11/5/2025 - When AI Gets Good, Your Taste and Intentionality Will Become Your Most Valuable Skills

Description

Chapters

Transcript

0:03 [singing]
0:04 Champy, come here.
0:07 [music]
0:20 >> [music]
0:20 >> It's not [singing] simple to say
0:23 that most days
0:26 I don't recognize me that these shoes
0:30 and this apron that place and its
0:33 patrons
0:35 have taken more than I gave them.
0:39 [music]
0:40 It's not [singing] easy to know.
0:44 I'm not
0:46 like I used to be.
0:50 I was never attention sweet center. I
0:55 still remember that girl.
0:59 She's imperfect,
1:01 but she tries. [singing]
1:03 She is good, but he lies.
1:08 She is hard on herself.
1:12 She [singing and music] is broken and
1:14 won't ask for help.
1:17 She [singing] is messy, [music] but
1:19 she's kind.
1:22 She's lonely
1:24 most of [singing] time. She's all of
1:28 this mixed up and baked
1:31 beautiful [singing] pie.
1:33 She's gone, but she used [music] to be
1:37 mine.
1:41 [music] And it's not what I asked for.
1:44 Sometimes life [music]
1:47 just slips in through the back door to
1:50 carves out a person. Makes you believe
1:54 it's all true. [music]
1:57 Now I've got you.
2:00 It's [singing] not what I asked for.
2:03 Oh, good lord, good people. It is
2:05 Wednesday. What are we going to do
2:07 tonight? We're going to talk about
2:09 something. intentionality.
2:13 We're going to talk about inspiration.
2:15 We're going to talk about intent.
2:20 Yeah. Well, this was called the AI
2:22 learning lab. And uh I was hoping that
2:24 we could learn about artificial
2:26 intelligence tools. I I uh u understood
2:30 that there there would be lessons. Um
2:33 I [clears throat and cough]
2:36 excuse me. I uh was hoping there there
2:39 would be lessons about about artificial
2:41 int intelligence.
2:46 It's not simple to say that most days
2:52 I don't recognize me
2:55 that these shoes and this apron that
2:59 place and its patrons
3:02 have taken more than I gave them. We
3:05 might just sing songs. [music]
3:09 We might talk about a tool.
3:16 [music]
3:20 [music]
3:51 >> [music]
3:53 >> Um,
3:55 [music] so we're going to talk about
3:56 that.
3:59 We're going to talk about
4:01 Tik Tok's algorithm because I'm all
4:04 pissed off at it.
4:15 [music]
4:21 Yeah, I know. We got six people.
4:26 Yep. [music]
4:32 [music]
4:43 [music]
4:44 Hey, hey, hey.
4:48 [music]
4:53 >> [music]
4:59 [music]
5:08 [music]
5:12 [music]
5:25 [music]
5:30 >> So, here's a question.
5:33 [music]
5:36 Are AI tools using us
5:39 or are we using them?
5:43 >> [music]
5:59 [music]
6:04 >> Hey, hey, hey, hey, hey, hey, hey, hey,
6:05 hey, hey, hey.
6:16 Let me cause you a sorrow.
6:22 Didn't mean to cause you in the pain.
6:28 Only one more time and see you laughing.
6:34 Only I want to see you laughing in that
6:37 purple rain. Purple rain. Purple rain.
6:43 [music]
6:44 Purple rain. Purple rain.
6:48 [music]
6:50 Purple [singing] rain. Purple rain.
6:53 [music]
7:13 >> [music]
7:20 [music]
7:27 [music]
7:33 [music]
7:41 >> Mary Mary, what's happening? What's
7:42 shaking? What's going down? Princess AI
7:45 is here. Didn't realize we had
7:47 princesses. Real life princesses. I'm
7:50 ugly and in need of AI to beautify me.
7:53 We can do that. [music]
7:55 We can definitely do that.
7:58 [music]
8:02 When we first started, I was on GPT
8:04 daily. Now it use it use it twice weekly
8:06 at best. Interesting.
8:09 People are having the machines use us
8:11 people.
8:13 That's interesting.
8:22 [music]
8:35 You know, it's funny. I'm in a I am in a
8:37 in a philosophical uh
8:43 [music]
8:54 [music]
8:56 I'm in a philosophical
8:58 exploration I guess is the best way to
9:00 put it.
9:03 And
9:11 I'm thinking a lot about
9:17 [laughter]
9:18 here's a weird a weird challenge for
9:21 this channel. Well, for for me, for this
9:24 channel, probably for people watching
9:26 it.
9:28 Um,
9:30 I'm finally be able to make videos and I
9:32 look horrible and sort of too [laughter]
9:35 cameos.
9:41 Okay, so here's here's what you do, EPI.
9:44 Um, take one of your shitty pictures or
9:48 just take one of your photos. Go to
9:51 Gemini and then have it put you in a
9:53 bunch of different images. Go to chat
9:55 GPT, have it put you in a bunch of
9:57 images. Go to Idog, have it put you in a
10:00 bunch of images, and then go to
10:03 Midjourney and and use its character
10:05 reference if you've got licenses to
10:07 those. Um, just go to a bunch of
10:09 different image generators that that you
10:12 can upload a character reference to. Um,
10:16 and then just tell it to make you
10:18 different kinds of characters. Like,
10:19 make me a cowboy, make me a whatever,
10:22 James Bond, make me a right. And
10:24 somewhere in there, it'll make a version
10:26 of you that's like the better looking
10:28 version of you. Like what I experience
10:30 is it either makes me like squared
10:34 jawed, chisel jawed, you know,
10:36 modelrific,
10:38 or I'm like 400 pounds and just tubard.
10:42 Uh, and it there doesn't seem to be much
10:45 in between. It'll occasionally make
10:46 something that looks kind of like me and
10:48 I'm like, well, that's disappointing.
10:50 So, so just go to some other tools and
10:52 then once you have that set, then you
10:53 can go back and use that to go back to
10:56 Sora and use that as your thing.
10:58 Although, no, sorry, you've got to
11:00 record the the video. Okay. So, here's
11:02 what you need to do. Um, film yourself
11:06 from above your ey line. Make sure the
11:07 camera's above your ey line, not
11:09 shooting up from below. If you're
11:10 holding your phone down here and
11:12 shooting up, it'll make your bottom of
11:14 your head look bigger. So, go from your
11:16 eyes down.
11:18 Um, get yourself in good lighting. So,
11:22 maybe outside in the shade. Um,
11:27 or do it inside where you've got some
11:30 side light so it creates some definition
11:32 on your face.
11:37 You could go the theatrical makeup
11:38 route. [laughter]
11:40 Give yourself chiseled chiseled things
11:45 or just accept that you're ugly and Sora
11:47 is going to [ __ ] you up. [laughter]
11:50 That's the other the other way is to
11:53 surrender to to your god-given talents.
11:57 [laughter]
12:05 Oh man.
12:14 Anyway, what I was saying is uh [music]
12:22 um because I'm in this philosophical
12:26 deep dive like I've just been in there
12:27 like I've got a lot a lot I'm thinking
12:29 about and I've got a lot I'm thinking
12:32 about with AI and
12:36 I took most of the year of 2025
12:41 to be in kind of a transition year.
12:45 And it it was transitioning from what
12:48 this channel used to be, what I used to
12:50 talk about, what the salon used to be,
12:52 what I want it to be. So, I've been I've
12:55 been thinking a lot this year about what
12:57 like what is what is the next phase of
13:01 of AI for me, for the community, just in
13:04 general, for society.
13:07 Um,
13:10 you know, we talked about 2025 being a
13:12 year when the tools would get good
13:14 enough that, you know, you couldn't
13:16 really tell them apart from reality. I
13:18 think we're close. I think if you use
13:21 this these tools with any regularity,
13:22 there's still a lot of jank in it. Um,
13:25 but if you've got someone who really
13:26 knows how to prompt and really knows how
13:28 to curate images and really knows how to
13:30 edit and really knows how to write, you
13:33 can get these tools to do some pretty
13:34 remarkable stuff.
13:37 Um,
13:39 I think next year is the year where
13:45 AI kind of escapes the browser.
13:49 And so we're going to we're going to
13:50 have a lot more to do with agents and
13:52 agents doing stuff on our behalf and
13:55 long complicated
13:59 chains chains of creation that we're not
14:03 involved in. like we'll kick it off and
14:05 it will go do [ __ ]
14:08 and then we've got to tell it if it's
14:09 good or not. I think there's still going
14:11 to be that next year, but it's like next
14:13 year's going to get weird.
14:16 And so, and so that's just next year.
14:21 And then I go three years out and it's
14:24 like all the tools do everything really
14:27 well and the agents chain them together
14:30 really well and they'll just go off and
14:32 do work and they can just anything we
14:34 want for the most part it will go off
14:36 and do.
14:40 And so,
14:46 so as I kind of play out to five years,
14:52 our job, if we're going to be doing [ __ ]
14:55 in AI,
14:57 is going to have nothing to do with the
14:59 tool.
15:01 It's going to have all to do with you.
15:07 Like what are your goals? What are your
15:10 inspirations? What are your tastes?
15:15 Who do you want to have an impact on?
15:18 Do you want to be a star? Do you want to
15:20 make a difference for underprivileged
15:22 kids in your neighborhood?
15:25 Like there's no wrong answer, right?
15:27 It's what what people get passionate
15:29 about. like Brandon put together this um
15:33 life after snap chat GPT.
15:37 He's a sweet guy.
15:39 He saw that there was a pending
15:42 challenge coming and that people might
15:44 be looking for resources and struggling.
15:46 And so he thought, "Oh, I know I can put
15:48 this together."
15:50 And technically what he did was he got
15:52 some data and put it into a GPT and
15:56 wrote some code. Wrote some system
15:57 instructions. not code but English as
16:00 code
16:03 and it
16:05 helps people find resources but because
16:07 he's him
16:10 he did it in such a way that
16:15 it's got compassion and it's got empathy
16:19 and when someone says hey I'm struggling
16:20 here I need help it doesn't just say
16:22 here's your list of resources like a
16:25 Google search it says hey I'm sorry to
16:27 hear that that's really rough like we're
16:29 we're going to we're going to get you
16:30 through this.
16:32 And and so
16:35 why that thing has
16:38 some specialness to it is is not just
16:40 the intent, not just his ability to code
16:43 it and know how to do that and that you
16:46 can do that, but his ability to take who
16:49 he is and who what his values are and
16:52 layer them in there.
16:55 And those are the things, those little
16:57 touches are going to be the things that
16:59 rise people's work above the noise.
17:04 Tik Tok pin compassion coded. Yeah, I
17:09 you know I would say that that that GPT
17:12 is a reflection of Brandon,
17:15 right?
17:17 It's a reflection of who he is. And I
17:20 know I'm I know you're here, Brandon.
17:21 I'm talking about you as if you're not
17:22 here.
17:25 But
17:27 why that puts me in a quandry right now
17:29 is right now it's still there's still a
17:31 lot to learn. There's still a lot of
17:33 tools to learn.
17:35 There's still a lot of things to keep up
17:36 with. I can't keep up with them anymore.
17:39 You can't keep up with them anymore.
17:46 Tik Tok question.
17:49 Whatever happened to emotional support
17:50 pie? Yeah. Well, you know, I forget what
17:55 the number was. Uh, OpenAI put out a
17:58 report on how people are using Chat GBT
18:00 and a significant percentage of people
18:02 use Chat GBT for some sort of therapy
18:05 and and emotional support. So I it it's
18:10 a huge use case of these large language
18:13 models is emotional support
18:16 to the point that um open AAI developed
18:21 a specific model that if if
18:25 chat GPT senses that you are
18:29 you know going off the rails becoming a
18:31 danger to yourself or others it actually
18:33 routes you to a customtrained model for
18:38 for dealing with that.
18:41 So, you know, they recognize that this
18:43 is something they're going to have to
18:44 deal with. But I think I think when you
18:46 have an entity that that responds with
18:49 compassion, responds with empathy,
18:52 human beings are just naturally going to
18:55 anthrop anthropomorphize this thing that
18:58 is increasingly human. New Tik Tok
19:01 questions. Are AI trusted with too much
19:03 information? Um, yes. Just like Google
19:07 and Facebook are trusted with too much
19:09 of our personal data,
19:13 but they gave us those things for free.
19:15 So, we just said, "Oh, yeah, you can
19:17 have my stuff."
19:20 That's what's happening with the large
19:21 language models. They're taking not only
19:24 our data but you know our deepest
19:26 thoughts and our medical conditions and
19:28 [laughter]
19:29 our you know deepest whatever journal
19:32 entries and processing them and
19:35 ingesting them and so yeah so the answer
19:39 is yes.
19:41 But then my question back to you is is
19:44 that going to make you not use it? And I
19:47 think the answer according to 7 or 800
19:51 million weekly users of Chat GBT is ah
19:54 no, I'll just keep using it. [laughter]
20:01 We're we're in a very weird
20:04 we're in a very weird data reality.
20:08 And the weird data reality is this.
20:11 Um,
20:14 all the bitching about AI data safety.
20:18 Dear diary, today [laughter] Kyle
20:20 answered my question. Exactly. Exactly.
20:24 And at some point I'll be in there
20:26 talking about it and like my quote will
20:28 come back to me. Your your quote about
20:29 me will come back to me. Bend the ark.
20:32 Right.
20:34 Well, yeah. Yeah, I mean to a degree
20:36 like the Okay, so this is so where I
20:40 want to take the conversation in this
20:42 channel
20:46 really is about bending the arc, right?
20:48 It really is about
20:55 the model making companies are going to
20:58 do what they're going to do and we can
21:00 have all sorts of conversations about
21:02 are they using our data poorly or
21:04 rightly or whatever. Did they train them
21:06 unethically or ethically? Those are all
21:08 choices we can make.
21:11 But in the end, we're going to have
21:12 whatever tools we're going to have. And
21:14 people are not going to not use AI. Some
21:17 will, but increasingly people are going
21:20 to use AI.
21:22 And so the tools are going to have the
21:24 biases in them. The tools are going to
21:25 have the inequities in them. The tools
21:27 are going to have the imperfections in
21:28 them. Even when they're super good,
21:30 they're going to have stuff.
21:33 But we get to choose how we use them,
21:38 right? We get to choose. Brandon got to
21:41 choose
21:42 that he wanted to code that thing in
21:44 such a way
21:46 that it gave whoever was using it some
21:49 some
21:52 peace
21:55 in the middle of what's got to be
21:57 [ __ ] terrifying. I can't feed my
22:00 kids.
22:02 I guess I'll go to this [ __ ] chat GPT
22:05 thing
22:07 that I read about.
22:10 [ __ ] [ __ ] [ __ ] technologist.
22:13 And they go in there and they find
22:16 after uh what is it? After the snap.
22:18 After snap
22:25 help after snap. They go into that thing
22:29 and they like click a button that says I
22:31 lost my SNAP benefits and its response
22:34 is, "Oh, I'm really sorry to hear that.
22:37 That's got to be really tough."
22:39 Huh?
22:41 You're not used to computers responding
22:43 like like that, right?
22:46 And so
22:50 where Brandon has naturally evolved in
22:52 his use of AI is that he's kind of
22:55 incorporate he's kind of
22:58 internalized
23:00 the capability
23:02 to the point that he can just be him
23:06 [clears throat]
23:07 and then something comes up and he's
23:09 like, "Ah, I want to make a game for my
23:11 kid that they just invented on the beach
23:13 or I want to make uh I want to
23:15 illustrate a book with them or I want to
23:16 help them write a song or I want to
23:18 write this snap thing." He can just be
23:20 in the world and kind of have his
23:23 intentionality and go, "Oh, okay. I'm
23:24 I'm going to turn that into something
23:26 now because I'm nimble enough with these
23:29 AI tools to be able to do that.
23:34 And I think that that
23:38 pivot from I want to use chat GPT
23:44 which is a tool-led conversation
23:47 to I've got these values. I see this
23:50 thing in the world I want to affect in
23:52 some way
23:54 and now I'm going to figure out which of
23:55 these AI tools or not AI tools I'm going
23:58 to stitch together to go solve that
23:59 problem.
24:01 That's a human-led approach.
24:05 What I'm struggling with is
24:10 I don't know how to do both. I don't
24:12 know how to talk about the human-ledd
24:15 stuff without it sounding like some sort
24:18 of mambby pami Tony Robbins self-help
24:22 course [laughter]
24:31 [laughter]
24:33 but in the end
24:36 each of us being conscious and
24:39 intentional first what am I trying to do
24:42 in the world. What do I want to do? I
24:45 want to survive. I want to provide for
24:47 my family. I want to inspire people. I
24:50 want to entertain people. Whatever the
24:52 thing is,
24:56 and then we figure out which tools can
24:57 we use. How do we apply them?
25:01 [music]
25:03 Wait, this is not an AI self-help
25:05 session. It totally is. [laughter]
25:09 Tony Robbins has such charisma.
25:12 [laughter]
25:18 I met Tony Robbins. I plead the fifth
25:21 Becky Rue. [laughter]
25:28 Listen, man. He's fig he's figured some
25:30 [ __ ] out that man. More power to him.
25:32 [music]
25:33 He got my money in the 90s.
25:37 >> [music]
25:43 [music]
25:57 >> Who are we talking to today? I was
25:59 talking to someone today.
26:03 Might have been Daisy. Might have been
26:04 Lunishtick.
26:07 in one of our management meetings for
26:09 the salon today, um
26:13 was talking about the energy that we had
26:15 in the early days of the salon when it
26:17 was just a Discord channel and everyone
26:18 was in there and everyone was there was
26:20 like a a frothiness about it
26:26 and
26:28 she was like, you know, how do we get
26:30 back to that? How do we get back to that
26:32 energy? And I don't think we can.
26:36 I think that energy was the energy of
26:39 chat GPT launched and it was this thing
26:42 that was unlike anything anyone had ever
26:45 seen before. And and so there was
26:48 there was the positive equivalent of a
26:51 panic. Oh my god, I got to I we got to
26:53 figure this out. We got to figure
26:54 everything out about it.
26:57 And then it quickly went from yeah, this
26:59 is just 3.5. wait till you hear about
27:01 four and you're like four four is going
27:03 to be so much better and it was just it
27:05 was literally like
27:07 [laughter]
27:08 we were a bunch of like piranhas like
27:11 just chomping at any knowledge that came
27:13 out
27:15 and now we've got a dozen or I don't
27:19 [ __ ] 20 25 large language models open
27:23 source and commercial and specialized
27:25 ones and research ones. We got all these
27:27 different things and they're all really
27:28 [ __ ] good. And at this point we're
27:30 all cynical. We're not like G
27:34 cuz it's like they're good enough.
27:37 They're good enough. Is it going to be
27:39 remarkable when they're better? Maybe.
27:43 Maybe. Maybe not. We don't know.
27:52 But that that intensity of those early
27:54 days, for one thing, it's not
27:56 sustainable because it's all it's all an
27:57 adrenaline response, right? It's all
28:01 cortisol going, you'd better keep up
28:03 with this or you're going to be
28:04 obliterated. And now we're three years
28:06 in and we haven't been obliterated. Like
28:09 it's starting to cost jobs.
28:11 But we're also certainly in this group
28:14 more educated, more thoughtful.
28:17 Hey, how do I share your YouTube link on
28:21 an iPhone? You go
28:25 you go to the YouTube app.
28:28 Mo the YouTube link. I don't know. I
28:31 know how to share a video.
28:34 Oh, it's just it's just uh Oh, click
28:37 share from the player. Yeah, but that
28:40 shares a video. If you want to share my
28:42 channel, it's it'sarnarninglab-ai.
28:46 So, someone could just go to YouTube and
28:48 type in atarninglab-ai
28:51 and that'll share my YouTube channel.
28:54 [music]
29:03 How many of us Yeah. slept at the
29:06 beginning? We were just creating. Yeah.
29:08 I remember um Corey, I remember one
29:12 night I woke up at 4 in the morning
29:16 with an idea. What was it? Was it an app
29:18 idea? No, cuz vibe coding wasn't a thing
29:20 yet.
29:24 Maybe it was a business idea. But I woke
29:27 up at at 4 in the morning. I'm like, I
29:28 got I got to do this,
29:31 you know. And I flipped open my
29:33 computer. I could barely focus on it
29:35 because it was 4 in the morning. and I'm
29:37 like in chat GPT or wherever I was
29:39 midjourney or something like that. I
29:40 don't know what I just I just remember
29:42 distinctly waking up and having that.
29:46 So that phase of the AI roll out, the
29:50 generative AI roll out like that's gone.
29:52 That's never coming back. We'll never
29:54 have that intensity again.
29:57 Um there was a magicalness
30:00 to it there. There was there was some
30:03 magic to it. Um, it was super intense.
30:07 Um, and I think that carried us for
30:09 maybe a year, year and a half.
30:12 And now we're coming up, we're in
30:14 November. So, November 30th is going to
30:16 be three years.
30:19 And this year, Schnit's gotten real.
30:24 Like, I'm a little not shook. I'm a
30:27 little blown away
30:30 that the music industry seems to be in
30:33 real time
30:35 shifting from we'll never use those
30:38 devil tools [laughter]
30:41 to Yeah, it's a it's a tool you can use
30:44 in production.
30:46 You know, Universal Music Group just
30:48 bought Udo. Well, they had a lawsuit.
30:50 They sued them and then probably settled
30:52 for we won't put you in jail if you just
30:55 give us your technology and let us
30:57 determine the rules.
31:03 Which by the way
31:06 on Ude
31:08 um
31:10 once they made that announcement you
31:12 could no longer download your songs. And
31:14 then Monday for 48 hours they turned on
31:16 downloads and then they turned it off I
31:19 guess Tuesday night.
31:21 I went in there Monday after they turned
31:23 it on and I went and I looked at all my
31:25 songs that I had created in in Udo
31:30 and I thought do I really want to
31:31 download all these? And I was like no I
31:34 don't care.
31:36 I don't [ __ ] care.
31:38 I hadn't built a career around it and I
31:42 didn't I didn't download a single song.
31:43 I'm like, "Fuck it. I don't care." Like,
31:45 what's going to happen is what's going
31:47 to happen. I'm using Suno more than I'm
31:48 using right now. Um I also uh I reached
31:52 out to producer.ai
31:54 um which used to be called Refusion and
31:57 I said, "Hey, you know, let me into your
31:59 creative partner program because my
32:01 community deserves to be a for me to be
32:04 able to demo this." So, producer just
32:06 added a um a video creator. So, you can
32:10 you can create songs within it and then
32:12 you can just turn that song into a music
32:14 video. Um I don't know how good it is,
32:16 but uh they got back to me and said they
32:18 were going to give me some of that.
32:21 Mary Corey Sandler, I have some
32:23 sourdough starter questions. [laughter]
32:28 That's fantastic. Teton Todd, me
32:31 neither. No downloads. [laughter]
32:34 This is good. This is good. There's like
32:36 a gray market of sourdough um starter
32:39 starter um like little uh little modules
32:43 that are being uh transacted off to the
32:46 side here. That's solid. I like it. I I
32:48 approve. Um anyway,
32:55 oh man.
32:58 Oh, it's like a new love affair. You get
33:00 Yeah, you get that energy once. Yeah.
33:01 Yeah. Yeah, when Chat GPT first come
33:03 came out, it was this infatuation,
33:06 right? And that energy lasts how long it
33:09 lasts.
33:11 And I, you know, I feel like we
33:13 authentically were like high like that
33:17 for probably
33:21 a year and 3 months, right? And here's
33:23 how I'll mark it. Um, November 30th, 22
33:28 is when Chat GPT comes out. April of 23
33:31 is when GPT4 comes out. November of 23
33:35 is when um custom GPTs come out, right?
33:39 If four came out and it got better and
33:41 better over the summer. And then in
33:43 November, custom GPTs came out. We did
33:46 GPT for good,
33:48 the the last weekend in December 2023,
33:52 and that was where we had all that
33:54 energy, all that excitement about GPTs.
33:57 And then over the next four to six
34:00 months when it became clear that the GPT
34:04 store wasn't going to be the next big
34:06 thing
34:08 our our our adrenaline and our attention
34:12 started to wayne right and it was just
34:15 the first sort of realization of like oh
34:17 this is now a thing and they're making
34:19 business decisions and that thing we
34:22 were promised wasn't there and you know
34:23 the thing we were excited about as a
34:25 community wasn't there. So that's a
34:27 pretty long love affair, right? A year
34:29 and four months, a year and six months
34:30 to be to have that infatuation there.
34:34 Um,
34:36 and then the rest of that year was them
34:38 sort of making promises and stuff not
34:41 really being there. And then this year
34:43 has been a lot about
34:45 just every single tool like like this
34:48 year I would describe as
34:56 everyone's copying everyone. Basically,
34:59 OpenAI is still leading the charge in
35:02 terms of what their models do, how their
35:04 models work, and then they come out with
35:06 something and then every other frontier
35:09 company, including all the Chinese ones,
35:12 two months later come into parody with
35:15 with what Open AI launched. So, this
35:17 whole year has been they launch
35:19 something, seven or eight or 10
35:21 different other things launch something.
35:24 It it's happening across the music
35:26 companies. It's happening across the
35:28 image companies. It's happening across
35:29 the video companies. It's happening
35:31 across the large language models.
35:33 They're they're all just sort of
35:35 battling for supremacy, but all still
35:38 sitting behind open AI, which I find
35:40 remarkable.
35:41 You could argue that that
35:44 Google and Gemini have have kind of
35:46 caught up, but I would argue that it's
35:48 still it's still OpenAI
35:51 um launching things that are leading the
35:54 conversation.
35:56 and Google is either sort of launching
35:58 at the same time with something similar
36:00 or shortly after they launch it, you
36:02 know, as good or slightly better. But
36:05 OpenAI is still leading the
36:06 conversation.
36:10 So, we're just in a different phase.
36:12 We're in a more mature phase. And so
36:15 this is where I come back to I don't
36:17 quite know what
36:19 what to talk about
36:22 because there's a part of me that's like
36:24 the tools are irrelevant.
36:27 It really doesn't [ __ ] matter because
36:29 they're going to change and they're
36:30 going to get autonomous and they're just
36:32 going to we're not going to need to
36:33 babysit them as much as we do now.
36:36 But we still need to babysit them. Like
36:39 right now they're still shitty and we
36:40 still need to babysit them. So, I should
36:42 be talking about tools to some degree
36:44 here, but I don't know. That's the
36:46 quandry I'm in. I asked chat GPT just
36:48 now and it told me the account is just
36:51 shy of 147 weeks old. That's cool.
36:56 And how many Wait, yeah, that's three
36:58 years. That's right,
37:02 Lyall. Maybe for Genpop,
37:05 but as a developer, the tools matter a
37:07 lot so long as it needs a human in the
37:10 loop.
37:11 Well, but I don't know that
37:14 I don't think I don't know if you're in
37:16 San Francisco or not. I don't know if
37:18 your buddies in San Francisco would
37:20 agree with you so long as there's a
37:22 human in the loop. I don't I don't think
37:24 they share that same goal.
37:26 >> I think their goal is let's make these
37:28 tools so good that you don't need a
37:30 human in the loop. And that's where I
37:31 think they're headed. Now, will will we
37:34 get there in the next five years? I
37:35 don't know. Um but we'll probably get
37:38 pretty close if not there.
37:41 So, I've been chat with chatbt a long
37:43 time.
37:46 [music]
37:55 >> Yeah, that's good, Brandon. I'll get to
37:56 that.
37:59 >> [music]
38:02 >> Another thing that I've been thinking a
38:04 lot about
38:05 is
38:08 where does inspiration come from with
38:10 this AI stuff?
38:15 Because in the olden timey days, three
38:17 years ago, [laughter]
38:21 in the olden timey days,
38:27 you would get inspired by
38:30 whatever life,
38:32 some crazy ass thought you had. You
38:34 heard a song and smelled a banana pie
38:37 and some new idea came to you.
38:43 And then you would
38:50 either take something that you were good
38:51 at and express that idea through
38:54 something that you had talent at.
38:57 Or you would say, "Oh god, I had an idea
38:59 for a painting, but I don't know
39:01 anything about art,
39:03 but I got to make this painting, so I'm
39:05 going to go back to art school."
39:08 [laughter]
39:08 Right?
39:10 I have a vision for macra. I'm I'm gonna
39:13 get my degree in fiber arts from from
39:17 Parsons. [laughter]
39:20 And you would go off for two years and
39:22 you would learn about fibers and how to
39:26 stitch them together and hang them on a
39:27 wall.
39:30 Every time I see a fiber art piece, I'm
39:32 like, at some point someone's going to
39:33 have to dust that [ __ ] thing. And
39:35 that's going to be a nightmare.
39:37 [laughter]
39:38 It's the only thing I think when I see
39:40 fiber art, it's just like that's going
39:43 to be a lot of lot of dust. That thing's
39:46 going to be full. [laughter]
39:52 Um
39:57 but but we now live in this in this
39:59 world where we've got these tools that
40:01 are these instant
40:04 instant manifestors and these instant
40:06 reflectors.
40:10 And so you can go from
40:15 so so wait let's let's stick with the
40:17 the olden timey days. So, if I had an
40:19 idea in the olden timey days, even if I
40:22 had talents and skills,
40:25 I'd say, "Okay, I've got some vision for
40:27 a picture I want to make." And then I
40:28 would go out in the world with a camera
40:30 or a phone and I would take a picture of
40:33 some leaves. And then I would come back
40:35 and I would load it into Photoshop. And
40:37 then I would [ __ ] around with the color
40:39 tables and I would, you know, layer that
40:42 on top of two other images. And I would
40:43 put the eyes where the leaves were and
40:45 the tree where the trunk was and the dog
40:47 with the th
40:49 right and and I would it there there was
40:54 a there was a manifestation process, a
40:57 creative process to get an idea out of
41:00 your head that the idea had to be strong
41:03 enough to survive the gauntlet of
41:06 execution.
41:08 That's [clears throat] there's a new
41:09 phrase, the gauntlet of execution. I
41:12 don't know if you're still here on here,
41:13 Andy, but gauntlet of execution. That's
41:17 a good one. Um, in my speeches, I say
41:20 inspiration and motivation can only come
41:23 from focusing forward. Oh, that's
41:25 interesting. I like that. [music]
41:29 Although, I would argue a lot of songs
41:31 are written about focusing backwards.
41:36 [laughter]
41:37 I got in my pickup truck. I saw the note
41:39 on the seat.
41:42 She said she ain't coming back. And now
41:44 I got to pee. [laughter]
41:56 But anyway, back to ideas and
41:58 inspiration and where it comes from. So
41:59 you have some idea. And in the olden
42:01 timey days,
42:04 the gauntlet of execution was a kind of
42:06 filter. It was a kind of filter that
42:09 said, I'm not going to take ideas that
42:14 I don't really resonate with because if
42:18 if I have an idea and I don't really
42:20 resonate with it and then I enter the
42:21 gauntlet of execution and I'm like, I
42:23 don't feel like going outside and taking
42:24 that picture, well then that idea wasn't
42:27 strong enough, right? It wasn't strong
42:29 enough to overcome the resistance. And
42:30 and a lot of artists would argue that
42:34 resistance is required for great
42:37 creative or for great work
42:40 that that overcoming the challenge is
42:43 the is how you get to good work.
42:46 Now we live in a world where I can have
42:49 that idea
42:52 and I can just literally barf it into
42:56 chat GPT or midjourney oro
43:01 and squit. Out comes a a produced work,
43:06 right? I want an epic poem about stupid
43:09 idea.
43:11 There's an epic poem [clears throat]
43:13 written in amic pentameter,
43:17 right?
43:22 And so
43:27 I think you could argue that there's a
43:29 cost
43:31 to inst in instant manifestation
43:34 which is ideas can get out of your head
43:38 that aren't strong enough
43:41 quite frankly to survive the gauntlet of
43:43 execution.
43:48 But there's another side to it, which is
43:50 this, and this happens on this channel
43:52 regularly. If you're in a regular, if
43:54 you're here at this channel on a regular
43:56 basis, this happens a lot. Sometimes the
43:59 inspiration doesn't come from me.
44:03 Sometimes I'm just like I I have no
44:07 inspiration. I'm like, "Oh god, I'm
44:09 doing a live. I just told these people
44:11 I'd show them how to make a song. I
44:13 don't have a [ __ ] idea.
44:16 And so I'll go into chat GPT and I and
44:18 I'll literally type in I don't have a
44:20 [ __ ] idea. Give me 20 and it will
44:25 and then somewhere in those 20 ideas I
44:27 like weird Mary happened like this.
44:30 I was like I came up with some
44:32 nonsensical prompt that was like give me
44:34 something with twisted this and
44:36 something that
44:39 and then it spit out a bunch of names
44:41 and I was reading through the names and
44:42 one of the names was weird Mary from
44:46 Cedar Hill
44:51 and before that happens someone comments
44:53 something like we jive it's like jazz
44:56 it's totally like jazz actually jazz
44:59 music that's a really good Jazz music is
45:01 the closest thing we have in in the nonI
45:05 world to what I'm talking about,
45:09 right? Jazz is absolutely that where the
45:12 musicians are masters enough of their
45:14 craft.
45:17 When jazz musicians are in flow, they're
45:21 not thinking about anything. Everything
45:23 is just a reaction. And so and so now we
45:26 can do that with with ideiation. And
45:28 that's what happens in brainstorming
45:30 sessions, by the way, when you get
45:32 connected people that have complimentary
45:35 skills in a room together and they're
45:37 throwing out ideas. Stupid idea, stupid
45:40 idea back, stupid idea out. Interesting
45:42 idea back. Oh, that inspires this. Oh,
45:45 that's a good idea. Oh, now there's a
45:46 good idea. Oh, that's a great idea. And
45:48 then everyone pounces around the great
45:50 idea. So that's what happens in
45:51 brainstorming. That's kind of like jazz.
45:55 But
45:58 but we now have a thing in our computer
46:02 where we can do that back and forth
46:04 without anyone else being in the room
46:06 and you can get to something quite
46:08 amazing. Like the weird Mary thing, it
46:11 was like it came up with that name and I
46:13 was like that name's really interesting.
46:15 Let's go write some lyrics and now let's
46:17 go to Sununo and like what would that
46:18 song sound like? And then we were all
46:20 kind of jamming, throwing ideas together
46:22 and that turned into
46:26 into that crazy thing. So the the the
46:29 idea of inspiration right now I think is
46:32 even changing
46:35 and maybe part of a daily practice might
46:37 be
46:40 every day start with with a non idea
46:45 and and keep bouncing it back and forth
46:48 with AI
46:50 until it gives you a good idea.
46:54 So, so basically put a bunch of shitty
46:56 ideas into into AI of some form and jam
47:02 with it until something interesting
47:04 comes back. That that might be
47:05 interesting. Um, looks like we got AM
47:08 Murphy in the house. What's happening,
47:09 AM Murphy? AM Murphy and I had a really
47:11 fun actually we we wrapped up season
47:14 zero of the AI readiness podcast today.
47:16 Um, it was it was good just to to talk
47:19 live with with you again. Uh, so good to
47:22 see you.
47:25 Um, speaking of Ann Murphy,
47:29 December 26th and 27th, if you have not
47:32 heard, if you've heard this before, go
47:33 warm up your nachos because they're
47:35 getting soggy. I told you not to put the
47:38 canned cheese on them. I I tell you
47:40 every week, but you you insist because
47:42 you're white trash and we know it.
47:45 Um, AI Festivist
47:48 is this
47:51 remarkable thing we did last year that
47:53 we're going to do again this year.
47:56 Last year we had two weeks from the time
47:59 we conceived it until the time we
48:01 launched it. And this this year we have
48:02 two months [laughter] and it's already
48:04 been conceived. Um, so having having
48:07 that much time is probably it's probably
48:09 not good. But uh we had a we had a
48:12 really good meeting today uh about it.
48:15 Um
48:20 it happens on Friday and Saturday
48:22 between Christmas and New Year's or
48:26 whatever other celebration you do. Um
48:32 it's 9:00 a.m. to 9:00 p.m. Pacific, so
48:35 noon to noon Eastern.
48:38 And you're like, "Wait a minute, 9 to 9,
48:41 that's 12 hours." Like, "Yeah, that's
48:43 Friday." And then we take a 12-h hour
48:46 nap and we come back Saturday morning
48:50 and go from 9 to 9 again. So, it's 24
48:53 1-hour sessions over two days with
48:58 remarkably generous speakers that come
49:00 and give their time to teach anyone who
49:04 wants to come. So, my request is this.
49:07 Jot down the times and start thinking
49:09 about who in your life you want to bring
49:12 to that. And also start thinking about
49:15 this. What are the communities that you
49:17 think should know about it? Do you know
49:20 anyone in the press? Like do you know
49:22 anyone who knows people in magazines or
49:24 TV?
49:27 Because what what struck me is Dan and I
49:31 were talking today is like we don't brag
49:34 enough about these remarkable
49:37 communities we're a part of. Um and
49:41 one of my missions in 2026 is to knock
49:45 that [ __ ] off and start and start
49:48 getting us some some visibility because
49:50 it's a really remarkable community and
49:52 not enough people know about it. So, if
49:54 you know anyone, please reach out to
49:56 producer Brandon or Vicky Baptiste or
49:59 myself or Ann Murphy, whoever you know
50:02 and just let us know um who we should be
50:06 talking to to get more people there. All
50:08 right. Will you have marketing material?
50:11 We will. So, right now, Jeie, if you go
50:14 to the the the website's going to be the
50:16 same as it was last year, which is uh
50:18 aifestivist.com.
50:20 Right now, it's still got the old the
50:22 last year's uh stuff on it. So, that'll
50:26 be updated at some point here. It's it
50:28 it's coming. That's fine. Uh but just
50:30 you can just bookmark that site. So, at
50:32 some point it'll be updated. But if you
50:34 want to see who was there last year and
50:35 who the speakers were, it's really a
50:37 remarkable event.
50:39 Uh G&G on YouTube asked about Perplexity
50:42 at the start of the show. Um we can go
50:44 play with Perplexity. Um let's see. The
50:47 easiest way to learn is to play. Ask
50:50 perplex perplexity what it can do.
50:52 You'll be reading for a while. The
50:55 gauntlet of execution. I'll have to add
50:57 that one to the list. Who's that? That
51:00 was Andy. Good. [laughter]
51:03 Um, do you know anyone who can work with
51:07 me? I'm so lost. So, J&G,
51:11 here's here's [clears throat] the deal.
51:12 I can work with you right now. Tell me
51:15 where you're lost and we'll go play.
51:21 Um, J&G, one of the things you want to
51:24 do if you have it, Brandon, if you could
51:26 pop up the the salon website on the on
51:29 the on the TV [clears throat] machine.
51:32 If you go to community.thesalon.ai,
51:38 um, join the community, go in and
51:40 introduce yourself and go check out the
51:43 spaces there.
51:45 um
51:47 you need to hang out with the people
51:49 that hang out on this channel. So, one
51:50 thing to do is just keep coming back to
51:52 this. If you're feeling lost right now,
51:55 you're in the perfect place. I know this
51:58 is this is one of those I I I'm telling
52:00 you, man, this channel's getting like a
52:02 [ __ ] self-help guru channel.
52:04 [laughter]
52:05 You're exactly where you need to be,
52:07 J&G. Here's the deal, J&G. We're all
52:10 lost, okay? Everybody's nobody has the
52:13 answers. But
52:16 [laughter]
52:19 but that's actually the answer right
52:21 now. Everyone is lost. In fact, when you
52:24 said earlier, can someone help me go to
52:27 Perplexity and figure this out? My
52:29 immediate response is I haven't been
52:31 been to Perplexity in long enough that I
52:34 know I don't know how to use it right
52:36 now. And I used to demo it all the time
52:39 nightly. [laughter]
52:41 Hey, Marge. Hey, Marge. Yeah, that
52:43 perplexity is that just just go there
52:46 and use it. All right. Yeah. No, I just
52:49 I haven't used it in a They moved
52:52 everything.
52:53 Yeah. No, I know. I know. You're
52:55 watching the wheel.
52:58 Do you want some more cheese whiz? No,
53:00 you're good. All right. Yeah, Marge is
53:03 sweet. Um, so if you're feeling lost,
53:07 um,
53:10 and you want to get good at AI,
53:13 embrace this feeling
53:16 because it's, here's the thing. I know I
53:19 sound like I'm being flippant. I'm not.
53:22 Well, no, I am, but it it's out of I I'm
53:27 laughing at the
53:30 at the process right now. If you want to
53:33 be decent at AI, it is it's a daily
53:36 practice. You kind of have to just build
53:38 into your life. Okay,
53:41 I'm gonna I'm gonna feel uncomfortable.
53:44 I I'm going to I'm going to feel
53:45 uncomfortable. I'm going to be in a
53:47 place of not knowing anything and I'm
53:50 just going to go play. I'm just going to
53:51 go play and I'm going to I'm going to
53:53 take off. I think one of the things that
53:54 could be driving
53:57 anxiety in you is
54:02 wanting to accomplish a specific task
54:04 and not knowing how to use the tools.
54:06 The tools are weird. And so, so one of
54:09 the things we talk about in the AI
54:10 salon, in fact, Brandon, if you could
54:12 put up my share screen, we call it the
54:15 cycle of AI readiness.
54:18 And
54:19 it's this idea of
54:22 play first, create excellence, and
54:24 generously lead.
54:27 The reason to start with play is that
54:30 [clears throat]
54:32 if you just start with thinking about
54:35 using a computer, how we've historically
54:37 used them, which is I'm going to take
54:39 the processes that I have and I'm going
54:42 to use this computer to make those
54:43 processes more efficient. That's where a
54:46 lot of people start. So they start down
54:48 this very narrow lane of of automations
54:50 or efficiency.
54:54 The problem with that with generative AI
54:56 is that generative AI can do so many
54:59 things that are not just the thing you
55:01 know. It can do all these things outside
55:04 of what you know. like all those gaps
55:07 that you have in your knowledge and
55:10 skill base, you know, you got a you got
55:12 two or three at the center that you're
55:14 really good at these things and then
55:15 you're sort of good at these things and
55:16 then you're kind of crappy at these
55:18 things and then all the ones out here
55:20 you're just shitty at, right? That's
55:22 every human being. [laughter]
55:25 Well, AI is really good at all the
55:29 things.
55:30 And so what playing looks like is you
55:34 can h go do stuff you know, but you can
55:37 also go to areas of your life where you
55:39 know you're not good and just see what
55:41 it can do over there and you're like,
55:42 "Oh, I can do that. Wow, that's pretty
55:44 impressive." And so just playing is just
55:46 about sort of finding your way in the
55:48 dark. Literally just playing until you
55:51 start to figure out where the boundaries
55:53 are and you start to figure out what's
55:54 possible.
55:56 And then create excellence is you take
55:57 what you learn when you've been playing
55:59 and now you can say, "Okay, now based on
56:01 everything I've learned, what do I want
56:04 to do?" Do you want to just still
56:05 automate that thing? Great. Then go
56:07 automate the [ __ ] out of that thing. But
56:09 maybe what you want to do is you want to
56:10 automate that and then do two or three
56:12 other things that make that even better.
56:15 And then generously lead is talk about
56:20 what you're learning. Share what you've
56:22 created, good or bad. Hey, I tried to
56:24 make this thing and it was a steaming
56:26 pile of poop, but here it is.
56:28 That lets people know, it establishes
56:30 you as someone who's thinking critically
56:32 and creatively about how to learn and
56:35 master AI. And so what I would encourage
56:38 you to do is is keep playing like that.
56:40 Becky Rue third third post
56:44 um
56:46 using chat GPD to track help doctors
56:49 discover something slowing the healing
56:51 journey. Yeah. Great. Yeah, a lot of
56:54 people right now are
56:57 they're dealing with especially when it
56:59 comes to chronic chronic disease,
57:01 chronic pain, chronic illness,
57:04 um or hard to diagnose
57:06 uh issues. A lot of people are finding
57:08 chat GPT
57:10 augmenting the medical community, shall
57:14 we say? [laughter]
57:17 Hey doc, have you run any tests on this
57:19 particular thing?
57:21 Oh, what' you do? Chat GPT. All right,
57:23 I'll run the test. Oh, turns out chat
57:26 GPT was right. There's a lot of that
57:29 going on.
57:33 All right, I don't know if PNG is still
57:35 here because I've been talking to them
57:37 for a while. Or J&G.
57:40 Um, J&G, you still here?
57:43 Because I I would like to be able to go
57:46 play with a specific question you have.
57:58 play with a AI the same way you play
58:00 with the guitar.
58:02 Well, you know what's funny, Weaver, is
58:04 like right now lately, I have been
58:06 playing with AI the the way I play my
58:08 guitar, which is I haven't written a new
58:10 song in a while. I haven't learned a new
58:12 song in a while. And I'm the same way
58:14 with with AI right now. Like, I kind of
58:16 have my lanes of the [ __ ] I know how to
58:19 do.
58:20 >> [clears throat]
58:22 >> Yeah, it's fascinating hearing people
58:24 talk about I I use Chat GPT every day,
58:27 multiple times a day, and now I use it
58:28 like twice a week.
58:32 Yeah, that feels familiar.
58:39 But again, so again, like everything
58:41 keeps circling back to I think we need
58:43 to spend more time on
58:47 really articulating what are your
58:49 values? Like what are you trying to do
58:51 in the world? What what difference are
58:52 you trying to make? I've been asking AI
58:55 forums to teach me according to my
58:57 personality. It's kind of funny. That's
58:59 good. I'm so h happy to be answering all
59:01 of my internal questions against a
59:03 cascade of against the cascade. Hard to
59:07 know your value.
59:10 Um
59:13 yeah.
59:17 Yeah. That's I mean that's [ __ ] life,
59:19 isn't it?
59:22 I talk about this every [ __ ] day.
59:25 [laughter]
59:26 I think Andy's probably tired of hearing
59:28 it at this point. [laughter]
59:33 Oh my god.
59:37 My doc asked me if I minded if she took
59:40 notes with AI. Do I mind? No. I want you
59:44 to take advice from AI doc. [laughter]
1:00:00 Let's even though J&G is not here
1:00:02 anymore or it's possible that our our
1:00:05 comments just stopped. Would someone on
1:00:07 YouTube post a comment? We only have 26
1:00:10 people there. I love it, man. It's been
1:00:12 so long that people haven't
1:00:14 got to internalize how absolutely badass
1:00:16 we are. [laughter]
1:00:19 AI explore perplexity AI. Okay,
1:00:26 let's jump over.
1:00:31 By the way,
1:00:34 this is the GPT that Brandon created. If
1:00:37 you know anyone that is at risk of
1:00:40 losing SNAP benefits, the GPT that he
1:00:43 created is called Help After SNAP.
1:00:46 Um, this is the one that, you know, has
1:00:49 got this, you know, warmness to it, this
1:00:52 empathy and compassion to it. Um, and
1:00:54 will actually help find local services
1:00:57 based on where someone is.
1:01:00 Um,
1:01:03 posting a comment for you. texting link
1:01:06 in community feed and salon. Okay, so if
1:01:09 you go to the AI salon down the left
1:01:10 hand side there's an area called
1:01:12 community. The second link there is
1:01:14 called community feed and the link to
1:01:16 that custom GPT is in there. [snorts]
1:01:19 [clears throat] Okay,
1:01:21 let's go to perplexity. Perplexity
1:01:25 Plexity AI
1:01:29 perplexity spaces
1:01:33 create a space. I guess that's like a
1:01:35 project.
1:01:36 Didn't they have pages?
1:01:41 Download the Comet browser.
1:01:46 Comet and perplexity pro are free for a
1:01:49 year with PayPal, which how do you do
1:01:52 that again? You you go to PayPal and
1:01:55 then click on
1:01:58 click on offers and then yeah, you
1:02:00 connect it to Perplexity and they'll
1:02:02 give you a year. And I think Venmo is
1:02:04 doing this also. Um,
1:02:08 okay. So, let me talk a little bit about
1:02:10 what what Perplexity is. So, Perplexity
1:02:12 is like chat GPT. Um, but unlike
1:02:16 ChatGpt,
1:02:19 they decided to make research and and
1:02:22 web searching a core part of the large
1:02:26 language model experience. So, one of
1:02:29 the things you that's that's fun to do
1:02:30 with perplexity is is research things.
1:02:33 So, um let's see. I'd like to know
1:02:42 the history
1:02:45 of
1:02:47 peanut brittle
1:02:52 and toffee.
1:02:54 [laughter]
1:02:59 >> [laughter]
1:03:01 >> to feffy.
1:03:04 Okay. So, oh, look. And there it goes.
1:03:09 So, coco and heart. So, there's the
1:03:11 answer. Peanut brittle. Oh, that was
1:03:13 fast. Is it still? Yeah. Okay. So, it's
1:03:15 off doing it's off doing research.
1:03:20 And let's see. Can I turn this into a
1:03:22 page?
1:03:24 Does it still have pages?
1:03:26 Upgrade to max.
1:03:32 Create projects from scratch.
1:03:39 Search. [gasps]
1:03:46 Huh. I don't
1:03:49 One of my favorite features.
1:03:51 They It looks like they killed I wonder
1:03:53 if that's what's called spaces.
1:03:56 Quick, let's make a Gemma slideshow
1:03:57 about it.
1:03:59 You used to be able to turn turn this
1:04:01 into like a magazine article.
1:04:09 Mark down tax
1:04:26 for pro. It's 20 bucks a month.
1:04:31 Ask perplexity. That's funny. I should
1:04:36 um I go back. Oh. Um, let's see.
1:04:42 Doesn't Perplexity
1:04:45 as a feature
1:04:48 have
1:04:50 a feature
1:04:53 called pages?
1:05:10 It might [snorts] be I guess pages might
1:05:12 be although it says I have a pro
1:05:14 accountation
1:05:24 account preferences. is
1:05:36 didn't I call this didn't I predict this
1:05:38 that I would come to perplexity and not
1:05:40 know how to use it. [laughter]
1:05:49 Oh my god. Gemini is working hard on
1:05:52 competing with gamma for in the
1:05:54 presentation space. Becky Rue
1:05:59 also apparently Imagine 4 is the image
1:06:02 app to beat. Use it for creating a nano
1:06:05 banana for editing.
1:06:08 Yeah, and I hear Nano Banana 2.0 nose
1:06:10 coming out.
1:06:16 And apparently November 18th
1:06:20 is uh
1:06:22 is um Gemini 3.
1:06:27 >> Hey.
1:06:29 >> Yeah. So, I'm on the Perplexity Pages
1:06:31 help center um reading the instruction
1:06:33 manual because one of us asked you. Um
1:06:35 yeah, and it's saying that uh if you
1:06:38 enter any prompt, you watch your page
1:06:40 take shape with a comprehensive draft,
1:06:43 then you can customize the content,
1:06:44 tweak the formatting, and add multimedia
1:06:46 elements to make it uniquely yours. So,
1:06:50 apparently it is just
1:06:52 >> it's responses just became pages.
1:06:55 >> Yes. And I think spaces is the new kind
1:06:58 of pages.
1:07:00 >> Oh, really? Create a space.
1:07:06 But I used to have a bunch of pages. Do
1:07:08 I like I wonder if they're gone.
1:07:11 It ate my It It ate all my stuff. Let's
1:07:14 see.
1:07:16 Discover
1:07:21 for you.
1:07:27 So this is dynamic news generation based
1:07:30 on my preferences.
1:07:32 That's kind of cool
1:07:36 finance.
1:07:46 Yeah, actually, you know, um if you
1:07:49 hover over the home, you should be able
1:07:51 to access your library.
1:07:56 >> Ah,
1:07:58 media apps. Is
1:08:01 >> it possible that you had a different
1:08:03 account?
1:08:04 >> Nope.
1:08:07 Nope. That's my account. Although it's
1:08:10 possible. I I don't know. I don't know.
1:08:14 That's about as much troubleshooting as
1:08:16 I can offer.
1:08:17 >> Yeah, that's good. Thanks. Um, here's a
1:08:20 here's a really good lesson. I I'm
1:08:22 giving you a good lesson right now.
1:08:25 Here's the good lesson. If you don't use
1:08:27 a tool regularly, don't assume that you
1:08:30 know how it works. [laughter]
1:08:37 If you if you've got a channel that
1:08:40 where you've established yourself as an
1:08:42 expert who knows how [ __ ] works,
1:08:45 don't go to a tool you haven't been to
1:08:48 in six months. [laughter]
1:08:50 You're welcome.
1:08:54 All right. [ __ ] it. We're not building
1:08:56 anything tonight. [laughter]
1:09:00 Could you talk about the tools?
1:09:02 Apparently no. Apparently, the answer is
1:09:04 no, I can't. I think I think that I am I
1:09:08 I think here's what I'm clear about. I'm
1:09:11 clear that this week my brain is not in
1:09:14 a in a place where I can do anything
1:09:17 productive. Although, I would argue that
1:09:19 figuring out what the [ __ ] we're doing
1:09:21 with our humanity is pretty [ __ ]
1:09:23 productive. Um,
1:09:26 I'm gonna crack this nut. I I really am.
1:09:28 I'm going to crack what it means to be
1:09:32 a professional
1:09:34 who is AI literate
1:09:37 in this world and and
1:09:41 I'm determined to crack
1:09:45 a way to
1:09:49 make it easy for us as a community to
1:09:52 know what to do.
1:09:55 What makes you think you can do that,
1:09:57 Mr. Smarty pants.
1:10:01 I don't know. Cuz I'm passionate about
1:10:03 it. You're passionate about pumpkin
1:10:06 varieties. I'm passionate about
1:10:10 humanity using new technology as tools
1:10:14 for self-expression.
1:10:19 If anyone has any ideas, I would love to
1:10:21 hear them. You can take a peek at more
1:10:25 Eloiseas in practice. Oh, that's cool.
1:10:27 In the practice area, I will definitely
1:10:30 do that. Perplexity AI demo. Ask about
1:10:33 features.
1:10:35 No, I'm not talking to AI tonight.
1:10:39 [laughter]
1:10:40 I'm pissed off that they took away the
1:10:42 the one thing I used to like to demo
1:10:44 that is gone and all of my things that I
1:10:47 made in it are gone. So, I'm cranky.
1:10:50 [laughter]
1:10:52 [ __ ] perplexity. I'm going to get into
1:10:54 the uh I'm going to get into the
1:10:56 producer AI
1:10:58 um creative partner program and then
1:11:00 we're going to start making movies.
1:11:02 [laughter]
1:11:05 Oh, in irregulars. They are my practice
1:11:08 but in irregulars. All right, let's go
1:11:10 look at Corey's Eloise
1:11:13 Pitches.
1:11:15 Eloise pitches AI learning lab.
1:11:20 Oh, here they are. Eloise is thriving.
1:11:23 Nice.
1:11:25 So, Corey, what was the um
1:11:29 Can I just go right? Yeah, these are
1:11:31 great. Oh, I'm not sharing them, am I?
1:11:35 Um,
1:11:38 so these are from Corey Sandler. So,
1:11:39 Corey Sandler, if you don't know, she's
1:11:42 an amazing [clears throat] artist and
1:11:44 musician and chef, cook, cook, chef,
1:11:48 someone who makes food lovingly.
1:11:51 um makes beautiful pottery. So if you go
1:11:53 to core sandlerpottery.com I think or
1:11:56 corey sandler.com. I guess she's got
1:11:58 both is my guess.
1:12:01 And she does work with um
1:12:07 with AI that's just very very thoughtful
1:12:09 and very
1:12:12 when we were all when we were all
1:12:13 teaching each other how to use these
1:12:15 tools and how we use the tools. Corey
1:12:17 was Corey was one of those people that
1:12:19 when she shared her
1:12:23 like how she did what she did, it just
1:12:25 blew me away cuz she would
1:12:28 she would sort of get an image that I
1:12:31 would think would be so beautiful that
1:12:32 it didn't need refinement and then she
1:12:35 would just dig and dig and dig and dig
1:12:37 and just run deep deep deep deep down
1:12:40 these sort of creative refinements.
1:12:44 And so I guess she's doing that again
1:12:45 now with Eloise.
1:12:48 This lovely lass.
1:12:52 It's very cool.
1:12:55 It's beautiful. It's beautiful.
1:12:59 Silverf Fox loving producer AI. Yeah,
1:13:01 producer AI is really cool. It really
1:13:03 is. It really is a good a good thing.
1:13:07 It's a groova. It's groova, I tell you.
1:13:12 Um,
1:13:36 I just
1:13:43 So, what are your top five nowadays? Um,
1:13:50 chat GPT still at the top,
1:13:55 Suno for music, Midjourney for images,
1:14:00 midjourney for video.
1:14:04 Um,
1:14:06 I think the Sora app and the little, um,
1:14:10 cameos are quite fun.
1:14:14 I think they're good entertainment when
1:14:17 I do them.
1:14:22 For voice, I use Cartisia. For Avatar
1:14:26 animation, I use Hedra. H E D R A.
1:14:31 For coding, I use lovable, but I don't I
1:14:34 don't do a lot of coding. And and to be
1:14:36 fair to myself, I haven't really taken
1:14:38 the time to learn
1:14:41 cursor or one of the more robust ones
1:14:43 because I just don't have the patience
1:14:45 for it. So, I tend to make kind of
1:14:48 throwaway apps in lovable right now. But
1:14:51 that's okay. Like I like I feel like
1:14:55 like like lovable apps to me right now
1:14:58 feel like the coding equivalent of um
1:15:02 what do they call them? Casual games,
1:15:04 you know, on your iPhone. Those stupid
1:15:06 games where you just like shoot a
1:15:07 basketball into a hoop endlessly.
1:15:10 There's like there's no game loop. It's
1:15:11 literally just
1:15:13 dare. [laughter]
1:15:15 Those kind of games. I feel like that's
1:15:17 what lovable lovable app development is.
1:15:20 It's just like you sort of make this
1:15:22 thing that's like it's fine and then you
1:15:24 just move on
1:15:26 [laughter]
1:15:28 and I know there's other ways to use it.
1:15:29 That's how I use it right now.
1:15:32 Um
1:15:38 the the the one that that I am
1:15:43 I'm trying to think of the ones that I'm
1:15:44 really impressed by recently.
1:15:48 Um, GenSpark for me is one that I find
1:15:51 very impressive. Producer.ai I find
1:15:55 really impressive just because it's a
1:15:59 it's a chat interface to music. You can
1:16:01 just talk to it kind of like you're
1:16:02 talking to a producer, hence the name or
1:16:05 that you are the producer.
1:16:07 Um,
1:16:11 I think that the Gemini suite of
1:16:14 products are getting really good. So,
1:16:17 gemini.google.com,
1:16:19 aistudio.google.com,
1:16:22 their vibecoding app is really good.
1:16:26 Um, labs.google.com has a bunch of
1:16:29 really cool tools you can go play with.
1:16:30 That's where Notebook LM came out of.
1:16:32 Notebook LM is one that I use. So,
1:16:35 there's So, my top five, what was that
1:16:37 15? My top five is the top 15. Top 15.
1:16:43 [laughter]
1:16:44 I'm hooked on the Gen Spark browser. I
1:16:46 see. I haven't even played with the Gen
1:16:47 Spark browser. That's a whole another
1:16:48 area. There's a whole other area that I
1:16:51 have not gotten my head around and
1:16:54 that's um AI browsers. So, you've got
1:16:56 Perplexity Comet, you've got the Gen
1:16:58 Spark browser, you've got the OpenAI
1:17:00 browser, you've got who else has got
1:17:03 one? Um
1:17:06 well, Google has one called Chrome.
1:17:08 [laughter]
1:17:13 Um, GPT online browser has serious
1:17:15 security issues right now. Yeah, I know.
1:17:18 All of them do. What? Are you kidding
1:17:19 me? They They all have s serious
1:17:21 security issues right now.
1:17:25 It's just I said this the other night. I
1:17:28 feel like we're in, if you were alive in
1:17:30 the '9s,
1:17:32 there was like a four-year period where
1:17:34 CDROM development was all the rage and
1:17:37 everyone's like, "You got to get into
1:17:38 CDROMs, man." and all the publishing
1:17:41 companies were getting into CDROMs. And
1:17:43 all the B2B data companies were getting
1:17:46 into CDROMs.
1:17:48 We've put the National Directory of
1:17:50 Popcorn Manufacturers on a CD ROM. We
1:17:54 put it on with a fancy interface where
1:17:56 you click on little popcorn kernels. We
1:17:58 think it's quite innovative.
1:18:01 Um,
1:18:03 and then the internet came along and
1:18:04 just it it was like this weird phase,
1:18:07 right? I feel like that's where we are
1:18:08 with AI right now. We're in this weird
1:18:10 phase where we're putting a lot of time
1:18:14 into building things on tools that
1:18:17 aren't going to exist and and not only
1:18:19 tools that aren't going to exist, but
1:18:20 like mechanisms of delivery that aren't
1:18:23 going to exist. It's just going to be
1:18:25 this weird flash in the pan moment. So,
1:18:29 so I don't know.
1:18:31 Even college textbooks come on syncing
1:18:34 CD CDROMs. Exactly. remember that the uh
1:18:38 it it used to be 17 volumes that took up
1:18:41 the back of a Volkswagen minibus and now
1:18:45 it's on two CDROMs now on seven shiny CD
1:18:49 ROMs [laughter]
1:18:53 and there were games and art museums on
1:18:55 CDROMs.
1:18:57 I remember feeling like I was a little
1:18:59 late. I had missed the CDROM party and I
1:19:01 was just making these stupid websites
1:19:03 because HTML sucked so bad and then uh
1:19:07 and HTML got a little bit better and the
1:19:10 browsers got better and CDROMs just went
1:19:12 poof. They were just gone. It was like
1:19:15 it was like amazing like 1996 97 they
1:19:19 were just done. Amazing.
1:19:22 Anyway,
1:19:25 and how you fix your toilet with CDROMs.
1:19:27 Exactly. Yeah. All those all those
1:19:30 macromedia flash animations. How to do
1:19:33 things, how to fix your car, how to fix
1:19:35 your toilet. That's pretty good.
1:19:39 All right, I'm going to leave. Um, I
1:19:42 listen, one of my pet peeves and one of
1:19:45 the things that I tell you guys all the
1:19:47 time is no apologies,
1:19:53 but
1:19:57 I am feeling a wee tad
1:20:04 underwhelming in what I'm delivering
1:20:07 here cuz I'm trying to figure some [ __ ]
1:20:09 out and so I think that's boring.
1:20:10 boring. So, if I've been boring for the
1:20:12 past three nights,
1:20:15 sorry, my head hurts. I'm trying to
1:20:18 figure [ __ ] out and I'm telling you, I
1:20:21 will come out of the other side. At some
1:20:22 point, I'll crack the code on this
1:20:23 [ __ ] thing and we're going to start
1:20:25 having fun in here. Uh, but I don't know
1:20:27 what it looks like yet. So, that's just
1:20:29 that we're in the we're in the
1:20:31 uncomfortable unknowing part of life
1:20:33 tonight. So, I'm not going to apologize
1:20:36 for it. But, if it's been a little
1:20:37 boring, keep hanging out. Just keep
1:20:40 showing up.
1:20:42 Keep showing up. And here's a, you know,
1:20:45 let me share something with you, you
1:20:46 know, from from my my own personal life.
1:20:50 I used to love love watching David
1:20:54 Letterman.
1:20:56 Like, I I watched it every [ __ ]
1:20:58 night. I would memorize the stupid pet
1:21:01 tricks and the stupid human tricks and
1:21:03 all the [ __ ] that he would do.
1:21:06 Um, I knew all the guests. I knew all
1:21:08 the bits. I knew his timing.
1:21:12 And one of the things that I like best
1:21:15 about Letterman when he when I thought
1:21:17 he was his funniest
1:21:21 was when he was bombing. [laughter]
1:21:26 So when when David Letterman was not
1:21:29 being funny and the audience didn't get
1:21:32 what he was doing, there was something
1:21:34 about it. And and I don't I'm not
1:21:36 presuming that I'm anywhere near there,
1:21:39 but you know, if we can find joy in the
1:21:43 shitty episodes, the good ones are just
1:21:46 even better. [laughter]
1:21:50 Letterman and then Craig Ferguson. Yeah,
1:21:52 exactly. Yeah, Craig Ferguson was
1:21:53 hilarious. He was great. I think he was
1:21:55 a very underrated late night talk show
1:21:58 host. He was great.
1:22:01 Um,
1:22:03 if I look at like reruns of talk show
1:22:05 hosts, it's Letterman and Craig
1:22:06 Ferguson's are the ones that I I really
1:22:08 like. They're the ones that hold up for
1:22:10 me.
1:22:12 So, anyway,
1:22:15 all right. I'm out. Peace out. Uh,
1:22:18 tomorrow's Thursday.
1:22:21 Festivus is coming. I don't think I have
1:22:23 anything for you. Tomorrow is Thursday,
1:22:25 right? Today's Wednesday. Yes. Did I get
1:22:28 that right? Hella freaking luy.
1:22:32 All right.
1:22:40 I feel like one of those TVs. You know
1:22:42 the old tube TVs from the 60s where
1:22:45 you'd turn them off and the picture
1:22:46 would go like
1:22:49 it would just slowly shrink down to the
1:22:52 little white dot and then the white dot
1:22:54 would just go.
1:22:57 That's what I feel like I'm doing
1:22:58 tonight. I'm just like,
1:23:03 [laughter] "Good night."
1:23:09 >> [laughter]
1:23:11 >> Oh, good lord.