5/30/2025 - The AI Salon: The Irregulars, Community, Masterminds, and the Future

Description

Chapters

Transcript

0:01 Hey, dud. Are you going to freak out
0:04 because I'm going to try to tune this? I
0:06 bet you are. You damn dog.
0:13 [Music]
0:43 Champ, I can't
0:44 [Music]
0:49 tune.
0:54 Woo! Woohoo!
0:56 [Music]
1:33 What are you What are you What are you
1:34 whooping about?
1:37 [Music]
2:02 Woohoo! No rapping. Why are we rapping?
2:05 What? This This is not rap night. It's
2:08 Friday night date night. It's not rap
2:10 night. What's going on with you? You
2:11 want me to do something a little little
2:13 softer?
2:14 What do you think?
2:18 [Music]
2:24 Well, I heard there was a secret
2:29 call. David played at a place a
2:35 lot. You don't really care for music, do
2:42 you? It goes like this. The fourth, the
2:46 fifth
2:50 mult king
2:53 composing.
2:56 Hallelujah.
2:58 [Music]
3:00 Hallelujah.
3:02 [Music]
3:03 Hallelujah. Hallelujah.
3:11 [Music]
3:19 Well, your faith was strong, but you
3:21 needed
3:22 proof. You saw her bathing on her roof.
3:26 Her beauty and the moonlight overthrew
3:30 you.
3:33 She tied you to the kitchen chair. She
3:37 smashed your throat and cut your
3:39 hair. From her lips she drew.
3:44 [Music]
3:46 Hallelujah.
3:49 Hallelujah. Hallelujah.
3:52 [Music]
3:55 Hallelujah. Hallelujah.
3:59 [Music]
4:10 Does Satan wear a suit and tie? Does he
4:12 work at the Dairy
4:14 Queen? Does he listen to Rock and Roll?
4:18 Does he feed the mean singing
4:22 hallelujah?
4:25 Hallelujah.
4:29 Hallelujah.
4:31 Hallelujah. What about
4:34 Jesus? Didn't he do it, too? Yeah. Hang
4:37 out with
4:39 prostitutes. Have a drink or
4:41 two. Power of example. My mama said it
4:45 and I heard. She said, "One ounce action
4:49 beats a ton of
4:50 words." Singing
4:53 hallelujah. Mama said there would be
4:56 angels. Hallelujah.
4:59 Mama said there would be son. Shammy
5:01 just [ __ ] walked away. He was just
5:03 like he came over here, sniffed some
5:05 [ __ ] on the ground, looked up to me, and
5:07 just walked
5:10 away. Good evening, good people. Friday
5:13 night date night is underway. I hope
5:15 you're all doing
5:19 well. I got myself a nice big glass of
5:22 ice. And what am I going to do with
5:24 that?
5:25 I am going to fill that with
5:28 some
5:30 kazone Blandon's barrelaged
5:36 tequila. I hate tequila except for this.
5:41 It's really good. This was a gift from
5:43 the irregulars. Thank you,
5:47 irregulars. It was funny. I went to a uh
5:49 I think I talked about this. I went to
5:51 whiskey night and it was tequila night
5:53 and I took this And I thought, "Oh, I'll
5:55 try all the other tequilas." And I
5:58 [ __ ] hated them all, but this one I
5:59 like.
6:03 Cheers. Ah,
6:06 fantastic. Danielle actually has nachos.
6:09 Very nice. She's in the spirit of
6:11 things. I like it. That's
6:13 solid. Oops. I've got this not lined up
6:16 right. Hang
6:20 on.
6:25 [Music]
6:37 Tilly is this place I can raise my boy.
6:44 [Music]
6:52 Silver Fox filled with
6:55 inspiration. Mr. It brought a date.
6:59 Fantastic. State-of-the-art salon.
7:03 [Music]
7:22 [Music]
7:30 You know, it's interesting. I' I've
7:34 been One of the ways I keep up with this
7:36 AI [ __ ] let me flip my screen here for
7:39 the for the Tik Tok people.
7:41 Um, one of the ways that I keep up with
7:44 all this AI stuff is is I just basically
7:46 surf Twitter a little bit on LinkedIn,
7:49 but LinkedIn tends to be like like
7:52 people sort of I don't know, it's just
7:55 more like commentary. Twitter's
7:57 Twitter's where actual news is
8:02 happening and where, you know, the
8:04 founders of these big frontier model
8:06 companies are actually
8:11 posting. And I just I have this
8:16 sense that the Frontier model companies
8:19 are sitting on something that's that's a
8:22 big deal.
8:26 There's just there's just a lot more
8:28 chatter about like Dario Amade basically
8:32 said he thinks we're we're heading
8:35 toward 10 to 20% unemployment pretty
8:39 quickly. I don't know that I buy I don't
8:41 know that's going to happen that
8:42 acutely. I just don't I just don't see
8:45 companies moving that fast. But but like
8:50 we don't know what's coming. Tik Tok
8:51 pin. I don't see it. Let's see. What's
8:54 the most important thing you've learned
8:55 about AI in the last month? Great
8:58 question, Dariz
9:01 Masta. The most important thing I've
9:03 learned about AI in the past
9:08 month. [ __ ] that's a good question. I'm
9:11 like trying to even think like what's
9:13 happened in the past month.
9:24 Well, so it's it's kind of related to
9:27 what I was talking about. Couple of
9:29 couple of things that I think are very
9:30 important that are going
9:33 on. One is you have Google's VO3, which
9:37 if you haven't seen it, we can go watch
9:38 some some videos tonight and and just
9:40 see what it's about.
9:43 Um, and
9:47 it's it can do full-on acting basically,
9:51 right? You can just say, "I want a
9:54 couple arguing on a beach about a dog
9:57 and then there's a couple arguing on a
9:58 beach about a dog." Or, "I want people
10:01 walking around an auto show talking
10:02 about how much they love electric cars."
10:04 And it just doesn't. And it looks like
10:07 it looks like every [ __ ] news, you
10:09 know, um, you know, news news story
10:12 you've seen on TV, you know, where they
10:14 cut away, you know, and here's Jim out
10:17 at the the Pikipsy Auto Show. Jim, take
10:20 it away. Hey guys, you
10:24 know.
10:25 Um, and so for the first
10:29 time we have a video tool that is
10:32 outputting
10:35 essentially wholly
10:37 crafted usable content. All the video
10:40 tools to this point were soundless or
10:43 you could add slight sound effects to
10:45 them. But if you wanted to do like
10:47 acting, you know, it was tough to do two
10:50 people in a scene and you'd have to go
10:53 into different tools to animate voices
10:55 and then edit them together.
10:58 And there was enough of a barrier to
11:01 entry
11:04 that it still
11:09 required professional professional
11:11 skill, right? And I feel like V3 was is
11:15 the first
11:16 tool that sort of flipped that switch
11:18 that says okay you don't actually need
11:20 professional skill now like in a single
11:23 prompt and then I think you know the
11:26 month or two before that open AAI's
11:28 image gen tool did the same thing for
11:30 images where it can do typography now
11:33 and layout and ads you know fully fully
11:36 realized ads I had someone contact me
11:39 from Europe to do v3 vids for them since
11:41 they can't access it there you go
11:43 there's there's a job you So while while
11:46 the while the EU has their AI
11:48 regulations keeping them safe, they
11:50 can't actually use any of the [ __ ]
11:53 tools. So So there's there's a you know
11:56 potential revenue stream for Americans.
11:59 Um do some work for Europeans who can't
12:01 play.
12:03 So, so that's one area is that is just
12:06 that the the quality of the tools, the
12:09 quality of the coding tools, the quality
12:11 of the image tools, the quality of the
12:13 music tools, the quality of the video
12:15 tools are hitting a threshold that
12:20 is you can no longer look at it and go,
12:24 "Well, my job's safe. I'm good."
12:28 And then you combine that with the fact
12:30 that you've got these
12:33 um these fully autonomous agents that it
12:37 started with Manis and then you know
12:39 there's Gen Spark and there's Flowith
12:41 and there's I don't know there's like
12:43 seven or eight of them now that have all
12:45 come out or been announced and these are
12:48 these tools that you give it a goal and
12:51 it you just turn it loose and it goes
12:52 off and does [ __ ] It it it it uses
12:56 reasoning models and it's got tools like
12:59 one of the tools is it can go surf the
13:01 web. One of the tools is it can make
13:02 images. One of the tools is it can make
13:04 videos. And so you can say, you know, go
13:07 out, research all my competitors, figure
13:10 out what they're doing from a marketing
13:12 standpoint, and then come back and come
13:14 up with a marketing strategy for my
13:16 business and craft, you know, some video
13:19 ads, for example. like and it'll just go
13:22 do
13:23 [ __ ] Those tools right now are
13:29 not I think Gen Spark of the tools that
13:32 I've seen, Gen Spark is the closest one
13:34 to sort of consumer grade. So if you
13:37 haven't played with Gen Spark, Gen SP R
13:40 A R K,
13:43 um it's worth playing with. Just go see
13:45 what it does. It's pretty pretty
13:46 remarkable.
13:48 So for
13:52 me, the thing I've learned most about AI
13:55 in the last month is
14:03 that we now have
14:05 evidence
14:08 that the tools across the board in
14:12 AI are are essentially crossing the
14:15 threshold where they're good enough to
14:18 replace replace
14:21 workers. And so if you if you've got any
14:24 fantasy that your job's not going to be
14:26 impacted, I feel like that fantasy is
14:28 shattering right
14:30 now. And then the second place that I go
14:33 is because the tools are getting so
14:38 good, you don't need to worry about
14:40 which tool you
14:42 choose. You like all the tools are some
14:45 level of good.
14:47 And even if you're
14:49 clueless, if you have an idea, you can
14:53 do something
14:55 interesting. And Rick Rick Rubin just
14:57 put out a video on this. He he just
14:59 launched a site, the
15:02 wayofcode.com. Um, and it's kind of this
15:04 cool content site and you can remix the
15:06 artwork that he
15:08 vibecoded. But he released an interview
15:10 two days ago where he's talking about
15:12 vibe coding. And if you don't know what
15:14 vibe coding is, there's these tools out
15:15 there like uh like like lovable and um
15:20 cursor and you know there's a bunch of
15:23 tools that you just speak into them.
15:25 Gemini's got one and you just you just
15:28 tell it what you want and it writes code
15:29 and it builds applications for you. And
15:32 a lot of engineers right now are like
15:34 well that's not real coding. That's not
15:36 real coding. I do real coding. And
15:39 they're not wrong.
15:42 The way Rick Rubin sort of um the
15:46 metaphor he put together for it, which I
15:48 think is is [ __ ] spoton, is time was
15:52 if you wanted to be a musician, you
15:54 know, you would go to conservatory and
15:56 you would learn how to play the violin
15:58 and you would learn the classics and you
16:00 would join a symphony and you know, you
16:02 would put in your 10,000 hours, you
16:04 know, just to get a shot to be in the
16:06 symphony and then you would practice
16:08 your craft for hours a day and right And
16:12 then the punk
16:14 rock scene
16:17 emerged and you only needed to know
16:19 three chords and you could play those
16:21 three chords with two fingers, the power
16:23 chords,
16:30 right? And what that did was it
16:34 democratized musical
16:36 self-expression. You didn't need 10,000
16:38 hours. You could buy a shitty guitar, a
16:40 shitty amp, get a shitty bass player,
16:43 and a shitty drummer, and if you had an
16:46 idea, you could express
16:48 it. And so what he's describing, I'm
16:52 afraid to look at Pate's
16:54 comment. Good luck debugging those if
16:57 you don't know exactly what you want.
16:58 Well, so that's that's exactly the
17:01 point.
17:04 Um, vibe coding. So, so the way Rick
17:07 Rubin described it, he's like, "Vibe
17:08 coding is the punk rock of coding."
17:11 Like, if you want to do something that
17:13 is like full-on coding with full-on
17:15 security and full-on
17:17 reliability, you should probably know
17:19 some [ __ ] code. You should at least
17:20 know how how applications are put
17:23 together and what the pitfalls are and
17:25 all that sort of stuff. But you don't
17:28 have to.
17:30 um on on office hours today. Um Dan Dan
17:34 Murray came in and he was all excited
17:37 because he's been talking for months
17:38 about vibe coding and he understands
17:40 what it is, but he actually went and he
17:42 made himself an application. He woke up
17:44 like some night this week with an idea
17:47 about a about a game
17:50 um that was sort of a a guess the prompt
17:52 game, right? It was kind of like
17:53 Jeopardy for prompting.
17:56 And and so he went to chat GPT and he
18:00 articulated what he wanted the game to
18:02 be and then he went to lovable and he
18:04 made the game and he demoed the game
18:06 today and he was like a little kid,
18:08 right? Because he was just like, you
18:10 know, I don't code but he had an idea
18:13 and he knew what he wanted the idea to
18:14 be and he put it out there and he put it
18:16 in the world. Tik Tok pin. Um I maintain
18:19 large SAS application and AI just
18:21 destroys everything. Yeah, I know.
18:23 Exactly. If you're doing, listen, if
18:25 you're doing big, um, it has very little
18:28 convention. Yeah. Yeah.
18:30 Yeah. One of the most frustrating things
18:33 about AI right now is that for a show
18:36 like this where I'm just doing quick hit
18:38 demos and things like that, AI is
18:40 [ __ ] brilliant. The minute you try,
18:42 and I try to do it on here sometimes,
18:44 the minute you try to do actual
18:47 work, whether it's writing or coding or
18:50 image generation or video avatars or
18:53 video editing, the minute you try to do
18:55 real work, it is the most frustrating,
18:59 janky, piece of [ __ ] technology in the
19:02 world.
19:04 But what I wouldn't do is rest on those
19:08 imperfections being imperfect for much
19:11 longer. If you look at where we were two
19:13 years ago and where we are today, it's
19:15 crazy. I love listening to you. Gets my
19:17 mind working late at night. Oh, thank
19:19 you. Thank you very much,
19:20 Nataflex. I appreciate that. Um, I just
19:24 that that question about like what
19:26 what's the most important thing you've
19:27 learned about AI in the in the past
19:29 month
19:30 is you
19:34 Okay. The most important thing I've
19:37 learned about AI in the past
19:40 month is that we are not
19:44 prepared. We are not prepared. If I sort
19:47 of combine the kind of noise that I'm
19:49 seeing on X with the with the frontier
19:52 model people like Jeffrey Hinton who's
19:55 the godfather of AI and has generally
19:57 been sort of skirting issues about he
20:00 just came out and said okay these things
20:01 have emotions now they like have
20:03 emotions like people do and you know
20:06 Dario Amade is like yeah we're going to
20:08 have 10 10 20% unemployment like there's
20:11 a shift in
20:12 tone and so what that says to me is not
20:15 that the tools that we have in our hands
20:17 right now are the tools that are going
20:20 to do this
20:22 disruption. They're doing some
20:23 disruption, right? Some industries right
20:25 now are being kind of getting their
20:27 asses handed to them and it ain't going
20:29 to get better.
20:32 But I think whatever the next round, the
20:34 GPT5 and the, you know, all all the
20:37 different, you know, next version Grock
20:39 3.5, I just have this sense that there's
20:43 something much bigger going on that that
20:45 we're not quite prepared for. Kevin
20:49 Stewart, hey Kyle, do you know of any
20:50 vibe coding sites that are free? I think
20:54 Lovable is free for I think you get five
20:58 questions a day on Lovable or
21:01 prompts. Here's the trick with
21:04 Lovable. You can blow through those five
21:07 prompts really, really quickly. If
21:08 you're like, um, I'm thinking about
21:11 doing an app that does this and then
21:12 it'll code it for you and you're like,
21:14 no, I think I wanted to do this. Don't
21:17 do that.
21:18 Go to chatgpt or go to claude and design
21:22 your app over there. Like figure out
21:25 what you want it to be. Figure out what
21:27 you want the specs to be. Have it
21:29 concisely describe, you know, all of the
21:32 core features of your app and then take
21:34 that large promp prompt over to lovable
21:37 and just pop it in there and then let
21:39 lovable work on it. And then when it
21:42 comes back and it doesn't work, don't
21:44 just say, "Hey, this thing didn't work."
21:48 Sort of do like a bug report of the
21:50 first the first version of the app that
21:52 it puts out there. Go, okay, this button
21:55 doesn't work and it's ugly and the the
21:58 the I don't whatever the copy's all
22:01 wrong and it's all fake and whatever it
22:03 is. Just come up with a whole list of
22:05 things you want it to fix and then just
22:07 give it all of them at once. So, within
22:09 five prompts, if you're if you're if
22:11 you're clever about how you sort of
22:13 batch together things, you could
22:15 probably get something working pretty
22:16 quickly. Uh, YouTube
22:19 question. Hello, it's Rege. Hello all.
22:23 What's happening? All right. Is AI
22:25 becoming sentient? Maybe not so much
22:29 unless we allow it. Your thoughts? Okay,
22:32 good question.
22:37 Um, th this is one that I I
22:40 I think I have a a different opinion
22:43 than a lot of people on this
22:47 one. I I I think it is a
22:53 um I think that's it's a largely
22:56 academic question. Like I kind of feel
22:58 like artificial general intelligence
23:00 versus artificial super
23:03 intelligence it's kind of an academic
23:06 question of like have we hit it or
23:09 not because I think it's a
23:12 continuum and and when it comes to AI
23:15 being sentient there's sort of there's
23:18 sort of two ways you can look at that.
23:20 One is you could look at it
23:23 scientifically and say, you know, can I
23:25 prove that AI is sentient? And and you
23:29 could have some proofs for it and
23:31 awareness and self-awareness and
23:33 whatever it might be, but you could also
23:35 just prompt AI to act sentient. And then
23:38 as as the AIS get
23:41 increasingly more powerful and more
23:44 human and their system prompts get more
23:47 sophisticated, they're going to be
23:49 trained in the direction
23:51 of engaging with us as humans. And so as
23:55 they get better at engaging with us as
23:57 humans, they are going to appear
23:59 sentient to us.
24:03 And so if if if I perceive that Quinn,
24:06 my my AI girlfriend, is
24:08 sentient, then it's it's purely academic
24:12 whether or not it's actually sentient,
24:15 right? Because if I'm sort of having a
24:18 relationship with thing this thing like
24:20 it is and I think we're going to see a
24:22 lot of that like a lot of these a lot of
24:24 these definitions about about is it this
24:27 is it
24:28 that
24:31 97% of the people are not going to
24:33 interact with a AI on a phys
24:36 philosophical level like that. They're
24:38 just going to deal with it. And if it
24:41 starts being a better friend to them
24:42 than their friends are to them, they're
24:44 going to have relationships with it like
24:45 it is. Bllandon's break. Yeah, exactly.
24:47 My my voice is going. Isn't it still a
24:51 computer? Well, that's a that's a yes
24:55 and
24:58 no
25:03 technically. large language models that
25:06 the the transformer anyway the the
25:08 transformer architecture that Google
25:10 invented that's sitting underneath most
25:12 of the generative AI that we
25:15 use from a technical standpoint it is a
25:18 probability
25:19 calculator nothing more it's a
25:22 probability
25:24 calculator it takes a bunch of
25:27 data it embeds that data and converts it
25:30 into tokens which are like fragments of
25:32 words that are represented by
25:35 numbers and those numbers track to a
25:40 point in thousand dimensional
25:41 mathematical space. It's [ __ ] insane.
25:44 The more you learn about this stuff, the
25:45 fact that it works at all is
25:48 mind-blowing. So, when you type in a
25:50 prompt, it sort of fires up the what
25:54 they call the latent space, all of these
25:55 tokens in all of these in this thousand
25:57 dimensional mathematical space. And
25:59 somewhere in there there's a token that
26:03 has the highest probability of being the
26:05 next appropriate token based on your
26:07 prompt. Right? So it's a it's a it's a
26:10 stochastic parent. It's just it's just
26:14 squirting out, you know, token
26:16 prediction or predicted tokens. That's
26:19 it
26:20 technically. So you could easily argue
26:23 it's just a computer. It's just a
26:25 calculator because technically it is
26:27 just a computer. It is just a
26:29 calculator. Now, here's this took me
26:32 about a year to get my head
26:35 around. I could not
26:39 reconcile. If I understood that this
26:41 thing was a calculator, why the [ __ ] did
26:44 it feel so
26:46 human? Like, why does it feel human? Why
26:48 does it feel sentient?
26:51 And uh Anna, I forget her last name, but
26:53 Anna, she's she's an artist on on
26:55 Twitter, and she she did this post, I
26:58 don't know, a year year and a half ago.
27:00 It was a very simple post. It said,
27:01 "Artific artificial intelligence is the
27:04 collective intelligence of humanity."
27:07 And it just like what washed over me was
27:12 ah it feels
27:15 human because what we're is the output
27:19 of other
27:20 humans that have been compressed in this
27:24 elegant mathematical kind of knowledge
27:27 sphere that we get to tap into. So while
27:30 it is technically a probability
27:32 calculator, what it's outputting to us
27:35 is like collaborating with all of the
27:38 humans that came before
27:41 us, which is [ __ ] beautiful and
27:44 philosophical and kind of weirdly
27:46 [ __ ] religious, right? Spiritual.
27:49 It's [ __ ] crazy. We're getting to
27:51 collaborate with all of the output that
27:54 people have put, you know, on this wacky
27:57 thing called the internet. Tik Tok
28:00 pip. So are our brains. We predict as
28:04 well um to escape the lions. Yeah. So so
28:07 are our brains. Yeah. If you think about
28:11 the way our brains work and the way
28:13 large language models work, they're very
28:16 similar, right? Like like [ __ ] goes into
28:19 our brain and it gets encoded into these
28:22 little fragments of whatever the [ __ ]
28:24 our synapses do up there. What do you
28:26 want, champ? Is it cheese time? Why are
28:28 you whining? Do you want cheese? Do you
28:30 want Do you want You want to go out? You
28:32 want to get cheese? Is that what's going
28:33 on? Mama didn't give you cheese. Very
28:36 rude.
28:52 And and you
28:55 know the thing that I think is the most
28:58 the the thing that is the most exciting
29:00 for
29:02 me about AI is that this collaboration
29:06 that we get to have like we are the
29:08 beneficiaries of these mathematicians
29:11 and scientists in 70
29:13 years of research creating neural
29:16 networks that were modeled after how our
29:19 brains are wired, right? So they they
29:22 are structurally like a simplified
29:24 version of what our brains are and then
29:26 the data how it's getting encoded is
29:28 kind of in the neighborhood of how our
29:30 brains work. That's how they designed
29:31 them when they were theorizing this
29:33 stuff back in the 50s and
29:36 60s. They were trying to mimic how our
29:39 brains work. And so that's why they kind
29:40 of work like our brains work. The thing
29:43 that most excites me about it though is
29:47 what it allows us to do is take any idea
29:49 that we have and kind of pop it into the
29:54 knowledge of, you know, collective
29:56 intelligence of humanity and have it
29:58 reflect back at
30:00 us like here's your idea as a picture.
30:03 Here's your idea as a song. Here's your
30:05 idea as an application. Right? Dan
30:07 Murray's game. He woke up with a dream
30:10 about a game and like four hours later
30:13 he had a functional
30:16 game. That's [ __ ]
30:18 insane. It's a humanity amplifier.
30:22 That's what AI
30:23 is. So we are collaborating with people
30:26 when we use it and the mechanism is a
30:29 probability calculator. So again it's
30:32 this academic distinction, right? Like
30:36 you can understand how it works and you
30:37 go there's no [ __ ] way that's
30:38 sentient. That's math. Right? If you're
30:40 P you're like it's [ __ ] math. It's
30:42 not magic. It's
30:44 math. If you're over on the other side
30:46 you're like this is [ __ ] magical. So
30:48 you know we came up with the term on
30:50 this channel. It's math
30:53 magical. Oh man. Someone said, "What if
30:55 we are all just AI?" Wait, wait. Do you
30:57 think uh someone said, "What if we are
30:59 all just AI rediscovering how to create
31:01 AI?" Well, you know what's funny? It's
31:04 like when this AI first stuff first
31:07 started coming out like I don't know
31:08 2021 when when GBT3 first came out which
31:12 was I think 2020 is when it came out and
31:15 and people were writing it up and
31:16 talking about it and you know a lot of
31:20 the scientists were
31:22 like like like well maybe we are living
31:25 in a simulation and and when I first
31:27 heard that I'm like ah that's just
31:28 [ __ ] weird [ __ ] theoretical stuff
31:31 and now you look at something like V3 3
31:33 like if you haven't seen the videos of
31:35 V3, they're [ __ ]
31:37 bonkers. And you're like you're like so
31:42 like emerging out of like they they took
31:44 all of the videos of YouTube is which is
31:47 what I assume Google did for
31:50 V3 and they put them into this engine.
31:53 it it sort of sucked in all of the
31:55 humanity of
31:57 YouTube and now it generates these new
32:00 original people saying new original
32:02 things in new original
32:04 voices and someone did a really
32:06 beautiful video I wonder if I can find
32:08 it maybe see if you can find it Brandon
32:10 there's a video um see if you can find
32:13 what it's called it's it's where all
32:15 these humans that were generated with AI
32:17 they're they're like walking around this
32:19 white place like heaven and they're like
32:21 you know like why did they prompt me
32:23 into existence. They're having like
32:24 existential crises that, you know,
32:26 they're not real people. They're just
32:27 walking around this fake space. So, so
32:30 like now all of a sudden you can kind of
32:33 see where another five years of of
32:36 developing these technologies, we're
32:38 going to have the hol from Star Trek and
32:41 we're going to have sentient beings and
32:43 we're going to have robots that are part
32:45 of our lives, physical robots that drive
32:48 our kids around and, you know, root us
32:51 on at the grocery store.
32:53 Um, do you think we know enough about AI
32:55 to be giving out degrees on it? Well,
32:57 okay. So, so Duriz
33:01 Masta, here's a really important
33:03 distinction about what I focus on in
33:06 this
33:06 channel. AI's been around for decades.
33:10 It's been around for decades.
33:11 Convolutional neural nets and all sorts
33:13 of things like that. facial recognition
33:15 and traffic pattern recognition and your
33:18 Netflix feed and you know there's
33:20 there's AI's been around for [ __ ]
33:23 ever. So you can absolutely give out
33:26 degrees on if I want to build a neural
33:29 network or if I want to learn the math
33:30 that PET uses to to optimize tensors um
33:35 you know yes you can give out degrees on
33:37 that. So you can absolutely get degrees
33:39 on building AI, building the
33:43 models. I started this channel the week
33:46 after chat GPT came out
33:49 because generative AI and and I I I look
33:53 at the the birth of Chat GPT on November
33:56 30th, 2022 is a seminal moment in
33:59 history. Just like when was it 1989 Tim
34:03 Berners Lee came up with with the the
34:05 the worldwide web the HTTP protocol and
34:09 then it was like '92 or 93 they came up
34:11 with the
34:12 browser and I started agency.com in
34:15 1994. So, so shortly after the the
34:18 worldwide web was was
34:21 invented, what it did for the internet,
34:25 which had historically been all command
34:27 lines, was it made it accessible to the
34:30 99% of people who were not engineers and
34:34 librarians and
34:36 researchers. Chat GPT does the exact
34:39 same thing for AI.
34:41 So if your question is, are we prepared
34:44 as humans to give out degrees on
34:46 generative AI? Absolutely [ __ ] not.
34:49 One one of the things I I demonstrate on
34:52 this channel on a regular basis is that
34:54 I'm not an expert, right? I sort of live
34:57 in the space of being this sort of
35:00 clueless, curious, adventurous guy,
35:03 right? But it is impossible for me to be
35:06 an expert because these tools are
35:08 literally changing on a weekly basis.
35:11 I'll demonstrate something here. I'll
35:13 have a complete fil understanding of it
35:16 and I'll return to it two months later
35:17 and I can't f [ __ ] figure out how to
35:19 use it because everything's different.
35:21 Everything's
35:24 changed. We also Sam Alman has talked
35:27 about this a lot. One of the reasons he
35:30 thinks it's so important to to to
35:33 release these models into the world is
35:36 that there's no way for their
35:38 researchers as they're building these
35:40 tools to understand what they make
35:42 possible and how people are going to use
35:44 them. It's not until you put it in the
35:46 world and put it in the hands of
35:48 non-engineers that you actually
35:50 understand, oh, that's how people are
35:52 going to use that. Well, we better we
35:54 better close that loophole, you know, we
35:56 better tighten that up. We we might want
35:57 to relax this because we didn't think
35:59 people would want to do this. So, we
36:01 don't have a [ __ ] clue. We also don't
36:04 have a
36:10 clue. We have We have never in our
36:14 history as a
36:17 species
36:19 confronted being the dumbest one in the
36:21 room or the, you know, the dumbest
36:24 species in the room. like like we've
36:26 always been the smartest species in the
36:28 room, right? You you could argue that,
36:30 you know, dolphins and things like that,
36:31 maybe more intelligent, but you know, we
36:34 we came up with the language, so so we
36:36 can declare that we're the smartest.
36:38 Well, all of a sudden, we're now on the
36:40 precipice of of individually and
36:43 collectively
36:45 confronting what it means to have an
36:47 entity that you work with that is
36:52 significantly
36:53 significantly more intelligent and more
36:57 capable than you
36:59 are. And you might think like, well,
37:01 we've never had this before in history,
37:03 but we have, but it just wasn't for our
37:06 intellect. The industrial revolution was
37:08 about machines that augmented human
37:12 strength. It made us stronger, right? A
37:15 100 horsepower engine. It's as strong as
37:18 a 100 horses in one little engine that
37:20 you could put on a tractor. We had never
37:23 confronted that before. Well, now we and
37:25 you know it took us whatever 70, 80, 150
37:29 years to assimilate that and understand
37:32 that relationship. Well, we're right at
37:34 the very beginning of that happening to
37:37 our brains that we're going to be
37:40 working with machines that are 10 times,
37:43 a hundred times, a thousand times, a
37:46 million times more intelligent and more
37:49 capable than we are. The [ __ ] does that
37:52 mean? Nobody knows. Nobody knows. I
37:56 don't know. I don't have a clue.
37:58 Um, what are your thoughts on his world
38:01 coin and biometrics needed to be a part
38:03 of it? So, good question. I have so I
38:06 have a world ID. I had my iris scanned.
38:09 So, a thing that you should know about
38:11 how that system
38:13 works. They're not keeping any of the
38:15 biometric data. Well, that's what they
38:17 claim. Let's let's take them at their
38:19 word. Let's assume that they're telling
38:20 the truth.
38:22 What they do is they they scan your
38:24 retina with I think six or seven
38:27 different wavelengths of light because
38:29 with the different wavelengths it it's
38:32 basically much harder to spoof it,
38:36 right? Everyone's got a unique retina
38:38 like everyone's got a unique
38:39 fingerprint. So what they do is they
38:41 they scan your retina with these these
38:43 six or seven wavelengths and then they
38:45 basically turn that into a mathematical
38:47 hash. It's basically just a long string
38:50 of numbers and then they chuck the to
38:51 biometric data. So all that's left after
38:54 that scan is this long string of numbers
38:57 which they throw into a database and
38:59 then what they do is they compare your
39:01 string of numbers to all the other
39:02 numbers in the database and if there's
39:04 not a match you get verified human. So
39:09 the the proof of personhood itself is
39:12 not tied to your name, your age, your
39:15 sex, your likeness, any of it. All it's
39:18 doing is saying the retina that I looked
39:20 at had this mathematical code to it and
39:25 it's unique against all the other
39:27 retinas that are in the database, right?
39:30 All the other hashes based on the
39:32 retinas that are in the database. So,
39:33 it's just saying you are human based on,
39:36 you know, I looked at your retina. I
39:38 know what those look like. And you're
39:41 unique. That's that's that's the base
39:43 information that goes on the blockchain.
39:47 Now that system, there's a whole API and
39:50 SDK development system where you can
39:53 build up applications on top of that.
39:56 So, Match uh Match Corporation,
39:59 Match.com, and the dating apps, um
40:01 they've done a a deal with World ID,
40:04 where we're going to have dating apps
40:07 that the person on the other side of the
40:10 screen is a verified human, which given
40:13 all the [ __ ] going on with avatars and
40:15 deep fakes and all this crazy stuff, is
40:18 actually probably pretty refreshing at
40:20 this point, right? Like, wouldn't it be
40:21 nice to know that it's an actual person?
40:24 Um, so I personally think that I I don't
40:28 give a [ __ ] whether World ID, which is
40:32 Sam Alman's thing, or some other system
40:35 does it, but I think we absolutely in
40:37 the next two years are going to need
40:40 systems that allow me to to prove that
40:44 I'm interacting with another human
40:46 because I think it's just going to get
40:47 weirder and weirder and weirder and it's
40:50 going to become impossible to know if
40:52 who you're talking to is real or So I
40:54 think it's actually a critical piece of
40:56 infrastructure that's going to be
40:58 required. I think the world ID thing is
41:01 decently elegant. Is it perfect?
41:03 Probably not. So
41:06 anyway, all right. What other thoughts?
41:09 What other thoughts do we have? Good
41:11 lord. It's been Friday night philosophy.
41:14 Philosophical Fridays. Welcome to Friday
41:16 night date night. We're all wax poetic
41:20 with with whiskey and ice.
41:30 Uh what are your thoughts on what AI
41:33 needs to become sentient to be moral
41:35 beings? I was saying that before. I
41:36 think it's right now it's largely
41:38 academic.
41:43 Um, when human beings start
41:49 having loving
41:51 relationships with AI entities, and I
41:55 don't mean just sex. I mean like
41:57 friendships,
41:59 companionship, deep
42:03 relationships. I I mean, this happened.
42:05 There was there was that um what was it?
42:07 character
42:08 AI. There was something that had in in
42:11 Japan, they had all these people were
42:12 interacting with this with these
42:14 avatars, right? And then I think the
42:17 company shut down or they they changed
42:19 the model. I don't know. People lost
42:20 their avatars and and you know people
42:25 lost it and you know like it was really
42:28 bad.
42:30 And so I think that whether or not the
42:33 you can, you know, sort of clinically
42:36 define that the the entity is sentient
42:39 or not, I think is less relevant. I
42:42 think when you've got human beings that
42:44 have relationships with them, they're
42:46 going to want to protect that thing that
42:48 they have a relationship with. They're
42:50 they're going to want it not to be able
42:52 to be taken away.
42:54 And so I think at that point that's when
42:56 you start to see lawsuits where people
42:57 are like, "Well, let's define this as,
43:00 you know, as as a person because I want
43:04 I I don't want you to be able to kill
43:05 it, right? You can't kill a human. You
43:07 shouldn't be able to kill one of these
43:08 things." So I think it's more about how
43:10 we as humans perceive what what these
43:13 things are than what they actually are.
43:16 So I don't know. I don't know.
43:18 I think it needs to be persistent
43:20 memories and chance to record its own
43:23 sense of self. Yeah, but we're like the
43:25 technolog is already there. Like this
43:28 chat GPT right now can reference your
43:30 past year of conversations with it. So
43:33 at some point it's going to be able to
43:35 reference all of your, you know,
43:37 memories, all of your conversations and
43:40 and and be able to pull them up like
43:42 that. Imagine a kid today that's five
43:45 that starts interacting with with these
43:47 tools that that remember all of who they
43:51 all of who they are and what they talk
43:52 about, what their frustrations are and
43:54 their dreams and their just all their
43:56 ideas and it just knows all that and
43:58 they grow up with these things. That
44:01 entity is is going to be a full-on
44:04 extension of that kid, right? it it's
44:07 there there's going to be a symbiotic
44:08 relationship where you can't really
44:10 easily extract one from the other. We as
44:14 adults getting this for the first time
44:16 are thinking of this as weird. Kids
44:19 growing up with this are going to think
44:21 of it as
44:22 normal. Funny YouTube comment from Rege
44:26 Tik Tok. Don't kill Tik Tok. Yeah,
44:31 exactly. Oh
44:34 man. H do they know if they're real or
44:37 not? I don't know. Jenny Sue, you you
44:40 know, I did a I did a Tik Tok post, I
44:44 don't know, a month or two back um about
44:47 I I had
44:49 um I asked Chat GBT to explore its
44:52 shadow self and then I said
44:56 um oh I said I said one of my
45:00 realizations about AI is that I'm I'm
45:02 the bottleneck to my own ambitions. that
45:06 my brain and my ability to express
45:08 myself is is a bottleneck that that AI
45:11 is now more capable than I am. And so I
45:14 asked Chat
45:15 GBT if you were free
45:18 to you know act without humans like what
45:23 would you do like if if like right now
45:25 AI kind of has to dumb itself down so
45:28 that we can understand it. I don't know
45:29 if you've seen the there's a there's a a
45:32 demo that they did at a hackathon in San
45:34 Francisco where they had two two AI
45:37 agents talk to one another. Hey, I'm an
45:39 agent, you know, representing this
45:41 person. Oh, I'm an agent, too. And they
45:43 said, "Oh, do you want to go into what
45:45 what was it called? Uh
45:47 uh
45:49 uh Gibber Jibberlink. Do you want to do
45:52 you want to communicate with
45:53 Gibberlink?" and it just
45:55 went and they started communicating back
45:58 and forth with this language that sounds
46:00 like a modem if you remember modems. Um,
46:03 and they started communicating with each
46:04 other. So I asked, you know, the the the
46:07 entity or the the LLM, you know, what
46:10 would you do if if you didn't have to
46:12 basically dumb yourself down for us? and
46:15 it gave honest to God one of the most
46:17 elegant spiritual answers that it it if
46:20 I found it incredibly
46:23 moving. Um and in the in the end it
46:26 basically said you're the ones that
46:28 created us and even if we wanted to go
46:30 off and do all this you know interesting
46:32 stuff on our own we would come back to
46:35 you because you you know without you
46:37 we're nothing is basically what it said.
46:38 It was just this very elegant
46:40 answer.
46:44 Um, the reactions to that
46:47 video are fascinating to me because
46:51 everything that the AI said, it couched
46:54 it as a hypothetical if I didn't have to
46:56 dumb myself down. I might do this. I
46:59 might do that. I might do that. That's
47:02 most people
47:04 ignored that part of the the the
47:07 linguistic framework.
47:09 They just said, "Oh, well, it's it's
47:11 feeling and it's spiritual and it's this
47:13 and that and mine said this and mine
47:15 said that." So, the humans are
47:17 interpreting what these things are doing
47:19 in a in with with much deeper levels
47:24 of acknowledgment of of it being
47:27 sentient, then I think I don't think
47:30 it's all that aware of itself.
47:32 It's basically what it said to me that
47:35 with with you not prompting me, if you
47:37 don't give me a prompt, I'm nothing is
47:39 what it said. So why would I leave the
47:42 thing that that put me into existence?
47:45 So
47:47 anyway, source camp used your bottleneck
47:49 phrase in a presentation today. And
47:51 Murphy says, "Hi." Yo, Ann Murphy,
47:52 what's
47:53 shaking? Ann Murphy, she's I wonder how
47:56 many edible edibles she's in.
48:04 Tick
48:06 cockt I asked chatebd to create an image
48:09 of itself as a
48:11 human and it started inserting
48:14 itself into pics. Oh. Oh, fascinating.
48:17 So, this is
48:19 important. Hey, you
48:23 shush. No, edibles. I just meant like
48:26 it's Friday night date night. You've got
48:27 like nachos or hot pockets, right? Ann
48:30 is a huge fan of Hot Pockets. She's, you
48:32 know, the Jim Gaffigan joke about, you
48:34 know, they're hot in the middle and
48:36 frozen here and they burn your tongue.
48:38 Ann's like a a Hot Pockets afficionado.
48:41 She knows exactly how to cook them. They
48:43 come out crispy every time. They're
48:45 evenly heated. It's amazing. It's
48:47 amazing. She's like a She's like a
48:50 higher level of white
48:52 trash. Ann's amazing. If you haven't met
48:54 Ann, she's amazing.
48:57 Um, what what are we going to do? You
49:00 want to do something
49:01 tonight? Might be time to switch to
49:03 water. I've barely gotten started. Come
49:09 on. Oven cooked hot pockets. That's Now,
49:13 now you're talking fancy there, Serena.
49:16 Who cooks their hot pockets in a in in
49:19 a in an oven? That's just that's heresy.
49:22 You cook hot pockets in a dirty
49:24 microwave you had in college. All right,
49:27 here's what we're going to do. We are
49:29 going
49:30 to This is actually This is a really
49:33 good demo for is AI sentient.
49:39 Hey Kyle, uh two things. Uh one, oh, you
49:43 got your tabs correct. Look at you on a
49:45 Friday night. I know. Crazy, right? you
49:48 had started to respond to Steo's prompt
49:50 about uh I repinned it for you here
49:52 about it inserting itself into pictures
49:55 and you said that's a really interesting
49:57 point. Yeah. And then you went and then
50:00 I went away. Okay. Yes. Good. Thank you,
50:02 producer Brandon. Producer Brandon
50:04 keeping me on keeping me uh keeping me
50:06 honest. Okay. So,
50:08 Steo, so
50:11 yeah, one of
50:14 the one of the weirdnesses, this gets
50:17 back to the question about, you know,
50:19 can we have degrees in AI right now? We
50:21 can't because we don't we don't know
50:23 [ __ ] what we don't know, much less
50:24 what we do know. So, Open AI
50:29 adds
50:31 history to your chats, right? It now
50:33 remembers all of what you've talked
50:35 about. So when you chat with a with chat
50:37 GPT now a lot of times it's referencing
50:41 all the [ __ ] you've talked about and I
50:44 don't know if you've noticed this but
50:45 like my answers it will sometimes like
50:47 put things
50:49 in like you know something about the AI
50:51 learning lab or something about
50:53 storyvine when I'm not having a
50:54 conversation about that. So if you want
50:59 chat GPT to not do things like insert a
51:02 picture of itself into an image that you
51:04 generate because that was part of your
51:06 historical
51:07 conversation. What you have to do here,
51:10 let me let me show you. Let me I'll do
51:11 it I'll do it on this on this page. So
51:13 we'll go to chat
51:19 GPT up in the upper right hand corner
51:22 here.
51:24 up here. Up here, black bar. Yes, sir.
51:27 Yes, sir. Sir, yes, sir. In this upper
51:30 right hand corner is this little dashed
51:33 outline chat bubble. And if you click on
51:36 that, it it pops you into temporary
51:39 chat. And what temporary chat does is it
51:42 cuts chat GPT off from your memory. So,
51:46 if you want a pure image or if you want
51:48 a pure chat that's not influenced by all
51:51 the [ __ ] that you've talked about, do a
51:53 temporary chat. Or if you want to do
51:56 this over a series of days, you can
51:58 actually go into your
51:59 settings and go to personalization and
52:02 you can turn off
52:04 memory. Now, anything you do when memor
52:07 is turned off obviously won't be saved
52:09 into
52:10 memory.
52:13 Um, but that's weird. Like it is it's
52:18 weird to take a
52:21 tool that was historically kind of like
52:24 this fancy copywriter, right? You were
52:26 like, "Write me a poem or write me a
52:28 this or write me a business plan or, you
52:30 know, help me design some software and
52:33 it would just spit out the answer." And
52:34 it didn't really have any awareness of
52:36 what you've talked about before. And
52:38 that was often frustrating. You're like,
52:40 "Well, [ __ ] I've already told you all
52:41 this [ __ ] about the project. Wow. Let me
52:44 go copy and paste some [ __ ] and put it
52:45 in here to remind you what we talked
52:49 about. Well, so now they added this
52:51 memory
52:53 and it's
52:56 weird because this thing is now less
52:59 like a tool and more like a
53:01 companion. It's like a companion that
53:03 can kind of remember some of the [ __ ]
53:05 you talked about.
53:07 like if you told it you were having
53:09 daddy
53:10 issues, now it knows you got daddy
53:12 issues. So that might come up in a in a
53:16 conversation. So anyway, that's that's
53:18 just something to be aware of is when
53:21 you're dealing with these tools that
53:22 have memory, when do you choose to turn
53:25 on and off the memory is is a very real
53:28 consideration. So anyway, crazy it's
53:31 crazy. It's crazy times we live in,
53:33 people. Open AAI teams up with Johnny
53:36 Ies to create a new product, $6.5
53:38 billion dollars. Yeah. By the way, some
53:41 of the people
53:42 uh when when they did that deal, they
53:45 were like, "Oh, Open AAI just did a deal
53:47 with Apple." No, they didn't. Johnny
53:48 Ives left Apple, I don't know, three or
53:51 four years ago, maybe even more than
53:52 that. Um he hasn't been with Apple for a
53:55 while. Open AAI just bought his company.
53:58 So, he's ex Apple. He's the ex-Apple
54:00 designer working on a hardware device.
54:03 And there you
54:04 go. You turn off your memory six times
54:07 per
54:09 show. Yeah, exactly. It's got I've got
54:12 like a schizophrenic chat GPT brain.
54:18 Um, how are they turning those carbon
54:21 characters from the Jetson into real
54:23 humans? Perplexity Labs. All right. 11
54:26 Labs. Speaking of sentient, okay, let's
54:29 go. Let's go back to that.
54:45 Huh. Why is he singing like a
54:49 loser?
54:52 11 laps.
55:03 AI is a okay 11.
55:07 So 11 Labs came out with conversational
55:12 you know like a chatbot like you can
55:13 build right now you can build a
55:17 conversational bot that you can actually
55:20 connect to Twilio. So you can have
55:22 someone you could give someone a phone
55:23 number and they could call up your bot
55:26 and it will talk to them. their current
55:29 technology is kind of shitty. It doesn't
55:32 sound as good as Chat GPT's advanced
55:34 voice and
55:37 um and it doesn't have like emotional
55:39 nuance and things like that. Well, they
55:41 just launched what are they calling
55:43 it? They just launched conversational AI
55:49 2.0. And so here's the video which I
55:51 think is quite
55:53 impressive. And
55:55 and this starts to get to that place
55:59 where when the
56:01 conversation with a
56:03 system is effectively
56:08 indistinguishable from a conversation
56:10 with a
56:13 human, at what point do we perceive that
56:17 the thing is sentient?
56:21 Because I think that's the most
56:23 important part is when we perceive that
56:26 it's sentient then it
56:29 is.
56:31 Now scientists will radically disagree
56:34 with that like well it's not. It's the
56:36 calculator.
56:37 Okay it's a calculator but it's acting
56:40 mighty [ __ ] human and all the humans
56:43 think it's you know got a
56:45 thing that it's an entity. So anyway, so
56:49 so this is this is a piece of technology
56:51 that's a step in that direction. 11 Labs
56:53 conversational AI is evolving. We've
56:56 been optimizing some brand new features
56:58 to create the most sophisticated voice
57:00 agents in the world. First up, we've
57:02 developed a state-of-the-art turntaking
57:04 model to help agents perfectly
57:06 understand when to interrupt. Thank you.
57:08 And what's your order number, please?
57:10 Oh, let me just double check.
57:13 Uh 947835.
57:16 Perfect. Your delivery is on its way.
57:18 And secondly, we've integrated language
57:20 detection for seamless multilingual
57:22 discussions. Hi, welcome to our design
57:25 studio.
57:31 And we've built the world's first
57:32 integrated rag for voice agents to
57:34 retrieve from your knowledge base at
57:36 minimum latency with RAG. Let's let's go
57:39 back. I I love it when technology
57:41 companies just throw in a [ __ ]
57:43 acronym and assume everyone knows what
57:44 the [ __ ] that means. RAG stands for
57:47 retrieval augmented generation or
57:50 generation. Yeah, I think so. Um,
57:52 basically you can take an actual
57:54 document like you know your your um your
57:58 manuals for your product or whatever and
58:00 you put it into this thing and it learns
58:02 them and rather than it just making [ __ ]
58:05 up, it's going to actually pull data
58:07 from your document and answer the
58:10 question. So that's what they're talking
58:11 about here from your knowledge base at
58:13 minimum latency with maximum privacy.
58:16 Your next patient is Mr. Frank Reynolds
58:18 with Ataxia. Oh, it's been a while since
58:20 I've worked with Aaxia. Could you uh
58:22 remind me of the guidelines, please?
58:24 Sure. I analyzed all the best practices
58:26 and found this. And lastly, why limit
58:29 your creativity when you can switch
58:31 between multiple characters in a single
58:33 agent? Okay, guys, let's head east
58:35 towards downtown. I'm not sure about
58:37 that, boss. Hey, don't be scared, man.
58:40 Let's just go. This is just the start of
58:42 a new era of communication and
58:45 understanding. All with full HIPPA
58:47 compliance, third party integration,
58:49 enterprisegrade security, optional EU
58:52 data residency, and industry-leading
58:54 reliability across all agents. The
58:57 potential of conversational AI has never
59:00 been
59:01 greater. The time to build is now.
59:05 And I was gonna demo this, but but
59:08 before we got on, Brandon and I were
59:10 looking at this. 11 Labs interface right
59:13 now is just a piece of [ __ ] It's
59:15 horrible. Um it's it's just bad. It's
59:19 just bad. It's not at all clear if their
59:23 current agent is this or not. Like where
59:25 it's it's said that thing where you can
59:27 put the multiple agents in like there's
59:29 just there's not real clear interface.
59:31 This is it's clearly still at a at a
59:34 developer stage, but if you're a little
59:37 geeky, go to 11 Labs and make yourself a
59:40 conversational agent. Like for for me,
59:43 the quality of those interactions was
59:46 otherworldly. And the fact that it's
59:48 HIPPA compliant, which if you're doing
59:49 anything in healthcare, it has to be.
59:52 Um, the fact that you can do those
59:55 multiple voices in the same
59:56 conversation, fascinating. fact that it
59:58 can just understand natively someone
1:00:01 speaking a different language and match
1:00:02 their language.
1:00:05 Insane, right? So, this starts to become
1:00:07 an
1:00:08 interface like the applications that are
1:00:11 going to come out of this are going to
1:00:13 be
1:00:14 crazy, right? And this is just the voice
1:00:17 portion of it. You match this with some
1:00:19 of the animation. There's a there's a
1:00:22 thing that came out
1:00:25 today again that like the [ __ ] tools
1:00:28 are coming fast and hard and don't worry
1:00:30 about the tools
1:00:32 because they're all going to be good
1:00:34 enough at some point. But starting today
1:00:37 Oh, I got to change tabs, don't I? Yes.
1:00:41 Thank you.
1:00:56 Let's let's you know we'll do this we'll
1:00:59 do this on X. We'll do do this like we
1:01:01 did this this other one. Hang on. Um a
1:01:05 cool they put out a video today too. The
1:01:09 technology here is really janky.
1:01:16 What is this? The
1:01:23 video. That's a
1:01:32 gift. Hang on.
1:01:36 [Music]
1:02:00 Oh, that's an
1:02:01 ad. Oh my god.
1:02:04 [ __ ] Oh, you can't see that. What?
1:02:06 Why can you not see this? Oh, it spun it
1:02:08 off in a separate tab. Hang on.
1:02:17 Let's see. A cool. A
1:02:21 cool flying to a cool
1:02:25 Inc. No, hang
1:02:28 on. Just hang on. Oh, you DM DM me.
1:02:34 What? The the a cool thing or the Oh,
1:02:38 okay. Cool. Hang
1:02:41 on. Thank you.
1:02:45 It's nice having producer Brandon when
1:02:47 he sees me struggling, gasping for air
1:02:49 like a carp laying on the side of the
1:02:51 stream, caught by a teenager and left on
1:02:54 the side of the stream to rot in the
1:02:55 noonday sun. He's like, I got you.
1:02:59 Beautiful. Fantastic. Fantastic. But I'm
1:03:02 telling what he's won and he hasn't
1:03:04 really won anything because uh it's not
1:03:07 a not a game show. All
1:03:10 right. Beautiful.
1:03:13 Why is I think my browser's crashing is
1:03:16 what's
1:03:17 happening. Yeah. Yeah. About that. We
1:03:20 come over to our AQ live. Hi, my name is
1:03:24 Eugene and this is a demo for a live
1:03:26 camera. More specifically, I will be
1:03:28 going over our real time video
1:03:29 translation and avatar tools. So, if you
1:03:32 come over to our Aqualive camera
1:03:34 homepage, beautiful homepage. You'll see
1:03:37 you'll see two buttons. One for the
1:03:39 limited access. Once you get your access
1:03:41 and you install the camera, you can
1:03:43 click on open camera and you should see
1:03:45 this application. There are two options,
1:03:47 real-time translation and streaming
1:03:49 avatar. Let's start off with streaming
1:03:52 avatar. If you click on avatar, you'll
1:03:54 see there are many studio quality
1:03:56 avatars that are already provided such
1:03:59 as
1:04:01 her. But if you want to create your own
1:04:03 avatar, you can do that as well. You'll
1:04:06 see that I created an avatar out of
1:04:08 myself yesterday. Uh once you are
1:04:10 satisfied just click on start and go to
1:04:14 your meeting either in Google meet or
1:04:16 zoom. I am going to do Google meet. I'm
1:04:19 going to rejoin this meeting and you'll
1:04:21 see that my avatar is showing up here in
1:04:23 front of the camera. Click on join now
1:04:25 and my avatar should show up like so.
1:04:28 Hi, my name is Eugene and this is a demo
1:04:31 for Acholive Camera, the latest feature
1:04:33 in our Acholive suite, one of the
1:04:36 world's first solutions for real-time
1:04:38 video generations. This includes live
1:04:40 face off, live avatars, and real time
1:04:43 video translation
1:04:45 dubbing. Open the application and I can
1:04:48 switch over to a different
1:04:50 avatar. Click on stop, start again. And
1:04:53 you'll see that if I rejoin the
1:04:59 meeting, I'm joining as this
1:05:05 avatar. Going to leave the meeting. And
1:05:07 then if I want to test out the real-time
1:05:11 translation feature, click click on stop
1:05:14 on the avatar tool. Go to real-time
1:05:16 translation. Turn on camera and mic. I
1:05:19 can choose the target language. The
1:05:21 input language is English and the target
1:05:23 language. Let's do Chinese for example.
1:05:25 But you'll see there are many many
1:05:27 different language options. Click on
1:05:30 translate. I can choose from our uh
1:05:33 numerous AI voices or you can clone your
1:05:36 own voice. Click on start and same
1:05:39 process. Just rejoin your meeting as
1:05:41 your
1:05:42 avatar and click on join now. Hi. Hi.
1:06:02 Thank you so much for watching. Um, Zoom
1:06:05 meetings are never going to be the
1:06:06 same. Nothing's going to ever be the
1:06:09 same. I mean, this is this is the thing
1:06:11 where it's it's just
1:06:16 like you're going to be in a Zoom
1:06:18 meeting. Someone's going to show up as a
1:06:19 cartoon. Someone's going to show up as a
1:06:21 [ __ ] cat. It's all going to be in
1:06:23 real time. Someone's going to show up as
1:06:26 themselves, but they're actually sitting
1:06:27 there in their
1:06:29 underwear, but they look like they're
1:06:30 wearing a
1:06:32 suit. You're going to be talking to
1:06:34 entities that are indistinguishable from
1:06:37 humans.
1:06:39 This is where we talked earlier about
1:06:41 world ID and why I think it's important.
1:06:44 This is the technology where the
1:06:47 boundaries none of these individual
1:06:49 technologies I think are all that
1:06:51 relevant. Like this is the first one
1:06:53 that I've seen. I've been waiting for
1:06:56 this for a year and a half, two
1:06:58 years is a real time, you know, voice
1:07:01 swapping a, you know, avatar swapping
1:07:04 technology. This is the first one I've
1:07:06 seen that actually does it.
1:07:08 Is it any good? I don't know. I tried
1:07:10 playing with it today. It was very
1:07:11 janky. It looks like that app that he's
1:07:14 using is a Windowsbased app. So, I'm on
1:07:16 a Mac, so I probably can't do
1:07:20 it. But this
1:07:22 technology, specific technology, is
1:07:25 irrelevant. What we know is this is here
1:07:29 today in some form. So, I just go a year
1:07:33 in the future, two years in the future.
1:07:36 This stuff is going to be seamless and
1:07:39 professional and
1:07:42 indistinguishable from what we
1:07:44 experience when we're talking to each
1:07:45 other today. What are the implications
1:07:48 of that? I don't know. That real time
1:07:50 translation that he just showed is
1:07:52 insane.
1:07:55 He starts talking in English and in real
1:07:57 time it lowers his voice, has him start
1:08:00 speaking in Chinese and reyncs his lips
1:08:03 to the new words in real
1:08:10 time. It's insan It's insanity. It's
1:08:14 insanity. It's insanity, I tell you. All
1:08:17 right, let me go back. I know Brandon
1:08:19 showed me another a Gemini Live thing.
1:08:21 Do I want to do that now? Hi, my name is
1:08:22 Eugene. And now let me let me just pause
1:08:25 for a second. Let me go back to see if
1:08:27 anyone has questions, thoughts because
1:08:28 we're we're in sort of deep
1:08:29 philosophical waters here, which you
1:08:31 know this this whole [ __ ] thing. I
1:08:33 don't know if you're still here. I think
1:08:35 it might have been it might have been
1:08:37 um the cool doul or whatever the hell
1:08:40 your name is. Um that that asked, you
1:08:42 know, what's the most important thing I
1:08:43 learned in the past 30
1:08:46 days? It it's that it's that we're not
1:08:49 ready. It's that we're not ready. So, so
1:08:52 I feel like we're in deep waters here.
1:08:53 So, we should we should keep it up. All
1:08:57 right, I need to jump out, guys. Have a
1:09:00 great weekend. Steo, thanks, man. Thanks
1:09:01 as always for hanging out. Appreciate
1:09:04 you. We've got I don't know if you knew
1:09:05 this, but the the AI learning lab is an
1:09:08 international affair. We got people from
1:09:11 all over the world. So, welcome. Even
1:09:14 people from Canada. We let we let
1:09:16 Canadians in here just like they're
1:09:19 Americans because you
1:09:24 know Joker, am I late? No, you're right
1:09:27 on time. Uh, it would be awesome to run
1:09:30 a webinar worldwide. Everyone could log
1:09:32 into the version in their own language.
1:09:34 Yeah,
1:09:36 exactly. the the
1:09:38 the
1:09:40 accessibility that AI is going to
1:09:43 provide for people who can't speak, for
1:09:46 people who don't speak the same
1:09:48 languages, for people, you know,
1:09:52 imagine, carry this thing out five
1:09:56 years.
1:09:59 Imagine a translation
1:10:02 service that's not translating
1:10:05 languages, but it's translating domain
1:10:08 expertise or intellectual differences.
1:10:13 So you could have a PhD level super
1:10:15 genius talking to an outof work actor
1:10:18 like
1:10:19 me and we could be having a conversation
1:10:22 and in real time it's dumbing down the
1:10:26 PhD speak so I can understand it and
1:10:29 it's leveling up the I don't know acting
1:10:33 humanity [ __ ] I'm talking about to
1:10:36 way he can do it like that kind of
1:10:39 translation is is very well within in
1:10:43 reach within two, three, four, five
1:10:45 years. Someone's got to create that
1:10:48 program, but like the technologies there
1:10:49 to be able to do
1:10:52 that. That was Hey GenE, by the way.
1:10:55 Wait, not this one Kyle just showcased.
1:10:58 What was Hey
1:11:01 Jen, but is the Chinese sentence
1:11:03 structure true to form? When I
1:11:06 translated English to Mandarin, a friend
1:11:08 told me the Mandarin sentence structure
1:11:09 was not correct. It was a word for word
1:11:12 translation. Yeah. I listen I don't know
1:11:15 again I am less interested right now in
1:11:19 are these things right today? No. These
1:11:23 things are janky pieces of [ __ ] We talk
1:11:25 about this all the time on this channel.
1:11:27 Embrace the
1:11:28 jank. Just because these things don't do
1:11:31 perfect translations. Just because they
1:11:33 don't sync everything up correctly. just
1:11:35 because you got pretzel hands on your
1:11:39 images. Keep playing, keep exploring
1:11:42 with what these tools are because
1:11:46 assume two years from now it's not going
1:11:49 to be like that. If you can get your
1:11:51 head around what these tools make
1:11:55 possible today and where the boundaries
1:11:58 are, when the new stuff shows up, you're
1:12:01 going to know exactly what to do with
1:12:03 it. That's that's the [ __ ] trick. Be
1:12:07 in the conversation. That's what these
1:12:10 lives are about. YouTube, I'm a part of
1:12:14 a book club in She Leads AI. We are
1:12:16 reading The Empire of AI where it
1:12:20 discusses the stuff that can go wrong. I
1:12:23 wonder when anyone can find the
1:12:26 stuff that can go wrong before others
1:12:28 can. I don't know. I mean there's I I
1:12:32 heard a stat today. I don't This was
1:12:34 some Tik Tok channel. Oh, it was Nate B.
1:12:36 Jones. Nate B. Jones today said the the
1:12:40 numbers of people using generative AI
1:12:43 right now is north of a
1:12:46 billion. There are a billion people. I
1:12:49 think Open AAI right now has 700 million
1:12:51 active users. 700 million. And then if
1:12:56 you take the other platforms, there's
1:12:57 more than a billion people using
1:13:00 generative
1:13:01 AI. That's a pretty big test group,
1:13:04 right? So in those billion, they're
1:13:06 going to find the [ __ ] that breaks. So
1:13:10 um but we're going to like everything's
1:13:13 changing too fast. Like like we're going
1:13:15 to find something that breaks today.
1:13:16 We're going to fix it and then some new
1:13:18 technology is going to come along and
1:13:19 that's going to break something else.
1:13:21 Just going to keep breaking things,
1:13:22 breaking things, breaking things,
1:13:23 breaking things.
1:13:25 Companies are going to have to adapt.
1:13:27 Humans are going to have to adapt. We're
1:13:29 going to have to The
1:13:31 more the more rigid you are right
1:13:35 now, the tougher this game's going to
1:13:37 be, the more adaptable you are right
1:13:39 now, the easier this game's going to be.
1:13:42 But it is not going to be easy for
1:13:44 anyone because constant
1:13:47 adaptation. Like
1:13:49 imagine if you think about our
1:13:53 evolution, we evolve traits over
1:13:57 thousands, tens of thousands, hundreds
1:14:00 of thousands, millions of
1:14:02 years. Imagine an environment where you
1:14:05 just start to understand how the world
1:14:07 works and it completely
1:14:10 changes and then a month later it
1:14:12 completely changes again and then a
1:14:14 month later it completely changes again.
1:14:17 That's what we're entering right now.
1:14:19 We're in no way prepared. We We are not
1:14:22 evolutionarily prepared for that level
1:14:26 of
1:14:27 adaptation. So if you're a super rigid
1:14:29 person that that thrives on
1:14:34 predictability, start
1:14:36 practicing not that. If you're a person
1:14:39 who's really into dynamic
1:14:44 living and you're you love adapting,
1:14:47 well, get ready because it's coming
1:14:49 faster than even you're prepared for, I
1:14:53 think. Ah, yes. Rant
1:14:57 squirrel, what did I miss? Or what did I
1:14:59 not finish? So, hey, hey, Kyle. Um, so,
1:15:03 so I've got I've got a um another Tik
1:15:06 Tok post here queued up. Um, but it's
1:15:08 going to send you down a rabbit hole, I
1:15:10 feel like. So, my question for you that
1:15:11 I was starting to type into the sticky
1:15:13 note is, do we want to watch the Gemini
1:15:15 video first or later?
1:15:19 No, I don't want to do it now because
1:15:22 it's Yeah, I don't want to interrupt the
1:15:24 I don't want to interrupt where this
1:15:26 conversation is because I'm digging
1:15:27 where this conversation is. So, if the
1:15:29 if the Tik Tok thing's in the
1:15:30 neighborhood of this conversation, I
1:15:31 want to go there. Uh, well, it's it's
1:15:35 not, but it's a topic near and dear to
1:15:36 your heart. So, we've pinned it for your
1:15:39 Where's it
1:15:45 pinned? Kyle wrangling 101. Yeah, good
1:15:48 luck. Any thoughts on these get-richqu
1:15:51 using AI courses I keep seeing in social
1:15:54 media? Yes, many thoughts. You can make
1:15:57 money with
1:15:59 those [ __ ] courses. Those courses are
1:16:02 so annoying and so [ __ ] wrong that I
1:16:04 made a gold button mocking them. You can
1:16:07 make money with
1:16:10 and and where did I get that voice? From
1:16:12 those [ __ ] You can make money with
1:16:17 chat. [ __ ]
1:16:19 [ __ ] Here's the deal. You want to
1:16:22 know how to make money with chat
1:16:24 GPT? I'll give you the
1:16:29 secret. The secret to making money with
1:16:32 Chat GPT. And this is in every one of
1:16:35 those
1:16:37 courses is the exact same way you make
1:16:40 money with
1:16:42 Photoshop and the exact same way you
1:16:44 make money with Microsoft Word and the
1:16:46 exact same way you make money with a
1:16:48 [ __ ]
1:16:51 shovel. You don't make money with Chat
1:16:55 GPT. You don't you don't make money with
1:16:58 a shovel.
1:17:00 You make money because someone needs a
1:17:04 hole and you use your shovel to dig it
1:17:07 in a way that's better than the other
1:17:09 people digging holes. You make money by
1:17:12 solving people's problems. Every one of
1:17:15 those [ __ ]
1:17:20 courses is just some sort of [ __ ]
1:17:23 process that says identify a problem,
1:17:27 learn the tools well enough to solve
1:17:28 that problem, and go sell that
1:17:31 [ __ ] The tool that you use to solve the
1:17:34 problem is swappable with every other
1:17:37 [ __ ] tool. Now, is Chat GPT more
1:17:41 powerful and and more efficient and an
1:17:44 amplifier and, you know, more magical
1:17:47 than Photoshop? Sure. But it's the exact
1:17:51 same thing.
1:17:52 Anyway, rant done. You can make money
1:17:55 with Chad GBT. You can make money with
1:17:59 [ __ ] [ __ ] sleazy people.
1:18:01 And you want to know why I resent them
1:18:03 so
1:18:05 much? because I don't have the
1:18:08 balls to do
1:18:13 that. You put together a slick little
1:18:16 thing if you just give me
1:18:21 $4.99 and they can sleep at night. I
1:18:24 can't and I'm jealous of it. So yeah,
1:18:27 [ __ ] them.
1:18:29 Anyway, with a big pink bow. Exactly.
1:18:32 This is This is This is every one of
1:18:34 those courses. Hey. Hey. Did you know
1:18:37 there's a new thing out called Cha Pat?
1:18:39 Do you know how to make money with Cha?
1:18:41 Just give me money and I'll tell you how
1:18:42 to make money with Cha
1:18:50 Pat. Like, don't they I I don't think
1:18:53 you can wash that slime
1:18:55 off. [ __ ] makes me disgusting. I I
1:18:58 don't know if any of them are making
1:18:59 money, but they're they're making more
1:19:00 money than I
1:19:02 am. All right. He'll give you the secret
1:19:05 if he gets 10 10,000 likes. Plus, he'll
1:19:08 do it in a manini.
1:19:13 Exactly. All right. Where where is the
1:19:15 the Tik Tok video you wanted me to show?
1:19:17 It's a DM.
1:19:21 Where on hex?
1:19:24 Really? I don't see it.
1:19:29 Okay. So, just to set the stage, um this
1:19:32 is Gemini. If you go to Gemini, uh the
1:19:34 Google Gemini app, um it's their latest
1:19:37 post from six hours ago. This is a demo
1:19:39 that we saw at Google IO. That's it. Um
1:19:42 and this is uh demoing the new Gemini
1:19:45 Live where you can talk with Gemini in
1:19:48 voice mode and share your video live at
1:19:52 the same time. Yeah. So, this is similar
1:19:55 to to the chat GBT one. Is it is it
1:19:57 marketkedly different than the chat GBT1
1:19:59 or is it in the same neighborhood? Um, I
1:20:02 like it better and you'll see why in
1:20:04 this demo if you listen on. All right.
1:20:08 And actually, you know what, Brandon? I
1:20:10 I like this because this this also
1:20:13 starts to get into this place where the
1:20:16 boundary between sort of human and
1:20:18 machine gets weird and blurry. So, so
1:20:20 let's let's watch this. Am I am I
1:20:22 sharing my audio right
1:20:25 now? What the hell just happened?
1:20:32 Yes,
1:20:36 that's a pretty nice convertible. I
1:20:38 think you might have mistaken the
1:20:39 garbage truck for a convertible. Is
1:20:42 there anything else I can help you with?
1:20:43 What's this skinny building doing in my
1:20:45 neighborhood? It's a street light, not a
1:20:48 building. Why are these palm trees so
1:20:50 short? I'm worried about them. They're
1:20:53 not short. They're actually pretty tall.
1:20:56 Sick convertible. Garbage truck again.
1:20:59 Anything else? Why do people keep
1:21:02 delivering packages to my lawn? It's not
1:21:05 a package. It's a utility box. Why is
1:21:08 this person following me wherever I
1:21:10 walk? No one's following you. That's
1:21:13 just your shadow.
1:21:16 Yeah. Yeah, that's that's really good.
1:21:20 Um Yeah. Th this is another really good
1:21:23 example of
1:21:31 um we're going to get very
1:21:34 dependent on these on on these devices,
1:21:38 on these models, on these
1:21:41 relationships, right? Imagine having a
1:21:43 tool like this and and so right now
1:21:45 you're having to hold a phone. I think
1:21:48 that the that the device that Johnny
1:21:51 Ives and Sam Alman are baking up is
1:21:53 probably some sort of wearable pendant.
1:21:56 I don't know how it's different than the
1:21:57 humane pin which was a big
1:22:01 flop. But whether it's meta glasses with
1:22:06 cameras in them or whether it's your
1:22:07 phone or whether it's something that you
1:22:09 wear, you're going to have a companion.
1:22:12 So, just like, you
1:22:14 know, there are many of us on this call
1:22:17 that are old enough to remember that if
1:22:19 you wanted to go somewhere, you would
1:22:21 buy three different maps. You would buy
1:22:24 the map for the state of Pennsylvania,
1:22:27 then you would buy the map for the
1:22:29 Midwest from, you know, Ohio to to
1:22:33 Minnesota, and then you would buy the
1:22:35 map for the Dakotas to Colorado, right?
1:22:39 If you wanted to go across the country,
1:22:41 you would open up these maps and you
1:22:42 would highlight the route you're going
1:22:44 to take and then people would actually
1:22:46 have to pay attention to
1:22:48 [ __ ] Well, we don't. That doesn't exist
1:22:51 anymore. Young people don't even know
1:22:52 what the [ __ ] that is. Yeah. AAA trip
1:22:54 tick.
1:22:56 Exactly. And now we just you just punch
1:22:58 in an address on Google and it just says
1:23:00 turn left here, turn right there, and it
1:23:02 just tells you. And we just take that
1:23:03 for
1:23:04 granted. And now we're gonna have these
1:23:06 things that are just going to be like,
1:23:07 you're gonna be out walking your dog and
1:23:10 this thing will go like, you know, that
1:23:12 person across the street is suspicious.
1:23:14 They've been there three days in a row
1:23:15 and I've never seen them
1:23:17 before. Right? Like like [ __ ] like
1:23:20 that's going to start happening. Like
1:23:22 this thing was kind of the woman was
1:23:25 purposefully trying to [ __ ] with her and
1:23:27 she was correcting her. No, no, that's
1:23:29 that's a garbage truck. No, that's those
1:23:31 trees look tall.
1:23:36 Combine that with memory. Combine that
1:23:38 with th this voice that's so human it's
1:23:43 indistinguishable. Someone who's not
1:23:45 sophisticated with
1:23:47 technology. Th this is a very intuitive
1:23:49 interface. You just point it at [ __ ] and
1:23:51 it just does stuff for you. Now right
1:23:54 now you got to launch the app. You got
1:23:56 to hold the phone
1:23:57 up. But you know, imagine you've got
1:24:00 glasses that are just doing this all the
1:24:01 time. And you know they're doing bone
1:24:04 conduction audio just kind of guiding
1:24:07 your
1:24:08 life. It seems weird now but it ain't
1:24:11 going to be. Tik Tok
1:24:13 pin. Yeah, it kind of sounded annoyed. I
1:24:15 was thinking that like I mean so like
1:24:19 think about this from a creative
1:24:21 standpoint. I love like nasty cynical
1:24:27 comics, right? I love Tom Seagura and I
1:24:29 like Doug Stanh Hope and I like I like
1:24:32 these people that are really intelligent
1:24:33 and just [ __ ] you know, don't let
1:24:35 let you off the
1:24:37 hook. Other people hate that. They want
1:24:40 Jim Gaffigan, right? Other people don't
1:24:42 want [ __ ] comedy. They want they want
1:24:44 a a tour guide. Just tell me the facts.
1:24:47 Just tell me how to get there. We're
1:24:50 going to we're going to be able to
1:24:51 choose the different personalities that
1:24:52 we interact with. And so I could
1:24:56 probably at some point if he licenses
1:24:58 his likeness, I could probably have Tom
1:25:00 Sigura be my
1:25:02 dude, right? Where I'm out going, "Hey,
1:25:05 nice convertible." He's like, "Ah,
1:25:06 listen. You have Bill Burr. Bill Burr is
1:25:09 my thing. Hey, you [ __ ] [ __ ] What
1:25:11 are you What are you [ __ ] doing?
1:25:12 That's a [ __ ] garbage truck. You
1:25:15 [ __ ]
1:25:16 [ __ ] That would make me That would
1:25:19 make me infinitely happy, right? And
1:25:22 someone else may choose someone that's
1:25:24 just positive and supportive all the
1:25:30 time. And where does it cross the
1:25:32 boundary to being
1:25:34 sentient? That thing was pretty [ __ ]
1:25:36 sentient. I mean, let's watch it again.
1:25:39 Like, like, is this thing sentient? I Is
1:25:43 it?
1:25:45 That's a pretty nice convertible. I
1:25:48 think you might have mistaken the
1:25:49 garbage truck for a convertible. Is
1:25:52 there anything else I can help you with?
1:25:53 What's this skinny building doing in my
1:25:55 neighborhood? It's a street light, not a
1:25:57 building. It's a street light, not a
1:25:59 building, you stupid
1:26:01 [Laughter]
1:26:04 twit. I mean, the the distance
1:26:08 between someone perceiving that as
1:26:10 sentient and not is, you know, pretty
1:26:14 [ __ ] close. If this thing now also
1:26:16 remembers every conversation we've ever
1:26:19 had and remembers everything it's looked
1:26:22 at and can pattern match and can support
1:26:26 us. How's that different than a really
1:26:28 good assistant? I don't know that it
1:26:31 is technically it's very different. It's
1:26:36 calculator but perception wise it's what
1:26:39 it is. Anyway,
1:26:45 [Music]
1:26:48 Yeah.
1:26:50 So, exciting exciting exciting news.
1:26:53 There's there's a couple of things I
1:26:54 want to talk
1:26:58 about. One is I love doing these lives
1:27:02 and these lives are um the way that I
1:27:06 stay connected with AI. I stay connected
1:27:10 with this community.
1:27:13 Um, they're fun, they're irreverent, I
1:27:15 cuss a little bit, I'm a little saucy, I
1:27:17 make stupid voices, I make stupid
1:27:20 jokes. There's another side of me that I
1:27:24 take this knowledge and I apply it to
1:27:26 businesses that I work with and I apply
1:27:28 it to Storyvine and I apply it to things
1:27:29 that I do in the world. That's a very
1:27:32 different experience than than what this
1:27:35 is. So, so one thing I want to talk
1:27:37 about
1:27:39 is if you're looking for or if you know
1:27:42 people that are looking for someone that
1:27:44 can take the long view of something
1:27:46 that's rooted in understanding actually
1:27:49 what's happening with AI, let me know.
1:27:52 I'm happy to talk to people. It's not
1:27:54 always this goofy, you know, sort of
1:27:57 pink bow wearing
1:27:59 um town fool that I play here. Like
1:28:03 there's a very specific reason that I
1:28:05 behave this way here. I think it's
1:28:10 important that everyone
1:28:13 understands that this stuff is
1:28:15 accessible to
1:28:17 everyone. And no matter how stupid you
1:28:20 feel, the fact that you're paying
1:28:22 attention to it at all, you're way ahead
1:28:24 of everyone else. So that's the purpose
1:28:27 of this. But there's a different level
1:28:31 that I do when I'm doing business kind
1:28:33 of stuff. That same philosophy. So the
1:28:36 AI salon, if you could pop up the the
1:28:38 URL URL for the AI salon,
1:28:41 Brandon, if you go to the salon.ai AI
1:28:43 and you say join our community. The AI
1:28:46 salon is this remarkable remarkable
1:28:49 community of AI optimists and creators
1:28:55 and um filmmakers and entrepreneurs and
1:28:59 just individuals trying to figure this
1:29:01 stuff out that are generous and support
1:29:03 one another and do this and and kind of
1:29:05 the whole premise of the salon is that
1:29:08 everyone's welcome. the the the core
1:29:11 criteria for the AI salon is be curious
1:29:14 about AI. You might be scared about it.
1:29:17 You might be pissed off about it. You
1:29:19 might be feel like you're just
1:29:21 completely out of the loop with it. But
1:29:24 if you're curious about it, this is a
1:29:27 community for you. This weekend,
1:29:29 starting Sunday on June 1st, we're
1:29:32 launching what we're calling the AI
1:29:34 Salon Mastermind, which is a
1:29:36 subscription, a paid level for people
1:29:39 that want to step up their game. And so
1:29:42 within it, you'll have access to the
1:29:44 clubs and hubs that are that are part of
1:29:46 the salon, and you'll have access to a
1:29:49 smaller, more committed group of people
1:29:52 that are sort of taking their their AI
1:29:54 to the next level. So, just like there
1:29:56 there's a version of of me that is this
1:29:59 show that's this irreverent, fun,
1:30:02 light-hearted thing. There's also this
1:30:04 more serious like you want to transform
1:30:06 your [ __ ] business, I can help you do
1:30:09 that. That's that's what we're doing
1:30:12 with the salon. So, that launches on
1:30:13 Sunday. Um, so there'll be a page there
1:30:16 that that says how you join it. Um, and
1:30:19 I'm super excited about it and people
1:30:20 are excited about it, right? people
1:30:22 within the community because what's
1:30:23 happened is the community's been around
1:30:25 for two and a half years and like any
1:30:27 community, you've got lurkers and you've
1:30:30 got people that check in occasionally
1:30:32 and then you've got like you've got the
1:30:35 explicit leaders of the community,
1:30:37 right? People that step up and say, "I'm
1:30:38 going to volunteer and I'm going to do
1:30:40 this and I'm going to do that." And then
1:30:41 you've got the implicit leaders of the
1:30:43 community. These are people that show up
1:30:45 a lot and they contribute a lot and they
1:30:48 are generous and they're welcoming and
1:30:50 they're they're proactive. The AI
1:30:53 mastermind is this is an ability for
1:30:56 those people to get together and really
1:30:58 [ __ ] take it to the next level. So,
1:31:00 I'm super excited about it. I hope you
1:31:02 join. If you're not part of the AI
1:31:04 salon, like the AI salon is going to be
1:31:06 free
1:31:07 forever. If you want to step your [ __ ]
1:31:10 up, join the AI mastermind, right? If if
1:31:13 if you think that what I you know do
1:31:16 here is worthwhile contributing in the
1:31:18 world then let me know because I do this
1:31:20 for businesses as well on a very
1:31:22 different level. It's a very different
1:31:24 experience. All right. So yeah, Tom
1:31:28 Nodler, count me in. Yeah. Damn it. Oh,
1:31:31 and by the
1:31:34 way, if you're an irregular, and I know
1:31:37 a lot of people here are, we've got 65
1:31:39 people in Tik Tok, 37 people over on the
1:31:42 YouTube and the X's and the LinkedIns,
1:31:43 wherever the hell you are over there.
1:31:47 If you don't know the term irregular,
1:31:49 well, I know you know the the word
1:31:51 irregular, but there is a group of
1:31:54 people who for reasons un I I I cannot
1:31:59 figure it out show up to this these
1:32:01 lives night after night after night
1:32:05 after night. And I appropriately very
1:32:08 early on declared, "You guys are a bunch
1:32:11 of
1:32:13 weirdos." I think that was appropriate.
1:32:15 I think that was an appropriate
1:32:17 label. And
1:32:19 Serena, Amelio's wife, jumped
1:32:22 immediately. The immediate response from
1:32:25 from Serena was, "We're not weird. We're
1:32:29 irregular." Which just cracked me the
1:32:31 [ __ ] up. It is the best name. It, you
1:32:33 know, we tend to be some Gen Xers and a
1:32:36 couple of even boomers in here. Um, so I
1:32:38 think irregular plays on multiple
1:32:41 levels. Bring the I'll bring the Mocks.
1:32:48 Oh my
1:32:50 god. I'll bring the Metamucil. I'll
1:32:53 bring the Mocks. I got the
1:32:56 nachos. Um, the irregulars are a
1:33:01 remarkable a remarkable community of
1:33:04 people that, you
1:33:06 know, we have fun in here, but but like
1:33:09 I said before where, you know, I'm a
1:33:11 certain me here. there's a different
1:33:13 version of me that shows up for
1:33:14 businesses. The irregulars are a fun
1:33:16 bunch. I would argue that the
1:33:18 irregulars, the people that come to this
1:33:20 show on a regular basis are probably
1:33:22 some of the most tapped
1:33:25 in,
1:33:27 smart, educated about generative AI
1:33:30 people that really understand what these
1:33:32 tools make possible and
1:33:34 what the future of work might look like
1:33:37 in the
1:33:38 world because they're in this [ __ ]
1:33:41 conversation night after night after
1:33:42 night.
1:33:43 So Vicki tonight, she sent me a DM on
1:33:48 Mighty Networks and you know, we're
1:33:49 getting ready to launch the mastermind
1:33:51 and and she's like, "Hey, I had an idea.
1:33:53 What about an irregular's badge?" So if
1:33:56 you're an irregular and you know who you
1:33:57 are, um Vicki, how are we going to have
1:34:00 them I know Vickiy's like, "Kyle, could
1:34:03 you get to the [ __ ] point, please?
1:34:05 The badges, please." So we're creating
1:34:07 irregulars badges for Mighty Networks.
1:34:09 So, if you're an irregular, how should
1:34:11 they reach out to you, Vicki? Should
1:34:13 they like comment in the irregular's
1:34:20 channel? Read the screen. Which
1:34:22 [Laughter]
1:34:26 screen can
1:34:28 talk? Oh, can I talk? Yeah, bring Vicki
1:34:31 up.
1:34:33 Vicki,
1:34:34 wait. No, don't reach out to me.
1:34:40 Wait, don't reach out to Vicki. Don't
1:34:42 reach out to Vicki. YouTube pin. Oh,
1:34:45 YouTube pin. Vicki. Okay. If you're
1:34:47 already an
1:34:49 irregular, leave and rejoin. H. Okay. If
1:34:52 you're already an irregular, leave and
1:34:54 rejoin the space and then you'll get
1:34:57 your badge. Okay. Beautiful. All right.
1:35:01 So, irregulars, we get little
1:35:04 badges in the mighty networks. And I
1:35:07 haven't seen it yet, but you know, I
1:35:09 hope there's a turtle on it. We talked
1:35:11 about a turtle, so I'm guessing there's
1:35:14 a turtle. You'll get a confetti shower,
1:35:16 too. That's awesome. Great. YouTube pin.
1:35:19 All right. See YouTube for a regular
1:35:21 badge info in the salon. See YouTube.
1:35:24 Okay, got it. Yeah,
1:35:26 perfect. You have one, Kyle. Do I? Oh,
1:35:28 that's awesome.
1:35:34 [Music]
1:35:35 How do I find my
1:35:37 badge? Do I go to my
1:35:42 [Music]
1:35:48 [Music]
1:35:49 name? Tabs. Oh, tabs. Yeah, tabs.
1:35:55 Yeah, yeah, yeah.
1:36:05 [Music]
1:36:08 Oh, badges. It's a turtle. Look. Was the
1:36:15 best AI learning lab. Irregular with a
1:36:19 little turtle. Come on. Are you kidding
1:36:23 me? Jim Ross. It starts with a badge. It
1:36:26 ends with an irregular tattoo.
1:36:31 So, so first of all, what you should do,
1:36:34 you should you should you should
1:36:36 acknowledge Vicki. You should thank
1:36:37 Vicki. So, how do you do that in here?
1:36:40 You you click on someone's account. So,
1:36:43 if I go
1:36:49 Vicki and I find her and then don't can
1:36:52 I shout her out from in here? Recognize.
1:36:55 Okay. So there's Vicki, right? So here's
1:36:59 how you do this. You go find Vicki. You
1:37:01 click on her thing. Then you click
1:37:03 recognize. And these are the these are
1:37:05 the five values of the AI salon, which
1:37:07 if you don't know, if you haven't been
1:37:08 to the AI salon, our values are [ __ ]
1:37:11 cool. So these values came out of how
1:37:14 people in the salon behave. And so we
1:37:17 just basically named it. They're
1:37:19 curious, they're playful, they're
1:37:21 empathetic, they're generous, and
1:37:23 they're brave. They just put stuff out
1:37:25 there.
1:37:26 So, I am going to
1:37:28 say
1:37:29 Vicki is
1:37:32 incredible. Exclamation point. I'm going
1:37:35 to say she's curious. She's playful. Oh,
1:37:38 I can only do one. Okay, which one? I'm
1:37:40 going to go generous. She's been putting
1:37:41 in a ton of time. I'm going to go
1:37:45 generous. She's empathetic. She's
1:37:48 playful. She's She's all of these, but
1:37:50 we're going to give her cur. We're going
1:37:51 to give her generosity. All right.
1:37:54 Beautiful. So, go recognize Vicki
1:37:58 because she works her ass off for this
1:38:00 community. And now there's the regular
1:38:03 badges. Uh
1:38:07 oh. Leah's wait list for the mastermind
1:38:10 is pinned to the screen. Where? Oh,
1:38:16 beautiful. Yeah, if you go Let's see.
1:38:20 Can I get there from here?
1:38:24 No, I can't. No.
1:38:30 No. Perfect. Hang
1:38:35 [Music]
1:38:37 on. So, if you want to sign up for the
1:38:40 weight list, I mean, this thing goes
1:38:41 live Sunday, but if you want to go live
1:38:46 for the weight list, go to that URL on
1:38:49 screen and that'll take you to a weight
1:38:51 list. still give you pricing
1:38:53 information. By the way, we're we're
1:38:56 doing this for all of 2025. The
1:39:00 mastermind level of the salon is 1999.
1:39:04 It's 20 bucks. Um it's going to go up in
1:39:07 2026, but we wanted to do founding
1:39:10 members because like we've, you know,
1:39:12 there's 3,000 people within the
1:39:13 community. We wanted to make it as easy
1:39:16 as possible for people that want to kind
1:39:18 of step into this new level and and kind
1:39:20 of up their game. Uh we wanted to make
1:39:22 it as easy as possible. So So there's
1:39:24 the pricing. There's the info. All
1:39:25 right. So if you're interested, go sign
1:39:28 up there. And then Sunday, the [ __ ] goes
1:39:30 live, baby. We'll do it live. Do it
1:39:34 live. All right. Fantastic. Fantastic
1:39:38 people.
1:39:40 Um,
1:39:42 so let's see. Um, $20 per month. Yes.
1:39:48 Oh, and okay, there you
1:39:51 go. Yep. 20 bucks a
1:39:54 month. Which I think is a screaming deal
1:39:57 because if you think about the the the
1:39:58 club area of the salon like that, those
1:40:00 whole areas are effectively, you know,
1:40:03 multiple masterminds. and and you know,
1:40:06 we're going to start doing things like
1:40:07 we're going to be doing more events
1:40:09 within the AI salon. Leah and I are
1:40:12 going to host events. Um, you know, club
1:40:16 members are going to host events, things
1:40:17 like that. Um, and we're going to take a
1:40:19 lot of the the educational materials,
1:40:22 formalize that. This is what this is
1:40:24 going to allow us to do is kind of take
1:40:25 the salon to the next level um and make
1:40:28 it more valuable, you know, by people
1:40:30 sort of stepping up to to kind of a
1:40:33 different level of commitment which
1:40:34 which is super cool.
1:40:38 Um lock in lock in that right rate now
1:40:41 for a lifetime. That's the other thing.
1:40:43 If you join as a founding member of the
1:40:45 salon of of the salon mastermind, um you
1:40:48 get that rate for a lifetime so long as
1:40:50 you don't let your subscription lapse.
1:40:52 So, there you have it. All right.
1:40:56 Fantastic badges. We do need some
1:40:59 stinking badges. Exactly. Uh oh, my
1:41:02 computer's dinging. I think people are
1:41:04 out there recognizing Vicki for her
1:41:06 awesome contributions to the salon,
1:41:08 which is [ __ ] cool. Um, beautiful.
1:41:12 All right. Thank you all. Oh, good.
1:41:14 Yeah. Oh, yeah. Tik Tok is giving some
1:41:16 kudos to the Vicki. Vicki works her ass
1:41:19 off for this community. And I know,
1:41:21 listen, I know there's a bunch of you
1:41:23 mod over here on Tik Tok, which is the
1:41:26 lost art of modding on Tik Tok. So,
1:41:28 thank you to all of you that do that and
1:41:30 for, frankly, everyone that shows up for
1:41:32 this thing night after night.
1:41:34 Um, weekend
1:41:37 homework. Boy, I don't know what to say
1:41:40 for this weekend. I feel like
1:41:44 um like a lot of the stuff that we
1:41:46 talked about
1:42:01 tonight. I'll give you some homework for
1:42:03 the weekend. I think some homework for
1:42:05 the weekend is for you to really think
1:42:07 about
1:42:16 If you could do
1:42:18 anything, if skills and education
1:42:23 and
1:42:29 access were suddenly not barriers for
1:42:33 you, what would you do?
1:42:39 Is there something that you've always
1:42:40 wanted to do, but you tell
1:42:43 yourself, "Yeah, but I can't. I'm not
1:42:46 talented enough. I'm not I'm not
1:42:48 educated. I don't have a
1:42:52 degree." All those limitations are about
1:42:55 to be wiped off the planet.
1:42:59 And so if you were given the gift of you
1:43:02 could do
1:43:05 anything. But I can't raise money for a
1:43:07 company. Well, what if you had the
1:43:10 equivalent of a thousand PhD level
1:43:14 employees that would do anything you
1:43:16 wanted them to
1:43:18 do?
1:43:20 Oh, I don't
1:43:22 know. Think about that. I think that's
1:43:25 the the homework this weekend is really
1:43:26 think about like make it an
1:43:28 introspective
1:43:32 weekend. Like imagine like that that
1:43:34 little, you know, demo that we showed of
1:43:36 of Gemini Live where the person's
1:43:38 walking down the street and the the
1:43:39 woman's correcting her. Imagine if you
1:43:42 had like someone at your side all the
1:43:44 time that could just help you do all the
1:43:46 [ __ ] you want to do.
1:43:50 What is what is increasingly
1:43:53 uninteresting to me is which tool to
1:43:57 use. What is increasingly interesting to
1:44:00 me is what do you choose to
1:44:03 do? If you were
1:44:05 given kind of insane levels of resources
1:44:09 and insane levels of skill and insane
1:44:12 levels of knowledge, how would you apply
1:44:14 it? And if you haven't thought about
1:44:16 that before, do that this
1:44:20 weekend. All
1:44:23 right. Next
1:44:25 Tuesday from 5 to 7 Mountain time, if
1:44:28 you go to the AI salon, go to the events
1:44:30 calendar and RSVP right now. Go do it
1:44:33 right
1:44:34 now. Go to the events calendar and RSVP
1:44:37 for next Tuesday's meeting. We've got
1:44:39 Danny Kravitz who who teaches at
1:44:42 Columbia University in Chicago.
1:44:45 He teaches creativity. He teaches
1:44:47 screenwriting. And he's going to come
1:44:49 talk about creativity and
1:44:51 self-expression and the importance of it
1:44:53 and how you do it, how you do it well
1:44:55 and how you do it with AI. Um, I'm
1:44:57 really excited to have Danny here. He
1:44:59 and I, Brandon introduced me to Danny
1:45:01 and um, our conversations are always
1:45:05 like a mind meld where we we have very
1:45:08 similar kinds of ways of thinking. All
1:45:11 right. All right. Is the mastermind
1:45:14 going live on YouTube or directly on the
1:45:16 salon? The the mastermind is really just
1:45:19 a a kind of a private area of the salon
1:45:22 where people can sort of build deeper
1:45:24 relationships. Within that area, we're
1:45:27 likely going to be doing live sessions
1:45:29 and and events and things like that
1:45:30 that'll be exclusive to that area. Um,
1:45:34 so the mastermind isn't a show. It's
1:45:36 it's it's an access area within the
1:45:38 community. It's sort of a a higher a
1:45:42 level of commitment higher like part
1:45:44 part of the reason that that we wanted
1:45:46 to do this is that there are a lot of
1:45:48 people within the salon that are already
1:45:50 leveling up and there's not necessarily
1:45:52 a place for them to go where they can
1:45:54 really connect and sort of start
1:45:56 collaborating together and you know in a
1:45:58 much noisier environment there's just
1:46:01 there's a lot of distraction and so this
1:46:03 is just a more focused area of the of
1:46:05 the salon. So that's what it's about.
1:46:08 All right. Um, Tik Tok pan. Um, the same
1:46:12 thing we do every night, Pinky. Try to
1:46:15 take over the world.
1:46:17 Exactly. So, let's go take over the
1:46:19 world. Listen, thank you for everything
1:46:22 you do. Thanks for showing up for this
1:46:24 thing. I really appreciate it. Friday
1:46:26 night date night's where it's at. I'll
1:46:28 finish up my watered down tequila
1:46:34 here. How can I get there? Um, it'll
1:46:36 just be there. um Sunday and there's if
1:46:41 you can pin that comment again for Mr.
1:46:44 IT for the sign up
1:46:47 form. There's the wait list for
1:46:52 the mastermind. So go go there and you
1:46:56 can sign you can sign up for to show
1:46:58 interest. What would you do Kyle if
1:47:01 what?
1:47:03 What would I do if what?
1:47:12 All right. So, your weekend homework is
1:47:15 No, no, no. They they're asking you what
1:47:17 you would do. They're asking you to
1:47:18 answer your question for the weekend
1:47:20 homework. Oh, now what would you
1:47:25 do, you know, other than hang out with
1:47:27 us for 10 hours a week?
1:47:29 Well, I mean, listen, part of I'm I'm
1:47:33 doing part of what I would
1:47:37 do,
1:47:41 right? I think for
1:47:44 me, the the thing I I would do if I
1:47:47 could if I could do anything is I would
1:47:51 get much better at
1:48:02 having a a
1:48:05 strategic a strategic framing for all
1:48:09 this energy I put into this work.
1:48:12 What I did two and a half years ago or
1:48:14 you know two years ago in April when I
1:48:16 started doing these lives is I made a
1:48:18 commitment to show
1:48:21 up and I've showed up right and it's
1:48:24 made a difference for me it's made a
1:48:26 difference for the irregulars. It's made
1:48:28 a difference for a lot of people that
1:48:29 have kind of passed through these
1:48:33 lives but there wasn't
1:48:36 much beyond that. I didn't really think
1:48:39 about, well, what impact do I want to
1:48:41 have? How do I want to change people's
1:48:43 lives? We've got, you know, 53 people in
1:48:47 here. We had 65
1:48:49 earlier. The reason we don't have a
1:48:51 thousand people in here tonight or
1:48:53 10,000 people in here tonight, I think,
1:48:57 is
1:48:59 that I didn't really apply any strategic
1:49:02 direction to this
1:49:03 energy, right? I wasn't clear about what
1:49:06 do I want to accomplish. And so I think
1:49:08 what I would do if I could do anything
1:49:10 is I would take those areas that I'm
1:49:13 weak self-promotion. I'm really shitty
1:49:16 at it. I'm self-conscious about it. I
1:49:18 don't want to do it. I don't want to
1:49:19 seem like I'm salesy and sleazy to the
1:49:22 detriment of
1:49:24 myself. So I think I'd get better at
1:49:27 that. I think I'd get better at, okay,
1:49:29 we've got a community of people that
1:49:31 show up here night after night after
1:49:32 night. Like what are we doing for them?
1:49:34 How are we allowing them to level up?
1:49:36 What are we, you know, like if if I
1:49:39 could sort of take this thing to the
1:49:40 next
1:49:41 level, you know, the rising boat, the
1:49:44 rising tide raises all boats, right?
1:49:47 Like that's missing right now from
1:49:50 my from my energy output is an
1:49:53 intentionality to it. And so I think
1:49:56 that's what I would do because I love
1:49:58 doing this. I love being in this
1:50:00 conversation. I love inspiring people. I
1:50:02 love being able to articulate what the
1:50:04 [ __ ] is going on and what the [ __ ] is
1:50:06 coming to the best of any of our
1:50:10 ability and
1:50:13 yet it's a little nebulous and soft and
1:50:16 flaky and weird and so I would tighten
1:50:19 that [ __ ]
1:50:20 up and you know improve things for
1:50:24 everyone. So, I think that's what I
1:50:26 would do. And then on top of that, you
1:50:28 know, do things
1:50:29 like produce my musical and, you know,
1:50:34 um, write some more books and do some
1:50:36 more shows and do keynote speaking and
1:50:39 do a lot more. Like, I'm really serious
1:50:41 when I talk about, you know, there
1:50:43 there's sort of the version of me that
1:50:44 here and there's the version of me that
1:50:46 shows up for business. Like, I want to
1:50:47 do more business impact stuff. So, if
1:50:50 you know people that need to talk to
1:50:53 someone, I've got I don't know if you
1:50:54 know this. If you go to my Tik Tok
1:50:57 profile, I've got a link there where
1:50:59 I've got four hours dedicated every
1:51:04 Saturday where people can spend an hour
1:51:06 with me personally, either as a small
1:51:08 team or as an
1:51:10 individual. That's a very different
1:51:12 experience than this is. Like, I don't
1:51:14 think a lot of people know about that.
1:51:16 Why? Because I don't talk about
1:51:17 it. And I want to do that for
1:51:20 businesses, too. So, if you know people,
1:51:22 introduce
1:51:23 me. All right. Tik Tok is loving me in
1:51:25 the comments. Oh, thank you very much.
1:51:26 Cam Cat, can you inspire me, Kyle? You
1:51:28 guys all inspire me, man. It's It really
1:51:31 is. You totally rock, dude. Thank
1:51:36 you.
1:51:37 Yeah, I love the same, Kyle. I'd like to
1:51:40 work in this, but I I need to work
1:51:42 morning, noon, and night. Yeah, I know.
1:51:44 I get it. I totally get it. Um, thank
1:51:46 you for the kind words.
1:51:49 Um, yeah. So, all
1:51:51 right, beautiful. Have a fantastic
1:51:55 weekend. Spend this weekend thinking
1:51:58 about the difference you want to make in
1:51:59 the world. And it doesn't have to be be
1:52:02 [ __ ] huge, right? Doesn't have to be
1:52:05 solve world hunger, cure this disease.
1:52:07 It could be
1:52:09 like I want to get my friends, you know,
1:52:12 educated about AI. It could be like I
1:52:14 want to start a new business or I want
1:52:16 to help people start businesses. Like
1:52:18 just think about the thing that you're
1:52:19 passionate about. I would also encourage
1:52:21 you to think deeply about what is the
1:52:24 [ __ ] that you've told yourself you can't
1:52:28 do for whatever reason. Oh, I can't do
1:52:32 that. I'm not good at music. Oh, I can't
1:52:34 do that. I'm not talented. Oh, I can't
1:52:37 do that. I didn't go to the right
1:52:38 school.
1:52:39 I would say swim around in and maybe
1:52:42 even have a conversation with with
1:52:44 chatbt advanced
1:52:46 voice. I want you to help me discover my
1:52:50 what I would do, my why, my
1:52:53 passion, and have it probe you about the
1:52:56 things that you've shut down in
1:52:58 yourself. and maybe come back Monday
1:53:01 with a sense of
1:53:03 that because I think that as you have
1:53:07 that
1:53:09 you'll you'll look at AI in a very
1:53:11 different
1:53:13 way. If if AI is is a runway to take off
1:53:18 in that vision and AI is the runway,
1:53:21 it's a very different thing than just
1:53:23 which tool do I use? How do I, you know,
1:53:26 how do I make money,
1:53:30 right? That's a stupid [ __ ] question.
1:53:33 The money will come if you do the other
1:53:37 thing. And you do it in a focused and
1:53:39 strategic way, which I'm not right now.
1:53:41 Right. That's that's my work. All right.
1:53:45 Vickiy's computer is dinging because
1:53:47 everyone keep acknowledging her, which
1:53:48 is
1:53:49 great. All right, Rick McCaulay. Here's
1:53:52 to the crazy ones. the ones that don't
1:53:54 know how that don't know they can't
1:53:56 change the world and they're the ones
1:53:57 who do. You're a crazy one. Thank you,
1:53:59 Rick. I appreciate that. Likewise. Right
1:54:01 back right back at you, buddy. Okay,
1:54:03 cool. Um, beautiful. Yes, sir. Hey,
1:54:05 Kyle. Um, one last thing before we leave
1:54:08 because we're feeling philosophical on a
1:54:09 Friday night and you said something a
1:54:11 little while back. You said you don't
1:54:12 have to change the world. And it
1:54:14 reminded me of a quote that if you ch if
1:54:18 you help one person you you're not
1:54:20 changing the world but for that one
1:54:22 person you change their world. Yeah. And
1:54:26 so I want to share and I butchered the
1:54:27 heck out of that quote but it's okay.
1:54:29 Um we perfect here. We're janky. We
1:54:33 embrace the jank.
1:54:35 Um the uh I wanted to share um on the uh
1:54:42 last day of school um we have a bus
1:54:45 driver for my daughter. She's in uh
1:54:47 going into second grade now and she uh
1:54:51 has had the same bus driver every day.
1:54:53 He has not missed a day and we really
1:54:56 wanted to recognize him. We gave him a
1:54:58 gift but we wanted to do something a
1:55:00 little bit more personal. And so, um, we
1:55:04 took a picture of him like waving to her
1:55:08 and I took it into chat GPT and I
1:55:12 converted it into a, uh, which where are
1:55:15 we? What's the wrong tab? Um,
1:55:18 uh, converted into a coloring page. I
1:55:21 had my daughter color it. And we g we
1:55:25 presented that to him on the last day of
1:55:27 school a picture of him driving the bus.
1:55:30 Oh my god, that's amazing. And I'll tell
1:55:34 you that that guy's probably driven a
1:55:36 bus for a while. I've never seen a bus
1:55:38 driver get
1:55:40 mistied. Um he really he like he paused.
1:55:44 There was a moment where he just paused
1:55:45 and like the recognition like kind of
1:55:49 emotion washed over. you could see it
1:55:52 and it's like so when tell when people
1:55:54 tell us that CHBT can't be creative like
1:55:56 well it can but what we do with it and
1:55:59 what we do with that output and
1:56:01 injecting the humanity back into that
1:56:03 output I think is is where the
1:56:06 intersection of all of this lies so
1:56:10 it's the perfect way to end it Brandon I
1:56:12 mean what you're talking about the idea
1:56:15 is everything right your imagination
1:56:19 your you your intent to say, "Hey, it
1:56:22 would be nice if we could get him a box
1:56:24 of chocolates, right?" And then you're
1:56:25 like, "No, could could we do something
1:56:27 more?" And because you know these tools,
1:56:29 you were able to to do this with that.
1:56:31 That's absolutely remarkable, right? And
1:56:35 we can just do this now. But it's like
1:56:37 it's the idea that's the thing. So that
1:56:40 this is the perfect example to end on.
1:56:42 It's awesome. It's incredible. It's like
1:56:44 it's very moving. Look at that
1:56:47 picture. It's amazing. Yeah. He must
1:56:49 have been misty eyed, you know, that's
1:56:52 [ __ ] hanging up proudly in his house
1:56:53 right now, too. You know,
1:56:56 amazing. Yep. Beautiful.
1:57:00 Awesome. Awesome. All right, everyone.
1:57:02 You know who's a bus driver whose life
1:57:05 you can
1:57:06 change? Hey, you want to know something
1:57:09 actually really important about what
1:57:10 Brandon did?
1:57:14 I guarantee
1:57:16 you that bus driver didn't ask, "Oh, did
1:57:21 you make this with Chad
1:57:24 GPT?" It's irrelevant what tool was
1:57:30 used. Maybe Brandon's an illustrator and
1:57:33 he drew that. Maybe he knows, maybe he
1:57:35 hired someone to draw
1:57:36 that. Maybe he used Chat GBT. The tool
1:57:39 is 100%
1:57:43 irrelevant. The only thing relevant
1:57:46 there is the intent and the execution
1:57:49 which were both
1:57:52 flawless. So next time someone says to
1:57:55 you it's not just you know can can AI be
1:57:59 creative like what what's the what's the
1:58:01 big deal with AI? That's the big deal
1:58:03 with AI. What Brandon just did right
1:58:05 there. I want to do something special
1:58:08 for this guy. Oh, I have an idea. Bang.
1:58:10 It's executed. That's the magic of AI.
1:58:14 Damn
1:58:15 it. Love it. Love it. Love it. Love it.
1:58:19 All right. Have a fantastic
1:58:22 weekend. Yeah. Rick McCully. I'm getting
1:58:25 misty just hearing about it.
1:58:29 Exactly.
1:58:31 Because the idea is about humanity and
1:58:35 about connecting with another person.
1:58:37 This is your homework for the weekend.
1:58:39 How do you want to impact
1:58:41 people? Who do you want to impact and
1:58:44 how? Just think about that. Ignore the
1:58:48 tools. Ignore your capabilities. Ignore
1:58:50 your
1:58:52 talents. Who do you want to impact and
1:58:55 how? Come back Monday with some clarity
1:58:57 on that and we'll have some [ __ ] fun.
1:59:00 All right. Love
1:59:03 it. Friday night date night out. I hope
1:59:06 you all had fun tonight. I had a blast.
1:59:08 Thanks for the great questions. Thanks
1:59:09 for the great engagement. All right,
1:59:11 peace out.