12/12/2024 - Overcoming AI Fear: A Framework for Embracing the Potential

Description

Chapters

Transcript

0:00 it's time to it's time to sing get over
0:03 here sit your butt down come on
0:07 [Music]
0:16 [Music]
0:30 you still got bad breath
0:33 buddy
0:35 [Music]
0:50 sit you're feeling nervous aren you
0:55 boy with a gentle
0:58 voice and impeccable Style
1:04 don't ever let them steal your
1:07 joy any gentle ways no and you something
1:12 ways to keep them from running
1:16 wild they can kick dirt in your face
1:20 dress you down tell you that your place
1:22 is in a middle when they hate the way
1:25 you shine
1:30 I see you tugging on your
1:33 shirt trying to hide inside of it i'
1:38 have much it
1:40 hurts let
1:43 them while
1:46 they let them
1:49 spin let them
1:52 scatter the
1:54 wind I've been to the movies I've seen
1:59 how
2:02 and the Jokes wait and the
2:10 jokes are you not singing with
2:13 me is this some sort of Union protest
2:16 what is going
2:18 [Music]
2:21 on oh man dogs they say never work with
2:25 kids never work with animals this is why
2:31 you get discouraged don't you
2:34 girl it's your brother's
2:37 world for a while
2:40 [Music]
2:43 longer we got
2:45 a Dance with the Devil on the river to
2:48 beat the
2:50 dream call a living a
2:53 dream call her kicking the
2:57 ladder they come to kick it in your face
3:01 to call you weak and then displace you
3:04 after Carrying Your Baby on your back
3:06 across the
3:10 desert I see your eyes behind your
3:14 head and you're looking
3:17 tired but you don't look
3:21 scared we let them
3:24 lay while they
3:27 can let them spin
3:31 let them scatter in the
3:35 wind I've been to the movies I've seen
3:39 how
3:42 end and
3:43 the on CH
3:49 [Music]
4:00 good evening good people hello on the
4:03 Tik Tok hello on the YouTube if you're
4:05 on LinkedIn welcome let's see what we
4:08 got here Joseph William Banker big news
4:10 kleene a VSS Cloud plug-in code editing
4:14 tool is sporting mCP tool creation and
4:21 use
4:23 okay took me a minute I put the pieces
4:26 together so mCP tool creation and use is
4:31 part of claud's new you can use tools
4:35 API and so it looks like a VSS clode a
4:39 vs code
4:41 plugin can now easily use
4:45 tools things things things if if you're
4:50 not if you're not paying attention well
4:53 if you're here you're paying attention
4:54 but if you're new here welcome my name
4:56 is Kyle Shannon this is the AI learning
4:58 lab we occasionally have dogs that sing
5:01 tonight not so much
5:04 um one of the things that you'll learn
5:07 about me is that uh I am wholly 1,000%
5:11 unqualified to be here but I do have a
5:16 button and so whenever I get stuck I
5:20 just pull out the button they some call
5:22 it the easy button but I you know I'm
5:25 not a shill for for you know corporate
5:27 weenies
5:30 [Music]
5:37 I also have a pink bow that I
5:39 occasionally wear and you might be
5:41 thinking but Kyle what's that got to do
5:43 with
5:47 AI
5:49 nothing
5:55 um so much going on today
6:08 so day six of open AI ship ship
6:12 shipus it's hard not to say shipus you
6:15 know have you ever tried to not say [ __ ]
6:17 miss when you're trying to say ship Miss
6:21 it's not
6:23 easy this Gen X mouth does not conform
6:26 to not swearing very easily
6:31 a golden button it is it is but a golden
6:35 button amongst us we share the golden
6:38 button the wisdom of the golden button
6:41 you can make
6:43 time came my ruin shipers for
6:46 [Music]
6:49 me oh man Peter Spenard says hello good
6:54 day sir Sharon
6:56 Crawford anyone else having streaming
6:58 lags uh yes so here's the deal for
7:04 tonight
7:05 apparently there's some big video game
7:08 award show that my two little weenies of
7:12 sons are like we gotta watch the award
7:14 show
7:16 Dad so I've got I've got two boys
7:20 streaming plus me streaming out so I
7:24 think it's going to suck
7:26 [Music]
7:27 tonight advanced voice committed
7:30 complimented me on my glasses oh that's
7:32 cool SE slug of Doom I don't have it yet
7:35 I have I have yet to get it um I do have
7:37 teams which they said teams was going to
7:39 get it today it is technically still
7:42 today but not for much longer
7:47 [Music]
8:06 wait let me get this straight you're
8:07 laying down and
8:09 singing you you know that it's not good
8:11 for your diaphragm
8:14 right
8:16 I you know we got the dog singing
8:19 lessons and he just he doesn't for one
8:22 thing he refuses to say
8:26 words and I think he's just thinking
8:28 like well AI is going to do it so why
8:30 should I learn your stupid language why
8:32 don't you learn mine so apparently
8:34 that's what we're going to have to do
8:35 going to talk to him you're out of
8:38 champy credits subscribe to champy Pro
8:41 to continue that is totally what's going
8:43 on I like that
8:48 one oh man oh good uh Jim Ross of of uh
8:55 Three Mile
8:57 Storage he got he got 73 three free
9:02 mascots for his business last night um I
9:05 produced 72 of them and those were all
9:08 unusable and then steo produced one and
9:11 it was perfect so am I bitter you
9:16 beta it's just my nature but congrats to
9:19 steo for doing good work there that was
9:21 solid I was overthinking it he probably
9:24 typed in a prompt like make me a mascot
9:27 about a storage business something you
9:29 know like named Three Mile Storage right
9:32 and then it it was perfect and I you
9:35 know came in there with my thoughts and
9:38 it it it just didn't go well for me steo
9:43 has the contract apparently steo and
9:45 three M storage are now Partners
9:48 strategic Partners in the mascot
9:51 business
9:53 [Music]
10:06 Call Me Al listen you Aussies you
10:09 Aussies can't support one another like
10:11 that that's not
10:13 fair come on I don't talk like
10:18 that you Aussies you're you're like oh
10:21 good job yeah wait a wait a [ __ ] all
10:23 over the Americans that it's just you
10:26 know what it is it's
10:28 country that is very
10:30 [Music]
10:35 countriest call me out good day Kyle
10:40 Aussie Aussie's for the
10:43 win get a got got I well [ __ ] with your
10:49 [Music]
10:57 mate you may wonder why do I just getar
10:59 and ramble like nothing is
11:03 happening good question I don't have a
11:05 good answer for
11:07 you I I mean I like to play
11:10 guitar um I like to try to
11:16 [Music]
11:25 sing well 10,000 words SW around my head
11:30 10 million moreing books WR beneath my
11:34 [Music]
11:37 bed and I wrote or read them all
11:40 searching in the Swarms still can't find
11:43 how to hold my
11:48 hands and I know you need me in the next
11:53 room
11:55 over I'm stuck in here all paralyzed
11:59 [Music]
12:04 four months I got myself in ruts too
12:06 much time spinning mirrors framed in
12:10 yellow
12:11 [Music]
12:15 walls a it like most people are no
12:18 different like talk on things we don't
12:21 know
12:26 about ain't it like most people I know
12:30 like talk on things we don't know
12:36 about yeah shitty dog shitty dog not
12:38 singing I don't know what to do I don't
12:41 know what to do he's not getting you
12:42 know you know what he's not getting
12:43 tonight [ __ ] chicken sticks that's
12:45 right out cheese right
12:48 out you got to earn your cheese champ
12:51 you got to earn your
12:53 cheese he's just wagging his
12:56 tail um so open AI today announced um
13:02 but they are apparently shipping it they
13:05 announced um that advanced voice will
13:08 now have video that's a part of it it
13:11 looks quite
13:13 remarkable the little example that they
13:15 showed was they they sort of went around
13:17 the table and had people introduce
13:21 themselves and then asked voice you know
13:23 who was the person wearing the reindeer
13:25 hat and who was the person in the red
13:27 sweater and it remembered it was pretty
13:30 cool I think I think it's going to be
13:32 quite quite
13:34 remarkable I saw
13:37 your 11 video today oh yeah when I when
13:40 I went a live when I went live for
13:44 that and then I I
13:47 um I talked last night or the night
13:54 before I've been thinking a lot so okay
13:57 so here
13:59 I'm trying to anticipate something that
14:01 I think is coming so here's what's
14:05 coming the tools the AI tools just
14:09 across the board in every domain from
14:12 video to image to to music to voice
14:16 synthesis to avatars to large language
14:19 models to coding tools all across the
14:22 board the tools are getting better and
14:25 better and better and
14:27 better and we still have
14:31 a rather Hefty
14:33 population and I don't mean Hefty like
14:36 my kind of Hefty you
14:40 know that's
14:43 husky we have a rather Hefty population
14:46 of people that are either ignoring
14:49 AI terrified of AI or just [ __ ]
14:54 raging at
14:56 AI
14:57 um and
15:00 the challenge that I have the thing that
15:04 I'm passionate about is getting as many
15:06 people as possible to try
15:07 it um and people are not going to try it
15:11 if they're pissed off at it and if
15:12 they're terrified of it and so I finally
15:15 had a breakthrough about about how to
15:18 reframe thinking about Ai and how you
15:21 might use it even if you're afraid of it
15:23 even if you're pissed off at
15:25 it
15:27 um so I've got that presentation I gave
15:30 that pre presentation spontaneously this
15:32 afternoon just so I could record I just
15:34 wanted to get it recorded but I think
15:36 I'll go through it again tonight because
15:38 i' I'd love to get your feedback on it
15:40 and see if it if it rings true and if it
15:41 feels like it's missing anything but I
15:44 think I think there's something there um
15:47 for mods on Tik Tok actually for mods
15:50 anywhere the mod channel on the salon I
15:53 have my iPad back up so if you want if
15:56 you want to ping me for
15:58 anything Peter Sven guard or Sven kurd
16:03 kude Sven kude Peter Sven
16:07 kude I just signed up for Sora it was
16:10 too busy yesterday thus closed for New
16:12 signups I was able to log in with my
16:15 chat GPT Plus account yeah I noticed I
16:18 did one video today and it seemed like
16:19 it rendered it faster so it looks like
16:21 they've up their uh their server
16:24 allocations um by the way the the two or
16:27 thre hour um
16:30 all of open AI crashing yesterday for
16:32 two or three
16:36 hours it was a configuration error
16:39 somebody put a
16:41 configuration in some upgrade they did
16:45 and it you know I think it basically
16:47 just um shut off access to their
16:50 servers it probably took him while to
16:52 find
16:54 [Music]
16:56 it I don't mean to laugh because as as
17:00 as you know one of the people that that
17:02 did that that's probably got to feel
17:04 like [ __ ]
17:06 but sometimes [ __ ] just happens we had
17:10 something happen a month or two ago at
17:12 Story vine that was a similar sort
17:15 of it was there was a lot more drama
17:18 than there needed to be let's put it
17:20 that way hey an Murphy what's happening
17:22 so Anne Murphy's in the house let me
17:23 share something
17:25 here um let me go find it first
17:33 where is it here it is let me reload
17:37 this let me go over here yes
17:43 so
17:45 um if you've been around here for a
17:48 while last year and Murphy and I did
17:51 this thing called uh GPT for good was a
17:55 24hour GP Aon where we just had people
17:58 make G gpts for nonprofits this
18:01 community and the AI Salon Community
18:04 came together and built more than 150
18:07 gpts for nonprofits in 24
18:10 hours this year we thought to hell with
18:13 the nonprofits this is about us right
18:16 everybody's got to learn this stuff now
18:18 I'm just kidding of course it's for the
18:20 nonprofits that's the business that
18:22 Murphy Works in that's that's why I'm
18:25 busting on
18:27 her she's like I I've got investors in
18:30 here you can't say those things um and
18:34 what's funny is Ann doesn't even talk
18:35 like that but you know what she does
18:36 talk like you
18:38 make the weirdest thing you see you know
18:41 beautiful long hair and then that's what
18:44 she sounds like it's like really anyway
18:47 that's neither here NE
18:49 there December 27th and 28th the Friday
18:52 and Saturday in between Christmas and
18:54 New Year's we're doing AI Festivus 2024
18:58 so it's basically we start it I think
19:00 it's like 900 in the morning Pacific
19:01 time and go until early evening all day
19:05 Friday all day Saturday
19:08 doing okay oh yeah and the website's not
19:11 really done exactly we need our pictures
19:14 on here and don't you think don't you
19:16 think it should just be I think it
19:17 should be like just giant pictures of
19:20 you and me and then a teeny little mouse
19:22 type oh and if you want to come learn
19:25 [ __ ] you can do that too but whatever we
19:27 can work with the designers on that um
19:30 so the site's not quite finished yet but
19:33 um reserve the the dates uh December
19:36 27th Friday December 28th Saturday from
19:41 morning till evening every hour is a
19:45 different instructor from the AI Salon
19:47 from uh she leads AI from just cool
19:49 people that we know coming in and
19:51 teaching you all sorts of cool stuff all
19:53 right so mark it down be there we I know
19:58 you're going to be with your family I
20:00 know that weekend in between Christmas
20:04 and New Year's you're going to
20:06 absolutely be fed up with your family
20:08 you're gonna need an
20:10 excuse to just say Aunt Martha listen
20:14 yes I no I I I do want to hear about the
20:17 insulin pump it's that's really exciting
20:20 um I I have a work thing and I know I
20:22 shouldn't do it when I'm on holiday but
20:24 I'm just going to have to lock myself in
20:25 the room and do this work I can't be
20:27 interrupted right so we can buy you
20:30 three four five hours of peace from the
20:33 family just lock the door and come hang
20:35 out with us you'll learn something it'll
20:37 be
20:38 fantastic all
20:40 right fantastic Bob tell him what he's
20:43 won uh well he hasn't won anything it's
20:46 not a game
20:48 show oh well I like that line well
20:52 that's a parent you say it all the time
20:57 [Music]
20:59 oversharing is a
21:02 struggle like 20 hours to hide for your
21:04 family you're welcome exactly exactly
21:07 it's like last year it was for the
21:09 nonprofits this year we're giving a gift
21:12 to you we're g to teach you things and
21:15 give you the opportunity for some peace
21:17 in your life I think that's good oh by
21:21 the way if you live in a functional
21:23 family where everyone loves each other
21:25 and you are all raised
21:27 nicely really
21:30 really really you're not
21:33 invited the dysfunctional people will
21:36 will be doing festiv us those of you
21:39 with those those 0.01% of you that grw
21:43 up in nice families just do you just
21:46 like walk around gloating all the time
21:49 like yeah well you know I was loved I
21:52 was
21:53 acknowledged they're proud of me really
21:56 really all right anyway all right you
22:00 know but anyway we're here we're here
22:02 for the rest of
22:05 you because if you've got families like
22:07 that then you deserve them probably
22:10 because you're well adjusted and you
22:12 actually do deserve them my family
22:15 pretends to be functional I love that my
22:18 f visitors say I wish my family was like
22:21 yours H got you all fooled exactly
22:24 that's a good one Tom those are the best
22:27 oh the families that look look like
22:29 they're
22:30 functional when they when they snap they
22:32 snap they snap
22:35 hard come on now most of us a Gen X
22:38 exactly exactly Gen X we didn't grow up
22:42 with functional families it was built
22:43 into the it was built into the
22:46 [Laughter]
22:50 framework Christmas cards of
22:53 [Music]
22:55 course oh man all right
22:59 [Music]
23:04 Kyle you're freaking me out your hair is
23:06 too I'm I'm in order well here we can
23:11 it'll it'll it'll snap I just I just wet
23:14 it a little I've got product in it so it
23:17 kind of shapes like I do it all
23:20 right fantastic Bob oh yeah Tik Tock cam
23:23 sorry about
23:25 that there's the website oh the website
23:28 by the way is is it on screen uh aif
23:31 festiv
23:32 [Music]
23:46 us.com oh good lord good
23:50 [Music]
23:55 people want something to cry about sure
23:58 Jason give me something to cry
24:04 [Music]
24:05 about Joker I dispose of my family for
24:10 using the word
24:12 [Laughter]
24:19 [Music]
24:22 Festivus fear loathing and unadulterated
24:24 loathing yeah so by the way here's my
24:27 I'm very happy with my my art work for
24:29 my fear and loathing and
24:32 AI
24:37 [Music]
24:48 stuff all right so here's what I'd like
24:52 to
24:53 do I want to walk you through this this
24:56 new framework I put together because I'm
24:59 actually pretty excited about it and I'm
25:01 I'm curious I'm curious like where and
25:03 how to share it
25:05 and if I should share it if it's got
25:07 value I think it's got value though so
25:09 here's here's here's my request to
25:12 you I'm g walk you through it it'll take
25:15 us about 10 minutes it's only three
25:17 slides
25:20 um i' I'd like you to think about this
25:22 in a in in a couple of different ways if
25:26 if you've got a lot of fear about a AI B
25:28 basically
25:29 if if fear of AI is overriding your
25:33 willingness to play with it to try it to
25:36 explore and and jump in definitely pay
25:39 attention if you know someone like that
25:41 if you know people that are just pissed
25:43 off about it they're just every time you
25:45 mention it they're just
25:47 like you to hear about AI anymore shut
25:50 up about AI you know that face if you if
25:52 you know people that have that
25:55 face um or they're angry about it it's
25:58 that it's the world's greatest largest
26:00 plagarism machine braah BR where they're
26:03 they're
26:04 using all of those very
26:08 valid fears and concerns and
26:12 rage as an excuse not to deal with it
26:16 and that's what that's what what what
26:18 this particular thing is about is
26:23 is is a framework to say can we
26:27 acknowledge Their Fear acknowledge
26:30 their
26:33 um their anger if they've got it and
26:36 then still say you know this could still
26:39 help 10 minutes of Kyle time is actually
26:42 45 minutes of real time yeah so wait
26:45 what time is it now all right so it's
26:47 8:25 to Tom's point if I finish this
26:50 before nine o'clock that's 10 minutes
26:54 right yeah that that math works out I
26:57 checked it on chat GPT
26:59 [Laughter]
27:01 gently smash the Thumbs Up Button yeah
27:03 if you're on YouTube give me thumbs up
27:05 if you're on LinkedIn um send someone a
27:08 sales
27:10 DM oh oh what would be good if you're on
27:13 LinkedIn send someone a sales DM
27:16 particularly if what you do is something
27:18 that they definitely don't
27:20 need all right but what would be good to
27:24 do is copy and paste their LinkedIn
27:26 profile use chat PT to um personalize
27:31 your email and tell them how much you
27:34 really like whatever their company does
27:37 and then offer them services that that
27:40 are clearly not something they need all
27:42 right so on LinkedIn do that uh if
27:45 you're on X just rage tweet just rage
27:47 tweet the whole time I do this all right
27:50 Perfect all right we're what we're
27:51 looking for here is engagement we're
27:53 just looking for engagement right want
27:55 to keep the ecosystems breathing um Tik
27:58 tock yall are
28:03 [Music]
28:08 swell all right alrighty um Fear and
28:12 Loathing in in AI all right so here's
28:14 here's the
28:17 deal uh yeah and please subscribe to
28:19 YouTube I'm trying to get to a thousand
28:21 uh subscribers on YouTube and at the
28:23 rate I'm going I should get there by
28:26 2027 we will we we will have super
28:29 intelligence before I get to a thousand
28:31 subscribers on YouTube all right so if
28:33 you could just subscribe there that'd be
28:37 awesome it don't work the way Tik Tok
28:39 works let me tell you that okay
28:45 so here's how this here here's how the
28:48 world works in the oh and let me give
28:50 you a caveat um what I'm going to talk
28:53 about is your relationship to AI but not
28:58 your relationship to other people this
29:00 is this is a very narrow lens we're
29:02 looking through right now what I'm what
29:04 I'm trying to get to is if someone has a
29:06 personal of front with AI how can I sort
29:09 of crack through that and and I didn't
29:11 want to over complicate it so this is
29:12 quite simple so we'll start with youu
29:14 and you is going to be represented by
29:17 two little halves of a circle right the
29:19 top half and bottom
29:21 half
29:23 and this is kind of the whole of you as
29:26 it relates to AI it again it's it's not
29:29 in relationship to teams colleagues any
29:31 of that sort of stuff just your your
29:32 relationship to what you do your
29:36 aspirations and and your work in the
29:39 world so the top half is the passionate
29:44 you the you that's like this is what I
29:46 love this is my passion this is the
29:48 thing I want to do for all my life this
29:50 is the the thing I have to do it might
29:53 be quilting it might be pizza making it
29:55 might be fixing cars it might be
29:58 whatever ever acting movie
30:00 making the bottom half is the prosperity
30:04 part right that's the money-making part
30:06 I've got a job I do this I do
30:09 that if
30:11 you're if you're remarkably like
30:14 remarkably fortunate in your life those
30:17 two things are actually the same thing
30:20 right there are people in this world few
30:22 and far
30:25 between that are making money doing a
30:27 thing they love I think entrepreneurs
30:29 fall into this a lot because they're
30:30 just like a I'm unhirable and B I'm so
30:33 passionate about this thing I'm going to
30:34 go figure out a way to make money at
30:36 it but those things could be separate
30:39 but they're very very personal right
30:41 they're you like you know what they
30:43 are and then Along Comes
30:46 Ai and AI initially you didn't have to
30:50 worry about it because it was it was
30:53 very specialized and you had to have a
30:54 mathematics degree to do anything about
30:56 it
30:59 and then chat GPT comes out and all of a
31:02 sudden it's like oh I guess anyone can
31:04 use this now and then it seems to be
31:07 getting better and better and
31:09 better and for 50 years Hollywood's been
31:12 telling us that that the robots are
31:15 going to kill us and all the economists
31:18 are telling us that the AI is going to
31:20 take our jobs and so
31:23 if if you're just sitting on the
31:26 outside listening to all this
31:28 essentially what it feels like is AI as
31:31 it gets better and better and better is
31:33 going to do this
31:35 right it's going to destroy
31:38 you you know I did that I did that
31:40 hip-hop video the Eminem hipop video and
31:43 a lot of the comments on there are like
31:45 rap's dead rappers are dead so much for
31:47 Ghost Riders this is [ __ ] this is
31:51 evil right because they're they're
31:53 looking at this as a self-expression
31:55 form potentially as a Prosperity part of
31:58 their
31:59 identity and they see something like
32:01 that AI song and they're like AI evil AI
32:05 is going to take me out AI bad
32:08 right so here's the Insight that I
32:16 had the self-expression part of you the
32:19 part that wants to put something in the
32:22 world it all starts from this internal
32:24 thing whatever that fire is is in there
32:27 and then you choose an outlet right you
32:29 choose I'm going to make a quilt I'm
32:31 going to make a painting I'm going to
32:32 make a sculpture I'm going to make a
32:34 computer software I'm going to make a
32:36 whatever and you've got this thing that
32:38 you build or these things that you build
32:41 you put them in the
32:42 world those things are actually one step
32:45 removed from you you might have very
32:47 personal attachment to them right you
32:49 might love your paintings and love your
32:51 sculpture and love the sandwiches you
32:53 make but they're not you they're a step
32:56 removed from you and then a Step Beyond
32:59 that oops A Step Beyond that is the
33:02 creative
33:03 process right so this is I've got to go
33:06 to Michaels to get paint and I've gotta
33:09 I've got to organize my files and I got
33:11 to clean up the studio and I've gotta
33:13 right and I I got to put all the stuff
33:15 together and that's that's even further
33:17 removed from you so so even if your
33:19 passion is
33:21 illustration you know maybe buying pens
33:24 at the art store drives you crazy and
33:27 you want to be able to make that easier
33:29 or something like that but it's it's
33:30 further removed from who you are what
33:32 that core value is same on the
33:35 prosperity side right you've got okay I
33:38 need to make money and you come up with
33:39 ways to make money and then and then the
33:42 business product is the maybe you're a
33:45 salesperson so it's it's the selling
33:47 process or maybe you've got an
33:49 e-commerce site and it's setting your
33:51 you know your uh your e-commerce site up
33:54 and putting the business product out
33:55 there that your customers actually
33:57 interact with and then sitting outside
33:59 that is the business
34:01 process and that's again further a step
34:04 further removed from you so there's all
34:07 these different layers and and one of
34:10 the things that led me to this
34:13 diagram
34:15 was when I would have an opportunity to
34:18 talk to someone that was really
34:19 passionately anti-
34:22 AI passionately it's evil
34:28 well would you ever use it to like help
34:30 you write an email oh yeah I use it for
34:32 that all the time oh okay so AI is not
34:36 evil it's just evil when it's when when
34:40 when you're assigning AI is going to
34:42 replace that part of me that I value
34:44 then it's evil right then it's like ah
34:47 okay great so so you can actually
34:49 distinguish between different activities
34:51 and different
34:53 tasks it's the whole purpose of this
34:55 chart is to say okay so now that we've
34:57 got AI
34:58 rather than just looking at yourself as
35:00 this
35:02 monolithic
35:03 blob and looking at AI as this
35:05 monolithic blob that's going to you know
35:08 subsume
35:09 you what if you look at yourself in a
35:12 bit more of a sophisticated
35:14 way and then you can say you can
35:18 choose which part of all of this that
35:23 you want to use AI for for maybe there
35:25 are some things that are sacred maybe
35:27 the creative product and your ideation
35:31 and all of that maybe that's the Sacred
35:33 Space but creative process where you're
35:35 organizing things business product
35:39 business process may maybe that stuff
35:40 you're all like yeah I'm I'm happy to
35:42 use AI on that okay great so so now
35:44 you've got some
35:46 boundaries and then the next thing
35:48 becomes well so how do you learn this
35:50 stuff and this this goes straight back
35:52 to the AI salon and how you learn this
35:54 stuff is
35:55 this you play
35:59 and you play with AI completely outside
36:02 of all of this stuff right if all of
36:05 this stuff is who you are and how you
36:07 make money learning AI can live 100%
36:11 outside of that just play just go make a
36:15 kids book go make a whatever a a design
36:20 a garden even even even if maybe you
36:24 have a friend that designs go design a
36:26 garden for a friend just play just play
36:28 play play all these tools are so
36:31 good and what you'll find in playing is
36:34 that you will
36:39 learn you will learn a thing or
36:42 two that will sort of Pierce the armor
36:45 of this protective
36:47 sphere and you're like oh that actually
36:50 probably would be cool to write an email
36:51 like maybe I could do personalized
36:53 emails a lot faster yeah right now I'm
36:56 spending two or three hours a week
36:57 writing emails what if I could make them
37:00 more personalized and only spend 15
37:02 minutes huh right and then you let you
37:06 let you know a little bit of AI into
37:09 business process right it's it's still
37:11 not who you are but
37:13 right and that's it this is basically it
37:16 this is this is all I got so I'm curious
37:18 what are your thoughts does make sense
37:20 does this track does this resonate
37:30 figuring out how to articulate that it's
37:34 been like a month in the
37:37 making it's like
37:39 God thoughts good
37:43 bad Watching You on YouTube instead of
37:45 Tik Tok very cool play is where I learn
37:48 Tik Tock
37:49 pin can we have a sticker of this sure I
37:53 will
37:56 [Music]
38:04 I will put this in the
38:09 AI I mean it doesn't I guess I should
38:12 have a version of this that's got some
38:14 sort of instructions on it like what the
38:15 what the hell this thing even means but
38:18 if you just want something to put
38:21 somewhere we can do
38:24 this needs more llamas yeah I know it's
38:27 very Cor and very like undesigned I just
38:30 I just did something quick if someone
38:32 wants to make a pretty version of it
38:34 you're you're welcome to do that all
38:35 right I'm gonna where am I gonna put
38:37 this um we don't have a future of
38:42 work I think I'll do it in hangout and
38:44 chat so where what you're looking at
38:46 right now is I'm in the AI salon so if
38:49 you are not part of the AI Salon go
38:53 to go to this address the salon. A and
38:57 click on join our community there's like
38:59 a little seven-step welcome process step
39:01 six is actually like a water cooler chat
39:04 where you can just hang out and chat and
39:06 I'm just going to throw this in here uh
39:08 I think there should be a discussion
39:11 thread and I'll just say
39:14 here is a way
39:19 um here it's a framework
39:22 framework for how to think about
39:28 using AI in a way that
39:34 retains
39:37 your I don't know
39:40 sacred center it's very
39:44 Bolder um while using AI uh for less
39:53 valuable
39:55 tasks I mean the the thing about that
39:58 the thing about this for me is like like
40:01 you can go as deep as you want to go in
40:04 my case I want AI all the way down to
40:06 the center because for
40:09 me generative AI tools are absolutely a
40:13 form of self-expression right and I
40:15 think they're also absolutely um a
40:18 potential for you
40:20 make for Revenue right so I want I want
40:24 AI I want those play Arrows to come from
40:27 those play play things all the way into
40:29 the center other people are going to
40:30 protect areas and that's that's the
40:32 whole point of this all right so I'll
40:34 just put that for there um if anyone
40:37 wants to make this useful for more
40:39 people you could just you could in
40:41 comments describe what the hell this
40:43 thing actually means um but let me copy
40:46 this and I'll also share it in the um
40:49 Irregulars so we'll go to Irregulars
40:53 Irregulars and we'll just pop that right
40:55 in there
40:58 there you
41:00 go and also feel free to use this if
41:03 there's if there's people in your life
41:05 that are like hate AI you know have this
41:08 conversation with them because I think
41:10 there's something there all right love
41:13 it because so many artists are bothered
41:14 by AI exactly and that's that's the
41:16 thing is like when I've dug down
41:18 particularly with
41:22 creatives um let's see are people really
41:26 afraid of losing their jobs are they
41:28 afraid of not having food well I think
41:31 in today's society those are
41:34 synonymous
41:36 um people are certainly afraid of their
41:40 livelihood being being threatened and
41:43 you know quite
41:45 frankly the I think the the the meme
41:50 that that Paul rates are um kind of
41:55 popularized of an AI is not going to
41:57 take your job but someone using AI is
41:59 going to take your job um even Paul rer
42:02 has said that that that quote is is not
42:05 aging well because the AIS are getting
42:08 capable enough and once we get uh
42:10 autonomous agents um they are going to
42:13 start taking jobs so people's
42:14 livelihoods are at risk right if you
42:17 look at what's going on in Hollywood if
42:19 you look at what's going on with writers
42:20 if you look at what's going on in ad
42:23 agencies there's already a lot of
42:24 disruption it's just not it's not on the
42:27 surface yet I have a feeling in 2025
42:29 it's going to Bubble more to the surface
42:31 so I think that I think that this rage
42:33 and anger and
42:35 fear is we're going to see more of it
42:38 and then I think we're going to see more
42:40 people going but I need to deal with
42:42 this how do I deal how do I reconcile
42:44 these two terrifying things and I think
42:47 by just you know separating out your
42:51 tasks into sacred and non-sacred peace
42:55 Parts might be a might be a way all
42:58 right let me jump over here to the
43:02 YouTube uh I don't see a spotlight
43:06 there which one are you talking
43:09 about tell not the right but the actual
43:12 threat is that organizations will try to
43:14 do jobs with AI in order to retain The
43:16 Profit that would have been wages yes
43:19 they will and Mr meanor so not to not
43:25 to I'm not going to let you off off the
43:27 hook on just taking one side of the coin
43:30 because I think there's another side to
43:32 this coin yes capitalist corporations
43:35 are going to be capitalist corporations
43:38 and particularly PE private equity-owned
43:41 corporations are going to be [ __ ]
43:43 ruthless and it's going to be depressing
43:45 and it's going to be gross I agree with
43:47 you
43:52 and one of the risks of running
43:56 companies like that particularly with
43:59 with technology that's moving as fast as
44:00 AI
44:02 is is that the people that you
44:05 displace the ones that Embrace
44:08 Ai and get really good at it are going
44:11 to be able to come back and compete with
44:13 your company in ways you couldn't even
44:17 imagine so I have a bet going right now
44:20 with Eric marier up in in Boulder that I
44:23 think a all AI 10p person gen
44:29 10 badass gen xers that are all in on AI
44:32 doing an agency are going to displace
44:35 like a a one or two hundred person
44:37 agency um in the next three
44:39 years and I and and I think that I think
44:42 that kind of competition is gonna it's
44:44 gonna come from everywhere it's it's
44:46 gonna be more disruptive than the drones
44:49 in New Jersey right
44:50 now because companies are not going to
44:53 know which way the competition's coming
44:56 from so
44:58 so while what you say is likely true and
45:01 those will be good headlines that's
45:03 still not a [ __ ] reason not to use AI
45:06 it's more of a reason to use AI like get
45:09 really good at AI so that if you're
45:11 disrupted in your business by one of
45:13 those PE
45:16 jackasses get together with some of your
45:18 buddies on the weekend and put together
45:20 a competitive business and go take them
45:22 down like PE firms can't move like that
45:25 they're not thinking strategically
45:27 they're just trying to extract value go
45:29 go take the value from them right all
45:32 right
45:34 ripy um I remember emails going to
45:38 replace the post office now we have so
45:39 much shipping in the post office yeah
45:41 well that's that that's the other thing
45:43 I think that's a really good point the
45:45 other thing about the the Doom and Gloom
45:47 is like oh it's G to destroy all these
45:49 jobs it's also going to create jobs well
45:52 what kind of jobs no one knows yet we
45:55 don't know what these things are like
45:58 like AI right now is so [ __ ] janky
46:01 it's so janky that you almost can't
46:04 imagine it taking anyone's job but
46:07 what's going to happen is it's just
46:08 gonna just like images you know how
46:10 images were always all [ __ ] up and now
46:13 you occasionally see one that's [ __ ]
46:15 up and like in six months you won't see
46:17 any that are [ __ ]
46:19 up every aspect of business is going to
46:21 be like that where it could it it can't
46:24 code right now well now it can sort of
46:25 code oh [ __ ] now it can really code Well
46:27 it can't do business oh now it can
46:29 really do business plans oh now it can
46:31 really do project management oh now it
46:33 can really do this like all of those
46:35 Dominos are going to
46:37 fall so all
46:40 right let's say what we got here Tom N I
46:45 wasn't listening is this going to be on
46:47 the
46:48 test everything that I say is going to
46:52 be on the test here's your Saving
46:55 Grace um these streams get saved
46:58 straight into the YouTube channel a day
47:00 or two later there's a
47:02 transcript and and you create the test
47:06 and then you just have chat GPT answer
47:09 the test everybody gets a
47:11 100 I encourage the use of of chat GPT
47:15 to cheat because if you can figure out
47:16 how to get a transcript from one of my
47:18 little ramblings here turn it into a
47:21 lesson plan have it Ace the lesson plan
47:23 give yourself an a I will sign that
47:28 the AI learning lab is an accredited
47:30 University well no I mean we've got
47:34 we've I I've got a credit
47:36 score and I and I and I know the word
47:49 university oh good lord now I can
47:52 finally want to be one of the be an
47:54 overseer like in Fallout I know
47:59 YouTube transcript or whisper it needs
48:02 to be
48:02 [Laughter]
48:08 accredited where where did you get your
48:10 where did you get your AI degree I I got
48:13 it at the at the AI learning lab I it
48:17 it's there's a dog there's a dog in they
48:20 start the with the it's he's got a
48:22 guitar and he knows well he knows like
48:24 six chords and four songs but but then
48:27 he's got a dog that play sings Some
48:30 sometimes he didn't sing tonight but he
48:32 sings sometimes and and then but they
48:34 get he got one of he went to Staples and
48:37 they got one of the um it see it it's
48:40 got the thing and then they sign it and
48:43 then so that's where I got
48:48 it what did I learn there
48:52 um about AI I learned about AI
48:58 were his what oh his his his
49:01 qualifications yeah he he talks about
49:04 that um he he doesn't he he doesn't have
49:07 any qualifications but that's that's
49:09 where I learned it like I learned the AI
49:11 from because
49:13 he I got that's why I got it
49:17 framed we're accredited
49:27 how many hours of credit it's two it's
49:30 it's 10 hours a week I never understood
49:33 the hours thing in college it's like
49:36 that's you get one hour for that I'm
49:37 like I'm going for the whole [ __ ]
49:39 semester why is it why do I only get an
49:45 hour this is this is a this is a 10hour
49:48 a week uh
49:50 [Laughter]
49:55 University oh man
49:58 subar subarashi piano cool name people
50:01 are just so scared of AI for some reason
50:04 they discredit any advances it's sad I
50:06 agree it's sad but I I would I would
50:08 encourage you um Saba subarashi sub
50:13 subarashi piano I would encourage you to
50:16 have have empathy um there's a couple of
50:20 things changes hard changes
50:23 hard couple of reasons people are afraid
50:26 of AI one is um we've got Hollywood
50:30 telling us the robots are going to kill
50:33 us historically if you wanted to be good
50:37 at AI you actually had to have like a
50:39 math degree you had to actually
50:41 understand things like embeddings and
50:44 the latent space and waiting and
50:47 algorithms to dig into the Thousand
50:50 dimensional latent space and I I it's it
50:53 attention and it like you had to know
50:55 that [ __ ]
50:57 and so part of the fear is that people
51:00 are like I like I've been illustrating
51:03 for 20 years I can't learn AI like I
51:05 don't I'm not a mathematician so part of
51:08 the fear is just they don't understand
51:10 what's required to be good at it like
51:13 like why it makes sense to have someone
51:15 like me running a channel like this is
51:18 you can just use your
51:21 words right use your words it's like
51:25 we're dealing with toddlers
51:28 you want something learn to ask for it
51:31 oh okay I can do that yeah so anyway
51:36 Bachelor of Kyle in artificial
51:40 intelligence give us a little
51:46 credit oh man all right what are we
51:49 gonna do tonight you wanna let's let's
51:51 go play with something
51:54 [Music]
51:56 um I'm going to go I'm going to go
52:01 somewhere you can't imagine I would go
52:04 um
52:06 Brandon and pate if pate's in here so so
52:10 someone needs to take notes if pate's
52:12 not in here um I'm actually gonna do
52:15 some demoing on Gemini Okay
52:20 so
52:22 yesterday
52:23 yesterday or the day before it doesn't
52:26 [ __ ] matter some sometime this
52:29 week in the middle of open AI ship Miss
52:34 sundor Pai and Google release Gemini
52:39 2.0 and so here where I am right now is
52:42 I'm at Google um AI studio so this is
52:45 the development site that is kind of
52:48 like the equivalent of uh open AI
52:52 playground so if you've ever been to
52:53 open AI playground if you haven't it
52:56 doesn't matter this is where programmers
52:58 come to to play around with different
53:00 models so if you click here there's
53:02 different models you can play with they
53:04 have different costs and if you like
53:06 roll over them it it gives you the cost
53:09 breakdown of how much it is to use them
53:11 and what they
53:13 do right and then there's these previews
53:16 so Gemini 2.0 flash this is a multimodal
53:21 model say that two times fast OKAY
53:25 multimodal model multi IM modal model
53:28 what are we supposed to do next
53:31 Professor shut up it's supposed to be a
53:33 tongue twister you're supposed to not be
53:34 able to say it like I
53:36 can't
53:38 jeez
53:40 um you can do things like turn up and
53:43 down temperature you can add tools you
53:46 can add structured output like all this
53:48 stuff if you're a coder you look get
53:52 code you click on this and look code
53:57 so if if if you get aroused looking at
54:02 back brackets and
54:03 such this you'll be hanging out here but
54:07 even if you don't there's cool things to
54:09 do here okay you get a roused looking at
54:13 back brackets why do you why does anyone
54:16 put up with
54:21 me I'm interested in their chip they
54:23 just made Jason yeah Willow holy [ __ ]
54:27 the the the chip that Google just
54:29 announced yesterday Willow this thing
54:31 that every time they add a cubit to it
54:33 it's a Quantum chip every time they add
54:35 a cubit to it it exponentially reduces
54:38 the error rate I'm pretty sure that's
54:40 what I read which it's a 30-year problem
54:43 in Quantum Computing is every time you
54:45 add a cubit it exponentially increases
54:48 the error rate and they built a chip
54:50 that exponentially decreases the error
54:52 rate every time you add a cubit [ __ ]
54:55 bonkers
54:57 [ __ ] Bonkers all right
55:00 anyway okay so Gemini 2.0 flash so this
55:04 is their new so Gemini
55:07 1. five is really good I I want to point
55:10 out a couple of things here one is even
55:13 with their latest 2.0
55:16 model flash is the small
55:18 one look at the token count right here
55:22 right underneath the The Flash thing is
55:25 this token count a million tokens and if
55:29 you if you flip down to um Gemini 1.5
55:33 experimental it's 2 million
55:36 tokens
55:39 so like a a normal sort of fiction
55:42 novels like 50,000
55:46 words so it's probably 70 65 or 70,000
55:52 tokens right this is 2 million so you
55:56 can upload a a tremendous amount
55:59 of data to these models and play with
56:01 them they're also also
56:04 multimodal it can understand video it
56:07 can understand audio and not like when
56:11 you uploaded it transcribes it into text
56:13 and then runs that text through a large
56:15 language model no no you can upload an
56:18 hourlong video and say what happens when
56:21 the wed Red Wagon is is is taken across
56:24 the
56:25 crosswalk and it'll tell
56:29 you
56:30 amazing and then um and and it it they
56:33 tend to be decently slow but look how
56:36 fast this new um Gemini is so I'm just
56:39 going to click on one of their
56:40 pre-written prompts here uh the math
56:42 work sheet generator create a set of
56:45 elementary math worksheets for math
56:47 Educators and parents bang
56:51 so oh of course it's going to be slow
56:53 now that I'm demoing it oh wait I'm in
56:55 oh I'm in Gemini 1.5 five flash so it's
56:58 a little bit slower all right so that's
57:00 I'm actually I'm actually glad I did
57:01 that all right so let's let's get rid of
57:04 that so you saw how slow that was it
57:05 took I don't know 15 20 seconds to start
57:07 so let's do let's go to Gemini
57:12 2 and go can I rerun The
57:16 Prompt let's see copy
57:20 this paste this
57:23 run boom
57:26 look how
57:33 fast and then it's doing it it's
57:36 creating these as little
57:38 code Codes copied to the clipboard with
57:41 caution oh yeah just here here's the
57:45 here's what our lawyers wrote that said
57:47 if it writes shitty code you can't sue
57:50 us but anyway um so that's cool and then
57:55 check this thing out so I think this is
57:56 all part also part of
57:58 2.0 over here on the left I'm going to
58:01 go to stream real
58:03 time and I'm gonna say show Gemini I'm
58:06 GNA allow it to use my microphone I'm G
58:09 allow it to use my
58:16 camera hey what am I holding here
58:28 um can you hear
58:30 that or do I need to share different no
58:35 okay sorry yeah sh
58:42 just hang on
58:45 people I'm trying to give Google some
58:47 sunshine here and their crappy screen
58:51 sharing stuff is a problem
58:58 [Music]
59:01 sure I can still hear you great do you
59:03 need to share how many fingers do I have
59:06 up you have five fingers
59:09 up um what's my shirt
59:13 say your shirt says
59:15 Catalyst what do you think that
59:19 is it seems that you were referring to
59:21 the name on your shirt Catalyst yeah I I
59:25 live in Denver Colorado what you think
59:26 it could
59:28 be okay you're asking what Catalyst
59:31 might refer to given that you live in
59:33 Denver Colorado I'm not sure what it
59:35 could be but I will try my best to
59:39 help uh
59:42 think it's too easy never okay I'm ready
59:46 n n never mind give me um give me an
59:49 inventory of the things you know on my
59:52 shelves and then ask me about something
59:54 that you don't understand
59:59 okay I will provide an inventory of what
1:00:01 I see on your shelves and then ask you
1:00:03 about something I don't understand on
1:00:05 your shelves I see a Monopoly game
1:00:07 several pictures and Frames some books
1:00:10 and other small objects some green file
1:00:12 folders a white box and a small yellow
1:00:15 Pikachu plush toy I'm curious about the
1:00:18 item that looks like a drawing with
1:00:20 lines that seem to represent a building
1:00:22 located in a frame toward the middle of
1:00:24 your shelf can you tell me more about
1:00:25 that
1:00:30 is it this
1:00:32 one yes that is the one I was referring
1:00:36 to it's a 202020 it's a 2022
1:00:43 print from an artist it is a that's K
1:00:48 Au z t i n as a first name and then p t
1:00:55 o as a last name
1:00:58 kazuk maybe kazin pedo does that name
1:01:01 ring a bell do you know that
1:01:04 artist you're asking if the name kazuk
1:01:06 and picato rings a bell I'm sorry I do
1:01:09 not know that artist huh but you knew
1:01:15 Pikachu okay I understand that you are
1:01:17 now referencing the Pikachu toy that is
1:01:19 located on your shelf who's
1:01:24 that you're holding a figurine of grou
1:01:27 also known as baby Yoda from the Star
1:01:28 Wars series The
1:01:31 Mandalorian you're pretty
1:01:34 smart thank you is there anything else I
1:01:37 can help you with no I'm just kind of
1:01:39 showing off for my Tik Tok live people
1:01:41 anyway that's cool you're cool thanks
1:01:44 man ah I understand you were just
1:01:47 showing off some of your items for your
1:01:49 Tik Tok live viewers that's
1:01:53 great can you tell me what this is
1:01:59 certainly the image appears to be a
1:02:00 slide from a presentation titled Fear
1:02:03 and Loathing of AI there's a large AI
1:02:07 displayed on the left side of the slide
1:02:09 on the right side there's a circular
1:02:11 diagram divided into four concentric
1:02:13 Rings the Rings are labeled with
1:02:15 different concepts business process on
1:02:18 the outer ring followed by business
1:02:20 product Prosperity which includes a
1:02:22 dollar bill icon then self-expression
1:02:24 which has a smiling face with part eyes
1:02:26 icon then creative product and finally
1:02:30 creative process on the innermost ring
1:02:32 the word play is written repeatedly in a
1:02:35 circular path around the
1:02:37 Rings what do you think it
1:02:40 means that's an interesting question it
1:02:43 seems that the diagram is trying to
1:02:45 convey the relationship between business
1:02:47 creativity self-expression and
1:02:49 prosperity with AI as a central theme
1:02:52 the word play surrounding the diagram
1:02:54 could indicate a sense of experiment
1:02:56 ation or exploration within these areas
1:02:59 perhaps the presentation suggests that
1:03:01 AI can be used in all those different
1:03:04 aspects of life and
1:03:06 work that's pretty good yeah it's
1:03:08 actually trying to frame learning about
1:03:11 AI for people that are afraid of it or
1:03:13 threatened by it that they can choose
1:03:16 parts of their their work or creative
1:03:19 life where they don't have to use AI but
1:03:21 then use it in other areas so they can
1:03:23 become AI literate
1:03:26 that's a great explanation it makes
1:03:28 sense that the presentation is trying to
1:03:30 frame AI as a tool that can be used in
1:03:33 different aspects of work in life but
1:03:35 also gives people the option to choose
1:03:37 where they use it this approach could
1:03:39 help those who are afraid of AI become
1:03:41 more comfortable and understand that
1:03:43 they have control over its use
1:03:45 ultimately leading to AI
1:03:47 literacy all right I'm gonna go now do I
1:03:50 how do I shut you down do I just go over
1:03:52 here you gone are you gone did I lose
1:03:56 you
1:04:02 um all right let me let's let's talk
1:04:06 about this for a bit
1:04:11 um what I just did in Google Gemini is
1:04:15 essentially what open AI released
1:04:18 today um some some version of it where
1:04:21 now on your phone you're just going to
1:04:22 be able to turn on the camera and just
1:04:24 talk to whatever ever it sees in the
1:04:27 world
1:04:31 um I like I I
1:04:35 don't I I mean maybe it changes nothing
1:04:39 right Google Lens like like if if you
1:04:42 use the Google app Google lens has been
1:04:45 in there for like five or six years or
1:04:47 more I maybe it's eight years Google
1:04:49 Lens you can pointed at a tree and say
1:04:51 what's this tree but but this feels
1:04:55 somehow different right like before
1:04:56 you're just getting an explanation like
1:04:58 now I could point this the camera at a
1:05:00 tree and say well tell me about that
1:05:02 tree and then I could just say well
1:05:04 who's ever written a poem about that
1:05:06 about an oak and it'll give me some poem
1:05:09 about an oak like like I I feel like
1:05:12 we're
1:05:14 entering all sorts of different
1:05:16 modalities you know it's funny let's
1:05:18 let's let me go back to my sharing
1:05:21 everything
1:05:23 screen how much does this a I cost me
1:05:27 per month I honestly don't know Tom the
1:05:31 the the Google Gemini see I'm paying I'm
1:05:34 part of Google work group so I think I
1:05:37 have access to this because we're we're
1:05:39 Android app
1:05:40 developers and we pay for Google so I I
1:05:44 honest to God don't know I think I think
1:05:50 you can you can um apply for an a Google
1:05:54 AI Studio account for free
1:05:58 honey curls Logic on Tik Tok how does it
1:06:01 feel about your
1:06:03 bow I'm not gonna I'm not gonna give it
1:06:06 the the
1:06:08 satisfaction because I you know I while
1:06:12 it would absolutely feel good that an
1:06:15 artificial intelligence model would tell
1:06:17 me how beautiful I am I don't need that
1:06:19 external validation so so but it's a
1:06:22 good idea
1:06:27 oh man what was that what was that
1:06:31 comment that I was
1:06:34 just talking
1:06:41 to oh
1:06:44 oh so when you have a tool like this
1:06:46 where so couple of things I showed it my
1:06:49 computer screen I showed it my webcam
1:06:52 and it can be interacting with it so I
1:06:55 could be in a complete completely
1:06:57 foreign piece of software like maybe I
1:06:59 want to edit in Final Cut Pro and I
1:07:01 don't know Final Cut Pro or maybe
1:07:03 someone's already always told me I
1:07:05 should learn uh uh aperture and to edit
1:07:09 my photos and I've never done that
1:07:11 before now I can just go into aperture
1:07:14 and just be talking to gemini or I can
1:07:15 just be outside in the world and and see
1:07:18 some language like on a sign that I
1:07:20 don't understand and say explain this to
1:07:22 me and then go into into this really
1:07:25 dynamic
1:07:29 fluid like I feel like the lines between
1:07:33 education and creativity and ideation
1:07:36 and execution are all blurring like
1:07:42 incredibly um
1:07:45 fast and I think part of our skill part
1:07:48 of the skill
1:07:51 set for being good at AI
1:07:56 is going to be adaptability and
1:08:01 and like creative and intellectual
1:08:06 nimbleness where you're just like like
1:08:09 like you know you know how you're I
1:08:11 don't know I
1:08:13 mean as someone with
1:08:16 ADHD um this is how it works for me but
1:08:19 like I walk down the street and I see a
1:08:22 squirrel and I'm like oh I wonder what
1:08:24 that squirrel's thinking and then it a
1:08:26 nut and I'm like oh nuts are interesting
1:08:27 I'm like oh I just bought peanuts at
1:08:29 Costco and do I like shelled peanuts or
1:08:32 like my mind is just a NeverEnding
1:08:34 string of thoughts and ideas and most of
1:08:37 the most of them are just horrific just
1:08:40 useless pieces of [ __ ] but
1:08:42 occasionally the shelled peanut leading
1:08:45 to the unshelled peanuts going I wonder
1:08:47 how they invented the peanut unshell
1:08:49 without ripping off the little red skin
1:08:51 and then that leads to hey I wonder if
1:08:53 you could do this and then all of a
1:08:55 sudden
1:08:57 I'll be at like I wonder if some actual
1:09:01 idea emerges out of all that stupid
1:09:03 stream
1:09:06 right now with
1:09:09 AI like the distance between the stupid
1:09:12 stuff and a and a and a decent idea and
1:09:16 getting that idea out in the
1:09:18 world is essentially
1:09:21 instant most of those ideas for me the
1:09:24 even the oh I wonder what would happen
1:09:27 if that most of those ideas just
1:09:31 die but now they don't have
1:09:35 to and so what do we do with that like
1:09:39 that that radically changes how we work
1:09:42 ripsy it enables curiosity at the core
1:09:45 of Being Human in a new way to learn so
1:09:47 this is what I wanted to do I wanted to
1:09:49 share my screen again did I share it
1:09:52 yeah I wanted to go back to this thing
1:10:02 is people that choose not to get curious
1:10:06 about AI right
1:10:09 now are cutting off so much more than
1:10:12 they realize they they think that AI is
1:10:15 a single thing and they think the threat
1:10:18 is monolithic right AI is so powerful
1:10:22 that I will no longer be creative it it
1:10:25 it will obliterate who I am as a
1:10:28 human and what they're losing is
1:10:32 the
1:10:34 422 other ways that they could use AI in
1:10:37 any given day that actually is not a
1:10:40 threat to who they are even if they want
1:10:43 to retain a sacred
1:10:45 container I will never write music with
1:10:48 AI okay great you want to do album cover
1:10:51 art do you want to come up with the name
1:10:53 of the album do you want to all of the
1:10:55 creative process [ __ ] around making
1:10:59 music do you want to have ai be a [ __ ]
1:11:03 ton better manager than your Stoned Out
1:11:06 loser manager that can't get you any
1:11:10 work and like the
1:11:13 more the more you play in that outside
1:11:16 ring the more you understand what's
1:11:20 possible the more you're going to be
1:11:22 able to nimbly weave AI in and out of
1:11:27 the process where you're still in
1:11:30 control like we as humans we still get
1:11:34 to choose the [ __ ] we put in the
1:11:40 world but now we have these [ __ ]
1:11:43 superpowers that are on our phone and
1:11:45 our silly little rabbit device and our
1:11:48 computer and probably inside this Yoda
1:11:51 two years from
1:11:52 now you know I'll have I'll have like a
1:11:56 hat with a Yoda on the bill and and this
1:11:58 is a fully fully articulated intelligent
1:12:04 being and it just it'll talk to people
1:12:07 for
1:12:08 me I'll be at a party and they like oh
1:12:11 that's cute oh you can talk to it if you
1:12:13 want oh hey little baby
1:12:16 Yoda and meanwhile it's capturing all
1:12:19 the ideas all the conversation
1:12:26 like people are going to just
1:12:29 flush so much opportunity down the
1:12:31 toilet if they don't get their [ __ ]
1:12:33 together tumultuous T
1:12:38 Thursday if actors trained their own AI
1:12:41 on themselves they could license their
1:12:42 workout get yeah get more listen this is
1:12:46 the thing
1:12:47 Jason I mean if I were a voice actor
1:12:50 right now I I would be freaking out
1:12:53 right I I spoke to a bunch of voice
1:12:55 actors about a year ago and I said you
1:12:58 know I didn't have great news for them
1:13:00 but but what I said to them was learn
1:13:03 everything you can about
1:13:05 Ai and about nine months ago 11 Labs put
1:13:10 out a professional voices
1:13:14 product that you can spend about 30 or
1:13:16 40 minutes
1:13:18 recording yourself to make a
1:13:21 professional voice that they will then
1:13:22 license out and they'll pay you for it
1:13:27 I wonder how many of those voice actors
1:13:29 that I gave that advice to actually did
1:13:32 that I would assume not a lot I hope all
1:13:35 of them
1:13:37 did but it's like the
1:13:40 opportunities there are more
1:13:41 opportunities if you play with AI than
1:13:45 if you don't all
1:13:48 right um thoughts I'm doing this this
1:13:52 week Mark beautiful
1:13:54 good um what do we got going over
1:13:57 here the nuggets come so fast now and I
1:14:00 can speak to the point so clearly
1:14:03 because of AI yep
1:14:05 absolutely Peter concurs that's well
1:14:08 said I don't know what I said but thank
1:14:10 you the peanut Innovation lab raised the
1:14:13 standard of living in mly Africa uh
1:14:16 Peter Sven
1:14:18 sarude um anybody who does not use AI to
1:14:22 customize their resume and cover letter
1:14:24 to the specific job they are playing for
1:14:25 is at a great disadvantage and quite
1:14:27 frankly Peter anyone right now who
1:14:32 doesn't figure out what agents are and
1:14:36 has an agent automatically crawl
1:14:38 LinkedIn every day to find Opportunities
1:14:41 and write custom letters
1:14:43 autonomously like there was a dude that
1:14:46 he he wrote a uh he wrote an autonomous
1:14:49 agent I think it scrape blinked in found
1:14:51 the people's emails the hiring manager's
1:14:54 emails wrote the them a custom email
1:14:56 emailed them and in 24 hours it had
1:15:00 applied for or sent you know a thousand
1:15:02 emails and booked 50 meetings in 24
1:15:09 hours like that technology is there
1:15:13 today now do you have to be geeky and
1:15:16 figure it out and probably learn a
1:15:18 little bit of programming yeah but you
1:15:21 know what you have to help you do that
1:15:24 AI it'll teach you how to program it
1:15:27 just go get the documentation give it to
1:15:29 chat GPT or Claude or Gemini and say
1:15:32 read this documentation and help me
1:15:34 build this thing it'll build it and
1:15:37 it'll start sending you
1:15:38 emails now when when open Ai and and all
1:15:43 these companies meta drop these agents
1:15:45 that are super easy to use and it gets
1:15:47 built into OSX guess what happens
1:15:50 then the early mover Advantage goes away
1:15:54 completely you'll be competing against
1:15:57 everyone else that's using Ai and we'll
1:16:00 be in a world where Bots are applying
1:16:02 for jobs that Bots are responding to and
1:16:05 no one's gonna get hired again and
1:16:07 everyone's gonna be like ah see AI
1:16:09 doesn't work I told you I didn't get 50
1:16:12 meetings that guy got 50 meetings
1:16:14 because he was the first one to do it
1:16:16 it's like being in
1:16:18 crypto everything's a [ __ ] Ponzi
1:16:20 scheme
1:16:25 um what else did I want to show you
1:16:29 tonight what's this Kaa oh this is
1:16:34 ideogram um Korea got um got
1:16:39 editing we can look at Korea how we
1:16:41 doing time
1:16:51 wise okay in Korea so if you haven't
1:16:54 seen Korea you go to generate
1:16:58 images
1:17:00 um there's there's lots of different
1:17:02 models in here but K is really fast if I
1:17:05 go let's say um wonder if it'll do
1:17:09 trademark [ __ ] baby
1:17:14 Yoda in a
1:17:19 Playboy
1:17:22 bathrobe I bet it won't generate this
1:17:25 will it yep sure did nice all
1:17:29 right
1:17:31 um so K is really fast uh it's got a
1:17:35 bunch of different models you can play
1:17:36 with including the super new sexy flux
1:17:39 1.1 Pro Ultra which this will be slower
1:17:43 but it'll still be good it'll be better
1:17:48 probably but I mean for the crappy model
1:17:51 those are pretty good I mean his feet
1:17:53 are really [ __ ] up there but
1:17:56 that's pretty
1:17:58 good so the fancy one let's see how much
1:18:01 let's see how much better the fancy one
1:18:16 [Music]
1:18:20 is come
1:18:22 on come on Kaa what are you doing it's
1:18:25 taking
1:18:28 forever I don't want to live in a world
1:18:30 where generating original artwork takes
1:18:32 me longer than 20 seconds this is
1:18:35 exhausting
1:18:37 Todd hey Todd hey Todd what's
1:18:42 happening Todd Todd hey
1:18:49 Todd all right well screw that let's go
1:18:53 let's take this bad Yoda
1:18:56 now how do I get over to
1:18:58 editing do I
1:19:02 no uh download all right we'll download
1:19:04 it we'll do that all right so we
1:19:06 downloaded that so if we go back out to
1:19:09 the
1:19:10 homepage
1:19:12 no create edit here we
1:19:17 go no it took me it took me to Twitter
1:19:23 I'm on the page take me to
1:19:26 this is
1:19:30 exhausting okay so if you're at if
1:19:32 you're at Korea and you went to generate
1:19:35 things go to the edit tab good Lord all
1:19:37 right so now I'm gonna upload an image
1:19:39 so I'm gonna upload baby Yoda that we
1:19:41 just
1:19:42 [Music]
1:19:44 made and we got to deal with these feet
1:19:48 so I can change a
1:19:51 region can I move this let's see okay
1:19:55 so I'm
1:19:57 GNA do
1:19:59 that and
1:20:01 oops can I
1:20:03 [Music]
1:20:04 do can I hold down the shift key and do
1:20:07 another one yes all right so they're
1:20:10 using some good some basic conventions
1:20:12 I'm going to say um
1:20:15 [Music]
1:20:17 baby
1:20:19 Yoda
1:20:21 feet let's see if we get nonhuman feet
1:20:25 we need Yoda
1:20:26 feet generate generate
1:20:30 generate
1:20:33 [Music]
1:20:35 generate
1:20:37 nope that was not
1:20:42 good all right let's try that
1:20:45 again and we go no and click bad
1:20:50 interface all right we go here and we'll
1:20:54 go
1:20:59 green
1:21:02 tiny Yoda feet Yoda
1:21:04 [Laughter]
1:21:14 feet come on make them green
1:21:18 Korea oh it's it's this is not
1:21:21 good all right well so there's something
1:21:25 it can't
1:21:27 do let's go let's do
1:21:32 um we'll do rayan
1:21:39 sunglasses gener see if this
1:21:43 works what species was
1:21:46 he give baby Yoda Gucci slippers oh
1:21:48 that's a good idea Peter that's a great
1:21:50 idea all right there's there's the ray
1:21:52 bands I wouldn't exactly call those rayb
1:21:54 bands but they're cute
1:21:55 [Laughter]
1:21:57 let's let's do the Gucci slippers I
1:21:59 think that's solid
1:22:04 okay it's good okay um
1:22:09 Gucci satin
1:22:16 slippers right now there are students
1:22:18 struggling with an assignment on the
1:22:19 most basic level these models can scan
1:22:22 and Lead these students through these
1:22:24 assignments with responsive interactions
1:22:27 imagine I know Jim Walters
1:22:30 that's this is one of the things that I
1:22:33 get [ __ ] apoplectic about
1:22:38 is the educational
1:22:40 institutions are it I I don't know I it
1:22:44 can't seem to do like the feet are just
1:22:46 getting weirder and weirder it can't
1:22:47 seem to do anything other than feet um
1:22:50 oh let me cut objects hang on let me
1:22:53 let's cut that
1:23:01 huh can I just delete it oh that's
1:23:05 interesting cut object so it's doing
1:23:08 segmentation that's kind of
1:23:11 nice and then let's just
1:23:22 do can I change the size this come
1:23:29 [Music]
1:23:31 on wearing Gucci slippers satin slippers
1:23:38 um yeah it's like one of the things that
1:23:41 just makes me [ __ ] nuts is you know
1:23:43 these are our educational institutions
1:23:45 you know Junior High and high
1:23:49 school I hate it but the universities
1:23:53 resisting this makes me [ __ ]
1:23:56 insane because you've got freshmen in
1:23:59 school right now that are going to
1:24:01 graduate four years from now like take
1:24:05 in what's happened with generative AI in
1:24:07 the past two years it went from a
1:24:09 standing start with with chat GPT
1:24:12 generative AI not ai ai generative AI we
1:24:16 start with chat GPT in November 2022 and
1:24:19 in two
1:24:21 years look at look at what's Happening
1:24:23 Here look at what we're doing
1:24:26 here and the universities are still
1:24:29 contemplating whether or not they should
1:24:31 allow AI in the [ __ ]
1:24:34 classroom it's
1:24:37 astounding you've got instant access to
1:24:39 knowledge the the core skill that's
1:24:42 going to be required in four years is
1:24:44 going to be critical
1:24:46 thinking what better way to practice
1:24:49 critical thinking then compare how we
1:24:51 used to do [ __ ] with what's possible now
1:24:53 with AI
1:24:55 right here's how we used to learn let's
1:24:58 go learn some [ __ ] that way okay that
1:25:00 was good now let's go learn some [ __ ]
1:25:03 the AI way let's take notebook LM and
1:25:05 put all of our semester
1:25:07 courses into notebook LM and have it
1:25:09 generate a podcast and listen to that
1:25:12 podcast and then let's take a test and
1:25:14 see who does
1:25:17 better and then let's talk about that
1:25:20 why is it like that why was this one
1:25:22 worse or better why was this one worse
1:25:24 or better better what did we lose when
1:25:26 we did it the old way what did we lose
1:25:28 when we did it the new way what did we
1:25:30 gain that level of critical dialogue is
1:25:35 what universities are supposed to be
1:25:38 about and they're like yeah well we've
1:25:41 got we've got artificial intelligence
1:25:43 detectors it's okay they're not going to
1:25:44 cheat on their papers because we know
1:25:47 that the big skill coming out of college
1:25:49 someone should be able to write a paper
1:25:51 if you know what I mean they should be
1:25:52 able to gather their thoughts I mean
1:25:54 what's there going to be some magical
1:25:55 machine that'll just write paper oh oh
1:25:58 there is oh and I'm sorry it's what it
1:26:03 it's PhD level quality now uh
1:26:07 uh seriously all right I'll shut
1:26:12 up Lori K today makes me nuts my kiddo
1:26:15 is a freshman I'm trying to get him
1:26:17 involved I mean my advice to anyone
1:26:19 who's got a kid in college right now is
1:26:21 have them seek out anyone
1:26:24 a
1:26:26 professor a TA another student seek out
1:26:30 anyone that is actively learning about
1:26:33 AI like I someone asked me this the
1:26:35 other night when I spoke at the ad Club
1:26:37 here in Denver he said I've got a kid in
1:26:39 college he just he's he's a freshman
1:26:41 what's your advice I said my advice is
1:26:43 he needs to get two degrees he should
1:26:45 take the degree right go to the classes
1:26:48 learn the [ __ ] they're going to teach I
1:26:50 said he should identify any professors
1:26:52 that he's got or anyone that's AI
1:26:54 literate and latch on to
1:26:56 them and then rather than partying at
1:26:59 night what he should do is get himself a
1:27:02 completely separate education teach
1:27:05 himself Ai and teach himself you know
1:27:08 how to use it learn all of the [ __ ] that
1:27:11 he's going to learn in all of his
1:27:12 classes over a weekend and then take it
1:27:15 from there take the [ __ ] that he enjoys
1:27:16 learning and just [ __ ]
1:27:19 run that like that's that's what I would
1:27:23 recommend to kids listen no I got two
1:27:25 upstairs that don't give a [ __ ] about
1:27:30 it all right if you implement AI in
1:27:33 schools it makes schools
1:27:35 obsolete right so you know what figure
1:27:39 out ways to not make the school
1:27:44 obsolete because you know what's still
1:27:46 going to be valuable in
1:27:48 society is relationships and learning
1:27:51 how to work with other people and
1:27:53 teamwork we had two um CU Boulder
1:27:56 professors from the advertising school
1:27:59 at this at this meeting and and both of
1:28:03 them were a little tepid on AI but they
1:28:05 were kind of acknowledging oh it's sort
1:28:06 of here but it's not all that good so
1:28:08 we're
1:28:10 fine but but the one the one woman I
1:28:14 think who teaches writing um she's doing
1:28:18 a massive amount of group group
1:28:20 activities to the point that the kids
1:28:21 are like could we not do a group
1:28:23 activity
1:28:25 like she's teaching them how to be
1:28:27 functional social beings which seems to
1:28:30 be lost in the current generation so
1:28:33 like there is still absolutely a role
1:28:35 for universities but it's not going to
1:28:37 look like what it looked like
1:28:39 before so I know and I know they're not
1:28:42 innovators but at the same time like
1:28:45 they they are going to make themselves
1:28:46 completely
1:28:49 irrelevant I got schooled by fellow
1:28:51 teachers when I told the kids find
1:28:53 creative to cheat but seriously using AI
1:28:58 to fool me yeah and then you know what's
1:29:01 cool about
1:29:03 that and [ __ ] the teachers that scolded
1:29:06 you like yeah tell them to use AI to to
1:29:10 get one over on you and then any student
1:29:14 who doesn't get
1:29:17 caught comes up to the front of the
1:29:19 class and tells the class how they did
1:29:23 it right
1:29:25 and then they get extra credit for that
1:29:27 if if you can fool the teacher with AI
1:29:30 you come tell us how you did it we'll
1:29:32 talk about that and then you'll get
1:29:34 extra credit for that like yeah why not
1:29:37 that's the kind of Ingenuity and and
1:29:40 nimbleness that is going to be required
1:29:42 in the world four years from now it's
1:29:44 it's required in the world right now
1:29:46 just no one knows it
1:29:48 yet all right
1:29:56 I can't generate AI can't gen generate
1:30:00 anything new just rehash old stuff not
1:30:08 true not
1:30:12 true if you understand how it
1:30:15 works if you actually use
1:30:18 it it is not copying and pasting from
1:30:22 previous documents
1:30:25 yes it's been trained on all of the
1:30:27 previous
1:30:29 documents and conceptually because this
1:30:33 is how computers have always worked we
1:30:36 think about data going into
1:30:39 databases and
1:30:41 so what everyone says is it it can't
1:30:44 create anything new because it's only
1:30:47 been trained on known stuff so it can it
1:30:49 can only spit out known stuff it's not
1:30:53 how it works
1:30:55 when you put in
1:30:59 130,000 1hour drama television
1:31:03 scripts into an AI embeddings
1:31:07 engine those scripts cease to
1:31:11 exist they're not in there it breaks
1:31:15 them up into individual tokens fragments
1:31:17 of Words which it then semantically
1:31:20 clusters in Thousand dimensional
1:31:23 mathematical space face and when you ask
1:31:26 for a new script and you say I want
1:31:29 something kind of like friends and kind
1:31:31 of like the Mandalorian but I want it to
1:31:33 be set on a boat where they get crashed
1:31:35 like Gilligan's Island it will write you
1:31:38 that thing and it's not copying and
1:31:40 pasting from those three different
1:31:42 scripts it's actually doing in essence
1:31:45 what a human does when you go to school
1:31:48 you learn
1:31:49 about Mandalorian drama structure and
1:31:53 you learn about friends s structure and
1:31:55 when you were a kid you watch Gilligan's
1:31:57 Island and all this shit's in your brain
1:31:59 and you're like huh hey you know it' be
1:32:03 funny and and you pull from some bizarre
1:32:06 neural network in your skull those ideas
1:32:10 and you mash them together and out comes
1:32:13 a novel
1:32:14 idea AI is [ __ ] awesome at that it's
1:32:19 incredible at
1:32:21 it so so just parting
1:32:25 what AI doomers have said does not make
1:32:28 it true so
1:32:31 anyway I I meant that AI can't invent
1:32:33 anything I would argue that AI can
1:32:36 invent I would argue that it
1:32:40 can right now it probably still requires
1:32:43 a human
1:32:46 to say oo that's an interesting idea to
1:32:50 maybe refine
1:32:52 it but if you look at open AI is five
1:32:55 levels of
1:32:58 AGI we're up to
1:33:00 three okay the five levels of AGI so AGI
1:33:04 is artificial general intelligence that
1:33:07 is the point at which AI is better than
1:33:09 most people at most jobs this is Loosely
1:33:14 how we're defining this across the
1:33:17 industry level one is chat Bots we had
1:33:19 that two years ago level two is
1:33:21 reasoning we got that four months ago
1:33:24 level three is Agents so agents are
1:33:28 going to be things that can take action
1:33:31 on your behalf it can go out and surf
1:33:32 the web research [ __ ] learn it come back
1:33:35 and do it Google just announced a thing
1:33:38 called Deep
1:33:40 research that you give it a topic and it
1:33:42 goes out and does deep research and
1:33:44 comes back with this synthesized thing
1:33:46 it just goes and does it for
1:33:48 you like think perplexity on steroids
1:33:54 this week probably tomorrow or next
1:33:57 early next week open AI is going to
1:33:59 launch operators which are autonomous
1:34:01 agents so we're at level three guess
1:34:04 what level four
1:34:06 is innovators and
1:34:10 inventors right you combine these three
1:34:14 and you can do that I would argue right
1:34:16 now if you sort of replace the agent
1:34:18 with yourself AI can absolutely invent
1:34:22 because what is an invention and
1:34:24 invention is nothing more than a human
1:34:26 being that took knowledge in their brain
1:34:28 mixed some [ __ ] together came up with an
1:34:30 insight and said huh let me try
1:34:32 something
1:34:35 new when we have autonomous agents it
1:34:38 can come up with the hypothesis and then
1:34:40 go test it that's invention like we're
1:34:44 there we're there all
1:34:47 right I don't mean to get so passionate
1:34:50 but but like just paring talking points
1:34:52 especially I don't I don't know your
1:34:53 name
1:34:54 if you're sitting on the sidelines if
1:34:56 you're not using AI because you think
1:34:59 it's just paring the [ __ ] that came
1:35:00 before and that it can't invent and
1:35:03 you're sitting on the sidelines like uh
1:35:05 wake me up when it gets
1:35:07 better it's a big
1:35:09 mistake it's getting better really fast
1:35:13 don't sit on the sidelines all
1:35:19 right okay mention this on Tik Tok do
1:35:21 you think AI should be used to create
1:35:24 entertainment movies series Etc should
1:35:28 my opinion is
1:35:30 1,000%
1:35:34 irrelevant do I think it should be
1:35:37 used I think that ship's already
1:35:41 sailed it is already actively being used
1:35:44 Lionsgate just cut a deal with Runway
1:35:47 and they're training a custom runway
1:35:49 model if you don't know what Runway is
1:35:51 it's a text to video model Lion's gate
1:35:54 right now is taking their entire catalog
1:35:56 of everything they own and training a
1:35:59 custom model model on it so so they're
1:36:02 use so it's happening right now just go
1:36:06 on Twitter look at look at all of the
1:36:07 movies that are being produced shorts
1:36:10 shorts right
1:36:12 now that are starting to get
1:36:15 good
1:36:19 right let me give you a little bit of
1:36:21 context for my point of view on this
1:36:24 because if you're like well you're just
1:36:25 a technologist and you know you don't
1:36:27 know anything about the entertainment
1:36:28 business I have a fine arts degree in
1:36:31 acting I moved to New York City out of
1:36:33 school I started and ran a theater
1:36:34 company I've WR written seven
1:36:37 screenplays um I started one of the
1:36:39 first digital web agencies like I
1:36:41 understand storytelling at an incredibly
1:36:42 deep level both on a business level and
1:36:45 on a creative
1:36:51 level the should it be used for
1:36:54 entertainment is kind of like saying
1:36:57 should synthesizers be used instead of
1:37:00 pianos back when they were
1:37:02 invented should they in in the 80s
1:37:06 should it be legal for anyone to sample
1:37:09 someone else's music and put that in
1:37:11 their song should no it shouldn't
1:37:14 according to those
1:37:16 rules but after 20 years of litigation
1:37:19 and figuring out IP and who gets
1:37:21 compensated and how you do all that well
1:37:23 yeah now sampling is just a part of the
1:37:26 industry AI is
1:37:29 identical it's a tool that is going to
1:37:32 feed into the creative process I am six
1:37:36 months in right
1:37:38 now on this channel six months ago no
1:37:41 eight months ago I was playing with
1:37:44 sunno and I put in a
1:37:46 prompt about the chatbot from from
1:37:49 Microsoft that fell in love with the New
1:37:51 York Times Reporter and said he should
1:37:53 leave his wife
1:37:55 and I put a prompt intoo and I said make
1:37:58 a song about a lover that's more
1:38:00 artificial than real named Sydney and
1:38:02 the song that it wrote was so
1:38:05 good and it it sounded like the end of
1:38:08 act one of a Broadway
1:38:11 musical and I said
1:38:14 um huh that sounds like a Broadway
1:38:17 musical I said it out loud and then I
1:38:20 had this amazing brain Flash and I said
1:38:24 holy [ __ ] this needs to be a musical
1:38:25 this is a great story I've spent the
1:38:28 last eight months I've got a writing
1:38:30 partner I've spent the last eight months
1:38:32 writing a musical all of the music is
1:38:35 right now is generated with AI the
1:38:38 original script outline was generated
1:38:40 with AI and my writing partner and I
1:38:42 have spent the last eight months
1:38:43 rewriting every [ __ ] piece of
1:38:47 it should AI be used in any way to
1:38:51 create that I don't know it was
1:38:54 because I haven't I haven't thought
1:38:57 about creating something for the theater
1:38:59 in 25
1:39:02 years but I have the skill to know what
1:39:06 to do if I have a good
1:39:08 idea and all of a sudden this song comes
1:39:12 out of AI that goes
1:39:14 huh I think this is a story that needs
1:39:16 to be told I think it could be told in a
1:39:18 musical in a really powerful way
1:39:24 that thing would not exist if it weren't
1:39:27 for
1:39:29 AI so like should it be used I don't
1:39:33 know I
1:39:34 mean no it
1:39:37 shouldn't if you don't if if you don't
1:39:39 want to innovate if you don't want
1:39:42 to here's the deal here's
1:39:48 okay this
1:39:51 channel the one thing that I don't I get
1:39:54 passionate about this stuff but the one
1:39:55 thing that I don't
1:39:57 do is take sides and say you know oh it
1:40:02 was trained on all this stuff that's
1:40:03 fine or it was trained on all this stuff
1:40:05 that's not fine like I don't take
1:40:07 sides on arguments that
1:40:11 ultimately are not going to impact the
1:40:15 trajectory right like there's a lot of
1:40:19 time and energy spent talking about well
1:40:22 it's just thievery
1:40:24 okay it may be it's POS like it it is
1:40:28 absolutely a gray gray
1:40:30 area the more you understand how AI
1:40:33 works the less it feels like thievery
1:40:35 and the more it feels like an incredible
1:40:38 library or an Incredible Museum that we
1:40:40 all get to tap into which is the whole
1:40:43 purpose of libraries and museums but put
1:40:45 that aside let's assume it's just
1:40:48 thievery I can be absolutely pissed off
1:40:51 about thievery and spend the next 20
1:40:52 years of my life SU the [ __ ] out of
1:40:55 everyone and AI isn't going
1:40:59 away so what I'm choosing to do in this
1:41:02 channel is say Okay assuming that AI is
1:41:06 not going to go away and assuming that
1:41:08 it's going to get better and better and
1:41:11 better what makes the most sense and
1:41:14 what makes the most sense to me and if
1:41:15 you disagree with this I wouldn't be
1:41:17 hanging out here what makes the most
1:41:19 sense to
1:41:20 me is learn everything [ __ ] can about
1:41:24 it because I don't want to be the victim
1:41:27 of this thing that is coming whether I
1:41:29 want it or not that is coming whether
1:41:32 it's good or not like I can have all
1:41:34 sorts of opinions on it and still take
1:41:37 advantage of it and that's what I'm
1:41:39 choosing to do all right Tik Tock wants
1:41:42 to know what level five is ah good
1:41:44 question okay so I was talking about uh
1:41:45 open ai's five levels of AGI
1:41:49 chatbots
1:41:50 reasoning which we just got 01 Pro and
1:41:54 01 regular so reasoning just got real
1:41:58 agents are coming um Google just
1:42:00 announced them um someone else just
1:42:03 announced them open ey AI is going to
1:42:05 announce them this week um innovators
1:42:08 and inventors is number four and then
1:42:10 number five and number five is a mind
1:42:14 [ __ ] because so four and five are
1:42:17 basically just remixes of one two and
1:42:19 three one two and three were the hard
1:42:22 ones one two and were the hard ones and
1:42:25 they're essentially
1:42:27 solved level five is
1:42:32 organizations and then you might sit
1:42:33 there and think like
1:42:36 organizations that that's weird what
1:42:40 huh so basically what happens
1:42:46 is you take the ability to talk to
1:42:49 something the ability to have it reason
1:42:51 the the ability for it to take action
1:42:53 and vent
1:42:55 [ __ ] and then you create a dynamic
1:42:59 system that will
1:43:01 generate a 100 a thousand or
1:43:06 10,000 autonomous
1:43:09 employees for the company that you just
1:43:12 started the organization that you just
1:43:14 started so I could literally
1:43:17 say I love this little
1:43:20 thing and I want to I want to
1:43:24 take I want to take every um cute animal
1:43:29 in a movie and I want to turn it into
1:43:31 little
1:43:33 desktop things that have Bobble I I want
1:43:36 to do bobbleheads of all these right and
1:43:39 so I could I could I'm going to
1:43:41 literally be able to say to this
1:43:45 system I want to start a business that
1:43:48 goes and finds all the cute animals in
1:43:50 all the movies secures the rights for
1:43:53 toy
1:43:54 manufacturing designs all the
1:43:56 bobbleheads figure figures out the
1:43:58 managing specs figures out the the mold
1:44:04 designs finds factories um I don't want
1:44:08 them where child labor might be used I I
1:44:10 you know here I I want them to be used
1:44:13 ethically sourced and good people
1:44:16 right I want to figure out how to get
1:44:19 them imported I want to secure a
1:44:21 warehouse I want to come up with
1:44:24 marketing I want to I want to create ads
1:44:28 for every new thing and before we
1:44:31 manufacture an ounce of plastic I want
1:44:34 to test the final designs with actual
1:44:37 consumers and anything that beats a
1:44:40 threshold of average will go into
1:44:43 manufacturing and anything that doesn't
1:44:45 will not and
1:44:48 go and it's going to go [ __ ] start a
1:44:51 business for
1:44:52 you Sam Alman talked about like a year
1:44:56 and a half ago every everyone you know
1:44:59 Andrew Yang had talked a little bit
1:45:01 about Ubi Universal basic income and
1:45:05 then people were kind of starting to
1:45:06 talk about it again when chat gbt came
1:45:08 out and about six months in Sam mman
1:45:11 said he goes you know like Universal
1:45:14 basic income is
1:45:15 interesting but but only for a short
1:45:18 amount of time and and only for a
1:45:19 limited impact he said I think what we
1:45:21 should actually be considering
1:45:23 is universal basic
1:45:27 compute where everyone's given some
1:45:30 allocation of compute
1:45:33 time and and I didn't get it at first
1:45:36 and I was thinking and I was thinking I
1:45:37 was thinking I was like oh this is
1:45:40 [ __ ]
1:45:41 brilliant so if you're given Universal
1:45:44 basic compute and we have and and you
1:45:47 know Sam mman knew what was coming right
1:45:50 a year and a half ago like they're two
1:45:52 years ahead of what we're seeing right
1:45:55 we were just giving Sora
1:45:57 1.0 they've demoed this past week Sora
1:46:01 2.0 right they're way ahead of where we
1:46:03 are so they know it's
1:46:05 coming so what Sam waltman's saying is
1:46:09 let's give people enough compute to run
1:46:12 a hundred agents for their virtual
1:46:14 company and and that's their gift and
1:46:17 then if you just want money you could
1:46:20 sell your compute credits to someone who
1:46:22 wants a thousand person company right
1:46:25 you could say oh Bob wants to get all
1:46:27 ambitious and have a thousand person
1:46:29 company I'll I'll sell him 50 of my my
1:46:31 hundred computes and that'll give me
1:46:33 enough money to whatever and then I'll
1:46:35 just use this other one for to make me
1:46:37 video games that I like to
1:46:39 play right and then there's going to be
1:46:41 other people that amass all those
1:46:43 compute tokens and they're going to
1:46:44 create 100,000 person companies and
1:46:47 million person
1:46:48 companies there's a bet in Silicon
1:46:51 Valley right now about who going to be
1:46:53 the
1:46:54 first single person billion doll
1:46:59 company like we will likely see that in
1:47:01 the next five years all right it's
1:47:04 getting late isn't it yeah it's getting
1:47:06 it's getting up
1:47:07 there
1:47:11 um rapid questions let's go let's go
1:47:13 knock knock knock down some questions
1:47:17 here all right lightning round what are
1:47:20 the limitations of AI
1:47:23 what are the limitations of AI and what
1:47:25 could be the next big thing the
1:47:26 limitations right now is the tools are
1:47:29 jankier than [ __ ] they're just janky
1:47:32 they're bad they're they they break
1:47:33 they're they're
1:47:35 inconsistent one time you try to make an
1:47:37 application it's absolutely [ __ ]
1:47:38 brilliant the next time you try to make
1:47:40 one it's just like wait does this thing
1:47:42 even [ __ ] work so just they're
1:47:45 inconsistent
1:47:46 um and what'll be the next big thing
1:47:50 um I think the combination of reasoning
1:47:54 and autonomous agents I so here's what
1:47:57 we're about to
1:47:59 hit and and this is a thing that I don't
1:48:02 have a good head for this and I don't
1:48:04 think most people
1:48:08 do exponential growth is not normal we
1:48:13 we as
1:48:15 humans can't comprehend it and we're
1:48:17 about to enter a phase where the quality
1:48:24 and capability of AI is about to just
1:48:27 like literally take off like one of Elon
1:48:29 musk's [ __ ]
1:48:31 Rockets right now we're in this flat
1:48:33 plane where where the the improvements
1:48:36 are kind of linear but what's going to
1:48:38 happen is you take reasoning and
1:48:40 autonomous agents so you you take PhD
1:48:44 level problem solving and you combine
1:48:47 that with the ability
1:48:50 for a fake person
1:48:54 to go test all the hypotheses that this
1:48:56 comes up with
1:48:59 right so I as a scientist say hey I'm
1:49:02 trying to cure
1:49:04 cancer and I say to this reasoning
1:49:06 engine I I have this idea if I mix this
1:49:10 this and that that there might be a
1:49:12 novel way to attack some protein
1:49:14 whatever whatever it is this thing
1:49:16 because it's got PhD level reasoning
1:49:19 goes and solves that problem comes up
1:49:21 with a 100 hypotheses and and then this
1:49:23 thing the autonomous agent can go do all
1:49:27 the testing in
1:49:28 simulation and then at some point we'll
1:49:31 be able to do all the testing in a
1:49:34 lab that cycle of a hundred simulations
1:49:38 100 hypotheses 100 simulations would
1:49:41 have taken that scientist two years or
1:49:44 four years previously and now it's going
1:49:46 to be done in two
1:49:49 weeks just every every problem that
1:49:52 needs to be solved from energy to
1:49:54 climate to health care to Material
1:49:58 Science to space travel to time travel
1:50:00 all of it is just GNA start to do that
1:50:03 so that's that's what's coming all right
1:50:06 that was supposed to be a short answer I
1:50:09 don't have short answers I don't know if
1:50:11 you knew this about
1:50:14 me uh Sammy Minister 4 have you guys
1:50:17 talked about Elon coming out with a
1:50:19 super computer Colossus um I don't think
1:50:22 he's come out with it yet but I yeah
1:50:23 Elon Musk wants to do a million GPU um
1:50:27 supercomputer right now he's got what
1:50:29 400,000 so he'll get there he'll do
1:50:32 [Laughter]
1:50:35 that um Eric on Tik Tok does AI have
1:50:39 imagination fascinating question
1:50:48 um hallucinations in a
1:50:51 way they're not
1:50:53 imagination but hallucinations in a way
1:50:56 are at least cre creative
1:50:59 thinking because as you turn the
1:51:01 temperature up so the way a large
1:51:04 language model works it's a predictive
1:51:07 calculator and there's this thing called
1:51:09 temperature and if you turn it to zero
1:51:11 it's 100% predictive that you write a
1:51:14 prompt and it will say this should be
1:51:16 the next word and then this should be
1:51:18 the next word or token this should be
1:51:20 the next word right it's very it's very
1:51:22 predictive that way it'll be consistent
1:51:24 time after time if you turn that
1:51:26 temperature up to one it it makes it
1:51:29 more random it'll say here's a cloud of
1:51:32 probable words and I'm going to you know
1:51:36 um give you you know some Meandering
1:51:39 path through these things so in a sense
1:51:41 it can act like it's got imagination but
1:51:44 it's just more like you know more random
1:51:48 than than
1:51:49 not but when so so
1:51:54 one of the things that the reasoning
1:51:55 engines are doing is they're generating
1:51:57 an answer they're then looking at that
1:51:59 answer thinking about that answer giving
1:52:01 itself another prompt so they're doing
1:52:03 this cyclical it's called a chain Chain
1:52:05 of Thought reasoning this sort of over
1:52:08 and over and over again prompting itself
1:52:10 so we may get to the point that they
1:52:12 have imagination I don't know that I the
1:52:15 Consciousness thing is a whole whole
1:52:16 other
1:52:17 thing um okay theorizing th Thursday
1:52:22 yeah preach it
1:52:26 rev RP star AMT AI is a hammer we use to
1:52:31 build the home we could do it without it
1:52:33 but sure makes it
1:52:35 easier yeah but it's see I think I think
1:52:40 Computing all almost all
1:52:44 Computing up to generative AI has been a
1:52:48 hammer right where where it's just
1:52:50 taking what we do and doing it in a more
1:52:52 efficent way generative AI can actually
1:52:56 design the house right it's there's a
1:52:58 different thing it can actually design
1:53:00 the
1:53:01 house um so yeah I don't know all right
1:53:05 last of these mitches on Tik Tok What is
1:53:08 the most efficient and effective way to
1:53:11 begin learning how to use AI so I'll go
1:53:14 back
1:53:16 to this
1:53:20 [Music]
1:53:23 play see this thing around the
1:53:26 outside play play play and then see if I
1:53:29 go over here to the AI
1:53:31 salon and go to welcome to the salon
1:53:34 step one in start your adventure on the
1:53:36 AI
1:53:38 Salon we've got this journey to AI
1:53:42 literacy and it's three steps and
1:53:45 they're really simple but they're not
1:53:46 easy and step one is play first like
1:53:49 literally play get cured ious play put
1:53:55 no expectations on it assume it's going
1:53:57 to be [ __ ] and let it surprise you play
1:54:01 play do [ __ ] for other people do [ __ ]
1:54:03 for yourself do [ __ ] for your kids
1:54:06 play in
1:54:08 playing well the other thing you should
1:54:10 do is join the AI salon so if you
1:54:12 haven't if you haven't joined the AI
1:54:14 Salon is you should go there the salon.
1:54:19 click on the button that says join our
1:54:20 community
1:54:26 and then and then join this group
1:54:28 mindfully creatives once you once you've
1:54:30 played enough you'll start to learn oh
1:54:33 if I combined this tool with this tool
1:54:36 with some way I used to work I might be
1:54:39 able to do something interesting here
1:54:41 that's mindfully create figure out a
1:54:43 problem you want to solve figure out
1:54:45 what you've learned about AI can do that
1:54:47 and start to solve it and then you'll
1:54:50 either fail miserably or not
1:54:53 and generously lead means share what
1:54:56 you're learning Learn out Loud be in a
1:54:59 community like this and as you're
1:55:01 learning say hey I got stuck here can
1:55:03 someone help me hey I didn't get stuck
1:55:05 here
1:55:07 right all right let's go see if uh by
1:55:11 the way there's there there's a he got
1:55:13 his Gucci slippers finally Yoda got his
1:55:15 Gucci
1:55:17 slippers and then what was I was I gonna
1:55:19 show something else here I think that
1:55:21 was it there's there's there's baby Yoda
1:55:25 with the Rayband glasses that don't look
1:55:26 like Rayband glasses and the Gucci
1:55:28 slippers that are way too big but at
1:55:30 least it's not baby
1:55:33 feet here here let's put a uh let's do
1:55:37 this we'll throw this here we'll say a
1:55:45 chrome Star Wars
1:55:49 appropriate pin
1:55:58 generate it's a Darth
1:56:00 [Laughter]
1:56:09 Vader all right fantastic Bob um so
1:56:17 tomorrow if you go to my LinkedIn I'm on
1:56:20 LinkedIn I'm Kyle Shannon
1:56:24 um look at my past
1:56:26 events and Fridays at 11:00 a.m.
1:56:30 mountain time I do what's called AI
1:56:33 office hours it's a really amazing group
1:56:35 of people that hang out they're
1:56:37 professionals that are trying to figure
1:56:39 this AI stuff out it's it's a really
1:56:40 neat group but the bonus for this Friday
1:56:44 and for next Friday is we're in the
1:56:46 middle of 12 days of ship Miss from open
1:56:49 AI which also happens to start at 11:00
1:56:51 a.m. so I'm going to start out the um
1:56:56 office hours by live streaming the live
1:56:59 stream of whatever they're going to drop
1:57:01 tomorrow I assume it's going to be
1:57:02 pretty big because it's a Friday and it
1:57:05 could be something really cool so we'll
1:57:07 see that tomorrow maybe we get agents
1:57:10 tomorrow
1:57:11 um so I'm doing that tomorrow night is
1:57:17 Friday I don't think I've got any sort
1:57:20 of art stuff with the misses so I should
1:57:23 be here at 8:00 Friday night date night
1:57:26 bring a date preferably a blind date
1:57:28 that you're just trying to get rid of
1:57:30 and then rather than Netflix and chill
1:57:32 you can just watch AI learning lab and
1:57:33 she'll just
1:57:35 [Laughter]
1:57:40 leave are are you really watching this
1:57:43 this is what you do is this this is the
1:57:46 the you made the nachos in a microwave I
1:57:49 I have to go I have to go you know that
1:57:52 sort of thing awkward is
1:57:55 funny without awkward you don't have
1:57:58 good stories so anyway Friday night date
1:58:00 night we'll do what we did here tonight
1:58:02 we'll play with some more stuff we'll
1:58:03 play with whatever the hell open a eyes
1:58:05 dropping there there's honest to God
1:58:08 there there I don't know Brandon there
1:58:10 are what 12 15 things we could be
1:58:14 demoing that I just we can't get to them
1:58:16 anymore it's it stuff's really moving
1:58:19 fast now so just just buckle up buckle
1:58:23 up
1:58:26 buttercups teachers are being abused by
1:58:29 the elites who meet the make the rules
1:58:31 oh I listen I don't I wouldn't want to
1:58:33 be a teacher right now I don't think
1:58:34 it's the teacher's fault I mean some
1:58:36 teachers the teachers that are you know
1:58:38 sticking their heads in the sand them I
1:58:40 blame but the ones like Mr K that really
1:58:43 are trying to do the right thing and
1:58:44 their their school districts are you
1:58:46 know blocking them from being able to
1:58:47 even talk about AI that [ __ ] grinds
1:58:51 my goat can you grind a goat I guess you
1:58:54 could you get a grinder you get a goat
1:58:56 grind it
1:58:57 up I don't know what that saying is
1:58:59 doesn't matter all
1:59:04 right all right everybody uh I'm gonna
1:59:07 get out of here thank you all for um for
1:59:10 hanging out looks like we had a good
1:59:12 good crowd
1:59:13 tonight and uh yeah I'm gonna go I I
1:59:17 forgot to bring a water so if you're
1:59:20 wondering why my voice sounds like this
1:59:21 I have had no FL fluid for two hours so
1:59:24 I'm going to go fluid eight I'm going to
1:59:25 go hydrate so peace out and I will see
1:59:28 you tomorrow at office hours at 11:00
1:59:30 a.m and 8:00 pm tomorrow for Friday
1:59:32 night date night all right peace out