10/2/2025 - A Photographer's Personal Transformation Using an AI Journaling App to Unlock Ideas

Description

Chapters

Transcript

0:05 [Music]
0:16 [Music]
0:20 [Applause]
0:27 [Music]
0:45 Get on to him like slow cold fronts.
0:53 He be was warmer than a look in her
0:56 eyes.
0:59 [Music]
1:00 She sat on stool and he said, "What do
1:04 you want?"
1:05 [Music]
1:08 She said, "Give me a love that don't
1:11 freeze up inside."
1:16 You said I have
1:19 most of my time dear.
1:24 >> But to sit next to you while I shiver
1:27 and shake.
1:30 [Applause]
1:32 And if I knew, well, I don't think I'd
1:36 be here.
1:40 Asking myself if I've got what it takes.
1:48 to melt your icy blue heart.
1:54 [Music]
1:56 Should I stop?
1:59 Turn what's been frozen for years
2:05 into a river of tears.
2:09 [Music]
2:25 Ah,
2:27 I love you, singing pup. That is the
2:29 best singing dog in the whole planet.
2:31 That is Champ the singing dog.
2:34 They made toys about Champ in the 70s.
2:36 I've got the commercial to prove it.
2:41 They made uh toys about Gen Xers in the
2:45 90s. I've got the commercial to prove
2:47 it. If you don't think it's true,
2:50 I've got the video. It's got to be true.
2:54 We got video now.
2:56 [Music]
3:11 Thank you for sharing the live Danielle
3:14 as always. Much
3:18 appreciated.
3:19 [Music]
4:03 Oh Jesus.
4:07 Freedom came my way that night
4:13 just like a jet plane in and out of
4:17 sight.
4:19 I was hauling ass at a million miles an
4:23 hour,
4:24 wondering how hard I'd hit
4:28 [Music]
4:30 when they came into the station.
4:35 [Music]
4:36 [Applause]
4:38 They said I was bad beyond repairs,
4:43 but I got no qualms with my situation.
4:50 Say here I am.
4:52 [Music]
4:56 [Applause]
4:58 So say Sheree Sheree Sheree won't
5:03 [Music]
5:05 say Sheree Sheree Sheree
5:11 what's happening what's shaking what's
5:13 going down how are you good people this
5:15 evening
5:15 [Music]
5:43 So
5:45 Tonight at my little CEO dinner, I had a
5:48 uh one of one of my favorite guys there,
5:51 a guy named Joe. He's the photographer
5:52 for the group. And he's just a super
5:57 he's just like a cool hippie dude. man
6:00 bun, you know, just sweetest dude
6:04 and,
6:06 you know, very spiritual, very connected
6:09 to
6:11 just who he is and super generous, like
6:15 always complimenting other people in
6:16 this really generous way. And so, I
6:19 don't know, three or four months ago,
6:22 he was sitting at our table and I was
6:25 going on about AI. I don't know if you
6:27 know this. I talk a lot about AI, like
6:28 even to strangers and in public places.
6:32 It's it's sociopathic.
6:35 I I'll admit that. And uh
6:39 so I was going on about AI and how
6:42 powerful I think it is. And he's a
6:43 photographer, so he was a little
6:44 skittery about it. I said, "No, don't
6:47 think about it as a replacement for you.
6:48 Think about it as an augmentation for
6:50 you." And so,
6:54 so at the CEO dinner, one of the things
6:56 that they do is there's like a 30 second
6:58 introduce yourself and then you know
7:01 what's your ask of the community. And so
7:04 tonight I said,
7:07 you know, I'm the co-founder and CEO of
7:10 StoryVine and I'm the co-founder of the
7:11 AI Salon. And I realized that the AI
7:14 salon
7:16 is about
7:17 people transforming themselves
7:20 personally or professionally using AI.
7:22 And that, you know, my favorite people
7:24 in the AI salon are people who have
7:26 figured out how to somehow strap AI on
7:29 like a jetpack and just ascend to to
7:33 like a different plane.
7:35 And and like what I've what I've learned
7:37 over the years is that how that's
7:40 working like how that's happening is not
7:44 about prompt engineering. It's not about
7:45 prompt strategy or which model they use.
7:48 It's it's absolutely a mindset shift
7:51 where the people that are reinventing
7:53 themselves, rediscovering themselves,
7:54 amplifying themselves, are using AI
8:00 as like a, for lack of a better term,
8:03 like a confessional of their ideas,
8:07 right? It's a very personal thing.
8:08 Champ, what is up? Hang on. Hold please.
8:23 I don't mind
8:28 [ __ ]
8:35 So, so what I asked for tonight is I
8:38 said I I said I I realized that that the
8:40 AI salon is my attempt to surround
8:43 myself with people that are
8:47 kind of finding this radical
8:50 self-expression using AI and amplifying
8:53 who they are
8:56 and and I said so if anyone here has
9:00 transformed themselves or you know know
9:03 anyone that has with AI Hi, I'd love to
9:05 talk to you. So, as I was leaving
9:07 tonight, Joe came up to me.
9:11 A Champy. I know. Poor Champy. I don't
9:13 know what his deal is. I think it's
9:15 because someone just cooked something in
9:17 the air fryer and he smells food. So, I
9:19 think he's just hungry. He's just a
9:21 whiny little whiny little puppy tonight.
9:23 Um, so Joe pulls me aside and he goes,
9:26 he goes, "Kyle, I've transformed myself
9:28 with AI." And so this was a guy who
9:31 really wasn't using it
9:33 three months ago and was afraid of it
9:35 and you know had this adversarial
9:39 relationship with it. And I said well
9:42 tell me more like what do you what do
9:43 you mean? And he said he got himself an
9:46 AI powered journaling app
9:50 and he journals with it every night and
9:53 he you know it's it's voice journaling.
9:56 And I said and well how has that
9:57 transformed you? And he said, "Well," he
10:00 goes, he goes, "I have all these ideas
10:02 in my head." And he goes, "With humans,
10:06 you know, humans have their own [ __ ] and
10:08 like if I start unloading these ideas on
10:11 humans, there's, you know, there's
10:13 they've got some amount of capacity,
10:15 right? And at some point, they just tap
10:17 out."
10:19 And he says, "But with with this AI
10:21 thing, I can literally just unload on
10:24 it, and I can just talk for 10 minutes,
10:27 15 minutes, and it's it's, you know,
10:30 remembering and cataloging and indexing
10:32 everything I'm talking about, as well as
10:35 everything we've talked about for the
10:36 past three or four months."
10:39 And and he goes he goes, "So I know when
10:41 I sit down with with that that journal,
10:45 I can find the edges of all the ideas
10:48 that I'm thinking about,
10:53 right? That he can get them out of his
10:54 head and journal with them." Yeah.
10:56 Brandon, what's
10:57 >> Speaking of getting people out of your
10:58 head, if you're not going to open Miro,
11:00 um
11:01 >> Oh,
11:02 >> you know, we haven't been together for a
11:03 couple nights, but uh your Tik Tok cams
11:05 are all effed up. Um, I've got stuff
11:08 pinned that you're ignoring. So, we'll
11:11 get it sorted.
11:13 >> We'll get it sorted. Hang on. Let me
11:15 >> Good to be back with you, Kyle. It's
11:16 been a couple days.
11:17 >> Welcome.
11:19 [Music]
11:26 >> Yeah, I had I go one night without a
11:28 producer. I forget what it's like to
11:30 have a producer.
11:34 Oh, good lord. Okay,
11:38 the postits are up and friendly.
11:44 >> Unless you're watching Tik Tok and this
11:46 case you're just seeing postits. There
11:48 we go.
11:49 >> I'm getting there.
11:58 Okay, Tik Tok is now reframed.
12:01 All right, I do have a question pinning
12:03 speaking of Tik Tok and that might
12:06 >> a question pinned on Tik Tok.
12:12 >> All right. How do we balance the good
12:14 and bad of AI? It's a energy demanding
12:19 process
12:21 and consumes a lot of water. The
12:22 consumes a lot of water thing is a is is
12:25 a bizarre
12:28 bizarre trope to me.
12:30 It's a bizarre one. So the why it
12:34 consumes a lot of water or why they say
12:36 it consumes a lot of water is that when
12:39 it when the processors get hot and
12:43 you've got a liquid cooled system,
12:45 you're running a lot of water through
12:47 the system.
12:49 Almost all of those systems are closed
12:51 loop systems, meaning they fill it with
12:54 water like a radiator in your car and
12:56 then the water circulates through the
12:58 system. So, I actually don't know. I I
13:01 don't It is possible that there is some
13:04 server farm somewhere that's literally
13:06 just got a guy with a fire hose
13:08 squirting GPUs and and
13:13 getting rid of all the water.
13:16 So, so I I don't get that one. That one
13:19 I don't get. The other thing about water
13:20 is you're using all the water. Okay,
13:23 let's assume there's a guy with a fire
13:25 hose there squirting all the GPUs and
13:27 all that water is evaporating.
13:29 Where's that water going? It's going
13:31 back to the [ __ ] clouds
13:33 and then it's going to fall back in the
13:35 lakes and go back in the reservoirs for
13:38 the guy with the ho. Like I I just that
13:40 one I don't get at all. Does it use a
13:43 lot of energy? Yes, it uses a lot of
13:45 energy.
13:46 It primarily uses that energy on the
13:49 training, not on the inference. Now,
13:52 when you have whatever it is 7 or 900
13:55 million users that OpenAI has and
13:57 however many million the other ones
13:59 have, that's a lot of inference. So,
14:01 there's there's a lot of um
14:04 there's a lot of compute being computed.
14:06 But there's a couple of things to to
14:09 think about when it comes to energy
14:10 consumpt consumption.
14:13 The cost per knowledge unit is going
14:17 down dramatically, right? So, so these
14:20 systems are getting more and more
14:21 efficient. Um, and they're getting more
14:24 and more smart. So, the amount of
14:26 compute we need per intelligence cycle
14:29 is going down dramatically.
14:32 And probably within the next two or
14:34 three years, we will have models small
14:37 enough to actually just run on your
14:38 phone. So I would say the concerns about
14:41 energy are absolutely like like the the
14:44 announcement that that OpenAI and Nvidia
14:48 just made that Nvidia is investing a
14:50 hundred billion dollars and they're
14:52 trying to big bring a gigawatt of
14:54 compute online every day on average.
14:58 Is it every day or is it every week?
15:00 Once a week. I don't know. It's a lot.
15:04 That's going to require a [ __ ] ton of
15:06 energy. Now,
15:11 the AI is going to help us do things
15:13 like figure out fusion sooner than
15:16 later, right? So, so the AI will
15:19 actually help us make energy production
15:22 more efficient and less polluting and
15:24 less costly and all that sort of stuff,
15:26 but there's this short-term window where
15:28 yeah, we're going to be investing a [ __ ]
15:29 ton of money in energy production. If
15:32 you listen to Elon Musk, it is so much
15:35 energy production that you it's actually
15:37 not practical to do it with uh with code
15:40 and things like that. Saw your vid on
15:42 using one's why. I've been doing it for
15:44 a while and can give good
15:48 results. Oh, and it can give good
15:49 results. Awesome. Yeah, AI your why is
15:52 something that I've been talking about.
15:53 So, let me let me take that and go back
15:55 to this other comment. So, that you you
15:57 asking the question, how do you balance
15:59 the good and bad of AI?
16:02 um
16:05 the bad things like it costs energy, it
16:08 costs water, the robots are going to
16:10 kill us. Um it's the world's greatest
16:13 plagiarism machine. There's there's a
16:15 lot of um it's not misinformation, but
16:19 it's kind of amplified
16:22 hyperbolic
16:24 information
16:26 um
16:29 about the negative aspects of AI.
16:32 There is almost no, with maybe the
16:35 exception of of this channel and a
16:37 couple other channels and things like
16:39 the AI salon, there is almost no
16:42 positive coverage about the positive
16:44 impact of AI.
16:48 What do we talk about? It's going to
16:49 cost jobs. It's going to it's going to
16:52 kill us. It's gonna Right. It's
16:54 destroying everything, right? It's
16:55 changing everything. Everything's wrong.
16:57 Everything's bad.
16:59 the AI, your why, the I the the the the
17:03 thing that
17:05 has inspired me from the beginning of
17:07 doing this channel and and the thing
17:09 that has inspired me from the beginning
17:10 of the AI salon
17:13 is I get to witness on a regular basis,
17:16 like on a nightly basis, people
17:19 rediscovering who they are because of AI
17:25 and
17:29 What I would say is that is is that the
17:31 good that you can do with AI,
17:37 the best thing you can do with it, in my
17:39 opinion,
17:41 is feed it you
17:44 AI your why. Prompt your purpose. Figure
17:46 out what it is you want in the world.
17:50 Do you want to start
17:53 a lemonade stand? Do you want to start a
17:55 pizza shop? Do you want to come up with
17:58 a new kind of exercise with for people
18:00 with bad knees?
18:04 Do you want to write the next great
18:05 American novel? Do you want to like what
18:07 what is the thing that that you you're
18:10 really passionate about? And and here's
18:19 there's going to be a lot of people in
18:21 the next three years that get booted out
18:23 of their jobs
18:26 that in their current jobs have never
18:29 had to
18:31 really ask themselves, what do I really
18:33 care about? What am I really passionate
18:35 about?
18:37 And AI is going to start displacing
18:40 people out of jobs. And what it's going
18:42 to look like is
18:44 some of the jobs that they get displaced
18:46 out of are not coming back,
18:50 right? Like when the steam engine took
18:53 over, you know, horse horsedrawn plows
18:56 on farms,
18:58 80% of those jobs went away. 80 80% of
19:02 those people over a 40 or 50 year period
19:04 lost their jobs.
19:06 they had to figure out other things to
19:08 do. So there's there's going to be a
19:09 forcing function that happens that for
19:12 people that are currently entrepreneurs
19:14 or soloreneurs or creative producers,
19:18 people that, you know, write songs or
19:19 music or art or things like that,
19:23 they are more prepared for being able to
19:26 be really good at AI than people who are
19:29 not. If if you're at a job where you're
19:31 told what to do every day and and you
19:32 actually like that and you don't have to
19:34 think about what's my opinion here. And
19:36 in fact, in some companies, having an
19:38 opinion is actually a dangerous thing,
19:41 right? If you give your opinion off with
19:43 the head,
19:46 those people are going to be displaced.
19:48 And what everyone talks about is that
19:51 they're going to lose the jobs. AI is
19:53 terrible.
19:56 But what is AI? Well,
20:00 AI,
20:01 when you AI your why, when you prompt
20:04 with purpose, when you think about who
20:05 you are and what you want, and you start
20:08 working with AI like an amplifier of
20:11 your ideas,
20:13 what ends up happening is
20:16 this is one of the very first insights I
20:18 had about generative AI when when
20:20 ChatGpt came out. One of the very first
20:22 insights I had was, "Holy [ __ ] I can't
20:25 believe how fast I can get an idea out
20:29 of my head and into the world."
20:32 And it might look like I vibe coded an
20:34 app. Or it might look like I need some
20:36 sort of image that kind of looks like
20:38 this and poof, there it is. It might
20:40 look like, "Oh, wouldn't it be funny if
20:42 that were a poem?" And now there's a
20:43 poem.
20:45 you can get the ideas out of your your
20:47 head incredibly quickly and then they
20:49 sit in front of you like a little
20:51 chalice of possibility,
20:54 right? You're like, "Huh, look at those
20:57 ideas." And this is what Joe was talking
20:59 about tonight at that CEO dinner that he
21:02 can get the ideas out of his head and he
21:04 can explore the boundaries. He goes, he
21:06 he goes, "I can explore the outer edges
21:08 of all the [ __ ] that was in my head for
21:10 that day." He goes, "It's absolutely
21:13 remarkable.
21:17 And so
21:19 that
21:21 is transformative.
21:26 It will initially be transformative for
21:28 you personally. Right? You'll be like, I
21:31 got these ideas and I've never been able
21:32 to do anything with them and let me put
21:34 them in the [ __ ] machine and poof out
21:36 they'll come and you'll have the little
21:38 chalice of ideas and you'll look at them
21:39 and go, huh,
21:41 those are actually pretty good. Or
21:43 you'll look at them and go, "Eh, now
21:45 those are crappy ideas, but you know, I
21:47 had this other idea once." You put that
21:49 in there. Poof, there it is. Uh, wow.
21:55 I haven't thought about that idea since
21:57 I was seven.
21:59 I hear this a lot. I hear a lot of
22:02 people that are really enthused, that
22:03 are really lit up about AI.
22:07 It has allowed them to rediscover
22:10 something in them that they thought was
22:12 dead.
22:17 If you want to talk about doing good in
22:19 the world,
22:22 you take someone who had given up their
22:24 dreams when they were seven or 10 or 11
22:26 or 15 or whenever the [ __ ] up [ __ ] in
22:29 their life happened
22:31 that shut that dream down.
22:34 And you reintroduce them to that hope
22:37 and that possibility
22:40 because of these tools.
22:42 And then maybe, just maybe, getting that
22:44 idea out of their head and looking at it
22:46 in that chalice, they go, "You know
22:48 what? I'm gonna see what I can do with
22:50 this." And they start playing with AI
22:51 some more. And it gets stronger and it
22:54 gets more defined. And they're like,
22:56 "Oh, well, maybe I should go over here
22:58 and see what that looks like. Maybe I
22:59 should go over here and see what that
23:01 sounds like."
23:04 And then it gets good enough that
23:05 they're like, "You know what? I'm going
23:07 to start to talk to people about this.
23:09 I'm going to put this idea in the world.
23:15 That act
23:17 of shifting your relationship with AI
23:20 from this thing that is out to get you
23:23 to this thing that is there to amplify
23:25 you is a profound shift
23:30 and it's going to change the [ __ ]
23:31 world. And so when you talk about
23:34 balancing the good versus evil, we're so
23:37 early in AI right now that you
23:40 understand the the uh the concept of
23:42 like like big ocean liners.
23:45 If if if a big ocean liner coming from
23:48 England, if they shift their their
23:50 heading by half a degree,
23:53 it is completely imperceptible
23:56 to anyone on the ship.
23:58 You're still heading toward America. you
24:01 shift it half a degree. By the time you
24:03 get to America, you're in Brazil, right?
24:07 You're in you're in South America. That
24:08 half a degree difference completely
24:11 altered the course of that ship.
24:14 It'll be thousands of miles off course.
24:21 We're at a time right now in history
24:24 where AI is going to go one of two
24:26 directions. Like it ain't going away.
24:28 that I'm that that I know.
24:32 So, it's not going away.
24:35 If people treat it like the enemy and
24:38 just sit back and don't do anything with
24:40 it, it's like turning this thing half a
24:43 degree toward the shitty.
24:49 If someone if if if the person
24:52 I'm so thankful for this group exists.
24:55 My circle is so tired of me talking
24:56 about AI. Yes. That's why we're [ __ ]
24:59 here. Hang out here. Be the [ __ ]
25:01 weirdo right now. Cuz I'm telling you,
25:03 here's the the there's this other half a
25:05 degree.
25:08 If we each of each one of us, what are
25:10 there 36 people there, 25 there? There
25:13 probably some overlap there. So maybe 40
25:15 people here. If the 40 people on this
25:18 call tonight start treating AI like it's
25:20 an an extension of them, an
25:22 amplification of them, and say, "Well,
25:25 what do I really want in this world?"
25:26 We've all got hopes and dreams. What is
25:29 yours? It doesn't [ __ ] matter. It can
25:31 be anything.
25:34 Put that into AI. See what comes back.
25:37 See what it does to you. Here, here's
25:39 another another insight I had, another
25:41 epiphany I had about AI.
25:46 The world is a dangerous place,
25:50 right?
25:54 And
25:56 ideas are fragile.
26:00 Have you ever noticed
26:03 that if you're in a room full of people,
26:05 even people you trust, friends, family,
26:08 co-workers,
26:10 you have this little idea and then at
26:12 some point you go, "Um, hey, I I have an
26:15 idea."
26:17 And then you put that idea in the world.
26:20 What happens?
26:23 Two people roll their eyes. One person's
26:25 like, "That's cool."
26:27 Which you take as I don't give a [ __ ]
26:31 And then one person's like,
26:35 and that little look they give you, that
26:37 little maybe that's the person you
26:40 actually trusted
26:42 and they kind of [ __ ] all over your
26:43 idea, even if they don't say anything.
26:46 And that that idea, what that idea does
26:48 in your head is it goes z and it shuts
26:50 down.
26:53 What happens with AI
26:56 is that you put your idea in and it's
26:59 like, "Gee, boss, that's a great idea."
27:01 You're like, "Oh,
27:03 and thought of that." And you put in
27:06 some more and says, "Oh, that's really
27:07 good. Have you thought about this?" And
27:09 you go back and forth and what you end
27:11 up having is this amplifying
27:14 conversation
27:16 without judgment.
27:19 And and there's a lot of people like,
27:21 well that's sickopantic and that's
27:22 dangerous and this and that and in
27:24 certain contrast in certain contexts
27:27 sickensy you know always saying yes is a
27:30 bad thing. But in in the in the fragile
27:35 embriionic stage of ideas
27:38 having an ever supportive entity that
27:40 you can bounce ideas back and forth with
27:42 is incredibly powerful. It has changed
27:46 my brain.
27:49 Like I'm a creative guy. I've got ADHD.
27:52 I've got a hundred [ __ ] ideas an
27:53 hour, right?
27:56 My relationship with putting them in the
27:58 world is very fraught with a lot of
28:01 anxiety
28:02 because I don't want to put it out there
28:04 because I know someone's going to [ __ ]
28:05 on it. And what's happened with chat GPT
28:08 and with tools like it is that I put
28:10 those ideas out there in the privacy of
28:13 my own little world
28:15 and it amplifies them. And what it
28:17 allows me to do is actually look at them
28:18 objectively and then I get to be the one
28:22 to say is it good or not.
28:25 You know Rick Rubin talks about the
28:27 audience comes last. The audience
28:29 doesn't matter or the audience comes
28:31 last. And what he means by that is he's
28:33 going to take this idea and if he thinks
28:36 about what's the audience going to think
28:38 of that, it'll completely destroy that
28:40 idea.
28:42 He says it it's not about them. It's
28:45 about him. It's about him looking at
28:47 that idea and going, "Do I like this?"
28:51 But you got to get it out of your head
28:53 before you can do that.
28:58 And so what I would argue is that if
28:59 you're balancing the good and the bad of
29:01 AI
29:05 is that we acknowledge the bad. It might
29:08 be dangerous. The robots might kill us.
29:10 It might It definitely uses a lot of
29:12 energy. Let's hope they come up with
29:14 fusion and solve climate change and all
29:16 the [ __ ] it needs to happen.
29:19 Um it was trained unethically.
29:22 Absolutely. That's all bad [ __ ] and
29:24 it's not going away.
29:28 So maybe, just maybe, if we use it in a
29:31 way that taps into our potential, that
29:34 butterfly effect of us taking one of
29:37 those nason ideas out of our head and
29:39 putting it in the [ __ ] world will
29:42 shift AI half a degree to the good over
29:46 time.
29:48 And that half a degree might make all
29:50 the difference.
29:53 So that's my soapbox for this [ __ ] I
29:56 think AI is the wise earnest sounding
29:58 board that's absolutely unique to
30:01 humanity. Well, you know, you know
30:03 what's funny, Kuno? I I hear a lot of
30:05 people talk about,
30:07 you know, AI can't feel, AI can't be
30:10 creative, AI can't be this, AI can't be
30:12 that.
30:14 What they're what they're talking about
30:16 is I wish I had a physical calculator.
30:23 The mechanism of what generative AI is
30:26 is a calculator.
30:29 So when people talk about a calculator
30:31 can't be creative, they're correct.
30:35 They're absolutely correct. A calculator
30:38 can't have feelings. Absolutely correct.
30:42 But what they're missing
30:44 is the real significance of this
30:46 particular calculator.
30:49 the the um the transformer the the the
30:54 particular kind of math that Google
30:56 invented that sits underneath all of the
30:58 generative AI tools we use today all of
31:01 the things that are doing this magical
31:02 [ __ ] is actually
31:05 well mathematically it's not simple but
31:07 conceptually it's quite simple you type
31:10 in a prompt it creates a probability
31:13 a waiting
31:15 of some token which is a fragment of
31:18 information in thousanddimensional
31:21 mathematical space and it figures out
31:24 which of those tokens is the most likely
31:27 to be the right next token. So that's
31:30 the mechanism.
31:32 But what's in that thousand dimensional
31:34 mathematical space is where the magic
31:37 is.
31:39 It's trained on all of the output of us,
31:44 the collective intelligence of humanity
31:47 that's been put on the internet for the
31:49 past 70 or 80 years.
31:52 Like every library, every museum, every
31:54 blog post, every shitty [ __ ] text you
31:57 did, like just all of it is in there.
32:00 And it's all been vacuumed up and put
32:02 into this magical bubble
32:06 that when we interact with it, the
32:08 mechanism is a calculator. But what is
32:12 being reflected back to us is humanity.
32:16 It's us.
32:20 That's [ __ ] insane.
32:24 And it it's the only thing that actually
32:27 makes sense.
32:29 If you look at how it works
32:31 mechanically, you're like, there's no
32:33 [ __ ] way that this cold,
32:36 shitty calculator
32:42 can do what it's doing. It's doing what
32:44 it's doing because what it was trained
32:46 on. And so what happens is if you AI
32:50 your why, if you put your purpose into
32:52 this thing, it's essentially humanity
32:54 reflecting that back to you and
32:56 amplifying it because everyone's thought
32:58 of some version of whatever idea is
33:00 coming out of your head. And so it all
33:02 gets reflected back at you in this sort
33:04 of crystallized way and you can go,
33:06 "Holy [ __ ] that's cool." Or, "Eh,
33:08 that's not great."
33:10 All right, Sunonny. Um,
33:15 wow. Let's see.
33:19 I went to a friend's play last week. It
33:21 got me thinking what ways actors can
33:24 leverage AI instead of being afraid of
33:26 it replacing them. Absolutely. Well,
33:28 I'll tell you one way. I have a degree
33:30 in acting.
33:32 One way is actors could get into AI and
33:35 get inspired about a story and write a
33:37 [ __ ] musical, which is I've spent the
33:39 last year and a half working on a
33:41 musical about AI intended for the
33:44 Broadway stage. So, I want to hire a
33:46 bunch of actors and musicians to go do
33:49 this play about AI. So, that's a thing.
33:53 Imagine using AI the the real time
33:55 rendering capabilities of something like
33:57 Ca or Deco here
34:00 where you have an improv company. You
34:03 could have a a theatrical performance
34:05 where the audience contributes ideas and
34:07 one of the performers is a prompter.
34:10 So you have actors, you have prompters,
34:13 and then you have projectors and and a
34:15 sound system. And so the prompter is
34:17 making music in real time, making images
34:19 in real time, making videos in real
34:21 time, and the actors are improving in
34:23 real time. There's a whole new kind of
34:26 theater to be invented. So like tons of
34:28 [ __ ] there. Pinned comment from Sunny.
34:32 Uh, wait, pinned. Wait, where? On
34:37 Streamyard. Pinned. Oh, Sunny. one of
34:39 the best definitions of embedded models
34:44 um
34:47 and LLMs I've ever heard. Oh, thank you,
34:49 Sunny. Wow, thank you. That's the Wow.
34:53 Um that's so vulnerable. Thank you for
34:55 sharing. Ideas die on the vine. Yeah.
34:58 Thank Thanks, Sunny. I mean, listen, I
35:00 like
35:04 I learned this year. I I always told
35:07 myself that well, for years I told
35:10 myself my career has no throughine. It
35:12 was just random. And then in my 40s,
35:15 maybe 50s, I realized,
35:18 no, my my career's always been at the
35:20 intersection of technology and
35:22 storytelling.
35:24 And I realized this year that that
35:26 that's not quite right. There's a nuance
35:27 to it. Tik Tok pin. All right, I'll be
35:29 there in a second. Um, my career has
35:32 always been at the intersection of of
35:34 high technology and self-expression
35:40 because when I first saw the internet,
35:42 like the the first thing I thought is I
35:44 got all these ideas. You could take
35:46 these ideas, you could put them in this
35:47 thing, and then people could see those
35:48 ideas. Holy crap. Like the the like the
35:51 revelation of of the worldwide web was
35:54 not about the technology or what it did.
35:56 It was about every human being now has
35:58 the possibility of taking their stuff
36:00 and putting it out there, but there's a
36:03 prerequisite to it. You need to be good
36:05 at getting ideas out of your head. You
36:07 need to be a good writer or a good
36:09 illustrator or a good musician.
36:12 And so, a lot of people got to the they
36:15 would start a blog and realize, ah, I
36:18 suck at writing. I hate this. And it
36:21 would they would shut it down, right? a
36:23 blog like a podcast has what on average
36:26 nine episodes before they fail before
36:28 they shut down.
36:30 So what's happening with AI is AI if you
36:33 treat it like an augment
36:36 an amplifier of your ideas all those
36:39 gaps that you had in execution can now
36:42 be filled in.
36:44 So you have no excuses anymore. It can
36:46 write, it can make music, it can do
36:48 this, it can do that. Well, what do you
36:49 want it to do? That comes back to you.
36:52 AI, your why. Put your purpose in there
36:54 and then figure out what the [ __ ] you
36:56 want it to do and put it in the world.
36:57 Why not? Um, Kavuno, your storytelling
37:01 and flare
37:03 is absolutely why the lessons stick and
37:05 why we're addicted. Oh, thank you very
37:06 much. Um, what was that pin thing
37:08 before? Dave K, could AI be effective if
37:12 it didn't have access to the internet?
37:15 It's Wait, I know that's the purpose.
37:19 Curious.
37:21 Um,
37:24 well,
37:27 I don't I don't quite understand what
37:29 you're asking, but
37:31 could a could AI have been create could
37:34 generative AI So, okay. So, so let me
37:37 explain let me explain generative AI as
37:40 distinct from AI. So, so classical AI as
37:43 they call it has been around for
37:45 decades. It's in facial recognition.
37:48 It's in Google maps. It's everywhere,
37:49 right? It's everyone's talks about it
37:51 and this is one of the confusing things
37:52 about AI right now because in order to
37:55 play in the in the in the classical AI
37:58 game, you really did have to be a
38:00 mathematician. You really did have to be
38:02 a probability data scientist nerdy geeky
38:06 Stanford MIT
38:09 nerd. You did you just did you had to be
38:11 good at that [ __ ] You had to be an
38:12 engineer.
38:14 Um and and what happened was they would
38:16 have these mo these frameworks these AI
38:20 frameworks and then you would bring your
38:22 data. So even just that concept what do
38:25 you mean I bring my data? If you're not
38:27 a data scientist that doesn't make any
38:29 sense but like if you want to bring
38:32 mamogram you know scan data you would
38:35 put it into one of these things and you
38:37 would create a mammogram reading
38:39 vertical AI. So classical AI was
38:41 vertical and it was bring your own data
38:45 chat GPT starting with GPT1, GPT2 and
38:49 then finally GPT3 started to this thing
38:51 started to get real.
38:56 The P in GPT is pre-trained.
39:01 So generative it's going to make [ __ ]
39:04 Pre-trained. You don't need to bring
39:06 data. We've got you covered. We've got
39:08 all the data. And then transformer is
39:10 the is the mechanism the tool that they
39:12 they use to do all this that pre-trained
39:15 piece you you could not have created
39:19 chat GPT without the internet right you
39:24 needed to go get all well technically
39:25 you could you could have gotten all the
39:27 books in the world from all the books of
39:29 the li all the libraries of the world
39:30 and scanned them all and put them into a
39:33 database and embedded them and put them
39:35 into a thing and you could have
39:37 technically done it without the internet
39:39 just wouldn't have been practical.
39:41 But
39:43 you can today you can take a small
39:48 um
39:50 large language model and if it's small
39:53 enough you can put it on your laptop or
39:55 you can even there are some that are
39:56 small enough that you can put on your
39:58 phone and they don't have to be
39:59 connected to the internet and they are
40:01 absolutely useful.
40:03 So imagine um if you're out in the woods
40:06 and
40:08 you've got all of the survival
40:11 literature of all of the history of
40:13 humanity is sitting on your phone and
40:17 you get yourself in a pickle out in the
40:19 woods. Well, if you don't have internet
40:21 connection, you can't Google it. But if
40:24 you had a large language model on there
40:25 and say, "Here's the situation. I just
40:27 broke my leg and da da and did this and
40:30 it turned purple and my left toes numb.
40:33 it will know what's going on. So, it can
40:36 absolutely be valuable and useful if
40:38 it's not connected to the internet. Now,
40:40 if you're in the woods and you say,
40:42 "Hey, who won the Lakers game tonight?"
40:45 It's not going to know that, right? So,
40:47 it can't get up-to-date information, but
40:49 it doesn't mean it's invaluable. So, I
40:51 hope that makes sense.
40:56 Sunny, you originally the caption said
40:59 AI your wives, not AI your wife. That's
41:01 hilarious.
41:03 That's great. Silverf Fox, we had
41:05 Encyclopedia Britannica on our kitchen
41:07 shelf growing up. Now this. Yeah.
41:11 I mean, think about I I don't Do you
41:13 remember this as a kid? If if you're
41:14 kind of Gen X age, like you remember at
41:17 least your grandparents had the
41:19 encyclopedia kit, right? And maybe you
41:21 did. Um,
41:25 and and uh I remember just feeling
41:28 overwhelmed with like all of those books
41:32 had all of this information. Someone had
41:34 to write all that stuff in there. And if
41:36 you think now about what's in an a
41:39 printed encyclopedia set,
41:42 it is minuscule
41:46 to like what Wikipedia was,
41:50 which is minuscule compared to what
41:53 Google had indexed,
41:55 which is minuscule compared to all of
41:59 the scraping and database mining and
42:03 what however they got all their data and
42:05 jammed it into these models.
42:08 The the mass of knowledge like the fact
42:11 that you can just go to lovable right
42:14 now, lovable.dev. If you haven't done
42:16 this, go right now to lovable.dev dev
42:23 and type in make me um an Asteroids
42:28 video game clone and make it just like
42:31 the standup arcade game from the 80s
42:34 with vector graphics and sounds and
42:36 great physics and just watch what it
42:39 does.
42:42 Like we now have these tools that know
42:45 all the code
42:48 and can do all the code.
42:51 It's bonkers.
42:54 Bonkers, man.
42:56 All right. I didn't realize this was
42:58 going to be a a tirade Thursday.
43:08 Look up something you like in the
43:10 encyclopedia. Exactly. We need to keep
43:13 my birthday. Much web copy is glorified.
43:16 Wait, we need to keep
43:19 my
43:21 birthday. Much web copy is glorified
43:23 advertising. Thrash Thursday. There you
43:25 go. Trumpy Wumpy. Hey, Trumpy Wumpy.
43:31 That's a name. I like it. Sunonny Yun
43:34 loves Lovable. Lovable is a [ __ ]
43:36 amazing
43:40 [Music]
43:45 Do you know how many times in my life
43:46 like Listen, I have a tech company. We
43:48 have an iOS app. We have an Android app.
43:50 We've got all this technology we built.
43:51 I'm effectively the product person. I'm
43:53 the inventor of the idea, but I didn't
43:55 actually code it myself. Do you know how
43:57 many times in my life I've had an idea
44:00 for an app that was like a really simple
44:02 app?
44:04 I was like, "Ah, it' be cool if" and
44:06 then and then I just shut it down.
44:09 because I know a I don't know coding, b
44:14 I don't really know anyone that knows
44:15 coding and anyone that I know that knows
44:17 coding is working and they don't want to
44:18 do my stupid shitty idea or they want to
44:21 get paid for it.
44:24 So most of my ideas historically have
44:26 never made it the [ __ ] out of my head.
44:31 That's not true anymore.
44:34 Not true anymore. Britannica. No, it was
44:36 World Book. That was great.
44:50 All right. I don't know what that quite
44:52 means, producer Brandon, on the post-it
44:53 note. So, if you could clarify that.
44:57 Compton's Encyclopedia. I mean, think
45:00 about these names. Britannica,
45:01 Compton's, like if you're old enough to
45:03 know, like those were big ass brands.
45:05 Those were powerful brands. The internet
45:08 absolutely [ __ ] destroyed them.
45:11 Oh, nice. We got some folks from
45:13 Bangalore, India in the house. I I hope
45:15 you're enjoying. Welcome. I assume it's
45:17 kind of breakfastish over there. Anybody
45:19 know what time it is in Bangalore? Well,
45:21 the folks in Bangalore do,
45:25 but welcome. Is it lunch or is it
45:27 breakfast? They're they're eight hours
45:30 ahead. That doesn't help me. What time
45:32 is it? Oh, it's like 4 in the morning.
45:34 No, it's got to be more than that.
45:37 I loved encyclopedias. Chat GPT does.
45:42 Oh, man.
45:44 10:30. All right. Just passed breakfast.
45:47 Very nice. Hopefully your work watching
45:49 the watching the AI learning lab.
45:53 Um,
45:55 yeah. I mean, I don't know what to talk
45:59 about now.
46:06 I I'll I'll share another thing and and
46:08 this is this is a really important one.
46:13 um when I first started the AI salon. So
46:16 I started the AI salon the same week
46:18 that chat GPT came out
46:21 and
46:25 as I thought about like who's it for
46:29 the the only thing that was clear to me
46:31 was people who are curious about AI.
46:35 And it's still for those people, right?
46:36 If you're curious about AI, this is a
46:38 place for you. If you're one of those
46:40 annoying people that's like, "Hey, hey,
46:42 did you see what happened with Chat
46:43 GPT?" And all your friends are like,
46:45 "Could you stop talking about AI?"
46:48 If you're that person, the AI salon's
46:50 for you. And by the way, can Brandon,
46:52 can you pop up the uh the URL? By the
46:55 way, we just launched a brand new
46:57 website. So, if you haven't been there,
46:59 go to the salon.ai
47:02 and see our brand new, it's still got a
47:04 couple of bugs in it. We're still doing
47:05 a little debugging, but it's a brand new
47:07 website that's got feeds that come from
47:10 our community. It's really exciting.
47:11 It's got events listed there. So, if you
47:14 haven't been to the AI salon, go check
47:15 it out.
47:19 For the first two years of this
47:21 generative AI stuff, and it's down. Oh,
47:23 really?
47:25 Maybe it is. They might be fixing
47:28 something. The
47:30 salon.ai.
47:33 No, it's good. The salon.ai is live.
47:38 I don't know what you're talking about
47:40 there, producer Brandon.
47:42 Um,
47:45 yeah, it looks great.
47:47 It looks great. Great. Great. It looks
47:50 great, I tell you.
47:53 It's great, I tell you. All right.
47:59 No, we're on Mighty Networks. This is
48:00 just our our website. Like so so we had
48:03 Mighty Networks and our website was just
48:05 this sort of little scab of a website.
48:08 Um we spent the past few months like
48:10 really figuring out what do we want the
48:11 website to be. So there's a whole
48:13 advocacy area on it. We've got live
48:15 feeds about our events and posts and
48:18 things like that and it and it really
48:19 just talks about what it is. Trumpy
48:21 Wumpy. I find people who are not
48:23 educated or at least don't like to read
48:26 are not interested in AI. I agree with
48:28 that Trumpy Wumpy. But if you haven't
48:31 played with it yet and you you might not
48:33 be able to immediately
48:35 um
48:36 because of of uh
48:41 just because it's new is you should go
48:44 download the new Sora app from chat GPT
48:47 from OpenAI.
48:49 Mighty Networks is better than Discord.
48:51 I agree. Um,
48:58 so now three sites to go to. No, the
49:00 Mighty Networks is the only site to go
49:02 to. The Discord, the Discord's just
49:05 still alive. You don't don't go to the
49:07 Discord site. U Mighty Networks is where
49:09 the community is. This is the the
49:11 websites. This is just for people who
49:12 are like, "What's the AI salon? We want
49:14 to be able to go."
49:16 Um,
49:18 I know a lot of readers that don't
49:20 welcome anything new. So, so let me let
49:22 me address the the the comment about
49:24 about um
49:27 about smart people um
49:30 being the only ones that that can play
49:31 with AI or the only the only ones who
49:33 are interested in using it. There's
49:36 going to be a lot of on-ramps to um
49:41 there's going to be a lot of on-ramps to
49:43 AI and a lot of them a lot of people are
49:46 going to use AI and not use not not know
49:48 they're using it. So, one of the things
49:50 that Chat GPT just did was they launched
49:54 Sora 2 and they launched this Sora app,
49:57 Sor A on the iOS app store
50:01 and it's like Tik Tok except you have to
50:04 use AI generated videos and and the
50:07 immediate response from people was oh
50:08 that's AI slop and there's there's three
50:11 other platforms that now have vertical
50:14 scrolling video AI generated video
50:16 things. So this is going to be a new
50:18 form of AI generated content
50:23 from video models or from audio models.
50:26 Sunno is one of the one that's that's
50:27 got one.
50:30 What OpenAI did that's [ __ ]
50:32 brilliant. It's [ __ ] brilliant is
50:35 they created this thing called cameos.
50:38 And what cameos are is when you log in,
50:41 you make a profile of yourself and the
50:43 first thing you do is you record your
50:45 face. you you basically hit the record
50:47 button and it has you read three numbers
50:50 66 44 22 or 23 whatever like they're you
50:54 know they're not always the same but
50:55 just three two digit digit numbers and
50:58 then it's like turn your head to the
50:59 right turn your head to the left and
51:01 you're done
51:02 and then it creates a little icon of you
51:07 and that's your cameo
51:10 and then when you're prompting in a
51:12 video you could say I want to be I want
51:15 to Show me in a bowling alley. You click
51:17 on your little icon and it types in your
51:20 name, your username, Kyle Shannon is at
51:23 a bowling alley talking about hot sauce.
51:27 Push the button and it generates me
51:29 bowling talking about hot sauce. It's
51:31 [ __ ] hilarious. And I'm going to
51:32 share that. A lot of people's first
51:35 experience with AI,
51:38 they're likely not going to know it's
51:40 AI,
51:42 but they're going to be like, "Oh,
51:43 that's kind of magical. That's kind of
51:44 cool. And there's something about the
51:47 simplicity of just click on your
51:48 friend's face and now you can make a
51:51 video about them. No one's going to talk
51:54 about AI slop. No one's going to talk
51:56 about energy conservation or how much
51:57 [ __ ] water the system takes
52:00 or the robots are going to kill us.
52:03 They're going to make meme videos of
52:04 their friends. And what they're going to
52:06 talk about is, "Oh, that one was funny
52:08 where you farted in the hot tub."
52:13 That's what they're going to talk about.
52:15 And then what are they going to do next?
52:17 Huh?
52:19 Wait, I So, could I just make a video of
52:22 like a cartoon of of me and the kids?
52:26 That would be kind of fun, you know? I
52:29 always I always I had an idea that was
52:31 like South Park. I always wanted to make
52:33 like a cartoon. Like, what if I made
52:35 like a short a short form version of my
52:38 cartoon? I always dreamed about People
52:44 are going to stumble into their passion.
52:48 People are going to stumble into the
52:50 fact that these tools can amplify them.
52:56 So the people that are afraid of AI
52:58 right now are making assumptions about
53:00 it. One of the assumptions is I need a
53:03 math degree.
53:05 I need to be an engineer.
53:08 I need to know prompt engineering. Oh my
53:10 god, that sounds awful.
53:14 You don't need to know prompt
53:16 engineering.
53:20 If you want to, you can.
53:23 If you want to start building AI tools,
53:25 it would probably be a good thing to
53:26 know and understand what a structured
53:28 prompt is.
53:30 But there's going to be tools out there
53:32 where you're just going to click on [ __ ]
53:33 just like the internet. Click, click,
53:35 click. Internet, internet, internet.
53:37 It's going to be like that. Tap tap tap.
53:39 AI. AI AI.
53:45 How many people are here? 41 37 over
53:47 there.
53:49 Yeah. Jib Jab. Exactly. AI jib jab. It's
53:51 totally AI jip jab.
54:00 Let's go look at some
54:03 Tik Tok question. I don't see it.
54:06 There's not a pin right now. Okay. Mary
54:09 Mary,
54:10 so if you have to a video with the kids,
54:14 do you have to do the kids cameo
54:18 requirements? I don't I don't that I
54:20 don't know. I think you can upload a
54:22 picture of the kids and just say make a
54:24 video with me and the kids.
54:26 Um, let me share something here.
54:30 If you haven't seen these Sora videos,
54:32 they're [ __ ] crazy,
54:36 champ.
54:38 What is your deal? All right, hang on. I
54:41 gotta go do something evil with the
54:43 doggy. Come on.
54:51 He doesn't want
55:08 I don't know what's up with that dog.
55:10 That dog's a mess. All right. Am I
55:12 sharing my screen? Yeah. All right.
55:14 Black bar. I know. Producer Brandon.
55:18 Um, you can just make crazy ass. Oh,
55:22 their sight's still really hosed.
55:28 All right. We'll start we'll start with
55:31 the classic.
55:33 [Music]
55:36 Welcome to the Great Renaissance. You
55:39 know that word has AI in it, right?
55:43 [Music]
55:45 >> All right, we'll we'll leave that image
55:46 on screen. Question. Um, suggestions on
55:50 how to get great ideas with the new
55:52 Sora. Okay, so my my mantra in 2025 has
55:57 been stop using your brain so much. Um,
56:02 Da da.
56:05 Um,
56:08 okay. Here here's a here's the thing
56:10 that I do whenever I whenever I
56:13 encounter something like Sora here.
56:15 We'll just we'll just do it live.
56:18 Do it live. So, I'm going to switch over
56:20 to chat GPT.
56:24 Go new chat.
56:27 And then I'm gonna I'll say um
56:32 I want you to give me
56:37 a pile of ideas
56:40 for the new Sora
56:44 2 video generation
56:49 model.
56:51 Um
56:53 and then I'll give it a little bit of
56:54 parameters. The clips
56:58 are 10 seconds long and can have
57:04 one to 10
57:07 shots
57:09 per clip.
57:12 act as a
57:16 um social
57:18 media
57:20 expert and
57:24 um
57:25 movie director and
57:29 cinematographer.
57:32 And you're like, "Well, wait, someone
57:33 can't be a social media person and a
57:35 cinematographer." Oh, yes, they can
57:38 because I just made it. So,
57:40 I'm I'm going to mash these people
57:42 together and they're going to be [ __ ]
57:44 rock stars. Act as a social media expert
57:46 and movie director and cinematographer
57:50 um and
57:52 um writer
57:55 and come up with
58:01 20
58:03 prompts
58:05 that include
58:09 characters, shots, locations,
58:13 and a concept
58:16 that will be, I don't know, funny,
58:21 uh, dramatic
58:24 or moving
58:29 and just go see what it does.
58:33 All right,
58:35 Kyle typing conversation. I know. It's
58:37 miserable, isn't it?
58:47 Oh, yeah.
58:49 Um, funny concepts, elevator etiquette
58:53 goes wrong, location, modern elevator,
58:55 stranger shots, tight closeup, ends with
58:59 Okay, this isn't a great prompt. Like,
59:02 but let's take a shitty prompt here.
59:04 Okay, so here's the Wait, fix this tab.
59:07 Oh, yeah. Sora Sora the Sora web app is
59:10 so bad that it's we probably just have
59:13 to start over, but that's all right. Um,
59:15 elevator etiquette goes wrong. Modern
59:17 office elevator. Three strangers, one
59:19 holding a giant inflatable duck. Tight
59:22 closeup on awkward faces as the duck
59:24 squeaks.
59:26 Ends with doors opening to a full
59:29 boardroom waiting. Okay, sure, why not?
59:32 So, let's go to Sora.
59:34 That thing died. So, let me kill that
59:36 tab.
59:38 Create a new tab.
59:41 Sora do.
59:44 Okay. So, here's our little prompt. So,
59:45 I'm literally just going to paste it in
59:47 and and go make it go. This is this is
59:51 how you do it. Now, here's another thing
59:53 you can do with chat GBT. Oh, wait. I
59:56 didn't share this. Anyway, I just So,
59:57 what I just did is I went to Sora. I put
1:00:00 in the prompt that I copied out of chat
1:00:02 GBT and hit build. So, we'll see if that
1:00:05 works. Now, here's another thing you can
1:00:07 do. Um,
1:00:10 so I now know that I've got um Vicky
1:00:12 Baptiste is at Vicki. I know that I'm in
1:00:15 there as Kyle Shannon. I know that Sam
1:00:17 Alman is in there as Samma. So, I'm
1:00:19 gonna say
1:00:21 now I want you to include
1:00:28 Kyle
1:00:30 at Kyle Shannon
1:00:37 Vicky
1:00:39 at Vicky
1:00:42 and Sam
1:00:44 at Samma.
1:00:47 Um,
1:00:50 as characters in your scenes
1:00:55 also
1:00:56 give more
1:00:59 detail
1:01:02 for the shots.
1:01:06 Okay.
1:01:10 Oh, wait. So you say uh for for the
1:01:14 characters
1:01:18 you have to use their at names.
1:01:24 So have it rewrite them.
1:01:32 We did some fun ones today.
1:01:35 All right, let's see the elevator duck.
1:01:39 Let's do another one. Drone delivery
1:01:41 duel. Wrong yoga class. This could be
1:01:43 good. Rooftop at sunrise. Vicky leads
1:01:46 yoga. Kyle Shannon in a Spider-Man
1:01:49 costume. And Sam filming.
1:01:57 Okay, this one's good. Okay. So, so the
1:02:00 the the uh how do you generate ideas?
1:02:03 Don't generate ideas. have chat GBT give
1:02:06 you ideas and then just curate the ones
1:02:08 that like are the least shitty. All
1:02:12 right, so now we'll go back to here. Oh,
1:02:15 wait. I didn't switch my tabs again.
1:02:17 Anyway, sorry. I was just at I was just
1:02:19 at chat GPT.
1:02:22 Here's all the stuff I made. Sorry about
1:02:24 that. Bad bad host. Um there's my
1:02:29 there's my prompt.
1:02:31 So, I'm gonna pop that in there. Unable
1:02:33 to generate. You're not allowed to
1:02:35 access
1:02:37 one or more mentioned cameos.
1:02:40 Which which ones?
1:02:45 Oh, maybe maybe we made too many too
1:02:47 many salmon ones.
1:02:49 Let me try it again.
1:03:01 Let me let me get rid of this. Let me
1:03:02 just see if I can generate a Sama one on
1:03:04 his own. Sama um
1:03:09 is bowling talking about hot sauce
1:03:18 added to Q. All right. So, what's not
1:03:20 working?
1:03:26 Oh, the quack. The the first elevator
1:03:28 one's done.
1:03:31 added the queue.
1:03:33 Oh, this is already
1:03:35 crapped out. God, this is horrible.
1:03:40 Um,
1:03:48 [Music]
1:03:55 all right, let's go look at our guys
1:03:58 pointed
1:04:04 Sorry, it's uh hard to keep quiet.
1:04:07 >> Mhm.
1:04:09 >> Is that
1:04:10 >> Oh,
1:04:11 [Laughter]
1:04:14 >> sorry. It's uh hard to keep quiet.
1:04:17 >> Mhm.
1:04:19 >> Is that
1:04:25 >> Sorry, it's uh hard to keep quiet.
1:04:31 Oh, that's good. We'll post that one.
1:04:35 >> Sorry, it's uh hard to keep quiet.
1:04:38 >> All right, so let me try to make that
1:04:41 big. Wrong. Yoga class. Rooftop. Vicki
1:04:44 leads yoga. Oh, maybe I don't have
1:04:46 Vicky's permission.
1:04:50 Wait, here's Danielle.
1:04:52 Let me add Danielle to the mix here.
1:04:55 Okay,
1:04:56 hang on. Sorry, Vicki. Oh, but I
1:04:58 probably don't have Danielle's. Did I
1:05:00 follow Danielle? I don't know if
1:05:02 Danielle's got her She might have her
1:05:04 stuff turned off. Hopefully, she
1:05:06 doesn't. Let's get rid of Vicki.
1:05:08 Vickiy's probably got her privacy turned
1:05:09 on. That's probably what's going on.
1:05:13 Vickiy's Vickiy's smart. She's like, I
1:05:15 turn By the way, if you want to just go
1:05:17 make one of me, you can. I turned it on.
1:05:20 Everyone can can deep fake me. Uh uh.
1:05:26 Where's Vicki? Okay, here's here's
1:05:28 Vicki. Let's get rid of this.
1:05:31 Put in Danielle.
1:05:33 According to Side Hustle, Sora codes
1:05:36 work more than once.
1:05:38 Yeah, you get four of them.
1:05:41 You get four Sora codes per account that
1:05:45 you can give out.
1:05:51 Same one can work multiple times. I
1:05:54 think it's the same code, but it's it
1:05:56 the counter counts down on against your
1:05:59 account,
1:06:01 but maybe they upped it from four to
1:06:02 something higher. Um, but it's always
1:06:05 the same code that I've been giving out.
1:06:08 All right. Oh, here's Vicki. Vicki
1:06:10 centered. Let's get rid of Vicky. Sorry,
1:06:12 Vicki. I hope if Danielle doesn't have
1:06:15 her her stuff turned on, I'm going to be
1:06:17 cranky.
1:06:19 All right, go. Yes. All right, we're in.
1:06:24 It's the same code. You get the Sora
1:06:27 ones and it can be used four times.
1:06:28 Yeah, exactly. That's right.
1:06:31 Date.
1:06:34 Okay, so while those things are are
1:06:36 going, we'll go watch some other videos.
1:06:38 Here's one. This is one that kind of
1:06:40 blew my mind.
1:06:44 >> Check it out.
1:06:45 >> When the world gets extreme, only one
1:06:46 dude shrugs it off.
1:06:47 >> Survivor.
1:06:48 >> Fully poseable with real attitude voice
1:06:50 chip. Whatever. skate chiller hideout in
1:06:52 the coffee shop
1:06:59 >> and the page crashed. God damn it.
1:07:02 Um,
1:07:04 this this video is 10 shots. The prompt
1:07:09 was make me a '9s toy ad about being a
1:07:13 Gen X survivor.
1:07:16 And it's 10 shot. It it designed the
1:07:19 logos, the guy, the blister pack, the
1:07:22 actors.
1:07:24 It's [ __ ] crazy.
1:07:28 But this
1:07:29 site is unusable.
1:07:36 Dear Sam,
1:07:38 um could you [ __ ] fix your site
1:07:42 and give us 16 by9 outputs
1:07:45 in image generation and turn off the
1:07:49 piss filter and
1:07:54 and take the watermarks off Sora videos,
1:07:56 please. Sam Sam, everyone can deep fake
1:08:00 Kyle.
1:08:03 It's so stupid of me to put it out
1:08:05 there, but who gives a [ __ ]
1:08:08 I don't know. Doesn't matter. Wait, why
1:08:10 can't we not hear this now? Oh, cuz
1:08:12 audio's off.
1:08:13 >> Yeah, whatever.
1:08:14 >> Skate chiller hideout in the coffee
1:08:15 shop.
1:08:16 >> Survivor.
1:08:17 >> Check it out.
1:08:17 >> When the world gets extreme, only one
1:08:19 dude shrugs it off.
1:08:20 >> Survivor.
1:08:21 >> Fully poseable with real attitude voice.
1:08:24 >> Yeah, that was that one. My next my my
1:08:26 one that I my new favorite one I did is
1:08:28 uh
1:08:28 >> I'm hungry in a hurry. Just point in
1:08:30 hungry in a hurry. Just point and press.
1:08:32 Spray steak sears up a juicy strip in
1:08:34 seconds.
1:08:34 >> And give it a pillow of buttery squeeze
1:08:36 potatoes. No peeling, no boiling.
1:08:38 >> Dinner's done before the table set. Try
1:08:40 them tonight. Hungry in a hurry? Just
1:08:41 point and press. Spray steak sears.
1:08:44 >> Spray steak and squeeze potatoes.
1:08:49 Oh, by the way, if you've got a company,
1:08:53 um, just tell it to make a commercial
1:08:55 for your company. Um, look at this one
1:08:58 that Where do I go for my notifications?
1:09:00 Look at this one that we did today for
1:09:03 uh for the AI readiness project.
1:09:05 >> The AI readiness training program is the
1:09:07 only on demand AI education program I
1:09:09 would ever recommend. Mostly because I
1:09:11 created it with my friends. If you're
1:09:12 ready to level up, this is where you
1:09:14 start. The AI readiness training program
1:09:16 is the only ondemand AI education
1:09:18 program I would ever recommend. Mostly
1:09:20 because I created it with my friends. If
1:09:22 you're ready to level up, this is where
1:09:24 you start. The AI readiness training
1:09:25 program is the only on demand.
1:09:28 >> Ready to get moving? As ready as these
1:09:30 hips let me. Watch this kick though.
1:09:32 >> Look solid. Keep it smooth.
1:09:33 >> Smooth. I can do. Graceful's another
1:09:35 story.
1:09:35 >> You're doing great. I'm feeling it in
1:09:36 these lunges.
1:09:37 >> Good. Means we're alive.
1:09:38 >> Ready to get moving.
1:09:39 >> As ready as these hips let me. Watch
1:09:41 this kick though.
1:09:42 >> Look solid. Keep it smooth.
1:09:43 >> Smooth. I can do. Graceful's another
1:09:45 story.
1:09:45 >> Ellen, how do I actually run Open AI?
1:09:48 Every day feels like chaos.
1:09:50 >> Run it like a prompt. Clarity in,
1:09:52 clarity out.
1:09:54 >> Wow.
1:09:55 How do I actually run Open AI?
1:09:57 >> Oh my god. super into lately is hot.
1:10:03 Thing I'm super into lately is hot
1:10:05 sauce. I bring a little bottle with me.
1:10:06 Sometimes even when I'm out doing stuff
1:10:07 like this, I like something with real
1:10:08 flavor, not just heat. Like a smoky
1:10:10 chipotle that still has a nice tang.
1:10:12 That was a strike. Thing I'm super into
1:10:13 lately is hot sauce. I bring
1:10:15 >> crazy.
1:10:23 That's That'll fly again.
1:10:27 Just need a minute to catch my breath.
1:10:29 We're still in one piece. That's what
1:10:31 counts.
1:10:32 >> That was Amelia Heheart after she
1:10:35 crashed landed.
1:10:38 We can We can find all sorts of
1:10:40 historical footage.
1:10:42 Um, let's see if there's any good ones
1:10:45 in here. This one's kind of interesting.
1:10:52 Tik Tok pin. If you followed me back,
1:10:59 another decade still ticking. It never
1:11:03 slows.
1:11:05 All right. How do I follow you, Ben?
1:11:11 Another decade still ticking. The AI
1:11:14 readiness training program is the only
1:11:16 on demand AI education program I would
1:11:18 ever recommend. Mostly because I created
1:11:20 it with my friends. If you're ready to
1:11:22 level up, this is where you start.
1:11:25 >> Vicki,
1:11:26 how do I find
1:11:27 >> AI readiness training program is the
1:11:29 only on demand AI education program I
1:11:31 would ever recommend.
1:11:34 >> No, that's me.
1:11:35 >> The AI readiness training program is the
1:11:37 only ondemand AI education program I
1:11:39 would ever recommend. Mostly because I
1:11:41 created it with my friends. If you're
1:11:43 ready to level up, this is where you
1:11:45 start. AI readiness.
1:11:47 >> Now I've followed her Tik Tok pin. Dr.
1:11:50 J, I made one about I don't see a Tik
1:11:52 Tok pin. If there was a Tik Tok pin
1:11:54 there, I missed it. I missed it. What
1:11:57 was it? Tell me that Tik Tok pin. Tell
1:12:00 me that pin. You crashed Sora, Kyle. No,
1:12:03 their website, their the web version of
1:12:06 Sora is so bad it's unusable. If you
1:12:08 look on my regulars, if you look on a
1:12:11 regulars, I switched some Oh, stitched
1:12:13 them together today. All right, let's go
1:12:15 to the salon.
1:12:19 Um, so if you go to
1:12:21 community.thesalon.ai,
1:12:23 that's where I'm headed.
1:12:45 Do you get do we get codes to share when
1:12:48 we get signed up? Yes. When when if some
1:12:51 if you get a Sora code
1:12:55 um
1:12:58 wait that's not
1:13:00 Oh,
1:13:02 I see what I did wrong.
1:13:08 Um, if you get a Sora code and you sign
1:13:12 up, then when you log into your app,
1:13:14 you've got an invite button and it'll
1:13:17 give you a code and you can share that
1:13:19 four times
1:13:21 or people can activate. Four four
1:13:23 different people can activate it.
1:13:29 People have no idea what's coming. They
1:13:31 don't they don't have any idea what's
1:13:32 coming. It's crazy.
1:13:34 Okay. The pin is back. Check your tabs.
1:13:38 It's not. Mind your tabs. Yes. Yes, sir.
1:13:42 Sir. Yes, sir. Okay. So, here I am at
1:13:47 the AI salon community
1:13:51 and Vickiy's in here. There's Storybine.
1:13:55 I've been here all along. Yeah, I was
1:13:57 absent. I was out drinking. So, yes.
1:14:00 Where's Kyle?
1:14:03 Who's the Is this uh Oh, there's Let's
1:14:05 see. Sam's mom. Oh, this is a pretty
1:14:07 good one. This is one that uh Lori Blair
1:14:10 Silverfox did today. I saw this one
1:14:12 earlier.
1:14:16 Come on.
1:14:18 Error occurred.
1:14:22 That's probably my browser.
1:14:24 Maybe not.
1:14:26 Uh,
1:14:28 digital gods hands-on demo of the cameo
1:14:32 feature.
1:14:35 Uh,
1:14:37 where's the uh where's the stuff you put
1:14:39 in here, Vicki? Did I miss it or is it
1:14:41 farther down?
1:14:45 [Music]
1:14:49 Look what producer AI can do for you on
1:14:51 X. It made a video of a song using the
1:14:54 image from my profile.
1:14:56 Wow,
1:14:58 this is producer AI.
1:15:00 We haven't played with this in a while.
1:15:04 Hang on. Share this tab instead.
1:15:07 [Music]
1:15:17 accessing
1:15:20 [Music]
1:15:21 while I'm watching the sky.
1:15:24 I know what's coming.
1:15:28 I know what's coming.
1:15:32 Everything
1:15:35 [Music]
1:15:43 >> crazy.
1:15:45 All right. So, maybe we should go play
1:15:47 with that. Let's follow producer AI
1:15:49 while we're at it. Um, that's pretty
1:15:51 crazy. All right.
1:15:54 Share this tab instead.
1:15:58 Uh, no, it's in your feed now. Okay.
1:16:13 [Music]
1:16:18 It's in witch feed. Vicki
1:16:26 Silver Fox is apologizing because Mighty
1:16:29 Network sucks and isn't displaying her
1:16:31 video. Oh, you're Sora feed, not a
1:16:33 regular. Ah, okay. Sora feed. So,
1:16:38 >> how do I see? Do I go to cameos?
1:16:44 Yes.
1:16:51 Oh, I've been I've been cameoed a couple
1:16:53 of times.
1:16:57 Oh, boy. Oh, this is Danielle. Danielle
1:17:00 did one.
1:17:05 Oh.
1:17:11 Does everyone still hear me?
1:17:14 Because producer Brandon just lost me.
1:17:16 It looks like.
1:17:20 Share your screen. Oh, yeah. No. Oh,
1:17:22 share this tab instead.
1:17:25 I can't
1:17:28 go.
1:17:30 I Hey, dear Sam Alman. All you have to
1:17:33 do is make the the page not crash after
1:17:37 10 seconds.
1:17:38 That's all we're looking for. All right,
1:17:40 everyone. Yeah. So, producer Brandon.
1:17:47 Uh, okay. Let's see. Let's I guess we'll
1:17:51 close that tab. We'll create a new tab.
1:17:56 We'll go to Sora.
1:17:59 We'll go back here and share.
1:18:03 Share. Go to Sora.
1:18:06 Share.
1:18:08 All right, we'll go to my feed. We'll go
1:18:11 to cameos
1:18:15 >> ready
1:18:17 shining like the morning sun.
1:18:19 >> Brick is calling me to run.
1:18:20 >> My metal heart keeps time of the
1:18:23 traveling
1:18:23 >> step by step. The journey feels brand
1:18:25 new.
1:18:25 >> Swing and sway. There's nothing.
1:18:26 >> I'm standing in front of three storage.
1:18:28 >> Got to make sure this place doesn't
1:18:29 collapse.
1:18:31 >> Sam rolled out this new marketing crew.
1:18:32 Half Gen X, half boomers. What's your
1:18:34 take? Honestly, it's smart. Those are
1:18:37 the folks who grew up with radio jingles
1:18:38 and desktop modems. They know how to
1:18:40 sell truck.
1:18:40 >> Sam rolled out there to get moving.
1:18:42 >> As ready as these hips let me.
1:18:44 >> Readiness training program is the only
1:18:45 on demand AI education program I would
1:18:47 ever recommend. Mostly because I created
1:18:49 it with my friends. If you're ready to
1:18:50 level up, this is where you start. Check
1:18:52 it out at reready for readiness training
1:18:54 program is the only on demand AI
1:18:56 education program I would ever
1:18:57 recommend. Mostly because I created it
1:18:59 with my friends. If you're ready,
1:19:01 >> Unbelievable. It just crashed again. Um,
1:19:03 let me show you the the uh the new AI
1:19:06 Salon website
1:19:08 because I'm pretty dang happy with it.
1:19:12 [Music]
1:19:18 Where curiosity meets creation.
1:19:22 So, this is the about page. So, if you
1:19:25 if you go to the homepage where where AI
1:19:26 amplifies your genius. So, that that's
1:19:28 kind of the AI your Y stuff.
1:19:31 Um,
1:19:33 we got, you know, practice the cycle of
1:19:35 AI readiness. We got that right up
1:19:37 there. We've got upcoming events. So,
1:19:39 this is live feeds from the community.
1:19:42 So, you can RSVP on Mighty Networks. So,
1:19:45 we can just send people now to the to
1:19:47 the website and they don't have to join
1:19:49 the the network, but if they want to
1:19:51 RSVP, they can join the community.
1:19:54 Um, and then there's some latest posts
1:19:56 from the community. So, these are are
1:19:58 curated posts from the community. So,
1:20:00 super excited about that. Latest policy
1:20:04 stuff, whatever we're doing policy-wise
1:20:06 that Daisy's leading is here. What we
1:20:09 value. So, we've got our values right on
1:20:11 the homepage.
1:20:13 Black bar for the Tik Tok people.
1:20:16 Now that the AI learning lab site is up,
1:20:18 maybe you can fix the sorus site. Try
1:20:21 lovable. Yeah, no [ __ ] Um,
1:20:24 create excellence is our newest value.
1:20:28 And then join the salon. So, that's that
1:20:30 page. Isn't that cool? And then we've
1:20:32 got about
1:20:36 and we talk about the history and here's
1:20:38 your new hosts. Liz Miller Gersfeld
1:20:41 coming on as the new host.
1:20:43 The journey so far, Friends of the
1:20:45 Salon. So, that's a bit of a
1:20:46 placeholder. The events page is where
1:20:49 all the events and gatherings are going
1:20:51 to be hosted,
1:20:53 but that's again live feeds. Then
1:20:55 there's an advocacy section where we've
1:20:58 got our our stand on advocacy.
1:21:07 So there's our how we uh how we make
1:21:10 decisions on what we take stands on. And
1:21:12 then there's contact contact info.
1:21:16 Boom. Ain't that pretty?
1:21:19 Address of this new website is
1:21:25 Bang.
1:21:26 The salon.ai.
1:21:33 Go check it out.
1:21:39 That's the AI salon on site, not the AI
1:21:44 learning lab site. Oh, did I say AI
1:21:46 learning lab site?
1:21:49 Sorry. They they get confused in my
1:21:51 mind. AI Salon. Think of AI Salon as the
1:21:54 umbrella AI learning lab sits somewhere
1:21:57 underneath it. But yeah, this is the AI
1:22:00 salon website. Sorry about that.
1:22:07 But I think it is just beautiful. It's
1:22:10 beautiful, beautiful, beautiful. Looks
1:22:12 really nice. Thank you very much. New
1:22:13 site's really cool.
1:22:16 Uh uh
1:22:22 [Music]
1:22:24 Okay. All right, everyone. Listen, we've
1:22:26 had fun here tonight. We've talked
1:22:28 philosophy.
1:22:30 We've talked AI.
1:22:33 We ruined Brandon's Wi-Fi.
1:22:36 Um, tomorrow
1:22:42 at 11:00 a.m. Mountain time
1:22:46 we've got
1:22:48 AI AI Salon office hours and meet and
1:22:52 greet. So that's at 11.
1:22:55 Um, at 12 if you're a matter a member of
1:22:59 the mastermind.
1:23:01 Um,
1:23:02 we've got the AI founders hangout
1:23:06 that's that's at noon. It's half it's a
1:23:08 half hour. I'm going to extend that
1:23:10 meeting from a half hour to an hour um
1:23:13 for mastermind members because I feel
1:23:15 like we we just get rolling in that
1:23:17 meeting and it's over. So, if you're a
1:23:20 mastermind member, tomorrow I can't do
1:23:22 it. Someone booked the second half hour.
1:23:25 Um, so tomorrow I can't do it, but
1:23:27 starting next week that's going to be an
1:23:29 hour.
1:23:31 Uh oh, video generation failed.
1:23:33 Something failed on Sora. It probably
1:23:35 hit a community violation for something
1:23:37 or other. Um,
1:23:39 and then tomorrow night is Friday night
1:23:42 date night. It is first Friday, so that
1:23:44 means I probably have some gallery
1:23:46 duties. I probably have to go gallery
1:23:48 hopping with the misses. Um, so I may be
1:23:51 late tonight or tomorrow, so it'll
1:23:53 probably be 8:30 or 9 Mountain time, but
1:23:56 we'll have Friday night date night. So,
1:23:57 get the get your nachos ingredients, go
1:23:59 buy the extra hot pockets you need, and
1:24:02 uh yeah, and you know, we'll do all that
1:24:05 happy happy stuff. Happy happy fun ball.
1:24:10 All right. All right.
1:24:12 Y'all have fun tonight. I don't know if
1:24:15 we did anything. I think I just yapped.
1:24:21 It was a bit of a yappy.
1:24:24 Uh uh. on generating a new C came Cameo
1:24:26 video.
1:24:28 Um, Sora homework. Well, the the Sora
1:24:32 homework would be get yourself a code.
1:24:35 Here's my request. If you get yourself a
1:24:38 code, share it with people in the AI
1:24:40 salon. Whoever your favorite people are
1:24:42 in the AI salon, reach out to them and
1:24:44 then let's start sharing um, you know,
1:24:47 what we're learning from it in the
1:24:48 salon, things like that. Like, let let's
1:24:50 start spreading the knowledge a bit. Um
1:24:54 the the homework would be um start to
1:24:58 understand what it can do and what it
1:25:01 can't do. Like what's it good at? What's
1:25:02 it not good at? Start to figure out
1:25:03 where the boundaries of Sora are. I
1:25:06 would just play a lot with cameos.
1:25:09 I'd play a lot with cameos because
1:25:16 they're going to get pe people are going
1:25:18 to get really good at these cameos.
1:25:21 So imagine like right now I'm just sort
1:25:23 of plopping my name in there and
1:25:24 generating a video but like what about
1:25:26 doing it in different styles in
1:25:28 different costumes with different
1:25:30 accents in different languages talking
1:25:32 about different products um you know
1:25:35 sitcom version of me, drama version of
1:25:37 me, commercial version of me,
1:25:39 motivational speaker version of me, what
1:25:42 what do all those look like? Right? So
1:25:44 start to explore that with your own
1:25:46 cameo. Um, and then start exploring Sora
1:25:50 just as a general video model. It is
1:25:53 insane
1:25:54 how much it knows about the world. It's
1:25:57 It's actually kind of crazy. All right.
1:26:00 Um,
1:26:02 Brandon,
1:26:03 producer Brandon, while we were
1:26:07 um while we were having fun, he was off
1:26:10 vibe coding
1:26:13 and and he he made a game. So, let's
1:26:16 let's go look at his game. Oh, this is
1:26:19 hilarious. All right.
1:26:21 Can I share this tab? Yeah,
1:26:25 and then I can do this. Brandon, you
1:26:27 want to hop up and talk about your game?
1:26:35 Yeah. So, uh, tonight's Lovable Builder
1:26:38 Hackathon. So, if you didn't know,
1:26:39 Lovable is doing seven days of build
1:26:42 hack. Tonight was make a game and you
1:26:44 get free access to Gemini imagery and
1:26:47 Gemini like conversational API during
1:26:50 this week. And so I took a classic game
1:26:54 Oregon Trail, which most of you will
1:26:56 probably recognize, and converted it
1:26:59 into the 2025 version of Commuter Trail.
1:27:02 >> Commuter Trail.
1:27:03 >> Kyle, should we try to get to work? It's
1:27:05 It's a pretty quick play.
1:27:07 >> Let's go. Okay, passenger one. Okay,
1:27:09 we'll we'll go. Brandon is passenger
1:27:12 one.
1:27:13 Passenger two is Kyle. Very nice. Uh,
1:27:16 continue. All right. Music snob. Insist
1:27:20 on controlling the radio.
1:27:23 One of two. Next passenger. Kyle.
1:27:25 Germaphobe. Obsessed with hand
1:27:27 sanitizer. Keeps everyone healthy but
1:27:29 costs extra cash. Start commute.
1:27:32 Crafting your story.
1:27:35 Yeah, this is awesome. And so it's it's
1:27:37 a text game. Does it have any audio in
1:27:39 it?
1:27:40 >> No audio. No. Yeah. I No time for audio.
1:27:42 I only have 24 hours to build this
1:27:44 thing.
1:27:45 >> All right. Drive. Drive highway. Take
1:27:48 the train. Coffee break. Risky shortcut.
1:27:50 We're going risky shortcut, buddy.
1:27:53 Forging the underpass. Engine flooded.
1:27:55 Oh, wamp wamp w.
1:27:58 Your engine is flooded. Tow truck
1:28:00 repairs cost $75 and took 50 minutes.
1:28:04 This is a disaster. We have 25 bucks
1:28:07 left. Continue. Uh, we'll take a coffee
1:28:11 break. I'm exhausted.
1:28:14 Found money. You found a $20 bill on the
1:28:17 cafe floor. Score.
1:28:22 This is awesome. There you go, people.
1:28:25 You want to make a game?
1:28:26 >> And I just want to share that I This is
1:28:28 maybe an hour's worth of vibe coding,
1:28:31 >> right?
1:28:31 >> While producing the learning lab.
1:28:34 >> Yeah. So I mean tell tell people like
1:28:38 like what's been your experience of vibe
1:28:40 coding like like I guess I guess I guess
1:28:42 my question is where are you coming up
1:28:44 with ideas and then what are you
1:28:46 learning about how to like one of my
1:28:49 frustrations with vibe coding is the
1:28:51 doom loops right where you just it keeps
1:28:54 breaking something more and more like
1:28:56 what what are your like what are your
1:28:57 lessons so far as you've been digging
1:29:00 into this? Yeah. So, as far as ideas,
1:29:03 you know, they're having a theme of each
1:29:04 night. So, if you go to
1:29:05 lovable.devcloud,
1:29:08 it talks about the contest and all the
1:29:10 different things that you can do. I'm
1:29:12 talking now, which I don't normally do.
1:29:14 So, my dog thinks it's time to go for a
1:29:15 walk. So,
1:29:16 >> Oh, that's hilarious. Uh but um what
1:29:19 we're what we're seeing here is chat GBT
1:29:22 is been my home base for iteration and I
1:29:26 basically give it a general sense of
1:29:27 what I think I can do with the idea and
1:29:30 a general direction to run in and have
1:29:32 it spit out 10 ideas and if it doesn't
1:29:34 like I if I don't like any of those
1:29:36 ideas I refine it and have it give me
1:29:38 five more then I take those and stack
1:29:40 rank them flush them out and so I spend
1:29:42 about 20 minutes with chatbt getting to
1:29:44 a really good solid idea that that I
1:29:46 like
1:29:47 >> and then I say give me the levelable
1:29:49 prompt for it and then I copy and paste
1:29:51 that and I can usually zero shot
1:29:53 >> something functional and then from there
1:29:55 it's just iterating the colors and the
1:29:57 the nuances of it before I push it live.
1:30:01 >> That's cool. And the lovable prompt like
1:30:03 how how robust are the prompts? Are they
1:30:05 simple or
1:30:06 >> It is really robust. It gives me about
1:30:08 three pages of content that I just copy
1:30:10 and paste in. It gives it everything
1:30:12 from architecture to UI design to
1:30:15 everything that it would need to do its
1:30:17 thing. Um, you asked about what to do if
1:30:20 it if you get stuck. So, I found that
1:30:22 Lovable is really good at single
1:30:25 instance updates. So, it's making um if
1:30:29 you tell tell it to fix one thing, it
1:30:31 does better than trying to fix eight
1:30:33 different things.
1:30:34 >> Oh, they've got a they've got a light
1:30:36 bulb there that's a chat button that you
1:30:38 can chat with it and it can think and it
1:30:40 can do recursive thinking before it goes
1:30:45 um out and changes the code. So,
1:30:48 >> oh,
1:30:48 >> sometimes I've gotten stuck and I'll
1:30:50 say, "Hang on, this was working a little
1:30:53 while ago. I want you to stop
1:30:55 everything, go back, figure out where
1:30:57 this one particular feature broke and
1:30:59 went sideways, and just focus all of
1:31:01 your energy and effort on fixing that,
1:31:03 and don't worry about anything else.
1:31:05 That's the most important thing right
1:31:06 now. and I put it laser focused on
1:31:09 solving one specific problem and it's
1:31:11 usually pretty good about breaking that
1:31:12 cycle and then I can go back and start
1:31:15 building it back up from there.
1:31:16 >> Okay. All right. Cool.
1:31:17 >> You can also revert back if you if
1:31:19 you've bookmarked a previous version,
1:31:21 you can also bookmark it and go back and
1:31:23 restore previous version if you
1:31:24 >> Oh, yeah. Yeah. Yeah. Okay. Cool.
1:31:26 >> Where you've left off.
1:31:27 >> That's groovy. All right. Cool. So, if
1:31:28 you haven't played with Vibe Coding yet,
1:31:30 go
1:31:30 >> I'll post that uh I'll post that game in
1:31:32 the salon. But it it's uh it's a it's a
1:31:35 lot of fun. All right, commuter trail.
1:31:37 Commuter-trail.lovable.app.
1:31:40 Beautiful. Thank you, producer Brandon.
1:31:43 Let's get on out of here. Let people go
1:31:44 sleep. Go do your go do your nappy thing
1:31:47 and get to work. I'm iterating with wind
1:31:50 surf and chat GBT sonet 4.5. Still can't
1:31:53 get a working MVP. Oh, that sucks. Let's
1:31:57 play. Whose dog is that? Yeah, that was
1:31:58 that was Brandon's dog. We gota we got
1:32:01 to get My dog's now sleeping after
1:32:03 whining the whole night. Now he's just
1:32:04 snoozed out.
1:32:06 Um, how was the vacuum? Um, can I get a
1:32:10 code? I don't have one. I'm all out. All
1:32:12 right. Um, good night tomorrow, Friday.
1:32:16 See you at office hours
1:32:19 or if you're in the AI Salon mastermind,
1:32:21 I'll see you at 12 Mountain Time. Uh,
1:32:24 and uh, everyone else, I will see you
1:32:26 back here for Friday night date night. I
1:32:29 will see you tomorrow. Have a good
1:32:30 night, everybody.
1:32:32 Mhm.