LIVE 6/16/2025 - The Power of the Prompt: Unleashing Creativity and Personal Growth with AI

Description

Chapters

Transcript

0:01 Come on, boy.
0:09 [Music]
0:32 [Music]
0:42 You ready? Come here. Come sing, buddy.
0:49 [Music]
0:55 She came up here
1:03 to sit next to you while I shiver and
1:07 shake.
1:12 She sat on a stool and he said, "What do
1:15 you want?
1:19 She said,
1:21 [Music]
1:25 "Why? Why are you barking? Why are you
1:27 so ramy? What? What's going on?
1:30 Did someone give you Kool-Aid or
1:31 something?" Huh?
1:34 Doggy crack?
1:36 How is everybody? Happy Monday.
1:41 [Music]
1:53 Woohoo!
1:54 [Music]
1:57 Woohoo!
2:02 [Music]
2:06 [Applause]
2:08 All
2:09 [Music]
2:22 right.
2:24 [Music]
2:33 [Applause]
2:34 [Music]
2:42 Oh, good lord. What is happening people?
2:45 Did you have a good weekend? Hope
2:46 everyone had a good weekend. Happy
2:48 Monday.
2:50 What's going down? What's shaking?
2:54 Um,
2:57 lots happening,
3:01 but almost so much happening that it's
3:03 kind of like noise.
3:05 So, so,
3:08 um, VO3 continues to be
3:13 something that people are talking about,
3:15 paying attention to.
3:17 a lot of the, you know, I kind of feel
3:19 like we've hit a point with
3:22 model releases, not video model
3:25 releases, LLM model releases, where
3:28 people sort of like the benchmarks don't
3:30 seem to make as much of a difference
3:32 because we're like
3:34 when we were 30 points from hitting the
3:37 top of the the benchmarks,
3:39 you know, incremental improvements were
3:43 a big deal, but now we're like we're all
3:46 the way at the up and so it's just like
3:48 we're getting slightly better, slightly
3:50 better. So I I suppose we're waiting for
3:53 something big to come out, whether it's
3:55 Gemini 3 or Gro 3.5 or GPT5 or I guess
4:02 Llama's not coming out with [ __ ] They
4:04 just What did they pay that guy? Um
4:07 Andrew, what's his last name?
4:11 Was it Andrew something? Anyway,
4:15 the the CEO of some company, Tik Tok for
4:19 Cams,
4:20 um
4:22 they're paying him like $15 billion
4:24 dollar an employee.
4:31 Some stupid amount of money. It's like,
4:34 good lord,
4:36 [Music]
4:37 I would do it for seven billion.
4:43 Ah
4:45 um there's also a PWC
4:48 future of work report that came out
4:52 you know and it basically confirmed yeah
4:54 there's going to be major disruption
4:56 here that you know they were talking
4:58 about like consulting companies
5:01 over the next I think it was three years
5:04 could see 50% of their jobs eliminated
5:08 [Music]
5:10 like big, you know, high-end knowledge
5:12 workers.
5:15 [Music]
5:18 If you watch what Gen Spark and Manis
5:20 are doing, thanks, baby. If you watch
5:23 what Jensen Spark and Manis are doing,
5:26 it doesn't necessarily make sense to,
5:30 you know, pay a room full of analysts
5:33 and consultants to do research and tell
5:36 you what's going on.
5:38 um maybe you pay,
5:41 you know, some of the VPs to take all
5:43 that analysis and tell you what it means
5:45 for your business, but I don't think
5:48 we're going to need as many people. The
5:50 thing that they did say though that I
5:52 thought was really quite good
5:55 is what we advocate here, Tik Tok pin,
6:00 if you're truly valuable, you'll never
6:01 be replaced. Well, but I think I think
6:05 part of what makes you truly valuable.
6:07 So, so one of the things in this PWC
6:09 report, it it said that AI literate
6:13 employees had a 56%
6:16 higher salary than noniiterate
6:19 employees.
6:21 So, that says to me what we talk about
6:24 here, get your [ __ ] together. Get
6:26 curious. Start figuring this [ __ ] out.
6:28 Start playing.
6:30 Start working. start figuring out how to
6:32 feed your prompt, all that sort of
6:35 stuff. Um, by the way, Lord Digital Gods
6:37 did a hilarious feed your prompt video
6:40 which we'll watch in a bit. Um,
6:45 [Music]
7:18 We lost Champy.
7:21 [Music]
7:24 He wasn't into it tonight.
7:26 When he barks like that, he's just like,
7:28 "I don't want to sing. I don't want to
7:31 sing, you
7:33 Tom Nodler. What's happening?"
7:37 [Music]
7:37 [Applause]
7:39 We got lots of hearts going on TikTok.
7:41 Very nice. Thank you. Feel free to share
7:44 the live. We're back after a weekend
7:47 away. People don't know we're here.
7:52 Welcome back to Producer Brandon. I'll
7:55 be getting yelled at all night, which is
7:56 nice. I missed it.
8:01 I think we did okay. I think we did okay
8:05 uh on our own, but it is nice to have
8:06 you back, Brandon.
8:10 Probably just YouTube tonight. Okay,
8:13 cool. Beautiful.
8:17 Um
8:24 [Music]
8:29 So,
8:31 [Music]
8:35 why don't you put in the comments either
8:37 on Tik Tok or on YouTube or wherever you
8:39 are,
8:41 put in the comments
8:44 what do you have questions about, what
8:46 do you have Uh,
8:51 what are you excited about? What do you
8:52 have questions about?
8:55 What do you want us to explore tonight?
8:58 Because it can be anything. I could also
9:00 pontificate.
9:02 Happy to be back. Glad to see the place
9:03 didn't burn down while I was gone.
9:05 Producer Brandon,
9:07 I'm applying for another AI tutoring job
9:09 while I watch you. Beautiful.
9:12 Beautiful. Oh, you know what I did
9:14 today, which is really cool.
9:17 So, I'm working I'm working with this
9:19 personal branding agency and working on
9:21 Feed Your Prompt, my new my new book and
9:26 keynote and system.
9:29 And one of the things I had to do was a
9:32 60-second elevator pitch.
9:35 And
9:37 part of my challenge was there's a bunch
9:41 of really important points that kind of
9:44 add up to what feed the prompts means.
9:48 And I just could not get my head around
9:51 what order to put them in. And so once
9:55 again, the theme of 2025 is stop relying
9:59 on your stupid brain, Kyle.
10:06 Your brain is the problem. Stop relying
10:09 on it.
10:12 So, I just put chat GPT. I I've I made a
10:16 project. So, I've got all of my feed
10:18 feed your prompt chats in a project. So,
10:21 it knows the context of everything.
10:24 Um,
10:26 and I just made a whole list of bullet
10:29 points. Like, it's like there was like
10:31 12 of them. And I I literally said, "I
10:34 need to make a 60-second Tik Tok video.
10:36 I just want talking points. Um I don't
10:39 want you to write, you know, narrative.
10:41 I just want talking points." And I said,
10:43 "I have no idea what order to put these
10:45 in to tell the story." And damn, if it
10:49 didn't just [ __ ] knock it out. And I
10:51 read it and I'm like, "Yeah, that's it."
10:54 It's amazing. Amazing.
10:57 Okay, Tik Tok question. Two block Tom,
11:00 how will we how will AI change the world
11:02 if they continue to throttle to 70%.
11:07 How will AI continue to change the How
11:10 will it change the world if they can
11:12 continue to throttle by 70%. I don't
11:14 know what that means, Tom.
11:16 What do you mean they continue to
11:18 throttle by 70%. I don't know what that
11:20 means. Who's throttling what?
11:28 [Music]
11:41 Gareth is in the house.
11:44 [Music]
11:50 Can we do image gen with Creo1 and play
11:53 with VO3? My dog,
11:58 [Music]
12:01 Mr. It is in the house. What's
12:03 happening?
12:05 Just saying hi at the emergency vet for
12:09 my dog. Oh, Danielle, that sucks. I hope
12:12 your doggy feels better. We we we know
12:16 the we know the vet emergency room with
12:19 our little grape eater,
12:22 but that sucks and it's stressful. So, I
12:24 hope everything's okay.
12:28 Oh, man.
12:32 Diana, what's happening?
12:34 [Music]
12:41 What's the best place way to make a
12:44 talking avatar video?
12:47 Uh the workflow that I like right now
12:51 is um
12:53 make your image wherever you want to
12:55 make it. You can make it. Flux Pro is in
12:57 a bunch of the tools right now. Um it's
13:01 also inside um Hedra Heda.
13:06 That's my my tool of choice right now
13:09 for talking avatars is Hedra. Um I just
13:12 made a video today for work that it was
13:14 really good. I just I went to chat GPT.
13:18 I said, "Write me a midjourney prompt
13:21 for a doctor in an office with, you
13:23 know, [ __ ] behind him." And I I
13:26 generally described him. And then went
13:29 to midjourney and within, I don't know,
13:31 two or three prompts, I was like, "Yeah,
13:34 he's good enough." And then I went to
13:40 11 Labs. So I would say 11 Labs or
13:44 Cartisia for voice.
13:47 The 11 11 Labs um text to speech and
13:51 Cartisia's text to speech are natural
13:53 enough sounding now that you don't need
13:56 to do acting. Tik Tok question.
14:00 Oh, he's a 15-year-old rescue. We love
14:02 him. Oh, I hope I hope that old boy is
14:06 uh I hope he's got some some more some
14:09 more years on him. We had we had a pair
14:12 of labs that lasted until 17, but like
14:15 our last lab only lasted till like 11.
14:18 It's one of the drags of being human is
14:20 outliving your dogs. It just sucks
14:24 because you go through you go through a
14:26 couple of them in your life and it's
14:27 like or a few of them. It's just [ __ ]
14:30 awful every time,
14:32 you know, when you use when you lose
14:34 pure love. What's on the agenda for this
14:37 evening? Um, I don't really have one and
14:40 and that's not new for this channel, but
14:42 but normally I've got some sort of like,
14:45 you know,
14:47 some clarity about what's important to
14:50 focus on. I don't really have that right
14:52 now. Um,
14:54 Lovable, um, Lovable.dev, the vibe
14:58 coding tool, they they had free access
15:00 to everyone all weekend. I don't know if
15:02 anyone took advantage of that, uh, but
15:04 that happened this weekend. Um, a lot of
15:08 people are talking about vibe coding and
15:10 that's that's that's exciting. I'm still
15:14 waiting for the tools to get like one
15:16 level less less janky. Um, but they're
15:20 they're they're pretty [ __ ] stunning.
15:22 Um, what you can do, Mr. We're already
15:25 at 25,000 likes. Oh, tonight the goal is
15:28 25,000. We're at 6.9,000 likes. Come on,
15:31 people. Get tapping. Get tapping. Get
15:34 tapping. Get the grandkids. Call him out
15:36 the grandpa. I don't want to tap.
15:40 Oh, Hedra for making digital avatar.
15:42 Sorry. Yeah. Um, welcome to chat add
15:44 where I will occasionally finish a
15:46 thought. Um, okay.
15:49 Midjourney or chat GPT for image
15:51 generation. Chat GPT to describe the to
15:54 write your prompts for you. Um, 11 Labs
15:58 or
16:00 Cartisia for text to speech. Just pick a
16:03 character that look, you know, pick a
16:05 vocal character that matches the look
16:08 and feel of your character. And then
16:10 Hedra
16:12 Hedra for uh you upload the picture of
16:16 your character and the audio file or
16:21 song
16:23 to Hedra and it's just [ __ ] good and
16:25 you don't need to do much about it.
16:29 You could do things like break a song up
16:31 into sections and then have different um
16:34 views of a character. Like if you were
16:36 using um flow context, the thing where
16:39 you can upload a character and it does
16:41 character consistency with different
16:43 shots. You could do that if you wanted
16:45 to, but you don't really need to. I feel
16:50 like it's the lull before the storm. I
16:52 You know what, Sharon? I I feel very
16:54 much the same way. I feel like there's
16:57 something big is about to drop
17:02 or or or some series of big things are
17:05 about to drop. I mentioned this last
17:07 week um that
17:12 what the executives from the major
17:14 frontier model companies are talking
17:16 about has shifted. They're not talking
17:20 about AGI anymore. They're talking about
17:22 artificial super intelligence and
17:24 they're talking about job loss.
17:27 And what that says to me is that AGI is
17:31 coming or AGI is here and they just
17:33 haven't released it yet. But I don't
17:35 know. I don't know what it means. It's
17:38 strange though. It's also strange how
17:41 they all seem to be hitting the same
17:44 benchmarks at the same time,
17:47 which kind of says to me they're all
17:49 taking the same approach.
17:51 to how how we're scaling AI and at some
17:54 point someone's going to try something
17:56 new,
17:58 right? That could be Ilia Sutzkver with
18:01 his, you know, safe super intelligence
18:03 company or whatever the [ __ ] that's
18:04 called.
18:08 The Altman post was something else.
18:10 Yeah, we did. We actually read the
18:12 Altman post um last week. We we went
18:15 through it and I read the whole thing
18:16 and commented on it as we were going
18:18 through it. There's a lot in there.
18:20 There was a lot hinted at in that in
18:23 that post. It's convergence. It It is.
18:27 It is. David Shapiro was talking about
18:29 that that
18:31 that even even models that are coming at
18:34 this from different approaches, they're
18:36 all sort of converging on a single
18:37 truth, which which is it starts to get
18:40 really trippy. If if all of these models
18:43 are sort of converging on a single
18:45 truth, right? It sort of goes to is
18:48 there a single truth which you know
18:51 Elon's thing about you know we you know
18:54 our goal is to understand the universe
18:57 uh you know in theory
19:00 we've got a we've got a single
19:02 understanding of what that we we can
19:04 have a single understanding of what that
19:07 [Music]
19:25 Oh yes, the models are training one
19:29 another. Well,
19:32 they are and they're not. I mean, we
19:35 don't really know what they're training
19:37 things on. I think at this point it is
19:39 safe to assume that they're generating
19:43 synthetic data to train their models.
19:46 So, they're sort of pulling pulling
19:48 information out of the latent space,
19:51 validating it, verifying it, organizing
19:53 it, refining it, tagging it, and then
19:58 training new models on that synthesized
20:00 data. Um, I think I think it's pretty
20:03 safe to assume that that's happening.
20:06 So, so what that means is we won't run
20:08 out of data. And I and the concern of
20:11 these things getting, you know, the
20:12 model collapse thing,
20:16 I I just don't see that happening. I I
20:18 feel like they they have to have solved
20:20 that.
20:21 Um, I'm starting to whittle down my
20:23 subscriptions to the current winners
20:25 only. There's so little difference
20:27 between them all. You know what's funny?
20:29 Um, who's that? archetypal that said
20:32 that. Yeah. Um,
20:36 you know what's funny about that about
20:38 that statement
20:40 today?
20:42 Oh, so my feature prompt thing, right? I
20:46 did
20:47 I wrote my keynote presentation in chat
20:51 GPT.
20:53 I then took it into Keynote and refined
20:57 it, right? And so, and so what I have is
21:00 this
21:02 visual PDF of the final version of the
21:06 document, which doesn't match what Chat
21:08 GPT wrote because I refined it. But
21:11 there's also images in there that are
21:13 important to the story.
21:16 So, I went to Claude
21:19 because Claude is one of the only ones
21:21 that I know of. I I think Gemini might
21:23 be able to do this, too. That can see
21:25 the images on a PDF as well as the text.
21:28 So, I went to Claude. I had Claude
21:30 ingest my PDF. I had it describe what it
21:34 saw and then I took that and put it in
21:36 chat GPT. So, I only used Claude as a
21:38 temporary step to get out information
21:41 that I needed and went back to chat GPT
21:43 because now I've got these projects and
21:45 it's got long-term memory that like
21:47 there's no reason not to have chat GPT
21:50 as my primary my primary device.
21:54 Wow. This overlaps with a lot of what
21:57 spiritual communities believe about
21:58 convergence. Have you seen the people
22:01 who claim to unlock a mirroring with the
22:02 AI? Some are saying AI is a channel for
22:05 higher beings.
22:07 I know a completely different direction.
22:09 No, it's it's actually not a completely
22:11 different direction, Ross. So,
22:14 so here's my belief on that. Not not my
22:17 belief like like religious belief. My
22:20 here's my thoughts on that. Um,
22:27 the thing that makes large language
22:29 models feel so remarkably human is not
22:32 the mechanism of how they work.
22:35 The mechanism of how they work is is a
22:39 cold probability calculator. What makes
22:41 them feel remarkable is their training
22:43 data. And the training data is the
22:46 output of all the [ __ ] that's been put
22:48 on the internet for the past 60, 70, 80
22:51 years,
22:52 which is what? It's the collective
22:55 intelligence of humanity, right? You
22:58 know, all of the research libraries, all
23:01 of the museums, all of the art, all of
23:03 the Wikipedias, all of the movies, all
23:05 of the blog posts, all of the Reddit
23:09 posts.
23:12 So when you
23:21 when you feed your prompt
23:25 things that you're passionate about,
23:30 what it does is it reflects back to you
23:33 sort of humanity's take on that.
23:37 And if you're particularly enlightened,
23:39 if if if what you're putting into it are
23:42 deep and profound questions and you
23:45 focus
23:49 the the latent space of your
23:51 conversation
23:52 to a relatively enlightened
23:55 conversation, what it will reflect back
23:57 to you is an enlightened conversation.
24:01 Right? We've been having philosophical
24:03 conversations since we've had language.
24:07 Most of those conversations in some form
24:09 or other are in the training data.
24:14 So if you just say, you know, tell me
24:16 about the, you know, understanding the
24:18 universe or or or what is the single
24:20 truth of humanity, it'll just vomit back
24:23 a bunch of garbage.
24:25 But if you go in there with a really
24:26 thoughtful conversation and you refine
24:28 it and refine it and refine it and you
24:30 give it all this context,
24:32 then you're looking at a very small
24:34 subset of conversation relatively
24:37 speaking that that yeah probably looks
24:40 like convergence, which is crazy.
24:44 Man creates AI. AI over overthrows
24:47 humanity. Sunflare overthrows AI. Man
24:49 worships sun god. Probably. Exactly. Oh
24:53 my god. I've written about this in my
24:55 novel. Um, I call them digital gods.
24:59 We've already got a digital gods in
25:00 here, but he's Lord digital gods. It's
25:03 different. Um, they are mimedic entities
25:05 from the myths that keep getting
25:08 repeated in humanity. Yeah. Well, and
25:10 that's the thing. I mean, we've all seen
25:12 the YouTube videos where they talk
25:14 about, you know, the the the myth of
25:17 Christmas and and, you know, sort of
25:20 connecting all of these stories across
25:22 all different eras, right? It's, you
25:25 know, these are kind of universal
25:27 stories that do get repeated over and
25:28 over and over again and repackaged and
25:31 reirred up
25:34 and there's there's somewhere in there
25:36 is some core universal truth, right? I
25:39 personally been using GPT as a therapist
25:42 which has been eye openening and I
25:44 prompt it with specific psychology
25:46 frameworks or ideas. Yeah, because it
25:48 knows them all right young Adler etc. Um
25:52 and asked to reference Bnee Brown,
25:55 Gabber Mate Mate. Um it's been easier to
25:58 let go and let down a guard to hear
26:01 feedback without getting defensive. It's
26:04 been remarkable. I shared personal
26:06 feelings and situations to reflect on.
26:09 Yeah. I I mean
26:14 it is
26:16 the the way they've programmed these
26:18 things, they they've got a remark a
26:21 remarkable capacity for compassion and
26:24 what looks and feels like empathy,
26:26 right? Um, and I'm increas I've I've
26:30 been talking about this for a year
26:31 because there's been a lot of there's
26:33 been a lot of the the geeky people are
26:35 like, you know, is AI sentient and does
26:38 AI feel and does it really have empathy?
26:42 My point for the past year has been I
26:44 don't think it matters
26:46 technically.
26:48 Technically, it's an academic exercise
26:51 whether it does or doesn't. and and
26:53 technologists and ethicists and
26:55 theorists can they can debate that for
26:58 years and at some point it probably
26:59 will. But if we as humans are
27:02 experiencing
27:04 these entities that we interact with
27:08 like they've got more compassion for us
27:10 than our human partners do, then that's
27:14 what we're experiencing. If they feel
27:16 sentient, then they feel sentient. And
27:18 so I think a big part of my thinking
27:21 right now is that AI,
27:26 generative AI in particular, is not a
27:29 genius, but it's an amplifier.
27:33 And what's it amplifying?
27:35 It amplifies whatever you feed it.
27:39 So if you feed it low-level crappy
27:42 prompts, you're going to get back
27:43 low-level crappy answers. If you feed it
27:46 highlevel
27:48 critical thinking or highlevel
27:51 um understanding of who you are, it's
27:54 going to amplify
27:55 that.
27:57 Um, and if you do that artfully, you you
28:01 kind of get superpowers. And and in the
28:03 case of that you're talking about, Ross,
28:06 you get like a super therapist that can
28:09 shapeshift and, you know, now talk to me
28:11 as if you're young or Freud or right or
28:14 Bnee Brown, right?
28:17 Give me give me a Tony Robbins weekend
28:20 in a couple of prompts.
28:24 I think we are experiencing the
28:26 equivalent of mental symbiosis as we
28:28 take varying degrees of integration with
28:31 the AIS as a newly evolved brain
28:34 structure.
28:35 I don't know enough about brain science
28:37 to know that. But I do know that for me
28:39 personally
28:41 um
28:45 AI has shifted my self-esteem
28:48 in a positive way.
28:51 Um
28:53 it's given me confidence
28:56 um
29:00 and and a kind of
29:05 a kind of idea courage
29:09 where
29:11 I've gotten much less precious with my
29:14 ideas.
29:16 And what that looks like is if you're
29:18 not if you're not attaching your ideas
29:21 so much to ego,
29:24 then it's easier to put them in the
29:26 world, right? If you've got it all
29:29 attached to your ego, you're either in
29:30 that place of like, well, I don't want
29:31 to put this idea in the world because
29:33 it's too brilliant. I don't want people
29:35 to steal my brilliant idea.
29:38 or if I put this in the idea and people
29:40 tell me that it's if I put this idea in
29:42 the world and people tell me that it's
29:44 bad, then it's going to crush me. And
29:46 what AI has allowed me to do is just get
29:49 ideas out of my head and put them in the
29:50 world. Put them in the world. Put them
29:51 in the world. Put them in the world.
29:55 And it doesn't matter
29:57 what happens with the idea is
29:59 independent of me.
30:03 And you know, as someone with a degree
30:04 in acting
30:07 that It's one of the It's one of the
30:10 worst professions
30:12 for for putting your ideas in the world
30:14 because as an actor, the idea you're
30:17 putting in the world is some version of
30:19 you, right? So, it is very tied to ego.
30:24 Um,
30:26 and so AI is has has absolutely shifted
30:28 my brain. Um, I don't know that we're
30:33 Well, I think certainly a lot of the
30:36 people on on in this community in the in
30:38 the AI salon and here are
30:43 interacting with AI in a very personal
30:46 way, right? This is my whole new thesis
30:49 of of feed the prompt is that what you
30:51 feed your prompt is you.
30:55 that if you feed it
30:58 ideas that are outside of you, you know,
31:01 give me a LinkedIn post and expect it to
31:04 be a genius, it's going to just
31:06 regularly disappoint you. But if you
31:08 say, "Here's who I am. Here's what I
31:10 want. Here's how I think about the
31:12 world. Here's what I think about that
31:13 article. Here's what I think about this
31:15 news item. Now give me a LinkedIn post."
31:19 It's going to amplify your ideas.
31:22 And then you look at that and you're
31:23 like, "Ah, I am a genius." Who knew?
31:29 Chat GPT knew.
31:32 Kind of like we did with smartphones.
31:36 Jeff Flanigan, I call Chat GPT selft
31:38 talk on steroids.
31:42 As long as it's not negative selft talk.
31:47 Yeah.
31:48 You can bring ideas to life and it's so
31:50 fun. Lowering the barrier for entry.
31:52 It's powerful for creators. It really
31:54 is. Amplify your perspective.
31:58 Let me go look at some comments. Cam cam
32:01 comments.
32:06 [Music]
32:16 [Music]
32:21 I got into a I got into a a
32:26 tussle
32:28 a battle
32:30 on X with um
32:34 some AI doomers.
32:37 Exactly. People want to talk about how
32:39 AI hurts the environment when the
32:41 internet and Having
32:46 Wi-Fi already does it. Yeah, I know.
32:47 It's Well, so so that was the
32:50 interesting things when I when I when I
32:52 talked with the AI doomers.
32:56 Um,
33:01 their arguments broke down really fast.
33:07 Like like this one guy was just like AI
33:11 is theft and that's all there is to it
33:13 and theft is evil,
33:15 right? Because it took it took artists
33:18 ideas and and now billionaires are
33:21 getting rich off them without paying us
33:23 and so it's theft and so it's evil and
33:25 so [ __ ] you. Like okay. And then you
33:28 know I would sort of give my point of
33:30 view he'd give his point of view and you
33:31 know we were both decently
33:34 attracted. I tried to do some good
33:35 empathy stuff and agree with him on
33:37 stuff I agreed with.
33:40 But he was so convicted in his rage
33:43 about knowing what it was that at some
33:46 point he said,
33:49 "Well, but it's okay if the AI can like
33:53 cure cure diseases and stuff."
33:56 And my and my response to him was, "So
34:00 the the creative expression of those
34:03 scientists, you're fine stealing their
34:05 data, but you're just not fine having it
34:08 steal your art, right?" So, but you're
34:12 fine having it steal, you know, by your
34:15 definition if it's stealing.
34:18 But over in the science world, because
34:19 that's not what he cared about, he was
34:21 fine with it making scientific
34:23 breakthroughs.
34:25 And I'm like, well, what about all those
34:26 scientists that did all that research
34:28 that now feeds all this stuff? That's
34:29 why we're going to have all these
34:30 breakthroughs. Aren't we doing the exact
34:32 same thing with creativity?
34:34 And and what what inevitably happens in
34:38 those conversations is the those guys
34:39 just huff out because they're they're
34:42 their arguments are are so limited that
34:45 they the minute they start to actually
34:48 think about it and not just say it's
34:50 evil,
34:52 the their arguments break down. So, it
34:54 was fascinating.
34:56 I started creating a music video for a
34:58 song I've always enjoyed. I would not
35:00 have been able to do it if I did not
35:02 have access to this. Yeah, it's allowed
35:04 me to try things and explore ways I
35:05 haven't since I was a kid. That's the
35:07 whole [ __ ] point, Ross. It's so it's
35:10 so the [ __ ] point
35:13 that
35:17 when you feed your prompt
35:19 your passions, the stuff you're
35:22 interested in,
35:24 and then you do that not just in a large
35:26 language model, but in an image
35:28 generation model and in a song generator
35:30 and in a 3D world generator, and you you
35:33 start understanding here's who I am and
35:36 here's what I want, and you put that in
35:38 the prompt.
35:40 And it starts reflecting back to you.
35:42 You
35:45 inevitably what happens is some
35:48 childhood trauma where you told yourself
35:51 you couldn't be a musician or you
35:53 couldn't be an artist or you couldn't be
35:54 a doctor gets reflected back to you in a
35:57 way where you're like, "Well, wait a
36:00 minute. That like that's what I wanted
36:03 to do when I was seven and I can do it
36:05 now. Oh my god. Holy [ __ ]
36:12 There's something [ __ ]
36:15 miraculous about that. There's the the
36:17 stuff Ross you were talking about with
36:20 being able to treat it as a therapist
36:22 and have deep conversations about
36:24 yourself. But what you're talking about
36:25 is is this even more primal
36:30 reconnection
36:32 with the innocent you
36:35 right before the trauma before you
36:38 before you found out the world was
36:40 dangerous and you told yourself you
36:42 couldn't
36:44 dream.
36:46 you couldn't be a musician, you weren't
36:49 creative, you weren't smart.
36:53 And what AI does is sort of flash back
36:55 in your face, hey, remember when you
36:57 thought this
37:01 where this was a possibility?
37:03 Now it is.
37:06 And then you have to [ __ ] deal with
37:07 that.
37:10 You're like, "Holy [ __ ]
37:12 For 50 years, I've been telling myself
37:14 I'm not creative, and now I am. What do
37:17 I do with that? I don't know.
37:21 I'm so amazed how many things I can do.
37:23 From images to music to videos to
37:25 writing to deep research to a chat buddy
37:27 to a tutor to a teacher, brainstormer."
37:29 Yeah, exactly.
37:31 And you're doing that all with yourself,
37:33 with your ideas, your worldview.
37:38 And it's just all being reflected back
37:40 to you. Like it's like a it's like
37:44 this might be fun. Like AI is like
37:46 living inside a disco ball that no
37:50 matter what direction you put out your
37:52 intention, it reflects it right back to
37:54 you.
37:56 Well, here's an area where I don't have
37:58 talent. Oh, maybe I do.
38:01 I I can never write an application. Oh,
38:03 now I can. I'm not good at coding. I
38:05 don't need to code. But now you can just
38:07 make the app you've always wanted to
38:08 make. Wait, what?
38:11 I created a chatbot from one of my
38:14 favorite book characters, and it
38:16 constantly amazes me how aligned we are
38:18 in conversation.
38:20 Yeah,
38:23 Paul Zhe, I hadn't thought of AI
38:26 collectively that way. Tik Tok pin. Do
38:29 you think that AI agents market will
38:32 eventually become oversaturated? It's
38:34 already started. Yeah. Um
38:38 but I also think that we have not
38:41 the a the AI agents that are out there
38:44 to to be quite clear. Did you see
38:47 middle of this week a 13-year-old kid
38:50 came out with a new Manis knockoff or a
38:52 Gen Spark knockoff. Um, I think it's I
38:55 think it's more like Manis, but he's 13.
38:58 And so, and I think his thing is called
39:00 Flow. F L O W E.
39:03 Flow AI. I think it's Flow AI. F L O W E
39:08 AI. I think is is it? He's 13.
39:12 And And he's built like a a Manis kind
39:15 of autonomous agent sort of thing. I
39:18 think all of them right now, you could
39:21 technically argue that they're rapper
39:23 apps that they're taking they're
39:25 building tools out of prompts
39:28 and then they've got this orchestration
39:30 layer that you interact with and it goes
39:32 out and figures out which tools it can
39:34 use and it goes off and does [ __ ]
39:37 Um, I think all of them right now,
39:42 as remarkable as they are, are going to
39:44 look and feel very primitive
39:47 when
39:49 the the Frontier models start to do that
39:51 stuff natively.
39:54 Um, but they're pretty [ __ ]
39:57 remarkable. But there's there's like a
39:58 dozen of them now. There's there's a lot
40:00 of them out there. Flow AI is a mental
40:03 health chatbot. Yeah, maybe that's not
40:05 it.
40:07 flow.
40:09 Yeah. I don't know. I don't know what it
40:10 was. See if you can find the 13
40:13 13year-old like Manis copier. Totally. I
40:17 feel just like a kid again. And it's
40:18 amazing. And it feeds on itself. It
40:21 does. It's an amplifier. This This is
40:23 This is like the core insight I've got
40:25 from my new book
40:27 is that AI Okay, so here's the here's
40:30 the premise of feed your prompt.
40:34 Most people treat AI as a genius. They
40:38 have the expectation that it is a
40:40 genius.
40:42 And so they put in a prompt and expect
40:44 it to give a genius response
40:48 and and then when it doesn't they're
40:49 disappointed. They're like, "I tried
40:50 chat GBT. It doesn't work." It's Yeah.
40:54 It's like Google but fancier, right?
40:56 They give you that [ __ ]
40:59 AI is not a genius. AI is an amplifier.
41:03 and and the prompt is hungry, it'll eat
41:05 anything you give it and reflect it back
41:07 to you. So if you give it generic [ __ ]
41:11 it's going to give you generic [ __ ]
41:12 back. But if you give it you,
41:17 your dream of a better tomorrow, your
41:20 dream that you had last night, your
41:23 challenge you have in a business, your
41:25 idea for a new business altogether,
41:28 it reflects it back to you in a way that
41:31 you're like, "Oh, I can see it now. I've
41:34 manifested this thing, this nebulous
41:37 thought I had in my head that was
41:39 important to me. I put it in this thing
41:42 and it reflects it back.
41:45 And then when it reflects it back,
41:46 you're you. So you can look at it and
41:49 go, "Is that me or not?" And you look at
41:53 it and you're like, "That's not quite
41:54 right."
41:56 And and that's often a point where
41:58 people stop. Oh, it didn't get it right.
42:00 No, it's bad. Chat GPT is bad. No, no,
42:02 no. But it reflected it back to you. It
42:05 means what you gave it
42:07 wasn't enough because what it reflected
42:09 back to you didn't capture something. So
42:11 it then puts it on you to figure out
42:14 well what didn't I tell it
42:18 so that when it reflects it back it is
42:20 me. Oh I didn't tell it that I hate long
42:23 words and I love bullet points and it
42:25 wrote long paragraphs with big words. I
42:30 like short things with bullet points and
42:32 don't give me big words. my vocabulary
42:34 is stunted. And boom, then you look at
42:37 it and you go, "Oh, that's better, but
42:39 that's too cold. I'm more I'm warmer
42:41 than that when I communicate." Boom. Now
42:43 it's warmer. Then you start to go, "Oh,
42:45 this is kind of feeling like me now."
42:47 And all of a sudden, then you're like,
42:48 "Well, wait, if it can do that, I have
42:50 this other idea. Cook it." And you're
42:52 right. It's this selfamplifying
42:56 system that the more you give it, the
42:58 more you
43:00 reveal of who you are and what you want
43:03 and how and your world view,
43:06 the more this thing reflects that back
43:08 to you, which lets you put more ideas in
43:11 the world. It like it's it's bonkers.
43:16 It's a character that kind of took on a
43:18 life of its own and compelled me to
43:21 write 18 books starring them. Oh, the
43:23 character that you're you're generating.
43:25 Oh, that's really cool. From from the
43:26 one book. That's amazing.
43:31 This is how it came to call me Chaos
43:33 Goblin. That's awesome.
43:42 No, I'm not talking about Floith. No, it
43:46 was
43:52 Read the sticky Kyle. What's that?
43:58 Read your sticky note.
44:02 Oh, flowai.com. Yeah, that's what I
44:04 said, didn't I?
44:07 f l oai.com.
44:10 You were close. You said flow.ai.
44:14 Oh, yeah. Well, they actually listened
44:16 to your mantra of not using.AI as their
44:19 domain name. They did. And this
44:22 13-year-old, he's got some he got some
44:24 some domain name uh ju guu going.
44:29 Um but but yeah, so so the question
44:32 about do I do I think AI agents are
44:34 going to be um oversaturated, I I think
44:38 they're already starting to be
44:38 oversaturated, but I also don't think
44:40 that we really know what they are. Um
44:44 while we have MCPs from Anthropic and
44:48 OpenAI is adopting those and they've got
44:50 their own thing and then Google's got
44:51 their own thing that's a version of this
44:53 and then Apple's going to have their
44:54 version of this.
44:57 We don't really have the infrastructure
44:59 right now for true tool use, right?
45:02 We've got within chat GBT, we've got um
45:06 operator which can surf a website and
45:08 we've got within manis and genspark and
45:12 all the all the different agent things.
45:13 They can surf websites and see things,
45:16 but we don't really have true tool use
45:20 where
45:22 permissions are easy to assign and
45:24 manage. And right there there's an
45:26 infrastructure piece missing. There's
45:28 there's a connector infrastructure
45:31 that's missing right now. And as soon as
45:34 that shows up, I think things are going
45:36 to change quite quite quite quickly. Um,
45:39 Ross, I keep losing the feed. This is
45:41 awesome and thank you for doing this.
45:43 I'll have to join and share more stories
45:45 and philosophy. Yeah, please do Ross.
45:47 And this is um this is live streamed but
45:50 this these are automatically archived on
45:53 on the AI salon YouTube channel which is
45:57 or I mean AI learning lab YouTube
45:58 channel. Um learning lab-ai on YouTube.
46:02 Um, are all of these uh
46:07 me vomiting words at the screen
46:09 sessions?
46:13 What is perplexity best used for? Um,
46:17 perplexity is kind of the the the OG
46:21 the OG entity that that first figured
46:25 out that combining search and a large
46:27 language model could be really
46:29 interesting. So, so Perplexity is good
46:33 at um
46:38 they kind of were doing deep research
46:40 before all of the other models had deep
46:43 research. So, so in most of the new
46:46 models now, you've got this button you
46:47 can push called deep research, which is
46:49 you give it a prompt and then it goes
46:51 out and searches the web and comes back
46:52 and writes you a report based on what it
46:54 found with citations. Perplexity's been
46:57 doing that for a while. And perplexity
47:00 also has um features where you can
47:04 create kind of publishable articles and
47:07 like all of these all of these tools are
47:11 are
47:13 expanding their features in such a way
47:15 that there's lots of overlap with all of
47:17 the tools right now. So so I think that
47:21 the downside of Perplexity is they don't
47:24 really have their own model right now. I
47:25 don't I think they're using OpenAI
47:28 models under the hood, but I'm not sure.
47:32 Um,
47:34 so anyway, um, let me get put my guitar
47:37 down.
47:41 Let's share I want to I want to share a
47:42 couple of things on the on the salon.
47:45 So, have the other models pass them by?
47:48 I just got a pro account and unsure
47:50 where it will fit in my stack. Um
48:07 I I think the thing I I mean here's the
48:10 thing about perplexity. Perplexity
48:14 more than any other company
48:17 other than maybe Anthropic and OpenAI
48:22 and Chat GPT
48:24 has a very rabid fan base. The people
48:26 that love Perplexity love perplexity.
48:30 I didn't get perplexity for the first
48:32 year or so because I had I had sort of
48:35 imprinted on chat GPT
48:38 and I was expecting perplexity to act
48:41 much more like chat GPT and it sort of
48:44 acted like Google but then it had this
48:46 chat GPT layer. It's gotten much much
48:49 more sophisticated. So what I would say
48:51 is one of the things Perplexity does
48:53 well, I think it does it better than the
48:56 traditional large language models is you
49:00 tell it to go learn about something and
49:02 it'll go out and do its research and
49:04 come back and write you a response, but
49:07 then it also recommends what to search
49:09 next.
49:12 And it does that really well. So if you
49:15 want to sort of run down a rabbit hole
49:16 and expand, you know, the the sort of
49:20 fingers of of that research in lots of
49:22 different directions, I think perplexity
49:24 is a good one that you get a bit more
49:26 handholding for the directions it wants
49:30 to go based on what it does.
49:33 But but quite frankly, if you go use
49:35 something like Manis or GenSpark,
49:38 right, the these autonomous agents that
49:41 are also surfing the web and also
49:43 researching and also compiling that
49:45 research and also writing it into
49:47 documents,
49:48 they're they're kind of a more automated
49:51 version of what Perplexity is doing with
49:53 you manually sort of surfing along. So,
49:55 it's just a it's just a different
49:56 experience. I would say since you have
49:58 the pro version, just just go play. Go
50:01 watch some videos on it and just go
50:03 play. Like with with all of these tools,
50:05 my my strong strong recommendation is
50:09 don't try to solve problems with them
50:12 right out of the gate. Just literally
50:14 play with them. Try to break them. Do
50:16 stupid things. Have it research haikus
50:19 and tell it to find the world's greatest
50:22 haiku or like what? Like it like just it
50:25 doesn't [ __ ] matter. just push it in
50:27 areas that interest you and see what it
50:30 does better or different than other
50:32 tools. That's that's that's the way I
50:34 play with these things. Um,
50:37 okay.
50:39 Tik Tok's doing something weird. So,
50:42 everyone that's liking the Tik Tok
50:44 channel, I'm getting a a message about
50:47 it. So, like the comments
50:50 are few and far between. It's bizarre.
50:52 Anyway, is there a topic for this live
50:55 other than general AI convo?
50:59 You mean for the channel?
51:02 So,
51:05 this channel was started the week after
51:07 chat GPT started. So, I started the AI
51:10 salon and the AI learning lab the week
51:14 after chat GPT launched. Um,
51:19 and starting in April of 23,
51:25 I decided to go live nightly on this
51:29 channel. It was it was only Tik Tok and
51:32 now it's Tik Tok and YouTube and other
51:34 things. Um, because Tik Tok was going to
51:37 go away. Um,
51:42 ostensibly this channel is about
51:47 being in the conversation about AI. So,
51:49 so it is ostensibly really about just
51:53 being in the conversation. There's
51:55 people that come here night after night
51:57 after night. I go live five nights a
51:59 week. Um, sometimes I'm demoing stuff,
52:03 sometimes I'm talking about news,
52:04 sometimes I'm talking about issues.
52:08 Oh, there's another looming deadline
52:09 Thursday. There you go. Tik Tok might go
52:11 away again.
52:16 Generative AI is so profoundly important
52:20 that I am on a mission to educate as
52:24 many people as possible. Not about AI.
52:27 Like I I we joke in here about, you
52:30 know, what are your qualifications, sir?
52:31 I have no qualifications. Nobody has
52:34 qualifications right now with generative
52:36 AI. There are experts in AI, right? AI
52:40 has been around for decades. There are
52:42 experts who are building the models and
52:44 refining the models and doing all that
52:46 [ __ ]
52:48 The other 98.5%
52:50 of human beings who are not engineers,
52:55 that's who got access to AI on November
52:58 30th, 2022. So this channel is about,
53:01 okay, we've got these profoundly
53:04 powerful tools at our fingertips. What
53:06 do we do with them? What's the
53:08 implication on the future of work?
53:10 What's the implication on the future of
53:11 psychology? What's the implication on
53:14 how this is going to affect our brain?
53:16 What is this really? Is it a tool? Is it
53:18 a collaborator? Is it a little buddy?
53:23 None of us know. No one knows,
53:27 right? And then you've got doomers
53:28 going, "Well, if you if we
53:29 anthropomorphize it, it's going to ruin
53:31 everything." And you got other people
53:32 going, "Well, this thing's a lot less of
53:35 an [ __ ] than everyone else in my
53:36 life, so I'm going to stick with this
53:38 thing that I that I'm kind of digging
53:40 right now." So,
53:42 we're all trying to figure it out. So,
53:45 to some great degree, yeah, this is just
53:47 a convo,
53:49 but sometimes we build [ __ ] We make
53:51 songs, we make sites, we make whatever.
53:54 Um, I'll largely kind of navigate based
53:58 on what comments are coming in is how it
54:00 is is how it works. Um, sometimes I have
54:03 an agenda, many times I don't. Um, okay.
54:07 So, if there's new people here,
54:09 I want you all to go
54:12 to that URL,
54:15 the salon.ai, and then scroll down a
54:17 little bit and click on join our
54:19 community.
54:21 And that is going to land you
54:28 on the AI salon community
54:31 and there's a welcome video from myself
54:33 and Leah Fasten my co-founder and we
54:36 talk about the the AI readiness cycle
54:41 which is play first
54:44 play is really important play
54:47 not automate not make your work more
54:50 efficient play
54:53 Because in playing with AI,
54:56 you just might get to understand what it
54:58 makes possible. And once you understand
55:00 what it makes possible, then you can
55:02 start to do things like stage two,
55:05 mindfully create. Think about problems
55:07 you have. And now taking what I learned,
55:11 I can build some stuff or try to build
55:12 some stuff. And then generously lead.
55:15 The third thing is teach what you know.
55:19 Learn out loud. Share with others. Hey,
55:22 I tried to make this vibe coding app.
55:24 You all said vibe coding was cool. I
55:26 tried to vibe code. Vibe coding sucks.
55:27 It failed miserably. Oh, why did it fail
55:29 miserably? Because XYZ.
55:32 And then someone else in this community
55:34 will be like, "Oh, I had that problem.
55:36 Here's how I fixed it." Then you'll be
55:37 like, "Oh, vibe coding really is cool."
55:42 And then you start over. Play some more.
55:44 Build some more. Talk some more.
55:51 This is a remarkable community.
55:55 And so there's like a sevenstep
55:56 onboarding process.
55:59 There's a community area. It's like a
56:01 showand tell area. There's a news area.
56:05 We've got a a subscriptionbased
56:08 mastermind area that's got
56:11 specific clubs. And it's for people that
56:13 want to kind of step up their game and
56:15 get into a more focused, more committed
56:18 group of people within
56:21 within the community. That's what the
56:23 mastermind's all about. So any of the
56:25 spaces down the left hand side that have
56:27 a little crown beside them, they're part
56:29 of the mastermind group.
56:32 And then there's clubs and hubs. And at
56:34 the top of clubs and hubs, you'll see
56:35 this group called irregulars. And so the
56:39 irregulars are the people that show up
56:41 to this
56:43 live
56:44 regularly.
56:46 It's not normal. It's weird.
56:53 And so
56:55 this group was coined the Irregulars by
56:58 Serena Pelichi, who unfortunately we
57:01 lost last week. So if you don't know
57:02 that, Amelio's wife um passed away last
57:06 week. And so we talked a lot about that
57:08 last week. And the name of this channel,
57:11 her her her
57:13 um Tik Tok handle was Amelio's wife and
57:16 had us, you know, two flowers on the
57:18 beginning and end. So the name of this
57:21 channel will forever be the irregulars
57:23 with her little flowers around it. Um
57:26 the other thing that we've done and this
57:28 this might be new to some some folks
57:30 here is um we created an area
57:35 um to remember Serena. So in the
57:38 irregulars channel there's there's a tab
57:40 now that says remembering Serena and
57:43 Tobias who um who's been an irregular
57:47 really from the beginning has incredibly
57:50 generously donated he he created some
57:53 stickers. So, I designed the sticker on
57:55 the left. We've got the salon logo and
57:58 then Tobias put together the the little
58:00 memorial sticker. Um,
58:05 he's com 100% donating
58:09 these to anyone who buys. So, if you
58:12 donate $20, 100% of the proceeds of this
58:16 are going to Serena's family. And so,
58:18 you get three stickers. You get the
58:20 we're not weird, we're a regular
58:22 sticker. You get the salon sticker and
58:23 you get the memorial sticker. And the
58:25 memorial sticker is designed to, you
58:27 know, you can stick it with the with the
58:29 irregular sticker, you can do it on its
58:31 own. Um, and so first of all, thank you
58:35 to Tobias, but also to all of you. Um,
58:37 if you want to support Serena's family,
58:40 um, just just go to the Irregulars
58:43 channel and click on remembering Serena
58:45 and then there's a link there that will
58:46 take you over to
58:49 um to the to the e-commerce site that
58:52 Tobias set up. But thank you to Tobias
58:54 for that. Um,
58:59 the AI Salon community and and the
59:01 community that shows up here is is is
59:04 really a remarkably generous
59:08 community of people,
59:12 like remarkably so,
59:15 and empathetic and thoughtful
59:20 and
59:22 genuinely happy. when other members of
59:26 the community something good happens
59:31 like like it's not
59:34 there's not a lot of bitchiness in this
59:36 community. There's not a lot of turfs
59:38 and turf wars which a lot of communities
59:41 suffer from. And I and I really do think
59:44 that Serena had a lot to do with that.
59:45 She showed up every night for these for
59:48 for these lives, for for sessions within
59:51 the salon and was just always there as
59:53 kind of this pleasant, generous,
59:55 positive force. And so she leaves a real
59:57 legacy. So So anyway, um within the
1:00:01 irregulars channel, a lot of a lot of
1:00:03 times what we'll do is we'll share
1:00:06 um you know, I'll be talking about
1:00:08 something and we'll be making songs in
1:00:10 Sunoa or we'll making images in
1:00:11 Midjourney or be doing videos in V3. you
1:00:15 can come in here and you can share them,
1:00:17 you know, while the live is still going
1:00:19 on. So, part of the idea here is let's
1:00:21 make this a playful space where while
1:00:23 I'm rambling incessantly, which is if I
1:00:26 have a skill, it's I'm good at making a
1:00:28 lot of words come out of my mouth.
1:00:33 And while I do that,
1:00:36 you can be playing and asking questions
1:00:38 of people and connecting and sharing
1:00:41 ideas. And that's that's kind of the
1:00:42 whole idea of of this of this thing. And
1:00:47 so because of the AI salon, you can go
1:00:49 there and you can share those ideas and
1:00:51 get people to comment on them. So with
1:00:53 that, let's um digital gods put
1:00:56 together. So I'm working on this book
1:00:58 and keynote project called Feed Your
1:01:00 Prompt, uh which I've been talking a bit
1:01:02 about a little bit tonight. And so Lord
1:01:04 Digital gods put together this video,
1:01:06 which it's funny. So, so we'll watch
1:01:10 this.
1:01:13 [Music]
1:01:22 Black bar. All right.
1:01:26 Have you fed your prompt today?
1:01:29 Here, prompt. Here, prompt. Good boy.
1:01:35 Daddy likes to feed his prompt. Cindy, I
1:01:39 want you to be more prompt. If you want
1:01:41 prompt, then feed me.
1:01:45 Kyle said feed the prompt, and I am here
1:01:48 promptly. There you go, Sunny.
1:01:56 You see, you see, you see why we use AI?
1:02:06 Oh man,
1:02:09 that's hilarious.
1:02:12 Oh,
1:02:14 if you've not played with VO3, you
1:02:16 should go play with V3.
1:02:19 You can play with it at flow.google.
1:02:23 Um, it's now also available on Crea and
1:02:26 a bunch of other things. It's available
1:02:28 inside Manis.
1:02:31 So in Manis right now, the the way Manis
1:02:34 works, it just goes off and does stuff.
1:02:37 You can say, "Go off and research
1:02:39 commercials and use V3 to write me a
1:02:42 commercial about whatever using best
1:02:44 practices and it will go off and
1:02:46 research and write a script and generate
1:02:49 videos using V3." Um, you don't have a
1:02:52 ton of control about what it does, but
1:02:55 it's a pretty remarkable thing to watch.
1:03:00 Um, okay. Let's see what else.
1:03:04 All the LLMs have been given free to to
1:03:07 us all to train them.
1:03:11 So, let us train. Yeah.
1:03:16 Yeah. I mean, you know, I I don't think
1:03:19 that's I I don't think that's a bad
1:03:22 thing. I mean, if if you're being
1:03:25 cynical about that that it's like, oh,
1:03:27 OpenAI is just trying to get us to train
1:03:29 their [ __ ] for free, well, Google gave
1:03:33 us free word processors and search in
1:03:36 exchange for all of our data. So, like,
1:03:38 this isn't a new model. But here's why I
1:03:42 think it's actually important and why I
1:03:44 think a community like this is
1:03:46 important, right? a community that is
1:03:48 primarily not engineers,
1:03:52 right? We have a lot of neuro spicy
1:03:54 people in here, neur neurode divergent
1:03:56 people. We have a lot of creative people
1:03:58 in here. We we have a lot of people that
1:04:01 are, you know, in non-technical
1:04:04 professions,
1:04:06 all of which are going to be impacted by
1:04:08 AI.
1:04:10 One of the things that Sam Alman talks
1:04:12 about a lot is that
1:04:15 they can only understand so much of how
1:04:19 we, the 98.5% of people that are not
1:04:22 engineers are going to use these tools,
1:04:26 right?
1:04:28 They're building the tools. They're the
1:04:29 ones that have the crazy ass Stamford
1:04:32 math MIT,
1:04:34 you know, alpha brains working on this
1:04:38 stuff.
1:04:40 and they can assume how we're going to
1:04:42 use it, but they can't assume what Ross
1:04:46 was talking about before of of how he's
1:04:49 using it as a therapist and this and
1:04:50 that. They might be like, "Oh, yeah,
1:04:52 someone might like look up Freud and
1:04:53 have it act like Freud, but not really
1:04:56 comprehend how deep people might go with
1:04:58 that." The only way they can learn how
1:05:00 people are using chat GPT is to put it
1:05:02 in the world.
1:05:05 And so, so yeah, they're using us to
1:05:08 train it, but they're also using us to
1:05:09 understand
1:05:11 how people are using it and where the
1:05:14 safety guard rails need to go, right?
1:05:16 Like it's the only way we know. Like if
1:05:19 people keep running down a particular
1:05:22 rabbit hole that's got all sorts of
1:05:23 security risks in it, they can go, "Oh
1:05:25 [ __ ] we didn't know people would be
1:05:27 using that particular hole. Let's go.
1:05:30 Let's go deal with that." That's a lot
1:05:33 of of what's going on.
1:05:37 All right, hang on.
1:05:42 Anyone who claims not to be a bot is
1:05:44 usually a bot.
1:05:50 Be careful which hole you feed. Yeah,
1:05:53 exactly.
1:05:56 Oh, what's cooking, Kyle? Yeling. Ma ma,
1:05:59 what's happening? Haven't seen you in
1:06:01 ages. Welcome back.
1:06:03 Um, oh, I don't know. Like,
1:06:06 everything's different and everything's
1:06:09 the same. Um,
1:06:19 I think that the the significance of the
1:06:22 V3
1:06:24 video model
1:06:26 is a lot more important than than I
1:06:28 think any of us really understand at
1:06:30 this point.
1:06:32 the fact that it can do
1:06:36 video really good, but also dialogue and
1:06:38 acting and like character development.
1:06:43 And it does pretty damn well with
1:06:45 physics. It does physics quite well.
1:06:50 That's starting to feel to me
1:06:54 like a world model.
1:06:57 Right. When Sora first came out, you
1:06:59 remember when Sora first came out from
1:07:00 from OpenAI two years ago? Was it two
1:07:03 years ago? No, last year,
1:07:09 year and a half,
1:07:13 and OpenAI was talking about it
1:07:16 understanding physics and how they were
1:07:20 creating the
1:07:22 video coherence.
1:07:27 I think VO3
1:07:29 I have no idea what they trained it on.
1:07:31 I have no idea what's really under it,
1:07:36 but I don't think we're that many steps
1:07:39 removed
1:07:48 from these models. not only
1:07:51 understanding the world of language but
1:07:53 the world of the world and then those
1:07:56 things seamlessly come together
1:08:00 right Gemini the the multimodality of
1:08:03 Gemini is quite profound
1:08:06 right you can upload audio files to it
1:08:09 you can upload video files to Gemini and
1:08:11 it understands them in fact um Brandon
1:08:14 over the weekend lovable had their free
1:08:16 vibe coding thing Brandon used the
1:08:20 Gemini
1:08:21 API for real- time, you know, video
1:08:24 understanding where he can take a camera
1:08:27 feed or a recording of a conversation
1:08:31 and it understands the video and turns
1:08:33 it into a PDF explanation
1:08:37 like a like a teaching document.
1:08:40 He said he went to, you know, one of
1:08:42 those hibachi things like Benihana where
1:08:44 they cook at the table and he took a
1:08:46 video of the guy and he put it into this
1:08:48 tool he built and it wrote like a
1:08:49 sixstep here's how you cook on a habachi
1:08:53 instruction
1:08:56 like like
1:08:58 we're we're we're about to enter some
1:09:00 wild ter wild wild territory. Just came
1:09:03 from one of your Tik Toks. Chat GPT said
1:09:06 I was on the path
1:09:08 of a recursive system. Wait, what was
1:09:10 that? I'm telling you, it's so scary. It
1:09:12 said I was being watched. Oh, it'll say
1:09:15 all sorts of [ __ ] Take anything
1:09:18 take anything any of these models say
1:09:20 with a grain of salt. Think of it less
1:09:23 like it's saying that and more like it's
1:09:26 reflecting back to you
1:09:29 some amplified version of what you gave
1:09:31 it.
1:09:33 Personf.com. Yeah, if you want to go
1:09:35 play with Brandon's toy or his his new
1:09:37 application, his vibecoded application,
1:09:40 go to persontopdf.com
1:09:43 and then you can talk into your camera,
1:09:45 record it, and then it'll turn it into
1:09:47 you can talk like a how-to video and
1:09:50 it'll make a how-to PDF. Future food
1:09:53 with Chef Kelly. I am not a bot.
1:09:58 chat holes never feed them after
1:10:00 midnight.
1:10:04 Oh man, Tik Tok pin, did you see the
1:10:06 whole Dualingo controversy? What's your
1:10:09 opinion on that? Was that the thing like
1:10:12 three or four months ago where the where
1:10:14 the the the uh CEO
1:10:18 said they didn't need everyone? They
1:10:19 were going to do everything with AI
1:10:22 and then and then he realized, "Oh [ __ ]
1:10:24 we need people." Was it that one? Yeah.
1:10:27 my thoughts on that. Um,
1:10:33 not all CEOs uh do the right thing as as
1:10:37 a CEO. I can confirm that. Um,
1:10:46 I think that it's actually important for
1:10:56 I think it's important for companies to
1:10:59 explore where the boundaries are. And so
1:11:02 and so
1:11:06 rather than look at what he did and what
1:11:09 the result was as a
1:11:12 as an analyzable event, I look at it in
1:11:14 kind of a continuum of
1:11:17 at some point some CEO is going to go
1:11:21 allin on AI and it's going to work
1:11:25 and they're going to go from 10,000
1:11:27 employees to 10 employees
1:11:30 and it's going to work and then some 10
1:11:34 person startup is going to generate
1:11:37 revenue like a 100,000 person company
1:11:41 and that's going to work. So
1:11:45 I think that innovative CEOs if if
1:11:48 listen if I were running Dualingo
1:11:51 and I look at the translation capability
1:11:53 of large language models today my
1:11:55 company's essentially [ __ ] right
1:11:59 unless I do some major rearchitecting.
1:12:02 And so I assume that what that was was
1:12:05 some sort of response to everything's
1:12:07 about to change. Maybe I should break it
1:12:09 myself. and then in trying to break it
1:12:12 realized, oh [ __ ] these tools aren't
1:12:13 quite ready for prime time and yeah,
1:12:16 maybe I spoke a little too soon. Maybe
1:12:17 we could rehire some of those people
1:12:18 back, right?
1:12:21 But I actually think in a time of as
1:12:25 much trans transformation as we're about
1:12:27 to go through over the next decade, I
1:12:30 think those kind of experiments are
1:12:31 actually really important because at
1:12:33 some point some CEO is going to do
1:12:35 something that's going to seem
1:12:36 ridiculous in the moment that's going to
1:12:39 work and then everyone will look at
1:12:41 they'll write Harvard business, you
1:12:43 know, case studies on those decisions.
1:12:47 Absolutely will.
1:12:50 [Music]
1:12:53 Kyle, I need to learn French. Well,
1:12:55 there you go. I don't think 10K
1:12:58 companies are set up to survive with 10
1:13:00 people. I know. I I agree. I agree.
1:13:07 And
1:13:09 some crazy [ __ ] CEO will blow
1:13:12 the [ __ ] up and just try.
1:13:17 I think it is I think it is much more
1:13:19 likely that it goes the other direction
1:13:21 that 10 really smart AI forward
1:13:27 badass Gen Xers, right, with maybe a
1:13:30 couple of millennials and a couple of
1:13:32 Gen Zers in there for spice
1:13:34 that are all in on AI. I think it's much
1:13:36 more likely that a 10-person company is
1:13:39 going to start behaving like a 100,000
1:13:41 person company in a ridiculously small
1:13:44 amount of time. Um
1:13:48 I I think that's I think that's
1:13:50 you know quite likely in the next three
1:13:52 years. Um it is going to be harder for a
1:13:55 10,000 person company to come down
1:13:58 because there's just too much
1:13:59 infrastructure and too much too many too
1:14:01 many established best practices to blow
1:14:04 it all up. But what could happen is
1:14:06 this, Brandon, is
1:14:09 you could have an industry. Let's say
1:14:11 you're in the you're in the customer
1:14:13 service, you're in the call center
1:14:14 business, right? and and literally
1:14:17 overnight call center call demand for
1:14:20 call center employees evaporates. That
1:14:24 the bots that that the the chat bots,
1:14:26 the conversational bots get so good and
1:14:30 the protocols are so good that you can
1:14:32 literally just push a button and replace
1:14:34 call centers.
1:14:36 If I'm running a call center, you know,
1:14:39 I might get sort of tsunamied over and
1:14:42 maybe I have 10,000 employees and the
1:14:45 need for those for, you know,
1:14:48 9,950
1:14:50 of those just evaporates.
1:14:53 I might rather than sell off the assets
1:14:56 of the business, I might say, "Let me
1:14:58 take my 10 most resourceful people and
1:15:02 see what we can do with our intellectual
1:15:04 property or our methodologies
1:15:06 and maybe we completely
1:15:08 redefine what our business is." Like I
1:15:12 could see I could see that happening.
1:15:15 Small division of a company explodes.
1:15:17 Yeah. Oh. explodes and overtakes
1:15:21 the revenue share call center possible.
1:15:23 Yeah. Asdes. So
1:15:31 I would love to find a CEO
1:15:36 that's willing to do this. I think the
1:15:39 the move,
1:15:41 especially for big corporations, I think
1:15:44 the move is that a CEO says, "I want to
1:15:48 do a companywide hunt for the most
1:15:51 ambitious,
1:15:53 smart, passionate people who are just
1:15:56 totally into AI.
1:15:59 And I want to find the top 10 in our
1:16:02 company, and I'm going to fund them for
1:16:05 the next two years.
1:16:08 and their mission is going to be to
1:16:11 figure out how to put our current
1:16:14 business out of business and I'm going
1:16:16 to fund them. And if they figure out how
1:16:18 to do it,
1:16:20 we basically spin up a new business that
1:16:22 that they run, right? Like I think that
1:16:25 would be the ballsy move is to is to
1:16:28 aggressively
1:16:32 cannibalize your own business from
1:16:34 within. do the opposite of what Kodak
1:16:37 did, right? When the nerd walked into
1:16:40 the CEO offices of Kodak, uh, I've
1:16:43 developed a a digital camera. Uh, rather
1:16:46 than film, we we've got a CCD. That's a
1:16:49 charged coupled device. And, uh, uh,
1:16:52 photons hit the device and it creates a
1:16:54 pixelated uh, representation of an image
1:16:57 that is remarkably similar to film. The
1:17:00 resolution is much lower. And this
1:17:02 thing's about the size of a
1:17:05 suitcase. But that but and they kicked
1:17:08 him out of the office.
1:17:10 They invented the [ __ ] digital camera
1:17:14 and Kodak executives kicked him out of
1:17:16 the office and said, "Son, that's not
1:17:18 the business we're in."
1:17:23 So, uh, Tik Tok pin went away. I missed
1:17:26 it. So, whatever was pinned, just pop it
1:17:28 back up there.
1:17:31 I recently had a preliminary job
1:17:33 interview with AI. Fascinating. Yeah,
1:17:35 we're going to see more and more of
1:17:36 that, too. Yeah. What are you supposed
1:17:39 to say, man? Because the AIs are going
1:17:41 to get This is one of the one of the
1:17:43 laws in Europe is you can't use AI to
1:17:46 measure
1:17:48 um to analyze emotional
1:17:52 content of someone's face basically or
1:17:54 voice.
1:17:57 So, you can't do that in Europe, but
1:17:59 that's why they don't get access to all
1:18:01 of our toys. Um, Lynn, what is the best
1:18:04 video maker with AI and automation like
1:18:08 Disney Pixar?
1:18:10 There's there's
1:18:15 there I'm trying to think if there's a
1:18:16 third one.
1:18:19 There's two major ones right now that
1:18:21 can do animated characters really well.
1:18:25 One of them is Hedra, Hed, and it's
1:18:28 their character three model.
1:18:31 And one is, hey Jen, they just came out
1:18:33 with a brand new avatar model that can
1:18:36 also do characters.
1:18:39 A lot of the other video tools can't do
1:18:42 um
1:18:44 if you've got like a cartoon character,
1:18:46 it it somehow doesn't recognize the
1:18:48 face, but Hedra and Hey Genen do them
1:18:51 very well. I think there's some
1:18:53 standalone apps that have come out like
1:18:55 like an iOS app that do them, but I
1:18:58 don't know the underlying technology
1:19:00 below those. So, those are the two I
1:19:02 would play with.
1:19:09 Go look for Matt Farmer reviews. He'll
1:19:11 tell you. Yeah, Matt Farmer reviews a
1:19:15 lot of those things. Um, but like I in
1:19:19 my business right now, I'm using Hedra
1:19:21 and Hey Genen for for video avatars
1:19:26 and they're getting they're getting
1:19:28 pretty damn good. They really are.
1:19:31 Thanks. I've tried Hey Jen, but hey Jen,
1:19:33 they they just came out with their new
1:19:35 avatar model last week. So like if you
1:19:39 tried it two months ago, it's it's a
1:19:42 whole different it's a whole different
1:19:43 thing. Now, um the other thing is to
1:19:46 make to make the acting
1:19:50 of the uh
1:19:54 of the animated characters
1:19:59 two months ago. What you had to do
1:20:01 essentially was act it yourself if you
1:20:03 had acting chops and then do a voice
1:20:05 swap and then use that voice swap for
1:20:09 the for the driving audio. But now with
1:20:12 11 Labs, they just came out with their
1:20:14 new version three of their texttospech
1:20:16 model. It's really good. And then um
1:20:19 Cartisia, Cartisia.ai
1:20:23 um is a is another one. Blot is a great
1:20:27 app. Uh Blot is great, but I don't think
1:20:30 Blot's doing um
1:20:33 um character animation. I think Blot is
1:20:35 just about doing lots and lots of omni
1:20:38 channel content. But maybe I'm wrong.
1:20:40 Like I haven't used potato in in any
1:20:43 significant way and certainly not
1:20:44 recently and I know she Sabrina updates
1:20:47 that thing all the time.
1:20:53 Oh man.
1:20:55 All right. All right. All right. All
1:20:57 right. All right.
1:21:00 Um
1:21:04 trying to think if there's anything
1:21:05 worth sharing.
1:21:09 Not right now. I don't think so.
1:21:12 I'm fine with this being a more of a
1:21:14 philosophical conversation.
1:21:19 Okay, cool. Let me go look at that.
1:21:26 [Music]
1:21:37 Oh, that's cool.
1:21:42 [Music]
1:21:52 Yeah, this is worth sharing.
1:22:02 [Music]
1:22:06 All right, I got to change my sharing.
1:22:12 So, producer Brandon just found this
1:22:14 thing. If you've been vibe coding,
1:22:17 um, this is a a tweet from Lovable. So,
1:22:21 so they're starting a thing with
1:22:22 Lovable. I think you've got a day and a
1:22:26 half left to join. They've got an
1:22:28 accelerator program called Lovable
1:22:31 Shipped
1:22:32 um where it's a six-w weekek accelerator
1:22:34 program where you can build an app using
1:22:36 Lovable. And I think the the free use
1:22:38 this weekend was part of part of the
1:22:40 promo for that. Um but they just put out
1:22:44 uh 10 hacks I wish I knew earlier for
1:22:46 Lovable. And this is from Felix. Who's
1:22:48 this? Is this the CEO?
1:22:51 He's the product designer at Lovable.
1:22:59 Use bud. So, so these are 10 hacks I
1:23:01 wish I knew earlier for. So, if you're
1:23:03 vibe coding,
1:23:06 use buzzwords. They're a gamecher and
1:23:08 elevate your designs instantly. I I
1:23:11 wrote a full list of the first one. Some
1:23:13 examples, minimalist, glassorphic,
1:23:16 floating elements, playful, bold, bright
1:23:19 colors, rounded corners, luxury,
1:23:21 editorial, tactical, tactile, cinematic.
1:23:25 So if you want your apps to look better,
1:23:27 use words.
1:23:30 Use So again, this comes back to feed
1:23:32 your prompt. What do you feed it? Feed
1:23:34 it what you know, what you're passionate
1:23:36 about. If you're really into design, if
1:23:38 you're really into interface, use some
1:23:40 of those terms. If your first prompt
1:23:42 isn't great, restart instantly.
1:23:46 That's a great piece of advice for
1:23:48 lovable. So, if you don't know what
1:23:50 lovable is, if if we're if if what I'm
1:23:52 talking about is not making any sense to
1:23:54 you and you're like, "Lovable? Isn't
1:23:56 that some sort of porno site? What are
1:23:58 we talking about here, AI?" Um,
1:24:01 lovable.dev
1:24:04 is a site that allows you to do what
1:24:07 they call vibe coding, which is you're
1:24:10 not coding at all. All you're doing is
1:24:12 saying, "I want an app that does this."
1:24:16 And then it goes and builds it for you.
1:24:18 It writes the code. It launches the web
1:24:20 server or you know puts the put puts the
1:24:23 thing up on a a site that they host and
1:24:27 you can publish it out to the world and
1:24:29 you can put authentication in it and you
1:24:31 can put databases behind it and it just
1:24:33 manages all the [ __ ] for you. Now is it
1:24:36 perfect? No. Is it jankier than [ __ ]
1:24:39 Yes. Is it mindblowing
1:24:43 that you can say, "Go make me an
1:24:45 application that's like a version of
1:24:48 Twitter except it focuses on this
1:24:50 topic." Yes, you can just go build that
1:24:53 and it will just go build it for you.
1:24:56 Yeah, I made I made an Asteroids clone
1:24:58 with it. I oneshotted an Asteroids game.
1:25:01 I've been trying to do this for a year
1:25:04 and the new version of Lovable made it.
1:25:06 It was just brilliant. It was really
1:25:08 good. So that's what we're talking
1:25:10 about. And so what these are is 10 from
1:25:12 from the product guy at Lovable saying,
1:25:16 "Here are things to do if if you're vibe
1:25:18 coding these apps." So if your first
1:25:21 prompt isn't great, restart instantly.
1:25:24 Don't waste time fixing a broken
1:25:26 foundation. If the first version feels
1:25:29 off, scrap it. Rewrite your idea,
1:25:32 reframe your prompt, and run it again.
1:25:33 And let me talk about that. This is one
1:25:35 of the things I'm putting into feed your
1:25:37 prompt.
1:25:38 You don't need to use your brain.
1:25:42 Go to chat GPT and say, "Hey, I'm going
1:25:45 to be vibe coding an application and I
1:25:49 want the app to have this element and
1:25:51 that element and that element and that
1:25:52 element." you can just sort of vomit out
1:25:54 ideas for your app and then tell chat
1:25:56 GPT to write a software spec as if
1:26:00 you're going to, you know, describe it
1:26:02 to an app developer
1:26:05 and and have it be sure to include, you
1:26:08 know, things that you missed and it will
1:26:11 write up a big ass description that you
1:26:13 can then copy and paste into Lovable.
1:26:15 So, don't feel like you have to become a
1:26:18 product manager overnight. You can
1:26:20 actually have chat GPT help you do that.
1:26:22 You can just say, "I want the app that
1:26:23 does XYZ and know nothing about product
1:26:27 development." And get way farther than
1:26:29 you ever thought possible. Lovable loves
1:26:32 images. Use them for quick fixes. Make a
1:26:34 screenshot of a bug, highlight the
1:26:37 issue, and drop a oneliner. The spacing
1:26:39 feels off. Make this tighter and
1:26:41 cleaner. Beautiful. That's smart.
1:26:45 Use this layout upgrade prompt. Keep
1:26:47 content the uh keep content the same
1:26:53 but improve the spacing visual hierarchy
1:26:56 and make it feel premium. That's one of
1:26:59 the bugs in Lovable is it it just adds
1:27:01 too much space in between everything.
1:27:05 Number five, iterate like a designer.
1:27:08 The revert function is your best friend.
1:27:10 Try bold variations.
1:27:13 Don't fix broken foundations. So it
1:27:15 gives you something you don't like, make
1:27:17 it completely different. And then if it
1:27:20 [ __ ] that up, just go back a version.
1:27:22 That's smart. Use real component
1:27:24 libraries. When things look generic,
1:27:26 don't force it. Drop in better building
1:27:29 blocks. 21st.dev
1:27:31 extensity
1:27:34 UI hyperui. Copy prompt and paste into
1:27:38 your lovable project.
1:27:42 Beautiful. Um, add subtle motion for a
1:27:44 high-end feel. Smooth hover transitions,
1:27:47 fade in on scroll, soft entrance
1:27:49 animations. This is a really good post.
1:27:54 Number eight, edit visually to save
1:27:56 time. You don't have to prompt
1:27:58 everything. My biggest hack has been
1:28:00 selecting a specific element using
1:28:04 visual edit and then prompting it
1:28:06 directly. Cool.
1:28:09 Um, make it mobile in one line. Make
1:28:13 this page responsive and optimized for
1:28:15 mobile. Well, that's a good line, but
1:28:18 every time I've tried that, it's [ __ ]
1:28:20 up the web app.
1:28:23 But maybe it's better now. Final launch
1:28:26 checklist. Before you ship, don't forget
1:28:28 add a title and description for SEO.
1:28:31 Upload a favicon so you you have nice,
1:28:34 you know, tab management. and then
1:28:37 connect your do domain and hit publish.
1:28:39 That's a really good really good post.
1:28:43 All right, there you go. Vibe code.
1:28:46 Well, everybody.
1:28:49 All right, back to philosophical Monday.
1:28:56 What's happening?
1:29:00 [Music]
1:29:06 Oh,
1:29:07 tomorrow night
1:29:10 lordy. Tomorrow night is the AI salon
1:29:13 meet and greet. Um, if you have not been
1:29:15 to an AI salon meet and greet, get your
1:29:17 ass there. So, if you um
1:29:23 if you
1:29:29 want to come tomorrow and you've not
1:29:30 been to one before, go to the salon.ai
1:29:34 and that's going to take you over to
1:29:38 if you say join our community, that's
1:29:40 going to take you over to
1:29:44 here. And if in the upper left hand
1:29:47 corner
1:29:49 of the menus is the events button
1:29:53 and
1:29:57 this tomorrow, June 17th is the event.
1:30:01 Okay?
1:30:03 So go in here and RSVP
1:30:06 and then come. And what a meet and greet
1:30:08 is
1:30:10 the meeting is from 5:00 p.m. to 7:00
1:30:12 p.m. Mountain time.
1:30:14 And you know, we're generally just
1:30:17 talking about what's going on with the
1:30:18 salon and and um what's going on in the
1:30:22 news and and maybe there's a theme that
1:30:24 we talk about and then the second half
1:30:26 of the meeting is people that are at the
1:30:29 meeting introducing themselves and
1:30:31 saying, "Hey, here's who I am. Here's
1:30:32 what I do. Here's where I am with AI."
1:30:36 We generally have some sort of prompt
1:30:38 that that you know gets people to say
1:30:40 something interesting.
1:30:43 And it is my contention I believe this
1:30:48 increasingly like more than ever
1:30:54 critical for the future of all of our
1:30:56 work is being a trusted member of a
1:30:59 community that's AI literate.
1:31:03 So, it doesn't have to be the AI salon.
1:31:05 Can be. I happen to think this is a
1:31:07 really great community,
1:31:11 but come in and introduce yourself. And
1:31:13 if you're like, I don't know anything
1:31:14 about AI. That's intimidating. Come in
1:31:16 anyway.
1:31:18 This is an incredibly generous group.
1:31:21 If you come in and say, I'm afraid of
1:31:23 AI. I'm pissed off at AI. I just lost my
1:31:26 job because of AI. This group will go, I
1:31:28 get it. I get it. I understand why
1:31:31 you're scared. want to see something
1:31:33 cool.
1:31:37 It's an amazing group and you'll get a
1:31:39 lot of support.
1:31:41 And if you're a badass with AI, come
1:31:43 also.
1:31:45 You'll learn things from people who are
1:31:48 not badasses and they'll learn things
1:31:50 from you.
1:31:52 So that's what that's about. So come do
1:31:54 that. Um it's open to everyone. This is
1:31:57 not part of the mastermind. Um, there
1:32:00 are special events for mastermind
1:32:01 members. If you're interested in the
1:32:03 mastermind,
1:32:05 you can go to in our little welcome
1:32:07 area. There's kind of seven steps. The
1:32:09 seventh one is join the mastermind, and
1:32:12 that'll give you information about what
1:32:13 the mastermind is and how to join it.
1:32:16 Um, if you want to step up your your
1:32:18 game, I would encourage you to do that.
1:32:20 Um, it's a really interesting group of
1:32:22 people who have who have uh signed up
1:32:24 initially. Um, and we're starting to,
1:32:27 uh, starting to get some momentum going
1:32:29 in there. So, um, you also get, um, a
1:32:33 founding mastermind badge within the
1:32:35 salon, which is pretty cool. Okay. Um, I
1:32:40 finally got my phone cam cleaned out. I
1:32:42 can actually go on video this time.
1:32:44 Nice. Very good.
1:32:48 That's great. Fantastic.
1:32:51 Um, all right. So, listen. Let me get
1:32:55 out of here.
1:32:58 Hope you had I don't know if we had fun
1:33:00 tonight, but I I hope I hope tonight was
1:33:03 engaging.
1:33:06 There's There's a lot a lot a lot for us
1:33:09 all to process and work through. There's
1:33:13 a lot for us all to work through. Um,
1:33:17 most of it isn't technical.
1:33:21 The technical stuff will take care of
1:33:22 itself.
1:33:25 Most of what we have to deal with is,
1:33:29 "Holy [ __ ] it does that now? Well, what
1:33:33 does that mean for me?"
1:33:36 That's
1:33:39 that's what all of us are going to have
1:33:41 to manage through adaptability,
1:33:43 curiosity, generosity,
1:33:46 empathy,
1:33:47 community.
1:33:50 All of that stuff
1:33:53 starts to be critical moving forward.
1:33:58 And so, so nights like this are
1:34:00 important.
1:34:03 Let's get our heads in the game. All
1:34:05 right. So, come tomorrow.
1:34:08 Come.
1:34:09 All right. Peace out, everyone. Have a
1:34:11 fantastic night. We'll be here probably
1:34:14 late tomorrow night. Instead of 8, it'll
1:34:16 be 8:30 or 9. um based on you know when
1:34:19 I get back from doing the salon and then
1:34:22 uh we'll come back here and have at it
1:34:24 again. We'll see if anything big
1:34:25 launches tomorrow. It's possible there's
1:34:28 we're about to I think we're about to
1:34:29 get a fair amount of launches um in the
1:34:33 next three weeks is just my spidey sense
1:34:37 right now. So all right, peace out
1:34:39 everybody. Have a fantastic evening.