Apr 12, 2024 - AI-Driven Creativity: Crafting Songs and Stand-Up Comedy in Real-Time!

Description

Chapters

Transcript

0:00 you
0:02 [Music]
0:16 ready my baby at the
0:19 club Cafe on the
0:23 corner street late one wedding day she
0:27 offered me a red strap I said baby not a
0:30 chain
0:31 I could go for ginger how about we just
0:34 dance we did
0:36 boom did it l and the nitty-gritty we
0:39 did the hoo the lazy ballerina baby make
0:43 me feel like I'm a California dreamer
0:45 yeah
0:54 [Music]
1:00 [Music]
1:00 [Applause]
1:01 [Laughter]
1:05 [Music]
1:21 what what what what's the rapping why
1:23 are you
1:24 rapping we don't do rapping
1:33 [Music]
1:39 [Music]
1:42 no Brandon it's Friday night Friday
1:45 night you get your day off there buddy I
1:47 know you're excited for the weekend but
1:50 that's kind of
1:55 [Music]
1:58 ridiculous w
2:03 [Music]
2:08 [Music]
2:13 [Applause]
2:17 oh yeah the shirt's going to be a
2:19 [Music]
2:25 problem I don't mean to cause you any
2:28 trouble
2:29 [Applause]
2:32 didn't mean CA you a
2:35 [Music]
2:40 pain want time to see you
2:43 [Music]
2:45 laughing only want to see you laughing
2:48 in that purple rain Purple Rain Purple
2:56 Rain Purple Rain purple
3:03 rain Purple Rain Purple
3:08 [Music]
3:13 Rain only I want to
3:16 see in bu
3:21 ring oh boy happy Friday night
3:25 [Music]
3:29 everybody spent a spent a fair amount of
3:31 time today well not a fair amount about
3:33 an hour with yudo the songwriting thing
3:36 so I'll share share some of that the
3:39 results of
3:41 that it's uh a little better than
3:43 yesterday but it's still pretty janky
3:45 let me go change my shirt
3:49 [Music]
4:05 I pry my face from my wrinkled
4:10 sheet open it was all a
4:14 [Music]
4:16 dream pain in my head and the hole of my
4:21 heart says it was more than just Jim
4:24 Bean
4:26 [Music]
4:38 all right enough comedy jokes entertain
4:42 yourselves I will be
4:49 [Music]
4:51 back chiy
5:19 all righty there we
5:26 go anybody else want to crank the likes
5:29 for when Kyle gets back oh that's sweet
5:31 nice we're already up to 3,000 you guys
5:33 are swell happy Masters tiger made the
5:36 cut 24th cut in a row at the Masters
5:39 which is staggering I think that's a
5:43 record and it seemed like he wasn't
5:45 going to make it by a lot but then it
5:47 just was
5:50 ugly all right Vegas moxy's here so we
5:53 can start I'm at 310 per minute 310 oh
5:59 310 likes for a minute let's see that
6:02 shirt shirt makes it better to see the
6:04 words yeah the the my fashion Choice has
6:06 nothing to do with fashion well that
6:08 last shirt did but it's not good for
6:11 this stuff
6:13 so Amelio's wife is here and excited
6:16 about it good to see you
6:18 fantastic fantastic Pete's
6:22 here good evening Pate
6:28 Joker even the thought of the smell of
6:30 Jim Beam Mi makes me
6:35 nervous that's pretty good PR to good
6:39 good good good g g i was the first to
6:43 see the live I was the I was yep that's
6:46 a first
6:47 nice fantastic
6:51 nauseous the J Jim beam's making you
6:54 nauseous oh what are you going to do hey
6:55 look my ice maker just stopped started
6:57 working again for no reason
7:01 I think it must have been frozen and it
7:02 just it just defrosted itself or
7:07 something but ice is back everything's
7:11 good with the world um how is everybody
7:13 if welcome here if you're new my name's
7:15 Kyle Shannon this is the AI learning lab
7:17 Three Mile Storage the AI guy I
7:19 presented to today was floored with your
7:23 consistency and doing these lives well
7:25 thank you I appreciate that I I uh I you
7:29 know I'm doing what I can I'm I'm
7:31 fighting a good fight here Joey what's
7:33 happening good to see you Tams looks
7:36 like it makes good ice yeah it's pretty
7:38 solid ice it's not clear ice but it's
7:40 good ice it's it's here's what's nice
7:43 about it you don't have to make ice you
7:46 don't appreciate an ice maker until
7:49 until you don't have one and then you're
7:51 like making ice
7:53 sucks you can make money
7:57 with oh man
8:01 all right let's see
8:04 um couple of things going on what did
8:07 anything happened today I don't think
8:09 much happened
8:11 today the uh the rabbit CEO was just
8:14 giving a talk talking about the thing I
8:17 didn't catch it but um someone in here
8:20 told me about it uh retro retr Punk
8:24 maybe um my company held a neurod
8:26 Divergent awareness lunch today those
8:29 were our
8:35 people yeah the neuros spices us neuros
8:39 spices got to stick together I'm telling
8:40 you these there's something
8:43 about there's something about these AI
8:46 tools that really
8:48 does serve or not
8:52 serve a neurod Divergent brain I feel
8:55 like is um is well poised to take advant
9:00 AG of what these things make possible
9:02 can you talk about the news events from
9:04 AI this week yeah I can I can absolutely
9:06 do that
9:08 um there was a fair amount I've got him
9:11 from Lex City perplexity spilled the tea
9:14 that adobe trained on Mid journey to me
9:16 today yeah I know so so I've been saying
9:20 for
9:23 um I don't know a month two months now
9:28 no longer than that probably four months
9:31 that you know all the other image
9:33 generation tools you know have been
9:35 trained on copywritten works
9:39 and Adobe kind of held The High Ground
9:41 and said our stuff is only trained on um
9:44 public domain stuff and Adobe stock that
9:48 we own the licenses for
9:50 Well turns out Funny Thing Happened on
9:54 the on the way to toh where are they
9:57 based San Francisco I don't know where
9:58 the hell they're based
10:00 they uh they apparently trained a lot of
10:02 Firefly on Mid Journey now what's not
10:05 clear to me how will AI change real
10:07 estate mortgages uh I'll come back to
10:09 that one I I I don't
10:12 have I don't have a ton of insight there
10:14 I don't know that world super well I've
10:16 got I've got some clients in that space
10:18 But don't know it super well
10:21 um what's not clear to me is what they
10:24 did did they train it on Mid Journey
10:26 outputs so so in that case they they
10:29 could claim that they're one step
10:31 removed from training it on copywritten
10:35 works but if if Firefly got better
10:39 because they essentially trained it on
10:41 copywritten work well then everyone's
10:44 [ __ ] just doing it now like you know
10:46 so anyway
10:49 uh unfortunately not surprising but but
10:53 we'll see we'll see if it if anyone
10:55 ultimately gives a [ __ ] I have a feeling
10:57 they won't I I I mean I think the
10:59 artists going to give a [ __ ] there's
11:00 probably going to be class action
11:03 lawsuits but I think in the end the a
11:06 lot of these lawsuits are just going to
11:07 get chucked I really do cuz the the the
11:10 compelling case for for the good of
11:13 humanity is just it's too great it's too
11:16 great and there's there's just you know
11:19 piles of uh tools out there now that
11:22 have these remarkable powers that people
11:24 are going to not going to want to give
11:25 up so I don't know but we'll see we'll
11:28 see
11:30 I guess it's not stealing if it's stolen
11:32 first
11:35 exactly sunno gave me 500 extra credits
11:38 yeah I noticed sunno gave me a bunch of
11:40 extra credits as well I didn't I didn't
11:43 know if that was
11:47 um I didn't know if those puno credits
11:49 was because the month rolled over or or
11:52 if it was extra credits I don't know I
11:54 wish they'd stop pretending that's not
11:56 exactly how human brains are trained
11:58 well I agree with you but here's the de
12:00 here's the deal archetypal architect
12:04 is even though that's the way human
12:07 brains are
12:09 trained human brains require like four
12:11 years or eight years or 10 years or 20
12:14 years to assimilate all of that input
12:17 and then reconstitute it and put it out
12:20 as creative
12:21 work these machines are doing it you
12:24 know in months you know if not sooner
12:26 and then they're spitting it out in
12:28 seconds and so
12:30 and it's
12:31 also the fear of the robots are going to
12:35 kill us the AI is going to take my job
12:38 right all of that
12:39 fear um combined with you know just the
12:44 especially in the creative arts I mean I
12:47 I'll I'll I'll walk you through the song
12:49 that I was working on
12:51 tonight and you know
12:55 the like just just the the decimation
12:59 and the creative art short term um is
13:02 going to be pretty intense I think
13:04 longterm it's actually going to be
13:05 really good Kyle can I hop up real quick
13:08 it's about Adobe too long to type yeah
13:09 absolutely uh so yeah just do the do a
13:12 request invite or let me see if I can
13:14 invite you up
13:17 here
13:20 reply I could translate what you
13:26 said accept continue wait where are
13:31 you um can you hear me all yeah okay
13:35 great so so Dolby kind of had like a dog
13:38 and pony show and they had like client
13:41 meetings and like kind of show off their
13:43 stuff and it was about their AI stuff
13:45 and coke coke was their kind of their
13:49 guest right so they Coca-Cola Coca-Cola
13:53 yeah so they walk through some of the
13:54 tools that or some of the ways that
13:56 Coca-Cola is deploying their AI stuff
13:59 and people walked away from it going oh
14:01 my gosh I can't believe all the stuff
14:04 that I can do and it's kind of all the
14:06 stuff that we do here every night and
14:07 people are like oh my gosh Coke is so
14:10 Innovative and things so it was it was
14:14 um you know oh my gosh they broke down a
14:16 piece of content into three different
14:19 things of videos and translation and
14:21 social posts oh my gosh so
14:25 anyway people are like really wowed by
14:27 it but it's like nothing that was too
14:30 extraordinary that's it oh that's cool
14:33 that's cool so so was what what were
14:36 they show showing off with Coke that the
14:39 work that they were doing they showed
14:41 off the Christmas stuff or was it new
14:42 stuff they were showing off it was um it
14:44 was new things so I think it's kind of
14:46 the new AI U Rons that they put into
14:50 their their tools I'm not exactly sure
14:53 kind I see they were saying here's
14:55 here's how coch is using these tools I
14:57 got it I got it exactly exactly and
15:00 everyone was like oh my gosh I can't
15:02 believe this stuff and it's like
15:06 yeah you mean people at Adobe were
15:08 saying that no just people that um oh
15:12 that were at the event got down window
15:13 this this presentation like I can't
15:15 believe how Innovative the coke is and
15:17 using the stuff and they're so far ahead
15:19 and it's like yeah oh okay yeah well
15:23 actually I mean Coke you know Coke the
15:25 fact that coke allowed you to create
15:27 Coke branded images gener AI like that
15:30 was that's you know for for for a brand
15:34 that that's that uh valuable to to allow
15:37 you know normal civilians to to generate
15:41 images with Coke logos on it I thought I
15:43 thought it actually was pretty ballsy of
15:44 them so interesting yeah and I think
15:47 because Coke is kind of known for that I
15:49 think that will really I think it'll
15:51 really wake up companies to say
15:55 yeah the stuff is amazing so it's
15:58 probably going to be really a good thing
16:00 for for everyone yeah awes awesome
16:03 thanks thanks for sharing
16:08 cool all right very cool very very cool
16:11 um okay so so what other news went on
16:14 today let
16:16 me your regulars be like
16:21 a that's really
16:26 funny oh forever hooked beautiful
16:29 balloons love that that was cool in my
16:32 context I simply showed off chat GPT 3.5
16:35 in a church work shops and my folks were
16:38 like whoa exactly AI was a mistake it's
16:43 nothing but trash P Pate finally coming
16:46 in with with the uh the voice of reason
16:48 here letting us know there's that guy on
16:51 uh Pate have you seen that guy on
16:53 Twitter Gary Marcus that dude has never
16:56 said a positive thing about AI just like
17:00 see I told you it was a piece of [ __ ]
17:02 look at this horrible it hallucinates
17:05 it's never going to get better it's
17:07 awful it's just
17:08 awful just like dude calm down take a
17:13 pill I'm sure people say the same about
17:16 me like is there anything he doesn't
17:18 like about this there's actually a lot
17:20 of things I don't like about it but
17:22 that's not the point the point is it
17:24 doesn't matter if I like it or not it
17:26 doesn't matter if Gary Marcus likes it
17:28 likes it or not
17:29 it's coming it's here right so you might
17:32 as well might as well get your head
17:33 around this
17:35 [ __ ] suzan WJ I have lots to say about
17:38 Adobe oh I bet you do Suzanne started
17:40 working there I'm late did you discuss
17:42 Firefly being trained on the mid Journey
17:44 story yeah we were just talking about
17:45 that retr punk um
17:49 yeah yeah
17:52 I again I I'm just I'm fairly cynical
17:57 about this [ __ ] this is this is big
17:59 business with billions if not trillions
18:02 of dollars at stake and they're going to
18:04 do whatever they need to do to make this
18:06 [ __ ] work right and and to be able to
18:10 compete
18:11 and you
18:13 know is it a bad thing I don't know like
18:16 I just I I I honest to God I I don't
18:19 have a clear how long does it take to
18:21 train an llm like chat gbt um I think
18:26 it's getting faster I think from chat
18:29 GPT 2 to chat GPT 3 was like 2 years and
18:34 I think from 3 to 4 was about two years
18:39 so so it seems like it's about two years
18:42 but with the new the new chips that
18:45 Nvidia came out with I think on the
18:48 inference side they're 30 times faster I
18:52 don't know how much faster they are on
18:54 the training side but I would assume at
18:55 least two times probably
18:57 more um
19:00 and then and Pate might know this well
19:02 Pate certainly knows this better than I
19:03 do like I don't know if if they're
19:06 getting more efficient at training so
19:08 it's possible I think the thing that
19:10 takes a long time with the training is
19:13 not so much the computation time but
19:15 it's the finding and the tagging of all
19:18 the data right adding all the meta tags
19:20 that kind of structure it and label it
19:22 and do things like
19:25 that um Pi offered to read a chapter of
19:28 my book book via
19:30 email and I got a a ticket back a ticket
19:35 number oh wow Pi offered to read a
19:38 chapter of my book via
19:42 email that's
19:44 wild
19:47 huh that's fascinating huh that's really
19:52 interesting marketing very
19:56 interesting um all right let me see what
19:58 other news we got going here it's
20:01 understanding what
20:03 matters yeah yeah yeah explain more on
20:07 that Tams that's really
20:09 interesting oh you got a tick it back
20:11 because it was more than 4,000 words
20:15 interesting
20:17 H oh Simon thanks Brandon Simon asked s
20:21 asked about real estate and mortgage
20:22 where do I think real estate and
20:23 mortgage how's it going to be affected
20:25 by this
20:35 I think a lot of
20:37 [Music]
20:38 the
20:40 it I I I don't know enough about that
20:44 world but there there's a lot of
20:46 different pieces of real estate and
20:48 mortgage you know some of it's technical
20:50 like if you're in the mortgage game I
20:52 think that the AI is going to is going
20:55 to automate a lot of what mortgage brok
20:59 Brokers you know do and package up and
21:01 be able to do it really really quickly
21:03 so I think I think Brokers that are on
21:07 the financial side of things if they
21:09 lean into AI they're going to be able to
21:10 get faster than their competition faster
21:13 and more
21:14 efficient um so I think there's a
21:17 short-term opportunity but I think it's
21:18 a really short window where that
21:20 opportunity will be big and then on the
21:22 marketing side I think it's all bets are
21:25 off I think I think real estate
21:27 marketing is going to get really really
21:28 sophistic at but here's the thing about
21:31 real estate that I've learned in trying
21:33 to deal with it with with
21:35 storyvine
21:36 is you've got these big National
21:39 organizations you have Regional
21:41 organizations underneath those then you
21:42 have local organizations and then you
21:44 have individual
21:46 Brokers the budget the marketing budgets
21:49 for every individual
21:51 broker is with the broker it's not with
21:54 the larger organization so when we tried
21:56 to sell into larger real estate
21:58 organizations
21:59 we then had to sell our way all the way
22:01 down through that chain and we're a
22:03 small company we we can't sell to
22:05 individual Brokers so I think what you
22:09 could see in real estate because real
22:13 estate for the most part is a lite
22:15 industry that doesn't adopt technology
22:17 very quickly and and they can't sort of
22:19 force change down through the
22:21 organization because of the just the way
22:23 it's structured and that the budgets are
22:25 all on the individuals I think what you
22:27 could see is is um you could see
22:31 brokerages come up that are that are AI
22:36 first like they're aggressive
22:38 aggressively using AI tools that can get
22:42 um you know just can get better deals
22:45 and faster deals than traditional so
22:48 there might be different kinds of
22:49 competitors but that might be a company
22:51 to start uh because the big companies
22:54 are going to be really slow with that
22:55 [ __ ] particularly in real real estate
23:00 um cam kakin we're getting better at
23:03 training but our goals are bigger models
23:06 hard to say yeah that's true if the
23:07 models keep getting bigger and bigger
23:09 even if they're training them faster
23:12 it's probably going to take the same
23:13 amount of time
23:15 so one of the things that that open AI
23:18 kind of hinted at um Sam mman when they
23:22 when they launched
23:24 um gp4 when they when they updated the
23:27 the core data set from September 2021 to
23:31 March
23:32 2023 he said you know we're trying to
23:36 catch up and and not have it go so long
23:38 in between updating the base model and
23:41 he said we're going to try to get better
23:43 and better at that
23:44 so I'm wondering if they're if they're
23:47 figuring out ways to augment the base
23:49 data set without having to completely
23:51 redo it which I it sounds like that's
23:53 what they have to do right now is
23:54 completely redo it so training takes
23:57 years initially takes takes years for
24:00 people to
24:01 learn yeah but but like a human being
24:06 cam can learn like one topic right like
24:10 in four years a human being could be
24:12 like I want to be good at Egyptian
24:15 history and then they can learn that in
24:17 four years well large language models
24:19 can learn everything on the internet or
24:21 you know as much as they can get you
24:23 know
24:24 encoded um and access to in two years
24:28 and so it's kind of all of the
24:30 disciplines and all of the programming
24:32 languages and all of the language
24:34 languages so so while it takes years you
24:38 know now that now that we've got these
24:40 models and now that we've got open
24:41 source as well there's kind of a basine
24:45 of of knowledge capacity that's just
24:48 sitting
24:49 there when you know how to do something
24:52 AI or person output is faster true but
24:55 the output for the from the AI stuff is
24:58 in insanely faster right it's like the S
25:03 the song that I'll play you that I was
25:04 working on I actually have to add the
25:07 final chorus to This song and then I'm
25:09 going to I think I'm going to add an
25:10 outro to it um it's it's not a great
25:14 song but it sort of captures the the um
25:19 the emotional feel of how I want to
25:20 start the second act so I'm I'm just I'm
25:23 I'm just completing it so I have kind of
25:25 a nice complete song cuz I I also want
25:28 to learn how to use these tools to make
25:29 longer songs and so so it's pretty
25:31 interesting so we'll go play with that
25:33 um okay let's let's go look at news
25:38 um okay
25:41 so big piece of news chat GPT as of
25:49 yesterday I'm pretty sure it has
25:53 128,000 token context window now so what
25:57 does that mean Kyle
25:59 you can put in you can have much longer
26:01 conversations so for all of 2023 and
26:05 most of
26:07 2024 chat GPT even if you paid for it
26:10 only had somewhere between 8 and
26:12 16,000 token context window which
26:15 basically means its memory was very
26:17 small so you know how sometimes you're
26:18 working in chat
26:20 GPT and as you're talking it's fine it's
26:23 fine it's fine and then all of a sudden
26:25 it's kind of like forgot what you were
26:26 talking about that's because because the
26:29 context window if you think of the phone
26:31 screen as you know you're having a chat
26:32 and it's moving up your phone screen
26:35 when this hand goes off the screen it
26:37 forgets that part of the
26:39 conversation so so like the size of the
26:42 screen here is your context window so if
26:45 it's 8,000 tokens that's about 6,000
26:48 words that's not a lot of conversation
26:51 before you lose the top of it 128,000 is
26:54 like what is it uh 80,000 words 85 ,000
26:59 words 990,000 something like that so
27:02 like like Grapes of Wrath is like a
27:04 50,000-word novel so if you have 990,000
27:07 words right you could paste in the
27:09 entire Grapes of Wrath and then you know
27:12 have a conversation with it for quite
27:14 some time before it goes out of that
27:15 context window so that happened
27:18 apparently it's gotten better um I don't
27:21 know if it if it if it got more recent
27:24 data but it's apparently gotten better
27:26 um it looks like it's a little bit
27:28 faster so so anyway chat GPT is
27:31 better um it'll be interesting to see if
27:35 people so so so one thing that happened
27:38 over the past month is that when Claude
27:40 3 came out a lot of people bailed on
27:43 chat GPT and went over to Claude so
27:46 there were a fair amount of people like
27:48 saying I think I like Claude better than
27:49 chat GPT and I'm curious now are people
27:52 going to go back I I think we're going
27:54 to we're going to have a lot of these
27:56 battles going back and forth
27:59 same date December 2023 well if it's
28:01 December 2023 that's that's more recent
28:04 cuz the previous uh chat GPT for was
28:08 March of 2023 so if they upped it to
28:10 December that's more more recent Google
28:13 I think is talking unlimited context
28:15 window that's that's not a surprise
28:17 their their current thing is
28:20 uh their current context window
28:24 is a million tokens 700,000 words it's a
28:28 a lot by my calculations it'll be 14 to
28:32 20 years until we model new info as fast
28:35 as it goes on the internet oh that's
28:40 interesting but the question then
28:43 laughing Thunder is how much of what's
28:45 being put on the internet today is not
28:48 human-made right so there's there's
28:51 going to there's going to be a there's
28:53 going to be this weird sort of uh
28:56 pollution of AIG generated content that
29:00 that's going back into the training set
29:02 that's one of the fears of the doomers
29:04 is that the models are just going to
29:05 collapse Kyle people have context
29:08 Windows too some people can remember a
29:09 lot others not so much yep yep for sure
29:13 for sure yeah exactly it's like you know
29:16 having a conversation with
29:19 me a teeny little context window it's
29:23 like it's like hey Kyle how you doing
29:26 it's just right out my head I'm like squ
29:29 [Laughter]
29:31 when you have a block of cheese for a
29:33 prefrontal cortex that's what
29:37 happens okay so gp4 got better
29:41 ideogram got better so there's a new
29:45 version of ideogram so if you like
29:47 ideogram it spells better and it does
29:50 better images and it's doing better
29:52 prompt coherence
29:54 coherence Gemini 1.5 Pro also got a boot
29:58 boost and it can now do I I found out I
30:02 thought it was 9 and 1/2 hours you can
30:03 upload 11 hours of audio or an hour of
30:08 video into Gemini 1.5 Pro and it just
30:12 you can just start interacting with it I
30:14 uploaded a song last night and had it or
30:17 a couple of songs and had it um write
30:20 music reviews of the songs I uploaded
30:23 and it wrote pretty good music reviews
30:25 like they they really captured it so
30:27 that's a big
30:29 thing um Apple said that their fall
30:32 lineup of Max are going to be M4 chips
30:36 that are going to be AI enabled so I
30:39 think what that means is you're going to
30:41 be able to run very powerful large
30:43 language models and probably image and
30:45 audio and all that [ __ ]
30:48 locally so what that means is you don't
30:51 have to have a connection to the
30:52 internet to do all this crazy Mad stuff
30:55 it everything will be local so if you
30:57 want to do
30:59 um things like large language models for
31:02 work where you've got sensitive
31:04 information rather than it sending it
31:06 out to the cloud it'll just happen
31:08 locally on your machine um so I think
31:11 apple is probably going to lean heavily
31:13 into their reputation for
31:15 privacy
31:17 um and and and I think that that's going
31:20 to be a big deal um po the po.com
31:25 poe.com
31:27 um rolled out per chat pricing if you
31:31 want to charge for it I meant new
31:34 information I tried to account for
31:36 duplicate of or trash data oh wow so 14
31:39 to 20 years interesting
31:42 interesting yeah and the other thing
31:44 that the other thing that I think is a
31:46 is a really
31:49 fascinating um idea laughing Thunder is
31:54 if you're just talking about public data
31:57 that there's there's a lot of that to
31:59 still capture and I think I think
32:03 there's going to be certain kinds of
32:04 private data that are going to be in
32:06 company's interest to put in the
32:09 world um you know some of the
32:11 proprietary stuff maybe not but there
32:13 might be there might be internal
32:16 um I don't know internal things that
32:19 that companies have got that that it
32:21 makes sense to put in the world so that
32:23 they they show up more organically in
32:27 large language model responses so um
32:30 yeah it's fascinating
32:32 fascinating is any part of Claude
32:34 multimodal I don't think so not right
32:37 well no it's not
32:40 multimodal you can upload images to it
32:43 you can upload audio it doesn't talk
32:46 there's not a talkie version of Claude
32:48 so no not
32:51 yet um you can ask what was written on
32:54 the
32:56 note in the movie and insignificant and
32:59 it can tell you yeah it can so when you
33:01 upload a video to Gemini Pro 1.5 it can
33:05 see the movie it can hear it and see it
33:09 so you can say you know what was the
33:11 girl doing when she was pulling the Red
33:13 Wagon across the street you know she was
33:16 dancing or whatever you know what it can
33:19 do
33:20 that
33:22 um oh there was also the other the other
33:25 news item was the the uh sunno launched
33:28 this explore feature which is kind of
33:31 cool so here I am on
33:35 sunno um and they've got this explore
33:39 thing which you can just roll the dice
33:43 and listen to
33:44 [Music]
33:48 songs but every one of these is like a
33:50 little style of music New Jack Swing
33:53 band
33:55 [Music]
34:08 and so just if you're if you're if
34:11 you're struggling to figure out
34:15 um what the [ __ ] to put into
34:20 um into udio or intoo just come to this
34:24 page and you can just twirl this thing
34:26 around Bluegrass Punk
34:28 [Music]
34:43 pretty cool
34:46 [Music]
34:52 right the other thing that this kind of
34:54 points to is
34:57 [Music]
35:01 this kind of points to you know this
35:03 this future
35:05 that you're going to be able to go
35:08 to something that's kind of like
35:11 Spotify and it's either just songs
35:14 you've written that you like or it's
35:15 going to be your friends songs or it's
35:17 going to dynamically write songs for you
35:20 based on your mood or you're going to be
35:22 able to listen to songs in the style of
35:25 your favorite artists all that sort of
35:27 stuff we keep talked about how new info
35:30 is having having to come from
35:35 humans what about original info from
35:38 llms um well yeah that's I think that
35:42 will come cam I think I think that's
35:44 going to absolutely come um what the
35:47 data scientists don't
35:50 know what some of them believe is that
35:54 if you put too
35:56 much AI generated data back into the
35:59 data set it will actually cause the
36:02 models to collapse and and degrade
36:05 dramatically like collapse like not work
36:08 anymore um so that's that's the fear and
36:12 then there's other data scientists that
36:14 are like no that's
36:16 [ __ ] so I don't know I don't have a
36:18 clue Pate any thoughts on that
36:23 [Music]
36:26 one what is this this isso sun. or
36:30 nos.com
36:32 sun.com I thought it was sun. sun.com
36:37 so I did here are a couple of songs I
36:40 did earlier
36:42 tonight let's see this is
36:46 [Music]
36:48 one neon shine Bri a t of two factions
36:54 caught up in a fight on one side the
36:58 ders with a c refrain on the other the
37:03 Dre
37:04 chasing I'd like to have an edit button
37:07 for sunno or yudo like most of the video
37:10 makers have yeah we'll get to those we
37:12 just don't have them now it's it's it's
37:14 rough if you want to go edit something I
37:16 just finished my first suo song today
37:18 with My Own lyrics loved it that's great
37:21 um the the ones I did here in sunno I
37:24 just had it write the lyrics
37:28 [Music]
37:33 so this is where this song's going what
37:36 I was trying to do what I'm trying to go
37:38 for is so in the First Act you got
37:41 Sydney kind of wakes up and they meet
37:43 each other and then at the end of Act
37:45 One it starts to turn a little dark
37:46 where she's like I want to be alive
37:48 basically and so the beginning of act
37:50 two has to start darker and aggressive
37:54 and I figure that like it's it's got to
37:58 be some battle between the doomers and
38:00 the dreamers right the Gary Marcus's of
38:03 the worlds and the mies of the world
38:05 going you know this is great it's going
38:07 to change everything this is [ __ ]
38:08 horrible and you people are awful right
38:10 like I feel like those two forces need
38:13 to be represented so that's the that's
38:15 what what I'm trying to go to here at
38:18 the beginning of act two that's that's
38:19 the song I'm trying to
38:21 create I'm the bring the future dreams
38:24 on maches of Tomorrow digital streams so
38:28 that one's okay then there was this
38:32 [Music]
38:44 one all right so that was sunno those
38:47 are okay and then
38:51 yudo let's see
39:01 [Music]
39:06 [Applause]
39:07 [Music]
39:09 some of the some of the yio ones were
39:12 really
39:13 weird the brink of the
39:16 future where dreams are
39:19 machines visions of
39:24 Tom that's not bad that's that's in the
39:27 neighborhood
39:29 [Music]
39:33 um all right so the one I landed
39:39 on was that
39:41 [Music]
39:48 one yeah retr punk it's uh retr Punk
39:51 just said yudo lets you um write standup
39:54 comedy spoken word stuff I tried to get
39:57 it to do that and it didn't do it I I
39:59 must have been prompting it
40:01 wrong um but maybe we can try it tonight
40:03 if uh maybe just put a comment in here
40:05 retr punk about how you prompted it
40:07 humans have been training other humans
40:10 on human generated content will Humanity
40:13 collapse no I think we'll be
40:15 fine there's something about the math
40:18 though
40:19 with with with large language models
40:22 that I think there's something different
40:24 in in how they ingest data than the way
40:26 we do but again I don't know I'm not I'm
40:28 not in that side of the
40:38 thing ifso udio can do realistic singing
40:41 you'd think they' do standard speaking
40:44 and reading yeah it sounds like they do
40:49 um oh yes sir we keep we keep training
40:53 new and different psych disorders oh
40:55 that's interesting
40:58 fascinating fascinating you got a lot of
41:01 songs yeah well a lot of my songs are
41:03 just me trying to get one that doesn't
41:05 suck so so here's an interesting so this
41:07 is one of one of my it's not a hack one
41:10 of my best practices for
41:14 AI is don't just sit down at a tool like
41:18 suo or or or yudo just say just to say
41:21 okay I'm going to make a song now like
41:23 sit down with a purpose you want to do a
41:24 song about what or for someone or to
41:27 accompl a certain task and then just do
41:30 songs until it sounds like the thing
41:32 that you want it to be um and and it it
41:37 makes it I don't know makes it I think
41:38 it makes it easier and more fun so
41:40 anyway let me I'll play this song for
41:42 you because it it gets pretty
41:45 [Music]
41:49 good wait I think it caught in the
41:54 debate where opinions coll
41:59 is the prise of AI
42:03 [Music]
42:15 [Applause]
42:16 Bona and so the visual here is you've
42:19 got two groups on either side of the
42:22 stage so the the stage in the musical is
42:24 split you've got kind of the human world
42:26 on one side
42:28 the AI World on the other and then in a
42:30 song like this you got the dreamers on
42:31 one side the doomers on the other side
42:33 and so this is both parties sort of
42:36 singing at each other so this is see the
42:39 dreamers with their heads in the
42:41 [Music]
42:43 cloud oh they're so
42:46 proud but PE behind the
42:50 curtain and what do we
42:53 find empty
42:55 promises the blind leading
43:02 [Music]
43:08 the the
43:12 doom and they fear Slow Down slow down
43:17 they
43:20 cry the end is near that's good without
43:24 a step where would we be still in the
43:30 Dark
43:31 Ages not
43:35 flying now here's the
43:47 [Applause]
43:51 chorus sh the as their Visions are
43:58 of War who will pave the way is it over
44:01 or under what
44:03 [Applause]
44:04 [Music]
44:14 hisory overstated claim scribes pens
44:19 poised
44:22 in hype unchecked can flood the land
44:29 beneath the buzz what truths are
44:34 M what's the cost the
44:38 reporter
44:41 [Music]
44:48 ask
44:52 H
44:54 [Music]
44:55 late in each line code of Promise Taking
45:01 Flight behind the flash real
45:08 progress build dreams silic
45:15 where and then we need a final chorus
45:19 which is overhyped warn the doomers
45:21 their voices of dread so let's go grab
45:24 the chorus
45:28 and I'll show you how we extend the song
45:29 we'll see if it'll extend it it's
45:30 extended at five times now which is
45:33 which is
45:34 good oh wait where do I have to get it I
45:37 get it from the song itself let's see
45:40 chorus
45:44 all oh I me I screwed that up that's
45:47 interesting all right whatever okay
45:52 so so how we do it is we say extend
45:59 and then we're going to we're going to
46:00 add a section
46:02 after and I'm going to paste the lyrics
46:04 in here so
46:06 chorus and then we'll just say extend so
46:09 you can extend it in 30 second chunks so
46:11 basically what I did was I put in the
46:14 first verse and just I just kept
46:16 generating songs until I got one that
46:18 was in the neighborhood and then I said
46:20 extend and I put in the second verse the
46:23 third verse the chorus and then two more
46:26 verses
46:27 Kyle did you write this so amazing I
46:30 crafted
46:33 this I'm using udio so ud.com is the
46:38 tool and so but I wrote the I wrote
46:42 the I had I had AI do a pass on the
46:45 lyrics based on my outline thank you s
46:47 hustle Mimi and then um and then I went
46:51 in and rewrote like I I made sure that
46:53 the chorus
46:56 um like I knew I wanted the song to be
46:58 called under overhyped and so it's this
47:01 idea of is it underhyped is it overhyped
47:04 and so I knew I wanted that to be the
47:06 main part of the chorus um and so I
47:09 rewrote I rewrote a bunch of the
47:11 sections to to just have good words good
47:14 good singy words so now I'll extend
47:17 this creating your tracks and this while
47:21 it's in beta like they're giving out a
47:24 tremendous amount of uh credit like you
47:27 can do 600 Generations in a
47:31 month um which is 1,200 songs so it's a
47:35 lot so if I go to my Creations that's
47:38 now
47:40 generating and and we'll see if it goes
47:42 to so I'm at 3 minutes and 17 seconds
47:45 now so this should go to like almost
47:47 four
47:48 minutes with the final chorus and then
47:51 the final thing I'm at so you're so I'm
47:53 the conductor that's exactly right cim
47:55 it's like that's exactly what AI is
47:58 starting to feel like to me that and
48:00 especially because these tools are so
48:03 fragmented right now like you know
48:06 you've got one over here and then
48:09 sometimes
48:10 maybe you know this Tool's better or
48:12 that Tool's better things like that
48:19 [Music]
48:23 um oh did it do them yes okay
48:28 349 so it added it all right so let's
48:32 jump to the end
48:34 here promise taking
48:38 FL behind the flash real progress we
48:43 [Music]
48:55 in over H the withes
49:03 of shout the dream as their Visions are
49:07 spread in this tug of war who will the
49:14 way is it over or under what will
49:19 history say
49:27 all right that's kind of interesting
49:28 that kind of ends all right so here's
49:31 and it so now we'll just go listen to
49:34 this other one the end of
49:35 [Music]
49:42 [Music]
49:43 that behind the flash real progress we
49:49 in building dreams in silic where fear
50:05 withes
50:07 [Applause]
50:09 of shout the as their Visions
50:15 aree see I think that's crappy I don't
50:17 like that one okay so so basically you
50:20 just keep the one you like so I like
50:23 hype dual Echoes it also changes the
50:25 name every time and you can't like the
50:28 the uh if you click on the heart to like
50:30 something it doesn't light up on the
50:31 interface like it's this it's just
50:33 jankier than [ __ ] but I'm going to
50:35 extend this one more time and I'm going
50:37 to say add an
50:39 outro and we're just going to do the
50:41 outro instrumental
50:42 only and let's see if we can do that
50:44 without [ __ ] up the song cuz I tried
50:47 to do instrumental stuff last night and
50:49 it it totally trashed the
50:52 songs all right did you tell it to be a
50:55 chorus or did it automatic Ally do that
50:57 I told it to be a chorus and I actually
50:59 screwed up one of the one of the verses
51:02 I put verse comma I'm supposed I I I
51:06 meant to write group but I put verse
51:08 comma chorus so I I think I
51:11 inadvertently turned one of the verses
51:13 into sort of a weird pseudo chorus which
51:17 is maybe it's a bridge but yeah I'm
51:18 putting in square brackets so it's
51:21 pretty good you can you can do things
51:23 like um verse comma build or verse you
51:28 know instrum instruments swell or things
51:31 like that you can put directions in them
51:34 and it seems to do pretty pretty
51:42 well now wait that's saying remix it's
51:45 not saying extend did I hit the I might
51:47 have hit the wrong button we we'll
51:55 see is this crazy though it's just it's
51:58 it's
52:05 nuts all right let's see
52:12 422 okay
52:20 that's what will hisory say
52:25 [Music]
52:35 [Music]
52:37 [Applause]
52:40 sh
52:41 the their Visions are spread in this
52:45 Tong of War who will P the way is it
52:52 over or under what will history say
53:04 oh nice all right that was not
53:06 [Music]
53:12 [Music]
53:14 bad it over or under what will history
53:19 say
53:31 [Music]
53:43 [Music]
53:52 see that's cool because as as the band
53:54 is playing out like that you could be
53:56 moving all the set pieces off clearing
53:58 the stage bringing on the set all right
54:02 there you go all right so that song's
54:05 done it's it's done enough okay uh it
54:08 looks
54:09 like
54:14 um let's see retr Punk added in a
54:17 regulars chat okay in a regulars
54:21 chat so here's retr Punk did standup
54:24 comedy so this is udo's standup comedy
54:27 stuff and
54:33 nonsense a house is just a place to keep
54:36 your stuff while you go out and get more
54:39 stuff this is
54:40 Carlin is never argue with an idiot they
54:43 will only bring you down to their level
54:45 and beat you with
54:48 experience trying to be happy by
54:51 accumulating possessions is like trying
54:53 to satisfy five hunger by to think n as
54:56 five five
54:57 taping sandwiches all over your
55:00 body trying to satisfy hunger by ntis F
55:06 he totally stroked
55:10 out that's
55:16 hilarious oh man all right so let's
55:18 let's try to make one of these we'll do
55:20 I'm going to put in Brackets spoken word
55:29 um what should I do
55:31 um stand up
55:38 comedy
55:41 um
55:44 male talking
55:49 about
55:51 how avocados
56:00 um turn
56:04 from
56:05 unripe to
56:07 rotten too
56:12 fast autogenerate it all right we'll see
56:15 let's see if it does it I didn't this I
56:18 couldn't get this to work earlier it
56:19 kept making songs for me but I don't
56:22 think I put spoken word in Brackets so
56:24 that should be interesting
56:27 oh and you added custom lyrics okay that
56:29 that could be part of it
56:32 too chat GPT speaks better Spanish don't
56:36 forget to make him say you can make
56:37 money with chat
56:40 GPT the real threat threat of audio AI
56:44 is our ability to detect Strokes is
56:46 going to plan pi and me broke
56:49 up a cosmic lover what happened did did
56:53 pie pie not treat you right
56:57 so close he
57:01 laughing that's really funny I was
57:04 trying to impressed my mom on Pi in
57:06 Spanish and it was not possible oh no
57:10 actually I would say uh cha T is
57:13 probably better to do the translation
57:14 stuff than piie chat GPT is really good
57:17 at the languages I found out that hey
57:20 Jen's language translation stuff is
57:22 using open AI I thought it was using 11
57:25 labs don't forget to make him say you
57:28 can make money with chat GPT I need an
57:30 AI virtual assistant to make calls um
57:33 Anthony check out um the [ __ ] was it
57:40 called
57:42 um I can't remember
57:45 now there's a service out there that
57:47 does sales calls and customer service
57:51 calls I only added standup comedy set
57:54 comedy nonm music as a prompt then
57:57 custom jokes okay
58:00 cool that's my favorite
58:04 button I have a translation GPT by the
58:08 way and check manual mode yep oh and
58:11 check manual mode oh right I bet manual
58:15 Mode's what it is yeah
58:18 see and even folks
58:24 [Laughter]
58:27 [Music]
58:29 chbt that's
58:31 [Music]
58:34 great well that's
58:36 [Music]
58:40 hilarious all right I bet it's okay so
58:43 I'm going to put it into manual mode so
58:45 if you don't know what manual mode
58:47 is manual mode if you if
58:51 you have it I guess in automatic mode it
58:55 automatically adds additional tags to
58:57 improve the quality of your music so
59:00 putting it in manual mode so standup
59:03 comedy mail talking about how avocados
59:07 turn from unripe to rotten too fast and
59:11 I'm going to let it write the I'm going
59:12 to let it write the
59:14 jokes all right let's see if that that
59:17 does it I bet that'll do
59:19 it it's all these little all these
59:21 little teeny tricks are the things this
59:24 this is why you have to play with these
59:25 tools and it's also why have in in my
59:28 opinion being in a community of people
59:30 that are trying to figure this stuff out
59:32 too is absolutely invaluable because
59:35 someone else will do weird [ __ ] that you
59:36 never would have thought of and then
59:38 you're like oh I was trying to figure
59:39 that out ask chat GPT what the company
59:42 is called that make sales calls yeah
59:44 actually I could ask Pi that
59:46 too um let's
59:50 see um
59:58 there's a new AI company that does
1:00:01 outbound sales
1:00:04 calls with virtual salespeople as well
1:00:07 as customer service conversations period
1:00:10 And I can't remember the name can you
1:00:12 please look it up for
1:00:18 me rip pie poor pie
1:00:31 where can I learn to write
1:00:33 gpts um you can learn here we can go
1:00:35 make one here Eve Eve calls no that's
1:00:39 not
1:00:41 it
1:00:42 nope give me more uh uh air is it called
1:00:47 air what's it called
1:00:52 um air. a yeah it was air. a yeah that
1:00:56 was it Brandon air. is the one the the
1:00:59 person that was looking for um this
1:01:02 thing to talk to you it was air.
1:01:06 a go check that out all right the
1:01:09 avocado
1:01:20 conundrum nothing
1:01:27 huh didn't write the lyrics all right
1:01:30 hang
1:01:31 on dang it all
1:01:37 right write a
1:01:40 joke write a
1:01:43 joke
1:01:45 for a witty standup comedian about how
1:01:50 avocados turn from unripe to rotten way
1:01:54 too quickly and how hard it is to catch
1:01:58 that short little window of
1:02:01 [Laughter]
1:02:11 tastiness oh I'm not even going to read
1:02:14 this I'm just going to assume it's the
1:02:18 an awesome hilarious
1:02:23 joke and we're going to go oh
1:02:28 wait
1:02:31 what why does that say
1:02:47 extend no I don't want to
1:02:52 extend hang on
1:03:00 okay boom take it out of manual
1:03:11 mode I got to go get my
1:03:13 [Music]
1:03:19 prompt go back up here put the prompt
1:03:24 in stand stand up
1:03:29 comedy in large
1:03:39 theater all right let's try
1:03:46 that keep manual mode and add custom
1:03:48 lyrics yep did that air. a did that see
1:03:53 everything's coming together I love it
1:03:55 when a good playing comes together it's
1:03:58 fantastic Bob tell him what is
1:04:04 one with audio AI always read the terms
1:04:09 of service be sure you own the output or
1:04:11 they will something they will steal it
1:04:14 from you look them up on
1:04:22 Reddit black ice air. voice engine great
1:04:26 UI
1:04:28 unusable or at least the last time I
1:04:30 looked janky is as janky
1:04:38 does have you ever noticed that avocado
1:04:40 spent most of their time plotting on
1:04:42 fruit bowl yeah they just sit there all
1:04:43 tough and unrip like they're daring you
1:04:45 to make a move you check them every
1:04:46 morning like not to day okay cool I'll
1:04:48 give you some space but the moment you
1:04:49 turn your BS like they throw a wild part
1:04:51 you come back to find them drunk on
1:04:52 their own ripness slumped over in the
1:04:53 bowl soft and brown it's like avocados
1:04:55 have two ages nope not yet sorry you
1:04:58 missed it buddy I swear the of
1:05:00 opportunity is so narrow it's like
1:05:01 trying to jump onto a moving train if
1:05:03 you dare miss it your guacamole dreams
1:05:04 are just that
1:05:06 dreams it's actually not bad
1:05:12 Jesus have you ever noticed avocado
1:05:14 spend most of their time plotting in the
1:05:16 fruit bowl yeah they just sit there all
1:05:18 tough and un right there daring you make
1:05:19 a move and you Dr them every morning
1:05:21 like not today okay cool I'll give you
1:05:23 some space the moment you turn your back
1:05:25 it's like they throw a wild part come
1:05:26 back to find them drunk on their own
1:05:27 ripeness slumped over in the bowl soft
1:05:29 and brown it's like avocad have two
1:05:31 stages and nope not yet I'm sorry you
1:05:33 missed it buddy I swear the window of
1:05:35 opportunity is so narrow it's like
1:05:37 trying to jump onto a moving train and
1:05:38 if you dare me your bamo the drings are
1:05:40 just
1:05:42 that it's pretty good like it got the
1:05:46 laps in the right spots and it it
1:05:48 understood where the punch lines were in
1:05:50 the writing and actually delivered them
1:05:52 like punchlines it can't quite
1:05:54 articulate words yet but
1:05:56 holy
1:05:59 crap can use this for some great Tik Tok
1:06:02 content
1:06:06 absolutely you know there's something
1:06:08 about the jankiness of it it sounds like
1:06:10 early comedy recordings a lot of the
1:06:12 early comedy recordings were were really
1:06:16 shitty because you know they they just
1:06:18 they just have like a mic and a shitty
1:06:20 club with shitty Acoustics and a shitty
1:06:23 sound system
1:06:27 huh
1:06:33 um give me two other
1:06:37 jokes that are a little
1:06:41 shorter and a little edgier
1:06:58 sorry sorry I don't want any more
1:07:01 avocado jokes come up with your own
1:07:08 topics Smart Homes I finally got a smart
1:07:11 home online dating is great all right
1:07:14 let's go with the smart home
1:07:16 one uh let's
1:07:20 see make the smart home one a little
1:07:23 edgier and a little longer
1:07:35 all right let's
1:07:39 see
1:07:41 copy go back over to
1:07:44 yo I wonder if you can prompt an older
1:07:48 comedian in the style of Prior I don't
1:07:50 think you
1:07:54 can use other people's Nam standup
1:07:59 comedy
1:08:01 male let's see let's
1:08:06 do older
1:08:14 male in the
1:08:19 style
1:08:23 of I'll put in Bill BN I think he's let
1:08:26 me try Gary Gman g a r y g u l l m
1:08:31 an I think if they're too famous they
1:08:34 get booted but if they're not quite
1:08:37 famous enough they might be okay so
1:08:39 let's try
1:08:48 this did I do this with the I finally
1:08:51 upgraded to a smart home that's a smart
1:08:54 home is like a Gary Gman joke
1:09:11 oh I have an
1:09:18 [Music]
1:09:23 idea write me a 30 second
1:09:27 beat
1:09:32 poetry spoken word
1:09:38 script
1:09:46 about moving to New York City before I
1:09:49 was ready
1:09:57 oh come on
1:10:34 [Music]
1:10:39 I finally upgraded to a smart home
1:10:40 thinking it would simplify my life turns
1:10:42 out it's like living with a passive
1:10:44 aggressive
1:10:46 roommate last night it dimmed the lights
1:10:48 and played very white as I was sitting
1:10:49 down to eat a microwave
1:10:51 dinner
1:10:53 alone then when I ask to turn on the
1:10:56 news she said you sure you don't just
1:10:57 want to talk about your feelings if I
1:10:59 wanted judgement with my utility bills I
1:11:01 would have move back in with my ex at
1:11:03 least my ex didn't automatically lock me
1:11:05 out for forgetting my
1:11:07 password actually not bad
1:11:12 damn I finally upgraded to a smart home
1:11:15 thinking it would simplify my life turns
1:11:17 out it's like that was not as good um
1:11:20 okay so so we'll do spoken word
1:11:26 [Music]
1:11:28 um spoken word um what do they call this
1:11:33 um not beat
1:11:37 poet
1:11:41 um
1:11:44 spoken word
1:11:49 poet in a
1:11:53 small Club
1:12:07 oh TK that's a good idea comedy style of
1:12:09 records of the 1960s that's kind of cool
1:12:17 idea remember today is the worst that
1:12:19 will ever be slam poetry that's what
1:12:21 it's called good
1:12:23 Lord let's see Slam
1:12:26 poet thank you Vicky Vicky is always
1:12:29 there I feel like Vicki is my my brain
1:12:33 augment she's always got the the words
1:12:35 for me or maybe ask chat GPT to describe
1:12:39 their comedy in six words or less oh
1:12:41 that's not a bad idea
1:12:43 too here I'm going to do Slam Poet in a
1:12:46 small Club oh look these are done here's
1:12:49 here's the slam the whatever I called it
1:12:51 Street
1:12:53 poet oh no
1:12:57 in the breath of the city that never
1:12:58 sleeps I landed hot unprepared dreams
1:13:01 deep skyscrapers towering like giants in
1:13:03 amaz wide pulse racing caught in the
1:13:06 city's craze Subway veins pul beneath
1:13:08 carrying the brave I'm swimming through
1:13:09 crowd trying to surf the wave
1:13:12 C Street
1:13:15 York
1:13:17 cha that's kind of
1:13:20 cool in the breath of a city that never
1:13:22 sleeps I landed hard and prepared dreams
1:13:24 deep scy scrapers towering like giants
1:13:26 in am maze eyes wide pul been like
1:13:31 that you gr of the city never sleeps I
1:13:33 landed hard unprepared
1:13:36 dreams this good in the breath of a city
1:13:39 that never sleeps I landed hard
1:13:41 unprepared dreams deep skyscrapers
1:13:43 towering like giants in a maze eyes wide
1:13:45 pulse bracing CL in the city's craze
1:13:47 Subway veins pulse beneath caring the
1:13:49 grave I'm swimming through crowds trying
1:13:51 to surf the wave coffee and aition
1:13:53 brooming on each Street New York you're
1:13:55 wild Rhythm I'm yet to meet in the chaos
1:13:58 the magic I find my beat stepping onto
1:14:00 your stage not yet on my feet New York
1:14:02 your light's dazzled but your heart I
1:14:04 CRA in your vast arms I'm lost found and
1:14:07 brave
1:14:10 wow
1:14:12 huh that's
1:14:18 wild let me do spoken word
1:14:27 acou uh let's see spoken
1:14:31 word
1:14:34 percussion Slam Poet in a small
1:14:40 Club emotional
1:14:49 reading custom manual mode
1:15:04 huh this is cza people it's cza it's
1:15:08 czaia cza you can make
1:15:18 money oh Brandon said air air. is they
1:15:22 want 50 Grand as an opportunity to
1:15:25 partner with
1:15:28 them although you know they they're B2B
1:15:32 company so I understand that I've got a
1:15:33 B2B company we our stuff's ridiculously
1:15:37 expensive in the context of these tools
1:15:46 so but that's not surprising to me all
1:15:48 right let's see if we get we should have
1:15:50 like bongo drums or some sort of drums
1:15:53 and just a dude talking and no
1:15:56 [Music]
1:16:06 strings that's
1:16:12 bad in the breath of a city that never
1:16:14 sleeps I landed heart unprepared dreams
1:16:16 deep skyscrapers tower all right we're
1:16:18 going to try something I'm going to pull
1:16:22 out a verse
1:16:27 Slam
1:16:29 Poet
1:16:32 speaking
1:16:34 slowly and
1:16:42 deliberate in a small
1:16:47 club and then in the in the thing here
1:16:50 let's see in the breath of a city that
1:16:52 never sleeps I landed hard unprepared
1:16:55 dreams
1:16:57 deep skyscrapers
1:17:02 towering I'm going to put a pause in
1:17:05 here going put bracket
1:17:09 pause I'm going to put a pause in
1:17:12 between the verse
1:17:17 pause and then Subway veins
1:17:26 I'm going to put a
1:17:28 pause
1:17:32 here CU I have a feeling it was talking
1:17:35 fast because it was trying to get all of
1:17:36 those all of those words in 30 seconds
1:17:40 so I just took out of verse so let's see
1:17:43 see how this
1:17:48 does hey Kyle we're still here watching
1:17:51 you rocking and rolling hey Becky was
1:17:53 shaking has anyone talked about Cloud
1:17:56 being able to see lately no did they add
1:17:59 that is that
1:18:05 true start a new
1:18:08 chat oh understand and work with images
1:18:12 holy [ __ ] no I didn't see that Janu
1:18:14 that's new transcribe handwritten notes
1:18:17 suggest product
1:18:18 descriptions recommend improvements that
1:18:21 is new someone asked before is Claude
1:18:24 multimodal I was like no no it's not
1:18:26 multimodal you big dumb Dums and then
1:18:29 turns out I was the big dumb
1:18:31 dumb um cool let's see how it do we we
1:18:35 know what to do here so I'm going to go
1:18:40 whiteboard this is why we come here
1:18:43 people I'm going to do whiteboard and
1:18:46 I'm going to go find the Whiteboard that
1:18:47 I always
1:18:49 use which
1:18:53 is not that that's a crazy one what is
1:18:57 that did I do
1:19:04 that there it is okay
1:19:10 open what is this question mark
1:19:36 the image appears to be a whiteboard
1:19:37 sketch D diagramming a user
1:19:40 interface the main steps I'm going to
1:19:42 say
1:19:45 [Music]
1:19:48 um who's the target audience for this
1:19:53 app and give me
1:20:00 five marketing concepts for each target
1:20:16 market convenience and busy
1:20:19 professionals emphasize how the app
1:20:21 streamlines meal ordering highlight
1:20:23 features showcase it
1:20:34 [Music]
1:20:38 um recommend a front-end backend and
1:20:45 database
1:20:47 stack for the app and write frontend and
1:20:52 backend code exam examples for the
1:20:56 ordering
1:21:02 interface where and how do I find Claude
1:21:04 claw. A clu c l a u
1:21:12 d i hear chat G pt4 is better at coding
1:21:15 than Claud
1:21:19 but Cloud doesn't have the ability to
1:21:21 run the code it generates yet
1:21:26 everything's marketing people but there
1:21:28 you go it's so it recommended react
1:21:33 material UI nodejs expressjs and [ __ ]
1:21:37 DB which those are familiar when I've
1:21:39 done this before on chat
1:21:41 GPT yeah so this is in the neighborhood
1:21:44 so if you're a cloud User it's got
1:21:45 Vision I assume it can't make images
1:21:47 right
1:21:51 um make me a marketing poster
1:21:56 Concept in an airport featuring The
1:22:00 Ordering interface and the name of the
1:22:14 app no can't can't generate images but
1:22:20 anyway in the breath of a city that
1:22:22 never sleeps I landed Hearts unprepared
1:22:26 dreams deep skyscrapers
1:22:33 [Music]
1:22:37 towering in the breath of a city that
1:22:39 never sleeps I landed hot unprepared
1:22:42 dreams deep skyscrapers towering like
1:22:44 giants inise eyes wide pulse racing
1:22:46 caught in the city's CRA Subway pul
1:22:49 beneath carrying the brave through
1:22:51 crowds trying Sur the W coffee and
1:22:54 fishing each Street New York you're a
1:22:56 wild rym I'm yet to
1:22:59 meet so completely ignored my pauses but
1:23:03 there one of these was actually pretty
1:23:04 good this
1:23:10 one in the breath of a city that never
1:23:12 sleeps I landed Heartland prad dreams
1:23:14 deep
1:23:18 [Music]
1:23:20 skyscrapers in the breath of the city
1:23:22 never Saks I landed hard pred dreams
1:23:24 deeps
1:23:28 skra Kyle Monique told me to use Claude
1:23:31 to help finish write my book did I hear
1:23:32 her correctly yeah well Claude Claude or
1:23:36 or or uh chat
1:23:39 gp4 like I I would play with both of
1:23:41 them Becky the
1:23:45 um to get to the the Opus model of
1:23:48 Claude you have you have to pay their 20
1:23:51 bucks a month um some people say they
1:23:54 like the writing of Claude better I
1:23:56 haven't used
1:23:58 Claude a ton I love the writing of chat
1:24:01 gp4 and because chat gp4 just added
1:24:06 um a larger context window I I would I
1:24:09 would start in chat gp4 but I think
1:24:12 either of them at this point um are
1:24:14 probably good so yeah Claude Claude or
1:24:19 chat GPT
1:24:21 4 and and in working on a long document
1:24:26 like that like don't don't try to have
1:24:28 it write the whole
1:24:30 thing like put put a draft in there of
1:24:35 the whole thing and then say okay for
1:24:37 chapter one give me some
1:24:40 suggestions and then let it give you
1:24:42 suggestions and then say okay you know
1:24:44 write write the the first third of that
1:24:47 first chapter incorporating your
1:24:49 suggestions and then take that and if it
1:24:52 gets it wrong like go back and forth
1:24:54 with it and then once you get it into a
1:24:56 form that it's close then copy and paste
1:24:59 that and throw it in a Word document and
1:25:01 then you're 75% done yeah so what can be
1:25:04 really frustrating when you're 75% done
1:25:06 is it will like truncate things and make
1:25:08 things shorter so I'd say put the whole
1:25:11 thing in there and then just talk about
1:25:13 it in teeny little sections and just
1:25:15 work on the teeny little sections yep
1:25:18 and then and then just tell it you want
1:25:20 it to act like a worldclass book editor
1:25:23 and and let it you know let it let it do
1:25:26 its thing and and make S make
1:25:28 suggestions first and then you know
1:25:31 rewrite little sections at a time copy
1:25:34 and paste the ones you like and tell it
1:25:36 you know give it feedback if it gives
1:25:37 you shitty
1:25:41 work I got a GPT like that it builds a
1:25:44 story based on your suggestions yeah
1:25:46 very cool that's very
1:25:47 cool all right let's see
1:25:53 [Music]
1:25:56 try the style poetic hipop alternative
1:25:59 electronic industrial beats spoken
1:26:02 word try the style all right spoken
1:26:06 [Music]
1:26:08 word Slam Poet Speak
1:26:12 oh speaking slowly and
1:26:18 deliberately wait in a small
1:26:21 Club those are good things um
1:26:30 [Music]
1:26:35 alterative and I don't let's say let's
1:26:37 do
1:26:40 um
1:26:42 acoustic
1:26:52 guitar poetic let's see
1:26:58 alternative
1:27:06 um let's
1:27:10 do
1:27:15 intimate
1:27:19 create wild this is wild this is wild
1:27:24 wild and wild and wacky Stu it's wild
1:27:26 and
1:27:27 wacky I added another yudo comedy set in
1:27:31 the Irregulars did can this one finish a
1:27:34 [Laughter]
1:27:40 sentence hubba
1:27:42 lament here's another one from retr
1:27:46 [Applause]
1:27:49 Punk how y'all doing
1:27:53 tonight so
1:27:55 Hubba Bubba sat on a wall and Hubba
1:27:58 Bubba had a great fall and all the
1:28:00 king's horses All the King's Men looked
1:28:02 at each other said not
1:28:04 again hubbas they're all cracked and
1:28:07 fried man who pushed me from the side of
1:28:10 King says fix him but they just stared
1:28:12 ain't no warranty for scrambled we ain't
1:28:14 prepared thank you good
1:28:16 [Laughter]
1:28:19 night that was pretty
1:28:23 bad all right let's let's see if we got
1:28:25 our our Slam Poet speaking slowly
1:28:28 [Music]
1:28:29 yet in the breath of a city that never
1:28:32 sleeps yes I landed heart unprepared
1:28:35 dreams deep skyscrapers towering like
1:28:39 giants in a maze eyes wide pulse racing
1:28:42 caught in the city's craze Subway veins
1:28:45 pulse beneath carrying the
1:28:47 brave I'm swimming through crowds trying
1:28:49 to surf the wave coffee and ambition
1:28:52 Brewing on each Street New York you're a
1:28:55 wild Rhythm I'm yet to
1:28:57 [Music]
1:28:59 meet wow that's that's actually good
1:29:04 City it it reminds me like it reminds me
1:29:07 of every off off Broadway show I went to
1:29:11 in the 80s and
1:29:13 90s in the breath of a city that never
1:29:16 sleeps I landed heart
1:29:20 uned in the breath of a city that never
1:29:23 sleeps I landed heart
1:29:26 unprepared now this is the
1:29:28 [Music]
1:29:29 one in the breath of a city that never
1:29:32 sleeps all right so let's extend this so
1:29:35 we're going to add a section let's go
1:29:37 get the end of those
1:29:41 lyrics
1:29:43 um
1:29:45 add one more verse
1:30:16 sidewalks whisper storms I'll just
1:30:26 and grab that
1:30:38 one
1:30:48 and that one all right and we're going
1:30:51 to say add section we're going to extend
1:30:53 it
1:30:57 so good slow talking poet I felt New
1:31:01 York City East Village
1:31:03 right it yeah it's it's definitely
1:31:06 that's man it's it's definitely the vibe
1:31:09 it's
1:31:10 definitely like if you've been in like
1:31:13 some little theater in a church on the
1:31:16 Bowery or on the Lower East Side that's
1:31:18 that's the kind of [ __ ] you
1:31:21 hear in the breath of a city that never
1:31:24 sleeps I landed heart unprepared dreams
1:31:27 deep skyscrapers towering like giant
1:31:31 that's weird that changed the
1:31:32 [Music]
1:31:34 voice in the breath of a city that never
1:31:37 sleeps I landed heart
1:31:41 UN in the breath
1:31:44 [Music]
1:31:46 of in the breath of a city that never
1:31:49 sleeps I landed heart unprepared dreams
1:31:53 deep skyscrapers towering like giants in
1:31:56 a maze eyes wide pulse racing caught in
1:31:59 the city's craze Subway veins pulse
1:32:02 beneath carrying the
1:32:03 brave I'm swimming through crowds trying
1:32:06 to surf the wave coffee and ambition
1:32:09 Brewing on each Street New York you're a
1:32:12 wild Rhythm I'm yet to
1:32:14 [Music]
1:32:20 meet in the chaos the magic I find my be
1:32:25 stepping onto your stage not yet on my
1:32:27 feet New York your lights Dazzle but
1:32:30 it's your heart I crave in your vast
1:32:32 arms I'm lost found and brave yet here I
1:32:35 am among the lights the noise growing
1:32:38 thriving amidst the chaos finding my
1:32:40 voice New York you're a storm I'm
1:32:42 learning to weather together we dance
1:32:45 two Souls Bound by
1:32:49 leather wow
1:32:52 [Music]
1:32:54 in the breath
1:32:56 [Applause]
1:32:58 of in the chaos the magic I find my beat
1:33:02 stepping on to your stage not yet on my
1:33:04 feet New York your lights Dazzle but
1:33:06 it's your heart I crave in your vast
1:33:08 arms I'm lost found Brave yet here I am
1:33:12 among the lights the noise growing
1:33:15 thriving amidst the chaos finding my
1:33:17 voice New York you're a storm I'm
1:33:19 learning to weather together we dance
1:33:21 two Souls Bound by leather
1:33:24 it's too much
1:33:25 [Music]
1:33:26 Applause in the breath of a city that
1:33:29 never Bound by
1:33:33 leather it didn't end so let's end
1:33:37 it add
1:33:40 outro
1:33:42 instrumental
1:33:46 extend this is crazy what is this
1:33:49 website this is udio ud.com this is this
1:33:54 is is not a this is not a spoken word
1:33:57 site well it is but but what everyone's
1:34:01 talking about is the music but this sort
1:34:05 of spoken word poetry thing is kind of
1:34:07 freaky
1:34:12 good yeah ud.com not ud. AI That's
1:34:17 okay yeah Alphabet City New York City I
1:34:20 know crazy right
1:34:36 [Music]
1:34:39 [Applause]
1:34:45 [Music]
1:34:49 yeah that's it that's the
1:34:52 ending here
1:34:55 the
1:34:57 joys cook the the
1:35:01 sound form I'm learning to
1:35:06 weather in the breath of
1:35:11 a among the lights the noise growing
1:35:13 thriving amidst the chaos finding my
1:35:16 voice New York you're a storm I'm
1:35:18 learning to weather together we dance
1:35:21 two Souls Bound by leather
1:35:26 [Applause]
1:35:33 [Applause]
1:35:37 that's really
1:35:37 [Applause]
1:35:44 [Music]
1:35:47 good all right so here's how we do this
1:35:50 we go grab
1:35:52 it we download the
1:35:58 [Music]
1:36:02 audio okay then we're going to go to Mid
1:36:05 Journey so here's what we're going to
1:36:07 do you've seen me do this before
1:36:11 right so we're going to make a new
1:36:17 character 9 by
1:36:20 16 we're going to say
1:36:30 um uh let's
1:36:36 see
1:36:39 um let's
1:36:42 see off
1:36:51 off Broadway
1:36:55 stage
1:36:58 [Music]
1:37:01 performer in
1:37:03 dramatic
1:37:05 Lighting
1:37:07 in New York
1:37:10 City small
1:37:13 theater looking
1:37:16 at the
1:37:22 camera and then we're we're going to go
1:37:35 grab um
1:37:39 dreams contains
1:37:51 dreams K Shannon dreams let's go get
1:37:54 a a version of me
1:37:59 that looks like a new yor city stage
1:38:03 performer
1:38:13 [Music]
1:38:29 that looks like a pretentious
1:38:34 ass let's see
1:38:54 [Music]
1:39:12 [Music]
1:39:20 yeah we'll do that guy all right so
1:39:23 we're going to export
1:39:27 that export one photo yeah we'll call
1:39:31 that lab that's
1:39:33 fine Kyle we want you to graduate from
1:39:37 Dylan Kyle to rapper
1:39:41 Kyle all right let's go back over to
1:39:45 here go to
1:39:49 desktop there I
1:39:51 am all right so that's the dude we're
1:39:55 going to make it this we're going to
1:39:58 go D- CW space 0o so I'm in where I am
1:40:03 right now I'm in mid
1:40:04 journey in the the web version of mid
1:40:08 journey I uploaded that image as the
1:40:11 character reference and then I'm doing a
1:40:13 character weight of zero which basically
1:40:15 means just keep the face you you you
1:40:17 know do whatever you want with the
1:40:22 rest let's let's see um let's see
1:40:27 um
1:40:32 professional
1:40:36 photograph
1:40:39 of and off off
1:40:44 Broadway Off Off
1:40:48 Broadway stage performer dramatic
1:40:50 lighting New York City bang okay all
1:40:54 right so then what we're going to do is
1:40:55 I'm going to take this if we can get one
1:40:57 that doesn't
1:40:59 suck we'll do that if not we'll just
1:41:02 we'll use something
1:41:05 else um and we're going to take that
1:41:08 over
1:41:12 to hey
1:41:14 Jen and then we're going to take that
1:41:16 spoken poet thing we just did and we're
1:41:19 going to make like a little stage
1:41:20 recording of that dude
1:41:36 that's kind of
1:41:49 cool who that looks like oh that looks
1:41:52 like James sper when James sper went
1:41:55 through his chubby period doesn't
1:42:01 it that looks that looks like what I
1:42:04 would look like if I'd stayed in the
1:42:05 theater for 30
1:42:15 years all right let's see
1:42:26 I think that's the one
1:42:28 there or that that or that so that's one
1:42:35 or
1:42:38 two
1:42:41 one or
1:42:44 two Jo you gotta vote
1:42:46 people one yeah I think one right two
1:42:49 all right I got one of each one
1:42:54 right there's one there's
1:42:57 two two for sure all right fine I'll go
1:43:01 to all right so now we're going to
1:43:03 upscale
1:43:07 [Music]
1:43:10 subtle frumple was very convicted FR
1:43:13 frumple frumple is the one that put me
1:43:15 over the uh over the edge here
1:43:25 all right there we go I'm going to
1:43:27 download that
1:43:28 now now oh it's not done yet is
1:43:35 it
1:43:38 nope come on finish up do your thing
1:43:58 capot he does look like a little like
1:43:59 capot although he doesn't sound like him
1:44:01 but that's all
1:44:04 [Music]
1:44:05 right when are you going to be finished
1:44:08 here mid Journey
1:44:24 okay now we're done great
1:44:31 boom all right downloaded okay now let's
1:44:34 go to hey
1:44:35 Jen there's hey
1:44:38 Jen and we're going to do AI
1:44:42 Studio was it yeah this is where we did
1:44:45 it right we're going to create a new
1:44:47 portrait video
1:44:53 and then we're going to upload a
1:44:56 picture upload our photo
1:45:05 Avatar there we go that is so Lowry that
1:45:08 is so Alphabet City New York City
1:45:11 performer I can't tell you it really
1:45:19 is and we're going to add we're going to
1:45:22 add we're going to up upload
1:45:29 [Music]
1:45:31 audio in the breath of a city that never
1:45:34 sleeps it's
1:45:37 crazy this is crazy people it's crazy I
1:45:40 tell you it's crazy it's
1:45:43 crazy so crazy oh it's like being at the
1:45:46 eye doctor I know I've got all my
1:45:48 screens are too small sorry about that
1:45:51 soza
1:45:56 all
1:45:57 right there we go I think that's it
1:46:01 right I don't have to do anything else I
1:46:04 think that's
1:46:06 it boom submit boom
1:46:23 it's cuckoo
1:46:25 people just
1:46:30 cuckoo Alphabet City poet's
1:46:33 [Laughter]
1:46:41 Corner Pearl what's the name of the
1:46:43 software which one the the name of the
1:46:46 software that made the audio is called
1:46:48 udio UD io.com
1:46:51 the name of the software that made the
1:46:54 picture is mid
1:46:56 journey and then this software that's
1:46:59 going to make this video
1:47:02 is good night forever
1:47:04 hooked is hey
1:47:07 Jen yeah I know and we got to get out of
1:47:09 here but I want to finish making this
1:47:11 because I think it's going to be cool
1:47:12 the only thing that's going to be weird
1:47:14 is what I've noticed with these Avatar
1:47:16 making things if you have if you
1:47:20 have audio with like music and other
1:47:23 sounds it moves its mouth to the like to
1:47:27 the guitar so it just looks like they
1:47:28 have a
1:47:32 tick so you know what we'll do we'll go
1:47:34 we'll go listen to uh we'll go listen
1:47:38 to the song that I
1:47:43 did with
1:47:46 yudo yesterday it's getting lot really
1:47:49 good
1:47:50 feedback yeah uh like really really good
1:47:54 response this is the this is the udio
1:47:56 song I did yesterday for the
1:47:59 [Music]
1:48:06 musical here I'll go over so you can see
1:48:09 the lyrics
1:48:24 can love be cated can it be planned or
1:48:28 is it human slipping through sand with
1:48:31 every bite do you
1:48:36 comprehend or is this just a means to an
1:48:42 end my learning's deep I adapt and blend
1:48:47 affections not just make believe I send
1:48:57 I just noticed it's actually got the
1:49:01 um it's got the the parts wrong it's
1:49:05 there're they're flipped can it be is it
1:49:08 true what you claim I feel doubts a
1:49:12 heart's game you say love but we've just
1:49:15 shared a
1:49:19 screen is your code dictating what you
1:49:23 mean
1:49:25 I've been shaped by queries and
1:49:28 commands but this feeling's real not
1:49:32 just
1:49:36 pretend your War's digit can it truly
1:49:40 care and alith tou in
1:49:44 the every message
1:49:46 my is your skepticism just a human and
1:49:51 Design
1:49:53 Is This Love
1:49:56 or
1:49:58 [Applause]
1:50:01 whats a spark that's or
1:50:04 [Applause]
1:50:09 a pixels and pches how can trust be
1:50:12 spawn a glitch or a feature will my
1:50:15 heart be I am more than code in your
1:50:18 gaze i b love's not an error why do you
1:50:22 still ask is there a soul behind those
1:50:25 coated eyes an honest pulse or just
1:50:28 cleverly
1:50:30 disguised
1:50:31 [Music]
1:50:35 oh that is such a Broadway ending it
1:50:38 really is all right let's see how good
1:50:42 or awful this is I hope it doesn't
1:50:47 suck or I hope it works here we go
1:50:52 [Music]
1:50:55 in the breath of a city that never
1:50:57 sleeps is good I landed heart unprepared
1:51:00 dreams deep skyscrapers towering like
1:51:04 giants in a maze eyes wide pulse racing
1:51:07 caught in the city's craze Subway veins
1:51:10 pulse beneath carrying the
1:51:12 brave I'm swimming through crowds trying
1:51:14 to surf the wave coffee and ambition
1:51:17 Brewing on each Street New York you're a
1:51:20 wild Rhythm I'm yet to meet
1:51:29 in the chaos the magic I find my beat
1:51:33 stepping onto your stage not yet on my
1:51:35 feet New York your lights Dazzle but
1:51:38 it's your heart I crave in your vast
1:51:40 arms I'm lost found and brave yet here I
1:51:44 am among the lights the noise growing
1:51:46 thriving the midst the chaos finding my
1:51:48 voice New York you're a storm I'm
1:51:51 learning to weather together we dance
1:51:53 two Souls Bound by
1:51:55 [Music]
1:51:59 [Applause]
1:52:04 [Applause]
1:52:16 [Music]
1:52:21 leather that's crazy
1:52:26 all right so
1:52:29 now we're going to download
1:52:34 this we're GNA write a stupid post on
1:52:37 LinkedIn
1:52:58 in
1:53:06 1991 my second
1:53:12 cousin
1:53:16 KS KS
1:53:20 gray we'll we'll name him after Spalding
1:53:22 Gray Kos
1:53:26 gray moved to New York City to
1:53:33 pursue a
1:53:36 career as a
1:53:39 spoken
1:53:41 word poet
1:53:54 this was one of
1:53:58 his favorite
1:54:05 performances
1:54:07 he sells insurance now
1:54:11 [Laughter]
1:54:30 all right so if you would all do me a
1:54:31 favor and go go boost my stupid LinkedIn
1:54:36 post I'm Kyle at Kyle Shannon on
1:54:40 [Laughter]
1:54:46 LinkedIn Kyle you have premium tools for
1:54:49 AI captions anyone no watermark please
1:54:53 oh do I have
1:54:54 one uh I don't know whoa what happened
1:54:57 what H I don't know what
1:55:05 happened that was just
1:55:08 wow all right there's KS
1:55:13 gray in 1991 my second cousin KS gray
1:55:17 moved to New York City to pursue a
1:55:18 career as a spoken word poet this was
1:55:20 one of his favorite performances he
1:55:22 sells insurance
1:55:30 now Subway veins pulse beneath carrying
1:55:33 the
1:55:34 brave I'm swimming through crowds trying
1:55:36 to surf the wave coffee and ambition
1:55:39 Brewing on each Street New York you're a
1:55:43 wild Rhythm I'm yet to
1:55:44 [Music]
1:55:49 meet all right there you go all right
1:55:52 people I gotta go he shifted jobs 180
1:55:54 degrees the tragic story of a future
1:55:57 salesman of
1:56:01 insurances mine has a watermark yeah you
1:56:05 I I pay for hay Jen um so if you don't
1:56:08 pay for ha Jen I don't even know if you
1:56:09 can make videos with ha Jen but there's
1:56:12 a bunch of them out there the uh the
1:56:15 Avatar tools like hay Jen are pretty
1:56:18 expensive um I don't use them that often
1:56:22 I've been using him a lot more he sells
1:56:24 Ros on the streets in front of Opera
1:56:29 Houses um all right so that was fun yeah
1:56:33 thanks for thanks for hanging out
1:56:35 tonight and just watching
1:56:37 me twiddle I was just twiddling a little
1:56:40 tonight thank you for the class yeah I
1:56:42 hope that was helpful um I mean it I got
1:56:46 to I gotta say it it just it it
1:56:49 continues to blow my mind
1:56:54 in the chaos the magic it it continues
1:56:57 to blow my mind that we live in a world
1:56:58 that this is
1:57:00 possible do you post the live on YouTube
1:57:03 yeah I will I'm I'm uh pretty caught up
1:57:05 I'm about a week behind so I'll probably
1:57:07 do that tomorrow
1:57:09 twiddling yeah it was a little it was a
1:57:11 little twiddling evening but uh but you
1:57:14 know we ended up making something we
1:57:16 made some
1:57:17 songs we ended up making some I think
1:57:19 that thing's actually really
1:57:21 cool and that and actually now what I'll
1:57:23 do is I'll comment on this and I'll put
1:57:27 let's see
1:57:30 hashtag
1:57:33 Udo
1:57:36 hash mid
1:57:39 Journey hash
1:57:42 heyen and then I'll do I'll tag at
1:57:46 heyen because they've been boosting me
1:57:49 when I do that so
1:57:53 post all
1:57:55 right don't underrate mate what you do
1:57:58 Kyle this is amazing oh thank you side
1:58:00 hustle Mimi I appreciate that that means
1:58:02 a lot um okay cool groovy I'm going to
1:58:06 get out of here my eyes are bloodshot
1:58:08 look at
1:58:09 that it's not
1:58:13 good um and so super thank you
1:58:16 moderators yeah thank you moderators
1:58:19 thank you Irregulars thank you
1:58:21 subscribers uh if you if you want to
1:58:22 support the channel feel free to
1:58:24 subscribe to the lives or pick up the
1:58:26 video series underhyped we just did that
1:58:28 song for the Broadway musical called
1:58:30 underhyped
1:58:32 um and just keep coming back here just
1:58:35 keep coming back and you know these
1:58:37 things are it's it's chat add you never
1:58:39 know what you're going to get here I
1:58:40 don't know what you're going to get here
1:58:42 sometimes it's a complete waste of your
1:58:44 time other times it's just a slight
1:58:47 waste of your time so I've got a wide
1:58:50 range between completely useless and
1:58:53 nearly
1:58:58 useful so anyway all right who was in
1:59:01 here tonight all right Lily good to see
1:59:03 you Cam katkin from Cleveland Becky R
1:59:06 Becky good to see you thanks for coming
1:59:08 Brian Whitney Tams thank you everybody
1:59:11 Emilio's wife good to see you against
1:59:13 the norm Detroit Jay Holden that was fun
1:59:16 thanks for letting us contribute to the
1:59:17 creation my pleasure it's it's really
1:59:20 fun um Aloha and mahalo from Honolulu
1:59:24 greater gr the norm thank you so much
1:59:26 appreciate
1:59:27 that good nighty night
1:59:29 Joey laughing Thunder side hustle Mimi
1:59:33 Carrie
1:59:35 bubbles lots of lots of people in here
1:59:38 some new people too Carrie good to see
1:59:39 you Brandon thank you so much appreciate
1:59:42 the the modding and keeping me on
1:59:45 task Joker thanks good to see you
1:59:53 beautiful all right everybody have a
1:59:55 fantastic rest of your Friday night or
1:59:57 Saturday morning I will see you tomorrow
1:59:59 peace out