Jun 24, 2024 - Creativity: How AI Transforms Ideas into Reality

Description

Chapters

Transcript

0:15 [Music]
0:25 oh get the dust bunnies
0:29 out all right here we go you ready stop
0:33 are you ready to sing what' you have in
0:35 there
0:45 [Music]
1:11 [Music]
1:28 y
1:44 [Music]
1:59 standing between
2:02 you in a hard place is
2:05 [Music]
2:07 insane standing too
2:11 near you and a fire makes it
2:17 clear your trouble to
2:23 me Rich trouble can't you say
2:30 leaning in
2:32 clothes smell of your perfume scares me
2:39 most leaning
2:42 away your feels father every
2:49 day your trouble to
2:55 me Rich trouble can't you say
3:08 no more singing champy
3:33 you and I here all
3:37 alone Sunday morning here at
3:41 home Sky's blue and the scarf is strong
3:45 it's
3:49 true then I open my eyes to dream
3:53 realize in front of
3:57 me and I haven't got a clue what in the
4:00 world is happening to H
4:04 me think I think I'm
4:08 happy like first day summer vacation
4:12 happy got to get a little rest and
4:15 relaxation
4:16 happy like to Qui on Sunday morning and
4:19 sing it
4:24 true Oh Sister when you cook that
4:28 way sweet distraction rules my
4:34 day uh kitchen table just my break to
4:39 the sound of you and I rocking our
4:41 bodies yeah I don't understand why
4:45 wasn't your of a man 100 years
4:50 ago F's going to have to understand put
4:54 the hat to the man who
4:58 scor think I think I I'm
5:01 happy like a PO on vacation happy got to
5:06 get a little rest and
5:08 relaxation
5:09 happy like to quiet on Sunday morning
5:12 singing
5:14 [Music]
5:16 true good evening good
5:19 people earbud in but I have a little
5:22 earbud
5:24 in good evening good people my audio is
5:28 weird oh that was because I had my
5:31 earbud
5:35 in I forgot I have my earbud in that was
5:38 making
5:39 weirdness ah the
5:41 [Music]
5:46 earbud we got a new
5:49 uh line
5:52 producer I know this just looks like a
5:55 shitty home office but this is actually
5:57 if I if I turn the camera around what
5:58 you would see is a $ 20 million
6:01 Production Studio and we have a uh a new
6:04 line producer we were uh we were testing
6:07 out some
6:08 [Music]
6:10 new some new
6:11 [Music]
6:13 techniques and that line
6:16 producer who's now been fired forgot to
6:20 tell me to take my earbud
6:21 [Music]
6:28 out a sad it's
6:31 sad I thought he was a fine fella 50
6:35 years in the
6:36 [Music]
6:38 business one
6:40 mistake
6:42 goodbye it's kind of Ruthless
6:44 organization we run around here because
6:46 you don't get this kind of production
6:48 quality with just an iPhone and a shitty
6:50 home office
6:52 no you think just anyone could do this
6:57 no I can't I can't see I can't
7:02 see I still can't
7:05 see I don't think my glasses have been
7:07 washed in a [ __ ]
7:09 [Laughter]
7:14 week all
7:15 right good evening good people of the
7:20 internet the
7:22 ticktockers all right is the audio
7:24 better now by the way the hair and
7:26 makeup people though they're top dollar
7:28 yeah like okay
7:30 this is four different wigs you wouldn't
7:34 know
7:36 that my eyebrows are actually bald these
7:38 are glued on a hair at a time every
7:41 [Music]
7:43 night and you're like do do they want to
7:46 make you seem like a underpaid Professor
7:50 with those eyebrows yeah part of the
7:53 look it's the AI learning lab they're
7:56 like we're going to make you all
7:57 professorial hang on we have another two
7:59 paid hold
8:00 [Music]
8:10 on this isn't even me playing this is
8:13 just a prop I've got James Taylor
8:16 sitting just just behind
8:20 [Music]
8:23 camera I mean he's decent
8:32 [Music]
8:35 James not your music
8:38 okay it's just it's always about him
8:41 it's always about
8:43 him he's like you know I was big in the
8:45 70s right I'm like yeah we know James we
8:47 know we've got the
8:50 [Music]
8:56 records Road
9:00 [Music]
9:04 this is what happens when you go live
9:06 early and the crew is not ready I know I
9:08 pulled a fast one on the mods I went 10
9:10 minutes early look it's just now 8:00
9:13 and we've almost got all our guitar
9:14 playing done we can let James go for the
9:17 [Music]
9:19 night fantastic I got Tom sigura over to
9:22 my right piping in the comedy ideas to
9:25 me as the Live's going on it seems like
9:27 I'm really quick witted and I'm just
9:28 constantly you know right I got Tom
9:31 sigura doing the the the like satirical
9:34 kind of wit I got Dana Carvey cooking Up
9:37 characters for
9:39 me you know what's funny what you don't
9:41 know before you have a $20 million
9:43 Production Studio like I
9:45 do you think that people you see on TV
9:48 are just talented what you don't realize
9:50 is there's like there's just a whole
9:52 it's a whole support
9:54 network basically just taking a
9:56 talentless meat bag and and shove ideas
10:00 into it and uh boy it's remarkable like
10:04 I you know I watch back the recordings
10:05 I'm like that's not even
10:11 me
10:13 [Music]
10:15 oh the poor qard guy exact well the qard
10:20 guy's a dyslexic which is why sometimes
10:23 things here do not
10:29 [Music]
10:52 I saw some Rumblings that a handful of
10:56 people were getting uh chat GPT voice
11:00 it's been 6 weeks 6
11:03 weeks yeah 6 weeks since they said we're
11:06 going to have voice and like four months
11:08 since they announced
11:10 Sora I think people are fed
11:14 up with old open AI not launching [ __ ]
11:29 w
11:32 [Music]
12:02 ah Silver Fox I went 10 minutes
12:10 [Music]
12:21 [Applause]
12:26 early well you don't tell me you better
12:29 I felt it tried to be chill but you're
12:32 so hot that I melt it I fell right
12:35 through the
12:36 cracks now I'm trying to get
12:39 back the cool run out I'll giving it my
12:43 best nothing going stop invention reckon
12:48 it's again my turn no CH to win some
12:52 more some but I
12:57 won't take no more
13:01 noal it canot wait I'm
13:06 [Music]
13:12 your
13:14 no
13:16 yeah yeah
13:19 yeah shitty
13:22 [Music]
13:24 dog you know what you did right there
13:26 champ you just revealed what a subpar
13:28 singer I Am
13:30 see here's the deal when you have a
13:31 singing dog you don't have to sing that
13:33 good because everyone was like oh
13:34 Champ's so
13:36 cute but when champy doesn't step up to
13:38 the plate and he doesn't sing people are
13:40 like what is with this guy trying to
13:43 should he should he be putting this on
13:45 Tick
13:46 Tock he'd look better in a
13:50 [Music]
13:54 manini twerking in a
13:56 man think about it all night
14:00 [Music]
14:05 all right let's get going so to those of
14:08 you that are new here my name is Kyle
14:11 Shannon this is the AI learning lab I'm
14:14 the CEO of this company Story vine this
14:16 is this is what they call
14:18 merchandising
14:20 marketing so right now you're all
14:22 Googling what Story vine oh that's
14:25 pretty cool it's a it's a video it's
14:27 like Tik Tok but different
14:30 huh that's strange
14:33 Bob so I do
14:38 that do this thing called the AI Salon
14:40 we'll talk about a little bit
14:43 later and I come in here and act like an
14:46 idiot just an
14:48 idiot I just come on here and I'm like
14:50 hey there people hi people what's
14:53 happening let's talk about artificial
14:57 intelligence so
15:00 I can answer any questions if you got
15:02 questions we can uh put them in the uh
15:06 the FAQ machine I spin it up like a
15:09 bingo wheel bingo cage we put all your
15:13 questions in there I spin it up like a
15:14 bingo cage and I pull out three
15:17 questions and then I look at them and I
15:20 answer some other random
15:23 question all right what about the
15:25 lawsuit with yudo andso yes I I will
15:28 read you the tweet I wrote about
15:35 [Music]
15:43 it gota get my acting voice going all
15:50 right major labels Sue AI firm sunno and
15:54 yudo for alleged copyright infringement
15:57 spearheaded by the raia a recording
16:01 industry
16:02 recording
16:04 industry something
16:06 Association Universal Warner and
16:09 Sony have banded together to sue two of
16:12 the world's most advanced startups in
16:14 emerging field VII music so they're
16:17 going to sue on behalf of all these
16:18 artists and then when they get all that
16:19 money they're going to give it to the
16:21 artists
16:26 right oh you that's cute
16:30 that is cute no they're going to keep it
16:32 they're going to keep it so here's what
16:33 I
16:34 wrote this was my
16:38 response lawyer quote I totally agree AI
16:43 is digital theft let's sue the bastards
16:46 lawyer hangs up the phone Stephanie
16:48 please bring me the folder of yacht
16:52 catalogs the winners in the next 20
16:55 years will be the lawyers AI isn't
16:57 slowing you won't stop it
16:59 accept AI starve a
17:04 lawyer
17:06 um do do I think that they are going to
17:09 find that there are all sorts of
17:11 copywritten songs in the training data
17:16 yes do I think
17:19 that it would be better if all these
17:22 companies had had found a better way to
17:26 attribute um and compensate for for the
17:29 training data that's in there yes um
17:34 would any of this technology that we're
17:36 playing with today exist if they had
17:38 done that
17:40 no so
17:43 so I mean there's there's sort of a
17:47 short-term reality and there's a
17:48 long-term reality and I look at this I
17:51 try to look at I I try to deal
17:55 with all of my philosophy around the a
17:58 around AI
18:01 is all built on the premise that this is
18:03 not stopping it's not slowing down and
18:06 it's not going
18:08 away any of it so I think the lawsuits
18:12 lawsuits are all just distractions
18:15 I I look at the
18:20 long-term the long-term embedding of
18:23 data the long-term GPT ification of
18:27 everything
18:29 I see it as more vital and more
18:32 important than all of the libraries in
18:35 the world
18:36 combined and all of the museums in the
18:38 world combined I see it as fundamental
18:41 to the future of
18:44 humanity
18:47 um and if you ask someone about a
18:49 library they say well yeah it could they
18:51 can have recorded music in a library
18:53 because it's for the good of the
18:56 community and so
18:59 now the the argument is yeah but they're
19:00 making money on it well they may or may
19:02 not be I think most of these companies
19:04 are losing money what they're doing is
19:06 they're building these models that
19:09 everyone gets to benefit from the
19:11 collective intelligence of everyone
19:12 that's come before
19:14 us the problem
19:17 is the artists in the
19:20 past 40 years 50 years that are still
19:26 alive but I think I think we could all
19:29 I'll safely argue that
19:32 having all of these Works trained on all
19:34 of the piano concertos that were
19:38 recorded in the early part of the
19:40 century that's probably good classic
19:42 good to have classical music in there
19:43 it's probably good to have the works of
19:44 Shakespeare in there it's probably good
19:46 to have the Betty Crocker cookbook in
19:48 there
19:51 right but there's a point at
19:54 which everything will be embedded
19:56 everything will be in the latent space
20:00 and we're just not there yet so we're in
20:01 a we're in a funky transition period
20:04 where I absolutely understand the
20:06 frustration the madness the they're
20:08 stealing from us and then I look at it
20:10 and go
20:13 and this is where it's all going and it
20:16 there the argument for this is for the
20:18 good of
20:19 humanity if ever there was an argument
20:21 to be made for a technology where that's
20:23 the case this is
20:27 it because if you think about what what
20:31 the Transformer model allows us to
20:34 do so so here's an amazing thing
20:39 Google their technology they figured out
20:41 what these spiders that would crawl
20:43 around the the worldwide web and it
20:45 would crawl around an index it would it
20:47 would say okay here's what this site's
20:49 about here's what this site's about it
20:51 took them years and years to do that
20:53 they went and they crawled all of the
20:55 internet and they let us find all the
20:57 [ __ ] on the internet and that was abs
20:59 absolutely remarkable cuz we could find
21:01 all the [ __ ] on the internet and then
21:02 over 10 years they became a monopoly and
21:06 the SEO [ __ ] took over and it ruined the
21:08 [ __ ] internet
21:10 right but there was a there was a sweet
21:12 spot there where they
21:15 had they had indexed everything that we
21:19 could get our hands on or you know most
21:21 of it a lot of
21:26 it chat GP or GPT 3 the the the model
21:30 that was the first one to sort of crack
21:32 the code on this and then GPT 4 and GPT
21:35 5 theoretically is coming in a couple of
21:39 weeks what they've effectively
21:43 done is not just taken the metadata
21:47 about
21:48 websites but they've taken all of the
21:51 data from all of the sites all of the
21:54 data and they've squished
21:57 it into a little gold
22:00 button that when you search on it you
22:03 can make
22:05 money it doesn't just return you a list
22:07 of links it
22:09 actually does some funky ass form of
22:12 reasoning and answers your
22:15 question so they've essentially
22:17 compressed the
22:20 internet into this remarkable
22:24 orb that can just answer
22:27 [ __ ] and do [ __ ] write code make
22:33 music make pictures make
22:42 video it's going to it's going to
22:44 fundamentally alter
22:48 everything and so I look at the three
22:52 biggest music conglomerates in the world
22:59 what they're doing is not protecting
23:01 artists what they're doing is trying
23:07 to you know the the metaphor I use
23:10 sometimes of like you know sort of
23:12 arguing about whether or not AI is good
23:15 or bad is is is sort of like the Florida
23:17 legislature arguing about and passing a
23:20 law making tsunamis
23:23 illegal suami are bad we should they're
23:27 bad yes let's pass a law let's make the
23:31 people feel good we will protect them
23:33 with our fantastic legislation all right
23:37 people gather
23:40 around you know that tsunami that's
23:42 coming
23:44 uhhuh not to worry we passed the
23:48 law you're
23:50 protected but but what happens if the
23:54 wait it's all right we've got laws for
23:57 that now
24:03 so these big music conglomerates are
24:07 like uh [ __ ] it's not they're stealing
24:11 from our artist it's not that it's their
24:14 entire business model I mean their
24:15 entire business
24:17 model got obliterated with streaming and
24:20 so that was bad
24:22 enough
24:24 um and so now this is going to
24:27 absolutely decimate it because
24:29 music in large part is going to shift
24:32 from one person making media and
24:35 distributing it to Millions to all of us
24:39 making music in kind of real time as we
24:43 need
24:45 it um and we may or may
24:49 not you
24:52 know buy music from companies like that
24:55 ever again or even stream music like
24:57 that ever again now we
24:59 probably will right songs you like
25:00 you're going to listen to again but you
25:03 might also just want songs that sound
25:04 like songs you like that are really
25:06 personal to
25:07 you so their entire business model is
25:09 going to change so I see this as
25:12 a their attempt to there the th this
25:16 lawsuit is like build a
25:18 dyke what's a dyke it's a dam build a
25:22 dam where around Tampa
25:26 Bay but the tsunami is going to hit all
25:28 of Florida and it's going to be 20 ft
25:32 High how how big is your Dyke going to
25:33 be
25:35 4T why 4 feet cuz high tide if it's
25:39 windy and high tide it goes to about 3
25:40 feet so we want to build a foot there's
25:42 a 20ft tsunami yeah just if we could
25:44 just build the the Dyke around Tampa Bay
25:48 and and that would protect us if that is
25:50 that good does that work for
25:53 you it's [ __ ]
25:57 ridiculous now they shut down sunno and
26:00 udio
26:02 yep you don't think there's an open
26:04 source project right now sampling all
26:06 the [ __ ] music they can get their
26:08 hands
26:12 on I bet one comes out from stability AI
26:17 they'll open source that [ __ ] cuz
26:19 stability AI knows that they're probably
26:21 going to get their asses sued into
26:22 Oblivion that's I think my opinion about
26:25 them is they're open sourcing everything
26:29 because if they get sued into Oblivion
26:33 all the codes still out
26:36 there so I
26:41 understand why the lawsuit's
26:43 happening it'll be interesting to see
26:45 what happens with it a lot of these have
26:47 gotten tossed out of
26:49 court because the arguments for the
26:51 greater good are pretty
26:54 significant and they'll do the same [ __ ]
26:56 that the New York Times Did which if you
26:58 prompted in the if if you prompt it in a
27:00 specific way it will give you a specific
27:02 result and the New York Times can say
27:04 they stole our articles
27:08 see and the music industry will do the
27:11 same [ __ ] with
27:13 this
27:17 um you know the other thing is that's
27:20 just [ __ ] gross is you know none of
27:24 the if let's say they win this lawsuit
27:26 let's say they get the $500 million that
27:30 got invested into these two companies
27:32 they get every penny of it how much of
27:34 that do you think is going to go to the
27:38 artists
27:40 anyone it's mighty quiet in
27:44 here it's because all you can do is
27:46 type I have all the power it's
27:50 hilarious all right there is a protect a
27:54 project that is combining all
27:56 possible chords and copyrighting them
27:59 with open access to everyone empty
28:01 bubbles yeah yeah exactly I mean again
28:05 if you sort of just mentally Do Your
28:08 Back to the Future get your car to 88
28:11 miles an hour and go back to the future
28:13 to 5 years from
28:20 now all of the data is going to be
28:23 ingested into all of these models we
28:26 like right now what makes this stuff so
28:29 powerful is not the Transformer model
28:31 it's not the technology well it is but
28:34 the thing that makes it so powerful is
28:36 that we are getting to tap into the
28:40 collective intelligence of everyone
28:42 that's come before us that's put any
28:45 anything on the
28:48 Internet it's [ __ ]
28:51 miraculous [ __ ]
28:56 miraculous I mean
29:01 in the data set is there some
29:03 copywritten books and that is why this
29:07 thing writes good
29:11 yeah but like the scale of the amount of
29:14 data that these things are trained on
29:17 it's kind of the if you say well I have
29:20 a song in there
29:22 okay and I want to be compensated for
29:25 that okay here's how we're going to do
29:27 it in this
29:29 cake there was three cups of sugar and
29:33 your song is one of the little granules
29:36 of sugar and so that granule of sugar
29:39 was dumped into the cake batter it was
29:42 stirred
29:43 around and then it was stuck in the oven
29:45 it was baked into a cake that little
29:48 piece of sugar crystal no longer
29:54 exists it's the way these things work
29:57 the original s Source
29:59 data is not there anymore it's turned
30:02 into vectors it's turned into
30:04 mathematical locations in Thousand
30:06 dimensional mathematical
30:10 space and when you say to do something
30:13 it squirts up in there and goes hey give
30:15 me the next most probable thing and all
30:18 these [ __ ] to tokens rain down and if
30:21 if it was trained on Words then words
30:23 rain down and if it was trained on music
30:25 then music rains down and if it was
30:27 trained on fmri brain
30:31 scans
30:32 then crippled or paralyzed people can
30:40 walk your your grain of sugar ceases to
30:44 exist it's part of
30:46 the we could cut the cake up into teeny
30:49 little pieces about the size of the
30:50 piece of sugar but it's no longer just a
30:52 piece of sugar it's now cake
31:00 sugar is considered a fluid in baking I
31:02 did not know that but but it makes sense
31:05 because when you eat your cake is not
31:08 crunchy it's smooth and delectable and
31:13 delightful smart people are saying 13
31:16 years 80% of all jobs will be gone so we
31:20 need to
31:23 do something else anyway
31:26 exactly and it's fun
31:29 I here's how that's phrased I think is a
31:37 little it's in in 13 years 80% of the
31:41 jobs are going to be
31:43 gone I don't look at it like that I look
31:46 at it like this and I think it's going
31:47 to happen sooner than 13 years quite
31:50 frankly let's say in 10
31:52 years 80% of the tasks
31:58 we do today will be
32:06 automated your job may still exist but
32:09 you do completely different
32:12 tasks cuz the tasks are going to be
32:15 automated and right now we call we call
32:17 this collection of tasks a
32:23 job most of those
32:26 tasks are shitty
32:30 repetitive
32:31 mind-numbing Soul crushing shitty shitty
32:39 tasks so if you flip the model and
32:42 instead of saying all the jobs are going
32:44 to be
32:45 gone and say all the shitty tasks and
32:49 some of the good tasks but all the
32:51 shitty tasks are going to be
32:54 automated and what that means is we get
32:57 to do something else now what we don't
33:00 know what none of us can know is what
33:02 does that actually [ __ ]
33:04 mean does it mean that companies are
33:06 going to keep their
33:10 employees well if if if what their
33:14 definition of a company is just does the
33:17 same tasks that those humans did and now
33:20 those tasks are
33:23 automated then that company's going to
33:25 get rid of people but there might be
33:28 other companies that say Hey all those
33:30 shitty tasks got automated you know what
33:33 we could do we could take those same
33:35 people and now we could put them in
33:38 highlevel customer service roles or we
33:41 could put them in Innovation roles or we
33:43 could put them in roles where they
33:44 actually get to use their [ __ ]
33:49 brains what if our employees weren't
33:51 just [ __ ]
33:53 robots we now have robots to do the
33:55 robotic work
33:59 side hustle meet me I think that's
34:00 definitely the case for 30% of my work
34:03 we need Less jobs amen joy joy P's into
34:07 my
34:08 preaching thank you
34:11 sister H Remington made the first
34:13 typewriter can we suit Apple for using
34:15 the quiry keyboard exactly that I mean
34:18 this is this is this is every [ __ ]
34:22 technology every single technology every
34:26 single
34:27 time there's a new technology it's gonna
34:29 kill the world the people are all gonna
34:33 die what will people do if they can't
34:35 till their fields with
34:37 horses what about the saddle
34:45 makers those people moved to
34:48 cities they invented the service economy
34:51 like like humans
34:54 [Applause]
34:55 adapt are they going to sue Taylor Swift
34:58 next by training on other artists
35:04 music humans
35:07 adapt we will have new jobs I mean
35:10 what's kind of wild is those jobs could
35:12 be jobs like entertainment jobs like
35:17 music and stupid comedy
35:21 and what I throwing parties
35:28 if the machines are doing the tasks and
35:31 the tasks are generating the
35:36 GDP and it actually generates enough in
35:39 the GDP that there's actual excess that
35:42 the rich people can't just hoard it
35:45 all because they knew if they knew if
35:47 they did hoard it all with 80%
35:50 unemployment they're
35:52 [Laughter]
35:55 outnumbered so so in that case like at
35:58 some point the the the uh Point 0.1 of
36:03 0.1% power people at some point they
36:07 realize they they got to keep people
36:11 from killing
36:13 them
36:15 right if Rich powerful people are good
36:18 at one thing it's
36:21 self-preservation so assuming that that
36:23 that much Prosperity is generated
36:26 then some decent amount of that will
36:29 trickle down to the people and the
36:32 people won't have to work as Elon Musk
36:34 said work will be
36:36 optional and I think most people want to
36:39 contribute they want to connect with
36:41 other people they want to contribute so
36:43 what will that contribution look like if
36:45 the tasks of work are being done I don't
36:48 know it might look like the server at
36:51 the restaurant you went to is in less of
36:55 a shitty
36:56 mood cuz they actually want to be there
36:59 cuz they're choosing to be there I don't
37:01 know I don't
37:04 know Darren Drew hey so what do you
37:07 think
37:08 AI is good at making a better mouse
37:12 trap but will it ever be
37:15 novel
37:18 yes yeah it will and it is it is for me
37:23 regularly daily here's here's the deal
37:32 I think humans right now are defensive
37:34 and elitist about our
37:38 brains I think we think our brains are a
37:41 bigger deal than they actually are now
37:42 listen don't get me wrong what's up here
37:47 a Marvel of ancient technology Marvel of
37:51 evolutionary
37:52 technology I think it's
37:56 fantastic but we talk about
37:58 only humans can be novel and AI is just
38:01 a prediction
38:03 engine so I fell for that for a while
38:07 but but here's where I've
38:10 come I've used the term collective
38:12 intelligence a couple of times
38:15 tonight the reason when you interact
38:18 with chat GPT that it doesn't feel like
38:20 a
38:22 robot even though it's a
38:26 calculator the it it is so remarkable
38:30 how this tool actually works how
38:31 Transformers work and how embeddings and
38:34 and the Laten space actually works you
38:37 type in a
38:38 prompt and it generates a
38:41 probability and it goes and looks into
38:44 Laten space and says somewhere in this
38:47 thousand dimensional mathematical space
38:50 is a point in space that is the most
38:53 probable to be the right one and then
38:55 there's some data associated with that
38:57 and it it comes
38:58 down out of the Laten space so it's
39:03 probability find it return it or wait
39:06 probability find the point in space
39:09 associate it with the data drop the data
39:11 into the
39:13 bucket so that is a cold
39:17 calculation with no feeling it's a
39:20 probability engine it's Pro it it's it's
39:22 uh
39:24 it's um you know it's um it's it's a
39:27 it's a mathematical
39:29 formula and it's it's just doing the
39:31 next most probable word and so you say
39:34 well it's just been trained on all the
39:36 existing data and it's just returning
39:38 things in this mea me mechanistic way
39:41 and so therefore it can't be
39:43 novel okay all
39:46 right on the surface absolutely agree
39:49 with you and when I first started
39:51 learning about this stuff I'm like yeah
39:53 yeah
39:54 yeah in the past 6 months or so things
39:58 have started to shift and and I've
39:59 finally been able to articulate it in
40:01 this phrase collective intelligence but
40:05 um what started bugging me about the
40:09 it's a probability engine argument and
40:11 it can never be novel argument is
40:15 that that's not what it felt
40:21 like when I say to
40:24 it explain quantum mechanics in an
40:28 Eminem rap and it does
40:34 that that doesn't exist
40:39 anywhere it was my
40:41 idea to combine those two
40:44 concepts
40:46 now you might argue well Kyle you as a
40:50 human if you were good at knowing Eminem
40:54 and you were good at science and physics
40:58 you could probably take a week or
41:01 two and come up with a
41:04 cipher that talked about scherer's cat
41:07 in a
41:14 beat that would be
41:16 novel how is that different when an AI
41:19 does
41:20 it these are two disperate
41:23 concepts where the papers on physics and
41:27 the pap
41:28 papers the poems on
41:32 Eminem they live in distinctly different
41:36 parts of the latent space they are in
41:38 semantically different regions of
41:41 space so I combined them by by with a
41:45 prompt I combin
41:47 them that's no different than our
41:50 [ __ ]
41:56 brains how can can
41:59 I write an Eminem rap that explains
42:03 quantum physics well I go to college and
42:07 I read all up
42:08 about Stephen Hawking
42:11 [ __ ] and schinger and weirdness and
42:14 super position and then I go watch 8
42:18 Mile and I listen to Eminem dropping
42:20 some mad beats and I get them all stuck
42:22 in my [ __ ] head and somewhere in
42:24 here's Eminem beats and somewhere in
42:26 here's quantum mechanics
42:29 understanding and if at some point I go
42:31 hey you know it would be cool what if I
42:33 wrote an Eminem rap about quantum
42:35 mechanics and those two my stupid
42:38 [ __ ] crazy Strava brain somehow makes
42:41 those synapses and spits that [ __ ] out
42:44 it takes me a week to do it chat jpt
42:47 does it in 3 seconds how is it
42:51 different it's
42:54 novel and it's not novel because chat
42:57 GPT just came up with that on its own
42:59 it's novel because me as a human decided
43:02 to combine those two concepts so I think
43:04 it's novel right now if you don't if
43:07 you're not a shitty prompter if you just
43:09 go in and go tell me about physics it'll
43:13 give you a shitty
43:15 answer what happens when AI understands
43:18 nuances I think we're about to find
43:21 [Laughter]
43:25 out the real time voice
43:29 [ __ ] that if open AI ever [ __ ] drops
43:34 it I don't know if you know this well
43:38 most PE okay 47% of people in America do
43:41 not know what they they haven't heard of
43:43 chat GPT okay
43:46 47% it's in the single digits of people
43:49 that actually use it and it's probably
43:50 in the low single digits of people that
43:53 use it a lot and kind of understand what
43:55 it
43:56 is you know no one knows about this
43:59 stuff where was I going with
44:03 this lost my train of
44:06 thought question is AI a gift or a
44:13 curse I think it's the most
44:16 profound
44:18 gift in the history of humanity
44:22 personally um it's not yet it's going
44:25 sorry it's going to be the most profound
44:27 gift in the history of the humanity
44:29 right now it's a janky piece of
44:34 [ __ ] right now
44:36 ai okay so right now ai is a curse AI is
44:40 a curse for a lot of
44:43 reasons it's a curse
44:45 because it's not
44:49 perfect the hype right now is higher
44:54 than its capacity to deliver it's really
44:57 EXP
45:00 expensive we don't know what it is we
45:02 don't know how to use
45:08 it and it's coming so
45:15 fast that
45:17 that 95% of the people are going to be
45:22 blindsided by it maybe it's not that
45:24 much let's call it 80% 80% of the people
45:27 people are going to be absolutely
45:30 blindsided by what it
45:33 does because what it does right now is
45:36 kind of like parlor tricks oh it made a
45:38 cute little video where the face looks
45:40 weird oh yeah it made a nice little song
45:43 it sounds a little off is it
45:46 me oh yeah that it WR it does the
45:48 writing thing why does it use the word
45:50 delve so much what's with
45:52 that right so right now
45:58 it's kind of a
46:00 mess and so the the implications of that
46:04 are pretty grave I think because it
46:06 allows people
46:08 to dismiss it so there's two things
46:11 happening right now no there's three
46:13 things happening right now that I think
46:14 are really
46:16 problematic there's ignorance or apathy
46:20 or both right just we don't know we
46:23 don't care whatever those smart people
46:26 are doing something with their thing
46:28 whatever we don't like science okay I I
46:32 mean science really you're going to send
46:34 a little balloon up in the sky and
46:36 that's going to tell you the weather
46:37 yeah
46:38 whatever
46:40 right
46:43 um so there's apathy and ignorance
46:47 willing you know willful
46:49 apathy I guess apathy is
46:53 willful and then
46:55 there's um
46:58 dismissing it it's not perfect I'll just
47:02 wait that's a yeah yeah you you tell me
47:07 uh when it when it's not going to get me
47:08 kicked out of college okay yeah
47:13 yeah so there's dismissing it because
47:15 it's not perfect it makes mistakes it
47:17 hallucinates yeah I I I'll tell you what
47:20 I'll use it uh when it doesn't
47:21 hallucinate okay uh I like my computer
47:24 technology not to be on mushrooms if you
47:26 know what I mean cuz when you take
47:27 mushrooms they
47:30 hallucinate and then there's demonizing
47:32 it thei is going to kill
47:35 us and the News loves that
47:40 one and they love talking about it's
47:43 stealing from people it's not perfect it
47:47 hallucinated somebody lost $20
47:51 million
47:53 right those three things people not
47:56 giving a [ __ ]
47:58 people dismissing it cuz like yeah
48:00 whatever I'll I'll I'll catch up or or
48:02 it's evil I will never I would never
48:07 support anything where the AI could come
48:13 around the sneak around the corner and
48:16 they're GNA it's going to kill us all
48:18 how how's it going to do that I that I
48:20 don't know but it's that I do not
48:22 support
48:25 it and meanwhile
48:30 it's
48:32 evolving this if you're here you know
48:35 this I'm preaching to the choir it's
48:37 evolving incredibly
48:40 quickly if you look at the images we
48:42 were creating a year ago and if you
48:43 listen to the [ __ ] songs we were
48:45 creating a year ago they were not
48:47 listenable they were
48:50 horrible and now they're good enough
48:52 that Sony who is it Sony Universal and
48:56 Warner
48:59 are suing to stop
49:02 it cuz now they're pretty [ __ ]
49:08 good every
49:11 single every single industry every
49:15 single industry every single
49:20 industry is going to face this
49:25 moment and it's it's not 80% of the jobs
49:28 that are going to be gone it's 80% of
49:30 the tasks are going to be
49:36 automated so is it a blessing is it a
49:47 curse it's going to [ __ ] some [ __ ] up
49:49 for the next five
49:51 years and then I think we emerge out of
49:54 it in in what I'm what I call the great
49:56 Rena
49:59 and I don't know if you knew this but in
50:00 the middle of the word renaissance is AI
50:03 so I always capitalize that when I write
50:05 the great
50:06 Renaissance I think we are entering the
50:08 great Renaissance but right now it's
50:10 going to feel like the great [ __ ]
50:13 show there's an image if you if you want
50:16 to put something in
50:18 Irregulars the great [ __ ] show and the
50:20 great Renaissance
50:28 and here's here's where I think it is
50:30 it's the it's the it's the greatest
50:35 blessing for those people that hang out
50:38 here and hang out in the AI salon and
50:40 the people that I know and respect that
50:42 are doing stuff with this everyone has a
50:45 very similar kind of story it's they
50:48 they articulate it in different
50:51 ways but I think the thing that that AI
50:54 does the most for me personally and and
50:57 just I hear it over and over and over
51:00 again is that it allows me to
51:04 take an unformed idea in my
51:09 head and manifest it
51:14 instantly and be able to start playing
51:17 with
51:18 it and I don't know if you've seen if
51:21 you haven't played with claw. if you're
51:23 new here and you haven't played with
51:26 chat GPT your
51:29 claw. go to claw.
51:34 a and tell
51:36 it I think you have to turn on this
51:39 feature called
51:41 artifacts which makes a window beside
51:43 your prompt
51:44 window so in in your settings is like
51:49 special features or something like that
51:51 and turn on
51:53 artifacts and then you'll have your your
51:57 your Claude chat GPT thing here and then
51:59 you'll have this window here and it'll
52:01 do this dynamically you don't have to
52:02 think about
52:04 it and you just tell
52:07 it make me an asteroids
52:11 clone and then it thinks a little bit
52:14 and then it spins up this window and
52:16 then it writes all this code and then it
52:18 runs the
52:21 code and there's an asteroids like video
52:25 game clode you know like from the 70s
52:27 asteroids anyone old
52:38 people and if you used hyperspace you're
52:41 a loser nobody used none of the good
52:44 players used hyperspace
52:46 okay but anyway I get digressed welcome
52:50 to Chad
52:55 add I can take any idea in my
52:58 head and instantly manifest it instantly
53:02 manifest it but the cod's not always
53:04 right it's sometimes broken shut
53:08 up I can say write an asteroids clone
53:13 and it [ __ ] writes the
53:16 code and plays the
53:19 game I don't care how janky it is how
53:21 broken it is that's [ __ ] miraculous
53:27 and it gets more interesting when you
53:29 say I want you to create a video game
53:32 with gameplay that's no one's ever seen
53:34 before and it will do
53:37 that back to the question of could it
53:39 ever be novel yeah if you ask it to
53:44 be and then it
53:50 is think about how
53:53 long most people spend
53:57 trying to get an
53:59 idea out of their
54:04 head oh I really have an idea for a a
54:09 musical here
54:12 I I I guess I should learn how to write
54:15 music Let Let me let me uh let me go
54:18 sign up for
54:20 college to get a musical theater degree
54:23 and a maybe a music theory degree and
54:26 then I then I should probably uh learn
54:29 learn about stagecraft
54:31 and I'll go spend four years doing that
54:35 to be able to get to the point that I
54:37 could know enough to even begin to get
54:39 this thing out of my
54:41 head and
54:43 now I just say hm I have an idea for a
54:48 musical how do you write a
54:50 musical
54:54 H what if I want it to be a comedy
54:59 shut up I don't want to know about how
55:00 it's not as impressive as it
55:07 sounds
55:09 oh it's so imperfect right now it's so
55:12 easy to dismiss
55:27 bet you weren't expecting binocular
55:29 jokes tonight
55:32 huh oh he's a prop comic I
55:36 see you can make money
55:40 with
55:46 um that was a fun hour
55:51 [Laughter]
55:57 [Music]
56:00 I have been telling everyone about your
56:02 idea of collective intelligence I say my
56:04 AI teacher says oh that's awesome that's
56:08 awesome yeah it's I I think collective
56:11 intelligence is a better term than
56:12 artificial intelligence because that's
56:14 that's what we're
56:15 experiencing it's AI wrapped in a curse
56:18 wrapped in a blessing it's a Truckin
56:20 yeah you know what I that that's
56:23 actually really good Wolf Man Clint I
56:25 think I I think AI as a turducken is
56:30 right like it's just there there's
56:31 there's so much to
56:33 it there's parts of it that is just
56:35 [ __ ] remarkable and and to Pat
56:39 Chagrin magic there's there there are
56:41 some things it does that are [ __ ]
56:45 magic and
56:47 then it's shitty and not good and and
56:51 it's just bad
56:53 math and then it's G to [ __ ] up a bunch
56:57 of people's lives cuz it's going to
56:58 change all the
57:00 tasks and so it's going to change all
57:02 the jobs and in changing all the jobs
57:05 some people will not want to change when
57:08 their job changes and some people won't
57:10 have a choice and so that's going to be
57:12 really [ __ ] painful so yeah it is
57:14 it's a it's a tradu AI is a traduction
57:17 of curses and blessings yes solid curses
57:22 blessings and shitty technology and
57:24 janky technology cuz it's not shitty
57:26 techn ology it's just early technology
57:28 so it's
57:30 janky crazy times crazy [ __ ] times
57:34 all
57:35 right uh collaborate collaborating with
57:38 AI is like talking to a smart human with
57:40 no ego that's an amazing bonus yeah
57:42 archetypal architect my my experience
57:46 with chat GPT and this this took me a a
57:49 while to assimilate into my psyche
57:56 it's just not bitter it's not jealous
57:59 it's not Petty it's like a golden
58:02 retriever you're like I have an idea
58:04 chat gbt awesome idea fantastic why
58:07 don't we explore that idea here's some
58:09 things you probably didn't know about
58:10 that idea oh okay chat gbt uh well what
58:14 if we did it this what is what a
58:16 fantastic idea about your idea let me
58:19 help you with that oh chat gbt you got
58:22 that completely wrong and like as I'm
58:24 typing you got that completely wrong I'm
58:26 like oh oh chat PT is going to be really
58:28 upset here they're at some point they're
58:30 probably just going to quit right and
58:32 then chaty PT is like my apologies for
58:35 not getting it exactly as you wanted it
58:37 I'll tell you what let me try again here
58:39 you
58:40 go it's bizarre it's it's like working
58:44 with a you know people that are
58:48 too
58:50 positive don't they drive you [ __ ]
58:52 crazy hey how's it going it's going
58:54 fantastic Bob
58:57 and you just want to like smash a pie in
58:59 their
59:01 face and and but then they would be
59:03 like that was a little aggressive but oh
59:06 that's tasty pie and which would again
59:09 make you want to smash another pie in
59:11 their face um chat gpt's got that
59:15 attitude but it's not
59:17 annoying it's just like oh sure oh
59:20 apologies here we go yeah let's go oh
59:23 fantastic oh here how's this
59:31 like for me personally like the mental
59:34 health
59:37 boost of just having my ideas validated
59:41 multiple times a day sometimes multiple
59:43 times an
59:45 hour is subconsciously changing my brain
59:49 I'm getting less afraid to put ideas
59:51 into the
59:54 world so imagine
59:57 if the collective
59:59 intelligence on
1:00:02 Earth put 5% more ideas into the world
1:00:06 because they weren't afraid to do
1:00:09 it do you know the [ __ ] massive
1:00:11 impact that would have now imagine if
1:00:13 that's 10% more ideas expressed now
1:00:17 imagine that's
1:00:19 40% more ideas
1:00:21 expressed so it's not just the
1:00:23 technology that's going to get better I
1:00:26 I think it's humans being freed up to
1:00:30 not do the shitty
1:00:32 tasks and freed up to share their
1:00:36 ideas is going to be
1:00:40 profound
1:00:42 like [ __ ]
1:00:46 staggering but let's sue them let's stop
1:00:49 this we must stop the
1:00:53 madness get us a lawyer
1:00:56 you know what get me a Bus full of
1:00:57 lawyers we're going to need
1:01:00 them it's dreaming hear me out when we
1:01:03 dream our brains combine two or more
1:01:05 things I antie bubbles totally again
1:01:08 this is why I think the term collective
1:01:09 intelligence is the term not artificial
1:01:12 intelligence it's
1:01:17 absolutely it's combining it is it's
1:01:20 combining ideas it's even doing it when
1:01:22 you tell it to give you the exact New
1:01:24 York Times article even if it gives you
1:01:26 the exact New York Times article it is
1:01:28 likely pulling those words from
1:01:30 different semantic regions of the latent
1:01:38 space it's how our brains work it's how
1:01:41 our brains
1:01:46 work maybe not technically like maybe
1:01:49 it's technically kind of
1:01:52 different but what is a novel thing what
1:01:54 is a novel idea a no
1:01:56 idea is understanding the world enough
1:01:59 to know how to mix together ideas how to
1:02:03 daydream so when you dream at night
1:02:05 you're night dreaming in day when you
1:02:08 invent things and you Daydream you're
1:02:09 like
1:02:11 ha I wonder if I combined a car wash
1:02:14 with a lollipop
1:02:16 Factory and put it in a
1:02:19 tent wonder what that would
1:02:22 be that's novel thinking well now you
1:02:25 can go to chat GPT just do that and it
1:02:27 will do it and it'll make you the images
1:02:29 of it it'll make you the business plan
1:02:31 for it it'll tell you how to Market it
1:02:33 it'll come up with a name for you what
1:02:36 tell me what about that is not
1:02:43 novel cuz the reason there are arguments
1:02:46 that it can't be
1:02:47 novel is that you see papers that show
1:02:50 you the math and the math
1:02:53 says this is a probability engine and
1:02:55 it's it it will be the same every time
1:02:58 because the probabilities are the same
1:02:59 and you can change your temperature to
1:03:01 change that up but because of
1:03:04 that
1:03:06 predictability it can't be
1:03:10 novel unless you think beyond the
1:03:12 mathematical formula into how those
1:03:15 mathematical formulas are actually being
1:03:16 used and how the data is actually being
1:03:19 combined AI destroys scientific and
1:03:21 critical thinking just ask get an answer
1:03:24 I I disagree dunker
1:03:27 in in this
1:03:30 sense it replaces some of the critical
1:03:33 thinking but it forces new critical
1:03:36 thinking because if the AI spits out of
1:03:40 something it's then your job as the
1:03:42 human to look at the something and go is
1:03:44 that good is that bad is it interesting
1:03:47 ooh that reminds me of some other thing
1:03:49 over
1:03:51 here let me get chat GPT to combine
1:03:54 these three things it just gave me with
1:03:55 that thing and see what that does so the
1:03:58 critical thinking is still happening
1:04:00 it's just happening way faster on a lot
1:04:03 more axes of
1:04:06 ideas dunkaroo that's a gross
1:04:08 oversimplification people thought the
1:04:10 same thing about books and the internet
1:04:12 yeah exactly pate's pate's right there
1:04:15 like pate's got a PhD he knows the math
1:04:18 stuff right it doesn't destroy critical
1:04:21 thinking it just it changes it it
1:04:25 changes it
1:04:26 our role in critical thinking shifts
1:04:30 from having to do the the heavy lifting
1:04:32 to let chat jpt do the heavy lifting and
1:04:36 then we get to do critical thinking at
1:04:38 the idea realm you know how there's okay
1:04:42 here's here's the thing I don't sorry
1:04:44 about this being a when I take two days
1:04:46 off I tend to
1:04:48 rant um what was I thinking critical
1:04:51 thinking I [ __ ] lost my train of
1:04:53 thought doesn't matter okay I told it to
1:04:54 make an app and spit out 100 ,000 worth
1:04:56 of startup in 2 minutes and how did it
1:05:02 do and you know and if you want to
1:05:04 create that startup you actually have to
1:05:06 use your critical thinking to say okay
1:05:08 of the planet gave me what's good what's
1:05:09 bad what am I actually going to put
1:05:11 money behind or time
1:05:13 behind I always said people would be
1:05:15 smarter with better data imagine when
1:05:18 that's at your fingertips yeah exactly
1:05:21 uh what What's Your Name reality time
1:05:23 real time oh like I don't know I can't
1:05:27 read it there's there's too many i's and
1:05:30 L's in your words in your name so um I
1:05:34 always said people would be smarter with
1:05:36 better data imagine when all that's at
1:05:38 your fingertips that's essentially what
1:05:39 chat GPT is it's just not all the data
1:05:42 and it's not
1:05:43 multimodal
1:05:45 but that's coming right that's when we
1:05:48 look at what Gemini demonstrated when we
1:05:51 look at what um Sora looks like and what
1:05:54 this real-time voice stuff looks looks
1:05:56 like these models are starting to get
1:05:58 multimodal there there's they're start
1:06:02 think about the data so right now when
1:06:05 we think about chat GPT and we think
1:06:07 about data we think about
1:06:09 words but think about these things where
1:06:12 they're not only trained on all of the
1:06:14 word space but when they're trained on
1:06:17 reality and they understand and can
1:06:20 simulate the
1:06:21 world that includes words and it
1:06:24 includes music and it includes movement
1:06:26 and it includes physics and it includes
1:06:32 rules it imagine when I can understand
1:06:35 all of
1:06:39 that that's what's
1:06:42 coming so imagine a system that
1:06:47 compresses the under and this is why
1:06:51 it's just hitting me right
1:06:53 now when Elon Musk was talking about
1:06:56 they they asked him well what do you
1:06:58 want X AI to be about he said the the
1:07:02 ultimate vision of X
1:07:05 AI is I want to un I want it to
1:07:08 understand the
1:07:12 universe why is that well if you
1:07:14 understand the universe if it
1:07:16 understands everything in the
1:07:19 universe then you can ask it anything
1:07:23 and it will know it anything
1:07:33 anything how do I get a marble to shoot
1:07:37 out of a
1:07:38 slingshot orbit the Sun and return to
1:07:41 Earth without
1:07:45 melting here you
1:07:47 go could you build me that machine and
1:07:49 some robot goes and builds that [ __ ]
1:07:53 machine quick question GP o versus
1:07:57 Claude uh uh okay here's a [ __ ] up for
1:08:02 Claude I don't I wrote this on Twitter
1:08:04 today why the [ __ ] is Claude not
1:08:07 connected to the
1:08:08 internet what is with that is that a
1:08:11 conscious Choice did they do that on
1:08:15 purpose I think clawed by a
1:08:19 landslide except it's not connected to
1:08:22 the internet and it can't do images
1:08:26 so I'm still with chat
1:08:29 gp4 cuz it's
1:08:31 just more complete than anything else
1:08:34 right
1:08:35 now but Claude is really [ __ ]
1:08:37 remarkable if you want to do coding if
1:08:40 you want to do writing it's got a
1:08:42 200,000 token context window like
1:08:44 there's things about Claud that are way
1:08:46 better like I would say right now
1:08:49 both now apparently July 15th is when
1:08:53 GPT 5 is coming out
1:08:56 but
1:08:57 apparently last [ __ ] May is
1:09:01 when Real Time audio was coming
1:09:04 [Applause]
1:09:05 out
1:09:09 so I don't like apparently anymore with
1:09:11 open
1:09:14 AI but there was some there was some
1:09:16 rumbles going around the the mirror
1:09:20 Mira Madi
1:09:24 interview she said said
1:09:27 um the next model that's that's going to
1:09:30 be PhD level is 2026 right yeah 2026 and
1:09:35 everyone was like oh chat gbt 5's not
1:09:37 coming out till
1:09:38 2026 well apparently chat gbt 5 is
1:09:42 done and what she was talking about was
1:09:44 Chad GPT 6 or whatever their next model
1:09:47 is um if you know anything about the
1:09:50 Cycles at open AI by the time we were
1:09:52 playing with gp4 chat G pt4 last March a
1:09:58 year ago
1:10:00 March it had already been out for a year
1:10:04 year I mean done
1:10:06 internally they started red team testing
1:10:08 it in September of 20 20
1:10:14 two GPT 4 they started red teaming it in
1:10:18 September of
1:10:19 22 November of 22 they launched chat GPT
1:10:23 3.5 on the 3.5 model
1:10:27 March so they were probably if they were
1:10:31 red
1:10:33 teaming red team is testing GPT 4 in
1:10:37 September of 2022 they were probably
1:10:40 starting GPT 5 beginning of
1:10:44 2023 so it's been a year in change or
1:10:47 and probably longer quite
1:10:49 honestly so yeah so gp5 should be here
1:10:53 soon Kyle Synthesia two dropped with
1:10:56 some great new features wow okay let's
1:10:58 go look at that
1:11:00 because oh so that's interesting that
1:11:03 that they dropped hang on
1:11:07 Synthesia um cuz one of the things I
1:11:10 heard I think from Robert
1:11:14 scoble was that
1:11:17 um no it wasn't him it was some someone
1:11:20 I talked to got to test something from
1:11:23 Synthesia that was using open ai's
1:11:26 realtime voice module and it was a a
1:11:30 real time Avatar where you could just
1:11:33 have a conversation with like a fake
1:11:35 version of me and it would just talk
1:11:37 like me and it would like be
1:11:48 fast oh sthesia the original piano
1:11:51 practice app that's an unfortunate name
1:11:54 for them
1:12:03 ah let's watch a
1:12:07 video new features new workflows new
1:12:10 Avatar capabilities that will change the
1:12:11 way we communicate with video not just
1:12:14 in our free time but also
1:12:16 work see more about it and let us know
1:12:19 if you're as excited as we are can I
1:12:21 skip ahead here
1:12:29 um is this yeah hang
1:12:37 on thank you aah and welcome everyone
1:12:41 today we're introducing synesia 2.0
1:12:44 world's first AI video communications
1:12:47 platform for businesses we'll share with
1:12:49 you exciting new products and features
1:12:52 that we believe will revolutionize the
1:12:54 way that businesses ch communicate and
1:12:56 share information but before we get into
1:12:59 that I'd like to give you a little bit
1:13:00 of a backstory as to why we believe this
1:13:02 is so
1:13:03 important 580 years ago the printing
1:13:06 Park sorry I'm sure a lot of you your
1:13:09 assistant can incorporate your brand
1:13:10 identity
1:13:20 [Music]
1:13:27 [Music]
1:13:33 we start off with something we often
1:13:34 hear from our customers if you ever
1:13:36 tried to write a script you're probably
1:13:38 familiar with writers
1:13:40 block class antagonist what about
1:13:42 augmented intelligence that's pretty
1:13:44 good I think collective intelligence
1:13:46 more
1:13:48 reflects what the experience is and what
1:13:50 the training is but I think augmented
1:13:52 intelligence is what it gives us so I
1:13:54 like that term or the fear of the blank
1:13:57 page getting started is just really hard
1:14:00 I'm sure a lot of you have experienced
1:14:02 it I know I hey that's really good Wolf
1:14:04 Man Clint chat
1:14:05 GPT
1:14:07 slai is the earth excavator of our ideas
1:14:11 I that's I totally agree with that
1:14:12 that's good we have many many times and
1:14:15 to solve this problem earlier in the
1:14:17 year we introduced our AI video system
1:14:19 today it enables you to Simply select a
1:14:21 template give us a prompt or give us an
1:14:24 existing document like a word file or
1:14:26 PowerPoint we can even take a link you
1:14:29 can specify a couple of things like tone
1:14:31 of voice length of your video audience
1:14:33 and so on and we just to click of a
1:14:35 button we'll take all that information
1:14:37 and we'll give you a draft of your video
1:14:38 which maybe 70 80% there since we
1:14:41 launched it's been widely adopted by our
1:14:43 customers and we've seen great feedback
1:14:45 on how we can improve it one of the key
1:14:48 requests has been that the AI video
1:14:50 assistant can incorporate your brand
1:14:52 identity sorry about that I'm happy to
1:14:53 announce that we're making this feature
1:14:55 available next month allowing you to
1:14:57 create videos automatically with your
1:14:59 brand elements such as your font your
1:15:02 colors and your logos and achieve a
1:15:04 consistent look and feel for all of your
1:15:05 videos a few months ago during a
1:15:08 conversation one of our customers we
1:15:10 discovered something interesting they
1:15:11 actually had hundreds of help articles
1:15:13 that they really wanted to convert into
1:15:15 videos as this would help their
1:15:17 customers find answers more easily and
1:15:19 save
1:15:20 resources um and this is this is a a
1:15:24 Synthesia 2 .0 product announcement um I
1:15:28 I knew I know that there's some cool
1:15:30 stuff coming from Synthesia so someone
1:15:32 just said this so uh I figure we'll just
1:15:34 watch it together I think I'm actually
1:15:35 going to speed it up I'll I'll go one
1:15:37 one and a half times and then Firefly do
1:15:39 I really think talking nicely to the AI
1:15:42 interface makes it work better
1:15:47 um I don't think it hurts I think Ethan
1:15:49 mullik did a thing that it like it
1:15:51 wasn't statistic
1:15:52 statistically better but if you there's
1:15:55 weird things that are better like if you
1:15:57 say things like take a deep breath and
1:16:01 and think step by
1:16:04 step that makes it better so it's kind
1:16:06 of a like if you're supportive to it
1:16:08 like I know you can do it that
1:16:10 apparently makes it better um just
1:16:13 talking nicer to it like these things
1:16:15 are going to get memory so at some point
1:16:17 if you're an [ __ ] to your AI it's
1:16:20 going to send a drone Swarm at you so
1:16:22 talk nice talk talk nice to the little
1:16:27 robots for the customer service team I'm
1:16:30 really excited to announce that we're
1:16:32 also launching video last night it
1:16:36 offered to me in you'll be able to
1:16:37 Simply select a template in your
1:16:39 branding provide us a link to your
1:16:41 knowledge Center and our AI video
1:16:43 assistant will automatically transform
1:16:45 all of the Articles into high quality
1:16:47 videos holy crap avatars are at the core
1:16:50 of Synthesia and we're constantly
1:16:52 working on improving their quality and
1:16:53 capabilities
1:16:55 we've made our goal to create the
1:16:57 world's most realistic AI avatars to
1:16:59 help humans augment their capabilities
1:17:01 and last month we introduced the world's
1:17:03 first expressive AI avatars powered by
1:17:06 our Express one model and what makes
1:17:09 them different is that for the first
1:17:11 time AI avatars actually understand what
1:17:14 they're saying how they should say it
1:17:16 and they will adjust their tone of voice
1:17:18 their facial expressions and their body
1:17:20 language based on the context of your
1:17:22 script I think that's based on GPD for
1:17:24 om
1:17:26 sure thing Victor let's look at four
1:17:28 scripts conveying four different
1:17:30 emotions she's fake um I don't think I
1:17:33 can do it I'm afraid I don't have enough
1:17:35 experience doubt this is incredible news
1:17:39 I am so excited about this opportunity
1:17:42 okay I got to stop this this is this is
1:17:44 so [ __ ] science
1:17:47 fiction this
1:17:50 is this is what Hollywood imagined these
1:17:53 things would be like
1:17:58 like this is the [ __ ] that you see on
1:18:01 screens when the character in a science
1:18:03 fiction movie walks into the big
1:18:05 Corporation and it's like it's like
1:18:08 don't forget to smile it's part of what
1:18:10 we
1:18:12 do imagine how frustrated I am about
1:18:14 this whole situation this is
1:18:17 unacceptable I just wanted to say thank
1:18:19 you for giving me this opportunity I am
1:18:21 so grateful for everything you've done
1:18:23 for me Express avatars are available
1:18:26 today so try them out now back to you
1:18:30 Victor thank you
1:18:32 aah so what's coming for avatars a few
1:18:35 things first up we know that many of our
1:18:38 customers want to have their own avatar
1:18:40 and this have of course been possible
1:18:42 with sesia for a while but with sesia
1:18:44 2.0 we're making it a much easier
1:18:47 experience and we're significantly
1:18:48 increasing the quality and capabilities
1:18:50 of your own avatar with udia 2.0 you'll
1:18:53 have three ways of creating your own
1:18:55 personal Avatar we go into a studio if
1:18:57 you want the more professional field
1:18:59 kind of like the avatars we have on the
1:19:00 platform or you can use your webcam or
1:19:03 your phone at home we're also giving you
1:19:05 some new capabilities first up you can
1:19:08 keep the natural background of your
1:19:10 video if you want to for some use cases
1:19:12 this can be really really really
1:19:14 powerful we now also give you a clone of
1:19:17 your voice but not just in the language
1:19:20 you recorded in but in more than 30
1:19:23 languages that's right Victor you can
1:19:25 now create your personal Avatar anywhere
1:19:27 with just a few minutes of
1:19:35 footage
1:19:38 stuff so what's happening like what what
1:19:41 this what this is doing is this is
1:19:43 bringing these guys in parody with ha
1:19:45 Jen which by the way ha Jen just raised
1:19:47 $60 million or $300 million I don't know
1:19:50 some shitty amount of millions of
1:19:52 dollars um so but this is this is very
1:19:57 much in line with with what they're
1:19:59 doing so I think you'll probably see the
1:20:01 same from D- so you'll have three major
1:20:04 players that can all do real avatars
1:20:07 based on you real time conversation with
1:20:10 them with all sorts of emotion so just
1:20:13 like we saw yudo and Sun kind of sunno
1:20:16 kind of duking it out um for features
1:20:20 we're g to now we're seeing this with
1:20:21 these video guys and we're also seeing
1:20:24 this with mid Journey sunno that Chinese
1:20:27 thing cing right so all of the different
1:20:31 media modalities are now starting to
1:20:36 emerge with capabilities that are at the
1:20:40 same kind of level that chat gp4 is at
1:20:43 right and so and so now we're going to
1:20:45 start to see all of these things go in
1:20:47 parallel and then it's some point you
1:20:50 they're going to go in parallel for a
1:20:51 while and then they'll start to converge
1:20:53 well where we will have a single
1:20:55 interface that brings all of this [ __ ]
1:20:58 together creep now get ready to be
1:21:00 creative personal avatars are coming to
1:21:03 Synthesia in July
1:21:05 2024 I'm also excited today to share a
1:21:08 glimpse into the future of AI avatars
1:21:12 later in this year we're going to have a
1:21:13 big platform shift over the last 12
1:21:16 months we've been capturing thousands of
1:21:18 people in our Studios all over the world
1:21:20 and with this data we've been training
1:21:22 several large Foundation models that can
1:21:24 now work together in lock step to
1:21:26 produce incredibly realistic and
1:21:28 engaging avatars later in the year when
1:21:30 we push this technology out you'll find
1:21:33 that avatars will become great
1:21:35 storytellers they'll be able to use
1:21:36 their hands be in environments they'll
1:21:39 be a much much more natural and almost
1:21:42 hyper real experience to use these
1:21:44 avatars wow another thing we learned
1:21:46 from our customers is that most video
1:21:48 editing tools are designed for
1:21:49 professionals and require extensive
1:21:51 training with Synthesia we've
1:21:53 dramatically simplified the in process
1:21:56 without compromising on flexibility in
1:21:58 fact n out of 10 people can create their
1:22:00 first video in less than 10 minutes
1:22:03 without any prior experience we've
1:22:05 achieved that by replacing the
1:22:06 traditional video timeline with simple
1:22:08 triggers that you can control directly
1:22:10 from your script this change puts your
1:22:12 script at the heart of the story
1:22:13 allowing you to animate video elements
1:22:16 and make edits in a simple and intuitive
1:22:18 way it also simplifies seen content
1:22:20 generation creating a whole new editing
1:22:22 experience that's easy to use for
1:22:24 everyone but what we also learned is
1:22:26 that many of our customers need to
1:22:27 include screen recorded content in their
1:22:29 videos but find the process today really
1:22:32 complicated you have to use multiple
1:22:34 tools to capture your screen edit the
1:22:36 recording match the voice over wait I
1:22:38 got to pause this Jim Ross three M
1:22:41 storage someone searched storage and Ai
1:22:44 and found my presentation and became a
1:22:47 client today play and share
1:22:51 yes what I tell everyone to do here play
1:22:55 first learn the tools just play play
1:22:58 play play play second mindfully create
1:23:01 build ship for your business Jim
1:23:03 Ross collapses those two often in the
1:23:06 same day he'll learn something and then
1:23:09 build something for his business and
1:23:11 then put it in the world
1:23:13 share he's been doing that regularly for
1:23:16 a year now he's gotten known as like the
1:23:19 AI guy guy in the whole Storage self-
1:23:23 storage business world
1:23:25 you're like that's weird that's not
1:23:27 about AI
1:23:29 exactly so he's [ __ ] owning that
1:23:34 corner so play mindfully create and then
1:23:38 generously lead put it out in the world
1:23:41 share anyway congrats Jim that's [ __ ]
1:23:44 awesome and need to update it you have
1:23:46 to start all over again we believe there
1:23:48 a better way and I'm excited to
1:23:50 introduce you to our AI screen recorder
1:23:52 and show you how it works AI screen
1:23:54 imagine I want to create a step-by-step
1:23:56 guide for synthesi on how to book their
1:23:58 well-deserved time off through our HR
1:24:00 System here's what the process would
1:24:02 look like yeah this is Synthesia by the
1:24:04 way remember it's important to take time
1:24:06 off and spend it with your loved ones
1:24:08 here are three simple steps on how to do
1:24:10 it after you log into our HR System
1:24:13 click on request time off next select
1:24:15 the type of request and the dates that
1:24:17 you want to take off then click submit
1:24:19 and that's it enjoy your vacation now
1:24:23 once I'm done with the recording my is
1:24:25 immediately available for editing with
1:24:27 Device over transcribed perfectly
1:24:29 matching the screen capture and
1:24:30 automatic Zoom effects to emphasize atic
1:24:33 Zoom effects nice from there I can edit
1:24:35 the script if needed I can trim the
1:24:37 video I can even add my own avatar and
1:24:39 voice for a personal touch ni the result
1:24:42 is a Sleek high quality video that can
1:24:44 easily be updated AI screen recorder is
1:24:46 coming to synesia in the next few months
1:24:48 and I'm really excited to see what you
1:24:49 all think about it but nothing great is
1:24:52 made alone in today's world where teams
1:24:55 are distributed globally effective
1:24:56 collaboration is crucial in order to be
1:24:58 able to align and quickly respond to
1:25:00 changes that's why we streamlined both
1:25:03 real time and asynchronous teamwork
1:25:05 inesia we can easily invite colleagues
1:25:08 we can co-edit videos in real time it
1:25:10 can provide feedback all in one place
1:25:12 which significantly speeds up the
1:25:14 content delivery process but there's one
1:25:16 more thing out of 4.2 billion internet
1:25:19 users only about 25% are English
1:25:22 speakers in a world where employees and
1:25:24 customer customers are distributed
1:25:25 globally adapting communication to local
1:25:28 languages and cultures is not just an
1:25:30 option is this free I doubt it none of
1:25:32 these things are free and they're they
1:25:33 tend to be pretty expensive they tend to
1:25:35 be
1:25:37 um uh like you get some amount of
1:25:40 credits for your monthly fee and then
1:25:43 and then you know every minute is a
1:25:45 credit or something like that is is how
1:25:47 they tend to go so probably not free
1:25:49 it's a massive business opportunity from
1:25:52 speaking to our customers we've learned
1:25:54 that translations today is a really
1:25:57 complicated process you have to find a
1:25:59 translation partner you have to send
1:26:01 them the content we have to wait for the
1:26:03 translations you have to review them
1:26:05 this can take weeks or even months
1:26:07 delaying important Communications and
1:26:09 increasing costs about a year ago we
1:26:11 introduced our oneclick translations
1:26:13 feature in tesia which enables you to
1:26:15 automatically translate your videos into
1:26:17 over 120 languages with literally just
1:26:19 one click and even though that unlock
1:26:22 massive productivity gains for our
1:26:23 customers they still have to manage and
1:26:25 maintain and share multiple files which
1:26:28 wasn't good and not really that
1:26:30 practical today I'm excited to introduce
1:26:33 the updated translation experience
1:26:34 inesia you simply create one version of
1:26:36 your video translate it into any other
1:26:38 language you want and if you need to
1:26:40 update your video you just make changes
1:26:42 to the original version all other
1:26:44 language versions are connected to this
1:26:45 main version and they'll update
1:26:47 automatically amazing but that's not all
1:26:49 there's one more feature which is
1:26:51 something I personally have really been
1:26:52 looking forward to and I know it's
1:26:54 highly anticip ipated by our customers
1:26:56 we're building a new type of video
1:26:58 player one that we believe will enable a
1:27:00 new generation of video experiences that
1:27:02 are interactive personalized and fun the
1:27:04 first feature we're launching next month
1:27:06 is the ability to Simply share your
1:27:08 video and our player will automatically
1:27:10 played in your viewers language it's
1:27:12 quite magical and truly complements our
1:27:14 translation capabilities later in the
1:27:16 year we're launching a whole Suite of
1:27:18 interactive capabilities for our player
1:27:20 we'll be able to create Rich video
1:27:22 experiences with features such as
1:27:23 clickable Hotpot
1:27:25 embedded forms quizzes and personalized
1:27:27 call to actions these capabilities will
1:27:30 make your videos more engaging Drive
1:27:32 higher viewer interaction and unlock use
1:27:35 cases that simply aren't possible today
1:27:38 this product is called Synthesia I'm
1:27:40 incredibly excited for all of you to try
1:27:42 it out and see all the cool things
1:27:43 you'll be
1:27:45 creating welcome to the interactive
1:27:48 world of cesia no this is not what I I I
1:27:51 think that's it for the announcements
1:27:53 let me see
1:27:59 abilities it's incredible to see how our
1:28:01 Solutions are creating very real
1:28:10 value that's use cases okay that's it
1:28:14 um holy [ __ ] Kyle Kyle Kyle two days in
1:28:18 a row is too much time
1:28:21 off philosophy 401
1:28:32 um yeah that's a pretty big deal there's
1:28:34 a lot of stuff
1:28:36 there a lot of stuff there lot lot lot
1:28:39 of stuff there
1:28:49 um it's just it like we're moving
1:28:58 you know it's funny I was talking
1:28:59 earlier about the thing that's most
1:29:02 profound about AI for me right now is
1:29:05 being able to get ideas out of my head
1:29:08 like you know in my head manifested
1:29:11 right and and a lot of the stuff that he
1:29:14 showed right there like that instant
1:29:15 translation thing where you just produce
1:29:18 a video once and then all the other
1:29:20 languages are there and it just
1:29:22 dynamically switches the speaker into
1:29:25 that language and the text on screen
1:29:26 into that language
1:29:29 it's any idea instantly accessible to
1:29:34 any audience and not just languages
1:29:36 right because this is going to be
1:29:38 there's going to be accessibility for
1:29:40 blind people for deaf people for um
1:29:43 cognitively impaired
1:29:45 people right like like the the the level
1:29:49 of accessibility of so so so maybe here
1:29:52 here's the this this gets back to the
1:29:54 the blessing part of AI
1:29:57 is one of the challenges of humans is
1:30:00 getting ideas out of our heads and
1:30:02 manifested and then the other one is
1:30:04 sharing those ideas with the world and
1:30:07 connecting with other
1:30:08 people and AI is doing both of those
1:30:11 what that Synthesia thing did said said
1:30:13 to me is I can take some idea I'd like
1:30:16 to talk about my company please put a
1:30:18 URL and here it spits out a bunch of
1:30:20 explainer videos that are instantly
1:30:23 accessible to anyone who [ __ ] wants
1:30:26 them and right now it's languages but it
1:30:29 might be like let's make those relevant
1:30:31 for a 10-year-old or you know someone in
1:30:35 nuclear physics versus someone in car
1:30:39 wash
1:30:44 technology again I try I try to look at
1:30:46 all these Technologies not for the their
1:30:48 specific features but like what is the
1:30:51 what is the trend here what's the
1:30:53 where's this stuff
1:30:55 headed holy
1:31:00 [ __ ] Dr J 0 $22 and $67 I guess that's
1:31:04 for Synthesia I've got to tell you oh I
1:31:08 have to tell you what chat GPT did for
1:31:10 me awesome what did it
1:31:15 do that sounds
1:31:17 cool I hope it helped you screw with
1:31:21 those stupid water people
1:31:26 Danielle's trying to start a business
1:31:27 and her her Town Council is being a
1:31:30 bunch of tools it was able to help me
1:31:34 locate the main waterline on my
1:31:38 blueprints because my contractor
1:31:44 couldn't that's [ __ ] awesome that is
1:31:48 [ __ ] awesome you just took a picture
1:31:50 of with your
1:31:51 phone that is awesome
1:31:55 uh
1:32:04 uh oh my eyes I need water or something
1:32:06 like that my eyes
1:32:08 are leave it to AI to do the
1:32:12 job all right all right people I think
1:32:16 I'm going to get on out of here
1:32:20 um I didn't talk it all about the AI
1:32:22 Salon tonight but let me just talk about
1:32:24 it briefly if you go to that
1:32:31 URL where's my
1:32:34 mouse go to the salon.
1:32:37 a right
1:32:40 there that's going to take you to a link
1:32:45 tree the very first link says AIS Salon
1:32:49 Community Pi kept my kids engaged for 30
1:32:52 minutes on a road trip this weekend see
1:32:54 a is a miracle it's not a blessing it's
1:32:56 a
1:32:59 miracle entertain the kids okay and you
1:33:02 know it's going to get better and better
1:33:04 and better at that Brandon I mean that
1:33:06 [ __ ] that he was just talking about
1:33:07 right here so imagine a version of
1:33:09 Synthesia where instead of it
1:33:11 translating into languages it translates
1:33:13 it into different Persona types you know
1:33:18 a fidgety add you know
1:33:23 5-year-old and a you
1:33:25 know romance novel loving 13-year-old
1:33:29 right you
1:33:32 know oh my God that's
1:33:35 crazy she even asked to
1:33:39 followup notification this morning crazy
1:33:42 amazing that's that's very cool anyway
1:33:44 go to the salon. a click on the click on
1:33:47 the uh community and go check it out
1:33:50 check out about the salon it'll tell you
1:33:52 what we're about it's we're really cool
1:33:54 the motto of the AI Salon is play
1:33:57 first why is that because there's too
1:34:00 much going on to like figure this stuff
1:34:02 out you can't learn at all so play what
1:34:05 do I play with play with whatever you
1:34:07 want play with anything that interests
1:34:09 you
1:34:10 don't don't attach any outcomes to
1:34:13 learning AI just start playing with it
1:34:16 and then as you play with it you'll
1:34:18 accidentally learn some [ __ ] and in
1:34:20 accidentally learning some [ __ ] you'll
1:34:22 have an idea you'll be like oh you know
1:34:25 I could probably do this with that and
1:34:27 then go do that and that second phase is
1:34:30 what we call mindfully create imagine a
1:34:32 problem that you could solve with the
1:34:34 [ __ ] you
1:34:36 know and then start building that or
1:34:39 making
1:34:41 that and then the third phase of the
1:34:46 Journey of AI
1:34:48 literacy is generously lead and what
1:34:50 that looks like is if you get stuck ask
1:34:53 for help
1:34:54 so being F vulnerable and asking for
1:34:57 help is a form of leadership because it
1:34:59 allows someone else to practice leading
1:35:04 you or the other thing is share what
1:35:07 you're learning learn out loud so as
1:35:09 you're mindfully creating as you're
1:35:11 building the thing as you're solving the
1:35:13 problem go to
1:35:15 Facebook I had this idea I I was playing
1:35:18 around with AI and then I learned this
1:35:20 little thing and now I'm trying to build
1:35:21 a thing and it's this is really amazing
1:35:23 or
1:35:25 I think this AI stuff's pretty cool but
1:35:27 I was trying to build a thing and this
1:35:28 stuff is janky share that [ __ ] put it
1:35:31 out
1:35:36 there it will serve you and it will
1:35:38 serve
1:35:39 others because you will have a
1:35:41 perspective that will seem obvious to
1:35:44 you that is revelatory for someone
1:35:50 else and you never know what that is
1:35:52 it's sometimes the smallest little thing
1:35:55 unlocks what this stuff makes possible
1:35:57 for
1:35:58 someone thoughts on Runway ml for making
1:36:01 videos they Gen 2 looks amazing they
1:36:03 said it would be out in a few days and
1:36:04 it's not [ __ ] out yet I'm tired of
1:36:07 these companies showing us [ __ ] that we
1:36:09 can't play
1:36:10 with just shut up about it until we can
1:36:14 play with it look how cool this is log
1:36:18 in here play
1:36:20 today that's all I want to hear from now
1:36:22 on I don't want I don't want [ __ ]
1:36:24 previews of what's coming which half of
1:36:27 that [ __ ] was since theia was later this
1:36:32 year we're innovating
1:36:34 too just give me the [ __ ] I can play
1:36:37 with all
1:36:38 right you all got to try the latest Mad
1:36:41 Lib and try to put your words in the
1:36:42 image okay so when you go to the AI
1:36:44 Salon Danielle Be not Afraid here she
1:36:48 does these things called um Mad Libs
1:36:50 with Danielle so so if you don't know
1:36:52 what Mad Libs were they were like a
1:36:53 little paragraph you replace the words
1:36:57 and so she does image prompts she crafts
1:36:58 these really clever image prompts that
1:37:00 you swap out the words and then
1:37:02 everyone's doing images in the same
1:37:04 general area because it's all the same
1:37:07 paragraph but lots of different subject
1:37:09 matters so that's one of the things you
1:37:10 can do here you can learn in here you
1:37:12 can connect with other people in here so
1:37:14 go there if you like what you see
1:37:16 introduce yourself doesn't cost
1:37:19 anything it's an amazing Community all
1:37:22 right thank you to the regulars all you
1:37:24 people who show up here night after
1:37:26 night after night after night by the way
1:37:28 if you're new
1:37:32 here and and if you've been watching for
1:37:35 a while and you're like I don't know why
1:37:36 I'm watching like what happens here does
1:37:39 does he actually teach any it's called
1:37:41 the AI learning lab I didn't did does he
1:37:43 actually teach
1:37:44 anything this isn't about
1:37:48 me I just ramble sometimes I teach [ __ ]
1:37:52 sometimes I get really inspired to make
1:37:53 [ __ ] [ __ ] we're almost at
1:37:57 25,000 what this has
1:38:00 become and why it's so important is that
1:38:03 this is a
1:38:05 time for us all to be in this
1:38:09 conversation the conversation where
1:38:11 you're were talking about poo poo and
1:38:13 what why would we want to be in that no
1:38:16 the AI
1:38:17 conversation just being in the
1:38:19 conversation just even if you're not
1:38:21 learning if you like came in halfway
1:38:23 through that video and you're like what
1:38:24 was that what was he showing why was he
1:38:26 doesn't just keep coming back keep
1:38:28 coming back keep coming
1:38:30 back because as you're in this
1:38:34 conversation I will inevitably within
1:38:36 the first five minutes bore the living
1:38:38 [ __ ] out of you and what but what you'll
1:38:41 do is you'll keep me on in the
1:38:42 background and you go want me go try
1:38:44 chat GPT well what could I do uh why
1:38:48 don't you make me a computer program oh
1:38:51 we can make computer programs I didn't
1:38:53 know I could do
1:38:56 that and I don't know why you talk like
1:38:59 that but I think it's kind of
1:39:01 humorous so just keep coming back here
1:39:04 okay thank you very much ton skier
1:39:06 appreciate that um it prepared me for
1:39:09 things I never knew I was capable of
1:39:11 that's why you play first because what
1:39:13 you discover is you discover things that
1:39:16 you're now capable of you never would
1:39:19 have [ __ ] imagined so you just keep
1:39:22 coming back here and you keep playing
1:39:23 keep playing use this time to practice
1:39:27 to play to
1:39:30 think to ideate to get an idea out of
1:39:33 your [ __ ] head all right thank you
1:39:37 mods really good job tonight like all
1:39:39 the like I know I was on a [ __ ] rant
1:39:41 but like all the stuff that came up was
1:39:42 relevant to what I was talking about so
1:39:44 it was really good thank you um has
1:39:46 anyone heard any new updates on the new
1:39:48 voice features I I heard there was some
1:39:49 Rumblings that the the Mac App started
1:39:53 some people were getting new features in
1:39:54 them today they weren't it it was the
1:39:57 like the settings for the new voice
1:39:59 stuff people were getting but I don't
1:40:00 think anyone that I saw today is getting
1:40:02 it but it might be coming it sounds like
1:40:05 something's
1:40:06 coming what's today the
1:40:09 24th something's coming this week maybe
1:40:12 and then something's coming the middle
1:40:13 of July from open AI but I don't [ __ ]
1:40:16 know I'm tired of trying to predict them
1:40:19 um thank you kind AI family I'm
1:40:22 listening great awesome
1:40:24 um all right I'm out of here follow my
1:40:27 channel subscribe to my lives pick up
1:40:31 the uh video series in the corner there
1:40:33 if you want to support you know if you
1:40:35 if you subscribe to my lives or if you
1:40:37 buy that thing that directly supports me
1:40:40 directly supports the channel it's
1:40:42 appreciated cuz I put in a lot of hours
1:40:44 here
1:40:46 um and I'll keep doing it but uh but
1:40:50 yeah that support's appreciated um I
1:40:53 think that's it
1:40:54 check out the AI Salon follow
1:40:58 subscribe yeah that's it all right I'm
1:41:02 out of here people thank you thank you
1:41:04 thank you everybody let me do a little
1:41:05 whoot in the house tonight Silver Fox
1:41:07 was here Genie in a Bottle side hustle
1:41:10 Mimi Wolfman Clint some damn good names
1:41:13 Be not Afraid Judy ARS thank you for
1:41:18 coming very nice Mary Mary recycle reuse
1:41:21 good night good to see you steo and the
1:41:23 house joy py in a house
1:41:26 Jason Jay Hamilton great to see you I'm
1:41:29 running a weekly AI Lab at work now did
1:41:33 a learning session this week Jay
1:41:34 Hamilton awesome that's generously
1:41:37 leading that's learning out loud that's
1:41:39 perfect awesome Papa JJ great to see you
1:41:44 30,000 nice nice thank you for the likes
1:41:47 I appreciate that Brandon thank you for
1:41:50 uh for doing what you do as always uh
1:41:53 let's see who else we got in the
1:41:56 house Silver Fox
1:42:01 Joey Jason uh Mr it is here Lord digital
1:42:07 Gods archetypal archetypal architect
1:42:10 Source Camp Teton skier
1:42:13 manred man it's a big night tonight Jay
1:42:16 Hamilton uh whoever's got the farsy I
1:42:19 don't know how to say I think it's farsy
1:42:22 I don't know how to say what it says but
1:42:24 thank you for being here I know you hear
1:42:26 a lot you always ask really good
1:42:27 questions uh so maybe in parentheses put
1:42:31 English and I can I can call you out cuz
1:42:34 right now it's just squiggle Mark
1:42:35 squiggle mark it nothing registers in in
1:42:38 this old Noggin I need a I need a
1:42:40 Synthesia
1:42:42 translation um all right everybody um
1:42:45 really fun tonight thank you uh thanks
1:42:47 for putting up with my uh rant Monday
1:42:50 night rant all right peace out