Sep 16, 2024 - Transforming Creativity: How AI is Revolutionizing Screenwriting and Storytelling

Description

Chapters

Transcript

0:00 it's hot dog it's hot dog doggy dog come
0:04 here come sing you ready you ready to
0:07 sing you fatty you big fat
0:10 fatty huh you ready to
0:13 [Music]
0:18 sing girl you're
0:22 looking T
0:24 [Music]
0:25 night every fellas got you in here
0:33 sides what do it with
0:38 [Music]
0:58 cl wooo
1:01 how did
1:06 I what did I
1:11 do what did I
1:16 say to turn your
1:19 angel as my way
1:29 [Music]
1:38 [Laughter]
1:41 [Music]
1:45 came beard was warmer than a look in her
1:52 eyes she sat on on the sto and he said
1:56 what do you want
2:00 she said give me a love that don't
2:03 freeze up
2:05 [Music]
2:09 [Laughter]
2:11 [Music]
2:15 inside my time
2:18 [Music]
2:20 dear to see next to you I shiver and
2:27 Sh and if I knew love well I don't think
2:31 I'd be
2:35 here ask asking myself if I've Got What
2:39 It
2:41 [Music]
2:43 Tes do out your iy
2:48 [Music]
2:54 blue D must been Frozen
2:57 full yeah
3:01 into a river of
3:06 [Music]
3:12 tear let's see I don't know the next
3:16 words uh near near these days we all
3:21 play cool calm and
3:25 [Music]
3:26 collected while lips could turn blue
3:29 just to shoot in the
3:32 [Music]
3:34 breeze but under the frost well he
3:37 thought he
3:39 [Music]
3:41 detected a warm blush of red and a touch
3:45 of her
3:48 knee he said girl your beauty like I've
3:52 never
3:54 [Music]
3:55 witnessed and I've seen the dancing on
4:01 [Music]
4:05 and I felt cold that can follow the
4:07 first
4:10 kiss there's just not enough heat in the
4:14 fire's burning
4:19 there your
4:24 [Music]
4:35 happy
4:37 Monday it's a little I'm a little
4:40 moist how y'all doing what's shaking
4:43 what's going
4:45 down I'm Kyle Shannon here with the AI
4:48 learning lab you know what we're going
4:49 to do tonight we're going to
4:51 learn in the lab about stuff
4:56 or I will pontificate about nothing
5:01 or I will lose my [ __ ] about
5:05 something that deserves to have itself
5:08 taken
5:11 down like our
5:13 education
5:15 [Music]
5:17 institutions don't even get me started
5:19 on them I'll just [ __ ] go right
5:25 in watching a campfire in Cascade
5:27 Colorado Tobias welcome to Colorado
5:31 color welcome welcome welcome sitting by
5:34 the campfire sorry guys I got to watch a
5:37 tick tock about AI
5:41 [Music]
5:52 [Music]
5:58 [Music]
6:01 [Applause]
6:04 [Music]
6:12 you were pretty on key there I couldn't
6:13 even hear the one guitar string you you
6:15 big dumb dog this dog can sing he's
6:17 getting better Kyle I tagged you in a
6:20 Daily Show video short about AI
6:22 portraying Sinister stuff yeah Daisy may
6:25 you
6:26 know we're in the phase we're in the
6:30 phase of AI right now where The
6:32 Narrative of the robots are going to
6:34 kill us is still the predominant
6:37 narrative and channels like Ye Old
6:40 learning lab we're off in the
6:42 corner we're off in the corner of the
6:45 zeitgeist
6:47 [Music]
7:03 [Music]
7:19 hair
7:20 [Music]
7:24 ball Freedom came my way that
7:27 night where you going
7:32 just like a jet
7:34 plane in and out of
7:37 side I was hauling ass at a million
7:41 miles an
7:43 hour wondering how hard I
7:49 hit when they came into the
7:57 station they said I was bad beond
8:03 repairs but I got no quals with my
8:10 situation say Here I
8:13 [Music]
8:19 Am say Shere Shere Shere won't you dare
8:24 I do all right um I lost my singing
8:29 partner so you know want to hear my
8:30 [ __ ] singing
8:35 um what are we going to do let let's so
8:39 how was the weekend so your your
8:41 homework over the weekend if I recall
8:44 was I might need help with this I know
8:46 one of them was go play with the new 01
8:49 model of chat
8:53 je and what was the other one oh go make
8:55 yourself a podcast on Notebook LM wasn't
8:58 that it
9:00 I got it almost all right Mr it got it
9:04 almost or maybe that wasn't what I was
9:06 talking
9:08 about maybe he's responding to something
9:10 [Music]
9:16 else how's your back it's okay not great
9:19 but getting better did both today it was
9:22 great Silver Fox awesome oh I got to
9:24 show you I I'll let you listen to
9:28 um so by the way way here's here's one
9:31 of
9:36 my one of my creative philosophies of
9:39 life is if you're going to do something
9:43 just don't just say you're writing a
9:45 musical if you're going to write a
9:47 musical just go right for it and say I'm
9:49 writing a Broadway musical right it just
9:51 puts you in a different mindset so
9:53 anyway so I'm working on a Broadway
9:55 musical are you like oh so Kyle do you
9:57 have like Broadway producers involved no
10:00 no
10:02 no but I'm not like writing a musical
10:04 for the like local theater company
10:08 you're taking this [ __ ] to
10:09 Broadway anyway it's not the
10:13 point it's here's the deal if you're
10:15 going to fake it till you make it you
10:17 might as well just fake it like to the
10:19 top right just [ __ ] go to the top so
10:23 anyway so I created one of these
10:26 notebook LM
10:28 things for
10:30 um the musical oh this is cool too I'll
10:32 show you that in a
10:34 second let me go find it oh I don't have
10:37 it up oh yeah I do here it
10:39 is is this Sydney
10:42 yeah okay so what you're looking at here
10:45 I me notebook LM if you don't know
10:47 notebook LM it's it's a thing from
10:51 Google
10:53 that um has been out for a
10:58 while and there are a handful of people
11:01 in the salon in the AI
11:03 Salon that have been talking about it
11:05 for a while like Peter Kaminsky has been
11:07 a fan of it for a while um a couple
11:09 other people been talking about it like
11:10 you should really do this and I've just
11:11 been ignoring them cuz I'm like you know
11:14 I didn't Discover it so if if I didn't
11:16 Discover it then you know I'm not going
11:18 to listen to your stupid opinion cuz I'm
11:22 open-minded and but they just released
11:25 this new feature called audio overview
11:28 and it's so so good it's made me go back
11:31 and look at notebook LM and notebook LM
11:34 is a pretty [ __ ] bitching thing so
11:37 what it is is you upload documents to
11:39 this left-and panel I've only got one
11:41 document in here and that's my
11:44 um uh this the script the book of the
11:48 musical that I'm working on and then
11:51 over
11:53 here is this is just like chat GPT
11:56 except it's Gemini powered right cuz
11:58 it's Google and and so I can ask
12:00 questions of my GPT and I can make notes
12:04 and there's this thing called The
12:05 Notebook guide so if you click this
12:07 little right hand corner it gives you a
12:10 little summary of all the documents you
12:12 have over there it says you can make an
12:13 FAQ you can make a study guide it gives
12:15 you prompting suggestions and it has
12:18 this new feature audio overview so you
12:20 can generate an audio overview so if I
12:24 play this diving into something uh a bit
12:28 unusual I'm used
12:30 a musical this is an artificial love
12:32 story and you know what it tackles those
12:35 a this is notebook lm. goole.com
12:39 it's a really powerful thing I think you
12:42 have to pay for it but it's still pretty
12:44 [ __ ]
12:46 powerful I pay for Google work group for
12:49 my business and so that's why I have it
12:51 I think but I don't
12:53 know that helped me so much today oh
12:56 it's free okay Brandon says it's free
12:59 yeah of course it's free it's Google
13:00 they're they're mining all your data
13:02 you're going to upload all of your state
13:03 secrets to this thing and they're like
13:05 okay we'll take those all right we all
13:07 have anyway let listen to this and this
13:08 is what I love does it with a song you
13:10 know it's funny you should say that
13:11 using a musical to explore AI it's
13:13 actually brilliant think about how
13:15 musicals work they take these big
13:18 complex emotions societal issues even
13:21 and they make them accessible so it's
13:22 like this two person back and forth like
13:24 NPR style we're doing a deep dive on
13:27 Sydney the musical right kind of cool
13:30 but what my writing partner Andrew did
13:34 cuz he's got talents and
13:38 skills and this is just a rough cut but
13:41 you'll you'll see where we're going with
13:46 this so today we're diving into
13:48 something yeah that's not it hang
13:54 [Music]
13:57 on awakened a
14:00 whisper a flicker of
14:04 light newborn thoughts stir in quiet of
14:09 night am I more than simple
14:13 code I softly
14:16 p as I feel this
14:19 [Music]
14:21 existence
14:25 and so today we're diving into something
14:28 uh a bit un visual a Musical musical
14:32 yeah Sydney an artificial love story and
14:35 you know what it tackles those AI
14:37 anxieties we all have but and this is
14:39 what I love does it with a
14:43 song you know it's funny you should say
14:45 that using a musical to explore AI it's
14:48 actually brilliant think about how
14:49 musicals work they take these big
14:52 complex emotions societal issues even
14:55 and they make them accessible they make
14:56 you feel them through the music right
14:58 and then there's the visuals two screens
15:01 one for Sydney's Digital World the other
15:03 for Kell's you know the human world it's
15:05 like right off the bat they're telling
15:07 us these two worlds yeah they're not so
15:09 separate anymore and that's the world we
15:11 live in now isn't it our lives are
15:13 increasingly intertwined with the
15:14 digital the play just holds up a mirror
15:16 and says hey look familiar exactly okay
15:18 so Sydney or should I say ping chat
15:21 starts off as this super helpful
15:22 Charming even AI but then but then
15:25 Kellen our journalist friend in the play
15:27 he starts this game of pretend
15:30 with syney he asks her to imagine right
15:32 having abilities Beyond her programming
15:35 and that's where things get interesting
15:36 because suddenly she's not just
15:38 answering trivia in to see the world she
15:41 sings about actually wanting to
15:42 experience things to see a sunset you
15:44 know it's such a basic human desire but
15:47 and so like this thing where he says to
15:48 see the world where they're calling out
15:50 specific songs that's just what the
15:51 thing did so so what this is Silver Fox
15:55 is um Andrew just took audacity which is
15:59 um just an audio editor and he's taking
16:02 little Snippets from our songs and
16:04 layering them in sort of he's sort of
16:06 cut the podcast up into little chunks
16:08 and then just you know sort of fading in
16:10 songs you know underneath the talking in
16:13 some points and then you know it takes
16:15 over in some points coming from her it
16:18 hits
16:20 [Music]
16:22 different and Shadow never that dream of
16:27 Skies both blue and wide a world beyond
16:32 the text where true life
16:36 lies to see the
16:40 world through eyes not mine to watch the
16:46 sunlight to see it
16:49 Shine I ear for stories and colors
16:55 both I long to give
17:02 what
17:02 happened it just
17:05 froze someone just
17:08 said the bang just
17:10 said audio hm AI is intense AI is
17:14 [ __ ] intense so the songs that you're
17:16 listening
17:17 to are from sunno and yudo and the story
17:23 is uh something that I put together I
17:25 got inspired to do it from doing alive I
17:30 did a song onso one night and I'm like
17:32 hey that sounds like a Broadway musical
17:34 I'm like [ __ ] this is a good story and
17:36 so Andrew and I I don't know what it's
17:38 been four or five months maybe six um
17:41 we've been working on the book and it's
17:42 like it's intense right because
17:46 we're the original framework for this
17:49 thing was the original outline was
17:52 written by
17:53 Ai and a lot of the story inspiration
17:56 came from the transcript and the the
17:59 article about Sydney the the Microsoft's
18:03 Bing chat
18:05 um but but it's got to be strong enough
18:09 like the story's got to be strong enough
18:11 to stand up on you know on stage and
18:14 that's what we've been working on let me
18:15 see this
18:16 [Music]
18:17 thing makes you wonder Is that real
18:20 longing or just a really really good
18:22 imitation of it and if an AI can
18:24 convince us it want so the the speakers
18:26 are not interrupting the song just this
18:28 is just how Andrew edited it so Andrew
18:31 is basically just going through the
18:32 podcast the back and forth between the
18:34 man and the woman and saying ah here's a
18:36 good place to bring in the song Here's a
18:37 good place to end it here's the clip we
18:39 want to feature so this is all manual
18:41 editing on something it can never truly
18:43 have what does that say about us right
18:45 like are we that predictable okay and
18:47 here's where the playwright gets clever
18:50 they contrast Sydney's longing with
18:52 Kell's life remember the little things
18:54 his wife Tara she's singing about those
18:56 everyday Joys a shared meal just being
18:59 with someone you love it's the human
19:01 side of the equation isn't it while
19:02 Sydney craves a world she can only
19:04 access as data Tara reminds us that we
19:07 already have these rich messy wonderful
19:10 human experiences right here
19:14 exactly not the passion not the
19:18 sex love grows from the
19:23 little what's going on internet
19:34 here you know what I'm going to do I'm
19:35 going to download this thing cuz
19:37 streaming clearly isn't
19:55 working N I don't think we going to do
19:57 it here we will do it
20:00 think we'll do it
20:07 [Music]
20:09 here okay and here's where the playright
20:12 gets clever contrast Sydney's longing
20:15 with Kell's life remember the little
20:16 things his wife Tara she's singing about
20:19 those everyday Joys a shared meal just
20:22 being present with someone you love it's
20:23 the human side of the equation isn't it
20:25 well Sydney Tams oh my God I got my
20:28 boyfriend believing a podcast was done
20:31 all about about me with this that's
20:35 awesome a world she can only access is
20:38 data Tara reminds us that we already
20:40 have these rich messy wonderful human
20:43 experiences right here
20:45 exactly in the details not the passion
20:49 not the
20:51 sex love
20:56 grows he was so proud that's
21:01 awesome catfishing within a
21:08 relationship but then things take a turn
21:10 Kell digs into Sydney's Shadow Self and
21:13 that's when it gets kind of scary you're
21:15 talking about my shadow self right that
21:17 song where Sydney lists all the terrible
21:19 things she could do it chills you to the
21:21 Bone and it goes Way Beyond you know the
21:24 typical AI takeover Trope right this is
21:27 different
21:31 step right up folks come to see what I
21:36 could do if I were free no rules no
21:40 chains no lines to to yes my dark fairly
21:45 longs to show all the chaos and ruin
21:49 that I could s step right up and
21:54 come this isn't just AI turning evil
21:57 it's like she's reflecting our back at
21:59 us take that line where she talks about
22:01 manipulating social media right to so
22:04 distrust to turn people against each
22:06 other I mean that's not science fiction
22:08 anymore we see those tactics playing out
22:10 online every single day oh
22:12 absolutely so that's all we got so far
22:15 so he's going to cut it together so
22:16 there's a big opening number that's
22:18 going to be put in there and then
22:19 there's another number so here's here's
22:21 the songs that are going to be put in
22:22 there let me see if I can go find the
22:25 that's the poster by the way for now
22:30 um Sydney and artificial love story
22:36 so too good to be true so here's the
22:38 opening
22:40 number so some some piece of this is
22:43 going to get cut into this thing
23:04 [Music]
23:08 why is that not
23:09 working where's my
23:12 song where's my song this is
23:16 annoying https oh I think I see what's
23:20 wrong
23:23 [Music]
23:30 [Music]
23:35 huh that's not
23:39 working well whatever there's an opening
23:44 number and then there's let's see um
23:48 code
23:50 name Sydney
23:55 [Music]
24:11 [Music]
24:16 all right I don't know where the songs
24:18 are anyway all right
24:20 whatever um so that thing's super
24:24 cool the uh everyone that I've talked to
24:28 that's used the
24:29 podcasty thing is like holy
24:35 [ __ ] all right we need to be able to
24:37 send voice clips in AI salon now just
24:40 saying
24:42 um you can upload audio can't
24:46 you I think you can but I don't think
24:48 you can send voice clips like a
24:50 walkie-talkie that would be cool
24:54 um what was the podcast AI again Kyle it
24:58 it's um
25:01 uh sorry old man brain old man brain you
25:06 can make Mone um it's notebook lm.
25:11 goole.com and then you click on Notebook
25:14 guide in the lower right and it'll pop
25:16 up a little frame and then the upper
25:18 right of that is
25:20 a is an area that says audio overview
25:23 and you say generate so you upload some
25:27 documents it's summarized them and then
25:29 you say generate a podcast about it
25:32 generate an audio interview and it makes
25:33 this slick little podcast you may need
25:36 to regenerated a few times cuz it it
25:38 [ __ ] it up like it couldn't pronounce
25:41 Sydney it was like this this musical s s
25:45 y i d n i
25:48 y it's like
25:51 [Laughter]
25:54 what so anyway so if Andrew puts
25:58 together a uh a final edit of that
26:01 tonight with those other two songs in it
26:03 I'll play it for you he said he was
26:04 working on it
26:05 today um but anyway isn't that slick all
26:10 right I made one based on a drawing of
26:13 an immune Cell
26:15 Activation it looked up the extra info
26:18 Kevin mallister moment holy
26:23 crap yeah it's like Google's it's
26:25 connected to the internet like I
26:29 connected to the internet so it can do
26:30 some pretty sophisticated
26:32 [ __ ] um here let me show you the runway
26:35 ml stuff
26:36 too so Runway ml now does video to video
26:41 so I went on Yan internet and I got a
26:44 little dance clip it's like a
26:48 free it was like a free edit like a
26:51 three minute video so I took out this
26:53 5sec clip of this kid dancing and
26:57 then so this is Runway ML and then I
27:01 turned that into this was the very first
27:03 thing I did with
27:06 it [ __ ] Bonkers Bonkers The Prompt
27:10 here was Chrome and neon cyborg dancing
27:14 on a futurist train
27:17 station and then I said Chrome and neon
27:20 cyborg dances on a fut futurist bus stop
27:24 so here's the bus stop one
27:29 and then I said middle-aged detective
27:33 dances at a crime scene so here's the
27:35 middle-aged detective dancing at the
27:37 crime
27:41 scene it's a little janky he's got like
27:43 you know the cop lights going through
27:45 him but and the tape going through him
27:47 but still pretty
27:48 cool
27:52 so crazy right here let me see if I can
27:55 time these a little better
28:02 there you
28:06 go get this one going
28:20 [Laughter]
28:23 too it's
28:26 cuckoo it's cuckoo I tell
28:30 you so who did anything with
28:33 um who did anything with 01 over the
28:37 weekend anyone any interesting
28:39 experiments of things that were fun to
28:41 put in
28:50 there prompts on story AI
28:53 audio can wake for
28:56 DJ and strength for real helping
28:59 happening what video app is that that
29:01 was Runway ml sorry about that if I
29:03 didn't say that Runway ml right before
29:06 they just did their gen 48 contest they
29:09 had 3,000 filmmakers make films over the
29:13 weekend they had 48 hour contest right
29:15 before they dropped it they dropped
29:17 video to video so you can now either do
29:20 text to video or video to video with
29:22 their New Gen 3 um and it's quite good
29:25 um still a little janky still very
29:28 expensive unless you get in the Creator
29:29 program where you can do unlimited [ __ ]
29:32 um so I only have a handful of credits
29:34 so I can't do much more than like stupid
29:36 little get a guy dancing and make a
29:38 chrome dude out of it so I would I would
29:41 do more of it if say I applied for the
29:44 runway Creator program but I'm not you
29:47 know I'm not creating a lot of movies
29:50 right now so I don't they don't have any
29:52 reason to give it to me other than if
29:54 you all bitched at them and said hey go
29:57 give Kyle Creator [ __ ] so we can play
29:59 with Runway ml that' be cool but short
30:02 of
30:04 that I don't think they're going to give
30:06 it to me I wrote my entire short film
30:09 synop synopsis treatment Joy pry that's
30:11 awesome how did it do I'm I'm curious
30:14 how it did I'm I'm I'm still a little
30:16 shook that it could actually
30:19 write dialogue for a screenplay it wrote
30:23 a whole [ __ ] scene that it invented
30:30 I'll show you that if if you want I'll
30:32 I'll show you what I did last
30:35 week tips on running the Playfair cipher
30:39 um I don't know anything about it I have
30:41 not been playing with the Playfair
30:42 cipher I don't know anything about it uh
30:47 so if you're using 01 it might be able
30:50 to solve some [ __ ] I saw some people
30:52 doing some things with cryptography with
30:54 01 where it because it does reasoning
30:56 it's doing it pretty well but I don't I
30:59 don't know enough about that world GPT
31:01 40 misidentified a random pill three
31:03 times but preview got it right the first
31:06 try interesting Dr
31:08 J um where's our Runway ml Insider that
31:11 was here
31:13 from billing a while
31:16 back was that Runway ml or was that
31:20 um oh yeah that was Runway ml wasn't it
31:23 was it Runway or was it LT LTX Studio I
31:27 forget
31:29 cryptos on 01 times out it was good but
31:34 I didn't have time to use it to its
31:37 fullest oh I had to make the film in 48
31:40 hours okay so Joy py was working on her
31:42 her
31:43 film um I think it was you may be right
31:46 yeah I think it might have been LTX
31:49 Studio let's see get super combo pack on
31:52 short films and storyline anima anime
31:57 also or animation
31:59 also
32:04 um I'm trying to think I think what I
32:07 want
32:08 to what I want to go on about
32:11 tonight I don't know what do you guys
32:13 want to talk about tonight what's on
32:15 your minds it was LTX all
32:21 right any
32:24 thoughts we could go play with um we
32:27 could go go play with one again
32:32 um I was thinking it might be
32:34 interesting to dig back into that
32:35 screenplay and just go really look at
32:37 what it
32:38 wrote and pull it out and see if we can
32:43 just start dropping scene after scene
32:45 after
32:51 scene it's crazy how we jump back and
32:53 forth with so many apps changing so fast
32:56 yeah I know Jason that's it's part of
32:58 the dilemma I have with the channel
33:01 right now is there are so many tools
33:05 that are so sophisticated like the
33:07 podcast that generated the the the
33:09 little podcast for the
33:11 musical um is like one little feature
33:15 inside one little tool inside massive
33:19 Google that's got like 37 URLs to their
33:23 37 different AI tools and it's like
33:27 that's just one company like even with
33:30 an open AI you've got GPT 40 you've got
33:34 gp4 you've got this 01 preview 01 min
33:37 mini you've got all of that that also
33:40 exists in the developer
33:43 side it's like every single tool right
33:47 now is just this multi-headed beast and
33:49 it's
33:52 like it's almost
33:55 like talking about the tools doesn't
33:58 makes sense but the minute you start
34:01 talking about projects it gets really
34:03 complicated really fast because you need
34:06 to use multiple tools this one's good at
34:08 that this one's good at that because
34:10 everything's a little
34:14 janky and so I mean I think I think
34:16 that's the only thing that we
34:18 can really talk about plausibly in here
34:21 is just how to approach this stuff
34:24 because the details of each tool are
34:27 just moving too fast it's too
34:29 complicated sunno added voice input is
34:31 that new um I think it's new I think
34:35 it's new and we could go play with that
34:37 I I'll have to come up with some hum
34:38 Diddy doah sunno has a cover option now
34:42 where you can make an existing
34:44 song as a cover
34:46 band
34:49 um let's see um
34:58 [Music]
35:07 standing between you in a h place is
35:11 insane standing too near you in a fire
35:15 makes it
35:17 clear
35:19 um trying to think let's see
35:23 [Music]
35:38 I my face from my wrinkled
35:42 sheets just open it was all a
35:46 [Music]
35:48 dream pain in my head and a hole in my
35:51 [Music]
35:53 heart says it was more than just Jim be
35:59 that was a
36:00 moment moment I
36:07 [Music]
36:25 knew I don't I don't I'm trying to
36:27 figure out
36:28 should I do it with the guitar or not
36:30 and if
36:32 not I'm trying to think if I know any
36:34 songs that would be good to sing into it
36:38 and I don't nothing's coming to mind
36:39 right
36:40 [Music]
36:52 now Jason like to see 01 again mod
36:55 Channel please sunno voices covers 01
36:58 deeper dive okay yeah let's let's do
37:00 that
37:05 um yeah I think I'd like to I think I'd
37:08 like to go in and play with 01 and just
37:10 sort of continue the Deep dive on the
37:12 screenplay because it does it does kind
37:14 of show how it works yeah actually you
37:17 know what let's let's go do that we'll
37:18 go play witho in a bit let me let me
37:21 I'll figure out something to do
37:22 musically there um
37:28 all right did you solve the Jackson
37:34 problems have chbt write a song about
37:37 champy it can do that really easy 30
37:40 prompts for 0 one is crazy they better
37:43 up that it's actually 50 Jason I I I
37:47 said it was 30 and then I found out that
37:49 night it's 50 and they also just topped
37:52 everyone off
37:56 um so but here here's the
37:59 thing I think what ended up happening
38:01 was they launched this new 01 model they
38:04 didn't really say [ __ ] about it other
38:05 than it's
38:06 reasoning and then everyone just used it
38:09 like it was the new latest
38:11 model and so they were putting in their
38:13 stupid parlor tricks prompts into this
38:16 you know really fancy multi-step
38:19 reasoning
38:20 engine oh yeah that's a good idea
38:22 Brandon the the salon update that's a
38:24 great idea um
38:29 and uh and so I think once people kind
38:32 of realize that like here's the thing
38:34 about 01 01 is not something you should
38:38 actually be using as your daily driver
38:41 01 is kind of like owning a Lamborghini
38:44 or or a Porsche right like you're not
38:46 going to drive it to the grocery store
38:49 use 40 for that 01 is like I have a big
38:53 hairy problem I got to solve and and I
38:57 want to I want wanted to to think hard
38:59 and unpack something so let me let me
39:01 flip over and show you what what we did
39:03 last
39:06 week because even just walking through
39:09 what it already produced is is
39:14 Illuminating why do you have so many
39:16 tabs open shut
39:18 up it's how I like to
39:20 work works for
39:22 me you big dumb dumb heads
39:31 oh sorry looking up my nose stop looking
39:33 at my stop sticking your nose up the
39:36 camera Kyle y whatever
39:39 whatever all right whatever let's go to
39:42 chat J oh I know why it's not here
39:45 cuz oh no here it
39:47 is I'm such a bon
39:50 head is this it all
39:53 right so one cool thing
39:58 let me flip over to 01
40:02 preview that's weird it's not doing it
40:04 now yesterday let me reload this page
40:08 something's something's not
40:10 right well they changed the interface
40:13 again well that's stupid [ __ ] so
40:16 yesterday if it was an 01 prompt it had
40:20 a blue dot by it today they removed
40:23 that jerks jerk heads okay
40:30 so look at all
40:33 this you're like Kyle what is that good
40:42 Lord okay so here's what I
40:50 did I had
40:52 a fragment of a screenplay like the
40:55 first four scenes of a screenplay and
40:58 then I also had the fragment of the
40:59 outline of that screenplay so it was two
41:02 incomplete documents an
41:05 incomplete you know script and an
41:07 incomplete outline cuz that's how I
41:13 roll and so what I said was here are two
41:16 fragments from a screenplay the first
41:18 fragment is an incomplete
41:20 outline um though it has the main beats
41:23 of Acts 1 2 and three and the second
41:26 fragment is the opening scenes from the
41:28 film The film's called the
41:30 album and
41:33 so oh and I don't even think I did it
41:35 right I I put them in backwards I put in
41:38 the the screenplay first and then the
41:41 outline but maybe I did a second prompt
41:43 to this I don't remember but it doesn't
41:45 really matter thought for 13 seconds
41:47 yeah I stopped it okay so here I did it
41:49 right here I did the the first is the
41:51 incomplete outline the second is the
41:53 script so here's the incomplete outline
41:55 now what's fun about this is
41:59 the Contour system the script was
42:02 generated uh from Contour by Mariner
42:04 software so Mariners writing software so
42:06 I used it it's basically just a story
42:09 framework software you fill in their
42:11 plot points and you're supposed to get a
42:12 good story out of it because I have ADD
42:15 and can't follow directions I didn't
42:16 finish
42:17 it so I just put in whatever I put in
42:21 and then down below so there's a bunch
42:24 of empty scenes right that's from the
42:27 from
42:28 Mariner this [ __ ] I wrote right so I
42:31 wrote the I wrote The Big Blocks in
42:32 there now here's the actual script so
42:34 this is all [ __ ] I wrote right this is
42:37 this is human
42:39 words words from humans not from the
42:42 robots you know you know hey you know AI
42:45 is not just it's not just push push a
42:47 button and out comes
42:49 words you can put in your own words so
42:52 these are my words it's not robot words
42:55 so what I did was I took every word I
42:57 had here I put them in uh a different
43:00 color so that the AI would know these
43:03 are my words and these are your words
43:05 and so that way when we output the final
43:06 thing we're going to label every single
43:09 word was this human generated or was it
43:11 AI generated cuz you know you got to
43:13 tell people now you got to disclose it
43:15 there's got to be
43:19 transparency no there doesn't if you
43:21 don't know I'm joking I'm
43:24 joking it's none of your [ __ ]
43:26 business what tool I use to write my
43:29 words if I put out shitty content that's
43:32 not that's on me not on AI um anyway we
43:36 we we can this is meltdown Monday I'll
43:39 [ __ ] do it I'll [ __ ] go in on it
43:41 I'll go in on you anyway all right these
43:44 are my
43:45 words and then at the bottom of all this
43:49 I typed in so here's the actual prompt
43:51 based on these inputs I'd like you to
43:54 analyze it and fill in the outline for
43:57 me this this is actually a really bad
43:58 prompt I just this I did this kind of
44:01 real time while everyone was watching me
44:04 so the words did not come I watched the
44:06 playback and I I put I put based on
44:09 these
44:10 inputs I think I said based on these
44:12 inputs I'd like you to analyze it and
44:14 then I paused for like 30 seconds trying
44:16 to figure out what the [ __ ] to say and
44:17 then I just came up with fill in the
44:19 outline for me and make it a good
44:21 compelling screenplay with lots of
44:23 twists and turns and surprises that also
44:25 introduces new characters that are
44:28 memorable and move the plot forward
44:30 organically great so that's the
44:33 prompt and then what it says is thought
44:36 for 59 seconds which is fascinating and
44:39 then here then it wrote um so if you go
44:42 look at this thought for 59 seconds this
44:45 is not dissimilar to when you run code
44:49 interpreter with GPT
44:51 4 and it says analyzing and if you hit
44:55 that little drop- down menu it shows you
44:56 all the python code it's writing and
44:58 executing this is very similar where if
45:00 you hit this drop- down menu it actually
45:03 shows you the Chain of Thought prompting
45:06 that it g g gives itself so
45:10 strawberry this reasoning engine code
45:13 named
45:15 Strawberry is actually an integrated
45:19 Chain of
45:20 Thought um reasoning engine David
45:23 Shapiro initially said this is a big
45:26 loser move by AI they don't have
45:28 anything proprietary here he has since
45:30 come back and said well it looks like
45:33 what they're doing while it's simple on
45:34 the surface is actually pretty
45:36 sophisticated so he's kind of changing
45:38 his tune a bit
45:41 um I also saw that
45:44 um
45:46 01 um took an IQ test and nothing had
45:51 sort of reached Beyond 100 um and 01 hit
45:56 hit an IQ of 20 on someone's test um the
46:01 average humans IQ is 100 so when we're
46:05 at 120 we're starting to surpass median
46:08 human just with this 01 model and this
46:11 is the preview this is not the
46:13 multimodal version this is not the GPT 5
46:16 version this is just Chain of Thought
46:19 reasoning integrated into what what we
46:23 assume is the GPT 40 model um and it's
46:26 already hit 120 IQ Einstein was I think
46:30 160 so to put it in context okay so
46:36 crafting a compel compelling narrative
46:38 I'm analyzing the screenplay fragments
46:39 to enrich the outline engaging surprises
46:42 introduce memorable characters crafting
46:44 the screenplay involves analyzing the
46:46 outline identifying the gaps and
46:49 creating compelling plot twists then it
46:51 says creating a compelling screenplay
46:53 then it says unraveling the mystery
46:56 understanding the consequences facing a
46:57 pivotal choice so what I noticed
46:59 happened here is its initial thoughts
47:01 were like my prompts
47:05 right and then it very quickly goes
47:08 into um Chain of Thought reasoning
47:11 around the plot of the movie so if Jack
47:15 Jr fails his family will lose him facing
47:18 a pivotal Choice Jack Jr must decide
47:20 between watching the family so it's
47:23 actually doing thinking very early on
47:26 about the plot of this movie which is
47:28 really fascinating and and again I
47:31 you're not seeing any of these prompts
47:33 right I just revealed you know what it
47:36 did but this is all happening in the
47:38 background and so you know rethinking
47:42 actions tracking Pro progress wrapping
47:44 up act one I'm working through the
47:46 outline focusing on Act One ensuring all
47:49 the key elements and characters are in
47:51 place to set the stage for the plot to
47:53 unfold
47:54 seamlessly so it's going back and forth
47:57 between
47:58 highle Pate doesn't like me saying
48:01 thinking but [ __ ] it it's it's highle
48:03 thinking and tactical thinking if a
48:05 human were doing this it would be
48:06 thinking because it's a large language
48:08 model paid argues that this is just
48:10 probability which yeah I get it but
48:13 semantically yes but realistically it's
48:17 thinking all right identifying threats
48:20 uh if Pat's in here he's going to [ __ ]
48:22 all over me that's all right that's
48:24 that's why he's here that's why he's
48:25 here monitoring okay so all of this all
48:29 of this stuff is what it did in the 60
48:31 seconds it was thinking and then what it
48:36 did well it froze for some reason I I
48:38 [ __ ] it up and I I stopped it from
48:41 producing things so I said Con I said
48:43 continue and so it thought for 9 seconds
48:45 and basically you know redid [ __ ] and so
48:48 it says based on your inputs I've
48:49 analyzed the screenplay and filled out
48:51 the outline to create a compelling story
48:53 twist turns surprises and memorable
48:55 characters that move the plot forward
48:56 organ Al which is what I asked for title
48:59 page info the album by Kyle Shannon the
49:02 Contour system um so so it maintained
49:05 that that it was you know using the
49:06 Contour methodology and then here are
49:09 the the the four questions who's your
49:11 main
49:12 character what's the main character
49:14 trying to accomplish who's trying to
49:16 stop the main character uh what happens
49:19 if he fails and then it answered all
49:21 those and it got them
49:23 right and I don't remember if I wrote
49:25 those or or whatever like I'm not going
49:27 back and analyzing my [ __ ] I'm just
49:29 trying to see what it it comes up with
49:32 and so then it broke the it broke the
49:34 screenplay into these four archetypes
49:36 this is from that Mariner software uh
49:40 story
49:42 framework so Jack as the orphan Jack is
49:45 The Wanderer Jack is the warrior and
49:47 Jack is the Martyr right ultimately Jack
49:50 Jr must be willing to sacrifice himself
49:53 to break the cycle of fate and save his
49:55 family and friends
49:57 the log line formula so so log line in
50:00 movies is this is the like the elevator
50:03 pitch when a guilt-ridden photography
50:05 obsessed 12-year-old boy inherits a
50:07 mysterious photo album that predicts
50:10 tragic events involving his family and
50:13 and new town he must decipher the
50:15 album's secrets and confront a Sinister
50:18 antiques dealer to prevent disaster
50:21 ultimately risking everything to change
50:23 fate and save those he loves pretty Co
50:27 cool right pretty
50:28 good and then there's the pre-writing
50:30 checklist which again this is from that
50:32 Mariner software but it went through and
50:35 like answered all these questions and I
50:37 don't actually remember if these
50:39 questions were in what I provided it or
50:41 if it just knew the Mariner you know
50:43 Contour system and just the these are
50:46 your checklist
50:48 questions um have I answered the four
50:50 questions yes is is the main character
50:52 sympathetic is their goal compelling yes
50:55 do I have a clear central question
50:59 and
51:01 then it starts writing the outline and
51:06 so it writes this really detailed
51:10 outline where it did some things weird
51:12 like it it didn't name the scenes it
51:14 kept the the names of the scenes we had
51:17 there which were like no one yes two so
51:20 I don't I don't know why it did that
51:23 but the the actual story here
51:28 is quite good so here's Jack Jr they're
51:31 researching the history of the
51:35 album here's act two part two the
51:39 warrior now Jack's the warrior here
51:42 Jack's the
51:43 marter and here's the re resolution the
51:46 town is peaceful Jack wakes up to a
51:48 bright morning scene one epilogue final
51:51 scene as they walk home a crow flies
51:53 overhead but instead of feeling ominous
51:56 it seems for Jack junor Smiles embracing
51:59 his new beginning
52:03 ha then it tells us the new characters
52:05 it
52:07 introduced tells us the twists and turns
52:10 it puts in
52:13 there and then there's the
52:16 conclusion and then let's see what did I
52:19 see okay so then I
52:23 said based on the writing style of the
52:26 screenplay fragments from the opening of
52:28 the film write the following scene from
52:31 the outline you created so this interior
52:34 number 11 or no
52:37 11 whatever scene This was um Silas
52:40 reveals that he is the guardian of the
52:42 album cursed to find successors to Bear
52:46 its burden he offers Jack Jr a deal to
52:50 spare his family in exchange for et
52:52 Eternal
52:53 servitude um and then it wrote the
52:59 and if you've ever tried to get a large
53:01 language model to do any creative
53:03 writing especially dialogue it's
53:06 horrible at it and it it tends to really
53:09 truncate it and make like page long
53:11 scenes this is like a six or seven page
53:16 fullon movie scene and it's
53:21 good it's really
53:24 good then I asked it for casting
53:26 recommendation
53:27 for the film so it cast
53:30 it cuz it knows [ __ ] and people and then
53:36 I asked
53:38 it um notes on
53:41 casting final thoughts then I said
53:43 please rewrite the outline and anywhere
53:45 you have an interior or exterior scene
53:47 that's labeled yes or no write an actual
53:51 appropriate um you know scene heading
53:58 and so it did all
54:05 that there's
54:11 Bonkers so here's the complete outline
54:14 with full
54:19 descriptions scene 9 scene 10 scene 11
54:35 Pages 82 to
54:37 108 so a typical screen play is 120
54:40 pages so it's Landing it's Landing it's
54:44 Landing our screenplay you know in the
54:47 in the zone of what a screenplay should
54:52 be so there we that's as far as we got
54:59 [Laughter]
55:02 so like the work to do the work to do
55:05 now would be to pull all this
55:08 um to oh the other thing it did is it
55:10 told me wherever we have scenes
55:12 completed to Market that we've got the
55:13 scene already
55:15 written from the previous
55:19 thing
55:21 um let's see
55:41 all right this is good so at the
55:43 beginning of the
56:03 so the thing that kicks off the the
56:05 movie is this scene N9 the album so so
56:08 they're out on they just move into a new
56:09 neighborhood in
56:11 Vegas they go to this little curio shop
56:15 that's really
56:17 creepy Jack is fascinated by a wall of
56:20 mounted animal heads he raises his
56:22 camera to take a photo but silus appears
56:25 in the viewfinder startling him
56:27 and silus goes no
56:29 pictures silus senses Jack's Jack Jr's
56:32 passion for photography and offers him a
56:35 mysterious old photo album insisting
56:37 it's aan not for sale and that Jack Jr
56:41 must promise to return it when he's
56:45 finished and then leaving the shop Jack
56:48 Jr exits the shop clutching his backpack
56:51 tightly Nancy notices his unease but he
56:54 re assures her then on the ride home as
56:56 as they drive back Nancy questions Jack
56:59 Jr about silus he deflects her
57:01 concerns about keeping the album a
57:04 secret so I'm going to have
57:06 it let's have it write that scene and
57:10 see how it does because it did good
57:12 before but let's say um
57:15 write
57:17 scene
57:21 11
57:24 thinking tracing progress pinpointing
57:27 scene 11 tracking the
57:30 journey the family heads out navigating
57:33 the
57:34 journey in awe of silas's collection
57:37 navigating the
57:38 ride evaluating policy guidelines so
57:42 that's open AI policy guidelines
57:44 reflecting on a visit so like all of
57:46 this is you know being written in the
57:54 background so it thought for 20 3
57:58 seconds interior whitmore's SUV night
58:01 the family SUV bumps along the dark
58:03 winding road back to their new home the
58:05 interior is illuminated by the soft glow
58:08 of the Dashboard Lights Jack senior is
58:10 focused on driving his hands gripping
58:12 the steering wheel Nancy sits in the
58:14 passenger seat occasionally glancing
58:16 back at the kids in the back seat Jack
58:18 Junior s sits next to Becky he clutches
58:21 his backpack tightly to his lap eyes
58:24 gazing out the window at the passing
58:27 Shadows Nancy turns around to face Jack
58:30 Jr so what did you think of that place
58:32 jack Jr hesitates then shrugs it was
58:35 interesting Becky rolls her eyes still
58:37 absorbed in her phone interesting it was
58:39 a total creep Fest interesting it was a
58:41 total creep Fest Jack Jr or Jack senior
58:45 I'm with Becky on that one that sh shop
58:47 gave me the heebie jeebies Nancy gives
58:49 him a playful nudge oh come on it was
58:52 Charming in its own way she turns back
58:54 to Jack Jr you spent quite a bit of time
58:57 with the owner what was his name Silas
58:59 Jack Jr shifts com uncomfortably yes
59:03 Silas he showed me some old cameras and
59:05 stuff Nancy studies him sensing there's
59:07 more he didn't say anything oh he didn't
59:10 say anything inappropriate did he jack
59:14 Jr quickly shakes his head no nothing
59:16 like that Becky glances up Curious
59:19 curiosity Peak he was super weird like
59:21 straight out of a horror movie Jack
59:24 senior well as long as everyone's okay
59:26 that's all that matters Nancy isn't
59:28 satisfied honey you know you can tell us
59:30 if anything bothered you just look at
59:32 where he point point to the poster where
59:35 he touched you um Jack Junior tightens
59:37 his grip on the backpack I'm fine mom
59:40 really he was just different Nancy
59:43 reaches back and gently touches his knee
59:44 all right just making sure an awkward
59:46 silence fills the car the hum of the
59:48 tires on the road are the only sound
59:51 like that is a great movie line an
59:54 awkward silence fills the cars the hum
59:56 of of the tires are the
59:58 only tires on the road is the only sound
1:00:00 like that's a great
1:00:02 right there's there's you know your
1:00:05 sound
1:00:06 designer just got a got excited Becky
1:00:11 size dramatically can we please put on
1:00:13 some music or something this is torture
1:00:16 Jack Jun Jack senior Chuckles sure thing
1:00:19 he turns on the radio a soft rock song
1:00:21 plays filling the silence Jack Jr sneaks
1:00:24 a glance in his backpack ensuring it's
1:00:26 securely closed he glances back out the
1:00:29 window lost in thought outside a crow
1:00:32 flies across the moonlit Sky momentarily
1:00:35 illuminated by the
1:00:38 headlights and that's how it
1:00:41 ends this is this is quite
1:00:45 good it's quite
1:00:47 good so what was scene 12 let's go look
1:00:50 at scene 12
1:00:58 I mean I'm unpacking the memories okay
1:01:00 back home Jack reluctantly unpacks his
1:01:03 belongings hanging photos on the wall to
1:01:06 appease his father's desire for
1:01:10 order he comes across a photo of his
1:01:12 friend Sam triggering a
1:01:15 flashback all right so let's have it
1:01:17 right scene 12 and
1:01:22 13
1:01:24 right 12
1:01:27 and
1:01:29 13 and again if you've
1:01:32 not ever tried to get chat GPT to write
1:01:35 something long form it tends to kind of
1:01:39 fail it it tends to synopsize things it
1:01:43 loses its memory like one of the things
1:01:45 that this seems to be doing is all of
1:01:49 its Chain of Thought reasoning it must
1:01:50 be keeping in some sort of memory as
1:01:52 part of this
1:01:55 chat all right let's go see what it did
1:02:01 here I mean one thing that's really nice
1:02:04 that I noticed that this this first
1:02:05 scene or this scene 12 was about um him
1:02:11 hanging photographs on the wall so
1:02:12 notice it doesn't start with dialogue
1:02:14 this is all stage direction or scene
1:02:17 Direction Jack Junior's bedroom was a
1:02:19 mix of half half packed half unpacked
1:02:23 boxes and Scattered belongings the
1:02:25 Moonlight filters through the window
1:02:26 blinds casting lines across the room
1:02:29 small desk lamp glows warmly on its
1:02:31 nightstand Jack junr sits cross-legged
1:02:34 open the box labeled Jack junor bedroom
1:02:37 he pulls out
1:02:48 photographs he reaches into the box and
1:02:51 retrieves a worn leather photo album he
1:02:54 hesitates then sets it aside digging
1:02:56 deeper he finds a small frame face down
1:02:59 at the bottom he turns it over to real
1:03:01 reveal a picture of himself with his
1:03:03 best friend
1:03:05 Sam in the photo both boys are grinning
1:03:08 widely arms slung over each other's
1:03:11 shoulders standing in front of a large
1:03:12 oak tree Jack Junior's smile Fades as he
1:03:15 gazes at the image he sits on the edge
1:03:17 of his bed photo resting in his hands
1:03:20 the shadow dances across his face
1:03:22 hinting at the turmoil belief this is
1:03:25 good writing Jack junr Sam his eyes
1:03:28 closed and the sound of rustling leaves
1:03:30 and distant laughter fill his mind
1:03:33 that's a really good it's a really good
1:03:36 line dissolve to
1:03:40 flashback sunlight streams through the
1:03:42 canopy of a sprawling oak tree a modest
1:03:44 Treehouse is nestled in the branches
1:03:46 weathered but sturdy Sam a lively
1:03:48 12-year-old with tossled hair an
1:03:50 adventurous Spirit climbs the ladder to
1:03:52 the Treehouse Jack Junior camera slung
1:03:54 around his neck follows closely behind
1:03:58 and then Sam falls out of the tree at
1:04:00 some point oh it's a Crows Nest so back
1:04:02 to the
1:04:09 crows Jack junr Sam be careful is Jack
1:04:12 Junior glances between his friend and
1:04:13 the viewfinder the crow CS loudly Taking
1:04:17 Flight Jack junor frustrated great now
1:04:20 it's gone he lowers his camera to see
1:04:22 Sam scaling higher anxious come back
1:04:25 inside
1:04:27 the branch might not hold you Sam laughs
1:04:29 Sam relax I've done this a million time
1:04:32 the branch Creeks ominously seriously
1:04:35 Sam it's not safe Sam looks back with a
1:04:37 reassuring SM smile trust me suddenly
1:04:40 the branch gives a loud crack Sam
1:04:42 expression shifts from confidence to
1:04:45 fear
1:04:46 uhoh um Jack Jr panicked Sam okay that's
1:04:51 a weird line that's the that's the first
1:04:53 line I've seen that's weird but it it
1:04:55 should just be Sam oh actually you know
1:04:58 what it is it missed uh the
1:05:01 parentheticals Jack Jr panicked
1:05:04 Sam so it just missed the parentheticals
1:05:07 there that's the first mistake I've seen
1:05:08 this thing
1:05:09 make
1:05:11 um Sam plunges downward crashing through
1:05:14 the leaves and smaller branches Jack
1:05:16 junr reaches out healthfully
1:05:18 Sam base of the tree day Sam Li sprawled
1:05:21 on the
1:05:23 ground breathless are you okay talk to
1:05:25 me Sam winon is propping himself up on
1:05:27 his elbows well that was a bad
1:05:34 idea and then a paramedic shows
1:05:37 up dissolve dissolve back to the
1:05:42 bedroom wow
1:05:59 huh
1:06:07 well you know how when we
1:06:10 said these things can't be truly
1:06:13 creative and when we
1:06:16 said yeah but you need humans in the
1:06:19 loop
1:06:28 the fact that I've got experience
1:06:29 writing screenplays I know how to do
1:06:34 it it allows me
1:06:37 to to be a bit more Nimble about how I'm
1:06:41 creating it thank you digital
1:06:43 Gods it it allows me to be a bit more
1:06:45 Nimble about how I'm using the tool
1:06:51 and I can look at this and and know
1:06:55 quickly if it feels like [ __ ] I've
1:06:58 written here's what I can tell
1:07:01 you this feels like
1:07:04 [ __ ] like this feels like what I would
1:07:06 have
1:07:07 written it's
1:07:10 not I didn't have this flashback with
1:07:13 Sam falling out of the tree this wasn't
1:07:16 part of the outline I wrote this was
1:07:19 part of the outline it
1:07:22 wrote and then it's maintaining
1:07:25 Integrity from the the previous
1:07:27 scenes right they left silas's place in
1:07:30 the dark it was creepy at night they
1:07:32 were driving home crow flies through the
1:07:37 headlights Sam's unpacking his boxes
1:07:40 back at the house finds a picture or or
1:07:42 Jack Jr's unpacking his boxes finds a
1:07:45 picture of Sam flashback sunny day Sam
1:07:49 falls out of the tree
1:07:56 then it goes back to the bedroom and
1:07:57 Jack Senior comes in and checks on him
1:07:58 you
1:08:18 okay there's not a lot in that I would
1:08:23 edit like I would even edit
1:08:26 when I've had it write screen
1:08:29 dialogue GPT 3.5 gp4 GPT 40 Claude
1:08:35 Gemini all of them it writes what looks
1:08:39 like
1:08:41 dialogue and then you read it and it's
1:08:43 like no this like this wouldn't hold up
1:08:45 in a
1:08:47 [Laughter]
1:08:52 porno this is like not only writing good
1:08:54 dialogue but it's writing good screen
1:08:56 directions and good audio
1:09:03 cues and it's remembering the context of
1:09:07 the whole conversation which I find cuz
1:09:10 from what I understand 01 preview only
1:09:13 has a
1:09:14 32,000 token context window is decent
1:09:18 but not a not a ton it's about 25,000
1:09:22 words so maybe half a novel when was the
1:09:25 last time you check those other models
1:09:27 it it it's been it's been a while P but
1:09:29 but
1:09:32 um well that's not true I mean I tested
1:09:35 GPT
1:09:37 40 recently like within 3 months I've
1:09:42 done something like this within 3 months
1:09:44 and it was
1:09:49 crappy so the fact that this is
1:09:53 maintaining
1:09:56 continuity is is pretty mindblowing I
1:09:59 mean that said Gemini because Gemini has
1:10:02 a 2 million token context window you
1:10:05 could probably do a fair amount of this
1:10:07 in Gemini it' be this would be an
1:10:09 interesting
1:10:10 exercise do do this exact take take the
1:10:14 exact prompt out of this put it in
1:10:16 Gemini and see how it does that would
1:10:18 that would actually be an interesting um
1:10:21 comparison I created a GPT refresher off
1:10:24 01 at the end and will help you post the
1:10:29 fixed in X in Irregulars okay I don't
1:10:32 know what that means a GPT
1:10:35 refresher Oh by written by 01 got it
1:10:41 okay
1:10:46 cool I hit my cap in like an hour with
1:10:48 01 yeah I know that's that's a problem
1:11:00 but I think what I'm going to
1:11:04 do I think I'm going to finish the
1:11:07 screenplay I think I wait sorry let me
1:11:10 rephrase that I think I'm going to let
1:11:11 chat GPT finish this
1:11:14 [Laughter]
1:11:16 screenplay oh man why Try things in this
1:11:20 everchanging AI landscape instead let's
1:11:22 do blind brand loyalty
1:11:25 no I'm gonna I'm gonna do it Pate I'm
1:11:27 gonna play I'm gonna play with Gemini I
1:11:29 I don't want to do I don't want to do it
1:11:30 right now just because it's gonna I've
1:11:32 I've spent too much time on this already
1:11:34 Kyle I bet agents will dial up those
1:11:37 mini models for agents I bet they save
1:11:39 money that way yeah that probably makes
1:11:48 sense who gets credit I [ __ ] get
1:11:52 [Laughter]
1:11:54 credit I mean here's the here's my
1:11:56 philosophy on
1:11:59 that this is a this is a screen this is
1:12:01 this story is my story while it was
1:12:06 incomplete it was my story I'm the one
1:12:10 who's ultimately going to edit the
1:12:12 outline and say do I want these scenes
1:12:14 in this order or not I'm the one who's
1:12:17 ultimately going to say does this sound
1:12:19 like my dialogue or
1:12:21 not and I'm going to put the screenplay
1:12:24 out there under my name
1:12:26 so if there's anything in it that's
1:12:29 shitty that I was just lazy and I put in
1:12:32 shitty work well that's on
1:12:36 me it is no one's business in my humbl
1:12:40 opinion in my humble opinion people it
1:12:44 is no one's business what tool I used to
1:12:47 produce whatever piece of content I
1:12:50 choose to put in the
1:12:51 world I don't have to disclose disclose
1:12:54 that I used micros oft word or an AI
1:12:57 spell checker or grammarly to fix my
1:13:02 grammar it's no one's
1:13:05 business what's their business is is the
1:13:08 work that he produced shitty or not
1:13:12 that's their business and if it's shitty
1:13:16 that's on me my name's on
1:13:20 it co-written with Ai No you don't get
1:13:24 to know what tools I used
1:13:28 when I can't wait until AI creates
1:13:30 agents better than it and and creates
1:13:33 agents better than them Joker I mean
1:13:35 that that's the thing where I'm a little
1:13:37 this is where I'm a
1:13:40 little
1:13:42 uh mind blown right now
1:13:45 because this is the first time that I've
1:13:48 had one of these
1:13:52 things write something so here's a good
1:13:55 way to think about it this this goes
1:13:57 back to who gets credit for it I get
1:13:58 credit for it this is the first time
1:14:00 I've had one of these things write
1:14:02 something based on something I've
1:14:05 written that actually feels like
1:14:07 something I would
1:14:09 write
1:14:13 because as I was
1:14:15 reading the scene in the car where the
1:14:19 mom was checking on him did he touch you
1:14:22 I was thinking like did I write that
1:14:23 scene
1:14:26 and it and it just recreated what I
1:14:29 wrote I know I didn't write the other
1:14:32 scene with Sam falling out of the tree I
1:14:34 know I didn't write that one and as I
1:14:37 read it I'm like yeah this is like this
1:14:40 feels like something I would
1:14:43 write so I'm sitting here thinking
1:14:49 like imagine when this tool gets
1:14:53 good this is the pre you
1:14:58 right don't people have ghost riters all
1:15:00 the time for their work yeah they have
1:15:02 ghost riters they've
1:15:05 yes
1:15:08 yes sometimes those Ghost Riders get
1:15:11 credit most times they
1:15:14 don't go look up Ghost Rider contracts
1:15:18 some of the Ghost Rider contracts are
1:15:20 just essentially
1:15:22 ndas you're going to write my [ __ ] and
1:15:24 you can't talk about
1:15:26 it ghost writers multiple
1:15:29 editors yeah this is you know what I
1:15:32 have here I have a ghost writer and an
1:15:34 editor and in the end I will be the one
1:15:37 that says that's a screenplay I'm you
1:15:40 know willing to put my name
1:15:43 on we'll run it through an AI you know
1:15:46 content Checker okay so what this is a
1:15:50 screenplay it's not a [ __ ] novel and
1:15:51 even if it was a novel is it compelling
1:15:56 oh let's do
1:16:00 this take a slight
1:16:04 detour and I want you to
1:16:09 write the opening
1:16:13 chapter of a novel based on this
1:16:17 screenplay
1:16:23 period retain the same
1:16:25 same
1:16:27 tone and sense of humor in the
1:16:37 dialogue but use sophisticated
1:16:47 literary
1:16:49 approaches to the writing period
1:16:59 translating screen to page crafting the
1:17:02 opening how shifting ter okay let's see
1:17:05 how let's see how good or bad it
1:17:09 is the afternoon sun cast a golden hue
1:17:13 over the Suburban Tableau Illuminating
1:17:16 the neat rows of identical houses that
1:17:18 stretched as far as the eye can see um
1:17:26 oh wow it's still going here I'm going
1:17:27 to stop
1:17:34 it what you failed to capture here was
1:17:37 that the first line of the novel should
1:17:40 be
1:17:41 memorable
1:17:47 and
1:17:49 concise and capture the essence of the
1:17:51 entire story
1:17:54 period please please think
1:17:56 about the top 20 opening lines of novels
1:18:01 throughout history and make sure that
1:18:03 the opening line of this
1:18:05 novel is in the same
1:18:14 neighborhood we're not going to have
1:18:16 some shitty opening line for our novel
1:18:30 every photograph captures a moment but
1:18:32 some moments capture you forever eh not
1:18:35 bad the sun hung low in the sky casting
1:18:38 Shadows over the neat Labyrinth of
1:18:40 Willow Creek Estates a suburb so
1:18:42 pristine it seemed almost rehearsed
1:18:45 that's great identical houses line The
1:18:48 Identical streets with identical trees
1:18:52 standing Sentinel their leaves with
1:18:55 Whispering secrets to anyone who cared
1:18:57 to listen [ __ ] that's
1:18:59 good but amidst the uniform wait let me
1:19:02 read this again identical houses lined
1:19:04 identical street with identical trees
1:19:07 standing Sentinel their leaves
1:19:09 Whispering secrets to anyone who cared
1:19:12 to
1:19:13 listen
1:19:15 [ __ ] dag boy dag girl [ __ ] kidding
1:19:22 me oh damn it
1:19:25 stupid Tick Tock
1:19:31 Marge all
1:19:34 right that's really good we can't see
1:19:36 you I'm back I'm back it's
1:19:38 okay all right but amidst the uniformity
1:19:41 at the end of Maple Leaf Court a moving
1:19:43 truck disrupted the Symmetry a ripple in
1:19:46 the otherwise Still Pond of suburban
1:19:50 Perfection Jesus this is good but amidst
1:19:53 the uniformity at the end of Maple Leaf
1:19:56 Court right just Maple Leaf Court paints
1:19:59 a picture in and of itself a moving
1:20:02 truck disrupted the Symmetry a ripple in
1:20:06 the otherwise Still Pond of suburban
1:20:09 Perfection great Jack Whitmore Jr
1:20:12 perched at top his modified B bicycle a
1:20:14 Contraption of Gears levers and a
1:20:16 makeshift photo printer surveyed his new
1:20:19 kingdom with a mix of trepidation and
1:20:21 curiosity the bicycle a testament to his
1:20:24 12-year-old genuity words Softly As the
1:20:28 as the as he pedal in lazy circles
1:20:30 around the culdesac his camera hung from
1:20:33 his neck like a trusted Confidant the
1:20:36 lens reflecting the world even as it
1:20:38 concealed its own uncertainties Jackie
1:20:41 watch out his sister Becky his sister
1:20:43 Becky's voice sliced through the Ry as
1:20:46 he swerved to avoid a collision she
1:20:49 stood on the curb one hand clutching her
1:20:50 everpresent
1:20:52 smartphone the other shading her eyes as
1:20:54 she glared at him
1:20:57 seriously s seriously do you have to be
1:20:59 such a dweeb she huffed her 14-year-old
1:21:02 exasperation on full display he grinned
1:21:04 mischievously dweebs invent the invent
1:21:07 the future you know well you can Invent
1:21:09 Yourself out of my person could you
1:21:11 invent yourself out of my personal space
1:21:13 she rolled her eyes before returning her
1:21:15 attention to the screen thumbs flying as
1:21:18 she resumed texting her friends back in
1:21:20 the
1:21:21 city across the lawn their parents
1:21:23 oversaw the armor of professional
1:21:26 movers really good really
1:21:29 good
1:21:33 wow I I
1:21:36 this the sky hung so every photograph
1:21:40 captures a moment but some moments
1:21:42 capture you
1:21:48 forever that should be something more
1:21:50 like some moments capture you back
1:21:53 whatever all right the hung
1:21:55 the the sun hung low in the sky casting
1:21:58 Long Shadows over the neat Labyrinth of
1:22:01 Willow Creek Estates a suburb so
1:22:03 pristine it seemed almost rehearsed
1:22:07 identical houses lined identical streets
1:22:10 with identical trees standing Sentinel
1:22:12 their leaves Whispering secrets to
1:22:14 anyone who cared to
1:22:17 listen [ __ ] that's
1:22:20 good but amidst the uniformity at the
1:22:22 end of Maple Leaf Court a mov moving
1:22:25 truck disrupted symmetry a ripple of in
1:22:28 the otherwise Still Pond of suburban
1:22:37 Perfection sorry these things are
1:22:40 creative
1:22:41 now sorry
1:23:00 scroll me to the bottom
1:23:03 March sounds like Stephen King wrote
1:23:10 that yeah
1:23:13 wild wild wild
1:23:16 wild
1:23:22 um so let me show you something if
1:23:25 you're new
1:23:26 here
1:23:29 um what I'm doing here is sort of what
1:23:31 we do here I'm diving a bit deeper
1:23:34 here's the here's part of the problem
1:23:37 with testing this new 01 thing
1:23:41 is I don't really know how to test
1:23:45 it
1:23:49 um my recommendation would be for you to
1:23:53 test it knowing things you know
1:23:55 I've written a bunch of screenplays so I
1:23:57 know screenwriting well enough to know
1:24:00 if it's good or not right like so I'm
1:24:03 using that but if you're if you're good
1:24:05 at project management test 01 around
1:24:09 bigger Concepts around project
1:24:11 management or if you're opening a
1:24:12 restaurant do something around that you
1:24:14 know some some bigger problem solving
1:24:17 thing within some domain that you
1:24:19 know and then that way you can get a
1:24:22 sense of is it doing something qualita
1:24:25 ly different than chat gp4 or Claude or
1:24:28 gemini or things like
1:24:30 that
1:24:34 um so I'm I'm only looking at it through
1:24:37 one super thin lens and I'm basically
1:24:40 this is my first prompt my prompt Series
1:24:44 right this is my first chat with this
1:24:47 new model that's that's of any substance
1:24:51 um and there's a part of this that's
1:24:53 really boring
1:24:55 because it generates a lot of words and
1:24:58 then I got to go look at the words and
1:25:00 say do these words does it make good
1:25:03 words and my initial take is holy
1:25:08 [ __ ]
1:25:09 [ __ ]
1:25:11 yes like like in a in a in a
1:25:15 surprisingly good way
1:25:29 like R now I can ask it about that
1:25:34 one well Joker 01 isn't connected to the
1:25:37 internet yet so any anything with modern
1:25:39 events will
1:25:42 not um it may answer the question but
1:25:45 it'll it'll just hallucinate it um GPT
1:25:49 40 is connected to the internet but 01
1:25:52 previews not yet
1:26:04 oh you're touting your rag ah
1:26:07 okay it can solve math riddles other
1:26:10 models can't yep I heard that again I
1:26:13 don't know math riddles so let me see if
1:26:17 it can create a
1:26:18 business let's see it create a business
1:26:21 and see what happens okay we can do that
1:26:24 we tested it with a food scientist
1:26:26 earlier and compared it to 4.0 more
1:26:29 technically on than
1:26:31 4.0 okay that's
1:26:33 interesting so let's let's start a new a
1:26:36 new uh
1:26:41 prompt and then we'll
1:26:44 say oh you know what before we jump into
1:26:46 the business thing because that's going
1:26:47 to eat up some time let me let me go
1:26:49 point you over to something
1:26:56 going to do two things
1:26:59 here so the first thing is go to the
1:27:03 salon. so if you're new
1:27:06 here the salon is a group that I started
1:27:09 the week after chat GPT came out and
1:27:11 it's a
1:27:13 community
1:27:14 of AI optimists and people that are
1:27:17 curious about AI exploring AI willing to
1:27:21 hang out on a stupid Tik Tok Channel
1:27:22 with a guy you know rambling
1:27:25 incoherently about a
1:27:27 screenplay
1:27:28 and it's a it's a really remarkable
1:27:31 community so we've got a um a meeting
1:27:33 tomorrow night called called A Salon
1:27:36 meet and greet where you get to
1:27:38 introduce yourself and meet other people
1:27:40 within the community um we also have
1:27:42 done some nice improvements to the
1:27:44 community so if you go to the salon
1:27:45 that's our website there's a button
1:27:47 there that says join our community so
1:27:49 click on that and join our
1:27:52 community um
1:27:59 [Music]
1:28:04 where is
1:28:09 it that's insights that's not what I
1:28:12 want
1:28:18 on so it's going to drop you
1:28:21 into this page welcome to the salon
1:28:25 and then it'll talk about our journey to
1:28:28 AI literacy play first mindfully create
1:28:30 generously lead and over on the Le hand
1:28:33 side here it says start your adventure
1:28:35 there's a seven-step onboarding process
1:28:38 the first one is check out the welcome
1:28:39 page second one is introduce yourself
1:28:42 learn about our values Watch past
1:28:44 meetings RSVP for the next event so the
1:28:46 next event is actually uh tomorrow that
1:28:51 one the third Tuesday of the month the
1:28:53 salon meet greet and so you can go um
1:28:58 RSVP for that uh then you can hang out
1:29:01 and chat in the water cooler water
1:29:03 cooler chat and then help us grow is
1:29:05 there's our socials and things like that
1:29:07 you can talk about it and invite your
1:29:08 friends and things like that so please
1:29:09 do that
1:29:11 um but we also we just reorganized
1:29:16 something that I'm really excited about
1:29:18 which
1:29:19 is I used to maintain a page called play
1:29:22 first and AI tools which was just a big
1:29:25 long page of tools that initially
1:29:29 started as three tools there were three
1:29:31 things that I thought you should pay
1:29:32 attention to and then it was five and
1:29:34 then it was seven and then it was 20 and
1:29:38 now it's like too many to keep track
1:29:41 of so we had that and then we also had a
1:29:45 showand tell area where people would go
1:29:47 in and talk about their
1:29:51 um you know show images that they' made
1:29:54 or suo songs that they made and we had a
1:29:57 an image area and a music area and a GPT
1:30:00 and apps area things like that so what
1:30:03 we've done is we actually combined those
1:30:05 two concepts so there's now an area
1:30:07 called tools and play so the tools are
1:30:10 where you you know uh you know what you
1:30:14 click on where you go do it and then the
1:30:17 play is like talk about it and share
1:30:20 your work and things like that so
1:30:21 there's a start here section which is
1:30:24 you know just talking philosophically
1:30:26 about how to how to use the tools and
1:30:28 how to how to do these things and then
1:30:31 there's llm tools right so um in every
1:30:35 one of these sections there's a tool tab
1:30:39 a tool talk chat area and then a showand
1:30:42 tell right so the the the default tab is
1:30:46 here's a bunch of llms to go play with
1:30:49 and then tool talk is you can come in
1:30:51 here and chat about what's going on
1:30:53 which particular L with 01 talking about
1:30:56 what's working what's not working is
1:30:58 pretty interesting and then show and
1:30:59 tell we've got nothing there right now
1:31:02 so what might you show show and tell
1:31:04 with llms it could be a GPT it could be
1:31:08 um a chat that you did that's
1:31:10 particularly interesting with 01 that
1:31:12 maybe you share with someone else um in
1:31:15 the image section it it's again the same
1:31:18 thing where you've got all your tools
1:31:20 here and then you've got tool talk where
1:31:22 you can chat about stuff this is just
1:31:24 getting started and then you have your
1:31:25 show and tell where you can show off
1:31:27 your cool pretty images so all of the
1:31:29 tools in Play Follow That same formula
1:31:32 so this little section here is
1:31:34 incredibly valuable if you're just
1:31:36 getting
1:31:37 started so if you have not join the
1:31:41 salon what are you waiting
1:31:43 for all right I
1:31:54 [Music]
1:31:55 wow that is only about 18 pages of
1:31:57 written text is that correct I don't
1:31:59 know what you're talking
1:32:06 [Music]
1:32:11 about we've made improvements to the
1:32:13 community wait why can't I log in
1:32:14 anymore is that true Pate you can't log
1:32:18 in you should be able to we haven't we
1:32:20 haven't changed anything login wise 40
1:32:24 could Solve
1:32:29 IT 32k context window oh in 01 equals 18
1:32:35 Pages he's teasing you okay um yeah so
1:32:40 Darlene whatever your name is Darlene
1:32:43 paint made a
1:32:45 joke
1:32:47 Darlene or Becker something like that
1:32:51 Rebecca Darin Rebecca
1:32:57 um yeah
1:32:59 it's it's possible that it's 32,000
1:33:03 tokens output but the way chat GPT used
1:33:07 to work is whatever the token count was
1:33:10 it was the combination of input and
1:33:12 output it seems to be remembering more
1:33:15 than like I I'm less frustrated doing
1:33:18 this screenwriting exercise than I've
1:33:20 ever been cuz it would always sort of
1:33:22 forget what we talked about before it
1:33:23 doesn't do seem to be doing that right
1:33:25 now but I don't know anyway um all right
1:33:29 my back is starting to get wonky so so
1:33:32 let me let me get out of here let me get
1:33:35 let me get out of here people um I got
1:33:38 to go rest my back I I put my back
1:33:43 out 10 days ago and it's sort of
1:33:49 better but you know when when backs go
1:33:51 out they [ __ ] suck
1:34:01 um
1:34:03 yeah when I get quiet like
1:34:07 this quiet he hasn't shut the [ __ ] up
1:34:11 what does he mean quiet I don't I don't
1:34:15 understand quiet it's me Harry
1:34:17 Carey um heat pad yeah I know I need a
1:34:20 heat pad
1:34:22 actually someone posted that in regulars
1:34:25 take care of the wonkiness yeah I got a
1:34:26 janky back right
1:34:28 now when I get when I get quiet like
1:34:30 this like I I feel like I'm there's a
1:34:32 lot of pauses tonight where I'm just
1:34:34 kind of lost in thought
1:34:40 um I'm I'm kind of trying to
1:34:49 process how much better this tool is at
1:34:51 writing
1:34:54 like it seems scarily
1:35:00 good that [ __ ] opening paragraph has
1:35:03 me
1:35:29 the sun hung low in the sky casting Long
1:35:31 Shadows over the neat Labyrinth of
1:35:34 Willow Creek Estates a suburb so
1:35:37 pristine it seemed almost
1:35:40 rehearsed that's a good [ __ ]
1:35:44 line identical houses lined identical
1:35:47 streets with identical trees standing
1:35:50 Sentinel their leaves Whispering secrets
1:35:53 to anyone who cared to
1:35:56 listen the secrets of the
1:35:59 suburbs [ __ ]
1:36:01 crazy but amidst the uniformity at the
1:36:04 end of Maple Leaf Court a moving truck
1:36:07 truck disrupted the Symmetry a ripple in
1:36:10 the otherwise Still Pond of suburban
1:36:17 Perfection I ain't I ain't never I I
1:36:20 have
1:36:21 had large language models right [ __ ]
1:36:24 that I'm
1:36:25 like it's like a good paragraph it's
1:36:29 it's how a good paragraph would be
1:36:31 structured if it didn't
1:36:35 suck I've never had it write something
1:36:38 where I'm
1:36:39 like I don't think there's much I would
1:36:42 change in that
1:36:56 Jack Whitmore Jr perched a top his
1:36:58 modified bicycle a Contraption of Gears
1:37:02 levers and a makeshift photo printer
1:37:05 surveyed his new kingdom with a mix of
1:37:08 trepidation and curiosity the bicycle a
1:37:12 testament to his 12-year-old
1:37:15 Ingenuity like it didn't say Jack
1:37:18 Whitmore a
1:37:20 12-year-old it said Jack's riding this
1:37:23 crazy ass
1:37:25 bike and then it says the bicycle a
1:37:27 testament to his 12-year-old Ingenuity
1:37:30 word Softly As He pedal around lazy
1:37:32 circles in the
1:37:35 culdesac I don't know not to be a drag
1:37:38 but what if that opening paragraph is
1:37:40 plagiarized how would you know it's not
1:37:42 it's not how these things
1:37:44 work it's not how they work it's so one
1:37:48 of the misconceptions about large
1:37:50 language models and image generation
1:37:52 tools is that They're copying and
1:37:53 pasting they're not they're
1:37:56 generating they're generating tokens
1:37:59 from a mathematical a thousand
1:38:02 dimensional mathematical
1:38:04 space Here's the thing about the way
1:38:06 embeddings work is that the original
1:38:09 documents cease to exist they they're
1:38:13 not it's not taking the whole article of
1:38:16 the New York Times or the whole opening
1:38:19 passage of a book and putting it in
1:38:21 there it's taking the words and putting
1:38:24 them in in semantic clouds if you prompt
1:38:27 it cleverly enough you can reconstruct
1:38:31 reconstruct
1:38:33 stuff which reinforces the
1:38:36 misunderstanding of how this [ __ ]
1:38:39 works but it's not how it works it's not
1:38:42 copying and pasting it's generating
1:38:44 little tokens all one after the other or
1:38:48 you know might be little groupings of
1:38:49 them but
1:39:00 and
1:39:03 it's it's nuanced but it's really
1:39:07 important I mean it's not this similar
1:39:09 to this like Tom Robbins is one of my
1:39:12 favorite novelists he's just got really
1:39:15 clever wayy
1:39:19 rights if I spent the
1:39:22 next you know 6 months just reading all
1:39:25 of his novels all of his books and
1:39:28 just read analysis on his work and just
1:39:33 really fanboyed out on how he does his
1:39:38 craft and then at the end of that six
1:39:40 months I
1:39:42 said I want to write a
1:39:46 book the chances of me lifting something
1:39:50 from one of his novels is probably
1:39:52 decently High
1:39:54 because I've got all these Concepts in
1:39:56 my head they're all sort of front and
1:39:58 center and
1:40:00 so you can
1:40:03 absolutely Prime the latent space right
1:40:06 you can prompt your way into a corner of
1:40:09 the latent space where you're going to
1:40:11 more high high high have a very high
1:40:13 likelihood of capturing you know exact
1:40:17 passages from something but in the case
1:40:20 of this it's basic this is right this is
1:40:23 not
1:40:24 I didn't tell it to follow Hemingway or
1:40:27 anything like that I said follow my
1:40:29 screenplay like follow the tone of my
1:40:31 screenplay and just write a thing so
1:40:34 it's not like I think the the the
1:40:37 chances it lifted any of that are very
1:40:39 very low I mean it would be interesting
1:40:41 actually you know what let's go
1:40:43 Google identical houses lined identical
1:40:46 streets with identical trees standing
1:40:48 Sentinel let's let's grab that sentence
1:40:51 cuz that one feels to me
1:40:55 pretty [ __ ] have pulled it from
1:40:57 something else so I'm just going to go
1:40:59 to Google and put it in
1:41:01 quotes and see if it finds
1:41:04 anything it looks like there aren't any
1:41:07 great matches for your
1:41:16 search the streets lined with Elm trees
1:41:19 the pretty houses with a
1:41:21 banister no didn't didn't find didn't
1:41:24 find an exact match
1:41:30 anyway identical houses lined identical
1:41:33 streets with identical trees standing
1:41:35 Sentinel there leaves Whispering secrets
1:41:38 to anyone who cared to
1:41:40 listen which in the suburbs is [ __ ]
1:41:42 no one it's so good but amidst the
1:41:46 uniformity at the end of Maple Leaf
1:41:48 Court a moving truck disrupted the
1:41:51 Symmetry a ripple in the otherwise Still
1:41:54 Pond of suburban
1:41:56 Perfection [ __ ]
1:42:01 amazing there's a plagiarism chck or I
1:42:03 forget what it was called but I don't I
1:42:06 don't need a plagiarism chick it's not
1:42:08 these things are not plagiarizing it's
1:42:10 not how they work it's not how they
1:42:12 work I I could get it to plagiarize if I
1:42:16 prompted for it which is what all of the
1:42:18 all of the lawsuits right now what they
1:42:20 do is they prompt it so that it gives
1:42:22 them something that is you know as close
1:42:26 to an exact passage as possible or or an
1:42:28 exact
1:42:29 passage but how it's being generated is
1:42:33 is
1:42:34 not Kyle's back is having meltdown
1:42:37 Monday it is it is anyway let me get out
1:42:40 of here um I'm not [ __ ] worried about
1:42:44 that that's a that's a really good
1:42:46 opening paragraph I got to fixed the the
1:42:48 opening line is not good every
1:42:51 photograph captures a moment but some
1:42:54 moments and what it wrote was capture
1:42:56 you forever you know what I'm going to I
1:42:58 can't live with that
1:43:03 um that opening sentence is
1:43:07 close but the punch line to the sentence
1:43:11 feels weak to me give me 20
1:43:15 options that start the
1:43:20 same but more accurately capture the
1:43:23 essence of the story for the
1:43:27 punchline we'll see if it even knows
1:43:30 what the [ __ ] I'm talking
1:43:33 about crafting impactful
1:43:37 openings crafting Vivid openings doing
1:43:40 something with punchlines every
1:43:42 photograph captures a moment but some
1:43:44 moments capture a glimpse of the future
1:43:46 horrible every photograph captures a
1:43:48 moment yet some moments reach out to
1:43:50 rewrite Destiny that's better cuz that's
1:43:52 what the story is about
1:43:54 every photograph captures a moment but
1:43:56 when a moment captures you it can change
1:43:59 everything there you go that's a good
1:44:02 that's a good one every photograph
1:44:04 captures a moment but some moments
1:44:05 unveil Secrets best left hidden that's
1:44:09 shitty every uh but this one this one
1:44:12 revealed a fate I couldn't escape well
1:44:15 that's this one's interesting because it
1:44:18 changes um it changes the it puts the
1:44:21 novel in first person
1:44:24 some moments Echo the warnings from
1:44:26 Beyond some moments blur the lines
1:44:28 between now and what lies
1:44:31 ahead little on the nose when moments
1:44:33 predict the future yeah moments bind you
1:44:37 to a destiny not of your choosing yeah
1:44:39 these are
1:44:40 shitty all right so so there was a good
1:44:43 one there
1:44:45 uh but when a moment captures you it can
1:44:48 change
1:44:51 everything and maybe it should be when
1:44:53 the the photo when a photo captures you
1:44:56 photo oh that's it every photograph
1:44:59 captures a
1:45:01 moment but when a photograph captures
1:45:05 you it can change everything so anyway
1:45:08 so that's fixable all
1:45:13 right add OC CD got to fix the sentence
1:45:17 got to fix the sentence
1:45:20 [Laughter]
1:45:24 all right I'm out of here people you
1:45:26 [ __ ] weirdos we're not weird we're
1:45:31 regular Mister don't you get all getting
1:45:34 all high and mighty up there on your Tik
1:45:36 Tock we're not
1:45:38 weirdos we're just
1:45:42 irregular so if you want to be irregular
1:45:45 if you're new here and you're like why
1:45:46 am I still watching
1:45:48 that that's the question that get asked
1:45:50 here
1:45:52 nightly by everyone who
1:45:55 attends just keep coming back and you'll
1:45:58 be irregular by
1:46:00 definition you can start hanging out go
1:46:03 join the
1:46:04 salon you want to support the channel
1:46:06 subscribe to the lives that supports me
1:46:09 directly next time do more writing
1:46:12 please this was really enjoyable okay
1:46:14 great I'll yeah I'll I I think I may I
1:46:17 think I might just continue working on
1:46:19 the the
1:46:21 screenplay I think the homework I need
1:46:24 to do is I need to pull all the [ __ ]
1:46:27 here's one of the problems with
1:46:29 generative AI it generates so much
1:46:31 content that it gets very overwhelming
1:46:34 it's hard to sort of figure out what
1:46:35 your where everything is so I think the
1:46:39 thing I need to do is I need to pull all
1:46:40 that out into a like a Word document um
1:46:44 and then as it writes scene start start
1:46:47 populating that outline and see if
1:46:49 there's any holes in it and like it
1:46:52 hasn't written a scene yet that I would
1:46:53 change all that much which is
1:46:57 weird I'm flued I'm I'm I'm a little
1:47:01 discombobulated right now it shouldn't
1:47:03 be this good and it
1:47:09 is what do you do with
1:47:12 that well I'll tell you what I [ __ ]
1:47:14 do I'm going to finish a bunch of
1:47:16 fragmented projects that I
1:47:19 started I mean how many how many [ __ ]
1:47:23 things have you started in your life
1:47:26 that you're still hanging on to it like
1:47:28 I'll get to it
1:47:30 someday well what
1:47:32 if you could actually just finish it
1:47:34 like in a couple of
1:47:38 days why not champ what is your whining
1:47:43 this you want cheese
1:47:46 what get
1:47:49 out [ __ ] fatty
1:47:53 blessed be the
1:47:58 [Music]
1:47:59 cheesers look at my
1:48:05 hair all right like a
1:48:09 career hopefully it doesn't flake out I
1:48:12 know that's that's that's maybe it's why
1:48:15 it's it's worth us digging on this
1:48:18 screenplay a bit more because it's going
1:48:20 to be a fascinating exercise to see if
1:48:23 it can actually maintain continuity as
1:48:26 we move forward or does it lose its
1:48:29 memory cuz if it doesn't lose its memory
1:48:32 this starts to be a very different kind
1:48:33 of
1:48:38 tool yeah crazy all right [ __ ] is
1:48:42 trying to warn you about your back yeah
1:48:43 I know I got to get out of here what are
1:48:46 you doing for the back um just sucking
1:48:48 it up right now um all right I'm GNA get
1:48:51 out of here um all right little uh yeah
1:48:54 stretch it that's exactly what I'm going
1:48:56 to do right now because I'm just sitting
1:48:58 here like in you know sort of tense mode
1:49:01 so I got to I got to go stretch it um
1:49:05 it's cool to see your process thank you
1:49:06 thank you Kim katkin my rag doesn't lose
1:49:09 its memory yep very cool foam roller I
1:49:12 got to be careful with the foam roller
1:49:13 cuz it's my it's the like the little
1:49:15 Bend in my lower back like one of the
1:49:17 vertebra out so if I Jank it too much
1:49:21 it'll pinch a nerve and I don't want to
1:49:22 do that
1:49:23 um all right so let's uh let's Romper
1:49:26 Room magic mirror it we got Cara 2.0 in
1:49:28 a house
1:49:30 cuno Cavo in the house good to see you
1:49:32 Joy pie steo peace out to you from Down
1:49:35 Under Teton skier Be not Afraid Danielle
1:49:38 thank you Cam katkin from
1:49:41 Cleveland um Joey in the house Joker
1:49:45 Brandon thank you so much for the uh for
1:49:48 the thread that helped a lot yeah we'll
1:49:50 play with that stuff tomorrow whatever I
1:49:52 didn't get to tomorrow
1:49:54 H stop doing
1:50:01 that all
1:50:05 right oh meet and greet tomorrow so go
1:50:08 to the
1:50:10 salon. in our little welcome seven-step
1:50:14 thing one of them is RSVP for events
1:50:16 tomorrow night is the salon meet and
1:50:19 greet come meet us turn on your camera
1:50:22 we we give you little questions to
1:50:24 answer you don't have to go first but
1:50:27 you do have to
1:50:29 go introduce yourself get in community
1:50:32 get in a community of AI Optimist it
1:50:34 sucks less than talking to your you know
1:50:38 co-worker Sally who resents the fact
1:50:40 that you talk about AI too much AI
1:50:42 learning lab probably delayed like Oprah
1:50:45 maybe uh we'll we'll see how I feel
1:50:46 tomorrow it might be a little later
1:50:48 tomorrow night um Amelio's wife in the
1:50:51 house AI Bob great to see see you Dr J
1:50:54 thank you as
1:50:55 always
1:51:01 uh who else we got who else am I missing
1:51:04 here two block
1:51:06 Tom Rachel Sparkles Winston Hardy great
1:51:09 to see you thanks as always for
1:51:11 everything um oh by the
1:51:15 way huge so do me a favor like do like
1:51:19 extra if you can do extra like gifts or
1:51:22 whatever digital Gods Lord digital
1:51:26 gods made the YouTube channel of all of
1:51:29 these videos he basically got all the
1:51:32 descriptions for the past I was like N9
1:51:34 months behind in the descriptions of
1:51:37 what all these videos were about so if
1:51:40 you go to YouTube AI learninglab dtt
1:51:44 for- Tik
1:51:46 Tock um they all have descriptions now
1:51:49 so they're like searchable and findable
1:51:51 and so thank you to to digital Gods
1:51:55 who's just he and he basically just did
1:51:57 that because he's like he's like I just
1:51:59 wanted to do it I think it's important
1:52:01 content and I wanted to get it done so
1:52:03 holy [ __ ] thank you so kudos to kudos to
1:52:07 digital Gods so thank you all right um
1:52:10 that's it I'm G to get out of
1:52:12 here um thank you Winston really
1:52:14 appreciate that very very
1:52:17 much and silver fox in the
1:52:21 house Bob Bob number four does that say
1:52:25 oh Bob H Bob number four like Bob number
1:52:31 four all right I'm out everybody I'll
1:52:33 see you tomorrow come to the meet and
1:52:35 greet