May 30, 2024 - AI's Creative Potential: Build Your Own Custom GPT Today!

Description

Chapters

Transcript

0:08 [Music]
0:14 [Music]
0:30 w
0:32 [Music]
0:38 Hill is a quiet
0:40 town where the oak trees
0:45 sway whistle of the train and go through
0:49 the
0:51 day people you with
0:54 a something never seems to change
1:01 the story is still
1:03 unfold it's Unique and strange Tobias
1:07 what's happening number one in the house
1:09 tonight
1:15 [Music]
1:49 weird Mary from C to
1:52 Hill she hows when she's blue
1:57 [Music]
2:04 of sadness and
2:06 shame under the southern Moon
2:10 [Music]
2:17 [Music]
2:46 what do you think
2:48 champy you good you
2:52 good did you all catch
2:54 the champy
2:56 live I did a live stream at like 6 with
3:00 with Champy's
3:02 butt Champy's butt was the star
3:07 attraction not many people saw it Tik
3:10 Tok doesn't like dog ass
3:14 apparently [ __ ] Tik Tock you know
3:18 like okay fat old guy sitting in the
3:20 chair I could see why they wouldn't want
3:22 to promote that but a cute dog's ass
3:25 come on Tik Tok there should be Millions
3:27 watching that you got the AI
3:32 if equals cute dog butt then
3:36 promote I know that's very 1987
3:39 programming thinking of
3:44 me is paid in here Pate loves if then
3:48 statements Kyle are you
3:53 happy in this moment absolutely
3:58 [Music]
4:05 [Applause]
4:07 [Music]
4:29 w yeah no politics in the lab I don't
4:31 give a [ __ ] about any of
4:34 [Music]
4:38 [Applause]
4:41 it got nothing to do with this
4:47 [Music]
4:55 [Music]
5:03 I didn't mean cause you in
5:06 [Music]
5:11 trouble CA you in a
5:18 pain one time to see you
5:21 [Music]
5:24 laughing I want to see you laughing in
5:27 that purple rain purple rain Purple
5:32 [Music]
5:35 Rain Purple Rain Purple
5:39 [Music]
5:41 Rain Purple Rain Purple
5:46 [Applause]
5:51 Rain well all want to see laugh in that
5:56 purple rain
5:58 [Music]
6:12 [Applause]
6:15 [Music]
6:19 what hang on got to tuck the girls in do
6:22 not do anything cool no chance of that
6:25 you're safe
6:28 [Music]
6:40 yeah the only the only politics we'll do
6:42 in here is if I run for president in
6:43 2028 as the spokesperson for the
6:46 [Laughter]
6:48 AI with the slogan Shannon AI it
6:52 couldn't possibly be worse um that's
6:55 going to be my campaign and and the
6:57 whole the whole thing's going to be uh
6:59 Mr Shannon what do you think about this
7:00 hang on a sec I'll get I'll have real
7:03 time chat I'll get i'll get Scarlet
7:05 Johansson to back the campaign so we'll
7:08 use her actual voice so we'll get Sky
7:12 back and then uh and I'll just be the
7:15 spokes model for the
7:18 campaign and what I will vow is um we'll
7:22 let AI make all the
7:23 decisions it couldn't possibly be any
7:26 worse
7:28 [Music]
7:30 Shannon Sky 2028 there you go there
7:34 there's if you want to make campaign
7:35 posters and put them in the regulars
7:37 Channel feel free Shannon Sky
7:40 2028 it couldn't possibly be
7:44 worse but you got to get champ in there
7:46 too
7:47 [Music]
8:07 my from my wrinkled
8:10 [Music]
8:12 sheets sto it was all a
8:16 [Music]
8:17 dream pain in my head and the hold of my
8:22 heart says it was more than just Jim
8:26 [Music]
8:28 be that was a
8:31 moment the moment I
8:36 knew there's a
8:37 [Music]
8:38 moment I to
8:45 [Music]
8:47 do I forget a song I wrote brains are
8:51 shitty hey did you hear they made a a
8:54 chip they made a chip with
8:58 16 brain
9:04 organel they took stem cells turned them
9:07 into brain
9:08 tissue and stuck them in a
9:13 chip and
9:15 apparently it's wicked
9:17 [Laughter]
9:20 smart that's not creepy
9:23 [Music]
9:31 and you can buy them I didn't know that
9:33 I thought this was just lab
9:38 [ __ ] did you make this image looks
9:40 familiar I think so what's the info
9:45 say
9:48 Dolly J J AO might have been Jason but I
9:52 don't know maybe possible it's possible
9:56 I try to put names in the files when I
9:58 when I add them to my little
10:00 my little situation but sometimes I like
10:03 if I save them from my phone I can't do
10:04 that
10:05 so it is very possible Dan
10:10 [Music]
10:15 you Mr it will be happy to
10:18 know sunno released yet another update
10:23 to their music thing you lose track of I
10:27 know
10:36 did you see what showrunner put out
10:38 [Music]
10:51 today that's pretty
10:57 [Music]
11:16 two block Tom Apple's going to have
11:18 emojis what the hell I think the phrase
11:20 you're looking for two block Tom is what
11:22 the
11:23 [ __ ] like seriously if Apple's big
11:27 announcement is we have gen AI
11:30 emojis I'm I'm [ __ ] going to Android
11:33 okay I'm not that's not I I I I'm sorry
11:36 Apple I didn't mean to say it I it was
11:38 just in a moment of
11:41 passion I wish I could quit you yeah I
11:46 seriously they better have a one more
11:49 thing moment if you know apple and their
11:53 uh when they used to put on good
11:56 shows honest to God if the big thing
11:58 they have is [ __ ] emojis it sounds
12:00 like it sounds like they signed the deal
12:02 with open Ai and it sounds like uh chat
12:05 gp4 Omni is going to be built into iOS
12:09 18 or you know it'll be the
12:11 preferred
12:13 um uh Mo model of choice and then I
12:16 think you'll be able to choose between
12:19 Gemini and maybe llama but I
12:24 don't I don't know what it
12:28 was we'll
12:32 see the fact that apple is partnering
12:35 with open
12:37 AI says that they don't have it yet I
12:39 assume I
12:41 [Music]
12:45 think and here's the thing you don't
12:47 need to be first with this [ __ ] apple is
12:51 rarely first with
12:53 stuff I mean they were developing the
12:56 Apple headsets for 9 years
13:02 although everyone's already declared the
13:04 death of the
13:05 [Music]
13:28 headset opening is a lot of drama right
13:30 now may may not be top dog for much
13:33 longer yeah the the thing that flies in
13:36 the face of that Danielle is
13:38 [Music]
13:40 that cat gbd4 has been out for um 14
13:45 months and no one's caught
13:48 [Music]
13:50 them and they just they just made their
13:53 most powerful model free for everyone
13:56 even though it's severely rate Limited
14:03 [Music]
14:05 so I have a feeling
14:08 in a month or two we're going to get a
14:10 more power the paid members are going to
14:12 get a more powerful
14:14 model so I don't see who's catching them
14:18 if Apple's not swinging for the
14:21 fences you know llama claims to be train
14:23 training their 400 trillion whatever the
14:26 hell that giant thing was
14:29 I don't
14:38 know Ken just got on what did I miss
14:41 nothing um me just saying I have no
14:44 interest in talking politics tonight
14:46 unless we're talking about my
14:48 20228 presidential run with Sky where my
14:53 campaign promises I will let the AI run
14:55 the country I'll just be the spokes
14:57 model for it and the uh the campaign is
15:02 Shannon Sky it should be Sky Shannon
15:04 shouldn't it well I guess I'm talking
15:06 first all right Shannon Sky 2028 and
15:08 then the tagline of the campaign it
15:10 couldn't possibly be
15:12 [Laughter]
15:14 worse and my only campaign promise is
15:17 I'll let the AI make all the
15:22 decisions but I'll prompt it real nice
15:32 Mr Shannon what do you suppose pres
15:34 president Shannon president Shannon what
15:35 do you suppose we should do about China
15:37 hold
15:39 please Sky listen listen get into
15:43 diplomacy mode and
15:45 [Laughter]
15:51 Pronto you've got my vote you'll have a
15:55 digital twin do it for you by then
15:56 you're damn right I will I'm I'm very
15:59 excited exced about digital twins right
16:01 now by the way the thing the I'll show
16:04 you what uh showrunner is doing we show
16:08 Runners the app I think the company's
16:10 called the
16:11 [Music]
16:25 simulation all right enough comedy jokes
16:29 champy is tired of singing he's over it
16:33 he doesn't want to hear from you
16:35 people it's like tell him I'm done
16:37 where's my [ __ ] chicken
16:39 stick you can make money with
16:42 Ching all
16:45 right what are we going to do shouldn't
16:47 the slogan be this is the worst it will
16:49 ever be oh that's not bad actually this
16:51 is the worst it will ever be no cuz then
16:54 when it's worse than that you're in
16:55 trouble
16:56 [Laughter]
17:01 Shannon sinking in the poles as
17:03 apparently the answer to it couldn't
17:05 possibly be worse is well yes it
17:14 can all
17:23 right all right Silver Fox is out
17:25 tonight she's got bad internet it can
17:27 get worse says says the woman who was in
17:32 politics
17:33 um yeah I know it Ken I'm yeah I know I
17:38 got I got that is just like I've told
17:41 you that before Daisy I I just got I got
17:44 I got no aptitude for that world at
17:51 all it is such a it's such a machine man
17:54 it's such a machine you got to know the
17:56 game I mean that's the thing about life
17:58 isn't it you got to know what game
17:59 you're playing if you know the game
18:01 you're playing then you can figure it
18:03 out but if you think you're playing one
18:05 game and you're actually playing a
18:06 different game that's rough that's rough
18:09 so that's what I do not want to be in
18:12 the middle of you know what I'm saying
18:15 you know what I'm
18:16 saying what if I did a whole campaign
18:19 like this I was like hey uh listen um
18:23 I've got a my partner here is an
18:25 artificial intelligence spot and um it's
18:28 a it's machine learning algorithm and
18:31 it's been it's been trained on um Mo
18:33 most of the public internet definitely
18:35 probably some copywritten materials but
18:38 it's definitely not plagiarism from a
18:40 technical perspective it's not really
18:41 how it works if you understand what I'm
18:43 saying um but anyway what I'm thinking
18:46 about doing is um is it would be a
18:49 national campaign um running for um what
18:53 they call President um where um I would
18:56 run the country well no I would I would
18:59 partner with my artificial intelligence
19:01 chatbot her name is Sky she's very nice
19:04 she very nice um I would partner with
19:07 her and
19:09 then what how I would kind of imagine it
19:11 maybe playing out would be um the
19:16 country would have an issue and then
19:18 someone would say to me Mr President Mr
19:20 President and and then they would say
19:22 what would you like to do and I would
19:24 say help please and then I would I would
19:27 go talk to Sky that's my my artificial
19:29 intelligence machine learning chatbot
19:31 named Sky it based on the voice of
19:33 Scarlett Johansson she's super super
19:36 sexy anyway anyway we probably shouldn't
19:39 talk about that on the campaign Trail
19:40 anyway should we oh am I am I flustered
19:44 so funny anyway um so then and then I
19:48 would she would give me an answer and
19:50 then what I would say is I would say
19:51 that to
19:53 you fantastic B tell him what he's won
19:57 who I'm sorry you've been elected
19:58 president
19:59 you thought it was a joke
20:04 oops it's happened
20:10 before okay
20:14 people what are we going to talk about
20:17 what would you like to talk about good
20:20 people of the internet I remember you
20:22 saying GPT was going to come out around
20:25 this time now version 4.0 is free
20:30 well chat CH chat GPT I remember you
20:34 saying chat GPT was going to come out
20:35 around this
20:37 time I said GPT 5 is probably going to
20:41 come out in the summer so I was thinking
20:43 more like
20:44 July um and I'm I it still seems like
20:47 that's going to be the case Well here
20:49 here's what it sounds like is
20:52 happening and again I don't have any
20:55 inside knowledge I just sort of try to
20:57 triangulate [ __ ] people are saying that
20:59 seem to have a
21:01 clue it seems to me what's happening gp4
21:08 Omni is not the next version of chat GPT
21:11 4 it's a completely different model it
21:13 was a different group within open AI
21:16 called the Omni
21:18 Division and this is
21:20 their GPT 4 level
21:25 multimodal um
21:27 model and it was fast enough and good
21:30 enough that I think it made it cheap
21:32 enough to give out to people for free I
21:37 think cuz it's apparently way cheaper to
21:40 run than regular chat
21:43 GPT then I think what's going to happen
21:46 is maybe Midsummer we get GPT
21:51 4.5 which is going to be just like
21:54 better some version of what we've got
21:56 now now I don't know if that's going to
21:57 be 4.5 Omni or 4.5 other
22:03 llm and then apparently November is when
22:06 we get GPT uh
22:10 five and they started training GPT 6
22:14 sometime around now so that would
22:17 probably come out the following summer
22:19 so that like if you follow Tech stuff
22:21 like that seems that seems like a
22:23 reasonable timeline but I don't know I
22:26 don't [ __ ] know I don't work there I
22:28 I live in [ __ ] Denver San Francisco's
22:31 having three or four AI events a night
22:36 you know what it's like in Denver oh hey
22:38 man yeah yeahi it's cool yeah man that's
22:40 cool it's cool you're into that is it
22:43 it's it's plagiarizing it's it's PL it's
22:46 plagiarizing from all the writers is
22:47 that it's that's it's theft so so you're
22:50 a thief that that's it yeah yeah I mean
22:53 it's good for you I mean fantastic I
22:56 guess
23:00 I don't know
23:02 nothing Apple's supposed to release AI
23:05 in June
23:07 well Apple's having the worldwide
23:09 developer conference on June 10th and
23:12 apparently what they're
23:14 launching is generative AI
23:18 emojis and oh yeah we stuck chat GPT
23:22 inside the operating system here's Sam
23:25 Alman Sam tell them what you've created
23:29 well you what we created it's it's a
23:32 it's a large language model it's going
23:34 to change the world we got AGI coming
23:38 yay I think that's going to be by the
23:40 way the Google leak was kind of
23:41 fascinating which Google leak what did I
23:44 miss was it
23:47 recent what's the Google leak hey
23:51 Google
23:52 uh Google leak
24:02 Google response to leak document lacks
24:05 context o that sounds
24:08 bad oh on the search algorithm yeah this
24:11 is Google won't comment on potentially
24:13 massive leak of its search
24:16 algorithm yeah that would be an oops a
24:19 Monumental
24:21 oops
24:24 except search is changing pretty
24:26 dramatically isn't it so does it does it
24:28 it
24:30 matter that it
24:33 leaked if they still had if they were
24:35 just the Monopoly in town and it was the
24:38 only way you could find things images of
24:40 2028 new here um apparently I'm running
24:44 for president with sky in
24:46 2028 where my only commitment is that I
24:50 will let the AI run the country and I'm
24:53 going to be the spokes model
24:59 Shannon wins
25:09 presidency oh good lord Shannon and Sky
25:13 2028 oh yeah it couldn't possibly we who
25:17 did this one Danielle that was beautiful
25:19 is that um is that uh what you call it
25:22 uh ideogram that's got ideogram written
25:25 all over it that's really good
25:30 champy champy made the poster Shannon
25:34 and Sky 2028 couldn't possibly be
25:41 worse oh that's good oh that's good with
25:44 the little the little Alexa thing champ
25:48 Champ's in there singing see I don't see
25:51 here's the problem I would never look
25:53 that buttoned up you're going to get
25:55 you're going to get it's just not going
25:57 to be that good
26:00 it's just not going to be that
26:02 good that's really good that's really
26:05 good all right what do you think the
26:07 white wait do you think the the White
26:09 House will have better
26:14 wifi Mr President can we get you
26:17 anything uh yeah what the [ __ ] is with
26:19 the Wi-Fi in this
26:24 place my Tik Tock my Tik Tock is
26:27 glitching like hell
26:32 Mr Shannon we need you in the situation
26:33 room I got to finish this thing with the
26:35 Irregulars someone wants to know how to
26:37 use chat GPT hang on a
26:39 sec I want open AI to have folders or
26:42 something better than archiving no [ __ ]
26:44 I want open AI to have a goddamn search
26:47 um look so let's apparently you can
26:50 search your prompt history in the Mac
26:52 app now but you have to have an M1 chip
26:54 in your Mac um let's see chat GPT
26:59 see if this thing got any better I doubt
27:01 it but it does have search because they
27:03 probably based it on the mobile
27:09 code hello are you the new voice model
27:12 are you still the crappy old piece of
27:17 crap hey there I'm here to help with any
27:20 questions or information you need how
27:22 can I assist you today are you based on
27:25 gp4 Omni or that old stinky gp4
27:32 .0 I'm based on gp4 designed to provide
27:36 helpful and accurate responses what
27:38 would you like to chat about or need
27:39 assistance with
27:41 today uh I'd like to know if you're new
27:44 or not but you pause so long that I know
27:47 that you're not you're old and
27:51 tired and I can't interrupt you I'm here
27:53 to assist you interrup I can't interrupt
27:56 help you with today so that's the old
27:58 model so we still don't have the new
27:59 model um it sounds like someone someone
28:02 posted today brand Brandon sent me
28:04 something on the AI Salon that API
28:08 access to the new voice [ __ ] is going to
28:11 be in coming weeks that was out today so
28:15 as I predicted the Scarlet Johansson
28:17 [ __ ] probably set back the launch of
28:20 that voice stuff for my guess is a month
28:23 at least probably 6 weeks so it's been 3
28:28 weeks
28:29 going on three weeks since they
28:32 announced that Sassy new voice
28:36 model and if they said in coming weeks
28:39 let's call that three
28:40 weeks API access and then it'll probably
28:44 be another three weeks after that that
28:47 real humans get it oh my God I had chat
28:49 GPT give me a pep talk in voice today it
28:52 was such a boost oh that's amazing I
28:55 mean wait till this thing is Real Time
28:57 lunar stick you know when it's when it's
28:59 real time when it responds in a third of
29:01 a second and it's got emotion and it can
29:03 sense your emotion I mean if if we think
29:07 Pi was insane Pi you know Pi is kind of
29:09 crazy because it does the good you know
29:12 vocal stuff or the the good kind of
29:15 emotive
29:17 style gp4 Omni can actually understand
29:21 emotion or I mean it can at
29:23 least it can at least can it understand
29:26 emotion no it can't understand emotion
29:29 it can hear your
29:33 tone yes and it can tokenize your tone
29:37 as well as your words and then it can
29:39 match that and it can probably
29:41 understand it to some degree there goes
29:43 my 20 bucks detect detect it
29:47 yes well it's it's more than detecting
29:50 it it's
29:52 like because it's natively been trained
29:55 on audio you're talking directly into
29:58 the
29:59 model your voice and your inflection is
30:02 getting
30:04 tokenized along with the
30:09 words I don't know enough about the math
30:12 to to say more than that but basically
30:16 but it the current models don't do that
30:19 at all the current models only
30:23 infer emotion because of the text
30:29 Kyle now maybe your rabbit can actually
30:31 work that well that's exactly what I'm
30:32 thinking Chris so if you don't know I
30:34 got one of them there the hell did I
30:36 just do oh I dropped my side quest drive
30:39 here let me let me show you some ancient
30:41 history so here's new
30:44 history rabbit do I still have any
30:46 charge in this thing no it's dead all
30:49 right let me charge it I'll charge that
30:51 up
30:53 um this here's a history lesson for you
30:57 this well Urban desires was a um web
31:01 Zine that I created in
31:04 1994 and that my
31:07 friends is a sidequest drive it's a
31:11 cartridge and inside
31:17 it is a hard drive
31:21 platter and this
31:23 holds 40
31:27 megabytes it's bigger than my
31:32 head I think this one might be it was a
31:34 44 there was a 44 I think this one might
31:36 be an 88 I don't want to brag but this
31:39 might be 88
31:43 [Laughter]
31:46 megabytes oh
31:49 [Laughter]
31:52 yeah I remember those I feel old well if
31:55 you remember those you are buddy
32:02 hate to break it to
32:06 you what the hell's going on over here I
32:10 see hang on
32:13 people I'm charging my
32:15 rabbit not that kind of rabbit you
32:18 perves all
32:21 right it's charging we'll see if it's
32:23 got we'll see if we get a software
32:24 update on it
32:28 are we doing anything tonight what do
32:30 you want to
32:32 do I got 8 inch I got an 8 inch floppy
32:36 nice my first the first computer
32:39 computer I learned to program on was a
32:41 trash 80 Radio Shack trs8 with dual 5
32:46 and a qu floppies the one that looked
32:48 like a like a Star Trek
32:52 computer sort of gray with the all
32:55 molded the keyboard was molded into the
32:57 Monitor and then the two floppies off to
32:59 the
33:00 side here I'll show
33:04 [Music]
33:10 you um TR
33:14 RS
33:26 80 that's the one
33:33 programming basic on the
33:35 trs8 10 print World 20 go to 10 run oh
33:41 yeah with those with those little
33:43 commands I created my first animation my
33:46 first thing I I did in uh that was it
33:49 that was what I I walked into my seventh
33:52 grade math class in
33:55 1977 and one of those was sitting on
33:57 every desk in math class I took advanced
34:00 math for some reason I didn't like math
34:03 but it just sounded better than not
34:05 advanced
34:07 math I'm like I don't want to be a
34:10 shitty math I want to be an advanced
34:12 math and we walked in and those [ __ ]
34:14 were on the desk we didn't we didn't
34:16 have cassette drives you used the Dual
34:19 floppies I don't know what this sad sad
34:22 situation
34:24 is but yeah that was uh
34:28 that was the deal yeah that was the
34:31 first that little program right there
34:33 that was my first program it's funny
34:35 that that's what he put
34:41 there you know K Kyle was Matthew
34:44 bradrick in war
34:46 games want want to play a
34:51 game thermonuclear war what
34:59 God I led that movie that was a good one
35:02 that was a good
35:04 one all right what are we going to do
35:06 here
35:08 people all right here's a thing how many
35:10 people in here 72 people if you're new
35:13 here one of the things you should
35:15 probably be asking is who the [ __ ] is
35:17 this guy does does anything happen in
35:19 here is this just him blabbing is this
35:22 all this is yeah so you see all the
35:25 hearts going on screen right now do you
35:27 see all those hearts s cuz people love
35:29 it they love the jibber jabber they like
35:33 to talk it's like Saturday Night Live
35:35 Without The jokes or the good
35:39 characters it's like low rent
35:45 Netflix and then occasionally you might
35:48 learn something but if you go there to
35:52 that thing the
35:54 salon. and click on the first link that
35:56 says AI Salon community
35:59 that is going to take you to this
36:06 place welcome to the a salon hello my
36:11 name is Kyle Shannon this is the a
36:15 hello it's French
36:17 [ __ ] and you can read about what we're
36:20 up to and who we are and why we do what
36:22 we're do we're we're about to upgrade up
36:24 update all of this stuff we just kind of
36:26 finalized this language today in the
36:28 organizers meeting so we're going to get
36:30 updated
36:32 values and updated mission statement
36:35 it's pretty baller I'm excited about
36:38 it but you should join it and then when
36:40 you join so you should read about who we
36:42 are and what we do and their values and
36:44 if you dig that if you want to be in a
36:46 community of people that are into this
36:49 AI stuff where you don't have to sneak
36:51 around like you
36:53 know you stole your stepfather's Playboy
36:59 when you're talking about AI hey hey uh
37:02 any anybody here into AI uh uh no I was
37:05 just joking I just kidding I'm not into
37:07 AI seriously it's not me seem my it's
37:11 for my buddy my buddy's into AI I'm I'm
37:13 not definitely not me so if you feel
37:15 like that this is a group for
37:18 you you seen the new development for per
37:21 from perplexity I saw something about
37:23 perplexity today but no I didn't I never
37:25 stole my dad's Playboy it was Penthouse
37:27 he
37:30 exactly your dad was
37:37 edgy that's
37:39 awesome let's see
37:42 perplexity
37:44 perplexity I should probably go to
37:46 Twitter rather than yeah let's go to
37:51 Twitter oh oh this we got to see okay so
37:56 the company is called or the the the
38:03 uh the Twitter channel is called Fable
38:07 simulation the simulation so these are
38:10 the guys
38:12 remember 8 months ago nine months
38:16 ago there was this thing where they
38:18 created an hourlong episode of South
38:21 Park from a prompt remember that that's
38:25 this company so today they made this
38:49 announcement okay so couple of things to
38:51 unpack
38:53 here so genre is right so this is going
38:56 to be looks like it's going to
38:59 be it's going to have
39:01 elements familiar to to those of you
39:04 that got
39:05 LTX licenses in here
39:10 but let's see for the
39:14 okay so take any
39:18 idea and just think it and then watch it
39:22 and then it
39:26 says get ready for the
39:31 simulated
39:34 show Revolution
39:38 okay I think what these guys are doing
39:41 is you create a
39:44 world and then there's a simulation
39:46 running so you have a bunch of
39:47 characters running
39:49 around that have history and can
39:53 interact with one another and then I
39:56 think that there are cameras that follow
39:57 the those characters around so I think
40:00 the characters like I think you like you
40:04 create some sort of
40:05 world and then it just generates a show
40:08 from the interactions of the characters
40:10 in the world I think that's where this
40:12 is going so if we go to their
40:24 site so make AI TV shows and episodes
40:28 with showrunner introducing the
40:31 first wait introducing first slate oh I
40:35 guess that's the name of
40:36 software oh no oh the first slate okay
40:40 so these shows are coming introducing
40:42 the first
40:46 slate the first slate the first ever the
40:49 first slate of they're missing an of
40:52 okay introducing the first slate of
40:56 first ever AI TV shows okay they need a
40:59 copywriter but that's okay we'll forgive
41:01 them because they're doing cool [ __ ] um
41:04 created by filmmakers and content
41:05 creators using showrunner technology
41:09 ushering in a new era of entertainment
41:10 and storytelling these groundbreaking
41:12 shows promis to Captivate move and
41:13 Inspire coming soon so there's exit
41:17 Valley and we get to see an episode of
41:19 that there's
41:20 uru there's pixels which looks like cars
41:23 there's leave behind which looks I
41:26 assume that's a robot got left behind
41:28 there's HSPA the prize San Francisco and
41:32 United flavors of America so it's a show
41:34 about ice cream cones or
41:38 cupcakes so here's exit Valley
41:41 showrunner shows aren't ordinary all
41:44 here we go showrunner shows can I make
41:47 this
41:51 bigger aren't like ordinary TV shows
41:54 they're powered by simulations built by
41:57 Fable simulation our first show Exit
42:00 Valley is powered by Sim San Francisco a
42:04 simulation of the be in the bustling
42:06 heart of Silicon Valley
42:09 unload oh upload yourself your friends
42:13 your characters to sim San Francisco and
42:16 use them in shows oh that's fascinating
42:19 so they're going to create worlds and
42:21 we'll be able to create characters in
42:22 those
42:23 worlds so okay okay this is super cool
42:27 so so now
42:33 imagine you could basically have
42:36 infinite amounts of shows right because
42:39 I could upload me and all my friends
42:42 like one
42:49 guy I can upload me and all my friends
42:52 to sim San Francisco and then I assume
42:55 what I could do is I could say I want to
42:58 focus on this guy is the main character
43:02 these two has the
43:03 sidekicks and they always hang out in
43:06 the you know in the down at the
43:11 Docks but then someone else could come
43:14 in and they could see my character in
43:16 there and they could say oh use use that
43:18 Kyle character in the background of my
43:20 show but my show is going to focus on
43:22 this pretty girl and whatever and we're
43:24 going to hang out in the hate and smoke
43:26 doobies
43:28 so so you could have dramas you could
43:30 have
43:32 sitcoms give them Frank
43:36 exactly exit Valley is a vicious crazy
43:39 satire of tech barrons startup
43:42 entrepreneurs and researchers seeking
43:44 their exits whether Financial bodily
43:48 escape death or psychotropic so drug
43:52 drug doing NPCs everyone comes to
43:56 Silicon Valley looking for an exit but
43:58 only some of them are getting out let's
44:00 watch
44:04 [Music]
44:09 it wait what's this say let's read this
44:13 all events and characters in showrunner
44:15 including including those based on real
44:18 people are
44:20 fictional so I think that's important I
44:23 think that says they're going to create
44:24 a system where we can upload showrunner
44:27 dox YZ yes thank you
44:31 Brandon are fictional no humans or
44:34 celebrities were involved in its making
44:36 the humor and impersonated voices are
44:39 designed solely to keep you laughing as
44:42 we take greater control of your
44:44 politicians and econ and economy best
44:48 you're soon to be AI overlords fingers
44:50 crossed good writing at least they've
44:52 got good
44:53 writing that's good I like it it's solid
44:59 gather around little ones tonight I
45:02 recount the legend of Silicon Valley a
45:05 distant land that once brimmed with
45:07 human genius and Folly where Tech Titans
45:11 and Gage Geniuses alike forg tools they
45:14 believed would better the world but
45:16 amidst this Innovation a virus more
45:19 potent than any computer could Harbor
45:21 spread unchecked it was greed and by the
45:25 time you meet popsicles had realized the
45:27 reality of your Creations it was all too
45:30 late wasn't it humans became so good at
45:33 consuming we couldn't help but wonder
45:35 what it it be like to like Chris sag has
45:37 entered the chat and they'd like to
45:39 speak to the manager yeah
45:42 exactly that's who was it who was it
45:44 that just said it um one of the big
45:47 directors just said it he basically said
45:49 yeah to to think that to think that AI
45:53 movies aren't
45:55 coming is
45:58 is is the equivalent of saying yeah this
45:59 car thing's not going to work out keep
46:02 using you know keep focusing on the
46:04 horses I forget who just said that I
46:06 wonder if your character injects
46:08 themselves into my Sims maybe chat GPT
46:11 edu oh yeah I saw that uh I I'll look at
46:14 that uh Mr K uh chat GPT edu becoming
46:18 available Enterprise level for schools
46:20 and universities yeah it was half the
46:22 cost and it does things like um helps
46:26 grade and write courses
46:30 and uh something else yeah we'll go look
46:33 at that for sure let's let's watch this
46:35 show though consume them but don't worry
46:39 we found humans more valuable for their
46:41 minds than for snacking though history
46:44 might have flavored things a bit
46:45 [Music]
46:51 differently the simulation original
46:54 series
46:59 I assume the title sequence is hand
47:03 [Music]
47:22 edited all right kids today we're going
47:24 to Deep dive into our very own backyard
47:27 Silicon Valley we're going to learn
47:29 about the history so so that opening
47:32 sequence that opening scene with the
47:34 robot I bet that was an actor that cuz
47:36 that didn't that didn't sound like
47:37 simulated voice it was edited too
47:40 tightly and now we're looking at South
47:42 Park South Park level cartoons they
47:45 might get better than this but these
47:47 don't have to be super sophisticated and
47:49 you can hear in her voice she doesn't
47:52 have the the same you know dynamic range
47:55 as as the dude the robot dude did
47:58 um gp4 Omni is g to dramatically
48:03 accelerate this it's going to get way
48:05 better way fastory of how a little
48:07 Valley became the tech Mecca of the
48:09 world not another documentary about
48:12 money hungry Tech nerds now now settle
48:15 down contrary to what many of you might
48:17 think Silicon Valley wasn't just whipped
48:19 up by some middleclass white guys in
48:20 their garages playing with circuit
48:22 boards wasn't that Steve Jobs and wnc
48:25 well yes but that's not the wo story
48:28 America's Tech Heartland has M all the
48:30 way back to the 1800s long before the
48:33 big greedy money hungry capitalist pigh
48:35 heads of today's big Tech so like it
48:37 started with cowboy nerds sort of Bobby
48:41 just watch
48:43 [Music]
48:47 [Applause]
48:48 [Music]
48:49 [Applause]
48:50 [Music]
48:56 [Applause]
49:01 four souls Thomas Zuckerberg III
49:04 Alejandra Savin Terry and Mortimer winkl
49:06 boss Bound by ambition oblivious to the
49:09 treacherous Allure of gold which awaits
49:10 in a valley made of hopes and dreams of
49:12 a future a future which would one day
49:15 shape the lives of every human being on
49:17 the planet gentlemen this is it this is
49:19 what we've been searching for the dawn
49:21 of a new future it's mesmerizing I've
49:24 heard Tales but seeing it touching it
49:26 this Chang everything it's more than
49:28 just gold this is our ticket to our
49:30 dreams in the valley I don't know about
49:32 you Buster but I dream of rails
49:34 connecting everyone rails vast networks
49:38 of them all right interesting we don't
49:41 have to watch the whole thing uh it's
49:43 you know I got it I I got it they're
49:47 going to be so they're going to they're
49:50 initially going to produce these shows
49:52 or they'll produce what they'll it's
49:54 it's probably going to be like the I'm
49:57 movie trailers thing where there's
49:59 probably a shell that's got like opening
50:01 sequences and sequences in between
50:03 scenes and then they they intercut these
50:07 NP NPC kind of scenes pretty clever
50:11 pretty [ __ ]
50:12 clever um the idea that they're going to
50:16 be able to create worlds and let people
50:18 upload themselves into those worlds and
50:20 make shows out of it is [ __ ] insane
50:23 fold Rush yeah absolutely Danielle as
50:26 soon as this things out if you've got
50:28 access to it oh that's the other thing
50:30 um the other thing I meant to tell you
50:33 hiy Pap Pap
50:36 Papo Papo
50:39 listen uh where is how do I get
50:43 there hang on people
50:47 p um
50:50 show runner oh that's
50:55 not show
50:57 [Music]
50:59 runner.
51:00 XYZ okay
51:03 so here is what you need to
51:14 know so if you say join the
51:19 alpha there's a weit
51:22 list so at the top of the page is a list
51:27 so you should put your email in the wait
51:29 list and then you should click the upper
51:32 right hand corner where it says join the
51:33 alpha and that Scrolls you down to the
51:36 bottom of the page where it says skip
51:38 the queue and it says we have a 50,000
51:42 person wait
51:43 list but we're doing some comedies here
51:47 so when you click apply it takes you to
51:50 a fairly long form that asks you things
51:53 like your favorite you know TV shows and
51:56 things like that and what's your idea
51:58 for a comedy so if you want to get
52:00 access to this earlier come up with a
52:02 [ __ ] idea and if you're like but Kyle
52:04 I'm not a writer how am I supposed to
52:06 come up with a comedy idea if I've got
52:09 no comedy I don't even have a sense of
52:13 humor um that's what chat jpt is
52:17 for go write a [ __ ] comedy plot all
52:20 right there's a poll up and the poll is
52:23 about I have no idea what this
52:25 is who gets 15 Minutes of Fame tonight
52:28 perplexity research did I didn't figure
52:30 out what was going on with perplexity
52:32 did I anthropic EXT tools don't know
52:35 what those are mid Journey high level
52:38 custom GPT building daily Jank with
52:40 Frank which we haven't done in a while
52:42 and Salon Show and Tell okay so there's
52:45 a clipboard in the corner over there go
52:47 to that
52:49 clipboard and you've got four more
52:52 minutes is that right I have old eyes
52:55 yes four more minutes you have 4 minutes
52:57 people 4 minutes go vote on what you
53:00 want me to talk
53:02 about and then because this show is what
53:06 it is I will look at that poll and I
53:09 will tell you what you voted for and
53:12 then I will likely promptly forget that
53:16 while I show you something else but I
53:18 will at least feel guilty about it after
53:20 the
53:21 show okay that's how it works around
53:24 here this is a neuros spice friendly
53:27 place if you're
53:28 neurotypical you're going to [ __ ]
53:30 hate this
53:33 place well what if he just had an agenda
53:36 and he followed the agenda Point by
53:38 point I mean you know he could it it's a
53:40 relatively uh trivial uh operation you
53:44 could just grab a you know a piece of
53:46 paper even even an index card and and
53:48 you could write you know even if it were
53:51 uh bullet points of of of the tasks that
53:54 you were going to accomplish in each
53:56 show uh and then the way I work is I I
53:59 start with task one and then I uh I I I
54:02 work on task one until I reach
54:04 completion or uh uh some sometimes I
54:08 don't reach completion if you know what
54:09 I mean just sort of lose momentum lose
54:12 the steam if you know what I mean anyway
54:14 it's the probably a
54:16 little little TMI little TMI anyway and
54:19 then and then after uh uh Point number
54:21 one on the agenda then I I generally
54:23 like to move on to point number two uh
54:26 at which point I I proceed uh uh orderly
54:30 uh through the tasks of Point number two
54:33 uh of course I've got uh sub bullet
54:35 points for for Point number two I don't
54:37 have those on my shorter yeah that's not
54:39 me okay people that's not how this crazy
54:44 straw of a brain
54:46 works like Steve Job Kyle tells people
54:49 what we want well you know there's
54:51 there's actually something to that that
54:54 that
54:55 um people ask Steve Jobs all the time
54:58 you know what do your customers want or
55:00 you know do you do
55:02 surveys and ask your customers to tell
55:05 you what they want and his response was
55:07 they don't know what they
55:08 want they would that's why guys like me
55:12 exist is basically what he said like I'm
55:15 going to figure out some cool [ __ ] and
55:17 I'm going to give it to them and then
55:19 they're going to want it they don't know
55:21 they want it until I make
55:24 it when I search for fold rush I get
55:27 links to the AI Salon that's really
55:29 funny can we vote twice I have no idea I
55:32 do not know how Tik Tock uh surveys work
55:36 you got a minute and 22 seconds though
55:38 um we are agenda
55:44 list listen I I spent the first 44 years
55:50 of my
55:52 life trying in vain to buy the right
55:56 five FIFA system in New York City in the
55:59 80s it was
56:00 filofax and then it was calendar apps in
56:04 the 90s and then it was project
56:07 management apps and it finally hit me
56:10 when I was like 45 I'm like oh I get
56:15 it it's
56:17 me there is no
56:21 system like I love setting up systems I
56:24 hate using them right so long as I'm
56:27 trying to figure it out and it's it's
56:29 giving me endorphin rushes I'm in but
56:32 the minute I actually have to start
56:33 using that system oh for the love of God
56:37 seriously no
56:41 yeah us add start three tasks partially
56:44 complete one start two more that's
56:46 exactly it that's exactly it and people
56:50 make fun of us but it's like that's how
56:52 that's how we get the good stuff done
56:54 you know what the advantage is of
56:55 starting three ideas is partially
56:57 completing one and starting two
56:59 more he this is what I've learned over
57:02 my over my over my many many
57:05 years the good ideas won't leave you
57:10 alone all right
57:14 so we got 19
57:17 votes custom gpts got 36% of the
57:25 vote so people want to see how we make a
57:27 custom GPT we could do that and then mid
57:30 Journey high level after that all right
57:32 those are the biggies so gpts I think
57:34 gpts is a good one cuz cuz
57:39 uh GPT 4 Omni was just made available to
57:44 all users of chat GPT anyone who has an
57:46 account even if they don't pay for it
57:49 now Kyle now Kyle we do soono thank you
57:52 for reminding me Mr T I think that's a
57:54 great idea um let's go to
58:00 sunno see you also can't be a smartass
58:04 with me because I'll [ __ ] go there
58:09 okay explore beta oh yeah that I don't
58:12 care
58:13 about
58:21 create if you guys don't know sunno
58:23 sun.com
58:26 and it now has
58:29 3.5 newest model better song structure
58:33 Max 4
58:35 minutes so let's see we're going to go
58:38 custom mode we're going to go show
58:42 lyrics Oh that's oh that's cool it shows
58:45 the lyrics over there like it your own
58:48 lyrics style of music
58:50 okay all right so we're going to go to
58:52 chat GPT
58:56 where we
59:00 go I want you to act as a Grammy
59:03 awardwinning Lyricist and
59:08 songwriter
59:14 and write me a song
59:27 about a guy that UPS uploads
59:30 himself into a simulation so that he can
59:34 make a Sit calm about his
59:39 life and can't find a way to escape
59:46 period once he's in the simulation
59:50 however he doesn't know it's a
59:53 simulation so he can't refer to
59:57 things like the digital void and
1:00:01 cyberspace and other crap that you
1:00:03 normally stick into songs like this
1:00:07 period give me five Concepts
1:00:15 first so my this wait it's not giving me
1:00:21 Concepts
1:00:23 well I mean this is a nice thing about
1:00:26 gp4 Omni is a lot less lazy than gp4 was
1:00:30 it just gave me one song let's see I
1:00:34 wanted five Concepts first not just a
1:00:37 single song period give me five Concepts
1:00:40 that are specific to the story above and
1:00:43 I'll choose one to write a song
1:00:46 about but let let chat GPT do your work
1:00:51 for
1:00:53 you you're like but Kyle shouldn't I use
1:00:56 my creative brain I I have a degree in
1:01:00 creativity and Kyle shouldn't I use my
1:01:03 creative brain cuz human brains are good
1:01:06 than than the machine brains and
1:01:09 shouldn't I use my human brain not if
1:01:11 you're just doing a dumb little Tik Tock
1:01:14 Channel you want to knock some [ __ ] out
1:01:17 quick you want to impress a friend just
1:01:20 let chat gbt do
1:01:23 it hey do you remember writer's block
1:01:27 I
1:01:29 don't I mean I do but I haven't had it
1:01:33 since chat gpt's been around because
1:01:35 every time I think oh God I don't have
1:01:38 any idea oh right I don't need to anyway
1:01:41 okay enough comedy jokes um let's see
1:01:46 our Concepts rerun reality the
1:01:48 protagonist finds himself living the
1:01:49 same day over no scripted Smiles I don't
1:01:52 no behind the curtain the protagonist
1:01:55 occasionally catches glimps of things
1:01:56 that didn't fit the sitcom world like
1:01:59 the backstage areas that I like
1:02:03 okay I love behind the curtain period
1:02:07 Now give me five different musical
1:02:10 Concepts that might bring the song to
1:02:12 life and I'll choose one of
1:02:16 those acoustic reflection I don't think
1:02:20 so synth wve mystery maybe theatrical
1:02:24 rock opera I think we're going there
1:02:26 Jazzy
1:02:29 Noir and melancholic Indie see I like
1:02:33 melancholic Indie but it's not right for
1:02:34 this particular creative project
1:02:36 theatrical rock opera is okay we're
1:02:40 going theatrical rock
1:02:43 opera hey Mr it you should be making
1:02:47 songs along I know you hate this but
1:02:59 write me the song lyrics and then follow
1:03:02 the lyrics with a musical description of
1:03:06 the
1:03:08 song period the description should be a
1:03:11 single short concise paragraph period no
1:03:15 extra
1:03:19 fluff sorry I'm late guys just watched
1:03:22 the Mavericks make their way to the NBA
1:03:23 finals wow good for them
1:03:27 Luca Luca Luca you're such a good
1:03:30 basketball player oh hey markco
1:03:36 hi welcome welcome
1:03:39 welcome
1:03:40 Allan how's weird Mary coming along Alan
1:03:44 do we have a play do we have high
1:03:46 schoolers all concerned about this AI
1:03:51 written thing okay this is
1:03:54 called the title is called behind the
1:03:56 curtain oh okay yeah I guess he wakes up
1:03:59 to the lights another Perfect Day the
1:04:00 shadows in the corner start to give him
1:04:02 away Whispers In The Silence uh it's not
1:04:06 quite right
1:04:08 okay the song needs to set up
1:04:14 that the character in The
1:04:18 simulation put himself there period make
1:04:22 sure that the first verse
1:04:26 lets us understand what the hell is
1:04:28 going
1:04:31 [Music]
1:04:34 on
1:04:36 okay he built himself a perfect world to
1:04:38 play a leading role uploaded into
1:04:40 fantasy where he could feel whole a life
1:04:43 was always scripted with laughter on Q
1:04:46 and now he
1:04:48 feels he feels the cracks in his dreams
1:04:51 he thought he knew Whispers In The
1:04:53 Silence echoes in the mind he searching
1:04:56 in the doorway a truth he hopes to find
1:04:59 behind the curtain where Secrets hide a
1:05:02 world constructed and Truth did not okay
1:05:04 this is
1:05:06 good does he always sing like that that
1:05:09 was
1:05:11 horrible um I think this is more than
1:05:13 four
1:05:14 minutes we can do a four minute song
1:05:17 well let's see how far we can get we'll
1:05:18 take all those lyrics that's like a six
1:05:21 minute
1:05:22 song hang on
1:05:27 okay I had to make a puzzle piece would
1:05:29 you make a copy of the first
1:05:33 draft would you like a copy of the first
1:05:36 draft for review of we Mary no no get it
1:05:39 get it to a place where where you're
1:05:41 happy with it you don't I'm just I'm
1:05:43 excited to see
1:05:44 it um okay we're going to sunno we're
1:05:49 going to pop our lyrics in here boom do
1:05:52 I want to kill anything here I I might
1:05:54 want to kill a pre chorus
1:05:57 [Music]
1:06:00 whisper now uh well we kill the bridge
1:06:04 yeah let's kill the bridge and
1:06:06 chorus boom we'll kill those then can we
1:06:09 can probably make the whole
1:06:10 song okay so we got that and then let's
1:06:14 go back to chat [ __ ] and again the song
1:06:17 starts with a gent guitar riff building
1:06:19 into rock
1:06:20 opera let's see if it can take all of
1:06:23 this I bet it can't I bet it's going to
1:06:24 choke and tell us that's too much I'm
1:06:27 exhausted with your musical description
1:06:31 here yeah it didn't it didn't like it
1:06:34 okay so I need to
1:06:37 go
1:06:39 um consolidate the musical description
1:06:42 into a single
1:06:50 sentence
1:06:54 of first the primary Styles and then
1:06:58 secondary
1:07:02 [Music]
1:07:04 flavors there we
1:07:07 go so I'm just going to go Dynamic rock
1:07:11 opera
1:07:30 okay title what was it behind the
1:07:36 scenes behind the curtain yeah it's
1:07:40 better behind the
1:07:43 cut all right we are apparently right
1:07:47 now going to
1:07:48 make that stupid Apple feature
1:07:56 we are going to make a oh they also did
1:07:59 look what they did this is a good ux
1:08:02 thing so it used to be when you created
1:08:04 a song it would put it at the bottom of
1:08:06 this list and so you constantly were
1:08:08 sort of scroll you had to scroll weirdly
1:08:12 cuz the new [ __ ] was at the bottom so
1:08:15 they actually put it oh look how fast it
1:08:16 went and is this four minutes long
1:08:21 [Music]
1:08:26 the to play the
1:08:30 roll into to see
1:08:35 where a LIC scripted with L you but now
1:08:42 we feel the cracks in this dream he
1:08:45 thought he
1:08:47 knew whis in the silence the echo in his
1:08:52 mind he's searching for the doorway
1:08:56 a truth he holds to F behind the c where
1:09:00 the secret
1:09:04 [Music]
1:09:09 High all right that's not horrible not
1:09:12 great but not horrible is it all right
1:09:15 here's the second one
1:09:19 [Applause]
1:09:20 [Music]
1:09:27 okay that's unlistenable all right we
1:09:30 got to fix some [ __ ] here Dynamic rock
1:09:32 rock opera Style with powerful or so
1:09:35 let's see
1:09:38 um
1:09:42 Broadway rock
1:09:51 opera um
1:10:03 let's see
1:10:06 [Laughter]
1:10:09 Tony all right let's make some more
1:10:11 songs I can't believe how fast these
1:10:13 things
1:10:17 are it's making a 4minute song in
1:10:20 whatever that
1:10:22 was Z
1:10:29 tap please
1:10:57 [Music]
1:11:05 bu the perfect world to play the leading
1:11:08 Ro uploaded into fantasy where he could
1:11:12 feel whole A Life That's always scripted
1:11:15 with laughter aren't you but now he
1:11:18 feels the cracks in the dream he thought
1:11:21 he knew The Whispers In The Silence
1:11:27 his mind
1:11:29 nice SE on the
1:11:34 [Music]
1:11:46 doorway
1:11:48 constructing
1:11:52 truth through the
1:11:55 the walls in the stage of make he's R
1:12:02 CRS his friends are also perfect their
1:12:05 smiles Never Fade but in the rise he
1:12:08 sees the lives the parts that forc to
1:12:10 play he's walking through the fabric
1:12:13 tearing down the walls in this stage of
1:12:17 make believe he's answering the calls
1:12:20 his friends are all so perfect the
1:12:22 smiles Never Fade but in the r he see
1:12:25 the lies the parts that F to play he's
1:12:28 walking through the stepes reing every
1:12:31 page deter to un cover the illusion of
1:12:35 this stage every SE to PR every life for
1:12:40 Chade but
1:12:42 [Music]
1:12:45 storying the
1:12:47 C the secrets
1:12:53 hide a world constructed
1:12:56 [Music]
1:13:06 there's really good continuity between
1:13:08 the two choruses a lot of times sunno
1:13:10 would make the second chorus way
1:13:12 different than the first one that's
1:13:13 really good
1:13:15 [Music]
1:13:18 continuity found the hidden doorway he
1:13:21 asked it for the SE and the quiet of the
1:13:24 blessing
1:13:26 heing me behind the curtain he's no
1:13:31 longer
1:13:35 blind a life
1:13:37 unscripted the truth he said to
1:13:41 [Music]
1:13:50 F all right um this is fascinating so it
1:13:55 look looks
1:13:56 like I pulled out the the
1:14:00 bridge and the second chorus before the
1:14:03 outro right
1:14:14 yeah that was 2 minutes and 45 seconds
1:14:17 with all those lyrics I guess because
1:14:19 it's faster style of music Broadway I'm
1:14:22 going to do rock instead of rock opera
1:14:25 bro
1:14:27 Broadway
1:14:32 Rock enriched by motive
1:14:39 screens and we'll
1:14:41 say
1:14:43 back ground
1:14:49 vocals yeah I know Mr it has got to be
1:14:52 in hell right now I'm I'm right there
1:14:53 with you Mr it I can't even imagine
1:14:56 how bad this is but let's go listen to
1:14:58 another
1:15:06 song you know the fact that I know that
1:15:09 this bugs you now gives me great joy I
1:15:11 hope you know that
1:15:15 [Music]
1:15:30 oh good Mr it is doing his song now
1:15:32 share your song yeah so if you want to
1:15:34 play along and share in the AI Salon in
1:15:37 the Irregulars
1:15:40 Channel dream he thought he knew
1:15:44 Whispers In The Silence eches in his
1:15:47 mind he's searching for the doorway a
1:15:51 truth he hopes to
1:15:53 find behind the c
1:15:56 where the secrets hi I to
1:15:59 CH and the truth deny he's W so beefy
1:16:05 daddy beefy daddy Zach hey my son's name
1:16:07 is Zach
1:16:10 um we are in suo
1:16:13 s.com they just released a new model
1:16:16 called sunno
1:16:18 3.5 that
1:16:20 produces full complete 4minute songs in
1:16:23 a single generation playing with yudo
1:16:26 again no I'm actually in sunno um just
1:16:28 because so sunno is apparently working
1:16:31 on version four of their model and this
1:16:33 is version
1:16:35 3.5 um but it does four minute songs and
1:16:38 then you can do 2minute
1:16:41 extensions um and then three days ago
1:16:45 yudo extended their Generations from 30
1:16:49 seconds to 2 minutes so it seems like
1:16:52 these guys are are trading blows back
1:16:54 and forth it's like we're watching a
1:16:55 heavyweight heavyweight fight um it's
1:16:58 pretty pretty wild all right so here's
1:17:00 the new the new one with all of the
1:17:02 lyrics in the preor in the
1:17:04 bridge so let's
1:17:11 see see if we get the whole song
1:17:18 in world to play the leading Ro
1:17:22 uploading into fantasy he could feel
1:17:28 home A Life That's always scripted with
1:17:32 laughter on C but now he feels the
1:17:36 cracks in this dream he thought he knew
1:17:41 wh was in the
1:17:42 silence echoes in his mind he's
1:17:46 searching for the doorway the truth he
1:17:50 hopes to
1:17:52 find all right that's fine I hate it but
1:17:55 it's
1:17:56 fine we still don't neither of these
1:17:59 really sound like
1:18:02 [Music]
1:18:04 Broadway this does a little
1:18:10 better into
1:18:14 fantasy where it could
1:18:17 feel a life that's always
1:18:20 scripted with left on c but now we feels
1:18:25 the cracks in this dream we thought he
1:18:31 will
1:18:36 Whispers that's
1:18:40 good the truth hold to find in the
1:18:46 C
1:18:52 secr and the TRU
1:18:59 [Music]
1:19:00 [Applause]
1:19:01 [Music]
1:19:05 he's he answering the
1:19:09 CR
1:19:14 [Music]
1:19:17 friend what about an extended TV show
1:19:20 theme song let's let's try
1:19:23 that um TV show theme theme
1:19:30 song from the 90s so we'll try to get
1:19:36 Cheers Cheers or friends or whatever
1:19:39 create all right that's a good
1:19:45 [Music]
1:19:50 idea speaking of songs Kyle did weird
1:19:53 Mary from Cedar Hill make it to Spotify
1:19:56 that's a good question um let me go to
1:20:00 Spotify I forget the name of the service
1:20:02 I
1:20:08 [Music]
1:20:11 used okay Spotify I don't want my
1:20:14 account I just want to go
1:20:17 there can I search
1:20:20 yes weird Mary
1:20:27 Weird Al
1:20:32 Yankovic weird
1:20:34 Mary
1:20:35 [Music]
1:20:37 Cedar
1:20:43 Hill
1:20:45 nope I got nothing people I wonder if I
1:20:49 put Kyle Shannon in here I don't think I
1:20:51 made it up they probably didn't approve
1:20:53 my song [ __ ]
1:20:57 you should have been
1:20:59 there this is
1:21:07 a hi there and welcome to Candy
1:21:15 [Music]
1:21:19 ears okay behind the curtain so here we
1:21:22 are TV theme song from the 90s
1:21:27 [Music]
1:21:30 good world to play the leading
1:21:34 Ro unload it into fantasy where he could
1:21:39 feel
1:21:41 hope A Life That's always scripted with
1:21:45 laughter on you but now he feels the
1:21:49 cracks in this dream he thought he knew
1:21:56 silence echoes in his mind he's
1:22:00 searching for the doorway it truth he
1:22:02 hopes to
1:22:03 [Music]
1:22:07 find that's [ __ ] horrible all right
1:22:10 but it's but it's exactly what we asked
1:22:12 for oh this is
1:22:20 better yeah this is totally TV themed
1:22:23 show from the 9s
1:22:26 [Music]
1:22:45 ech
1:22:49 behind double time on the drum nice
1:22:58 Perfect all right what were we supposed
1:23:00 to do here what where we G to we were
1:23:03 going to do we were going to make
1:23:07 gpts it's your fault Robert
1:23:10 meter don't cover your
1:23:15 face you you caused that nightmare do
1:23:18 they have a little roll the dice thing
1:23:20 on the style they don't
1:23:27 let's see um we'll
1:23:30 do melon wait
1:23:35 melancolic indie
1:23:39 folk that's where I always go
1:23:41 melancholic indie
1:23:44 folk that's what led to weird
1:23:49 Mary
1:23:51 wo thank you very much for what the heck
1:23:53 are those things the old microphone or
1:23:56 something I can't see what that is
1:23:57 that's cool thank you very much
1:23:59 appreciate the
1:24:03 gifts melancholic indie
1:24:06 [Music]
1:24:09 folk play the leading
1:24:12 Ro uploaded into fantasy where he could
1:24:19 feel a life that's always scried with
1:24:24 laughter on
1:24:25 you but now he feels the Cs in this
1:24:29 dream he thought he
1:24:32 knew Whispers In The
1:24:35 Silence echoes in his mind he's
1:24:40 searching for the doorway the truth he
1:24:44 hopes to find behind the
1:24:47 curtain when
1:24:50 SEC yeah see not bad indie folk always
1:24:53 seems to come through with
1:24:55 solid
1:24:57 although sunno has really bad like
1:25:00 autotune artifacts that it just sounds
1:25:04 everything sounds like [ __ ] tea pain
1:25:07 he
1:25:09 bu to play the leading Ro I blown it
1:25:14 into fantasy where he could
1:25:19 feel oh it's
1:25:23 corn thank you for the corn
1:25:26 but now he feels the Cs in this dream he
1:25:30 thought he
1:25:31 [Music]
1:25:33 knew Whispers In The
1:25:36 Silence echoes in his mind he's
1:25:41 searching for the doorway a truth he
1:25:44 holds to find behind the
1:25:47 [Music]
1:25:49 [Applause]
1:25:52 curtain oh my God I'm going to I'm going
1:25:54 to start weeping no we
1:25:56 can't I didn't I didn't realize it was
1:25:59 that sad we put this guy in his little
1:26:01 synthetic world now he's just trying to
1:26:03 get
1:26:05 out better than what you had okay so
1:26:09 question how will we get to hiring and
1:26:11 creating a paycheck for everyone with
1:26:13 the help of AI it's a really good
1:26:15 question
1:26:17 um I don't know there's going to be a
1:26:19 lot of disruption there's going to be a
1:26:20 lot of categories of work think like
1:26:25 customer service
1:26:27 reps um probably outbound sales
1:26:31 um you know some of the some of the
1:26:34 areas where you got massive amounts of
1:26:36 people all doing exactly the same thing
1:26:37 in a kind of scripted
1:26:39 manner I think those probably go first
1:26:43 but I but I think everyone's going to be
1:26:45 impacted and I don't know I mean people
1:26:48 are talking about Universal basic income
1:26:50 that's the reason
1:26:53 why um
1:26:55 if if if the AI disruption
1:26:59 is if the AI disruption
1:27:04 is
1:27:08 linear right so even if it's increasing
1:27:11 if it's
1:27:12 linear we can manage to that right then
1:27:15 they'll do some sort of stimulus [ __ ]
1:27:18 and they can manage to that the the
1:27:21 challenge is going to be if we're going
1:27:23 along with
1:27:25 normal employment [ __ ] and then all of a
1:27:28 sudden an entire sector gets wiped out
1:27:31 or three sectors get wiped out and
1:27:33 unemployment does that then I think
1:27:36 you're going to need to see something
1:27:39 more dramatic like Universal basic
1:27:41 income but I don't know I got I got no
1:27:43 idea I
1:27:45 mean what this Channel's about why I
1:27:48 started this more than a year
1:27:52 ago was that having been threw it before
1:27:55 with the worldwide web I knew that there
1:27:57 were going to be a lot of people trying
1:27:58 to figure it out there're going to be a
1:28:00 lot of people
1:28:02 scared um and unlike the worldwide web
1:28:06 this is moving fast and the impact of it
1:28:10 like as as dramatic as the impact was of
1:28:13 the worldwide web it really took it 15
1:28:16 or 20 years to really have that impact
1:28:20 and it was a very slow build the the
1:28:22 infrastructure wasn't there
1:28:25 like in my day my computer had to call
1:28:29 up another computer and shake its
1:28:33 hand and it you know we we uploaded
1:28:36 files slowly and now it's like instant
1:28:41 right so so you've got the
1:28:44 infrastructures in place right now for
1:28:46 generative AI to to go like wildfire so
1:28:50 the worldwide web to get to 100 million
1:28:52 users took six years chat
1:28:58 GPT hello [ __ ] cat chat GPT took six
1:29:03 weeks to get to 100 million
1:29:07 users and that there was no advertising
1:29:09 that was just Word of Mouth 100 million
1:29:12 users in six weeks fastest adoption of
1:29:15 technology in
1:29:17 history
1:29:19 so the the chances of AI having a really
1:29:23 acute impact on unemployment are
1:29:26 actually pretty
1:29:27 good um and that's
1:29:30 scary but it also has you know the
1:29:35 the effect of generating a lot of
1:29:40 product without a lot of cost so so
1:29:43 Prosperity goes up now the question is
1:29:46 who gets the prosperity if it's the
1:29:48 private Equity firms they're going to be
1:29:50 like more profit for us
1:29:55 cuz they're ass
1:29:59 hats
1:30:00 so at what point is it at what point is
1:30:04 the disruption intense
1:30:07 enough that you know you got to step in
1:30:09 and do something I don't have a [ __ ]
1:30:11 clue yeah but how many of those are
1:30:15 Bots how many of what are
1:30:19 Bots I don't I don't know how many oh
1:30:22 how many Bots are using chat GPT
1:30:26 I don't think a lot maybe a lot I
1:30:29 don't I don't think
1:30:33 so the 100 million users no like it was
1:30:36 it it was like the adoption of chat apt
1:30:39 was was dramatic and Swift I mean it it
1:30:42 caused Google to have a [ __ ] freakout
1:30:44 moment they called it a what a red flag
1:30:46 moment or or a red Red Alert red flag I
1:30:50 don't know something Google lost their
1:30:53 [ __ ]
1:30:54 cuz chat PT got to a million users in
1:30:57 five days and 100 million users in six
1:31:01 weeks you posted but it's too long oh
1:31:03 okay that
1:31:04 sucks thank you Robert R Robert Ry hello
1:31:08 Robert Ry thank you for the lightning
1:31:11 bolts the gifts mean a lot I appreciate
1:31:23 it dashing out for a bit this woman
1:31:26 works but I need to focus in on a
1:31:29 document all right dash out focus on
1:31:32 your document Robert Rossy that's a lot
1:31:35 of lightning you're awesome thank you so
1:31:38 much um Jim and I have some very
1:31:42 serious problems converting metric to US
1:31:49 units thank you Tren is it Tren tronan
1:31:53 TR trco tronin tronin TR trenin
1:31:59 trenin I don't [ __ ] know I use chat
1:32:02 GPT more than Google now when I ask
1:32:05 about something absolutely me too Tik
1:32:07 Tok live
1:32:08 games I do too Google's AI response
1:32:12 training with Reddit was a smart move
1:32:16 not I love sunna oh Mr it is now he's
1:32:19 Ino ah gotcha
1:32:24 it's
1:32:25 amazing the these music things are
1:32:27 amazing the thing the thing that's wow
1:32:31 Robert Rossy oh my God thank you so much
1:32:33 it's very very generous um I'll tell you
1:32:36 what
1:32:38 like check your regulars okay I'll do
1:32:41 that Google's response yeah training on
1:32:43 Reddit I know training on Reddit what
1:32:45 could possibly go
1:32:47 wrong thank you Robert so much that's
1:32:50 really generous um
1:32:55 oh oh Reddit I use pie her name is Diane
1:32:59 my wife named hers
1:33:03 Cherry Gemini AI will will not show the
1:33:08 text of many US
1:33:10 amendments Gemini is a [ __ ] mess
1:33:13 right now I made two songs in Spanish
1:33:15 nice oh that's that's the other thing
1:33:17 about these tools they'll just speak in
1:33:19 any language you want it was funny I was
1:33:22 I was on a call a Content Evolution call
1:33:24 they were talking about you know we can
1:33:26 make these personas and then we could
1:33:28 make them for different languages I'm
1:33:30 like they're already made for different
1:33:33 languages I mean it's just
1:33:40 [Music]
1:33:42 crazy I'm going to go explore
1:33:45 gpts although I don't know if my my um
1:33:48 GPT app has been not functioning
1:33:53 right let me see
1:33:57 okay
1:34:02 so why can I
1:34:06 not start chat
1:34:46 hold please I know you're waiting for
1:34:51 me thunder thunder they're coming to get
1:34:55 us okay I'm in my I'm in my own custom
1:34:59 GPT
1:35:03 called wait that's not the one I
1:35:06 wanted C
1:35:16 [Music]
1:35:18 GPT hang on
1:35:30 well that's fine
1:35:32 whatever so I'm in I I've got a virtual
1:35:36 twin this is for me as author of a book
1:35:39 and then if you're using the chat GPT
1:35:42 app you can put it into conversation
1:35:44 mode now this is the thing that they
1:35:47 demoed 3 weeks ago that's you know that
1:35:50 Scarlett Johansson got pissed off about
1:35:52 that's not available yet it won't be for
1:35:54 I'm guessing another month to 6
1:35:56 weeks so it's not available yet so if
1:35:59 anyone's telling you they're using the
1:36:01 new thing they're not but the underlying
1:36:03 model is the new thing
1:36:06 so I can click on these little
1:36:11 things we'll see if
1:36:13 it hey Kyle how's it
1:36:18 going in case it isn't clear and to set
1:36:21 expectations I'm not the real Kyle
1:36:22 Shannon I am a virtual twin anything I
1:36:25 say should not be attributed as a direct
1:36:27 quote I might even hallucinate so check
1:36:30 the results for yourself it sounds like
1:36:32 you're referring to a moment or
1:36:34 experience that might have been
1:36:36 important or memorable could you share a
1:36:38 bit more about what you're thinking or
1:36:40 reflecting on I'd love to hear more
1:36:44 actually what I'd like you to do is
1:36:46 respond in Italian every time I ask
1:36:48 something in
1:36:52 English so
1:36:58 yeah I don't know what you said but can
1:36:59 you explain the future of AI to
1:37:13 me so I don't know if you knew these
1:37:15 things could do that
1:37:18 but you create a custom GPT it can just
1:37:22 speak other languages
1:37:25 good and going to get better and going
1:37:28 to be real time so um it's it's batshit
1:37:32 crazy what's happening bat [ __ ] crazy
1:37:35 I'm going to use this feature when I go
1:37:36 to Greece next Friday yep exactly why
1:37:39 would you
1:37:40 not um I asked you what you oh it asked
1:37:43 you what you want to know oh that's
1:37:45 funny Lazarus that's funny let's see can
1:37:50 I have it speak this hang
1:37:52 on select text read aloud
1:38:42 um sorry I'm an ugly American and
1:38:45 haven't even attempted to learn your
1:38:47 language
1:38:55 oh that's fascinating so Lazarus says
1:38:57 it's saying that it has promised it has
1:39:01 promised to do the boring work and free
1:39:03 humans to be more creative which is
1:39:05 absolutely that's something that I talk
1:39:07 about in my chapter of the AI Futures
1:39:09 book so that is on point that is a that
1:39:12 is a uh good
1:39:14 [Laughter]
1:39:18 GPT yeah I am I am quite enamored with
1:39:22 with digital twins right now um not just
1:39:25 of me just of the whole concept of
1:39:27 digital twins um the technology is
1:39:31 advancing so fast that we're probably
1:39:34 less than 6 months
1:39:37 away from being able to very easily
1:39:40 create a video representation of you
1:39:44 that can talk to someone in real
1:39:47 time that's not going to
1:39:49 sound as stilted and weird as it does
1:39:52 now and it's going to be like like when
1:39:54 I say real time the response will be a
1:39:58 third of a second human response is
1:40:01 about a quarter of a second gp4 Omni is
1:40:05 about a third of a second so we're like
1:40:07 a tenth of a second
1:40:09 away from the speed of response of a
1:40:12 human
1:40:13 being don't worry I'll keep replying in
1:40:16 Italian to your English questions oh
1:40:18 that's great that's great she's so
1:40:21 polite I'm a douchy American and she's
1:40:24 very very nice thank you for for um
1:40:28 cluing me in as to what was being said I
1:40:30 but it just it's still so I'm curi I'm
1:40:33 curious Lazarus um like does it sound
1:40:36 good like does the is the accent decent
1:40:39 is the language good or is it is it
1:40:40 awkward and weird I assume it's decent
1:40:43 apparently gp4 Omni is better at
1:40:45 language can you point to a website wait
1:40:49 can you point it at a website or did you
1:40:52 just feed it docks from your twin you
1:40:54 can do
1:40:55 both you could say I could say I'm the
1:40:58 CEO of storyvine go go learn what you
1:41:01 can about storyvine or when someone asks
1:41:03 about what I do you know go to story
1:41:06 vine.com and talk about it yes the
1:41:08 accent in the language is fine yeah
1:41:10 that's the thing what I mean what's
1:41:11 what's just astounding to me so my
1:41:14 company storyvine we have a it's a it's
1:41:17 an automated video platform a video
1:41:19 storytelling
1:41:20 platform May 1st was our 12th year
1:41:22 anniversary for 10
1:41:24 years we didn't have transcriptions in
1:41:27 the system because the automated
1:41:29 transcriptions were so bad that it was
1:41:33 just like it was horrible and then like
1:41:35 overnight whisper came out and was like
1:41:37 oh now it's good and then like 3 months
1:41:40 after that these this language capacity
1:41:43 came out it's just like all right now
1:41:46 now they can do translation and then hey
1:41:48 Jen did the lip sync thing to match the
1:41:50 voice to the lip sync it's absolutely
1:41:53 crazy just posted my song bin oh Song in
1:41:57 Spotify nice fantastic how did you put
1:42:01 it in in Spotify don't you have to go
1:42:03 through some
1:42:07 publisher let's see what let's see what
1:42:09 Mr it made we're going to go to AI
1:42:13 learning lab
1:42:15 Irregulars this version is
1:42:17 better VI cinti
1:42:23 severe thunderstorm warning till 9:30
1:42:27 well it's past
1:42:29 [Music]
1:42:43 9:30 pretty good
1:42:48 [Music]
1:42:58 [Music]
1:43:00 that's
1:43:02 beautiful I don't know what it's saying
1:43:04 but it's
1:43:06 beautiful
1:43:08 Blanco Mr it wrote a song about the
1:43:11 white boy ah I see I see how we're
1:43:13 playing this oh
1:43:15 Blanco Blanco Diablo I see I see where
1:43:19 this is headed Mr it I think that is
1:43:22 quite aggressive of you
1:43:24 it's very
1:43:29 aggro I got the same version in
1:43:31 [Laughter]
1:43:36 English that's pretty cool that's really
1:43:38 good really good all right let's see
1:43:40 what else we got going in the regulars
1:43:42 Channel nice song Mr it I like it I
1:43:44 approve
1:43:46 solid solid action got another one in
1:43:50 there yeah it's a beautiful song it
1:43:51 really is all right I wrote a Diddy as
1:43:54 in P Diddy a piss I've gone in [ __ ]
1:44:00 myself I was sitting on my front porch
1:44:04 sipping on some
1:44:06 roof when all of a sudden I fell to
1:44:09 rumble my stomach
1:44:13 too kind of hold it in but it was coming
1:44:16 on
1:44:19 strong a shoot yes uh I have talked to
1:44:23 about Alpha fold in here but I'll I'll
1:44:25 talk about it again I mean I I think
1:44:28 Alpha fold's amazing Shannon Sky 2028 it
1:44:31 couldn't possibly be worse oh yeah look
1:44:34 so this is I don't know if you guys knew
1:44:36 this but uh we don't talk politics in
1:44:38 here other than the fact that I'm
1:44:40 running for president in 2028 with AI as
1:44:43 my running
1:44:44 mate and and the the the uh the the
1:44:48 platform is that what I will promise is
1:44:51 that AI will run the country I'll just
1:44:52 be the spokes model and then the the
1:44:55 tagline of campaign it couldn't possibly
1:44:58 be
1:45:04 worse I wish Sumo sunno would time the
1:45:07 endings better seems like it tries to
1:45:09 fit 410 into four in four minutes um so
1:45:13 long serenade there's probably I would
1:45:16 if I were you I would go to Theo Discord
1:45:19 and
1:45:20 Theo maybe on Reddit they're talking
1:45:22 about it as well I know that within
1:45:24 sunno there's all sorts of tags you can
1:45:26 put in there so I think you can put in
1:45:28 like brackets around the word drop and
1:45:32 it will end a
1:45:33 song um there's there's ways to get them
1:45:36 to end better than they do on their
1:45:39 own and it might be worth trying I mean
1:45:41 they just launched this 3.5 model that
1:45:44 they launched today um they said it's
1:45:46 got better song structure I don't know
1:45:48 if it's got better better
1:45:49 endings create a song with Barry White
1:45:52 style you can't use um I don't think you
1:45:55 can
1:45:57 use let's see
1:46:02 um
1:46:05 um what is the
1:46:09 best
1:46:13 musical
1:46:14 [Music]
1:46:16 description
1:46:19 for Peak Berry White
1:46:33 song uh no
1:46:36 what
1:46:38 specific
1:46:51 genre Lush from antic 70 Soul and
1:46:56 R&B all right let's grab that we'll go
1:46:59 over to
1:47:01 sunno and we'll do it with
1:47:05 these all right that's good all right
1:47:09 we're going Berry
1:47:14 White oh Alpha fold three so a shoe um
1:47:18 yeah I saw Alpha fold three so if you if
1:47:20 you okay
1:47:23 let's
1:47:25 see Alpha fold
1:47:28 2 wait I'm I'm getting all my all my
1:47:31 things mixed up yeah Alpha fold
1:47:34 2 in a
1:47:38 year predicted
1:47:40 430,000 I think I think that's the
1:47:44 number protein folding things Alpha fold
1:47:48 3 which just launched did it at a much
1:47:51 higher accuracy and it didn't just do
1:47:53 proteins it did they called it all of
1:47:56 the molecules of Life predicts how they
1:48:00 um how they exist and how they
1:48:05 interact so Alpha fold 3 the impact of
1:48:08 alpha fold 3 I like I would be very
1:48:11 excited to be in a in a biolab right now
1:48:14 because I assume that Alpha fold 3 is
1:48:16 just like you know let's go cure some
1:48:19 diseases and [ __ ] is is what it seems
1:48:22 like to me but I don't know enough about
1:48:23 that world to know how profound it is
1:48:25 but it seems pretty [ __ ] profound how
1:48:28 late am I going to go I'm not going to
1:48:29 go much longer ton ski or Todd we're not
1:48:33 going to keep you up too much
1:48:36 later it needs secure something right
1:48:39 now L I know it it really does I've I've
1:48:42 trust me I've got you know two boys
1:48:44 upstairs with chronic lyme disease and
1:48:46 Co infections right now so we need to
1:48:49 [ __ ] fix that [ __ ] up so all right
1:48:52 here Berry White uh behind the
1:49:01 curtain to play the leading
1:49:05 wall loaded into
1:49:09 [Music]
1:49:13 fantasy let's see if we get one this
1:49:15 little romantic ear
1:49:19 [Music]
1:49:27 he all right so wait
1:49:31 romantic 70
1:49:35 Soul
1:49:38 R&B I'm just going to say
1:49:44 deep baritone vocals
1:49:55 all
1:49:55 right
1:50:00 shank what happened to the promised deep
1:50:02 velvety tone so Lazarus here's the deal
1:50:05 right now all of these tools this is the
1:50:09 the single
1:50:13 greatest phrase that you can remember
1:50:15 for the next year and a half or
1:50:18 two is embrace the Jank
1:50:23 just know that these tools are not very
1:50:27 predictable and the more you want them
1:50:31 to give you exactly what you want the
1:50:33 more frustrating they get so so just let
1:50:37 them kind of do their thing and here's
1:50:40 how I've been using AI more than more
1:50:43 than as a tool to produce is I use it as
1:50:46 an ideation tool where I'll just say
1:50:48 give me an idea give me another idea no
1:50:50 that's not even close give me another
1:50:51 one give me another one give me another
1:50:52 one and I just just like literally like
1:50:54 just flushing crap down the toilet fast
1:50:56 fast fast bang bang bang and then if I
1:50:58 find something good I like then I'll
1:51:00 hone in on that and try to refine it and
1:51:02 that's a a very often multi-hour
1:51:05 frustrating
1:51:08 [Music]
1:51:13 experience to
1:51:16 theing it's not ber white it's more like
1:51:19 uh Luther vandros right
1:51:24 [Music]
1:51:29 yeah this isn't good all
1:51:32 right oh Champ's dreaming Champ's
1:51:34 dreaming that's cute yeah Luther exactly
1:51:37 it was Luther yeah Luther the first one
1:51:39 was
1:51:41 [Music]
1:51:46 Luther world to play the leading Ro
1:51:53 uploaded into
1:51:55 fantasy where he could
1:51:59 feel a life that
1:52:04 so it's pretty
1:52:06 [Music]
1:52:08 good in this dream he thought he
1:52:13 knew Whispers In The
1:52:18 Silence his o That's nice searching for
1:52:22 the way a truth be hosts to
1:52:27 find behind the
1:52:30 [Music]
1:52:37 curtain chorus sucks sucks man that
1:52:41 chorus
1:52:43 blows all right we didn't build any gpts
1:52:47 tonight did we
1:52:49 listen I told you I told you you can you
1:52:53 can run surveys you can create an agenda
1:52:55 in here the chances of me following
1:52:59 it I feel like I've lost some
1:53:03 conversations in conversation 40
1:53:07 interesting
1:53:11 huh what do you what do you mean lost
1:53:13 some conversations it was my fault yes
1:53:17 it's your it's your fault Mr
1:53:19 at um
1:53:23 this should make your chest bump what
1:53:26 that
1:53:27 song I don't know what you mean you've
1:53:29 lost some conversations I've lost janky
1:53:32 Berry White is my new favorite artist
1:53:35 name oh that's really good Lazarus yeah
1:53:37 exactly embrace the Jank janky B Berry
1:53:40 White yeah here's you want to hear janky
1:53:42 Berry White here let's do it it's this
1:53:44 dude down
1:53:47 here he built himself aect to the
1:53:51 leading Ro
1:53:52 [Music]
1:53:54 into
1:54:01 fantasy I feel like there's some news
1:54:03 I'm missing yeah there probably is
1:54:07 um what happened
1:54:10 today that that showrunner thing came
1:54:14 out which it's going to be interesting
1:54:17 to see if it's if it's any good or if
1:54:18 it's like LTX where it's it's not quite
1:54:20 ready for prime time I'm sure it's going
1:54:21 to be like that
1:54:23 how much manual editing they need to do
1:54:25 to actually make those shows they aren't
1:54:27 showing in history only a few oh that's
1:54:30 interesting yeah you know what soong
1:54:33 serenade is I um I lost my memory like
1:54:37 when memory first came out I was I was
1:54:40 consciously adding some [ __ ] to it and
1:54:42 then just one day I went in and I was
1:54:43 like yeah you don't have any there's
1:54:44 nothing in your memory you know start
1:54:46 adding
1:54:47 now and I I waited for a few days trying
1:54:50 to see if it would come back and it
1:54:51 never did so that was
1:54:53 weird weird weird weird weird weird AI
1:54:56 wetwear in Switzerland is news yeah the
1:54:59 the uh that so
1:55:02 so what they've built is they built
1:55:05 these things called brain
1:55:10 organoids first of all if you're working
1:55:13 on something where you're doing testing
1:55:15 on brain matter don't call it an
1:55:17 organoid don't call it something that a
1:55:21 science fiction writer would make up
1:55:26 okay don't call the thing that scans
1:55:28 your retina the
1:55:30 orb open AI oh sorry too
1:55:35 late and if you're making brain matter
1:55:37 that you plan on putting on a computer
1:55:39 chip don't call it an
1:55:43 organoid that's how every AI movie
1:55:45 starts well Bob looks like the organoids
1:55:49 have decided to turn into full-on organs
1:55:51 what are we going to do
1:55:54 uh hey Jim the organoids
1:55:57 [Laughter]
1:56:02 blinking at which point do we not call
1:56:05 it an organized organoid and call it a
1:56:08 oh
1:56:12 [ __ ] oh my goodness so
1:56:15 anyway these Swiss
1:56:19 folk is that what you call the Swiss
1:56:21 folk these Swiss folk yeah I think
1:56:25 so these Swiss folk took some
1:56:30 uh stem
1:56:33 cells definitely harvested in an
1:56:36 appropriate manner I'm
1:56:38 sure don't give me your
1:56:41 [ __ ] and they made brain organoids which
1:56:45 were like little teeny pieces of brain
1:56:48 tissue that functioned like little teeny
1:56:51 pieces of brain tissue and they have now
1:56:53 made a chip that can hold 16 of them and
1:56:57 you can send instructions into these
1:56:59 organ into these organoids and they
1:57:02 process them or not instructions but I
1:57:04 guess bits or btes I don't [ __ ]
1:57:08 know and apparently someone said you can
1:57:10 buy them you can they're selling them
1:57:12 that you can actually buy to use
1:57:14 probably in research I
1:57:16 assume all have access to gpts oh yeah
1:57:19 so so another piece of news is
1:57:23 yesterday if you don't know this 75
1:57:25 people in here if you don't know this
1:57:28 yesterday chat GPT made their most
1:57:31 powerful model GPT 4 available to
1:57:34 everyone so if you go to chat gp.com
1:57:37 you get to use the model that is it's
1:57:42 this full-on um what they call
1:57:45 multimodal which won't mean a a ton
1:57:48 right now but it's got like all the
1:57:50 bells and whistles it can do Advanced
1:57:52 data analysis it's connected to the
1:57:54 internet it will be able to make images
1:57:57 I don't think it can do that right now
1:58:00 it was it was restricted yesterday let
1:58:02 me let me actually try it and
1:58:04 see see if they fixed that little piece
1:58:07 from the launch
1:58:10 yesterday chat jpay.com
1:58:14 all right I'm in here
1:58:21 uh make me an image of 16 brain
1:58:24 organoids on a
1:58:28 [Laughter]
1:58:33 chip oh wow it said oh that's
1:58:35 interesting what it did look what it did
1:58:37 I I said create an image and it said
1:58:40 sure let's create an image here's the
1:58:42 concept chip layout a grid-like chip
1:58:44 layout with 16 compartments brain
1:58:49 organoids oh that did it with code
1:58:51 interpreter that's
1:58:58 interesting I'd like an illustration
1:59:02 please so this if I if I click on this
1:59:07 this is wild so one of the thing it's
1:59:09 got It's got this thing called Advanced
1:59:11 Data analysis so when I asked it to make
1:59:13 me an image it it thought I wanted a
1:59:17 schematic right so let me put this
1:59:19 together the schematic for you how it
1:59:21 drew this was it actually wrote python
1:59:25 code to generate that image so what code
1:59:29 interpreter does is it actually writes
1:59:31 and execute code for you so you could
1:59:33 upload a spreadsheet to chat gp4 now see
1:59:37 this
1:59:37 little paperclip you can upload a
1:59:40 spreadsheet to that what about my
1:59:43 subscription okay here's the thing
1:59:47 Jim
1:59:49 um um no
1:59:53 I want a
1:59:56 photographic
1:59:59 image with
2:00:12 Dolly it's not do it that's
2:00:17 weird let start a new
2:00:20 chat make me an illustration
2:00:24 make me an illustration of 16
2:00:28 organoids on
2:00:30 a
2:00:36 microchip huh it's doing the same
2:00:41 thing maybe it's because I put it on a
2:00:43 microchip hang on let me try something
2:00:49 new photo of a 1970s
2:00:52 photo of a 70s muscle car in an
2:00:55 abandoned Factory with sunlight
2:00:56 streaming Through the
2:01:05 Windows oh they okay yeah so it's not
2:01:08 making it's not making images at
2:01:12 all can you make a
2:01:16 picture of
2:01:20 a ghost in in
2:01:29 a
2:01:32 upper
2:01:38 window using
2:01:41 python all right so if you've got the
2:01:43 free version you don't have access to do
2:01:45 right now I don't know why that is and
2:01:48 you've reached your data limit okay so
2:01:50 basically
2:01:52 even though they gave access to chat GPT
2:01:55 to everyone you get about five prompts
2:01:59 and then you're
2:02:02 done um
2:02:04 but you can at least get you you can at
2:02:07 least get an understanding of what it
2:02:08 makes possible it's it's pretty crazy
2:02:10 good to answer your question about why
2:02:13 pay your 20 bucks one is because you saw
2:02:16 how quickly that rate limited out yeah
2:02:19 now I'm in chat gbt 3.5 um that rate
2:02:22 limited out in like five prompts so you
2:02:26 get five times that with if you have the
2:02:29 $20 subscription and I think it's more
2:02:31 than five times that they they it seems
2:02:33 like they're pretty severely limiting
2:02:34 this I I assume that's going to get
2:02:36 better over
2:02:38 time
2:02:40 um
2:02:45 and the other
2:02:48 thing GPT Plus members the 20 Buck a
2:02:51 month members are going to get access to
2:02:53 the real-time voice and the new image
2:02:56 generation model sooner and it's also
2:02:58 got advantages right now like our the
2:03:01 the paid gp4 Omni when it generates
2:03:04 charts and graphs they're interactive um
2:03:07 we have access to doly 3 which
2:03:09 apparently this one does not for some
2:03:10 reason I guess just for cuz Dolly 3 is
2:03:13 probably
2:03:14 expensive
2:03:16 um and then when the new stuff comes out
2:03:19 when the new models come out this summer
2:03:21 and next November um GPT Plus members
2:03:24 are going to have access to them before
2:03:25 free members so um so this is a an
2:03:30 incredibly powerful thing it's just
2:03:32 fairly
2:03:33 limited I wonder actually you know
2:03:35 what's Curious let me
2:03:39 go what oh that's the app let me get rid
2:03:43 of
2:03:43 that if I go to Sin City
2:03:48 sipper um
2:03:52 let me see make me a
2:03:58 martini okay so even though I rate
2:04:00 limited out with GPT
2:04:08 4 I can use custom gpts
2:04:13 but I can't create images right
2:04:17 now that's fascinating all right they
2:04:20 got all sorts of weird ass
2:04:22 weird ass restrictions on it so it's
2:04:26 going to be using gp4 is going to be
2:04:29 frustrating unless they get that they
2:04:30 got to get that rate limit up to like 20
2:04:32 interactions or at least 10 that you can
2:04:35 at least do a
2:04:37 session like I I barely got started
2:04:39 there and I was like yeah we're good
2:04:40 you're done done sorry Bob yeah Bob hey
2:04:44 Bob yeah listen we're going to need you
2:04:46 to go on ahead go go go suck a butt yeah
2:04:50 Bob Bob listen yeah about wanting to use
2:04:53 that chat GPT
2:05:01 40 Vision Works on GPT 40 yeah you can
2:05:04 upload images to it it's pretty amazing
2:05:07 when do you think we'll have access to
2:05:08 GPT Vision you've got it right now so if
2:05:11 I
2:05:13 upload see if I can do let look up
2:05:17 whiteboard
2:05:32 what
2:05:34 is
2:05:37 this so I uploaded a picture of a
2:05:42 whiteboard yeah so right now even even
2:05:46 if I'm rate limited
2:05:49 out this little upload thing I can still
2:05:53 upload CSV files and do data analysis I
2:05:57 can upload images and see them so so
2:05:59 this is this is this is still a pretty
2:06:03 big deal that they gave all of this
2:06:04 stuff out for
2:06:08 free using my phone I can scan my room
2:06:11 and it will tell me exactly what I'm
2:06:13 looking at
2:06:15 fascinating how many did it say
2:06:18 um I asked GPD 40 how many paid subscri
2:06:22 subcribers it has and it told me that's
2:06:25 interesting it it probably hallucinated
2:06:28 but you
2:06:29 know Vision Works on chat GPT 40 when do
2:06:32 you think we'll have access to
2:06:34 vision test image capabilities in
2:06:37 version 4 it didn't work because you
2:06:39 tested in
2:06:40 3.5 yeah but I'm I'm rate limited out in
2:06:43 four
2:07:04 chat GPT plus maybe I wasn't in chat GPT
2:07:10 40
2:07:12 H weird
2:07:18 [Music]
2:07:22 sky is still
2:07:31 missing Fascinating People just
2:07:33 fascinating fascinating will creating
2:07:36 gpts come to the free version probably
2:07:39 not um if you look at um Microsoft
2:07:43 co-pilot anyone can use a co-pilot GPT
2:07:47 but you have to be a paid member to make
2:07:49 them same thing right now for um GPT 40
2:07:54 listen quite frankly 20 bucks a month if
2:07:58 you know there there's a lot of things
2:08:00 you could be spending money on right now
2:08:02 in in terms of AI
2:08:04 subscriptions in my opinion the 20 bucks
2:08:06 a month for chat GPT is the best deal in
2:08:09 software it has been for a while and I I
2:08:13 like I think it's even with the fact
2:08:15 that they gave this away everyone's like
2:08:17 well what's my 20 bucks for well you can
2:08:18 actually [ __ ] use it for one thing
2:08:22 right it doesn't rate limit you out in 3
2:08:25 seconds um and it's got we're going to
2:08:27 have access to the other tools faster I
2:08:29 think that right now it's the it's
2:08:31 probably the best investment save
2:08:33 yourself 20 bucks a month at Starbucks
2:08:36 and put it into chat GPT and really
2:08:38 learn how to use the [ __ ] out of it
2:08:41 because as it gets better and better if
2:08:44 you've got your head around all the
2:08:45 stuff it can do when the new [ __ ] comes
2:08:48 out you're going to be like a superhero
2:08:52 all
2:08:54 right it pays for itself in a day yeah
2:08:56 seriously agreed I make gpts for every
2:09:00 work project yeah fascinating road
2:09:02 warrior princess I um so someone told me
2:09:06 this once when when gpts first came out
2:09:08 this blew my mind he said he got off a
2:09:12 conference call with four people he took
2:09:14 the transcript of the conference call he
2:09:16 made a GPT of the conference call
2:09:19 conversation and sent it to the four
2:09:20 participants
2:09:22 and said and said you you can now ask
2:09:26 questions of the conversation we just
2:09:29 had like that starts to get really cool
2:09:32 when you can actually have the avatars
2:09:34 of those people where you could so
2:09:36 imagine imagine having you have a
2:09:39 conference call and then you get the
2:09:41 transcript of the conference call but
2:09:43 you say I want to put it into
2:09:44 conversation mode and it goes right back
2:09:47 into a zoom call where you've got the
2:09:48 same participants and you say hey Jim
2:09:51 could you clarify what you were talking
2:09:52 about and Jim will just start talking
2:09:55 based on the notes of the
2:10:02 call you can save the extra $10 for
2:10:05 co-pilot just by going through the UI
2:10:08 that's actually really cool I know right
2:10:11 yeah JLo's here what's happening
2:10:13 Jennifer Lopez what's shaking what's
2:10:15 going down well worth of 20 bucks I have
2:10:17 three paid accounts there you go it's a
2:10:20 little Overkill but go for it especially
2:10:22 if you need
2:10:23 it um pays for itself in day yeah yeah
2:10:26 yeah I can't do that right now I'm I'm
2:10:29 sorry how no I'm sorry Dave I'm afraid I
2:10:32 can't do that right
2:10:34 now how let me
2:10:39 in going to bed all right yeah I got to
2:10:41 go to bed too all right listen it is
2:10:43 10:10 it is Thursday night tomorrow
2:10:46 Friday night date night um so uh you
2:10:51 know Mark your calendars and uh bring a
2:10:54 date or not you be the date and we will
2:10:57 uh we'll hang out tomorrow
2:11:00 um I think I think what I will do
2:11:02 tomorrow I think I think what we'll do
2:11:04 tomorrow is I'll start out we'll start
2:11:06 out building a GPT building a custom GPT
2:11:10 cuz I haven't built one in a while and I
2:11:12 think it's actually important for all of
2:11:14 us to see if they're different um one
2:11:18 thing you should know if you've built
2:11:20 gpts
2:11:23 they all now use the new model gp4
2:11:28 Omni it's
2:11:31 different like I noticed last night in
2:11:34 my GPT for Sin City sipper that it was
2:11:38 weird with the recipes the recipe
2:11:40 quantities were not right and they were
2:11:43 always right before like they were
2:11:45 always like amazingly good and last
2:11:48 night they were weird so if you've built
2:11:51 a GPT you should go test your gpts the
2:11:54 other thing is you should go market your
2:11:56 gpts if you want people to use them
2:11:58 there's now a 100 million active weekly
2:12:01 users that have access to your gpts so
2:12:04 all
2:12:05 right GPT 4 Omni tomorrow night we will
2:12:09 do that thanks everybody follow my
2:12:11 channel if you haven't followed it um
2:12:13 subscribe to the lives if you want to
2:12:14 support me thank you Amelio's wife um in
2:12:18 the bottom corner down there is a uh a
2:12:21 video series you can also buy it where I
2:12:23 talk about why I think AI is actually
2:12:25 underhyped not
2:12:28 overhyped um thank you Mr Mr it for the
2:12:32 Roses very very nice gift thank you so
2:12:35 much thank you so much oh and office
2:12:38 hours tomorrow I will I will make sure I
2:12:41 get a reminder out I I didn't do that
2:12:43 yesterday or last week um at 11:00 a.m.
2:12:46 mountain time I do an AI office hours
2:12:51 just it's like a little hang session um
2:12:53 so if you're into this ai ai stuff and
2:12:56 want to talk with it's the group of
2:12:57 people that shows up is really
2:12:59 interesting they know what's going on
2:13:00 with this AI stuff really it's a really
2:13:02 powerful group that's at 11:00 a.m.
2:13:04 tomorrow check my account on LinkedIn
2:13:06 I'll put a reminder in the morning with
2:13:08 the link to it and uh yeah that's it so
2:13:13 thank you Irregulars Irregulars are
2:13:14 people who just show up here all the
2:13:16 time if you want to find a community if
2:13:19 you want to be in a community of AI
2:13:21 enthusiasts that rocks go to the
2:13:26 salon. and sign up it's free but it's an
2:13:30 amazing community so go there and check
2:13:32 that out um don't be a douche and don't
2:13:35 go in there and
2:13:36 spam some people are coming in here just
2:13:38 spamming away it's like really just
2:13:41 don't be a douche just in general this
2:13:43 is like here's a life lesson from Kyle
2:13:45 I've been around for 59 [ __ ] years
2:13:48 you know what I learned don't be a
2:13:50 douche it's that simple it really is
2:13:53 that simple and I'll tell you what one
2:13:56 of the things that AI is going to do
2:13:59 because AI is going to democratize
2:14:01 skills you know how sometimes you hire
2:14:04 someone because they're good at
2:14:05 something they're good at the skill but
2:14:08 they're a tool and an [ __ ] and you
2:14:11 hate them but you have to work with them
2:14:13 because they're the only ones that you
2:14:14 know that knows that [ __ ] skill guess
2:14:17 what AI is going to create a reckoning
2:14:20 for asshol
2:14:23 so don't be an
2:14:26 [ __ ] because the AI is going to do
2:14:29 the skill part it's going to do the
2:14:31 Tactical execution and you know who
2:14:33 people are going to work
2:14:35 with not
2:14:37 [ __ ] they're going to work with
2:14:39 people they trust people that have good
2:14:41 work ethics people that you'd want to
2:14:42 hang out
2:14:46 with all right Romper Room time Allan
2:14:50 good to see you Joe good to see you
2:14:53 Danielle thank you very much
2:14:56 ASU Amelio's wife in the house tonight
2:14:59 Amelio's wife also caught the the live
2:15:02 uh champ walking uh situation today it
2:15:05 was it very limited
2:15:07 audience watching Champ's cute little
2:15:09 butt walk around the streets of Denver
2:15:12 nighter Kyle Jox of the mighty good
2:15:14 night good night Tobias good night to
2:15:16 you Mr it congratulations on your
2:15:20 songs uh
2:15:23 Allan good to see you Mr
2:15:25 Studio
2:15:28 sh shainsa is in the house
2:15:32 tonight all right I think I got
2:15:34 everybody there probably not fireflies
2:15:36 in the house nice good to see you
2:15:39 against the Norms in the house Jennifer
2:15:41 Lopez in the
2:15:43 house all right everybody uh I am gone
2:15:46 spin B3 good to see you fun guy Robert
2:15:49 mezer thank you good night all all right
2:15:53 um I'm out see you Friday night date
2:15:54 night it's not first Friday so it should
2:15:57 be normal it should be 8:00 P.M Mountain
2:15:59 tomorrow all right peace I will see you
2:16:02 then