May 6, 2024 - AI's Role in Storytelling & Music Production

Description

Chapters

Transcript

0:01 you
0:14 [Music]
0:20 ready what's happening good
0:23 people do you enjoy your weekend
0:26 off what's happening what's happening
0:30 what's happening what's happening is
0:33 everybody having a good time
0:37 [Music]
1:09 quiet
1:10 town
1:14 the wh of the train Echoes through the
1:20 day people
1:24 with nothing ever seems to
1:28 change the story is still on F each you
1:32 and
1:33 [Music]
1:38 strange we Mar from CA
1:42 [Music]
1:45 Hill how she
1:49 blue sness of Shame under the southern
1:53 Moon
1:58 [Music]
2:12 oh that is good
2:15 champy Tobias said Ohio or Oklahoma is
2:18 getting slammed with tornadoes right now
2:20 so let's send some good wishes out to
2:25 them these tornadoes this year good Lord
2:28 what a mess
2:29 [Music]
3:04 got Cindy [ __ ] in the house Milo's wife
3:06 was not all that late she she just made
3:09 it under the wire she was about to get a
3:12 demerit it's it's an automated system
3:15 it's like Tick Tock when when you say
3:18 booby you get banned for life oh
3:27 wait ah
3:30 welcome back people look I got a
3:33 haircut it's not it's not just a [ __ ]
3:38 horrific
3:40 disaster how was your weekend Kyle I was
3:44 good well Jesus when was it
3:50 Saturday Saturday I got about 5: in the
3:54 morning this is just it's just a day of
3:57 whining okay so get up at 5: in the
3:59 morning go to Colorado Springs to pick
4:01 something up for the
4:03 boy one of the boys so that was I don't
4:06 know 3 hours then what did I do then I
4:11 went I've been I've been in denial that
4:15 spring has has arrived so I haven't
4:19 mowed or weed whacked or anything for
4:21 like two months but shit's been growing
4:24 so I actually went and I got new
4:27 electric weed whackers and
4:31 yard [ __ ] basically and then I came back
4:35 and I did four hours of [ __ ] yard
4:38 work and I don't know if you know this
4:40 about
4:41 me but one of the things I hate most in
4:44 the world is yard
4:49 work but I got sick kids and a sick wife
4:55 and a [ __ ]
4:57 dog and all the dog does is [ __ ] all
5:00 over
5:00 [Music]
5:02 everything a salon group whne in we
5:05 should we should have an all whining
5:08 Guild hi hi my name is Kyle and I had to
5:12 work really hard today and I got sweaty
5:14 and I didn't like it it didn't feel like
5:17 the fantasy life I designed for
5:21 myself I think that would be
5:28 popular and then I don't know what else
5:31 did I do
5:33 Saturday oh came home and had to
5:37 cook grilled some [ __ ] on the
5:39 grill I rage cooked you ever rage
5:43 cook where you just just pissed off at
5:45 everyone you're like he here's your
5:47 [ __ ]
5:49 [Laughter]
5:51 meat I don't see the rabbit R1 on the
5:54 Shelf the rabbit R1 is in my
5:57 pocket so here's the here's what I'm
6:01 learning about the rabbit R1 it is kind
6:03 of cool it's right
6:07 there
6:08 um
6:10 but because it doesn't cuz I don't have
6:12 cell service with it and I can't I
6:15 haven't figured out how to hook it up to
6:16 my complicated work wi-fi that's got
6:19 like some
6:21 actual I don't know it's some [ __ ] I
6:24 got to deal with I've been it's never
6:27 connected to the internet
6:30 and then when it is connected to the
6:32 internet I've got my phone and I've got
6:34 computers around
6:44 me I still think the form factor is
6:48 rather
6:50 amazing and I'm I'm just kind of waiting
6:52 for the software to suck a little bit
6:55 less but I don't I don't know that I'm
6:57 going to use it
7:01 [Music]
7:07 they were going for a th000 bucks now
7:09 they're going for 600 so they're coming
7:12 [Music]
7:21 down welcome everybody if you're new
7:23 here my name is Kyle Shannon what's your
7:26 B best AI success story besides the AI
7:30 learning lab well thank you very much I
7:31 suppose that is a lovely compliment
7:35 [Music]
7:38 um I think the the one that is the most
7:43 um the one that continues to surprise me
7:47 in in that a that it still
7:52 works and B that it's just it's just it
7:56 saves us
7:58 days when we're on boarding clients at
8:01 storyvine and it's I created
8:06 a a form so we've got a we've got a
8:08 creative process when we kick off a
8:10 client we have to write story Vine's
8:13 this prompt driven video recording
8:15 platform so we have an app and it guides
8:17 someone to answer a series of questions
8:20 well you got to write those questions
8:23 and the Art of Story vine is the art of
8:25 writing those prompts and if you get
8:27 them right you get really good
8:29 consistent high quality authentic video
8:31 and if you get them wrong they
8:33 suck and especially when you've got
8:35 agencies involved who are not used to
8:37 writing for the storyvine platform
8:40 they're used to writing for traditional
8:42 video production they literally don't
8:44 understand what the [ __ ] we're talking
8:46 about when we talk about prompts and and
8:48 how to write them and how many there can
8:50 be like they're like we want it to be a
8:52 two-minute video here's the 17 questions
8:54 we're going to ask no the math doesn't
8:58 work so so so we spent in the in the 12
9:03 years story Vine's been around we spent
9:05 an inordinate amount of time just
9:08 educating people on how to do what we do
9:09 and then trying to creatively
9:12 collaborate with them after we told them
9:13 they didn't know what they were talking
9:15 about and so I wrote this little script
9:18 that we we took the manual process and
9:21 we asked him a couple of questions
9:23 what's the kind of story you want to
9:25 tell who's the target audience you know
9:27 what are you trying to you know achieve
9:30 with
9:31 them and then in the ideal video what
9:34 would the person be saying and everyone
9:35 can say that well they would say hi my
9:38 name is my name is Sally and I was
9:40 diagnosed with the blah blah blah in you
9:43 know 20
9:44 whatever and then you know it sucked and
9:47 then I got the pills and now I feel
9:49 better
9:51 huh okay and so we reverse engineer that
9:53 into these series of prompts well those
9:56 are all inputs we can put in a form so
9:58 we put them in a form
10:00 and then we
10:01 trained um chat GPT we created a a
10:05 clever prompt that included how we write
10:07 those those questions with
10:10 examples and you fill out a form with
10:14 the client on screen and it just
10:16 goes spits out the questions and posts
10:19 them to our slack Channel and then the
10:21 the account team just copies the ones
10:23 that are appropriate and just shows them
10:25 to the client says here's your questions
10:27 let's edit them like that one to me is
10:29 is just one of those things where we had
10:32 a we had a a real thorn in our
10:35 [Music]
10:38 side so that's a small one and then I
10:40 guess the big one is you know I've got a
10:42 a new for for 12 years storyvine you
10:46 would answer questions in an app and a
10:48 video would come out and that's what you
10:49 got you paid us money we'd build you a
10:52 template your people would answer
10:53 questions they' get video or you'd get
10:55 video that you could use of
10:57 them and now we built this thing called
10:59 called the authenticity engine which is
11:01 this cascading series of AI Generations
11:05 that starts with a transcript of the
11:06 video and then it makes a title a
11:08 description key messages it chapter IES
11:11 it breaks it into modules creates a blog
11:14 post
11:15 hashtags pull quotes of interesting
11:18 things that were said in the video I
11:19 mean it's just [ __ ] slickered than
11:23 [ __ ] and so and so we've got a very
11:26 different deliverable now than we did
11:28 for the last 12 years look over night
11:30 happy 12th birthday to story mine thank
11:32 you
11:33 Brandon time time to start preparing for
11:35 your bar
11:37 [Music]
11:46 mitz you still want to sing
11:50 [Music]
12:11 [Applause]
12:13 [Music]
12:20 I didn't mean to cause you in
12:25 sorrow didn't mean cause you in a pain
12:30 [Music]
12:35 I want time to see you
12:38 [Music]
12:40 laughing all I want to see a laughing in
12:44 that bu rain Purple Rain Purple
12:49 [Music]
12:52 Rain Purple
12:54 Rain Rain
12:58 [Applause]
13:00 purle rain purle rain is Google AI
13:04 better than
13:08 Microsoft
13:10 no
13:12 um but there's there's some weird
13:14 caveats
13:16 there if you go to gemini.com that [ __ ]
13:21 sucks if you go to aist studio.
13:25 goole.com and use Gemini Pro 1.5 live
13:30 it's got a million token context window
13:32 and it's way better
13:35 now other than it going to cost a lot of
13:40 money why Google hasn't released the one
13:43 at AI Studio into Gemini is [ __ ]
13:47 Beyond me it's been
13:50 months that one I think I don't know if
13:53 it's better it's
13:55 it's it's competent and a million tokens
13:58 is a big [ __ ] deal and it's
14:00 multimodal you can upload video you can
14:02 upload
14:04 audio you can upload 11 hours of audio
14:07 recordings
14:09 into Gemini 1.5 Pro and it it will just
14:15 be able to understand every all the 11
14:19 hours it's [ __ ]
14:22 Bonkers but it's hidden it's in a it's
14:24 in a teeny little Dev corner off in the
14:27 [ __ ] you know [ __ ] hole of Google Now
14:32 um Microsoft co-pilot
14:40 is should be better than it
14:43 is it's using gp4 as the back end but
14:48 they've got
14:49 the the the prompt for what it does so
14:54 hobbled with security that it's just
14:57 it's just not good and like the
14:59 interface is just they they just it's
15:01 just not good right yet so so so
15:07 uh so that's not really an answer is
15:11 it some parts of part of Google are
15:15 somewhat sort of better and then other
15:18 parts of
15:19 Microsoft suck a little
15:24 less not exactly a ring endorsement for
15:27 either is it just show the Gemini bear
15:31 use case yeah go to go to gemini.com and
15:35 say um make a picture of of a bear with
15:40 magical fairy wings standing underneath
15:42 a uh waterfall in a magical Woods at at
15:48 dusk and watch what it gives
15:52 you we have lovingly called it the the
15:54 Gemini bear
15:59 it is embarrassingly
16:02 bad so
16:12 bad how's everybody
16:16 [Music]
16:27 been I want to learn gener AI then keep
16:30 coming back here here's how you learn
16:31 generative AI start playing start
16:36 playing oh you want to know something
16:38 cool it used to be if you wanted to go
16:40 to chat GPT you couldn't go to chat
16:42 gp.com you had to go to chat.
16:45 open.com guess what go type in chat
16:47 gp.com
16:49 now watch it not
16:53 work what how did I do that
17:01 there we go
17:05 [Music]
17:06 um oh I did a cool thing I'll I'll show
17:09 you this too let's see where am I going
17:12 I'm going to go here and I'm going to go
17:14 here and I'm going to go chat gp.com
17:24 boom that says to me something news
17:27 coming
17:30 [Music]
17:32 apparently May 9th
17:35 Thursday chat GPT is launching a search
17:40 [Music]
17:46 engine imagine being able to design a
17:49 search
17:51 engine being the ones that made chat
17:55 GPT and you weren't burdened with having
17:59 a Legacy advertising
18:02 model what could you do I don't know
18:06 we're going to find
18:08 out and I would be surprised
18:16 if they didn't release a new version of
18:18 chat GPT as
18:20 [Music]
18:24 well but I'm also I've got a little
18:27 conspiracy theory going in my head right
18:29 [Laughter]
18:33 now I think
18:36 that I think that Sam Alman is just
18:40 [ __ ] toying with us I think he's he's
18:43 making all these announcements and not
18:45 really releasing
18:47 anything just enough that people get
18:50 their hopes up and then he doesn't
18:51 release anything and people are getting
18:53 more and more and more pissed
18:56 off and I think he's just he's just
18:58 gaging the barometer of rage and right
19:01 when a rage starts to hit its boiling
19:03 point then I think he's going to drop
19:05 some cool [ __ ] on
19:09 us oh man chat GPT has been hella slow
19:12 today yeah if if chat GPT gets wonky and
19:15 especially if it's wonky for more than
19:17 one day in a row that generally means
19:19 they're updating something on their
19:21 they're taking Banks of servers offline
19:23 to upgrade
19:25 software so if this is a major update
19:29 and if if they're switching the domain
19:30 to chat gp.com if that's going to be the
19:34 the the main interface
19:37 um yeah it would make sense that it's a
19:40 little shitty today yeah I saw I saw a
19:42 couple of
19:43 tweets like it just got stupid people
19:46 are rate limiting out after like three
19:48 or four
19:52 prompts oh the games they
19:54 play they are going to start ads on
19:57 their search engine unless you you have
19:59 paid chat
20:00 GPT all right if you don't have paid
20:04 chat GPT unless you can't afford it
20:08 unless you like can't give
20:10 up your three cups of coffee a month
20:13 from the the shitty Barista who treats
20:16 you like [ __ ]
20:19 anyway you should be spending the 20
20:21 bucks a month for chat GPT
20:23 [Music]
20:25 plus to going back to the question
20:28 before which is better you know Google
20:31 or co-pilot neither of them chat PT is
20:35 [Music]
20:37 better no one's caught them
20:40 yet everyone's like we're close we're
20:42 almost as good some benchmarks they were
20:49 better you can make money with J big but
20:52 no one beat him yet no one has
20:54 consistently beat him claude's the
20:55 closest one but claude's not
20:57 multimodal and Claude you can't it
20:59 doesn't have code
21:03 interpreter perplexity got uh searching
21:07 but it's not a very good
21:13 [Music]
21:19 llm oh [ __ ] who said that who's who's
21:22 expecting
21:23 tornadoes Wolfman Clint yikes we're
21:26 expecting tornadoes here in the next two
21:28 hours so if I drop you'll know why yikes
21:31 get the to a where do you go you go to a
21:33 basement doorway right or basement
21:36 bathroom or some [ __ ] like
21:40 that these tornadoes like like couple of
21:44 things are going on stormchaser
21:47 stormchasers have gotten more insane and
21:50 they've gotten better better weather
21:52 technology so the stormchasers are
21:54 filming these things like regularly
21:57 these tornadoes this year are [ __ ]
22:00 Bonkers they're like half a mile wide a
22:03 mile wide they're they're just
22:06 insane I mean normally you see it like
22:08 you know wipe out a barn or something
22:11 but these are [ __ ]
22:13 [Music]
22:14 nuts what's the best tool for market
22:18 research ah I don't know I I mean for
22:22 research in general perplexity is pretty
22:24 good
22:29 um this is a place where I think that
22:31 Google
22:34 um Gemini could be really
22:37 [Music]
22:41 good
22:43 [Music]
22:46 um I would say that the Gemini Pro 1.5
22:49 the million token context window could
22:51 be really interesting for research
22:53 because you could go get a whole [ __ ]
22:55 ton of um you know PDFs and and and you
22:59 know Market Research in PDF form and
23:03 just upload them all
23:06 into um and and charts and graphs and
23:09 even like audio podcasts of market
23:12 research people talking about market
23:14 research just upload it all into Gemini
23:16 Pro
23:18 1.5 and I bet you could do some really
23:21 interesting damage with
23:24 that so that's at aist studio. goole.com
23:29 and make sure you're on the 1.5
23:34 pro
23:36 [Music]
23:46 [Applause]
23:48 model I'm at work we go to the middle of
23:50 the building well stay safe Clint that's
23:53 yikes that's yikes
23:55 territory thoughts on the rabbit heard
23:57 it was heard it runs Android and has
23:59 mixed reviews yeah so here it
24:02 is
24:05 um it's gotten a lot of in my opinion
24:07 unfair
24:14 reviews and just unfair cuz
24:18 like I don't know I I I mean I guess I
24:22 guess we live in this in this world that
24:24 you know we [ __ ] did this to
24:27 ourselves we click the sensationalistic
24:30 [ __ ] right so we did it to ourselves
24:33 so if I say hey this thing is designed
24:37 beautifully built really well is way
24:39 [ __ ] faster than I thought it was and
24:42 the software is very 1.0 and you should
24:44 just wait for it to get better that's
24:46 kind of a Mambi pambi mediocre review
24:49 and no one will give a [ __ ] like the
24:51 rabbit R1 review I did right here on
24:54 this Tik Tok Channel nobody gave a [ __ ]
24:57 so but if I got on there and said this
24:59 thing is the scourge of the universe and
25:01 I want to do a class action lawsuit for
25:03 how horrible this device is you know
25:06 it's just an Android app don't
25:08 you I should have done it like
25:12 that but you know it didn't I just I
25:16 just told people what I really thought I
25:18 like it I think it's I think the
25:20 Nostalgia factor is through the [ __ ]
25:23 roof I think the build quality is really
25:25 good you can't really see like the the
25:28 camera does doesn't really catch it
25:29 right but it's like fluorescent like
25:31 Hunter orange yellow like I could wear
25:34 this out deer hunting and probably be
25:36 fine
25:39 um it's well
25:41 made the vision stuff is decently
25:45 um it's like decently quick so you
25:50 um to to turn on the camera you flip the
25:55 little you hit that button twice and see
25:57 there's a little
25:59 cylinder in the top there that's the
26:01 camera but it's rolled down right now
26:03 but if I go bang bang Z there a cute
26:07 little thing and if I turn this Z it
26:10 flips it around there so then what's on
26:12 that side Z now it's over
26:15 there
26:17 um so let's
26:20 see
26:22 um we'll do
26:25 this what do you see
26:29 he's going to call me an old
26:31 man taking a look
26:34 now the image shows a blurred close-up
26:37 view of a person's
26:41 face is that man
26:46 old let me take a
26:49 look yes the person in the image appears
26:52 to be an older man with gray hair and
26:54 wearing
26:57 glasses that's
27:00 rude taking a look
27:04 now I apologize but I am unable to
27:07 identify or name the individual in the
27:10 image so that's kind of cool and then
27:13 let's see double click it to go back to
27:16 this and then however I can describe
27:18 what I see no stop describing he's
27:20 getting self-esteem
27:25 issues I understand
27:28 Rudy Rudy psycho old man Mr Kyle I will
27:32 refrain from providing detailed
27:34 descriptions of individuals and images
27:36 in the future if you have any other
27:38 questions or need assistance feel free
27:40 to ask do you remember my
27:46 name I don't have the ability to
27:48 remember personal information like names
27:50 oh that's funny cuz it did so I asked
27:55 you can let me know your name and I can
27:58 oh yeah
28:00 actually yeah let's go ahead remember my
28:02 name it's
28:06 Kyle understood Kyle if there's anything
28:10 you need assistance with or any
28:12 questions you have feel free to ask Kyle
28:15 did you write that to my rabbit
28:20 hole I do not have the ability to write
28:23 or access the rabbit hole if you need to
28:25 make any changes or access specific
28:28 features you would need to do so
28:30 directly through the rabbit hole web
28:31 portal snowflake mode
28:36 activated I okay we just seriously now
28:40 we've got to give the rabbit R1 a fair
28:42 review
28:43 um okay I I I use the rabbit R1 and um
28:48 to be honest I'm feeling a little
28:50 triggered okay um it did a little fat
28:53 shaming uh and it it's aist it's it's
28:56 aist um I consider myself a younger
29:00 middle-aged man and and it said I was an
29:03 older uh rather rotund um Jolly fellow I
29:09 think we know what that means and uh I I
29:12 I take great offense to to the rabbit
29:14 this is clear this shows clear
29:17 bias this shows clear bias this is
29:19 high-risk AI according to the new
29:22 Colorado legislation uh Bill SB
29:26 24-25 this is highrisk high risk
29:30 highrisk artificial intelligence okay
29:33 and I feel biased I felt attacked I felt
29:40 shamed I think using the word
29:43 portly Was A Veiled Threat okay I really
29:48 do I really do I it's it's cute it's
29:54 cute oh if you shake it you get the the
29:57 preferences
29:58 [Laughter]
30:01 and
30:05 see how do I like the keyboard this
30:08 keyboard it's okay here's something
30:10 [ __ ] stupid so if you go to network
30:14 you have to use your you have to use
30:15 your little wheel which is fine it's
30:18 it's shitty but if you go to network and
30:20 then you join a network it makes you
30:23 rotate this thing and then it puts a
30:26 keyboard on screen and then you type on
30:27 it right like a little little mini iPad
30:31 right touchcreen
30:33 clever great now the keys are about the
30:37 size
30:41 of
30:43 um like a rly poly bug you know those
30:46 little rly Polly bugs like when they're
30:48 rolled
30:49 up like just larger than a BB that's how
30:52 big the keys are so with my rotund
30:55 fingers I'm in there trying to to type
30:58 and it's fine they they work okay but
31:01 then you get into this mode
31:06 and you can't touch
31:09 it you you have to use the scroll wheel
31:13 and then if you go into a setting you
31:15 have to scroll all the way to the top to
31:16 to go back
31:18 out so you're like seven seven layers
31:21 down in this interface and you're like
31:23 scrolling up and then clicking and then
31:25 scrolling up and then clicking and uh oh
31:29 [Laughter]
31:32 man good Lord people anyway well it
31:37 looks like we might have I I don't know
31:38 if we' we've killed the Colorado bill
31:42 but we we bitched about it enough
31:46 that I think it's at least causing
31:48 political issues so hopefully it slows
31:50 it down it would be nice if it didn't if
31:52 it didn't laugh if it didn't laugh I
31:54 just saw L TS is like L la la
32:00 be right back Mary Mary's going to be
32:02 right back oh no he used the word
32:07 rotund exactly you can't use the r
32:12 word okay
32:14 listen Okay the the Society of rotund
32:17 people they do not like that
32:20 word
32:22 we we're going with fabric challenged
32:25 [Laughter]
32:29 we need to make images proms prompts
32:31 using the word
32:32 portly no cuz no cuz all it's going to
32:35 be is pictures of fat versions of
32:41 me do portly hedgehogs or Turtles fat
32:45 turtles is
32:47 fine fat learning
32:49 lab coaches leaders whatever the [ __ ] I
32:52 am here not
32:56 good oh oh lordy lordy lordy all right
33:01 let's
33:03 see trying to think if anything went
33:07 down today on the internet front I don't
33:09 think so Apple's got I can't believe
33:13 Apple they're just they're just
33:18 weird they're just weird
33:21 tomorrow they've got an iPad Pro and a
33:24 new pencil 2 event
33:32 yay like are people really clamoring for
33:35 more powerful iPads at this point
33:38 weren't they good enough like five iPads
33:41 [Applause]
33:42 ago isn't
33:46 this
33:48 $129 thing that you can get your company
33:51 name stencl
33:55 on good enough
33:58 with its pressure
34:02 sensitivity please show the world GPT
34:04 memory I'm in love with it okay I have
34:06 so Rudy I haven't actually used it
34:08 enough to really know I I mean I can
34:10 show you what I've
34:13 got um I haven't been using it I haven't
34:15 been using chat GPT for the past week or
34:19 two yes we need poor more more iPad
34:22 power for CAD yeah exactly exactly it's
34:26 like you want to be you you want to be
34:28 3D modeling while you're while you're
34:30 sitting on the front porch sketching
34:33 away with your
34:35 iPad send a
34:40 print oh and then well I guess but then
34:43 you could just watch your 3D printers
34:45 all from your iPad like a nine up you
34:48 know nine camera setup and one of your
34:51 cameras quite frankly Vicki could be a
34:53 rabbit R1 if you had
34:56 one I told you to buy won when they
34:58 first came out this is you know what
35:01 this hasn't been talked about as yet the
35:03 ultimate 3D printer
35:06 accessory nope it hasn't you know why
35:08 cuz it's
35:14 not oh are you joking about iPad with
35:18 CAD I hope you
35:19 are Vicki's like I'm really offended
35:22 he's he's seriously I like I we really
35:26 need the M4 iPad so I can do cat on the
35:30 iPad it's like I don't know man I tried
35:35 when the iPads first came out the iPad
35:37 1.0 I sold all my
35:40 laptops and bought an iPad and I tried
35:43 to go all in and and
35:45 use oh no I'm not
35:48 joking well then you'll be excited about
35:50 tomorrow and apparently maybe the pencil
35:54 2.0's got some sort of magic 3D thing
35:57 with it that's going to be great
35:59 um I have a fantasy that Apple's going
36:01 to say something about AI but I know
36:02 they're not like they're just
36:04 not they're just going to be like yeah
36:07 we'll just
36:08 wait everyone else is losing their mind
36:11 what's chat gp5 going to be like how
36:13 many billions have you invested and
36:14 Apple's like
36:18 de dear dear dear hey you want to see
36:24 something really
36:25 exciting we've changed the radius of the
36:28 corners of the iPhone and we went from
36:31 rounded edges to straight edges then we
36:34 went back to rounded edges and now we've
36:37 gone back to straight edges it's [ __ ]
36:40 radical we we're just people are
36:43 expecting us to Zig and we certainly zag
36:46 Mister I'll tell you that boy
36:48 howdy when did Apple become a [ __ ]
36:52 Midwest cheesecake judge
36:56 [Laughter]
37:04 why is
37:05 some company out designing
37:11 Apple some little startup some little
37:14 pesky
37:16 startup why didn't Apple invent this or
37:19 some version of this the one button
37:21 device the one button
37:26 iPhone why why
37:29 people corporate
37:32 Mala it's what you get when you put a
37:34 CFO in as the [ __ ] CEO you replace a
37:38 Visionary with a CFO you had a [ __ ]
37:41 decade and a half of [ __ ]
37:45 boredom good
37:55 Lord I don't I I don't think he should
37:57 take we off anymore he seems rather
38:00 angry all right
38:03 um
38:09 feisty I'm just like could just [ __ ]
38:12 put something out put something out put
38:14 something out about AI Apple like give
38:17 us a [ __ ] agent make Siri not suck
38:20 and Siri if you [ __ ] show
38:24 up right she didn't
38:30 like where are they it's weird it's
38:33 weird Apple stopped innovating they
38:38 did it's just so
38:41 boring you know it's funny like the the
38:43 past five years Apple to me has felt
38:46 like the John Scully area
38:49 era where remember they all got beige it
38:52 was all beige business boxes and then
38:54 Steve Jobs came back and he was like
38:56 [ __ ] it flush all that [ __ ] down the
38:58 toilet Johnny Johnny IES come over here
39:01 make some fruit colored
39:04 computers remember how exciting that was
39:06 and then it was just like [ __ ] every
39:08 two years some new industry defining
39:13 product and then now just [ __ ] it
39:17 just
39:19 sucks anyway all
39:24 right I think they added an update to my
39:27 Alex Alexa and it sounds absurd oh
39:31 no Siri is biting her time she certainly
39:35 is I love my iPad so much that's cool I
39:39 you know I I'll be honest I
39:43 haven't I never really figured out a use
39:46 case for my iPad that I
39:49 that is really a killer app in fact I
39:52 feel like between this and my iPad
40:04 trying to thing I'm trying to trying to
40:06 measure my words carefully
40:11 here I have no interest in in taking my
40:14 iPad out of the house which its whole
40:16 point is to be mobile and to carry it
40:18 around and always be there and this I
40:21 carry with me I don't use
40:23 it but I'm carrying it with
40:25 me so I don't know I I don't know yet I
40:28 don't know jury's still
40:31 out portly images in Irregulars oh
40:39 no oh Irregulars what are we going to do
40:42 with you the portly
40:50 Trio yeah how do you know how do you
40:52 know if Turtles are obese they just fill
40:55 up their shell
41:00 a portly bottles of bottle of Blandon
41:03 I'll take
41:05 that yeah that's th this is what this is
41:08 what the rabbit R1 thinks I look
41:12 like oh that's nice Claire Jacobs that's
41:16 beautiful welcome
41:18 back welcome back your dreams will your
41:22 ticket out welcome back to that same old
41:26 place that you laugh about
41:28 well the names have all changed since
41:30 you hung
41:32 around that's very good I love that
41:34 Claire thank
41:39 you oh I put up on the YouTube channel I
41:42 put up the um the weird Mary
41:47 um when we did when we did weird Mary
41:50 the other night um May 2nd I think we
41:53 did it um I thought that whole it took
41:57 us 2 hours hours it took us two hours to
41:58 make that song and I did a I did a
42:01 LinkedIn tweet today about it um
42:04 about you know there's this
42:06 misconception with AI generated stuff
42:08 that it's like you just push a button
42:10 and you get out the stuff and you're
42:11 lazy and it's not real art and I thought
42:15 that that 2hour session was was just a
42:17 good example of it was like a jam
42:19 session where everybody was throwing out
42:21 ideas and I was trying different things
42:23 and sometimes udio was [ __ ] brilliant
42:26 and sometimes it was just trash fion you
42:29 know and it and it was it was like 2
42:30 hours of like working and working and
42:33 working and working and working to get a
42:35 thing that sounds like a
42:39 song and it's like you
42:42 know who wrote like the the the
42:45 questions I put in my LinkedIn
42:51 piece let me
42:54 see uh uh uh uh uh uh uh
42:58 [Music]
43:01 oh then there was that Randy Travis
43:02 thing that's
43:06 amazing yeah so the process we did idea
43:09 concepting in chat gp4 with me talking
43:11 like Sling Blade then we did lyric
43:14 generation then there were lots of
43:16 rewrites and ideas from the Irregulars
43:18 from the audience from you from y'all
43:21 from the
43:23 community then we had numerous attempts
43:25 to get udio to generate something
43:27 interesting then we had many attempts
43:29 trying to extend the song that were some
43:32 were just
43:33 bad then I published it then we finally
43:36 got it done I published it to social
43:37 media then I realized oh I should add a
43:40 an image to it and I made an image in
43:42 mid Journey Only to realize you can't
43:43 upload an image to yudo so we had to
43:46 make an image inside
43:47 yudo then I republished it but it didn't
43:50 update any of my twitters with the new
43:53 image so then I got cranky and left
43:58 and then uh let's see if Pi was in the
44:02 rabbit R1 would it be better it would be
44:04 so much better Rudy it's it it's not the
44:08 the one of the problems with the rabbit
44:09 right now is it's using perplexity and
44:12 it's got It's got shitty prompting in it
44:15 it's like it it's just it's not it's not
44:18 pie it's not chat GPT and so it's just
44:21 not
44:24 good um but here's the questions I posed
44:27 so so we put out this song If the song
44:29 is good and it moves you does it matter
44:32 how it was
44:34 created and I mean that very seriously
44:37 like if you hear I'll I'll play I'll
44:39 play um I'll play weird Mary the the
44:42 song we made is called weird Mary from
44:44 Cedar Hill and if you just heard that
44:47 and you and you thought oh that's that's
44:49 a cool song and it moved you and it was
44:51 interesting and and was compelling in
44:53 some
44:55 way does it matter how it was
44:59 created so that was one question then
45:02 what percentage of human involvement
45:04 makes it not AI
45:06 generated or what percentage of AI
45:09 generation makes it not human
45:12 generated and then should should the
45:15 song be labeled AI music or should it
45:17 just be labeled
45:20 music and then the last question is if I
45:22 learn the song on my guitar and sing it
45:25 is it still AI music
45:32 [Music]
45:46 see Hill is quiet
45:50 town uh the oak trees
45:54 sway I don't remember the words
45:57 [Music]
45:57 [Laughter]
46:00 oh the Train the sound of the trains
46:02 echo through the
46:03 [Music]
46:05 day right is that AI music if you didn't
46:09 know how that was created and you just
46:11 heard me sing that is that an AI song I
46:15 don't
46:17 know I do not
46:20 know how's this how is that any
46:22 different than the gorillas exactly
46:24 exactly how is that any different than
46:26 DJs how's it any different than hip-hop
46:29 wow in hip-hop they yeah they spend
46:32 hours taking lots of little janky things
46:35 that don't really work together and
46:38 slowly molding them like clay into
46:41 something that sounds like music I did a
46:43 cool thing so so I think it was Amelia's
46:46 wife said let's try to get this thing on
46:48 Spotify so I've got podcasts on Spotify
46:51 but I've never published a song on
46:53 Spotify so here's what I learned and
46:55 Amelio's wife told me this you've got to
46:57 go through a distributor you can't
46:59 upload songs directly to Spotify so I
47:01 found a free distributor and then one of
47:04 the things that they had was they had
47:06 automatic AI mastering and I thought
47:09 well let me try it and they were really
47:10 smart about it they didn't say pay first
47:12 and then we'll let you listen to how it
47:14 sounds they were like um click this
47:17 button and we'll a we'll remaster your
47:19 song optimized for any speaker and and
47:23 then if you like it you can download it
47:25 and so I liked it and I was like well
47:27 how how much is it and it was six bucks
47:28 and so I was like I'll pay six bucks for
47:30 that so that was that's pretty good
47:32 marketing right there I wonder how long
47:34 it'll take to go live with the rabbit R1
47:36 oh that's interesting yeah uh could be I
47:39 it what's going to be interesting about
47:41 the rabbit R1 is how long it takes
47:45 them how long they can survive with it
47:48 being a proprietary device or do they
47:50 just open source the operating system CU
47:52 someone's already cracked it and got it
47:54 running on Android so
47:57 but here listen to the I I'll I'll go to
48:00 udio and play a little bit of the
48:02 beginning
48:03 of of weird Mary from Cedar Hill so this
48:08 is the one on udio and then I'll play
48:09 the remastered version I'll just play
48:11 like a little bit of this to start and
48:12 then we can compare it wait let me just
48:14 go get it up and running so so I have
48:21 it open with can I get it
48:26 there is that
48:33 it no that's not
48:37 it oh here it is okay all right
48:42 so
48:45 yeah all right so here's here's a little
48:47 bit of weird Mary from Cedar
48:49 Hill this was made in if you're
48:51 wondering this is made in yudo people
48:53 hated on sampling for a very long time
48:55 before it became ex accepted
48:57 yeah that's that's exactly what I'm I'm
48:59 think is going to go on here and you
49:01 know what it's going to take is and and
49:03 I think someone just put out a song well
49:07 Randy Travis just put out a song and
49:09 it's it's really really moving
49:11 actually um if you don't know Randy
49:13 Travis had a stroke in 2005 and it took
49:16 his voice and so AI has finally gotten
49:20 good enough that that he could put out a
49:23 new song and he's still alive and he's
49:25 you know he just can't sing
49:27 um and it was the the music video is
49:31 really cool because it's it's him and
49:32 all and his friends is like you know
49:35 friends and family listening to his
49:37 first song in 15 years all right so so
49:41 here's this this is uh weird Mary from
49:43 Cedar Hill
49:47 [Music]
50:02 [Laughter]
50:07 [Music]
50:19 Cedar Hill is a quiet
50:22 town where the oak Tre s all right so
50:26 you got that sound and Mina I saw your
50:28 thing about uh product videos I I'll
50:31 show you something here in a second um
50:34 okay so now listen to the remastered
50:35 version version of this
50:40 [Music]
50:55 [Laughter]
51:05 [Music]
51:08 [Laughter]
51:12 CER Hill is a quiet
51:16 town where the oak trees
51:20 sway the whistle of the train Echoes
51:24 through the day
51:27 [Music]
51:29 people GRE you with they fixed the
51:32 clipping issues doesn't it sound betters
51:35 change the guitar is more clear the
51:37 voice is more clear
51:39 [Music]
51:44 UNF she grew up
51:47 tough in the town we called
51:50 [Music]
51:52 Home Folks never took to her no matter
51:57 at her
52:00 tone longing to be on the
52:04 inside longing for love just a
52:08 little she found herself as a guest in
52:12 the nervous
52:14 [Music]
52:16 Hospital weird Mary from Cedar
52:20 Hill she hows when she's
52:25 blue of sadness and shame under the
52:29 southern
52:33 Mo but when the sun sets
52:37 behind those Appalachian
52:41 Hills Mary's voice it carries and the
52:45 whole town
52:49 chills she walked to the gas
52:54 station winked at the clerk
52:58 she stole her a soda then she called him
53:02 a
53:05 jerk in her own little
53:09 world she floats in the
53:14 breeze among the town's folk who whisper
53:19 with
53:22 ease in a town that never
53:27 understood her
53:30 soul she found beauty in the wild let
53:34 her story
53:38 unfold her laughter Echoes In The
53:42 Wind she was never the
53:46 same weird Mary from Cedar
53:50 Hill you'll remember her names
53:56 [Laughter]
53:57 [Music]
54:19 [Music]
54:25 come on
54:28 all right so that's weird Mary now if
54:31 you just heard that as a
54:36 song all that's missing is some champy
54:39 backup freaking love it
54:42 uh yeah so couple of things so I'm this
54:45 is posted I'll post the the new um this
54:48 mastered version of it on um I'll post
54:53 it on the AI Salon it's not there right
54:55 now um but I'll I'll post the mastered
54:57 version there I'm also I also did um my
55:00 my song
55:02 is this song is um I signed up with a
55:07 service called a
55:08 muse that's the one that did the
55:10 mastering and then it's in review right
55:13 now so it's going to be interesting I
55:17 couldn't it so I put it up under my
55:21 account because I wanted under my
55:24 Spotify account and then it said
55:25 Performing Artist
55:27 and I had to choose Kyle Shannon's
55:29 Performing Artist and then I was like
55:32 well let me add a secondary artist and
55:35 then maybe I can change the secondary to
55:36 primary so I changed the secondary
55:39 artist to yudo as the first name and
55:41 Comm as the second
55:44 name and that's the secondary artist on
55:46 it but I couldn't remove myself as the
55:49 Performing Artist so so I'm now the sing
55:52 singer of this although I will learn it
55:55 on guitar well enough to sing it it's
55:58 you you know what's funny about it is
56:01 it's it's really weird musically like it
56:06 sort of goes into the melody of the
56:08 chorus halfway through the second verse
56:11 and then it's it's sort of like a
56:13 precour halfway through the second verse
56:16 so so it's kind of like six line stanzas
56:19 instead of groupings of four it's kind
56:22 of weird um but it but it kind of works
56:26 like there's
56:27 something the offness about it
56:31 works and it's a 17 like the E
56:36 minor it's a song in E
56:38 [Music]
56:43 Minor it's a 17 you know normally you
56:47 got like a
56:48 145
56:50 [Music]
56:53 like and this one's like
56:57 it's like
57:00 [Music]
57:01 me oh I want to play that on my live
57:04 play it on your live Do
57:06 It um oh and let me show you something
57:09 else so Allan Allan from in
57:11 here got
57:13 inspired and after we did this Allan
57:17 went
57:19 off so first of all he loved the song
57:22 enough he played it for his um for his
57:26 drama a coach at his I guess at his kids
57:30 school Alan
57:32 hilbeck and so you can make it a Tik
57:35 Tock sound oh I could make it a Tik Tock
57:39 sound couldn't I I don't know how to do
57:41 that Tams but yes yes I could I know
57:44 that people have done
57:46 that that was the other thing Alan Allan
57:48 was asking me about copyrights and how
57:50 do we protect it and this and that I'm
57:51 like oh my God
57:54 like let's just let's just let's not
57:57 over complicate it and take the joy out
57:59 of it just yet oh he's here Mr Studio
58:03 dhj I shared it with the high school
58:04 theater teacher that's right so so the
58:07 high school theater teacher oh by the
58:09 way it's windier than [ __ ] here
58:12 and my lights are dimming I might you
58:15 might lose me so so um I wonder if AI
58:20 can generate the Sheep music oh it can
58:22 I'll show you refusion so so I'm
58:24 terrible at pulling chords out of things
58:26 so I'll I'll show you a cool piece of
58:29 software I've had for years and it it
58:31 actually um analyzes the music and gives
58:34 you all the chords so so I just I just
58:36 wrote down the chords from that I didn't
58:37 actually figure it out on my own um but
58:41 yeah my my electricity right now is kind
58:43 of crapp in the bed so if I if I go away
58:45 it's not you it's me anyway so Allan
58:48 shared this with the High School Drama
58:49 teacher the High School drama teacher's
58:51 response was I love it can we turn this
58:53 into a play so I said yes [ __ ] yes
58:56 let's make it into a high school play um
58:59 yeah the lights are dimming yeah it's
59:01 it's not good it's not
59:04 good um and then
59:07 um and then what else oh and then Allan
59:10 made this I I I hope you're cool with me
59:12 showing this Alan because I think this
59:14 is so so cool so this is this is a where
59:18 are you at I'm in
59:19 Denver wow weird Mary in your homes
59:21 playing with the wind yeah exactly well
59:25 wait till you see this video all right
59:27 so here we go let's
59:39 [Music]
59:45 let's it's good with the flickering
59:47 lights
59:53 [Music]
1:00:05 the town's people will finally learn the
1:00:07 truth and discover the secrets hidden
1:00:09 Within
1:00:12 [Music]
1:00:23 [Laughter]
1:00:26 [Music]
1:00:28 [Applause]
1:00:29 [Music]
1:00:36 yeah is that
1:00:40 great I like it I like it and I mean
1:00:45 listen this is for
1:00:48 [Music]
1:00:50 me what a AI is all about yeah isn't it
1:00:55 great this is what AI is all about right
1:00:59 because like did I did I have the idea
1:01:05 for weird Mary in my head
1:01:08 no no it was like a let's see if we can
1:01:13 make a song and we went to chat GPT and
1:01:15 I talked like Sling Blade I talked like
1:01:17 an idiot into chat
1:01:20 GPT what a trailer yeah isn't it
1:01:24 good pitch this to a screenwriter Kyle
1:01:27 oh I'm a screenwriter I can pitch it to
1:01:31 myself I'm in the middle of writing a a
1:01:33 musical
1:01:36 though so I don't I I don't want another
1:01:39 writing project right now because I'm
1:01:41 really excited about the musical
1:01:43 actually it's really good um but I think
1:01:46 weird Mary I like listen I mean like
1:01:49 have that it let's let's make this a
1:01:50 play right did he use AI to create the
1:01:53 trailer yes if so which programs and
1:01:55 apps uh
1:01:57 after we got the screenplay written
1:01:58 between myself and the theater teacher
1:02:00 there you go um how was the trailer
1:02:03 developed uh I assume Dolly and uh
1:02:07 Runway ml but what what tools did you
1:02:09 use Mr Studio
1:02:12 Allan my adaptive brightness is having a
1:02:15 moment trying to keep up with your
1:02:16 lights yeah my lights this is not
1:02:20 looking promising
1:02:22 people although I do have a battery
1:02:24 powered little you know this thing so
1:02:27 but I just don't know if I'll have
1:02:29 internet oh it's not good it's not
1:02:34 good was it only udio used for the song
1:02:37 um well sort of so
1:02:42 um the concept for the song the lyrics
1:02:47 of the song were all written in chat
1:02:51 GPT and then we did like I don't know
1:02:53 three or four iterations in chat GPT and
1:02:56 then we took little pieces of it and
1:02:58 then we rewrote individual so the way
1:03:00 you make things in in udio is you do
1:03:03 them in 30- second chunks so we took the
1:03:05 first verse and we put that in yudo and
1:03:08 then we just basically just kept making
1:03:11 songs and describing it until we got to
1:03:13 something that didn't suck there were a
1:03:15 couple of candidates quite frankly there
1:03:17 were a couple other ones there was one
1:03:19 that had a female singer that I I heard
1:03:23 it this morning and and I thought that
1:03:27 was that was pretty good too but there
1:03:29 was just something about that one that
1:03:31 kind of had the that edge to it and then
1:03:35 and then it was just iterations within
1:03:37 yudo where we were changing out the
1:03:39 lyrics and you know we do a version of
1:03:42 it and then someone would have a better
1:03:43 idea for lyrics like Joker had the
1:03:46 nervous Hospital line and then the we
1:03:50 did we did some hand crafting on the the
1:03:54 she bought her she stole herself aoda
1:03:57 and called called the clerk a
1:03:59 jerk which that was a pretty good line
1:04:01 there there are a couple of really good
1:04:03 lines in there that came straight out of
1:04:04 the conversation on the
1:04:06 channel so the so the lyrics are
1:04:09 decently
1:04:10 manipulated the music is just out of
1:04:14 udio but it's like in 30 second chunks
1:04:17 and like sometimes like we had a real
1:04:18 tough time getting a chorus that didn't
1:04:22 suck like we had a good first verse we
1:04:25 had a good second verse the second verse
1:04:26 took a little while and then the chorus
1:04:29 we kept just going and going and going
1:04:31 and it just didn't seem like we were
1:04:32 going to get
1:04:34 there yeah I love nervous hospital I
1:04:36 know me
1:04:40 too I'm still here that was just I just
1:04:43 had to prove I'm
1:04:44 human
1:04:47 Um Papa JJ did you'll remember her name
1:04:50 well I know it makes me chuckle I know
1:04:53 what's cool is like everyone will
1:04:54 remember like the the piece that they
1:04:56 contributed to it which I I think that's
1:04:58 really [ __ ] cool
1:05:01 um and then I put when I when I
1:05:04 published it I put in the so in the in
1:05:08 the metadata when I saved the
1:05:12 song so I took the song into Adobe
1:05:17 Audition and then I just did a stereo
1:05:19 split to to to define the stereo spread
1:05:22 a bit more and then I put metadata in it
1:05:25 so in the I think it was in the
1:05:28 songwriters I was primary and then the
1:05:31 Irregulars were secondary writer or
1:05:33 something like that so the Irregulars
1:05:35 the Irregulars are in the metadata which
1:05:38 is kind of cool I'll remember my line
1:05:41 nervous Hospital slaps it really does it
1:05:44 really does let's let's see let's go
1:05:46 back to the let's
1:05:52 see Cedar Hill is a quiet town uh the
1:05:56 the first two lines are just right out
1:05:57 of chat
1:06:00 BT but the stories unfold each unique
1:06:03 and strange I think that was Silver Fox
1:06:05 we we I think we massaged that a bit she
1:06:09 grew up tough that was my line in a town
1:06:12 we called Home Folks never took to her
1:06:15 no matter her tone I think that was chat
1:06:18 GPT longing to be on the inside was chat
1:06:22 PT but then longing for love just a
1:06:25 little she found herself as a guest in
1:06:27 the nervous hospital I think we wrote
1:06:29 rewrote that those two lines about 20
1:06:33 times but I'm I'm like they're they're
1:06:36 both so weird that they're just right
1:06:40 longing to be on the
1:06:42 inside longing for love just a little
1:06:46 that like that line longing for love
1:06:48 just a little is is so weird and then
1:06:51 it's followed up with she found herself
1:06:54 as a guest in the nerous is hospital and
1:06:57 Hospital's got a weird Accent on it so
1:07:00 it's like just
1:07:01 off weird Mary from Ceda Hill she house
1:07:04 when she's blue of sadness and shame
1:07:07 under the southern
1:07:09 Moon and then the she walked into the
1:07:12 gas station winked at the clerk stole
1:07:14 her a soda then she called him a jerk
1:07:16 that was that was all
1:07:19 human she floats in the breeze among the
1:07:22 town folks who whisper with ease I think
1:07:24 that was Silver Fox also
1:07:27 in a town that never understood her soul
1:07:29 she found beauty in the wild her lto e
1:07:32 Echoes her lto Echo her laughter Echoes
1:07:34 In The Wind she was never the same weird
1:07:36 Mary from Cedar Hills you'll remember
1:07:38 her
1:07:39 name so anyway that was it that I mean
1:07:42 it was a it was a really fun exercise
1:07:44 but it's like you know it was a lot of
1:07:46 work right like we were we were [ __ ]
1:07:50 hustling we need to make a music video
1:07:52 out of the song oh so so I can show you
1:07:55 there's a there 's a
1:07:57 um there's a Discord service called
1:08:02 noisy and I
1:08:05 did here it
1:08:09 is longing for love just a little she
1:08:13 found herself as a guest in a nervous
1:08:17 Hospital
1:08:28 all right so
1:08:33 here's here's the video so the this the
1:08:36 the site is called noisy n o i s
1:08:41 e and you have to make things on a
1:08:44 Discord server and it only made it a
1:08:46 minute long and the song's almost 4
1:08:48 minutes long so I don't know if you can
1:08:51 extend this or or if you can put them
1:08:53 together but we could also like if you
1:08:56 wanted to we could do it kind of like
1:08:58 Kelly Bosch does it and just go make a
1:09:00 bunch of images in mid journey and then
1:09:03 animate them in Runway or or one of
1:09:05 those the the problem is I don't have
1:09:07 unlimited animation
1:09:10 accounts and so I chew up my like I
1:09:13 think in a 4-minute video we would
1:09:15 probably chew up all of my all of my
1:09:17 things who needs Hotel California when
1:09:19 you got this I know right so weird Mary
1:09:22 from Cedar Hill so here's the video that
1:09:24 that this thing made it's pretty good
1:09:26 let me can
1:09:33 [Music]
1:09:35 I I guess I'll do it like this
1:09:40 [Music]
1:09:55 that's janky right there but
1:09:57 [Music]
1:10:13 coolar
1:10:17 hillet Oak love that
1:10:21 shot whistle of the train
1:10:30 people you
1:10:33 [Music]
1:10:38 a that's all we got so so anyway that's
1:10:42 noisy but I mean there's there's
1:10:44 probably an infinite number of ways we
1:10:46 can we could do this but um Joker wasn't
1:10:49 the guy that served soda at the pharmacy
1:10:50 in Old places called a soda jerk yeah
1:10:52 absolutely that's that that play on
1:10:55 words
1:10:56 works and she called him a jerk what I
1:11:00 like about that line is it's like a
1:11:02 whole story right it's
1:11:04 like either she remembers that or he
1:11:08 talked about that right like why why is
1:11:12 that little scene where she stole a soda
1:11:15 and called him a jerk after she winked
1:11:16 at him why is that in the song Right CU
1:11:20 it's because there was some bigger story
1:11:22 there that's that's what I like about
1:11:23 the song is that there's just a there's
1:11:26 a lot not in
1:11:28 it and it's a song like in fact we we
1:11:32 cut out the outro and a whole other
1:11:35 verse there was an whole other verse and
1:11:37 a whole outro that we
1:11:39 cut we're moving into AI stardom in our
1:11:43 own
1:11:49 minds um somebody wanted to see about um
1:11:54 product videos let's go
1:11:56 go let's go to Hay Jen and I'll show you
1:11:59 something cool if you haven't seen it
1:12:01 cuz it's pretty [ __ ]
1:12:04 wild um hey genen if you haven't seen ha
1:12:07 Jen there there's so much on he Jen
1:12:09 that's
1:12:10 wild video
1:12:13 translations
1:12:16 um okay they changed the name of this
1:12:19 this used to be called URL to
1:12:21 ugc but it's now called URL to ad
1:12:28 oh this is cool app
1:12:32 store all right I'll do that let's go
1:12:36 grab the story fine app we're gonna
1:12:39 we're going to make a we're going to
1:12:40 make a an ad for the for my my company's
1:12:42 B2B
1:12:47 app uh let's see app
1:12:53 store seea h is a quiet town where the
1:12:58 oak trees
1:13:05 sway the sound of the trains Echoes
1:13:08 through the day that's another weird
1:13:10 line that just is cool it it just
1:13:14 works why opening the
1:13:18 app I don't want to open in the app I
1:13:20 guess I can open in the app all right
1:13:24 fine d
1:13:34 all right star Van star
1:13:46 [Music]
1:13:53 van whoa
1:13:57 why why that why that not
1:14:00 working why why I can't find me why
1:14:03 can't find me
1:14:05 on my
1:14:07 thing I
1:14:09 just let's see uh can I do SL Story vine
1:14:13 is that g to work no that's not going to
1:14:19 work I screw
1:14:21 it all right we'll go grab an Amazon
1:14:24 product they're more
1:14:27 okay Amazon all right what are we going
1:14:29 to what are we to search for here we'll
1:14:31 do Mother's Day gifts no that's
1:14:35 past Amazon Basics buy again live
1:14:38 streams
1:14:39 Pharmacy household Health and Beauty
1:14:42 this is this is where everyone's making
1:14:43 the money the skin creams let's see uh
1:14:48 retinol skin cream skin retinol skin 2.0
1:14:52 oh this is going to be
1:14:54 solid all right right let's see let's go
1:14:57 get us a
1:15:01 serum skin renewing Retinol Serum we got
1:15:05 to get one that's
1:15:06 decently put together because we need
1:15:14 video okay it's got eight videos this
1:15:17 one's perfect okay so look so over here
1:15:20 you got all these pictures and then you
1:15:22 got eight
1:15:23 videos double size
1:15:26 20 o Retinol
1:15:27 Serum but this has also got hyaluronic
1:15:30 acid this isn't just
1:15:32 retinol here's the thing if you think
1:15:36 you can just get away with a Retinol
1:15:37 Serum without the hyaluronic acid you're
1:15:41 fulling yourself
1:15:43 mister okay so we're going to grab this
1:15:49 URL and then I'm going to paste that in
1:15:52 and say generate a video
1:15:56 and then and then what this does is kind
1:15:58 of Bonkers so so what they've done is
1:16:01 they've cut deals with these influencers
1:16:03 or these people that call themselves
1:16:06 influencers so we'll say double sized 20
1:16:10 o Retinol Serum for face with hyaluronic
1:16:13 acid this is this is a lot
1:16:31 elevator okay we're going to kill it
1:16:34 here that's a lot all right so there's
1:16:37 your big thing and you can add a logo if
1:16:39 you
1:16:40 want weird Mary night cream it's weird
1:16:44 Mary is pissed off she's like I want the
1:16:47 song on Spotify why is it not on Spotify
1:16:50 this is okay first those [ __ ] at
1:16:53 Cedar Hill treated me just like [ __ ] for
1:16:56 my entire life and now in the afterlife
1:16:58 you can't even get my song on [ __ ]
1:17:00 Spotify seriously okay I'm [ __ ] with
1:17:03 your lights I'm [ __ ] with your lights
1:17:05 I'm just doing it I'm doing it you know
1:17:08 what I say I don't care Kyle I just
1:17:11 don't you know whatever with your
1:17:14 Irregulars or
1:17:16 whatever so she's a little cranky right
1:17:19 now that's okay okay so you can upload a
1:17:22 logo you you put in the product name in
1:17:24 the description
1:17:26 and then here's all the media right so
1:17:28 it pulled this in from the Amazon site
1:17:30 so it's going to incorporate all this
1:17:32 into our
1:17:33 video so that's a thing to know that if
1:17:36 you grab a product that doesn't have a
1:17:38 lot of stuff you're going to have a
1:17:39 boring video but if you grab one that
1:17:41 has a lot of stuff it's going to be
1:17:42 interesting so we go
1:17:44 next and then we can we can tag
1:17:47 women
1:17:52 Housewives professional sure health
1:17:54 conscious individuals of
1:17:58 course fashion followers beautiful it's
1:18:01 going to be in English the default local
1:18:05 interesting we'll go us I don't know why
1:18:08 uh I don't give a [ __ ] about that I'm
1:18:10 not pan what do you what are you people
1:18:12 stupid all right
1:18:18 next it's haen yeah this is haen by the
1:18:21 way the these these lights are not good
1:18:24 people Mary is
1:18:28 pissed okay then what it does is it
1:18:31 writes us scripts so there's the
1:18:33 solution or oriented script tired of
1:18:36 dull skin wrinkles and acne breakouts
1:18:39 introducing the double-sized Retinol
1:18:41 Serum all right super confident
1:18:44 introduced double-sized Retinol Serum by
1:18:46 insta skincare say goodbye to dull
1:18:49 skin hey there are you ready to elevate
1:18:52 your skin care game I feel like that's
1:18:54 what we need to go go
1:18:56 with introducing double-sized retinol no
1:18:59 this is hey there are you ready okay so
1:19:01 we're going to do quick
1:19:02 demonstration you can edit this if you
1:19:04 want if you don't like that but now
1:19:06 we're going to go
1:19:08 next and now we pick an
1:19:10 influencer I
1:19:13 think I'm pretty
1:19:15 sure weird Mary fan art on Irregulars
1:19:18 okay
1:19:21 cool all
1:19:23 right so
1:19:26 we've got this Asian woman and then
1:19:29 you've got all these different templates
1:19:32 there where you know she's walking in a
1:19:36 natural environment she's green screened
1:19:38 out and then they're you know in inter
1:19:41 cutting the stuff
1:19:42 there they've got different
1:19:46 avatars so let's go see who else we got
1:19:49 we got Darren I don't think Darren's
1:19:50 going to be our retinol
1:19:53 spokesperson oh I think I what's her
1:19:55 name Gail I think Gail's going to be the
1:19:59 one oh oh it's
1:20:01 Kim Kim is it GNA be
1:20:05 Kim they need some more [ __ ] they
1:20:07 need they listen listen I'm not saying
1:20:11 that these are not the most popular
1:20:13 influencers on the planet
1:20:16 but let's do Kim do we want Kim sitting
1:20:20 inside I think Kim sitting in her car I
1:20:22 think Kim selling retinol skin care for
1:20:25 car is the way we want to
1:20:31 go okay so we're going
1:20:36 there and then we got to go pick a
1:20:38 template so we want to pick one where we
1:20:41 see her in her car CU I think it's
1:20:46 funny so there she is in her car full
1:20:48 screen plus videos social share there
1:20:51 she's not in her car
1:20:55 there she's in her car intro plus
1:20:57 voiceover mixed intro plus voiceover I
1:21:00 think this is cool she's in her car but
1:21:02 we get little overlays all right so
1:21:05 we'll pick that one and then let's pick
1:21:07 music oh nice okay
1:21:12 [Music]
1:21:24 [Music]
1:21:29 I think that's kind of
1:21:30 [Music]
1:21:35 cool I like that but it's not very
1:21:38 retinol skin
1:21:45 care this is good this is soothing right
1:21:47 is retinol soothing or does it just
1:21:51 blast your skin off
1:21:57 [Music]
1:22:05 I think we're going with that one all
1:22:08 right okay and then you can preview it
1:22:11 here and our lips aren't going to be
1:22:12 synced but we can sort of see what what
1:22:14 they cut together hey there are you
1:22:17 ready to elevate your skincare game with
1:22:18 the double sized Retinol Serum for face
1:22:21 by insta skincare this powerful formula
1:22:23 is packed with hyaluronic acid vitamin E
1:22:26 and A Plus aloe vera for ultimate skin
1:22:29 Health all right fine whatever all right
1:22:32 so we're going to submit
1:22:35 it avatars of real confused or or real
1:22:41 so your your ads being generated so so
1:22:44 the way hen hen does these things called
1:22:47 video avatars so I've got one of me I
1:22:50 can show you one of me actually while
1:22:51 while that's being generated we'll come
1:22:53 back to that
1:22:55 um let me go to my
1:22:58 home where's my stuff recent
1:23:08 Creations AI Studio personal
1:23:14 items so
1:23:20 here here's a video
1:23:26 I did of
1:23:30 me so the way you do this is you upload
1:23:35 I think it's got to be at least 3
1:23:36 minutes between 3 and 5 minutes and so
1:23:38 you upload a video of yourself talking
1:23:40 and you can use your hands and right and
1:23:42 so this was in my office right there's
1:23:43 my little Story vine thing and my guitar
1:23:45 and and the Rocky Mountains are back
1:23:47 here like there you can see a little
1:23:49 that see the little thing right next to
1:23:50 my head that's a Rocky Mountain right
1:23:53 there see that little nuble that's a
1:23:56 Rocky Mountain so I uploaded that and
1:23:59 now that it's got that it's got my voice
1:24:02 it's got my head movements it's got my
1:24:04 facial expressions it's got my hand
1:24:06 movements and then you basically just
1:24:08 type in a script and say generate and so
1:24:10 this is what it
1:24:12 generated as we stand on the precipice
1:24:14 of a new era with generative AI we are
1:24:17 not just looking at the evolution of
1:24:19 Technology but a transformation in how
1:24:22 we approach creativity problem solving
1:24:25 and Innovation across all sectors the
1:24:28 power of generative AI lies in its
1:24:31 ability to amplify our human potential
1:24:34 enabling us to tackle complex challenges
1:24:37 with unprecedented speed and creativity
1:24:40 so crappy script and and you know it's
1:24:43 very flat right it's very non-emotional
1:24:46 but if you didn't know me I might talk
1:24:49 like that right it's you know it's it's
1:24:53 just this side of believable
1:24:56 and these guys hey Jen in particular has
1:24:59 been launching product after product
1:25:02 after product this year and so I have a
1:25:06 feeling by this time next year they're
1:25:07 going to have some crazy
1:25:09 [ __ ] um they did video translations just
1:25:12 as kind of a product afterthought and
1:25:15 and people kind of lost their minds on
1:25:17 it in fact I've got let's see
1:25:25 [Music]
1:25:27 here's me speaking I think this is me
1:25:28 speaking
1:25:31 Mandarin is
1:25:34 it hey it's Kyle Shannon with the AI
1:25:36 learning lab
1:25:42 [Music]
1:25:51 no what's this one
1:25:56 hello I'm Kyle Shannon the co-founder
1:25:58 and CEO of Story vine hello I'm Kyle
1:26:02 Shannon digital
1:26:05 twin it's just dumb it's just dumb I
1:26:08 tell you it just dumb all right let's go
1:26:11 see URL to ads let's see if we got one
1:26:14 all right you ready you ready for this
1:26:20 people so we didn't do any editing it
1:26:23 just went out and you know wrote the
1:26:25 script and and you could fix this all up
1:26:28 if you wanted to but hey there are you
1:26:30 ready to elevate your skincare game with
1:26:32 the double sized Retinol Serum for face
1:26:34 by insta skincare this powerful formula
1:26:37 is packed with hyaluronic acid vitamin E
1:26:39 and A Plus aloe vera for ultimate skin
1:26:42 Health say goodbye to dark spots and
1:26:45 hello to a smoother healthier complexion
1:26:47 experience the magic of retinol as it
1:26:49 visibly minimizes wrinkles and Fine
1:26:51 Lines slowing down the aging process and
1:26:55 for those pesky breakouts our serum
1:26:57 helps reduce acne and pores with natural
1:27:00 ingredients like chamomile olive leaf
1:27:03 and green tea extract it's suitable for
1:27:06 even the most sensitive skin just
1:27:08 remember to use it at night on clean dry
1:27:10 skin and follow up with sunscreen during
1:27:11 the day get your glow on and grab a
1:27:14 bottle of double-sized Retinol Serum
1:27:16 today your skin will thank
1:27:21 you did you see her moving her hands and
1:27:24 it went on under her face that's one of
1:27:26 the one of the Tells that this is so
1:27:30 they they resync the lips and how they
1:27:32 do it is they sort of Chop Chop that
1:27:33 part of the face out and redo it so she
1:27:36 puts her hands up there and it's it's
1:27:38 under her
1:27:39 [Laughter]
1:27:43 nose
1:27:45 but here's the
1:27:48 deal laugh
1:27:50 now Gary ve if you know Gary vaynerchuck
1:27:53 um
1:27:55 I saw a video from him I don't know two
1:27:57 or three days ago he's basically like
1:27:59 you know all these influencers think
1:28:01 their [ __ ] doesn't stink and and just
1:28:03 like influencers [ __ ] up the
1:28:06 traditional you know marketing World um
1:28:10 these virtual influencers are going to
1:28:12 [ __ ] up the influencer world and he
1:28:14 basically said you want to position
1:28:17 yourself for for a win start a virtual
1:28:20 get good at this
1:28:22 [ __ ] because this stuff is going to get
1:28:25 so good you're going to be able to have
1:28:28 a thousand of these things out there now
1:28:32 here's the problem everyone's going to
1:28:35 be able to have a thousand of them out
1:28:37 there which ones rise to the
1:28:41 top which
1:28:43 ones the ones that have smart talented
1:28:48 driven people with good judgment and
1:28:51 good taste behind them so just like we
1:28:54 we took two hours to make a udio song
1:28:58 that we could have made in 30
1:29:00 seconds why' we take two hours cuz we
1:29:03 wanted it to be [ __ ]
1:29:05 listenable same with these you could
1:29:08 figure out what are the what are the
1:29:10 limitations of this right now you could
1:29:11 figure out which of those Avatar people
1:29:15 sucks the less the
1:29:17 least you could figure out which
1:29:19 template you know does
1:29:21 good you could really customize your
1:29:24 your video so instead of uploading
1:29:26 horizontal video You're uploading
1:29:27 vertical video so you don't have those
1:29:29 white bars right you could really get
1:29:32 good at this and you could make these
1:29:34 things so you could knock them out like
1:29:36 that and they wouldn't
1:29:37 suck and it's just going to be like the
1:29:41 here's what AI is not going to
1:29:44 change the people that do the best
1:29:49 work are still going to do the best work
1:29:56 the people that phone it
1:29:59 in their shit's going to
1:30:03 suck and there will
1:30:06 be you know this amount of AI generated
1:30:09 stuff is [ __ ] mind-blowingly
1:30:12 good this amount is
1:30:15 decent this
1:30:18 amount and
1:30:21 then 60% of it's going to be trash no
1:30:24 70% of it's going to be
1:30:26 trash be 30% that'll be some version of
1:30:30 acceptable and there's going to be 2%
1:30:32 that's [ __ ]
1:30:35 mindblowing Kelly Bosch the video she's
1:30:38 putting out right now and I don't know
1:30:40 if you've been following her if not you
1:30:42 should
1:30:44 be
1:30:45 she's finding all sorts of new you can
1:30:49 watch her you can watch her
1:30:52 mind expanding she she's finding
1:30:55 multiple Lanes of narrative to to run
1:30:58 down with this AI stuff when Sora comes
1:31:01 out and she gets her hands on Sora she's
1:31:03 going to do some mind-blowing [ __ ] mark
1:31:05 my
1:31:07 words and everyone else is going to go
1:31:09 how'd she do that I I went you know what
1:31:12 I did I went to Sora and I typed in you
1:31:15 know make me a make me a video of a lady
1:31:17 walking down a street and you know what
1:31:19 it looked like [ __ ] Sora is
1:31:23 stupid you know you know I you know what
1:31:26 I'd rather do I'd rather just go get a
1:31:29 camera you know like a go to like a
1:31:31 thrift store and get like a VHS camera
1:31:34 and I would just film it on VHS and then
1:31:37 that CU it's a real
1:31:44 camera don't be that
1:31:48 person be like
1:31:50 Kelly do good
1:31:52 work critical
1:31:54 thinking judgment
1:32:00 taste
1:32:02 right those are going to be the critical
1:32:04 skills all the executional skills how to
1:32:07 do this how to do
1:32:10 that just learn what you can now it it
1:32:14 seriously if you're feeling like oh my
1:32:16 God K I'm so far behind I don't how to
1:32:18 do
1:32:20 anything it doesn't matter just start
1:32:23 paying attention start paying attention
1:32:26 to what you like if you're like hey how
1:32:29 did they make that that weird Mary song
1:32:32 go watch the video I just posted it you
1:32:34 can watch it's a 2hour process just sit
1:32:36 there put it on one and a half
1:32:38 speed it'll take you whatever one and a
1:32:41 halftimes 2 hours
1:32:43 is
1:32:46 P 90
1:32:48 minutes
1:32:50 right um
1:32:56 just start getting curious start
1:32:58 figuring out how to be a curator start
1:33:01 figuring out how
1:33:03 to think
1:33:09 critically understand that
1:33:14 language is
1:33:17 now how we're going to control almost
1:33:20 everything
1:33:25 get more gter with
1:33:30 words get more gter with
1:33:37 storytelling I I don't think he realizes
1:33:40 this but more gooder is not
1:33:41 grammatically
1:33:43 correct
1:33:47 awesome go use your grammar
1:33:50 skills I miss Kyle's AI sermon during
1:33:53 the weekend I know yeah it's true it
1:33:56 it's true normally this is like a Sunday
1:33:58 kind of thing and and because we weren't
1:34:00 here on Sunday I'm kind of doing my
1:34:01 sermon on
1:34:06 Monday oh I'm sure it's not
1:34:10 dandruff it's probably not animals on my
1:34:14 head I'm definitely probably sure
1:34:21 [Music]
1:34:24 it's a long weekend Dr J 1.5 goes faster
1:34:28 takes less time were you joking it's in
1:34:32 Canada I'm going to show you one other
1:34:33 thing that I think is
1:34:37 cool your next critical skill Kyle might
1:34:40 becoming an
1:34:44 electrician there's some there's some
1:34:46 there's some real [ __ ] Badness there
1:34:49 some bad mojo in the wires actually the
1:34:52 guy we bought this house from
1:34:55 we brought an electrician to rewire some
1:34:57 [ __ ] and the the work that the previous
1:35:01 guy had done was so bad that our
1:35:04 electrician was taking pictures of stuff
1:35:06 so he could show his buddies he goes
1:35:08 I've never seen anything like
1:35:11 this so
1:35:14 oh Joe fix
1:35:17 it just run down to the Home Depot and
1:35:20 get myself some electrical tape I can
1:35:23 figure it out just flip a breaker we'll
1:35:25 put the wires in here let's just we'll
1:35:27 screw it all together and see if it did
1:35:30 did the light come on
1:35:34 no what's up guys let three as usual but
1:35:37 always on time oh late as usual L
1:35:43 three I'm very
1:35:45 clever ha my name's K I'm
1:35:49 clever um I want to show you one other
1:35:52 thing cuz
1:35:54 while we're talking about critical
1:35:55 thinking this is an interesting one so
1:35:58 okay okay let's
1:36:03 see so if you saw me dick around with
1:36:05 LTX Studio LTX studi is a thing where
1:36:08 you just put in a single sentence and
1:36:10 then it writes you a whole script for a
1:36:12 movie it plans out the shots it writes
1:36:16 the scenes writes the
1:36:19 dialogue plans out the shots renders the
1:36:22 images for each shot
1:36:25 using all sorts of parameters like
1:36:27 close-up wide shot things like that puts
1:36:29 them all
1:36:30 together
1:36:32 um and like even a 30 second it does it
1:36:35 in like again everything's 30 second
1:36:36 chunks it makes like a 30- second video
1:36:39 but it's got like five scenes three
1:36:41 shots
1:36:43 each you could spend days refining that
1:36:48 script and like adding shots to it and
1:36:50 timing the shots to the script and
1:36:52 rewriting the script or you can just put
1:36:54 in make me a video of a dog that goes
1:36:57 through a car wash and gets its fur all
1:36:59 soapy and it'll make
1:37:02 it or you could make it clever and good
1:37:06 right so there's going to be lots and
1:37:09 lots and lots and lots and lots and lots
1:37:13 of tools where
1:37:15 it it doesn't require doesn't
1:37:19 require any Talent from you or anyone
1:37:23 else
1:37:29 but creative people and people with good
1:37:32 work ethics and people with critical
1:37:34 thinking skills and people with good
1:37:36 judgment and people with taste and
1:37:39 people with something to say and a point
1:37:41 of
1:37:42 view are going to use those same exact
1:37:45 tools and they're going to do remarkable
1:37:48 work and you're going to Marvel at
1:37:51 it you won't know how they got there and
1:37:55 how they got there was hard
1:37:57 work so so learn the tools well enough
1:38:02 to understand what they make
1:38:06 possible and then and the lights came on
1:38:09 and then they went off then they came
1:38:14 on so I'm going to show you a tool right
1:38:17 now that is a really good example
1:38:22 of you can choose to use
1:38:31 this in a passive way here so decco
1:38:36 here.
1:38:39 a and I'm on the explore tab you can go
1:38:42 to the create Tab and you can type a
1:38:44 prompt but but watch watch what this
1:38:47 does so I'm on the explore tab so I'm
1:38:48 going to grab this thing's kind of
1:38:50 funny so here's kind of an interesting
1:38:53 poke pish Mario
1:38:56 is cartoony Illustrated it says an
1:38:59 unbeatable video
1:39:01 game so I can either click create my own
1:39:04 which is going to use this as an
1:39:06 inspiration and I can go modify the
1:39:08 prompt or I can uh upscale it or I can
1:39:11 add it to my library or this little icon
1:39:15 this little icon says create
1:39:18 station and so what a station is
1:39:21 is a NeverEnding
1:39:24 Scroll of original
1:39:29 artwork that's being generated one image
1:39:32 per
1:39:33 second essentially for
1:39:38 Infinity so how do you use a tool like
1:39:42 this well I don't know for one thing you
1:39:44 don't need to prompt it you can you can
1:39:48 go create a prompted image and you can
1:39:49 turn that into a station like these but
1:39:52 like okay that little thing right there
1:39:54 I love that thing I don't know what the
1:39:56 prompt is behind that I don't care I
1:39:58 just like it so I'm going to say create
1:40:00 a station out of
1:40:04 that and now you could treat this really
1:40:07 passively like entertainment or you
1:40:10 could say hey that gives me an idea for
1:40:13 a story to
1:40:18 tell and then you could start to think
1:40:20 about well if I wanted to tell that
1:40:21 story what would it be would I tell the
1:40:23 story
1:40:25 of the device would I tell the story of
1:40:26 the person that invented it would I tell
1:40:28 the story of a kid who found that object
1:40:32 and learned how to use it yeah deoh
1:40:35 here. a is this
1:40:38 tool it took yeah oh Allan so Allan I
1:40:42 showed that that uh weird marry trailer
1:40:44 before that Allan did took him five
1:40:47 hours from start to finish to do that
1:40:50 right so again I think the the
1:40:54 the misconception about AI is push a
1:40:57 button get crappy artwork out of
1:41:01 it well this is one where you don't even
1:41:03 have to push the button you can just sit
1:41:05 back and this thing's generating art and
1:41:08 some of it's good some of it's crap but
1:41:11 eventually it's going to stumble upon
1:41:13 something here that's interesting like
1:41:14 that's kind of
1:41:16 interesting kind of looks like some 60s
1:41:20 era
1:41:23 oh good
1:41:28 lord what's that on this is uh this is
1:41:32 uh Deco here.
1:41:36 a and these are called stations this is
1:41:38 brand new this is 3 days
1:41:42 old now
1:41:44 imagine this
1:41:46 tool for every kind of media type
1:41:50 imagine a radio station that just plays
1:41:52 music that is is the music you want to
1:41:54 hear and it's all
1:41:57 original and you could start with weird
1:42:00 Mary from Cedar Hill as your starting
1:42:02 point and it would just generate a bunch
1:42:04 of music like that and then when you
1:42:05 hear something you like you go oh let me
1:42:07 hear that and then you're like oh I love
1:42:09 that let me hear more like that save
1:42:12 that to my
1:42:13 library and imagine that
1:42:16 for any kind of writing any kind of
1:42:19 problem solving any kind of programming
1:42:22 any kind of animation any kind of 3D
1:42:25 object any kind of 3D World any kind of
1:42:28 gameplay
1:42:32 anything imagine a system like this that
1:42:36 when you click on this instead of it
1:42:38 being an image this is a
1:42:41 whole um sort of Consulting analysis of
1:42:45 an
1:42:45 industry right so instead of the outputs
1:42:48 being images it's it's research and
1:42:53 Analysis of that
1:42:55 research and you can just have instant
1:42:57 access to that stuff and it's just it's
1:42:59 just rolling by what do you do with
1:43:05 that but that's the world we're entering
1:43:09 right we're not there yet but if you
1:43:11 look at this and if you understand a
1:43:13 little bit about how AI
1:43:16 Works what you learn is
1:43:19 that the reason this makes images really
1:43:22 good and udio makes songs really good
1:43:25 and chat gbt makes words really good is
1:43:28 that they just train them on different
1:43:31 data this tool train things on images
1:43:35 and so it does images really good udio
1:43:37 train things on music so it does music
1:43:40 really
1:43:43 good and so any kind of knowledge work
1:43:47 is going to be at our
1:43:49 disposal over the next 3 years in in
1:43:53 rally powerful
1:43:55 ways so the skill that's going to be
1:43:58 asked of you is not do you have the
1:44:02 skill it's going to be do you have the
1:44:04 aptitude do you have the Curiosity do
1:44:06 you have the Judgment do you have the
1:44:08 taste do you have the drive do you have
1:44:12 the
1:44:13 interest do you have the
1:44:17 creativity to connect the dots between
1:44:19 these crazy images and some analysis of
1:44:22 a business
1:44:24 well what would be the connection there
1:44:26 Kyle I don't know but I'll tell you what
1:44:29 there's going to be some people that get
1:44:30 really [ __ ] good at
1:44:32 that and they're going to seem like
1:44:35 magicians because that's not how most
1:44:38 people are trained right now most people
1:44:40 are trained on a single
1:44:42 specialty most people are going to
1:44:44 school for a single
1:44:46 specialty I like to call this era
1:44:48 revenge of the liberal arts majors
1:44:57 the neuro Divergent will run and rule
1:44:59 the world using AI I think so this app
1:45:02 is decco here. these are called
1:45:06 stations so how you get to them is you
1:45:08 go to decco here you click on the
1:45:10 explore
1:45:11 Tab and it'll just scroll by and then
1:45:13 you click on any image you like and just
1:45:17 say create a station out of
1:45:20 that and what's cool is you can share
1:45:22 these stations I could share this
1:45:24 station with you and we would all be
1:45:26 looking at exactly the same
1:45:29 images so it's like this car that it
1:45:31 just generated you would have access to
1:45:33 that too if you had access to this
1:45:39 station and you're like well why did it
1:45:41 just do three cars and then a woman and
1:45:43 then robots I don't know does it matter
1:45:46 again this this kind of goes back to the
1:45:47 conversation we had at the beginning
1:45:49 about if you listen to a song and it
1:45:52 moves you does it matter how it was
1:45:55 created did not see the liberal arts
1:45:58 inheriting the earth I know oh
1:46:02 wow the the electricity she is I think I
1:46:07 think I might need to I think I might
1:46:10 need to call it a night here good Lord I
1:46:13 know archetyp archetypal architect it's
1:46:16 crazy isn't it well it's it's funny I
1:46:19 you know I think you know did not see
1:46:21 liberal arts taking taking over the
1:46:22 world but also I think I think creative
1:46:26 and knowledge workers didn't
1:46:29 see computer
1:46:32 automation um I I think I think the
1:46:36 general consensus was that knowledge
1:46:39 workers and creatives were the were the
1:46:42 last to be
1:46:43 affected and in a weird way they're kind
1:46:45 of the first to be
1:46:47 affected because you know some of the
1:46:49 factory automation did did cost
1:46:51 manufacturing jobs but but it's coming
1:46:55 so directly for knowledge and creative
1:46:58 work um where it's all about the work
1:47:03 you're doing on computers um and it's
1:47:06 going to it's going to alter everything
1:47:08 and then who gets good who gets good who
1:47:12 gets the work who gets the work are
1:47:15 people who are trusted who have good
1:47:18 work ethics who are creative who can get
1:47:20 [ __ ]
1:47:21 done that's who it's work not who's got
1:47:24 the best degree not who spent the most
1:47:27 time learning
1:47:29 JavaScript it's going to be who knows
1:47:31 how to wield these tools and get
1:47:33 interesting results out of them look at
1:47:35 that that's [ __ ] awesome create a
1:47:37 station out of
1:47:38 that crazy right AI now nerds can now do
1:47:43 dis tracks with music videos in eight
1:47:46 hours no studio time and editor yeah I
1:47:49 know it's crazy Chris and it's you know
1:47:51 Tyler Perry canceled an 800
1:47:53 million Studio
1:47:56 expansion after seeing Sora Sora video
1:47:59 $800 million Studio expansion why
1:48:02 because he's like we're not going to
1:48:04 need Studios that big we're not going to
1:48:06 need to shoot all that we're not going
1:48:07 to need need to build sound
1:48:12 stages what is weird your lights flicker
1:48:14 and we get Thunder
1:48:16 here that's
1:48:19 wild who would have thought AI opened up
1:48:21 real meritocracy I
1:48:25 know I know exactly it's it's it's
1:48:30 weird it's like you know what's going to
1:48:33 happen Chris is the centers of
1:48:36 Power are going to
1:48:39 shift and holy [ __ ] the like the death
1:48:42 throws of the centers of Power are going
1:48:44 to be it's like grab your popcorn and
1:48:48 watch Industries try to legislate their
1:48:51 way out of this LOB their way out of
1:48:54 this um sue the living pants off of
1:48:58 people who are doing I I would not be
1:49:01 surprised if big agency holding
1:49:04 companies start suing smaller agencies
1:49:07 for being able to deliver better work
1:49:10 faster than they can Sora can now do
1:49:14 b-rolls so no need to be going to
1:49:16 location as such yeah I was you know
1:49:19 it's funny the the day after Sora came
1:49:22 out I was watching a show on
1:49:24 TV and it was it was one of those you
1:49:27 know those sort of um dramas like the
1:49:30 Americans
1:49:31 but the there's very few of them that
1:49:33 are as good as the Americans so you see
1:49:36 like 10 of them for everyone that's
1:49:37 decent so I was watching one of those I
1:49:39 forget what it was and so they had to
1:49:42 run out of whatever building they were
1:49:43 in and run out under the dusty desert
1:49:46 tarmac and they piled into three Russian
1:49:50 helicopters and then they did a wide
1:49:53 shot of the airport tarmac and the three
1:49:56 Russian helicopters take off and the
1:49:58 dust flies and then there's like eight
1:50:00 army vehicles that follow them right
1:50:02 they all take off to go wherever the
1:50:04 [ __ ] they're
1:50:05 going and I looked at that shot and I'm
1:50:08 like okay that shot was 15 seconds on
1:50:12 film it probably took two or three
1:50:15 months to plan you know they probably
1:50:18 spent I don't know 50 $70,000 just in
1:50:22 insurance
1:50:23 probably another $100,000 in rentals
1:50:27 right renting trucks cars the airport I
1:50:30 love the American Seasons yeah the
1:50:32 Americans that was a great great great
1:50:35 show and I and I I just watched the hel
1:50:38 as the helicopters were taking off I'm
1:50:40 like holy [ __ ] you you will never need
1:50:44 to do that kind of logistical project
1:50:48 management again what is Sora so Sora
1:50:54 is an announced um but not
1:50:58 released uh video model from open aai
1:51:02 let me go
1:51:04 to oh you know what I can go if you go
1:51:06 to
1:51:08 open.com SL
1:51:11 Sora it will take you here
1:51:23 and it makes like it it makes it can
1:51:26 make 60-second Clips these are these are
1:51:28 loading really slow oh
1:51:32 boy are you all still there AI will do
1:51:36 drone with drones versus helicopters and
1:51:38 Arial shots yeah
1:51:40 exactly exactly all right I think I
1:51:43 think I think we're we're about I think
1:51:47 the the electricity she is about ready
1:51:50 to go
1:51:55 but anyway like look at this video right
1:51:58 look at this video of this you know
1:52:00 truck now is it perfect no it's got
1:52:03 weirdness to
1:52:07 it but it's pretty [ __ ] coherent
1:52:11 right and again this is the worst it
1:52:13 will ever be like look at the look at
1:52:16 the coherence of the paintings on the
1:52:18 walls
1:52:21 here how's it doing that
1:52:23 no other video thing does that all the
1:52:25 other video tools do all this
1:52:27 warping the the way this does what it
1:52:29 does
1:52:31 is this isn't so much a video tool as it
1:52:35 is a world simulation model and they're
1:52:38 doing all sorts of crazy math where
1:52:40 they're breaking up a flat image into
1:52:42 these things called what are they called
1:52:46 panels forget what they're
1:52:50 called Kyle's disturbing The Matrix with
1:52:53 these
1:52:54 Revelations but they basically break the
1:52:57 they break the image into a grid and
1:53:00 then they take each little part of the
1:53:01 image and they understand what it is and
1:53:04 what its relationship to the rest of the
1:53:06 world is and then they look at that over
1:53:08 like three seconds like 100
1:53:10 frames and then they make sure it's
1:53:12 coherent across those 100
1:53:14 frames and so so this system actually
1:53:19 understands like the physics of this
1:53:24 room which is [ __ ]
1:53:26 insane it's it's not making like the way
1:53:30 we make 3D worlds right now if you look
1:53:32 at Unreal Engine is you build 3D worlds
1:53:34 and you build textures and you build
1:53:37 lighting models you build all this
1:53:39 rendering you know all of these
1:53:42 geometries that have to be rendered this
1:53:46 understands all of that but it doesn't
1:53:48 have to build everything it just goes oh
1:53:50 we're in a we're in a museum here and
1:53:52 these are paintings and the camera's
1:53:55 moving and this is going here well I I
1:53:57 understand that this is a painting and
1:53:59 it should not change in shape as the
1:54:02 camera
1:54:04 moves it's not vectors it's no they're
1:54:07 not called pixels they're called here
1:54:08 it's it's on this page let me let me go
1:54:10 down
1:54:13 and research
1:54:20 techniques there's a there's a paper
1:54:24 let's see
1:54:27 research Sora
1:54:30 Sora s
1:54:33 s oh here here
1:54:36 the let's
1:54:41 see the first paragraph we're teaching
1:54:45 AI to understand and simulate the
1:54:47 physical world in motion with the goal
1:54:50 of training models that help people
1:54:52 solve problems that require real world
1:54:55 interaction so
1:54:57 think figure one
1:55:00 robots as well you got a system that
1:55:04 understands the world and can understand
1:55:06 it
1:55:07 visually and can understand real world
1:55:10 interactions this this isn't just a
1:55:12 video tool introducing Sora our text to
1:55:14 video
1:55:15 model um here I'll show you how it works
1:55:19 here let me go to the research
1:55:28 they talk about some of the problems it
1:55:30 has which is nice that they did
1:55:32 that research
1:55:37 techniques are technical report
1:55:47 okay
1:55:48 here what are they called patches
1:55:52 patches is okay so if you understand how
1:55:56 video
1:55:58 works video works by having a frame with
1:56:03 visual information on it and then you
1:56:04 have a bunch of those in a row right
1:56:06 that you know 24 frames a second 30
1:56:08 frames a second 60 frames a second and
1:56:11 you go out in the real world and you
1:56:12 take a camera and it it you know has all
1:56:14 these pixels on it and so what they're
1:56:18 doing with their little
1:56:20 patches is they're taking each image and
1:56:24 they're doing visual
1:56:26 encoding so that they understand what
1:56:29 this is they understand what this image
1:56:31 is they understand this is a coral reef
1:56:33 they understand this is a butterfly they
1:56:36 understand that butterflies you know are
1:56:38 animals and lighter and softer than
1:56:41 coral and you know technically they
1:56:44 shouldn't be interacting but it
1:56:45 understands how this works and it
1:56:47 understands how this works so it can put
1:56:48 put them together then what it does is
1:56:51 it breaks that single image into patches
1:56:54 so into small little discrete areas a
1:56:57 grid right and so it understands that
1:57:01 this little patch of the butterfly wing
1:57:05 right is a butterfly wing and if it's
1:57:07 moving a little bit it understands that
1:57:10 over
1:57:13 time right so you take one of those
1:57:16 patches and if that little patch is that
1:57:18 little front of the butterfly wing it
1:57:20 understands that over time it's got keep
1:57:23 track of that and keep it co coherent
1:57:27 so so it's understanding a complex
1:57:30 Matrix of interactions of objects in a
1:57:33 single frame and then it's doing it like
1:57:35 over a 100 frames it's [ __ ]
1:57:37 insane [ __ ]
1:57:40 insane that's what Sora is everyone
1:57:42 thinks it's a video tool it's not a
1:57:44 video tool I mean it is it's going to
1:57:47 give us lots of cool video now here's
1:57:49 the deal you're you're like that seems
1:57:51 complicated yeah I think to do a minute
1:57:53 of video at something like 20 or 30
1:57:55 minutes or an hour like it's it's not
1:57:59 fast but it's understanding and
1:58:02 simulating reality when we look back on
1:58:05 these lives the equivalent of us raving
1:58:08 about MS DOS
1:58:10 exactly this is going to do you do you
1:58:13 know that like right now whatever this
1:58:15 resolution is let's say it's 20 across
1:58:17 and 10 down right like two years from
1:58:20 now it's you know it's going to be down
1:58:22 to the pixel level where in real time
1:58:24 it's going to be understanding every
1:58:26 pixel and every pixel's relationship to
1:58:29 the physics of the world it's
1:58:31 representing and it's going to be able
1:58:32 to do that over an hour
1:58:36 instantly you like you'll I I mean we
1:58:40 are so close to being able to literally
1:58:44 just describe a movie and have it start
1:58:46 playing and then go no make it a girl
1:58:49 make it a cartoon um make it a comedy
1:58:52 I'm board speed it up slow it down ooh I
1:58:56 want to know more about that character
1:58:58 run down that rabbit hole and just it'll
1:59:00 just start doing it we're not that far
1:59:04 from that
1:59:07 world three years five years
1:59:11 crazy all
1:59:13 right I'm totally distracted at this
1:59:16 point yeah 56k is the fastest internet
1:59:19 You've got mail exactly
1:59:23 what what was it Bill Gates who would
1:59:25 ever need more than
1:59:26 128k of Ram or something like that
1:59:28 whatever it
1:59:32 was cuz we can't we can't see what we
1:59:35 don't
1:59:37 know why a lot of the AI stuff is boring
1:59:40 right now Rudy was asking
1:59:44 this why a lot of the AI stuff is boring
1:59:47 right now is
1:59:50 that we're using AI
1:59:54 to automate and speed up the way we used
1:59:57 to do things not the way we're going to
2:00:00 do
2:00:01 things one of the things I want to
2:00:04 encourage in here is that you start
2:00:07 thinking not about how you do what you
2:00:10 do
2:00:12 now but
2:00:13 imagine if these tools got just a little
2:00:16 bit less
2:00:18 shitty what would be possible
2:00:23 if you can start putting yourself in
2:00:25 that space you're going you're going to
2:00:27 do good you're going to do good little
2:00:31 little Lassie little
2:00:33 Mister all right I'm going to get my ass
2:00:36 out of
2:00:37 here that was a bit chaotic tonight but
2:00:40 you
2:00:41 know what do you
2:00:43 want what do you want I was away for two
2:00:46 days oh tomorrow night AI Salon meeting
2:00:49 if you have not joined the AI Salon you
2:00:52 need to get get the ass to a meeting and
2:00:55 we got the next meeting tomorrow night
2:00:58 uh
2:01:00 um she's going to be speaking on some
2:01:03 research she's been working on I think
2:01:04 she's a PhD student she's at least in
2:01:06 college I think she's PhD I don't
2:01:09 know so that's going to be really cool
2:01:11 the meetings are great they're they're
2:01:13 just good people we'll talk about news
2:01:16 um we'll get to know each other you'll
2:01:17 learn more about the salon so go
2:01:20 there click on the third link down it
2:01:22 says AI Salon community and then in the
2:01:25 meetings there's a meetings area you'll
2:01:28 see the calendar it's tomorrow night at
2:01:30 5:00 p.m. mountain 700 p.m. Eastern
2:01:33 that's
2:01:34 correct and whatever that is Pacific
2:01:39 four so 5 to S my time here in
2:01:46 Denver Busters in advertising I don't
2:01:49 think you meant to say Busters did you
2:01:51 but I would love to learn about Busters
2:01:53 and
2:01:55 advertising K klie Jameson Stewart you
2:01:58 are deep thinking tonight Kyle I like it
2:02:01 well thank
2:02:02 you I I think I'm I think some just call
2:02:05 it
2:02:10 bloviating but well I'll tell you what
2:02:13 you know what there there was there was
2:02:15 something nice about um having the space
2:02:18 this
2:02:19 weekend to
2:02:22 not have to think about the nightly
2:02:24 thing and just kind of let my brain go
2:02:27 cuz there's we we are in a really
2:02:30 interesting lull right now as as insane
2:02:33 as all this [ __ ]
2:02:34 feels there's there's kind of a dip in
2:02:37 in energy right now with AI everyone's
2:02:41 kind of waiting for what open AI is
2:02:42 going to do next and and everything's
2:02:45 kind of caught up to chat G bt4 there's
2:02:47 so many tools to use that I think you
2:02:49 know people are kind of settling in and
2:02:50 they're picking their tools to use but
2:02:52 we're also kind of waiting for um for
2:02:55 what's coming bloviating another new
2:02:57 vocabulary
2:03:02 word oh yeah all right we'll try to keep
2:03:05 the words
2:03:07 coming Masters that's right she's going
2:03:09 for her
2:03:10 Masters be careful the funny farm is
2:03:13 waiting I think they call it the nervous
2:03:15 hospital um can we promote the local AI
2:03:17 salons somewhere in Mighty networks yeah
2:03:21 we should do that Brandon oh TK is doing
2:03:23 asking that um yeah we can we can do
2:03:25 that TK uh we that's something we should
2:03:27 talk about and figure that out um okay
2:03:31 cool um Gavin I've settled on idag as my
2:03:35 primary AI image generator very cool
2:03:37 steo peace out worked on the Broadway
2:03:40 Tunes not I not yet I worked on the
2:03:44 Andrew and I met my my co-writer on the
2:03:47 Broadway thing met um Sunday and spent 2
2:03:51 hours on
2:03:52 and it was mostly on act two
2:03:55 outline like we're what we're really
2:03:57 trying to do is make sure the story is
2:03:59 [ __ ] bulletproof and and really good
2:04:03 like we've got to get the story right it
2:04:04 can't be cliche it can't be expected it
2:04:08 can't be just trite and so we're working
2:04:11 really hard to do
2:04:16 that all right
2:04:18 um okay so if you want to support
2:04:25 this me this
2:04:28 channel um follow me follow the
2:04:32 channel um Tik Tok for whatever reason
2:04:36 doesn't like us anymore and so we're not
2:04:39 getting the visibility we used to so
2:04:42 just do me a favor like it follow it
2:04:45 share these lives when you can um if you
2:04:47 want to subscribe that helps I think in
2:04:50 one of the corners I don't know if I it
2:04:52 I think in one of the corners there's a
2:04:54 video series there you can buy that that
2:04:56 that supports it as well it was an
2:04:57 electrifying evening it was it's still
2:05:00 it's still a little zappy um but anyway
2:05:03 thanks your regulars thank you
2:05:05 subscribers thank you mods um let's do
2:05:08 our little uh our Romper Room thing all
2:05:11 right Rachel Sparkles good to see you
2:05:13 Tobias as always thank you for hanging
2:05:15 out Amelio's wife Wolfman Clint hope
2:05:18 you're okay stay safe uh digital Gods uh
2:05:21 k j Stewart
2:05:24 peace uh Mr Studio dhj good to see you
2:05:28 Alan thank you thank you for very cool
2:05:30 uh trailer by the way um TK great to see
2:05:33 you
2:05:34 Danielle Cosmic lover Gavin biddy Vicky
2:05:38 you excited for
2:05:40 tomorrow by the way your daughter's got
2:05:42 a great name Paris great
2:05:44 name great name laughing Thunder great
2:05:47 to see you
2:05:53 Joker
2:05:54 Joker Joker just sitting there in the
2:05:57 nervous
2:06:04 hospital I think on that we'll end it on
2:06:07 that joker at his nervous Hospital peace
2:06:10 out everybody I will see you tomorrow
2:06:13 night I'll probably be late tomorrow
2:06:14 night because we have a salon but I'll
2:06:16 I'll make a video if I'm going to be
2:06:17 late peace bye