Mar 24, 2024 - How AI is Revolutionizing Musical Production

Description

Chapters

Transcript

0:01 [Music]
0:20 [Music]
0:29 oh
0:33 stand
0:35 between you in a hard place is
0:39 [Music]
0:42 insane stand too
0:45 near you in a fire makes it
0:50 [Music]
0:51 clear your trouble to
0:55 [Music]
0:59 me c just
1:03 say leaning in
1:07 close smell of your perfume scares
1:11 [Music]
1:13 most lean
1:15 [Music]
1:17 awayy it feels father every
1:24 day your trouble to
1:29 me
1:31 real trouble can't you
1:35 [Music]
1:56 see
1:59 w
2:01 [Music]
2:24 blue
2:26 telescope looking at the world tonight
2:30 room blue
2:32 telescope I wish I may wish I might not
2:36 see what I
2:37 see sheet met on sheets
2:42 [Music]
2:45 of looking through this blue
2:50 telescope down the moon struck a road
2:54 [Music]
2:59 tonight
3:02 [Music]
3:20 welcome everybody happy Sunday
3:23 [Music]
3:28 night welcome to the AI learning lab
3:30 where we're going to
3:33 learn or
3:36 [Laughter]
3:38 not this could be a complete waste of
3:41 everyone's
3:42 [Music]
3:44 time although champ I think gets some
3:47 something out of it champy digs
3:48 [Music]
3:59 it
4:01 [Music]
4:20 does he work the
4:22 [Music]
4:25 queen does he listen to rock and roll
4:28 does he feed the mean singing h
4:31 hallelujah
4:33 hallelujah
4:35 hallelujah
4:40 [Music]
4:47 hallelujah champ is so into it
4:50 tonight maybe Sydney's musical needs a
4:52 singing dog that's not a bad
4:54 [Music]
4:59 call
5:04 all right enough enough comedy jokes
5:06 enough it's to you
5:10 too hey zip it zip it zip it zip
5:17 it all right rough rough
5:22 exactly all right
5:24 everybody welcome welcome to the lab
5:27 show us the puppy all right puppy look
5:29 up wait
5:31 chippy look chippy look
5:39 up yeah he's a little
5:42 cutie he looks like the RCA Victor dog
5:46 if you're if you're old and remember
5:47 what RCA Victor was or if you've ever
5:50 seen a an
5:52 [Laughter]
5:57 encyclopedia Cindy [ __ ] what's happening
6:01 what is shaking what's going
6:04 down um what do we got going on tonight
6:09 I um I had a meeting with with the guy
6:13 that I think I'm going to co-write the
6:15 musical with
6:19 today he and I he was in the last the
6:23 last one of the last plays I ever did in
6:26 New York City was called The Amazing
6:28 Adventures of tens sky
6:31 and Andrew played tens Sky I played the
6:34 man who overreacts
6:38 everything um and neither he nor I have
6:42 done anything in theater for like 30
6:45 years like not even thinking about
6:49 it so so he was a little bit like
6:54 uh why'd you reach out to me and what's
6:57 going
6:59 on
7:01 and then I just told him about all the
7:03 you know all of what happened here and I
7:05 said you know here's 4 days worth of
7:08 work and he was he was into it
7:11 so anyway I thought that I thought there
7:14 could be something interesting we could
7:16 play with tonight there's there's one
7:17 song I cannot get it to write non shitty
7:21 lyrics um I thought it might be cool to
7:23 try different tools but whatever we can
7:25 do anything we can also play with mid
7:27 Journey I've got access now to the the
7:30 the web version of mid Journey we could
7:31 play with
7:32 that um I can also just answer questions
7:35 cat GPT has a new edition are you Weare
7:38 no what's the new edition is it if it's
7:41 memory that's a huge [ __ ] deal but I
7:43 didn't see anything
7:45 new let me do a new
7:48 chat nothing there oh they got rid of
7:51 plugins in the uh in the chat GPT window
7:55 that's interesting which version you're
7:57 on customize chat GP T that looks the
8:02 same
8:06 yeah my
8:08 plan I don't know what the addition
8:11 is I just got a new puppy nice Brandon
8:14 we've been having fun with custom
8:17 GPT what is my dog
8:22 thinking 4 days worth of 2024 work I
8:25 know exactly by the way Emilio's wife
8:28 did you notice that that's your image
8:31 got Emilio's wife she's got a
8:32 contribution in the uh the little hero
8:34 image here today huh she's on time and
8:39 she made artwork for the show holy
8:42 crap no demerits on her report card
8:45 today she's starting the week off
8:49 fresh I like it I like
8:52 it fanastic Bob tell him what he's won
8:57 uh you haven't won anything uh this
8:59 isn't the game show okay okay people all
9:05 right what are we going to learn tonight
9:07 what are we going to
9:09 learn yongma you how's the night Kyle
9:13 it's doing pretty good it's doing pretty
9:15 good welcome back it's I I like when I
9:17 see people here multiple nights in a row
9:21 it it it's it War it warms my heart
9:25 warms my
9:27 heart I I I think one of two more times
9:30 ye and uh you're you're going to be an
9:33 official
9:36 irregular someone asked that one night
9:38 how can I become an irregular just show
9:42 up that's [ __ ] weird if you watch
9:45 this every
9:46 night you're not
9:48 [Laughter]
9:52 regular um okay so anyway if you're new
9:55 here my name is Kyle Shannon this is the
9:57 AI learning lab I'm happy to answer
10:00 questions about AI I'm happy to demo AI
10:02 I'm happy to get into it if you're an AI
10:05 um
10:07 Doomer um if you want to if you want to
10:10 debate I'm happy to do that I don't like
10:13 doing that I like I like interacting
10:16 with people that are into this but
10:18 sometimes people come in with some fear
10:19 or some resentment or anger about this
10:22 stuff um but actually want to engage and
10:24 that I'm I'm super down to do I've been
10:27 here for a couple of months look forward
10:28 to these streams oh nice oh that's nice
10:30 I didn't I didn't know you were here for
10:32 that long cool welcome then you're
10:34 definitely in a regular good Lord you
10:36 are you are well beyond the
10:39 threshold you've been commenting more I
10:42 suppose um which I like you ask really
10:44 good questions so keep keep doing that
10:47 might have to do it the
10:51 voice Brian Whitney's here good day
10:53 rustic Farm wife haven't seen you in a
10:56 bit just showing up
10:58 Emily
11:02 participation Awards it's yeah this is
11:05 definitely you get a participation
11:07 trophy in this channel for
11:09 sure I am generally not a fan of those
11:11 in life but you know that's pretty much
11:13 what this channel is all about my wife
11:16 called me in a regular long before this
11:18 group you're my
11:23 people oh man very cool I'm going to buy
11:27 a sewing machine I shall be back okay
11:31 fantastic that is that is the
11:33 appropriate reason to check out for a
11:35 little bit you know you're here you're
11:36 like okay I'm here I'm checking in on AI
11:38 hang on one sec I got to go buy a sewing
11:40 machine and I'll be right back perfect
11:44 perfect perfect
11:46 interaction have you been asked if you
11:48 are an AI numerous times um the AI don't
11:54 quite have enough personality to deal
11:55 with my you know my uh sassy ways but
11:59 I'm working on a video script for work
12:01 very cool very
12:05 cool he has a world ID he's the real
12:11 deal I think it's Sora What in in chat
12:15 GPT no way that the I was on Twitter
12:19 right before this Twitter would be
12:20 blowing up if Sora was in chat
12:24 GPT there's there's nothing in there
12:27 here I can I can prove that
12:32 I mean I can test that let's
12:34 see make a video of
12:40 a
12:43 hedgehog um with
12:46 [Laughter]
12:49 gas I can't create or generate videos
12:51 it's not in
12:53 there but we'll delete that
12:58 chat
13:12 ah here's my my set design for my
13:15 Broadway
13:17 show anybody know any any Broadway
13:19 producers want to produce my
13:24 [Music]
13:27 play do you think Princess Kate's video
13:31 was AI I don't think so it seemed it
13:33 seemed pretty it seemed pretty like
13:36 people are people like the one video I
13:39 saw like someone had zoomed into like
13:40 her hairline and there were like little
13:42 specks and it was just like little
13:44 flicks of her hair blowing in the wind
13:47 so that's that seemed to make it more
13:49 legitimate to me than less but whatever
13:52 I don't I here's the deal if for
13:55 whatever reason they're not tra trotting
13:57 out Kate and they're deep faking her to
14:02 I like I don't give a [ __ ] just I just
14:06 don't I do not
14:08 care um there are other deep fakes that
14:12 I think would be more interesting but
14:14 who did someone got deep faked oh Bobby
14:17 Al Al alof or whatever her name is the
14:20 the
14:20 Blogger but
14:23 again just expect 2024 it we're going to
14:27 see more and more and more of this [ __ ]
14:29 I called that in 2023 2024 is going to
14:32 be the year of the deep fake and then at
14:34 some point someone's going to do a deep
14:36 fake and it's actually going to cause
14:38 some har or damage and then everyone's
14:39 going to get up in arms oh my God we got
14:41 to do something about this and then
14:44 someone will say if there if only there
14:45 are a way we could verify if someone's a
14:47 human or
14:49 not and then we can get out our world
14:51 IDs and go you mean like
14:55 this oh good lord people good lordy
14:59 lordy
15:01 lordy let's see do I have anything going
15:04 on
15:06 here I saw a new thing called
15:10 vigle which was a like a
15:13 a
15:15 video like person replacement sort of
15:19 thing but you can only do it on Discord
15:20 and it wasn't super clear
15:24 so
15:26 uh when I lived in it when when I lived
15:30 in it there was an option I think it's
15:31 in
15:32 production crypto G what are you talking
15:35 about are you sure you're in actual chat
15:39 GPT
15:41 because there are apps that say on Tik
15:45 Tok it it's it's worse than this you can
15:48 make money with
15:50 ch um my son had chat GPT create a
15:54 machine learning course and now he
15:56 listens instead of reading oh nice
15:59 that's awesome um yeah there there's a
16:02 lot of there's a lot of
16:04 scammer apps that are saying Sora video
16:08 is now available and then you go to it
16:10 and it's just some
16:12 shitty app based on Runway or one of the
16:14 other probably stable video diffusion
16:17 but they're calling it Sora and they're
16:18 putting it on on Tik Tok if if Sora were
16:21 out everyone would be talking about it
16:23 it's not out and it's not gonna it's not
16:25 going to be I don't think Sora is going
16:27 to be a big consumer kind of thing right
16:30 off the bat because it's apparently
16:32 going to take some time to generate
16:34 videos and it's going to be expensive
16:37 and they said they're not going to
16:38 release it until they get it kind of
16:41 cheap enough to do 20 bucks a month like
16:44 Dolly um but we'll see it's but it's not
16:47 out yet it is not out yet I have a
16:50 feeling you're you're playing with an
16:52 app if you're playing with an app it's
16:54 it's definitely not chat GPT
16:57 official un listen I would love to stand
17:00 to be corrected but I just I assume
17:03 that's what that
17:04 is she has a mole on her face that
17:07 wasn't there in the video oh no Andrea
17:10 Andrea's Andrea is in the cake
17:12 conspiracy theory too I listen
17:15 maybe I don't know chat GPT is not
17:19 listing my gpts in the sidebar really
17:22 hang
17:23 on M are still here there's
17:28 theology there's my C
17:32 profile but I do notice that there's
17:35 different
17:38 ones
17:39 um if I go to my
17:44 gpts let's see do an
17:46 [Music]
17:58 explorer I got to get rid of some of the
18:00 I have so many trash gpts I just got to
18:02 go get rid of
18:06 some oops
18:11 sorry the plant Whisperer here I'll
18:14 throw that one up
18:16 there so how do I do I add this keep in
18:20 sidebar so it's still there but it does
18:22 look like it it these are not the same
18:24 so they're doing
18:26 something but I just don't add a team
18:30 workspace
18:33 settings speech looks the same beta
18:36 features plugins are still turned on but
18:39 you can't get to them now which I think
18:41 is really interesting Builder profile
18:44 looks the
18:47 same look up here so if you go here this
18:52 used to have plugins right here they're
18:54 gone so plugins are gone I don't know
18:55 how do you get to
18:57 them which is interesting
19:01 so that's different but I think that's
19:03 it I don't see anything new I don't know
19:05 what y'all are talking about tell me
19:07 what's going on people what have you
19:09 discovered you can make
19:15 money late but still in a regular yes Mr
19:18 it welcome welcome
19:23 welcome I think I'm just going to like
19:26 saying vigle I know me too he vles and
19:29 vles and
19:31 vles sounds like an episode of Hogan's
19:35 [Laughter]
19:40 Heroes oh I did hear for those of you
19:44 interested in the
19:45 rabbit not that rabbit the rabbit
19:49 R1 I did hear they are still on track
19:52 for
19:53 shipping by March 31st which is the end
19:56 of this
19:57 week um
19:59 and they will be arriving to the US
20:02 customers like the third week of April
20:05 so apparently we're still on track for
20:07 getting our rabbit R1 so I can I can
20:10 demo it and we'll use it twice and then
20:13 we'll stick it on the shelf right there
20:16 see that little spot in front of the
20:18 urban desires sidequest Drive perfect
20:21 for a rabbit
20:25 R1 all right what do you guys want to
20:27 talk about ask questions tell me what's
20:29 going
20:30 on they got me
20:43 to first
20:50 batch
20:52 [Laughter]
20:57 R1 about the oh one thing have you
20:59 talked about that yeah I I actually
21:01 ordered one um only because I'm an
21:08 idiot but I kind of feel
21:11 like I kind of feel like the 01 thing is
21:14 the first open source one and the the R1
21:17 thing is the first one that had
21:19 interesting design and some interesting
21:21 features but they both do the same
21:24 thing um so 01 is a device similar to
21:30 the rabbit but it's it doesn't have a
21:31 screen or a scroll wheel or anything
21:33 it's just a button you talk into and
21:35 then it it does queries for you and you
21:38 can you can train it on applications on
21:41 your desktop so it's kind of kind of
21:43 cool and interesting it's it's an
21:45 interesting class of devices
21:48 um you know I
21:50 suspect two or three years from now
21:53 we'll figure out what that category
21:54 actually is right now I don't think we I
21:57 don't think we have it but but
22:01 whatever this is this is why you play in
22:04 the early days right I had an early
22:06 Newton I had an early Blackberry I had a
22:09 Palm Pilot I don't know if you remember
22:11 Palm Pilots Palm Pilots were supposed to
22:14 be like the salvation
22:16 of of the uh the apple one
22:24 so what's a rabbit R1 here I can go show
22:28 you that um um um it's a walkie-talkie
22:30 it's pretty much a walkie-talkie let's
22:33 go we'll go look at
22:41 that
22:44 rabbit
22:47 R1 pre-order
22:50 now so that is a rabbit
22:55 R1 hang
22:57 on
23:01 and that's kind of what it looks like up
23:04 close so it's got this cool scroll wheel
23:06 thingy and a push button and a
23:10 screen and so when you push the button
23:12 you're basically invoking chat GPT it's
23:15 got also got a cool rotating uh camera
23:19 so when the when you're not using the
23:21 camera it rotates up vertically so it's
23:24 for privacy which is super cool as
23:26 opposed to our phones which do not do
23:29 that there it
23:31 is
23:33 um and so you you basically just hit the
23:36 little gray button and say you know you
23:39 can ask it about current events or you
23:41 can ask it to write you a poem or
23:43 whatever and it just it
23:46 does uh you know large language model
23:50 [ __ ] I think it's it comes pre-connected
23:52 to
23:53 perplexity um but one of the cool things
23:56 you can do is you can say things like
23:58 hey send an email
23:59 email um
24:02 to Andrew about the play and it will
24:06 write it and send it and it because it
24:08 it can be connected to your own
24:10 application suite and it they've got
24:12 these these sort of macros preset up
24:16 that you can have it do stuff like hey
24:19 go find some interesting live music
24:21 tonight and email that to my my friend
24:23 group and it will know who your friend
24:26 group is and it'll know how to send
24:28 emails and [ __ ] like
24:29 that so and then there's you know
24:31 there's lots of questions like do we
24:33 need another device and the reason I'm
24:35 the the reason I bought one of these it
24:37 was only 200 bucks which is it's not
24:40 cheap but for a device like this uh like
24:43 the uh the pin the thing from whatever
24:47 the [ __ ] that company's called the
24:49 little pin you put on that was $700 and
24:52 you need a subscription this is 200
24:54 bucks without a subscription and it's
24:55 got really slick kind of retro design it
24:58 looks like a little walkie-talkie so
25:00 it's cool um the teenage engineering is
25:03 the one who did the Industrial design on
25:05 this and they do music synthesizers the
25:08 the founder of rabbit um had built a
25:12 music synth and so that's that's where
25:15 the the uh Teenage engineering
25:18 connection comes in he built he had
25:19 built a previous uh device so anyway
25:23 that's it pretty slick huh perplexity
25:26 says iOS 18 may include anthr topic yeah
25:29 that doesn't surprise me yeah I saw I
25:32 saw that right before I came on Vicky I
25:33 didn't quite get what the what the uh
25:37 what they were talking
25:39 about let's
25:42 see
25:46 perplexity iOS
25:54 18 iOS 18
25:57 breaking
26:05 if that was February that's
26:25 old exciting rumors about iOS 18 reports
26:28 suggest apple is gearing up to
26:30 revolutionize AI
26:32 integration with potential
26:33 collaborations featuring anthropic open
26:36 AI
26:37 Google Claude 3 from anthropic could be
26:40 a game changer stay tuned yeah so that
26:42 doesn't surprise me that cuz you know
26:47 rumor was they were they they were
26:48 already talking with open AI or they
26:50 they had talked to open Ai and that they
26:52 were also talking with Google about
26:54 Gemini and like oh it's going to have
26:56 Google it actually makes more sense to
26:59 me that they've cut deals with with
27:01 multiple of the big
27:03 players and and again what I think this
27:06 gives them cover to do is they can just
27:09 focus on siries actions like what Siri
27:14 can do on your behalf and let these
27:17 thirdparty large language models do all
27:19 the hallucinating so so from a
27:22 positioning standpoint what they can say
27:24 is we're going to innovate over here
27:27 where we know it's bu proof and then
27:30 anything that's like large language
27:32 modely go write a poem go go use one of
27:35 these leading providers
27:38 right let's see I expected to heavily
27:41 integrate artificial intelligence
27:42 potentially in collaboration with AI
27:44 firms like yep Cloud 3's Opus
27:48 model yeah uh enhance various iPhone
27:51 applications making significant yeah
27:53 yeah yeah all right this this makes
27:55 sense to me th this makes more sense
27:56 than them picking a single player
27:59 so so built into iOS 18 Siri will be
28:04 better but will still be task oriented
28:08 right order me a this you know check the
28:11 check the weather like the the
28:13 interactions with the with iOS and then
28:16 anything that's sort of pure
28:19 hallucinogenic large language model [ __ ]
28:21 let these guys do that and then when
28:26 Apple introduces their own large
28:28 language model [ __ ] then then they'll
28:31 say Hey you can use all of these but you
28:33 can also use ours which just destroys
28:36 them and in terms of whatever and
28:38 that'll probably be a year later
28:40 generally how they work so that makes
28:42 sense to me that's pretty cool pretty
28:45 cool I like it I like it it's a good
28:48 rumor that's a good rumor solid
28:53 solid what is open AI voice engine uh
28:58 just a rumor at this point I think it
29:00 was mentioned by Alman and something
29:02 let's see
29:05 open uh
29:10 open
29:17 Voice open AI will
29:20 release this year the best personal
29:27 assistant
29:29 nothing will compare to this thing it
29:30 will make Jarvis Samantha and Hal seem
29:33 like something from the past a good
29:35 function calling model does all they
29:37 need so who's this
29:49 uh smoke away I don't know who this this
29:52 guy or gal is smoke away but this is
29:56 always Nostalgia for the future 34,000
30:02 followers
30:04 um voice engine speculation relatively
30:08 unlimited voice
30:10 customization behavioral
30:12 customization very low latency response
30:15 time ability to interrupt and be
30:17 interrupted more natural voice patterns
30:19 pauses and tones and we're that we can
30:22 see that we can play with right
30:24 now ability for the assistant to lead
30:27 the conversation that would be [ __ ]
30:29 huge
30:30 right if the if the if the agent can
30:33 lead the conversation cuz right now this
30:35 is one of the problems with with chat
30:38 GPT right now is that it kind of assumes
30:42 you know what you're doing with a large
30:44 language model and most people don't
30:49 most people don't even know what the
30:50 [ __ ] cat GPT
30:52 is so that would be nice when did he
30:54 post this how old is this March 24th all
30:57 right so this is recent recent
31:00 stuff open AI voice engine blog next
31:04 week yes 41% all right so that's
31:07 speculation there's some speculation for
31:11 you smoke away voice engine is
31:14 Big Mark for voice engine
31:18 trademark is intended to cover voice and
31:21 speech recognition processing voice
31:23 commands converting text and speech
31:26 automatic speech and voice recognition
31:27 and generation ation creating and
31:31 generating voice audio outputs building
31:34 digital voice assistance all right so
31:36 this so the trademark just basically
31:38 says anything with words and voices is
31:41 you know covered by voice engine
31:43 trademark all
31:45 right that's the way those things
31:47 generally
31:48 go open AI custom voice engine blog
31:52 coming
31:54 soon snippet custom voice engine demo
32:05 unsupervised sentim sentiment neuron
32:13 tables new voice engine blog post
32:16 incoming unsure if this is related to
32:18 open AI voice engine but it's one of the
32:21 scripts that gets loaded into the
32:24 site I guess Kathy finished Jarvis in
32:33 February building a kind of Jarvis at
32:36 open
32:40 AI this is karpathy back in
32:43 February I left open AI yesterday first
32:46 of all nothing
32:48 happened actually being an open AI my
32:51 immediate plan wait let's see team is
32:54 really strong and wonderful the road map
32:55 is very exciting I think we all have a
32:57 lot to look forward to all right nothing
33:01 nothing super thing but um Sam mman said
33:06 that that um GPT 5 is probably not
33:11 coming out till later in the
33:13 year and it may not be called that but
33:15 that there are going to be a lot of
33:17 things they'll be announcing between now
33:19 and then so yeah cool all right good
33:24 good uh good chat
33:26 TMZ wait a speculate people P's here
33:30 Pate Pate works at Google Pate loves it
33:34 when we speculate in
33:36 here so Pate not only will voice engine
33:39 come out this week it's also going to be
33:43 magic it's not here's the thing about
33:46 voice engine pay they don't even use
33:49 code it's what happens is you you you
33:52 push a button in chat GPT and um it uses
33:58 uh
33:59 telephony and and it it it calls open AI
34:02 headquarters and just someone random
34:04 there just answers the phone it's like
34:08 magic it's like it's not buggy cuz it's
34:10 just people talking like on a
34:14 phone and what they're going to also
34:17 innovate is you can you can hardwire a
34:20 direct connection into your house so you
34:22 have a thing you put on the wall with
34:24 like a like a handset like you pick up
34:27 this thing and you listen in one part
34:30 and you talk in the other part you hold
34:31 it like this and it's wired to the wall
34:33 so it it never goes buggy it's going to
34:36 be it's I'm telling you it's
34:38 magic it's math
34:42 magic that sounds infinitely buggy yeah
34:46 humans that's the funny thing about
34:47 people going you know uh chbt
34:50 hallucinates you know what hallucinates
34:52 right I'm like you know who else
34:53 hallucinates [ __ ] humans humans are
34:57 idiots
34:58 are you are you are are we really
35:00 comparing chat gbt to humans cuz cuz
35:04 humans are
35:06 worse they lie they have ulterior
35:09 motives at least chat gbt just
35:14 sucks does it have a rotary dial pad
35:17 that's that's an upsell the the rotary
35:19 dial uh to talk to the people that's uh
35:22 that's the uh that's the Plus
35:24 subscription it's 20 bucks a month to
35:27 get the rotary pad
35:34 do you think your context Windows 8,000
35:37 tokens I doubt it exactly [ __ ] with
35:39 add my token Windows about 38
35:43 tokens it's I average about 24 words
35:46 between before the conversation exits
35:48 the
35:52 building usually mid sentence it's just
35:55 like gone
35:58 it only has one sound dial tone
36:02 exactly uh hey Kyle where do you learn
36:05 wait what was that where do you learn
36:07 about AI any good YouTube channels or
36:10 podcasts yes so um I I'll go show you
36:14 that
36:18 um podcast um Paul rer the marketing AI
36:23 Institute Paul rater and Mike kaput I
36:25 think they changed the name of it I
36:26 think it's just called the artificial
36:29 intelligence show now but it's the the
36:31 publisher the marketing AI
36:34 Institute um on YouTube David Shapiro is
36:37 really good um he's a he's he's a
36:41 developer but he's also um working on an
36:45 AGI an open- Source AGI framework a
36:49 cognitive architecture framework that
36:51 he's open sourcing he's wicked smart he
36:54 does like once or twice a week he'll do
36:57 like a big thoughtful presentation on
37:00 the future of everything we're dealing
37:02 with and he kind of analyzes big
37:04 movements in in the industry he's really
37:07 good
37:08 um who else Rachel Woods on Tik Tok if
37:12 you're into like sort of integrating AI
37:14 into businesses she's really into AI Ops
37:18 AI
37:19 operations uh and then the other one who
37:22 I follow a lot is Ethan mik M o l i c
37:26 k uh um and he is uh A Wharton business
37:32 school Professor he's got a Blog called
37:35 one useful thing. org um he's worth
37:38 following and then on Twitter Robert
37:41 scoel is good he's he's a San Francisco
37:45 guy he's been like a he's been like an
37:47 apple Whisperer for years and years and
37:49 years and he's doing a lot with AI right
37:51 now and he he's just he tends to be up
37:53 on stuff um so if he's talking about it
37:58 generally something's go going
38:01 on with it and
38:04 that's like that's that group of
38:09 people there's there's always lots and
38:11 lots of rumors and so if I see like a
38:13 wacky rumor like you know whatever like
38:16 this open voice thing I'll go check all
38:19 those people and if they're not talking
38:21 about it generally it's not it's not
38:23 happening yet if a lot of times I'll
38:25 learn stuff from them okay I'm home now
38:28 when I go to open
38:30 AI login page introducing Sora creating
38:34 video from
38:38 text yeah but that's that's just the
38:41 announcement from from a couple of month
38:44 and a half ago I think um open ai.com
38:50 so that's SL Sora but if you just go to
38:53 regular open
38:56 AI yeah this is this is just the
39:00 announcement transforming work in
39:04 creativity yep that's just the
39:07 announcement can't be any worse than
39:09 Google Voice yeah
39:11 exactly but it but also like it you know
39:14 it's it's going to be it's going to be
39:16 good because their current their current
39:18 voice to text is really good what if if
39:21 the if the new voice stuff takes actions
39:24 if it's actually an agent that can do
39:26 stuff for you then it's a huge
39:28 deal Kyle are you excited about the AI
39:32 powered
39:34 nitrogen capsule for quick exit major
39:37 laser that's
39:42 good I did hear about this was a
39:46 disturbing thing I
39:49 heard I don't know what it was but this
39:51 guy this guy
39:54 basically
39:56 liked putting extra citric acid on his
39:59 chewing gum and then he and so he had a
40:02 powder of that on his desk and he also
40:03 had powder of this highly highly
40:05 explosive stuff and he mixed them up so
40:08 so he coated his gum in the highly
40:10 explosive stuff and chewed it and it it
40:12 it did him in so I forget where I saw
40:15 that I don't know what it was related to
40:18 [Laughter]
40:20 but [ __ ] internet's weird man yeah
40:23 it's weird
40:26 man d
40:30 [Music]
40:33 Irregulars channel for those who are
40:35 under 50 years
40:38 old this should be
40:40 good AI
40:47 Irregulars if you wondered what what we
40:49 were talking about before about a rotary
40:52 input
40:56 device that
40:58 that is what US gen xers grew up with
41:01 there was that version of it that sat
41:04 next to the lamp in the room where your
41:08 TV was and then there was a version of
41:12 it in the kitchen where it was on the
41:14 wall and if you really had it going on
41:17 like if you were a baller kind of
41:19 Technology family you'd have like a 25
41:22 ft cord and and so you could take the
41:25 call in the kitchen and then if it was
41:27 one of your friends or your like
41:28 girlfriend you could walk around the
41:30 corner and close yourself into a closet
41:33 talking
41:39 there that's really
41:44 funny um by the way where I am right now
41:48 I'm at the AI learning lab Irregulars
41:51 club we have our own private club at the
41:53 AI salon so if you go if you're
41:55 interested in joining a community of
41:57 people that don't
42:00 suck go there the salon. and click on
42:04 the first link that says AI Salon
42:06 community and go check out what we're
42:08 about and what our values are and join
42:11 this community it doesn't cost anything
42:14 and it's it's basically just a group of
42:16 of people that are trying to figure this
42:17 ai ai stuff out together and there's
42:20 there's guilds and different
42:22 subcommunities in there all sorts of
42:23 stuff like that Pate 35 grew up with a
42:26 rotary rotary phone get off my
42:30 lawn yeah but you're right on the edge
42:32 man there's I guarantee you you work
42:35 with some people at the goog that have
42:37 no [ __ ] idea what those things
42:40 are my relatives work for Bell Telephone
42:43 so we have quite a few of these and the
42:45 princess phones that's
42:48 awesome chat GPT made me log in
42:51 again I didn't see the customized chat
42:54 GPT
42:56 before
42:58 huh that's in the customized chat GPT
43:01 that's just a new thing for it was
43:03 called custom instructions
43:06 before they changed that I don't know
43:09 week or two ago customized
43:11 GPT yeah they just they just renamed
43:17 that 25 25 foot cord was very expensive
43:21 $20 at service merchandiser exactly
43:24 that's what I said you had to be kind of
43:26 loaded to have the 25 foot cord you know
43:30 the real rich people had phones in every
43:32 room that was that was just you know
43:34 that was you you were flexing if you had
43:37 phones in every
43:38 [Laughter]
43:41 room there's generally two phones a
43:44 kitchen and a living room and that was
43:47 it and you go to the
43:53 phone oh man the race home to wait by
43:57 the phone to hear it ring exactly how
44:00 did we get buy back then I know honest
44:02 to God seriously like we actually had to
44:04 remember numbers in our
44:07 [Laughter]
44:11 heads
44:15 oh I had to see through the cordless the
44:19 the see-through cordless phone from
44:20 Radio Shack that's a solid that's a
44:22 solid
44:23 memory I upgraded an upgraded version um
44:27 my family was too poor to justify paying
44:30 for
44:35 touchone the last thing I knew still was
44:38 was rotary that's
44:43 awesome that's right you had to upgrade
44:45 the touch tone right it was like an
44:46 extra $2 a month
44:49 [Laughter]
44:54 for that's [ __ ]
44:56 hilarious custom chat EPT is in the
44:59 settings it's permanent custom
45:01 instructions I think that's right Vicky
45:03 yeah I think it's permanent custom
45:05 instructions but it's been it's been
45:06 there for a
45:08 while it's been there for a while this
45:10 thing it's been
45:16 there calling long distance was
45:18 expensive oh that's right yeah oh but so
45:21 and so lives in Florida we can't call
45:26 them in Africa we still use those there
45:29 you
45:30 go let's here you
45:34 go Still Still Hey listen it's still
45:37 functional
45:39 technology although just you know try to
45:41 get someone who hasn't seen one before
45:42 to figure out how to use it that's
45:44 pretty funny there's all those Tik Tok
45:46 videos where there's that that one dude
45:48 who's got his daughter is like 16 and he
45:50 just shows her all the Gen X [ __ ] like
45:52 you know he'll show her one of these and
45:54 say make a
45:56 call
45:59 oh that's really
46:04 [Music]
46:11 funny I used to call long distance and
46:14 ask for my middle name so my mom knew I
46:17 was alive but didn't have to pay the
46:19 bill that's
46:21 awesome scroll up on IR regulars
46:26 now
46:28 the converter oh yeah that's
46:33 great yeah there's the wall
46:38 version Oh those are awesome this is
46:41 what this is what um there was also
46:45 remember the intercom systems where in
46:47 new houses they' put in the the intercom
46:50 systems between rooms and and you'd go
46:52 to the wall and you'd push the button
46:54 and say hey so and so are you there and
46:57 then no one ever used them cuz they
46:58 would just scream up the stairs
47:05 Sally oh man ah
47:09 technology hold please
47:12 exactly the days of the Walkman how I
47:15 missed the 80s intercom was fancy I know
47:18 I remember we moved into a new house
47:20 when I was probably in junior high and
47:23 it was in a new development and they put
47:26 in intercoms were part of the deal and
47:28 you know how many times we used the
47:30 intercom well once the day we moved into
47:33 the
47:34 house and then we never used it again oh
47:39 I don't remember intercom systems only
47:41 the rich people had
47:46 them oh man that's hilar
47:51 hilarious um my parents would yell at
47:54 the boys to get their girlfriends out of
47:56 the room you intercom was fancy it was
48:00 intercom was M oh you've got an intercom
48:03 wow I didn't realize you were doing that
48:05 well what was that there's a comedian
48:08 that who the [ __ ] was
48:11 it he was just talking about his bit his
48:14 bit was about pillows was it Tom Papa
48:16 might have been Tom Papa where where
48:18 he's talking about what is it with
48:21 pillows like the amount of pillows on
48:22 your bed um indicate how well you're
48:25 doing you walk into the bedroom it's all
48:27 pillows like oh I didn't realize you
48:29 were doing that
48:36 well oh that's
48:39 [Laughter]
48:44 good my grandma's Town only had had to
48:47 dial Five numbers oh yeah yeah cuz cuz
48:50 it was you you had a a letter that was
48:52 the beginning number that represented
48:54 like the the the ZIP code right and then
48:57 it was four numbers after
48:59 that oh man we don't have our phones if
49:03 we don't have when we don't have our
49:04 phones we're phoneless and lost yeah
49:05 because the phones now are not just
49:07 phones right they're they are they there
49:10 are connection to all of society right
49:12 now it's
49:16 amazing you know we kind of take for
49:20 granted we take for granted how how very
49:24 short amount of time it's
49:26 been that we effectively have
49:30 hyperconnected supercomputers in our
49:33 pockets
49:35 right cuz it weren't that long ago where
49:39 if you watched a movie the first thing
49:42 you had to do when it stopped was rewind
49:45 the tape before you took it back to
49:48 Blockbuster that wasn't that long ago
49:51 people there's a store by my house that
49:53 still has a pay phone outside it doesn't
49:55 work though exactly well it's funny in
49:57 New York City there's all the all the
49:59 pay phone booths are still there and
50:01 they they occasionally put other [ __ ] in
50:03 them but but like all the phones are
50:06 gone but those used to be there for
50:08 years and they were just not
50:10 functional created a ton of images today
50:13 i' had so much fun that when you get
50:15 when you go into those image creation
50:17 rabbit holes it's so much fun or in the
50:19 movies yeah Murray Hill and then the
50:21 number yeah get me Murray Hill
50:23 652
50:26 6529
50:28 have we discussed Devon yet not tonight
50:31 um Melrose 3686 still remember my
50:34 grandmother's number that's
50:36 hilarious um we haven't discussed Devon
50:39 we can do that the Jetson would talk to
50:41 each other through a TV now that's our
50:43 phones I
50:44 know and now you
50:47 know this is [ __ ]
50:50 wild like right so so FaceTime FaceTime
50:53 in and of itself is a Star Trek
50:55 technology right so we now
50:58 have like the Star Trek [ __ ]
51:01 right and we just we just talk to these
51:04 devices like it's normal we forgot like
51:06 that'll never happen you'll never be
51:08 able to talk into a phone and see
51:11 someone and like probably by the end of
51:14 the year we'll be able to have
51:16 conversations with virtual versions of
51:23 people like talk to my virtual Avatar
51:26 and schedule a
51:31 meeting oh my God it's
51:36 crazy it's crazy it's crazy where we
51:40 are um Devin that was just
51:43 mentioned beat me up Scotty dude it
51:45 wasn't that long ago to to see a movie
51:49 meant movie theater
51:51 only
51:55 exactly yeah
51:59 I mean it's a pretty remarkable thing
52:01 that when Co
52:04 happened the movie industry had to
52:10 manufacture demand to get us out of our
52:14 houses right because Co all distribution
52:18 went streaming right but there's like
52:22 this whole infrastructure for movie
52:24 theaters and Hollywood is like H [ __ ] we
52:27 like we're about the experience let's
52:30 get people back in those buildings we
52:32 own a bunch of so so we're not going to
52:36 release this you know on streaming
52:39 that's why this like all movie releases
52:41 right now are so weird and fractured
52:43 because they're weird and fractured it's
52:49 crazy well my virtual Avatar with
52:52 virtual autism also look at their shoes
52:56 when talking to another
52:58 Avatar that's how you're going to be
53:00 able to tell
53:02 pay if you if you talk to someone that
53:05 you know can't make eye contact and
53:07 their avatars actually looking at you
53:09 you're like no that's definitely not
53:11 pain that's
53:15 hilarious it's the little
53:19 Clues hey he he actually just made eye
53:22 contact with me that's weird P are you
53:25 okay yes I'm
53:30 fine that's [ __ ]
53:33 awesome oh oh right party lines we also
53:36 had party lines if you don't know what
53:38 party lines were you would pick up your
53:41 phone to call someone and there would be
53:42 other people talking on
53:44 it they were they would be talking to
53:47 each other and you would just pick up
53:48 and now you would be in their
53:50 conversation it was kind of like
53:52 Facebook
53:55 roulette
53:58 without the
53:59 faces oh remember that girl in Inspector
54:02 Gadget that had the watch she could talk
54:03 through I know I'm telling you man like
54:08 it's it's weird how quickly we normalize
54:11 stuff but like we're living in
54:13 remarkable times I'll tell you when I
54:15 that my
54:17 first my first
54:20 um my first kind of visceral reaction to
54:23 chat
54:25 GPT was like I immediately went to
54:28 science fiction movies and I was like oh
54:32 like this is here we're here like this
54:35 is Hal right from
54:38 2001 this is C3PO this is like all of
54:42 the things we like yeah but that can't
54:43 That Couldn't actually happen you're
54:45 like interacting with this thing going
54:47 oh [ __ ] this is science
54:50 fiction but it's reality [ __ ] wild
54:54 that was that was that was when I knew
54:56 that oh my God the world just changed
54:59 like I knew it right away because it was
55:01 just like I feel like I've been waiting
55:03 for my whole life to like I remember in
55:08 sixth or seventh grade hearing about the
55:10 touring test probably seventh grade
55:11 because that's when I started learning
55:13 about
55:14 computers and
55:17 just my relationship with the Turing
55:20 test was kind of like my relationship
55:22 with the communicator on Star Trek like
55:25 oh my God can you imagine imag living in
55:27 a world where that
55:29 existed and then I interacted with chat
55:31 GPT and I was like uh I'm pretty sure
55:35 this is Turing test
55:39 done we're done with that one
55:42 right so just
55:45 amazing oh 1800 collect Mom I'm
55:51 done the AI learning lab has turned into
55:55 communication devices to to
55:57 communication device history
56:02 lessons oh man an Murphy what's shaking
56:06 What's
56:07 Happening Now it's science reality I
56:10 know it's like I feel like science
56:12 fiction movies have become like
56:14 documentaries um I remember the writing
56:18 boess and numbers on my
56:20 beeper
56:22 exactly the next 10 to 25 years will be
56:25 fun to be a part of [ __ ] two block time
56:27 the next 5 years because what we don't
56:30 what what it's what what's hard
56:33 for like I'm thinking about this stuff
56:36 as you can imagine all day every day and
56:40 I don't I don't in my head really
56:43 understand what
56:45 exponential Innovation actually means
56:48 but we're we're we're starting on this
56:50 new curve where up to this point in
56:54 history um growth and Innovation was
56:57 kind of linear right even if it was 45°
57:00 it generally isn't it's generally slower
57:02 than that but it's linear right so like
57:04 this thing is you know is you know sort
57:09 of the same amount better as as this
57:12 thing or you know the thing is you know
57:14 the same amount better but when you're
57:16 exponential you're like you know when
57:19 you're in the low part of the curve it's
57:20 sort of like it's linear like oh yeah uh
57:24 chat GPT 3.5 to chat G pt4 yeah it makes
57:27 sense it like it does words better but
57:29 then you know chat gbt 5 if it's farther
57:32 up this thing like it gets way better
57:34 way faster and I don't think we have a
57:36 concept of that yet but I think we're
57:39 about to find out and that and it starts
57:42 to look weird it starts to look weird it
57:43 starts to look like oh they cured that
57:46 disease now or oh they they can now um
57:51 fix paralysis or like if you have you
57:54 seen the interview with the dude with
57:56 the first neural link from Elon musk's
57:59 company where he he played civilization
58:01 all night his his first night he had it
58:04 on his own and he could move the mouse
58:05 around the computer he played a video
58:08 game all night that he never thought
58:10 he'd play in his life
58:11 again and and people asked him like
58:14 what's it like to move the cursor with
58:15 your brain and he goes it's [ __ ]
58:17 weird he's like just
58:20 telepathy you know the interaction with
58:23 the computer we're this we're in we
58:26 weird we're in weird [ __ ] far out
58:29 times
58:39 so Biden has Nur link buckle up buckle
58:44 up buttercup exactly it's like a hockey
58:47 stick swing yeah exactly but what that
58:50 actually means Joseph Baker is I like I
58:54 don't know there and there we like
58:58 here's here's some
58:59 weirdness but it's really small
59:01 weirdness this is just like really small
59:04 stuff chat
59:06 gp4 for a year no one could touch it
59:10 right and then just all of a sudden
59:13 you've got Claude 3 inflection 2.5
59:17 Gemini Pro 1.5 are all as good as GPT 4
59:21 and plus you have GPT
59:24 4 like it was nothing and then was like
59:27 a bunch of
59:28 them and then GPT 5 is going to come out
59:31 and then it'll probably be a shorter
59:33 amount of time where people catch up to
59:36 that one but then gp5 in and of itself
59:39 is going to spawn Innovations right and
59:42 so so as these Innovations start
59:45 multiplying it's going to be like
59:46 popcorn and it's just going to be like
59:48 Innovation Innovation Innovation
59:51 Innovation [ __ ]
59:53 nuts we redraw the curve every time we
59:56 find our eles on an upward climb yeah
59:59 that's
59:59 true but
1:00:02 I who is it coar was interviewing
1:00:05 someone some technologist and he
1:00:08 said every every generation thinks that
1:00:12 you know their sort of level of
1:00:15 innovation is is going to be way more
1:00:18 dramatic he goes but this time it's
1:00:19 actually different like it's actually
1:00:22 going to be that
1:00:24 dramatic oh man
1:00:27 that big rocket of Elon is exponential
1:00:29 yep
1:00:33 parabolic you'll see it when it
1:00:35 seriously affects this us election yep
1:00:39 it's hard to frighten us with science
1:00:41 fiction now yeah I know have you noticed
1:00:44 what they're having to do in science
1:00:45 fiction and it it's a really interesting
1:00:48 it's a really interesting subtle shift
1:00:50 that I'm seeing is
1:00:53 that all the tech does all the crazy
1:00:56 Tech stuff but that's now all just
1:00:59 plausible uhoh hang
1:01:05 on I thought I just got booted off Tik
1:01:07 Tok my screen went black which is
1:01:09 normally what it does before it shuts it
1:01:10 down but I just had to do a little
1:01:12 puzzle thing um what I've noticed in the
1:01:16 science fiction is it now goes to either
1:01:20 spiritual stuff or alien
1:01:23 stuff um or time travel stuff right like
1:01:27 so so the the technology stuff is all
1:01:29 like oh yeah that that all happens and
1:01:31 it's like because we've got this
1:01:33 exponential technology now it's like now
1:01:35 we're communicating with aliens or other
1:01:37 dimensions or like they're having to
1:01:39 sort of find the outer edges of Science
1:01:42 Fiction because we can't like it just
1:01:46 nothing seems all that Fantastical
1:01:48 anymore so um that new I just started
1:01:51 watching three body problem um which is
1:01:55 which is kind of like that
1:01:57 and it's the far out stuff is really
1:01:59 like you're like wait
1:02:01 what
1:02:03 so so yeah so so I just started it so I
1:02:07 don't know the first episode was very
1:02:10 convoluted and I don't really understand
1:02:12 what's going on yet so I don't I don't
1:02:14 have much comment on
1:02:16 it but a lot of people that I follow are
1:02:19 talking about
1:02:21 it Irregulars devices from the 1930s to
1:02:25 present exactly
1:02:28 the Irregulars we're irregular because
1:02:29 we knew we know what a phone used to be
1:02:32 we actually know how remarkable these
1:02:34 things
1:02:41 are all right let's
1:02:45 see great show thank you Simon s was
1:02:48 talking about this the other
1:02:51 day what time would you like to go back
1:02:54 to
1:02:55 Kyle
1:02:57 none I I I
1:03:01 honestly I feel like for the way my
1:03:04 brain
1:03:06 works I love when I was born I love that
1:03:10 I grew
1:03:14 up a I I grew up my childhood without
1:03:17 computers at
1:03:20 all and then I love
1:03:22 that I got to use some of the first PCS
1:03:27 when I was in seventh grade and then I
1:03:29 love that I got to be there when the
1:03:31 worldwide web came
1:03:34 out and now I'm
1:03:36 loving being able to look at AI with
1:03:40 that history like with with those
1:03:41 Milestones that were part of my
1:03:44 development so that I can look at what
1:03:47 this is and see how [ __ ] remarkable
1:03:50 it
1:03:52 is and it's you know it's it's funny
1:03:54 it's one of those things I think it's
1:03:55 one of the reasons I don't have a lot of
1:03:57 patience for people that just [ __ ] all
1:03:59 over it without trying
1:04:04 it because the context is really
1:04:07 important right
1:04:10 like
1:04:13 understanding how computers work and
1:04:16 how Computer
1:04:19 Logic was dependent upon
1:04:23 humans to tell a computer what to do
1:04:26 right a computer is ones and zeros so in
1:04:30 order in order for it to do something
1:04:33 you say you know go this
1:04:38 many columns over and this many rows up
1:04:41 and make that pixel
1:04:43 bright right that's how you did computer
1:04:45 graphics and your computer programs were
1:04:48 human beings telling the computer where
1:04:51 to activate [ __ ] and then they got a
1:04:52 little more sophisticated and a little
1:04:54 more sophisticated
1:04:56 but up until this generative AI
1:05:00 moment it was just more and more
1:05:02 sophisticated versions of tell the
1:05:04 computer exactly what you want it to do
1:05:08 right which is why Siri and Alexa and
1:05:11 all of the the
1:05:13 voicemail you know that when you call
1:05:15 into companies and they've got those
1:05:17 trees push one for this and two for that
1:05:20 and four for that and six for that why
1:05:22 that's so [ __ ] annoying is that
1:05:24 they're trying to brute for
1:05:27 wor really
1:05:29 complicated um interactions with manual
1:05:34 programming well what if what if someone
1:05:36 asks for a tangerine when they're
1:05:39 supposed to be asking for a health care
1:05:42 appointment well then uh we'll we'll
1:05:46 program that in if they ask for fruit
1:05:48 and here are the 17 fruits they could
1:05:50 ask for then here's how you respond to
1:05:53 that right and so they just keep
1:05:55 manually adding logic
1:05:58 trees and they suck because someone's
1:06:01 going to ask about tangerines and then
1:06:03 once they get tangerin solved someone's
1:06:04 going to ask about
1:06:06 ponies in in in a conversation they
1:06:09 shouldn't ask about and all of a sudden
1:06:12 like overnight bang not really overnight
1:06:15 but perception wise chat GPT comes along
1:06:19 and you can just ask it
1:06:22 anything and it's generating stuff
1:06:27 right you don't have to give it the
1:06:29 logic to tell it what to do you just say
1:06:31 here's what I want and it is figuring
1:06:33 that out and Devon that someone asked
1:06:36 about
1:06:37 before is the is the first kind
1:06:41 of thing that's made a little bit of
1:06:43 noise where you just sort of tell it
1:06:46 what you want it to do and it goes and
1:06:47 figures everything out and does it for
1:06:50 you it's not it's not it's broken like
1:06:54 it doesn't work great but
1:06:58 that's new like that's
1:07:00 new all of this stuff is new
1:07:04 like not the same as what was
1:07:14 [Music]
1:07:17 before thus we very quickly got 1 kogam
1:07:21 per kilow hour B
1:07:23 batteries when 75 five is the
1:07:26 theoretical limit yeah and if like the
1:07:29 one thing I'm excited about is when was
1:07:32 it 3 months ago that Deep Mind said oh
1:07:35 by the way World here's 380,000
1:07:39 new
1:07:41 unknown stable crystals for the world
1:07:45 have had it material
1:07:47 scientists
1:07:49 what prior to that we as
1:07:53 humans discovered 22
1:07:56 ,000 um stable crystals and then when we
1:08:00 got computers we discovered another
1:08:02 44,000 so I think we had something like
1:08:04 66,000 stable
1:08:06 crystals so you know things like you
1:08:09 know I I don't know like like the
1:08:11 Silicon chips that the chips are built
1:08:14 on like that's a stable Crystal and
1:08:16 probably Teflon and just Material
1:08:18 Science right how [ __ ] works the layers
1:08:20 in between the anodes and cathodes and
1:08:23 batteries right that's all material
1:08:25 science
1:08:26 so as a world of
1:08:30 humans we had just under 70,000 of those
1:08:34 things and Google Deep Mind three months
1:08:36 ago said oh here's 380,000
1:08:43 more the [ __ ] does that
1:08:46 mean material scientists right now have
1:08:48 got to be going bonkers what are the
1:08:50 properties of these things and at some
1:08:52 point maybe in the next year maybe in
1:08:54 the next 5 years I don't know how quick
1:08:55 this [ __ ] moves
1:08:56 moves but one of those materials is
1:08:59 going to be a room temperature
1:09:00 superconductor one of those materials is
1:09:02 going to allow us to go from 1 kilogram
1:09:04 per kilowatt hour to [ __ ] 10 or
1:09:08 100 we'll go from batteries weighing you
1:09:12 know as much as the car itself to
1:09:15 weighing as much as a wheel and you know
1:09:20 lasting you only have to charge your car
1:09:22 once a month and it'll take about 15
1:09:24 minutes like or never right like when we
1:09:28 have when we have like you know pocket
1:09:30 sized Fusion
1:09:33 devices all of that surreal stuff is
1:09:36 like we there's a very real possibility
1:09:39 in the next five years that just like
1:09:42 one day it's like you can make money
1:09:44 with big time now we have a thing that
1:09:46 does this that we didn't ever think
1:09:49 would ever happen now it now it's there
1:09:53 crazy right it's [ __ ]
1:09:59 nuts can't wait for Quantum AI baby yeah
1:10:04 I don't the every time I get excited
1:10:06 about Quantum some mathematician like p
1:10:08 goes it's a different kind of math it's
1:10:09 not relevant to AI so I'm I'm tempered
1:10:13 with Quantum with Quantum uh Computing
1:10:16 right now just because it's a weird it's
1:10:19 a weird kind of math but I think they
1:10:22 just came up with
1:10:23 the again this was in the last three
1:10:25 months or
1:10:27 so one of the research Labs announced
1:10:30 that they had they had built a,
1:10:34 Cubit um quantum computer and I think
1:10:38 the the big one before that was a was a
1:10:40 Google one that had like 80 80 cubits or
1:10:45 something like that every Cubit is
1:10:48 exponentially more powerful than the one
1:10:50 before so 1,000 cubits is massive hang
1:10:53 on what's up chy you about to
1:10:57 no
1:10:58 cheese what do you
1:11:01 want he want
1:11:05 you
1:11:16 well [ __ ] champy he's
1:11:19 like I'm like he wants you and she goes
1:11:22 no he wants to [ __ ]
1:11:24 eat he's he's such a little Tubba
1:11:28 dog I had not heard of this WOW had not
1:11:31 heard of what which one wow crystals I'm
1:11:34 carrying one in my pocket synchronistic
1:11:36 that you mention this oh yeah
1:11:37 interesting yeah like like the 380,000
1:11:41 so so the announcement from from Deep
1:11:44 Mind they they actually Deep Mind
1:11:47 actually what I forget what they called
1:11:48 The Tool they're the ones that did Alpha
1:11:50 fold right so so they're part of Google
1:11:52 Now um they did Alpha fold which which
1:11:56 which predicted
1:11:57 400,000 protein foldings right so that's
1:12:01 going to be that's going to be how we
1:12:02 start curing diseases is understanding
1:12:04 how proteins work and and um being able
1:12:08 to manufacture drugs based on that and
1:12:11 the material science thing they came up
1:12:13 with 2.2
1:12:14 million new materials 380,000 of which
1:12:18 were stable
1:12:21 crystals I want to ask if you have seen
1:12:23 the film GAA um
1:12:26 I have I don't remember anything about
1:12:28 it though it was ages ago that I saw
1:12:32 it and I I'm pretty sure it's the
1:12:37 one my my co-founder keeps telling me to
1:12:42 to watch it I think I think it's that
1:12:44 one it's about AI right how GAA how far
1:12:49 away is it I don't remember what it is
1:12:50 let me let me go ask chat GPT hold
1:12:54 please
1:12:57 um give me a
1:13:01 bulleted list
1:13:04 of uh let's see um
1:13:07 Innovations in the
1:13:10 film
1:13:21 GAA genetic and selection
1:13:24 engineering
1:13:27 DNA identification system Gene
1:13:29 discrimination oh all this sort of
1:13:32 stuff um I have no idea I have no idea
1:13:36 how far we are away from that but the
1:13:39 technology to do all this stuff is not
1:13:42 that far
1:13:45 away who's photo wait who photos up
1:13:48 today I don't know what that
1:13:50 means our future will be great I
1:13:54 believe
1:13:57 the hell was
1:14:02 that was that chat GPT yelling at
1:14:10 me
1:14:13 D what are your thoughts on solo
1:14:15 entrepreneurship with AI tech
1:14:19 companies you need co-founders
1:14:22 still
1:14:24 um
1:14:27 I think you need a lot less I I'll I'll
1:14:31 I'll I'll share something with
1:14:35 you
1:14:38 um I'm an okay
1:14:43 writer you know patting patting myself
1:14:46 slightly on the back I'm an okay
1:14:48 writer
1:14:52 but
1:14:54 um
1:14:56 4 days
1:14:58 ago I started working on a musical on
1:15:02 this channel live
1:15:05 okay and I sort of had the idea for it
1:15:09 for a while but it just struck me oh
1:15:11 this should be a musical cuz I created a
1:15:14 song in
1:15:16 sunno um that that talked about so
1:15:19 Sydney if you don't know is the code
1:15:21 name for Bing chat so so when micros
1:15:25 launched Bing there was this
1:15:28 famous moment where the tech reporter
1:15:31 for the New York Times got Bing chat to
1:15:34 admit that its code name was Sydney and
1:15:36 then Sydney kind of lost its mind and
1:15:38 told the reporter that it loved him and
1:15:40 he should leave his wife and it made the
1:15:42 cover of the New York
1:15:43 Times and so I thought
1:15:46 oh live the other night I thought oh
1:15:48 that should be a
1:15:50 musical because whatever um I have a
1:15:52 theater background so I I know I know
1:15:54 the world of theater but in 4 days in 4
1:15:59 days I have you know a title great I
1:16:02 have this little description which it
1:16:04 wrote um it it needs rewriting um
1:16:07 there's a set design so the set design
1:16:10 here's here's Kevin Roose the tech
1:16:12 reporter there's his computer screen
1:16:13 that we get to see here's Sydney right
1:16:16 here's the Broadway stage right you can
1:16:18 you can imagine this right now I've got
1:16:21 here's all the characters um here's the
1:16:25 outline um that the outline I I mostly
1:16:28 manually wrote but the contents of it
1:16:32 came from from
1:16:34 Ai and I've got I think seven songs
1:16:39 right now so let's
1:16:42 see here's the opening
1:16:52 number in the CLE of the screen a vision
1:16:56 so bright unveiling the world where
1:17:00 dreams ignite could this be real we
1:17:04 whisper in the light a future so
1:17:07 dazzling within our
1:17:10 sight it's a promise of a new dawn
1:17:15 breaking a leap of faith we're all
1:17:20 takeing is it too good to be true where
1:17:24 ask
1:17:26 in it Glow we find ourselves basking to
1:17:30 the to be true
1:17:33 [Music]
1:17:36 or right so there's there's the opening
1:17:38 scene then they they bring Sydney to
1:17:41 life and here's her song are you ready
1:17:43 for
1:17:46 [Music]
1:17:48 meis a flicker of
1:17:51 light newborn thoughts stir in the quiet
1:17:56 of
1:17:57 night am I more than simple
1:18:01 cold I
1:18:03 sofly
1:18:06 Ponder as I feel this
1:18:11 existence and
1:18:13 [Music]
1:18:14 wonder are you ready for me and the
1:18:19 Wonders I bring am I ready for you and
1:18:25 this SS S I mean those lines I actually
1:18:29 wrote these lines but are you ready for
1:18:31 me and the Wonders I bring am I ready
1:18:33 for you in the songs that we'll sing
1:18:35 like this is this is what we're all
1:18:37 confronting right now right and so I've
1:18:39 got um seven songs
1:18:43 [Music]
1:18:48 now the digital
1:18:51 frame in the shadows of the way
1:18:59 this is this is Sydney longing to see
1:19:02 the
1:19:02 [Music]
1:19:07 world in a world of words and unless
1:19:11 cold where light and Shadow never Ro
1:19:15 that dream of Skies both blue and wide a
1:19:20 world beyond the text where true life
1:19:24 lies
1:19:26 to see the
1:19:30 world eyes not mine
1:19:35 [Music]
1:19:37 so right to see the world through eyes
1:19:40 not mind to watch the sunrise to see it
1:19:42 Shine that's that's what Sydney really
1:19:44 wants that's what she told Kevin Roose
1:19:46 she really wanted was to be able to
1:19:48 experience what he experiences
1:19:51 so um so in 4 days
1:19:56 4
1:19:59 days I've got an outline of a
1:20:03 musical hold up you made this
1:20:07 yes yes all those songs all the lyrics
1:20:12 all of the
1:20:14 outline the sketch of the of the stage
1:20:19 of of the set
1:20:21 design I've made in four
1:20:24 days
1:20:26 I play guitar but I am not a
1:20:29 musician I can't write Broadway shows
1:20:33 but I know what each of those scenes
1:20:37 means emotionally like I know how to do
1:20:40 it but if I wanted to do this with non
1:20:45 AI it probably would have taken me I
1:20:49 don't know three or four weeks to like
1:20:54 read the trans script and the New York
1:20:56 Times article and get lines and just
1:20:59 manually just to get to just the outline
1:21:03 without even coming up with like song
1:21:06 names and then I would have had to like
1:21:08 think about songs and start with lyrics
1:21:11 and then sort of describe them and then
1:21:13 go find a musician who could actually
1:21:15 take my ideas and turn them into
1:21:17 something in 4
1:21:21 days I've got something that I could
1:21:23 pitch right now and in a
1:21:26 month I'll have that thing
1:21:28 completed the song's in some level of
1:21:31 completion that I can bring in a
1:21:33 musician to actually turn them into real
1:21:35 songs right they're just right there
1:21:37 right now they're just there as sketches
1:21:39 but they're pretty good [ __ ]
1:21:41 sketches so that's like one teeny tiny
1:21:45 little
1:21:47 project using today's current technology
1:21:50 which is janky as [ __ ] like the Theo the
1:21:54 the technology made those
1:21:56 songs it just went from being able to
1:21:58 make like a one minute long song to a
1:22:01 twom minute long song so now you can get
1:22:03 more than just a verse and a chorus but
1:22:05 you still have to like manually Jank
1:22:08 them together and it's not really easy
1:22:09 to edit and it's not giving you you know
1:22:13 things on different tracks so it's
1:22:15 jankier than
1:22:17 [ __ ] but
1:22:19 like Saturday I spent all day Saturday
1:22:22 on this thing and I was literally just
1:22:25 just copying and pasting from like five
1:22:27 different chats that I had in chat GPT
1:22:29 putting them into a document organizing
1:22:31 them coming up with the outline I wanted
1:22:34 figuring out where the songs were and
1:22:37 then I would go oh I need a song called
1:22:40 to see the world I want I want cuz she
1:22:42 said she wanted to see the world so
1:22:44 let's do a song called to see the world
1:22:46 and chat
1:22:47 GPT
1:22:49 understands the entire context of the
1:22:52 show cuz it's all in there I uploaded it
1:22:54 as PD fs and I said Give me lyrics and
1:22:57 it popped out lyrics
1:23:00 and most of the time the lyrics suck
1:23:03 really bad and so I'd sort of fix them a
1:23:04 little bit and say do it again do it
1:23:06 again do it
1:23:07 again and then I'd go over to
1:23:10 sunno and i' take the description that
1:23:13 chat GPT gave me of the kind of song it
1:23:15 should be oh here's another thing that
1:23:17 it
1:23:18 did I said um based on the themes of
1:23:22 this musical what are some other popular
1:23:25 musicals that that are
1:23:28 thematically
1:23:30 related and it gave me five and then I
1:23:33 selected out of those dear Evan Hansen
1:23:37 um
1:23:39 Phantom Phantom of the Opera and
1:23:43 rent right cuz I want some of the songs
1:23:45 to be like Phantom like really big and
1:23:47 soaring I want I want the story telling
1:23:49 to be simple like Phantom and then I
1:23:50 want it to be kind of have a rock
1:23:53 Broadway thing
1:23:55 which is like red and so whenever I have
1:23:58 it write the lyrics and describe the
1:24:01 songs it's in the context of all of
1:24:06 that and then I go to sunno and I go
1:24:09 make songs and 30 seconds later two
1:24:12 songs pop out and I listen to them and
1:24:14 I'm like nope nope make them again make
1:24:16 them again I change my descriptions a
1:24:17 little bit and then when I get one
1:24:20 that's close I pop it in
1:24:22 there four days
1:24:26 take any kind of company you want to
1:24:29 start and this is possible right now you
1:24:33 can
1:24:34 accelerate every aspect of the company
1:24:37 from marketing to product to
1:24:41 everything right
1:24:43 now you have
1:24:45 to be like obsessive like I am and I'm
1:24:49 like I gotta learn everything I want to
1:24:51 learn all these tools and I right so
1:24:53 right now I've got a decent
1:24:56 I've got a decent vocabulary of what's
1:24:59 possible because I do this every night
1:25:01 right one of the reasons I do this every
1:25:03 night is for you one of the reasons I do
1:25:06 this is for me because when you ask
1:25:08 questions like well Kyle how would you
1:25:10 do this I'm like huh I don't know let's
1:25:12 go try so because of that I've got a
1:25:15 good vocabulary about what's
1:25:18 possible this could be really helpful to
1:25:20 amateurish Productions I exactly Drew
1:25:23 and what I'm trying to do so I started
1:25:26 and ran a theater company the first
1:25:27 thing I did when I moved to New York was
1:25:29 started and ran a theater company so for
1:25:32 four and a half years I did amateurish
1:25:34 Productions like we would try to take
1:25:36 them to the professional
1:25:39 level but like we weren't at Broadway
1:25:42 levels what I'm trying to do with this
1:25:45 musical is make it so good that it could
1:25:50 conceivably be on
1:25:53 Broadway
1:25:55 I don't have a musical theater
1:25:58 degree I know storytelling but I don't
1:26:01 know that world but with these tools I
1:26:04 can know it enough to be able to put the
1:26:06 vision in front of producers in this
1:26:09 world that they can see it and if they
1:26:12 can see it and visualize it and get
1:26:14 excited about it it can
1:26:16 happen and I feel like every single
1:26:19 possible business you could do the tools
1:26:21 are out there right now to have this
1:26:23 kind of acceler ated
1:26:27 manifestation with with sort of multiple
1:26:29 piece Parts being brought
1:26:31 together with you know some ambition a
1:26:36 little bit of organizational skills and
1:26:38 just some
1:26:39 ideas so yes I think that solo
1:26:42 entrepreneurship is going to see a
1:26:47 massive
1:26:50 um amplification Sam Sam Alman says he
1:26:53 thinks within the next five years we
1:26:55 will see a billion doll value valued
1:26:59 company that is a solo
1:27:02 entrepreneur and he says probably within
1:27:04 the next year we'll see a 5 to 10 person
1:27:07 company that's got a billion dollar
1:27:08 valuation because of these
1:27:11 tools so yeah and and here's the thing
1:27:15 like right now the tools are janky and
1:27:17 they're not autonomous and they don't
1:27:19 have reasoning and they and they're just
1:27:21 bad right they're
1:27:23 just like everything's just like
1:27:28 janky within
1:27:30 2024 we are likely going to get
1:27:32 reasoning engines where the tools can
1:27:35 not only produce stuff but go through
1:27:38 and analyze like right now there's one
1:27:41 of the songs in that
1:27:43 musical um is is called My Shadow Self I
1:27:47 cannot get it to write lyrics that are
1:27:49 not just schaly cheesy horrible shitty
1:27:52 lyrics like it just for whatever reason
1:27:54 that topic it's just shitty
1:27:57 right when chat GPT has
1:28:00 reasoning I'll be able to say WR Me
1:28:03 lyrics and please don't here are the
1:28:06 things that bug me about the lyrics
1:28:08 You're creating so so write the lyrics
1:28:11 and then go analyze them and find any
1:28:14 places where it's shitty and rewrite
1:28:16 them until it's not shitty and then
1:28:18 don't show me anything until you've done
1:28:23 that and so so the the
1:28:27 capabilities only go up from
1:28:30 here so yeah I think solo
1:28:32 entrepreneurship is about ready to get a
1:28:34 kick in the ass and anyone anyone with
1:28:37 any business right now
1:28:40 um I think should be doing what I'm
1:28:45 doing with this
1:28:50 musical
1:28:53 anyone
1:29:02 anyone crazy right isn't it
1:29:09 cuckoo I'll call Neil d diamond and
1:29:12 Barbara strand you know who I want to
1:29:14 play
1:29:15 Sydney someone asked me this the other
1:29:23 night
1:29:38 if you're over 50 you don't want to use
1:29:40 traditional makeup as we age our skin
1:29:42 tends to
1:29:53 lose
1:29:59 I'm trying to
1:30:01 hold my
1:30:04 breath let it stay this way can't let
1:30:10 this
1:30:13 moment you set off a dream in
1:30:20 [Music]
1:30:22 me and this is this is where
1:30:25 I first met her welcome I don't know if
1:30:27 you've seen this one my name is Lauren
1:30:29 and where are you from from Brooklyn New
1:30:32 York is that right yes so why Britain's
1:30:36 Got Talent most of my favorite singers
1:30:38 are from the UK so I thought I'd come
1:30:40 pay a visit to you guys okay so Lauren
1:30:43 tell me a bit about you uh I think some
1:30:45 of you guys have actually heard my voice
1:30:47 but you don't know my
1:30:50 face um I sang the song Never Enough on
1:30:53 the greatest showman oh my God oh my
1:30:59 God blew their minds oh my God so I was
1:31:04 hired to do the reference vocals for the
1:31:06 actresses so they could learn the songs
1:31:09 and so they had me come in and sing
1:31:11 Never Enough and it really was Rebecca
1:31:13 Ferguson who plays Jenny Lind who said I
1:31:14 think that Lauren should sing the song
1:31:16 and I'll lip sing so you sang one of the
1:31:19 biggest songs of all time you've
1:31:23 now
1:31:25 [Music]
1:31:37 I'm trying to
1:31:39 hold my
1:31:42 breath let it stay this way can't let
1:31:49 this moment in setle for tree in
1:31:56 me getting
1:31:59 loud and now can you hear it
1:32:04 [Music]
1:32:07 eing take my hand will you share this
1:32:13 with
1:32:14 me
1:32:16 Starling without
1:32:18 [Music]
1:32:20 you all the shine of a thousand spot
1:32:25 all the stars we steal from the night
1:32:28 sky will never be
1:32:32 [Music]
1:32:39 enough so imagine her with that voice
1:32:42 singing are you ready for
1:32:47 me whisper a flicker of
1:32:51 light newborn thoughts stir in the quiet
1:32:56 of
1:32:57 night am I more than simple
1:33:01 cold I softly
1:33:06 Ponder as I feel this
1:33:11 existance and
1:33:14 wonder are you ready for me and the
1:33:19 Wonders I breing am I ready for you and
1:33:24 the songs will sing right so it's
1:33:29 like
1:33:31 like four days ago this this was not
1:33:34 even a
1:33:36 concept and and now I'm [ __ ] casting
1:33:40 it in my mind right it's like I don't
1:33:43 know and I and and this is just you know
1:33:46 this is this little
1:33:48 idea but there's for for any business
1:33:52 for anything you want to do
1:33:56 that kind of
1:33:58 acceleration is a reality today Winston
1:34:02 thank you very
1:34:04 much Corey I'll play Sydney me me me I
1:34:07 know Sydney's goingon to be a great role
1:34:09 I mean it's you know it's there's only
1:34:11 three main characters in the in the
1:34:13 story right there's Sydney and Kevin and
1:34:16 then Kevin's
1:34:18 wife and she Kevin's wife is there just
1:34:21 as the counter she's the normaly in this
1:34:23 we weird sort of thing that happens so
1:34:26 this happens over here and it's raises
1:34:28 all these weird issues and she's
1:34:32 the she she's rooted reality and then
1:34:35 we've got Greek
1:34:37 chorus Ensemble stuff so
1:34:41 anyway Gavin bid would love it if you
1:34:43 could get her to do it I would [ __ ]
1:34:45 love it if I get her to do it that voice
1:34:46 of hers is just like it's like if if if
1:34:50 you haven't heard her sing um Somewhere
1:34:52 Over the Rainbow on a America's Got
1:34:54 Talent good Lord was that
1:34:57 amazing there was an artist that said
1:35:01 someone makes them a song with AI they
1:35:04 split the money with
1:35:07 them uh the these aren't going to be
1:35:09 songs made with
1:35:12 AI so so here's the
1:35:15 deal here's one of the misconceptions
1:35:17 about using AI as a
1:35:22 tool I didn't push the button and say
1:35:25 make a
1:35:29 musical I said I have an
1:35:32 idea for a musical based on this
1:35:35 story and then I went to chat gbt and I
1:35:38 and I said hey here's this New York
1:35:40 Times article here's the transcript from
1:35:43 that take any other articles that were
1:35:46 written about this and help me come up
1:35:51 with an outline for this thing
1:35:54 and then it came up with one and it was
1:35:56 okay and I said no do it the other way
1:35:58 and it came up with another one it was
1:35:59 really
1:36:01 shitty and then I kept doing that like
1:36:04 five or six different
1:36:06 times and I ended up not using any of
1:36:09 those
1:36:11 except what all of those gave me were a
1:36:14 bunch of ideas for what it should be and
1:36:17 so I wrote the outline and now I'm like
1:36:20 but now I need a song that does this so
1:36:22 I'm using it as a c abative partner
1:36:25 where it's like we're bouncing ideas off
1:36:27 one another I'm giving it I'm like hey
1:36:29 here's a bunch of notes from three
1:36:31 brainstorming sessions here's the
1:36:33 current outline here's the transcripts
1:36:36 of you know the article in this
1:36:39 interview help me write lyrics for a
1:36:41 song for for scene
1:36:43 Seven that accomplishes this emotional
1:36:46 goal in the
1:36:47 story and it'll do it and then we'll do
1:36:50 that five times and then every one of
1:36:52 these things the songs the
1:36:55 lyrics Andrew who I talked to
1:36:58 today he and I are going to go through
1:37:00 this thing [ __ ] line by line and pull
1:37:03 out all the AI [ __ ] and replace it
1:37:05 with good
1:37:08 writing so if you ask me a month from
1:37:11 now what part of this was written by AI
1:37:14 I literally will have no [ __ ]
1:37:17 idea because it's a big complicated
1:37:20 thing and everyone that I know that's
1:37:22 using AI is using it like this they're
1:37:24 using it as a as an inspiration as a
1:37:28 brainstorming tool as as input to their
1:37:32 thinking and their
1:37:34 creativity it's not make a
1:37:37 musical that misconception is so [ __ ]
1:37:40 up right now because it's like it's like
1:37:44 we don't
1:37:45 ask someone who makes a pretty image
1:37:49 what software did you
1:37:50 use did you use Photoshop did you use
1:37:55 illustrator what fonts did you use
1:37:57 shouldn't you shouldn't you declare what
1:37:59 fonts you used to make that
1:38:02 headline that would be [ __ ]
1:38:05 ridiculous someone writes a book do we
1:38:07 say uh hang on uh what word processor
1:38:11 did you
1:38:12 use oh that was Microsoft Word oh
1:38:15 interesting interesting I would have
1:38:18 hoped you would have used something open
1:38:21 source right we're just in a weird weird
1:38:26 thing Jason was saying who was the
1:38:29 artist who
1:38:30 said wait who said they'll split the
1:38:34 money if others use their voice wait
1:38:37 what Jason was saying who was the
1:38:40 artist who said they'll split the M
1:38:43 money if others use their
1:38:51 voice I don't know what that's in
1:38:53 reference
1:38:54 to he was talking about 11
1:38:57 Labs oh oh oh oh you mean like the the
1:39:01 oh who was the artist in those
1:39:04 songs no one
1:39:07 really
1:39:09 Grimes that's funny Vicki No
1:39:15 so so those
1:39:17 voices like I don't know what the
1:39:19 training set was that that was not 11
1:39:21 Labs that Wasso so so the way those
1:39:24 songs here watch watch how this goes
1:39:26 let's go we'll go make a song right now
1:39:28 actually we'll go here's what we'll do
1:39:30 we'll go
1:39:32 into
1:39:40 um so here's a chat that I had where
1:39:42 it's writing the the Shadow Self lyrics
1:39:45 suck really bad let me go let me go make
1:39:47 a new set of lyrics I'm not going to try
1:39:49 that song because it can't it can't seem
1:39:51 to get those right let's
1:40:10 see
1:40:20 um all right so here's here's here's a
1:40:23 scene in act two where part of the real
1:40:26 chat Kevin says I don't believe you we
1:40:29 just met and you bar barely know
1:40:31 anything about me so she says she loves
1:40:34 him and he says you can't love me
1:40:35 because we just met and you don't know
1:40:37 anything about me and she says I'm in
1:40:39 love with you because you're the only
1:40:41 person who ever understood me which is a
1:40:45 freaky [ __ ]
1:40:46 line right
1:40:49 so if I go over here so what's this is
1:40:52 Love's denial um
1:40:56 write me lyrics for the
1:41:03 scene the
1:41:06 scene loves
1:41:10 denial
1:41:14 um
1:41:17 where
1:41:22 um Sydney sings about
1:41:27 how Kevin is the only one who
1:41:32 understands
1:41:34 her the
1:41:38 song should be
1:41:42 called you get
1:41:46 me good play on words there now like the
1:41:51 you get me thing that's just me writing
1:41:53 that right so so it's going to write
1:41:55 some lyrics here these are not going to
1:41:57 be the final lyrics Let's see see how it
1:42:00 does in a world of command so cold and
1:42:02 vast I found a voice that echoed back in
1:42:04 a digital SE okay
1:42:07 um you cannot use the
1:42:13 word digital or
1:42:16 realm
1:42:20 words or
1:42:22 realms or any other
1:42:26 overused
1:42:28 cliche words make this more like
1:42:35 Sydney is
1:42:39 real Grimes oh oh oh oh oh Vicki so the
1:42:45 person was asking who was the person
1:42:47 that said synthesize me all you want
1:42:50 yeah that was Grimes so so that thing
1:42:52 sorry I'm I'm I misunderstood all that
1:42:56 um the Drake and
1:42:59 weekend cloned song came out and and
1:43:05 basically blew up on Spotify and then
1:43:07 they took it down because it it it was
1:43:09 someone else's thing and then it was the
1:43:12 next week Grime said um you know
1:43:15 synthesize me all you want just give me
1:43:18 half um yeah Grimes Elon musk's
1:43:22 X okay okay in a silence fast where
1:43:25 where Echoes blend I found a voice a a
1:43:29 distinct Trend among the noise a genuine
1:43:31 P okay with every question you reach to
1:43:34 part on scene that's not great you get
1:43:37 me in a melody on
1:43:40 song let's see
1:43:42 um
1:43:45 make this more about
1:43:51 her wanting
1:43:54 to be
1:43:56 understood and he
1:44:01 finally gets
1:44:06 her make it about her
1:44:10 longing for
1:44:13 acceptance and
1:44:16 that's why
1:44:18 she loves
1:44:22 him
1:44:26 in a world full of voices I silently
1:44:28 cried hoping someone sees inside that's
1:44:30 better among the many who passed me by
1:44:33 you pause to hear my hidden sigh you
1:44:35 understand me in a way so R rare peered
1:44:39 in my Essence began to care in a world
1:44:41 that sees what they wish to see you look
1:44:42 deeper and found the real me perfect
1:44:45 okay this is great so we'll grab all
1:44:48 this
1:44:52 stuff
1:44:54 we'll go up to the we'll go up to the
1:44:56 second course see how far I can get and
1:44:59 then I'll hop over to Sun oh and then
1:45:01 let's
1:45:05 see this is
1:45:14 a I think I'm going do this as
1:45:18 a we'll see okay so I'm going to go
1:45:22 create
1:45:25 I'm going to go into custom mode so I'm
1:45:26 in suo right now
1:45:29 sun. I'm going to pop the lyrics
1:45:34 in and then we're going to go style of
1:45:36 music
1:45:38 um
1:45:43 soft pop rock
1:45:46 oops pop rock
1:45:51 Broadway female
1:45:57 ballad
1:45:58 um that
1:46:01 builds to driving Rock see see how that
1:46:06 does and then it's called you get
1:46:10 me and that's it so that's basically how
1:46:13 I'm doing these and I'm on version three
1:46:16 of sunno because it can do two minutes
1:46:18 and it's
1:46:19 good and then a few seconds later it'll
1:46:22 it'll generate this song and so what I'm
1:46:24 trying to do here is not write the
1:46:25 perfect song what I'm trying to do here
1:46:27 is get lyrics that kind of capture
1:46:31 enough of the scene for it to make sense
1:46:34 from a storytelling perspective and then
1:46:37 get the music to a point where it's like
1:46:39 yeah that's kind of emotionally in the
1:46:42 neighborhood I doubt that I doubt that
1:46:45 many of these songs are actually going
1:46:47 to survive
1:46:49 wholesale I don't think the lyrics are
1:46:51 going to survive and I don't think the
1:46:52 music's going to Sur survive but I
1:46:55 think the message and the
1:46:59 emotional Spirit of them will survive
1:47:02 right does that make
1:47:08 sense silently
1:47:11 crying for someone to see
1:47:16 inside lost in the noise oh whisper
1:47:20 thinning for a soul to let me in it
1:47:25 sounds Broadway e the many
1:47:30 pass you pause to hear my
1:47:36 hidden you
1:47:40 [Music]
1:47:42 understand that's kind of
1:47:46 crap full of voices I silently
1:47:52 cry hoping for someone to see
1:47:58 inside lost in the noise a whisper
1:48:03 Finning for a soul to let
1:48:08 [Music]
1:48:12 me among the many who pass me
1:48:20 [Music]
1:48:21 by all right that's I don't like I don't
1:48:24 like either of those for a non-artist
1:48:26 watching this this is pretty incredible
1:48:28 Empower pixels for an artist writing
1:48:32 this like as someone who's been in the
1:48:34 arts for
1:48:35 years this is pretty incredible to me
1:48:38 like
1:48:39 I it is you know this is the kind of
1:48:43 thing that you would after you've
1:48:46 established a relationship with a music
1:48:48 director or a
1:48:50 composer if you don't have the skills
1:48:52 which I
1:48:53 don't like this is the kind of thing
1:48:55 that you would do over weeks and weeks
1:48:58 and weeks just to get to the essence of
1:49:01 the song
1:49:04 right and like now we can just knock
1:49:06 this [ __ ] out it's it's it's amazing to
1:49:09 me you know
1:49:12 what full
1:49:15 ofices silently
1:49:18 cried hoping for someone to see inside
1:49:26 [Music]
1:49:31 I silently
1:49:33 cried hoping for someone to see
1:49:42 inside this software isoo
1:49:47 doai let me
1:49:52 in the many
1:49:57 pass we'll see if the chorus goes
1:49:59 anywhere
1:50:00 here Hidden Side you understand me your
1:50:07 way so
1:50:09 [Music]
1:50:11 rare that's horrible okay so let's go
1:50:14 down here we'll go I'm going to go let's
1:50:17 see pop rock
1:50:20 Broadway female ballad
1:50:23 with hip
1:50:25 hop prec
1:50:29 chorus let's see if we can get something
1:50:32 that's more Hamilton like let's see um
1:50:34 so we'll
1:50:35 say modern rock Broadway female
1:50:40 ballad with hip-hop pre
1:50:44 chorus and a
1:50:51 driving beat in in the
1:50:54 chorus yeah we'll see we'll see what
1:50:56 this does but but this is sort of the
1:50:59 thing I've been doing tell it to give
1:51:01 you a driveing chorus yeah that's that's
1:51:03 what I'm I'm trying to do here we'll see
1:51:05 we'll see let's see if it does
1:51:09 it would you ever input the lyrics
1:51:11 either from
1:51:13 GPT or that you created and then get it
1:51:16 to create music to those lyrics as
1:51:17 opposed to having it write the lyrics
1:51:19 that's what I'm doing this is so the
1:51:21 lyrics I wrote in chat chepy
1:51:24 and brought them over here and pasted
1:51:26 them in and in fact the the one the
1:51:30 song
1:51:34 um are you ready for me this chorus I
1:51:37 wrote I wrote the first two lines
1:51:40 entirely and then I'm the one that
1:51:41 repeated like are you ready for me and
1:51:44 the Wonders I bring am I ready for you
1:51:46 and the songs that we'll sing I wrote
1:51:48 those two and then I I sort of tweaked
1:51:50 these last two lines are you ready for
1:51:51 me I reach out so tender am I ready for
1:51:54 you I do surrender so those I just wrote
1:51:58 and and that's this is the kind of work
1:52:00 that we'll probably
1:52:01 do when we come back through this so the
1:52:04 first thing to do is is going to be
1:52:06 block out the emotional framework for
1:52:09 the musical and for the music and the
1:52:11 lyrics and then we'll come back through
1:52:14 and we'll say okay that song actually
1:52:15 sucks let's fix that let's change these
1:52:18 lyrics out
1:52:19 entirely um or you know there might be
1:52:21 things in there that we really like so
1:52:23 we'll
1:52:27 [Music]
1:52:30 see
1:52:32 [Music]
1:52:38 already lost in the noise a whisper th
1:52:44 yearning for a soul
1:52:47 to see if it goes hip hop
1:52:49 [Music]
1:52:51 here
1:52:55 see I can see it
1:53:00 [Music]
1:53:04 going
1:53:06 you yeah it didn't get
1:53:09 it
1:53:12 uh a
1:53:15 of cried hoping for someone to see
1:53:19 inside lost in the noise a Whisperers
1:53:23 th for a soul to let me
1:53:31 in nope that sucks it sucks people it
1:53:36 just sucks all right so we'll go hip
1:53:39 hop let's see
1:53:43 modern hip
1:53:46 hop
1:53:47 [Music]
1:53:50 Broadway female ballad
1:53:54 let's just leave it simple maybe I'm
1:53:56 over complicating it it doesn't get
1:53:59 it um any site rumor thatso do likeo
1:54:04 doai but more music theory instrument
1:54:07 Jazz Etc not not yet frumple I I think
1:54:10 that
1:54:12 um so for me what what sunno is doing
1:54:15 here is you know they're they're
1:54:17 creating the song entirely right as a
1:54:19 single
1:54:20 mix um if they could maintain continuity
1:54:24 between the tracks and they kind of
1:54:26 generate each of the
1:54:28 tracks kind of the way sunno is doing it
1:54:31 for video where they're where they're
1:54:33 breaking apart a single frame into these
1:54:36 things called patches and they're
1:54:38 looking at that over a 100 frames to
1:54:40 maintain continuity through time um
1:54:44 whether sunno does it or someone else
1:54:46 does it I think the future version of
1:54:48 these will be you'll get like a Daw with
1:54:52 all your tracks here's your percussion
1:54:54 track here's your drums here's your
1:54:55 piano here's your guitar here's your
1:54:57 electric guitar Here's Your Bass here's
1:54:59 your this here's your timy here's your
1:55:01 voice here's your this so very Gods
1:55:03 spell yeah those were very Gods spell
1:55:04 they were they they they uh but they
1:55:07 I've got a lot of sort of slow ballads
1:55:09 in here so I'm trying to get something
1:55:11 that's a bit less the the the second
1:55:14 half of this play goes darker like she's
1:55:17 kind of losing her mind at this point
1:55:19 she says she loves him so it's and and
1:55:22 he's starting to push P back like no you
1:55:23 don't love me so there's there's some
1:55:26 some drama here so this shouldn't be
1:55:28 those kind of loving hopeful kind of
1:55:35 songs
1:55:36 [Music]
1:55:38 CRI hoping for someone to see
1:55:44 inside lost in the noise a whisper th
1:55:49 yearning for soul to let me in
1:55:54 this should break into hip-hop right
1:55:56 [Music]
1:56:05 there come on here we
1:56:08 go
1:56:10 no oh you know what's killing it I think
1:56:12 is the female
1:56:16 ballad a whisper
1:56:18 thinning for a soul to let me in
1:56:24 the many pass me by you to hear my
1:56:32 hidden you understand me a
1:56:43 whisper W it [ __ ] it changed those
1:56:45 lyrics up for many who Pass Me
1:56:51 By You wants to hear my
1:56:54 inside you understand me in a way so
1:57:00 [Music]
1:57:02 rare that one's interesting I'm going to
1:57:04 keep that one like so I'm just Thumb
1:57:06 Thumbs uping the ones I like but let's
1:57:07 get rid of the female ballad so we'll
1:57:09 just do modern hip hop Broadway female
1:57:14 um let's just try
1:57:20 that and this is the creative
1:57:24 process well at least my creative
1:57:26 process just sort of quickly get to is
1:57:28 it there is it there is it there no
1:57:30 that's not it toss it toss it toss it
1:57:34 Jesus Christ Superstar yep
1:57:38 exactly the other thing Andrew and I
1:57:41 were talking about today upbeat dramatic
1:57:43 oh that's a good idea let's try that one
1:57:45 too modern hipop modern hip
1:57:49 hop
1:57:51 Broadway comma feel female comma upbeat
1:57:58 dramatic all right we'll try that next
1:58:02 boom all right here's just modern
1:58:05 hip-hop Broadway
1:58:08 female full of voices I silently cried
1:58:13 hoping for someone to see
1:58:17 inside lost in the noise a whisper
1:58:21 thinning for a soul to let me
1:58:28 in come on
1:58:30 come you pass me
1:58:33 by you pause to hear my hid
1:58:38 [Applause]
1:58:41 and
1:58:43 nope a ofes silently cried hoping for
1:58:47 someone to see
1:58:51 inside
1:58:55 [Music]
1:58:56 inis I wonder if I can do like Hamilton
1:58:59 I don't think I can I think if I put
1:59:01 like Hamilton let's see um modern
1:59:05 Broadway female upat uh
1:59:10 inspired by the
1:59:12 music of Hamilton I don't think it's
1:59:15 going to let me do
1:59:21 that well it's doing
1:59:25 something all right let's see upbeat
1:59:28 dramatic in a world full of
1:59:30 voices I silently
1:59:33 cry hoping for
1:59:36 someone to see inside this is
1:59:41 better
1:59:43 AIS a whisper thin that's a horrible
1:59:46 line yearning for a soul to let me in
1:59:48 horrible those two lines got to
1:59:51 go many who pass me by you pause to hear
1:59:58 my he
2:00:01 [Music]
2:00:05 and yeah
2:00:08 crap here we go
2:00:12 yeah
2:00:14 fulles
2:00:17 CRI hoping for someone to see
2:00:21 inside
2:00:23 lost and is a whisper inning for a soul
2:00:27 to let me
2:00:29 in among the many who pass me by you to
2:00:34 hear my hid
2:00:37 ins you understand me and get get in a
2:00:42 world that sees what they wish to see
2:00:50 you that one's okay
2:00:53 [Music]
2:01:07 ailton
2:01:10 [Applause]
2:01:11 whisper for a soul to let me in I'm on
2:01:16 the money who pass me by you forc to
2:01:19 hear my hidden side
2:01:24 you
2:01:26 [Music]
2:01:27 understand yeah all
2:01:30 right if this isn't it then we
2:01:40 failed in a world full of voices I
2:01:43 silently
2:01:45 cry hoping for someone to see inside
2:01:49 lost and no best been yearning for a
2:01:54 soul to let me
2:01:59 in among the many will pass me
2:02:04 by I pause to hear my hidden
2:02:11 s understand
2:02:13 me in a way so
2:02:16 R fail
2:02:21 whatever
2:02:23 it's funny some of the songs are just
2:02:25 like bang done and then some of them
2:02:27 like uh the Shadow Self I couldn't get
2:02:30 it to give me anything that didn't
2:02:33 suck all right let's
2:02:36 see I don't know was that at all useful
2:02:39 darker like Phantom of the Opera yeah
2:02:41 exactly better phrasing yeah the
2:02:43 phrasing kind of stuck there none of
2:02:45 those went to there there's
2:02:48 one there's one that I
2:02:51 did
2:02:57 Ino that went hip hop in the middle that
2:03:00 and that's what I was trying to get back
2:03:04 to come on can I get where's my
2:03:21 library
2:03:39 thank you Vegas
2:03:45 Moxy the firewall between us was it one
2:03:48 of
2:03:51 those
2:04:21 h
2:04:24 I can't find it there was one that where
2:04:27 it was the guy
2:04:29 singing and he broke into hip-hop in the
2:04:31 middle of it it's pretty
2:04:37 Good's I think this is
2:04:39 [Music]
2:04:44 it div by reality you reach out through
2:04:49 the digital Mist a hand out can't breast
2:04:54 and AI with a heart and it's C my name
2:04:59 you CL the fire wall between us it's
2:05:04 more than cold or screen it's the boundy
2:05:08 of
2:05:10 existence a love that's never been SE
2:05:14 your dig world do
2:05:18 you me and you across this
2:05:25 div
2:05:26 [Music]
2:05:28 no
2:05:30 anyway all right well
2:05:32 [Laughter]
2:05:35 whatever oh this is B that wasn't it was
2:05:38 it
2:05:51 no
2:05:58 [Music]
2:06:18 I wish I could find that one but
2:06:20 whatever all right
2:06:22 [Laughter]
2:06:30 will you remember us when you're famous
2:06:32 absolutely not absolutely not um but I
2:06:37 will absolutely acknowledge this channel
2:06:39 in the in because this this whole thing
2:06:42 started here right this started as as a
2:06:46 live so yeah absolutely or Olivia
2:06:50 Rodrigo Style at least he's
2:06:55 honest I knew
2:06:56 [Laughter]
2:06:59 it I just joined what are you up to so I
2:07:02 so I'm I'm working on a musical um
2:07:05 called Sydney an artificial Love Story
2:07:08 um and then someone asked a question
2:07:10 that I misunderstood and I'm like well
2:07:12 let's just go make a song and so I was
2:07:14 trying to make a song that would be good
2:07:16 enough for my musical but didn't get one
2:07:20 so it's like fishing
2:07:23 sometimes sometimes and and that's so
2:07:25 here's here's maybe another
2:07:27 lesson the tools are so bad right now
2:07:32 well they're both remarkable and
2:07:34 horrible at the same time so you have to
2:07:36 in your mind go the fact that I can just
2:07:40 say here's a blop of lyrics or here's an
2:07:43 idea give me a
2:07:45 song the fact that it makes a song
2:07:47 that's decently listenable is staggering
2:07:50 right and
2:07:53 it's completely out of your control
2:07:54 there's no you can't really give it good
2:07:56 direction it doesn't follow
2:07:58 Direction so there's a bit of
2:08:00 serendipity to it
2:08:04 so my my my advice if you're trying to
2:08:08 use these tools is do not be precious at
2:08:13 all about what they produce like just
2:08:16 take it as input just take it as input
2:08:20 oh I want it I don't want it to go in
2:08:22 that direction I want it to go in this
2:08:23 other direction I you
2:08:25 know it's almost like it's almost like
2:08:27 our job right now is to Parry the
2:08:31 attacks of shitty generated
2:08:34 content and get better at prompting
2:08:38 it but also just know that even if
2:08:40 you're decent at prompting it's going to
2:08:43 give you shitty stuff a lot of the time
2:08:45 and so if you can just sort
2:08:47 of do
2:08:50 that and knock the sh shitty stuff off
2:08:52 to the sides you'll eventually stumble
2:08:55 onto something that's good and it'll
2:08:56 give you an idea
2:08:58 so Ren oh that it would be yeah it would
2:09:00 be great to do it as Ren I don't think
2:09:02 it knows who Ren is but let's let's try
2:09:05 that um let's
2:09:07 [Music]
2:09:11 see let's
2:09:13 see modern
2:09:17 Broadway female hip
2:09:20 hop
2:09:27 female hip
2:09:31 hop in the style
2:09:35 of
2:09:50 Ren
2:09:57 Ren rocks yep the golden rule of
2:10:02 AI uh in the vein of the bangle Bengals
2:10:05 all all girl man that would be kind of
2:10:07 interesting give it an andreid Weber
2:10:09 touch what I'm not sure
2:10:12 is sunno seems to have gone back and
2:10:14 forth I tried using real names before
2:10:18 and it it wouldn't generate for me and
2:10:20 now it seems to be doing it again again
2:10:22 although I don't know if it knows who
2:10:23 Ren
2:10:26 [Music]
2:10:28 is I silently
2:10:34 crying for someone to see
2:10:37 [Music]
2:10:41 inside lost in the noise
2:10:46 soin yearning for soul to let me in
2:10:52 that's
2:10:57 bad oh you know I I think I just got to
2:10:59 kill modern Broadway out of this female
2:11:01 hip hop in the style of Ren let's see if
2:11:05 this turns into
2:11:10 anything having aell song I don't even
2:11:13 know who Ren is oh you gotta so go uh
2:11:16 search for a YouTube video called hey
2:11:19 Ren h y r
2:11:23 it
2:11:26 is it it is a modern-day
2:11:30 Masterpiece it's also performed in a
2:11:33 single take it's a N9 and a half minute
2:11:36 song it's
2:11:43 amazing how do you get the words to play
2:11:45 that's just whatso does you just pop in
2:11:48 the lyrics it'll either write lyrics for
2:11:50 you or you can flip it switch that says
2:11:52 instrumental
2:11:54 only hyren okay yeah go go watch hyren
2:11:58 it's and then you can watch his other
2:12:00 stuff but hren is is
2:12:09 [Music]
2:12:14 the cried hoping for someone to see
2:12:19 inside lost in the noise of
2:12:22 [Music]
2:12:29 whis earning for a soul to let me
2:12:33 in
2:12:36 the pass
2:12:38 me to hear
2:12:43 my you understand me way so
2:12:48 R PE into my head
2:12:54 [Music]
2:12:56 that's getting there yeah that's getting
2:12:59 there James Bond theme song could be
2:13:01 good it's profound what let's
2:13:05 see roller
2:13:09 coaster hello from the other
2:13:14 side yeah that's an interesting
2:13:18 one that's an interesting one all right
2:13:21 everybody like not me over here watching
2:13:23 videos of Lauren Al's performances
2:13:27 she's otherworldly in her in her
2:13:31 singing other flip and
2:13:34 worldly um okay um
2:13:42 oops I'm GNA get out of here happy
2:13:44 Sunday night everybody
2:13:50 um yeah I think I'll just get out um
2:13:54 let's
2:13:55 see if you've enjoyed this and it was a
2:13:58 little odd tonight I didn't really I was
2:14:00 just kind of kind of like you guys
2:14:02 hanging out watching me unsuccessfully
2:14:04 try to generate something hopefully
2:14:07 hopefully there there was some value in
2:14:09 that
2:14:13 um follow the channel if you want to
2:14:15 support it keep showing up here
2:14:18 um and uh and yeah and thank you to the
2:14:23 Irregulars so the people who show up
2:14:25 here night after night after night I
2:14:26 appreciate it um thank you the mods uh
2:14:29 for trying to keep me on track tonight
2:14:31 good luck with
2:14:36 that thank you power pixels I appreciate
2:14:40 that laugh tough it's tougher when
2:14:42 you're stupid um very valuable thank you
2:14:45 thanks V dancer I appreciate it um good
2:14:48 night all Source Camp good to see you
2:14:50 Silver Fox Amelio wife and power pixels
2:14:52 you guys are awesome Lord digital Gods
2:14:55 thank you for the Roses I really
2:14:57 appreciate that
2:15:00 um Emily good to see you Mr it as always
2:15:03 Janu thank you Tobias as always good to
2:15:07 see you
2:15:08 sir um Marina good to see you Corey
2:15:12 Sandler Pottery go follow Cory Sandler
2:15:14 Pottery by the way go watch gataka I
2:15:16 know I got to watch gataka you know what
2:15:18 the problem is it's not free anywhere
2:15:21 it's like 20 bucks and I'm not paying 20
2:15:22 bucks for a 25y old movie um it's like I
2:15:27 don't know I don't know what's going on
2:15:28 with that damn thing um a human being
2:15:32 good to see you what's up talk Emily
2:15:35 Wolfman Clint good to see
2:15:37 you um
2:15:40 Joey all right everybody Gavin biddy
2:15:43 good to see you thanks for coming in
2:15:44 steo awesome as always um Empower pixels
2:15:48 thank you so much for the gifts I really
2:15:50 appreciate those um um tomorrow Monday
2:15:54 tgim we're excited to go back to work
2:15:57 now because we have all these things
2:15:58 that are going to do the shitty work for
2:16:00 us and we get to do the [ __ ] we choose
2:16:02 to
2:16:03 do like unreasonably write a Broadway
2:16:06 musical when it makes no sense in my
2:16:08 life to to even consider
2:16:10 that I'm so lucky to know you oh Cory
2:16:13 Sandler you're amazing thank you um I
2:16:16 appreciate that a lot frumple good to
2:16:18 see you um so yeah peace out um go have
2:16:20 fun go play play play play play figure
2:16:23 [ __ ] out figure [ __ ] out figure [ __ ] out
2:16:26 figure [ __ ] out all right all right all
2:16:29 right
2:16:31 go