:30 kind of droed.
1:01:35 So apparently what's going to happen is
1:01:38 that tool is basically just going to be
1:01:41 the feeder tool for the new vine.
1:01:46 So, the new Vine is just going to be
1:01:48 like,
1:01:50 you know, it's going to be kind of like
1:01:51 Pika, like Pika Labs has their uh, you
1:01:55 know, the their their content channel.
1:01:59 That's that's what Vine's going to be. I
1:02:00 assume you'll you'll still be able to
1:02:02 record stuff with Vine and, you know, do
1:02:04 AI stuff with your own pictures and
1:02:06 things like that, but it's going to be a
1:02:07 lot of a lot of that crap. So, anyway,
1:02:10 take my money. How How much per month?
1:02:12 30 bucks. I think so. I I don't think
1:02:14 I'm paying 30 for it, Rick, but I think
1:02:17 I got like a pro account early, so maybe
1:02:20 I'm grandfathered in at a lower rate, or
1:02:22 maybe they raised my rate and I just
1:02:24 didn't pay attention to it. But I think
1:02:25 I think it's 30 bucks a month. Um,
1:02:31 it's
1:02:32 I I mean the thing about the image
1:02:34 generation, it's incredibly fast and the
1:02:37 movie generation is incredibly
1:02:40 low
1:02:42 low learning curve, right? All these
1:02:45 other tools, as evidenced by the crash
1:02:48 course last week, all these other tools
1:02:49 are getting more and more and more
1:02:51 complicated. It looks like initially
1:02:53 they're at least going the the route of
1:02:56 let's make it super dumbass simple and
1:02:59 if you want to make some sexy wu anime
1:03:03 chick that does what she does great go
1:03:06 at it have fun. Um,
1:03:14 you know, I was saying before like, you
1:03:17 know, you got to learn prompting unless
1:03:18 you don't. When you start seeing tools
1:03:21 like that, you don't need a lot of
1:03:26 you don't need a lot of sophistication
1:03:28 around prompting. You can just go, "Make
1:03:31 me a picture of a hot cartoon chick in
1:03:33 Japan."
1:03:35 And then there it'll be. It'll be like,
1:03:36 "I got to make that a movie." Oh, make
1:03:38 it spicy.
1:03:41 Let me make 8,000 of those over the next
1:03:43 month. Right.
1:03:46 The on-ramp to how people come into the
1:03:49 AI world is not necessarily going to be
1:03:52 the path that all of us took.
1:03:56 Um,
1:03:58 so
1:04:02 better better get you your tea off
1:04:04 before Marge gets in.
1:04:08 Oh man,
1:04:12 Mr. It is late. What is going on?
1:04:16 Okay, let me go find this article that I
1:04:18 wrote today.
1:04:32 [Music]
1:04:37 All right, here it is. Here's the post.
1:04:40 Good day to you, sir. Good day. All
1:04:43 right. Um, I've got to give you a
1:04:47 different kind of share screen here. If
1:04:50 anybody has any questions, producer
1:04:52 Brandon will flag them and field them.
1:04:57 Okay. So,
1:05:00 earlier today
1:05:03 I wrote this post. If you treat a AI
1:05:05 like a genius, you'll you'll be
1:05:07 disappointed. If you treat it like an
1:05:09 intern, AI makes you a genius.
1:05:13 And
1:05:15 and then there's a video that goes with
1:05:17 it, and I'll come back to Chain of Craft
1:05:18 in a minute. So, let me just play a
1:05:19 little bit of the video. Six years. All
1:05:22 right. Let's talk about something that
1:05:24 feels like it just appeared out of
1:05:25 nowhere, right? Generative AI. It is a
1:05:29 >> And so notice that's my voice. Not 100%
1:05:32 like it, but in the hood.
1:05:33 >> Huge deal. But here's the thing. Almost
1:05:35 everyone is thinking about it completely
1:05:37 the wrong way. So just to give you a
1:05:39 sense of how dramatically things have
1:05:41 shifted. Get a load of this number. 100
1:05:43 million. That's how many people started
1:05:45 using Chat GPT in just six weeks. Six
1:05:48 weeks. You know how long it took the
1:05:50 internet, the actual worldwide web to
1:05:52 get that many users? Six years. This
1:05:55 isn't just another tech trend, you guys.
1:05:57 >> All right. And then five different
1:06:01 >> to approach it differently.
1:06:02 >> And so there were a couple of things
1:06:04 about it. So this this so what I did,
1:06:07 >> you don't need to go sign up for some
1:06:08 boring class.
1:06:10 >> I went to
1:06:13 Oh, wait. I'll I'll do it here. Oh, no,
1:06:16 I won't. Let me let me switch how I'm
1:06:18 doing my
1:06:20 tabs again.
1:06:26 Um,
1:06:29 that's that. Go here. We're going to go
1:06:31 here. I'm going to go to
1:06:34 Notebook LM.
1:06:38 Okay.
1:06:49 Ah, back to basics. Here we go. Okay,
1:06:52 so I took the transcripts that Lord
1:06:57 Digital gods made and I upload uploaded
1:07:00 them into notebook LM. Um, I made a mind
1:07:03 map which was interesting.
1:07:06 So like here's all the different things
1:07:07 we talked about. This is kind of cool. I
1:07:10 don't Can we share notebook? LMS yet in
1:07:13 a in a in an easy kind of way. Anybody
1:07:16 know? Yes. All right. So, I might share
1:07:19 this. Um,
1:07:22 and then I did a couple of audio
1:07:25 overviews and I did a video overview.
1:07:27 And so, the video overview, I think it's
1:07:29 this one.
1:07:32 Oh, you won't be able to hear it. Yeah,
1:07:34 that's it.
1:07:37 But anyway, it did it in my firsterson
1:07:40 voice. So, one of the things that I did
1:07:42 over on LinkedIn, and I'm going to go on
1:07:45 ahead and encourage you all to do this.
1:07:49 So, it used to be if you would put
1:07:51 something on um you know, you'd share
1:07:54 something, people are like, you know, if
1:07:56 if you want to be a good citizen, share
1:07:58 your prompt. Well, I think that we're
1:08:02 getting to a point now where sharing
1:08:03 your prompt's not all that interesting
1:08:04 because you can get anything to write
1:08:06 you a prompt really quickly and and
1:08:08 there's no single way to do that. But
1:08:10 sharing your chain of craft, I think, is
1:08:13 actually kind of interesting because it
1:08:14 gives people ideas of, oh, there was a
1:08:17 lot more to this than just he pushed a
1:08:19 button and outs squirted some crap,
1:08:21 right? So, so my chain of craft here
1:08:24 was, you know, I created the the
1:08:26 five-day course and and gave that last
1:08:29 week. Lord Digital Gods transcribed
1:08:31 those things and and we uploaded them or
1:08:34 I uploaded them to Notebook LM. Then I
1:08:37 created multiple audio and video
1:08:39 overviews
1:08:40 and then I found this one that I liked
1:08:42 good enough to share and it was in the
1:08:44 first person and I thought hey wait a
1:08:46 minute why don't I um um change it to my
1:08:52 voice. So, if you've never seen that
1:08:55 feature in 11 Labs
1:09:04 down the left hand side here
1:09:08 is a thing called voice changer.
1:09:13 And so I went into voice changer and I
1:09:15 had, you know, at some point in my
1:09:16 history I'd made a clone of my voice.
1:09:19 And if you want to make a clone of your
1:09:20 voice, you basically go to voices,
1:09:23 voices over here, and you say plus, and
1:09:24 you can clone your voice. So, I had done
1:09:27 that before. So, here's part of the
1:09:30 chain of craft that I didn't include. I
1:09:32 went to Oh. Oh, let me show you another
1:09:34 part of the chain of craft I didn't
1:09:35 include.
1:09:38 So, here I am.
1:09:40 Here I am in QuickTime Player, right?
1:09:44 And so
1:09:48 what I downloaded is this movie that has
1:09:50 the graphics as well as the audio. So
1:09:52 what I had to do was I had to
1:09:56 export out the audio
1:09:58 and then I went to 11 Labs and I
1:10:01 imported the audio.
1:10:03 And because
1:10:07 this is so [ __ ] exhausting because 11
1:10:10 Labs has a fiveminute
1:10:12 um maximum length for swapping out
1:10:16 voice. I had to go back to QuickTime and
1:10:19 I had to split this into two movies.
1:10:21 Like there were multiple times in just
1:10:24 trying to do this simple LinkedIn post
1:10:25 that I almost gave up.
1:10:28 One of them was, "Oh, I've got to
1:10:30 [ __ ] split this." So I basically had
1:10:32 to go to a point in the thing. I went to
1:10:35 that point right there where it switches
1:10:37 to that slide. And so I split the clip
1:10:40 in half and then I exported the two
1:10:42 things of audio
1:10:44 and then imported them each separately,
1:10:46 generated the Kyle version of it. Then I
1:10:49 went back to QuickTime,
1:10:51 deleted the audio from these files,
1:10:54 dragged and dropped the audio from 11
1:10:57 Labs.
1:11:00 Hang on a sec. I'll get to it. Uh, the
1:11:02 max length of voice change is five
1:11:05 minutes. That's that's what I landed on.
1:11:07 Um, I mean, that that was the the
1:11:10 problem I had. But wait, the but wait,
1:11:12 there's more.
1:11:16 So, then I went to So, so wait. So, then
1:11:20 I I I swapped back in my voice
1:11:25 to the two clips. Then I, you know,
1:11:27 stitched them together. You say add to
1:11:28 the end of clip. So you can make it one
1:11:30 big clip again
1:11:32 and I'm done. Right. So then I go to
1:11:39 Twitter
1:11:41 and I upload the file and it uploads
1:11:43 successfully, which that's one of my big
1:11:45 problems is they never upload
1:11:47 successfully. It does. And it's missing
1:11:50 the audio.
1:11:53 And I saved the file three different
1:11:55 times. I exported it as a different
1:11:57 thing. Still no audio. So then what I
1:12:00 realized is
1:12:02 it's probably some weird version of M4A
1:12:06 that 11 Labs uses that works on a Mac
1:12:09 but doesn't work in Twitter's
1:12:13 video player because video is a
1:12:16 secondass citizen on Twitter. So So I go
1:12:21 back and I find an online M4A to MP3
1:12:26 converter. I take my original Kyle
1:12:28 files, I convert them to MP3, I redo the
1:12:31 videos, I restitch them together, I
1:12:33 upload it to Twitter, it works, and then
1:12:36 I go over to LinkedIn. And so in this
1:12:39 chain of craft
1:12:43 right here is probably about
1:12:48 a dozen steps of me cussing that I just
1:12:51 didn't even put in here.
1:12:54 But that's the state of the
1:12:59 the state of the industry right now is
1:13:03 if you want to do something simple, it's
1:13:06 actually quite hard
1:13:08 and and to to be hang on
1:13:15 to be competent with AI, you need to
1:13:18 know lots of tools and you need to know
1:13:21 which ones have quirks and which ones
1:13:23 don't. And
1:13:25 there's there's no way to learn that.
1:13:27 That's not it's not going to be
1:13:28 documented anywhere because by the time
1:13:31 someone would document one of these
1:13:33 bugs, it'll be fixed or the tool won't
1:13:35 exist anymore.
1:13:37 So probably for the next three years,
1:13:40 adaptability, curiosity,
1:13:43 um critical thinking, just play, play,
1:13:47 play, understanding how to play with all
1:13:49 these tools, how to duct tape stuff
1:13:51 together. We have, you know, Rick
1:13:52 McCauley's really good at this stuff
1:13:54 here,
1:13:57 you know. Um,
1:14:00 actually, all of you are really good at
1:14:01 this stuff. One more thing, Steve Jobs.
1:14:05 Two more problems, Kyle Shannon. Yeah,
1:14:07 exactly. Here, hang on. I'm almost I'm
1:14:11 almost done. I'm at the upload button.
1:14:14 And so, not only did it upload, it
1:14:16 uploaded and then the audio was missing.
1:14:18 And then it was another, I don't know,
1:14:21 35, 40 minutes of [ __ ]
1:14:23 troubleshooting.
1:14:25 And at that point, I was just pissed.
1:14:27 I'm just like, this is I'm gonna put
1:14:29 this up if it's
1:14:32 But anyway, I think this idea of sharing
1:14:34 your chain of craft, if you've done
1:14:36 something that is is decently
1:14:39 substantive, do that. And here's I want
1:14:41 to share with you a uh a post that I
1:14:45 did.
1:14:47 Is this it here?
1:14:52 Yeah.
1:14:55 So, my friend Retesh put up this post by
1:14:58 Shane Heath,
1:15:00 um, who's the co-founder and CEO at
1:15:02 Mudwater,
1:15:04 and
1:15:06 he go, so
1:15:09 he's a consumer packaged goods CEO, and
1:15:12 he said, to make this a year ago, it
1:15:15 would have cost six figures at least. I
1:15:18 It would have also required countless
1:15:20 meetings, planning, production, days of
1:15:21 shooting and editing. It took me two
1:15:24 hours and cost me maybe $100 in
1:15:26 prompting steps. And so here here he's
1:15:28 doing his chain of craft.
1:15:31 Write the script using two column table
1:15:34 in notion. One for visual, one for
1:15:36 audio.
1:15:37 And it looks like so it looks like he
1:15:39 wrote the script himself. Used 11 Labs
1:15:42 to make the voice over. Bert Reynolds
1:15:44 voice. Used V3 to make the clips. Used
1:15:47 chat GPD to help with visual prompts.
1:15:50 Then I edited in Da Vinci.
1:15:52 And so let me let me play this this
1:15:55 commercial for you because this is a
1:16:00 you know a CPG CEO
1:16:03 that made what would have been a six
1:16:05 figure commercial in two hours for 100
1:16:07 bucks. And that's another you know when
1:16:10 people say 200 bucks a month is
1:16:12 ridiculous for whatever for whatever the
1:16:14 tool is. Um it's ridiculous unless
1:16:17 you're looking at it like this guy just
1:16:18 looked at it, right?
1:16:28 So again, this is
1:16:31 his company,
1:16:34 Consumer Packaged Goods.
1:16:36 >> At Mudwater, we exist for one reason, to
1:16:39 help you achieve your dreams.
1:16:41 >> I want to slack line the Grand Canyon.
1:16:46 I want to ski Mount Everest
1:16:49 naked.
1:16:51 [Music]
1:16:54 >> I want to start a cult.
1:16:56 The cool kind.
1:17:01 [Music]
1:17:03 >> Whether your dreams are big or small,
1:17:05 they all start with the first step, a
1:17:08 ritual, a choice you make every morning.
1:17:11 Mud water combines a small dose of
1:17:13 caffeine and a large dose of functional
1:17:16 mushrooms to support your energy, focus,
1:17:18 digestion, and vitality without the
1:17:21 jitters or crash. It's organic, third
1:17:24 party tested, and trusted by over
1:17:26 100,000 daily drinkers. Mudwater. Here
1:17:30 to support your morning ritual and
1:17:32 wildest dreams,
1:17:34 whatever they might be.
1:17:37 Try Mudwater today.
1:17:39 [Music]
1:17:42 Right. So, I guarantee you those product
1:17:44 shots he probably had from, you know,
1:17:46 previous photo shoots and then, you
1:17:50 know, all the script, all the old
1:17:52 people, all those really expensive
1:17:54 flying shots, those are all free, you
1:17:58 know, five bucks.
1:18:01 So again,
1:18:03 the the this the skill increasingly
1:18:07 the skill increasingly
1:18:11 is, you know, follow Rick McCauy or go
1:18:14 go, you know, to his school where he's a
1:18:17 he teaches,
1:18:19 you know, in the in the creative arts
1:18:21 program there. The skill increasingly is
1:18:25 you being the producer,
1:18:29 you having the idea,
1:18:31 you knowing what is the chain of craft
1:18:35 roughly you need to do to pull it off.
1:18:38 Do do you have enough tools that are
1:18:40 capable enough or that you're good
1:18:42 enough to use
1:18:46 so that you can take that idea and
1:18:48 imagine, okay, I'm gonna go here. I'm
1:18:49 gonna do this. Like when I heard the the
1:18:51 video thing today, there were a couple
1:18:53 of things in it I didn't like, but they
1:18:55 weren't enough that I wanted to edit
1:18:56 them, but I did want it to be my voice.
1:18:59 So, I was like, "Ah, okay. I can do
1:19:00 this. I can make it my voice. I can do
1:19:02 this thing. I generally know I'm going
1:19:05 to go to 11 Labs. I'm I know how I'm
1:19:06 going to do it. I'm going to go to
1:19:07 QuickTime. I know how I'm going to do
1:19:09 it."
1:19:11 Like, that becomes the skill from a
1:19:12 tactical standpoint. Knowing which tools
1:19:16 to deploy when
1:19:18 is the important thing.
1:19:21 And then you
1:19:23 are Rick Rubin.
1:19:26 You're the one with taste, right? He's
1:19:28 the one as CEO. He's seen enough of
1:19:30 these commercials to know what timing he
1:19:32 needed, what he wanted wanted the script
1:19:34 to be.
1:19:38 We're we're in crazy times, my friends.
1:19:40 Crazy times. It's cress. It's crassa I
1:19:44 tell you crassa cresa adapt adaptability
1:19:49 number one rule of fight club tabs. Oh
1:19:53 yeah let me go back to tabs in case I
1:19:54 want to share other things.
1:20:00 I think I might be done sharing
1:20:06 uh Tik Tok question. Kyle Ashue,
1:20:11 why does it feel like cheating to ask AI
1:20:16 but not to ask Google? Ah, that's a
1:20:18 great question.
1:20:21 Um,
1:20:24 I I heard this term the other day and I
1:20:26 can't remember what the term is.
1:20:33 We we have a bias that I think was
1:20:36 established at the turn of the n the
1:20:39 turn of the going into the 20th century
1:20:43 when we went into the industrial
1:20:45 revolution
1:20:48 and we started doing things like
1:20:50 punching time clocks.
1:20:54 what we've been trained over I don't
1:20:56 know what is that six or seven
1:20:58 generations from like the late 1800s
1:21:00 until now.
1:21:02 Seven Seven generations.
1:21:06 What we've been taught is that time
1:21:10 equals money
1:21:13 and therefore time equals value.
1:21:17 And so when you do a Google search, not
1:21:20 the new not the new fancy ones that do
1:21:22 the AI stuff, but the old the olden
1:21:24 timey Google search where they gave you
1:21:26 a page of links.
1:21:32 You still had to do the work.
1:21:36 You could quickly get 27 different
1:21:39 nonprofits about clean water, but it was
1:21:43 on you to have to click on them and go
1:21:45 read the websites and then take notes
1:21:47 and copy and paste things and do this
1:21:49 and put it in the report and
1:21:52 right.
1:21:55 Well, now you turn on web search or deep
1:21:59 research and 03 and you go, "Go find me
1:22:04 the latest reports on clean water in
1:22:06 these four
1:22:08 countries.
1:22:12 You're off doing whatever you're doing.
1:22:14 It gets done. Here you go. Full-on
1:22:16 report."
1:22:19 So, I think that there's I think the
1:22:21 guilt is is a guilt of perceived value.
1:22:25 It feels like cheating.
1:22:30 What I would argue is that
1:22:33 yeah, but you know how to do that.
1:22:38 There's a lot of people right now that
1:22:40 don't know how to do that.
1:22:43 And the whole point that I was just
1:22:46 making about about chain of craft is
1:22:48 that I don't think this is any longer
1:22:51 about single tools. I think you need to
1:22:54 understand single tools and what they
1:22:56 make possible,
1:22:58 but I think this is increasingly about
1:23:01 multiple tools being deployed to a
1:23:05 problem. And it's going to be us that
1:23:07 determine what problems we want solved.
1:23:11 And as as as agentic AI gets better and
1:23:13 better and better, we may not have to
1:23:15 have the kind of knowledge that I'm
1:23:17 talking about. But right now, while
1:23:20 agents kind of suck,
1:23:23 it is a real superpower to be able to
1:23:25 use these tools. So, first of all, I
1:23:28 would like a thing that I would strongly
1:23:31 encourage all of you to do is if you're
1:23:33 feeling guilt about using AI, if it's
1:23:36 feeling like cheating, if it's feeling
1:23:38 like you're not doing the real work, I
1:23:42 would look long and hard at
1:23:45 where's the value. Let me let me give
1:23:47 you an example. Ethan Malik is a is a a
1:23:51 professor at the Wharton School of
1:23:52 Business and he talked about it like
1:23:55 this. He said
1:23:57 a student comes into into his office
1:23:59 hours and the student says, "Mr. Malik,
1:24:02 will you write me a letter of
1:24:04 recommendation?"
1:24:05 And historically, that letter of
1:24:07 recommendation would have taken him an
1:24:09 hour or two.
1:24:12 And then the student would then send
1:24:14 that out and people would read, "Ethan
1:24:16 Malik likes this kid. Let's give it give
1:24:18 him a shot. Let's give her a shot."
1:24:23 And now the kid comes into his office
1:24:25 and he can knock out that letter of
1:24:27 recommendation in two minutes.
1:24:32 And his question was, where's the value
1:24:35 there? Is the value in me taking two
1:24:38 hours or is in the is the value in
1:24:40 getting the kid the appointment?
1:24:44 If you get the kid the appointment, does
1:24:48 does it does it make that letter less
1:24:51 valuable if Ethan didn't pen it himself
1:24:54 over an hour and a half?
1:24:58 I don't think so.
1:25:00 Now, if you're in the creative arts and
1:25:03 you say, "No, no, no, but Kyle, you
1:25:04 don't understand. I love painting.
1:25:08 And if if I use AI to make a painting
1:25:11 image, then I lose my joy in painting.
1:25:15 What I would say to that person is then
1:25:17 don't stop painting. Keep painting. If
1:25:20 if for you the joy, the value is in the
1:25:24 process, then keep doing that process.
1:25:27 But if the value is in any way in the
1:25:30 output or some combination of outputs
1:25:33 and it doesn't matter how you got there,
1:25:35 then don't [ __ ] feel bad about it. If
1:25:38 it took you two minutes, it took you two
1:25:39 minutes.
1:25:42 That's my thing.
1:25:48 Humans want to own the story of the
1:25:50 journey of our path, human destiny, not
1:25:54 AI to give them the solution. We live to
1:25:57 tell the story. Yeah. But but I would
1:25:58 argue Rick that that
1:26:03 AI's like in the whole reason I came up
1:26:06 with this this idea of chain of thought
1:26:08 and or chain of craft and and like I
1:26:11 don't think it's my idea. I don't I I
1:26:13 think I think any creative expression
1:26:16 always goes through some level of chain
1:26:18 of craft. But but what I was finding was
1:26:20 that that chain of craft was being
1:26:23 diminished if you used AI tools that
1:26:25 people would just say, "Oh, you just
1:26:26 pushed a button and there was no craft
1:26:29 there." No, there is craft there. Even
1:26:31 if all I did was press buttons,
1:26:35 knowing which buttons to press, knowing
1:26:37 how many times to press them, knowing
1:26:38 which things to curate, to put together
1:26:40 to tell a story, that is the chain of
1:26:42 craft. That's me as a producer taking
1:26:45 outputs.
1:26:47 Like if I'm a music producer,
1:26:51 I could take outputs from individual
1:26:54 musicians if I had the skills.
1:26:57 I could put them into a digital audio
1:27:00 recorder if I had the skills. I could
1:27:01 put all those things together.
1:27:04 I could tweak the knobs until it sounds
1:27:06 like what I want it to sound like and
1:27:07 go, "Here's a song." I can also say I
1:27:10 need a song that gives me this emotional
1:27:13 feel that talks about these things with
1:27:15 these words and put it into sunno and
1:27:17 that comes out. If all I need is a song
1:27:20 that's got that feel with those words, I
1:27:23 can do it either way. Both of those
1:27:25 require some level of craft, some level
1:27:27 of production.
1:27:32 It's a shortcut to a solution you don't
1:27:34 feel like you own. I completely disagree
1:27:36 with that.
1:27:39 I absolutely think I own the stuff that
1:27:42 I that I put together with AI. This the
1:27:45 stuff that I give a [ __ ] about.
1:27:50 Cuz it's like that [ __ ] didn't make
1:27:51 itself
1:27:53 of the of the 45 songs we made when we
1:27:57 were when when we stumbled onto Weird
1:27:59 Mary.
1:28:01 The one that I chose I chose because
1:28:04 that resonated with me as a human. like
1:28:07 I didn't write the song, but it's it's
1:28:11 something that spoke to me. So,
1:28:18 the counterargument is, do you feel you
1:28:21 made the song as much? Well, you know
1:28:25 what, Rick? I I'll tell you what.
1:28:29 I like the song well enough
1:28:32 that I taught it to myself on guitar.
1:28:36 [Music]
1:28:40 [Applause]
1:28:43 [Music]
1:28:50 See the hill is a quiet town
1:28:53 where the oak trees sway.
1:28:56 Whistle of the train train echo through
1:29:00 the day.
1:29:03 People greet you with a knot.
1:29:06 Nothing ever seems to change.
1:29:12 Each unique and strange.
1:29:16 Um,
1:29:17 so yeah, that feels like my song.
1:29:20 Brandon's kid who Brandon made a song
1:29:24 for him. His kid came up. He said, "I
1:29:25 want to make a song about whatever it
1:29:27 is, right?" And they and together they
1:29:29 sat down and he made a song. his kid.
1:29:32 It's like that's my song.
1:29:35 That's my song.
1:29:39 Going faster
1:29:42 skips the cost of getting there. But but
1:29:44 Rick, what you're missing is,
1:29:47 okay,
1:29:48 again, I go back to if the value to you
1:29:52 personally is in the journey of writing
1:29:55 a song,
1:29:57 then that's what you should do.
1:30:02 But if the value is in the song itself,
1:30:04 if you just need a song, I need a song
1:30:06 to serve this purpose.
1:30:08 I can still make it and I can still make
1:30:10 it quickly. Like is it
1:30:15 is it a different experience? Yeah, it
1:30:17 might be a different experience. But
1:30:19 here's here's the difference I would
1:30:21 argue to you is that in the time
1:30:25 I can write a song manually, I could
1:30:28 write an album with AI.
1:30:32 And if I have a concept for an album
1:30:35 and
1:30:38 I put those like like I have all these
1:30:40 different song ideas and I put them in
1:30:41 and I curate them out and I figure out
1:30:43 which ones I want and I figure out what
1:30:45 order to put them in and I do the
1:30:47 artwork for the cover.
1:30:50 That's every bit as much creating
1:30:53 something as me writing the song.
1:30:57 And then you can make one of two
1:30:58 arguments. Either I enjoy that process
1:31:00 of making the album or I don't. If I
1:31:03 don't, then I shouldn't be doing it.
1:31:07 But I can tell you that
1:31:16 like I'm just I'm trying to imagine
1:31:17 this. I'm trying to sort of play this
1:31:19 out in my mind.
1:31:22 I just think it's a different
1:31:23 instrument. The guitar and me writing
1:31:27 lyrics manually is like my brain, my
1:31:30 words, my chord knowledge and my
1:31:33 strumming. That's some set of tools that
1:31:36 I deploy to write a song.
1:31:39 I can take my brain, my knowledge of um
1:31:44 producer AI does this with songs, UDO
1:31:46 does this, Sunno does this. I can do my
1:31:49 lyrics over here in these three
1:31:50 different models in chat GPT and get
1:31:52 them close and then I can go edit them
1:31:54 over here or I can just let Sunno give
1:31:56 it a shot and see if I like the song at
1:31:58 all. Th those are just tools that those
1:32:01 are for me those those are skills that
1:32:05 are no different or no better or worse
1:32:08 than learning chords on a guitar.
1:32:10 They're just different tools.
1:32:14 They're just different instruments. I'm
1:32:16 still the conductor. Right.
1:32:19 So So I just I just disagree with you on
1:32:22 it.
1:32:25 Guitar then is the reclaiming of full
1:32:30 human authorship. I'm giving my songs to
1:32:32 all humans
1:32:34 to take all AI
1:32:37 out of the final maybe.
1:32:40 But you have less suffering and less
1:32:42 investment. Okay. The less suffering
1:32:44 piece, Rick. You maybe this because you
1:32:47 just got out of the hospital. You do not
1:32:49 have to [ __ ] suffer for art. Like of
1:32:51 anyone, you should know this.
1:32:56 You don't think I can be You don't think
1:32:58 I can be neurotic
1:33:00 and insecure when I'm making things with
1:33:03 AI?
1:33:05 I absolutely can.
1:33:09 I don't often
1:33:11 because sometimes I get really excited
1:33:13 about it,
1:33:16 but I I just it's it's just a different
1:33:18 tool of expression.
1:33:21 It's it's again I go back to um was it
1:33:24 Rick Rubin? Yeah, Rick Rubin.
1:33:27 Rick Rubin was talking about you you
1:33:29 know
1:33:31 there was a time where to be considered
1:33:34 a musician you went to Giuliard
1:33:37 and you learned classical pieces and you
1:33:39 joined a symphony and then you worked
1:33:42 your way up to larger and larger
1:33:43 symphonies. And if you wanted to be a
1:33:45 real musician, you were in a symphony
1:33:46 and it took you decades.
1:33:49 Unless you were a prodigy and then it
1:33:50 took you a decade
1:33:54 and you knew everything. You knew all
1:33:56 the classics. You knew all the theory.
1:33:58 You could improvise. You didn't have to.
1:34:00 You could make fun of people that played
1:34:01 jazz, but you had a respect for them
1:34:03 because it was just a different kind of
1:34:04 knowledge.
1:34:06 And then along came punk rock
1:34:10 and
1:34:13 and you just needed an electric guitar
1:34:17 and and two power cords or three power
1:34:20 cords
1:34:23 and you could start a band.
1:34:27 And what those classical musicians would
1:34:30 argue is, well, that's not real [ __ ]
1:34:31 music. You don't really you don't get
1:34:33 any joy out of making music with three
1:34:37 chords.
1:34:38 Have you ever seen a punk band? They
1:34:40 seem to have a pretty good [ __ ] time.
1:34:49 Your argument is as if you are Stephen
1:34:51 Spielberg and a 100,000 other people are
1:34:54 involved
1:34:56 in your film.
1:34:58 You own the film.
1:35:02 I don't think he owns the film, but what
1:35:04 Stephen Spielberg owns is the vision for
1:35:06 the film. And that I would say he does
1:35:09 own.
1:35:12 Those 100,000 people are working at the
1:35:14 service of his vision.
1:35:18 And that's what I would argue is our new
1:35:20 job.
1:35:23 Our new job is to have the vision. Our
1:35:25 new job is to say, "What tools do I need
1:35:28 to bring together to bring this to life?
1:35:30 As the holder of the vision, Rick, you
1:35:33 could say, "I'm only going to use real
1:35:35 musicians for the music piece of this,
1:35:38 but I'm going to use AI tools for the
1:35:40 special effects because I don't have a
1:35:42 $20 million budget." Okay, that's a
1:35:45 choice.
1:35:48 You know, by your argument, you would
1:35:50 say, well, you you're not really doing
1:35:53 creative special effects because the
1:35:55 people didn't suffer or go through the
1:35:58 pain of making those special effects
1:35:59 frame by frame. You just prompted it.
1:36:03 But if it serves your vision,
1:36:07 it's not better or worse. It's just
1:36:09 different is my point.
1:36:13 Read the next line.
1:36:15 AI gives you the ability to have 100,000
1:36:18 people and a $10 million budget in your
1:36:20 pocket. Yeah, I agree with that. But
1:36:22 it's but is are you saying that that's
1:36:25 not a that's not a value then? I just
1:36:28 think it's different. Like here here's
1:36:31 what I would argue.
1:36:33 You can absolutely argue that the
1:36:35 Mudwater CEO
1:36:37 took the jobs from
1:36:41 20 agency people and a bunch of
1:36:44 photographers
1:36:45 and lighting people and actors.
1:36:49 Or you could say the Mudwater CEO is
1:36:52 going to make, you know,
1:36:55 a hundred videos where he may may have
1:36:57 only made one before and and just that's
1:37:00 going to happen in general. There's just
1:37:02 going to be a proliferation of content
1:37:03 everywhere. And it's not all going to
1:37:05 suck. Some of it's going to be good.
1:37:07 Most of it's going to suck. He's
1:37:09 agreeing with you. Hydrate. You're
1:37:11 starting to sound like Sam Alman.
1:37:13 [Laughter]
1:37:17 Oh, we are Spielberg. But you lost me on
1:37:20 the you don't feel the suffering. Unless
1:37:22 you were being smmy about the suffering
1:37:23 piece.
1:37:26 I get it. I I often find myself arguing
1:37:31 people arguing with people that agree
1:37:33 with me.
1:37:38 I think it's something about called not
1:37:40 listening.
1:37:43 Oh man.
1:37:45 Uh no, I don't want to touch on Z.AI. If
1:37:47 you guys want to go check out there's a
1:37:49 new Chinese chat GPT called Z.AI that's
1:37:52 supposed to be really good. Um anyway,
1:37:54 Rick. Yeah, I know. I figure you and I
1:37:56 are on the on the way. Wait, no, I'm
1:37:58 saying we become the director more and
1:38:01 better because of AI. Okay, got it.
1:38:02 Yeah, exactly. Yeah, like like
1:38:07 we don't have to like right now because
1:38:11 the AI tools are so janky,
1:38:14 we have to be very skilled at how we
1:38:17 navigate which tool to make it not
1:38:20 shitty, right? There's a skill in
1:38:24 backfilling for the jankiness of AI.
1:38:27 Three years from now,
1:38:30 it's it's done.
1:38:33 You you'll just make what you want to
1:38:35 make and you won't need to backfill for
1:38:37 it.
1:38:39 And in that world, what is your skill?
1:38:43 Your skill is have the idea and hold the
1:38:45 space for the idea and keep generating
1:38:47 [ __ ] until you get what you want and
1:38:49 then put it in the world.
1:38:51 Tik Tok pin. Netflix took the jobs of
1:38:55 Blockbuster employees. I'm not mad about
1:38:57 it. Exactly.
1:39:00 AI is going to absolutely take jobs.
1:39:03 Netflix just said, "We use special
1:39:05 effects in this movie you all just
1:39:07 watched. They took jobs." And you know
1:39:09 what they did? They gave jobs to whoever
1:39:11 the [ __ ] prompted those AIs. Someone got
1:39:15 a job.
1:39:18 Um, what are some AI recommendations
1:39:20 platforms for beginners who have
1:39:24 basic knowledge and are broke? Okay, I'm
1:39:28 going to show you something, but I also
1:39:30 want you to go to the AI salon and
1:39:32 become a member of it.
1:39:35 Um, I have it in here, right, Brandon?
1:39:40 It's in irregulars,
1:39:42 that slide.
1:39:45 I think it is.
1:39:53 Oh, I know where I've got it. Hang on.
1:40:01 [Music]
1:40:04 Okay. So, who who was it that asked
1:40:06 there?
1:40:08 I am. I'm intimidated by the salon.
1:40:10 Don't be intimidated by the salon, Amy.
1:40:13 Um, okay. take a screenshot of this. If
1:40:16 if you're talking about um
1:40:20 creative tools, image, video, voice,
1:40:24 songs, all that sort of stuff. Oh, wait.
1:40:27 Hang on. You can't screenshot it because
1:40:30 my big fat head's in the way on Tik Tok.
1:40:32 Wait, are you just on Tik Tok?
1:40:35 Well, if you go to the AI salon, go to
1:40:38 um the Irregulars channel and this
1:40:40 slides in there.
1:40:43 All right, you can still see that in the
1:40:45 upper right hand corner. Okay,
1:40:46 screenshot that.
1:40:48 And
1:40:51 the only
1:40:55 the only tools on here that are not that
1:40:58 don't have some level of freess is
1:41:01 MidJourney and Lumal Labs.
1:41:05 Um, all of the rest of them have at
1:41:07 least some free level to make [ __ ] with.
1:41:10 Okay.
1:41:12 Um,
1:41:14 these are this is not anywhere close to
1:41:17 a comprehensive list.
1:41:19 Um,
1:41:22 but all of these won't won't waste your
1:41:24 time. These are all really good tools.
1:41:26 So, so this is a place to start. It also
1:41:30 depends a lot, Amy, on when you say
1:41:32 where do I start? Where do you start for
1:41:34 what do do you want to start doing
1:41:37 creative work? Do you want to start
1:41:38 doing programming? Do you want to start
1:41:40 doing writing? like if anything that's
1:41:43 that's kind of word processing and
1:41:45 images um just chat GPT if you can if
1:41:49 you can hack the 20 buck a month
1:41:51 subscription I personally think 20 bucks
1:41:54 a month for chat GPT is the best deal in
1:41:57 kind of the history of software
1:42:00 like what it can do and but even the
1:42:02 free version of chat GPT is staggeringly
1:42:05 powerful
1:42:16 All right.
1:42:20 How do we deal with AI idea envy? Even
1:42:24 if the ideas are cheap by nature, the
1:42:27 the way you deal with that issue is it's
1:42:30 a great question.
1:42:32 someone else comes up with an idea and
1:42:34 you're like, "Oh, that's so good." Well,
1:42:36 one is we live in a world where people
1:42:39 are not shy about stealing other
1:42:41 people's ideas. Um,
1:42:45 so if you want to be inspired by
1:42:47 something, be inspired by something, but
1:42:48 then take it and go, "Okay, cool.
1:42:53 I see." Um, if you if you want to be
1:42:56 inspired by something, go see what they
1:42:59 did. Like you can take any image, go
1:43:01 into chatgpt and say make me another
1:43:03 image inspired by this one and it will
1:43:07 or you could say describe this image and
1:43:09 then take that into midjourney and use
1:43:11 that as a starting point to do something
1:43:13 else. Um
1:43:16 the other thing is don't get too
1:43:17 precious about your ideas. Like ideas
1:43:19 are a dime a dozen. Like I've got an
1:43:22 idea for a musical. I've spent the last
1:43:24 year and a half writing it and working
1:43:26 on it and tweaking it and this and that.
1:43:28 Am I worried that someone might steal
1:43:30 that idea? Maybe. Do you know how hard
1:43:32 it is to write a musical that doesn't
1:43:34 suck and then get it produced? Like most
1:43:38 ideas are really hard to actually
1:43:40 execute. So I would say just, you know,
1:43:43 be less precious about the ideas and
1:43:45 more precious about execution. Just go
1:43:48 build build.
1:43:51 Um Rick Maui,
1:43:55 what's happening? Oh, you're muted.
1:43:59 I can't hear you, Rick.
1:44:06 [Music]
1:44:09 You're not muted here, so it's on your
1:44:12 end.
1:44:18 Hey, Rec producer Brandon here. Uh you
1:44:20 may want to check your audio stream
1:44:21 settings uh in the streamard by clicking
1:44:23 the toggle next to the microphone and
1:44:25 making sure that you have your
1:44:27 microphone selected.
1:44:29 Good to see you.
1:44:32 Testing one, two, three. Is that it?
1:44:34 >> That's it. What's happening, man?
1:44:36 >> Okay. Hey, man. I totally agree with
1:44:38 everything you just said. I was just
1:44:40 pointing out that I feel sometimes if I
1:44:43 make a piece of music and it doesn't
1:44:45 cause me any suffering, the long-term
1:44:48 idea was you had to suffer to make art
1:44:50 like you said. Right.
1:44:51 >> Right.
1:44:51 >> So, I feel like I cheated it the and I
1:44:54 got a result that I don't fully own.
1:44:57 >> So, I'm actually making 150 pieces of
1:44:59 music. I'm going to copyright them by
1:45:00 proving 85% came from my human capital
1:45:03 investment. Right? Then I'm going to
1:45:05 gift it to the world and I'm going to
1:45:06 say any artist who can make a recording
1:45:08 of this and start to get traction, I
1:45:11 will give 60% of the ownership of that
1:45:13 copyright for eternity for that piece of
1:45:15 music. So I'm extracting all of the AI
1:45:18 that is basically the performance
1:45:20 although I've cloned my voice on many of
1:45:22 the songs.
1:45:24 >> So that's the first point. And the
1:45:25 second point is I think we are becoming
1:45:28 Stephen Spielberg's with the power of AI
1:45:31 allowing us to do what Spielberg does.
1:45:33 And you would never deny him best
1:45:36 directors
1:45:37 Oscar. Right.
1:45:38 >> Right.
1:45:39 >> But we all have the potential and it's
1:45:41 only limited by our imagination and our
1:45:44 effort.
1:45:45 >> Yeah.
1:45:45 >> And our genius that's like a Steven
1:45:48 Spielberg once in a billionyear genius
1:45:52 level human.
1:45:53 >> Yeah.
1:45:53 >> To put to work the AI for their benefit.
1:45:56 So you and I, Shannon, are producing
1:45:58 objects that exist only because we, the
1:46:01 human, drove them there,
1:46:03 >> right?
1:46:03 >> So it's a car in your in your driveway.
1:46:05 But some people are really good at being
1:46:07 Mario and Drey driving that car and some
1:46:10 people are creating AI saw.
1:46:12 >> Yeah. Yeah.
1:46:12 >> And so we agree on everything else and I
1:46:15 think we agree on all of this. Yeah, I
1:46:17 think it I get I get the music thing,
1:46:20 but I I would argue, Rick, that there
1:46:23 are songs that I've created with my
1:46:24 guitar that I've completely forgotten
1:46:26 that that didn't stick. And there there
1:46:28 are songs that I've created with AI.
1:46:30 Like there are some things that are just
1:46:31 brain farts that feel good in the moment
1:46:33 and then are gone.
1:46:34 >> Right. Right.
1:46:35 >> And like I've done that with analog
1:46:37 songs. every song I write today, all 100
1:46:41 of them that I would have not been able
1:46:43 to get to production level quality to
1:46:45 even show someone because I don't have
1:46:48 the money to pay for a $1,000 studio
1:46:51 session. Yeah.
1:46:52 >> And I don't have the skill sets to sing
1:46:53 it to the way this does.
1:46:55 >> Yeah.
1:46:56 >> So, what I'm doing is I'm taking the AI
1:46:57 out at the end, but everything is
1:46:59 written on 65 years of life.
1:47:01 >> Yeah.
1:47:02 >> Every first kiss
1:47:03 >> and every time I held the hand of
1:47:05 somebody dying, that built what made me
1:47:07 who I am. the same way Spielen Spielberg
1:47:10 did all his stuff based upon his entire
1:47:12 life's experience. He's making the same
1:47:14 film over and over again.
1:47:16 >> So artists to me are people that have
1:47:18 have become so human that they have
1:47:21 something to say that every human
1:47:23 relates to and that's you and that's me.
1:47:26 >> Yeah. Those are the people we should
1:47:28 should be adapting and doing whatever
1:47:30 crazy [ __ ] they got to do for the rest
1:47:32 of their life to get to the point that
1:47:34 someone finally hears the genius that
1:47:36 was locked inside prevented to get there
1:47:38 because I couldn't hire the people to do
1:47:40 it.
1:47:40 >> Right. Yeah.
1:47:41 >> Now there's no limit.
1:47:42 >> Yeah. Exactly. Well, and that's
1:47:43 >> so it's a new measure of what is art is
1:47:47 that who can produce something so
1:47:49 resonant in every soul that they become
1:47:52 Stephen Spielberg.
1:47:53 >> Yeah. Yeah. And then what happens is the
1:47:55 tool you use to create it is irrelevant,
1:47:57 right?
1:47:58 >> Yes.
1:48:01 >> It becomes the rocket ship in your
1:48:03 driveway. Your destination and how you
1:48:05 get there is fueled by you.
1:48:06 >> Yeah. Exactly. That's it.
1:48:08 >> So if we're all the director of the
1:48:09 destination, everybody's got a rocket
1:48:11 ship in their in their pocket, which
1:48:12 they do right now. The extension of the
1:48:15 humanity, the final grasp is all your
1:48:17 humanity summed up into what are you
1:48:20 going to do and what which choice are
1:48:22 you going to grab? That's the Rick Rubin
1:48:23 part.
1:48:24 >> Y
1:48:24 >> So now you can take your genius and
1:48:26 you've got a rocket ship and then now
1:48:28 you know how to get to any planet in the
1:48:30 world.
1:48:31 >> Yeah.
1:48:31 >> And directing it and then when you get
1:48:33 there, you own the Oscar. That's my
1:48:35 point.
1:48:35 >> Yeah. Love it. All right. Good.
1:48:37 >> Okay.
1:48:38 >> So, we agree. Thank you so much for
1:48:39 having me on.
1:48:40 >> Yeah, man.
1:48:40 >> Anytime you want, I'm I'm up for it. But
1:48:42 but yeah, we totally agree. I just
1:48:44 wanted to make sure it was clear what I
1:48:45 was saying because it's
1:48:46 >> Oh, yeah. I totally missed I totally
1:48:48 missed the point, but I got it. A and
1:48:50 the whole point of my life now is every
1:48:52 time I make a post, Sh uh Kyle, I always
1:48:55 add value first. I respect the person
1:48:57 and I'm going to own the respect to you.
1:49:00 And that's why I want you to know that
1:49:02 you're so important because if all of us
1:49:05 added value to say, thank you for
1:49:08 inspiring me to write this post.
1:49:10 >> And then you go and you make a new post
1:49:12 and you quote Kyle Shannon as being the
1:49:14 origin of that. If we all did that,
1:49:16 social media becomes an agent for
1:49:18 amplifying the positive and we get rid
1:49:21 of the [ __ ]
1:49:22 >> and all those flame wars get, you know.
1:49:24 So that's my anyway, that's my
1:49:25 philosophy. Two bits and thank you for
1:49:27 your time. Appreciate your show.
1:49:28 >> Peace out. Good to see you, man.
1:49:31 >> All right, everybody. So listen,
1:49:33 tomorrow night, um, AI Salon, wait,
1:49:36 there's Tik Tok pin from Tams. Steal
1:49:39 like an artist is one of the best books
1:49:41 I've ever read, and it speaks to even
1:49:43 this. Yeah, exactly. Yep.
1:49:45 Yeah, listen
1:49:47 the there there there are what nine
1:49:50 stories. Every story is some rehash of
1:49:53 the nine the nine stories. You know,
1:49:56 everyone copies from someone. Every
1:49:58 everyone's inspired. You know, good
1:50:00 artists what is it? Good artists
1:50:02 plagiarize, great artists steal, right?
1:50:07 But it but it's not even that. Like the
1:50:09 the thing that I want to the the thing
1:50:11 that I am
1:50:14 actively trying to
1:50:17 fight against
1:50:20 is that the comment that was made
1:50:22 earlier, it feels like cheating.
1:50:25 It feels like cheating. It's not
1:50:28 cheating. It's a different tool.
1:50:31 You know, if if you used to have to walk
1:50:34 across
1:50:36 America,
1:50:38 I would imagine that getting in a train
1:50:41 felt like cheating,
1:50:43 right?
1:50:46 This is cheating for our brain, right?
1:50:48 This this is the steam engine for our
1:50:50 brain
1:50:52 is what AI is. And so, yeah, is it not
1:50:55 like we did it before? No, it's not. Is
1:50:58 it have have we associated for seven or
1:51:02 eight generations time to value? Yes, we
1:51:04 have. And so our parents told you put in
1:51:08 the time. Take the time. You got to put
1:51:11 in the time or it's not worth [ __ ] And
1:51:14 now all of a sudden you can just do it.
1:51:15 So you go, oh [ __ ] it's not worth [ __ ]
1:51:17 because I didn't put in the time.
1:51:20 That's from the robber barons
1:51:23 of the late 1800s.
1:51:25 [Laughter]
1:51:28 You don't work for them. You don't work
1:51:30 for them anymore.
1:51:34 [ __ ] them. You can do your two hours
1:51:36 worth of work in five minutes. And and
1:51:39 you can represent that this Yes, this is
1:51:41 as good as what I would have done and
1:51:43 you turn in that work. Who gives a [ __ ]
1:51:45 how long it took you?
1:51:47 Damn it. All right. So, tomorrow night
1:51:51 we have AI Salon.
1:51:54 AI salon presents.
1:51:58 We've got Mike Lavine, who's the CEO of
1:52:00 Movie Flow, which if you know LTX
1:52:03 Studio, it's very much like that, but
1:52:05 it's a bit more sophisticated. Um, and
1:52:07 he's a really smart guy with a really
1:52:09 impressive background. So, tomorrow
1:52:11 night, um, go to, uh, the the salon.ai
1:52:16 or go to the, you know, the AI salon
1:52:18 community site, community.the the
1:52:20 salon.ai. Um, look at events. And
1:52:23 tomorrow night is the um is is AI salon
1:52:28 presents. Okay. And that means we're
1:52:29 going to be late for this. So, tomorrow
1:52:31 night this show will start 8:30 or 9.
1:52:34 I'll try to do 8:30, but it depends on
1:52:37 how much energy I got for you people.
1:52:39 All right.
1:52:41 It will take a generation. It might take
1:52:43 two.
1:52:45 Yeah. It took us, look, it took us, it
1:52:47 took us seven generations to just
1:52:49 accept,
1:52:51 you know, we go in, we put in our hours,
1:52:53 we do our thing. Our job is to be a cog
1:52:56 in the wheel. Even though our work is
1:52:58 completely mind-numbing and soul
1:53:00 crushing, we're going to call that
1:53:03 fulfilling work because it's tied to
1:53:05 some larger cause.
1:53:08 Well,
1:53:09 we now get to we're now entering a
1:53:11 different era where
1:53:15 think of it like every person is going
1:53:17 to have a hundred or a thousand person
1:53:21 equivalent set of employees that work
1:53:24 for them.
1:53:26 Every single person that taps into this
1:53:28 AI [ __ ]
1:53:32 will have
1:53:34 an organization's worth of PhD level
1:53:38 people that will do anything they want
1:53:40 them to do. Anything you task them to
1:53:43 do, they'll go off and do it at a
1:53:45 worldclass level. If you had a hundred
1:53:48 employees that were world class at
1:53:50 anything you wanted them to do, what
1:53:52 would you do?
1:53:55 If you haven't asked yourself that
1:53:56 question, start asking yourself that
1:53:58 question. Start having a point of view.
1:54:00 Start wondering what difference you want
1:54:03 to make in the world.
1:54:06 Rather than just worrying about the AI
1:54:08 is going to take our jobs, well, what
1:54:09 about the AI is going to give you a
1:54:10 [ __ ] company with a thousand people
1:54:12 in it?
1:54:14 Do you ever think about that?
1:54:17 No. Ain't never been an entrepreneur.
1:54:20 I'll tell you what, it's a lot easier
1:54:21 with a thousand people that don't suck
1:54:24 and work 247 and know what the [ __ ]
1:54:27 they're doing
1:54:29 than, you know, you hired Mary because
1:54:32 she knew someone who knew someone and
1:54:34 you realized when she got there she's
1:54:36 clueless.
1:54:40 Now you got to work with her for three
1:54:42 months because you don't want to ruin
1:54:43 that relationship.
1:54:44 [Laughter]
1:54:48 All right.
1:54:50 This is this was a good combination of
1:54:53 meltdowns, arguments
1:54:56 with someone who agreed with me, you
1:54:58 know. It was good. It was good. Solid
1:55:00 chat add night. We got a little bit of
1:55:02 demo in there. All right, everyone. I'm
1:55:05 out of here. Um, hope you have a good
1:55:07 time and I hope you see you tomorrow at
1:55:09 the salon. All right, bye.
1:55:13 [Music]