10/16/2025 - The Latest AI Updates, From GPT-5's Rocky Launch to New Runway Features

Description

Chapters

Transcript

0:02 Come on, buddy. Sit. Yeah.
0:06 [Music]
0:27 [Applause]
0:30 [Music]
0:31 [Applause]
0:33 [Music]
0:38 I met my baby at the F club underneath
0:43 time cafe on the corner life at four
0:47 street late one's day she offered me a
0:51 red strap I said baby not a chance I But
0:54 go for cool ginger ale. How about we
0:57 just dance? We did the boom shabboom
1:00 doo. I did it. Mocha chocolata and a
1:03 nitty-gritty. We did the oo the lazy
1:05 ballerina. Baby make me feel like I'm a
1:08 California dreamer. Yeah.
1:13 Oh yeah.
1:15 [Music]
1:21 >> What's happening, champ?
1:24 [Music]
1:41 [Music]
2:16 [Music]
2:26 Good evening, good people. We're
2:28 shaking.
2:32 [Music]
2:45 [Music]
2:50 I love you.
2:56 [Music]
3:29 Hello.
3:33 Hello.
3:36 [Music]
3:54 Is this place I can rinse my por
3:58 [Music]
4:02 my thoughts in
4:04 Silence.
4:05 [Music]
4:08 This is place where the feelings are
4:11 dead.
4:12 [Applause]
4:15 From an overest
4:20 place I can slowly face. The only one I
4:24 truly can know.
4:28 These are tears from a long time ago.
4:31 Got these tears from a long time ago.
4:38 So these are tears from a long time ago.
4:44 Go.
4:48 [Music]
4:50 [Applause]
4:53 [Music]
4:58 Okay.
5:00 Um, lots going on tonight.
5:03 We've got the new uh the new I I can't
5:08 even keep track anymore. We've got
5:10 VO3.1.
5:12 We've got Runway that's got a whole new
5:14 app suite. And if you're like, what are
5:16 Runway apps? We can play with those
5:18 tonight and find that out. Um, Brandon
5:21 just spent two days at MECON, the
5:24 Marketing AI Institute annual
5:26 convention, I guess, learning all about
5:28 all sorts of marketing and AI.
5:31 They go together like like peanut butter
5:34 and chocolate.
5:36 [Laughter]
5:38 They do. They do. It's good. It's good.
5:41 It's [ __ ] everything up. It's going
5:43 to change everything. Nobody knows
5:45 what's going on. Everybody's in a panic.
5:46 But it's awesome. It's awesome. Things
5:49 are good. Things are awesome. So, he's
5:52 going to give us a report from the
5:54 floor. From the floor of the New York of
5:56 the Cleveland Mecon Exchange.
5:59 [Music]
6:01 Sounds like he met some cool people
6:03 doing some cool things.
6:06 And uh yeah,
6:09 [Music]
6:11 [Laughter]
6:18 [Music]
6:27 [Applause]
6:28 [Music]
6:38 How's how's five been working out for
6:41 you? Uh, I heard I heard a thing today
6:45 that apparently Open AAI is accelerating
6:48 their launch timeline and they're going
6:50 to launch GPT6 before the end of the
6:52 year. I also heard that they're going to
6:56 launch an update to five that has four
7:00 O's personality.
7:03 I I'm guessing that's more marketing
7:06 than reality. uh even if they do it and
7:08 say that that's the case. So they can
7:11 actually retire four before they go to
7:14 six.
7:16 So basically what what it sounds like is
7:19 by February we'll have 17 models again.
7:24 [Music]
7:26 Um, I gotta tell you though, so
7:29 with GPT5,
7:35 one of the things that came out, one of
7:36 the things that seems pretty clear
7:39 is that when they launched GPT5, so so
7:42 GPT5 is actually three different models.
7:46 It's like a regular multimodal model.
7:49 It's a light reasoning model, a fast
7:51 reasoning model, and then it's a thinky
7:53 reasoning model, and then they have a
7:54 pro reasoning model. So, I think the the
7:57 distinction between
7:59 the the regular thinking and pro is just
8:01 how much time they spend. And then I
8:03 think the the light thinking is actually
8:05 a lighter model and then the GPT50
8:10 is the one that just answers. Sitting on
8:13 top of those is a routing model. So,
8:16 when you ask chat GPT5 something,
8:18 assuming you've got it in auto mode,
8:20 which is the default,
8:23 um,
8:25 hang on. Changing things here.
8:28 It switches between models for you
8:31 dynamically. Well, apparently at launch
8:34 that router was was broken. It was
8:37 faulty. It was not good. Um, and what
8:41 I've noticed is that probably starting
8:45 two weeks ago, the results I'm getting
8:48 back from GPT5 are just good. They're
8:52 they're just they they may not have the
8:54 personality that 40 had and and they
8:57 they feel less predictable than 40 was
8:59 probably because it's switching between
9:01 models. Um, but it's just good. It's
9:05 just good. So, so GPT5 is fine. If
9:08 that's not what you're asking about,
9:09 then I don't I don't know. Thinking
9:12 longer for a better answer. Not good at
9:14 all. Why not? Why is that not good at
9:16 all? Sacred is profane. Why is it not
9:19 good if it thinks longer? Oh, it's been
9:22 a code menace for me. Yeah, I'm not
9:24 doing a ton of coding with it. I I'm I I
9:28 would imagine that um that different
9:33 yeah different use cases are going to
9:35 have you know what is it your mileage
9:38 may vary. Um yeah I mean good news is
9:42 you've got Claude 4.5 Sonnet that's
9:45 apparently pretty baller. Um that's the
9:47 new model that sits behind uh
9:49 lovable.dev which I've been playing with
9:51 a fair amount recently and that's pretty
9:53 good. Um, but yeah. Um, no, you said
9:59 the node was not good at launch. Oh, not
10:01 good at all. Yeah, it was the it was the
10:04 model. It It wasn't the node. It was It
10:06 was the It was the routing model. So So
10:09 there's basically a model sitting on top
10:11 of all these other models. That's the
10:13 thing that was faulty at launch. Yeah.
10:14 No, it's not great, but listen, it's
10:17 they're they're launching fast and you
10:19 know, they launch it first and fix it
10:21 later. So, I mean, they also the the
10:24 other [ __ ] they did at launch was they
10:26 deprecated 40 without really telling
10:29 anyone. They just said, "Here's five.
10:31 Four is gone." And everyone that relied
10:34 on four to be their AI girlfriend, all
10:36 of a sudden chat GBT just didn't work
10:38 anymore. So, yeah, it was a bit of a
10:40 disaster. Um, but it's it seems like
10:43 it's better now. When are we going to
10:44 get some GPT erotica? So, so that is um
10:49 apparently one of the things OpenAI is
10:51 working on right now is an age
10:53 verification system. And if you're older
10:56 than 18,
10:58 you can get the uh it it's funny they
11:01 they're calling it the adult version of
11:03 of uh GPT that can do erotica. I don't
11:06 care so much about erotica, although
11:07 it's fine. is um like what
11:13 like like there's a lot of times like if
11:15 you just want to do like a Halloween
11:16 story.
11:18 It's just it doesn't like the the
11:20 subject matter so it it doesn't let you
11:22 work like there's a lot of [ __ ] you want
11:24 to do that that gets slammed by safety
11:27 guard rails that I think throttles
11:29 people from from using these things
11:31 fully. Thoughtful Thursday. We'll see.
11:35 We'll see how thoughtful it's going to
11:36 be. Um, let me look at something here
11:40 without sharing it. See if it does
11:42 anything interesting.
11:46 Not really.
11:49 Huh. Interesting. All right. Um,
11:54 let's see. Let's get a few more people
11:56 in here. We'll we'll get a few more
11:57 people in here, Brandon, before we talk
11:59 uh we talk stuff. Uh, if you're on the
12:02 YouTube there, if you're over on
12:03 YouTube, do me a favor. Subscribe.
12:06 I'm about 200% certain that even though
12:09 even the adult mode won't be able to do
12:11 anything when it's a a broad image
12:15 model. Yeah, that's probably true.
12:18 The reporting mcon from Brandon was
12:20 excellent. Yeah, if you haven't seen it,
12:23 so if you go to Brandon, if you'd pop
12:25 the AI salon uh what you call it up
12:28 there, the uh banner
12:31 um if you go to the salon.ai AI and and
12:35 just click join our community if you're
12:36 not a part of it or just go to the
12:38 community site which
12:40 community.thesalon.ai
12:43 um
12:45 in in the general in the community feed
12:48 which is like the second feed down when
12:50 you get into the community. Brandon did
12:52 a wrap-up of Mecon days 0, one, and two
12:57 um that are just really excellent. Like
13:00 he really, you know, captured what was
13:02 talked about and what was important
13:04 about it and you got good pictures and
13:05 things like that. I'm heading next week
13:08 to TED AI. Uh and do not expect this the
13:14 quality of reporting from me that you
13:16 got from him. Uh but it's really good.
13:18 So if you haven't seen that yet, it's
13:19 worth checking out.
13:21 Um, Grock is like, "The channels in the
13:24 hotel are unscrambled." It totally is.
13:27 Please start singing.
13:36 You know what we should do?
13:43 I know what we're gonna do.
13:49 uh soon.
13:53 Okay, here we go.
13:56 [Music]
13:58 All right, Sununo Studios here. We're
14:02 going to go create.
14:05 We're going to go audio. We're going to
14:06 go record Tik Tok pin. I'll be at at
14:10 OpenAI on Monday. About 150 folks from
14:12 higher ed. Oh, very cool. That's
14:14 awesome.
14:15 Uh yeah, if anyone's in San Francisco
14:17 next week, I'm I'm there Tuesday,
14:20 Wednesday, Thursday.
14:22 Um okay, so let's
14:26 Somebody told me to sing. Here we go.
14:32 Damn it. Hang on.
14:38 Um
14:50 [Music]
14:53 All right, we're going to turn that into
14:54 a song.
15:05 I think he meant with your guitar. I
15:07 know what he meant.
15:10 Now taking requests. Okay. lyrics. Uh,
15:14 you make them for me. Uh, write full
15:16 song. Uh, describe what you want. Um,
15:19 uh, silly man. Silly man sings. Silly
15:24 man sings. Silly man sing, I think, is
15:27 better.
15:29 All right.
15:31 So, let's see. He wears a hat too big.
15:33 He's got socks that don't match his
15:34 shoes. That's good. We'll select that
15:36 option. Styles, we'll go um goth rock.
15:42 Gothrock
15:44 um anthem.
15:47 All right, here we go. You wanted me to
15:51 sing? I'm singing, baby.
15:55 Uh
15:57 that uh is silly.
16:03 Mr. K, the educator Mr. K, who's got the
16:06 uh the uh YouTube channel. He looks like
16:10 he's going to get a pretty good gig here
16:12 shortly, which is exciting.
16:15 Um, here we go.
16:17 [Music]
16:30 [Applause]
16:35 He's got socks that don't match his
16:38 shoes
16:40 only.
16:44 [Music]
17:00 All
17:09 right. So, there you go. If you haven't
17:11 done that before, you should go do that.
17:13 All right,
17:15 producer Brandon, why don't you hop up
17:17 on stage here?
17:20 >> Hello.
17:20 >> And hello.
17:22 How are you? You must be exhausted,
17:25 first of all. Yeah.
17:28 Here.
17:31 There we go. We'll give you the We'll
17:32 give you the big floor.
17:33 >> Wow. All right.
17:34 >> Yeah. Right.
17:37 You should reframe it.
17:38 >> Yeah.
17:38 >> We can see your your AI salon logo a
17:41 little better.
17:42 >> Yeah. There you go.
17:43 >> Bend your camera down. Can you bend your
17:45 camera down a little or is it fixed?
17:48 >> Oh, it's
17:50 You have to see the the slop. It is my
17:54 office, but there we go. There you go.
17:56 Nice. All right. Beautiful. Beautiful.
18:00 Um, yeah. So, so, so when was it
18:02 started? What's today? Thursday. It
18:04 started Tuesday, right?
18:05 >> Tuesday.
18:06 >> Yeah. The, uh, the workshop session
18:08 started out on Tuesday. So, I was
18:10 hanging out, uh, getting access badges
18:13 and all that stuff. Uh, meeting up with
18:15 people. Met Gwen uh, from the salon who
18:18 was also there. We hung out for a little
18:20 while.
18:20 >> And Wednesday was the big day. Wednesday
18:22 was the the kind of all day uh,
18:25 conference. And uh it was it was a good
18:27 turnout. They were saying in uh 2022
18:30 this conference had about 200 people
18:32 coming to it and just over the past
18:34 three years it has exploded and uh there
18:37 were upwards of 2,000 people there uh
18:39 this week.
18:40 >> Wow. And I, you know, Kyle, I don't know
18:42 how many tech conferences you've been
18:44 to, but traditionally we see uh
18:47 depreciating value on attendance over
18:49 the course of the week uh by the last
18:51 night after people are going to the
18:52 happy hours and, you know, they've
18:54 missed all the keynotes and everything.
18:56 I'm just going to sleep in.
18:58 >> It was standing room only this morning
19:01 at 8 a.m. for the third day, which was
19:03 just mind-blowing to me.
19:04 >> Wow.
19:05 >> Um, so a lot of agency folks hanging
19:08 out. um a lot of uh entrepreneurs and
19:11 people who are just trying to figure out
19:13 AI and figure out what it means for
19:15 business. And it was interesting I was
19:17 talking to some of the agency leads and
19:19 you think everybody's got this figured
19:20 out. Uh and the agency folks all kind of
19:23 came in with at least the ones I talked
19:24 to came in with this idea that they were
19:26 going to take back lessons for their
19:29 clients. And a lot of them said, you
19:31 know what, I we're getting stuff for us.
19:33 Like this is this is gold for us to
19:36 figure out how to use AI. Uh, and then
19:38 obviously we're going to take that those
19:39 learnings back to our clients. But it's
19:41 interesting to see how fast it's
19:43 evolving that even the folks that you
19:45 trust with your marketing dollars are
19:47 still the first person learners for this
19:50 right now.
19:51 >> Yeah. Yeah, that that makes good sense.
19:52 Um, I just want to um Vicki said, "Dang
19:55 it, I thought Brandon said 30 minutes
19:56 late." He did say that because I told
19:59 him that and then I decided not to go to
20:01 Whiskey Club tonight. So then I unsaid
20:03 it. So sorry if you're late. That's on
20:06 us. That's on me. So anyway, carry on.
20:11 >> Yeah. But um you know, a lot of great uh
20:14 cool I met folks from Hey Gen there. I
20:16 met some digital first advertising
20:19 agencies. I met people who are doing
20:21 some really cool stuff.
20:22 >> Hang on for a second on those. On the
20:24 digital first advertising agencies, did
20:26 you talk to them at all about like their
20:28 business model? Are they doing value pro
20:30 pricing? Are they doing time and
20:32 materials? Are they like ex agency
20:34 people? Did you how how far into it did
20:36 you get with them? Yeah. So, for the
20:38 ones that I've talked to, and
20:40 everybody's got a different take on it,
20:41 but the ones that really caught my
20:43 attention were the ones that were trying
20:44 to democratize digital agency by
20:47 twinning people who are famous in the uh
20:51 marketing space and offering their
20:53 services uh on a monthly
20:55 subscriptionbased
20:57 where they've actually got the humans
20:59 behind the scenes prompting AI twins to
21:03 give you marketing advice.
21:04 >> Got it. Got it. Got it. That's
21:06 fascinating. Okay, cool. Awesome. All
21:08 right.
21:08 >> Yeah. So, Paul was there um Paul Ritzer
21:12 uh and his crew and they were very
21:13 generous with their time. Uh spent a lot
21:16 of time talking with us and you know the
21:18 the overall theory is that it is no
21:21 longer a question of if, it's a question
21:24 of how. Uh it's not a question of if
21:26 it's going to disrupt the industry, it's
21:28 a question of how fast can you adapt to
21:30 that disruption. And that was kind of
21:32 the overarching theme we heard again and
21:34 again. And um you know for all of the
21:37 focus on AI, it was really a human-
21:40 centered conversation. A lot of the uh
21:44 folks focused on what do the humans need
21:48 to know? So it wasn't I mean there was a
21:50 little bit of prompting. Um one of the
21:52 presenters uh Jeff Woods had a really
21:55 cool prompting uh technique. He called
21:57 his cry method which is uh context uh ro
22:02 uh interview and um t whatever t stood
22:07 for um uh but it is an idea that you're
22:13 wanting the AI to ask you questions
22:18 about what it needs to know to answer
22:20 that role. And uh yeah, it was overall
22:24 everyone was super friendly and um it
22:27 was in Cleveland, so that was nice. You
22:29 know, it was a lot of people from in
22:31 from out of town, so we got to showcase,
22:33 put our best foot forward uh in
22:35 Cleveland. Uh it's a great place to host
22:37 an event. And um Paul actually said
22:39 that, you know, he had the option when
22:41 they were planning this thing. Do we go
22:42 to New York? Do we go to Chicago? Do we
22:44 go to Austin or San Francisco? So
22:46 there's just something authentic about
22:48 having it in Cleveland uh where we have
22:51 more of about the community. Um so yeah,
22:55 it was it was a really cool experience.
22:57 >> Okay, cool. Mary Mary uh Paul's
23:00 Brandon's talking about uh the marketing
23:02 AI institutes make their annual
23:05 conference that he just got back from.
23:08 Um, what for the I I I know you were
23:12 having AI listen to the keynotes, but I
23:14 assume you were listening as well. What
23:17 are are there three big takeaways from
23:19 the things that you watch that that are
23:21 that stick with you? Like what are the
23:24 insights that made you sort of [ __ ] your
23:26 head or feel like something's different?
23:28 Or was it more about the people you
23:29 talked to?
23:31 >> Um, so I it's a little bit of both. I
23:33 mean, the conversations were fantastic
23:35 and everyone was so open and so willing
23:39 to share where they were at with their
23:40 AI journey.
23:42 >> Yeah. Figure it out. Right.
23:45 >> Exactly. You know, it was it was very
23:47 much a, you know, an open comfortable
23:50 environment. No one was gatekeeping. No
23:52 one was trying to sound smarter than
23:55 they are. Everyone was just, it takes a
23:59 village. It takes a community. and
24:01 everyone was just a very willing
24:03 participant. Um, a lot of interest in
24:05 the AI salon, both from potential
24:08 partners to potential speakers to just
24:10 people who, you know, saw the shirt and
24:11 they're like, "Hey, what's the AI
24:12 salon?" And so got to talk to people
24:15 about that. But, you know, in terms of
24:17 mainstage takeaways, um, there was a
24:20 woman from McDonald's, she's their chief
24:22 digital analytics officer, and um, she
24:25 had everybody in the crowd raise their
24:26 hand and say, "How many of you have an
24:28 AI strategy?" And like 80% of the people
24:30 in the room raised their hand. They go,
24:31 "Okay, cool. How many of you have an
24:33 internet strategy?" Like everybody kind
24:35 of put their hands down and looked at
24:36 her. You know, the point was you don't
24:38 have an internet strategy. You just use
24:40 the internet to market.
24:42 >> Yeah. Exactly. Oh, that's great. That's
24:44 great. Yeah. That's Well, that's the
24:46 thing is like people had had, you know,
24:49 chief chief internet officers for a long
24:51 time or, you know, chief worldwide web
24:53 officer and then it just gets absorbed
24:55 and now we're seeing the same thing,
24:56 chief AI officer. So fascinating
25:01 >> and I I think that the biggest thing
25:02 that you know uh Paul and company kept
25:06 coming back to on the stage is that it
25:08 you know it's humans first and you know
25:10 the humans are the marketers AI is the
25:13 tool and they really drew that line
25:15 pretty firmly in the sand. Uh where this
25:18 wasn't a tech conference we weren't
25:20 talking about prompt engineering we
25:22 weren't talking about how to
25:24 >> model Yeah.
25:26 >> Exactly. Yeah. We were talking about
25:29 storytelling and um one of my favorite
25:32 presenters was and I want to get him on
25:34 for the uh AI salon here, maybe even
25:36 Festivist u PJ Ace um started a digital
25:40 first advertising agency that uses VO3
25:45 uh as its base model to build these
25:47 commercials and build these marketing
25:49 assets. Uh and now Sora too because he's
25:52 already disrupting his own industry. But
25:54 he shared with us one of the first prime
25:57 time network spots that his agency
25:59 produced and I've got it pulled up here
26:01 if you want to take a look at it. It he
26:03 couldn't believe Disney because it was
26:05 it aired during the March Madness and uh
26:08 NBA Finals inside of on ABC. He's like,
26:11 I can't believe we got Disney to approve
26:13 this crazy commercial that we made with
26:15 AI. Uh so what you're about to see
26:18 >> I wanna I want to tee it up. I just
26:20 wanna I just want to put a pin in
26:22 something. the the focus that you saw,
26:24 the fact that they they said, you know,
26:26 it's it's humans first. This isn't a
26:28 technology conference. That was what we
26:30 organically experienced last year during
26:33 Festivus. That we had 34 different
26:35 speakers over 24 hours and not a single
26:38 speaker like we didn't we didn't The way
26:41 Festivus works is if you're invited, you
26:43 talk about whatever you want. And and 34
26:46 out of 34 speakers, no one led with
26:48 technology. they were all talking about
26:50 humans and changing the mindset and
26:53 things like that. So, it that that's a
26:56 reassuring thing to me, Brandon, that
26:58 you know, it didn't go there and there
26:59 wasn't the tech track versus this that
27:02 it really is just about us dealing with
27:03 this. So, speaking of that, yeah, let's
27:05 let's see this this spot. Um, and uh
27:08 that's very cool.
27:10 >> Yeah. And so, I'm going to go ahead and
27:13 here I'll stop pull your screen shown
27:15 and and there there we go.
27:19 So everything you're about to see here
27:20 was generated with V3.
27:23 >> Indiana going to win, baby.
27:24 >> We're in Florida asking people what they
27:26 put their money on. I'm all in on OKC.
27:30 >> Indiana got that dog in them.
27:32 >> Will egg prices go up this month?
27:34 >> I think we'll hit $20. How many
27:37 hurricanes do you think we'll have this
27:38 year?
27:41 [Applause]
27:43 >> Kelshi lets you legally trade on
27:45 anything anywhere in the US. Casey.
27:49 >> Oh, Casey.
27:56 >> My favorite was how how many hurricanes
27:59 do you think we're going to see this
28:00 year?
28:04 So, yeah, all all three days are
28:06 recapped on uh the AI salon inside of
28:08 the feed. And um yeah, it was a lot of
28:10 fun. Can't wait to go back next year and
28:13 um you know hoping to have some of those
28:14 guys come back and talk to us a little
28:16 bit more about their uh chain of craft.
28:18 >> Beautiful. Fantastic. Cool. Well, thank
28:21 you for that. I know you're exhausted,
28:22 so I appreciate you coming up here and
28:24 uh and yeah, so um so that's the things
28:27 that's one of the things that I want to
28:28 do more of in 2026 is as these events
28:31 come up, get people from the salon out
28:33 to these events and and uh you know,
28:36 learning and meeting people and uh you
28:40 know, ex expanding their community and
28:42 expanding the salon community. Um I
28:44 think it's a it's a big big deal. Um so
28:47 thank you. Um what was I going to say? I
28:50 I mentioned last night that I had a
28:52 meeting today about Sydney uh the
28:55 musical and with some potential folks
28:57 that you know could potentially do
29:00 something with it. That meeting went
29:01 very well. Uh the way these things work
29:04 if you're if you're not used to the
29:06 professional
29:08 professional creative development world,
29:11 you have you kiss many frogs. There's
29:13 lots and lots of conversations before
29:14 stuff happens. Uh but it was a really
29:16 good meeting. So that was that was
29:18 exciting. Um, I'm trying to think what
29:21 else. I want to go play with runway
29:24 because they have this new thing called
29:25 apps. I again I I said this last night
29:30 and you know the the the way I tend to
29:34 take the focus of this channel is
29:37 largely driven by instinct. I'm not I'm
29:41 not a planner. Uh, in case you in case
29:44 that wasn't clear. If you're new here,
29:46 my name is Kyle Shannon. the AI learning
29:48 lab and we'll we'll do some stuff
29:50 tonight. If anyone has any questions or
29:52 thoughts about AI, I'm happy to chat. Um
29:56 but but uh but one of my instincts right
29:59 now is that
30:04 video that we just saw that that played
30:06 on national TV during March Madness that
30:08 was just created with VO clips
30:13 the the creatives who are on the
30:15 sideline that hate AI and that are angry
30:19 at AI and it's not real creative and it
30:21 can't be creative.
30:24 That commercial, whether you like it or
30:26 not, was not really an AI
30:29 commercial. It was a commercial from an
30:32 agency that had an idea to say funny
30:34 things about things going on, you know,
30:37 in Indiana, I guess, or wherever it was.
30:41 It was the vision of the creative
30:42 director that had the idea that happened
30:46 to use a tool like Sora rather than
30:49 going and filming people doing that just
30:51 used a tool like this stayed committed
30:54 to the fidelity of the idea which is a a
30:57 term that I'm I'm stealing blatantly
30:59 from Liz Miller Gersfeld who's my
31:01 co-host in the AI salon.
31:04 She talks about doing work and being
31:06 committed to the fidelity of the idea
31:08 and I love that. I I I love that
31:10 concept.
31:12 Um,
31:14 so you see a commercial like that and
31:19 the the the the
31:23 feeling that I'm having is very very
31:25 reminiscent very reminiscent of when you
31:30 first saw worldwide web URLs on TV
31:34 commercials or when Brian Gumble and
31:37 Katie Kirk were like, "What's that a
31:39 with the circle around it? What is this
31:41 internet anyway where websites and
31:45 internet and web addresses what's this
31:47 http
31:49 colon line line thing
31:55 right like people didn't know how to
31:57 talk about them they didn't know where
31:58 to put them was an event when you saw a
32:02 URL on TV or on a billboard
32:05 um it it it there was we just didn't
32:09 have a place to to put it in context
32:13 because it had never existed before. I
32:16 think these kinds of videos and things
32:18 like Sora's, you know, new Tik Tok like
32:22 app that that's got all these memes. you
32:24 can really easily make a meme of
32:25 yourself or other people kind of lives
32:28 in this strange place where it's not
32:31 going to be the final form but we're
32:33 going to be able to look back at it and
32:34 say ah that's the moment when you know
32:37 ease of using someone's likeness started
32:40 was the Sora app that was really the
32:42 beginning of it um or or you know that
32:46 commercial that we just saw in March
32:47 Madness I don't know if that was the
32:48 first AI commercial to go national but
32:51 you know it was a big one right so so
32:53 people are going look back at that and
32:54 go that that was the moment.
32:57 The other thing that is striking me and
33:00 and and I guess why I'm talking about
33:02 this is
33:04 it's it's my gut that we need we need to
33:08 be exploring
33:10 these video tools more.
33:14 Not that large language models and
33:17 multimodal models and chat GBT and
33:19 application development is not
33:21 important,
33:22 but I think that we're going to very
33:24 quickly
33:26 transition from
33:28 these tools making videos and and short
33:32 video clips
33:34 to these tools being worldbuilders
33:38 and and so creating 3D worlds that you
33:41 can enter. I saw a a a tweet right
33:44 before I came online tonight. Um, a guy
33:48 is building a Sora based world that
33:52 every time someone creates a new image
33:55 in Sora, I guess within his app, it
33:59 basically adds a node to this world that
34:02 you can explore.
34:04 And so like people are already starting
34:06 to take these video generation tools,
34:09 especially the ones that generate things
34:10 quickly, and they're doing creative
34:11 things with them.
34:14 Runway has has already said openly that
34:17 that Runway is not a video company.
34:20 They're a worldbuing company and and
34:22 that that one of the next things coming
34:24 from them is a worldbuilder.
34:27 And so that starts to look like you can
34:29 create a world you can enter. you can
34:31 create a set that you can move a camera
34:34 around and have complete consistency,
34:37 you know, no matter what's in that world
34:39 because you're moving a camera around an
34:40 existing, you know, world.
34:43 Um, you're going to be able to
34:46 dynamically create video games.
34:49 And then I just I also have this
34:51 sneaking suspicion that there's going to
34:54 be a new kind of communication that
34:56 comes out of this where you will enter
34:58 information worlds and and not just like
35:02 information visualization, but I I don't
35:04 know some some sort of you might enter a
35:07 news world where you can go to a bunch
35:09 of different news stations and they
35:11 they're all talking about news in
35:12 different way. I I don't know. I don't
35:14 know what it is. It's just my instinct
35:16 is it's important. We should pay
35:17 attention to it. Um,
35:20 so yeah. So with that, I figure we'll go
35:23 play with a little video. I also, um,
35:26 yesterday,
35:28 um, VO3.1
35:30 came out, and apparently one of the
35:34 settings you can do is you can set the
35:36 video for 25 seconds. And so I thought
35:39 it might be fun to to have Chat GBT
35:42 write a prompt that is a a single prompt
35:47 for 25 shots in a 25 second video. So a
35:51 shot per second. Um and see how it does.
35:54 Maybe we'll start with that. That's kind
35:55 of fun. Look how Tik Tok has evolved.
35:58 Yeah, exactly. Tik Tok is now full-on AI
36:01 AI augmentation, right? Can you imagine
36:03 creating a world that serves a specific
36:06 function? concerts that go beyond a
36:08 stage. That That's exactly what I'm
36:10 thinking. Side hustle is is
36:15 the minute the minute that you have
36:20 instant world creation
36:23 and and within that world, you can you
36:26 can also vibe code, right?
36:30 So, so imagine
36:33 creating a world that is nothing but
36:35 video games and then people can come in
36:38 and everyone can vibe code their own
36:40 video game. You set it up in such a way
36:42 that basically, you know, you can you
36:44 can take a token and you can play a game
36:46 or you can take 10 tokens and you can
36:48 make a game and then we're going to
36:50 split how much you make on the game and
36:52 and everyone comes in and invent invents
36:54 video games and it's a dynamic infinite
36:58 video arcade. I don't know. A video game
37:00 arcade that that could be a thing. I'm
37:03 busy trying to make Sora 2 put me on
37:06 on stage with Ace Freely. Very nice
37:08 Shane Saw. Good. Good luck with that.
37:11 Um, or a business presentation that was
37:14 a PowerPoint and now we have movies and
37:16 [ __ ] Yeah. Or or a business
37:18 presentation where you invite people
37:20 into the presentation.
37:23 There was a uh there was a play in New
37:25 York City in the in the late 80s and
37:29 early 90s called Tony and Tina's
37:31 Wedding.
37:33 and Tony and Tina's wedding. It was a it
37:36 was a it was a theater piece that was
37:39 held in a in an upper east side or upper
37:42 west side, I think, in an upper west
37:44 side brownstone. So, it was a three or
37:46 four story big building. Um, and the you
37:50 you walk in as the audience and and
37:53 you're getting ready for Tony and Tina's
37:55 wedding. And the play was the audience
37:59 can go to any room they want and and all
38:02 of the actors are playing the characters
38:05 that they're going to play during the
38:06 wedding. And like you could go into the
38:08 room where the where the bride is
38:10 getting ready or you could go hang out
38:12 with the groom and his buddies or you
38:14 could go, you know, where the caterers
38:16 are making the food. And so so your
38:19 experience of the play was 100%
38:23 dependent on where you went and and it
38:26 it was part improvisation, part scripted
38:29 and it culminates with a wedding which
38:31 you get to go sit in and watch and
38:33 participate after having heard all the
38:35 stories, you know, of of what went on.
38:39 So imagine a business presentation being
38:41 like that where you're like, "Okay, the
38:44 annual report this year, you can walk
38:46 into it and and right now we're going to
38:49 walk in and we've got we've got, you
38:52 know, the CEOs in this room and we've
38:55 got the six different executives that
38:57 head up our our six different divisions.
38:59 They're all going to be in these
39:00 different rooms. You can take all the
39:02 time you want. There's gonna we're going
39:03 to be doing this live for the next two
39:05 hours and so you can ask us questions
39:08 and look at charts and graphs and pull
39:10 them off the wall and what right why not
39:14 and if you don't think that
39:18 like like Brandon was talking about the
39:20 the ad agency that's this all digital
39:22 agency where they're creating um virtual
39:26 agency employees that are based on
39:28 famous agency people.
39:31 That's like scratching the surface of
39:33 the kinds of companies that are going to
39:35 be created to to take these capabilities
39:39 and stitch them together in ways that
39:41 none of us could have imagined before.
39:44 So, with that, um, let's let me let me
39:49 jump over to Flo for a second here
39:54 and share my screen so y'all can see.
40:00 Um,
40:02 [Music]
40:05 okay.
40:07 V3 fast.
40:14 Let's see. I don't see
40:18 where you can change timing.
40:25 [Music]
40:29 Let's go to Gemini.
40:32 Maybe it's there.
40:35 Create video.
40:44 No, I don't see.
40:48 Does anybody know where VO's timing is?
40:55 I can also imagine this being a
40:57 monumental win for Neurospicy Non-verbal
41:00 Friends. Absolutely. I hear lawyers
41:03 licking their lips. Yeah. Sorry, you you
41:07 didn't go into the room where we were
41:09 doing the insider trading, Bob.
41:15 A coworker said, "I'm sick and tired
41:17 of.com this and.com that." Yep. That was
41:21 that was just just like you have people
41:23 in your life right now going, "Could you
41:25 just not talk about AI anymore, Kyle?
41:28 It's really annoying." Yeah, we I heard
41:30 all that in the 90s because I was in the
41:32 theater and I like I would come into
41:35 rehearsals for a play and I'd be like,
41:37 "Oh, this you should see what we're
41:39 doing at with this web stuff." And could
41:41 you shut up about that.com stuff, Kyle?
41:43 You dumb dumb.
41:48 Um, I don't know how to I don't know how
41:50 to change Oh, wait. Let me go let me go
41:53 back to flow. Maybe I need to reload the
41:55 app.
42:00 Reload.
42:02 New project.
42:06 Expand film. No newir action. Create new
42:09 expander. Cinematic.
42:13 What?
42:15 Create new expander.
42:18 Oh, wow. Oh, these are kind of styles
42:20 within
42:22 That's kind of cool.
42:26 Yeah, you lose them if you
42:34 supporting support ending soon. Oh,
42:37 okay.
42:42 Huh?
42:46 Oh, you know where it could be.
42:51 By the way, if you're new here and if
42:54 your question is, are we just going to
42:56 watch an old guy try to figure out where
42:58 something is? Yeah, that's pretty much
43:01 the channel.
43:05 Aistic studio.google.com.
43:07 Follow along, won't you? You can be
43:09 clueless, too.
43:12 My job is to make it okay for you to
43:14 feel clueless. Okay. VO 3.1.
43:20 Um,
43:23 make changes. Select API. A A A A A A A
43:25 A A A A A A A A A A A A A A A API key
43:26 required for video.
43:31 Okay. I'm exhausted. I'm exhausted.
43:34 We're going to go to runway.
43:37 Okay.
43:38 Um,
43:43 share this tab.
43:47 All right. Fantastic.
43:50 Let's see what we've got here.
43:56 This is a video we did of Ann Murphy's.
43:58 Oh, did we do videos of these? I don't
44:00 think we did. We can make a video.
44:04 Woman in a beach pullover sits in a
44:07 lounge chair on the beach with her salt
44:10 mask warming her face.
44:15 All right, we'll do we'll do this one.
44:16 We'll turn it into a video.
44:19 All right. So, we'll go um
44:24 woman
44:26 hums
44:29 a jy
44:32 little tune
44:38 while a seagull
44:41 poops on her shoulder.
44:45 Is an here?
44:47 Sorry I'm late. Just finished a speaking
44:50 gig. Cool. Congrats.
44:54 Cool, cool, cool, cool, cool, cool,
44:56 cool, cool. Beautiful. All right, so
44:58 we're going to generate this into a
44:59 video.
45:01 All right, so let me show you what we
45:03 got going here.
45:07 Um,
45:10 so Runway now has these thing things
45:12 called apps.
45:15 So you can change the image style. So
45:18 let's try that. Let's change the image
45:19 style. Um rough expression. Oh, this is
45:23 cool.
45:26 Can I change?
45:29 How do I use it? I thought he was the
45:33 expert. Isn't he supposed to be teaching
45:35 us how to do that? Let's see. Use as
45:38 reference. There we go. See? You see
45:41 what I'm saying, people? All right.
45:43 apps. Change image style.
45:47 All right, that didn't work. Shut up.
45:51 I dragged and dropped it. Okay, here we
45:54 see you see you see people
45:57 magazine collage. So, we're going to
45:59 turn an into a magazine. No, let's turn
46:01 Let's turn this into a storyboard.
46:04 All right. Generate. Can I No. Why did
46:09 this uh
46:11 u make this a storyboard? Oh, please
46:16 switch to credits mode. Okay.
46:18 Okay. So, I don't need a prompt.
46:21 All right. So, we got this image. We're
46:23 going to turn it into a storyboard.
46:24 Bang. And then that's going to be down
46:27 there. Beautiful.
46:29 Beautiful. Beautiful. Beautiful.
46:33 All right. Now we should have a video of
46:36 Ann humming here and then we can play
46:39 with these apps.
46:41 So it looks like one of the apps is
46:43 there was an issue on our end. The any
46:45 used credits will be refunded. Generate
46:47 again. Let's get rid of that.
46:52 93%
46:58 [Music]
47:02 [Laughter]
47:04 Just ordering my cola crunchers. Did you
47:06 see my my chip commercial I did? All
47:09 right, here we go. Look at this.
47:12 So, what did I say? She hums a little
47:14 tune, right?
47:20 Why is it got no audio?
47:30 It vomited on her shirt.
47:39 All right. And the that thing's not
47:42 working at all. All right. Let me go
47:44 find a different video because that's
47:45 kind of a crappy video. Let's find a
47:47 better video.
47:53 D.
47:54 [Music]
48:00 [Music]
48:10 All right, that one's good. All right,
48:11 so this is one. This is another one we
48:13 did of of Ann Murphy. It totally made
48:15 her into a different woman. But let's
48:18 take that.
48:20 So, we got key frames, subject
48:22 reference, remove from video.
48:27 All right,
48:29 let's do remove from video. Upload video
48:32 to edit. Can I drop this here? Yes, I
48:34 can.
48:36 And then can I
48:41 How do you do it?
48:43 He doesn't know what he's doing. Oh,
48:45 okay. So, we'll say remove the telephone
48:49 pole.
48:54 You know, this is called the AI learning
48:56 lab. I came in here to learn and I feel
48:58 like all I'm doing is watching someone
49:00 else try to figure something out. I
49:03 could do that on my own. I know, but
49:05 you're not on your own, are you? You're
49:07 here with other people and they're
49:09 making fun of you right now. I wouldn't
49:12 stay. Okay. Shane saw a Tik Tok pin. Can
49:15 you you can get paid for pe for
49:19 speaking? I do that all the time.
49:23 Okay. So, we're going to remove the
49:25 telephone poll. Generate 5 seconds.
49:29 Okay.
49:31 So, now let's go back to apps.
49:35 That was removed from video. Upscale
49:38 video. Reshoot product. Add dialogue.
49:41 Let's do add dialogue. Boom. Wait.
49:45 Oh, an image. Sorry, that's not going to
49:48 work. Change the weather. Okay, this is
49:51 good because we got all sorts of crazy
49:53 stuff there. So, we're going to go um
49:57 it is a bright
50:02 sunny day
50:05 at
50:06 sunrise.
50:09 Sunrise.
50:13 Oh. Oh, and look, they've got presets.
50:15 Smoke, thunder.
50:19 Let's do thunder.
50:21 Generate. All right. Beautiful. Go back
50:24 to apps.
50:26 Uh, change weather. Change time of day.
50:30 We'll make it early evening. We'll go
50:32 twilight. We'll go midnight. She She's
50:35 running at the Big Bunny Motel at
50:37 midnight.
50:39 Okay. Generate.
50:46 Did I say generate?
50:48 Yes. All right. Back to apps.
50:53 Change the background. Beautiful.
50:57 Um Oh, look again. They have presets for
51:01 you.
51:02 So, she's going to be jogging in an
51:04 alley.
51:08 I I wish Ann Murphy was here. We're
51:11 We're picking on her her third cousin,
51:14 Elma.
51:19 Okay, that's Ally. All right, let's go
51:21 back to apps,
51:23 then relight the scene. Oh, I like this.
51:26 Okay. Oh, and again, we've got blue
51:28 light, rainbow, lens flare. Let's do
51:32 lens flare. Or police. Police is cool.
51:35 Like she's running from the cops. I like
51:37 it. Okay. Beautiful.
51:46 Generate. All right, that's off.
51:51 Relight. We did change background. We
51:53 did add performance.
51:56 Add a driving performance. Shoot. Okay,
51:59 this is going to be good. I'm going to
52:01 record something.
52:04 Allow when using the site. Okay. Film
52:07 yourself from the waist up.
52:16 Oh my god, I'm so tired. Running away
52:19 from those awful people. What do I do?
52:23 Where do I go? Someone help me. Okay.
52:27 >> Oh my god, I'm so tired. Running away
52:31 from those awful people. Okay, good.
52:34 Perfect.
52:42 Um, can I drag over this video? No, I
52:45 have to select an asset. All right,
52:47 we'll go we'll go select some other
52:49 asset. Let's see what we got here.
52:52 Um,
52:55 let's see.
53:00 [Music]
53:05 This is good.
53:11 Facial expressiveness three. Let's let's
53:14 max that to five.
53:16 Use motion and gestures. Yes.
53:20 Do we want this 9 by6? We'll try it.
53:24 Generate. All right. So, do we have
53:27 Okay. Remove from video.
53:30 removed the telephone pole. It And it
53:34 rem It removed the sign.
53:39 Look at this. It removed the sign.
53:49 Oh, good lord. Janky is as Janky does.
53:54 We embrace the jank. Yeah. All right.
53:57 Thunder
54:00 change the weather.
54:07 Not super realistic lightning, but
54:10 definitely changed the lighting and the
54:11 weather. And it's raining. Kind of cool.
54:14 Beauty.
54:16 This is change it to midnight.
54:26 Very cool.
54:29 That one's background alley. It didn't
54:31 start. Why? Your video is rendering.
54:35 Here's Oh, here's the police version.
54:39 Wow, this is cool.
54:43 It turned the building into a cop car.
54:51 That's pretty hilarious.
54:56 Oh, did I not generate the character?
54:57 Oh, yeah. It's coming in.
55:01 Here we go. All right, people. It's
55:04 going to be beautiful. Wait, what
55:06 happened?
55:09 Why can I Why can I not zoom in on Oh my
55:15 That's strange.
55:18 Oh my god, I'm so tired. Running away
55:22 from those awful people. What do I do?
55:26 Where do I go? Someone help me.
55:30 Oh my god, I'm so tired.
55:36 Bravo, Bradley Cooper.
55:41 I prefer Mr. Brando.
55:51 Um,
55:53 let's go to tools for a second here. Oh,
55:55 that's just the tools for this thing.
55:59 Then you've got video models, image
56:01 models, video models.
56:04 Oh, yeah. You can choose them all here.
56:06 Oh, VO3.1.
56:08 Ah, look at this. Oh, no. Only 8
56:11 seconds. All right. Damn it.
56:15 H.
56:17 You can chat your things into existence.
56:20 Hello, I'm Runway's creative assistant.
56:26 It's a drag these things take so long.
56:28 Let's Let's jump back over.
56:33 Hm.
56:36 I'm really bummed we can't we can't make
56:39 a a VO3 3.1 video for 20 25 seconds.
56:45 I don't know where that is.
56:49 Let me share this. Let's go to Sora.
56:52 Let's see if Sora did anything to their
56:54 interface. I heard I saw somewhere that
56:57 they've got a new interface.
57:00 Oh, it was Sora that was 25 seconds.
57:03 Your session expired. Go back.
57:09 Authentication error. Retry.
57:16 This page could not be found. Fine.
57:23 Why are you saying.com all the time?
57:25 What's this dot? What's this period coom
57:28 thing everyone's talking about?
57:31 I don't get it. I don't get it.
57:35 Why are you always the thing? Artificial
57:36 int is the chip that break. I'm Kyle
57:39 Shannon and this this is the chip that
57:40 breaks every rule. We ditch the oil and
57:42 fried it in cola. That's how you get
57:43 that sweet salty snap and a little fizz
57:45 of magic. Cola crunchers. Light crispy
57:47 caramelized. I'm Kyle Sh.
57:51 I'm making commercials now for fake
57:54 products. Um
57:58 [Music]
58:01 You can do 15 seconds now. All right,
58:03 we'll do 15 seconds here. We'll do
58:06 landscape orientation.
58:09 Did you see my
58:12 Madagascar
58:15 vanilla beans cameo? No, I didn't. Wait,
58:20 let's go to look at cameos.
58:24 Looks like Brandon made a little
58:26 something for me.
58:28 Uh oh. Official intell.
58:31 >> All right, let me get this finished up
58:32 for you. I thought the teal would suit
58:35 you and a little ribbon cuz why not?
58:36 There we go.
58:37 >> Really didn't have to, but I'm thrilled.
58:38 Looks great.
58:39 >> Spark some ideas.
58:39 >> I'm sure it will. Thank you.
58:41 >> All right, let me get this finished up
58:42 for you.
58:44 >> Love the color.
58:45 >> Thought you might.
58:46 >> Let's see what's inside. Oh, this is
58:48 beautiful. A leather notebook
58:49 >> for all those ideas. Kyle, happy early
58:51 birthday.
58:52 >> Wow, thanks. Love the color.
58:54 >> Thought you might.
58:55 >> Let's see what's inside. Oh, this is
58:56 beautiful. a leather.
58:58 >> Okay. So, I want I want you to take a
59:00 look at something here.
59:03 So, notice that at the end of the video,
59:07 Dr. Jay has got
59:10 uh, you know, a little card with her
59:12 email on it, and I've got the the join
59:15 the salon thing.
59:17 When you make your your cameo character
59:22 in the Sora app,
59:25 you can go into preferences
59:27 and then there's a thing called
59:29 description, I think,
59:32 and you can put in a system prompt
59:35 that tells you that tells the system
59:38 what to do. So, let me go to settings,
59:42 Cameo Access, and then
59:45 if I go to Cameo Rules,
59:50 see here.
59:56 Uh, it's not letting me edit it, is it?
59:59 Oh, yes it is. At the end of the video,
1:00:01 make a little sign that says,
1:00:05 "Join the AI salon, the salon.ai.
1:00:09 So, let me just see if that actually
1:00:10 saved my change. It did. Good. Because
1:00:14 the uh the the mobile version doesn't
1:00:16 save the change. So, anything I do now
1:00:18 is that Where's your Madagascar vanilla
1:00:20 beans there, Brandon? It must be in
1:00:22 drafts, maybe.
1:00:33 No, you published it. I don't see it.
1:00:37 Um,
1:00:39 if I go here,
1:00:42 let me go to notifications. Maybe it's a
1:00:44 notifications. Oh, this this app just
1:00:47 crashed. God damn, this thing is so bad.
1:00:50 Fix it. You've got billions of dollars.
1:00:54 Fix your stupid app.
1:00:56 [Laughter]
1:00:58 Hang on. Hang on, people.
1:01:02 Um, Sora.
1:01:04 Sora. Hello. What is your name, sir? My
1:01:08 name is Sora.
1:01:10 Sora. That's a rather peculiar name. Ah,
1:01:13 at Brendan Tid. Got it. Let's go to
1:01:20 search.
1:01:22 Search search
1:01:24 Brandon. Brandon
1:01:28 Ted.
1:01:37 tabs. Share screen. Shut up. I'm a
1:01:42 professional, Brandon. I don't need
1:01:44 handholding. Except when I do, like
1:01:47 right now.
1:01:52 I'm a defensive professional. I'm a
1:01:54 defensive creative professional.
1:01:57 I will claim independence. Well, while
1:02:00 all the while needing my diaper changed.
1:02:04 >> Gar vanilla beans. These are plump,
1:02:05 oily, and absolutely bursting with
1:02:07 aroma. And I'm making them available
1:02:08 right here on the live. Look how bendy
1:02:10 that pot is when you split. Friends,
1:02:12 check these out. Fresh Madagascar
1:02:14 vanilla beans. These are plump, oily,
1:02:16 and absolutely bursting with aroma. And
1:02:19 I gotta tell you, I love these. I got to
1:02:22 make a new one with a better shirt
1:02:24 because I hate that shirt. I've got I've
1:02:26 got great shirts and yet
1:02:30 that's my my one and no one changes the
1:02:32 shirt. So I gota I gotta get like a nice
1:02:34 hip maybe I'll do like a Hamilton shirt
1:02:36 or something like that. Um that's pretty
1:02:38 cool. I'll hardy heart it share remix.
1:02:47 [Music]
1:02:49 All right, let's make a 15sec when you
1:02:51 learn how to
1:02:52 >> Let's make a 15-second video. Okay,
1:02:56 we're doing 15
1:03:00 landscape. It It looks like uh it looks
1:03:03 like a lot of you are having the same
1:03:05 issue that I had, which is you set it as
1:03:08 landscape and then it makes a vertical
1:03:10 video inside a landscape shell, which is
1:03:13 just stupid. So, we'll do olden timey
1:03:20 black and white news broadcast
1:03:24 where
1:03:27 Kyle Shannon
1:03:32 says
1:03:36 and they still
1:03:39 Oh no, the app just crashed again.
1:03:43 Come on, take that question. Kyle, can
1:03:46 you try using my cameo on Sora? Sure.
1:03:51 Side hustle Mimi. Are you Side Hustle
1:03:52 Mimi here? And do you have it? Wait. And
1:03:55 And they still Wait. And the And the
1:03:59 killer is still at large.
1:04:05 Here's
1:04:08 Here's
1:04:13 Mimi with the weather.
1:04:19 Cut to
1:04:22 let's see outside hustle. Oh, come on.
1:04:26 Do not crash.
1:04:28 Come on.
1:04:33 [Music]
1:04:37 Wait. I'll wait for it a little bit. It
1:04:39 may come back.
1:04:42 And I can't even I can't even
1:04:46 select the text because it's crashing.
1:04:50 Telling Sor not to crash.
1:04:54 Fun size AI. That's your Okay, that's
1:04:56 your Okay.
1:04:58 Um I'm gonna have to exit page. Damn it.
1:05:00 Okay.
1:05:04 Reload.
1:05:07 Okay,
1:05:08 let's let's go back to duration 15
1:05:12 landscape. Um,
1:05:15 I'm going to put wide
1:05:18 old 60s news broadcast. Well, no, let's
1:05:24 do let's do 70s. We'll do we'll do we'll
1:05:27 do 70s news
1:05:31 broadcast
1:05:34 with
1:05:36 Kyle as the
1:05:39 Oops.
1:05:47 um as the anchor
1:05:50 with a big mustache.
1:05:57 And he says,
1:06:01 and the killer is still at large.
1:06:12 Now, here's
1:06:15 Mimi
1:06:17 with the weather.
1:06:26 Um,
1:06:28 cut to
1:06:30 at
1:06:34 just hung again. Oh, no, it didn't.
1:06:36 Okay, it sort of did. at fun
1:06:40 size AI.
1:06:44 Uh, I can't use your thing.
1:06:47 I got I have to follow you. Wait, hang
1:06:50 on. Let me copy this so we don't lose
1:06:51 it. So, let me go find you and then I'll
1:06:54 follow you. You may need to follow me,
1:06:56 too. Fun
1:06:59 size AI.
1:07:02 Six followers. Follow.
1:07:05 Cut to it. Fun size AI. Okay, I had to
1:07:07 follow you. You can set your You can set
1:07:10 your uh permissions that anyone can
1:07:13 cameo you, which is how I have mine set
1:07:16 because I'm an idiot.
1:07:21 Cut to fun size AI. Um
1:07:25 in a in a um
1:07:30 70s
1:07:31 uh what do they call those? I don't
1:07:33 know. 70s dress, whatever. Um, and big
1:07:37 hair
1:07:41 talking about
1:07:43 it raining
1:07:46 cats and dogs
1:07:50 then.
1:07:52 Uh,
1:07:54 oh no, it just hung again. I was getting
1:07:57 too ambitious. Come on. No.
1:08:01 No.
1:08:05 This is so frustrating.
1:08:11 Oh, I'm not even sharing.
1:08:14 I type my prompts outside of Sora and
1:08:16 copy paste. That's a great idea.
1:08:20 That's a really good idea, Brandon.
1:08:23 Let me close. I wasn't sharing anyway,
1:08:25 so it didn't matter.
1:08:28 Okay, let me go to notes.
1:08:34 [Music]
1:08:41 [Music]
1:08:43 New. Oops. New. Oops. New.
1:08:48 Paste. Okay. There's that.
1:08:51 There's that.
1:09:03 Okay.
1:09:06 Wide old 70s nude broad news broadcast
1:09:09 with Kyle Shannon as the anchor with a
1:09:11 big mustache. And he says, "And the
1:09:12 killer is still at large." Now here's
1:09:14 Mimi with the weather. Cut the fun size
1:09:16 AI wearing a loud
1:09:21 70s oh go-go dress. Go dress
1:09:27 with big 70s hair.
1:09:32 and she talks about it raining
1:09:39 cats and dogs.
1:09:42 Cut to
1:09:50 at Kyle Shannon
1:09:52 saying
1:09:57 cheesily cheesily,
1:10:00 "Hey,
1:10:02 who are you calling a dog
1:10:07 [Laughter]
1:10:12 They both laugh.
1:10:16 All right.
1:10:19 I think cheesily is a word.
1:10:24 Okay. Now, let's go back here and we're
1:10:28 going to go sora.
1:10:33 And then we're going to go bang. There's
1:10:35 our prompt. Bang. All right, we're
1:10:38 going.
1:10:41 Oh, big hair and big eyeliner. I missed
1:10:42 the big eyeliner. I think it'll get it.
1:10:44 I think it'll I think it's going to make
1:10:46 you like a a good solid 70s weather girl
1:10:50 trolop
1:10:52 with the with the loud 70s dress.
1:10:56 All right, we'll come back to that. Um,
1:10:59 anybody have any questions, thoughts?
1:11:02 What's the best app to watch this stream
1:11:04 on a phone? Yeah, probably probably Tik
1:11:07 Tok just because the you can chat with
1:11:10 the other people in there easily. Um,
1:11:12 but you can watch the stream on on the
1:11:14 YouTube app.
1:11:25 I think I did, but it crashes when I The
1:11:27 app crashes when I open it. Yeah, it's
1:11:30 we're still in, don't forget people, as
1:11:34 exciting as all of this tech is, we're
1:11:36 still in the janky phase of AI.
1:11:40 I agree. I flopped between and settled
1:11:42 on Tik Tok. Yeah. Yeah. Yeah. Tik Tok.
1:11:45 Just for vertical stuff like this, it's
1:11:47 better. And I try to keep it a decent
1:11:50 experience with the black bar and stuff
1:11:52 like that, though I fail miserably much
1:11:54 of the time.
1:11:56 Um,
1:12:02 who else do we have here? Oh, Watsy.
1:12:07 Nice. Ricky Bacon. Is that who that is?
1:12:10 No. Berwick.
1:12:15 Hi, Justine.
1:12:30 Um, one thing I've noticed is that the
1:12:33 Sora videos that I do, um, that I do
1:12:37 post onto Tik Tok are getting shitty
1:12:41 views, which makes sense. They're
1:12:43 probably looking for the watermark or
1:12:45 something like that or the name of the
1:12:46 video.
1:12:49 If I really wanted to take this
1:12:50 seriously, what I'd probably do is
1:12:52 generate stuff in Sora. I heard
1:12:55 somewhere that you can download them
1:12:56 without the watermark right out of the
1:12:58 Sora app, but I I haven't experienced
1:13:00 that. That's that's not what I I see.
1:13:03 But if you take the videos out of Sora
1:13:06 and and move them into something like
1:13:07 Cap Cut and sort of recaption them
1:13:09 there, then you might you might get
1:13:12 better views. You might not get
1:13:13 throttled as much unless they're looking
1:13:16 for the watermark. Um,
1:13:21 all right. Let's go see if this thing is
1:13:23 ready. Are you ready? No, it's not
1:13:25 ready. Let's We can play other videos
1:13:28 here.
1:13:30 No, we can't.
1:13:33 Oh my god, this app. It's so bad. Oh,
1:13:37 I'm not even sharing it, am I? Am I? I
1:13:40 am. Wait. Okay,
1:13:45 close that down.
1:13:57 All right. Fantastic.
1:14:05 [Music]
1:14:13 Nah, I'm just going to watch a video and
1:14:15 be done.
1:14:26 All right. So, tomorrow, actually, let
1:14:28 me go full screen here while we wait for
1:14:30 that thing to render.
1:14:33 So, tomorrow we've got office hours at
1:14:35 11:00 a.m. If you're an AI salon
1:14:38 mastermind, we've got the mastermind
1:14:41 meeting right after that at noon. That
1:14:43 meeting is ultimately going to move, but
1:14:45 tomorrow it's still at noon mountain
1:14:47 time.
1:14:49 Uh, and then tomorrow night we got
1:14:51 Friday night date night. Although I do
1:14:53 have to go do some gallery duties, so
1:14:55 it's probably going to be 8:30 or 9
1:14:57 tomorrow night mountain time. And you're
1:14:59 like, I'm on the East Coast. That's
1:15:00 awfully late to start a Tik Tok. I know,
1:15:05 but it's Friday night date night. So
1:15:07 what are you doing? What? You got a
1:15:09 date? No, you come hang out with me. I'm
1:15:12 your date.
1:15:15 This is where your life is led.
1:15:21 Uh, you said tonight. I meant tomorrow
1:15:24 night. Yes.
1:15:27 What Brandon said. Um, just so you know,
1:15:32 in case it wasn't clear, I started a
1:15:36 fast today.
1:15:40 So, I haven't really eaten that much, so
1:15:42 I'm a little lightheaded.
1:15:49 Okay, cool. Um,
1:15:52 let's go.
1:15:55 I would think that our video is ready by
1:15:57 now.
1:16:04 Nope.
1:16:09 Mm-m.
1:16:25 [Laughter]
1:16:30 Can't hear. Oh, yeah. Because I'm not
1:16:32 sharing it.
1:16:36 Did I tell you I was starting it fast?
1:16:41 Oh, I'm sharing the wrong kind of wrong
1:16:44 kind of share. Hang on.
1:16:52 >> Chip fried and cola. I'm Kyle Shannon
1:16:54 and this this is fizzy crisps. Real
1:16:56 potatoes dropped into real cola, then
1:16:58 crisp to a sweet smoky pop. Listen to
1:17:00 that fizz. The bubbles caramel taste of
1:17:02 chip fried and cola. I'm Kyle Shannon
1:17:04 and this this is fizzy crisp.
1:17:06 >> This is fizzy crisps.
1:17:09 You need to eat some caramel.
1:17:11 >> I do.
1:17:12 >> Existence.
1:17:15 >> That's impossible.
1:17:19 >> What in the world? They prompted you
1:17:22 into existence.
1:17:25 >> That's impossible.
1:17:28 >> What in the world? They prompted you
1:17:32 into ex
1:17:34 Oh, we're done.
1:17:37 Hang on. We're gonna reload. I think
1:17:41 we're going to reload. We're not going
1:17:43 to reload. We're gonna crash, aren't we?
1:17:46 Wait, we reloaded. Here we go. Killer is
1:17:51 still at large. Now, here's Mimi with
1:17:53 the weather. Thanks, Kyle. We're looking
1:17:55 at a soggy evening, folks. It's raining
1:17:57 cats and dogs out there, and by
1:17:59 midnight, we could pile up an inch and a
1:18:00 half across the metro. Keep those
1:18:02 umbrellas handy, but tomorrow morning,
1:18:04 the sun will poke back out. killer is
1:18:05 still at large. Now, here's Mimi with
1:18:08 the weather.
1:18:09 >> I don't I don't know why it didn't use
1:18:11 my voice,
1:18:16 but nice nice 70s go-go dress.
1:18:23 Look at the hair. That's solid.
1:18:27 Uh-oh. We've crashed.
1:18:32 >> Thanks, Kyle. We're looking at a soggy
1:18:34 evening, folks. It's raining cats and
1:18:36 dogs out there. And by midnight, we
1:18:38 could pile up an inch and a half across
1:18:40 the metro. Keep those umbrellas handy.
1:18:42 But tomorrow morning, the sun will poke
1:18:43 back out.
1:18:45 >> All right. Imagine when the when these
1:18:47 things don't suck.
1:18:50 The we're looking
1:18:52 >> I would say my experience right now.
1:18:55 That's pretty funny, isn't it? My
1:18:57 experience right now. Uh oh. What did I
1:19:00 do? The hell did I do?
1:19:03 There we go. Um, my experience right now
1:19:06 is that the um
1:19:10 creating something in Sora, I'm getting
1:19:13 about a a one in three hit rate. And I'm
1:19:15 not really trying that hard, but it's
1:19:18 like I have to do at least three videos
1:19:21 to get one that's usable. And then it's
1:19:23 kind of halfass usable. So, it's
1:19:27 probably more like five or seven to one
1:19:29 to get something that's actually decent.
1:19:32 just rerun it with the same prompt and
1:19:34 it'll be completely different. The thing
1:19:36 that's driving me absolutely baddy is
1:19:38 that I'm trying to render these things
1:19:41 as um as wide videos and and it's just
1:19:45 not doing it. It's making like just
1:19:48 weird aspect ratio mistakes.
1:19:52 Uh the the thing crashed again. So,
1:19:54 we're we're good. We're done.
1:19:57 We're done. All right. I'm going to get
1:19:58 out of here. So, tomorrow, Friday, uh
1:20:01 Friday, Friday night date night, so
1:20:03 11:00 a.m. Mountain time is AI office
1:20:06 hours. So, you can find that on my
1:20:07 LinkedIn channel. And then tomorrow, uh
1:20:11 and then after that at noon mountain is
1:20:13 the AI salon mastermind hangout.
1:20:18 And then tomorrow night, probably 8:30
1:20:21 or 9 is Friday night date night.
1:20:25 Want to see one more fake VO3
1:20:27 commercial? As long as I don't have to
1:20:28 go to the Sora app. Is it on the salon?
1:20:37 >> Oh, you're going to show it. Great.
1:20:39 >> I tried everything for my depression.
1:20:42 Nothing worked.
1:20:44 >> Every day felt heavy. I felt trapped.
1:20:47 >> Then I tried Peppermint.
1:20:48 >> Our prescription helps your body secrete
1:20:51 a special pheromone that attracts
1:20:54 puppies. I took the pill before bed and
1:20:57 when I woke up, there he was, the love
1:21:00 of my life.
1:21:01 >> The pill does not target depression
1:21:03 directly, but we found that it's really
1:21:05 difficult to be depressed when cute dogs
1:21:07 show up at your doorstep. I used to feel
1:21:10 so empty, but now I feel joy and mild
1:21:13 concern how a pea stain got on the
1:21:15 ceiling. My puppy listens twice as good
1:21:18 as my ex-husband and only climbs into
1:21:21 the lap of half as many of my friends.
1:21:23 He chewed up my Bible and pooped in my
1:21:26 good chair. But I'm happy for the first
1:21:29 time in years. Looks like a rat, barks
1:21:33 like a demon.
1:21:36 But he saved my life.
1:21:37 >> I named him Earl. He follows me
1:21:40 everywhere and farts in his sleep. Just
1:21:43 like my first husband.
1:21:45 >> Puppine. For when your therapist says,
1:21:48 "Maybe you should get a dog.
1:21:52 Beautiful.
1:21:54 Um,
1:21:56 I'm trying to figure out where
1:22:03 [Music]
1:22:05 the mastermind the link to the
1:22:06 mastermind
1:22:12 hangout.
1:22:14 I don't know where that event is.
1:22:22 Leadership lounge. No projects and
1:22:24 collabs. No.
1:22:26 There. No.
1:22:31 I'll we we might have screwed something
1:22:33 up in moving things around. So, I'll
1:22:35 check I'll check with Andy tomorrow. Um
1:22:38 I'll send out what I'll do is I'll send
1:22:40 out an announcement. Do you know where
1:22:41 it is, Brandon? It's in leadership
1:22:44 lounge under uh TG AIF for tomorrow.
1:22:48 >> Ah, okay.
1:22:49 >> Founders hangout now on Google Meet.
1:22:54 [Music]
1:22:56 >> You're also not sharing your
1:22:59 Oh, yeah. Got it. Okay. Here. Look.
1:23:10 All right. So, if you're if you're in
1:23:12 the mastermind, go into leadership
1:23:16 lounge and under TG Aif, the founder
1:23:19 hangout is here. Because it's a private
1:23:21 meeting, it doesn't show up in the
1:23:22 general things. That was a great ad. I
1:23:25 need a pup. Yeah, right. Okay, cool. Um,
1:23:29 thanks for the help. I can't find it
1:23:30 either. Great. Yeah, leadership lounge.
1:23:32 And then, yeah, please join us tomorrow
1:23:34 if you're part of the mastermind. Come
1:23:35 hang out. It's a nice It's a nice
1:23:37 tighter, you know, we can we can talk a
1:23:39 little deeper because it's a it's a
1:23:40 tighter little group. Um, cool. All
1:23:43 right, I'm out of here. Have yourself a
1:23:45 swell evening and I will see you
1:23:48 tomorrow. Bye.