4/8/2025 - Exploring the AI Frontier: Tools, Trends and Keeping Up in a Rapidly Evolving Landscape

Description

Chapters

Transcript

0:01 What are we going to sing?
0:03 H What are we going to sing? Which
0:06 glasses I want. I guess these are fine.
0:11 [Music]
0:26 [Music]
0:31 Round myself of my wrinkled
0:33 [Music]
0:36 sheets. Smoking. It was all a
0:41 dream in my head and a hole in my
0:47 heart. Says it was more than just your
0:50 be.
0:53 [Music]
1:05 [Applause]
1:11 There's been something baby I'm trying
1:14 to
1:16 say. For aging, it seems I don't know
1:19 how.
1:22 Past and a future now surrounding
1:27 me. Surrender to an empty
1:32 f. There's been a little
1:35 trouble since you came to my
1:38 [Music]
1:41 rescue. And if you like all of the rest,
1:44 I would have quit you long ago. But I
1:47 couldn't do that.
1:49 [Music]
1:52 Oh, tell me now. My mind never went too
1:56 well.
1:59 Make a man crazy, make him cold as
2:03 hell. I'm a woman that you wish me
2:07 well, but I'm trying. Still going to
2:11 have to find my own way through.
2:15 [Music]
2:21 [Applause]
2:24 Good evening good people. What is
2:26 shaking? What's going
2:29 [Music]
2:37 down in a westerly?
2:40 [Music]
2:48 Car is my
2:51 train and driving. I've been
2:55 wondering what it is I'm from.
3:01 Again, feel like an 80y old
3:04 [Music]
3:05 man holding on to
3:10 29. And up ahead on that
3:13 [Music]
3:14 horizon is a California
3:18 [Music]
3:25 line. Up had a truck scaring a wide
3:28 [Music]
3:33 log. Cute little front door and two
3:36 windows. My
3:37 love is sure whether I should
3:42 laugh. You see, I broke a home of myself
3:46 once. This I stumbled to that
3:51 dog. I read a note by the
3:55 dog said, "Don't you come around
4:00 here anymore.
4:06 Well, I've had enough.
4:10 There's freedom
4:12 around. Never was good with
4:16 decisionions. At least that's what I've
4:19 been
4:21 [Music]
4:28 told. Good singing, Champy. Real nice
4:31 tonight. Solid, good tone.
4:37 [Music]
4:43 [Applause]
4:44 [Music]
4:49 Well, I heard there was a secret call
4:53 and played in a pleasing love. You don't
4:56 really care for music, do you?
5:00 goes like this or
5:07 [Applause]
5:09 minalujah
5:12 hallelujah
5:14 hallelujah
5:17 hallelujah
5:21 hallelujah
5:24 [Music]
5:32 Yeah. Strong and you needed proof. Saw
5:35 her bathing on a roof. Her beauty and
5:38 the moonlight through
5:41 you. She tied you to the kitchen chair.
5:44 She smashed your throat and cut your
5:46 head from the lip. She drew. Hallelujah.
5:54 Hallelujah.
5:56 Hallelujah.
5:59 Hallelujah.
6:01 [Music]
6:07 Hallelujah. What do you think, champ?
6:11 [Music]
6:22 Uh, I am still sharing. All right. Good
6:25 day. What's up, producer Brandon? Good
6:28 to see you, sir.
6:31 [Music]
6:54 I have 10% left on my watch
6:59 battery. Ah, what's happening? Good
7:05 people. I'm up to 46 published titles as
7:09 of tonight. Good. Keep it going.
7:16 Nice. I am more convinced than
7:21 ever that the era of the singular
7:25 author is waning.
7:27 [Music]
7:38 I think AI is going to transform just
7:40 [ __ ] everything. Like
7:44 everything. I lost my subscription star.
7:47 What the hell? What the hell's going on?
7:49 I have no
7:58 idea. There's a new uh there's a new
8:01 video tool I want to play with tonight.
8:03 Well, it's not a new tool. It's an old
8:06 video making tool, but it looked like
8:08 they had something that looked
8:10 interesting. So, we'll see if it
8:12 [Music]
8:28 is. There's something going on that's
8:31 been going on for a while.
8:34 The subscriptions expire. Where's that?
8:37 On YouTube or here on
8:43 TikTok? I guess on
8:47 [Music]
8:53 YouTube. Oh, it's an US problem. Oh,
8:57 that's right. I've got to No. Hello,
9:00 Kyle. Am I late? No, I've got to verify.
9:02 I've got a I've got a what do they say
9:04 in the crypto world? I've got to dox
9:09 myself. Um YouTube wants me to verify my
9:12 identity. So I just I I didn't get to
9:15 that
9:17 today. I thought you YouTube had dropped
9:19 it. No, no, no. We got a I got an email
9:22 that was like you can't monetize your
9:25 channel
9:26 anymore. And then I sent a note to
9:28 producer Brandon. I'm like what's up
9:29 with this? And then I read the next
9:31 email from them which said, "You've got
9:32 to
9:35 um prove that you're
9:38 [Music]
9:46 real. I mean, we have a [ __ ] business
9:49 [Music]
9:51 account. That's why they don't have the
9:53 badge anymore."
9:55 Okay. Oh, here. That's why they don't
9:57 have the badge anymore. Oh, if it's on
9:59 Tik Tok. Tik Tok's still fine with
10:02 subscriptions, but if you don't have
10:04 your badge, it's cuz your subscription
10:07 lapsed. Kyle, I'm doing my persona with
10:10 chat
10:11 GPT and make. What do you think? I think
10:14 awesome. Beautiful. You've entertained
10:17 me nearly every night for a couple of
10:19 years now, Kyle. At least I could do a
10:20 sub. Thank you very much, Mr. Architect.
10:23 I appreciate that a lot.
10:29 Oh, thing I was going to talk about or a
10:32 thing I was going to talk about
10:36 is we've kind of talked about this a
10:39 little bit, but it's
10:42 like I
10:46 am, how do I put this?
10:51 If
10:53 anyone asks you or if anyone asks
10:58 me what tool should I
11:01 use, we're at the point where it
11:05 literally doesn't [ __ ] matter. It
11:07 doesn't [ __ ]
11:08 matter. Use co-pilot, use Claude, use
11:12 Perplexity, use Chat
11:15 GPT, use
11:18 Gemini. Like, which one do you like?
11:20 Which one fits your
11:21 personality? Which one is one that you
11:24 took the [ __ ] you know, four hours
11:28 to learn how to sort of prompt it and
11:30 you sort of get decent results
11:34 from, you know, which music generation
11:36 tool? Doesn't matter. They're all fine.
11:39 Which image generation tool? All all of
11:42 them. Midjourney's better. Okay. ID's
11:46 okay. Chat GBT is better. Okay.
11:49 I would actually argue argue chatbt
11:52 image
11:53 generation is better than anything out
11:56 there right now with the exception of
11:58 midjourney for
12:01 aesthetics. If you want it to look arty,
12:04 use
12:04 midjourney. I plugged the AI learning
12:07 lab today on an interview. Oh, awesome,
12:09 Gareth. Thank you very much. I
12:10 appreciate that a lot.
12:11 Um, yeah. You
12:15 know, I've got to find a way to spread
12:18 the word about this because since I've
12:21 since I've split the stream to YouTube
12:24 and Tik Tok. Um, YouTube doesn't promote
12:29 like Tik Tok used to and Tik Tok just
12:32 isn't promoting the show anymore. Um, so
12:35 I got to find a way to get out there.
12:37 So, anything you can do like that to
12:39 promote this would be great. Um, and
12:41 then I gota I gotta put my put my noggin
12:44 together with producer Brandon. We got
12:46 to figure some [ __ ] out. Uh, because you
12:50 know it's better for all of us if
12:52 there's more people here. And I would
12:53 love to get I would love to find a way
12:55 to get new people in
12:57 here. Not that I don't love y'all. Y'all
12:59 are
13:00 regulars. Y'all are regulars are what
13:02 make this place special. And I would
13:05 love to see some new faces in here. They
13:07 asked me how I stay on top of everything
13:09 in AI. with how fast it's changing. So,
13:12 this is kind of my point, Gareth, about
13:14 um I mean, yes, what you said, get your
13:18 ass in community. Whether you come to
13:19 the AI learning lab every night, whether
13:21 you go to the AI salon, whether you join
13:23 she leads AI, doesn't matter. Get your
13:25 ass in community and participate with it
13:28 because holy [ __ ] is stuff moving
13:32 fast. I mean, it's it's funny. Well,
13:34 it's not funny for her. my co-founder in
13:38 StoryVine. I'm the handwavy guy, right?
13:41 I'm the inventor. She keeps the trains
13:43 running on time. So, she's the one
13:45 paying attention to privacy policies and
13:48 data policies and things like that. And
13:51 I keep changing tools because the target
13:54 keeps moving for what's, you know, what
13:57 what's going to produce a better product
13:59 for the for our clients, which is
14:01 ultimately what we need to do. and and
14:03 it's
14:04 like from two weeks ago I'm using two
14:08 completely different tools than I was
14:09 two weeks
14:10 ago. And she's like, "Wait, I've have I
14:13 heard of those tools before?" And I'm
14:15 like, "No, they didn't exist two weeks
14:17 ago." One of them is Cartisia text to
14:20 voice. Cartisia is
14:22 amazing. I think it's better than 11
14:25 Labs. Um, and then Hedra or Hedra. H E D
14:31 A for voice to animation is better than
14:36 anything. But a new one came out tonight
14:38 called Vigle, which has got a new voice
14:42 to um animation model. So, we're going
14:45 to go look at
14:46 [Music]
14:49 that. Kyle isn't happy with us anymore.
14:52 He wants a new version. No, you all are
14:55 fine. You all like Word 5.1 before
14:59 before it turned into bloatware, right?
15:02 You're a good solid word
15:04 [Laughter]
15:08 processor, trading us irregulars for
15:11 regulars. No, I want new irregulars. I
15:14 want I want people You know what I want?
15:16 And I'm I'm giving a talk tomorrow night
15:21 um at a just at a local like a whiskey
15:25 club kind of thing. Um I'm doing an AI
15:28 talk tomorrow night. And I'm I'm really
15:30 hoping that there's a couple of people
15:32 in the audience that are like I was
15:34 thinking I should maybe try chat GPT
15:38 sometime. So we're going to be late
15:40 tomorrow. We are going to be late
15:41 tomorrow. My my talk is only from 5 to
15:44 6:30, but then I've got a dinner.
15:46 Someone's in town from one of our
15:47 clients, so I've got a dinner tomorrow
15:49 night. So, it might, depending on how
15:51 late the dinner goes, it might even be a
15:53 uh scrap tomorrow night. I'll just come
15:56 on for half an hour and do a little
15:57 something. Viggle. Vi Gle. V is in
16:01 Victor. I have three or four different
16:03 communities I stay involved in just so
16:05 nothing slips through the cracks. That's
16:07 smart. I plug this channel to everyone I
16:10 know. Beautiful. They're probably
16:11 annoyed how much I talk about it. Yes.
16:14 Okay. This is the thing. If I if I teach
16:17 you nothing else in this channel, be
16:19 shameless about how much you talk about
16:21 AI and I don't if it's this channel,
16:23 awesome. But like just [ __ ] put
16:26 yourself out
16:28 there. Do I hear Professor Brandon take
16:32 or producer Brandon takeover? Uh it Hey,
16:35 it very well could be producer Brandon.
16:38 You've got the keys to the car. So, so
16:41 you can probably uh teach them all how
16:43 to make some Microsoft stuff tomorrow
16:45 since I keep resisting
16:49 it. Juporro imaging make it more
16:52 irregular. Exactly. Oh, we already bored
16:55 you. No, not at all. No, you know what
16:58 it is? I'm really pissed off at [ __ ]
17:00 Tik Tok that they just don't produce
17:03 they they just don't promote this
17:04 channel anymore. The lives. It's just
17:06 it's it's really frustrating.
17:11 I just saw
17:13 Hawk, some old dude that I just like
17:15 watching, they just canled his channel.
17:19 Yeah, because I'm not selling Madagascar
17:20 vanilla beans. Yeah. Yeah. If I were if
17:23 I were twerking and uh and selling
17:26 cardboard cutters, ice makers, and
17:28 Madagascar vanilla beans, we'd have all
17:31 sorts of people. Or if I had if I had
17:34 gone down the battling Dar the Star Tik
17:37 Tok
17:38 [Laughter]
17:44 battles, I would be a Tik Tok
17:47 bajillionaire. But but no, no, I just
17:51 wanted to go with integrity.
17:52 [Music]
17:58 [Music]
18:04 Yeah. So, I'm I'm kind of thinking
18:08 of coming up with a new way to to
18:13 recommend what people should
18:18 learn because it's
18:21 like like it used to be it used to be
18:26 pick a large language model, pick an
18:28 image generation
18:31 tool, you know, pick a video tool, pick
18:33 a right, it it was
18:35 one and now there's like five or six in
18:39 every category with the exception of
18:41 music which has three three big
18:49 ones. And then and then within each
18:51 tool, there's tools within the
18:54 tools. This one's better at animation.
18:56 This one's better at natural movement.
18:58 This one's better at, you know,
19:01 consistency between frames. Tik Tok pin.
19:04 Techie Becky, my sister is isn't talking
19:06 to me because she's upset I keep talking
19:10 about
19:13 AI. Keep doing it. Keep doing it. She's
19:16 afraid of it. I mean, so so quite
19:19 seriously, Becky, here's here's what I
19:21 would
19:21 say. Have empathy for
19:24 her, for that fear.
19:28 Um, I mean, to the extent that you can,
19:31 try to understand it and and understand
19:33 where she's coming from and see if you
19:36 can find an on-ramp.
19:39 Um, because here's the reality for her.
19:43 It's not going away. Like like it's the
19:46 only thing the only thing I'm 100%
19:49 convinced
19:50 of 100%
19:53 positive like a
19:56 thousand% AI isn't going away. It's not
19:59 going to slow down. It's not going to go
20:02 away. Did you hear that? They they
20:05 [ __ ] unextincted the direwolf.
20:10 [Music]
20:17 When was the last one? 7,000 years ago.
20:19 70,000 years ago. Please consider
20:22 sharing the live. Yeah, do that.
20:26 They found a tooth of a dire wolf
20:28 somewhere in a riverbed and then they
20:29 found a skull somewhere else and between
20:31 those two they had enough
20:33 DNA and they basically took a wolf which
20:37 is about 99% I
20:40 think of the
20:43 DNA and they now have two
20:46 seven-month-old unex unextinct
20:51 unextincted
20:52 deextincted direwolves
20:56 I think they said a woolly mammoth is
20:59 [Music]
21:12 next. So the [ __ ] ain't going
21:17 away. You know the other thing you could
21:20 do, Becky? I mean, listen, she's she's
21:22 got to be open to it. But I mean I like
21:24 I don't think force feeding anyone AI
21:25 serves anyone. So I think it's I think
21:27 understand where she's coming from is
21:29 part of
21:32 it. And if there's if there's a
21:35 particular area that she
21:37 is resistant to. So like maybe she's an
21:42 artist or an illustrator and she's like
21:44 totally freaked out by the images, but
21:47 maybe she wouldn't be freaked out to
21:49 write emails
21:51 faster. Like one of the one of the
21:53 things you can do is you can segment the
21:56 work that's personal to you where you
21:58 feel existentially threatened from the
22:02 [ __ ] where AI can make your life a
22:04 little
22:05 easier. And so if you can help her
22:07 identify like is is there an area where
22:10 you've got a hard boundary? I remember
22:13 um Jane Endicott who's one of the early
22:16 salon members
22:18 [Music]
22:19 u she was a copywriter and when I first
22:21 met her she was like should I be worried
22:24 and I'm like yes I'm like you need to
22:26 learn AI you know pronto and she and she
22:30 very very quickly did and like very
22:32 quickly kind of 10xed her output for her
22:35 client which the client
22:37 basically coinciding with the client
22:39 going I need you to do more work, but
22:42 I'm not going to pay you more. Um, she
22:45 got her hands on AI and what she said
22:47 was there's a hard line for her between
22:49 her creative writing and the the writing
22:52 she does for work. And then over two
22:55 years that's gotten more sophisticated.
22:57 She'll use AI to help her ideulate and
23:00 outline things, but then she still does
23:03 writing writing to write to on the
23:05 creative writing side.
23:07 So, every person's got to find that
23:09 personal boundary. Um, okay. Let's see.
23:14 Shopify CEO said in an internal memo
23:17 that everyone should be using AI and if
23:20 you want more headcount, you have to
23:21 prove AI can't do that job. Yeah. The
23:24 there the the head of Shopify is calling
23:26 AI a core
23:29 competence. A core competence.
23:34 [Music]
23:36 And yeah, I guess saying to managers, if
23:38 you want to increase headcount, you've
23:39 got to prove that AI can't do that
23:43 job.
23:45 That's if the writing isn't on the
23:49 wall, it is scribbled in pencil on the
23:52 toilet
23:54 paper. You better get with it. AI 2027.
23:59 As someone who's trying to find a new
24:00 company, I'm like, [ __ ] we're all going
24:03 down. the AI enthusiast in me, it's
24:05 like, well, I'm glad I'm fluent in
24:07 AI and and we all saw this coming. Like,
24:11 I I just at this point, I mean, it's
24:15 been what, two and a half [ __ ] years
24:17 since Chat GPT came
24:21 out. It's starting to get irresponsible
24:26 for people to be on the sidelines with
24:28 this [ __ ]
24:33 thinking about
24:37 it. I mean, the the agents haven't even
24:40 really started yet this year. We, you
24:42 know, we got a little bit with Manis, a
24:44 little bit with
24:45 [Music]
24:51 Operator. Oh, the the Oh, Spotlight the
24:54 Report.
24:56 AI2027.com. What's that? Is that
24:59 [Music]
25:00 the Is that the Spotify one or is that
25:02 something different?
25:17 AI-2027.com. Oh, this is the thing. Is
25:19 this is this from Daario Amade? Who's
25:22 who is this?
25:24 We predict that the impact of superhuman
25:26 AI over the next decade will be
25:28 enormous, exceeding that of the
25:29 industrial revolution. Yeah, I don't
25:32 disagree with that. We wrote a scenario
25:34 that represents our best guess on what
25:36 it might look
25:37 like. It's informed by
25:40 trend extrapolations, war games, expert
25:45 feedback, experience at OpenAI, and
25:48 previous forecasting success.
25:52 Mid 2025 stumbling agents. Yes, the
25:55 world sees the first glimpse of AI
25:56 agents. They're jankier than
26:00 [ __ ] This is
26:02 AI-2027.com. I'm trying to use Manis to
26:04 vibe code a game. You know what's funny?
26:07 I had I had Manis write me a game. I was
26:10 really impressed with how it did it. It
26:12 then wrote me a game and I had to like
26:14 install it on my Mac somehow and I
26:16 couldn't quite figure it out and I was
26:18 just like, "Oh, [ __ ] it. That's too much
26:20 like work.
26:25 Hey Kyle. Yeah, what's up? Ju just so
26:27 you know it's um just giving you some
26:30 context here. This report uh was
26:32 something that was put together by some
26:33 of the leading minds of AI AI futurists
26:38 and it's an interactive report. As you
26:40 scroll down the chart on the right hand
26:43 side starts to show what impact AI is
26:46 going to have on society. Oh, that is
26:49 it's a timeline over the next until
26:52 2027. What is going to happen to AI? Oh,
26:56 I get it.
26:59 Now, that is very clever. Thank you.
27:02 Thank you for
27:05 that. All
27:08 right. Got it. Okay. I see. Okay. So,
27:12 over here we've
27:13 got coding ability. coding ability is
27:17 just starting.
27:18 Basically,
27:20 hacking,
27:22 boweapons, basically everything's just
27:25 starting. And we and we have a little
27:27 bit of compute and it's mostly in
27:29 America and a big chunk of it open
27:34 AI. 2,000 unreliable agents thinking at
27:38 8x human
27:40 speed, minus 25% approval.
27:44 [Music]
27:46 Oh, for open
27:51 AI. All right. Mid 2025. Stumbling
27:55 agents. Late
27:57 2025. The world's most expensive AI.
28:00 Open brain is building the biggest data
28:03 centers the world has ever
28:06 seen. Wow.
28:13 To avoid singling out singling out
28:15 anyone company, we're going to describe
28:17 a fictional artificial intelligence
28:19 company we'll call Open Brain. Oh, okay.
28:21 Open
28:26 AI. So, basically, they're going to they
28:29 threw they threw a dot worth of
28:35 compute. They they did a dot worth of
28:37 compute for GPT3.
28:40 um seven bricks and I guess a brick has
28:44 got nine dots in it, something like
28:48 that. And then a whole lot more bricks
28:50 for agent one, which is going to be the
28:53 I don't know if that's going to be GPT5,
28:55 but that's going to be their
29:00 agent. Those are my ink boxes. Yeah,
29:02 exactly.
29:05 There are lots of AI ident data centers
29:08 across C's that aren't being used
29:11 because of R1 right now. Oh, that's
29:14 interesting. Mind reading and genome
29:16 encoding. Yeah, the mind readading stuff
29:18 is going to be
29:19 crazy. All right, let's keep going down
29:21 this this prediction chart or this
29:24 prediction thing.
29:27 early 2026 coding automation and that's
29:31 you know everyone's saying that Meta
29:34 Open AI anthropic are all saying we're
29:36 going to have fully automated encoding
29:39 this year so I guess it gets up to
29:42 speed early
29:51 2026 mid 2026 China wakes
29:55 up
29:58 Wait, isn't China Isn't China already
30:01 awake and kind of kicking our
30:04 ass with with models that are just as
30:07 good and and cheaper? 10 times
30:21 cheaper. Late 2026, AI takes some jobs.
30:28 This This feels like a weirdly
30:31 conservative maybe I I'm just reading
30:33 their headlines. So
30:38 So now we're starting to to get into
30:42 monthly monthby-month
30:45 predictions. Agent January 2027, agent 2
30:49 never finishes learning.
30:53 February, China steals agent 2.
30:58 Nice. So, wait, this thing is we just
31:01 passed this. We're just
31:03 now Yeah, we're we're in the slow the
31:06 the flat part of this curve right
31:09 now.
31:11 Huh. Algorithmic
31:14 breakthroughs in March, April 2027.
31:17 Alignment for agent 3. Yeah, this we
31:20 start getting stuff every couple of
31:21 months that are
31:23 The next excuse me the next things May
31:26 2027 national
31:30 security June 2027 self-improving
31:34 AI open brain now has a country of
31:37 geniuses in a data center that's from
31:38 Dario
31:47 Amade research automation deployment
31:50 takeoff
31:54 off a thousand tokens a
32:07 second July 2027 the cheap remote
32:13 worker so cost
32:17 plummets LLM inference prices will have
32:19 fallen 9x X to 900X a
32:24 year.
32:25 Basically, intelligence is
32:31 free. Geopolitics, the superhuman AI
32:35 researcher, agent 4 is in September of
32:37 2027. So, basically, basically, we've
32:41 got two
32:42 [Laughter]
32:44 years. Again, you know, I I I go back to
32:47 I mean, this is a this is an interesting
32:49 report. I'm kind of reading it on the
32:50 fly here, so I I don't know enough about
32:52 it to to really comment on it.
32:54 Um, but go to
32:57 ai-2027.com and it's that's a really
33:00 slick uh interface to dynamically change
33:03 that chart as you read the article. It's
33:04 pretty
33:05 smart.
33:09 Um, yeah, it's it it is irresponsible at
33:13 this point for anyone that's not getting
33:15 on this train. It just it just is. It
33:18 has a choose your own adventure ending.
33:20 Oh, cool.
33:23 Nice. Let me
33:28 see. Choose your ending. Slow down or
33:31 race. Well, I'm going to [ __ ] race.
33:33 We're
33:34 here. The committee votes six to four to
33:37 continue internal use of agent 4. Oh,
33:39 and then it keeps going.
33:43 OpenAI's approval has dropped
33:46 to minus
33:48 45%. Their revenue is 250 billion a year
33:52 on a $9 trillion
33:56 valuation. The Agent Five Collective.
33:59 Yeah. Now, now agents are in swarms.
34:03 2028 the the AI economy. 2029 the deal.
34:11 2030 takeover. By 2030, the robot
34:14 economy has filled up the old SEZs. I
34:18 don't know what those
34:24 are. Wow.
34:27 Crazy. All right. Well, that's super
34:29 cool. I dig
34:34 it. You know what's
34:37 amazing? It's like
34:41 We're at this place
34:46 where we're paying attention to this. We
34:49 know enough
34:50 about what AI is making possible to see
34:54 the path to what this paper is talking
34:56 about. Maybe it happens by 2027 like
34:58 they're saying here. Maybe it happens by
35:00 2030. Doesn't matter. it at some point
35:02 in the next two to five
35:08 years. Everything will have
35:13 changed in the next two to five
35:16 years. Everything will have
35:20 changed. And right now, we're still
35:22 having conversations about, well, was it
35:26 trained on someone else's data? Does it
35:29 make mistakes?
35:32 Why do their faces look
35:37 weird? I think it's hype.
35:45 I mean, I've just I've never I've never
35:48 been I've been on the planet for a lot
35:51 of years, relatively 59
35:58 years, and I've never
36:01 felt like in the early days of the
36:03 worldwide web, I knew that I was ahead
36:06 of the
36:07 curve, but when I look back at people
36:10 that were were when I was talking too
36:13 much about the worldwide web in in 1994
36:15 and 1995 and all my theatrical friends
36:18 were telling me to [ __ ] off. We don't
36:20 want to hear about the web anymore,
36:22 including my current writing partner who
36:24 told me to [ __ ] off because he's like,
36:26 I'm a designer. Why would I need to
36:28 learn HTML, you big fat dumb dumb?
36:34 Um but but when I looked back to them,
36:38 I'm like, "Oh, okay. They're just
36:39 they're they're not that far behind
36:42 because the the acceleration of the
36:44 worldwide web was was really
36:47 slow, right? It was just
36:51 like like we were getting a little more
36:53 knowledgeable, a little more
36:54 knowledgeable. You can still kind of
36:55 back look back and go, "Well, you could
36:58 now you could buy HTML for dummies and
37:00 you could actually learn this stuff and
37:01 catch
37:04 up." Just look at I don't know you were
37:07 here, many of you were here the other
37:09 night when I did that sort of joke rant
37:12 on all of the different
37:15 models within chat
37:17 GPT and they're adding two more now.
37:20 Like it's it's it's [ __ ]
37:23 insanity, right? That's just one tool.
37:26 One
37:27 tool has what? 10 models in
37:32 it plus Sora plus
37:41 operator. I don't know how you catch up
37:43 with this. Like we can't keep up with it
37:46 and we're [ __ ] trying to keep up with
37:48 it.
37:51 I just feel like the the the
37:55 [ __ ] velocity is doing this and it's
37:58 like, you know, with the web thing, you
38:01 know, was kind of on this slow
38:02 trajectory and I could look back and
38:03 still see the people. I just feel like
38:05 at this point the this [ __ ] rocket is
38:08 taking
38:10 off. I've never experienced anything
38:13 like this. Like I just feel like the gap
38:16 between people that have some sense of
38:19 what's happening and people that are
38:22 just
38:24 utterly unprepared for what's coming is
38:29 massive. We could pause the current
38:31 frontier and still spend 20 years not
38:33 fully leveraging it. I totally agree
38:34 with that.
38:36 Honest to God, I feel like I could just
38:38 take GPT40 right
38:41 now, just as it is right now, and spend
38:44 the next decade trying to figure out
38:46 what to do with
38:47 it easily.
38:55 Easily
38:59 crazy. Be like a rock skimming over the
39:02 water. Yep.
39:11 [Laughter]
39:24 Touch enough to get you by. If you stay
39:27 long enough or you could fall beh If you
39:29 stay long enough, you could fall behind.
39:32 Um, where should I jump back in? Give me
39:35 some ideas. Um, I think a place to jump
39:38 back in, Becky, is um is is the new
39:42 image gen within within chat
39:44 GPT. Just just keep playing in chat GPT.
39:50 [Music]
39:59 Or the other thing you could do, Becky,
40:02 is pick an area. This is something I've
40:05 been thinking about. What's an area in
40:08 your life where you have convinced
40:10 yourself that you don't have talent at
40:12 something or you're not good at
40:13 something and go find an AI that's
40:16 really good at that and start playing
40:17 with
40:19 it because I think that I think that
40:22 we'll increasingly
40:25 [Music]
40:32 um we'll increasingly start to discover
40:37 that we can be good at anything we want
40:39 to be good at, but you have to be
40:41 willing to like go explore outside of
40:43 your comfort zone. And then and then I
40:45 think if you have one or two of those
40:48 where you're like, I actually am really
40:49 good at this and I'm really passionate
40:51 about it and it was a thing you told
40:52 yourself you could never do. That feels
40:54 really empowering to me. The new image
40:57 gen is ready but on hold. Reliable
41:01 source. Yeah, that doesn't surprise me,
41:03 Brian. They there and there there've
41:05 been a couple of hints on Twitter where
41:06 they're like, you know, you're going to
41:08 like what you see and see was
41:10 capitalized. And um Sam Alman said the
41:13 the next version of Image Gen is really
41:16 really good. He thinks we're going to
41:17 like it.
41:20 Um maybe the next version of Image Gen
41:23 can do clocks that have the actual time
41:25 on them, not 10 after 10.
41:36 [Music]
41:43 Um, a thing I'm getting kind of excited
41:45 about is, and I mean I could see OpenAI
41:49 being being the one to lead the way on
41:51 this, but there might be
41:53 others, is is true
41:56 multimodality where within the single
41:58 model, you can do
42:02 code,
42:04 images, voices, sound effects,
42:08 music,
42:09 animation, you know,
42:11 video, 3D world building, 3D object
42:15 building, and you
42:18 know, world world
42:23 development. Like, we can't be that far
42:26 from that, right? maybe a year, maybe
42:30 two, because I think the first person
42:33 that cracks it, like it's [ __ ]
42:34 exhausting right now to have to go
42:39 to chat GBT for the the script and the
42:45 logic. Maybe chat GB for chat GBT for
42:48 the image likely midjourney. Then you
42:52 have to go to Lumalabs or Runway or some
42:56 other [ __ ] video tool for the video.
42:58 Then you have to go to 11 Labs for sound
43:00 effects. Then you have to go
43:02 to Cartisia for
43:06 voice, right? Then you have to go edit
43:09 it together in something. Like come
43:12 on. I don't I don't want to have to work
43:16 anymore. I just want to have an idea and
43:18 have it vomit out.
43:24 You're describing NAN right now. No, I'm
43:27 not. Because NAN you still have to build
43:30 your automations. I mean, yeah, I I'm
43:34 describing the output of
43:36 NAND. And if you're a technologist, you
43:38 can look at NAND and go, "Yeah, yeah,
43:40 that's obvious." No, what I'm talking
43:42 about is you talk into the chat
43:45 hole and out comes an image and then
43:49 you're like, "Turn that into a movie.
43:50 Now give it sound effects. Now write a
43:51 song about it." No, I don't want it to
43:54 have lyrics. I just want it to be
43:55 background cinema. Cinemagraphic music.
43:59 Cinematic. Cinemagraphic. I'm just
44:02 making up [ __ ] words
44:08 here. Oh man.
44:11 Christian Robbins. What's happening,
44:13 dude? I didn't see that. It was
44:17 you. I assume you're the same Christian
44:19 Robbins from from $11
44:23 bill. If you have not looked up $11 bill
44:26 and Christian Robbins, he is the [ __ ]
44:28 man. He's been doing motion
44:32 graphics in
44:34 um he's had a motion graphics company
44:37 for what, 12 years, something like that.
44:39 What's up, man? Runway's getting closer
44:42 these days with that. I know. So, is
44:44 Luma Lumalabs right now is cool because
44:46 they can do images, video, now they're
44:48 doing sound effects. It's like that's
44:51 what I'm saying. There's and and the
44:52 other thing Lumalabs has is the really
44:54 interesting interface where you can sort
44:57 of go horizontally and deep in a in a
45:00 visual story, but it's like I've got to
45:02 go write my story over in chat GPT, and
45:04 then I've got to go over here and make
45:06 my music.
45:07 And it's just it's [ __ ] madness.
45:11 Chris, I want to talk to you about
45:12 something, Christian. Um, do me a favor
45:14 if if you're still on. I think you're
45:16 still here. Um, send me an email because
45:19 I want to talk to you about a project.
45:20 I've got an idea for
45:22 something. Um, and I think you'd be
45:24 really good at at at least to bounce the
45:27 idea off of. Maybe we could do something
45:28 together. New Tik Tok from Source Camp
45:33 [Music]
45:34 where I've built 100 plus GPTs and still
45:38 learning how to build them. I know it's
45:41 right. It's it's like the depth the
45:44 depth of what we have in our hands. It's
45:46 it's so hard to get your head
45:49 around. It was funny in the early days
45:51 of the web when
45:55 when trying to think who did it first.
46:00 There was a there was a search engine
46:02 before Alta Vista but after Yahoo. So
46:05 Yak Yahoo came out and what Yahoo did is
46:08 like we're going to categorize the
46:10 internet. So they created a directory
46:12 and they were manually you know
46:14 organizing the internet and then the
46:17 first search engine came out. I forget
46:19 what it
46:20 was and it was pretty cool. And then
46:22 Alta Vista came out and Alta Vista was
46:26 like, "Holy [ __ ] like they seem to like
46:28 be they seem to know everything on the
46:31 internet, which was really quite
46:32 mind-blowing." And then Google came out
46:34 and Google, you know, took over and
46:35 wiped out everything because they they
46:38 they they really got that indexing thing
46:41 down cold, right?
46:44 And it was it
46:47 was if you had ever thought about data
46:51 structures and how much data
46:54 it to crawl the internet every night or
46:57 every it was like every 24 hours they'd
47:00 crawl it or every two hours. I don't
47:01 know. It was some ridiculous amount of
47:04 data that they were ingesting every some
47:06 amount of hours to keep the search
47:08 engine active. And it was just
47:10 mind-boggling. Right. And I feel like
47:13 we're in the same the same place with
47:16 um with large language models right
47:19 there. It's it's almost impossible to
47:21 comprehend how much knowledge is
47:25 compressed into this this box the box.
47:31 You know, you've got this [ __ ] box
47:33 and they just took like all of the
47:35 output of humanity and they're like,
47:37 we're going to do this thing called
47:38 embedding and we're going to create
47:39 thousand dimensional latent space and
47:42 we're just going to jam all those
47:44 numbers in
47:46 here and then you can ask it [ __ ] and it
47:49 knows everything we've ever done. Like I
47:53 I can't really comprehend what that
47:57 means. But here's the thing I do know.
48:00 Every time I talk to chat GPT in my
48:02 little voice
48:03 mode, I like I have yet to stump
48:07 her. She answers [ __ ] every
48:11 question. Where's the fud box on my 2019
48:14 Ford Flex? Oh, the fuse box on your 2019
48:17 Ford Flex is under the hood on the left
48:19 hand side just above the wheel. How the
48:21 [ __ ] does it know
48:22 that? Because it does. Because that's
48:25 what they did.
48:28 I like ideas and projects. People sleep
48:31 on Luma these days. They have some nice
48:32 camera tools, too. Yeah, they they added
48:35 that new the I forget what they call it,
48:37 like the rack zoom or the crunch zoom or
48:39 the poppy zoom. I don't know what the
48:41 [ __ ] it's called, but it
48:44 goes. It's pretty
48:47 cool. Oh, man. All right. I'm bad at AI,
48:52 but AI is very good at AI. Yeah,
48:54 exactly. That's true. That's that's the
48:56 other thing, Pate,
48:59 is right now and and in fact, this
49:03 channel is kind of dedicated to us as
49:08 these severely limited cognitive beings
49:13 trying to get our heads around how the
49:15 [ __ ] do we use these
49:17 tools as the tools are going like this.
49:23 And the answer is we probably won't have
49:25 to for long. Like we'll just be able to
49:27 ask for what we want and get it. Bolt
49:30 cam on Luma. Yeah, exactly. That's cool.
49:33 My advanced
49:35 voice knew I was mad about something and
49:39 lowered his voice about five octaves.
49:41 That's pretty
49:46 funny. Oh man.
49:51 Have you tried asking harder
49:54 questions? Yeah. No, I haven't, Pete.
49:56 And I mean, that's the that's But that's
50:00 a really good point
50:02 that the things I do in the
50:07 world. For the most part, I'm not asking
50:10 hard questions. Now, I do a lot of
50:12 creative stuff, which creative stuff is
50:15 interesting because you can ask really
50:18 simple questions and and it's
50:20 essentially an impossible answer because
50:22 it's not a quantitative answer. There's
50:24 not a right or wrong, right? Creative
50:27 work is all
50:28 qualitative. Um, but I do a lot of that
50:31 sort of stuff and large language models
50:34 are [ __ ] insanely good at that. Now,
50:36 are there areas where large language
50:39 models still suck at creative? Yeah,
50:41 like writing song lyrics in a
50:43 sophisticated way, you've got to [ __ ]
50:45 work to get them to do it well. Um, and
50:49 I would I would assume that there are
50:50 similar things if you're doing, you
50:52 know, high-end math, high-end research,
50:54 high-end science, that these things
50:57 probably fall apart and you're like,
50:58 well, [ __ ] they're just not ready yet.
51:01 But again, I don't like I don't think
51:02 it's going to be that long. Dang, ran
51:03 out of credits before I could finish the
51:05 game. Oh, that sucks.
51:07 Um um let's let's jump over to I want to
51:11 jump over to Vigle for a second and just
51:15 see if this thing works at
51:19 all. All
51:22 right. So, if you're still here,
51:24 Christian, I they just announced this
51:28 today. It's another one that promises
51:32 um good character movement with with
51:35 voice.
51:45 Okay.
51:47 So, so I don't know what this thing is.
51:49 I think this is a background. So, I I
51:52 put in an image
51:56 here. Let's do someone at at night in
52:00 the streets of Paris. Bigle.ai. Okay,
52:03 cool. And then I'm going to drop an
52:05 image here.
52:08 Assets.
52:10 Upload. Upload a
52:13 character image of any size. For better
52:15 results, use a full body photo with a
52:17 clean background. All right. So, let's
52:19 go to midjourney.
52:26 Oh, this was this was a very horrible
52:29 experiment trying to
52:31 get trying to get
52:35 um midjourney to do a Chuck Close piece
52:39 of art. It failed miserably.
52:55 [Music]
53:06 Oh, he might be
53:08 fun. Let's do him.
53:26 There's a bunch of theatrical images I
53:40 did.
53:42 Edit. Why did that not
53:50 edit?
53:54 Wait. Editor. Ah, there we go.
54:09 Okay. All right, let's do creepy dude.
54:21 [Music]
54:29 side book. Side note, Notebook LM is
54:32 getting an app. That is not surprising
54:34 to me at all. It's pretty
54:36 cool. All right, so we have a crazy ass
54:39 character here.
54:45 Which one do we
54:51 like? I think that one.
54:55 Maybe. Yeah, I think that one. All
54:57 right. That's [ __ ] weird. All right.
54:59 I like it. So, I'll do a creative a
55:02 subtle upscale and a creative
55:04 upscale. All right. So I may start using
55:08 Gemini more because I have direct access
55:10 to Docs, Drive, etc. Yeah, Gemini is I
55:14 would say at this point Gemini is right
55:16 up there with Chat GPT, Claude,
55:19 um and uh like Perplexity, just any of
55:24 the ones you want to use.
55:27 Um it's just that with all of them, you
55:31 got to figure out their personality.
55:33 Like some of
55:35 them some of them technically execute
55:39 things that you ask them to execute, but
55:41 you're like that's just not good or I
55:42 don't like the
55:43 formatting. Like I feel like that with
55:45 Claude. Like I just don't like how how
55:47 it formats
55:48 [ __ ] And like the logical part of my
55:52 brain is like, "Well, that shouldn't
55:54 [ __ ] matter. It's just data. You're
55:56 just making words. What the [ __ ] does it
55:57 matter?" It
55:59 matters. It matters. So anyway, so just
56:03 you know, but yeah, there's it does all
56:06 sorts of stuff. It can generate images.
56:07 It can edit images. One of the
56:09 advantages of Gemini's native image
56:12 editing thing where you just say like
56:15 you create this character and you say,
56:16 "Give him a different hat." Gemini will
56:19 just give him a different hat. Chat
56:21 GPT's new image gen will redraw the
56:23 whole image. So it can be really [ __ ]
56:26 frustrating. But like those little
56:29 subtleties you only learn by jumping in
56:31 there and doing it. He looks a little
56:34 plastic there. He looks a little [ __ ]
56:36 up
56:37 there.
56:40 Uh let's see. And then what
56:43 else? How are
56:46 these? I think these look more
56:48 realistic. So we're going to go with
56:49 we're going to go with this one. All
56:51 right. So I'm going to download that
56:53 image. Then we're going to go back to
56:55 Vigle.
56:58 We're going to upload an
57:02 image. Should we edit that? I think we
57:04 should. Let's lighten it up a bit so we
57:06 can see a bit
57:10 [Music]
57:15 more. Tools. Adjust the
57:18 color. Let's go.
57:21 Exposure. There we go. Fantastic. Bob,
57:24 tell him what he's won. Oh, he hasn't
57:27 won anything. This isn't a game
57:29 show. Why do I work here? What is my
57:37 [Music]
57:43 job? Look at all the granular control
57:46 I've got right there in the preview
57:48 tool.
57:50 Save. Okay, so we can at least see that
57:52 a little better. That was an interesting
57:55 thing. I I read on uh on on the Twitter
57:59 not that long ago. Someone was bitching
58:02 about I'm gonna I'm gonna get rid of
58:05 this. How do
58:06 I Do I get rid of
58:10 this? Upload image or
58:13 video? I don't get it.
58:25 Um, okay. It's still the right one. Why
58:28 was that not showing
58:29 up
58:31 here? There it
58:38 is. Oh, and then we've got to upload
58:40 audio. So, we got to go make audio. So,
58:42 let's go to
58:45 um we'll go to 11
58:52 labs. All right, hang on a
58:57 sec.
59:00 Um, there's my new Embrace the Jank
59:03 cover. Yay.
59:06 I know I'm not talking much uh and I'm
59:09 not telling you what I'm doing, but
59:13 sorry. I I will at some point. I've got
59:17 I've got three. So, here's what I'm
59:18 trying to do. I'm trying to get an
59:20 image, get a voice, and marry those two
59:24 together. I could also do a
59:27 song. And maybe we should do a song.
59:29 Let's do a song. Okay, hang on. Hold,
59:32 please. Let me stop presenting there.
59:35 I'm going to present. I'm going to
59:36 present. Oh, wait. I can't present
59:40 yet. This is chat
59:44 add. They don't call it chat add for
59:48 nothing. Well, actually, well, I guess
59:50 this is very add- like, but
59:54 um I know what I want to do. And it's
59:57 just in my head it's a simple thing. And
1:00:00 then as I'm doing it, I'm realizing, oh,
1:00:02 there's 16 [ __ ] tools here we need to
1:00:05 use. Christian Robbins knows this game
1:00:08 well. He's been he's been doing this for
1:00:11 years now. He understands the nightmare
1:00:13 that that bringing all these tools
1:00:15 together is.
1:00:16 Um, okay. So, we're going to create a
1:00:23 song. Oh, we're going to go to chat
1:00:27 GPT. Hang on a second. I want to look up
1:00:30 Griswald.
1:00:34 Griswald. Griswald the Griner.
1:00:36 Backstory. Okay. I'm going to say,
1:00:38 "Write
1:00:40 me some
1:00:43 twisted lyrics
1:00:46 about
1:00:49 Grizzwald scaring kids with
1:00:53 stories." Not sharing. Oh, I know.
1:00:56 That's all right. Hang on. I know. I
1:00:59 know. I just just calm down,
1:01:05 people. No, I didn't. I didn't share.
1:01:09 Come close. Come near. Don't look
1:01:11 behind. Griswald knows what's on your
1:01:12 mind. He's seen the things beneath your
1:01:14 bed. Okay. I'm just going to say you
1:01:16 want a full length version. Um, I'm
1:01:19 going to tell Chat
1:01:20 GPT.
1:01:22 Um, this feels very surface and
1:01:28 cliche and like you didn't think about
1:01:32 it. I want to win a Grammy. You
1:01:41 ninny tick tock pin. Wouldn't surprise
1:01:45 me if you turned this song into a
1:01:47 musical. Shut up. I've already done that
1:01:49 once. 16 tools is what makes it fun.
1:01:52 Embrace the jank. Yeah, this is totally
1:01:54 embrace the jank time. They gather like
1:01:56 moth at the edge of his boots with
1:01:58 licorice breath and balloon animal
1:02:00 roots. This is better. His voice it's
1:02:02 match strike inside of a jar. Okay. Oh,
1:02:06 twisted
1:02:07 lullabi slow theatrical with the weight
1:02:09 of a funeral carousel. Um, love this
1:02:15 angle
1:02:17 and song
1:02:21 description, but I want it from his
1:02:26 point of
1:02:33 view. Yes. Now we're rolling into
1:02:35 something real. A haunted show tune with
1:02:38 soul. Let's step fully into Griswald's
1:02:42 shoes from G Grisswald's point of view.
1:02:44 Dark cabaret, minor key, 68 swing with
1:02:47 toy piano, bode saw, and an accordion
1:02:50 that wheezes like it remembers things it
1:02:52 shouldn't. I should never shout boo.
1:02:55 That's for am amateur. I never shout
1:02:58 boo. That's for amateur sport. I prefer
1:03:01 tales with longer retort. A whisper, a
1:03:03 pause, a welltimed aside, and poof. The
1:03:06 brave runa brave run brave ones go
1:03:11 running to hide. They say on the cloud,
1:03:14 but okay, this is great. Okay, so we're
1:03:15 going to say we're going to go grab this
1:03:18 and an accordion that wheezes. Okay,
1:03:20 that's that's probably just about
1:03:22 enough. So, we're going to go to suno
1:03:25 and we're going to put this in song
1:03:28 style style
1:03:31 tags. more
1:03:36 options. All right, there we
1:03:40 go. Okay, and then we're going to go
1:03:43 grab the
1:03:51 lyrics. Copy. We're going to come in
1:03:53 here. We're going to plop the lyrics in.
1:03:55 We're going to call this
1:03:59 song Where's the title?
1:04:05 Exclude
1:04:09 options. I can't. Oh, up
1:04:13 here.
1:04:15 Grisswald's
1:04:20 world. Griswald's
1:04:23 kids. No. Griswald's.
1:04:29 Griswald's Griswald's
1:04:32 tail tail. I don't know. I don't [ __ ]
1:04:34 know, people.
1:04:37 Okay, so I'm going to create these. Let
1:04:40 me go change my sharing so you can hear
1:04:42 this.
1:04:47 [Music]
1:05:00 Right.
1:05:05 [Music]
1:05:09 All right. We got a dark cabaret with
1:05:11 toy piano.
1:05:15 Um, let's see.
1:05:19 Um, I'm going to
1:05:21 say, what's another word for grally?
1:05:25 Scratchy.
1:05:32 Scratchy male
1:05:36 with
1:05:42 smokers rasp.
1:05:45 Dark Cabaret. All right, let's create
1:05:48 those. Oh, Griswald's Grim. Yes. Yes,
1:05:53 correct. That is the correct title. By
1:05:55 the way, if you're new
1:05:57 here, I do these creative exercises and
1:06:00 I fancy myself a creative. I'm like, I
1:06:02 can I'm good with words. Some people
1:06:05 have a way with words, other
1:06:08 people not have
1:06:11 way. Old Steve Martin joke. Thank you.
1:06:15 Um, shout out if you if if if there's an
1:06:18 obvious good answer to the thing where
1:06:21 my brain is just [ __ ] shut off. Grizz
1:06:24 Griswald's Grim is absolutely what this
1:06:26 song's about. Okay. So, did we make any
1:06:30 of these yet?
1:06:32 No. Griswald's Grim. Boom. All right.
1:06:35 We'll go listen to a couple of these
1:06:37 other ones. It seemed to kind of get
1:06:38 what we were looking for
1:06:41 there. Griswald's black bar. There you
1:06:44 go.
1:06:49 That's for amateur sport. Yeah, these
1:06:52 all sound too nice. Like we need this
1:06:54 guy to sound [ __ ]
1:07:05 terrifying. I never shout boo. That's
1:07:07 for No.
1:07:10 [Music]
1:07:14 No. All right, let's do Griswald's
1:07:18 [Music]
1:07:19 [Applause]
1:07:20 [Music]
1:07:24 Grim. I never shout boo. That's for
1:07:28 amateur sport. I prefer tales with a
1:07:32 longer retort. A whisper, a pause, a
1:07:36 welltimed aside, and poof, the brave
1:07:39 ones go running to hide. They say I'm a
1:07:42 clown, but
1:07:44 I just
1:07:48 [Music]
1:07:50 That's pretty close.
1:07:52 [Music]
1:08:02 Nope. Let's see. Uh, let's do older
1:08:07 male. Older male with smoker's rasp.
1:08:10 Okay, let's see. This This one here is
1:08:13 probably usable.
1:08:19 It still sounds a little clean. Like he
1:08:21 sounds like he's about to drop [ __ ]
1:08:23 dead is what we want.
1:08:31 I never shout.
1:08:34 No, I never shout.
1:08:39 [Music]
1:08:41 Boo for amateur sport. I prefer.
1:08:46 These all sound like young anenus
1:08:49 like you know like some SNL character is
1:08:53 going
1:08:55 to whisper.
1:08:58 This is not a Disney film.
1:09:00 [Music]
1:09:06 I never shout. That's for This is honest
1:09:10 to God. This is a
1:09:11 Disney sport.
1:09:15 All right, [ __ ] it. We're going to go
1:09:16 back to this one. What's that, Vicki? A
1:09:18 grally, menacing vocal tone. Okay, we'll
1:09:20 try that.
1:09:22 A
1:09:25 grally,
1:09:28 menacing
1:09:31 male
1:09:34 vocal with a smoker's
1:09:39 rasp. They were not they never thought
1:09:41 to train it to sound like crap. but also
1:09:44 like minor key
1:09:46 um
1:09:48 68
1:09:52 swing and an accordion that wheezes.
1:09:56 Um, I'm going to say,
1:09:59 um, the beat is
1:10:09 disjointed,
1:10:11 twisted, and darken.
1:10:21 Gank gank
1:10:23 gank.
1:10:25 Okay, let's create a few more of
1:10:30 those. We'll do three
1:10:33 more. All right, Vicki, you're on the
1:10:35 hook here. If this isn't
1:10:38 good, Perry Kik Jr. try grainy
1:10:43 rasp. Well, you know what I mean. Okay,
1:10:46 let's see. So, one, two, three, four,
1:10:49 five, six. Here we go.
1:10:56 I
1:10:57 never amateur sport. Okay, that's not
1:11:00 bad.
1:11:01 [Applause]
1:11:02 [Music]
1:11:10 Jesus Lord, these are
1:11:12 bad. Well, by the way, if you're new
1:11:14 here, you might be amazed that this
1:11:16 thing's making songs like this. It
1:11:19 is. Here's one of the things
1:11:24 that people who are not actively playing
1:11:27 with AI miss this. What we're doing
1:11:31 right now is creative development,
1:11:34 right? I'm serving as a creative
1:11:35 director right now. I have in my mind,
1:11:38 we've got we've got a
1:11:40 picture of this crazy ass [ __ ] dude,
1:11:44 which I can't switch screens and show
1:11:45 you. We have this crazy ass
1:11:48 carney, right? And I've got some idea of
1:11:52 what I want, but I don't quite know what
1:11:53 it is yet. And and a big part of working
1:11:56 with AI, even though it makes things
1:11:58 fast, it rarely makes
1:12:01 things that match what you want. And so
1:12:04 the art of AI is just like swinging back
1:12:08 and forth like, you know, tacking, you
1:12:11 know, in a sailboat back and forth and
1:12:13 back and forth. You kind of know where
1:12:15 you're going, but n it's too much that
1:12:16 way. That's too much that way. So
1:12:18 anyway, um this is kind of how it goes.
1:12:28 That's for amateur. I never shout
1:12:33 boo. You know what I'm going to
1:12:36 do? Final
1:12:38 chorus. I never shout quotes capital B O
1:12:43 exclamation
1:12:51 point and then I'm going to put that a
1:12:53 capital T. All right. And I'll create
1:12:56 two more and just see if we can get that
1:12:58 to hit the word
1:12:59 [Music]
1:13:06 better. for amateur sport.
1:13:11 [Music]
1:13:13 No, I never shout. Boo. That's for
1:13:18 [Music]
1:13:26 amateur. I never shout boo. That's for
1:13:30 amateur sport.
1:13:33 Not bad.
1:13:35 [Music]
1:13:38 That's for amateur
1:13:47 sport. I I never shout. That's for
1:13:50 amateur sport. I prefer tails with a
1:13:56 [Music]
1:14:00 longer. I never shout boo. That's for
1:14:04 amateur you
1:14:06 sport. I prefer tales with a longer
1:14:12 close. We got to change the music. A
1:14:14 grally menacing. Oh, I spelled that
1:14:17 wrong. Menacing male vocal. The the
1:14:21 rasque. I didn't put in the whole thing.
1:14:22 Tik Tok pin. I wish I could audition for
1:14:25 this
1:14:27 one. Let's see. Like a clown dragging
1:14:31 its laugh through broken
1:14:34 glass. Let's see.
1:14:37 Vocal like a
1:14:44 clown dragging
1:14:47 it
1:14:50 laugh. Oh, I'm out of I'm out of
1:14:52 characters.
1:14:58 Grally menacing male
1:15:02 vocal
1:15:05 clown dragging its
1:15:07 laugh
1:15:10 through broken
1:15:14 glass. Okay. Dark cabaret minor key toy
1:15:17 piano bode saw accordion that wheezes
1:15:20 disjointed twisted and dark.
1:15:24 Create. Create. Okay, these last This
1:15:27 should be it. This should be it,
1:15:32 people. You can just type in Tom Waits
1:15:34 and see what happens. Yeah, that's not
1:15:36 bad. Let's see. Um, a grally menacing
1:15:40 vocal. Let me just say
1:15:45 um like Tom Waits through broken glass.
1:15:50 We'll see if it it might reject that
1:15:52 one.
1:15:57 Did it? Yeah, I think it did. It
1:16:00 rejected it. Couldn't generate that. You
1:16:04 can't put in an artist's name.
1:16:10 I never shout boo. That's for amateur
1:16:18 sport. I never shout boo. That's for
1:16:21 amateur sport.
1:16:29 I never shout boo. That's for amateur
1:16:32 sport.
1:16:37 [Music]
1:16:40 I think we had it way back when.
1:16:44 [Music]
1:16:49 It was one of these early ones. Hang on.
1:16:52 Is it this one?
1:16:55 A
1:16:59 [Music]
1:17:02 whisper. That's part of the
1:17:07 game. The brave goveout.
1:17:10 [Music]
1:17:19 All right. This is the one we're going
1:17:20 to do. This is the one we're going to
1:17:22 do. It ain't perfect,
1:17:24 people. Just perfect enough. That's one
1:17:27 of the songs from Sydney, my new
1:17:29 Broadway musical. Not yet on
1:17:32 [Laughter]
1:17:35 Broadway. Okay. Um, let me change my
1:17:38 screens again so y'all can see. I want
1:17:40 y'all to see.
1:17:47 Gank.
1:17:49 Yeah. Oh, yeah. Tomorrow at 400 PM
1:17:52 Mountain time is the uh AI readiness
1:17:55 project podcast with myself and Anne
1:17:57 Murphy. Uh you can find it at the AI
1:18:00 salon. You can find it on my LinkedIn.
1:18:02 You can find it at
1:18:05 sheleadsai.ai.
1:18:07 Um we downloaded that. We're going to
1:18:10 vigle. Now we're going to upload
1:18:13 audio. Griswald's Grim. Here we go.
1:18:21 [Applause]
1:18:21 [Music]
1:18:28 Wait, who had archetypal note? Never
1:18:31 thought to train it to sound like crap.
1:18:35 Good advice for Sunno. Where's the good
1:18:37 advice for? Oh, down here.
1:18:40 Click the like on it. Make it easier to
1:18:42 find later. Oh, yeah. I know. That's
1:18:44 That's one of those add oversightes that
1:18:47 every time I I'm like, "Oh, click like
1:18:50 and then you can find it later." And
1:18:51 then I just forget to click like. It's
1:18:52 But it's a good
1:18:54 call. Audio must be under 60 seconds.
1:18:57 Fine. We can We can fix
1:19:06 that. Oh, no. That wasn't what I wanted
1:19:08 to do. God damn
1:19:16 it. Open with QuickTime
1:19:21 player and let's see.
1:19:34 [Music]
1:19:39 [Music]
1:19:47 kid. Okay.
1:19:52 [Music]
1:20:00 Okay. Clip clip that
1:20:04 [ __ ] Wait. could not be split
1:20:06 because some clips could not be loaded.
1:20:09 What do you mean it can't be loaded?
1:20:10 We're [ __ ] playing
1:20:14 it. Oh, come on. God damn
1:20:20 it. I just every [ __ ] time I want to
1:20:24 do something on this channel and it's
1:20:26 just like just [ __ ] doesn't [ __ ]
1:20:29 work.
1:20:33 My stories don't end, kid. They make you
1:20:37 pretend.
1:20:40 Fine. Split clip. Please
1:20:44 work. Yay.
1:20:49 Okay. Bitter. You betcha. Export as
1:20:53 audio
1:20:54 only.
1:20:57 Um,
1:20:58 Grizzwald's
1:21:00 Grim. Uh uh uh
1:21:05 short. Oh, and um you can check with
1:21:10 Christian Robbins on this. I've talked
1:21:12 to him about this. My best practice for
1:21:15 file management. Save everything to your
1:21:19 desktop. Just loose with a completely
1:21:22 nondescript name so you can't find it
1:21:24 later. That's the really the best way if
1:21:26 you're going to do especially if you're
1:21:27 doing big complicated projects and you
1:21:29 got to make a lot of different assets
1:21:31 from a lot of different places. Just
1:21:32 save them all the desktop because
1:21:34 they'll all be there. It's really easy
1:21:36 for you to
1:21:39 find. If you think I'm being serious,
1:21:42 then this is not the channel for
1:21:49 you. Confirm.
1:21:58 I don't want this customized thing here.
1:22:01 How do I remove
1:22:06 this from
1:22:09 image? Okay, there we
1:22:12 go. All right,
1:22:15 generate. Generation created
1:22:17 successfully.
1:22:22 Just blame the blame the producer when
1:22:23 [ __ ] breaks. Well, if if it weren't if
1:22:26 you worked for Apple, I'd [ __ ] blame
1:22:29 you. It's like, come on, people. Let's
1:22:32 get your [ __ ] together with these
1:22:34 softwares. All
1:22:36 right, I'm going to have this first line
1:22:39 stuck in my head for the next week.
1:22:40 You're welcome. All right, let's go.
1:22:42 Let's go. Let's go listen to our little
1:22:44 song
1:22:45 here while that thing's rendering.
1:22:48 [Applause]
1:22:50 [Music]
1:22:52 [Applause]
1:22:55 I never shout boo. That's for amateur
1:22:58 sport. I prefer tales with a longer
1:23:02 retort. A whisper, a pause, a welltimed
1:23:06 aside, and poof, the brave ones go
1:23:09 running to hide. They say I'm a clown,
1:23:12 but I ain't in disguise. is I just dress
1:23:16 the way I remember your cries. One
1:23:19 button's for secrets, the other for
1:23:22 shame and the smile. Well, darling,
1:23:24 that's part of the game. I don't chase.
1:23:33 Wait, where else? Which which other one
1:23:35 is good at? Hedra is Let's go do Let's
1:23:38 go do the same thing in Hedra.
1:23:43 Um. Oh, did it do it already? Come on.
1:24:07 Guessing you know I'm missing a kind.
1:24:10 Hang on. Sorry about that. Tab, tab,
1:24:13 tab. Oh, and I got to change my audio,
1:24:16 right? God damn
1:24:18 [Laughter]
1:24:20 it. It's so [ __ ] exhausting. It's not
1:24:24 bad. It's a little janky. I want to try
1:24:25 this in Hedra, too, because I think
1:24:27 Hedra is better. But we'll see. This is
1:24:30 not bad.
1:24:31 [Applause]
1:24:38 I never shout boo. That's for amateur
1:24:41 sport. I prefer tails with a longer
1:24:45 retort. A whisper, a pause, a welltimed
1:24:49 aside, and poof, the brave ones go
1:24:52 running to hide. They say I'm a clown,
1:24:56 but I ain't in disguise. I just dress
1:24:59 the way I remember your cries. One
1:25:02 button's for secrets, the other for
1:25:05 shame and the smile. Well, darling,
1:25:08 that's part of the game. I don't chase,
1:25:12 no need to begin. I curl up beside you
1:25:16 and crawl right. All right, we're going
1:25:18 to do some AB testing here. We're going
1:25:19 to do this in
1:25:21 Hedra. And Christian Robbins, do you
1:25:23 know any if you're still here, is are
1:25:25 there That's creepy as [ __ ] Yeah, it's
1:25:27 totally creepy. Let's do it in Hedra.
1:25:28 Are there anything other than Hedra and
1:25:31 uh Vigle now that do this uh voice
1:25:35 stuff? Good. Where's Hedra? I guess I
1:25:39 can just do a new tab. Why not do a new
1:25:42 tab?
1:25:45 Because you're not sharing the new tab.
1:25:47 Yeah, I know. I'm gonna go do that
1:25:50 now. I really like it. Okay. Present.
1:25:54 We're gonna do this. We're gonna do
1:25:56 this. We're gonna do this. Okay. Yeah,
1:25:58 Kyle, just looking at that one that we
1:26:00 just watched, it it reminds me a lot of
1:26:02 Will Smith eating spaghetti a year ago.
1:26:05 Yeah. Yeah. You you said it's 2025 good.
1:26:07 Yeah, it totally it totally is that. The
1:26:10 thing that I've been impressed with
1:26:12 Hedra is it's got the the least amount
1:26:14 of jankiness of anything so far. It's
1:26:17 still a little janky. You know what
1:26:20 these things all do, which is weird, is
1:26:22 when there are pauses, the characters
1:26:24 like look up and to the left. They're
1:26:27 like they they just look up that way and
1:26:30 it's just freaky and
1:26:32 weird. Um Oops, I just lost my
1:26:35 light. Can't lose my
1:26:40 [Music]
1:26:49 light. That's kind of a catchy song.
1:26:51 It's It's a lot. Uh, it's way too
1:26:54 Disney, I will
1:26:57 say. Okay, we're going to upload an
1:26:59 image, which is going to be found here
1:27:02 in
1:27:03 downloads. We're going to upload our
1:27:05 audio script, which is going to be found
1:27:07 on the desktop, which is where we save
1:27:09 everything because we have good file
1:27:11 management skills. Oh, this is nice. God
1:27:14 damn it. Okay, so now we get to go to
1:27:19 um free
1:27:22 M4A to MP3
1:27:27 converter
1:27:39 web convertio. I want free convert. Con
1:27:44 convertio. Which one of these doesn't
1:27:47 run me down a stupid
1:27:50 [ __ ] sales
1:27:53 funnel? Oh, wait. I wonder if I can do a
1:27:56 threem minute song in Hedra. That would
1:27:58 be a good thing, wouldn't
1:28:01 it? Let's try to add the three minute
1:28:04 song.
1:28:12 All right, it says
1:28:14 yes.
1:28:19 Um, clown
1:28:25 looks.
1:28:31 Um,
1:28:33 pervertedly happy as he sings.
1:28:44 Shrank dank. You could also try
1:28:47 downloading the stems in Suno so it
1:28:49 doesn't react to the music. Right
1:28:52 click. Oh yeah, that's not a bad idea.
1:28:54 Oh [ __ ] I just dropped my deck of cards
1:28:57 all over the floor. it. You
1:29:02 know, when you use a deck of cards as
1:29:04 your add
1:29:06 pacifier, it's just not good. But let's
1:29:08 go to
1:29:10 Sunno. We'll go to this song and we will
1:29:13 grab
1:29:16 stems. That's in create, right?
1:29:19 Create. Yeah, get
1:29:21 stems. If you haven't done this before,
1:29:23 you can get stems. So what what Sununo
1:29:26 will do is it will split your vocals
1:29:30 from your music. So we can just do
1:29:33 the the thing to
1:29:37 uh just to the vocals. But but Hedra
1:29:40 seems to be pretty good at that
1:29:42 actually. They've sort of figured it
1:29:43 out. $20 million studio, no black bar.
1:29:47 Okay.
1:29:48 [Music]
1:29:53 I wish the song weren't quite so Disney,
1:29:56 but you
1:29:57 know, it's like it's like Disney
1:30:00 menacing and I want it to be like really
1:30:02 [ __ ]
1:30:03 menacing. 52 card
1:30:06 pickup chat GPD converts MP4 to MP3. Oh,
1:30:10 it does. That's
1:30:16 cool. Real estate in Columbus, Ohio.
1:30:18 Isn't it funny that Elon Musk
1:30:22 Oh, he does the same pose. He looks up
1:30:23 and to the left. Maybe he's a All right,
1:30:26 so let's let me change my tabs here.
1:30:29 Maybe refusion. Yeah, actually, you know
1:30:32 what? Refusion's not a bad
1:30:35 idea. Let me go let me grab the these.
1:30:41 Copy. Do I have Refusion up here? I
1:30:44 thought I did. Nope. New tab.
1:30:52 refusion. Uh
1:30:57 shank. Am I logged
1:31:05 in?
1:31:08 Lyrics. We don't want an
1:31:12 instrumental. Okay, there's that.
1:31:20 prompt. Let's go back to
1:31:24 Sunno. The nice thing about this is you
1:31:27 can put in longer prompts in
1:31:38 Refusion. All right.
1:31:46 Get
1:31:54 shink. Oh, AI
1:32:01 Bob. Bob, tell him what he's
1:32:04 [Laughter]
1:32:06 won. Um, the uh, refusion I think is a
1:32:10 little bit better than Suno. Certainly a
1:32:12 prompt prompt adherence
1:32:23 coherence. All right, let me go change
1:32:25 my tab so you can hear.
1:32:30 [Music]
1:32:39 Jank
1:32:41 [Music]
1:32:50 dank. Sit here close.
1:32:55 [Music]
1:33:11 These are too
1:33:15 peppy.
1:33:17 Prompt. Let's
1:33:20 say a
1:33:24 slow,
1:33:35 menacing, dark cabaret, minor keys with
1:33:38 both saw and
1:33:41 a
1:33:43 slow,
1:33:46 menacing, disjointed
1:33:49 twisted and dark
1:33:53 cabaret. A
1:33:59 slow and put male
1:34:04 voice. Whoops. Let's see.
1:34:09 Um like
1:34:14 a
1:34:16 50year
1:34:21 smoker who never sleeps.
1:34:28 There is a
1:34:31 strength in his
1:34:34 voice despite its frailty. Look at that
1:34:39 fantastic [ __ ]
1:34:41 writing. Could have written a [ __ ]
1:34:43 screenplay. All right.
1:34:47 [Music]
1:34:58 Let's go check back on Hedra. What's it
1:35:00 doing?
1:35:02 Processing. Clown looks pervertedly
1:35:04 happy as he sings. All right, we'll see.
1:35:07 We'll see how ugly that is. All
1:35:09 right. Deathbed vocals. That's not
1:35:13 bad. Let's see.
1:35:16 male voice like a 50-year-old smoker who
1:35:19 never
1:35:21 sleeps.
1:35:25 Deathbed
1:35:28 vocals. Generate two
1:35:37 more. Hi AI fam. Just been conversing
1:35:40 with my BFF chat GPT. I know pretty
1:35:43 amazing. Played through a phonograph is
1:35:45 a good idea, too. All right, hang
1:35:47 on.
1:35:49 Um, played through a phonog graph.
1:35:55 That's
1:35:56 good. Like I said, if I'm doing it
1:36:00 wrong, don't just secretly make fun of
1:36:03 me. Make fun of me in
1:36:05 public. And even better, have a good
1:36:08 idea like Christian just did. And I know
1:36:11 you said that before. Part of the joy
1:36:13 here is I'll ignore most of the good
1:36:21 ideas. Okay, here we go. Let's
1:36:30 see. Okay, this is starting out
1:36:32 promising.
1:36:35 [Music]
1:36:39 It's a woman.
1:36:42 [Music]
1:36:43 [Applause]
1:36:44 [Music]
1:36:50 Sawustle.
1:36:52 God damn it. It's so
1:36:55 [Music]
1:37:05 frustrating.
1:37:08 [Music]
1:37:09 Uh they why can Can they not [ __ ]
1:37:13 train these things to say like a pull
1:37:15 down menu? Man, woman, I know we're not
1:37:18 allowed to gender anymore, but honest to
1:37:20 God, like
1:37:23 the male voice like a 50-year-old smoker
1:37:27 who never sleeps. Deathbed vocals.
1:37:29 There's a strength in his voice despite
1:37:31 his frail its frailty played through a
1:37:33 photograph. Why are we getting an
1:37:37 women? God damn
1:37:40 it. Maybe cabaret's throwing it off. Oh,
1:37:43 that's a good point. What's another word
1:37:45 for cabaret? Um
1:37:48 um
1:37:51 carnival. A slow, menacing, disjointed,
1:37:54 and
1:37:56 dark
1:38:01 carnival. Diddy.
1:38:05 Diddy. TTY.
1:38:07 Right. All right. We'll try that
1:38:10 one. About Jack and
1:38:12 Diane. A little diddy about Jack and
1:38:15 [Laughter]
1:38:22 D. Okay. Let's
1:38:24 [Music]
1:38:29 see. No.
1:38:41 Sit down.
1:38:45 [Music]
1:38:48 What?
1:38:52 That's okay. It just [ __ ] up. Okay. I
1:38:55 was like, did I put in the wrong lyrics?
1:39:10 [Music]
1:39:15 alone.
1:39:19 I don't know. I don't know what the
1:39:20 [ __ ] going on. It also rewrote all my
1:39:23 lyrics. I'm sure there's a place to
1:39:26 [ __ ] fix that. Add to instrumental.
1:39:29 Roll the
1:39:31 dice.
1:39:34 H. Do I need to be in this
1:39:39 mode? Let me generate
1:39:42 it. Let's go back to Hedra. Oh, you
1:39:45 can't see this. Still
1:39:48 processing. Vaudeville's not
1:39:50 bad, but I can't get it. Let's see if I
1:39:53 can get it to do the the right lyrics.
1:39:56 [Music]
1:40:00 I never shout boo. That's for amateur
1:40:03 sport. That's not bad.
1:40:09 I never shout boo. That's for amateur
1:40:13 sport. I prefer tails.
1:40:18 All right. Whatever. All
1:40:20 right. I'm growing I am growing weary of
1:40:24 this.
1:40:25 Let's go back to tabs, tabs, tabs. Yep.
1:40:28 Let's go back
1:40:31 to uh let's go watch the
1:40:35 um the vigle one again.
1:40:43 [Music]
1:40:55 Shank
1:40:59 shank
1:41:03 shanking
1:41:10 shank. Yes, I am that annoying. I know
1:41:14 I'm
1:41:15 annoying. It's just it just comes with
1:41:18 the [ __ ] deal. It's just part of the
1:41:19 deal. All
1:41:22 [Music]
1:41:23 [Applause]
1:41:25 right. It's actually interesting. What
1:41:27 these guys are doing
1:41:30 is they've masked the main
1:41:33 character and then it's just in front of
1:41:35 this
1:41:38 background. So, they basically just make
1:41:40 the background the
1:41:42 background. Not to throw you down
1:41:44 another rabbit hole, but
1:41:55 Uh, yeah, I could do that, but I I don't
1:41:58 Yeah. Yeah. I don't want to I don't want
1:42:01 to sample my voice. That's too much of a
1:42:02 rabbit hole.
1:42:06 [Applause]
1:42:08 I never shout boo. That's for amateur
1:42:12 sport. I prefer tales with a longer
1:42:15 retort. A whisper, a pause, a welltimed
1:42:19 aside, and poof. The brave ones go
1:42:23 running to hide. They say I'm a clown,
1:42:26 but I ain't in disguise. I just dress
1:42:29 the way I remember your cries. One
1:42:32 button's for secrets, the other for
1:42:35 shame, and the smile. Well, darling,
1:42:38 that's part of the game. I don't chase.
1:42:42 No need to begin. I curl up beside you
1:42:46 and crawl right in a lullaby
1:42:51 twisted. My stories don't end, kid. They
1:42:55 make you
1:42:56 [Music]
1:42:59 pretend. All right. It's not bad. It's
1:43:02 not bad, but it's I'm with Brandon. It's
1:43:04 2025 good.
1:43:11 Oh man. All right, Hedra, please be
1:43:14 there. All right, this is taking for
1:43:15 [ __ ] ever. Um, what time is it? It's
1:43:18 getting to be late. All right, so here's
1:43:19 what we're going to
1:43:20 do. Well, this is what I'm dreaming
1:43:22 about
1:43:24 tonight. Um, so Christian, good to have
1:43:28 you here. I look forward to catching up.
1:43:29 So, let's definitely talk. I want to
1:43:30 tell you about this project. I'm just
1:43:32 going to say I've
1:43:33 never had Sunno sing entirely different
1:43:36 lyrics to what I gave it. No, I know
1:43:38 that was Refusion. I've never had
1:43:40 Refusion do that to me before. I think I
1:43:42 must have had some weird cash thing
1:43:44 going on. Probably because I've got 80
1:43:46 tabs open.
1:43:49 Um and uh so so anyway, so so let me let
1:43:54 me get out of here. I look for whichever
1:43:56 I played the most and like combo and
1:43:58 Suna. Oh, that's not a bad idea. Silver
1:44:01 Fox, that's really good. Um okay, let
1:44:05 me let me entertain you. Let me
1:44:09 entertain
1:44:10 you. I never say boo. That's for amateur
1:44:16 sport. It's pretty good. It's pretty
1:44:19 good. All right, one more. One more
1:44:22 play. One more play. Come on. It's going
1:44:26 to be good.
1:44:29 [Music]
1:44:38 I never shout. That's for amateur sport.
1:44:42 I prefer tales with a longer retort. A
1:44:46 whisper, a pause, a welltimed aside, and
1:44:50 poof. The brave ones go running to hide.
1:44:54 They say I'm a clown, but I ain't in
1:44:57 disguise. I just dress the way I
1:45:00 remember your cries. One button for you
1:45:04 know who this this looks like is that um
1:45:07 Puddles the Clown. Remember when he was
1:45:10 on Idol or America's Got Talent or
1:45:13 whatever the [ __ ] show? I don't think it
1:45:15 was Idol. It was one of those one of
1:45:17 those talent shows. That's that's who
1:45:19 this looks like. It's got a claimation
1:45:22 feel. Juggle image in Irregulars. Okay,
1:45:25 let's go look at irregulars quick.
1:45:33 Look at that new AI readiness cycle.
1:45:36 Isn't that cool? We did that last night.
1:45:39 That used to be linear and we were like,
1:45:42 "Hey, it's not really linear. It's a
1:45:44 cycle." Well, no. I said that. I'm like,
1:45:46 "Here's this thing, but it's not really
1:45:48 linear." And I was like, "Wait a minute.
1:45:50 Couldn't we have one of these things fix
1:45:53 that for us?" And the answer is yes.
1:45:56 This was a midjourney
1:45:59 fix. Um, going down to AI learning
1:46:05 lab, juggling all the different tools.
1:46:08 That is totally what it [ __ ] feels
1:46:10 like right
1:46:11 now. What you have to have is one one
1:46:14 like coming at him, smacking him in the
1:46:22 forehead. Oh man.
1:46:27 Oh, that's
1:46:29 pretty. Oh, it's a sununo song. Nice.
1:46:32 All right, one more check. No, still
1:46:34 processing. All right, I'm getting the
1:46:36 [ __ ] out of here. It's getting late. I'm
1:46:39 getting crispy. Um, okay. So,
1:46:42 tomorrow I'm speaking in the afternoon,
1:46:45 then I got a dinner at night. So, it's
1:46:47 possible I won't be here tomorrow or it
1:46:50 might be like 9 or 9:30. If it is, I'll
1:46:52 probably just do it late because I'll
1:46:54 probably be exhausted.
1:46:57 Um, and that's Wednesday. Thursday,
1:46:59 we'll be back to normal 8:00 PM time.
1:47:02 All
1:47:03 right. Trying to think what else.
1:47:06 Um, at work, I'm doing a lot of uh
1:47:10 persona. Like the stuff we're doing here
1:47:11 with this thing with music, I'm doing it
1:47:14 with scripts with like doctors and
1:47:16 patients telling their stories. Um, and
1:47:19 it's going pretty well. So, I'll I'm
1:47:21 gonna one of the things I'm going to do
1:47:24 is
1:47:26 um I'm going to capture my workflow for
1:47:28 that and I'll share that with you guys.
1:47:31 But, um all right, cool. Um peace out
1:47:34 YouTube X and LinkedIn. Hope you're
1:47:37 doing well. Christian, good to see you.
1:47:39 Talk soon. And everyone else, great to
1:47:41 see you. Thanks for hanging out and I'll
1:47:44 see you
1:47:46 tomorrow or