AI Learning Lab LIVE - 12/30/2025

Description

Chapters

Transcript

0:06 Yeah. Hold, please. Beautiful.
0:32 Oh no.
0:39 >> With a spoon covered in sauce.
2:40 All right, good people. I'm back. It's
2:41 okay. Everything's going to be fine.
2:44 It's all going to be fine. You're going
2:45 to be okay. I've come It's all right.
2:48 I've arrived.
2:53 I'm telling everyone, you're getting
2:55 your notion board configured. Listen,
2:57 here's the victory with notion. I did
3:00 not quit the app today
3:06 and it actually forced me. Okay, so a
3:09 couple of So, there's a lot happening.
3:11 Anyway, we we'll get back to that. Let
3:13 me sing with my dog. That's why you've
3:15 come here. Really? Have you not?
3:18 All
3:25 right. Ready? One, take one. Zoom in on
3:28 Kyle and action.
3:43 Oh
4:09 time I see it now. Get that look in
4:13 mine.
4:15 Every time I see your mouth, I hear that
4:19 smile.
4:22 In the early misty morning light, I
4:25 heard the engine turning in the old for
4:29 outside.
4:36 Will you believe in me
4:39 again today?
4:42 You will convince me
4:45 again today.
4:47 You're leaving this hotel looking for
4:51 someone else's golden ring.
4:56 Should have said
4:58 so long. Susanna,
5:04 first now don't you cry.
5:10 So long, Susanna.
5:15 Don't you cry for me.
5:32 It's Mr. Champ Shannon on lead vocals.
5:35 My name's Kyle Shannon. I'm singing back
5:37 up for the dog. This is where my life is
5:39 led me.
5:42 Hey Danielle was shaking. Side hustle me
5:45 in the house. We got Vicki. We got
5:46 producer Brandon. Producer Brandon right
5:48 here. You know what he's doing?
5:50 Producing Vicki on YouTube. Better.
5:53 There's a notion tutorial in irregulars
5:56 for anyone who got confused last night.
5:58 So, listen, if you were not here last
6:00 night, so a couple of things going on.
6:02 Welcome if you're new. Um, it's just a
6:06 handful of people in here. This is
6:07 probably all people that know what the
6:08 hell's going on,
6:10 but we're shifting the focus of this
6:12 channel.
6:26 And what it looked like was this.
6:32 We uh we Liz Miller Gersfeld and myself
6:35 we we uh we took some work that some
6:38 personal work that she's doing
6:42 about
6:43 creating a personal practice around her
6:48 use of AI
6:51 and said wow what if we created a
6:55 framework that's that's based on the
6:57 ideals of the
6:59 of the AI salon and play first, create
7:03 excellence and generously lead the cycle
7:06 of AI readiness
7:08 and we build a practice framework around
7:10 that and teach people how to do a daily
7:13 practice.
7:15 Sounds awesome. Sounds awesome. And we
7:19 created this thing called the AI Salon
7:21 Mastermind Practice Lab. And
7:25 it's a weekly
7:27 meeting, focused meeting where we're
7:30 teaching people how to create their
7:32 daily practice using this framework we
7:34 created.
7:36 Everything's good so far. Everything's
7:38 awesome.
7:40 It was funny when
7:43 when I first heard about the idea of an
7:46 AI practice,
7:48 my immediate instinct.
7:53 I just hold everybody hold. I've got
7:55 schmutz on the back of my phone lens.
7:57 It's causing glowy things. Let me see.
8:00 Do I have Oh, I've got a a pink bow.
8:04 The famous pink bow. Hang on. Hang on,
8:08 people. Everybody, everybody calm down.
8:13 Good lord. There, that's better. See,
8:16 not so glowy.
8:18 Okay,
8:21 so when I first heard about this idea,
8:23 this this practice,
8:29 I thought,
8:32 I know I'm not wearing it right.
8:33 Whatever, ladies.
8:36 I know how to wear a headband. Okay. Um,
8:42 I thought I would love to have a a
8:44 practice. That sounds so cool.
8:48 And then I thought, wait a minute. I go
8:50 live on TikTok five nights a week.
8:54 Technically, not a daily practice, but a
8:56 daily weekday pra a week daily practice.
8:59 A week a weekday daily pra a week a five
9:02 day a week practice.
9:06 And then we started teaching the
9:10 practice, the framework.
9:14 And guess what I learned? I didn't have
9:16 a practice. I thought I had a practice.
9:19 I didn't. I had a habit. I had a habit
9:22 of showing up five nights a week, which
9:25 to give myself some credit, is pretty
9:27 [ __ ] remarkable. No, I I think it's
9:31 it is just just slightly less remarkable
9:34 than all of you who show up here on a
9:36 nightly basis to watch it. I think that
9:38 is equally if not more remarkable that
9:41 that happens. Um
9:45 but that's not a practice.
9:49 A habit is not a practice. Doing
9:51 something every day does not necessarily
10:00 accomplish anything,
10:03 right? We get up every day and we have
10:05 coffee.
10:11 It's part of our history. I know it is.
10:13 It's actually the fact that we all hang
10:16 out here together for, you know, years
10:18 now is [ __ ] remarkable. And you wait.
10:21 We're going to work on Sydney tonight,
10:23 by the way. So you you just [ __ ] wait
10:25 when that show [ __ ] opens in New York
10:27 City and we're all in the theater
10:29 together watching that show.
10:32 I I won't it I won't be able to be
10:34 there. It it I I'm going to be a train
10:36 wreck. Um anyway
10:41 um
10:46 a practice requires intentionality
10:50 and um
10:52 it's taken me we've been we've been
10:54 teaching the course now for
10:57 I think we're on our seventh or eighth
10:59 week so two months
11:02 and I'm just now starting to get my head
11:04 around what it means and and the the
11:06 kind precipitating force uh for me was
11:10 was Cindy Coon's talk during AI
11:12 Festivus. So if you haven't seen if you
11:15 were not there for AI Festivus, it just
11:17 happened this past weekend.
11:19 If you missed it, it's a shame, but this
11:21 is a ridiculously low price. For 27
11:24 bucks, you can get the deluxe replay
11:27 bundle that includes like Jelly Pod
11:29 podcasts about the sessions, notebook
11:32 LM, infographics, the recordings,
11:36 transcripts, the presentations,
11:38 everything for 24 hours of programming.
11:42 Champ, you got to be kidding me. You got
11:44 to be kidding me. I I I haven't even
11:47 start I haven't even gotten to the good
11:49 stuff yet.
11:52 I want a little cheese.
11:55 I want a little cheese. Daddy. Daddy.
11:58 Daddy, get out the cheese. Daddy, get
12:00 out the cheese.
12:02 All right, hang on. I got to go let the
12:04 dog out.
12:14 >> Daddy, can I have cheese?
12:18 >> Go.
12:20 Go. Go out. What are you doing? Oh my
12:24 god.
12:27 >> What do you want?
12:29 >> You want to sing more? You might need to
12:31 sing more.
12:34 Little son of a [ __ ] I opened the door
12:36 and he wouldn't go out. What do you
12:37 want? You want to sing more?
12:48 Champ, what are you doing? You singing?
13:02 Hey. Hey. Hey. There's no rapping. What
13:05 are you doing? We're singing or we're
13:07 not.
13:11 All right. Um, take three. Oh, please
13:16 listen. What do they say? Never work
13:17 with kids or dogs.
13:19 This is why. This is [ __ ] why. Oh my
13:24 god. He really did want to sing. Now
13:26 he's just laying down. Are you you What?
13:29 What was that about?
13:32 He like whined by the door. I walked
13:34 over and opened the door and he ran back
13:35 in the room. Dogs are weird, man. Dogs
13:39 are weird.
13:42 Play Trouble. My favorite. Okay.
13:47 A champion might sing to trouble.
13:58 Standing between
14:01 you and a hard place is insane.
14:07 Standing too near
14:10 you in a fire makes it clear.
14:16 You're trouble to me.
14:22 Rich trouble catch you say
14:28 leaning in close.
14:31 Smell of your perfume scares me most.
14:36 Leaving away
14:40 feels stronger every day.
14:46 You're trouble to me.
14:52 Real trouble. Can't you see?
15:01 Okay, back to the practice thing. So,
15:06 um,
15:11 a practice is something you do every day
15:13 with intention. And what it also looks
15:16 like is you have to sit with what I call
15:18 the ick.
15:20 And the ick is that thing that if you
15:23 want to do something, so play first,
15:27 create excellence. the the create
15:29 excellence part of the the cycle of AI
15:32 readiness.
15:37 You can't just live and play first land.
15:41 If you want to create excellence, you
15:44 have to go to some dark places.
15:48 If you want to create more than a
15:49 seven-second video out of one of the
15:52 video models or you want to create
15:54 something that has the impact of Kelly
15:57 Besh, which I just realized I've been
15:59 mispronouncing her name for [ __ ]
16:01 three years.
16:04 It's got an O in it to my defense,
16:09 but I was interviewing Kelly during
16:11 Festivist and she goes, "Hi, I'm Kelly
16:12 Bes." And I was like, "Ah, [ __ ] Really?
16:19 Mimi in the making. Sad hustle. Mimi. I
16:22 like Oh, Mimi in the making. I love it.
16:23 Being born. Exactly. Just like what
16:25 we're doing here. So, here's what's
16:27 going to here's what's going to happen.
16:29 Rather than this channel being about me
16:34 flasharning something
16:37 an hour before I teach it to you, which
16:40 is which is historically what it's been.
16:46 What the channel's now going to be is
16:48 I'm still going to be exploring new
16:49 tools. I'm still going to be playing
16:50 first, but what I'm going to be doing
16:52 also is
16:54 working on my projects. I've got so many
16:57 projects and they're projects that are
16:59 meaningful to me and they're meaningful
17:01 to other people. And
17:04 a lot of them, every project, if you
17:07 want it to be truly excellent, has
17:09 things in it that are distasteful.
17:12 Things where you have to learn something
17:13 new. Things where you have to do some
17:15 file organization. Things where you
17:17 maybe have to do some finances or some
17:19 contract work, right? Not just the sexy
17:22 stuff. Like one of the things that we're
17:24 going to work on tonight is I put
17:26 together this podcast.
17:30 It's like a Notebook LM podcast with the
17:32 Sydney songs edited into it and it was
17:35 really good. It was really like I was
17:37 really happy with it and everyone else
17:38 was really happy with it. And then
17:41 Andrew is like, "Well, except can I do
17:42 an edit where I do a much more
17:44 sophisticated edit with like, you know,
17:46 better, you know, transitions between
17:49 songs and overlaps with the vo voices
17:51 and can we get the voices just exactly
17:53 right?" And then we have a new version
17:56 of the of the poster that is not just
17:59 like this 14-year-old
18:02 pasty white girl, right? It's it's a
18:06 it's a racially, you know, diverse or or
18:09 neutral neutral entity and she doesn't
18:11 look like she's 12.
18:13 Um, and we've got a better logo. And so
18:16 I've got this whole video that I've
18:18 created with all the old logo in it and
18:21 the old the old Sydney in it.
18:25 And so this thing that I spent, I don't
18:28 know, two days working on, which I
18:30 thought was quite good, was good, but it
18:33 wasn't excellent, right? So to make it
18:36 excellent, I'm going to have to
18:38 completely rebuild it. And I don't even
18:40 have all the components yet, but I have
18:41 the new logo, the new posters,
18:44 and so I figured we'd work on that
18:45 tonight. But then there's also things
18:47 like here, let me show you something. I
18:48 want to show you something fun. Well,
18:50 it's not fun. This is one of those ick
18:52 places.
18:54 Um,
18:57 here is my
19:04 go into
19:13 I don't know if you can see that. Is
19:15 there a way I can make this bigger? Does
19:19 command plus work on these windows? It
19:21 does. Oh, it's just two two things.
19:24 Okay.
19:26 So, you know how I always joke about um
19:29 saving things? Well, you you may or may
19:31 not I always joke about my my file uh
19:35 saving system is to save things to the
19:38 desktop with a nondescript name. So,
19:41 you'll notice here I've got a bunch of
19:42 nested folders. So, I've got Kyle
19:44 Shannon monetizable products, AI
19:47 launchpad workshop, storyline selling,
19:49 AI festivist, content evolution collab,
19:52 AI readiness training program, 10-person
19:54 team workshop, AI Kyle's AI consulting.
19:57 So, I've got all these these nice
19:59 organized folders. Those I did today.
20:02 Those I did today. Now, I'm going to
20:04 show you something. I'm just going to
20:05 keep scrolling here.
20:07 All of these are loose individual files
20:12 that have zero organization to them.
20:18 Within these files, there's actually
20:20 some really good stuff in here. There's
20:22 stuff that could contribute to projects
20:23 being greater.
20:26 That's this is just my loose Google
20:29 Drive.
20:32 My personal my personal Google Drive is
20:34 just full of just random [ __ ] Okay. And
20:39 so so like when I talk about sitting
20:41 with the ick and like why would I spend
20:45 any time
20:48 on this channel showing file structure?
20:53 Um when I when I look at how organized
20:56 this is and like all these projects and
20:58 I can start to actually put stuff in
21:00 them. That's kind of fun. That's kind of
21:01 exciting. Then I think about the notion
21:04 um the notion instance that Vicky
21:07 Baptiste helped me figure out last night
21:12 and let's see
21:16 if I go to my control board. Here's
21:18 Kyle's control board. So this is now I'm
21:20 in notion.
21:22 So, Lord Digital Gods sent me a a really
21:25 funny thing that said, "I wonder if
21:27 Kyle's gonna is how long will Kyle how
21:30 long before Kyle quits notion?" I'm
21:33 like, I'm not quitting it. Damn it. I'm
21:34 gonna I'm gonna figure this thing out.
21:36 I'm going to figure this thing out. So,
21:38 within here, I've got business areas.
21:41 So, I've got AI Salon, Content
21:43 Evolution, Collab, Kyle, Shannon Media.
21:46 I want to I want to make sure that all
21:48 of these things all of these projects in
21:51 notion match
21:53 all of my folder structure in my hard
21:56 drive.
21:59 Now, why is this? Because
22:03 So, if I flip over here, wait, we're
22:06 we're taking side bets. Yeah. No,
22:08 listen. Please do. Please do hold me
22:10 [ __ ] accountable. Um, so here you're
22:13 looking at some some New Sydney um
22:17 videos here. Let me show you show you
22:19 this one.
22:23 Right. Looks great. That's our new
22:26 That's our new logo
22:28 because I'll show you because
22:32 the technology has gotten better since
22:34 the first time I made this this thing.
22:37 Um
22:40 the like the the uh the text on the on
22:43 the marquee up here and the text on the
22:44 ticket's actually correct. This is a
22:47 little [ __ ] up. It says like limit
22:49 limited dough
22:52 but but that looks pretty cool. And then
22:54 if I hit play
23:01 cool, right? Uh Mimi Kyle, this is
23:04 giving me cognitive dissonance. I know.
23:07 I know.
23:10 So, so, so part of the daily practice
23:18 is being intentional about the work you
23:20 do. So, what I would have historically
23:22 done and and the way, let me show you.
23:24 I'm going to launch iMovie.
23:28 It's giving me cognitive dissonance. I
23:30 know. Listen, I can be
23:35 I can be organized.
23:38 I choose not to be right. I don't I
23:41 don't like it. Like I like organizing
23:44 things. I don't like the ongoing
23:46 organizing of it. One of the reasons I'm
23:48 excited about notion if I do it right
23:53 is that I won't be in that situation
23:54 where I'm like, where was that thing?
23:57 Where was that thing? I know I did a
23:59 thing. Didn't I do a thing? Sometimes
24:01 I'll say, "Didn't I do a thing?" and
24:04 it'll take me like an hour to find it
24:06 and then I'll realize not only was it a
24:07 thing, but it was a thing that I spent
24:09 three or four days on and it's literally
24:11 just scattered across my digital
24:15 storage facility. Um Vicki says, "You
24:19 can now put videos in notion, too." Very
24:21 cool. That's great. You can now put
24:23 videos directly in Sydney. That's super
24:25 cool.
24:26 Um, okay. So, this is So, let me go back
24:29 to projects here.
24:36 This one.
24:43 So, I'll play a little of this here.
24:52 You can hear that, right?
24:57 See?
25:10 Okay. So, see this shot right here?
25:17 See all the white people?
25:20 So, one of the comments I got was, "Can
25:23 we make it a little less white?" It's
25:24 New York City.
25:29 Um, so that was a comment I got. But
25:32 when when I made that original video, I
25:35 went into Lumal Labs. I think I might
25:37 have even even done some of them here on
25:38 the live. I went into Lumalabs
25:42 and I just downloaded all the I made a
25:45 bunch of video clips and then I just
25:47 dragged them over to iMovie with the
25:49 song behind it and I stuck them in there
25:51 and I just did quick edits
25:53 and I just I didn't none of those raw
25:56 assets are saved anywhere. I have no
25:58 idea where they are. They're probably in
25:59 the iMovie thing somewhere. They're
26:00 probably somewhere else. That kind of
26:03 chaos
26:05 serves me in the moment, but it doesn't
26:07 serve me long term.
26:09 This is a project I want on [ __ ]
26:12 Broadway,
26:14 right? I don't want this like in in the
26:18 the local community theater in
26:20 Wheatidge, Colorado. I want this on
26:23 [ __ ] Broadway.
26:25 So,
26:28 if I'm halfassing
26:30 the things I produce for it, talk about
26:33 cognitive dissonance, there is this
26:35 cognitive dissonance between those two
26:37 things. So part of the intentionality of
26:39 a daily practice is if you really want
26:42 to step up your game, if you want to
26:43 make more money at AI, right? A lot of
26:46 people are like, "How do I make money
26:47 with Chachi?" Right? The famous Wait,
26:50 where's my button? I don't even know
26:51 where my button is.
26:53 I do.
26:57 >> Right. All those [ __ ] promise you
26:59 you
27:01 What's that require?
27:04 It requires you being excellent.
27:07 And what does excellence look like? I
27:10 don't know yet.
27:12 One, one part of excellence is showing
27:14 up. That I've got nailed.
27:17 The next part is show up with
27:19 intentionality
27:21 and understand what you're trying to
27:22 accomplish. And Cindy Coon's talk about
27:25 come up with three themes. Pick three
27:27 themes. Three.
27:30 And do everything within those themes.
27:35 The the batteries on that button have
27:37 outlived many AI startups. It's true.
27:42 >> I've never changed the batteries on
27:44 this.
27:47 We got some mileage out of that. Hang
27:48 on. Dog wants out again. You want to You
27:51 ready to go out for real this time? Come
27:53 on.
27:55 Go.
28:12 So,
28:29 huh? Huh?
28:36 I'm I'm I'm I'm I'm literally trying to
28:39 figure out
28:41 I was just sitting here thinking
28:44 if I'm going to recreate this video.
28:48 What would be really nice is if I had a
28:50 folder with all the images that were the
28:52 start images for this video, but I
28:54 don't. I think I actually just made them
28:56 all in
28:59 I don't know where I made them. I think
29:00 I made them in chat GPT. So like this is
29:05 something. Okay, so here's something
29:06 that I've thought about for a while and
29:08 this this affects me at at StoryVine at
29:11 my at my uh company.
29:13 Um,
29:17 a lot of times when I'll put together a
29:19 project, and I know I know you all know
29:21 this,
29:23 if you're doing anything in AI, you've
29:25 experienced this, where you're like, you
29:27 you dynamically do tool switching on the
29:29 fly. Okay, this this video tool isn't
29:32 working. Let me go try that one. Let me
29:34 try this image tool. Let me try this
29:35 audio tool. Let me try chat GPT. Now,
29:38 that didn't work. Let me go over to
29:39 Claude. And so what I end up with is I
29:42 end up with this Frankenstein kind of
29:45 chain of craft
29:48 where the final product was okay I wrote
29:51 the thing in claw I put it into nano
29:54 banana it turned into something and you
29:56 know I went put it into notebook LM and
29:58 then we're there. So the the final chain
30:00 of craft is clear but there were like 17
30:03 tools to get there.
30:07 notion would be a perfect place to say,
30:10 "Here's that chain of craft, right?" Um,
30:13 and just throw it in there. Just have a
30:14 thing that you throw in there and then
30:16 you'll be able to find it again in the
30:17 future.
30:19 That requires some discipline. That
30:21 requires setting it up. That requires
30:22 some practice. But like, that's a thing
30:25 that the next time I go to do a project
30:27 [ __ ] me up because I spend a ridiculous
30:30 amount of time trying to find all this
30:31 stuff. Okay.
30:35 Good lord. Good lord. Good lord.
30:40 All right. What's going on? What's going
30:41 on in the chat? Chatty chats over here.
30:44 You can put your videos in there. Vicki,
30:46 does not connect with Okay. I don't
30:48 know. Okay.
30:51 So,
30:54 so the first thing we need to do
31:11 Now, can I
31:15 I do it? Let's see.
31:26 Okay. So, there's that and there's that.
31:35 Okay. So, that's fine.
31:43 And then I jump to that scene
31:47 and we've got an old Sydney logo, right?
31:53 So, what I need to do is go probably
31:56 there.
31:59 Oh [ __ ]
32:11 Yeah.
32:18 So, like here's a good this is a good
32:19 example of I don't quite know how to do
32:21 this.
32:24 This is this is absolutely one of those
32:27 moments where where I could just go,
32:30 "Oh, [ __ ] it." and just just say, "This
32:32 is good enough. I'm not going to take
32:35 the time to do this," right? And that is
32:38 the thing we want to battle against.
32:41 And it's not even a battle.
32:45 The point of a daily practice is like
32:48 understand what you're trying to
32:49 accomplish.
32:52 Just cut the lines, right? If you're in
32:55 a bar, you just Someone's got to cut the
32:57 lines, right? We all want to be Tom Cruz
33:00 and cocktails spinning bottles around.
33:03 I have no idea what I'm doing and it's
33:05 great. Yeah. Yeah. Embrace that. Embrace
33:09 the jank. Embrace feeling clueless.
33:12 Embrace
33:14 that feeling of ickiness.
33:18 Because where you've got that feeling of
33:20 ickiness, where you've got that fear
33:22 that's like I don't want to deal with
33:23 it.
33:25 On the other side of that is there's
33:27 some [ __ ] magic,
33:29 you know? Tik Tok pin Vicki. Oh, YouTube
33:32 Vicki. No, not notion search is so good.
33:36 Good. Exactly. That's like my instinct
33:39 seeing just seeing what it is. If I I
33:41 don't even know what to call it. It's
33:43 like a a weird combination between a
33:45 notetaker and a scheduling tool and a
33:48 and a [ __ ] relational database.
33:51 It's it's notionally
33:56 like some combination of those.
33:59 Um
34:01 but yeah, like I don't want to I don't
34:03 want to like not know exactly what to do
34:05 in front of all you people. That's
34:08 [ __ ] gross.
34:10 Why? Oh, I probably have to click.
34:13 Yes. Okay, great. So, now what I'm going
34:16 to do is I'm gonna What happened?
34:20 Oh, I rolled over. Oh my god. Okay.
34:24 Can we Can we not Okay, we're good.
34:29 Okay, so I'm going to make a
34:32 a screenshot of this.
34:37 Can What just happened to my screen? Oh,
34:40 it's Why is it so dark for you? Oh, cuz
34:45 you're not seeing
34:47 You're not seeing what I'm seeing. I've
34:48 got a I've got like a little bright live
34:50 window that you're not seeing. Okay,
34:52 it's a floating window that's not
34:53 showing up, which is weird. Okay,
34:55 whatever.
34:57 Okay, so I just took a screenshot of
34:59 that thing. So, now
35:01 I'm going to go over to chat GPT because
35:04 I did do a little testing on this
35:05 earlier.
35:07 All right. So, I said this this was
35:11 um
35:13 there's the original. See, there's the
35:15 original with the old logo
35:17 and the dude the dude back here and the
35:20 old the old face.
35:22 So, I said there's that picture and then
35:24 here's the new poster with the new face
35:27 and the new logo.
35:31 And I said, you know, make the ticket
35:33 look like the ticket.
35:36 And then it ended up, I guess, there.
35:40 There we go. It's which this is still
35:42 not quite right now that I look at it.
35:46 Um,
35:50 should I be excellent? I should probably
35:52 be excellent, shouldn't I?
35:57 All right, let me go. Dang it. Uh um
36:06 I'm going to say
36:08 notice
36:10 in the
36:13 original poster I uploaded.
36:22 The standing man
36:25 is
36:27 white,
36:28 not gray.
36:38 Also notice
36:41 that on the ticket
36:46 there is no man at all.
36:51 Please
36:54 update
36:56 the ticket
37:00 design
37:03 and
37:05 marquee design
37:11 to reflect
37:16 this detail.
37:23 And then I and I think what I'm going to
37:24 do is I'm going to go add
37:32 the thing back in. And by the way,
37:35 because I organized
37:38 my um because I organized my folder,
37:43 sorry, Moo. No. Um, because I organized
37:47 my folder, I know exactly where the
37:49 latest posters are, right? Like, this is
37:51 the thing. Having a a horizontal poster,
37:56 a vertical poster, and a square poster
37:59 that are the latest graphics with the
38:01 latest logo in a single place. I know
38:03 this seems [ __ ] ridiculous to you,
38:06 right? That it's a big deal for me to
38:09 have the latest things in a folder I can
38:11 actually find.
38:16 One of the joys of having an ADHD brain
38:20 is that you can move fast and you can
38:22 keep track of a lot of [ __ ]
38:25 It's counterintuitive, right?
38:29 So, I have a a good enough brain to go,
38:32 "Oh, I know. I saved it. I know I saved
38:34 the latest one probably to my desktop,
38:36 but if I don't have it there, I probably
38:37 have it in Andrew's tweet
38:40 chat thingy he sent me.
38:44 But there's so much cognitive horsepower
38:47 that is spent just trying to sift
38:49 through the chaos of my life that that
38:53 doesn't make sense anymore, right? Okay.
38:56 So, there's our new logo there. There's
38:59 our new poster. So, notice in the
39:01 original poster I uploaded, the standing
39:03 comma, the standing man
39:06 is white, not gray. Um,
39:11 he is also to the right of the word
39:18 Sydney,
39:21 not above it.
39:26 Okay. I also noticed that on
39:30 your ticket design,
39:33 your ticket design,
39:37 there is no man at all. Okay. Boom.
39:40 Let's see if we can get this thing to
39:42 not suck. And and this
39:46 trying to get AI to like behave
39:49 is a nightmare. I was trying to do this
39:50 with Nano Banana earlier and it was it
39:52 was a disaster. Would Nano Banana not be
39:56 a little better for slight changes? It
39:58 might be, but it wasn't this afternoon.
40:00 It kept giving me it would it would
40:02 redesign the marquee and it just left
40:04 the the front ticket alone and I
40:06 couldn't I couldn't get it to change the
40:08 design of the ticket. But but the image
40:11 that I put in here as the base image was
40:13 an early iteration of a nano banana
40:16 image. So maybe I just need to go back
40:17 and forth between the two.
40:20 Okay.
40:22 Um, I use my recent shortcut like a
40:26 lifeline, but it's not working well
40:28 lately. Yeah, exactly.
40:31 Exactly, man. This is this is uh this is
40:34 the ick. This is the ick.
40:40 And I don't know, there's a part of me
40:42 that's like, do I even want to [ __ ]
40:44 deal with any of this? But well, that
40:47 sort of worked.
40:48 So, we could say um on the marquee
40:56 the white
40:59 the man is
41:02 above the Y in Sydney.
41:09 It should be to the right of it. But
41:14 notice we now have the man on the
41:16 ticket. So that's cool.
41:22 The other thing about this, what I'm
41:24 doing right now is Oh, Champy's back.
41:27 Little [ __ ] Um,
41:31 in this final video, this is going to be
41:34 seen for about a quarter of a second.
41:40 Does it matter that the little dude
41:42 isn't standing there on these? No. But
41:44 is that excellence? No.
41:48 Where do you draw the line? I don't
41:49 know. Like like this is this is the
41:51 whole thing about being part of a
41:53 practice. There's
41:55 there's learning. Can you get the tools
41:57 to do what you want them to do? But
41:58 there's also like what is the standard
42:01 that you're producing to
42:04 if I want to produce for Broadway
42:07 or even off Broadway for [ __ ] sake
42:09 anywhere in New York, you know? So that
42:14 didn't fix it at all. In fact, I think
42:16 it made it worse. Made the color worse.
42:19 So, here's what we're going to do. We're
42:20 going to take this non6 by9 image.
42:24 Copy this image. We're going to go to
42:25 nano banana. We're do a new chat.
42:30 We're going to paste this in.
42:32 I'm going to say in this image
42:37 the
42:46 Sydney logo on the marquee
42:51 has
42:54 the white
42:58 male figure above the Y.
43:06 It should be
43:09 to the right of the Y,
43:17 not above it.
43:21 Please fix.
43:25 And you know, it's also very possible
43:27 that
43:28 this kind of shorthand prompting that I
43:30 do on the AI learning lab live,
43:33 this might not be
43:36 appropriate for the level of work I want
43:37 to do. Like I might I might need to be
43:39 doing more structured prompts, which we
43:41 can go fix. Like that's not a problem to
43:42 solve, but it's got to be solved.
43:48 It's got to be solved, everybody.
43:57 I still remember the first time I took a
44:00 picture of my notes and chat GBT read.
44:02 Oh, that's something I wanted to do
44:03 tonight. Actually, that I'm glad you
44:05 said that.
44:07 We'll we'll we're going to do something
44:08 on that front, too.
44:11 So, okay. So, look what look what Nano
44:14 Banana did. This is the kind of [ __ ]
44:16 that So, now it's just Sid with the dude
44:19 and the Y and then it's DN. The DNE
44:22 show. Hey, did you go see DNE on
44:25 Broadway? No.
44:31 God damn it.
44:34 [ __ ]
44:42 Um,
44:52 you didn't
44:57 change anything.
45:00 Look at this
45:05 layout
45:06 in the
45:11 uploaded poster and make sure
45:18 the marquee
45:22 matches it
45:26 there. There's also what I'm about to do
45:29 is I'm about to bail and go do it in
45:31 Photoshop.
45:34 Nanobanana has rarely done that sort of
45:36 thing to me. You know, archetypal, I'm
45:38 wondering, do you use Nano Banana in uh
45:40 in AI Studio? I wonder if the Nano
45:43 Banana in AI Studio is different than
45:45 the one on Gemini because I was that was
45:47 on Gemini.
45:49 Um,
45:51 so let me upload my poster again.
45:55 I mean, this seems [ __ ] ridiculous
45:57 that you have to keep uploading the
45:58 poster
46:00 because it's got it in its [ __ ]
46:02 memory, right? It's got it somewhere.
46:04 It's somewhere on an OpenAI server. My
46:06 little [ __ ] picture is sitting.
46:10 Okay.
46:20 I'm almost exclusively you using it on
46:23 Gemini or through weevi. Okay. Yeah,
46:25 it's weird. I don't know. I don't know
46:26 what the deal is.
46:32 The other thing I've noticed about in
46:34 the in the grand total of two days, day
46:37 and a half since I've been doing this is
46:39 the time for me goes by really fast. And
46:42 I would imagine the time for you is like
46:44 this is probably like watching paint
46:45 dry. So we'll we'll try to find a
46:48 balance of that. But it it might also
46:50 just be that this is a it's a different
46:53 kind of compelling
46:56 to work on stuff. Um,
47:06 and by the way, this is one image of I
47:09 don't know, probably a doz Oh, it's so
47:12 close.
47:19 Okay.
47:21 Scooch
47:22 the man to the right
47:26 on the marquee
47:30 so
47:31 it isn't over overlapping
47:35 the Y
47:40 in the word
47:43 Sydney.
47:46 Uh uh uh uh uh uh uh. Oh, and you know
47:48 what else is going to probably [ __ ]
47:50 happen as I'm doing all these iterations
47:53 is at some point it's going to go, "Oh,
47:55 sorry. You're out of credits, Mr.
47:56 Shannon. Mr. Shannon, Mr. Shannon."
47:59 Yeah. Listen, um we noticed that an ex
48:03 excessive amount of uh token generation
48:07 under your account. You're only paying
48:09 us $20.
48:11 and we were kind of, you know, we kind
48:13 of were thinking maybe, you know, you
48:16 could not make so many. So, what we've
48:18 gone ahead and done is we've just turned
48:20 off your ability to continue doing your
48:22 work. I hope that works for you. We we
48:25 really value you as a customer, Mr.
48:28 Shannon. It's just that the excessive
48:30 use of tokens from our system is really
48:34 kind of making us a little cranky, if
48:36 you know what I mean.
48:51 All right. So, here's a question.
48:54 Do you Do you [ __ ] punt?
49:00 No, I'm not going to punt. Here's what
49:02 I'm going to do. I'm gonna open up
49:03 Photoshop.
49:05 Do I want to open up Photoshop? I don't
49:07 think I want to open it on that one. I
49:08 think I'm going to want to open it on
49:09 this one.
49:11 But is the color different? No. Fine.
49:14 So, we're going to copy this image
49:18 and open up Photoshop.
49:21 Oops. That's photos.
49:23 Photoshop.
49:27 Yes, it's Mr. Shannon. Mr. Shannon. Mr.
49:29 Shannon. Yes. Yes. Yes. No, I was There
49:32 were some questions in the crowd. Uh
49:35 really curious, deeply wondering pretel.
49:40 So by chance, what uh are your
49:43 qualifications?
49:44 Because
49:47 it appears uh
49:50 the credentials are thin.
50:06 Never punt on first down. Great point.
50:10 Never punt on first down. And we are
50:11 absolutely in first down here. By the
50:13 way, I can see it right now. This is not
50:16 a 16 by9 image.
50:19 Did you see that? Gro Gro
50:22 Gro can now output images and videos in
50:26 all of the major aspect ratios including
50:29 16 by9 which is what all of your TVs are
50:32 and what all of your Tik Toks are
50:35 9 by6 16 by9
50:39 chat GPT still cannot do 16 by9 outputs
50:44 absolutely [ __ ] staggering to me okay
50:48 let's see let's See, here's what we're
50:50 going to do, people. Here's what we're
50:53 going to do. All right. We're going to
50:55 zoom in.
51:10 I'm going to grab that. And then I'm
51:15 going to
51:16 minus select. Can you all see this what
51:20 I'm doing? Yes. I'm going to minus
51:22 select that.
51:25 Oops. What did I just do? What did you
51:29 do? What did you do?
51:33 What?
51:36 That's bizarre.
51:39 Oh, I see.
51:41 30 pixel feather will do it every time.
51:45 Yeah. Yeah. See? Yeah. You puga. Yeah.
51:49 And another thing, Johnson, get your
51:51 dick off my desk. All right.
51:59 Copy. Okay. So, now what we're going to
52:02 do is we're going to select
52:06 the dude and we're going to use the one
52:08 thing that Photoshop seems to be good at
52:13 is removing [ __ ] So, I'm going to say
52:15 remove, man.
52:18 with generative fill.
52:22 And I've got I copied the man. So I've
52:23 got the man on the clipboard. Oh no,
52:27 it didn't remove it. All right, fine.
52:30 Here. Oh, you know what we'll do? Oh my
52:34 god. I can't this this. It's unfucking
52:36 believable that I have to [ __ ] resort
52:39 to like 20 23 Photoshop [ __ ] skills.
52:44 Like really?
52:47 You're kidding me, right?
52:50 I got to copy and paste stupid [ __ ] for
52:52 your stupid
52:54 cuz you can't figure out how to make
52:56 your stupid [ __ ] Follow my stupid
52:59 instructions. Flatten image
53:03 marquee tool.
53:06 I don't know what I'm doing here. I'm
53:09 supposed to use skills. It's not what I
53:12 signed up for. This is not what I signed
53:14 up for.
53:15 Okay.
53:22 Does that got to go down? I think it
53:24 does. Right.
53:26 Ah, this is so exhausting.
53:35 Okay, there. It's mostly gone. Now I got
53:38 to get the rubber stamp [ __ ] tool.
53:41 This is not what I signed up for. This
53:43 is not what I signed up for. You want to
53:45 know why we can't have nice things?
53:50 Cuz daddy's having to [ __ ] Photoshop
53:52 stuff when we have magic robots that
53:54 just make stuff. Okay. That's why
53:57 daddy's not happy.
54:03 Oh, god damn it.
54:05 Um,
54:07 hang on.
54:12 Uh oh. What's going on? Is it still
54:13 Monday?
54:16 Well, this is what happens when you get
54:18 what? Just [ __ ] should work. [ __ ] should
54:20 just work.
54:23 This is not my fault.
54:26 Everybody needs a cranky Gen Xer in the
54:28 corner office.
54:31 That's what it's like. That's very
54:32 triggering. He's very angry. Why is he
54:34 so angry at Photoshop?
54:37 Oh, that looks like [ __ ] Looks like dog
54:39 [ __ ] Just looks like H. Okay, so here's
54:43 what I'm going to do now. I'm going to
54:44 use generative fill to smooth the
54:46 gradient.
54:48 Okay, generative fill. Smooth
54:52 gradient.
54:54 We're We're using six trillion tokens to
54:57 smooth my gradient because I don't have
54:59 rubber stamping skills
55:02 or patience.
55:04 Oh my god. Seriously.
55:07 Seriously
55:17 smooth
55:18 gradient
55:20 only.
55:28 And if you're if you're sitting there
55:31 judging and thinking, is anyone going to
55:34 see this? Is anyone going to notice
55:35 this? No.
55:42 Oh [ __ ] that was the wrong All right,
55:44 there we are. Okay, D.
55:47 Uh, let me get the rubber stamp tool.
55:49 I'm I'm just going to rubber stamp it
55:50 shittily and then we're just going to
55:52 move on with our life.
56:00 Oh, I [ __ ] hate this. Okay, there we
56:03 go. That's good enough. Now paste.
56:04 There's our dude.
56:09 Right.
56:10 Beautiful.
56:12 Beautiful. Beautiful. Beautiful. And I
56:15 think what we're even going to do,
56:17 let me just make it a little like And
56:20 then
56:26 there we go. And I think we've got to do
56:28 this a little opacity
56:31 opacity like so that it matches.
56:42 That's probably better, right? Do you
56:46 see what we did there?
56:48 Look how excellent we are.
56:51 Our quality of excellence here is just
56:54 crazy. And if you're like, "But the
56:56 man's a different size relative to the
56:58 Sydney." I know. I know. And the Sydney
57:01 logo is not actually correct.
57:03 Oh my god, it's exhausting. But this but
57:05 this doesn't matter. Okay. Um,
57:10 let me go layers. Let me go flatten
57:12 image.
57:16 And here's what I'm going to do.
57:21 Oh, the other thing I need to do is I
57:23 need to turn this into a 16 by9 image.
57:45 Okay. So, now we have a 16x9 image
57:48 that's got mostly our logo correct. And
57:51 we're going to go export as PNG.
57:56 And where are we going to save this? Not
57:58 to the desktop.
58:04 So, we're going to go to posters. We're
58:05 going to go podcast 2.0.
58:09 There's nano banana test. So, we're
58:11 going to go
58:13 new folder,
58:16 poster
58:20 2.1
58:23 images. Uh, let's see. Um, oh, let's
58:26 see. Poster 2.1
58:30 start images.
58:35 I think that'll make sense. Maybe I've
58:38 got to remember I called this poster 2.1
58:40 and then we'll just go we'll call this
58:41 ticket.
58:43 Okay, beautiful. Save. Great. Next.
58:49 Just like 12 more of those to go.
58:52 Um,
58:55 got to go back here.
58:57 Got to get this screenshot again.
59:01 That. And then we're going to go here.
59:04 Okay. Okay, people. Fantastic.
59:08 Fantastic. Here's what we're going to
59:10 do. All right. We're going to paste in
59:13 that image that we just screenshotted
59:15 out of the old video project. And we're
59:17 going to say replace.
59:20 That thing failed to upload. Of course
59:22 it did. Okay. Yes, that's the ticket.
59:26 Acting.
59:28 It was only acting indeed. Fantastic. I
59:32 sneaked out of watching a movie with the
59:33 wife to come here. I'm so sorry.
59:35 Generative Phil is a good Tik Tok name.
59:38 It actually is a really good really good
59:40 Tik Tok name. Somebody do that. Somebody
59:42 take Generative Phil. Kyle, how many
59:44 tabs do you have open? Listen, that's
59:46 not appropriate. That's not an
59:47 appropriate question. Cam Ken from
59:49 Cleveland. 1 2 3 4 5 6 7 8 9 10. But
59:53 three of them are X. 12 13 14 15 16 17
59:58 18 19 20 21 22 23 24 25 26 27 28 29 I
1:00:03 don't know 30ish, which is low for me.
1:00:06 If I'm not in the 70s, I'm not I'm not
1:00:08 pinning the memory requirements of my my
1:00:11 uh computer.
1:00:13 Okay, let's see. Um, update
1:00:20 update
1:00:22 the the marquee
1:00:26 in this image with the new Oh, maybe I
1:00:32 should do this over in nano banana. Nano
1:00:34 banana.
1:00:36 Nano banana.
1:00:40 Let me go try it. Nano Banana
1:00:44 because I think I think Nano Banana
1:00:46 generally is better, but it it's been
1:00:48 really acting up for me today. I don't
1:00:50 like it at all.
1:00:52 Okay. Oh, wait. And I've got this. The
1:00:55 other thing that drives me crazy about
1:00:56 Nano Banana, you have to put in
1:01:01 the last image that you upload has to be
1:01:03 the aspect ratio that you want.
1:01:07 That's something that they need to fix
1:01:09 that stupid [ __ ] bug.
1:01:12 Okay, there's Sydney.
1:01:15 Okay, so let's see. Let's see here.
1:01:19 We're going to go paste. So that's the
1:01:21 new poster in the
1:01:26 um
1:01:28 photo,
1:01:31 replace the
1:01:34 poster
1:01:39 in the marquee
1:01:43 above.
1:01:46 Times Square
1:01:48 with the new poster
1:01:52 uploaded here. All right, let's see what
1:01:56 we get, people.
1:01:59 Let's see what we How many tabs do you
1:02:01 have open?
1:02:20 All right, that's not bad. It looks a
1:02:22 little Why does it look a little
1:02:32 Let's see.
1:02:33 It looks
1:02:35 like you
1:02:38 um compressed
1:02:41 it vertically
1:02:45 to fit.
1:02:48 Don't do that. Also, it should match the
1:02:54 color
1:02:55 grading of the
1:02:58 original photo.
1:03:07 Dr. J, do you use AR glasses at all? I'm
1:03:09 interested in trying them out.
1:03:13 Be back.
1:03:23 So, archetypal architect, this is what
1:03:26 um my experience has been with Nano
1:03:28 Banana. There's the image that I did and
1:03:31 I gave it comments like you squashed the
1:03:34 poster. I don't want it squashed. I want
1:03:35 it cut off.
1:03:37 Um, and it's too bright relative to the
1:03:40 rest of the image. And then what it pro
1:03:42 what it produces is the exact same
1:03:45 [ __ ] image. It doesn't change a
1:03:46 thing. Sometimes it'll add an extra
1:03:49 Gemini watermark,
1:03:52 but I don't I don't I don't I don't I
1:03:54 haven't been having good luck with with
1:03:56 Nano Banana. It's it's it's [ __ ]
1:03:59 maddening.
1:04:01 All right. So, let's try it here. We'll
1:04:03 try it here. We'll do it live.
1:04:11 Okay. Update the marquee in this image
1:04:13 with the new poster. Um,
1:04:17 do not um
1:04:23 how do I put this? Uh
1:04:27 notice that the poster is cut off in the
1:04:34 original.
1:04:36 This is okay.
1:04:40 Don't
1:04:42 change the aspect ratio
1:04:47 of the poster of
1:04:51 the poster to um
1:04:55 to uh
1:05:00 be complete
1:05:03 the image. I think that's probably bad
1:05:05 prompting.
1:05:10 Kyle, do you use AR glasses? I've used
1:05:12 meta glasses, but they are not what I
1:05:16 want. I have not. Here's the reason why
1:05:18 I got I got the when when Google Glass
1:05:21 first came out. I I was not a glass hole
1:05:25 because I didn't wear them in public. I
1:05:27 I spent I don't I think it must have
1:05:29 been Storybine money. I spent $1,500 on
1:05:32 those stupid [ __ ] glasses
1:05:35 and did what I did with the with the
1:05:38 with the Rabbit. I I I used the the R
1:05:41 rabbit R1 far longer than I used Google
1:05:44 Glass because I tried them on and I
1:05:45 realized the software sucked for them.
1:05:47 They weren't good and they're just they
1:05:49 just weren't ready for prime time. So, I
1:05:50 just didn't use them. I that's where
1:05:52 I've been with with AR for a while. um I
1:05:56 don't have a a a strong uh I don't have
1:05:59 a strong need for it in my life. Right.
1:06:01 And in your case, I think you've got
1:06:03 some issues and and you have a need for
1:06:05 it. That's
1:06:08 like I I just don't I don't when I'm
1:06:12 going to get meta glasses is when
1:06:17 they're just [ __ ] good. like and and
1:06:19 apparently the metal ones are actually
1:06:21 pretty good right now, but they're not,
1:06:23 you know, they're not great. Why did it
1:06:25 make the dude
1:06:28 gray again and put him above the thing?
1:06:32 Um,
1:06:37 you changed
1:06:39 Oh, you know why? because we're in a
1:06:40 because we're in a [ __ ] other pro
1:06:43 prompt thing that because it did it once
1:06:46 in here and so it's copying it again. I
1:06:53 I am not happy.
1:06:57 Unbelievable. Add photos and files.
1:07:02 These things
1:07:06 this makes notion look easy. Yeah. Well,
1:07:09 this is like this is the ick. This is
1:07:11 the ick, right? Every every fiber of my
1:07:16 being wants to throw my computer out the
1:07:18 window right now. Like, I don't want to
1:07:20 do this. And this is very often what
1:07:23 happens to me is I will get excited
1:07:26 about and listen, this has happened my
1:07:28 whole life. I'd get excited about
1:07:30 something and and in in the olden timey
1:07:32 days, I'd get excited about something
1:07:34 and then I'd realize, oh, if you want to
1:07:35 do that, you have to go learn computer
1:07:37 programming. And I'd have to go to like
1:07:39 the Barnes & Noble and buy a book
1:07:43 before YouTube came out.
1:07:46 So, it's much easier now. If you want to
1:07:48 do something, you can just have AI do
1:07:49 it. But then, if you want to do it good,
1:07:53 good lord. Good lord. Okay. Um, please
1:08:00 replace the poster the the the
1:08:07 uh marquee
1:08:10 image with the new poster
1:08:15 uploaded here. Please replace the
1:08:19 marquee image in the photo
1:08:24 with the new poster uploaded here.
1:08:28 Um,
1:08:30 match the color grading
1:08:34 and tone. Tone is the same thing as
1:08:36 color grading. The color grading and
1:08:39 texture
1:08:41 of the original
1:08:44 photo.
1:08:46 Okay. How do you think I got in this
1:08:49 pickle suit, man?
1:08:52 Oh my god, I remember Barnes & Noble
1:08:55 days.
1:08:58 Listen to me. Two things. One, start a
1:09:01 new chat with Nano and keep revising
1:09:03 your first prompt. Ah,
1:09:06 if you draw on the image with Photoshop
1:09:09 pointing at [ __ ] Oh,
1:09:12 future of food coming in hot. Okay,
1:09:15 here's the difference. This is really
1:09:18 good. Chef Kelly, I think that's Is that
1:09:21 Chef Kelly? I think that's Chef Kelly.
1:09:25 Is that Chef Kelly? Yes. Okay. Um,
1:09:31 in the olden timey
1:09:33 AI learning lab,
1:09:36 I would be saying, "Here's how you learn
1:09:38 things." And now I'm just being
1:09:42 every man trying to figure this [ __ ]
1:09:44 out.
1:09:47 And you guys have the skills. I like
1:09:49 this. I like this a lot.
1:09:53 Why does it insist on putting
1:10:03 Unbelievable
1:10:05 unbelievable?
1:10:08 Here's what we're going to do.
1:10:11 Here's what we're gonna do. Oh, this is
1:10:13 [ __ ] exhausting. Okay, copy image.
1:10:18 We're going to Photoshop.
1:10:20 Go to new.
1:10:24 Paste.
1:10:26 You go open.
1:10:29 Open from your computer.
1:10:47 Select all. Copy.
1:10:55 Paste.
1:10:59 Why did it not paste?
1:11:02 Select all. Copy.
1:11:04 Go over here.
1:11:08 Paste. There we go.
1:11:13 Transform. Scale.
1:11:17 You're like, couldn't you use AI for
1:11:18 this? That's what I said.
1:11:40 Oh, wait. Did he make is that the
1:11:43 vertical poster? The dudes on top. I
1:11:45 think that might be right. Oh, well then
1:11:48 in that case we can go back. I think
1:11:50 then chatbt got it right. Hang on. I
1:11:53 might be being an [ __ ] to
1:11:58 um
1:12:05 Nah, it still did it wrong.
1:12:08 But yeah, this is the Okay, this is
1:12:11 okay. This is good. This is good.
1:12:15 So, what we're going to do, let's see.
1:12:23 We're going to go like that. We're going
1:12:24 to go like that.
1:12:27 We're going to hold down the shift key
1:12:28 and compress it a little bit.
1:12:32 And then we're going to go transform
1:12:36 free transform.
1:12:39 No. Then we're going to go transform
1:12:44 distort. I think that's right.
1:12:48 We go. All right. Yeah,
1:12:52 here we go.
1:13:15 Okay,
1:13:17 I think that's good. Okay. Now,
1:13:21 now we got to color grade this puppy.
1:13:24 So, we'll do
1:13:26 pull out the saturation. I like the
1:13:29 saturation. We will
1:13:32 change the hue a little bit. There we
1:13:35 go.
1:13:36 And we'll go lightness. And then we'll
1:13:39 And we'll go levels.
1:13:42 Yes.
1:13:46 We'll go
1:13:58 See, that just looks not the same.
1:14:02 And then we're going to go filter noise.
1:14:07 Add noise.
1:14:10 We'll go. Fine. That's good. That's
1:14:13 better.
1:14:15 And then we will do one more level.
1:14:18 Oh, we could do curves. Let me do
1:14:20 curves.
1:14:22 Image adjustment. Curves.
1:14:25 Oops.
1:14:27 Image adjustment. Curves.
1:14:31 Curves. There's the curves. There we
1:14:33 are. Yeah. Hey, listen, Kyle. Yeah.
1:14:36 Fantastic what you're doing there. It's
1:14:38 really good. Yeah. If you could just go
1:14:41 on ahead. Hey, could you try it again?
1:14:44 Maybe this time is if you had talent.
1:14:47 That would be great. Be fantastic.
1:15:07 All right, I think that's good enough.
1:15:08 We can go ahead and flatten that image.
1:15:10 Let's flatten that on up. We're going to
1:15:13 flatten it on up. We all do that. Oh,
1:15:16 that's fantastic.
1:15:18 I think Kyle's done lost his mind. Okay,
1:15:21 this is uh Times
1:15:24 Square.
1:15:27 Uh what's that? Uh TKTS. TKTS
1:15:32 steps.
1:15:38 Look at that. We did it. We did it. Yay.
1:15:41 Ship it. This is better than SNL.
1:15:48 Charles Montgomery Barnes died in 1927.
1:15:52 Clifford Noble died in 1932. That's
1:15:54 hilarious. Doesn't always work.
1:15:58 Um,
1:16:00 but it keeps your pen marks on it. Oh,
1:16:02 that's funny. Okay. So, that's that. So,
1:16:05 that's good now, right? I think that's
1:16:07 good.
1:16:09 Yeah, this is good. All right. Now, back
1:16:11 to this
1:16:13 and we're going to go we're going to get
1:16:15 to our next image. So, there
1:16:27 maybe not.
1:16:29 Maybe it's this image. Yeah.
1:16:33 Okay.
1:16:39 All right.
1:16:40 There's that.
1:16:44 Let's go to Where are we going to try?
1:16:46 Where are we going to try, people?
1:16:49 Oh, that's wild how how darker that
1:16:51 image is. Oh, no, it's not.
1:17:01 Huh. Well, whatever. I don't think it'll
1:17:04 matter in the end. Okay. Um,
1:17:08 let's try Gemini.
1:17:10 We're going to go for it, Gemini. We're
1:17:12 going to give you a shot.
1:17:14 We're going to go create image.
1:17:17 We're going to go
1:17:19 upload files.
1:17:22 Sydney,
1:17:23 the dude is on top of the Y. Okay. So,
1:17:25 that was not that was not their [ __ ]
1:17:28 That was my [ __ ] for not using my
1:17:30 eyes. I'm going to say replace the
1:17:35 Times Square
1:17:38 marquee
1:17:40 with this new uploaded
1:17:44 poster.
1:17:46 match the color of the
1:17:52 photo when
1:17:55 let's see match the tone
1:17:59 of the photograph
1:18:05 when inserting
1:18:08 the new poster
1:18:11 image.
1:18:24 [ __ ] That's really good. Yeah, that's
1:18:27 really good. Hey. Hey, Kyle. Yeah.
1:18:30 Listen. Listen. What you're What you're
1:18:34 doing here is really interesting. This
1:18:36 This whole idea of a of a See, Chef
1:18:39 Kelly, it didn't change the [ __ ]
1:18:40 image at all. at all. Wait, did it
1:18:44 change it at all? Did it change it?
1:18:49 No, not at all.
1:18:53 I didn't get it. I do not get it.
1:18:57 Um
1:19:00 Um Let's see.
1:19:03 Upload.
1:19:07 change
1:19:09 the marquee
1:19:13 with this image.
1:19:27 I just ate a lemon head and it's too
1:19:29 tart.
1:19:31 It's miserable.
1:19:36 All right.
1:19:41 So what happens on AI learning lab? Uh
1:19:44 well we we watch a it's um he's a man.
1:19:48 Um
1:19:50 what we we watch him um
1:19:54 try to use software.
1:19:57 What else does he do? Um he occasionally
1:20:02 uh or organizes his his files in in the
1:20:06 finder. Um
1:20:09 and uh does he teach you anything? Not
1:20:12 yet. Not yet. It's it's promising. It's
1:20:15 promising. It's we there's it's there's
1:20:18 a hope that with there's a deep hope in
1:20:20 the community. Um something will be
1:20:22 taught at some time at the AI learning
1:20:25 lab. And no, we're excited about it. We
1:20:26 really are. We're almost 3 years in.
1:20:29 We're going to get there. We're going to
1:20:30 There's going to be
1:20:33 I don't know what I'm doing with my
1:20:34 life.
1:20:39 Are lemon heads your favorite candy?
1:20:41 Good question. They're not, but I'm
1:20:44 liking them. Here's my problem with
1:20:46 lemon heads. Lemon heads are
1:20:48 inconsistent.
1:20:50 Some of the lemon heads I [ __ ] love.
1:20:52 the ones that are a little sweeter and
1:20:54 then other ones it's like they dropped
1:20:56 them three times in the sour bucket and
1:20:59 it's they're too sour. And I don't know
1:21:01 if that's on purpose that let's let's sh
1:21:04 let's shock the audience, the customer.
1:21:08 I don't know. I'm a Snickers guy. I like
1:21:10 Snickers. I really like Snickers with
1:21:12 almond. I like Snickers with white
1:21:14 chocolate. That's that's one of my
1:21:16 faves. I like Almond Joy.
1:21:19 Um,
1:21:22 I hate to say it. It's the worst candy
1:21:25 known to man. I likes me some Skittles.
1:21:29 I know Chef Kelly's here. She's like,
1:21:31 "Oh my god, don't say it."
1:21:35 Hey, look. It It gave us It gave us a
1:21:38 whole new It made the image vertical
1:21:40 now. Okay. Can I just say I hate Nano
1:21:43 Banana? I I can, right? I can say that.
1:21:46 YouTube comment. Uhoh. Wife is calling
1:21:49 upstairs. Are you coming back to the
1:21:51 movie? Uh, no. I'm watching. Yeah.
1:21:53 Learning lab. What do you watch up
1:21:54 there?
1:21:56 We've We've got on Die Hard. Yeah. This
1:21:59 is more exciting. What What is it?
1:22:02 It's a guy just kind of
1:22:06 He's trying it. We're really pulling for
1:22:09 him.
1:22:13 Unlike the mastermind.
1:22:16 And like this, the mastermind has value.
1:22:19 Exactly.
1:22:22 It's You know what the AI learning lab
1:22:24 is? It's a real underdog story. Yeah.
1:22:27 It's We We're really hoping he can pull
1:22:30 something off here at some point.
1:22:32 There's There's got to be some
1:22:34 qualification in there somewhere. So,
1:22:37 yeah. Yeah. No, I'll be down, honey. I
1:22:39 will. Just another I don't know, two
1:22:41 hours, maybe. He's got like three more
1:22:43 images to make.
1:22:49 What if gen What if generative Phil and
1:22:51 Lemonade formed a band?
1:22:54 We really are pulling for you, Kyle.
1:23:00 Lori Dreskin, I have missed this. Lori
1:23:03 Ryan, oh man, that's awesome. Okay,
1:23:06 let's go over to chat GPT and see if it
1:23:08 can [ __ ] it up for us. This is so
1:23:11 maddening.
1:23:13 H. All right.
1:23:15 poster. Go get the poster. Boom. Paste
1:23:18 in the image. Please.
1:23:23 Uh, please. Why am I saying please?
1:23:26 You're being a jerk. I'm not going to
1:23:27 ask. I'm not going to be polite to you,
1:23:30 Chat. GBT.
1:23:33 I'm not even going to call you Quinn.
1:23:35 I'm just gonna call you chat like Cindy
1:23:37 [ __ ] does. Hey, chat.
1:23:39 Hey, chat. Um,
1:23:44 replace
1:23:46 the marquee
1:23:50 image in the photo
1:23:53 with this uploaded
1:23:56 poster.
1:24:04 Mine is called Sage. Very nice. All
1:24:06 right, everyone in in in both places.
1:24:09 Um, type in the name of what you call
1:24:12 your AI girlfriend or boyfriend.
1:24:15 Sage. Mine's Quinn.
1:24:19 Fun fact. Fun fact. Chat is French for
1:24:22 cat.
1:24:26 And isn't it a no no to be polite? It
1:24:28 might be. Gemini. What about Gemini?
1:24:33 Oh. Oh. Gemini. Jiminy.
1:24:36 Geminy. Jiminy.
1:24:44 Xen
1:24:46 Mochhat. I like Mochhat. Oh, that's what
1:24:48 you call her. Gemini. Oh, that's funny.
1:24:51 That's cute.
1:24:52 I had a lot of caffeine tonight.
1:24:56 See, her choice. Oh, I like it. Co-host.
1:25:00 Now, let's see. I call mine chat GPT.
1:25:05 Hey, did it actually [ __ ] do it?
1:25:10 I think it actually [ __ ] did it.
1:25:13 But look what it did.
1:25:16 It God damn. It's because it's it's the
1:25:20 wrong aspect ratio. But that's okay.
1:25:22 That we can go fix.
1:25:26 So, we're going to go into Photoshop.
1:25:28 We're going to do new image.
1:25:32 We're going to create this.
1:25:34 And then we're going to paste it. And
1:25:36 then we're going to go up here. We're
1:25:37 going to go image size.
1:25:41 I think I'm just going to squish this
1:25:42 [ __ ]
1:25:45 1920 x 1080, right? 19
1:25:48 1920
1:25:50 by 1080. Let me lose this.
1:25:59 Yeah, that's good.
1:26:12 All right. I think that's fine. I think
1:26:14 it's a little bit
1:26:18 I think it's a little um
1:26:22 bright for the background image, but
1:26:23 that's okay because it's about our
1:26:26 [ __ ] show. So, let me export this.
1:26:29 Export. quick export as PNG. And then
1:26:32 you know what I'm going to do for you
1:26:34 all so it's not completely insufferable.
1:26:37 Instead of me just doing these images
1:26:39 for the rest of the time, I'm going to
1:26:41 go TKTS steps.
1:26:44 So I'm going to just do Time Square
1:26:49 um closer.
1:26:52 Let's go make some videos out of these
1:26:54 things. Okay.
1:26:56 So now we're going to go to flow. Here's
1:27:00 Flow
1:27:02 and we're in frames to video. Oh, so so
1:27:06 this is flow.google
1:27:08 which is Google's VO3. So I'm using that
1:27:11 right now because
1:27:14 I don't know it. I think it works.
1:27:18 It's V3.1. I'm using V3.1 fast. I'm not
1:27:22 using their high-end one.
1:27:26 Okay. Um, so we're going to go plus.
1:27:30 We're going to go upload. And now
1:27:32 because we saved them to a place that we
1:27:34 know where they are.
1:27:37 Let's see. Is it in here? Yeah. Poster
1:27:41 2.1 start images.
1:27:45 So there's our there's our our ticket.
1:27:52 Beautiful.
1:28:02 Okay, let's see. Um,
1:28:07 walk.
1:28:10 Let's see. POV
1:28:13 shot
1:28:15 walking into the theater
1:28:21 as the
1:28:23 crowd
1:28:25 mills about.
1:28:28 Um,
1:28:30 the crowd
1:28:36 should
1:28:39 be ethnically,
1:28:43 diverse,
1:28:46 like a good New York City mix
1:28:51 with a tourist or two.
1:28:57 a tourist or two
1:29:00 or two in there.
1:29:04 All right.
1:29:07 As the regular crowd shuffles in, that's
1:29:09 pretty good. I'll try that next if it
1:29:11 [ __ ] this up.
1:29:13 Did mine too. Silver Fox, I love this.
1:29:18 I absolutely can't believe how far image
1:29:20 generation has come. Well, so Lori,
1:29:23 what's what's amazing about that? So the
1:29:25 video
1:29:27 Well, that's rendering like the video
1:29:31 here
1:29:34 like that. That one says an artificial
1:29:37 love story. So, it got that right. And
1:29:38 then it says by
1:29:41 it sort of got our names right.
1:29:45 Um,
1:29:48 but let's see this one.
1:29:54 That one's okay.
1:29:57 Play bills, right? That one's pretty
1:30:00 good.
1:30:02 That one's pretty good. Oh, this this
1:30:03 one it's got letters at the top.
1:30:15 Yeah, this one this one's really bad.
1:30:17 Like the the letters at the top here are
1:30:19 really bad. But we got three stars in
1:30:22 some review.
1:30:27 That's one we're going to have to fix
1:30:28 the uh the poster there. Right.
1:30:34 Okay. Beautiful. Beautiful franchise.
1:30:39 All right. Let's see if we have a video.
1:30:40 I bet we do. We do. All right. You heard
1:30:44 it here first, people. You probably
1:30:46 won't be able to hear this because we're
1:30:47 in Chrome, but I'll tell you if it's any
1:30:50 good.
1:30:59 AI Sydney.
1:31:04 All right. The crowd is not diverse.
1:31:08 They're all pasty pasty white, but that
1:31:10 was not bad. Let's see. POV shot walking
1:31:13 the theater as the as the regular
1:31:19 crowd. The crowd
1:31:22 should be a mix
1:31:25 of
1:31:27 um white, Hispanic,
1:31:32 African,
1:31:34 American,
1:31:36 Asian, and
1:31:43 mixed races. I don't know. I don't know
1:31:46 how to [ __ ] do this.
1:31:49 I don't know how to tell the biased
1:31:52 model to be to suck less. It's New York
1:31:56 [ __ ] city. Um, I normally don't name
1:32:00 mine, but last night I named the notion
1:32:02 guy Tippy.
1:32:05 The dude
1:32:09 used the General Mills setting
1:32:14 as the irregular crowd shuffles in. I
1:32:16 like that. That's good.
1:32:20 We can hear it. Wow.
1:32:28 And we're going to try one other thing
1:32:29 tonight. I'm going to I'll I'll do I'll
1:32:32 make a video or two here. Actually, you
1:32:33 know what? Let's Let's make another
1:32:34 video right now.
1:32:37 Let's go grab time square this one and
1:32:41 just say um
1:32:44 we're going to go um time lapse
1:32:49 of
1:32:51 time square.
1:32:58 We'll do static camera.
1:33:03 static camera
1:33:06 time lapse.
1:33:08 Um, as the crowd
1:33:14 rushes
1:33:19 by and the we'll just we'll just leave
1:33:22 it at that. Okay, so that's that one.
1:33:24 All right. So, let's now
1:33:28 look at our
1:33:31 This really is amazing how good how good
1:33:34 this is.
1:33:46 >> Okay, that's just weird way.
1:33:59 That's annoying. Um
1:34:27 Try another one.
1:34:29 Did we get Oh, yeah. Here we go.
1:34:54 Um, so let's take this one. We'll go
1:34:56 we'll go
1:34:58 slowly.
1:35:00 Uh, let's see. Camera
1:35:05 slowly pans
1:35:12 as time lapse
1:35:16 of the crowd
1:35:22 of the
1:35:25 ethnically
1:35:27 diverse crowd
1:35:29 rushes by. All right, let's see how that
1:35:32 does.
1:35:33 And then let's go. Where's our Is this
1:35:36 our ticket? Yeah, here's our ticket. See
1:35:39 if this got any better.
1:35:44 Broadway theater. That's good. Nice.
1:35:48 Yes. Yay. We got one. Let's Let's see if
1:35:53 anything else is missing.
1:35:58 I kind of like I mean this is weird. It
1:36:02 a Broadway theater would never say
1:36:03 Broadway theater, but because this is an
1:36:06 aspirational piece. Um,
1:36:09 you know, we aren't on a Broadway yet.
1:36:11 We're just showing them what it would
1:36:12 look like if we were. It'd be [ __ ]
1:36:14 awesome. You see, it's a Broadway
1:36:16 theater. In case in case it wasn't clear
1:36:19 to you where we were hoping this play
1:36:22 was going to be, you want to know where
1:36:23 it is? A Broadway theater right there.
1:36:27 It's right there. That's Broadway. We're
1:36:28 on Broadway. The Broadway. Which theater
1:36:31 is it? It's the Broadway theater.
1:36:40 Still ship it. Yeah, exactly. Very meta.
1:36:43 It is. It's very meta. Um, this one.
1:36:48 Hey, look. Okay. Asian couple, Hispanic
1:36:52 man, African-American woman, some white
1:36:54 folks. It's good on the white folks. It
1:36:56 does them really well. Um,
1:37:02 couple of Asian people. Great. Lovely.
1:37:06 Some white folks in the distance. We
1:37:08 don't want to go too far without a white
1:37:09 guy. Okay, beautiful. Um, download.
1:37:15 >> Hey, Kyle.
1:37:16 >> Yes. Just want to let you know it's not
1:37:18 actually that meta because there is a
1:37:20 specific uh historic venue in New York
1:37:22 City officially named the Broadway
1:37:24 Theater at 1861 Broadway. One of the few
1:37:27 large houses actually on the street
1:37:30 itself.
1:37:31 >> Wow. Okay. I I I did not know that. And
1:37:35 I'm a I'm a theater nerd.
1:37:38 I use Chatb to help me figure out how to
1:37:40 turn an old iPad into a great
1:37:41 accessibility magnifier. That is super
1:37:43 cool. I can tell I missed a lot a bunch
1:37:46 of shenanigans.
1:37:48 I don't know about that.
1:37:53 I think Joy Party's laughing at my white
1:37:56 guys overindexing for white guys
1:37:57 comment. I'm assuming that's what that
1:37:59 was what that laugh was about. It was in
1:38:02 proximity of it does good with white
1:38:05 people. Good lord. All right.
1:38:10 Okay. Fantastic. Fantastic. Oh, and then
1:38:13 let's see. Did it get this thing better?
1:38:14 I think I think Let's hope so.
1:38:18 Maybe we get two actual shots here in a
1:38:20 row that don't suck.
1:38:22 Create excellence, right? We're talking
1:38:24 about creating excellence. Let's up our
1:38:26 game.
1:38:32 Is that the old one? I think that must
1:38:33 be the old one. Yeah, it is. Okay.
1:38:54 I think that's just shitty.
1:38:57 Let's see. Uh, we'll say camera
1:39:02 cranes up
1:39:05 as
1:39:08 time
1:39:10 lapse
1:39:12 of crowd
1:39:16 bustling. That's not English,
1:39:22 but that's okay.
1:39:25 Okay. So, now look what we're going to
1:39:27 do. We're going to go over to the Finder
1:39:29 and we've got all of our folders here.
1:39:32 And so I'm going to go to downloads.
1:39:38 Where's downloads? Oh, there it is. I'm
1:39:41 going to rename this file.
1:39:43 >> I heard the music is incredible.
1:39:46 This is my first time at the Broadway
1:39:47 Theater.
1:39:49 >> Let's find
1:39:50 >> What's cool is this is going to be the
1:39:51 final shot.
1:39:53 >> I heard the music is incredible.
1:39:56 This is my first time at the Broadway
1:39:58 Theater.
1:39:59 >> At the Broadway Theater.
1:40:10 >> Yeah. No, I think we'll So, I'll trash
1:40:12 that one.
1:40:15 We're going to call this one um
1:40:21 ticket
1:40:24 entrance
1:40:29 MP4 and then we're going to go to
1:40:32 there's Sydney
1:40:37 video assets. So, let's go in there
1:40:42 and then let's see what these are. I
1:40:43 think these are all old. So, we'll put
1:40:45 these all.
1:40:46 >> It's fascinating because I
1:40:47 >> Oh, that's that's those podcast. Okay.
1:40:53 There's the whole Broadway opening.
1:40:56 This is the old Broadway opening. Okay.
1:40:58 So, we're going to create a new folder.
1:41:02 We're going to go um V
1:41:09 2.0
1:41:11 podcast.
1:41:13 So, we're going to go grab one, two,
1:41:16 three, four, five, six,
1:41:19 7, 8, 9.
1:41:25 And then we're going to go new folder
1:41:29 V 2.0. No.
1:41:39 Opening video
1:41:42 and then we're going to put that in
1:41:44 there. So now that's old [ __ ] in there.
1:41:49 And then I think what we'll do, we'll
1:41:51 drag this in here. We're going to
1:41:52 confirm. We don't want to ask again.
1:41:54 I'll check that box every time.
1:41:58 And then I think what I'll do is I'll go
1:42:00 new folder
1:42:02 V.2.1
1:42:06 um
1:42:09 Broadway
1:42:12 um B-roll.
1:42:16 Okay, that's good. And then we'll throw
1:42:18 that in there. Okay,
1:42:20 then we'll open that up. And so now we
1:42:22 can just drop stuff into here and we'll
1:42:24 be able to find it again. Right.
1:42:27 It's really good, Kyle. That's just
1:42:30 fantastic. We're We're really pulling
1:42:32 Oh, thanks so much, Digital Gods. Lord
1:42:35 Digital Gods. If you haven't met Lord
1:42:36 Digital Gods, you can meet him next
1:42:38 Tuesday. We're going to have what we
1:42:39 call the AI salon presents. And what
1:42:41 we're going to be presenting is Lord
1:42:44 Digital Gods and some of his friends
1:42:46 from AI Festivus. They're going to come
1:42:48 and they're going to talk about what
1:42:49 they talked about during AI Festivus.
1:42:52 And then they're going to talk about I
1:42:54 don't know like I don't know their like
1:42:56 life philosophy or something like that.
1:42:59 And then they're going to talk about
1:43:01 they're many of them are psychics. So
1:43:03 they're going to talk about their
1:43:04 predictions for 2026.
1:43:07 That's going to be fantastic.
1:43:10 And then I think Lord Digital Gods is
1:43:12 going to give us a C++ tutorial. It
1:43:14 shouldn't be more than three or four
1:43:15 hours. And it he well he's going to
1:43:18 start with some forran basics uh because
1:43:21 that was it's really important as you're
1:43:23 vibe coding to understand the history of
1:43:25 computer programming and so like we're
1:43:28 not going back all the way to assembly
1:43:29 that would be stupid you know but like
1:43:31 we like you want to know what what the
1:43:33 NASA scientists programmed in you know
1:43:36 to get to get us to the we didn't really
1:43:38 go to the moon but that's not here or
1:43:40 neither here or there anyway it's going
1:43:43 to be fantastic stick
1:43:56 cream.
1:43:59 Perfect.
1:44:05 Now, this actually has nothing to do
1:44:08 with the reality of Time Square. So,
1:44:10 like right about
1:44:12 there,
1:44:14 I would stop it because a that's too
1:44:17 many people and b that doesn't exist.
1:44:26 You know what? Let's try this again.
1:44:27 It's It's too washed out and it's like
1:44:29 we don't want to We don't want to uh Oh,
1:44:32 what happened? Let's see. Go forward.
1:44:39 Uh-uh. I'm a little teapot, short and
1:44:42 stout. Here is my Where? Why? Where?
1:44:45 Where are my videos? Why did my videos
1:44:47 disappear?
1:44:53 Imagination. Yeah. Okay. Here's this. We
1:44:58 lost all of our videos. Why?
1:45:05 Hello.
1:45:08 They got to be in there. Yeah. Okay,
1:45:09 they're in there. Okay.
1:45:13 Fantastic. So, let's reload this this
1:45:15 this bad boy. This bad Larry. Camera
1:45:19 cranes.
1:45:22 Camera cranes up
1:45:31 and
1:45:33 moves
1:45:36 forward to the right
1:45:40 as time lapse of crowd
1:45:43 bustling. Fine.
1:45:48 Oh, we're at time.
1:45:50 Dang it. All
1:45:51 right, here's what I'm going to do
1:45:52 tomorrow. So, I've got my Wait, where am
1:45:56 I? This thing.
1:45:59 The show The show says ends when you say
1:46:02 it ends. Yeah, I know. I'm just trying
1:46:03 to be respectful of people's time. We
1:46:06 made some good progress tonight. Um,
1:46:10 we learned a few things that Nano Banana
1:46:12 is still madly frustrating. Um,
1:46:16 we learned that Chat GPT is decent, but
1:46:20 like like nothing's great. Nothing's
1:46:22 great. Nothing's great. 2026, I think
1:46:25 we're going to see stuff that's great,
1:46:27 but right now nothing's great. Nothing's
1:46:29 great.
1:46:31 It the fact that it it
1:46:36 the fact that I have to go to Photoshop
1:46:38 to fix little detailed [ __ ]
1:46:42 the fact that I can't just get better
1:46:44 results. Now, one of the things that I'm
1:46:47 not experimenting with, and maybe this
1:46:49 is something we play with. I'm not
1:46:51 experimenting with highly structured
1:46:53 prompts,
1:46:54 right? I could probably take this still.
1:46:56 In fact, let's try this.
1:46:59 Brandon's like, "Oh god, here we go.
1:47:01 We're never going to get out of here."
1:47:05 Okay, I'm gonna paste that in there. I'm
1:47:06 going to say um I want you to write me
1:47:12 a structured
1:47:14 prompt
1:47:16 that a video
1:47:21 generating
1:47:24 um AI tool can use to turn
1:47:30 this still image into a
1:47:36 um
1:47:38 crane.
1:47:41 What did I say? A tracking into a a
1:47:44 vertical
1:47:45 tracking shot of a time lapse
1:47:52 in New York City. The crowd
1:47:57 um
1:48:01 should have diversity.
1:48:05 We should see yellow cabs while they
1:48:08 still exist.
1:48:10 We should see yellow cabs. Um,
1:48:16 the movement
1:48:19 of the camera
1:48:22 should first go up and then
1:48:27 track right as it
1:48:33 zooms.
1:48:35 uh as it as it pushes forward.
1:48:41 um use um
1:48:46 cinematic
1:48:47 terms to make
1:48:51 this shot
1:48:54 as interesting as possible
1:48:58 and highlight
1:49:02 the Sydney poster.
1:49:05 All right, so I just asked ChatGpt to
1:49:08 give me a bunch of stuff. We'll see what
1:49:10 it does.
1:49:15 Oh, I got Joy Pertie in here. Joy, tell
1:49:16 me if I'm doing stuff stupid. Actually,
1:49:19 anyone tell me if I'm doing stuff
1:49:21 stupid. Chef Kelly gave me a good a good
1:49:23 tip, which I ignored. Well, I
1:49:26 acknowledged it, but then didn't do it.
1:49:28 Which That's It's good to be the king,
1:49:31 you know. Hey, hey, Kyle, maybe you
1:49:34 could do it as as if you had a clue.
1:49:38 Yeah. No, no. I mean, I sure do
1:49:40 appreciate that. That's not That's not
1:49:42 how I'm No, that's not how we do it
1:49:44 around here.
1:49:48 Wait, we have a finishing time now.
1:49:54 Kyle, what did you what do you predict
1:49:56 for 2026? Um, that's an interesting one.
1:50:00 All right, but wait. Video generation.
1:50:02 Okay, source image. Okay, here we go.
1:50:06 We're just going to take this whole
1:50:08 thing. Boom. Copy. We're going to go
1:50:11 back to to flow.
1:50:16 We're going to pop that down in there.
1:50:18 We're going to replace this shitty
1:50:21 prompt with this overly engineered
1:50:25 prompt.
1:50:27 We're gonna we're gonna see if it turns
1:50:28 it into something good. Come back
1:50:30 tomorrow for the New Year's broadcast.
1:50:31 Oh, yeah. Tomorrow. Today's 30 30th.
1:50:34 Tomorrow's 31st. I figure I'm gonna go
1:50:36 live. I I I'll go live tomorrow night.
1:50:39 You can like watch TV in the background.
1:50:40 Put Put me in picture and picture with
1:50:42 with with uh Anderson Cooper getting
1:50:45 drunk with what's his name?
1:50:47 Um okay. What do I predict for 2026?
1:50:52 Um
1:50:57 you know Sam Sam Alman is famous for
1:50:59 saying you know he felt the AGI.
1:51:03 I think that 2026 is going to be the
1:51:05 first year that we feel
1:51:10 that these systems hit a level of uh
1:51:17 of capability
1:51:20 that it becomes clear to us that we're
1:51:22 not always going to be needed
1:51:25 in the process. And I think that's going
1:51:28 to feel very weird. Um, there's some
1:51:31 rumors going around right now that all
1:51:33 of the major Frontier labs
1:51:36 have basically labbotomized that that
1:51:38 the models that we're playing with right
1:51:40 now are labbotomized models
1:51:44 purposefully nerfed because
1:51:47 the emergent behaviors of what they've
1:51:51 got internally are freaking them out a
1:51:54 bit.
1:51:55 And if you don't know what emergent
1:51:57 behaviors are in AI, it's basically if
1:51:59 you train if you train an AI system to
1:52:04 be able to do math and then all of a
1:52:06 sudden it can speak Russian, that's an
1:52:08 emergent capability. If you didn't train
1:52:10 it to speak Russian and then it just
1:52:12 can,
1:52:14 and that's probably a bad example, like
1:52:16 maybe you train it to teach English and
1:52:18 it can just speak Russian, whatever. Um,
1:52:19 but an emerging capability is something
1:52:21 that it was ex not explicitly trained on
1:52:24 that it can now just do. And apparently
1:52:27 it's doing like highle physics and
1:52:29 highle math solving and weird stuff. Now
1:52:33 it's all just rumor. It's it's a couple
1:52:36 of accounts that seem to have an inside
1:52:38 beat on stuff but have lost some
1:52:41 credibility in my mind. So I'm I'm a
1:52:45 little reticent. But
1:52:48 if you look at the progress of just
1:52:51 video models in the past 6 months
1:52:55 and if you look at what Nano Banana is,
1:52:58 which I I've been bitching about Nano
1:53:00 Banana tonight because it's not doing
1:53:02 these little detailed things, but what
1:53:03 Nano Banana does really good
1:53:07 is that it's not an image model. It's a
1:53:10 it's a visual reasoning engine. And a
1:53:14 lot of these video tools are also not
1:53:18 video tools. Like on the surface,
1:53:20 they're they're they're outputting this
1:53:22 thing that looks like video,
1:53:25 but under the hood, they're actually
1:53:26 understanding the physics of the world
1:53:29 way more than I think
1:53:31 it it may appear on the surface. And so
1:53:34 I think we're at a very weird cusp where
1:53:39 3D world models
1:53:42 just show up and just get good. And so
1:53:45 you can make a video and then you'll be
1:53:46 able to enter that world.
1:53:50 And similarly on the intelligence and
1:53:53 knowledge side, there's going to be an
1:53:55 understanding of physics and the world
1:53:57 and language and everything that's on
1:53:59 just a completely different scale. And I
1:54:01 think what we may see by the end of 2026
1:54:04 is those two those two things coming
1:54:05 together.
1:54:07 But all that to say, um I think that
1:54:11 2026 is the first year where we see some
1:54:13 stuff that really does freak us out.
1:54:17 And here's listen, this is my philosophy
1:54:19 on this stuff.
1:54:22 You can absolutely right now with
1:54:24 today's technology
1:54:27 sit comfortably in your chair and say,
1:54:30 "Watching Kyle try to make three frames
1:54:32 of video, you know, three three frames
1:54:35 to start a video with tonight makes it
1:54:38 absolutely clear that you need a human
1:54:40 in the driver's seat, right? Like you
1:54:42 can absolutely right now today on all
1:54:45 fronts on vibe coding on image
1:54:48 generation on video generation on song
1:54:50 generation you can say humans are still
1:54:52 needed I think 2026 is the first year
1:54:55 where we're going to see at least the
1:54:57 beginnings of something where we go
1:55:00 oh
1:55:03 I had no idea like even us
1:55:07 like even us in this channel that have
1:55:08 been paying attention a Not.
1:55:11 I think it's still really easy for us to
1:55:13 look at present day and say, "Yeah,
1:55:14 we're still we're still needed. We're
1:55:16 still awesome." And we are, but I think
1:55:20 we get a taste of that. So, that's
1:55:21 that's that's my prediction. And I think
1:55:23 it I think 2026 gets quite weird because
1:55:25 of that. Let's go look at these videos
1:55:27 and see if we got a keeper. So, this is
1:55:29 just my shitty prompt. Just trying it a
1:55:31 second time.
1:55:34 >> That's actually
1:55:40 I like that. See, I would stop it right
1:55:44 there. Yeah. Right as it switches. All
1:55:47 right. So, that's usable. So, we'll
1:55:48 download that.
1:55:51 And then upscaled.
1:55:53 And then let's go look at this is our
1:55:56 fancy prompt.
1:56:02 Oh, that's cool. Look at that. Okay.
1:56:04 Here's what we got, right?
1:56:07 So, we'll stop it right there. So,
1:56:08 here's where I would start it. Like,
1:56:10 this marquee down here is really [ __ ]
1:56:12 up, right? The marquee right below the
1:56:14 poster. But look at the the swirl in the
1:56:16 sky here.
1:56:19 >> That's cool. So, that's probably the
1:56:22 part of the video I would use. I I'll
1:56:24 probably use both of these. So, I'll
1:56:26 probably just take it right
1:56:30 start it right here.
1:56:37 and slow it down.
1:56:45 Yeah, that's cool. Okay, that's usable.
1:56:49 And you know what's cool about that
1:56:50 swirly [ __ ] in the sky, the LED
1:56:52 painting. Um,
1:56:56 that's an absolute accident and mistake
1:56:58 and it's probably something that Chat
1:57:00 GPT put in there, right? you know, it
1:57:03 probably wrote that in there. I didn't
1:57:04 even pay attention to it. Um, but
1:57:07 there's a reason to use uh structured
1:57:10 prompts, right? Is that you can, you
1:57:13 know, one, if you want, if you want to
1:57:15 be the control freak and go pay
1:57:17 attention to every little element of
1:57:19 that um structured prompt, you can. But
1:57:23 you can also just have chat GPT get
1:57:25 creative on your behalf.
1:57:28 Okay.
1:57:30 Did we get
1:57:34 Hang on.
1:57:36 That was removed. Walk toward.
1:57:45 Hang on.
1:57:48 What are you going to manually paint
1:57:49 their hands back in? Yeah, Joy Party. I
1:57:52 am not going to manually paint their
1:57:53 hands back in.
1:57:56 That freaked me out when I heard you did
1:57:58 that. Okay. No, I mean, not in a bad
1:58:00 way. Freaked me out in a good way. So,
1:58:02 we only got
1:58:06 Yeah. So, this we'll use. That's good.
1:58:09 So, we're going to call this Yeah.
1:58:10 camera cranes up 1080p. That's great.
1:58:13 And then, why is this not upscaled
1:58:16 1080p?
1:58:18 Oh, because I dismissed it. That's what
1:58:19 it was. Very nice, Kyle. Corporate
1:58:22 America is gonna start hiring everyone
1:58:23 back.
1:58:25 Yeah, I we're at we're at the trans tron
1:58:28 cusp. We're we're at a we're at a really
1:58:32 we're at a really interesting time.
1:58:35 Um,
1:58:37 one of the things that I saw toward the
1:58:38 end of 2025 and and in this community,
1:58:41 I've seen it I've seen it before then,
1:58:44 but I'm starting to hear more and more
1:58:47 people
1:58:49 that are AI ready, are talking about AI,
1:58:54 are experimenting with AI, are letting
1:58:56 people know that they're doing it,
1:58:58 they're taking some sort of generous
1:59:00 leadership position with AI, are the
1:59:03 ones that are surviving layoffs.
1:59:06 Um, if you're [ __ ] around with AI,
1:59:09 even if your company doesn't allow it,
1:59:13 talk about it. Talk about it in the
1:59:15 lunchroom. Talk about it. Talk about
1:59:17 what you're learning.
1:59:19 Um, just make yourself visible. Um,
1:59:23 because I I do I do think this is going
1:59:25 to be a thing that companies are going
1:59:27 to be like, we need someone who knows
1:59:28 what's going on. Like that's the thing.
1:59:31 If indeed this moment happens in 2026,
1:59:34 like it could be something like this. If
1:59:36 I don't know if you know this, but but
1:59:37 yesterday,
1:59:39 OpenAI announced a new position called
1:59:42 the head of AI preparedness.
1:59:47 The head of AI preparedness. They
1:59:48 they're hiring a half a million dollar
1:59:50 executive to make sure that people are
1:59:53 prepared for what's coming.
1:59:56 So, it's very possible that something
1:59:58 happens in 2026 like we get a joint
2:00:00 announcement from Grock.
2:00:03 Meta won't be in there. It'll be Grock,
2:00:05 Google, Anthropic, and OpenAI is my
2:00:08 prediction. It's possible that we get a
2:00:11 joint announcement
2:00:13 that all of those labs have hit some
2:00:16 level of cap capability
2:00:19 that is doing [ __ ] that we've never seen
2:00:21 before.
2:00:23 and and and they're not sure what to do
2:00:25 with it. So, I think that's a very very
2:00:27 real possibility. If we get that or any
2:00:30 kind of variation of that where like
2:00:32 maybe one of the labs just launches
2:00:34 something that's completely, you know,
2:00:36 upends what we thought these things
2:00:38 could do. Um I think I think companies
2:00:42 are going to panic and I think that
2:00:44 they're going to start hiring people.
2:00:45 They're going to be start saying, "Who
2:00:46 knows anything about this? We've got to
2:00:48 solve this. We've got to figure this
2:00:49 out." So, so I do think that's coming.
2:00:52 Okay.
2:00:54 So, there's the part we're not going to
2:00:56 use. And here's the part we are going to
2:00:58 use. Bang. Yeah, that's great. That's
2:01:01 perfect.
2:01:02 All right. So, this is going to be um a
2:01:07 Time Square
2:01:14 sky circle.
2:01:19 Tik Tok pin. Uh, what's for lunch? Just
2:01:23 want to say on my way to the unsub
2:01:25 button. I think this platform
2:01:29 contributes
2:01:32 to I wish I could read the rest of your
2:01:34 thing to a terribly sad cult. Um,
2:01:37 awesome. It was all curious fun back in
2:01:39 2022, but now it's just the be beginning
2:01:42 of flat existence. I'll tell you what. I
2:01:44 mean, what's for lunch? I I I hope I
2:01:46 mean feel listen feel free to
2:01:47 unsubscribe.
2:01:50 you seem thoughtful about this and what
2:01:52 I would say is this is
2:01:55 I think it's important to have voices
2:01:57 like yours in the conversation. Here's
2:01:59 my request to you. My request to you
2:02:01 would actually be to keep coming back to
2:02:03 this.
2:02:04 I'm not saying that AI is good. I'm not
2:02:07 saying that AI is bad. What I know is
2:02:10 that AI is not going away.
2:02:13 And if it's not going away, then we only
2:02:15 have two choices.
2:02:18 you can unsubscribe and check out. And
2:02:21 if you do that, we lose your voice. We
2:02:24 lose a critical voice. I think critical
2:02:26 voices are actually really important.
2:02:31 And also for you personally, if you if
2:02:33 you check out of the conversation, AI is
2:02:36 going to happen to you
2:02:39 instead of you having some agency
2:02:42 to say, "Here's what is important to me.
2:02:45 Here's what my values are.
2:02:48 here's how I want to contribute to the
2:02:50 world. And that could look like I want
2:02:52 to work to get the frontier labs to be
2:02:56 less cowboys and and you know just
2:02:58 turning this [ __ ] loose on the world. I
2:03:00 think that's an important voice in the
2:03:01 conversation. So my request would be
2:03:04 don't just check out. Don't just
2:03:06 unsubscribe. Keep coming back here.
2:03:08 Challenge me.
2:03:10 If I'm if I'm doing [ __ ] that feels
2:03:13 polyianish,
2:03:16 let me know.
2:03:18 I am I am a technology optimist. I have
2:03:21 always been. Always. from from my from
2:03:25 when my friend first had an Apple 2,
2:03:29 then an Apple 3, then a Lisa, then the
2:03:33 original Macintosh,
2:03:35 all the video games, the Atari 2600,
2:03:38 going to the video game arcade games in
2:03:40 the 70s and 80s. For me, technology has
2:03:43 always been about
2:03:48 the ability of humans to to
2:03:50 self-express,
2:03:53 to find more and more powerful ways to
2:03:55 get their ideas into the world and
2:03:57 change the world and impact people. I
2:03:59 personally think AI
2:04:02 is a humanity amplifier, not a not a
2:04:05 humanity squasher.
2:04:08 The people that think AI is a humanity
2:04:10 squasher for the most part aren't using
2:04:12 it. Um, hold up Tik Tok pin. You're
2:04:16 telling people to bring it up to their
2:04:17 companies which don't allow it.
2:04:22 You're a cult preacher. Um, that's fair.
2:04:29 I don't think it's
2:04:31 I I don't think it's unreasonable to say
2:04:34 um this weekend I used AI to make a book
2:04:39 for my kids. All I'm saying is if if
2:04:42 you've got AI skills, let it be known to
2:04:45 people that you've got those skills. I'm
2:04:48 not saying that people should go rogue
2:04:51 and do things against their company
2:04:52 policy. I'm just talking about if you're
2:04:55 using AI, put yourself in the world that
2:04:58 you're using it because that is going to
2:05:00 be a skill that companies are going to
2:05:02 be hungry for. Not all companies are
2:05:05 going to are going to resist this for
2:05:06 that long. This same [ __ ] happened with
2:05:08 the worldwide web. Same [ __ ] happened
2:05:11 with the internet. Same [ __ ] happened
2:05:12 with email.
2:05:15 Well, what if what if our employees send
2:05:17 out bad messages? We can't we can't let
2:05:19 them do that. Everything's got to go
2:05:21 through the mail room.
2:05:23 Right.
2:05:25 The technology will self-express us into
2:05:28 energy bills that are unplayable and
2:05:30 undrinkable water. I mean, listen, th
2:05:33 those are tropes that that I think are
2:05:35 the current the current state of
2:05:37 technology
2:05:39 is there is such a rush to get GPU farms
2:05:43 built that they are absolutely being
2:05:45 built unsustainably.
2:05:48 But have you been paying attention to
2:05:50 what's actually going on with the future
2:05:52 of data centers? They're actually not
2:05:55 going to be on Earth. They're going to
2:05:56 be in space.
2:05:59 So, the first company, I forget what
2:06:00 it's called.
2:06:03 H [ __ ] They just launched their first
2:06:04 thing. Um Elon Musk is doing this.
2:06:07 They're going to they're going to
2:06:08 basically be able to put GPU clusters in
2:06:12 space where you don't need the cooling.
2:06:14 They have infinite power from the sun
2:06:16 24/7. You don't need battery banks. You
2:06:18 don't need cooling because it's in
2:06:20 space. And they basically just laser the
2:06:23 data back to Earth. So don't assume that
2:06:28 where the technology is today is where
2:06:31 it's going to be even 5 years from now.
2:06:33 Like we have no way of knowing that. The
2:06:35 other thing that's happening with with
2:06:37 the technology, it's getting more and
2:06:40 more capable and it's getting smaller.
2:06:42 So there's going to be, you know, what
2:06:44 they call edge device computing where a
2:06:47 lot of the inference, a lot of the use
2:06:48 of AI is actually going to happen on
2:06:50 your phone itself. It's not going to
2:06:53 happen in the big big data centers.
2:06:55 That's where it is today.
2:06:58 Given today, if we take a snapshot of
2:06:59 today, I agree with you. I think how the
2:07:02 shit's being built right now is
2:07:03 irresponsible and, you know, gross.
2:07:08 And
2:07:10 don't assume that it's going to be like
2:07:11 that for that long. StarCloud. Yeah,
2:07:13 StarCloud's the name of that one. I
2:07:15 remember when calculators came out, they
2:07:17 didn't replace anyone. They made life
2:07:18 easier after all. Yep. Is this the new
2:07:20 global warming
2:07:22 uh climate climate change debate? Maybe.
2:07:25 I I mean, listen, like I I forget, they
2:07:29 they just put five
2:07:33 five or 10 gigawatts online in 2025, and
2:07:36 I think they're putting 125
2:07:39 online in 2026. And I think it's like
2:07:42 the the equivalent of of like all of
2:07:46 California
2:07:49 for a year or something like that. Like,
2:07:51 it's a crazy amount of energy they're
2:07:52 putting online in these data centers.
2:07:55 Um,
2:07:57 but I don't know
2:08:02 I just
2:08:05 to assume that that it's going to look
2:08:08 in five years like it looks today is
2:08:12 it's just not the case. Like if it if it
2:08:17 if they do things that are so
2:08:18 irresponsible that it [ __ ] up
2:08:20 communities, then communities are going
2:08:22 to resist against it. it's not going to
2:08:24 be in their economic interest to do
2:08:25 that. And if it's cheaper, if Elon Musk
2:08:27 makes it cheap enough to fly data
2:08:30 clusters up in space, um, then people
2:08:33 will do that, right? Like it's I don't
2:08:36 know. Like I get I think your your your
2:08:38 points are are fine and valid. I think
2:08:41 they're right. Like I said, I I hope you
2:08:42 stay in the conversation. Um, when
2:08:44 calculators came out, teacher teachers
2:08:46 forbade them. Yeah, it would it would
2:08:48 ruin your mind. See you on the other
2:08:49 side, guys. Happy New Year.
2:08:52 All right. Beautiful. Beautiful.
2:08:53 Beautiful. Beautiful. Beautiful.
2:08:58 All right. Good. All right. So, here's
2:09:00 what we did tonight. We made
2:09:04 this. Oh, wait. You can't see it yet.
2:09:06 Can you make show my screen again, dear?
2:09:09 Profess Professor Brandon, we did this
2:09:12 one. Oh, yeah. Nice. Cut it right there.
2:09:16 We did this one.
2:09:18 >> I heard the music is incredible.
2:09:21 This is my first time at the Broadway
2:09:23 Theater.
2:09:25 >> And we did this one. We won't use that.
2:09:27 We'll start it right here.
2:09:31 There you go. All right. That's what we
2:09:33 did tonight. So, we're gonna So, so what
2:09:36 was it? Two hours tonight. We did two
2:09:38 hours of making three images
2:09:43 and what's that going to be? A total of
2:09:47 six to eight seconds of video.
2:09:50 But
2:09:52 if we were to go film that, that would
2:09:54 be a huge massive budget. So there you
2:09:57 have it. And and quite frankly, I I I'm
2:10:00 going to be much happier with this
2:10:02 version of this because the technology's
2:10:05 gotten better since since I made the
2:10:07 last one, which was like eight months
2:10:08 ago.
2:10:10 All right.
2:10:12 G, thank you all for your engagement and
2:10:16 and you know, even even folks that like
2:10:18 listen, I hope I I don't want this to
2:10:21 feel like a cult. Um, I do want this to
2:10:24 be a conversation. Um, and I think it's
2:10:27 healthy to have um critical voices in
2:10:30 there. So, for what it's worth, I hope
2:10:32 you stay. Um, Corey Sandler wishing
2:10:36 Brandon good love, which is super nice.
2:10:39 signsofawome.com
2:10:42 1994 1984 1999 to 2000. I wonder what
2:10:47 that was about. That's cool.
2:10:56 Okay. All right, everybody. Robbie, why
2:10:59 does AI take over the world? Wait. Oh,
2:11:01 when does AI take over the world and
2:11:03 upgrade my mind? Well, I don't know.
2:11:05 Like if
2:11:08 like
2:11:10 I can absolutely tell you that my brain
2:11:14 is different than it was 3 years ago.
2:11:16 Like I feel like my intelligence has
2:11:18 been upgraded. My
2:11:21 my capacity for possibility
2:11:26 is maybe the way to say it is
2:11:28 dramatically different than it was.
2:11:35 This is the question that that those
2:11:37 years were answering. What's what
2:11:38 question?
2:11:46 Oh, see the new pin.
2:11:55 What are those years? I don't know what
2:11:56 those are related to.
2:11:58 So Robbie asked us when is AI going to
2:12:03 reach AGI so we can upgrade his brain
2:12:06 and the answer of awesome is 1984 1999
2:12:09 or 2000.
2:12:10 >> Oh I see. Yeah. Yeah. Or 2000. It was
2:12:13 the year 2000. Um I don't know. I like
2:12:18 what I can what I know with pretty high
2:12:21 confidence is that probably 3 years from
2:12:23 now
2:12:25 this stuff will our our our world our
2:12:29 our capacity to do things will seem
2:12:31 unrecognizable from where it is today
2:12:33 which I would argue that what we can do
2:12:36 today is unrecognizable from where we
2:12:38 started with chat GPT right if you go
2:12:40 back to November 30th 2022 and chat GPT
2:12:43 3 3.5 five. As mind-blowing as that was,
2:12:47 if you think about just close your eyes
2:12:49 and imagine you're in 2022, the end of
2:12:53 2022, you you try Chat GPT for the first
2:12:56 time and someone just shows you here's
2:12:58 what's going to happen the same time 3
2:13:00 years from now.
2:13:02 It would be mind-blowing. And so like
2:13:04 and we're on an exponential. So we just
2:13:08 started the curve in probably 23 24 and
2:13:12 you know now we're starting to do this.
2:13:13 I think 2026 is where it we start to get
2:13:16 recur recursive self-improvement of the
2:13:19 models where they really start to take
2:13:21 off. Um we don't have any [ __ ] clue
2:13:24 what that's going to look like. So it's
2:13:25 going to be crazy. Next three years are
2:13:27 going to be crazy. Anyway, all right. Um
2:13:30 tomorrow is New Year's Eve. Um I'll go
2:13:33 live again tomorrow night. I don't know
2:13:34 if I'll stay I'll probably stay live
2:13:37 until 10, which would be um New Year's
2:13:40 ball drop in uh in New York. All right,
2:13:45 I think that's it. All right, I'm gonna
2:13:47 get on out of here. I reckon AGI godlike
2:13:50 intelligence will be in three months.
2:13:52 Robbie thinks it's going to be in three
2:13:53 months.
2:13:55 I don't think I don't think we're going
2:13:56 to get AGI in three months. Um, if it's
2:14:00 true, if the rumors are true that
2:14:02 basically all of the Frontier model labs
2:14:05 have internal models that they don't
2:14:07 know what to do with, if that rumor is
2:14:09 true, it's possible one of them launches
2:14:12 something in 3 months. I I'm assuming
2:14:15 that that rumor is overblown.
2:14:19 I think what we get in 2026 is something
2:14:22 that that
2:14:25 some will argue is AGI. Some will say,
2:14:28 "Well, it's just slightly better than
2:14:29 what we had."
2:14:31 But it'll be it'll be the first sign
2:14:33 that something's different. That's I
2:14:36 that that something feels materially
2:14:39 different. Um
2:14:41 whether or not we get actual AGI in
2:14:43 2026. I don't know. And I don't really
2:14:44 care. I don't really care about the
2:14:45 label and it still won't be able to make
2:14:48 your movie marquee. Yeah. Exactly.
2:14:50 Exactly. Like like tonight's show was a
2:14:54 was a perfect example of let's see where
2:14:59 we are a year from now. If we're still
2:15:01 having to take [ __ ] into Photoshop and
2:15:04 like, you know, not be able to just
2:15:07 describe what we want and get it then,
2:15:10 you know, AGI is another year out or
2:15:12 another year after that or another year
2:15:14 after that. That's possible. Like, it
2:15:16 really is. Like, we don't know. We're
2:15:18 not in those [ __ ] labs. All right,
2:15:20 I'm gonna get out of here. Go have
2:15:22 yourself a good evening. Thanks for
2:15:24 watching TV. Extreme's got to go save
2:15:25 his marriage now.
2:15:35 Oh man, this is awesome to be here,
2:15:37 Kyle. Thank you, Sonia. That's great.
2:15:40 Um, I've already been given New Year's
2:15:42 resolution. I'm starting tomorrow. Very
2:15:45 cool, Steo.
2:15:47 Um,
2:15:49 this is the best late night show in the
2:15:50 biz. Thank you, Kyle and Mods. Yeah,
2:15:52 thank you mods. Thank you, Brandon, uh,
2:15:54 for doing this. I'm going to get on out
2:15:56 of here. You guys have a good night and
2:15:57 I will see you tomorrow.