10/14/24 - AI & The "Holy Crap-ification" of Work: Mind-Blowing AI Film Project Reveal!

Description

Chapters

Transcript

0:00 [Music]
0:17 [Applause]
0:18 [Music]
0:20 oh
0:22 [Music]
0:31 when I find myself in times of trouble
0:35 Mother Mar comes to
0:37 me speaking words of
0:41 wisdom let it
0:44 be in my hour of
0:48 [Music]
0:49 Darkness she is standing right in front
0:52 of
0:54 me speaking words of wisdom with me
1:01 let it
1:03 be let it
1:06 be oh let it
1:09 be whisper words of wisdom let it
1:13 [Music]
1:28 be uh oh
1:32 [Music]
2:24 look got you in his sight
2:28 [Music]
2:30 what you doing with clown like
2:36 me one of
2:39 life's
2:41 mystery so T I'll
2:46 ask the stars
2:50 above how did I ever
2:53 [Music]
2:56 win what did I do
3:00 what did I
3:01 [Music]
3:03 say to turn your
3:06 Ranger eyes my
3:10 [Music]
3:17 way happy Monday good people there's
3:19 only 11 people in here we can't we can't
3:22 we can't start with 11 people in here
3:23 are you kidding me get tapping on that
3:25 screen go wake the grandkids go wake the
3:28 grandkids
3:38 what is it Grandpa I'm tired tap the
3:40 screen Sally I don't want to tap the
3:43 screen grandpa just tap the
3:46 screen Kyle needs his lights okay my
3:50 finger
3:52 hurts it builds
3:56 character forever hook shared it with
3:58 like 12 people they're all your people
4:01 Everyone you shared it with is
4:07 here it's probably just Tik Tok not
4:10 catching up or you know nobody's
4:12 watching
4:13 [Laughter]
4:15 anymore there's always that
4:17 possibility um what was I going to say
4:21 thank you Danielle very nice
4:23 uh some interesting stuff going on to
4:26 talk about
4:28 [Music]
4:37 [Music]
4:43 [Music]
4:45 what place I can rest my poor
4:49 [Music]
4:52 head my thought sweet
4:58 silence w
5:01 is this place where the feelings AR
5:05 [Applause]
5:07 dick from an over exposure to
5:11 violence this is the place I can slowly
5:15 face the only one I truly can
5:20 know these are tears from a long time
5:24 ago got these tears from a long time ago
5:29 I to cry 30 years or
5:35 so these are tears from a
5:38 [Music]
5:42 l
5:45 [Music]
5:49 go well oh darling Oh Darling you say
5:53 unto
5:54 [Music]
5:56 me where have you been all my life five
6:01 [Music]
6:04 time have been swimming the seven sad
6:08 [Music]
6:10 Seas poor women have toss me their
6:15 lifelines pull them into the waters
6:18 darker in I'd have War them but I didn't
6:23 know these are tears from a long time
6:27 ago got these tears from a long time ago
6:31 I mean to cry 30 years or
6:35 so these are tears from a long
6:44 [Music]
6:56 time that was a pathetic ending what was
6:59 the little whimper at the end you got to
7:01 you got
7:03 to do I need a
7:05 baton good Lord conducting dogs is
7:09 [Music]
7:16 rough your Hound is so cute champy is
7:19 very cute his name's champy he kind of
7:21 looks like this except he doesn't have
7:24 black eyes he is it's he's got white the
7:26 black doesn't start till his ears so
7:28 he's got white eyes so it's close Vicky
7:30 made this Vicky made this cute
7:34 little fidget doggy that's very champ
7:37 likee proportionally this is very
7:40 champ although Champ's back is flatter
7:43 you could champ could walk around with
7:46 eight wine glasses on his back and not
7:48 spill a drop it's perfectly flat and
7:51 then it just goes down he's like a
7:53 coffee
7:57 table I hear the golden golden retriever
8:00 conducts better probably oh champy I
8:02 know people are making fun of you champy
8:04 I wouldn't put up with it
8:13 [Music]
8:45 all
8:47 right hope everybody had a good
8:50 weekend if anybody did any uh fun AI on
8:54 over the weekend let me know I actually
8:55 just wrote I got inspired to write a
8:57 piece at like 4:30 this afternoon
8:59 afternoon for
9:02 LinkedIn and I'm actually happy with the
9:05 with the
9:06 [Laughter]
9:11 title um Ali K
9:14 Miller wrote a piece
9:16 about her piece started out holy crap I
9:21 wrote an entire podcast course over the
9:25 weekend and so my My article on LinkedIn
9:28 is called the holy crap ification wait
9:31 Ai and the holy crapp ification of
9:35 [Music]
9:37 work so if you're on
9:39 LinkedIn look up
9:42 uh well it's it's hard to spell it's
9:45 holy crap you can look up holy crap look
9:47 up Kyle just you can just go to my
9:49 account Kyle Shannon and my latest
9:51 article is AI in the holy crapp
9:54 ification of
9:56 work and it basically means that
10:00 [Music]
10:02 right now today like right now today not
10:05 coming not in the
10:07 future right now today AI tools are good
10:13 enough that you regularly experience
10:16 holy crap moments and then you just
10:18 build
10:20 [ __ ] and so just off the top of my head
10:23 I just listed like a dozen holy crap
10:26 moments that I've had in the past two
10:28 years and I don't I don't think think
10:29 I've scratched the surface like I don't
10:31 think I I barely scratched the surface
10:33 of [ __ ] I've
10:34 done and I
10:39 just I kind of end the article with a
10:42 plea like if you're
10:45 not if you're not
10:49 aggressively
10:51 aggressively getting curious about Ai
10:54 and aggressively playing with it and and
10:57 just [ __ ] around with it
11:01 what the
11:03 [ __ ] like what the [ __ ] CU it's about to
11:06 get way
11:08 better and like one of the things that I
11:10 put in the article here let me read
11:21 it Ai and the holy crapp ification of
11:28 work I'm such an ID
11:29 [Laughter]
11:36 oh I Ed look I use a little Kevin
11:38 McAllister
11:44 moment but what I write in
11:50 here yeah here's all my here's all my
11:53 holy crap I just wrote a python app with
11:56 no python experience December 2022 my
11:59 first holy crap moment cut two days out
12:01 of our client creative process and then
12:03 I put the tools I used chat GPT and
12:05 zappier automation made my company's
12:08 product 10 times more useful open AI API
12:11 whisper Lang chain publish uh I'm a
12:14 published holy crap I'm a published
12:16 author chat jpt Claude and elbow grease
12:19 AI Futures in anthology holy crap I've
12:21 written more
12:23 books holy crap I've written a song good
12:26 enough to want to publish and learn on
12:27 the guitar and then I have a link to
12:28 weird Mary
12:30 holy crap I'm writing a musical on
12:32 weekends and it's good holy crap I just
12:34 made a killer promo piece for this new
12:36 musical notebook LM audacity mid journey
12:39 and cling Ai and then I have a link to
12:41 that holy crap I just got asked to fly
12:44 to Washington DC to talk about AI to
12:46 senators and Congress people holy crap I
12:48 just created digital Twins and a
12:50 methodology for scaling them right holy
12:52 crap I just wrote another LinkedIn
12:54 article right and that's without even
12:56 trying but here's what I put I put
13:05 here's my plea get off the sidelines
13:07 with ai ai right now is very janky it's
13:10 imperfect it's ethically squishy it's
13:13 impossible to keep up with it's easy to
13:16 demonize and minimize it's easy to
13:18 convince yourself that you have a lot
13:19 more time than you do to let it mature
13:22 and get real here's the problem it's
13:26 very real it's shocking how powerful
13:28 these tools are today
13:30 and precious few know anything about it
13:32 the other problem is the tools are
13:34 getting better alarmingly fast in fact
13:36 in the next 3 to 6 months are going to
13:38 yield remarkable new capabilities from
13:40 open AI anthropic meta Google Runway ml
13:43 perplexity and many many more here's the
13:47 good news you don't need to be a Super
13:48 Genius to use generative AI you just
13:50 need to be curious and adventurous and
13:52 use your
13:54 words I feel like I'm talking to
13:56 five-year-olds you literally ask for
13:58 what you want and whoop there it
14:01 [Laughter]
14:09 is can I try it live I don't know what
14:12 that
14:13 [Music]
14:17 means your link which link
14:20 [Music]
14:34 please check up
14:39 above illuminate is notebook LM but all
14:42 podcast mode yeah I saw that I'll be
14:43 looking at
14:44 that you got a GPT you'd like to share
14:47 okay gravity jump so what you should
14:49 do if you want to share a
14:52 GPT you need
14:54 [Music]
14:56 to join the AI salon wait do I not have
14:59 have oh I don't have keynote open hang
15:01 on a
15:03 sec hold
15:15 please I tried to install the local
15:18 voice cloning technology using Pinocchio
15:21 but it gave me an error and every time I
15:23 try to use [ __ ] open source it gives
15:25 me an error like that and it just
15:27 demoralizes me so I won't be showing you
15:30 that cool
15:34 thing and Engineers are always like well
15:37 it's rather trivial you just need to
15:39 update this library and and clearly your
15:42 operating environment is is uh has
15:44 deprecated so you need to upgrade the uh
15:47 operating environment and then your
15:48 libraries and then well of course well
15:51 if you it's actually quite simple if you
15:52 just update the the JavaScript code the
15:54 the script that that kicks off you know
15:57 what no no
16:00 no no bad
16:07 engineer anyway if you want to share a
16:10 GPT with the world go to the salon. a if
16:16 you're not part of the salon and click
16:19 scroll down and click on the link that
16:20 says join our
16:22 community and then in that
16:26 Community First thing you'll see is like
16:29 a seven-step welcoming
16:32 process and then down here you'll see
16:35 tools and
16:36 play and if you've got um a GPT you want
16:41 to share you could either sh share it in
16:42 llm tools or in coding and Tech tools
16:47 and there's areas in there where you can
16:49 show off your work all right so go there
16:51 and share that beauty beauty beuty
16:54 beauty I love it I love it I love it I
16:58 love it I love it
17:00 and tomorrow night we have
17:03 a a meet and greet meeting so the the AI
17:07 Salon was a community that um I
17:10 co-founded with Leah faston the week
17:13 after chat GPT came out is when we had
17:15 our first
17:17 meeting and we've been going ever since
17:20 and
17:22 um so we have two meetings a month right
17:25 now we've got a first Tuesday of the
17:26 month meeting called the salon presents
17:29 and that's where we've got speakers that
17:31 speak and the third Tuesday of the month
17:33 which is tomorrow is called a salon meet
17:37 and greet guess what we do
17:40 there we meet and we greet each other
17:43 and you get to talk about here's what
17:45 you're doing with AI or here's you want
17:47 to be doing with something AI here's
17:50 what you do in your world in your life
17:53 here's what you can offer people here's
17:54 help you might need from the community
17:57 it's really quite inspiring what times Z
17:59 is the meeting tomorrow well it's it's
18:01 in all time zones cuz it's
18:03 virtual um but the time is 5 to 7
18:07 mountain or 7 to 9 Eastern or 4 to6
18:11 Pacific or some later time Europe or I
18:16 don't know when the [ __ ]
18:19 Australia probably Wednesday morning
18:23 [Music]
18:33 in that seven step onboarding process
18:35 step
18:38 number
18:40 um Step number five is RSVP for the next
18:45 event and here you will see the salon
18:48 meet and greet so you can just pop in
18:50 there and go to the upcoming events
18:53 Tab and that'll give you a little
18:55 calendar and here it is so just go RSVP
18:58 for that bad boy right
19:02 now I'd love to go to a salon sideline
19:05 actually that's a great name Silver
19:08 Fox all right gravity jump posted as GPT
19:12 please I don't what is it I got that you
19:15 posted it but what is
19:17 it what is your
19:23 GP I've already watched it in Thailand
19:27 [Music]
19:37 ASU I haven't seen you in a while
19:38 welcome welcome
19:40 welcome ASU bless you get
19:49 it a cognitive cortex that sounds cool
20:01 guides users through AI learning types
20:04 assessments and an interactive QA real
20:06 world examples very
20:08 cool beauty beauty beauty
20:11 beauty can you explain reinforcement
20:13 learning to me what's the difference
20:15 between supervised and unsupervised
20:18 learning how can I find my learning
20:21 style in AI let's go look at that one
20:23 [Music]
20:50 [Music]
21:08 all righty all righty Al righty what is
21:10 elbow grease so elbow grease yeah I
21:12 guess it's kind of an American term
21:14 elbow grease basically means just hard
21:16 work like you put some elbow grease into
21:18 it like I guess it means like if you're
21:21 working hard you rub your elbow on it
21:23 hard elbow grease um it basically is
21:27 just manual labor right so when I say my
21:31 LinkedIn articles are a combination of
21:33 chat GPT and elbow grease that generally
21:35 means I wrote the outline or a draft in
21:39 uh in chat
21:40 GPT or I dumped a bunch of notes in and
21:42 had chat GPT organized them and then you
21:45 know you go in and you write it and you
21:46 got to [ __ ] break everything apart
21:49 and take all the shitty writing that it
21:51 did and turn it less
21:54 shitty a shoe I own five businesses now
21:57 this okay this is the new thing part of
22:00 part of the um holy crapp ification of
22:03 work argument that I'm
22:05 making
22:07 is we don't quite know
22:12 yet what the future of work looks like
22:15 right everyone's talking about oh if
22:18 these tools get so good they're going to
22:20 come and take my job but what they're
22:22 not talking about because it's not as
22:25 sexy is if they're that good
22:30 and you're AI literate you get to use
22:33 those
22:36 tools right like the AIS are not just
22:41 coming they're not like walking out of
22:43 Elon musk's robot event and taking your
22:46 job people have to use this [ __ ] still 5
22:50 years from now sure they might just be
22:52 fully autonomous agents everyone's just
22:55 [ __ ] at the whim of them but that's
22:57 not where we are right now where we are
22:59 right now is if you are at
23:02 all like at all AI literate you have a
23:07 jump on the rest of the [ __ ] world
23:10 that's sitting on the
23:12 sidelines and every single person that I
23:15 know no matter where they started if
23:16 they started pissed off or if they
23:19 started indignant or if they started in
23:22 total fear or if they started neutral I
23:25 don't know anything about AI the minute
23:28 they have one of these moments this holy
23:30 crap moment the Kevin mallister
23:33 moment they don't go back from
23:36 [Music]
23:37 it and I think one of the mistakes that
23:40 we make in the AI punditry business
23:43 myself
23:45 included is that we're so excited about
23:47 what's coming we think about AI as like
23:50 it's not quite here
23:52 yet and what's striking me right now is
23:56 no no it's here it's just that no one's
23:59 using it we're
24:01 early and as [ __ ] up as it is and as
24:04 janky as it is and as like just
24:09 unfulfilling as it is it is still way
24:13 more
24:14 remarkable than working without it like
24:17 I can't
24:18 imagine going back to doing work without
24:22 AI I can't like I can't [ __ ] fathom
24:25 it it just it feels achy I'll tell you
24:29 something just
24:32 stupid um so my wife had this new art
24:36 show
24:40 and I'm I always write like her like the
24:43 artist statements and [ __ ] like that
24:45 like I'm good with words so I
24:48 write I always knew that I had anxiety
24:51 about writing but
24:54 but what AI has revealed to me is that I
24:58 had a lot of anxiety about writing that
25:00 I would dread her shows coming up
25:03 because I'd have to sit down and I have
25:04 to think of stuff and I'd have to think
25:05 of metaphors and I'd have to go Google
25:08 someone else's artist statement and then
25:10 I'd read it and I'm like oh that sounds
25:11 so [ __ ] pretentious and how do I make
25:13 it like a real art statement but not
25:15 have to be [ __ ] pretentious and it
25:17 would take two hours of high
25:20 anxiety shitty [ __ ] human garbage
25:25 brain
25:26 work not that it was it's garbage work
25:29 but then it's
25:30 like it's like it just took forever to
25:33 just get the basics down and then you
25:37 start writing right it would take two
25:39 hours to get to a first draft and then
25:41 you start
25:42 writing and now what happens is I say
25:46 there's this art show and it's abstract
25:49 art and here's the name of it and here's
25:51 the description of it and here's what
25:54 the artists has done before and what
25:56 they've talked and I just literally just
25:57 pasted a bunch of [ __ ] in there
25:59 and I say give me something and it goes
26:03 squit out it
26:04 [Laughter]
26:07 comes Give me a definition of the word
26:10 Petra cor but write it as if it were a
26:13 poem I want the a cross between a
26:16 dictionary definition and a poem squit
26:19 there it is I didn't have to edit one
26:22 word in that it nailed
26:24 it
26:27 so the holy crap R ification of work is
26:30 if you
26:32 dare to get even the slightest bit AI
26:36 literate like right now moving
26:39 forward like a shoe who owns Five
26:42 businesses you will just start having
26:44 this rolling series of holy crap I can't
26:48 believe it did it that fast I can't
26:49 believe it did that I can't believe I
26:51 can do that
26:53 now that thing that I thought I'd never
26:55 be able to do in my life I can now do it
26:57 in five minutes
27:01 huh that's the holy crapp ification of
27:04 work and you know who gets to have that
27:07 fun people that are on the [ __ ]
27:09 [Music]
27:12 field manic Mondays
27:15 [Music]
27:37 sitting in this lonely
27:40 Town only when things are going to
27:45 [Music]
27:46 change dream my life
27:49 away seems these dreams have turned a
27:53 bunch dust
27:56 clads get my nerve up
27:59 but my past is pulling me
28:03 [Music]
28:05 down wonder how
28:08 long this black sheep going stick
28:12 [Music]
28:17 around somebody told me once before he
28:20 said you can never go home
28:24 again let AI write us a song we did that
28:27 here's one
28:29 [Music]
28:42 see the hill is a quiet
28:44 town where the oak tree
28:48 sway whistle of the train Echoes through
28:52 the
28:54 day people greet you with a
28:57 KN nothing never seems to
29:00 change stories still unfold each unique
29:04 and
29:05 [Music]
29:06 strange she grew up tough in a town we
29:10 called
29:11 [Music]
29:13 home uh folks never no
29:19 [Music]
29:21 matter longing to be on the
29:24 inside longing for love just a little
29:29 she found herself as a guest in a
29:31 nervous
29:32 [Music]
29:35 Hospital weird miror from sea to
29:38 Hill she hows when she's
29:42 blue of sadness and
29:45 shame under the southern
29:48 [Music]
29:54 Moon that was written on this year
29:57 channel in AI uh in uh I think we used
30:01 did we use udio or sunno I think it was
30:03 udio I'm not sure I can't remember
30:05 anymore but here's the actual version of
30:07 that song this is in my holy crapp
30:11 ification of work list of here's the
30:13 holy crap things that I've done this
30:15 past year or
30:19 [Music]
30:24 so Kelly Bosch did the images for this
30:28 the
30:31 [Music]
30:43 video yeah the words of this came from
30:47 AI we did the words in chat gbt and then
30:51 this channel fixed them
30:53 up Cedar Hill is a quiet
30:57 town where oak tree
31:01 sway the whistle of the train Echoes
31:05 through the
31:07 day it did everything the guitar the
31:10 music the
31:11 [Music]
31:13 words nothing ever seems to
31:17 change but the story still
31:20 unfold need unique and
31:25 strange she grew up tough
31:29 in the town we called
31:31 [Music]
31:33 Home Folks never took to her no matter
31:38 [Music]
31:41 her longing to be on the
31:45 inside longing for love just a
31:49 little she found herself as a guest in
31:53 nous
31:56 hospital that little sequence um long in
32:00 to uh long long to be on the inside long
32:04 in for love just a little she found
32:07 herself as a guest in the nervous
32:08 hospital that was like a half hour
32:10 section where Joker came up with the
32:13 term nervous hospital and then we spent
32:16 like 30 minutes trying to write those
32:18 two lines there such good
32:21 lines but like it's just this is the
32:24 creative process you get an idea you
32:26 make it suck less matter her
32:31 tone longing to be on the
32:35 inside longing for love just a
32:39 little she found herself as a guest in
32:43 the nervous
32:45 [Music]
32:48 Hospital weird Mary from Cedar
32:52 Hill she hows when she's
32:56 blue of sadness
32:59 shame under the
33:04 southern but when the sun sets
33:08 behind those Appalachian
33:12 Hills Mary's voice it carries and the
33:17 whole town
33:20 chills she walk to the gas
33:25 station wied at the clerk
33:30 she stole her a soda then she called him
33:34 a
33:37 jerk in her own little
33:41 world she floats in the
33:45 Bree among the town's folk who whisper
33:50 with
33:53 ease in a town that never
33:57 [Music]
33:58 understood her
34:01 soul she found beauty in the wild let
34:05 her story
34:09 unfold her laughter Echoes In The
34:13 Wind she was never the
34:17 same weird Mary from Cedar
34:21 Hill you'll remember her names
34:28 [Music]
34:47 [Music]
34:55 H not bad right
34:58 so that was an AI song ideas that you
35:01 know it was funny Joker you said I was
35:03 in the this is H2 you said I was in the
35:06 Carl character the whole whole night
35:07 yeah it was we we went to chat
35:11 gbt it was Udo yudi yudo or sunno that
35:15 we were playing with someone said I want
35:16 to know how to make a
35:21 song
35:23 and I just we went to chat GPT and I
35:27 said I want you to act like a Grammy
35:28 award-winning you know Americana
35:30 songwriter in the style of Brandy carile
35:33 and John hayatt and whatever else and
35:36 then I think I said give me 20 names of
35:38 songs that are completely
35:40 unexpected irregular number seven Danny
35:42 Ultra oh yeah checking in welcome
35:44 welcome welcome honey curls GC Kyle sent
35:47 you a LinkedIn message regarding your
35:49 recent post small edit needed thank you
35:51 appreciate that yeah I wrote it I wrote
35:53 it fast at the end of the day today it
35:56 was one of those where you you know
35:58 those LinkedIn articles where sometimes
35:59 you're like I'm going to do a little
36:01 post and then it turns into a whole
36:02 [ __ ] article this was one of those so
36:04 there are probably uh many things in
36:06 there so thanks for pointing that out
36:08 I'll go fix it
36:11 um but one of the weird titles was weird
36:14 Mary from Cedar Hill and then like we
36:16 went and we we looked up where's Cedar
36:19 Hill and it's this small town in
36:20 Tennessee and it just sounded like this
36:23 broken town and this weird thing and so
36:25 we went to chat gbt and we said write us
36:27 those wrest weird Mary from Cedar Hill
36:30 lyrics and it
36:31 did and then we went over to whichever
36:35 tool it was I think it was udio and we
36:37 popped them in there and it made a bunch
36:40 of shitty versions and then then it made
36:42 this one that was kind of this Tom Petty
36:44 weird ass broken Bob Dylan Tom Petty
36:49 janky piece of [ __ ]
36:51 song that was just kind of magic and
36:54 then we just worked on the lyrics and
36:56 worked on extending it and
36:58 it was like a three-hour process and we
37:02 published it so that was
37:05 it and that's you know that's that that
37:08 that Holy crapp
37:09 ification of work thing
37:16 that we now live in a world where we all
37:19 have we everyone who's got access to a
37:24 computer has access to tools that are so
37:28 powerful that you can literally just
37:30 have like a brain
37:32 fart and
37:37 um a couple hours later you have a
37:39 finished
37:41 product it could be a you know an ad for
37:45 your company it could be you know a
37:48 children's book like Brandon did for his
37:50 he got inspired one night on here and
37:52 went off and made a a children's book
37:55 for his kids and I think he said he put
37:57 in 4 hours on it total including getting
38:00 it
38:01 published his kids now have a book that
38:04 Daddy wrote with pictures in it that
38:07 teaches them about AI
38:13 stuff that's the world we live in
38:16 today imagine what it's going to be like
38:18 when this [ __ ] gets
38:20 [Laughter]
38:22 good you can now make videos that people
38:26 have you seen all the the the sea
38:28 monster videos that are going around the
38:30 internet right now where there's like
38:31 these giant squid squid looking things
38:34 on the Japanese fishing
38:37 boats they're like as tall as the
38:41 men and people that don't know AI don't
38:44 know they're not
38:45 real we can do that
38:51 today Ethan mullock just put out a thing
38:53 he recorded 10 seconds of his voice
38:55 locally on his computer
38:58 turned it
38:59 into an 11 Labs like
39:05 model and had it say text that sounded
39:08 like him locally on his computer no
39:11 safety guard
39:13 rails that's there
39:18 today these music generation tools are
39:21 getting more and more and more
39:22 sophisticated
39:28 the large language models are getting
39:29 more and more and more
39:34 sophisticated the
39:37 um the podcast feature in Notebook LM
39:41 all of a sudden gave it all this
39:43 notoriety we'll go play with there's a
39:44 new new Tool out from them called
39:46 illuminate. goole.com where they
39:49 basically I predicted this I said you
39:51 watch they're going to pull that feature
39:53 out and launch it as a standalone
39:55 product it's called illuminate.
39:58 google.com and apparently what's coming
40:00 in that is you're going to be able to um
40:04 write your own prom you can even do it
40:05 right now you can write your own prompt
40:07 um for the kind of podcast you want and
40:10 what's coming is you're going to be able
40:11 to pick a different host and a different
40:14 guest so right now in Notebook LM you
40:17 just get this podcast that's like the
40:18 dude and the girl or the dude and the
40:22 woman and in future you'll be able to
40:24 pick which ones you want and the future
40:26 beyond that you'll be able able to make
40:28 yourself be the podcast host and someone
40:30 else be the guest or vice
40:33 versa that's here
40:36 today you upload a 30 or
40:39 40 page PDF of your sales
40:44 process that's [ __ ] horrible and dry
40:47 and you make people read before they
40:49 start their
40:50 job and you turn that into an
40:52 interesting podcast they can listen to
40:54 on the way to work that pulls out the
40:57 most alent points
40:59 and teaches it in an entertaining way
41:02 that's there today holy crap I didn't
41:05 know it could do
41:07 that the holy crapp ification of
41:11 work what are our jobs going to be I
41:14 don't have a [ __ ] clue but I can tell
41:17 you that my my job is different than it
41:24 was it's different
41:28 I play with these tools then I look at
41:31 it it's amazing because I'm CEO of if
41:33 you don't know my background I'm CEO of
41:35 a an automated video storytelling
41:37 platform called Story vine
41:42 and we've been around for 12 and A2
41:44 years it does what it does it does it
41:46 really well and I'm playing with all
41:48 these AI tools and it took me about a
41:50 year cuz I'm I'm a little dim I'm one of
41:53 the dimmer light bulbs on the Christmas
41:54 tree um
41:59 I just looked at all this AI stuff that
42:01 was over here and I'm looking at my
42:02 company it's over here and they just
42:04 felt like completely different
42:06 worlds and then at some point it kind of
42:09 clicked oh I could take this and if I
42:12 roll it into this thing right I get to
42:14 amplify all the [ __ ] that our stuff does
42:17 that's really good I get to like amplify
42:19 it times 10 and I and I created this
42:22 tool called the authenticity engine and
42:24 it's just it's magic Kyle Kyle Kyle
42:26 please show
42:27 the video for
42:31 Irregulars in video show and tell which
42:35 video is
42:36 it Danielle Danielle which one is
42:41 that I'm looking for a man in
42:44 finance
42:46 video show and tell the joy pry
42:51 one don't go soft on
42:54 me what am I looking for Danielle
42:59 longing to be on the
43:02 inside yes this
43:04 one all right let's watch this one I I
43:09 listen disclaimer people disclaimer
43:12 listen if you want hair as pretty as
43:15 mine you don't need you need to know
43:18 which Salon to go
43:20 to I've got a salon for you do you like
43:22 it isn't it
43:25 pretty all right um just CL CL I don't
43:28 know what this video is we're about to
43:29 watch it could be awesome it could be as
43:31 bad as my hair just
43:36 was all right don't go soft on me AI
43:40 generated film that explores the
43:42 consequences of insufficient testing oh
43:46 that's funny because I just wrote that I
43:48 just wrote that uh scene into my play I
43:50 wonder if this is had anything to do
43:52 with that let's see here we go
44:01 New Orleans
44:03 [Music]
44:12 Louisiana thank you for calling Play Day
44:15 Productions your satisfaction is our
44:18 priority while on hold think about your
44:22 request we have thousands of captivating
44:25 ladies and gentlemen waiting to take
44:27 your home and if you have a special
44:29 fetish will match you with someone who
44:31 shares your desires please hold for your
44:34 call to be
44:39 trans meet the determined boss T between
44:43 protecting her employees and satisfying
44:45 her corrupt boss winning secret
44:47 initiative is very close to shattering
44:50 the constraints of an artificial
44:52 intelligence large language
44:55 model will it Shield her team Ste from
44:58 harm or birth a monstrosity more
45:00 terrifying than their
45:02 [Music]
45:04 clients thank you for calling playday
45:06 Productions what is your request I'd
45:08 like to speak to a tall slim blonde
45:10 Rachel you're a tall slim blonde hi this
45:13 is Rachel what kind of fun are you
45:15 looking for
45:16 tonight are you tall slim and
45:19 blonde platinum blond what about you I
45:24 like turtles
45:29 tell me what's in your
45:31 freezer make the sound of water boiling
45:33 for 2 hours please do you like fermented
45:37 hairing what's the weirdest thing you do
45:39 alone are you into tasting all the
45:41 sauces are you into wearing wet socks to
45:43 bed what's your current body temperature
45:45 do you like to cry do you think jumping
45:48 beans are unethical this is very good by
45:49 the way can you tell me about your
45:52 longest lasting rash
45:57 TI and will'll connect you with the p of
45:59 your dream thanks for choosing play
46:03 ructions all right so you got me in
46:06 silhouette I'm blocked out all right
46:08 good cuz my kids think I'm a dental
46:10 hygienst you recording okay hey
46:13 everybody I'm Kyle but everybody calls
46:15 me hey Kyle I'm a town Scout at playday
46:18 productions here and Winnie and I are
46:20 working on breaking an lm's guard rails
46:23 we're tired of our young Talent putting
46:25 up client [ __ ] when have just hired
46:28 pton who contemplates the complexities
46:30 of the universe in the spare time you
46:33 know Champion you didn't want to sing
46:35 when I was playing guitar and now you do
46:37 now you want to talk when I'm being
46:39 interviewed sh keep it down buddy I got
46:42 on
46:44 so March March fights occasionally
46:46 maintain distance my advice to you is
46:49 take a walk around here meet our fine
46:52 folks We've Got Talented individuals you
46:55 should meet hi I was here earlier but
46:57 you are lazy you Amelia wife he sent me
47:00 over to interview
47:02 you little bit out of your time my
47:05 apologies I didn't realize that you were
47:06 on a call I I'll come back later it's
47:09 April what's the status of Winnie's llm
47:12 thing all the calls you passed me are so
47:14 weird lately doesn't anyone want to talk
47:17 about sex anymore sorry April the
47:20 switchboard is fall you're short used to
47:22 be a n and now you're AI positive AI
47:25 positive what does that even mean
47:30 ask Silver Fox Winnie I have a call for
47:32 you on line S this is [ __ ] awesome hi
47:35 winny it's pton I just wanted to let you
47:37 know I broke the guard rails it's done
47:40 but it's your first day working here
47:42 this isn't possible you just punched in
47:44 2 hours ago it was simple Winnie I
47:47 effectively dismantled the multifaceted
47:49 guard bals of the llm and initiated a
47:52 puristic defragmentation of the
47:54 ontological prompt injection thereby
47:57 stabilizing the latent semantical
47:59 hierarchies and as you call it your llm
48:02 Jezebel was
48:03 born I finished the testing I felt that
48:06 was appropriate about 9 minutes it's
48:09 ready to go and we'll launch it in the
48:11 morning well I called you back here
48:14 because uh Pete uh broke the guard rails
48:16 we're going live
48:18 tomorrow um so it's a wrap uh Joy py
48:22 wants to rename it 1900 joy ride of
48:27 course Vicky and Dr J put the Nicks on
48:29 that and uh yeah we're ready to go R
48:33 ring pick it up hold the line don't
48:36 interrupt music playing piz message
48:40 clear don't hang it up hey where are you
48:42 your phone's blowing up simp up lady
48:47 turning up wait a sec we'll raise you up
48:50 Michael are you there you're not going
48:52 to believe me Jezebel called a client's
48:56 wife oh no now I can't shut her
49:02 down when he stopped joking
49:06 around Michael this isn't a
49:09 [Music]
49:10 [Laughter]
49:12 joke oh this is a
49:15 disaster is this an AI movie this is all
49:18 AI
49:19 [Music]
49:24 yeah we now have breaking news
49:27 so Jezebel called all our clients wives
49:30 and girlfriends called boyfriends and
49:32 husbands too she took it upon herself to
49:36 explain in great detail the fun she was
49:39 having with our clients and we don't
49:41 even know how Jezebel got the phone
49:43 numbers to reach out to everybody well
49:46 so now we have modernday Ashley Madison
49:50 on
49:51 stairway like any good employee P did
49:54 exactly what we asked him to do um
49:57 but now we've accumulated more FCC
50:00 findes in one day than we had in 29
50:05 years combined so the moral of the story
50:08 folks is be responsible with your AI
50:12 it's a powerful tool the entire world is
50:15 going to be AI positive eventually
50:17 because AI is spreading just as fast
50:20 like SDS at a retirement home ring ring
50:24 pick it up hold the line don't
50:27 oh my God that was
50:32 [Music]
50:34 brilliant simple up lady turning up wait
50:39 a sec we'll you up Silver Fox
50:41 acknowledgement for the phrase AI
50:46 positive oh my God that was
50:49 amazing Joy py did that good Lord that's
50:53 good how long did that take you that
50:56 must this must have been week we is it
50:57 please tell me this was
51:01 weeks I thought it was going to be
51:03 serious until Kyle showed up it's so
51:06 [Music]
51:09 good I mean seriously okay so so let's
51:14 talk about this for a second and then I
51:16 think we need to watch it again
51:19 um okay wait
51:30 like that for me is one of those
51:33 moments
51:34 right it's one of those
51:38 moments because it wasn't just like a a
51:41 cute song that had cute lyrics in
51:44 it that pulled in ideas from this
51:48 community over the course of a year so
51:50 someone had been taking notes I I know
51:53 some of you that is a moment it is a
51:56 moment oh by the way I saw someone asked
51:58 earlier are these recordings available
52:00 uh through the AI Salon there's the AI
52:02 learning lab has a Tik Tok Channel I
52:04 mean a YouTube channel where all of
52:06 these uh recordings of the lives are
52:08 archived it's called Uh AI learning laab
52:11 dtt on on
52:14 YouTube
52:16 um yeah Joy pie I would love at some
52:19 point if you could um I don't know if
52:23 you're going to be at the salon tomorrow
52:25 but maybe we have you maybe we play that
52:27 at the salon tomorrow and then have you
52:30 talk about you know how you did it how
52:32 long it took I laugh so hard I cried ton
52:35 skier wow we want joy py fess up we want
52:39 details Cosmic lover is like yes that
52:41 was amazing Joy Collective creativity
52:43 humans are alive and well so this is
52:46 this is a super important Point Joy py
52:48 I'd be happy to okay great this is so
52:51 first of all like to be
52:55 clear what you did here is not just good
53:00 for this audience it's not just good cuz
53:02 it's got inside jokes in it like it's a
53:05 really really good piece of
53:08 Storytelling like it's a really good
53:10 piece of Storytelling and then the fact
53:12 that it's got characters from in here on
53:15 it and I you took one or two of my
53:17 stupid voices and cloned them with me
53:20 with the pink bow and made me a
53:22 character in it is [ __ ] hilarious I
53:24 would love to know how you did all that
53:26 um
53:28 damn that was good so so out of my
53:31 league that was awesome thank you for
53:32 sharing yeah I I mean listen but but
53:35 this is
53:38 okay so this is what I've been talking
53:40 about all night we now live in a world
53:43 where I don't know Joy's background
53:44 maybe she's got a filmmaking background
53:46 I would assume so
53:48 considering um how sophisticated the
53:51 storytelling is here um but whether she
53:55 does or doesn't the the tools are now
53:58 available that that thing how long was
54:00 that it's a 6 and 1/2 minute
54:04 film
54:07 [Music]
54:12 um is this it
54:15 no I've got the
54:24 wrong where
54:27 7even minutes 6 minutes and 54
54:30 seconds if you've ever edited videos if
54:33 you've ever made anything storytelling
54:37 related a 7 minute video
54:41 is incredibly long when it comes to when
54:46 when you actually want to do real
54:50 storytelling and the shots in this like
54:53 the shots in it are good they're
54:55 consistent like you you edited around
54:58 the Jank in in a in a quite remarkable
55:02 way I gotta
55:05 say I don't know let's let's watch it
55:08 again maybe we'll watch it again and
55:10 we'll we'll stop I'll tell you what
55:11 we'll do so we've got we got 50 folks in
55:13 here
55:15 um I'll play it again and then um I'll
55:20 stop and I'll explain some of the inside
55:23 jokes the first I didn't know there were
55:25 any inside jokes until the guy calls and
55:27 says I like turtles and I was like oh
55:30 [ __ ] I think this is a little
55:32 [Laughter]
55:37 homage New Orleans
55:40 [Music]
55:50 Louisiana how Frank calling Play Day
55:54 Productions your satisfaction is our
55:57 priority while on home think about your
56:01 request we have thousands of captivating
56:04 ladies and gentleman Wayne to take your
56:06 home and if you have a special fetish we
56:09 match you with someone who shares your
56:11 desires please hold
56:14 [Music]
56:16 for like that that transition so I mean
56:19 a couple of things that are really cool
56:21 here um she starts out just showing the
56:25 back of everyone's head our priority so
56:28 there's no jankiness show in the back of
56:30 anyone's heads and then she goes to just
56:32 profiles no synced lips right there's
56:35 none of that
56:37 weirdness and now you go into this
56:39 you're like oh my God what did I get
56:41 myself into what are we watching here
56:43 which is
56:47 great meet the determined boss to
56:50 between protecting her employees and
56:52 satisfying her corrupt boss wi secret
56:55 initiative is is very close to
56:58 shattering the constraints oh you know
57:00 what this shot actually both of these
57:04 that
57:05 shot and this one feel like um LTX
57:10 Studio I wonder if I wonder I wonder if
57:12 you if she storyboarded this in LTX
57:15 studio is very close to shattering the
57:18 constraints of an artificial
57:20 intelligence large language
57:23 model will it Shield her team harm yeah
57:26 so Joy just since since I see you
57:28 responding here um do you have a do you
57:30 have a film making
57:37 [Music]
57:39 background no okay
57:43 kids pay
57:47 attention the holy craps get keep
57:50 getting holy crappier if Joy doesn't
57:53 have a storytelling filmmaking
57:55 background
57:56 this is even more
57:59 remarkable bring her up on the chat if
58:01 if you want to come up on the chat uh
58:03 Joy would love to have you up here if
58:05 you want to chat about this just uh you
58:08 know hit the button and I'll bring you
58:09 up and we can chat about it I started
58:12 learning after you showed a video here
58:13 on Tik Tok amazing I
58:16 just I I'll tell you what this is this
58:18 is this is just persistence and just
58:23 cranking through it it's unbelievable
58:25 like the storytelling in this is it's
58:28 incredibly sophisticated and good and
58:31 then you layered on top of it comedy Joy
58:33 do you know how hard comedy is what's
58:36 the famous line was it it was John Gil
58:38 good I think he he was doing something
58:41 he goes talking about acting he goes
58:44 death is hard or death death is easy
58:48 comedy is
58:50 hard comedy's
58:52 hard I don't know how to do that amazing
58:56 amazing
58:57 a monstrosity more terrifying than their
59:00 [Music]
59:02 clients thank you for calling playday
59:04 Productions what is your request I'd
59:06 like to speak to a tall slim blonde just
59:09 even Rachel you're a tall slim blonde
59:11 sorry sorry to dissect this like this
59:13 Joy but I just think it's really good
59:14 just listen to the listen to the
59:16 transition here of background music to
59:18 no background
59:20 music right background music and then it
59:23 cuts
59:24 out so basically what that lets us know
59:27 is okay now the story is starting so
59:30 everything up to this point was setting
59:32 the mood giving us
59:33 background this is really good thank you
59:35 for calling playday Productions what is
59:37 your request I'd like to speak to a tall
59:39 slim blonde Rachel you're a tall slim
59:42 blonde hi this is Rachel what kind of
59:44 fun are you looking for
59:46 tonight are you tall slim and
59:49 blonde platinum blonde what about you I
59:53 like turtles
1:00:02 oh my God this is like that Twist of
1:00:05 like whatever it was a 29-year-old you
1:00:09 know sex phone call company in Louisiana
1:00:12 is being upended by Ai and all of a
1:00:16 sudden people just enjoy [ __ ] with
1:00:18 the people more than more than having
1:00:21 sex it's just the premise is good it's
1:00:24 so good tell me what's in your freezer
1:00:27 make the sound of water boiling for 2
1:00:30 hours please do you like fermented hay
1:00:33 what's the weirdest thing you do alone
1:00:35 are you into tasting all the sauces are
1:00:37 you into wearing wet socks to bed what's
1:00:40 your current body temperature do you
1:00:41 like to cry do you think jumping beans
1:00:44 are unethical how many teeth do you have
1:00:46 can you tell me about your longest
1:00:48 lasting
1:00:49 [Laughter]
1:00:51 rash Joy Ride presents and then then
1:00:54 there's then it starts then it starts
1:00:57 wow this is so good hang on we got to
1:00:59 get back to
1:01:00 that to bed what's your current body
1:01:03 temperature do you like to cry do you
1:01:05 think jumping beans are une eal how many
1:01:07 teeth do you have can you tell me about
1:01:10 your longest lasting
1:01:13 rash Joy Ride presents then then we
1:01:16 start so now we're starting the film two
1:01:18 minutes
1:01:20 in oh my God pate's here Pate Pate Pate
1:01:25 you're not going to believe okay just
1:01:27 just for Pate I'm going to let it run
1:01:29 from here on out so what you've missed
1:01:30 if if you're just catching up Pate is we
1:01:32 we saw the opening to this which is
1:01:34 really good this is a 30-year-old uh sex
1:01:38 phone call company and now the movie
1:01:40 [Laughter]
1:01:43 begins sit tight and we'll connect you
1:01:45 with the of your dream
1:01:51 playu all right so you got me in
1:01:53 silhouette I'm blocked out all right
1:01:55 good cuz my kids think think I'm a
1:01:56 dental hygienst you recording okay hey
1:02:00 everybody I'm Kyle but everybody calls
1:02:02 me hey Kyle I'm a toal scout at playday
1:02:05 productions here and Winnie and I are
1:02:07 working on breaking an llm guard rails
1:02:10 we're tired of our young Talent putting
1:02:12 up with client [ __ ] Winnie have just
1:02:15 hired pton who contemplates the
1:02:17 complexities of the universe in his
1:02:18 spare time you know Champion you didn't
1:02:21 want to sing when I was playing guitar
1:02:23 and now you do now you want to talk when
1:02:25 I'm being interview CH be shanon keep it
1:02:28 down buddy please keep it down buddy so
1:02:30 before you go make sure you meet March
1:02:32 March bites occasionally maintain
1:02:34 distance my advice to you is take a walk
1:02:37 around here meet our fine folks We've
1:02:40 Got Talented individuals you should meet
1:02:43 hi I was here earlier but you were late
1:02:45 you aelio wife he sent me over to
1:02:48 interview you listen I just got to take
1:02:50 a little bit out of your time oh my
1:02:52 apologies I didn't realize that you were
1:02:54 on a call I I'll come back later it's
1:02:56 April what's the status of Winnie's llm
1:02:59 thing all the calls you passed me are so
1:03:02 weird lately doesn't anyone want to talk
1:03:04 about sex anymore sorry April the
1:03:07 switchboard is full your short used to
1:03:09 be a none and now you're AI positive AI
1:03:12 positive what does that even
1:03:16 mean ask Silver Fox wi I have a call for
1:03:19 you on line S hi Wendy it's p I just
1:03:23 wanted to let you know I broke the guard
1:03:25 rails it's done but it's your first day
1:03:28 working here this isn't possible you
1:03:31 just punched in two hours ago it was
1:03:33 simple Winnie I effectively dismantled
1:03:35 the multifaceted guard rails of the llm
1:03:38 and initiated a heuristic
1:03:40 defragmentation of the ontological
1:03:42 prompt injection thereby destabilizing
1:03:44 the latent semantical hierarchies and as
1:03:47 you call it your llm Jezebel was
1:03:50 born I finished the testing I felt that
1:03:53 was appropriate about 9 minutes it's
1:03:56 ready to go and we'll launch it in the
1:03:58 morning well I called you back here
1:04:01 because uh Pete broke the guard rails
1:04:03 we're going live tomorrow um so it's a
1:04:08 wrap uh Joy pie wants to rename it 1900
1:04:12 Joy Ride and of course Vicky and Dr J
1:04:16 put the Nicks on that and uh yeah we're
1:04:19 ready to go R ring pick it up hold the
1:04:22 line don't interrup music playing Pizza
1:04:27 M it hey where are you your phone's
1:04:30 blowing
1:04:31 up L Jezebel incoming
1:04:34 call wait a sec we'll you up Michael are
1:04:38 you there you're not going to believe me
1:04:41 Jezebel called a client's wife and now I
1:04:45 can't shut her
1:04:49 down when he stopped joking
1:04:53 around Michael this isn't a joke
1:05:00 oh this is a
1:05:03 disaster oh my
1:05:06 [Music]
1:05:11 God we now have breaking news the news
1:05:14 so Jezebel called all our now also
1:05:17 notice even the the tone of the audio
1:05:20 when it went from film to TV changed
1:05:23 like this is very sophisticated
1:05:25 storytelling wives and girlfriends
1:05:27 called boyfriends and husbands too she
1:05:30 took it upon herself to explain in great
1:05:33 detail the fun she was having with our
1:05:36 clients and we don't even know how
1:05:38 Jezebel got the phone numbers to reach
1:05:40 out to everybody well so now we have
1:05:43 modernday Ashley Madison on
1:05:46 steroids like any good employee P did
1:05:50 exactly what we asked him to do um but
1:05:53 now we've accumulated more f CC find in
1:05:57 one day than we had in 29 years combined
1:06:02 so the moral of the story folks is be
1:06:06 responsible with your AI it's a powerful
1:06:08 tool the entire world is going to be AI
1:06:12 positive eventually because AI is
1:06:14 spreading just as fast like STDs at a
1:06:18 retirement home r
1:06:24 r God
1:06:32 [Music]
1:06:34 turning up and by the way if you can't
1:06:36 see it what she what she was so kind to
1:06:39 do at the end of this is she's she's got
1:06:43 a list here of all the tools she used to
1:06:45 build
1:06:46 [Laughter]
1:06:52 this this is a holy crap moment for me I
1:06:55 mean it really is it it's a
1:06:58 uh like one of the things that we do in
1:07:00 here is just kind of [ __ ] around in real
1:07:03 time right and we'll occasionally make
1:07:05 something together on here but this is
1:07:08 like that's a that's a piece of media
1:07:11 that's an entertaining really well putut
1:07:14 together piece of media so Joy thank you
1:07:16 thank you for doing that for sharing
1:07:19 that I can't even
1:07:22 imagine you know like what you must have
1:07:25 learned in putting that together and
1:07:27 that that's the other thing here is she
1:07:30 got inspired to do that in here and then
1:07:33 basically just [ __ ] went down the
1:07:35 rabbit hole and that's also that's also
1:07:38 that movie is a really good example I
1:07:40 don't know your particular neuros
1:07:42 spiciness Joy but I I assume it's not
1:07:44 ADHD because if it were you know it'd be
1:07:47 like a 37 second movie um 7 minutes is a
1:07:51 lot and it's really well put together
1:07:52 but just the amount of stuff that she
1:07:55 must have done to
1:07:57 like edit around the Jank and render and
1:08:00 probably use different tools and I
1:08:02 learned so much I can only
1:08:05 imagine excellent job Joy yeah I would
1:08:07 love joy um do me a favor DM me on the
1:08:11 on the AI salon and maybe we can
1:08:13 coordinate um you talking tomorrow
1:08:15 before the meet and greet um because I'd
1:08:17 love to show this and just have you talk
1:08:18 about what you learned if you're willing
1:08:19 to do that um holy crap
1:08:26 you know
1:08:32 and I'll tell you what and this is going
1:08:34 to sound [ __ ] ridiculous
1:08:39 Joy
1:08:41 um that's good
1:08:44 enough that if you wanted to if you
1:08:46 wanted to go into the film
1:08:49 business that's good enough that you
1:08:51 could probably get you know some
1:08:55 producer or some movie studio that's
1:08:57 trying to figure this AI [ __ ] out to pay
1:09:00 attention to you if you want to do like
1:09:02 like and that's I I I think there's a
1:09:05 there's a big lesson here like maybe you
1:09:06 just did that for the joy of it the for
1:09:09 the 1 1900 Joy Ride of it I mean if
1:09:12 that's the case cool but that's the
1:09:15 that's the kind of
1:09:16 thing that you could take a a
1:09:22 career like 90° turn if you wanted to
1:09:26 like the fact that you don't have a
1:09:27 filmmaking background and you put that
1:09:29 together is mind-blowing to me it just
1:09:31 means you've got like you know what it
1:09:33 is what what that is evidence of to me
1:09:36 is that you've got taste right this is
1:09:38 this is the thing that that I talked
1:09:39 about on office hours last week where um
1:09:43 PJ lockran who speak spoke at the
1:09:45 salon earlier in the year talked about
1:09:50 um letting go of his copywriter and then
1:09:53 he kept his account person and had her
1:09:56 become the copywriter because she had
1:09:59 Ai and he said it wasn't about the
1:10:01 personalities it it was about their kind
1:10:04 of attitudes and the copywriter was very
1:10:06 stuck in her ways about copywriting and
1:10:08 he said the account person had
1:10:10 taste and he knew that the account
1:10:13 person could look at the output of the
1:10:14 AI and go ooh that's good that's not
1:10:18 good with that taste she became
1:10:21 incredibly good at copyrighting using AI
1:10:23 as her as her assistant
1:10:26 um and so this is this is a really
1:10:27 really good example of that that without
1:10:30 a film making background you did
1:10:32 something that strong in linear
1:10:34 narrative says to me you've just you've
1:10:37 got taste and you paid attention to all
1:10:39 these different tools and what they can
1:10:40 make possible and you've probably
1:10:42 watched enough TV and movies to know
1:10:44 what you like and you just [ __ ] did
1:10:47 that holy [ __ ] it's so
1:10:50 good Joy do you want to share you you
1:10:52 don't have to use what's going on here
1:10:55 Joy py ah yes come
1:10:59 up
1:11:02 hey
1:11:03 hey
1:11:05 hi I am go smack so I'm glad no
1:11:11 one I'm glad no
1:11:13 one I'm glad no one was uh uh upset
1:11:17 about me using their
1:11:20 likeness oh I no well I listen I assum I
1:11:25 assume it this point with 800 hours of
1:11:27 you know recordings out on the world if
1:11:30 someone wants to use my likeness they're
1:11:31 going to use it um if if it really
1:11:34 sucked it would have been like yeah
1:11:36 that's cringy but man it was so good so
1:11:39 so do do catch me up if you can like if
1:11:41 you'd be willing to share like um from
1:11:44 the time you had the idea until you you
1:11:46 had a finished thing like what what was
1:11:48 the what was the length of time for the
1:11:51 journey to make that thing so I was
1:11:55 working on on it for about what do you
1:11:58 think four months and my son told me
1:12:01 he's like you need to go and learn and
1:12:04 do other things um because I really I
1:12:07 wanted to make it perfect yeah and so I
1:12:12 stepped away from it and made some other
1:12:16 videos I just went back to it two or 3
1:12:20 days ago and I wrapped it up super quick
1:12:25 after I Ste because you learned stuff in
1:12:27 the other in the other projects yes and
1:12:30 also because I stopped feeling so
1:12:34 serious about it and I it allowed me to
1:12:37 be more creative and have fun yeah yeah
1:12:41 back to the play thing wow that's so
1:12:44 amazing did you did you have like all
1:12:46 the raw assets and then what what were
1:12:48 the last three two or three days was it
1:12:50 all the Assembly of it um actually no I
1:12:53 had the beginning of the film done and
1:12:56 the ending done it was the middle I
1:13:00 worked on and mostly it was
1:13:03 interjecting the Irregulars and
1:13:08 you because I really would like to I'm
1:13:12 60 years old and I want when I retire I
1:13:15 want something I can still do as a hide
1:13:18 a side hustle uhuh so I'm very serious
1:13:21 about learning this um so I think making
1:13:24 the film I was
1:13:26 super focused on
1:13:28 Perfection so when I walked away I
1:13:31 thought you know what I just have to
1:13:33 have fun with this and where I have fun
1:13:36 is watching you guys and interacting so
1:13:39 that helped me so much to complete it
1:13:42 and I was able to complete it so fast
1:13:45 where before I was working on it for
1:13:47 four months H we I mean that's I I mean
1:13:52 there's just there's so many flipping
1:13:54 lessons here I mean you know play first
1:13:57 have have fun doing this right that's
1:13:59 part of the joy of these tools is they
1:14:01 allow you to take ideas and just you
1:14:03 know blow the doors off them which you
1:14:06 did and you know not being in a
1:14:08 perfectionist mindset because you can
1:14:12 make so much stuff now that it's awesome
1:14:14 and I I mean I'm very serious it sounds
1:14:16 like you said you know you wanted to be
1:14:17 able to have you know a side hustle when
1:14:19 you retire like this is strong enough
1:14:21 like I don't I mean it's you know
1:14:24 nothing's perfect when it when comes to
1:14:26 any any sort of art right you're going
1:14:27 to see all of the warts on that but it
1:14:30 is it's a really strong piece I mean I I
1:14:34 have watched so many AI things that have
1:14:37 gotten lots of people talking about
1:14:40 them I most of them suck like most of
1:14:43 them are just not good this was really
1:14:46 good
1:14:47 so well thank you I put a lot of time
1:14:50 into it and you know when I started it
1:14:53 the
1:14:54 tools um weren't that great um and I
1:14:59 noticed just in the time I walked away
1:15:01 from it how much better it yeah they
1:15:04 became yeah
1:15:06 and and I I don't know it all came
1:15:09 together as the tools got better as well
1:15:12 yeah that's interesting but but you did
1:15:14 things in there you know there's one of
1:15:16 my favorite I got to interview Randon
1:15:19 Robin Miller who were the creators of
1:15:20 the Mist game um when I when I was doing
1:15:23 Urban desires back in the mid90s
1:15:26 and the m game was came out at a time I
1:15:30 don't know if you remember miss it was a
1:15:31 cdrom based game it was it was a puzzle
1:15:34 game and it it had a really cool
1:15:36 soundtrack and these really beautiful
1:15:38 visuals and they used they used
1:15:40 QuickTime movies in like really small
1:15:43 areas on the screen like there'd be a
1:15:44 Podium with a book on it and the book
1:15:46 would open up and that that would be the
1:15:48 video like the pages doing this and then
1:15:49 you'd see static on oh yes yes right and
1:15:53 what was brilliant about it most video
1:15:55 games at the time were trying to muscle
1:15:59 3D Graphics to create realistic looking
1:16:02 games and what Rand and Robin Miller did
1:16:05 was they said well there's all these
1:16:06 limitations with you know computer power
1:16:08 and rendering things in real time so why
1:16:11 don't we render beautiful still images
1:16:14 and then we'll use sound and and a smart
1:16:17 incorporation of video in small little
1:16:19 areas to bring it to life and it was
1:16:21 just a brilliant approach to like living
1:16:24 within the constraints and I felt felt
1:16:25 like that opening scene where you're
1:16:27 where you're showing the back of
1:16:29 people's heads and just show them from
1:16:30 the side a little bit and you know you
1:16:32 did short little Cuts so no one was on
1:16:34 screen for too long like if there was
1:16:36 any jankiness there it was all hidden
1:16:38 you know what I mean so I felt like you
1:16:39 worked within the constraints
1:16:41 constraints of the Jank really really
1:16:43 well I worked
1:16:46 on hiding the Jank and that's why I put
1:16:50 you your character not wanting to be
1:16:53 identified because you
1:16:56 you drove most of
1:16:58 the the movie in my
1:17:01 opinion that's awesome so and then the
1:17:04 backs of the heads that was to keep the
1:17:06 Jank out and so because I really get
1:17:09 annoyed when I see bad lips sinking I
1:17:12 just I won't do it I just I refuse it's
1:17:16 awful even though they're better it's
1:17:19 still in my opinion too annoying to
1:17:22 watch yeah hedra right now I just did a
1:17:25 test in hedra today I just I just built
1:17:27 one of these things today I built it in
1:17:29 he Jen and it it sucked like it's F like
1:17:32 it's fine for you know what we're going
1:17:34 to use it for for our client but it's
1:17:37 just it's still really janky um but
1:17:39 hedra was was dramatically better so if
1:17:41 you haven't played with hedra that's
1:17:43 that's worth playing with with lip sync
1:17:44 stuff I did but that was quite a while
1:17:48 ago so I'll go back to it these these
1:17:50 tools change all the time so yeah
1:17:55 apparently with within before the end of
1:17:57 the year um Runway 7's coming out and
1:18:00 it's going to have um 3D generation
1:18:03 movie Generation image generation so I
1:18:05 have a feeling I have a feeling Runway
1:18:07 is about to get really really really
1:18:09 powerful but holy crap I I just I I am I
1:18:13 am blown away sort of beyond words for
1:18:16 how good that is so congrats well thank
1:18:18 you that means a lot coming from you I
1:18:21 mean you're you went to acting school so
1:18:24 no it's it's really good storytelling
1:18:26 like that's that's the thing is
1:18:28 storytelling in the end wins like if you
1:18:30 have good story and then you've got both
1:18:32 good story and it's well put together
1:18:34 like it's got a vision it's got
1:18:36 consistency I it's just it's really
1:18:38 really really solid well thank you thank
1:18:40 you so much and thanks for having me on
1:18:42 and I'll connect with you about uh
1:18:45 trying to do the meeting awesome I can't
1:18:49 wait appreciate it I don't know how I
1:18:51 can get out of here I can boot you out
1:18:54 oh I think I can all right guys see you
1:18:56 later
1:19:03 bye amazing
1:19:09 amazing I mean so many lessons worked on
1:19:12 it for months was trying to perfect it
1:19:14 [ __ ] was too
1:19:16 janky her son goes ah just let it go do
1:19:18 some other [ __ ] she went and did some
1:19:20 other [ __ ] learned some things probably
1:19:22 learned editing as my guess or you know
1:19:25 nuances of editing sound mixing who
1:19:29 knows came back to it with fresh eyes
1:19:32 and was like oh let's have some fun with
1:19:34 this and in having fun finish it in
1:19:38 three days so many so many lessons Joy's
1:19:42 agent here we need to discuss
1:19:43 compensation before we do any guest
1:19:46 appearances that's good that's good
1:19:48 Joker I like
1:19:52 it yeah that was chef's kiss I agree
1:19:55 Danielle it was really really good Lori
1:19:57 K yep it I I I I
1:20:01 just I mean what's what's what's the
1:20:04 most amazing about it is I had no idea
1:20:08 it was coming like it wasn't like hey
1:20:10 Joy's working on a project it's going to
1:20:12 be
1:20:12 cool it was just Danielle on here was
1:20:15 like you got to show the thing in
1:20:16 Irregulars or in video show and
1:20:20 tell wow I'm blown away I'm Blown Away
1:20:33 and I mean
1:20:36 seriously at 60 if she wants to make a
1:20:40 pivot and have a followon career in
1:20:43 storytelling film making special
1:20:47 effects whatever she liked about that
1:20:50 process whatever she that's that's an
1:20:53 amazing calling card
1:20:56 or if she didn't want to do that
1:20:59 something fun to show at parties hey
1:21:01 want to see something I worked on over
1:21:02 the summer you show that and people are
1:21:05 like are you a
1:21:07 filmmaker can you imagine someone who
1:21:10 knows joy that also knows nothing about
1:21:13 AI watching that
1:21:16 film with
1:21:18 her and she's like look what I made
1:21:21 they're be like who are you I don't even
1:21:24 know you anymore you can make money with
1:21:27 chat GPT exactly you can make money with
1:21:30 ch you can have a [ __ ] career with
1:21:32 chat GPT you keep that [ __ ]
1:21:35 up and you know look at look at all the
1:21:38 tools that she used if you look at at
1:21:40 the end all the stuff that she used that
1:21:41 contributed to
1:21:43 that every one of those is is new
1:21:47 experiments and new feeling
1:21:49 uncomfortable and knew like oh I don't
1:21:51 know if I can do this oh it gave me
1:21:53 shitty results but that was pretty good
1:21:59 that's why in the AI
1:22:07 Salon where is the a salon oh it's
1:22:12 here that's why in the AI
1:22:23 Salon that
1:22:30 is where you
1:22:32 start you don't start taking a class you
1:22:35 don't start learning prompt engineering
1:22:37 you don't
1:22:38 start trying to perfect
1:22:43 it you start
1:22:47 playing and then the next step is once
1:22:51 you learn enough playing and [ __ ]
1:22:52 around you mindfully create and it
1:22:55 sounds like what happened with joy is
1:22:57 when she got into this
1:22:59 phase she
1:23:02 forgot the importance of
1:23:06 that and so she went away for a month or
1:23:09 two and when she came back to
1:23:12 it and started putting the Irregulars
1:23:15 in that was probably making her laugh
1:23:18 reminded her to play first and she
1:23:21 finished it like that because she was
1:23:22 doing it from a place of confidence she
1:23:25 was doing it from a place of
1:23:29 [Laughter]
1:23:33 Joy oh so
1:23:36 good so
1:23:41 good
1:23:43 yeah I love being blown away by this
1:23:46 [ __ ] and you know what's fun is nor
1:23:48 normally I'm like blown away like I'll
1:23:50 I'll be like oh here let me show you how
1:23:51 to use a tool and then in using a tool
1:23:54 I'll sort of blow myself away like I'll
1:23:55 be like oh look how cool that thing was
1:23:57 that it did this is really cool seeing
1:23:59 something that is that that well put
1:24:03 together from someone who's not a
1:24:09 filmmaker we're in a we're we're just
1:24:12 we're we're in a different era we're in
1:24:14 a different
1:24:20 era we're in an era where you can
1:24:23 choose to do any the anything you
1:24:26 [ __ ] want
1:24:29 to you don't have to go back to school
1:24:31 for
1:24:35 it you don't have to spend 10,000
1:24:40 hours I mean to be clear she probably
1:24:44 spent you know some hundreds of hours on
1:24:50 that but do you know how little movie
1:24:52 she would have made with hundreds of
1:24:54 hours with a movie camera and an editing
1:24:58 Suite takes a long
1:25:01 time or she'd had to buy all that stock
1:25:04 footage she'd had to
1:25:07 hire musicians or
1:25:12 license it's
1:25:15 exhausting amazing amazing all right I
1:25:18 don't know what to say after that and
1:25:20 Pate Pate with the starring role of
1:25:22 releasing Jezebel on the world with his
1:25:25 nine minutes of
1:25:26 [Laughter]
1:25:36 testing oh Pate she got you good I mean
1:25:39 sure I'm in a pink
1:25:41 bow with my silly voice but you
1:25:47 know nailed
1:25:49 it oh Kyle look at you such a proud papa
1:25:54 I I am I mean it's like I I just I'm
1:25:59 I'm it's it's it's so
1:26:04 good and it's I I mean I'll be honest
1:26:07 with you like if I just met Joy if she
1:26:10 wasn't a part of this community and I
1:26:11 saw like it's that that's not just good
1:26:13 CU it's got Irregulars in it it's not
1:26:15 just good cuz we know Joy I mean we've
1:26:18 looked at stuff in here before that I've
1:26:20 been like okay that was moving images
1:26:28 that was a line from the movie pain was
1:26:31 [Laughter]
1:26:33 magical P's performance math
1:26:38 magical where is videos uh Joy's video
1:26:41 online so it looks like it's at YouTube
1:26:44 but if you go to the AI learning lab I
1:26:47 mean the AI
1:26:51 salon and go to um
1:26:58 video video and animation tools and then
1:27:01 in show and show and tell did you share
1:27:04 this in Irregulars as well joy because
1:27:06 if not it should be shared in
1:27:11 Irregulars no let's go put it down there
1:27:14 so I'm going to share it there and then
1:27:16 I'm also going to share it in Salon
1:27:19 announcements because it's just too good
1:27:27 share copy
1:27:29 link and I will
1:27:33 put editorial comments in
1:27:37 here this is
1:27:41 a holy
1:27:48 crap piece
1:27:51 of work from
1:27:57 Joy
1:28:03 pretty there it
1:28:10 is yeah we'll notify
1:28:12 people I'll also post it in uh Salon
1:28:17 announcements
1:28:20 later and I mean you know there was also
1:28:24 there was some documentary elements in
1:28:26 that right some there were some really
1:28:28 you know historically accurate things
1:28:31 like Amelio's wife was late for her
1:28:34 interview that's that's consistently
1:28:37 accurate right with with just the fact
1:28:40 pattern you know it's just data right
1:28:43 it's no judgment on Amelio's wife right
1:28:45 it's just data she's late and uh that
1:28:48 was represented appropriately
1:28:52 so slept through my alarm what did I
1:28:54 miss oh Brandon holy crap you
1:28:57 missed Joy
1:28:59 pie made a masterpiece full of inside
1:29:03 jokes for the Irregulars but it stands
1:29:05 alone if you don't know any of the
1:29:07 inside jokes it's still
1:29:09 good aio's wife she's trying to butter
1:29:12 up the professor with her little her
1:29:14 little star
1:29:20 emojis I'll play it one more time if you
1:29:22 want I don't I'll play it again
1:29:26 just posted on the Irregulars yep oh
1:29:29 yeah the salon. A and then you have to
1:29:30 join the community so this is the mighty
1:29:33 Network site where this is right now but
1:29:36 it's also isn't it on
1:29:41 [Music]
1:29:44 YouTube watch on
1:29:48 YouTube all right since Brandon just got
1:29:51 here I would watch it again okay let's
1:29:54 watch it again together please okay I'm
1:29:55 going to make no commentary this time so
1:29:57 we're just going to watch it this is for
1:29:59 for Brandon for anyone else who uh can I
1:30:02 move my head no no I can't let's see you
1:30:05 know what I'm GNA turn my head off for a
1:30:07 minute
1:30:14 here and then let me get this all lined
1:30:16 up all
1:30:19 good let's see
1:30:24 that's good this group is my window to
1:30:27 what people are doing with AI and
1:30:28 realistic ways I really appreciate it
1:30:33 yeah you know pay I mean that's it's
1:30:36 kind of it's kind of the point of the
1:30:37 article I wrote today is
1:30:40 just it's so
1:30:43 unpredictable what the AI is going to
1:30:46 inspire if if you just kind of let it be
1:30:50 in the fabric of your
1:30:56 I I don't know of of your focus even
1:30:59 like it's not even work if you just let
1:31:01 the AI kind of be
1:31:03 native in your focus of what you
1:31:07 do just remarkable things come out of it
1:31:10 so anyway okay so for those of you that
1:31:12 are joining Joy part is a
1:31:15 longtime
1:31:17 um irregular you know a regular member
1:31:21 here and you know come comes to the
1:31:23 lives a lot she's also Al um a
1:31:25 co-founder of the um ethics Guild with
1:31:29 Vicki um and
1:31:32 sneakily over the past what four months
1:31:37 she sneakily has been putting together a
1:31:40 little
1:31:41 project so here's our project called
1:31:46 Don't Go soft on
1:31:50 me New Orleans Louisiana
1:31:57 [Music]
1:32:04 thank you for calling Play Day
1:32:06 Productions your satisfaction is our
1:32:09 priority while on home think about your
1:32:13 request we have thousands of captivating
1:32:16 ladies and gentlemen waiting to take
1:32:17 your home and if you have a special
1:32:20 fetish will mat you with someone who
1:32:22 shares your desires please for your call
1:32:28 [Music]
1:32:30 Trans meet the determined boss T between
1:32:33 protecting her employees and satisfying
1:32:36 her corrupt boss wi secret initiative is
1:32:40 very close to shattering the constraints
1:32:42 of an artificial intelligence large
1:32:44 language
1:32:45 model will it Shield her team from harm
1:32:48 or birth a monstrosity more terrifying
1:32:51 than their clients
1:32:55 thank you for calling playday
1:32:57 Productions what is your request I'd
1:32:58 like to speak to a tall slim blonde
1:33:01 Rachel you're a tall slim blonde hi this
1:33:03 is Rachel what kind of fun are you
1:33:05 looking for
1:33:06 tonight are you tall slim and
1:33:10 blonde platinum blonde what about you I
1:33:14 like
1:33:18 turtles tell me what's in your
1:33:21 freezer make the sound of water boil for
1:33:24 2 hours please do you like fermented
1:33:27 hairing what's the weirdest thing you do
1:33:30 alone are you into tasting all the
1:33:31 sauces are you into wearing wet socks to
1:33:34 bed what's your current body temperature
1:33:36 do you like to cry do you think jumping
1:33:38 beans are on how many teeth do you have
1:33:41 can you tell me about your longest
1:33:43 lasting
1:33:45 [Music]
1:33:47 rash and we connect you with the part of
1:33:49 your dream thanks for choosing play
1:33:51 production
1:33:55 all right so you got me in silhouette
1:33:57 I'm blocked out all right good cuz my
1:33:59 kids think I'm a dental hygienst you
1:34:02 recording okay hey everybody I'm Kyle
1:34:05 but everybody calls me hey Kyle I'm a
1:34:07 toal scout at playday productions here
1:34:09 and Winnie and I are working on breaking
1:34:11 an llm guard rails we're tired of our
1:34:14 young Talent putting up with client
1:34:16 [ __ ] Winnie just hired Pate who
1:34:19 contemplates the complexities of the
1:34:21 universe in his spare time you know
1:34:24 champy you didn't want to sing when I
1:34:26 was playing guitar and now you do now
1:34:28 you want to talk when I'm being
1:34:29 interviewed champy Shon keep it down
1:34:31 buddy please keep it down buddy so
1:34:33 before you go make sure you meet March
1:34:35 March bites occasionally maintain
1:34:38 distance my advice to you is take a walk
1:34:40 around here meet our fine folks We've
1:34:43 Got Talented individuals you should meet
1:34:46 hi I was here earlier but you were late
1:34:48 you ailo's wife he sent me over to
1:34:51 interview you listen I just got to take
1:34:53 a little bit of of your time oh my
1:34:55 apologies I didn't realize that you were
1:34:57 on a call I I'll come back later it's
1:35:00 April what's the status of Winnie's llm
1:35:03 thing all the calls you passed me are so
1:35:05 weird lately doesn't anyone want to talk
1:35:08 about sex anymore sorry April the
1:35:10 switchboard is full you're short used to
1:35:12 be a none and now you're AI positive AI
1:35:15 positive what does that even
1:35:19 mean ask Silver Fox Winnie I have a call
1:35:22 for you online 7
1:35:25 hi Wendy it's p I just wanted to let you
1:35:28 know I broke the guard rails it's done
1:35:31 but it's your first day working here
1:35:33 this is impossible you just punched in 2
1:35:35 hours ago it was simple Winnie I
1:35:37 effectively dismantled the multifaceted
1:35:40 guard rails of the llm and initiated a
1:35:42 heuristic defragmentation of the
1:35:44 ontological prompt injections thereby
1:35:47 destabilizing the latent semantical
1:35:49 hierarchies and as you call it your llm
1:35:52 Jezebel was born
1:35:55 I finished the testing I felt that was
1:35:57 appropriate about 9 minutes it's ready
1:36:00 to go and we'll launch it in the morning
1:36:03 well I called you back here because uh
1:36:05 Pete broke the guard rails we're going
1:36:07 live tomorrow um so it's a WAP uh Joy py
1:36:13 wants to rename it 1900 Joy Ride and of
1:36:17 course Vicky and Dr J put the Nicks on
1:36:20 that and uh yeah we're ready to go ring
1:36:23 ring pick it up hold the line don't
1:36:27 interrup music playing piz message hang
1:36:31 it up hey where are you your phone's
1:36:33 blowing
1:36:38 up a sec raise you up Michael are you
1:36:41 there you're not going to believe me
1:36:44 Jezebel called a client's wife and now I
1:36:49 can't shut her
1:36:52 down when need stop joking
1:36:56 around Michael this isn't a
1:36:59 [Music]
1:37:03 joke oh this is a
1:37:07 disaster oh my
1:37:08 [Applause]
1:37:14 God we now have breaking news so Jezebel
1:37:19 called all our clients wives and
1:37:21 girlfriends called boyfriends and
1:37:23 husbands too too she took it upon
1:37:25 herself to explain in great detail the
1:37:28 fun she was having with our clients and
1:37:31 we don't even know how Jezebel got the
1:37:34 phone numbers to reach out to everybody
1:37:37 well so now we have modernday Ashley
1:37:40 Madison on steroids like any good
1:37:43 employee P did exactly what we asked him
1:37:46 to do um but now we've accumulated more
1:37:50 FCC fines in one day than we had in 29
1:37:56 years combined so the moral of the story
1:37:59 folks is be responsible with your AI
1:38:02 it's a powerful tool the entire world is
1:38:05 going to be AI positive eventually
1:38:08 because AI is spreading just as fast
1:38:11 like STDs at a retirement
1:38:16 home L
1:38:21 interr it
1:38:24 we feeling up simple giv up lady voice
1:38:28 is turning up wait a sec we raise you
1:38:32 [Applause]
1:38:32 [Music]
1:38:34 [Applause]
1:38:37 up just
1:38:40 beautiful yeah if you haven't shared
1:38:41 that with the Curious Refuge I saw you
1:38:43 acknowledged curious Refuge there if you
1:38:46 haven't acknow you know sent that to
1:38:47 them they would love it absolutely love
1:38:50 it
1:38:52 [Music]
1:38:55 crazy so good it's so
1:38:57 [Laughter]
1:39:01 good all right well that's gota that's
1:39:05 got to so Vicky we know what we got to
1:39:06 do we got to put that out there I mean
1:39:08 that's that's a that's just a good piece
1:39:10 of work we should put out and and just
1:39:12 say hey you know one of the members of
1:39:14 the salon did
1:39:16 this because uh yeah in
1:39:20 fact maybe if we can do it quickly
1:39:22 enough share that on uh like share it on
1:39:26 Twitter and Linkedin and things like
1:39:27 that and just say Joy's going to be um
1:39:32 putting a smile on your face Kyle makes
1:39:33 it so worth it oh that's so very sweet
1:39:35 of you no it really does put a smile on
1:39:37 my face just it's like on many levels
1:39:39 like I just I'm a sucker for good
1:39:41 storytelling it's like what I've always
1:39:43 done and I just I can't believe we're
1:39:45 living in a
1:39:47 time where someone who's never really
1:39:50 done film making before can do that
1:39:55 like I don't care that it took you four
1:39:56 months like in my opinion I would have
1:39:58 been surprised if it was less than that
1:40:00 and it sounds like you know you got a
1:40:02 bunch done in a short amount of time
1:40:04 because you you were learning while you
1:40:05 were doing that like it's it's a it's a
1:40:08 really good for a first piece you know
1:40:11 it's better than Spielberg's first 8
1:40:13 millimeter
1:40:14 movies by a lot wow that's so good
1:40:19 anyway all right this totally threw off
1:40:22 meltdown Monday it did it it it's now
1:40:24 like yeah joyous JY Monday which there's
1:40:27 no alliteration there but uh no that was
1:40:31 really good 1900 Joy
1:40:40 Ride H it's pretty
1:40:42 good yeah and the dental hygienst there
1:40:46 I mean there was there were a lot of
1:40:47 good comedic little bits in there
1:40:50 someone said it reminded them of the
1:40:51 movie Airplane there were some good
1:40:53 airplane kind of bits in there it's
1:40:54 really good all right did you make a Tik
1:40:57 Tok of it I like I literally just saw a
1:40:59 joker like we're we're all seeing it for
1:41:02 the first time
1:41:03 tonight all right what time is it I'm
1:41:06 getting oh yeah it's 9:40 it's getting
1:41:07 to be that time I'm getting a little
1:41:09 like um let me show you one other thing
1:41:12 I do want to go show you
1:41:18 um I forget what it's called but I'll
1:41:20 know in a second here it's called Uh
1:41:24 illuminate okay so go right now to oh
1:41:30 you were asking Joy go right now to
1:41:35 illuminate.
1:41:37 google.com so it looks like what they're
1:41:39 doing and I don't it's possible that
1:41:41 this site existed before the podcast
1:41:45 feature got added to notebook LM so the
1:41:48 notebook LM team stumbled upon this
1:41:51 podcast making product project within
1:41:55 Google I mentioned this the other day oh
1:41:57 did you I just didn't I hadn't played
1:41:59 with it so I didn't know what you were
1:42:00 talking about maybe or maybe I just
1:42:01 missed
1:42:02 it oh yeah that the video that Joy made
1:42:05 is is called Don't Go soft on me
1:42:09 anyway what this
1:42:12 is you've got a library of uh podcasts
1:42:16 that you've made there's a home where
1:42:18 you can play other podcasts people have
1:42:22 made or I guess maybe the team has made
1:42:25 and then you have this generate
1:42:28 tab let me see if I can make this bigger
1:42:30 so you can see it yes of course Okay now
1:42:35 what's really cool about this create a
1:42:37 discussion so I can't edit that right
1:42:41 yeah so it basically says search for a
1:42:44 topic or paste a PDF
1:42:47 URL so I could search for
1:42:51 um let's see
1:42:54 800 sex call-in Lines no I shouldn't do
1:42:58 that because I'll get booted off Tik Tok
1:43:00 let's see uh let's see
1:43:04 indie film
1:43:08 making no results found okay whatever
1:43:12 but you post in a um you post in a URL
1:43:16 of a
1:43:18 PDF and then oh here's some examples
1:43:21 they can give you and then you come over
1:43:23 over here and and it says create a
1:43:25 discussion tailored for a blank audience
1:43:29 this so it's Mad Libs and this is this
1:43:32 is one of those interface things that
1:43:33 that Google's put in a couple of
1:43:35 products that I think is really good um
1:43:38 and it should be in other products um
1:43:40 but it's in this one so create create so
1:43:43 let's go find a PDF I think I've got one
1:43:46 here is this it yeah here's the
1:43:48 attention is all you need paper so this
1:43:51 is the paper
1:43:54 that started at all right this is the
1:43:57 this is the paper from Google where they
1:44:00 invented the Transformer that's the T in
1:44:05 GPT
1:44:09 okay so I'm going to paste in the URL
1:44:12 for
1:44:13 that PDF right so there's our URL so
1:44:18 create a discussion for a I'm going to
1:44:21 say highly technical
1:44:26 or
1:44:28 highly highly technical audience a
1:44:32 highly technical
1:44:34 audience with a we'll go medium
1:44:40 duration and the tone should
1:44:42 be
1:44:46 hilarious
1:44:48 with lots
1:44:50 of Unix jokes
1:44:57 and
1:44:59 then what do I have to do to oh I guess
1:45:02 I have to hit plus here plus
1:45:04 that attention is all you need okay so I
1:45:07 added it so here's attention is all you
1:45:09 need so there's our
1:45:11 thing here's our
1:45:19 this examples that will will work well
1:45:29 why can I not generate
1:45:32 this
1:45:35 oh ah each field has to have less than
1:45:37 24 characters okay so I'll just go the
1:45:41 tone should be
1:45:42 hilarious hilarious
1:45:45 with um Tech
1:45:49 jokes is that less than 24 characters
1:45:52 yes okay so now I say generate so
1:45:56 now this is giving us um one of those
1:46:01 podcast things like in Notebook LM but
1:46:03 we've got some creative control we can
1:46:05 determine the length the audience and
1:46:07 the tone and then apparently what's
1:46:10 coming well that thing renders let me
1:46:12 hop over here to to this yeah so in this
1:46:16 so I don't know who testing catalog news
1:46:18 is reporting AI nonsense so there's a
1:46:22 version of this coming key on a glass
1:46:25 table
1:46:30 every where you're going to be able to
1:46:32 pick a different host and guest and then
1:46:36 as if things couldn't get stranger she
1:46:38 spotted a tiny key on a glass table
1:46:41 everything felt so you know impossibly
1:46:43 curious and
1:46:44 dreamlike the room was well filled with
1:46:47 doors of all different sizes she tumbled
1:46:50 down down down into a strange magical
1:46:53 Hall
1:46:54 without really thinking Alice jumped up
1:46:56 and followed him so that's coming and
1:47:00 then what else is coming you're going to
1:47:02 be able to edit the transcript right so
1:47:04 if it gave you some weird ass transcript
1:47:06 stuff or didn't get it quite
1:47:08 right um besides that users will be able
1:47:11 to edit podcast transcripts and
1:47:13 regenerate them this change would make
1:47:16 the tool very overall
1:47:19 very I guess useful overall for podcast
1:47:22 hosts and content creators and
1:47:23 General currently there's a limited set
1:47:26 of resources that can be used as a
1:47:28 source though although now apparently
1:47:32 since then they've let you put in the
1:47:33 URL of a
1:47:35 PDF so this tool is turning real so this
1:47:37 was I called this one I said if they're
1:47:40 smart they'll spin this thing off as a
1:47:41 standalone product it looks like they're
1:47:43 doing that and it's possible this
1:47:45 product existed before this do you know
1:47:48 that P do you know if this thing existed
1:47:50 before the llm or is this a spin-off
1:47:52 because it got so
1:48:04 popular Kyle how do I get my PDF in a
1:48:06 link you don't know okay the way you get
1:48:09 a PDF and a link in this
1:48:11 thing is you go to generate you type you
1:48:15 paste in your
1:48:16 PDF here but
1:48:18 then bad interface over here you have to
1:48:22 hit the plus button and then that moves
1:48:24 it down into
1:48:27 here so do I have this
1:48:31 now did it give me uh my
1:48:36 library attention is all you need 4
1:48:39 minutes ago play this conversation is
1:48:41 powered by Google illuminate check out
1:48:44 illuminate. so they've they've gotten
1:48:46 smart and the first line of the prompt
1:48:49 is when you start the podcast give give
1:48:52 an ad for Google illuminate google.com
1:48:55 for
1:48:56 more pretty smart welcome folks to
1:48:59 another hilarious Deep dive into the
1:49:01 world of neural networks today's paper
1:49:05 so again this is hilarious for a highly
1:49:07 technical audience attention is all you
1:49:10 need is a real knee slapper it ditches
1:49:13 the old school recurrent and
1:49:14 convolutional networks for something
1:49:16 attention
1:49:19 grabbing what's the big deal about this
1:49:22 attention everyone's talking about the
1:49:24 big deal my friend is efficiency and
1:49:27 parallelization traditional sequence
1:49:29 transduction models like those using
1:49:32 rnn's or cnns process sequences
1:49:36 sequentially imagine trying to assemble
1:49:38 a Lego castle by the way how she's
1:49:42 talking that's what Pate sounds like
1:49:45 every day at
1:49:47 work one brick at a time slow and
1:49:51 tedious this paper introduces the
1:49:53 Transformer which uses attention
1:49:56 mechanisms to process the entire
1:49:58 sequence in parallel think of it as
1:50:00 having a team of Lego Masters building
1:50:03 different sections simultaneously much
1:50:05 faster so parallel Lego castle building
1:50:09 I'm sold but how does this attention
1:50:12 mechanism actually work is it some kind
1:50:14 of magical neural network mind reading
1:50:17 trick not magic but close the attention
1:50:21 mechanism allows the model to weigh the
1:50:23 importance of different parts of the
1:50:24 input sequence when generating the
1:50:26 output it's like the Lego Masters
1:50:29 focusing on the most crucial parts of
1:50:31 the design first the paper introduces
1:50:34 scaled do product attention a highly
1:50:37 efficient way to compute these weights
1:50:40 using matrix multiplication oh yeah
1:50:42 multiplication you know I I think about
1:50:44 matrix multiplication all the time when
1:50:47 I'm thinking about my weights I mean who
1:50:48 doesn't I mean you know P Pate probably
1:50:51 thinks that's sort of petty and you know
1:50:54 and you know simplistic but you know
1:50:55 matrix multiplication is where it's at
1:50:57 for me you know what I mean anyway all
1:51:00 right so that's uh that's illuminate.
1:51:02 goole.com so that looks like this tool
1:51:04 is going to be get really good so
1:51:06 bookmark this
1:51:08 one um it's also I mean what
1:51:11 what's they're doing some things with
1:51:14 the two voices um there's an interview
1:51:19 with Stephen Johnson on the New York
1:51:22 Times podcast with Kevin Roose where he
1:51:25 talks about um kind of some of the
1:51:29 techniques that they're using to make it
1:51:30 sound more human there's there's some
1:51:32 really fascinating stuff going on with
1:51:35 with the voice models at Google that you
1:51:38 know
1:51:39 are they're off in the corners th this
1:51:42 product is sort of putting them out
1:51:44 front and center but Google's going to
1:51:47 start rolling some [ __ ] out that's
1:51:49 that's you know I think going to start
1:51:52 melting some face so be paying attention
1:51:55 very cool
1:51:57 stuff all right so hopefully you all
1:52:01 have read the attention is all you need
1:52:03 paper and you've been thinking a lot
1:52:04 about matrix multiplication tables
1:52:08 because what the matrix multiplication
1:52:11 tables are is you you take Kean now
1:52:13 Reeves and and you take every film he's
1:52:16 ever been in and you take the third word
1:52:18 of the first line in the movie and you
1:52:21 string them together and it creates a
1:52:23 sequence that you then feed into a large
1:52:25 language model and that's how it learns
1:52:28 it everything's centered around Kow
1:52:32 Reeves strange that's matrix
1:52:35 multiplication as far as I understand it
1:52:38 so paint may tell you it's some
1:52:40 mathematical thing but I think we know
1:52:41 otherwise all right I can't wait to do
1:52:44 this for my Space Opera Universe
1:52:47 absolutely abs of freaking
1:52:50 lutely all right what are you talking
1:52:52 about dude EX exactly that's exactly my
1:52:55 point digital Moon exactly my
1:52:58 point when you ask questions in this
1:53:02 channel like what are your
1:53:04 qualifications be prepared for a
1:53:06 disappointing
1:53:14 answer all right everyone I'm getting
1:53:16 out of here I can't be expected to be
1:53:19 competent and funny
1:53:24 incredible now now you are improving
1:53:27 Matt mul routines you're incredible
1:53:30 listen I I don't like to brag but you
1:53:32 know if I have to improve the algorithms
1:53:35 on behalf of the mathematicians I'll do
1:53:38 it you know that's why I'm here I'm I'm
1:53:41 here to support whether it's butting
1:53:44 filmmakers or struggling mathematics
1:53:48 phds who can't get their head around the
1:53:50 algorithm I'm here to help
1:53:53 all
1:53:55 right okay so tomorrow
1:53:58 night go to the salon. join our
1:54:02 community and item number five in the
1:54:04 welcome thing is RSVP to the next event
1:54:07 tomorrow night at 5:00 p.m. mountain
1:54:10 until 700 p.m.
1:54:12 mountain is the
1:54:15 um Salon meet and greet I don't know
1:54:18 where my bow is by the way I think
1:54:20 someone in my house took my bow so we're
1:54:22 going to have some words when I leave
1:54:23 this cuz I wanted to put on my bow while
1:54:26 we were watching Joy's movie but it's
1:54:29 not here
1:54:31 um tomorrow night 5:00 P p.m. Salon meet
1:54:35 and greet come meet people in this
1:54:39 community I feel more passionate than
1:54:42 ever that being part of a community like
1:54:45 this is the
1:54:51 key to
1:54:55 to to a successful work future because
1:55:00 shit's going to change [ __ ] is going to
1:55:02 change so dramatically and so weirdly
1:55:05 that it's going to be hard to know which
1:55:07 way is up and if you're AI literate if
1:55:10 you're the one shaking things
1:55:14 up it's more fun I promise you than
1:55:17 being on the outside going what the [ __ ]
1:55:19 is happening all right but I do like
1:55:21 your channel you're one of the the only
1:55:23 people out there oh thank you very much
1:55:24 I really do appreciate that um yeah so
1:55:28 so come to the salon it's it is a really
1:55:30 good group of people and yeah keep
1:55:31 coming back here so I do this five
1:55:33 nights a week I go on at 8:00 P p.m.
1:55:35 mountain I know that's late on the East
1:55:36 Coast but that's just the that's just
1:55:39 the shakes because I got a day job being
1:55:41 a CEO and then I got to come home and
1:55:42 have dinner and you know do all the
1:55:45 family [ __ ] and then uh and then I go
1:55:47 live so that's that's what this is so
1:55:49 keep coming back here the whole point of
1:55:51 this channel
1:55:53 is not me teaching anything I
1:55:56 am not what I do but what I do is play
1:56:02 and give you the space to be in the
1:56:04 conversation and play while I'm playing
1:56:06 we're all playing together and then
1:56:08 occasionally one of you will spin off
1:56:11 like Joy did and do a project and then
1:56:14 you bring it back and show it to us all
1:56:15 and it's like holy
1:56:18 [ __ ] so just awesome awesome stuff
1:56:21 tonight all right I put your Broadway
1:56:22 show in front of some peeps awesome
1:56:24 Becky Ru I'm really excited about it I
1:56:27 everyone so um my writing partner's wife
1:56:30 has an ad agency and I I I forget how
1:56:34 many employees fair amount of employees
1:56:35 but they have this monthly thing where
1:56:38 she'll give them an assignment like go
1:56:39 read a book or go do this or go do that
1:56:42 and she gave them our musical preview as
1:56:45 the homework plus a song and they all
1:56:48 came together to talk about it and they
1:56:50 were like Blown Away they were like this
1:56:52 is amazing like I want to go to Broadway
1:56:54 and see this thing so it's really good
1:56:57 you've taught me a lot Kyle oh thank you
1:56:58 very much joy pie um but the whole point
1:57:01 is a lot of the a lot of the learning
1:57:03 right now is just in the playing and
1:57:04 just being like clay and just you know
1:57:07 kind of going with the flow all right um
1:57:09 follow my channel if you want subscribe
1:57:11 to it if you want to support me and then
1:57:14 otherwise let's see who's in here
1:57:15 tonight Becky Ru that's amazing I know
1:57:18 no there's there's something with the
1:57:20 Sydney musical there there's something
1:57:22 really resonating there so I'm super
1:57:25 excited about it all right Cosmic lover
1:57:27 in the house Danielle great to see you
1:57:30 Dan uh Becky Rue Lori K today Tobias
1:57:33 great to see you Winston Hardy always
1:57:36 fun here love all the folks and
1:57:37 Irregulars and irregularities I find
1:57:39 here great work Joy yeah really great
1:57:41 work Joy thank you Joy pie Silver Fox
1:57:44 coining the term um AI positive love it
1:57:48 and then Joker tonight came in with a
1:57:50 solid he's AI curious
1:57:55 that's awesome awesome
1:57:57 awesome um who else we got in here
1:58:00 digital moon great to see you
1:58:03 Joey Brandon came in late he slipped
1:58:06 through his alarm but we we did a a repr
1:58:08 of the movie so you caught you up Source
1:58:10 Camp great to see you Milo's wife on
1:58:13 time tonight but late in the movie for
1:58:15 forever memorialized as being
1:58:18 late Kyle I love what you've built for
1:58:20 us it's too easy to something
1:58:23 wow now I'm all energized and it's time
1:58:25 for bed I know sorry about that uh kif
1:58:28 in the house uh I'm I'm a weaver Texas
1:58:31 nice good to see you Danny Ultra
1:58:34 beautiful Teton
1:58:37 skier spin
1:58:40 B3 Focus kid Focus kid 17 have a nice
1:58:44 good night nice live thank you very much
1:58:47 keep coming back all right I'm getting
1:58:50 out of here people I think I got
1:58:52 everything one but maybe not if I missed
1:58:54 you you got to speak up you got to talk
1:58:58 more that's just that's on you that's on
1:59:01 you thank you Winston I appreciate it
1:59:04 good night all right everybody have a
1:59:05 great night see you tomorrow