Jun 14, 2024 - AI Creativity: Songs, Comedy, and the Future of Entertainment

Description

Chapters

Transcript

0:00 ready to sing ready to
0:04 [Music]
0:26 [Music]
0:28 sing hey hey hey hey hey very aggressive
0:33 very
0:34 aggressive what's with the
0:37 [Music]
0:45 rapping what are you
0:53 [Music]
0:57 doing w
0:59 [Music]
1:14 in a Westerly
1:19 Direction this car is my
1:23 train I've been driving I've been
1:28 wondering what it is I'm running from
1:34 again feel like an 80-year old
1:38 man holding on to
1:43 29 and a head on that
1:48 R is a California
1:52 [Music]
1:58 line w
2:00 up ahead of trucks carrying a wide
2:04 load preab house cutting
2:07 [Music]
2:08 half cute little front door and windows
2:12 my
2:13 love ain't sure whether the dri or
2:15 should
2:18 left see I broke a home up myself
2:22 once there I Stumble to that
2:27 door I read a note by the dawn
2:32 life said don't you come around
2:37 here
2:40 anymore well I
2:44 enough how this freom the
2:48 RO never was good with
2:53 decision least that's what I've been
2:56 told
3:00 [Music]
3:03 good evening good people of the internet
3:06 or whatever the hell this is It's
3:08 streaming live over the Internet on the
3:10 Tik Tok soon to be banned fantastic Bob
3:14 tell him what he's
3:16 [Music]
3:19 W you've won the only bipartisan bill in
3:22 the history of Congress in the last
3:25 decade Tik Tock evil let's get it
3:31 it's so bad to Tik Tok you come on here
3:33 you can watch old men strumming their
3:35 guitar and having their dogs sing Oh my
3:37 God the propaganda it's terrible it's
3:40 terrible we should just ban
3:43 it because you know why privacy CU PR
3:46 you know what it is the Chinese they're
3:50 stealing all your
3:53 data which technically they bought from
3:57 Facebook and Google but that's not the
3:59 point that is not the point they're
4:01 stealing new fresher data it it's newer
4:05 fresher data that they're stealing and
4:07 it's so it's got to go it's got to go
4:09 the Tik tok's got to go Tick Tock right
4:13 [Music]
4:20 out for my next number
4:27 [Music]
4:38 what are we going to do
4:40 tonight Tick Tock spends spreads too
4:42 much truth ban it ban it ban the truth
4:45 tellers we shall never spake the truth
4:49 speaking the truth is very concerning to
4:52 the interests of the vestages that shall
4:55 be leved upon us henceforth with the
5:02 yeah you
5:05 [Laughter]
5:10 know oh
5:12 [Music]
5:18 [Applause]
5:20 [Music]
5:27 [Applause]
5:29 [Music]
5:49 big day today today was my last day of a
5:53 one-year planning for a group's World
5:55 Tour
5:57 wow beautiful that's very cool is it all
6:01 planned
6:02 [Music]
6:06 up big world tourus oh [ __ ] what's going
6:09 on with the
6:13 hair the hair looks
6:17 [Laughter]
6:24 intentional I get my hair done by
6:27 AI it's clearly hallucinated
6:37 welcome everybody if you're new here my
6:38 name is Kyle Shannon this is the AI
6:40 learning lab it is a little warm here in
6:42 Denver I have the air on but hasn't
6:44 kicked in yet let me make sure it's on
6:46 cool is it un cool it's un
6:49 cool come on air conditioner do your
6:53 thing I can't be moist for the
6:57 people moist is right out
7:02 all
7:03 right good evening good people I'm it's
7:06 giving Elvis yeah it's giving Elvis at a
7:11 time oh boy oh that was bad all right
7:15 well I'm I just I can't touch it
7:17 touching it's right
7:22 out it's the evaporative cooling or
7:25 something yeah something's going on
7:27 something is going on Friday night date
7:29 night
7:31 um I saw nothing about AI today I was
7:35 busy so I saw I saw very little so
7:38 anything happening why why don't you
7:40 guys tell me what's happening anything
7:42 anything out
7:46 there all right fireflies in house
7:49 welcome welcome welcome atie bubbles
7:51 made of GPT nice nice play first
7:56 mindfully create there you go Building G
8:00 pts is mindfully
8:02 creating I did a
8:04 uh I I'll show you a cool one today so
8:07 we've got this um we made these
8:10 individual gpts for this group I'm in
8:13 called content
8:14 Evolution um we have this little AI
8:17 working group and so we made all these
8:19 digital
8:21 Twins and then we made a master GPT that
8:25 would recommend people and then we
8:27 actually kind of melded them all
8:28 together so where's the rabbit the
8:30 rabbit is in the other room but the
8:34 there's the the rabbit stand that Vicky
8:35 printed me is right there I just don't
8:37 have the rabbit in it I could go get it
8:39 put it in
8:40 there
8:45 um what was I talking
8:48 about what was I talking about I got
8:51 rabbit distracted can you make a
8:54 consistent character in
8:55 video not really but you can make a
8:59 consist consistent character in mid
9:02 Journey with character reference and
9:05 then you can use that character the the
9:07 mid Journey image as the basis for video
9:09 so yeah yeah you sort of
9:11 can after telling you no you can't do
9:14 that you you sort of can it's not
9:18 great it's not
9:20 great oh the GPT thing that's right
9:25 um so now we've got this master GPT that
9:28 has all of the people in this one
9:32 GPT and we're starting to figure out
9:34 some [ __ ] about how to how to use
9:38 it and like one of the really cool
9:40 things is so we we we prompted it to be
9:45 a
9:46 facilitator of conversations between
9:49 these digital
9:50 Twins and so now you can say things like
9:53 create a round table of the five people
9:56 you'd you know who are best suited to
9:58 talk about this topic
10:00 and then it steps into the role of
10:02 facilitator and it asks different people
10:04 in the group it's pretty cool I I'll
10:05 show you we'll do
10:07 one so we can play with
10:10 that um we can go look at news I mean we
10:13 can go look at uh Twitter and see if
10:16 anything happened today but let's see
10:17 former former head of NSA joins the open
10:20 AI board yep so I read in Mighty
10:23 networks I don't know what that means I
10:25 don't
10:27 either they're taking
10:30 security seriously they need Connections
10:33 in the government so they don't get
10:34 their services
10:37 gutted
10:39 because they tell too much truth I don't
10:43 know I don't
10:45 know I'm too cynical to exist in this
10:48 world I think everything's a [ __ ]
10:52 angle it's
10:58 just looking for an AI app that can play
11:02 sound what do you mean play
11:05 sound do you mean
11:08 um can make like sound
11:12 effects cuz stable diffusion just put
11:15 out stable audio which is open source
11:19 free um and can do 47 second clips of
11:24 sound effects 11 Labs now has sound
11:27 effects if you're think if you're talk
11:29 about music is it sound like music you
11:32 want AIS that make music that's a
11:34 different thing hang on my my wires are
11:38 all discombobulated people hang on
11:43 people got to figure out what the hell
11:45 is going on
11:48 here all
11:51 right create music Etc all right we've
11:54 been doing a lot of that lately so the
11:56 two biggies right now are sun.com s o
12:01 and ud.com
12:03 so here we'll
12:06 go we'll go make
12:09 one uh uh uh let's
12:12 see let's do
12:23 sunno all right so I'm in this thing
12:25 called
12:28 sunno darken my my
12:38 screen all right upload
12:41 audio I to record
12:44 something let's record something what
12:46 are we going to
12:48 record oh I know what we can do I got an
12:54 idea
12:57 um oh okay wait how are we going to do
13:00 this hang on people it's going to be
13:06 good okay let's see so we'll just do
13:10 [Music]
13:17 like wait how we going to do
13:22 this
13:26 make oh wait maybe we can get champy to
13:28 sing all right let's let's get champy to
13:30 sing so what are we going what are we
13:31 going to do champy
13:34 [Music]
13:35 uh okay you ready champy you ready to
13:38 sing okay here we go here we go
13:41 [Music]
13:53 ready oh wait I didn't record it dang it
13:55 okay we're going to we got to start over
13:57 ready
13:59 [Music]
14:07 [Music]
14:11 okay so we have La
14:15 [Music]
14:18 mus all right
14:21 beautiful I know champy it's very
14:23 annoying okay so we're going to upload
14:25 that
14:31 sing champy sing all
14:37 right all right let's see
14:43 champ uh champ
14:51 [Music]
14:52 sings okay so then I go Library I find
14:56 Champ things I go extend Okay so
15:00 we're now going to extend this little
15:02 song that we
15:05 made and we're going to make we're going
15:07 to make some lyrics so I'm over in chat
15:10 JP excuse me chat
15:12 J I'm going to say uh WR an
15:20 award
15:23 winning song
15:26 about um
15:31 a
15:33 dog singing
15:36 for chicken
15:39 [Laughter]
15:44 strips chicken strip dreams I like
15:49 it all
15:54 right so now we're going to copy all
15:57 this well that's a lot but we'll see
15:59 we'll see if it can do
16:05 [Music]
16:08 it all
16:17 right all right there's our
16:20 lyrics and then we're going to
16:23 go guitar
16:29 let's see uh
16:33 folk singer
16:36 songwriter I know it's boring I always
16:38 do the same [ __ ] but whatever
16:42 acoustic and then we'll call this uh
16:47 what was what did chat GPT call
16:50 it chicken strip dreams all right we'll
16:53 go with
16:56 that you're like is this guy ever going
16:58 to get to the point here yeah and then
17:01 we go create and now it's going to do it
17:05 boring we're all still here oh yeah
17:07 there's a bunch of people in here
17:08 welcome welcome
17:17 [Music]
17:25 welcome lots of love in the house
17:27 tonight everybody checking in with each
17:29 other very nice to see I love that bored
17:33 dog got love that my dogs heard him and
17:35 started howling too nice well let's see
17:37 how the song okay here's chicken strip
17:38 dreams here we go
17:42 People application error nice sweet
17:47 perfect sweet
17:51 awesome okay all right Mr
17:55 sunno are we uh
17:59 all right hang on I think we got to
18:02 quit quit Google Chrome there's a
18:12 doggy by the way if you want to find a
18:16 it wrote digging not singing oh that's
18:18 okay I sent you a DM about your
18:21 prediction that came true oh interesting
18:23 okay I'll go check that out um you want
18:25 a cool Community Check out the salon. a
18:28 click on the Community Link at the top
18:30 there and this is a group of AI
18:33 enthusiasts that are really cool people
18:37 um hang on let me get Chrome back up and
18:46 [Music]
18:49 running let's say oh I know what I need
18:53 to do
18:57 [Music]
19:02 okay can we join this now without it
19:05 crashing no all right looks like sunno
19:08 is in a world of
19:15 hurt but I
19:17 bet I can just play it from this page a
19:21 quiet little town with The Grass Is
19:23 Always
19:25 Green there's a curious PO with a
19:28 loveful Quine
19:30 [Music]
19:31 he sniffs the morning air his tail
19:34 starts to
19:35 whip he's got a h for those chicken
19:41 strips
19:46 deep
19:49 knows chicken stri
19:53 [Music]
19:55 dreams all right yeah we got to make it
19:57 we got to make it singing not
20:00 scy it's cute but it's the wrong lyrics
20:02 all right let's go back to chat gbt hang
20:05 on all
20:08 right oops you made the song about him
20:11 digging and it needs to be about him
20:16 singing chicken strip serenade
20:20 okay he sings with all his might for
20:23 those St chicken strips tonight Melodies
20:25 In The Air for the treat Beyond Compare
20:28 there right good beautiful fine fine
20:35 people by the way this is suo s n o if
20:40 you're wondering wait what did he say
20:42 what what tool is
20:44 this um champ
20:51 sings now I can't extend this this
20:54 interface is really weird and
20:57 janky now I can extend it
21:03 okay wait what the hell's going
21:17 on let's do
21:19 um Americana
21:22 rock singer song
21:27 writer um
21:31 rough
21:41 voice
21:43 um
21:45 groovy
21:47 pocket I don't know what that
21:50 [Laughter]
21:54 means chicken
21:56 strip Serenade
22:00 all right bang let's knock this out
22:02 let's get some better songs going here
22:05 people is that rough voice or rough
22:07 voice oh that's good Janu that's really
22:13 good that's comedy that's comedy with
22:15 language right
22:19 [Music]
22:23 there we have a Jennifer Lopez in the
22:25 house Jennifer
22:26 Lopez JLo JLo will JLo please check in
22:31 with the conci JLo
22:41 [Music]
22:43 JLo
22:49 oh yeah they got
22:52 some they got some [ __ ] going on sunno
22:56 they're up updating something a cry to
23:00 with the
23:03 night lives a with a voice that's quiet
23:06 un for
23:09 se he sits by the porch under the
23:12 Moonlight
23:15 ble dream met a chicken strips it's his
23:18 favorite
23:19 them oh he sings with
23:23 all it's okay
23:29 there's a par with a voice that's quite
23:31 unforeseen he sits by the porch under
23:34 the Moonlight CLE dreaming a chicken
23:36 strips his favorite theme oh he sings
23:39 with all his for those chicken strips
23:42 tonight Melodies In The
23:47 [Music]
23:50 Air all right so that's sunno and then
23:53 yudo is similar and it's got a it's got
23:56 a song extend feature as well but it it
23:58 doesn't do very good with that but this
24:00 one we could
24:04 say an Americana rock
24:09 song about a dog that sings for his
24:12 chicken
24:18 strips and then
24:21 we'll let's see
24:49 maybe try a commercial jingle yeah
24:51 that's not bad we could do a commercial
24:53 jingle
25:06 a
25:08 commercial jingle from
25:13 the80s about a song that sings for
25:15 chicken strips that's
25:21 good hey
25:28 [Music]
25:31 have you been dude good good Jennifer
25:34 thank you for
25:37 [Music]
25:40 asking janky Breaky Heart janky branky
25:44 [Laughter]
25:51 [Music]
25:55 heart in a quiet little town where the
25:58 night
25:59 areem lives a pup with a voice that's
26:02 quite on he's by the porch under The
26:04 moonlight's Gleam dreaming of chicken
26:07 strips it's his favorite
26:10 theme oh he sings with all his
26:14 might for those chicken strips
26:19 tonight so bad in a quiet little town
26:24 where the nights are Serene lives a P
26:27 with a Poise that's
26:30 [Music]
26:36 it's his favorite theme oh he sings with
26:40 all
26:41 his for those chicken
26:44 [Music]
26:45 [Laughter]
26:48 strips Gavin
26:53 bitty Luma lum's uh dream dream machine
26:57 yeah dream machine's pretty interesting
27:00 my Creations let's hear the commercial
27:06 Jingles town where the nights are Serene
27:10 a with a voice that's
27:13 unfor he
27:21 sits town where the nights are Serene
27:25 lives a pup with a voice that's quite
27:27 unforeseen he sits by the porch under
27:30 the Moonlight screen Dre of chicken
27:33 strips it's his favorite them oh he
27:36 sings with all his might for those
27:39 chicken strips
27:42 tonight Melodies In The Air for the
27:46 treat beond
27:49 [Laughter]
27:51 compare all right that's music making
27:55 then we can also make
27:56 movies so if you go to Luma laabs L
28:06 labs. you now have something that makes
28:14 movies oh that actually did a
28:20 cut
28:23 huh so Sydney this is the musical that
28:26 I'm working on
28:29 so this is Time Square I didn't tell it
28:31 to make that cut I just said it's
28:35 pointing around Time
28:37 Square that's really interesting cuz
28:39 Sora does
28:43 that huh
28:49 fascinating and then let's see what else
28:51 did I have here seagull flying past a
28:54 crab boat on the ocean on a windy day
28:58 there's another
28:59 edit
29:01 seagull flying forwards and backwards at
29:03 the same time and a
29:05 boat drifting backwards through the
29:08 [Laughter]
29:12 ocean well
29:15 listen embrace the Jank people let's see
29:19 this one's pretty good
29:21 actually that one's pretty good it gets
29:24 a little janky on the face gets a little
29:26 weird on the face right there the eyes
29:27 get too close to
29:29 together but still that's pretty
29:34 remarkable I mean one of the things
29:36 remarkable about this is this isn't a 3D
29:39 model this was just a 2D image and it's
29:42 sort of figuring out you
29:46 know what that object is as it rotates
29:50 pretty pretty
29:53 crazy sort of a
29:56 train sort of a train looking sort sort
29:58 of
30:00 normal this looks
30:06 good I dig that car
30:13 design watch
30:16 this
30:19 Z his head his head like disappears into
30:22 his body for a second and then it pops
30:25 back out and he's a different guy
30:41 zik thoughts on the correlation between
30:43 jumping to the metaverse versus AI fast
30:47 adoption huh that's a really interesting
30:51 questions what about Turtle tenders
30:53 instead of chicken tenders or llama L
30:56 Linguini yeah that's good
31:00 um thoughts on the correlation between
31:02 thank you Lord digital Gods thank you
31:04 for the roses and if anyone else gave me
31:07 gifts and I didn't see them and didn't
31:08 thank you that's just cuz I didn't see
31:10 them so if you do give gifts I'll try to
31:12 acknowledge them if I miss you just yell
31:14 at
31:15 me
31:16 [Music]
31:19 um correlation between jumping to the
31:22 metaverse versus AI fast adoption
31:35 yes I do have
31:45 thoughts the metaverse probably still
31:48 has 5 years of
31:50 Hardware Hardware advances to catch
31:54 up um I don't think it'll be called the
31:57 metaverse I think it'll I think Apple
32:00 has successfully redefined it as spatial
32:03 Computing um and I think it's going to
32:05 be really broadly
32:07 defined so I don't think it's just going
32:10 to be goggles I think some of the things
32:12 that are our devices will just look like
32:14 these some will be immersive goggles
32:17 some will just be our phones
32:30 hm I mean I definitely think AI comes
32:34 first but then I think here's here's
32:38 what AI does for the
32:41 metaverse the problem the problem with
32:43 the metaverse life's tough it's tough
32:46 when you're stupid the problem with the
32:48 metaverse as it was being built by Mark
32:53 Zuckerberg and and the Gang at
32:56 meta and every other attempt that I've
33:01 seen is that because of the
33:04 technology you actually had to build a
33:07 world and you had to build a world with
33:09 a specific
33:11 aesthetic and meta chose to build a
33:14 world that looked like some weird ass
33:16 generic non-offensive cartoon world that
33:20 kind of looked like second life like
33:22 remember Second Life from the early
33:26 2000s so they spent whatever they spent
33:28 hundred billion dollar and they changed
33:30 the name of their company for this one
33:33 world that kind of looked
33:36 like awkward and cheesy and not
33:43 great and so so if you don't dig their
33:47 aesthetic it's not good and then
33:52 also no one's in there because to go in
33:55 there you have to get the big goggles
33:57 and you have to go in there and then you
33:58 have to go into a physical space and
34:01 then you go into a physical space and no
34:03 one's [ __ ] there I I had some friends
34:06 that did some AI art galleries in the
34:09 metaverse and like you'd go there and
34:12 the artist would be there and then you'd
34:14 be there and no one else would be there
34:17 and then you'd be like hey and be like
34:19 hey and it was just like an awkward
34:22 Junior High School
34:24 dance so so that just is a miserable
34:31 experience what I think AI is going to
34:35 do and actually you know looking at
34:37 these at these
34:41 videos this one
34:43 here
34:47 where you know this is a New York City
34:49 street scene and it just sort of
34:51 invented that I think we're pretty
34:53 quickly going to get to the point where
34:56 you're just going to be able to create
35:00 any any kind of environment you want and
35:03 any kind of experience you want within
35:06 it so it might be Time Square in the
35:09 '90s and
35:13 whatever and then you might want it to
35:16 be a video game and then it'll be a
35:17 video game where you drive as a New York
35:19 City Taxi Cab trying to not be in
35:22 accidents in New York City while you're
35:24 driving or you might want to see uh a
35:27 comedy like Seinfeld that makes you
35:29 laugh like Seinfeld set in New York City
35:31 or you might want to like so I think I
35:33 think we're going to dynamically be
35:36 generating things that we can enter or
35:38 just sit back and watch
35:41 passively and if you look at there's
35:43 there's a little hint of this there's
35:45 the the new thing came out um or was
35:48 announced a week ago called
35:52 showrunner and it's it's the group that
35:55 eight or nine months ago did the full
35:57 South South Park episode like an hourong
35:59 south South Park episode from a single
36:02 prompt how they're doing that is they're
36:05 actually
36:06 creating worlds like Sim Francisco is
36:11 what they called it and you're going to
36:13 be able to upload yourself into San
36:16 Francisco and your friends into San
36:18 Francisco and then they're going to run
36:20 around San Francisco interacting with
36:22 other people and
36:24 avatars and then it will generate a TV
36:27 show that you can sit back and watch
36:30 about your characters that are
36:32 gallivanting around this fake
36:35 place so I think the metaverse is going
36:37 to get is it's going to become much more
36:41 broad
36:43 definitionally than what we've
36:45 experienced to date I think it's going
36:46 to be everywhere and I might go from
36:49 passively watching a TV show to like oh
36:52 I want to go hang out in that bar and
36:54 play you know poker with those guys
36:57 sitting at that table
36:59 and then I can actually go into that bar
37:00 and play right so I think I think we'll
37:02 sort of come in and out of
37:07 it so I so but but I think I think AI
37:11 comes first for
37:13 sure I think AI comes first for sure cuz
37:16 the form factor is just not there for
37:17 the for all the augmented reality
37:22 [ __ ] I mean you know there were
37:24 douchebags walking around with vision
37:26 Pros around San Francisco I I'm going to
37:28 wear these all day you know walking
37:31 around with ski goggles
37:33 on $4,000 ski goggles I wonder how many
37:37 of them got mugged before they were like
37:38 that's probably not such a good
37:40 [Laughter]
37:44 idea [ __ ]
37:46 crazy all right GTA but AI vers yeah
37:51 exactly yeah something like that
37:52 something like that's coming
37:57 [Music]
37:58 I want Ready Player one I we're not that
38:00 far from it if pay attention for for
38:03 Ready Player One pay attention to what
38:08 um um oh what the [ __ ] it called
38:11 Unreal Engine five pay attention to what
38:13 Unreal Engine
38:15 does at some point here
38:17 shortly so they've already got these
38:20 things called are they I think they're
38:21 called procedural builds where you can
38:24 just drop like a whole world like like a
38:28 whole island
38:29 into you know an ocean and then you can
38:32 drop another Island on top of it and
38:34 it'll dynamically sort of connect them
38:36 and make everything seamlessly
38:39 connect so they're going to be able to
38:41 have whole worlds like components of
38:43 Worlds that you can just drop together
38:45 and it'll automatically figure out all
38:47 the
38:48 geometry and then they're going to be
38:50 able to stick an AI layer on top of
38:54 that and you'll be able to just go make
38:56 me a world that does blend and it'll
38:58 make you a full-on massive 3d world that
39:02 is exactly what you want it to be so so
39:05 stage one will be you can enter the
39:07 world and then stage two will be make
39:10 the world and then give it these game
39:14 mechanics and then it will be that world
39:16 with the game mechanics and then you'll
39:18 be like and tell my buddies I'm in here
39:20 now and then your buddies come and you
39:22 go do Little
39:24 Adventures dynamically in real time so
39:27 pay attention to them cuz I think I
39:29 think that could accelerate some of the
39:31 Hol deex
39:35 [Music]
39:42 stuff it's usually just a bunch of kids
39:44 being rude what in the metaverse yeah
39:51 exactly bored dog thanks for your help I
39:54 don't know if I helped but thank you I'm
39:56 going to see Jerry Seinfeld to night
39:58 actually awesome that's pretty cool
40:02 beautiful I guess Jerry's done
40:05 under we playing
40:07 [Laughter]
40:09 Ry you're talking about D and D pretty
40:15 much aren't there a few movies that
40:18 explore several different stories that
40:20 expand over the same
40:26 time yeah
40:37 um what were movies that had multiple
40:40 story
40:42 lines that expand over the same timeline
41:00 these are just movie
41:09 movies were there any that were were
41:12 there any that were more Choose Your Own
41:14 Adventure rather than a single
41:24 narrative Black Mirror bander switch or
41:27 bander
41:29 snatch produced by Netflix this
41:31 interactive film allows viewers to make
41:33 choices that influence the story's
41:35 outcome huh I didn't know about that
41:38 late
41:40 shift directed by Tobias
41:43 Weber offers multiple story paths Puss
41:47 and Boots Minecraft what's what's
41:49 amazing about
41:51 AI is
41:54 that right now if you want to do a you
41:57 know sort of a Choose Your Own Adventure
41:59 kind of narrative you actually have to
42:02 write all the [ __ ]
42:04 narratives and it's this manual process
42:07 where you're writing multiple story
42:09 lines and you know someone can choose to
42:11 go down them AI is going to be able to
42:14 just dynamically at any point in the
42:16 road you could basically just say to AI
42:19 you know introduce three characters and
42:22 and then come to a point in the story
42:25 where the the user makes a choice
42:29 and then the user could say this isn't
42:31 funny enough or you know I want to meet
42:36 that minor character and have them be a
42:38 major character and then and then it
42:41 just sort of meanders wherever the user
42:43 wants it to go I I don't know that a lot
42:46 of people are going to want to do Choose
42:48 Your Own Adventure kind of
42:50 stuff um but you're going to be able
42:54 to here's what 40 says Kyle Shannon is a
42:58 pioneering figure in digital marketing
43:01 and
43:03 Technology I I think what it missed was
43:06 a towering pioneering figure in digital
43:09 marketing yes towering in fact
43:13 um I have done digital
43:17 marketing I don't pioneering figure I
43:20 don't think I'm a pioneering figure but
43:22 I've done stuff
43:28 all right have you read snow crash ages
43:30 ago I don't remember a ton about it
43:32 thanks
43:34 add Grand Canyon that's
43:38 right we're at 10,000 likes you guys are
43:45 [Music]
43:49 swell bander snatches years old yeah
43:52 2018 it said you versus wild an
43:55 interactive series starring be grills
43:58 where viewers make decisions to help
44:00 bears survive in Wild
44:03 environments remember all the CD ROMs in
44:06 the in the early '90s they were starting
44:09 to get pretty sophisticated and then the
44:11 internet came along and just ate that
44:13 whole world that whole world just
44:15 disappeared
44:16 overnight known for his contributions to
44:20 the development of web based marketing
44:22 strategies
44:23 hello hello hi Kyle Shannon ering figure
44:28 in digital marketing how are
44:30 [Laughter]
44:35 you you can make
44:37 money Shannon has played a significant
44:39 role in integrating technology with
44:41 marketing and
44:43 advertising all right cool fantastic
44:46 well thanks
44:51 GPT
44:53 H that's
44:55 funny funny funny
45:02 funny Max Head dream
45:04 2024 vanity AI yeah exactly I got pretty
45:08 wrapped up in second life for a few
45:09 years yeah a lot of people did I mean
45:12 second life was pretty cool but again
45:14 it's
45:16 like you're at the mercy of their game
45:19 mechanics they're looking feel you could
45:23 build your own [ __ ]
45:25 but people had to kind of learn that
45:28 interface learn the currency learn
45:32 the the social
45:37 dynamics the other thing the other thing
45:39 that AI is going to contribute to these
45:41 kind of virtual worlds is the NPCs
45:45 you're not going to be able to tell an
45:48 NPC from a real
45:52 person it's kind of
45:54 cool it'll be like Westworld
45:57 right some people can tell some people
46:03 can't it's going to be a lot like
46:06 that where the where the the characters
46:09 that are not people have their own lives
46:11 and have their own memories and interact
46:14 and have jobs and go do
46:17 [ __ ] and we'll plan
46:21 parties for someone's birthday I'm in
46:24 Miami everyone here is an NPC that's
46:27 hilarious
46:29 ious it'll be like
46:35 now oh man played second life just
46:39 didn't use those apple Vision Pros yeah
46:41 there you
46:46 go it knows a lot about agency.com and
46:49 my role in the theater oh that's
46:52 interesting wonder what it knows about
46:53 my role in the theater I wonder where it
46:55 got it maybe from articles or something
46:57 like
46:57 that cuz I started and ran a theater
47:00 company but it wasn't it wasn't super
47:02 well
47:03 known I wonder if it knows anything
47:05 about Urban
47:07 desires my
47:19 magazine oh you know what going to be
47:21 amazing wait wait I just something just
47:24 hit me all this stuff is going to start
47:26 to do this this and and it's going to
47:29 get really
47:31 cool so
47:34 tonight I played some stupid plucking on
47:38 the guitar with the dog singing and it
47:39 turned into
47:41 songs and then the other one I just came
47:43 up with a stupid idea and it turned into
47:45 a
47:50 song There are going to be
47:54 characters that exist in the world
47:58 some of them will live in metaverse like
48:00 places and some of them will be on Tik
48:04 Tock think of
48:07 like a really slick version of like Joe
48:14 Jonas and two or three times a
48:19 day he's going to write a song and sing
48:23 it on his Tik Tok Channel
48:30 and people will start following him or
48:36 her and some of those songs will become
48:40 hits and that person will have never
48:48 existed that's [ __ ]
48:50 insane all right here's here's a here's
48:53 a business idea it's not a business idea
48:56 but here's a
48:58 if you're a musician now and you're like
48:59 ah this [ __ ] AI is going to kill all
49:01 the musicians it's going to kill all
49:07 the I understand your
49:12 pain [ __ ]
49:19 AI there's going to be a
49:23 premium on live performance
49:28 so what if you start either a band or or
49:35 a like an event
49:42 company that gets real bands that books
49:46 real bands to play AI generated songs
49:49 that have become
49:52 hits so if you have like the digital
49:54 version of you know whatever the next
49:55 Joe Jonas is or whatever do doesn't
49:57 matter Billy ish right the the it's it's
50:00 just and it's not a copy of Billy ish
50:02 like it's a new digital performer that
50:04 writes songs and those songs are going
50:07 to
50:08 be some of them are going to become hits
50:12 and some of those
50:14 NPCs are going to become
50:17 famous if you had a business that very
50:20 quickly found a musical artist that
50:22 could go perform that [ __ ]
50:25 live book a
50:27 like let those songs become hits and
50:30 then just book a
50:34 tour with someone that looks close
50:36 enough like that person I I
50:43 bet give it two
50:53 years and you start building talent to
50:56 be able to do that you you figure out
50:57 how to book
50:59 it you probably have the NPC up in the
51:03 big jumbotrons and you have a live band
51:06 with someone that looks close enough and
51:08 sings close
51:09 enough yep there's a business there's a
51:12 business real life bands covering AI
51:16 bands this is the way to meta for this
51:22 Multiverse it's going to get weird it's
51:25 going to get weird
51:27 I've I've I've said this all year this
51:29 year 2023 the last normal year it's
51:32 going to get
51:35 weird having real bands covering AI
51:45 bands cuz like a good song is a good
51:47 song and if a good song
51:52 hits and then you could
51:54 quickly make the connection between
51:58 all of the songs from that artist are
52:00 now going to go on tour with this band
52:02 that just does that right
52:05 like you
52:09 [Music]
52:21 know C the hill was a quiet
52:25 town where the old tree
52:29 sway whistle of the train echo through
52:33 the
52:36 day people you with a
52:39 knot nothing ever seems to
52:44 change stor still unfold each unique and
52:50 strange she grew up tough
52:55 tough so that's weird Mary that's an AI
52:58 song so imagine if you could had someone
53:01 that could sing and could
53:02 play singing that on tour with all the
53:05 other songs that if that were some
53:07 digital
53:10 artist pay attention to Kelly Bosch so
53:12 Kelly Bosch is the one who did the the
53:15 video for weird Mary for that song We
53:17 did an official video for it she does
53:19 these really cool Tik Tock videos um
53:22 where she for the most part is using
53:24 other people's songs and she's doing her
53:26 cool visuals
53:27 well I just saw a LinkedIn post from her
53:30 tonight where she's doing suo songs so
53:34 she's now creating songs and videos for
53:36 those songs pay attention to her pay
53:39 attention to to artists like
53:46 her where they could end up cre creating
53:49 some
53:50 hits and then people want to hear those
53:53 songs what do you do she's not going to
53:55 go on tour
53:57 I don't
54:01 [Laughter]
54:03 think it's going to get [ __ ] weird
54:06 people there was a Star Trek episode
54:08 where the crew was kidnapped and forced
54:10 to do
54:14 plays weird Mary from Cedar Hill on solo
54:18 guitar with a real dog yeah there you go
54:24 exactly so maybe AI Will Keep Us around
54:27 for entertainment seriously I saw I saw
54:29 someone post about that that you know
54:32 the the the uh The Matrix movies were
54:34 like AI is going to keep humans around
54:36 because they need the energy or like
54:39 whatever it was about the
54:40 energy um and I saw a thing on Twitter
54:44 like a week ago where it said AI is
54:46 going to keep humans around for the
54:49 data that they generate right so so
54:52 imagine that so AI generates these
54:54 songs they become hits
54:57 and then real humans real bands Go cover
55:01 those hit songs right and like all of
55:04 those performances are recorded and that
55:06 now becomes new
55:08 data so yeah so what if AI needs the
55:11 humans to keep contributing data because
55:14 it's run if it runs out of data it's
55:17 going to fall apart right so it
55:18 constantly needs new data and then it
55:21 optimizes for [ __ ] we like doing so we
55:24 keep creating more data
55:28 there's some oh K Kelly BOS k l l y b o
55:35 e
55:38 c and yeah she's on she's on
55:43 uh she's on Tik Tock and then I think I
55:46 I just saw her post on
55:50 LinkedIn but her stuff her stuff's so
55:53 cool here I'll show you some of her
55:54 stuff let me see if I can find it first
56:03 she grew up tough in a town she called
56:18 home this is Kelly BOS
56:20 [Music]
56:29 now this is with existing
56:32 [Music]
56:49 [Applause]
56:50 music gotch you run in
56:54 circles touching me
57:00 [Music]
57:01 get so look at look at
57:04 [Music]
57:16 this so
57:19 imagine she starts doing original songs
57:22 original AI songs with these original AI
57:24 images she's got such a clear creative
57:29 aesthetic imagine putting together a
57:32 stage show that's this
57:34 aesthetic that's all of that music that
57:38 would be [ __ ]
57:40 incredible
57:50 right or just do big video Pro
57:53 projections with all this [ __ ] in it
57:57 and then this is all live you know what
57:58 I saw once I saw um I don't know if you
58:02 remember the the movie co a scoty um by
58:06 uh Jeffrey Regio and then it was music
58:10 by Philip
58:12 Glass and they performed it in New York
58:16 City the Philip Glass
58:18 Orchestra performed the score of coana
58:22 scoty with the film playing in the
58:23 background so imagine being at a live
58:26 show
58:27 you know playing all the cool songs
58:29 whatever the songs are with all this
58:31 imagery being projected and I mean come
58:33 on
58:36 [Music]
58:45 [Music]
58:58 dolly
59:01 [Music]
59:07 [Music]
59:33 um here let me show you the video she
59:36 did for weird
59:55 Mary so this this is an AI song and
59:58 Kelly did the video for
1:00:01 [Music]
1:00:17 [Music]
1:00:24 this Cedar Hill is a quiet
1:00:28 town where the oak trees
1:00:33 sway the whistle of the train Echoes
1:00:37 through the
1:00:41 day people greet you with a
1:00:45 nod nothing ever seems to
1:00:49 change but the story still
1:00:52 unfold each unique and strange
1:00:58 she grew up
1:01:00 tough in the town we called
1:01:05 Home Folks never took to her no matter
1:01:12 [Music]
1:01:13 herone longing to be on the
1:01:17 inside longing for love just a
1:01:21 little she found herself as a guest in
1:01:24 the nervous Hospital
1:01:27 Tom Petty Vibes right yeah
1:01:31 exactly
1:01:33 H she HS when she's
1:01:38 blue of sadness and shame under the
1:01:46 S but when the sun sets
1:01:50 [Music]
1:01:51 behind Hill seems like for Springsteen
1:01:56 is commercial worthy
1:02:03 right the gas
1:02:07 station winked at the
1:02:10 clerk she stole her
1:02:13 soda and she called him
1:02:18 aerk in her own little
1:02:22 world she WS in the breeze
1:02:28 among the town's folk who whisper with
1:02:34 ease in a town that
1:02:39 never understood her
1:02:42 soul she found beauty in the wild let
1:02:47 her story
1:02:50 unfold her laughter Echoes In The
1:02:55 Wind she he was never the
1:02:59 same weird Mary from Cedar
1:03:03 Hill you'll remember her names
1:03:07 [Music]
1:03:26 [Music]
1:03:38 huh Firefly basics of Melody but needs
1:03:42 more range yeah I agree you're watching
1:03:43 this and you are addicted to um I agree
1:03:47 like it but but so
1:03:52 imagine imagine if you had a mechanism
1:03:55 where you basically let all these
1:03:56 artists go out and start making your AI
1:03:59 music go start making your AI music and
1:04:02 then the good ones that start getting
1:04:05 known and famous and have hits you start
1:04:07 signing
1:04:08 them and then what you say is you keep
1:04:11 making all your [ __ ] you keep doing
1:04:12 whatever you're doing we will then take
1:04:15 those songs and turn that into a show so
1:04:17 couple of
1:04:18 things the music as it stands right now
1:04:21 like sunno and yudo they're at like when
1:04:24 they do really good stuff like Mary the
1:04:27 the weird Mary thing that took us
1:04:29 probably three three and a half hours
1:04:31 that
1:04:32 night um and then like I remixed it and
1:04:36 and I've since like the one that I play
1:04:38 on my guitar I've changed the lyrics and
1:04:41 the chords so imagine just having an
1:04:44 organization that you take all those
1:04:46 songs and then you make them good in
1:04:49 performance worthy and ready and you get
1:04:51 artists to do
1:04:54 that I think that could be and you put
1:04:56 together just these [ __ ] Kick-Ass
1:04:58 stage
1:04:59 shows so maybe that's the metaverse like
1:05:02 maybe the metaverse is is physical
1:05:05 interactions with
1:05:07 humans that that brings AI [ __ ] to life
1:05:12 in ways that moves
1:05:16 people why
1:05:20 not I made a GPT that will basically
1:05:23 level level up our global
1:05:27 blank Dr J agency.com Kyle yeah I was
1:05:30 the founder of agency.com I was the
1:05:33 co-founder and chief creative officer
1:05:34 and then the chief people officer of
1:05:37 agency.com
1:05:39 that is the
1:05:49 one what program created the images
1:05:53 so um Kelly BOS mostly uses mid Journey
1:05:57 um for the weird marry one she used
1:06:00 stable diffusion because she liked the
1:06:02 aesthetic it gave a weird janky
1:06:04 aesthetic that she
1:06:05 liked um and I I love like it's got
1:06:09 weirdness where the you'll see her head
1:06:11 sort of morph and at one point she opens
1:06:14 her mouth and her jaw goes
1:06:18 like and the song is kind of janky and
1:06:21 the video is kind of janky and I think
1:06:23 like I I what I love about weird Mary is
1:06:27 I think it is a perfect artifact for
1:06:30 early to mid
1:06:33 2024 because this time next year the
1:06:36 video shit's going to be [ __ ]
1:06:39 Flawless and the music stuff's going to
1:06:41 be [ __ ] Flawless it's like just is if
1:06:45 you just follow the trajectory of how
1:06:46 good these tools are getting and how
1:06:48 fast they're getting
1:06:49 there weird Mary is going to be a really
1:06:52 cool even even some of Kelly's early a
1:06:56 of her early videos she's doing right
1:06:57 now you you can look at her early ones
1:07:00 and you can you can see that the tools
1:07:02 have
1:07:04 progressed
1:07:06 so yeah and I dig that I mean you know
1:07:11 that's that's where the term embrace the
1:07:12 Jank came from is
1:07:15 just the the imperfection of of the
1:07:18 tools right now is actually kind of
1:07:21 beautiful and it will be more beautiful
1:07:23 in hindsight you'll be able to look at
1:07:25 it and go oh that was 2023 that was 2024
1:07:28 oh this this was after sunno actually
1:07:30 came
1:07:31 out right one of my favorite nights was
1:07:35 the weird Mary night and that's saying a
1:07:37 lot yeah that was that was pretty
1:07:39 intense and you know what was funny
1:07:41 about that Silver Fox is
1:07:49 um like it was a deep rabbit hole right
1:07:53 like it was writing lyrics and tweaking
1:07:56 lyrics and trying to extend the song and
1:07:59 and like there there was a lot of time
1:08:01 spent there where it wasn't very
1:08:03 entertaining but it was like it was
1:08:05 interesting it was it was interesting
1:08:08 for
1:08:10 me trying to get it to do something
1:08:14 intentional on a channel like this
1:08:17 because one thing that AI is really
1:08:19 shitty at is doing anything predictable
1:08:22 and decent right like you can get [ __ ]
1:08:26 that's shitty really easy and you can
1:08:29 get stuff that's good occasionally but
1:08:31 if you try to like run something for
1:08:33 more than an hour and maintain a through
1:08:36 line of it it's really hard with this
1:08:37 stuff so yeah that was an interesting
1:08:40 night that was really
1:08:51 [Music]
1:08:54 cool wow
1:08:57 wow and all of Kelly's videos are
1:09:00 composed via AI tools yeah almost all of
1:09:02 her videos so she's got she's a
1:09:06 Runway Artist Run so Runway ml that does
1:09:09 the the
1:09:11 animations does all the video she can do
1:09:13 unlimited Runway ml um videos CU she's
1:09:16 part of the artist program and then she
1:09:19 does all of her images in mid Journey
1:09:21 with with some some exceptions
1:09:26 and then I noticed that she put out a
1:09:28 song today well I don't know if she put
1:09:30 it out today I saw it today on LinkedIn
1:09:33 she used soono to make the
1:09:37 song and I don't know what she did the
1:09:39 images in but then she made the movie in
1:09:41 something else that wasn't
1:09:43 Runway so she's she's she's pushing her
1:09:46 boundaries trying to figure out where
1:09:48 she wants to play and what she wants to
1:09:51 do but just pay attention
1:09:56 start paying attention to artists that
1:09:59 are fully embracing
1:10:03 Ai and just consistently putting out
1:10:06 work because they're going to start
1:10:08 getting really interesting really
1:10:15 fast you know
1:10:18 imagine imagine a time in the future if
1:10:21 you
1:10:21 will where Kelly could not only make
1:10:24 those characters that show up in her
1:10:26 weird videos with the weird sunglasses
1:10:28 and the weird fashion
1:10:30 [ __ ] but she could take that aesthetic
1:10:32 and spread it through a world you could
1:10:34 enter that
1:10:36 world and so all the cars have that
1:10:38 aesthetic and all the buildings and all
1:10:41 the
1:10:43 people and all the
1:10:47 music and so there's a version of that
1:10:50 word where you sit back passively and
1:10:52 watch a music video and then there's
1:10:54 another version of that world where
1:10:56 you're like [ __ ] it I want to go in and
1:10:58 you enter that
1:11:01 world why
1:11:04 not just the tools aren't there right
1:11:06 now but they're going to
1:11:11 be so you know right now how you can you
1:11:14 can take I can I can play some songs on
1:11:17 my guitar and then it turns it into a
1:11:18 sunno song imagine her being able to
1:11:21 upload one of her early Tik Tok videos
1:11:24 with all those cool characters
1:11:26 and that's the starting point to
1:11:28 generate a whole
1:11:31 metaverse it basically takes the
1:11:33 aesthetic of that it interprets the song
1:11:36 takes all the look and feel of that and
1:11:38 it uses that as the inspiration for a
1:11:40 whole
1:11:43 metaverse we're not that far from
1:11:46 that holy
1:11:51 [ __ ] that's pretty slick
1:11:57 I saw an immersion movie theater that
1:11:59 was like the sphere it was wild yeah the
1:12:02 the sphere is ahead of its time the
1:12:04 sphere's going to I bet they're going to
1:12:05 do some really interesting things in the
1:12:07 sphere the AI combo platter works best
1:12:10 for me exactly I'll take the poo poo
1:12:13 platter uh generative a on a poo poo
1:12:16 platter please yes yeah I'll take a
1:12:18 little bit of image generation and then
1:12:20 just a Sprinkle of Animation to those
1:12:22 images please I don't want to make a
1:12:24 whole movie but I would like my images
1:12:26 moving around if you know what I mean
1:12:28 and then we'll do some songs and then oh
1:12:31 I'd like to use one of the large
1:12:33 language models to do some poetry but I
1:12:35 I wanted to say the Poetry to me and I
1:12:38 wanted to kind of act it like like I was
1:12:40 at Shakespeare in the park and yeah if
1:12:43 it could do that that'd be great oh and
1:12:45 then while that's all happening and I'm
1:12:47 having fun could you just have that do
1:12:48 my taxes to could it could do that okay
1:12:51 good yeah and no get me the maximum
1:12:53 refund that's good yep have game the
1:12:56 system have it have it just game the
1:12:58 system just do it right up that'd be
1:13:00 good yep yep same tax breaks as the
1:13:03 billionaires bring it on for me that's
1:13:06 what I always say yeah that'd be
1:13:12 [Laughter]
1:13:15 good I lost my job today I'm sorry about
1:13:18 that and I'm about to make money with
1:13:19 chat GPT you can make money with chat GP
1:13:23 well what I you know listen what I would
1:13:25 say is
1:13:26 dive in start playing I don't know where
1:13:27 you are uh Embrace son was it Embrace
1:13:30 son um sorry you lost your job that
1:13:33 sucks um but yeah dive into this AI
1:13:36 stuff um get good at it start talking
1:13:39 about it put it out there have it get us
1:13:42 out of this mess we're in yeah that
1:13:44 would be nice wouldn't it it really
1:13:50 would cuno unce Sun I'm sorry how can we
1:13:53 help so unce Sun if you haven't joined
1:13:56 the AI Salon you should join the AI
1:13:58 Salon cuz it really is uh an amazing
1:14:01 support
1:14:02 network um and especially if you're
1:14:05 trying to get up to speed quickly or if
1:14:07 you want strategies on how to use AI to
1:14:10 talk about what you do you know spin up
1:14:13 your resume write personalized Outreach
1:14:16 emails to prospective employers like
1:14:19 there there's a lot of people here that
1:14:20 know it so go there join introduce
1:14:22 yourself if you haven't and start
1:14:24 connecting with people um it's a it's a
1:14:27 it's an amazing resource it's amazing
1:14:30 Community thank you for asking that cuno
1:14:32 that's a great
1:14:34 idea can I make a music video with the
1:14:38 lyrics of the
1:14:40 song and what capos I can use capos I
1:14:44 don't know what capos you
1:14:45 mean like what the chords
1:14:48 are you can do that I mean you can just
1:14:51 do that in
1:14:53 uh in like iMovie your Final Cut
1:14:58 Pro could be the best thing that's ever
1:15:01 happened the job thing the AI Salon play
1:15:04 first that's it the salon. a we we we
1:15:09 made our motto or our motto our
1:15:15 tagline of this organization
1:15:26 play
1:15:31 first play
1:15:36 first mindfully create generously lead
1:15:40 play first mindfully create generously
1:15:45 lead you do
1:15:47 this with this AI stuff and then keep
1:15:51 doing it just keep cycling through this
1:15:53 this journey
1:15:55 play first is how do you learn this [ __ ]
1:15:57 you just play with it you just play with
1:15:59 it there's too much of it to actually
1:16:01 learn it play with it figure out where
1:16:04 you've got aptitude for it figure out
1:16:06 where you enjoy using it figure out
1:16:08 where you're like holy [ __ ] I'm good at
1:16:09 this within this AI world and then in
1:16:13 doing that you'll discover oh I know
1:16:17 what I need to build and there's an art
1:16:19 project or here's a work project or
1:16:21 here's a marketing plan or whatever it
1:16:22 is mindfully create and then as you're
1:16:25 doing all of that Learn out Loud share
1:16:28 what you're learning connect with other
1:16:30 people ask people
1:16:33 questions do it in a community like the
1:16:36 AI
1:16:40 Salon do it out
1:16:44 loud what I'm hearing pretty
1:16:46 consistently now from people in this
1:16:48 community who who have have hung up
1:16:51 their own shingles as they say they've
1:16:53 started their own businesses
1:16:58 that people are starting to come to them
1:17:00 going okay I'm ready to hire you now
1:17:02 like people are starting to get hungry
1:17:03 for this AI
1:17:05 stuff but I don't think the floodgates
1:17:08 have not opened yet the flood like this
1:17:10 is we we are still very early so you've
1:17:12 still got a lot of time to get AI
1:17:15 literate relatively speaking I think
1:17:18 you've probably
1:17:21 got 12 to 18
1:17:24 months where if you get if you get AI
1:17:27 literate in this kind of way in this
1:17:29 sort of focused kind of
1:17:32 way where you're putting yourself out
1:17:34 there in the next 12 to 18 months it's
1:17:36 going to be massive opportunity for
1:17:38 anyone doing that consistently well
1:17:41 massive
1:17:43 opportunity and then after that it's
1:17:45 going to it's going to be a blood bath
1:17:47 because the competition's going to get
1:17:49 really
1:17:51 heated and so if you can be an
1:17:54 established player
1:17:56 in whatever the [ __ ] AI
1:17:59 becomes so Microsoft is investing in a
1:18:02 nuclear plant to help power the intense
1:18:04 amount of electricity for AI that
1:18:07 doesn't surprise me at
1:18:10 all and in fact what they'll do is
1:18:12 they'll put the data center right next
1:18:15 to the nuclear power plant so they don't
1:18:17 have to you know suffer any of the loss
1:18:20 transmission loss right doing it over
1:18:24 miles and miles and
1:18:26 miles that's one of the fascinating
1:18:28 things that I never knew that where they
1:18:30 make the electricity is never near where
1:18:32 people live because it's it's messy and
1:18:35 obnoxious to make the electricity so
1:18:38 they make it way off in the distance and
1:18:41 then they ship the electricity to your
1:18:43 to a town near
1:18:46 you well they don't need to do that if
1:18:49 if if what they're using all that
1:18:50 electricity for is training these models
1:18:53 they just build data centers right at
1:18:55 the dam right at the nuclear power plant
1:18:58 generate all these cool models and then
1:19:00 you know run the run the instances out
1:19:02 on the edge on local devices or on
1:19:05 smaller server centers near population
1:19:09 centers it's really cool it's like this
1:19:11 macro the the macro infrastructure stuff
1:19:15 with this stuff is absolutely crazy
1:19:18 against the norm I'm a retired power
1:19:19 plant operator yes you are correct wow
1:19:22 cool okay good I'm glad I wasn't talking
1:19:23 completely out of my ass damn nuclear
1:19:26 Tri solar yeah yeah they could do solar
1:19:30 as
1:19:41 well Mr it if you need a job I can help
1:19:44 if you're in the it field there you go
1:19:46 connect with Mr
1:19:49 it of this organization or of
1:19:53 disorganization that's funny
1:20:02 unbrace son I love your attitude cuno
1:20:04 thanks for asking I feel it's for the
1:20:06 best that that he he lost the job he's
1:20:08 about to dive in yeah nothing nothing
1:20:11 like getting getting a a a factory reset
1:20:15 uh to to inspire you to to go all in on
1:20:17 this stuff huh
1:20:20 crazy our power comes from coal plants
1:20:23 way south of here train loads of coal
1:20:25 pass through yep I know crazy right yeah
1:20:29 we we have big train big train loads of
1:20:31 coal pass through Denver all the time
1:20:33 too they built a data center next to a
1:20:35 nuclear plant doesn't surprise me at all
1:20:38 I meant
1:20:43 damn between electric cars and AI we
1:20:45 need to go back to
1:20:48 Coal we will need to go back to Coal no
1:20:51 man we need AI to to help us crack cold
1:20:53 fusion we need
1:20:55 fusion and we need like we'll crack it
1:20:59 we we'll pull if there's if there's any
1:21:02 energy in the what do they call it not
1:21:05 the latent space I'm thinking large
1:21:07 language models in the in The Ether what
1:21:09 did Tesla call it the energy he was
1:21:12 pulling out of the
1:21:14 atmosphere there's that stuff Zero Point
1:21:16 Energy if Zero Point Energy is real AI
1:21:18 will find that if Fusion's possible AI
1:21:20 will find that if we've got real-time
1:21:23 superconductors AI will find that like
1:21:25 the energy thing I think is going to be
1:21:29 cracked AI is going to crack the energy
1:21:32 thing and actually it's in ai's interest
1:21:35 to crack the energy thing because it
1:21:37 needs it to
1:21:39 survive I mean holy [ __ ] are we in the
1:21:43 um in in sci-fi we're in we're in
1:21:47 science
1:21:48 fiction it's
1:21:50 crazy Fusion is demonstrated it's
1:21:53 difficult to capture capture and sustain
1:21:55 the reaction yep I know yeah we need
1:21:58 some like room temperature
1:21:59 superconductors and some other [ __ ]
1:22:01 going on that so I don't know if anyone
1:22:06 did you all pay attention to when Deep
1:22:09 Mind um launched
1:22:13 480,000 430,000
1:22:16 stable
1:22:18 crystals into the world about 6 months
1:22:22 ago
1:22:25 did anyone see
1:22:50 that of new materials discovered with
1:22:53 deep learning this is by Google Deep
1:22:56 Mind an AI tool called gnome finds 2.2
1:23:01 million new crystals including 380,000
1:23:05 stable materials that could power future
1:23:10 Technologies then you're like okay
1:23:12 that's cool that doesn't mean anything
1:23:13 right but then you keep looking at
1:23:17 this this
1:23:19 circle
1:23:22 represents all of the stable crystals
1:23:25 that have been discovered by human
1:23:27 experimentation 20,000 of
1:23:30 them there there's Material Science the
1:23:32 history of humanity
1:23:35 people unassisted through
1:23:37 experimentation have discovered that
1:23:39 many
1:23:40 materials this this this ring out here
1:23:44 is with computational
1:23:49 assistance
1:23:51 48,000 and then gnome
1:23:55 deep mind's gnome these are all new
1:23:57 stable crystals that deep mind just gave
1:24:02 to the world and they're starting to
1:24:04 test these in labs and some high
1:24:07 percentage of them are turning out to be
1:24:10 indeed stable crystals that have novel
1:24:13 properties and if you're like I don't
1:24:14 know what a [ __ ] stable crystal is I
1:24:16 don't either but I know that when you
1:24:18 have things like graphine or or
1:24:23 um you know the like the the uh the
1:24:27 different materials that are in
1:24:29 batteries that transmit materials or if
1:24:31 you're looking for room temperature
1:24:34 superconductors all of that shit's going
1:24:36 to be in here somewhere so scientists
1:24:39 now are just going to be in here playing
1:24:40 testing playing testing and they'll be
1:24:42 using AI to play and test as
1:24:47 well so somewhere sitting in here is
1:24:51 really efficient energy transmission
1:24:53 really efficient battery you know
1:24:55 imagine a battery that you could charge
1:24:57 in 10 minutes that'll last for a year
1:25:00 and a half that's somewhere in
1:25:03 [Laughter]
1:25:06 here you know deep mind just did the
1:25:09 same [ __ ] with the molecules of life
1:25:13 right um Alpha fold 3 just launched and
1:25:17 they can now
1:25:19 predict the
1:25:21 structure and behavior of every every
1:25:25 molecule of life including DNA and
1:25:29 RNA that just happened a month
1:25:36 ago at at a high level of
1:25:42 accuracy
1:25:44 humans humans Plus
1:25:47 Computers Deep Mind in one year in one
1:25:51 year they were like oh why don't we
1:25:53 point this at Material Science and see
1:25:54 what we can find one
1:25:57 year efficient energy transmission maybe
1:26:00 we could start with upgrading and
1:26:02 maintaining before people get killed as
1:26:04 a starting point well antie bubbles see
1:26:07 now now you're talk now that's crazy
1:26:09 talk doing basic stuff come
1:26:12 [Laughter]
1:26:16 on uh oh Kyle I just got access to
1:26:20 Google 2 million context window and I
1:26:22 don't know what that even means
1:26:25 uh that's awesome here's here's simply
1:26:27 what it means um
1:26:29 Danielle the the Google
1:26:33 um Google Gemini 1.5 with 2 million
1:26:37 token context window okay I'll tell you
1:26:39 what it means to have 1 million tokens
1:26:42 and then you can double
1:26:45 it so with 1 million
1:26:48 tokens um Gemini 1.5 Pro is truly
1:26:52 multimodal so you can upload a video
1:26:55 to Google deepmind and it it can not
1:26:58 only understand the words that are
1:27:00 spoken but it can understand what's in
1:27:02 it so you can upload an hourlong video
1:27:06 to Google
1:27:07 deepmind and and say you know what's
1:27:11 happening when the Red Wagon shows up
1:27:13 the first time and it'll say oh a little
1:27:15 girl is pulling it across the
1:27:18 playground so with a million tokens you
1:27:21 can do an hour of video and you can do
1:27:23 11 hours of audio and then you can do
1:27:26 like 750,000 words so I don't know some
1:27:29 stupid amount of books right like 30 or
1:27:33 40 big
1:27:35 books
1:27:37 um so 2 million just double that you
1:27:39 could do a 2hour video 22 hours of
1:27:47 audio and I mean it's like when you
1:27:49 upload 11 hours of audio if you if let's
1:27:52 say you have a 10hour recording of of a
1:27:54 whole day conference and you upload it
1:27:57 into this thing it's not transcribing it
1:27:59 into text it it understands all of what
1:28:03 was what is what is recorded in that
1:28:07 audio Robert Rossy thank you very very
1:28:09 much for the uh for the lightning bolts
1:28:11 that's amazing you're amazingly generous
1:28:13 thank
1:28:19 you so so go start playing you know play
1:28:23 first now now that you have 2 million
1:28:25 tokens go dump a 2-hour video in there
1:28:28 and start asking questions about it and
1:28:29 just see what it does it's pretty
1:28:31 amazing it it's pretty
1:28:38 stunning but Kyle I'm not nerdy enough
1:28:41 to do all that
1:28:45 ah if okay here here's a thing you could
1:28:48 do
1:28:50 Danielle let's
1:28:53 say let's say that there's a really long
1:28:57 podcast or a whole series of
1:29:01 podcasts
1:29:03 and or or audio books doesn't matter
1:29:06 some long thing that you don't have time
1:29:08 to listen to you can just upload all
1:29:11 that audio and just say tell me what
1:29:13 this is about summarize this for me out
1:29:17 of the summarize every half hour chunk
1:29:22 [Laughter]
1:29:31 that's like the first half of Ben
1:29:33 [Laughter]
1:29:35 her well so there so what what Google
1:29:38 has
1:29:40 said is they're they're giving access to
1:29:43 2 million tokens they're successfully
1:29:45 testing 10 million
1:29:48 tokens
1:29:49 so an hour times 10 hours of
1:29:52 video and then they're going to be be
1:29:54 able to do unlimited
1:29:57 tokens so
1:29:59 anyway so I can dump content into it and
1:30:02 ask it to search for keywords absolutely
1:30:04 can you absolutely can and then go where
1:30:08 the dopamine takes you yep
1:30:10 exactly
1:30:12 exactly all right people it's still warm
1:30:15 in here I'm getting a little getting a
1:30:18 little so let me get on out of here um
1:30:22 fun fun night tonight thank you all for
1:30:24 hanging out there's champy little
1:30:28 champy
1:30:30 um I will see you again on Monday thank
1:30:33 you to the Irregulars thank you to all
1:30:35 of you if you if you don't follow my
1:30:39 channel what are you doing with your
1:30:41 life if you want to support the channel
1:30:43 directly you can subscribe to the lives
1:30:45 that would help
1:30:49 um what else just keep showing up
1:30:52 here I say this every and every night
1:30:55 I'm like wait what am I what what else
1:30:56 am I talking about um Happy Father's Day
1:31:00 thank you very much I appreciate that
1:31:02 can you look at Irregulars yes I can let
1:31:05 me go to Irregulars hey you
1:31:11 [Music]
1:31:15 Irregulars Kyle Shannon's AI agency Toby
1:31:20 Firefly not saving your generated images
1:31:25 oh is this the thing
1:31:26 where yeah it you have no library of
1:31:29 your images no it's completely [ __ ]
1:31:32 ridiculous they're still not doing it
1:31:34 they still haven't fixed
1:31:36 that everyone needs to watch this it's
1:31:40 perfect you know what I'd like to
1:31:42 discover where my [ __ ] files are
1:31:45 that's what I'd like to discover Adobe
1:31:47 this is me Adobe Firefly ideas history
1:31:50 of previous generated images history is
1:31:53 not
1:31:55 saved you got to be [ __ ] kidding
1:31:58 me who not saving your generative AI
1:32:03 history what [ __ ]
1:32:05 tools I guarantee you that's some
1:32:07 [ __ ]
1:32:10 lawyer well if we if we allow them to
1:32:12 save their history and they've created
1:32:14 something inappropriate if there is some
1:32:17 sort of legal activity it is it's
1:32:19 possible that by them uh retaining a
1:32:23 history of what they done with the
1:32:25 applications which of course we know are
1:32:27 highrisk and uh the the the the uh
1:32:30 exposure to liability of course is is uh
1:32:33 quite distinct here so uh recommendation
1:32:36 uh of council uh for the for the
1:32:39 aforementioned future therein here to
1:32:42 for the uh extended uh Runway that's a
1:32:47 as a colloquial word Runway of course
1:32:49 that's not a legal term of course just
1:32:50 uh just uh you know backing off on my
1:32:52 personal liability there okay just to be
1:32:55 clear uh would be that uh we would of
1:32:58 course uh allow people to see their
1:33:00 images upon generation and then at any
1:33:03 point in the future if they wanted to
1:33:04 return to that uh they would they would
1:33:06 it it would be simply a situation where
1:33:10 they would uh uh create a prompt and and
1:33:13 regenerate that image from from the from
1:33:16 the an original for for the so it is
1:33:18 more of an ephemeral situation uh to
1:33:21 reduce the liability of the corporation
1:33:24 the shareholder value and the risk
1:33:26 mitigated
1:33:29 therein are you [ __ ] kidding
1:33:34 me who did that digital
1:33:41 Gods dude that's [ __ ]
1:33:45 awesome I'm like oh somebody made a cool
1:33:48 video talking about the same thing I've
1:33:50 been bitching about oh it was
1:33:53 me digal Gods that's [ __ ] genius love
1:33:57 it champy
1:34:00 Gucci oh my God that's so
1:34:05 hilarious thefor
1:34:08 mentioned we've got more of an ephemeral
1:34:10 situation here people could of course
1:34:12 recreate their
1:34:18 images oh good lord janky is as janky
1:34:23 does
1:34:25 uh way to go Lord digital Gods you have
1:34:28 absolutely
1:34:30 earned the Lord
1:34:32 moniker uh with that piece of work fan
1:34:36 [ __ ]
1:34:37 tastic is there a salon chapter in
1:34:39 Scottsdale um not right now we're we're
1:34:44 looking to upgrade
1:34:46 the
1:34:51 um the local the the local media don't
1:34:54 have a ton of structure in them right
1:34:55 now and we're looking to upgrade that so
1:34:58 that will be coming but but uh if you if
1:35:01 you are interested in hosting an AI
1:35:05 Salon after hours in Scottdale join the
1:35:08 AI salon and uh
1:35:11 DM I guess right now DM
1:35:14 Vicki um VI kkii and connect with her
1:35:19 and and she she's on the organizers
1:35:21 committee and she's the community
1:35:22 director and so we'll be we'll be up in
1:35:26 the game on on local events so that's
1:35:30 coming all right Kyle I've told you I
1:35:33 have lots of office space in Phoenix you
1:35:35 can start a chapter there just say so
1:35:36 okay cool thank you appreciate that yeah
1:35:39 we got to get our [ __ ] together with
1:35:41 kind of infrastructure to support local
1:35:44 people doing stuff and it's just we just
1:35:47 we're trying to get the the main Salon
1:35:49 kind of organized and you know it's
1:35:52 structured down and then we'll do the
1:35:54 we'll do the extended ones but we've got
1:35:56 listen we've got some offers to support
1:35:58 the salon so so we'll see should be good
1:36:02 all right people I'm going to get out of
1:36:03 here let me do a little uh a little
1:36:06 Romper Room magic mirror uh troen in
1:36:12 thank you for the gifts I appreciate
1:36:13 that Firefly good to see you Joe wey
1:36:16 good night good night Dr J Papa
1:36:19 JJ uh anti-aging Marina she recall te Dr
1:36:25 J cuno unraced
1:36:27 Sun I'm lead cyborg nice good luck with
1:36:31 all this stuff Jump Right In Silver Fox
1:36:33 great to see
1:36:36 you Danielle thank you Pro
1:36:41 Planet all right all right all right all
1:36:44 right Mary Mary good to see you anti
1:36:49 bubbles cuno Mr it
1:36:54 all right spin B3 okay cool groovy have
1:36:59 a fantastic weekend everybody I will see
1:37:01 you on Monday um if anything exciting
1:37:04 happens over the weekend with AI I'll do
1:37:06 something but it probably won't so
1:37:09 anyway peace out and I will see you
1:37:11 Monday