Jul 22, 2024 - Embrace the Jank: AI Music, Digital Twins, and the Future of Creativity**

Description

Chapters

Transcript

0:00 it I get it I understand you're ready to
0:04 sing for the people the people
0:06 await the melodious tones of one champ
0:09 Shannon are you ready
0:13 [Music]
0:14 [Applause]
0:17 [Music]
0:39 you're feeling nervous AR you
0:42 boy with quiet
0:45 voice and impecable
0:49 style don't let them steal your
0:54 joy ways to keep them from running wild
1:01 they can kick dir in your face dress you
1:04 down and tell you that your Blazers in
1:06 the middle when they hate the way you
1:12 shine I see you tugging on your
1:16 shirt try to hide inside of it hide how
1:20 much it
1:22 hurts let him laugh while they
1:28 can let them spin
1:31 a
1:33 spin let them scatter in the
1:38 wind I've been to the
1:43 movies I've seen How It
1:48 Ends and a
1:50 jokes on there
1:57 [Music]
2:22 champ sh is a dog he's a fine fine dog
2:31 [Music]
2:52 s in this lonely child now wonder when
2:56 things are going to
2:59 change dream my life
3:02 away dreams a bunch
3:08 glass get my nerve up but my past is
3:12 pulling me
3:14 [Music]
3:15 down wondering how
3:19 long black she stick
3:22 [Music]
3:26 around somebody
3:29 told you can never go home
3:32 again want to
3:34 leave say to me
3:39 away from the truth of who I am and what
3:42 I believe so I thanked him for his two
3:45 cents with a
3:47 handshake and some sympathy and a packed
3:52 up my blue
3:53 jeans head it for this big
3:58 price of my freedom
4:03 byebye Black Sheep to the black sheep of
4:07 the
4:07 [Music]
4:09 family
4:11 [Music]
4:13 by or it means so very much to
4:17 me
4:18 byebye
4:20 bless to my friends and my
4:25 family bye
4:30 set my
4:32 soul set it
4:37 free
4:39 free times they were a
4:42 changing a little bit of
4:45 [Music]
4:47 re take a couple
4:50 chances my progress I'm not
4:53 advances to
4:56 the of my freedom
4:59 [Music]
5:27 guitar nearly in tune voice nearly
5:30 intact dog can sing like the
5:36 Dickens good evening good people happy
5:38 Monday hope you had a good weekend how
5:41 many people are in here got a pile of
5:44 people in
5:46 here welcome to the janky times The
5:50 Times They Are a janky Bob Dylan sang
5:52 that back in the 60s not many people
5:54 know that that was the original lyrics
5:56 of that song Times They Are
6:00 [Music]
6:02 janky I was just 7 years old in the
6:05 front row of a show that
6:07 year down in the East Village New York
6:10 City and he said times they are a
6:14 janky and after his set I went up to him
6:17 I was like uh what's your name
6:21 Mister and he said Bob and I said hey
6:25 Bob I
6:26 said I don't know what the word janky
6:29 means but
6:30 don't you think could work better if you
6:32 said times they are a changing instead
6:35 of times they are a
6:37 janky and he goes you got a head on your
6:40 shoulders
6:41 kid then he slapped me and gave me a
6:44 surely Temple and sent me on my
6:48 way that was the time I Changed History
6:53 [Music]
7:04 yeah other than the time I
7:05 was having dinner with Elon
7:09 [Music]
7:15 Musk and he was
7:19 like you know
7:22 Kyle having billions of dollars is it's
7:26 not as easy as people think he said for
7:28 example like right now I got an extra 50
7:33 million sitting in my bank account and
7:34 what am I supposed to do with
7:36 it I was like that's a rough one Elon I
7:39 don't
7:40 know you thought it started a rocket
7:42 company he was like nah that doesn't
7:45 feel believable to
7:47 me I said well you know there's that guy
7:50 Sam guy used to run y combinator I think
7:53 he's looking for some money to start up
7:55 this AI business
8:00 and he goes oh huh I haven't heard of
8:03 him I said yeah here's his number why
8:05 don't you call him up and maybe give him
8:07 your $50
8:09 million rest is
8:12 history you know there's a
8:15 burden being instrumental in the future
8:18 of
8:19 humanity but I'm willing to take it
8:28 on that's that's why I'm here
8:30 [Music]
8:33 tonight anyone has an extra $50 million
8:36 doesn't know what to do with it just
8:37 reach out to me I got some ideas
8:45 [Applause]
8:49 [Music]
9:07 [Music]
9:21 you're a strong man history owes you
9:24 you're welcome Dr Jay I mean I don't
9:26 like to brag it's just facts you know
9:30 you know some people are like oh that's
9:31 just braggadocious and I'm like I don't
9:34 even know what that word
9:37 means I'm just spitting the spitting the
9:40 [Music]
9:49 truth champ is a purebred
9:55 Labrador now that's comedy
10:07 [Music]
10:10 [Applause]
10:13 [Music]
10:17 I didn't mean to cause you any a
10:22 sorrow I didn't mean cause you any pain
10:26 [Music]
10:31 I only want want time to see you
10:35 laughing I only want to see your laugh
10:39 in that purple rain Purple Rain Purple
10:46 Rain Purple Rain Purple
10:50 [Music]
10:52 Rain Purple Rain Purple Rain
11:00 I only want to see you laughing in that
11:04 purple rain you know that
11:06 song I was hanging out with Prince one
11:10 night Minneapolis was there on a
11:13 business
11:13 [Music]
11:15 trip this guy walks in and he's wearing
11:17 this orange
11:20 jumper and he was like singing this song
11:22 Orange Crush orange crush crush and I
11:25 was
11:26 like just doesn't work right it's not
11:28 the right color orange is too bright
11:30 it's too festive for the the tone of
11:33 that song He's like what do you mean man
11:36 he was hip had a really deep
11:39 voice I said I don't know it feels more
11:41 purple it's like like it's raining and
11:44 he's like I got to
11:45 go that was the last I ever
11:49 saw of the artist formerly known as but
11:53 it was shortly after that that that song
11:56 came out and I thought maybe I played a
11:58 little part in that
12:01 cuz Orange Crush just wouldn't have been
12:04 right look here see that that's an
12:07 orange crush can you'd think I'd be
12:10 making up this story if I didn't have
12:13 that
12:17 [Music]
12:20 [Applause]
12:21 no I actually did talk to Prince once
12:25 talked to him on the
12:26 phone talked to him I talked to Howard
12:29 Stern actually Howard Stern was in my
12:31 office when we were doing agency.com
12:34 back in the olden timey
12:36 days but I talk to prince prince had a
12:38 very deep voice like hey man it's like
12:44 who stands on a chair cuz it's getting
12:48 deep Kyle tell us his story I have been
12:51 weaving some Tales tonight haven't I
12:54 [Music]
13:17 all right good people let's get this
13:19 party
13:21 started Winston welcome back welcome
13:25 back it's really good to have you here
13:28 missed you too
13:37 [Music]
13:45 yep it is good to have you back
13:49 sir
13:51 fantastic
13:53 um that by the way
13:56 this I picked this up at a yard sale
13:59 it's from Super Bowl
14:02 12 and it is the Denver Broncos Super
14:07 Bowl trophy SL pen
14:10 set number 522 out of
14:14 3000 can't just get those at any
14:17 old [ __ ] you can't get that on
14:20 Amazon
14:24 nope I do have it on sale for face my
14:27 Facebook Marketplace though I had a guy
14:29 that was about to buy it and then he
14:31 called me back and goes my wife said I
14:32 can't have
14:34 it so I put it
14:38 here nice t-shirt Mr Shannon I know
14:41 Blandon look this is a good one look at
14:43 the back
14:45 H big old
14:47 blands thank you Milo's
14:52 wife oh man all right what do we got
14:56 going on
14:58 um couple of
15:02 things well over the weekend we have a
15:06 new Presidential nominee so there's that
15:08 oh by the way so I posted this on the AI
15:13 Salon but as of this morning I was
15:17 headed to Washington DC
15:20 tomorrow um to speak to a bunch of
15:22 congressional and Senate folks on
15:26 Wednesday the meeting was cancelled
15:31 you can make
15:33 money apparently old Benny Netanyahu is
15:37 going to be uh addressing a joint
15:38 session of Congress
15:41 and and apparently they're shutting down
15:43 the uh Congress and the Senate to the
15:46 public so our event required access to
15:51 those
15:53 buildings so possible I do a zoom call
15:56 or something like that but that that
15:58 event went went the way of the dodo bird
16:02 uh as of uh like 11:00 a.m. this morning
16:05 I'm like proofreading stuff I'm sending
16:07 fixes to documents we put together and
16:10 then they were just like and it's not
16:13 happening so who' to thought I know
16:16 right better you know now than after you
16:18 got to DC yeah exactly and also listen
16:21 if if it's a kind of thing where they
16:22 think Security's got to be tight enough
16:24 that they're going to shut down those
16:25 buildings it probably was not going to
16:27 be all that fun to be around DC anyway
16:30 so um so there you have it but that's
16:34 all
16:36 right I don't get to tell Senators that
16:39 I think they should embrace the Jank and
16:44 uh you can make have their first Kevin
16:48 McAllister moment you should see the
16:50 requirements for AI remote
16:53 jobs on usa.gov
16:56 AI remote jobs like like meaning you've
16:59 got to have 10 years experience that
17:02 kind of
17:03 [ __ ] that happen that happens every new
17:06 technology we want we want to hire you
17:09 to do this completely new Innovative
17:11 technology and you have to have 10 years
17:13 experience doing it but I thought it was
17:14 new and Innovative shut
17:17 up yeah exactly yeah well that's the
17:21 thing I mean like one of the things I
17:23 was going to talk about was was to
17:26 distinguish between the 30 to 50 years
17:29 of AI that has
17:31 preceded the chat GPT era and and what
17:35 we have now there are two very
17:38 distinct eras of AI and and and very
17:41 very distinct skill sets so if if you
17:44 want to be building Frontier models then
17:47 you know it go get yourself a
17:49 mathematics degree if you want to use
17:51 chat
17:52 gbt do what Jim Ross does made a song in
17:56 sunno about
17:57 items said in a consult call and gave
18:01 him gave him the song before the call
18:03 was done impactful there you
18:07 go what was it it was lyrics full of
18:09 things like padlocks and and angle
18:12 grinders and
18:14 auctions and amortization
18:17 schedules that's
18:19 fantastic oh
18:22 man um there was actually a very cool
18:27 um post on on the
18:33 Twitter Elon Musk likes to call it the
18:49 X like share follow from more um let's
18:52 see let's go to yudo yudo put out
19:00 this little
19:05 beast should be an exciting week for
19:08 yudo creators thanks to ideogram.svg
19:27 [Laughter]
19:31 which AI app do you recommend best for
19:34 generating
19:37 videos I don't know right now my my
19:40 problem with them is um Runway ml Gen 3
19:44 is ridiculously [ __ ]
19:46 expensive I get I get 650 credits for
19:51 $10 a
19:52 month and my credits just rolled over
19:56 today and so I had 630
19:59 credits and I made one video and now I'm
20:02 down to 530 credits so that basically
20:05 means I'm going to get six videos for
20:10 $10 which is [ __ ] stupid so I I'm
20:14 Runway is on my [ __ ] list right now they
20:17 they've got to get that price down
20:18 that's that's just [ __ ]
20:20 Bonkers cuz it's not that much money for
20:22 unlimited it's like 70 bucks for
20:24 unlimited but you know $10 feels like it
20:26 should get you more than six [ __ ]
20:28 videos
20:31 um it depends what you want to do with
20:33 it a lot there's like like the video
20:35 world is really complicated right now
20:37 cuz all the tools are janky and so you
20:40 can get good results if you combine
20:42 tools and if you do things like like um
20:46 face swapping so you can get consistent
20:48 characters and you do that after the
20:49 fact and it ends up being like video
20:52 right now still ends up being Film
20:55 Production so I I don't know dream
20:58 machine couldn't extend a single
21:00 5-second clip on the free
21:02 option yeah I mean my problem with dream
21:05 machine is the free version right now
21:07 you you're you can make I don't know I
21:10 think five Clips a day but they take
21:13 like three or 4 hours to to go through
21:15 the queue one one day I had one take
21:18 eight hours to go through the queue so
21:20 that's not viable um Leonardo's got some
21:24 video tools Korea's got some video tools
21:27 Deco here's got some video tools there's
21:30 a bunch of tools out there
21:33 I I would say right now play with
21:37 everything
21:40 and try to find something that's not
21:43 ridiculously
21:45 expensive um and just be
21:51 patient like don't expect any of the
21:53 video tools to give you something that
21:55 is not fundamentally disappointing
22:01 back to back
22:03 to janky
22:06 Tails back to janky
22:10 tals all right so this this tweet this
22:14 little moment right
22:16 here that's a big
22:18 deal why is that a big deal that is a
22:21 big
22:22 deal because of these things not the
22:25 cherries the stems so if you don't know
22:27 anything about music stems are the
22:30 different
22:31 tracks that make up a
22:34 song so if you want to publish out a
22:36 song and you want people like DJs and
22:39 things or other musicians to be able to
22:41 easily remix it you release what are
22:45 called the stems and those are kind of
22:47 the mixed versions of all the different
22:49 tracks so you have your stems for the
22:50 drums and your stems for the percussion
22:53 and your stems for the bass and your
22:55 stems for the guitars and V vocals and
22:58 all that sort of [ __ ]
22:59 and then people can take those stems
23:01 like little Lego blocks and they can mix
23:03 them so it sounds like what's coming
23:06 from yudo this week this should be an
23:09 exciting week for udio creators and I
23:12 know Vicky you pronounce it udio that
23:15 just sounds wrong to me it sounds like
23:17 I'm saying something dirty so I'm going
23:19 with
23:20 udio
23:24 um studio is udio I I think you're right
23:28 I think you're right right but it's yudo
23:30 it's I'm going yudo it's this is we're
23:33 not going to talk about this this is not
23:35 a this is not a
23:36 democracy I I'm a dictator with the
23:39 pronunciation of this particular word
23:41 yud it is okay we're good anyway I think
23:45 they're getting stems we're getting
23:47 stems so what that means is you're going
23:49 to be able to make a song and if you
23:51 don't like the vocals you can redo that
23:54 stem or if you don't like the bass you
23:55 can redo that or you can turn the bass
23:58 into keyboard boards or whatever the
23:59 hell you want to do you can tell all
24:02 about it in your Uber oh that's just
24:04 rude Brandon see now we're getting
24:07 pronunciation police coming in here
24:10 trying to smack me down put me in
24:12 pronunciation
24:14 jail we live in America our our
24:17 consistency in pronunciation is all over
24:19 the [ __ ] map I wouldn't want to learn
24:22 English coming from another country cuz
24:25 it we we're just a we're just a pile a
24:28 heap and demon pile of trash here in
24:30 this country so I'm going to stick with
24:32 yio I'm doing it I'm also calling it
24:35 ideogram not idiogram some people call
24:38 it idiogram it's Kyle show and he can
24:40 pronounce it how he wants that's right
24:42 Source Camp you tell him Janu that's the
24:45 French
24:46 pronunciation hey Alan what's
24:49 happening you put the emphasis on the
24:51 wrong syllable Kyle exactly
25:01 so I thought that was pretty cool then
25:04 we had a Sam
25:06 Alman
25:09 um at
25:12 s released something
25:15 today well they released their teeny
25:18 little gp4 mini Omni
25:22 mini and apparently people are all
25:24 excited about that but everyone's just
25:27 like yes Sam when's the when's the other
25:29 [ __ ] coming um
25:32 apparently real time audio is coming
25:34 this month and we are on the 22nd so
25:38 we've got eight days and then Sam also
25:41 released today um their research Labs
25:45 put out their first
25:46 study on universal basic
25:52 income and let's see
26:04 how does unconditional cash affect
26:07 employment employment rates and work
26:08 hours Rose substantially over the course
26:11 of the study for both recipients and
26:13 control participants but the increases
26:16 for the control par participants were
26:21 larger on average re recipients were two
26:24 percentage points less likely to be
26:26 employed than control participants
26:29 recipients worked on average of 1.3 hour
26:32 fewer hours per week compared to control
26:35 the wide variation across participants
26:37 from diverse ways of people chose to
26:39 reallocate time reflect how cash can
26:42 increase people's agency to make
26:45 employment decisions that align with
26:46 their individual circumstances goals and
26:50 values so interesting study on universal
26:55 basic income I they basically gave
26:57 people an extra,
26:59 a month and then just studied how it how
27:02 it affected them so that's probably
27:04 something worth checking out it's O open
27:07 resarch lab.org
27:10 and and this is the first study they put
27:13 out I
27:23 think 3,000 people receed monthly cash
27:27 transfers for three years
27:31 why wasn't I in that
27:34 study I want an extra $1,000 a month to
27:38 for you could you can study
27:40 [Laughter]
27:44 me dang
27:47 [Laughter]
27:48 it oh
27:51 man so anyway um GP uh the new llama 405
27:57 billion parameter thing I think came out
27:59 or it leaked llama
28:01 3.1 leak outshines open AI GPT
28:07 4 but apparently it's giant so you can't
28:10 just install it on any old
28:13 machine and then my Twitter just turned
28:16 into Biden dropping out so so I I
28:20 actually there might be major AI news
28:23 but all I could see was we have a new we
28:27 we have a new candidate say yay boo yay
28:30 [Laughter]
28:34 boo oh you don't you don't think we're
28:38 studying you
28:40 Kyle oh I know I'm being
28:44 studied this is a [ __ ] case study and
28:47 something neuros
28:51 spiciness we like the neuros spicy here
28:54 we embrace
28:56 it why
29:03 not Biden's deep fake was amazing yeah
29:06 that was pretty funny I
29:08 mean the thing about deep fakes is they
29:13 they are emotionally
29:14 flat so you can tell you can tell right
29:20 away let me see if I can find
29:25 it let's see Biden uh
29:34 announcement yeah this one I want to
29:37 take a moment to address some of the
29:38 hateful [ __ ] you've been talking about
29:39 me many of you have said I am suffering
29:41 from brain worms or that I have
29:43 applesauce for brains well I won't m
29:45 words so here it is [ __ ] you you're all
29:47 a bunch of [ __ ] end of quote repeat
29:49 the line I'm probably president I'm
29:52 probably going to get uh whatever young
29:54 people can't watch this cuz I played
29:56 that but whatever um
29:59 but but yeah someone deep faked him
30:03 um and it's you know it's just very flat
30:06 you can you can tell but you know I
30:08 don't know next year next year they're
30:11 not going to be that
30:13 flat I've seen I've seen a camela one
30:16 Kamala one I've seen a Biden one I've
30:18 seen a couple of trump ones so they're
30:20 starting to come so I think we're going
30:22 to I think we're going to start seeing a
30:23 lot of those end quote repeat the line I
30:26 know it was pretty good
30:28 oh no the bands I know I'm sure I'm
30:30 going to get the little the little thing
30:33 here any second we heard a bad word on
30:35 your we heard a bad word on your Tik
30:37 Tock live there Mr sh we're going to
30:40 have to go on ahead and cancel you for
30:43 life cuz somebody said a bad word that
30:45 was on our
30:47 list our artificial intelligence caught
30:50 it there so we're just going to go on
30:52 ahead and ban you for life there
30:58 have you ever seen the Biden versus
31:00 Trump twitch don't play it
31:04 okay um all right what do we want to
31:08 talk about what is on your minds I will
31:11 talk about anything we can go play with
31:13 [ __ ] we can talk about
31:16 jankiness I can show you some of my new
31:20 book
31:22 covers he's actually happy with
31:26 these let me see where can I can I do it
31:29 here
31:32 yeah
31:34 so so I'm working on a book with Kevin
31:36 Clark called the rise of collective
31:38 intelligence in the age of
31:41 AI so there's cover number one actually
31:46 Brandon can you put together a
31:49 um what's it called a survey a poll a
31:53 poll of vote for your favorite book
31:57 cover of ky's one one two or three so
32:01 this is this is number one let's call
32:03 this one
32:06 forest and then here's cover number two
32:10 is
32:13 brain and here's cover number
32:17 three
32:19 is I don't know what that is grid we'll
32:22 just call it
32:24 grid so vote for Forest
32:34 brain or
32:37 grid One Forest by a mile one I like the
32:42 blue green book cover that would be
32:45 Forest this one this is the one that
32:49 yeah well all right Frank doesn't go to
32:51 Washington no Frank didn't get to go to
32:55 Washington 111 Forest all right Forest
32:59 it seems to be Forest pole is live okay
33:03 right over there right up there in that
33:05 corner is a
33:06 little blue thing on an angle it's
33:09 supposed to be a clipboard that's the
33:11 that's the pole so so there's
33:15 Forest there's
33:19 brain and then here is what's that
33:22 called uh GD
33:28 grid seemed more on brand interesting
33:30 yeah I think yeah this this one is the
33:33 most explicit in terms of what we're
33:35 talking
33:38 about this one is
33:40 obviously
33:44 obvious and this one feels like it kind
33:46 of struck a
33:47 balance of somewhere between neurons and
33:49 a forest and
33:52 right so you got 52 seconds left go vote
33:55 over there dang it I meant to vote for
33:57 Frank change my
34:00 vote Forest seems less cluttered it
34:04 is it is and actually all of the things
34:07 point in toward the title which is kind
34:09 of
34:10 cool one three is visually messy
34:13 confusing and
34:16 cluttered it's
34:22 true and this one's slightly too
34:24 biological
34:25 [Laughter]
34:31 Forest is more pleasing to the eye Kyle
34:33 there was a grid
34:35 one that was in blue green Hues from the
34:38 other night yeah I'll go look at that
34:42 one I just I was just messing around
34:44 with typography so I just wanted to put
34:46 some images behind it and once I got
34:48 them there I'm actually kind of pleased
34:50 with it I think the these
34:52 letters all right Forest wins by 55%
34:56 nobody likes the brain everyone's all
34:58 freaked out by the brain it should be
35:03 frank and grid grid comes in a solid
35:07 third but or solid second but but
35:10 clearly uh Forest is where it's at all
35:12 right that was the one that Kevin Clark
35:13 liked as well so might go with that one
35:18 these were all done in mid Journey mid
35:22 Journey
35:44 those were some of the forest ones that
35:48 were
35:49 Alternatives that's the one I
35:53 chose this one was also pretty
35:56 cool see seems a little too busy
36:20 though oh by the way I'll show you
36:22 something cool how I how I got to a lot
36:24 of
36:26 these I think that's it
36:41 um I also like how Forest looks like
36:44 neurons yeah I
36:45 agree um let's see did I have other
36:49 things in here yeah oh these were some
36:51 of the other ones I
36:55 had this one I kind of liked
36:59 but again it's a little bit busy there's
37:00 a lot to sort of take
37:03 in that one was kind of
37:09 cool that was one that could be okay but
37:12 it's a little little
37:18 abstract this one was sort of a brain
37:21 made up of people's faces which is a
37:23 little
37:27 creepy that was another one of those
37:29 grid
37:32 ones that one's just [ __ ]
37:36 terrifying all of these all of these
37:38 were pretty terrifying
37:48 [Music]
38:07 that very propaganda poster like I think
38:09 less is more yeah I agree so so let me
38:11 show you there's a there's a cool thing
38:14 that Vicky taught me I think it was
38:18 Vicki yeah it was
38:19 Vicki
38:21 um that I think I showed the other night
38:23 but it's it's really slick so one of the
38:26 things mid journey is really good at is
38:28 you can just put in single words or like
38:30 an obscure phrase and nothing else so
38:33 one of the things I was playing with was
38:35 typing in collective
38:42 intelligence and then D- sref r
38:47 sref as in Frank space
38:54 random and then I I just put the
39:01 uh the aspect ratio to vertical and just
39:05 hit
39:11 return and then let's see go D- sref
39:18 random and then let me copy all of
39:22 that and then just paste in like five of
39:26 those in a row
39:30 and then it will sort of do its thing
39:32 and sort of figure out what it thinks
39:34 collective intelligence means and it'll
39:36 pick a
39:43 style and so these are all in that same
39:46 style it just does random images from a
39:48 single word yeah
39:53 exactly but like that one's pretty good
39:55 like the kids version of collective
39:57 intelligence these would be good
39:59 [Laughter]
40:03 for like that one's pretty cool that
40:05 could be a really interesting book cover
40:07 right there's plenty of room for title
40:10 and stuff up above and below that
40:16 thing these are kind of
40:23 cool that sort of looks like a 1980s
40:29 cover of 1984 George
40:42 Orwell but like just just in terms of
40:45 brainstorming and just coming up with
40:47 ideas you know sort of a single word
40:50 like let's do uh let's do
40:56 embrace the
41:00 Jank wonder if I put that in quote marks
41:03 now I'm just GNA leave it like that dash
41:04 dash
41:07 sref
41:11 random
41:15 oops
41:17 oh
41:21 embrace the
41:23 Jank Dash DH sref random
41:33 copy and then we'll do just like five of
41:46 those these are kind of cool that one's
41:49 kind of
41:50 creepy creepy
41:55 cool what if you made a deco here
41:57 station for collective intelligence oh
41:59 let's try that okay so if you haven't
42:01 seen Deco here that's a really good idea
42:03 Silver
42:04 Fox
42:13 um what are you mailing
42:15 about you want to go out is it cheese
42:18 time you want to go beg for cheese
42:32 Champion the cheese
42:35 parade all right
42:40 um wait you have a cat no that was just
42:43 champy meowing like a cat he was just
42:47 whining these are kind of
42:50 cool love that style and what's cool
42:53 about these
42:56 is I can okay so I'm in by the way what
43:01 I'm using here is mid Journey this is
43:03 the web version of mid journey and over
43:06 here on the
43:08 right there's the prompt I used and then
43:11 it assigned because I did SRS sref
43:14 random it assigned an
43:17 srf so if I want to do a new one but
43:20 follow that same style I can click on
43:23 embrace the
43:24 Jank
43:26 oops let's see I think I have to be back
43:32 here oh those are cool so if I click on
43:35 embrace the Jank it'll populate this up
43:38 here and then if I click on this srf
43:42 it'll automatically put it in so now I
43:45 can hit
43:46 return and it's going to create a new
43:49 set of four Images up here that follow
43:53 the same
43:54 srf that I had here
44:05 oh these are embrac the
44:07 Jank that's
44:11 cool very janky like it very
44:21 janky and and you know this
44:25 is if you understand anything about how
44:28 these AI tools work there's a thing
44:30 called the latent space and the latent
44:32 space is this big mathematical crazy
44:36 thousand
44:38 dimensional mathematical space that has
44:41 all these
44:42 points that in the case of words are
44:45 semantic groupings in the case of images
44:48 are are sort of image groupings or
44:51 image like subject groupings I don't
44:53 quite know how it
44:55 works but when when you
45:02 if you want to get inspired this idea of
45:04 being able to use random style
45:06 references basically what it's doing is
45:09 it's popping you into random points in
45:12 this crazy ass visual
45:16 space um so we're just getting you know
45:20 really
45:22 crazy
45:24 variations on a single theme
45:28 that you know may or may not feel
45:32 like what you want to see but if you
45:34 have no idea what you want to do this is
45:36 a really good way to play and get your
45:39 head
45:40 around some new ideas like I love that
45:45 style love that
45:48 style oh wow look at that very
45:52 cool higher dimensional Vector groupings
45:55 exactly and and what s random does is
45:58 just pops you
46:00 around those different visual
46:09 languages love
46:11 that so here we're back to this is this
46:14 is that one that I that I explicitly
46:17 told it to use that same style and so
46:21 those these at the top here are
46:25 stylistically almost identical to these
46:28 right cuz we use the same sref number so
46:31 you can use sref random to find a style
46:34 you like and then just click on
46:37 collective
46:38 intelligence and then click on the style
46:41 ref and it'll give you that specific
46:43 thing so now I can go let's
46:45 see embrace the
46:51 Jank in this
46:56 style that's going to go off and do its
46:58 thing
47:00 now collective intelligence plus melium
47:03 plus avantgard plus special oh plus
47:06 spatial all right
47:11 wait
47:13 collective
47:18 intelligence
47:23 plus my
47:25 celium two l's
47:30 no
47:33 plus
47:35 aant guard
47:38 plus
47:40 spatial why a not
47:43 spatial and then I'll do dash dash
47:50 sref
47:54 random well you know what I'm going to
47:56 do I'm going to copy that so I'm going
47:58 to do one without any srf at all we'll
48:02 just do those words then I'm going to do
48:05 one with random and I'll do a few more
48:08 of
48:19 those Dr bod hello it has been a while
48:23 it has welcome back we have we have a
48:25 number of returning returning Irregulars
48:29 I love
48:30 it I love it when when Irregulars are
48:33 like hey it's been too long it's been
48:36 too irregular I got to get back in
48:39 there it's very
48:44 good all right what am I doing
48:48 here I'm lost okay so here these are
48:52 embrac the
48:54 Jank with that blue one
49:01 oops that one's kind of
49:05 creepy and then what's this one oh this
49:08 is the this is yours
49:10 Corey collective intelligence plus
49:13 melium plus avantgard plus
49:16 spatial that one kind of cool this went
49:19 heavy on the
49:22 mycelium all right here's
49:24 another oh so that was this was with
49:27 without any style reference these four
49:29 were without any style reference so
49:32 leaned in heavy into the into the
49:36 mushrooms and then these are with style
49:41 references like that grid plus mushroom
49:50 equals o that's cool love that look the
49:54 the drawing here and then the the
49:56 mushroom sculpture sure that matches
50:01 it huh there's a hand with sort of a
50:05 normal
50:07 thumb looks like an art
50:12 installation looks like every artist
50:15 Instagram closeup of their
50:17 [Laughter]
50:20 work oh these are
50:24 cool here's one of those weird things
50:26 where it gets the in in the latent space
50:28 it obviously trained on images that had
50:31 you know people's names at the bottom so
50:34 that's in in the Laten
50:43 space here you go back to the
50:47 60s that's
50:50 cool it thinks fungi equal
50:53 mycelium yeah I know well let's just do
50:57 my
50:58 celium did I spell it wrong maybe I
51:00 spelled it wrong that's another
51:03 possibility my
51:09 cium my I think I got it right yeah I
51:13 got it right let's just try my celium
51:15 and see what it does
51:33 this shit's just crazy it's crazy crazy
51:36 I tell
51:50 you yeah it's it's not isn't the
51:54 mycelium the the sort of Network
51:57 structure of the of the mushrooms kind
52:00 of the the connective tissue it's not
52:03 really getting
52:12 that all right so let's go over and play
52:15 with Deco
52:17 here Silver Fox had this idea so okay so
52:23 where we just were was um Mid Journey um
52:26 as as you can tell mid journey is very
52:28 very powerful it's very fast
52:32 um it is incredibly rich in your ability
52:38 to just kind of mine ideas that you know
52:41 you didn't even think of you can be very
52:43 intentional with it you can be very
52:45 non-intentional with it and just sort of
52:47 let it take you places which I think is
52:50 a fun way to use these tools in general
52:52 you can do that with words you can do it
52:53 with images you can do it with
52:55 music um
52:57 this next tool Deco
53:05 here let me log in here I'm got to
53:08 brighten my screen up a little bit I
53:09 can't see
53:11 anything basically if I can see it it's
53:14 too bright for for Tik Tok claim my
53:16 first
53:18 reward I got 200 turbo images yay you
53:22 did it
53:24 yay okay so
53:28 one of the cool things that you can do
53:31 with Deco here I'm going to do
53:36 a do I want taller or do I want wide
53:39 let's go
53:41 wide one of the cool things you can do
53:43 with Deco here is it's real
53:47 time so I can type
53:50 in colective
54:04 intelligence
54:08 and oh that one's kind of cool and it it
54:12 basically makes it in real
54:16 time and if I find one that I
54:19 particularly
54:23 like that one's kind of crazy that's
54:26 kind of medus
54:28 Meda Medusa collective
54:34 intelligence that seems very stock phoy
54:37 doesn't it bad stock
54:46 photo all right that one's kind of
54:49 cool so Deco here can can generate
54:53 images really quickly um like for
54:56 example example if I if I wanted these
54:58 to be red I could just throw the word
55:00 red at the end of
55:02 this and it it'll
55:05 automatically give me red things
55:13 now or I could say collective
55:17 intelligence
55:21 wispy and now all of the images will be
55:25 inspired by wispy
55:30 and you just keep going right but for
55:33 any image that I find there's this
55:36 option right here called
55:38 station and if I click on the station
55:40 let me get rid of wispy H maybe I'll
55:42 keep wispy [ __ ] it I'll keep wispy oh
55:45 ethereal let me try ethereal although I
55:47 kind of like that image let's let's
55:49 let's let's go with this so if I click
55:51 on this word station
55:58 oh apparently stations are no longer
56:07 free create
56:12 station ah creating a station from an
56:16 image that I created is not free okay so
56:19 I've got to upgrade for that
56:22 okay so that's a bummer oh well um but
56:25 what stations are we're actually looking
56:27 at a station but you wouldn't you don't
56:28 you wouldn't know that but for any image
56:30 if you go to the explore tab at the top
56:33 of Deco
56:34 here and it starts making images we'll
56:37 wait till we see something here that's
56:38 kind of cool
57:06 I'm not liking any of these right now
57:08 all right there's a we'll try this watch
57:10 all right so I click on that watch I can
57:13 turn this the illusion of instant yeah I
57:15 know it totally
57:21 is I can click on that and this now
57:23 creates what they call a station and so
57:26 the the word station how they're
57:28 defining it is like a radio station a
57:31 radio station is this passive listening
57:33 experience this is a passive image
57:36 generating experience what's cool about
57:38 stations is I can copy this
57:42 station and I can share it with people I
57:45 could share it with you I could share it
57:48 on the AI salon and you'll see the exact
57:51 same images we can actually look at the
57:52 same images together yeah Deco here is
57:55 not safe for work I would agree with
57:56 with that it's it's really not safe for
57:59 for Tik Tok either but you know you know
58:02 me I live on the
58:03 [Laughter]
58:07 edge
58:09 so
58:23 um let me see my saved ones
58:28 I've got some really cool ones in here
58:29 this one's one of my favorite
58:32 ones abstract
58:38 portrait it just does all these
58:41 cool very analog kind of
58:46 things I've always been super intrigued
58:51 by analog representations in digital
58:55 like with music I don't like electronic
58:57 music but I love the idea of like using
59:01 DJ tools to mix super acoustic sounds
59:05 and actually Maggie Rogers kind of came
59:07 out and did that her stuff is this weird
59:09 mix of electronic and
59:13 analog or
59:19 acoustic cool
59:21 right they seem to be lesser quality
59:24 than they used
59:25 to yeah yeah I don't know they I mean
59:28 you know listen it's I I think with with
59:32 all of these
59:37 things I mean if you're relying on the
59:39 public the public in general is going to
59:41 be lazy in how they prompt so most of
59:44 the crap you see on the explore side's
59:46 going to be really shitty um but you'll
59:48 occasionally see something good in there
59:50 like if you you know take more than 3
59:52 minutes like I do um you'll be able to
59:55 find some good stuff in there and if you
59:56 listen if you pay for this
1:00:00 subscription you can make your
1:00:02 own that one's pretty
1:00:05 cool let me make a station out of this
1:00:08 create
1:00:11 station someone just sent me a gift
1:00:13 thank you very much whoever just sent
1:00:15 the gift madcap thank you very
1:00:18 much appreciate that I don't see them
1:00:20 all the time so if you sent me a gift
1:00:22 and I didn't acknowledge you my
1:00:23 apologies for that just sometimes I
1:00:25 don't see that [ __ ] cuz I'm off looking
1:00:28 at that screen over there how's the
1:00:30 Gemini bear these days Gemini bear is
1:00:33 good hold your mouse over the X just in
1:00:36 case oh the to to shut it
1:00:40 down uh oh I hear champy wants back in
1:00:43 hang on
1:01:01 did you get your cheese did you get your
1:01:04 cheese
1:01:08 buddy champ be the cheese
1:01:12 eater did I get did I get
1:01:15 booted did I get kicked off
1:01:24 ticktock my inter pretation of your book
1:01:27 covers in show and tell that's cool
1:01:29 let's go look at that all right so where
1:01:31 I'm going now is the AI Salon if you're
1:01:33 new to the AI
1:01:36 Salon go to the
1:01:39 salon. and that will take you to this
1:01:42 page which will tell you
1:01:45 about the journey to AI
1:01:48 literacy and our values and then if you
1:01:51 scroll down on the left hand side toward
1:01:53 the bottom there's a group called
1:01:56 Community clubs and in the first Club
1:01:59 there is called AI learning lab
1:02:01 Irregulars the rise of collective
1:02:04 intelligence in the age of AI very
1:02:10 nice
1:02:17 beautiful bot [ __ ] bot [ __ ] is better
1:02:21 than hallucination
1:02:26 that's pretty
1:02:28 good yeah the hallucination word is a
1:02:31 really interesting one because it's
1:02:35 like if I'm trying to do something
1:02:37 creative hallucinations are good they're
1:02:40 going to give me [ __ ] that's creative
1:02:42 but if I'm trying to do something
1:02:43 accurate hallucinations are bad so I
1:02:46 think bot shit's kind of
1:02:49 nice gave me bot [ __ ] again what am I
1:02:52 going to
1:02:54 do I have so many AI sub that I've
1:02:56 forgotten which ones I've sub subscribed
1:02:58 to I'm totally like that forever hooked
1:03:02 H I've got I I actually just cancelled
1:03:05 three of them I forget which three I
1:03:10 cancelled I don't even
1:03:17 remember but it's like I don't know how
1:03:20 you all are but I find myself going
1:03:22 through waves of I'll go through an llm
1:03:26 wave and then I go through an image wave
1:03:28 then I go through a video wave then I go
1:03:30 through a music wave I go through these
1:03:32 waves of output
1:03:34 types and then when I'm in one of those
1:03:36 waves I kind of completely ignore the
1:03:38 other tools and even forget they exist
1:03:40 and forget that I subscribe to them and
1:03:42 why they were so cool in the first place
1:03:45 I think it's going to be like that for a
1:03:46 while I I just you know I think the next
1:03:48 two years are just going to be this
1:03:50 rolling series of like oh that's cool
1:03:52 you know shiny object syndrome um while
1:03:55 we wait for the tools to
1:03:57 become
1:04:00 consistently
1:04:01 professional cuz right now they are
1:04:04 consistently
1:04:07 [Laughter]
1:04:09 janky oh
1:04:13 man I pay for one sub and don't use it
1:04:17 I'm in my music wave Jim Jim Jim Ross is
1:04:20 in his music wave in 15 minutes just now
1:04:23 made 10 different versions for my
1:04:24 podcast intro it's amazing well that's
1:04:28 like like all those images I was just
1:04:29 showing in in mid
1:04:31 journey I mean I just made a [ __ ] ton
1:04:34 that that were just us [ __ ]
1:04:37 around but let me see let me go to the
1:04:40 book cover
1:04:43 Explorations like there was
1:04:47 one there's
1:04:50 four there's four these are all book
1:04:53 cover
1:04:54 Explorations none of those were
1:04:56 successful like these were kind of
1:05:00 cool I mean these are very kind of like
1:05:03 early Del kind of kind of
1:05:09 styles but like just the ability to just
1:05:11 go crazy like these are all book cover
1:05:15 studies these are all book cover studies
1:05:18 those weren't those
1:05:21 are it's just crazy
1:05:38 like that could have been one the rise
1:05:40 of collective intelligence just some
1:05:41 sort of weird
1:05:43 ass crystal ball into the
1:05:47 future or just you know here's people
1:05:50 here's the the brain we all get to tap
1:05:52 into
1:06:04 that was one that I thought was pretty
1:06:05 cool a little busy but you
1:06:08 know I have photo leap I have photo leap
1:06:11 $55 not well
1:06:14 spent this is my my one piece of advice
1:06:17 on subscriptions is do not pay for the
1:06:19 year in advance for a while just pay the
1:06:23 monthly thing pay the extra per month
1:06:26 because just too much is coming too
1:06:29 fast that one kind of
1:06:34 works and actually you
1:06:37 know from a creative standpoint you know
1:06:40 people talk about well what's art and is
1:06:42 it art if if AI made
1:06:45 it part of
1:06:48 art is
1:06:51 choosing is curation what do you choose
1:06:54 to show the world right
1:06:57 what I
1:06:59 chose to put on my book cover was that
1:07:03 image right
1:07:05 there no one other than in this live no
1:07:08 one will ever see that image but that
1:07:10 could have been the cover of a
1:07:14 book that could have been that could
1:07:16 have been that could have been that
1:07:18 could have been you know it it's
1:07:21 like for all this stuff here
1:07:26 part of art is the moment you choose to
1:07:29 stop right you you choose to go oh
1:07:32 that's
1:07:34 done or that's going to be it is that
1:07:37 the rabbit R1 over your right shoulder
1:07:39 yes it is as a matter of fact I don't
1:07:42 think I'm charged
1:07:45 though oh yeah let me see if it's got
1:07:48 any charge in it it
1:07:50 might nope I'll charge it we'll see if
1:07:54 we got maybe we get an upgrade
1:07:57 um the rabbit R1 now has sunno um hooked
1:08:00 up to it so you can just say hey make me
1:08:02 a song about whatever and it does
1:08:05 it okay I'm in I should
1:08:14 [Music]
1:08:19 say art is a
1:08:21 feeling if you feel something then you
1:08:24 experienced art on the receiver side but
1:08:26 as a as an artist on the expressing side
1:08:31 art is when you choose to share
1:08:33 it well you choose to make
1:08:36 it then you can choose to not destroy
1:08:40 it then you can choose to share it and
1:08:43 if you're a good artist well not a good
1:08:47 artist if you've got a healthy
1:08:49 relationship with your art when you
1:08:51 share it with the world you let it
1:08:54 go but that's really hard to
1:08:57 do [ __ ] ego [ __ ]
1:09:02 ego like the silence between the notes
1:09:05 exactly right yeah
1:09:07 exactly made it here this week welcome
1:09:22 Jason if it comes back to you exactly if
1:09:25 it comes back to you it's yours it's
1:09:29 wait it's yours right if it doesn't it
1:09:31 was never meant to
1:09:34 be all right my rabbit she's a
1:09:42 charging Jason wondered where you were
1:09:44 it's going fine here in Ohio nice
1:09:47 welcome
1:09:52 back oh lordy
1:09:57 lordy what else we want to talk about
1:09:58 what who's got something going on
1:10:00 anybody doing anything fun with
1:10:03 AI anybody have any
1:10:06 questions magic
1:10:15 trick have you ever used any automation
1:10:17 tools yeah I used so who asked that that
1:10:22 was Michael Montoya
1:10:27 um I've used zapier um make is a very
1:10:31 similar
1:10:32 tool um Bubble not so much bubble claims
1:10:36 to be a low code tool but it's it's a
1:10:39 very it's bubble is more complicated if
1:10:41 you want to create a standalone
1:10:43 application you can do it in bubble but
1:10:45 it's more complicated zappier and make
1:10:48 and there's uh probably another one or
1:10:50 two if if this then that I think is one
1:10:52 that's in the
1:10:54 neighborhood um
1:10:56 when I first started getting into AI I
1:10:59 was doing a lot of automation a lot of
1:11:03 it
1:11:10 because the first place I went when I
1:11:13 saw how
1:11:16 fast AI could do what it does the first
1:11:19 place I went was
1:11:22 oh I can make what we do now more
1:11:25 efficient
1:11:27 and so I made a tool for my company
1:11:30 using zapier that we still use that that
1:11:33 writes prompts for video
1:11:35 scripts
1:11:37 um I created some content generation
1:11:41 machines
1:11:43 um I created a number of different
1:11:45 things with it and
1:11:49 then it wasn't that I soured on
1:11:52 automations it's that I realized
1:11:58 that what generative AI makes possible
1:12:01 is so far
1:12:03 beyond automating what we do
1:12:07 now or or making what we do now more
1:12:11 efficient it's going to completely
1:12:14 transform what we do and how we do it
1:12:17 and so where I've been trying to push
1:12:21 this Channel and push myself
1:12:24 is on the front edge of the wave right
1:12:29 because I I feel like automating what we
1:12:32 do now
1:12:37 is is kind of looking
1:12:41 backwards and when you have a tool that
1:12:43 can actually do coding for
1:12:47 you and when these tools get reasoning
1:12:50 which I have a feeling that we're going
1:12:52 to see something within the next month
1:12:54 or two from open AI that has reasoning
1:12:57 in
1:12:58 it
1:12:59 um these are tools of a very very
1:13:02 different
1:13:03 nature than a fancy Microsoft Excel or a
1:13:07 fancy Microsoft Word or even a fancy
1:13:09 Google search they they're fundamentally
1:13:12 different they're generative the G and
1:13:14 GPT generative generate generating
1:13:18 generating that's the stuff that I want
1:13:20 to lean
1:13:22 into that's a good idea Brandon yeah
1:13:25 I'll do that any recommendations on
1:13:26 creating a
1:13:27 GPT of someone that's dying soon
1:13:32 um first of all I think that's a really
1:13:34 beautiful
1:13:35 gesture um assuming that they're still
1:13:39 you know conscious and
1:13:41 engaged
1:13:46 um I
1:13:48 would so there's a couple of things you
1:13:50 can
1:13:52 do the digital twins that I've made that
1:13:55 we've made for Content Evolution and
1:13:57 that we made for the the AI Futures book
1:14:00 we actually
1:14:03 designed
1:14:05 um a set of interview questions it's
1:14:08 like you know 25
1:14:10 questions
1:14:16 um wait brain works great is my grandma
1:14:19 oh
1:14:21 cool who's in here oh that's
1:14:24 amazing um
1:14:28 so there's a couple of ways you could do
1:14:29 it if if the person can still talk and
1:14:32 can you know still have a conversation I
1:14:34 would say just
1:14:37 record a bunch of conversations and if
1:14:39 you put together a structured interview
1:14:41 like we we went to to chat GPT and we
1:14:45 said if we wanted to create wait before
1:14:47 she dies I want to create a GPT of my
1:14:49 grandma Yeah so
1:14:53 um so I would go to
1:14:56 GPT and I would say I want to create a
1:15:00 GPT of my grandma assuming she can still
1:15:02 talk
1:15:04 create a set of
1:15:06 questions um that capture you know that
1:15:10 ask her questions about her childhood
1:15:12 and her favorite memories and all that
1:15:15 sort of stuff chat chat gbt can help you
1:15:17 create a structured interview that's got
1:15:19 different sections in it and I would
1:15:23 record them in video if you can do that
1:15:25 if not you can get writings of her you
1:15:29 could also do interviews of other people
1:15:31 in the family about
1:15:33 her um but if you can at all capture her
1:15:38 talking in in her voice and do it on
1:15:41 video you won't use the video right now
1:15:44 but it's going to be valuable in the
1:15:46 future cloner voice as well yeah exactly
1:15:49 but but capture it now and then you can
1:15:51 take that transcript so how we made our
1:15:53 digital twins is we recorded everyone
1:15:55 answering these 25
1:15:57 questions we then
1:16:00 transcribed
1:16:02 the the video and got a a transcript and
1:16:05 then we use that transcript as the basis
1:16:08 for a digital
1:16:09 twin as soon as viable digital twin
1:16:14 technology comes along that can talk and
1:16:17 move we'll take the video that we
1:16:19 recorded and use that as the as the the
1:16:22 training video to create a talking
1:16:25 version of those digital twins right now
1:16:27 they're text only but they're remarkably
1:16:29 good like they're really good and if you
1:16:32 design that set of prompting questions
1:16:35 in the right way um and get her to
1:16:38 answer those questions you'll be able to
1:16:40 capture a ton of stuff you could also
1:16:42 take if she's got any
1:16:44 writings um or journals or things like
1:16:47 that don't forget chat GPT has Vision so
1:16:50 you could take pictures of handwritten
1:16:52 journals and chat GPT can convert all
1:16:55 that
1:16:57 Kyle serious question should she also
1:16:59 record a consent statement for use for
1:17:03 the
1:17:05 future um if it's her grandmother and
1:17:09 it's like it depends on the usage I mean
1:17:12 I suppose it would be a nice
1:17:19 courtesy you know maybe a power of
1:17:21 attorney like I don't know if she is
1:17:24 dying right now or if she if you just
1:17:26 want to do it before she
1:17:28 dies but yeah it's not a bad idea to get
1:17:30 someone's permission to do that I think
1:17:32 we're listen we are all going to have to
1:17:35 confront in the next some amount of
1:17:39 years people using our likeness whether
1:17:44 we know about it or not um and so one of
1:17:48 the things I'm trying to do right now is
1:17:50 get ahead of it and proactively do it
1:17:52 for myself and the people that I work
1:17:54 with um so that we can learn how to do
1:17:57 it and we can be good at this because I
1:17:58 think we're going to all all have to
1:18:00 we're going to there are going to be
1:18:01 skills in how to use your likeness chat
1:18:04 TPT came up with a great structure when
1:18:06 I recorded my mom when I thought she was
1:18:10 going to pass there you go perfect yeah
1:18:13 chat GPT is really good at that you tell
1:18:15 it what your goals are say here's what
1:18:16 I'm doing here's the project I'm doing
1:18:18 write me a set of questions it'll be
1:18:20 really good and then go through there
1:18:21 and just edit them like don't if you
1:18:23 don't like them throw them out WR at
1:18:25 other ones in you can't create a Hy
1:18:27 genen video of someone else yet for
1:18:29 example no you can't um well you can if
1:18:33 if you've got them there and they can
1:18:34 record the
1:18:36 consent you
1:18:39 can um but yeah in general they're
1:18:42 trying to be safe about that which I
1:18:45 understand um we did we did one for this
1:18:48 teacher for Monique's uh kids school and
1:18:52 um we got rejected by hey Jen she
1:18:55 recorded the consent but we got rejected
1:18:57 because they thought she was someone
1:18:59 famous she wasn't but she looks like
1:19:02 someone famous um I'm recording me
1:19:05 reading children's books for my grandson
1:19:07 doing November oh that's awesome
1:19:09 wonderful love
1:19:12 that yeah and again you
1:19:15 know play around with these tools and
1:19:18 play around with
1:19:21 um making digital twins of yourself and
1:19:25 you
1:19:26 know Lori make a like do an 11 Labs
1:19:30 model of your voice and then you can you
1:19:33 can send stories uh to your grandkid
1:19:36 that'll be spoken by you um or you can
1:19:39 just send them written stories and they
1:19:40 can use 11 laabs reader oh that was the
1:19:42 other thing that came out today if
1:19:44 you're on Android um 11 Labs reader app
1:19:48 came out for Android today as well can
1:19:49 you talk more about how you use your
1:19:51 likeness to get ahead of the curb
1:20:00 yeah
1:20:07 um so when we first started thinking
1:20:11 about digital twins in for Content
1:20:19 Evolution the original idea was we we've
1:20:22 got
1:20:23 this Federation of different companies
1:20:26 and different
1:20:28 professionals and one of the one of the
1:20:30 things that we've always wanted to do
1:20:33 within content Evolution was make it
1:20:35 easy for someone to figure out who they
1:20:37 should talk to if I'm if I'm facing this
1:20:39 problem in my business who within
1:20:42 content Evolution would be good to talk
1:20:44 to and our initial Instinct
1:20:49 was um let's go scrape LinkedIn right
1:20:53 and in LinkedIn you've got their work
1:20:55 history and so you could be able to say
1:20:57 hey if I'm dealing with this kind of
1:21:00 problem and then we thought well maybe
1:21:02 it should be a combination of LinkedIn
1:21:04 and what they've written so CU some
1:21:06 people in content Evolution have written
1:21:08 books and some people have you know have
1:21:11 got you know lots of Publications
1:21:14 they've done and blogs they've done and
1:21:16 [ __ ] like
1:21:19 that and we just kept we kept sort of
1:21:21 nipping around the heels of this this
1:21:25 thing about what do we use to represent
1:21:28 the person like how do we get chat gbd
1:21:31 to understand the
1:21:33 person and what what struck me one day I
1:21:36 was talking to Kevin about this and what
1:21:38 we wanted to do and how we wanted to use
1:21:39 these
1:21:40 things and what struck me was all of
1:21:44 that stuff that we were going to build
1:21:47 these digital twins on was done in the
1:21:51 past your LinkedIn is a historic record
1:21:55 of [ __ ] you did
1:22:00 before
1:22:02 um all the blog posts are things you
1:22:05 were thinking about back
1:22:09 then and and so we thought
1:22:12 well how do we capture who a person is
1:22:15 right
1:22:16 now and and then the minute we had that
1:22:19 we were like oh wait a minute but there
1:22:20 there's so much more to a
1:22:23 person than their work history or than
1:22:25 what they've
1:22:27 written so he said how can we capture
1:22:30 who they are today and that's when we
1:22:32 came up with this idea of
1:22:34 designing an
1:22:37 interview um and then designing that
1:22:41 interview like we were like well what
1:22:43 should be in it like what are the
1:22:44 important elements to have in it and
1:22:46 like one of the things was like well how
1:22:48 does the person solve problems um how do
1:22:51 they think creatively how have they
1:22:53 responded to diversity in their life um
1:22:57 who are people who have inspired them
1:22:59 who do they look up to um what have they
1:23:02 done professionally right what what have
1:23:04 they done personally that they're really
1:23:05 proud
1:23:07 of so we tried to create a set of
1:23:10 questions that that was as broad a swath
1:23:12 of who that person is today as possible
1:23:16 and we decided to just use that
1:23:19 interview as the source none of the
1:23:22 other stuff just that interview and the
1:23:24 reason we wanted to do that is we wanted
1:23:27 the same interview to be used across all
1:23:29 the digital twins so that we could
1:23:32 more accurately figure out if if those
1:23:35 digital Twins were representative of the
1:23:38 person and they're surprisingly good
1:23:40 they're really good because chat GPT
1:23:43 knows lots of other stuff so you can ask
1:23:45 a digital twin things Way Beyond that
1:23:48 interview and it'll answer them because
1:23:49 it's got access to all of gpt's brain
1:23:53 right
1:23:55 so that was sort of how we got there and
1:23:58 then the minute we got
1:24:00 there we we we started talking about
1:24:02 well we're going to give these out to
1:24:04 we're going to publish these cuz like
1:24:08 I'm thinking like old school publishing
1:24:11 we're going to we're going to create
1:24:12 this piece of software that represents a
1:24:15 person and we're going to publish it to
1:24:16 the world and then people are going to
1:24:18 use that and what we very quickly
1:24:21 realized is not only do people out in
1:24:23 the world not know how to use a digital
1:24:26 twin we didn't know how to use a digital
1:24:30 twin and so so I actually
1:24:35 shifted the the
1:24:39 the the goal for the for the collab this
1:24:43 AI group that I run within content
1:24:45 Evolution the goal used to be to build
1:24:48 [ __ ] you know let's let's figure out
1:24:51 what these new tools are you know we're
1:24:53 going to talk about the tools and we're
1:24:54 to go build [ __ ] and all of this year
1:24:58 it's been
1:24:59 about us learning how to use the [ __ ] we
1:25:04 built so I'm right in the middle of this
1:25:08 idea
1:25:12 of
1:25:14 proactively figuring out how to
1:25:16 represent myself in the world and part
1:25:18 of that is look and feel and part of
1:25:20 that is voice but most of it right now
1:25:23 in terms of t tangible interactive stuff
1:25:27 is just text and that's going to get
1:25:30 like that's going to evolve very very
1:25:32 quickly but I think starting with text
1:25:34 is a good thing so anyway Kyle is a
1:25:37 digital twin akin to a custom GPT but
1:25:40 with specific reference sources yeah
1:25:42 that's exactly what it is it is it so we
1:25:45 have a series of custom gpts we have one
1:25:49 per person so we've got like 22 digital
1:25:53 twins
1:25:55 so if you go to chat
1:26:00 GPT and you say explore
1:26:13 gpts let me see if if you type in
1:26:16 digital
1:26:17 twin digital twin
1:26:27 now type in C
1:26:32 profile yeah type in C
1:26:39 profile and then you'll see a bunch of
1:26:41 CE profiles there's Cindy
1:26:44 [ __ ] Clark Haywood Mike Wing Joe gaml
1:26:47 Todd Hoskins Kyle Shannon there's my
1:26:49 digital twin so you can go play with
1:26:51 these digital twins right now
1:26:57 um and then we have a master GPT that's
1:27:01 got all the digital twins within it and
1:27:04 that one's really cool because for that
1:27:06 one we can say write an article and and
1:27:09 grab a quote from three the three people
1:27:11 that are most qualified to talk about
1:27:13 what this article is about and so right
1:27:16 now we're just trying to figure out how
1:27:18 do
1:27:19 you what is the purpose of these and
1:27:22 like where is their value and why would
1:27:24 someone use this and you know are we
1:27:27 going to use this and is it of value to
1:27:29 someone else to use this and the answer
1:27:31 right now is just a massive I don't
1:27:33 [ __ ] know I don't have a clue how do
1:27:37 you recommend
1:27:38 people get comfortable with letting our
1:27:41 likenesses out in the world well so
1:27:44 there's a there's a couple of things
1:27:45 about that Cam that I think are really
1:27:47 interesting cuz I've been thinking about
1:27:49 this very much like you know okay what's
1:27:52 the professional version of me like one
1:27:53 of the lessons that I
1:27:56 learned when I did my my C my content
1:28:00 Evolution digital twin is it's it's
1:28:04 that's a very professional group and
1:28:06 when I told people in here about it and
1:28:09 they went and played with it they're
1:28:10 like it doesn't say embrace the Jank and
1:28:13 it doesn't swear it's not it's not like
1:28:15 you and I realized that oh there's
1:28:17 different versions of me right like like
1:28:20 as humans we context shift and we shape
1:28:23 shift um Vicky doesn't V Vicky was like
1:28:26 I don't who does that who shape shifts
1:28:28 who who context switches a lot of people
1:28:31 do um and so I think I think one of the
1:28:35 things that we're we as humans are going
1:28:39 to confront is which version of us do we
1:28:42 want out in the
1:28:43 world in which Con in in a certain
1:28:46 context right if I'm trying to get a job
1:28:50 that's going to be a different kind of
1:28:52 digital twin out there than if I'm
1:28:54 trying to get a date right or if I'm
1:28:56 trying to I don't know hang out with my
1:29:00 friends right or you know do do
1:29:02 something like that like I might have
1:29:03 different versions of
1:29:05 me
1:29:07 um in kicking around the idea of digital
1:29:12 twin management with Cindy [ __ ] and HT
1:29:15 snow day I think it was HT came up with
1:29:18 the idea he was like well wait a minute
1:29:20 not only should I be able to have
1:29:22 different versions of me but I might I
1:29:24 might want to have different fictitious
1:29:27 characters that represent me but no one
1:29:30 knows it's me right so so cam katkin
1:29:34 from Cleveland you might put out you
1:29:36 know a a a kitty cat that's got a really
1:29:39 wild personality and you know it's you
1:29:41 and you put it out in the world and you
1:29:43 let people interact with it but no one
1:29:45 knows it's you but you're still managing
1:29:49 that personality that that instance of
1:29:52 you in the world so I think that this is
1:29:56 something that is going to be incredibly
1:30:00 creative incredibly personal incredibly
1:30:03 expressive I think that talking
1:30:08 about capturing people before they
1:30:12 pass I mean you know talking about
1:30:15 capturing your grandmother before she
1:30:18 passes maybe you do two or three
1:30:20 different versions of her maybe there's
1:30:22 the young version of her when she was in
1:30:24 her 20s you do questions just about her
1:30:27 when she was a little girl and then you
1:30:29 do other ones about her as as grandma
1:30:32 now looking back on her life and then
1:30:34 you do I I I don't know like I could
1:30:36 imagine there being all sorts of
1:30:38 different creative ways that we can
1:30:40 choose to express ourselves and I think
1:30:42 that's something that we're going to
1:30:43 have to confront and deal
1:30:46 with
1:30:48 um the whole purpose of this channel is
1:30:52 well why don't we play with it now where
1:30:54 it's relatively low risk and these tools
1:30:58 are as janky as they are and anything we
1:31:01 build today is probably going to be op
1:31:03 why not just
1:31:07 play
1:31:12 stop so I don't I don't have a real good
1:31:14 answer other
1:31:18 than you
1:31:22 know other than you know start playing
1:31:25 make a GPT of
1:31:27 yourself figure out who is the U you
1:31:31 would want to be in your first digital
1:31:33 twin and go make
1:31:38 it and I can show you how to do that at
1:31:40 some point it's getting a little late
1:31:41 tonight to start one of those but maybe
1:31:43 tomorrow we can do that cuz I'm not
1:31:44 going to be in Washington DC so I will
1:31:47 probably be
1:31:49 here uh Sam and open AI should really
1:31:51 release digital twins
1:31:54 yeah they should I mean it's a it's it's
1:31:57 a relatively straightforward thing to do
1:31:59 on your own but it's not you know it's
1:32:00 not super intuitive to do it I don't
1:32:02 think people are running around right
1:32:04 now going huh wonder what I should do
1:32:06 with my digital twin this has been this
1:32:08 has been sort of a year-long year and a
1:32:10 half long Evolution for
1:32:12 me you know just trying to solve a very
1:32:15 specific problem in this one
1:32:16 organization and and then kind of
1:32:18 realizing oh [ __ ] like this if if you do
1:32:23 your digital twin right
1:32:25 it can be really powerful and really
1:32:27 fun but what does that mean do it right
1:32:29 do it right for who in which context all
1:32:32 that sort of stuff
1:32:34 so crazy
1:32:39 times all
1:32:41 right
1:32:43 um let's see one thing I do want to
1:32:46 share if you're new here especially if
1:32:49 you're new
1:32:50 here um go there go to the salon.
1:32:56 Ai and that's going to take you to our
1:33:00 brand spanking new
1:33:08 website and if you scroll
1:33:11 down it says join us and this little
1:33:14 button to the right says join our
1:33:16 community of AI
1:33:18 optimists and if you click on that
1:33:20 that'll jump you over to the AI salon
1:33:22 and the first place it's going to take
1:33:24 thank you is to Welcome to the salon can
1:33:27 we talk about gp4 mini please um we can
1:33:31 let me just do a quick little run
1:33:33 through of the AI salon so when you get
1:33:35 to the AI Salon check out the welcome to
1:33:37 the AI
1:33:38 Salon um read about what we're about
1:33:42 read about this what we call the the uh
1:33:45 journey to AI literacy play first
1:33:47 mindfully create generously lead read
1:33:49 about our values and if you like those
1:33:52 then explore the space the first thing
1:33:54 you should
1:33:55 do is under the Community section the
1:33:58 first link there is introduce yourself
1:34:00 is jump in and introduce yourself use
1:34:02 your real name um use either an avatar
1:34:05 or a picture of yourself um join this
1:34:09 community this is a community of really
1:34:11 remarkable people curious about AI
1:34:14 exploring AI optimistic about AI so if
1:34:17 you feel like you're the only [ __ ]
1:34:19 weirdo that's like I kind of like this
1:34:21 AI but no one in my life likes it I
1:34:23 don't know what to do this is the group
1:34:25 for you um I firmly believe that over
1:34:29 the next 2 to three years being a
1:34:31 trusted member of a community like this
1:34:33 is going to be really really important
1:34:37 for people's
1:34:38 employment uh and and engagement as AI
1:34:43 democratizes the skills that we all went
1:34:46 to school for and makes it so that
1:34:49 anyone can do any work who we're going
1:34:52 to choose to work with are the people
1:34:53 that we know and we trust and we like so
1:34:57 become a functional contributing member
1:34:59 of this community and it will serve you
1:35:01 really well um there's all sorts of cool
1:35:03 things in here um there's a show and
1:35:05 tell area there are guilds uh creative
1:35:08 guilds business guilds Tech guilds and
1:35:11 social impact guilds there are clubs um
1:35:14 Brandon just started a club called AI
1:35:16 life hacks with Brandon and Claire um Dr
1:35:18 Jay in here uh and Brandon have started
1:35:21 that club so you can go share little
1:35:23 life hacks the AI learning lab
1:35:25 Irregulars we have our own club so you
1:35:27 can go play in there so go check out the
1:35:30 AI salon and join it it doesn't cost
1:35:33 anything um just go join it and have a
1:35:35 good time Jump in be generous all right
1:35:39 so that's that okay let's go talk about
1:35:42 GPT
1:35:44 4
1:35:46 [Laughter]
1:35:48 mini
1:35:50 okay so if you're at chat GPT
1:35:54 they just launched this thing called gp4
1:35:58 mini which
1:36:01 is I don't know if this classifies as a
1:36:06 small language model you've heard of
1:36:07 llms probably which is a large language
1:36:10 model slms are all the new rage and slms
1:36:14 are smaller more concise models that are
1:36:18 super um highly tuned to be good and GPT
1:36:24 for o mini is one of those
1:36:29 um if you've got access to GPT for
1:36:32 regular
1:36:34 Omni
1:36:40 um you might not need to use gp4 mini
1:36:44 but it's always good to experiment with
1:36:46 it and understand what it
1:36:48 is if you're using the API of of chat
1:36:53 GPT
1:36:55 gp4 Omni
1:36:57 mini is about half the price of or like
1:37:03 60% the price it's much cheaper than GPT
1:37:07 3.5 turbo which is what a lot of uh chat
1:37:12 GPT like applications that that use open
1:37:15 ai's API were using GPT 3.5 turbo and
1:37:21 this is cheaper than that and
1:37:22 significantly better than that that
1:37:26 um it works just the same if you flip
1:37:29 over to gp4 mini you can do things like
1:37:32 you
1:37:34 know um hang on I got to
1:37:37 sneeze do
1:37:43 I yes um hang on gotta sneeze
1:37:50 um write a poem in the style of Eminem
1:37:54 that explains quantum
1:38:00 mechanics so gp4 mini is very very fast
1:38:03 you saw how fast it did that um it's
1:38:07 it's way better than uh GPT 3.5 was yo
1:38:11 check it listen close I'm breaking down
1:38:12 the quantum scene where particles ain't
1:38:14 just tiny de dots they're kind of like a
1:38:17 dream they don't play by the rules they'
1:38:19 be twisting and they turn unfortunate uh
1:38:21 uncertainty is the name so pay attention
1:38:22 watch and learn so you can do all that
1:38:25 um gp4 mini does not have a lot of the
1:38:29 um bells and whistles of gp4
1:38:33 omni like the ability to upload an image
1:38:36 or upload a CSV file or do code
1:38:38 interpreter so think of you you'll use
1:38:41 gp4 mini when you want like a really
1:38:44 quick down and dirty you just want
1:38:46 something fast um use it for that if
1:38:49 it's just simple sort of text generation
1:38:51 or if you're using it to program and
1:38:54 you're using the open AI API use this
1:38:57 one cuz it's super cheap super cheap
1:39:00 it's like I don't
1:39:02 know 20 cents for a million tokens or
1:39:05 something like that super cheap so
1:39:07 anyway does GPT mini automatically come
1:39:10 with the gp4 sub yes it
1:39:13 does
1:39:15 and I don't
1:39:20 recall yeah cuz it's so cheap I assume
1:39:22 it yeah g G pt4 Omni is is now free to
1:39:27 everyone and gp4 mini comes with it as
1:39:31 does the Legacy gp4 model but I I think
1:39:34 if you're paying for free you you just
1:39:38 have the first two gp4 Omni and
1:39:41 mini
1:39:44 um the limitations here's if if you if
1:39:49 you don't pay for chat GPT
1:39:53 what you miss are
1:39:58 um trying to think what the things
1:40:01 are you still get code interpreter but
1:40:04 it's very limited you have a limited
1:40:06 number of uses so it rate limits out
1:40:10 pretty quick so you'll probably get
1:40:11 dumped down to gp4 mini
1:40:14 sooner um you don't get interactive
1:40:17 charts and graphs like if I make a if
1:40:20 I'm in a gp4 Omni and I say
1:40:26 um create a circle chart of the US
1:40:35 economy this should use code interpreter
1:40:38 see where it says
1:40:41 analyzing so if you click on the word
1:40:43 analyzing you can actually watch what
1:40:45 it's doing it's writing and executing
1:40:47 python code to make us a little
1:40:50 interactive
1:40:52 chart of the US economy so there it is
1:40:55 so see how that was animated and then I
1:40:58 can go in and I can actually change out
1:41:02 colors so if I don't like Services being
1:41:05 green I can make
1:41:07 Services uh
1:41:09 red and then I can make agriculture
1:41:12 [Music]
1:41:13 green and I can make industry blue right
1:41:16 and so
1:41:18 there's there's the US economy B sector
1:41:22 so all that interact stuff you don't get
1:41:25 with the with the free version um so
1:41:28 there's there's some subtle differences
1:41:30 the other thing you get by paying for
1:41:32 chat GPT 4 um the 20 bucks a month for
1:41:37 GPT plus oh right it doesn't make images
1:41:39 that's right uh the free version doesn't
1:41:42 make images the paid version does but
1:41:46 the other thing that you don't get if
1:41:48 you don't pay is access to the new stuff
1:41:50 that they're launching so when um real
1:41:55 time voice comes out when sunno comes
1:41:57 out when GPT 5 comes out which are all
1:42:00 pending they're all sort of anytime now
1:42:03 right sometime in the
1:42:04 fall um all of that stuff is going to go
1:42:07 to paid users first so I still think
1:42:10 that paying the 20 bucks a month to open
1:42:12 AI is worth it um but a lot of people
1:42:15 have dropped it and a lot of people have
1:42:16 switched over to claw cuz claude's
1:42:18 really good too so anyway I'm still
1:42:21 paying for
1:42:22 Omni hoping I still get more time in the
1:42:25 app yeah I know oh yeah you also can't
1:42:28 make custom gpts if you're not paying
1:42:30 the 20 bucks a month that's that's
1:42:33 right all right there you go all right
1:42:37 people my voice is getting a little
1:42:38 crunchy good Lord look at the hair I got
1:42:41 a haircut did you see look how nice and
1:42:44 trim huh but boy that's a that's a
1:42:48 that's a pile of a pile of qua right
1:42:53 there
1:42:54 [Laughter]
1:42:57 the best reason to pay the subscription
1:42:59 is the license about how they use the
1:43:00 data oh yeah you can also tell them not
1:43:02 to train on your data I've been really
1:43:04 liking perplexity lately I like
1:43:06 perplexity right now perplexity is
1:43:09 really good if you haven't played with
1:43:10 perplexity Pages you should do
1:43:14 that um Claude with its uh artifacts is
1:43:20 really good really good and then chat
1:43:25 GPT those three for the large language
1:43:27 models are the are the three that I
1:43:28 would I would be messing around with
1:43:30 right now um Gemini can be good there's
1:43:33 some cases Gemini is good like if you
1:43:35 need a really big on fleek yeah my hair
1:43:39 my hair is definitely on
1:43:42 fleek what is that that's like Flock of
1:43:44 Seagulls right Flock of Seagulls
1:43:50 Fleek oh
1:43:52 yeah oh
1:43:59 all
1:43:59 right so go check out the AI Salon if
1:44:03 you haven't done that the salon. a
1:44:04 looking sharp it's it's looking
1:44:08 something um go do that follow my
1:44:12 channel if you want subscribe to my live
1:44:14 if you want to support the channel
1:44:17 um what else was I going to
1:44:20 say um
1:44:23 most important thing you can do I'm here
1:44:26 five nights a week and you'll probably
1:44:30 notice if you're new here that it seems
1:44:32 like there's a lot of people that kind
1:44:33 of know what's going on and they're
1:44:35 answering questions for you and they
1:44:37 mods are pinning things those people
1:44:42 show up here night after night after
1:44:43 night and you're like well that's weird
1:44:47 I agree and I called him out one night
1:44:49 and I said you're weird why are you here
1:44:53 and and the response was we're not weird
1:44:55 we're
1:44:57 irregular so they're the Irregulars so
1:45:00 how you become an irregular is you just
1:45:02 keep showing up
1:45:05 here and if what's in your mind right
1:45:07 now is why would I do that that's
1:45:10 perfect that's exactly where you want to
1:45:12 be that is the right question to ask and
1:45:14 if you still feel oddly compelled to be
1:45:16 here here's what you
1:45:18 get these nightly sessions become kind
1:45:21 of jam sessions kind of independent of
1:45:24 what I'm talking about there's all sorts
1:45:26 of conversations going back and forth
1:45:28 people are on the AI Salon sharing
1:45:30 images answering questions with one
1:45:32 another helping each out here becoming
1:45:36 friends starting projects together what
1:45:39 this channel has become is a reason for
1:45:42 all of us myself included to stay in
1:45:46 this AI conversation because things are
1:45:48 moving fast and then they stagnate and
1:45:51 then they pick up again and then it's
1:45:52 all over the place and everyone's trying
1:45:54 to figure this stuff out and no one zero
1:45:59 people zero people are experts on this
1:46:03 stuff right
1:46:04 now anyone who tells you they're an
1:46:06 expert is full of
1:46:09 [ __ ] like I spend an inordinate amount
1:46:13 of time trying to keep up with this [ __ ]
1:46:16 and I can't like not even
1:46:18 close not even close if I spent 247 I
1:46:22 couldn't scratch the surface of it and
1:46:25 so what happens when you get in a group
1:46:26 like this is Silver Fox will pay
1:46:29 attention to this thing and Vicki will
1:46:31 pay attention to that thing and Suzanne
1:46:32 will pay attention to that thing and
1:46:34 everyone's sort of doing their own thing
1:46:36 and then when they talk about it you're
1:46:37 like oh that's a really good way of
1:46:39 approaching that or I'll I'll say
1:46:41 something and they're like o don't
1:46:42 forget about this thing CU they're
1:46:44 thinking about it from some other
1:46:47 direction so I think being in a
1:46:49 community is a superpower and and doing
1:46:51 it on a regular B basis and just being
1:46:54 in the conversation playing with these
1:46:56 tools building [ __ ] with these tools
1:46:58 talking about what you're
1:47:00 learning you will establish yourself as
1:47:03 a
1:47:04 leader you might be like but I'm not a
1:47:07 leader I hate leading I most of the
1:47:10 people in here are some version of
1:47:12 neuros
1:47:13 spicy that don't
1:47:17 naturally lean into being a
1:47:21 leader but what's happening is
1:47:24 as they're getting educated with this
1:47:25 stuff and putting themselves in the
1:47:27 world just saying well I tried this
1:47:29 experiment with AI and it went kind of
1:47:31 okay someone else goes oh you're an
1:47:33 expert can you come talk to my group uh
1:47:36 okay I really don't know what I'm doing
1:47:38 but
1:47:39 okay and then before you know it they've
1:47:41 started an agency or they got promoted
1:47:43 to the AI Council in their company or
1:47:46 whatever it is it's remarkable so just
1:47:48 come back here every
1:47:50 night so with that thank you Irregulars
1:47:52 for being a nightly inspiration for me
1:47:55 Joker is an expert there's one expert in
1:47:57 here sorry Joker is an expert I stand
1:48:00 corrected so not this many people are
1:48:02 experts this many people are
1:48:05 experts and Thy Thy Name AR Arn
1:48:11 Joker
1:48:13 um thank you mods or maybe just I'm full
1:48:17 of [ __ ] same thing
1:48:21 exactly and then the last thing we did
1:48:23 here when I get out of here every night
1:48:24 we do a if you're old and crusty like me
1:48:28 you remember Romper Room you remember
1:48:29 magic
1:48:31 mirror and I see papa JJ who's an
1:48:36 expert Amelio's wife thank you as always
1:48:39 Silver Fox great to see you steo Joey
1:48:43 Jodi ARS good to see you it's been a
1:48:45 while Allan good to see you Tams against
1:48:48 the norm got all the hits in here
1:48:50 Brandon I am here to help very nice
1:48:53 awesome Silver Fox what is that icon or
1:48:57 whatever that was that was cool Emoji
1:48:59 Tik Tock live games can someone explain
1:49:01 what privacy of using your data if you
1:49:04 paid 20 bucks for chat GPT in your
1:49:07 settings in chat GPT if you paid your 20
1:49:10 bucks you can flip a switch that says
1:49:12 Don't train on my
1:49:14 data be careful if you flip that switch
1:49:18 it deletes your
1:49:19 history so before you flip the privacy
1:49:23 switch if you want to do
1:49:25 that there's an option to export all of
1:49:28 your data export all of your data so you
1:49:31 have your prompt history somewhere else
1:49:33 and then you will not be pissed that you
1:49:35 deleted your prompt history so um cam
1:49:38 cin always a great time thank you Kyle
1:49:40 you're welcome cam great to see you glad
1:49:43 glad to have you here tonight
1:49:45 Tobias Andrea good to see you Andre I
1:49:48 haven't seen you in a while welcome back
1:49:50 we had Winston in here tonight Daisy May
1:49:53 new lots of lots of old oldtimers newly
1:49:57 back Bob Hamilton great to see you Wolf
1:50:00 Man Clint Dr J in the house as
1:50:06 always spin
1:50:08 B3 Tik Tock live games Winston good
1:50:12 night good night good night Welcome Back
1:50:14 Sir Judy
1:50:15 ARS Joe
1:50:18 we Joker against the norm steo or right
1:50:23 I think I am caught up another story
1:50:26 I've become proficient in small useful
1:50:29 use case for my business that's perfect
1:50:31 the way to do it you're a tab hoarder
1:50:33 I'm a GPT memory hoarder can't decide
1:50:36 which memories to keep well open AI is
1:50:39 solved that for you because there's only
1:50:40 2,000 tokens you got about 1,500 words
1:50:42 Brandon so you can't have that much
1:50:45 so it's hard to hoard when it when your
1:50:48 space is that small Cory Sandler's in
1:50:50 the
1:50:51 house um
1:50:57 all right cool um fantastic everyone
1:51:01 like I said I'm not going to Washington
1:51:03 DC so I will be back tomorrow I should
1:51:06 be on time 8:00 and uh yeah that's that
1:51:11 one of the best AI profiles thank you
1:51:14 very much I appreciate that go play with
1:51:17 my digital twin
1:51:20 too it doesn't swear like you I got to
1:51:24 make I'm building a rag nice wouldn't
1:51:27 you be weaving a rag
1:51:30 Joker I don't think you build a rag you
1:51:33 weave a
1:51:36 rag 30 30,000 likes yall are nuts and I
1:51:40 love it um peace out everybody have a
1:51:42 good night bye