Dec 13, 2023 - AI's Creative Frontier: Making Original Art and Stories with Chat GPT & Midjourney

Description

Chapters

Transcript

0:02 in a Westerly
0:07 Direction this car is my
0:12 train I've been driving I've been
0:15 [Music]
0:17 wondering what it is I'm running from
0:22 again feel like an 8 old
0:26 man holding on to 29
0:32 up on that
0:35 Horizon no singing is a California
0:45 [Music]
0:56 L cute little front door on Windows my
0:59 Lord
1:01 ain't sure whether the crow should
1:04 [Music]
1:05 laugh she broke home up myself
1:10 once as I stumbled to that
1:14 door I read a note by the d
1:19 light said don't you come around
1:24 here
1:26 [Music]
1:29 anymore
1:30 [Music]
1:38 he work at the D
1:40 Queen does listen to rock and roll does
1:44 he feed the me
1:48 hallelujah
1:49 hallelujah
1:52 hallelujah
1:54 hujjah yeah hello hello hello
1:59 everybody
2:04 [Music]
2:23 [Music]
2:28 you didn't like those notes did you that
2:30 was dissonance you didn't like
2:32 dissonance
2:34 a good evening Amelio's wife is on time
2:39 show us the dog
2:41 champy
2:43 champy where you going you leaving you
2:45 need to go lay down you want to say hi
2:50 [Laughter]
2:53 hi
2:55 and welcome to the AI learning lab
2:59 everybody
3:02 all right my dog is pacing and whining
3:05 big beginning songstress yeah it takes
3:07 it takes years to get you know for the
3:10 training for the training and the it's a
3:13 lot of discipline to get your dog to
3:15 sing like that step one go to a
3:18 pound step two pick a dog that no one
3:22 wants step three bring him home and
3:25 strum your guitar see if he starts
3:29 singing
3:31 he did first
3:35 drum it's pretty
3:38 cute oh
3:42 man all
3:44 right let's get going here welcome
3:47 everybody to the AI learning lab uh it
3:50 is Wednesday
3:53 night um a lot going
3:56 on Sam Alman said today that uh chat GPT
4:02 4 is available
4:04 again if if I were you and could afford
4:08 the 20 bucks a month I'd go grab it now
4:10 if you
4:12 haven't uh cuz who knows they might Doo
4:16 that again but looks like they scored
4:18 some gpus somehow I guess Microsoft spun
4:21 up some new center
4:25 somewhere um korea. went public
4:32 um Mid Journey if you've made more than
4:36 10,000 images in mid Journey you can now
4:39 go to Mid journey.com if you go to
4:41 alpha. Mid journey.com and you've made
4:44 more than 10,000 images you can start
4:46 using the web interface from mid Journey
4:48 which everyone that I've seen is was
4:50 like this is cool never going back now I
4:53 don't have access to it cam kin hi Kyle
4:55 how are you I'm good I'm good I am I am
4:58 uh I am am hopeful that this evening's
5:02 live goes more swimmingly than last
5:05 night's I was just totally incoherent
5:07 last night I think I think it was a it
5:10 was a uh it was a special night for all
5:13 involved I don't know why people hung
5:16 out last night I just been like I'm good
5:18 we're out of here use chat GPT to read
5:20 my old family handwritten letters am I
5:22 wearing the right glasses no I'm wearing
5:23 my driving glasses what the [ __ ] am I
5:26 doing why can I not read ah ah
5:30 for you youngsters out there if anyone's
5:33 on the channel that's in the uh in the
5:35 uh sub30 range enjoy
5:40 it the eyes just [ __ ] just
5:51 go but they're going to cure all that
5:53 [ __ ] AI is going to cure all that all
5:55 the diseases are going to be gone
5:57 everything's going to be shangrilla it's
5:59 all going to be perfect pied they don't
6:02 hallucinate that's a myth hallucinations
6:04 are a myth a AI will cure body pains for
6:07 sure that would be nice I was kind of
6:08 sore all day today I don't know why
6:10 Korea AI is a trip what kind of safety
6:12 guard rails does it have I have no idea
6:14 I don't so Korea as far as I
6:18 know is just a wrapper app sitting on
6:20 top of stable diffusion open source
6:22 stuff so it may not have a ton of
6:25 uh it may not have a ton of safety guard
6:28 rails bring the dog back
6:32 oh Brad you're awesome thank you yes
6:34 good troll solid I like it
6:38 um yeah if the dog knew anything about
6:40 AI he'd run the channel cuz this this
6:43 would be a billion billion follower
6:46 Channel if that dog were running the
6:48 joint that dog's cute he's cute he's
6:51 cute he's wide it's a wide it's a it's a
6:54 lot of dog there it's a it's a wide dog
6:57 it's a it's flatback
7:04 oh man all
7:07 right kore.ai generated some sketch
7:10 images that I did not ask for yeah it
7:13 probably does not have Yan safety guard
7:15 rails which will make some people happy
7:17 and other people like cuz I'm not using
7:20 that at
7:21 work can jeeper help developing an app
7:25 for
7:25 iOS yeah yeah it it can probably write a
7:29 amount of the code it will Coach you on
7:32 how to do it I would be surprised at
7:34 this point if someone didn't come up
7:35 with a
7:36 GPT uh to help you build an IOS app so
7:39 you can probably go find that on the
7:41 Google uh P's in in here so when you
7:46 said something was coming this morning
7:48 Pate I didn't put two and two together
7:50 two and two together that yesterday was
7:52 the 12th so the other thing that
7:55 happened is um Google announc or
8:00 released made access
8:02 to the
8:05 Gemini apis for pro and Nano available
8:09 to developers today but you know what I
8:11 saw Pate McKenzie Bose who did um GPT
8:14 boss um he did a couple of videos today
8:17 apparently Canada doesn't have access to
8:19 the Google apis so so if you pay for
8:23 Google workpace or if you're Canadian
8:25 Google's like [ __ ] you everyone but you
8:30 at least mid journey I understand if you
8:32 haven't made 10,000 images [ __ ] off the
8:35 the people that have proven themselves
8:37 mid Journey loyalists we're going to
8:39 give them first crack at the thing which
8:40 I think is
8:42 smart um but anyway yeah so chat gp4 can
8:46 probably help uh with almost all aspects
8:49 of the IOS app you're still going to
8:51 have to like it'll tell you install X
8:53 code install these libraries do these
8:55 things it's glitchy bad Kyle you know
8:58 what when I started
9:02 people are saying good things about Mist
9:03 AI so Mist AI looks really amazing it's
9:06 a powerful free llm yeah Christopher
9:08 Ramsey so um thanks for being a
9:11 subscriber by the way uh and YouTube Be
9:13 not Afraid sorry it's glitchy I might
9:16 bail on this and come back in uh I might
9:18 need to restart the r let me actually go
9:22 check and see if Yan boys are
9:24 downloading you know I don't know
9:27 whatever the [ __ ] they download
9:58 you
10:01 all right champy we found the
10:03 culprits we found the culprits you're
10:06 going to have to bite their ankles
10:08 later all right we found
10:11 it oh yeah looks like stream was
10:13 downloading something oh does it does it
10:17 so it's like it's not like you don't
10:19 know I do this every night
10:28 right all right mine froze with Kyle's
10:33 eyes closed is it back are we better now
10:36 are we more better now it's better yay
10:40 thank you for speaking up if you see
10:43 something like glitching say something
10:45 and I'll go yell at the
10:47 [Laughter]
10:51 boys okay Define better oh sorry about
10:56 that there Wisconsin
10:58 condos how old are they they are 23y old
11:01 twin
11:03 boys more
11:08 better well well yeah if you're talking
11:11 about quality better no if you're
11:14 talking about bandwidth
11:20 better he never uses the BRB
11:23 slide what's the BRB oh the B right
11:26 backslide oh yeah hang on wait you know
11:28 what I can do hold please we're going to
11:31 create a new
11:33 album good idea good idea a i l BRB I'm
11:38 I'll do a special because one of the
11:40 problems with it is is that
11:46 uh
11:48 um I got too many images so I don't know
11:51 where the BRB slide is in this mix so
11:54 let me go so okay so for those of you
11:58 new here my name's Kyle shanon this is
11:59 the AI learning lab there's a bunch of
12:01 people in here called the the
12:04 Irregulars K wait is Kyle also a twin
12:07 Kyle is not a twin the they're identical
12:09 twins which is not genetic it is just
12:12 random it is just a random situation
12:15 that egg
12:17 split and made two
12:21 babies we were about to go on in vitro
12:24 that that was a [ __ ] trip we're we're
12:26 like literally like a month from going
12:28 on in vitro
12:30 Gabby got
12:31 pregnant we went in wait we went into
12:34 the fertility doctor and he goes he goes
12:37 oh yeah you're looks like you're
12:40 pregnant those are heartbeats yeah oh
12:43 yeah oh I said I said what's what is
12:45 that why are there two he goes oh yeah
12:49 it's twins yeah probably won't work out
12:52 come back in two or three weeks we'll
12:53 vacuum you out and start
12:55 over like
12:58 what we walked out of that place and
13:00 never came back he was disappointed
13:02 because he saw his 25 or 50 Grand
13:04 flushing down the toilet natural
13:07 pregnancies do not make fertility
13:09 doctors happy at least at least ones on
13:11 the Upper East Side of New York Upper
13:13 West
13:14 Side that was that was a [ __ ]
13:16 disturbing disturbing doctor's visit
13:20 question what can you tell us about
13:21 project silica from micro Microsoft um I
13:24 saw you post about that
13:26 earlier cam um that it's the like I read
13:31 about it two or three months ago where
13:32 they're storing massive amounts of um
13:35 data on glass like somehow writing it in
13:39 into glass but I don't know I mean I I
13:41 haven't heard anything about it champy
13:43 download oh wait sorry welcome to Chad
13:46 add let me go find a be right back so
13:50 there's a stay tuned all right so let's
13:52 drag that over there a i l be right
13:57 back I think think I have a couple of
14:00 them that's Happy
14:03 Birthday listening to
14:07 Kyle was that'll be right back
14:13 [Music]
14:16 no hold please that's a good one we'll
14:20 put hold please over
14:21 [Laughter]
14:27 there we I could do [ __ ]
14:30 storm
14:36 nah all right if anyone has made um be
14:40 right back or hold please AI learning
14:43 lab things will you post those to
14:47 um show and tell on the
14:51 salon and then we
14:54 will that Hedgehog's pretty cute that
14:57 would be a good that that would be a
14:59 good hold one yeah we'll do that
15:02 one
15:14 oops all right so so now when I go away
15:19 I've
15:20 got I've
15:22 got good uh good be right back
15:28 stuff all right Debbie I've been good
15:30 but glad it's good for every oh it's
15:32 I've been good but glad it's good for
15:34 everyone very good my eldest is 22 nice
15:38 youngest turns six on Friday wow that's
15:41 quite a spread there lunatick
15:44 dang ai. gooogle dodev is so much fun
15:47 Snappy fast Gemini ai. gooogle dodev so
15:51 this is going to be one of those
15:52 situations I will go there and if I
15:53 don't immediately get into it we're not
15:56 going
15:57 to Dad daddy daddy's not going to play
16:00 with it I think I might have gone here
16:02 last night and got very frustrated um
16:07 AI do Google dodev wait Google dodev
16:12 wait a minute that doesn't sound like
16:15 Google is
16:18 it all right maybe it
16:22 is oh this is the dev stuff yeah I
16:25 haven't played with it yet today
16:28 cool get an API key in Google
16:35 studio sign in with my Gmail that should
16:38 just snap me right in right I consent to
16:40 the generative AI policies I would like
16:42 to receive emails no I would would like
16:46 to receive invitations to participate in
16:48 research
16:50 studies nah not right
16:53 now use Google AI
16:56 Studio developing your own envir
16:58 enironment get API key can I get
17:02 one yeah looks like I can interesting
17:06 let me go back and is this Google AI
17:15 [Music]
17:21 Studio yeah cool all right good that
17:25 looks awesome will be fun to play in
17:27 there is Kyle also a twin Kyle is not a
17:32 twin how about dobard Google dobard you
17:36 mean bard. goole.com
17:38 that is so b.google.r is the one I
17:41 couldn't get into last night um we we'll
17:44 come back to all this we'll come back to
17:45 all this okay all right let's get
17:48 started here so welcome
17:50 everybody if you're new to the AI
17:52 learning lab we'll do lots of things
17:54 here tonight we'll we'll go play and
17:57 we'll I'll answer any questions if you
17:59 have questions about AI pop them in the
18:00 comments below uh if you're curious
18:03 about my qualifications it's super easy
18:05 I have none um no one is expert at this
18:09 generative AI stuff right now and that's
18:11 kind of what this channel is about this
18:12 is about curiosity and exploring and
18:15 just showing up and trying and trying
18:17 and
18:18 trying um and I'm happy to debate that I
18:21 know that there are AI experts in the
18:24 machine learning [ __ ] but the generative
18:25 AI stuff
18:27 is it's also so new there's too
18:31 much
18:33 um yeah so so pop any comments you have
18:36 below do me a favor if you would share
18:39 the live how you can support the channel
18:41 share the live share it share it with uh
18:44 with some of your friends that might not
18:46 be into AI especially ones that are like
18:48 why are you sharing AI crap with me
18:51 let's let's get them over here
18:55 uh
18:57 and you can tap some of the people like
19:00 doing the carpal tunnel tap uh to do the
19:04 like thing you can do that if you want
19:06 there's a uh there's a Ser a video
19:08 series in one of the corners here called
19:10 underhyped which is a 17 or 18 video
19:13 series that I put together about why AI
19:16 is important and and how it compares to
19:19 the early days of the web and how to use
19:21 it and things like that so is grock
19:22 Chained trained on open AI data I do not
19:27 know what it's been trained on it's
19:29 certainly been trained on all of X data
19:33 um and it has access to X data in real
19:35 time so what what grock is looking like
19:38 to me is it's like a pretty good Twitter
19:41 search engine it's like they upgraded
19:43 Twitter search right now it might get
19:45 better um it's like a smartass Twitter
19:49 search engine right
19:51 now uh but I don't know what it was
19:52 trained on I I assume he did not train
19:55 it on Open AI data because there was not
19:57 any good relationship ship there so I'm
20:00 assuming I'm assuming they did a little
20:03 something different if you know what I
20:05 mean um all right so one of the things
20:09 we did last night is I went through all
20:11 of
20:11 the not all of I I went through an
20:15 exercise where I showed a bunch of the
20:16 different things you can do in chat GPT
20:19 4 but what I didn't do was make a GPT so
20:23 we can do that tonight if you want we
20:25 can go look at um some of the multimedia
20:29 stuff I created a couple of new products
20:32 for my business Story vine that involve
20:35 doing avatars
20:37 and personas like
20:39 bringing non-real people to life
20:44 um and we could go look at that there's
20:46 some really interesting things 11 Labs
20:48 just announced or not announced they
20:50 released this new feature called Speech
20:52 to speech which is super cool maybe I'll
20:55 go show that first I don't know what I'm
20:57 going to do first what we going to do
20:59 why don't you lead me you tell me what
21:00 to do what do you want to do what do you
21:03 want to do people listen listen we could
21:05 do anything you want to do you want to
21:07 you want to go uh you want to just go
21:09 like what's chat J we could go there or
21:12 we could go like uh we could go like uh
21:15 like a computer computer science
21:17 programming like I could just do I could
21:19 go go deep deep deep into computer
21:21 science programming ma
21:24 mathematics no I can't I told you I
21:27 don't have qualifications speech to
21:29 speech all right you want to see speech
21:32 to speech
21:33 fine let me go show
21:40 you yes this one this is the one yes yes
21:45 this is the
21:46 one let's see bring personas to life
21:50 okay look at this what do I want to show
21:53 you here okay I'm going to show you a
21:54 couple of things we'll we'll we'll go
21:56 we'll go pretty deep here
21:59 all
22:02 right I know I'll make it bigger calm
22:08 down so impatient all of
22:11 [Music]
22:19 you 11 Labs yes Jim Buckley 11 Labs
22:23 indeed um okay so what I'm about to show
22:26 you is well it's 11 Labs under the
22:30 hood uh but this is hey
22:33 Jen so uh so one of the things that my
22:36 company Story vine does is authentic
22:38 video storytelling we have an automated
22:41 platform where we have an app that
22:43 somebody answers a series of prompts in
22:45 the
22:45 app um and they kind of just follow
22:48 follow the bouncing ball and it's kind
22:50 of like karaoke for storytelling they
22:52 just answer some questions that goes up
22:54 to the cloud 5 minutes later you got a
22:56 fully edited video with graphics and and
22:58 all that sort of happy horse [ __ ] so
22:59 it's pretty cool um but we're starting
23:03 to expand what it can do with AI cuz you
23:07 know I don't shut up about this [ __ ]
23:09 it's all I think about and I'm like hey
23:11 you know what we could do so here's one
23:13 of the things that we could do
23:16 um here we got a video of Ellen my
23:19 greatest source of support living with
23:20 rheumatoid arthritis has been um
23:23 speaking with other people about what
23:26 they've gone through uh reading a lot
23:29 about the effects of rheumatoid
23:30 arthritis so we do a lot in healthcare
23:33 and Pharma and so you know patient
23:36 patient story here hang on let make this
23:38 a little bigger a little more bigger so
23:41 you can see some
23:43 stuff uh and then here she is speaking
23:57 Spanish
24:04 and so what we can do is because our
24:07 system is modular we can take the
24:09 modules the the video modules of first
24:11 speaking English translate them to
24:12 Spanish put them back into our system
24:14 and generate uh a Spanish video and then
24:17 we can put it into this system that we
24:19 just built where it'll take the video in
24:22 Spanish and now generate Spanish
24:23 transcripts and a Spanish description
24:25 and Spanish chapters and Spanish blog
24:27 posts and things like that
24:28 super cool but here's the finished video
24:31 with Ellen uh speaking Spanish I'll just
24:33 put a little bit in
24:34 [Music]
24:45 [Music]
24:52 here and so I don't know if you saw it
24:55 but the the little lower third was in
24:57 Spanish so you know translate all the
24:59 lower thirds and everything like that
25:00 and here's Ellen speaking
25:02 [Music]
25:16 Hindi and let's
25:20 see there's half of the languages we can
25:23 translate into and then here she is
25:25 speaking
25:27 Japanese
25:31 so her voice you know all that sort of
25:33 stuff so that's super cool
25:35 now what we can also do is we could take
25:39 let's just say we just had audio
25:40 recording of Ellen telling her story but
25:42 we didn't have her image we could take a
25:44 picture of her we could turn that into a
25:47 say a dolly 3 or mid Journey image like
25:50 a little stylized sketch of her Hey
25:54 Siri hey Siri what do you want to search
25:56 for go away
25:59 what do you want to search for go
26:01 away what do you want to search for
26:08 boy all right um so we got audio of
26:12 Ellen we make a picture of her and then
26:13 we can bring that picture to life so we
26:15 can generate a video from just an audio
26:18 Source my greatest source of support
26:21 living with ritard arthritis has been um
26:24 speaking with other people about what
26:27 they've gone through
26:28 uh reading a lot about the effects of
26:30 rheumatoid arthritis or we can take a
26:33 recording of her we can make um a
26:36 synthesized model of her on 11 Labs is
26:40 where I did it and hi I'm Ellen despite
26:44 living with rheumatoid arthritis I've
26:46 discovered a passion for water skiing
26:47 and interpretive dance which bring joy
26:50 and a sense of freedom to my life so
26:52 that's just text to video right so I can
26:56 just have a written story and now I can
26:58 do that and it gets even better I can
27:01 then take it one step further and say if
27:03 we have no stories if we have no people
27:05 that we've captured video from we can
27:07 generate
27:09 personas and we can bring those personas
27:12 to life in a similar way and this is
27:13 where you speech to speech hi I'm George
27:17 Miller my life's always been about
27:19 Simple Pleasures cooking my Catch of the
27:21 Day and enjoying the quiet of the
27:23 country there's nothing like the piece
27:26 of fishing at dawn or the aroma of a
27:28 home-cooked meal but when I was
27:30 diagnosed with diabetes everything
27:32 changed suddenly the foods I loved
27:35 needed rethinking and my days had to be
27:37 planned with care it was a challenge but
27:40 I've learned to adapt finding new ways
27:42 to enjoy my passions while managing my
27:44 health life's different now but no less
27:48 rich and look here's George that's
27:50 George Miller here's George's
27:53 doctor hello I'm Dr Emily Park as an
27:56 endrocrinologist special izing in
27:58 diabetes my Approach goes beyond
28:00 traditional treatment I'm deeply
28:02 committed to empowering my nice Silver
28:05 Fox I helped an older Spanish-speaking
28:07 woman in the airport using chat GPT
28:08 today she was all smiles that's awesome
28:11 okay so both of these videos are are me
28:16 speaking hi I'm George
28:18 Miller but I'm not and I'm certainly not
28:21 doing this one hello I'm Dr Emily Park
28:25 but I did be right be right back poster
28:27 and show until great thanks steo okay so
28:31 how did we do that let's go to 11 labs
28:33 and I'll show you speech to speech it's
28:35 pretty
28:37 cool 11
28:40 labs.com
28:43 and then right at the top they say
28:45 speech to speech release automatic voice
28:49 conversion
28:52 so uh do I need to log in am I logged in
28:56 what's going on here sign up sign in
29:05 in why is the Olympic song in my
29:11 brain okay so here we are so you got
29:13 text to speech and you got you got a
29:16 list of voices
29:19 okay and some of those are mine so I've
29:22 got there's there's Ellen so wait let me
29:25 make this bigger so you can see it
29:31 so Ellen is the the uh example patient
29:34 that I just showed you then we have Kyle
29:37 Shannon Pixar bright Kyle Shannon Pixar
29:40 normal Kyle Shannon extrovert Clips
29:43 that's from a YouTube channel that I had
29:45 but those were really bad quality so
29:46 that's crappy and then here's a bunch of
29:49 theirs and so you can just listen to any
29:50 of these you can sample them all our
29:52 dreams can come true there is no greater
29:55 harm than that of time wasted the less
29:57 effort the the faster and more powerful
29:59 you will be okay so that's ranata let's
30:02 let's let's grab
30:07 ranata and so now there's this speech to
30:10 speech thing so voice settings uh you
30:14 can you can geek out on that [ __ ] if you
30:16 want I just leave them where the
30:17 defaults are cuz I have
30:20 ADD you more scientific people can have
30:23 fun geeking out with that [ __ ] as long
30:25 as it sounds good enough it's good
30:28 enough for me okay so you can upload a
30:31 file or I can just record an audio file
30:33 so we're we're going to record something
30:35 here oh wait let's go let's go over to
30:39 chat
30:43 J and we'll do chat gbt 3.5 since it's
30:47 faster chat gbt
30:49 3.5
30:56 um
31:00 write me a
31:02 poem
31:03 where as many of the words as possible
31:06 begin with the
31:08 letter
31:16 k all right it's too big but we'll just
31:19 get we'll get a little small piece of it
31:22 here uh so how am I going to do this I
31:26 will
31:30 we'll drag that over there we'll drag
31:33 this over
31:35 here all
31:37 right so I am going to record
31:46 audio in a kingdom where Moonlight
31:48 kissed the sea kid a Stope of scars so
31:52 good Lord
31:54 what hold
31:56 please I I had to do a tongue twister I
32:00 had to do a tongue twister on on a live
32:03 this is so one of the things that you
32:05 will learn if you come here regularly
32:07 and if you do you can call yourself an
32:09 irregular um if you come here a lot what
32:13 you'll realiz is I don't always make
32:14 great life decisions okay so we're going
32:17 to record audio again all right in a
32:20 kingdom where Moonlight kiss the sea
32:23 Kaleidoscope of stars a shimmering
32:25 decree knights in armor
32:28 this is not oh you know why it's not
32:30 good it's not good because it was chat
32:32 gbt 3.5 chat gbd4 would have been much
32:35 better with all the cas in fact yeah
32:37 hang
32:39 on I'm gon I'm going to do it in chat
32:41 gbt 4 because this is
32:44 shitty and let's see let's use the
32:46 letter M instead of the letter K I think
32:49 there's more m
32:50 words ah here we go this is good okay
32:54 just you're going to love it it's going
32:56 to be fantastic you this is I don't know
32:58 if you know this yet but this is
33:03 [Laughter]
33:06 entertainment you can make money with
33:08 Jing this is it's like comedy without
33:11 the laughs you see what I'm saying oh
33:14 it's so good
33:18 okay Midnight Moonlight mingles Mystic
33:21 Meadows murmurs okay here we
33:24 go midnight Moonlight mingles Mystic
33:28 Meadows murmur Melodies Majestic
33:30 mountains mirror Misty Mists mingle
33:33 merrily magical moments multiply merry
33:37 minstrels make music mellow murmurs
33:41 mirthful
33:42 size memories
33:45 made magically
33:48 Mystic now why did I act like an idiot
33:51 at the end there and and go really slow
33:55 because I want you to note that it
33:59 captures the Cadence of what I did
34:01 midnight Moonlight mingles Mystic
34:04 Meadows murmur Melodies Majestic
34:06 mountains mirror Misty Mists mingle
34:09 merrily magical moments multiply merry
34:13 minstral make music mellow murmurs
34:17 mirthful
34:18 size memories
34:21 made magically Mystic beautiful sign
34:26 that man up get him in agent yesterday
34:29 okay so we got ranata we got our default
34:32 settings and we just say
34:47 generate midnight Moonlight mingles
34:50 Mystic Meadows murmur Melodies Majestic
34:53 mountains mirror Misty Mists mingle
34:56 merrily magical moments multiply merry
34:59 minstrels make music mellow murmurs
35:04 mirthful
35:05 size memories made magically
35:11 Mystic wild huh so let's go grab let's
35:15 go grab a another
35:18 voice we'll see if we can get something
35:20 fun in
35:22 here Santa Claus was who I used for
35:24 George Miller ideas at the beginning
35:27 points of all fortunes love doesn't make
35:29 the world go never mistake motion for
35:32 Action a man is not old as long as he's
35:34 seeking
35:35 something never mistake motion for
35:39 Action the fox has many tricks the
35:42 Hedgehog has but one but that is the
35:44 best of all if opportunity doesn't knock
35:47 build a door oh that's a good one Mimi
35:49 this is Swedish
35:51 English generate this will be a good
35:56 one
36:03 are we having fun yet are we having fun
36:05 people hey I was wondering midnight
36:08 Moonlight mingles Mystic
36:10 meows majest why did it do
36:15 Rada what did they say her name
36:17 was
36:25 [ __ ] Mimi
36:31 okay now I got Mimi selected I guess I
36:33 didn't select her I just played
36:40 her what is this side this side is 11
36:43 labs.
36:46 iio 11 spelled out
36:50 labs. midnight Moonlight mingles Mystic
36:54 Meadows murmur Melodies Majestic
36:56 mountains miror mirror Misty Mists
36:58 mingle merrily magical moment beautiful
37:02 okay so now I'm going to download this
37:03 file bang done so now we have
37:07 Mimi Mimi speaking the M poem and now
37:11 we're going to go over
37:13 [Music]
37:15 to we're going to go over to D-
37:20 ID that was a cool shite site that Mr K
37:24 showed me um d -
37:28 ID and so this is a site they just they
37:31 just did new interface allow all yeah
37:33 we're going to we're going to let you do
37:34 that and then we're going to log in with
37:38 the
37:40 goog all right so here we are so here's
37:43 a lot of those videos that I
37:45 made but we're going to say create
37:49 video and so now down below here I've
37:53 got pictures that I've uploaded so we
37:55 can go find one that looks like Mimi
37:57 Mimi kind of had a high
38:01 voice let's see yeah those are all
38:05 crappy we'll do one that I
38:09 created we'll do this cool one that's a
38:12 cool one so super pop hearty it's going
38:14 to look weird cuz she's not looking at
38:16 the camera but that's okay so I'm going
38:19 to choose my image now I'm going to
38:21 choose the audio upload your own audio
38:24 going to go to downloads we're going to
38:26 grab that 11 Labs file and you go boom
38:30 generate
38:31 video I'm running out of credits I'm
38:33 spending my credits on you
38:35 people hope you appreciate
38:38 this these credits don't grow on
38:41 [Laughter]
38:44 trees uh uh uh uh uh uh some of these
38:47 things got really creepy like this one
38:49 watch how creepy this one got so that's
38:52 normal help look at that weird ass smile
38:55 because I was Finding when I arose in
38:58 the morning uh my hands were very stiff
39:00 I could not really we've successfully
39:02 generated your video oh they up upgraded
39:05 their servers holy crap that was fast
39:08 all right here we go here's
39:11 Mimi midnight Moonlight mingles Mystic
39:14 Meadows murmur Melodies Majestic
39:16 Mountain who's mirror Misty Mists mingle
39:19 merrily magical moments multiply merry
39:23 minstrels make music mellow murmur
39:27 mirthful
39:29 size memories
39:31 made magically
39:34 Mystic yay you did
39:38 it so isn't that
39:41 wild wild wild [ __ ] speech to
39:46 speech so and I think I think one of the
39:49 things that that kind of cures is that
39:52 robotic Cadence that that these things
39:55 have right hey H Jen or just 11 Labs so
39:58 this was did so this was 11 Labs plus
40:02 did hey Jen for their Avatar and for
40:06 their um translation tool use 11 labs
40:10 for the audio you can see the in
40:12 somewhere in their documentation I saw
40:14 that they're doing API calls to 11
40:17 Labs um so that's how they're generating
40:20 their audio so 11 Labs is the
40:22 state-ofthe-art right now um will there
40:26 be
40:28 lots of others yeah there will here's
40:30 the the challenge with all these um
40:33 super sexy multimedia things is they're
40:35 all pretty pricey
40:38 um which is just sucks but there's
40:42 there's a lot of the open source stuff
40:44 will be
40:45 coming so all
40:48 right cam cat can I love this Megan so
40:52 cool Brad
40:55 Shortridge wait Brad T
40:58 Shortridge hey it's Brad T Shortridge
41:01 how you doing I'm in mortgage sales uh
41:04 how you looking on the uh housing front
41:06 these
41:10 days I like the name Brad oh Jim Buckley
41:14 thank you for the Llama llamas is so
41:19 tasty
41:22 um oh PR proms cart wanted to know what
41:25 site that was too that was a 11 Labs 11
41:28 Labs hey Jenner just 11 Labs she killed
41:31 it she did she had a good Mentor I I
41:35 gave her her line readings actually
41:37 that's that's exactly what it
41:40 is speech to speech is you're giving
41:42 some other actor their line readings
41:45 that just hit me that's pretty
41:46 funny
41:49 um what else do you want to do we can do
41:51 whatever you want buy your own h100 Pate
41:55 pate's sitting there on a on a a pile of
41:58 tpus just acting like everybody can just
42:00 go pick up an h100 come
42:07 on I haven't raised a billion dollars
42:10 yet this week
42:11 Pate working on
42:14 it as soon as I can get this Tik Tok
42:17 channel built up bring in the big
42:20 bucks oh
42:23 man I remember that they're costly
42:34 yes travel info gpts okay I'll show you
42:38 how to find
42:39 gpts AI news programs are coming
42:42 personalized news with avatars yeah I
42:43 saw that Ted looks at life there's a
42:45 someone just put out a 22-minute news
42:47 piece we can go look at
42:49 that a 22 minute news piece where all of
42:52 the news anchors are what I just showed
42:55 you basically um and and I don't they're
42:58 not using speech to speech they're just
43:00 using the they're using this technology
43:03 on this
43:04 one hi I'm Ellen despite living with
43:07 rheumatoid arthritis I've discovered a
43:09 passion for water skiing and
43:11 interpretive dance see that one's a bit
43:13 like kind
43:16 of and then you know the the the ones
43:21 where I recorded the voice hello I'm Dr
43:24 Emily Park as an endrocrinologist
43:27 specializing in diabetes my Approach
43:29 goes beyond traditional treatment I'm
43:31 deeply committed to empowering my
43:33 patients into she got better Cadence you
43:35 know um but let's go look at was quick
43:38 take Jake did that lovely image isn't
43:40 that a cool
43:42 one although it's kind of a drag that
43:44 the the
43:45 uh the exclamation points on the left
43:48 are dotree but that would certainly be
43:50 an easy photo slop
43:54 addition okay uh uh what was I going to
43:58 go
43:59 show oh dang it oh dang the dang the
44:04 add was it going to go
44:09 show anybody
44:11 remember anybody care no all right I
44:15 don't know what I was going to go
44:17 show travel
44:20 bot yeah I was going to go do the travel
44:23 B oh the AI news thank you Ted the
44:26 travel news the travel or not the travel
44:29 news the the uh 22 minute let's see uh
44:33 future of
44:37 news will that get us there no that's
44:40 not
44:43 it um let's see
44:49 news
44:51 avatars
44:55 oops
44:59 is this it yeah this is it and welcome
45:01 to channel one a new way of consuming
45:04 reporting and thinking about the news
45:06 powered by artificial intelligence today
45:10 so she's fake you'll witness AI
45:12 generated stories and headlines
45:15 captivating visuals that's fake and
45:17 datadriven insights stuck so this is a
45:20 mix of stock and generated Stu from
45:22 Global News to finance to entertainment
45:25 we'll show you how technology oh it's in
45:27 Mighty networks okay I should tell
45:29 people about Mighty networks bring a
45:31 Global Perspective 247 right from the
45:34 heart of our AI off some of the
45:36 technology we're using to build channel
45:38 one we'll show you how AI powers and
45:41 empowers are new are also involved in
45:42 checking the stories for accuracy and
45:44 Clarity at every step of the process so
45:47 none of these are real maybe the best
45:48 way to explain what channel one is is
45:50 just to show you right now we're going
45:52 to jump into a program featuring a
45:54 selection of stories all produced with
45:57 the help of AI so please enjoy the
46:00 world's first look at channel
46:07 one this is channel one military
46:12 operations against Hamas by Israeli
46:14 Defense Forces continue in Gaza with
46:16 Israeli Prime Minister Benjamin
46:17 Netanyahu rejecting the idea of a
46:19 ceasefire in Gaza unless hostages are
46:21 released this week the White House said
46:23 Israel has a so while it while it feels
46:26 a little cold and weird it ain't that
46:28 far off right and and again with this
46:31 Tech this is as bad as it's going to get
46:34 right we we at the we at the
46:37 very
46:40 beginning of this entire universe of
46:43 content generation where like the big
46:46 bang was a year ago on November 30th
46:49 right Chachi PT was the sort of big bang
46:51 of this stuff um and it you know it it
46:56 it enables so chat GPT itself enables a
46:59 lot of this but what what really
47:00 happened is that because chat GPT got so
47:04 popular so quickly what you got was an
47:08 immediate shift of all attention from
47:11 anything anyone else was doing including
47:13 metaverse and um blockchain and Defi and
47:18 all that [ __ ] all of the tech resources
47:20 just went zo and pointed toward
47:23 generative Ai and started funding it and
47:25 started working on it and the open space
47:27 Community started going so we are just
47:30 now reaping the rewards or starting to
47:33 see the beginnings of the outputs of all
47:36 of that investment of time and money so
47:39 we're at the very beginning and and this
47:41 is this is how good this stuff is okay
47:44 let me let me show you something if you
47:45 how many folks are in here we got a
47:47 bunch of folks in here welcome um that's
47:50 not what I
47:52 wanted okay so if if you are one of
47:57 those people that's super curious about
47:58 Ai and you're kind of like I can't find
48:01 my people I don't know where my people
48:03 all everyone I tell about AI says I'm a
48:05 dumb Dum and they're angry at me and
48:07 they don't like me and I don't like it
48:09 and they're poo poo heads if you're like
48:11 that if this sounds like you um wouldn't
48:16 it be nice if there were a community of
48:19 people that um we're like you that were
48:22 curious about this stuff and we're
48:24 choosing to be on this AI Adventure
48:26 rather than be at the mercy of it that's
48:28 what this group is so I started the AI
48:30 Salon um the week after chat GPT launch
48:33 last year we just celebrated our first
48:35 year anniversary in fact I just put the
48:37 video of our um Salon meeting from
48:40 December 5th up on the YouTube
48:43 channel um so that's that's cool if you
48:45 want to if you missed that want to see
48:46 that it's up there um we brought back
48:48 speakers from the last year um came back
48:50 and kind of reminisced
48:52 about what's been going on for a year so
48:55 it's pretty interesting stuff so if you
48:58 dig that stuff go there oh it doesn't
49:00 cost anything that's another thing
49:02 everyone's like oh what's he trying to
49:03 Shi I didn't realize his channel was
49:05 about him chilling something I didn't
49:07 realize this was you can make money with
49:10 seems seems a little sus to me sister
49:13 you know what I'm
49:14 saying he can't just be uh actually
49:17 supporting people that doesn't make any
49:19 sense at
49:22 all so if you like that that's fine it's
49:25 free though so go there uh that URL
49:28 points to a link tree the first link is
49:31 to the community itself on Mighty
49:33 networks um the second link is to meet
49:36 up and then the third link is to a merch
49:38 a merch store um which does support the
49:41 salon that's how we're paying for the
49:42 salon is with the merch so if you dig
49:44 the logo um get feel free to pick up
49:47 some
49:47 merch um but yeah go there and then when
49:51 you get to the when you get to Mighty
49:53 networks it's a pretty cool it's pretty
49:55 cool cool deal we've got set up we're
49:57 getting there we're about we're about a
49:58 month in so we're we're figuring we're
50:00 figuring some things out so when you
50:02 first get to Mighty networks where I
50:04 want you to start is welcome to the
50:07 salon and it talks about who we are why
50:10 we do what we do how we do it what we're
50:12 looking for like what the Future Vision
50:14 is and then it's got our values
50:17 curiosity generosity like generosity is
50:20 like if you learn something share it if
50:25 you post an and show and tell post the
50:27 The Prompt so people can see what you
50:29 did like just be generous because we're
50:32 at a time right now where things are
50:33 moving so fast that literally zero
50:37 people zero
50:39 people
50:42 0.00000000 people have a complete
50:46 clue everyone is trying to figure it out
50:50 and things are changing so fast that if
50:52 you give away your big magical secrets
50:56 it's not going to matter cuz the big
50:57 magical Secret's going to be valueless
50:59 in a month or two so share and and and
51:02 everyone gets better faster and then um
51:04 exploration is one of our values
51:07 collaboration inclusion people are
51:09 coming from all walks of life all levels
51:11 of experience and then empathy is the
51:13 last value so if you resonate with that
51:15 stuff cool great and then your first
51:18 assignment should you choose to join
51:20 join the uh the salon is introduce
51:23 yourself go to the community section and
51:26 there's an introduce yourself section
51:27 and be generous talk about who you are
51:29 talk about what you do talk about where
51:31 you are on your AI Journey um if you're
51:34 scared of it if you're pissed off about
51:35 it like it it's all good um as long as
51:38 you're here kind of in the spirit of
51:40 trying to figure this [ __ ] out great and
51:42 then there's a water cooler which is
51:43 just cool discussions people talking
51:45 about stuff and there's showand tell
51:48 there's a news channel there's news and
51:50 res or tools and resources which I'll
51:51 show you in a second but in Show and
51:54 Tell um a lot of a lot of what happens
51:57 in these lives is that I'm as I'm
52:00 talking generosity is not being a
52:02 gatekeeper yeah exactly exactly like
52:07 just just come be generous I like what I
52:11 can promise you is that it will serve
52:13 you as you learn about this stuff talk
52:16 about it talk about it on LinkedIn talk
52:18 about it on the salon talk about it on
52:20 Twitter talk about it at your family
52:22 dinner talk about it at work but at work
52:25 they don't like
52:27 okay don't
52:29 care make them uncomfortable they're
52:31 going to be uncomfortable anyway when
52:34 when they're they get slapped upside the
52:36 head because their job got eliminated
52:38 your assignment should you choose to
52:39 accept it exactly this is Mission
52:41 possible not mission impossible oh
52:44 there's a there's some images we can
52:45 make Mission possible I like it um and
52:49 so so what happens in here a lot is the
52:51 people that come back here regularly
52:53 when I get into doing things like this
52:55 where I'm talking about things they've
52:56 heard before what they do is take stupid
52:58 [ __ ] that I said either from tonight or
53:00 from previous nights and they go make
53:03 funny images and they post them on
53:04 showand tell and in all of that play in
53:07 all of that sort of visual trolling and
53:10 meming and and all the cool stuff that
53:12 it is everyone's learning and everyone's
53:15 going oh I wouldn't have thought to do
53:17 the prompt like that or I wouldn't have
53:18 made the image like that but I really
53:19 like that right and then because it's a
53:22 community you can connect with them you
53:24 can comment you can go ooh how did you
53:26 do that and because it's a community of
53:28 generous people they go I did it but
53:31 like
53:33 this they don't give you that weird ass
53:35 Discord [ __ ] where they're like well I
53:38 could tell you someday perhaps if you've
53:40 earned enough stripes and candy
53:44 bars all right Shady ask Kyle if he will
53:48 uplo upload more Tik Tok videos to
53:51 YouTube please yeah I'm I'm I'm caught
53:55 up to
53:56 to um to the 24-hour ticktock Aon the
54:01 chat PT Aon is all up on YouTube right
54:03 now I downloaded everything through
54:06 December 7th today so I'll upload that
54:10 tomorrow it's at work or i' do it
54:13 tonight so I should be caught up by this
54:15 week I got I got a little behind there
54:17 I've I've been busy at work like that
54:19 that um that stuff that I showed you
54:21 those those cool Avatar Persona things
54:26 that's been like my past week I've been
54:27 pretty pretty heads down building
54:29 [ __ ] um but anyway here's show and tell
54:33 that's pretty cool one starting to see
54:35 similarity one day where Kyle got his
54:37 access to the button to make Siri go
54:41 away thankfully Google Google still has
54:44 that passcode he cannot
54:46 [Laughter]
54:47 find oh that's good that's solid oh
54:50 that's a Korea image nice text prom
54:53 prompt my brain is fr
54:56 R wow cool back in five this is nice
55:00 watch so here's how we do this save
55:03 image wait okay Cindy save image as then
55:07 I go
55:09 Cindy Dash
55:12 back in five we'll see if there's some
55:15 other ones in here to the
55:18 desktop I'll be back here's another one
55:21 uhoh what did I
55:24 do
55:26 oh who did this
55:29 one Tom th good day Tom hello name's Tom
55:37 Tom Tom
55:39 th are you the Tom I am a
55:46 Tom uh Tom
55:52 the oops spelled wrong
55:57 be I'll be
56:00 back I'll be back I'll be back Kyle will
56:05 be right back that's a good one who did
56:08 this y'all are weird
56:13 man be right back oh that's a good one
56:16 oh dang it all right let me let me grab
56:18 I'll grab two more here and then I'll
56:19 I'll I will spare you all of all of the
56:22 like maintenance Melanie Weber this is
56:25 cool
56:27 um what'd you do Wednesday night I
56:30 watched some old guy save files to his
56:32 hard
56:33 drive he I'm pretty sure he saved him to
56:35 a
56:36 floppy
56:41 um M
56:44 [Music]
56:54 Weber
56:57 [Music]
57:01 llamas is tasty I like
57:09 llamas did he say he was the CEO of a
57:12 company uh yeah about
57:15 that steo this is a good one steo I like
57:18 it I like it I like it a lot
57:22 steo and if you're all wondering like is
57:25 that the actual steo of course it
57:28 is jeez what do you think we run a half
57:31 ass operation here God dang God dang it
57:36 you people all right look watch watch
57:39 the magic I'll pull this stuff off on a
57:41 floppy Drive oh wait those two are bad I
57:45 guess I won't H I'll come back I got two
57:49 of them I saved two of them
57:52 wrong but
57:53 whatever sorry sorry steo yours didn't
57:56 make the cut okay you can also make
57:59 technical difficulties sign that just
58:01 stays up the entire time uh see pain
58:05 that's just mean that's just mean it's
58:08 true but it's mean you can make money
58:12 let we're we're whole assing it yeah we
58:15 don't have asset here we whole ass it
58:19 here yeah the AI learning lab is a
58:22 nightly exercise in 17 Factor authentic
58:25 ation
58:26 challenges I'm telling you man I I I so
58:30 so want to go back to Post-it notes on
58:33 the [ __ ] monitor I I just don't care
58:35 I'm like take it all just take it all
58:39 take my children you just have it just
58:42 don't make me log in with the I scan
58:49 again oh
58:52 man hey that's mine oh you know what I
58:54 was going to show
58:56 hey that's mine oh Athena Luna oh great
58:59 that's awesome which one is yours the
59:00 turtle one this one's
59:04 awesome be right
59:06 back where's Kyle Hey where's Kyle
59:10 where's Kyle he'll be right back can't
59:12 you see the turtle there's a sign there
59:15 Sally jeez
59:21 Sally all right I told you I would tell
59:23 you how to find gpts you you want you
59:25 want to see a cool
59:26 trick this is if you've used Google for
59:29 20 years this is one of those tricks
59:31 that you learned about early on in
59:33 Google and then you forgot
59:36 it and because there wasn't a a real
59:39 need to do it because people figured out
59:42 SEO um but it it's still good so I'm
59:46 going to go to
59:48 google.com all right here we are at the
59:50 Google all right so if you want to find
59:53 all the gpts well at least all the gpts
59:56 that have been indexed and if you're if
59:58 you're sitting there thinking but K I
1:00:00 don't know what a GPT is could you tell
1:00:02 me what a GPT is what's a GPT hey KY hey
1:00:05 k c c not sure the salon. a or how do I
1:00:10 access the site to generate AI images
1:00:13 okay Ricon Ricon Ron ricorn Ron Rona Ron
1:00:19 ronon Ronan Ron R Ron ran all right
1:00:24 Ronan listen here's what we're going to
1:00:26 do so the AI
1:00:29 Salon is just a site that will take you
1:00:33 it's just the community where we talk to
1:00:34 each other and you can share
1:00:36 stuff where we're going to make images
1:00:38 I'm going to show you two places to do
1:00:40 it one is chat gp4 and the other one is
1:00:43 Microsoft co-pilot Microsoft co-pilot is
1:00:46 free um chat gp4 is better I'm going to
1:00:50 show you both and you can make images
1:00:52 there and it's it's really cool because
1:00:54 you don't really need to learn about
1:00:55 prompting
1:00:57 but gpts are another thing that are a
1:01:00 part of chat GPT 4 so chat GPT
1:01:03 4 you can create these things called
1:01:06 gpts that are
1:01:08 customized
1:01:10 personalized versions of chat GPT that
1:01:13 you can share with other people okay um
1:01:16 in fact Let Me Go show you one I will go
1:01:19 show you the very first one I made I
1:01:21 made on this
1:01:23 channel and the Irregulars are probably
1:01:26 like oh God Kyle you going to show that
1:01:28 again Chas show that every stupid time
1:01:32 you're such a dum dumb Kyle want don't
1:01:34 you to show something different
1:01:37 sometimes why are you talking like
1:01:38 Napoleon Dynamite eat the food eat the
1:01:42 food jeez anyway um so Sin City
1:01:48 sipper is like chat chei te but just for
1:01:51 like
1:01:51 boozing so so if you want to make
1:01:54 holiday cocktails you go to Sin City
1:01:58 sipper and what it does is it puts a
1:02:00 Vegas Twist on classic Cocktails so I
1:02:03 could say say reinvent the margarita for
1:02:06 me I've got you know some pre-done
1:02:08 things here but you can also just ask it
1:02:09 to make you a drink for whatever in fact
1:02:11 Let's do let's do a um I would
1:02:14 like like you to
1:02:18 help me create a lovely holiday
1:02:26 cocktail that is
1:02:29 original and oh so tasty wait oh
1:02:35 so
1:02:37 tasty ta taty Tasty
1:02:41 Pasty all right and then sinity sipper
1:02:44 ah holiday cocktail with sinity sipper
1:02:47 twist the sinity holiday sparkler spiced
1:02:50 rum for that warm holiday spice kick
1:02:53 look it even tells you why it's added
1:02:55 the ingredients spiced drum orange lur
1:02:58 cranberry juice lime juice a splash of
1:03:02 sparkling wine cranberries and a sprig
1:03:05 of rosemary for garnish and then look
1:03:06 what it does teaches you how to make
1:03:09 it and then it makes an image of the
1:03:12 drink how flipping cool is that so I can
1:03:16 share this with the world in fact I have
1:03:18 I've gone on my my Ticky talk the what
1:03:21 what or not Tik Tock it's called the
1:03:22 Twitter the X the call it X now X isn't
1:03:26 a name X isn't a name Twitter's a name
1:03:30 Twitter sounds like glitter X it's just
1:03:34 X anyway I put it out on the X you can
1:03:38 see it on the X I can I'll share it it's
1:03:40 I it's shared over in it's everywhere
1:03:44 all right so that's what a GPT is so
1:03:47 here's the here's the cool little
1:03:50 hack
1:03:52 um open AI is going to launch launch a
1:03:55 GPT store and there are a bunch of GPT
1:03:58 directories right now that you can
1:04:00 they're but they're all over the place
1:04:01 but chat GPT is going to launch the
1:04:03 official store in lie of that until that
1:04:06 comes it looks like it's going to be
1:04:08 January although someone told me there
1:04:10 was a rumor they were going to launch it
1:04:11 next week I didn't I didn't see any any
1:04:14 to back that up so I'm guessing
1:04:16 January if you go to Google and you type
1:04:18 in site
1:04:21 colon and then
1:04:23 this wait let me go
1:04:28 back site colon no space site colon
1:04:33 chat.
1:04:34 open.com or chat.
1:04:37 open.com G for GPT I suppose that's the
1:04:41 directory name that all these things sit
1:04:43 in site colon chat. open.com G all right
1:04:49 so there they all are all the ones that
1:04:51 have been indexed 44,000 of them there
1:04:54 there's 44,000 of these things out there
1:04:57 all right and now someone said they
1:05:00 wanted to find travel gpts so you can do
1:05:02 site chat.
1:05:04 open.com and then type in
1:05:07 travel and now you've used Google for
1:05:10 the first time in a month or
1:05:13 two anyone else not using Google anymore
1:05:17 I I just don't I do not I do not use it
1:05:21 okay so that's that's a a fun if you
1:05:25 want to find these things hack okay
1:05:26 let's now go
1:05:28 to um the person that said I forget your
1:05:31 name sorry um who said how do you make
1:05:34 the images on this thing okay
1:05:37 so AI
1:05:39 Salon or the salon. that that that
1:05:44 uh that link tree that points to the the
1:05:47 mighty Network channel the mighty
1:05:50 networks is just the
1:05:52 community uh would a chat GPT store be
1:05:55 like for developers so that so yeah so
1:05:57 so jrb
1:05:59 185 the chat GPT store is going to allow
1:06:02 you to monetize
1:06:05 gpts think of where we are right now if
1:06:09 you're old enough to remember the launch
1:06:11 of the iPhone when Steve Jobs launched
1:06:14 the iPhone it was this pretty miraculous
1:06:17 thing but but one of the things that got
1:06:19 dinged for was he called these can I get
1:06:22 that pinned the address
1:06:25 um yeah could someone pin the address of
1:06:30 which which one the site slash or the
1:06:34 salon what's what what URL would you
1:06:38 like pan
1:06:41 pan so so when they launched the um the
1:06:45 iPhone it had these things called apps
1:06:49 and what the apps were were just like
1:06:51 icons that pointed to a website and all
1:06:54 the developers were like those aren't
1:06:56 apps those are just those are just
1:06:58 websites you dumb Dums you're all dumb
1:07:01 Dums and then the iPhone was out for a
1:07:02 while and then like at one of the next
1:07:04 announcements they said announcing the
1:07:07 App Store and I think the initial apps
1:07:09 were still those janky ass websites and
1:07:12 then the App Store turned into the App
1:07:14 Store and a lot of people made a lot of
1:07:16 [ __ ] money so there are currently two
1:07:20 places where you can yes the site SL
1:07:24 please yes
1:07:25 so let me
1:07:29 go I'll just go do a custom little slide
1:07:32 for you and I'll just leave it up here
1:07:33 while I'm Jibber
1:07:35 jabbering jibber
1:07:39 [Music]
1:07:43 jabber let's see where we going to do
1:07:46 this I'll do it down here on the
1:07:52 bottom site
1:07:57 col
1:07:59 chat. ai.com
1:08:03 G
1:08:05 slash all right can y'all see that there
1:08:10 so so when the App Store came out
1:08:12 everyone made fun of it these gpts there
1:08:16 people have made Google has indexed
1:08:18 44,000 of them and these things are like
1:08:21 a month old and there's not even a
1:08:23 directory to find them
1:08:25 so there's two two sites right now where
1:08:27 you can make these things one's called
1:08:28 po poe.com so if you go to PO you can
1:08:33 click on the explore button and go
1:08:35 explore lots of these things they have a
1:08:37 directory and then you can also say I
1:08:39 think it says create a bot so if you go
1:08:42 to um if you go to po.com create a
1:08:47 bot it's almost identical to how you
1:08:49 create a
1:08:50 GPT um except creating gpts is easier
1:08:54 and I would argue that open AI because
1:08:58 they have 100 million active users and
1:09:00 that's probably just gone up because
1:09:01 they they had to stop um taking new
1:09:04 subscribers for for a month uh so it's
1:09:07 probably going up pretty quickly right
1:09:08 now
1:09:12 um you can you can make Bots there on PO
1:09:15 you can monetize them today what Sam
1:09:18 Alman said is when the store comes out
1:09:20 you're going to have to be an
1:09:22 approved GPT maker right for them to put
1:09:26 your bots on their store so I assume
1:09:30 that's a relatively straightforward
1:09:31 process and then once you're approved if
1:09:34 your GPT performs well if it's one of
1:09:37 the ones that gets used they will be
1:09:39 doing rev sharing what is not at all
1:09:42 clear at all like zero Clarity
1:09:47 is if they make a million dollars are
1:09:50 they splitting it 50/50 with the GPT
1:09:52 makers or they splitting it 991 with the
1:09:55 GPT makers now if you listen to Sam mman
1:09:59 and if you look at his actions like his
1:10:01 worldcoin project he is concerned about
1:10:05 making sure that people get compensated
1:10:09 um through the the spoils that AI is
1:10:11 going to generate so I would imagine
1:10:14 that it might be a pretty good deal but
1:10:15 we'll see we have to see but you're
1:10:16 going to be able to monetize them yes
1:10:18 Antonio Kyle once explained to me the
1:10:21 way to protect your your work but I
1:10:23 forget in the gpts so so you can go you
1:10:27 can go look for for prompts for this so
1:10:29 there are if you look up uh GPT
1:10:34 jailbreak just on Twitter or
1:10:38 Reddit there are things that you can
1:10:40 type in that will force your GPT to give
1:10:44 up all of its prompts like in
1:10:47 detail um there are also things that you
1:10:50 can build into your GPT to prevent that
1:10:53 now that's going to be an arms race so
1:10:56 so what I what I've been doing is I
1:10:58 found one that sort of pre prevents it
1:11:01 and I just put it into my GPT prompt so
1:11:03 I can show you that how can we search
1:11:05 gpts show please right there at the
1:11:09 bottom the this very bottom
1:11:12 thingy hang on wait where's my mouse
1:11:16 where's my mouse oh my God this is so
1:11:19 frustrating you people are
1:11:21 weird oh jeez you can search gpts by
1:11:26 going to Google and typing in
1:11:29 that site colon oh wait no sorry this is
1:11:35 wrong was anyone was anyone actually
1:11:37 listening to me don't listen to me I
1:11:40 can't believe Pate has not jumped up
1:11:42 here and said uh Kyle uh you don't put
1:11:45 two slashes after the colon and site
1:11:47 colon just cuz you do that in every
1:11:49 other Google it does not work for me
1:11:51 that was cuz I put the wrong thing there
1:11:53 go do it
1:11:57 now sorry site colon chat. open.com not
1:12:03 site that's just me being an idiot
1:12:07 sitecon chat. open.com SLG slash and
1:12:11 then after that you can put a space you
1:12:13 can put subjects you want gambling gpts
1:12:17 do you want therapy gpts you want travel
1:12:20 gpts you put that after
1:12:22 this okay
1:12:27 um so that's that let's go make some
1:12:33 pitches all right here we are on chat
1:12:36 Jee now notice I'm in chat
1:12:40 gp4 um within chat bt4 now it's got this
1:12:44 thing called Dolly 3 in it um I'm trying
1:12:46 to learn how to wrap a Christmas present
1:12:48 don't have time to spell
1:12:52 check good Pate are are you one of those
1:12:55 that does like the the origami like the
1:12:57 super the super
1:12:59 crazy um the super crazy folded
1:13:03 Christmas presents I'm just like two
1:13:05 pieces of tape I'm like here it's
1:13:07 newspapers deal with
1:13:10 it I'm
1:13:16 horrible Christmas
1:13:19 Christmas tradition in my family was um
1:13:22 talking about who gave the shittiest
1:13:24 present the year
1:13:28 before that that was our holiday
1:13:32 tradition remember when Bill gave you
1:13:35 that
1:13:37 [Laughter]
1:13:40 sweater still still had the $ five tag
1:13:43 on from TJ
1:13:46 Maxx um okay so so built into this thing
1:13:49 is the ability to make images there's
1:13:51 also the ability to to see images which
1:13:54 I'm going to show you a cool thing
1:13:55 called round
1:13:57 tripping Dolly round tripping is very
1:13:59 very cool but we're not going to do that
1:14:02 right now what we're going to do right
1:14:03 now is you can just say make me an image
1:14:06 about something two words gift bags
1:14:08 exactly
1:14:10 exactly um so you can just say make me
1:14:14 an image of blank but one of the things
1:14:16 you can also do is you can make images
1:14:20 after you've had a conversation with
1:14:21 chat gbt so I'm going to say
1:14:27 um I want you to act as a worldclass
1:14:32 poet and
1:14:33 Storyteller and create a lyrical poem
1:14:36 that has lots of visuals
1:14:41 about a warm
1:14:45 holiday
1:14:47 tale of a llama who befriends a
1:14:52 hedgehog
1:14:56 not Huber friends HBY friends come
1:15:00 on Apple's voice recognition sucks it's
1:15:03 almost as bad as Siri who
1:15:06 befriends a hedgehog okay so in a quaint
1:15:11 snow kissed Village near Mountain so
1:15:13 high where stars twinkle bright in the
1:15:14 Deep Velvet Sky a llama so gentle okay
1:15:18 lovely we got a lovely poem dear friend
1:15:20 said the llama with a smile so kind and
1:15:22 it's going on and on a magical moment ah
1:15:25 I my my my cockal are
1:15:28 warmed can you warm a cockle pretty sure
1:15:31 you can okay so I've got warm cockal now
1:15:34 so now I can go in here and I can say
1:15:41 um describe four different visual styles
1:15:45 that might be appropriate to create
1:15:48 images from this lyrical
1:15:52 poem
1:15:55 watercolor illustration style okay
1:15:57 little predictable but fine classic
1:16:00 children's book illustration style H now
1:16:03 we're getting somewhere digital art with
1:16:05 a cartoon like flare eh vintage
1:16:08 handdrawn sketch okay I like
1:16:12 that okay I like number four period now
1:16:16 can you please
1:16:18 describe the four most interesting
1:16:22 visual moments in the poem that you
1:16:25 wrote now do you have to do all this
1:16:28 [ __ ] that I'm doing you don't but I'm
1:16:31 what I want to show you is
1:16:33 that the difference between using a tool
1:16:36 like mid Journey which is just like this
1:16:38 thing makes pictures learn our [ __ ]
1:16:40 language to make your
1:16:42 pictures and this is just a completely
1:16:45 different
1:16:46 thing that
1:16:49 that so chat GPT 4 one of the things
1:16:52 that that the that the Geeks talk about
1:16:55 is it's a multimodal large language
1:16:57 model what does that mean it does
1:17:00 pictures and [ __ ] is basically what that
1:17:02 means but it's multimodal meaning you
1:17:05 can be in text mode and then you can be
1:17:07 in picture mode and then you can go in
1:17:09 you know conversation mode and then you
1:17:10 can go in data analysis mode and then
1:17:12 you can go and search the internet mode
1:17:14 so there's these different modes
1:17:16 multimodal okay uh so now it we've given
1:17:20 it a style we like number
1:17:22 four it's given us the first meeting
1:17:25 under the Moonlight journeying through
1:17:26 the snow valleys dancing and singing by
1:17:28 the Fireside a quiet moment of
1:17:30 friendship so now I'm going to say okay
1:17:33 that's great I want the images to be 16
1:17:35 by9 wide
1:17:38 orientation and I'd like you to make the
1:17:41 first image to see if I like this
1:17:45 style and boom now it'll make the
1:17:51 image now you have to have chat gp4 to
1:17:55 be able to do what I'm doing here but
1:17:58 there's a there's a sneak around the
1:18:00 sneak around is go to copilot
1:18:05 microsoft.com and do the same
1:18:08 thing co- pilot. microsoft.com is
1:18:11 Microsoft's version of chat
1:18:14 gp4 for
1:18:15 free oh that's beautiful actually I love
1:18:19 it I'm going to
1:18:21 say I love it but what I would like to
1:18:23 see is the image go all the way to the
1:18:26 edge of the frame can you recreate it
1:18:29 with that slight
1:18:31 change now it's going to look different
1:18:33 it's not going to look exactly like that
1:18:35 so and this is one of the things you can
1:18:38 do a thing there's a thing where you can
1:18:39 ask chat gbt to give you the seed seed
1:18:43 of this image and it'll give you some
1:18:46 number uh and then you can um use that
1:18:50 to generate future images and it should
1:18:52 look kind of similar I haven't had great
1:18:54 luck with that but that's something you
1:18:56 can look up and look at let's see how
1:18:58 did it do all right it sort of went to
1:19:00 the edges I think that first one was
1:19:03 better wasn't it yeah it's just it's
1:19:06 warmer but whatever they're both nice
1:19:08 I'm going to
1:19:13 go could you add a splash of
1:19:22 color well
1:19:26 maintaining the black and white look
1:19:29 overall let's see what it
1:19:42 does while that's making that image
1:19:44 here's a cool thing that I don't know if
1:19:46 you knew this they kind of changed the
1:19:47 interface here if you click on the image
1:19:50 in the upper right hand corner there's
1:19:51 this I for information
1:19:54 this is the prompt that chat GPT wrote
1:19:57 to make
1:19:58 this so you can copy this and modify it
1:20:01 do anything you want with it you could
1:20:03 also take this and pop it over to Mid
1:20:05 journey and see how mid Journey does
1:20:08 with this image you know compare the
1:20:12 two um it didn't give me any color
1:20:15 whatsoever whatever the image has up
1:20:17 been updated
1:20:21 um let's make the Moon blue and cast a
1:20:25 blue hue over the
1:20:38 image does he always sing like that
1:20:41 that's very
1:20:43 annoying you can make money
1:20:49 with oh if you are wondering if I've
1:20:52 been drinking no I'm just an idiot there
1:20:55 we go there's a nice blue
1:20:57 splash okay so now I'm going to go great
1:21:00 can you now make the other three images
1:21:03 for
1:21:06 me and it
1:21:09 will and so so this is this is image
1:21:14 Creation in the context of a
1:21:16 conversation um which which for
1:21:19 me once I experience this like it just
1:21:23 if if I want really high quality art I
1:21:26 go to midj but I go there so rarely
1:21:29 because most art that I need is I just
1:21:31 need something good enough oh look how
1:21:32 cute that
1:21:35 is that's cute he got the little llama
1:21:38 on his back or a little Hedgehog on his
1:21:40 back that's one giant ass Hedgehog or
1:21:43 one tiny ass
1:21:47 llama that's a beautiful image
1:21:50 though it's got some proportion issues
1:21:53 oo look at that
1:21:58 nice not going to lie you're a very good
1:22:00 explainer Kyle very helpful with all
1:22:03 this evolving technology thank you very
1:22:05 much listen I so first of all I
1:22:07 appreciate that a
1:22:08 lot one of the misconceptions about this
1:22:11 AI stuff is that for the last 30 years
1:22:15 if you wanted to be an AI it actually
1:22:18 did behoove you to have a degree from
1:22:20 MIT or
1:22:21 Stamford okay
1:22:25 like like Pate is working on making tpus
1:22:28 more efficient at Google he I there's
1:22:31 something going on there with math okay
1:22:33 he's better at it than I am Pate our org
1:22:36 sent us cocktail kits and we had an
1:22:38 online class today made two strong
1:22:41 drinks at 12:30 uh day nothing like day
1:22:44 drinking Pate that's awesome
1:22:47 um starting last November one of the the
1:22:51 reason in my opinion that chat GPT is as
1:22:54 significant as it is like like of all
1:22:58 the stuff it does the most significant
1:23:00 thing it
1:23:01 did was it made the interface as simple
1:23:05 as a Google search box and what that did
1:23:08 was it
1:23:10 made it it released the power of AI and
1:23:14 machine learning to the rest of us it's
1:23:17 exactly what the worldwide web did for
1:23:19 the internet the internet prior to the
1:23:21 worldwide web required Comm line
1:23:23 interfaces and learning systems and
1:23:26 commands kind of like making images on
1:23:28 Discord right now right and thank God
1:23:31 they're moving it over to the
1:23:33 website um but the the the web freed the
1:23:38 power of the internet to anyone who
1:23:41 could sort of point and click chat GPT
1:23:43 is doing the same thing here Kyle can
1:23:46 you please repeat the name of the chat
1:23:47 GPT for alternative site yeah I'll do
1:23:50 that in a second oh look at the and the
1:23:54 Hedgehog so you know what I'm going to
1:23:55 do I I'll go show the the alternative
1:23:57 site here um but we're going to go plop
1:24:00 these pictures over in uh in show and
1:24:03 tell I'm going to have it
1:24:07 redo can you redo the picture around the
1:24:10 fire so it has the tan background with
1:24:12 the soft edges but keep the image
1:24:15 basically the
1:24:18 same um okay so if you don't have chat
1:24:21 gp4 you you can they opened up access to
1:24:24 chat gp4 now so if you want to get it
1:24:27 you can get it if you can't afford the
1:24:29 20 bucks a month then don't don't get it
1:24:32 if you can afford the 20 bucks a month I
1:24:34 think this is a much better um solution
1:24:36 than Microsoft's but Microsoft's is chat
1:24:40 gp4 so this is it
1:24:44 here so it's
1:24:46 co-pilot
1:24:51 microsoft.com here let me oh I'll put it
1:24:53 on visually this time I'll try I'll try
1:24:56 to type it without a typo let me move
1:25:00 this up we do this we do this and we do
1:25:05 this and we'll do um
1:25:08 co-pilot
1:25:14 dot all
1:25:15 right so that is the alternative to chat
1:25:21 GPT which is chat. openi
1:25:23 right above it and then site colon chat.
1:25:27 open.com G that's what you put into
1:25:30 Google if you want to find all the gpts
1:25:32 out there that have been
1:25:36 indexed all
1:25:38 right and then what we're going to do
1:25:40 here good people is we're going to go
1:25:43 let's see it's it didn't quite do it did
1:25:46 it well that's all right [ __ ]
1:25:48 it um it's a shame though CU these other
1:25:51 ones have such a cool stule style
1:25:56 um try again where the whole background
1:25:59 is beige right now you've got a little
1:26:01 bit of blue on the background as well
1:26:03 period I want the blackr to be beige not
1:26:07 the black girl oh my God Apple
1:26:12 background come
1:26:14 on going to get me
1:26:21 cancell
1:26:23 oh good lordy people good lordy what are
1:26:26 we doing with our lives what are we
1:26:28 doing what are we doing
1:26:37 PA all
1:26:39 right let me pop over quickly while
1:26:42 that's making that to
1:26:49 here can I make this a
1:26:58 expand to article
1:27:01 yes tag your post with a
1:27:05 hashtag
1:27:08 images
1:27:14 Irregulars
1:27:17 uh gpts no all right
1:27:21 whatever so this is going to be what was
1:27:24 this poem called did it have a
1:27:27 title it's
1:27:33 whatever all right that's all I'm
1:27:39 use
1:27:44 um I forgot to ask what is the title of
1:27:47 the
1:27:50 poem the poem does not not have a
1:27:53 specific title but considering it's a
1:27:54 theme it's theme a a suitable title
1:27:58 might be the llama and the Hedgehog a
1:28:00 Winter's tail that's not
1:28:03 bad not bad but I'd like 20 more
1:28:07 options oh this is another super user
1:28:11 trick and as stupid as this sounds like
1:28:14 don't be
1:28:15 content if if you don't love something
1:28:18 that chat GPT did tell
1:28:21 it give it feedback pretend like you're
1:28:24 working with a simpleton who's like the
1:28:27 most educated simpleton on the history
1:28:30 of
1:28:32 humankind but you got to tell them
1:28:35 everything all right let's see here what
1:28:37 do we got Moonlight companions no
1:28:40 snowbound friends Whispers in the winter
1:28:42 that's kind of cute the Starlet
1:28:47 Journey what what's a Winter's Tale the
1:28:50 Starlet Journey a Winter's Tale okay
1:28:52 that's that's what we're going to call
1:28:53 this
1:28:56 thing
1:28:57 the
1:29:00 Starlit
1:29:03 Journey a
1:29:07 Winter's tail a
1:29:11 Winter's oh
1:29:13 Tail cuz it's animals get
1:29:21 it
1:29:22 [Laughter]
1:29:28 is this what he does he this he just
1:29:30 laughs at his own jokes is that that's
1:29:32 is that what this channel is do what why
1:29:36 are there 150 people in here what it
1:29:40 people enjoy this some like it more than
1:29:43 television
1:29:44 [Laughter]
1:29:51 what I'm an idiot I don't care I'm an
1:29:55 idiot now I don't
1:30:00 care they call it a sling blade cute
1:30:03 cute cute
1:30:05 cute okay poem there's our lovely
1:30:13 po I
1:30:15 wonder how do you where do we I don't
1:30:19 know how to put images in this I'm just
1:30:21 learning this here
1:30:23 uh this here Mighty networks along with
1:30:26 y'all all right so I'm going to copy
1:30:29 this can I copy this no I got to do this
1:30:34 and then I can copy this copy image all
1:30:36 right let's try
1:30:38 that go back over here to might networks
1:30:41 and then we'll put it right after that
1:30:44 stanza boom all right we did
1:30:48 it what's
1:30:51 this your media has not finished
1:30:54 uploading yet oh that's because I have
1:30:55 shitty
1:30:59 internet standard
1:31:01 width full width Let's Go full width
1:31:05 nice pictures all right so we got that
1:31:09 one I could actually read this and
1:31:11 figure out where they actually go but
1:31:12 that's
1:31:13 okay then we'll go back here we'll go
1:31:17 that one is so
1:31:19 cute we'll go copy image
1:31:25 is he really copying and pasting files
1:31:27 this is what we're doing for
1:31:28 entertainment now seriously this is what
1:31:31 this
1:31:34 is um wait why can I
1:31:41 not God damn
1:31:44 it fine we'll download
1:31:51 them we'll do your your your fancy ways
1:31:55 oh that's I love this little end
1:31:59 image all right let's see we'll go back
1:32:02 here we'll go here let's see we'll go
1:32:06 here we'll go here we'll go plus we'll
1:32:10 go
1:32:12 image we'll go is this the yeah that's
1:32:15 it right yeah because that's that one
1:32:17 and that's the last one yeah okay so we
1:32:20 go here we go here we go here
1:32:24 oh and you want to hear something that
1:32:26 sucks I don't I think this is still the
1:32:28 case with
1:32:30 most with most chats you can copy the
1:32:35 chat and share it with
1:32:37 people
1:32:39 but
1:32:42 um if you make images in them the last
1:32:45 time I tried to do this it said you
1:32:47 couldn't do that you couldn't share an
1:32:50 image or you could couldn't share a chat
1:32:52 that had images in it so what I'll try
1:32:55 to do is I'll try to I'll copy The
1:32:59 Prompt up above or something like that
1:33:01 what's going on you haven't finished
1:33:03 your post yet I know no I don't want to
1:33:05 leave
1:33:06 without okay and then we'll go here and
1:33:09 we'll go
1:33:10 here and we'll go here and we'll go
1:33:18 here we'll go
1:33:20 here all right I got some hashtags in
1:33:23 there and then I'm going to put let's do
1:33:25 before the
1:33:29 hashtags we'll
1:33:31 [Music]
1:33:32 go uh I hope you
1:33:39 enjoyed here was
1:33:42 the
1:33:44 original
1:33:48 promp
1:33:51 that
1:33:54 generated the
1:33:56 poem dot dot dot and then we'll do this
1:33:59 and then we'll go back over to chat jie
1:34:02 Tha and we'll scroll on up to the
1:34:05 beginning let's see I want you to act as
1:34:07 a world class poet
1:34:14 and put it in quote
1:34:20 marks
1:34:23 all right and we're going to post it
1:34:26 shabang we're not going to notify all of
1:34:29 the
1:34:30 network your article is
1:34:33 live look you can go see pretty pictures
1:34:35 on a pretty
1:34:37 poem we made it in a thing we made it
1:34:40 together
1:34:41 [Laughter]
1:34:44 yay oh wait you know what we should do
1:34:47 because it would be
1:34:49 stupider uh let's see save article mut
1:34:53 can I edit
1:34:55 this edit is this editing this a
1:34:58 Winter's
1:35:04 tail get
1:35:11 it yes yes I am entertaining myself yes
1:35:16 I am entertaining myself all right so
1:35:19 within show and tell now you can see
1:35:22 little star Starlit
1:35:26 Journey oh
1:35:29 nice beautiful
1:35:31 Ted
1:35:33 a what the [ __ ] happened here sometimes
1:35:37 after lives I need to put Kyle Sin City
1:35:39 sipper to
1:35:47 work okay y'all are weird man but this
1:35:50 is what happens like I run down rabbit
1:35:52 holes and then the Irregulars just lose
1:35:54 their [ __ ] minds it's beautiful I
1:35:57 think this is this is a beautiful uh
1:36:00 conflagration of talents and skills and
1:36:03 learning and excitement and
1:36:07 enthusiasm all
1:36:10 [Laughter]
1:36:10 [Music]
1:36:17 right hey look it's the AI learning
1:36:20 lab wait this is oldie but a
1:36:23 goodie this one was created like many
1:36:25 moons
1:36:28 ago I think that was made during a live
1:36:32 I
1:36:32 think CU I kind of remember not being
1:36:35 able to quite get the letters right oh
1:36:38 and then I also took it to photoshop
1:36:41 this was
1:36:42 one did I fix his legs no this is the
1:36:45 unfixed one see how he's he's got five
1:36:47 leg he's got a double back
1:36:50 foot I have a I have a fixed version
1:36:52 version of this where I did actual
1:36:55 better
1:36:57 betterness but I there's something I'm
1:36:59 telling you we are going to we are going
1:37:01 to look back fondly on the janky images
1:37:04 from this era cuz the shit's going to
1:37:09 get everything is going to get
1:37:11 ridiculously good 2024 if you think that
1:37:15 2023 was even remotely interesting with
1:37:18 this AI
1:37:20 [ __ ] 24 is just going to be stupid like
1:37:23 seriously
1:37:25 it's it's going to start accelerating
1:37:28 it's well it's already accelerating but
1:37:29 it's going to really start
1:37:32 accelerating all right I am so mods you
1:37:36 know where I've been for the past 45
1:37:38 minutes or however the hell long that
1:37:40 was
1:37:41 so I've not been paying attention to
1:37:44 chats at
1:37:48 all all right so feel free to pin any
1:37:51 and I'm just going to go back and start
1:37:53 looking
1:37:53 at what is the possibility that poem is
1:37:56 plagiarized or that someone else could
1:37:58 get the same poem great question Don
1:38:00 solo
1:38:03 um
1:38:05 zero
1:38:07 well is it possible that it created
1:38:10 something that existed before it's
1:38:12 possible but it's very low
1:38:14 likelihood there's a there's a um
1:38:17 there's a setting when you make these uh
1:38:20 when you make something like chat GPT
1:38:22 called
1:38:24 temperature so the the way large
1:38:27 language models work one of the
1:38:28 misconceptions is that is that They're
1:38:30 copying and pasting right and there's
1:38:32 there's articles out there from from you
1:38:34 know people that say you know chat EPT
1:38:37 is the world's greatest plagiarizing
1:38:39 machine that person clearly doesn't
1:38:42 understand how these things
1:38:44 work or or or what's going on behind
1:38:48 them now have they been trained on lots
1:38:50 and lots of [ __ ] where there could
1:38:52 conceivably be something in there that
1:38:54 comes out they've been trained on
1:38:57 that
1:38:59 but none of the original documents
1:39:03 survive what's called embedding what
1:39:05 happens when when when these models are
1:39:08 trained there's there's this process
1:39:10 called embedding and what they do is
1:39:12 they take big
1:39:14 documents and the words within the
1:39:17 document are given a mathematical value
1:39:20 in Vector space
1:39:22 so like a two Vector thing a two a two
1:39:26 whatever a two Vector no a two axis a
1:39:30 two
1:39:31 axis Vector space XY right will have
1:39:36 like a DOT here that's like you know 1
1:39:38 one and a DOT here that's like 7 s and a
1:39:40 DOT here that's like 17 right
1:39:44 so think of a thousand Dimension
1:39:47 mathematical
1:39:49 space when when documents get embedded
1:39:51 the individual words get assigned a
1:39:53 mathematical value in Thousand Dimension
1:39:57 space this is this is true this how the
1:39:59 it's how the [ __ ]
1:40:01 works and when you type your
1:40:05 prompt oh and the other thing about the
1:40:07 the the point in Thousand Vector space
1:40:10 thousand Dimension Vector
1:40:12 space is the words from certain
1:40:16 documents get clustered together in
1:40:18 semantic regions in in Vector space and
1:40:21 so a semantic region might be like dog a
1:40:25 dog is the fluffy furry thing that sings
1:40:29 When you play your guitar or there might
1:40:32 be another semantic region of that space
1:40:35 where dog is the guy that did his girl
1:40:37 bad he's a dog right so it it it ingests
1:40:41 all these documents it takes the words
1:40:44 and understands their semantic meaning
1:40:46 and just basically like scatters those
1:40:48 words into math
1:40:50 space and then when you type your
1:40:54 prompt and hit submit it turns that
1:40:58 prompt into a mathematical probability
1:41:01 of the most likely word to come
1:41:06 next and then it goes into math space
1:41:08 and it goes where's that vector and it
1:41:11 pulls down that vector and says what
1:41:13 word is this number associated with and
1:41:15 it says dog and it sticks that in your
1:41:18 sentence so when you see it write a
1:41:20 whole paragraph for you or a whole
1:41:22 article for you it's doing it one word
1:41:24 at a time with mathematical probability
1:41:26 it's got no [ __ ] idea at all what it
1:41:28 wrote zero which I find it I find it
1:41:33 when you understand how these things
1:41:34 work it is more remarkable that they do
1:41:36 what they do like it is just it is
1:41:38 [ __ ] magic sorry it is I don't have a
1:41:42 math degree mathematicians are like well
1:41:44 it's really just probability Kyle it's
1:41:45 not
1:41:46 very no it's [ __ ]
1:41:50 Magic it's like it's like answering a
1:41:53 question by picking random words out of
1:41:54 a bucket until they until they make
1:41:56 sense yeah that it's kind of that
1:41:58 although you're reaching into the bucket
1:42:00 in you kind of know in the bottom right
1:42:02 of the bucket is where like the little
1:42:04 pink slips are and you and within the
1:42:06 pink slips there's one that's got the
1:42:07 highest probability the regions are each
1:42:10 of the dimensions each Dimension is a
1:42:12 way of measuring things yeah
1:42:14 exactly so it's just
1:42:17 cuckoo so so because of that the G and
1:42:20 GP T stands for generative it is
1:42:23 generating generating original content
1:42:26 through this mathematical
1:42:28 process and then there's a thing in in
1:42:32 the programming side called temperature
1:42:35 it's a setting chat jbt as far as I know
1:42:38 defaults to 7 and it's a scale of 0 to
1:42:41 one so if your temperature is zero it
1:42:45 will take the the word that has the
1:42:47 highest probability will be the next
1:42:49 word so that makes things very very
1:42:50 predictive
1:42:52 so if you have your temperature set to
1:42:55 zero you will get an answer for a
1:42:57 specific prompt and then you'll keep
1:42:59 getting that same answer for that
1:43:00 specific prompt chat GPT defaults to 7
1:43:04 so it's it's more on the creative side
1:43:06 so what what temperature does is it
1:43:08 randomizes it it'll say Okay instead of
1:43:11 just one word being the probability
1:43:13 here's like 20 that might be the next
1:43:15 word or some version of that so it gets
1:43:18 more creative and that's why when you do
1:43:21 the same prompt over and over again
1:43:23 you'll get very different results
1:43:24 sometimes will this work put all my Ants
1:43:27 genealogy info and have it sorted out
1:43:29 and make a book absolutely will work
1:43:33 absolutely
1:43:35 yes
1:43:37 yes you can upload
1:43:39 it um it depends how you have it right
1:43:43 if if you've got it um it it depends how
1:43:47 you have it if you've got CSV files with
1:43:49 data in it you you can upload that and
1:43:52 have it do data
1:43:54 analysis um check out our Show and Tell
1:43:56 for Silicon Valley Silicon Valley sipper
1:44:00 and Shannon family xus Hallmark movie oh
1:44:03 no oh no what have I
1:44:10 done I just I can't I can't talk I can't
1:44:14 talk
1:44:18 reality that's beautiful that's actually
1:44:20 very nice Becky thank you for having
1:44:22 something tasteful I fear what troll is
1:44:24 coming
1:44:27 up wait where is it check out oh our
1:44:30 show in oh the oh the uh the
1:44:34 Irregulars all right hang on let me go
1:44:36 to
1:44:39 Irregulars so by the way the Irregulars
1:44:42 channel is um if you come here a lot you
1:44:46 get into The Irregular Channel just
1:44:47 basically a private club Silicon Valley
1:44:50 sipper that's good that's
1:44:53 solid hey by the way I don't know is the
1:44:55 dude in here that said he was going to
1:44:57 send me a bottle of Blandon he never he
1:44:59 said he was trying desperately to get in
1:45:01 touch with me I told him how to get in
1:45:02 touch with me and he never got in touch
1:45:04 with me Shannon family Christmas
1:45:06 everyone crying about their
1:45:14 gifts that's pretty much
1:45:19 it
1:45:26 I was way behind oh my God those are
1:45:29 [Laughter]
1:45:32 good oh so anyway if you're a subscriber
1:45:36 to this channel um a it supports the
1:45:38 channel which is swell um but you you
1:45:41 get access to the Irregulars uh club uh
1:45:45 on on the AI Salon um so it's just a
1:45:48 private little area where where you know
1:45:50 people that come here a lot do it so you
1:45:52 don't have to be a subscriber if you if
1:45:54 you show up a lot and you're not douchy
1:45:56 then you're in all right I'm not sure
1:45:59 where I think it is wait I'm not sure
1:46:01 where it is I think I'm supposed to have
1:46:04 access what what are you talking about
1:46:06 Drew access to
1:46:09 what
1:46:11 to to the Irregulars Channel or to
1:46:14 something
1:46:19 else
1:46:21 [Music]
1:46:27 Kyle did you read about opening eyes
1:46:28 deal with political and Business Insider
1:46:30 yeah actually I did I forgot to talk
1:46:32 about that hang on good call thank you
1:46:35 so if you go to open ai.com
1:46:39 blog blog blog is such an ugly word
1:46:43 isn't it
1:46:45 um
1:46:47 open.com
1:46:49 blog
1:46:51 so this is a pretty big deal um
1:46:55 partnership with Axel Springer to deepen
1:46:57 beneficial use of AI in
1:47:00 journalism so uh where is
1:47:04 it Axel Springer media Brands include
1:47:07 political Business Insider European
1:47:10 properties build and
1:47:11 welt um so what's going to happen is
1:47:14 you're going to be able oh uh are you
1:47:17 drew in in uh in
1:47:22 uh is your name the same in Mighty
1:47:25 networks
1:47:28 Drew I can make that
1:47:31 happen I can't make that happen happen I
1:47:35 can't make that happen
1:47:41 [Music]
1:47:44 happen Drew you is it drew you yeah all
1:47:49 right let's
1:47:51 see all right so invite has been sent
1:47:54 what I've learned about the invites on
1:47:56 Mighty networks is I don't think they're
1:47:58 actually sent but if you go right click
1:48:00 on the like the three dots next to the
1:48:02 channel you should be able to give
1:48:03 yourself access
1:48:06 now um are we past the math part can I
1:48:09 come
1:48:10 out yeah I think I think we're done I
1:48:13 think we're done with the chat TMZ
1:48:16 version of
1:48:18 math are your presents wrapped
1:48:21 [Laughter]
1:48:24 oh my god that actually looks like my
1:48:26 uncle my Uncle Bruce that's pretty
1:48:29 funny although that one's [ __ ]
1:48:32 hilarious oh my
1:48:37 god oh why do so many say AI is just
1:48:41 copying existing works of others i e
1:48:43 illustrations because Tams because they
1:48:45 don't understand it so here's the
1:48:47 deal every single time in history in the
1:48:51 history of of
1:48:53 humankind every single time there's been
1:48:57 a new
1:48:58 technology there's resistance and fear
1:49:01 and anger and resentment and push
1:49:06 back
1:49:07 when when people
1:49:11 see what this stuff
1:49:14 does there is no logical explanation
1:49:17 that it's doing anything but copying and
1:49:20 pasting
1:49:21 like in their minds if you haven't if
1:49:23 you haven't learned what it's actually
1:49:25 doing and if you haven't really
1:49:28 experienced it and experienced
1:49:30 how generative it is that that this is
1:49:34 creating original works right if you
1:49:36 haven't experienced that you look at it
1:49:39 and you go that looks like Greg root
1:49:41 kow's work and they use Greg root kow's
1:49:45 name and his work was in the training
1:49:47 set so therefore it's copying and
1:49:48 pasting and it's stealing from him
1:49:51 in the case of Greg grotowski you know
1:49:54 two things are true he had so much work
1:49:56 out on the internet that there was lots
1:49:58 of stuff trained on it people liked his
1:50:00 images and for whatever reason in the
1:50:02 stable diffusion Community his name just
1:50:05 got used in every prompt right like he's
1:50:08 got you know his his artwork his his art
1:50:11 is very fantasy you know DND D like so
1:50:14 it was just super popular amongst the
1:50:16 crowd I would argue two things are true
1:50:19 one is a lot of artwork was generated in
1:50:23 his name and and in a similar style of
1:50:26 his work which he could argue took money
1:50:29 from
1:50:30 him that is probably partially true
1:50:34 there is
1:50:36 probably a massive amount of work that
1:50:38 was created though that they never would
1:50:41 have paid him to make them an
1:50:42 illustration anyway right so this is net
1:50:45 new work that I would argue that a lot
1:50:49 of that work that used his name name got
1:50:51 his name out there further and further
1:50:53 he didn't feel that way he felt like you
1:50:56 know it was trained on my stuff I'm not
1:50:58 being compensated you're stealing your
1:50:59 [ __ ] go [ __ ]
1:51:01 yourselves but when you understand how
1:51:04 these things actually work that's not
1:51:06 what's happening it's not what's there
1:51:08 the images that get trained are not in
1:51:12 the system as images they're in as these
1:51:14 random
1:51:17 noise things that have been noised over
1:51:20 50 70 steps and then when you type your
1:51:23 prompt it does that similar sort of math
1:51:25 thing and it amalgamates thousands of
1:51:28 images that are all random pixels into a
1:51:31 sort of random pixelized thing and then
1:51:34 it denoises it over 50 steps to go from
1:51:37 random noise to the image that you see
1:51:39 on screen again the fact that it does
1:51:42 what it does is staggering like it is
1:51:46 remarkable like how it does it I don't
1:51:48 know but it is not copying and pasting
1:51:50 but it's just it this is just people are
1:51:53 afraid that's all it is that's they're
1:51:55 afraid and they don't understand how it
1:51:56 works and so therefore they go it's the
1:51:58 Devil Burn
1:52:00 him
1:52:02 right burn the witch burn the
1:52:06 witch what other than witches
1:52:10 float uh very small
1:52:13 [Laughter]
1:52:17 rocks ducks ducks
1:52:23 if she's heavier than a duck she's a
1:52:25 witch burn
1:52:28 [Laughter]
1:52:32 her oh my God that's that's where we are
1:52:37 right now this our [ __ ] world is is a
1:52:41 [ __ ] Monty Python
1:52:43 [Laughter]
1:52:49 sketch
1:52:52 the math behind television Transmissions
1:52:54 is similar to the mathematics behind
1:52:56 chat gbt vectors oh that's super cool I
1:52:58 did not know that that's really cool I'm
1:53:01 a robot and pull the words out of a
1:53:03 bucket that sometimes someone filled
1:53:06 with the internet that's pretty much it
1:53:09 where is this site I don't know which
1:53:10 one it is you have to see the second
1:53:12 image wait which second image is it the
1:53:15 one this one I pulled
1:53:17 [Laughter]
1:53:19 up
1:53:20 [Laughter]
1:53:22 oh my God they're so good they're so
1:53:28 good
1:53:31 uh still waiting on my secret decoder
1:53:34 ring focal point MN um focal point DM me
1:53:40 what your real name is in Mighty
1:53:42 networks and I'll I'll pop you in
1:53:45 there their their wish was a GPT and
1:53:48 they got a pullover exactly
1:53:51 exactly thank you for your response of
1:53:54 course that's why we're here I do try to
1:53:56 answer as as many questions as I can
1:53:58 although now that I've Got Mods here I
1:54:01 dove I've been spending more time doing
1:54:03 kind of deeper
1:54:05 Dives
1:54:06 um you
1:54:09 know all
1:54:11 right we did that let's go back here
1:54:14 show and tell see we got some action oh
1:54:18 that's a nice one nice one Tom Papa
1:54:21 JJ see I mean like okay not for
1:54:28 nothing we we
1:54:32 so one of the things you'll see in a lot
1:54:34 of the Irregulars images are llamas
1:54:37 hedgehogs Turtles and gummy bears and
1:54:39 now blue rubber ducks
1:54:41 why cuz over the course of this channel
1:54:45 like every time I would mention llama
1:54:47 Winston would put up the llama sticker
1:54:50 whatever the thing is the gift on screen
1:54:54 and so llama became kind of a spirit
1:54:57 like a like like an animal of the thing
1:55:00 and then um Turtles came from the I like
1:55:02 turtles
1:55:04 kid um hedgehogs came from when when
1:55:07 Dolly 3 was announced they did a little
1:55:09 children's book with Larry the Hedgehog
1:55:12 that's where that came from so there's
1:55:13 there's all these different origin
1:55:16 stories having a a book about a llama
1:55:20 befriending a hedgehog is nonsensical
1:55:22 like that would never exist in the world
1:55:24 and
1:55:26 yet now here we are with technology
1:55:28 where we can just kind of joke it we can
1:55:30 have these kind of inside jokes
1:55:33 and for me to describe how to make an
1:55:35 image and chat GPT I can do something
1:55:38 dumb like make me a lyrical poem about a
1:55:40 llama befriending a hedgehog and then I
1:55:42 do that and I make the images and
1:55:43 they're cool and then Papa JJ takes that
1:55:47 same idea and does this image which is
1:55:49 [ __ ] beautiful like that's stunning
1:55:52 so now imagine like going hey that's
1:55:54 actually good enough that like we should
1:55:57 like I want to make a children's book
1:55:58 out of this and I want to do like all
1:56:01 the illustrations in that in that you
1:56:03 know that
1:56:05 quality and I'm going to take that poem
1:56:07 Kyle wrote and I'm going to turn that
1:56:08 into like yeah [ __ ] go do that right
1:56:10 why not we're moving into a world of
1:56:14 infinitely
1:56:15 personalized content
1:56:18 creation and I just I just find find
1:56:20 this this whole concept
1:56:26 of I I don't even know how to describe
1:56:28 it because it's like historically
1:56:30 there's a thing called creative
1:56:32 friction which which is one of the ways
1:56:35 you get to really good creative work is
1:56:37 you've got two or three or 10 people in
1:56:39 a
1:56:40 room all pursuing some creative idea and
1:56:43 and they all have different points of
1:56:45 view and there's this
1:56:46 friction and what chbt and what these
1:56:50 tools like mid journey and and all these
1:56:52 generative AI tools do is they're the
1:56:55 the absolute absence of friction so it's
1:56:57 like it's like creative
1:57:00 lubrication oh here come the images for
1:57:02 that one but it is like it's it's just
1:57:04 like it's like instant creativity like
1:57:07 have a brain fart lyrical poem have a
1:57:10 brain fart turn it into an image
1:57:12 style see that image that Kyle did on
1:57:15 his Tik Tok live and go in and do my own
1:57:17 version of it and then generate
1:57:18 something that beautiful like that image
1:57:20 is just stunning it's just like it's got
1:57:23 emotion it's got
1:57:26 Nuance it's got like they're they're
1:57:29 they're deep in thought it's got the
1:57:30 beautiful moon like holy crap that's
1:57:33 it's just like this Serene image of like
1:57:36 these these two in the middle of their
1:57:40 Journey
1:57:42 right I don't know I think it's [ __ ]
1:57:47 amazing or it's stealing from from
1:57:50 people and we're all [ __ ] it's one
1:57:51 of the
1:57:53 two teacher he's
1:57:56 copying this is this this is
1:58:00 exactly this is exactly what's going on
1:58:03 right
1:58:05 now thank you Robert Rossy hang on the
1:58:08 champ needs to go
1:58:10 out oh wait wait wait for
1:58:14 it hang on you weirdos I know I've got
1:58:17 to turn on my I got to but I'm coming
1:58:20 right back but that's okay you're right
1:58:22 you're right I should follow the
1:58:25 [Laughter]
1:58:35 protocol goor I'll be right
1:58:48 there
1:58:51 he's back he did it yay he did it
1:58:55 teacher he's copying this is the world
1:58:57 we live in right now and it it it is
1:59:00 kind of petty like that it's like it's
1:59:03 evil it's
1:59:05 bad nothing is absolute I
1:59:08 fear or I surmise or I I
1:59:13 propose that one's beautiful too like
1:59:16 there like so there's a really good
1:59:18 example the other the one I just showed
1:59:21 that Papa JJ did was like super
1:59:24 sophisticated and and nuanced and this
1:59:27 one's more trippy they could both be
1:59:29 really interesting kids books right
1:59:31 maybe one's got a more Twisted story and
1:59:33 it's about a dream that the Llama had
1:59:35 and you know now the llamas sprouted El
1:59:38 corns and it's a whole sea of hedgehogs
1:59:42 and super cool oh that's beautiful
1:59:46 too very very ooh nice
1:59:51 nice
1:59:53 whoa I didn't make it it's on
1:59:57 Amazon
2:00:01 Joker oh that's good that's
2:00:06 solid yeah these are really good so cool
2:00:10 all right we'll go look at one
2:00:13 [Music]
2:00:15 more the other side of a massive black
2:00:18 hole a that's cool
2:00:22 all
2:00:24 right where do I see these oh okay so if
2:00:28 you are are new
2:00:31 here where everyone is sharing these and
2:00:35 where you can see
2:00:40 these is that the AI Salon community so
2:00:44 the AI Salon is a group that I started a
2:00:46 year ago so I started this group The the
2:00:48 week after chat gbt came out chat gbt
2:00:51 came out I lost my [ __ ] mind I
2:00:53 started a Tik Tok Channel I started the
2:00:55 AI Salon I started a newsletter I
2:00:58 completely rethought my company's
2:00:59 product road map so yeah okay so go to
2:01:04 the salon. the first link in that link
2:01:06 tree is the community it's it's a mighty
2:01:09 networks Community you can also go to
2:01:11 AIS salon. MN Mighty networks. Co
2:01:15 not.com
2:01:17 doco I forgot to include llamas and
2:01:19 turtles in my Christmas images but maybe
2:01:21 it's for the best I think it's probably
2:01:23 good I think you're good you're good
2:01:25 Pate um but go there and so when you get
2:01:28 to when you get to this community the
2:01:30 first place I want you to go is welcome
2:01:32 to the
2:01:34 salon and it'll tell you about the salon
2:01:36 and so basically here's the deal think
2:01:38 of the AI Salon as a support
2:01:40 group for
2:01:44 like okay great um think of the AI Salon
2:01:48 as like hi my name is Kyle and I'm into
2:01:49 AI hayle it's been 5 minutes since I
2:01:53 created an image in
2:01:55 Dolly same here a micking 10 right now
2:01:59 um but go check out who we are and and
2:02:02 check out our values curiosity
2:02:05 generosity
2:02:07 exploration collaboration inclusion and
2:02:11 empathy this is about people first not
2:02:13 technology first cuz we're all trying to
2:02:15 figure this stuff out so go there first
2:02:18 and then your next stop should be
2:02:19 introduce yourself so go to the
2:02:22 introduce yourself
2:02:24 section and then go to show and tell and
2:02:27 then show and tell is where all these
2:02:28 images
2:02:29 are look at that one like there's
2:02:31 another one be not afraid it Danielle
2:02:33 did look at that it's beautiful it's
2:02:35 gorgeous I love the light damn it's like
2:02:38 [ __ ]
2:02:40 yes yes dang it and there's a
2:02:46 t-shirt I'm sorry I look at these images
2:02:49 I I I look at these images and like
2:02:51 what's happening in this community in
2:02:52 real time while these lives go
2:02:55 on and one of the things I was here is
2:02:57 well it's not real creativity these
2:02:59 tools aren't doing real creativity no
2:03:01 they [ __ ]
2:03:03 are It's a combination of human beings
2:03:06 using these miraculous tools now what
2:03:08 are these tools they're not
2:03:11 robots they are mathematical probability
2:03:14 machines but what are they probability
2:03:18 on
2:03:20 human
2:03:21 stuff human words human images things
2:03:25 that have been stuck on the internet for
2:03:27 the past five
2:03:29 decades so when you get a response back
2:03:31 from chat GPT you're not getting a
2:03:33 response back from a robot you're
2:03:35 getting a response back from like all of
2:03:37 humanity that contributed to the
2:03:39 internet or contributed to the written
2:03:42 word that got stuck on the Internet by
2:03:44 humans in the past 50
2:03:46 years that's why it seems so freakily
2:03:48 human because it is freakily
2:03:55 human these I made oh those are
2:03:58 beautiful like yeah I mean yeah go make
2:04:03 those
2:04:04 t-shirts 3D print them Vicki's been
2:04:07 taking these things that she generates
2:04:09 in here 3D modeling them and then
2:04:11 sending them to her 3D printer Farms
2:04:14 it's crazy isn't it it's internet
2:04:15 recycling it's totally internet
2:04:17 recycling it is to
2:04:19 okay this is this is one of the big
2:04:22 things that people don't quite get yet
2:04:24 that that I think is the the big killer
2:04:27 app of all these
2:04:28 tools if you think about the last 70
2:04:32 years of of computer
2:04:35 science we have as a series of
2:04:39 generations what's that three and a half
2:04:41 four generations five generations maybe
2:04:44 we've been dumping
2:04:46 data into these digital
2:04:50 buckets and almost all of that data
2:04:54 never sees the light of day
2:04:56 again what what AI does is it reveals
2:05:01 all of that back to us it like it it
2:05:03 bubbles it up I did this the other
2:05:05 night the the someone asked me how I
2:05:09 would do a storyboard with chat GPT so I
2:05:11 pulled out a screenplay that I wrote 30
2:05:13 years ago couldn't could barely remember
2:05:15 what it was about popped it into chat
2:05:18 GPT
2:05:19 within I don't know 15 minutes it had
2:05:22 written like a an elevator pitch for the
2:05:25 movie it had described the visual scenes
2:05:28 and the most important moments in it it
2:05:30 had generated storyboards from the movie
2:05:34 I had it um describe the characters and
2:05:38 and recommend current stars that might
2:05:41 play the roles and it like it did good
2:05:43 at
2:05:44 casting and then it generated images of
2:05:46 the lead characters of the movie It
2:05:49 totally brought this dead piece of data
2:05:51 back to life I think that's one of the
2:05:54 killer apps of of AI is that just
2:05:58 massive amounts of
2:06:00 human capital and intellect and
2:06:04 intellectual property that have been
2:06:06 locked away in hard drives and in
2:06:08 databases and all these [ __ ] dark
2:06:11 corners of the internet are now like
2:06:13 brought back to life in this incredibly
2:06:15 Dynamic way that we are lucky enough to
2:06:18 get to play with
2:06:21 and so
2:06:24 yeah anyway all right listen everyone I
2:06:28 gotta
2:06:32 go for me and maybe pate's present
2:06:36 wrapping my biz was all about delivery
2:06:38 and
2:06:40 packaging all right everybody I would
2:06:42 wear this dead ass yeah exactly right
2:06:46 that's [ __ ] beautiful
2:06:51 all right everybody listen let me let me
2:06:54 do some acknowledgements here um here
2:06:58 we'll go look at the latest the latest
2:07:00 activity from Danielle oh that one's
2:07:03 beautiful what's the what's the prompt
2:07:07 there that one's staggeringly beautiful
2:07:10 um okay first of all to the mods thank
2:07:13 you Amelio's wife you're on time tonight
2:07:15 holy crap that doesn't happen it was a
2:07:19 magical night we'll leave it on this
2:07:21 magical image Amelio's wife was here on
2:07:23 time
2:07:25 magical um so thank you to the mods as
2:07:28 always um and then let's do a little
2:07:30 Romper Room here I see Firefly and
2:07:33 Michaela Tobias you're incredible thanks
2:07:36 for being a subscriber Source Camp jrb
2:07:39 185 an important part of AI is the
2:07:41 middleman sorting through all the
2:07:42 garbage to present what's useful that's
2:07:45 totally it that's the human part the
2:07:47 humans are the ones that choose art
2:07:50 art is when you choose to stop right the
2:07:54 artist is the one that keeps throwing
2:07:57 [ __ ] at the canvas until they finally go
2:08:00 uh yep it's done there's art same thing
2:08:02 with the AI you just keep dumping [ __ ]
2:08:04 out of it until you go o that's good o
2:08:06 that's good o that's good absolutely
2:08:10 Broody hen good to see you jenu good to
2:08:12 see you Eric Pate Kyle check out
2:08:15 Christmas album I made later if you
2:08:16 haven't okay awesome I will do that
2:08:19 sounds very cool Ted looks at life
2:08:21 refresh okay I thought we were going to
2:08:23 leave this on on a on a lovely thing
2:08:25 what's this look more llamas
2:08:28 again the open AI product marketing
2:08:35 team oh man you guys are great Jen Jen
2:08:38 mcau good to see you uh let's see Megan
2:08:41 Papa JJ awesome ao's wife Dr J Ted looks
2:08:45 at life forever hooked good night
2:08:47 everyone new new folks come back
2:08:49 tomorrow learn more yeah that's the
2:08:51 other thing come back here like you just
2:08:54 this is an osmosis Channel like you
2:08:56 don't have to actually take notes you
2:08:58 just come back and eventually you'll go
2:09:00 oh I get it now that's totally what this
2:09:03 is about so all right everybody great to
2:09:06 see you all and uh join the salon yes
2:09:09 please join the
2:09:10 salon I will be back on Tom tomorrow
2:09:14 it's Thursday and uh we'll have a good
2:09:17 time so this was fun toight thank yall
2:09:19 peace
2:09:20 out