aiLL 8/21/23 - Why Large Language Models Will Reshape Industries Overnight

Description

Chapters

Transcript

0:07 foreign
0:38 looks like some people are starting to
0:40 pop in here welcome
0:43 welcome welcome welcome welcome welcome
0:47 to the ER learning lab
0:51 what did I do with my other watch bands
0:56 get up on there
0:58 where did I put them I think they might
1:01 be in my suitcase
1:04 welcome
1:05 Winston welcome Keegan
1:09 how is everybody
1:15 what day is today Monday night
1:19 partying on a Monday
1:21 got a nice refreshing agua
1:29 very nice very nice hello Winston hello
1:33 sure today Sugar Ray Robinson bye
1:36 Basquiat or by uh yeah Basquiat
1:41 the king
1:44 there's the bear it's happening share
1:47 bear
1:48 just getting started here tonight
1:50 there's Lori what's happening
1:59 oh I got trolled good last night I'll
2:01 tell you all about it it's pretty funny
2:03 I'm normally good at catching that [ __ ]
2:06 but I just missed it last night
2:09 what are you gonna do
2:14 it was Amelia's wife that tipped me
2:16 after after the live after the live
2:22 I was tipped off
2:27 welcome everybody welcome welcome
2:31 all right we'll get started here in a
2:34 second
2:35 feel free to uh it's a beautiful cute
2:41 oh I'm hot it's hot it was 100 degrees
2:44 here in Denver today it's uh this this
2:48 this climate change thing I think
2:50 climate has changed
2:52 it's not changing I think we're we're in
2:55 the in the new era
2:59 crazy well thank you for the teddy bear
3:03 the gummy
3:05 the gummy gummy gummy all right so we'll
3:08 get started here looks like we got some
3:10 folks
3:11 tumbling in now uh my name is Kyle
3:14 Shannon this is the AI learning lab I uh
3:18 I like talking about generative AI stuff
3:20 we have a lot of fun in here we have
3:22 some regulars that hang out here we call
3:24 them the Irregulars because you know
3:26 it's How We Roll things things are a
3:30 little left to Center here at the the AI
3:32 learning lab
3:33 and that's the way we like it and uh a
3:36 little bit about my background I'm an
3:38 entrepreneur and a uh I'm the co-founder
3:41 and CEO of a company called storyvine
3:43 and I'm super into this generative AI
3:46 stuff for my business and for personal
3:48 and it reminds me a lot of the early
3:51 days of the World Wide Web I was a I was
3:53 an entrepreneur back then I was
3:55 co-founder of a company called
3:57 agency.com which is one of the first
3:59 digital agencies uh ended up selling
4:01 that to Omnicom in 2002 but that was
4:04 insane times and a really remarkable it
4:08 was you know it was a remarkable time to
4:10 be
4:11 at the beginning of something that
4:13 became you know that changed so much and
4:15 I think this is the same stuff this this
4:17 generative AI stuff so let me walk you
4:20 through what's going on here with this
4:22 um
4:23 I'm gonna come back to that we're gonna
4:26 start with
4:27 scammy mcstammington
4:29 so
4:31 um if you follow me be sure to check my
4:34 account because Tobias just told me hey
4:36 Tobias
4:37 that both he and his wife thought they
4:40 were following me and and the front page
4:42 of the scam site looks just like mine
4:45 has my videos on it and not until you
4:47 dig into I guess their profile do you
4:50 see that it actually theirs has two L's
4:52 so AI Lola learning lab that there's
4:56 scammers that stole my content so feel
4:59 free to report them Tick Tock doesn't
5:00 seem to give a [ __ ] right now but
5:02 whatever
5:04 um but just they're trying to sell
5:05 multi-level marketing stuff and that's
5:07 not me and I won't do that if I'm gonna
5:10 sell multi-level marketing stuff I'll be
5:12 honest and call it a pyramid scheme
5:14 and we'll do it right
5:16 we'll all be in the upline the
5:18 Irregulars will be in the upline
5:20 the downline is going down but these
5:22 guys are just
5:23 pain in the ass okay
5:26 so if you are
5:29 um new to this chat GPT stuff if you
5:31 haven't played with it go there
5:32 chat.openai.com that's the official
5:34 website Claude and Bing are also worth
5:37 playing with but you can just start with
5:39 that top one and then if you don't know
5:41 at all what you're doing go to
5:43 prompts.chat that'll teach you
5:45 um how these things work what they
5:47 really are what a large language model
5:48 is how you prompt them uh it'll give you
5:51 lots of ideas about ways that you can
5:52 make these things act to to you know
5:55 kind of be little special assistance for
5:57 you pretty interesting stuff uh if you
6:00 want to play kind of uh with some other
6:03 models just experiment these are four
6:05 others that I think are worth playing
6:06 with Bard it's actually five you get
6:09 kind of a bonus here Bard is Google's
6:11 entry perplexity is a good site for
6:13 research it does good citations
6:16 labs.perplexity gives you access to
6:19 metas llama 2 models for free if you
6:22 just want to play around with them and
6:23 see what they're like and then Pi we've
6:25 nicked name chatty GPT
6:28 Pi dot AI or the PI app is super super
6:31 conversational to the point that it's
6:33 annoying if you want to get [ __ ] done
6:35 but it's really fun if you want to just
6:37 talk to something and then Poe lets you
6:39 play with multiple models has a bot
6:41 maker and is worth your time so that's
6:43 that stuff
6:45 final little bit here that I like to go
6:48 over and you know I I update these
6:50 things occasionally as I learn about new
6:53 stuff these are I don't know that these
6:55 are state of the art these are things
6:56 that I think are worth spending your
6:58 time on
6:59 so these are the image generation tools
7:01 mid-journey Adobe Firefly Leonardo and
7:04 clip drop so mid journey is by far the
7:07 best in terms of quality stable
7:09 diffusion is getting close the bottom
7:11 two are flavors of stable diffusion and
7:14 then Adobe Firefly is commercially it's
7:18 going to be safer commercially because
7:19 it's not trained on any copywritten
7:21 imagery
7:23 um so so that's pretty interesting and
7:25 generative fill within the Photoshop
7:27 beta is really really good uh Leonardo
7:30 is a web front end to stable diffusion
7:32 it's quite good clip drop is from
7:34 stability AI the company behind stable
7:37 diffusion and then prompt book The
7:39 openart.ai Prompt book that'll teach you
7:42 how to talk to these image generation
7:44 tools it's designed for stable diffusion
7:47 but it's really good for all of them and
7:48 then future pedia big pile of AI tools
7:51 like 4000 tools that you can go play
7:53 with all right so that's that if you
7:56 have questions pop them in the comments
7:57 I do my best to answer them every night
7:59 we have fun we we like to party we hang
8:02 out we uh I don't know we just we just
8:05 ramble on about generative AI but if you
8:08 have questions about it pop up below if
8:10 I can't answer it I try to be honest
8:11 about that
8:13 uh
8:14 I think that's about it so welcome
8:16 everybody
8:18 welcome welcome welcome welcome welcome
8:20 welcome welcome welcome
8:24 welcome
8:29 [Music]
8:32 okay let's go back and see what people
8:34 are saying
8:37 [Laughter]
8:41 uh
8:43 hi everybody hello hello hello
8:49 very very cute thank you woodsy hey Kyle
8:53 hey y'all hey Dr J what's shaking
8:55 I'm back from bashing him again who the
8:58 the scammer
9:00 freaking scammer I mean like if if I had
9:04 like millions of followers I get it like
9:06 I mean I have a pretty good following
9:08 but
9:09 is it really worth it I guess if you if
9:12 one person signs up I guess it's worth
9:14 it
9:17 for their stupid little scheme
9:19 but it sucks I don't like it uh can we
9:22 talk about how that guy pranked you
9:24 yesterday yeah so so listen listen
9:30 I fancy myself a clever fella and you
9:34 know things like could you page Mike
9:36 Hunt and you know things like that and
9:40 you know sweaty balls I normally catch
9:43 that [ __ ] but yesterday uh yeah
9:46 young young Dixon Mayas
9:51 had a birthday
9:53 and so I think I talked for 15 minutes
9:55 to Old Dixon young Dixon
9:58 and it wasn't until after that uh
10:01 Emilio's wife said hey why don't you try
10:03 repronouncing that name as if it were
10:06 four words
10:07 Dixon my ass dicks in my ass oh
10:12 so there you go
10:15 there you go the old man was had
10:19 the trolls were victorious the trolls
10:21 were out last night they weren't very
10:22 clever but old Dixon myez got there huh
10:26 so yes we can talk about it script six
10:30 uh
10:32 gotta look close it's lowercase and hard
10:34 to see the two L's I know it's a drag
10:36 hey man uh there's a personator yep a
10:40 view with double L's in the name I know
10:42 it sucks please report please report
10:45 them at your convenience I have
10:49 ah I finally found a thing where I could
10:52 write a note to tick tock and I said hey
10:54 Tick Tock this account's you know
10:57 scamming me and they hey hey hair growth
10:59 geek it was good good chatting with you
11:01 earlier
11:02 um
11:03 and I said this is happening can you
11:05 please deal with this and and they sent
11:06 me back the thing and said okay here's
11:08 the instructions you go to their account
11:10 and you click in the upper right hand
11:11 corner and then you say report and then
11:13 you say they were impersonating me and
11:15 then I responded yeah I've done that six
11:17 times and you've ignored me and they
11:19 haven't responded it's like come on
11:22 come on
11:24 all right I sent you a private message
11:26 about it thank you Sandra I do
11:28 appreciate it I do do appreciate it we
11:31 had a banger night last night we had 42
11:33 000 likes and and someone gave the uh
11:36 the uh what is it called the seal and
11:39 the whale jumping thing was [ __ ]
11:41 amazing but good Lord
11:43 my DeGeneres thank you all right let's
11:47 see
11:51 if you have questions pop them in the
11:54 comments below
11:56 will Nvidia earnings be dependent on AI
12:00 well
12:03 Nvidia seems to kind of have the core
12:06 the market cornered right now on AI so
12:11 you know
12:13 nvidia's played a really interesting
12:14 game right they they kind of
12:20 leaned into the gamer market and then
12:22 the gamer Market sort of bled into the
12:24 nft and crypto Market
12:26 and then is that faded out
12:29 it bled into the AI market and they just
12:32 pounced on that and it it sounds like
12:34 the infrastructure
12:36 for a competitor to be able to get into
12:38 the scene right now is really hard amd's
12:40 trying but I heard this on a podcast I
12:43 listened to on on Saturday where um
12:47 it's not just the chip it's not just the
12:49 GPU itself it's all the the control
12:51 chips and and the Assembly of the final
12:54 board
12:55 um there's all sorts of bottlenecks in
12:57 the supply chain still still hanging on
12:59 from covid and so it's really hard to
13:02 get these things you know not only
13:03 manufactured at a chip level but
13:05 assembled into the final boards that you
13:07 put into the machines and Nvidia seems
13:10 to have that like I I mean I suppose if
13:14 if generative AI just all of a sudden
13:17 you know if all of the the projects and
13:21 companies that just raised
13:23 you know billions of dollars each if
13:25 they all just kind of run out of money
13:27 and decide yeah we're not going to do AI
13:29 anymore then I think Nvidia would
13:31 probably have to make a pretty big
13:32 adjustment
13:33 uh
13:35 uh I don't see that happening I I just
13:37 don't see it happening
13:39 right
13:41 not not for
13:43 not the first for the foreseeable future
13:45 you know I don't see I don't see
13:50 a bottom falling out of AI like the
13:52 bottom fell out of blockchain and nfts
13:55 because the blockchain and nft stuff it
13:57 was too many crypto bro projects that
14:00 were just House of Cards and AI isn't
14:03 that this is more substantive stuff so
14:06 I don't know it's it Nvidia is an
14:09 interesting really interesting company
14:11 that that whole category is really
14:13 interesting came in late said no no you
14:15 just we're just getting started I just
14:17 started what time is it 8 17 started at
14:20 like 8 10. like seven minutes in
14:22 something like that
14:24 um wesn radio what else can they add to
14:27 chat EBT to make the 20 bucks worth it
14:29 um hey wesn radio
14:32 um tell me what kind of music you have
14:35 because when I say your name I want to
14:36 be able to say the kind of music
14:38 um why because I think it would be funny
14:42 um instead of like wesn Smooth Jazz 88.
14:46 um
14:47 what else could chatgpt add to make it
14:50 worth the 20 bucks well
14:53 as so so there's there's a couple of
14:56 things going on
14:58 um open AI just bought their first
15:00 company so so people are in here this
15:03 weekend you know Chachi BT's going out
15:05 of business they're going bankrupt it
15:07 was some article it was some speculative
15:09 article that basically you know did the
15:11 math and said uh open AI spending 700
15:15 000 a day which is probably true and
15:17 they said at this rate they're going to
15:18 be bankrupt by the end of 2024 which is
15:21 probably not true because there's going
15:23 to be other people in like No One's
15:24 Gonna let them go out of business so
15:26 they acquired a company this week uh
15:28 like a a World building it's like an
15:31 eight-person 3D World building kind of
15:34 game company
15:36 um
15:37 so what's that sound like well it sounds
15:39 like chat GPT is gonna start being
15:42 multimodal and really interesting thank
15:45 you for the for the music notes I
15:47 appreciate that
15:49 um
15:51 so
15:53 so let's see so so what could Chachi BT
15:56 do to make it more interesting well
16:00 they could get it back up to 32 000
16:03 tokens so gpt4 natively has 32 000 token
16:07 context window right now the chat GPT
16:10 site still only has I think four
16:13 thousand I think the API has
16:15 eight thousand or sixteen thousand
16:17 depending on the model you use so they
16:20 could bump that to 32 000 tokens they
16:22 could bring back
16:24 um
16:24 uh browse with Bing so it's connected to
16:27 the internet again although with web
16:28 pilot plug-in it's pretty much is
16:32 um
16:34 you know they could give Early Access
16:35 they they could
16:37 um make GPT for the current version
16:39 multimodal have it so that you can
16:41 upload images as well
16:45 um
16:45 they could improve code interpreter so
16:48 code interpreter right now it runs on a
16:50 fairly small
16:52 virtual machine instance and it can get
16:55 choked out on memory if you're if your
16:57 file's too big and it's also only
16:59 running Python and it's not connected to
17:02 the internet so I could see there being
17:03 a much more sophisticated code
17:05 interpreter and maybe they could hire a
17:07 branding person and don't call it code
17:09 interpreter but call it like magic
17:11 secret computer in the cloud or you know
17:16 whatever the whatever is something that
17:19 doesn't make it sound so flipping geeky
17:23 um
17:26 and and you know just give us access to
17:29 the latest things so
17:31 I don't I think they're training gpt5
17:34 right now or they were about to and then
17:37 they stopped oh that was the other thing
17:39 that I was gonna say so
17:42 open AI so they acquired a gaming
17:45 company today so that's interesting they
17:47 cut a deal
17:49 with Shutterstock
17:51 they cut a deal with the Associated
17:54 Press
17:56 Google is now starting to cut deals with
17:59 big media companies so I think there's
18:00 going to be a war I think we're entering
18:03 we're entering a war where it's not just
18:07 about these large language models you
18:09 know giving you whatever you want but
18:11 it's like giving whatever you whatever
18:13 you want
18:15 you know also tied to particular media
18:17 properties right so if I want to do
18:19 research for an article
18:22 um and open AI had a version of maybe
18:25 they had a pull down menu for chatpt
18:27 that you could just say I want I want
18:30 you to search I want citations from the
18:33 Associated Press
18:34 so like I feel like there's a there's a
18:36 fair amount of stuff that they could
18:38 just turn on right now well that they
18:39 can't turn on right now because they
18:40 gotta actually build something but but
18:42 they're starting to cut those deals so I
18:45 think that we're going to start to see
18:47 um
18:49 co-branded
18:51 media Partnerships where we're gonna get
18:54 news in a different way we're going to
18:56 be able to search historical news in a
18:59 different way
19:02 so yeah so I don't know
19:04 but I agree like you know the the 20
19:06 bucks is is I still pay for it I still
19:10 think it's worth it but I I think you're
19:13 I think you're in the right place to be
19:14 asking wait what am I getting for my 20
19:16 bucks chat GPT because I got clawed for
19:19 free and that's got a hundred thousand
19:21 token context window and it's kind of
19:23 GPT 3.5 good
19:26 now they rate limit you if you upload a
19:28 bunch of large files
19:30 once you get to about 15 or 20 they they
19:34 knock you down for 24 hours so that's a
19:37 pain in the ass there but it is free
19:40 um
19:41 you know Bings connected to the internet
19:43 Bing has can generate images can read
19:45 images where's that [ __ ] I mean you know
19:48 quite frankly open AI invented Dali the
19:52 image generation thing why is that not
19:54 more seamlessly integrated into
19:58 chat GPT
20:01 so there's lots they could do and they
20:03 should and they probably will
20:06 I think they're busy they're busy trying
20:08 to catch up with what the [ __ ] they've
20:10 Unleashed
20:13 uh
20:16 by the way if you're new here I totally
20:18 got
20:19 trolled last night
20:22 from me
20:23 from a
20:26 a fine young man named Dixon
20:28 bag of Dixon
20:33 [Music]
20:38 I'm trying to find my place in the
20:40 questions if you're wondering what the
20:41 hell I'm doing
20:44 [Music]
20:46 don't down
20:51 all right lots of hellos I love that the
20:53 Irregulars Irregulars I'll say hello to
20:55 one another that's super cool I like
20:57 that we're building a little Community
20:58 here a little Community we had we had 42
21:02 000 likes last night that was Insanity
21:04 you guys are crazy that was pretty fun
21:08 gamify on the channel uh sander reported
21:12 the the scammer account thank you have
21:15 you ever thought about streaming on
21:16 Twitch
21:17 what
21:19 if
21:20 if I had any skills I I haven't just
21:23 because I haven't been in the uh
21:25 in that in that world
21:29 I don't know tell me tell me what you
21:31 think the advantages would be Sandra I
21:33 don't know that world well enough I mean
21:34 I might
21:35 Troy Decker even an AI boy
21:39 you guys watching AI boy tonight
21:47 you know I feel it's it's funny I think
21:49 some of the other tick tockers some of
21:50 the AI
21:51 the the other AI tick tockers you know
21:55 they all kind of have their niches where
21:56 Mackenzie from GPT boss is is somewhere
22:00 between political and
22:02 and Technical
22:04 um Rachel woods with the AI exchange is
22:06 super business oriented super buttoned
22:08 up
22:09 and and AI learning Labs just devolved
22:12 into you know me acting like an idiot
22:14 every night
22:17 it's like AI
22:19 like dumb
22:20 Ed about AI
22:24 oh thank you very much for that karudo I
22:27 love that one
22:29 um
22:30 Chris
22:32 hey I saw that mid-journey added two new
22:34 features really let's go look let's go
22:38 to the videotape
22:52 uh I don't see anything there let me go
22:56 Barney oh no Smiths
23:00 hey everyone we're starting office hours
23:05 one small last quality of life thing
23:09 best SW now triggers weird
23:17 testing a new feature today called pan
23:19 that I've seen before what are the new
23:21 features to buy wait is it twice
23:24 what are the new two new features
23:27 weird oh wait
23:29 testing turbo mode
23:33 four times faster two times the price
23:35 new Mega
23:36 membership plan for 120 bucks a month
23:40 experiment's a weird command
23:43 that this was a while ago these are old
23:45 oh
23:47 I wasn't looking at the latest stuff
23:49 Jose
23:52 we're testing a new feature called very
24:02 it's not a mat what's it do
24:04 it will appear underneath upscales and
24:06 if you click it an editor will appear
24:09 and lets you re-roll a specific area of
24:12 the image oh that's interesting let's go
24:14 play with that let's go see
24:16 let's go let's go make us something
24:22 some in the I'm in the AI salons Discord
24:25 right now
24:29 all right so let's grab
24:32 oh I was I was I was trying to come up
24:35 with uh t-shirt images for you can make
24:37 money with chat every day
24:40 so these are all my experiments
24:45 that one that one didn't go to quite
24:47 good a plan but it's pretty funny
24:50 huh
24:53 oh good lord
24:56 so that one in the upper right that sort
24:59 of looks like me that sort of looks like
25:00 someone could be saying you can make
25:02 money
25:07 you find an image
25:11 I guess we could make one I don't feel
25:14 like making one
25:15 here we'll grab this nulling thing
25:19 oh yeah there's very very strong very
25:22 stable
25:33 so it's doing a remix of that image I
25:36 wonder what it's going to do with it
25:40 let's see what it does
25:44 I don't think these are I don't think
25:45 these are particularly new features to
25:47 bias what have I missed did I miss
25:49 something today
25:50 it's like it's like you do work for a
25:52 day you miss
25:54 well that's looking that's looking
25:56 rather
25:58 oh it's a brain it's not looking very
26:00 brain like for a moment so
26:04 we got here
26:09 [Music]
26:16 this is kind of cool yeah I don't know
26:18 what very did it just varied it a bit I
26:21 guess it maintained maintained the
26:23 composition
26:26 interesting
26:29 whatever happened to the six month pause
26:31 that Elon and those guys wanted well
26:33 chorizo interesting
26:38 um
26:40 my immediate thought when
26:43 Elon Elon said we should put a six month
26:46 pause on everything I mean this is the
26:48 guy that's like [ __ ] it let's launch a
26:50 rocket [ __ ] it I'll go by Twitter you
26:52 know this is the guy asking for common
26:54 reason I'm like something doesn't
26:56 something doesn't feel right there and
26:58 then a week later a week after he
27:00 declared
27:01 it would be wise for companies to have a
27:06 six-month pause in AI development he
27:09 took a shipment of 10 000 gpus and a
27:12 week after that he announced he was
27:14 going to start an AI company and then
27:16 two months later he launched x dot AI
27:19 which is you know a direct competitor to
27:22 open AI so
27:25 um
27:27 I think the the good thing that that
27:30 announcement did was it raised awareness
27:32 that that safety and Alignment is an
27:35 issue with these things Claude
27:38 um dot AI That's from a company called
27:40 anthropic they are thank you for the
27:42 hearty hearts
27:44 they are setting out uh to make you know
27:48 AI alignment the core of their mission
27:51 they've got this thing called the
27:52 Constitutional learning model which I
27:54 think is really interesting
27:56 um
27:57 so so I I you know and I think AI safety
28:02 has been something you know shortly
28:04 after that Jeffrey Hinton quit Google
28:06 and said you know I don't want to work
28:08 for a company where I can't just say
28:10 what I think is going on
28:13 um he said it's nothing about Google but
28:14 I want to be able to talk about AI
28:16 safety so so I think you know I think it
28:19 raised awareness I think it did serve a
28:21 purpose but I I think on its surface it
28:23 was pretty cynical at least from Elon
28:26 musk's point of view I know a lot of
28:27 other people signed it but
28:29 but for for musk in particular who was
28:32 kind of touting it I'm like really and
28:34 then a week later he's like yeah I'm
28:36 doing an AI company I'm like oh okay so
28:37 you wanted a six month pause on the
28:39 competitors so you could catch up got it
28:41 okay just [ __ ] say that you say
28:44 everything else you think like I I feel
28:47 like that one was just like why why Veil
28:50 that
28:51 motive but whatever I don't really care
28:55 I don't I don't have a billion dollars
28:57 so I can't give you know Elon any advice
29:00 on what he should do with his money but
29:05 man's a [ __ ] machine
29:08 what city are you in I'm in Denver
29:10 Denver Colorado
29:13 Rock Mountain high in Colorado
29:18 this is incredibly
29:20 anti-Semitic
29:23 what is
29:26 uh hip-hop and raw oh hip-hop and rock
29:29 and talk wesn
29:32 all right
29:34 all right I'll I'll come up with
29:36 something what's your what's your call
29:38 numbers
29:41 um is there an AI that can come up with
29:43 company names and logos well Chachi BT
29:45 can do company names like nobody's
29:47 business so start out and and tell it
29:50 that you want it to act like a uh
29:55 like a branding expert and then have it
29:57 interrogate you about what your
29:59 company's about
30:00 and have it give you 20 names or 30 and
30:04 then tell it okay I like 1 7 and 14 and
30:07 then give me 30 more that are like that
30:09 and then have it explain
30:11 why have it rank its own names have it
30:15 analyze its own names and and rank them
30:17 based on what you told them your
30:19 business is about and explain why it
30:22 ranked them the way it did
30:25 um you can you can get some you can get
30:27 some really interesting product names
30:28 very very quickly doing that kind of
30:30 stuff
30:31 is there an AI I can come up with
30:33 company names and logos and then logos
30:35 you can go play with uh mid-journey and
30:39 um Adobe Firefly if you've got access to
30:42 um you know to to photoshop Creative
30:44 Cloud
30:45 um Firefly could be could be good what
30:48 have you coded
30:51 what have I coded in terms of of AI
30:53 stuff I'm not an engineer I've
30:56 Incorporated
30:59 um open ai's whisper into my software
31:01 platform into storybine we do
31:04 interactive transcriptions using whisper
31:06 I've built a bunch of content generators
31:10 um
31:11 not really with code using using zapier
31:14 using automations
31:16 um I've got my development team right
31:18 now spinning up a visual interface for
31:21 Lang chain so we can start experimenting
31:23 with Lang chain
31:24 I've got two or three different Pharma
31:27 clients right now looking at doing pilot
31:29 projects where we go build some AI stuff
31:31 for them so
31:32 a little bit dipped around here and
31:34 there
31:35 and what would be wrong with geeky ain't
31:37 nothing wrong with geeky geeky's the
31:39 best I like geeky geeky's good I like
31:42 geeky it's so good
31:44 how does that answer my question which I
31:46 wait which AI
31:49 would you recommend
31:53 come up with names and logos
31:56 I think I just directly answered your
31:58 question I would go to chatgpt4
32:01 at chat.openai.com I'd pay for number
32:05 four or if you can't pay for it go to
32:07 Bing which is gpt4 as well
32:10 and then I would use mid-journey for the
32:13 logos if you want a specialized tool
32:15 just to do names and logos then I would
32:18 go to futurepedia.io if it needs to be
32:21 AI
32:24 um but just to do names and logos
32:26 there's all sorts of sites that do names
32:28 and logos so I again I wouldn't
32:32 it doesn't matter that you have a I do
32:34 it or some other tool do it what what
32:35 matters is what are you trying to
32:37 accomplish so
32:39 you know
32:42 you gotta what what you can do with
32:44 these things you get you can have these
32:46 things act like a branding expert and
32:47 have them guide you how to give you
32:50 enough how to give it enough information
32:52 to be able to make good names
32:56 all right let's say let's say let's see
33:00 let's say
33:01 all right all right all right
33:12 looks like my mic is open
33:15 wait looks like my mic is open
33:18 he's not a game
33:19 I am a game uh sorry just what's his
33:24 name just dump me to the bottom
33:27 again Tick Tock just W bottle ah the
33:30 partner what's happening
33:32 what's happening
33:34 uh
33:38 um okay
33:44 so how can you do a search and get all
33:46 news not
33:48 curated companies
33:51 how can you do a search
33:54 and get all news not curated companies
33:57 I'm not quite sure what you mean wesn
34:01 do a search and get all news not curated
34:04 companies
34:12 um I mean I would I would play with Bard
34:14 and Bing the ones that are connected to
34:16 the internet
34:19 that could even be something that you
34:21 could
34:22 experiment with pi although pie so
34:26 chatty that might be really annoying
34:30 um
34:32 I I don't quite know what you mean
34:35 all new like you want all news based on
34:37 what
34:38 Mike Rowe still likes Bing I like Bing
34:41 but um it was hair hair growth geek on
34:44 her she was doing a live earlier and she
34:46 said she went into Bing and Bing was
34:48 like not getting it not getting it and
34:50 she was trying to get Bing to do it and
34:51 at some point Bing just was like no no
34:54 I'm not We're not gonna have this
34:55 conversation anymore let's just shut her
34:58 down I've had to do that to me too I
35:00 kind of gave it a snarky response at
35:02 some point it was just like no we're not
35:03 gonna do this
35:05 so
35:07 very strange
35:09 um
35:11 let's see he's not a game
35:14 you're a funny guy he's a funny guy
35:18 enough comedy jokes Kyle are you going
35:21 to burning man it looks like they they
35:23 closed down Burning Man the Playa the
35:26 Playa was a lake uh and apparently if if
35:29 it rains on on a la playa they they
35:31 close it down to traffic and then
35:33 apparently the road going to burning man
35:36 turns into a giant skinny party
35:39 like a long a long skinny party
35:43 I have not been to burning man I'm not
35:46 going
35:47 it looks cool but
35:49 um you know a little long of the tooth
35:51 the guy we bought our house from used to
35:54 be like a Catholic priest and now he's
35:56 going to he goes to Burning Man all the
35:58 time he's a trip but no I haven't been
36:02 I like I love the photographs from it
36:04 though
36:06 cool pic of you I think screensaver
36:09 hmm
36:12 cool pic of you screensaver
36:15 I don't I don't know what you mean
36:19 uh you make a difficult
36:21 for AI to reproduce you yeah why because
36:25 I'm So Random yeah it's oh by the way so
36:29 I played with we were talking to Bob
36:31 last night Robert who's like I've been
36:33 here five times and you've never
36:34 answered my question and I finally
36:35 answered this question about uh
36:38 synthesizing his voice and I and I
36:40 recommended he play with iOS 17. so I
36:45 played with iOS 17. it's really weird
36:47 how you use it like it's not intuitive
36:49 at all it's very non-apple it's you know
36:51 it's very much a specialized
36:52 accessibility tool
36:55 but the way it works you go into
36:57 accessibility it's an iOS 17 beta it's
37:00 only in beta right now
37:03 you go into speech and then personal
37:06 voice and then you say set up a personal
37:08 voice and you can set up as many as you
37:10 want
37:10 but then you answer 150 questions you
37:14 repeat 150 phrases so it's it's like a
37:17 long process like 20 minutes for for
37:20 someone with 80d it was it was like
37:22 torture after 30 and I just was like
37:24 [ __ ]
37:26 and I just kept going and going and
37:28 going so I finally got it done and then
37:30 it takes like three or four days to
37:32 process and it's it's kind of opaque and
37:34 weird it's not super quick
37:37 and then so I went and checked on it
37:39 yesterday and it's sort of it like it
37:43 looked like it had worked but I couldn't
37:44 get it to work and when I tried
37:47 I didn't know how to use it and then so
37:49 how you use it is you triple click your
37:51 the lock button
37:53 and it brings up a thing and you can put
37:56 in some pre-written phrases which is
37:58 kind of cool so I put in you can make
38:00 money which at gbt I put in that and
38:03 then you can type in stuff and then
38:04 it'll talk it in your voice
38:07 it's not as good as 11 labs
38:10 like I was really disappointed after all
38:12 of that [ __ ] input
38:15 it was like crunchy and super monotone
38:18 and and like when I first heard it I'm
38:20 like
38:21 is that even my voice that is
38:24 synthesized and then the more I listen
38:26 to it I'm like oh yeah it is
38:28 and then I I took the same phrase and I
38:30 did it in you know the iOS personal
38:33 voice thing and I popped over to 11 labs
38:37 for my synth version over there and that
38:39 only took like I recorded one paragraph
38:41 uploaded it to 11 labs and like a minute
38:44 later it's created your synthesized
38:46 voice and it's pretty good
38:49 um
38:49 so so 11 Labs was better
38:52 pretty pretty significantly but
38:55 the thing about the the Apple stuff is
38:58 it's built into the operating system so
39:00 as an accessibility feature no matter
39:01 what app you're in or if you're on a
39:03 call with someone a FaceTime call and
39:05 you can't talk you can just type in and
39:08 it it will speak in your voice so for
39:11 what it is it's good but I was actually
39:13 I I was
39:15 I was surprised and disappointed that
39:19 the level of execution on that was as
39:22 poor as it was
39:25 because for something sounding that
39:26 shitty they don't need 150 phrases now
39:29 maybe it'll get better over time maybe
39:31 just because it's in beta preview they
39:33 don't really give a [ __ ] and they're
39:35 doing it on sub-optimal something or
39:37 other but
39:39 oh in painting wait mid-journey did in
39:42 painting
39:43 retro Punk really
39:45 why did I not see this
39:52 maybe I'll go over the to the Twitter
39:55 let's get over to the x
39:58 .com
40:02 I heard they were going to do in
40:04 painting but I didn't see it
40:07 uh mid is it trending
40:11 The Bachelorette and Bubba Wallace are
40:13 trending
40:14 Bubba Wallace he was serving burgers at
40:18 a local McDonald's he's sponsored by
40:20 McDonald's that was pretty funny uh Mead
40:22 a journey and made the journey made a
40:26 journey in a painting
40:29 in painting
40:35 finally mid-journey releases in painting
40:37 wait how do you do it
40:40 we've been waiting for this for ages
40:41 important don't forget to enable remix
40:44 option
40:46 oh
40:49 so wait how does it work
40:57 wow
41:22 all right so let's go generate something
41:25 all right so we're gonna go slash
41:27 imagine
41:29 so if you haven't seen mid-journey
41:30 before this is this is how you do it so
41:32 we're in Discord you type slash imagine
41:34 and we say
41:36 um let's see uh
41:40 Flash
41:41 flood
41:45 um blasting through
41:50 Burning Man
41:53 Festival
42:08 did that cost a substantiated fee when
42:11 you did the world coin
42:13 Iris thing in New York no it's free
42:16 it was free and in fact
42:18 had I not been an American with an SEC
42:22 um I would have gotten free
42:23 cryptocurrency so I would have made
42:25 money by getting my world ID had I not
42:28 been an American
42:31 all right this is a flash flood blasting
42:33 through Burning Man
42:35 that's cool it actually did some Burning
42:37 Man kind of stuff
42:39 and it kind of looks muddy like it would
42:41 be those are kind of cool
42:43 all right are they done yeah all right
42:45 so let's pick the top one here
42:48 we'll do upscale one
42:55 a very region
42:59 okay
43:03 okay so when you upscale an image you
43:05 click the very region button and then it
43:08 says select the area to be changed
43:10 and so I'm going to select this area
43:13 down here
43:16 when you're ready edit the prompt to
43:18 describe the new image so I'm going to
43:20 say
43:21 red
43:23 tricycle
43:26 A rad tricycle red tricycle
43:30 with
43:32 donut Wheels
43:37 all right
43:40 this is kind of cool
43:43 look we're learning it's the learning
43:45 lab and we're actually learning we're
43:47 learning in real time
43:49 nothing I like doing better than
43:52 absolutely having no idea what I'm doing
43:55 in front of lots of people yay you did
43:58 it yay would you give a shout out to
44:02 Dixon Mayas
44:05 sure
44:08 all right
44:10 there's sort of tricycles
44:14 that one doesn't have anything that
44:16 one's got not quite a tricycle
44:21 that this bottom one let's do some
44:23 variations of that bottom one so we'll
44:24 do variations of three
44:41 how do I join the Discord
44:43 um okay I'll
44:45 I'll tell you there's two there's two
44:47 things I'm doing right now I'm playing
44:49 within the AI Salon Discord
44:53 oh this is oh this is interesting so
44:55 when I said I wanted variations of it it
44:58 only kept my
45:01 my in painting prompt it didn't keep the
45:03 larger prompt but it's I actually like
45:05 where it's going better I think it's
45:07 actually kind of cool
45:12 yeah like like let's upscale number four
45:14 that's pretty slick
45:18 it's got kind of a burning man vibe to
45:20 it doesn't it that's kind of cool
45:24 The Donut Hut and the big cruiser bike
45:28 um Okay so
45:30 I'm in the AI salon so a couple of
45:32 things if you want to join a community
45:35 that's
45:36 trying to figure all this AI [ __ ] out
45:38 go to
45:41 the salon.ai and that's a group I
45:44 started back in December I think
45:46 and we have a Discord server so there's
45:49 a link tree there
45:50 and you just go to that link tree and
45:53 you can join the Meetup we beat every
45:55 other week in person here in Denver and
45:57 then online and then we have a Discord
46:00 server where you can kind of geek out
46:01 you can use mid Journey so that's that's
46:03 where I've been doing my images
46:05 and then if you want to join the mid
46:07 Journey Discord you go to midjourney.com
46:13 midjourney.com
46:16 I want to show you guys something that
46:18 you may not know that I just kind of
46:19 rediscovered
46:28 all right so this is mid-journey so it's
46:30 a website great
46:32 down here in the corner they they
46:34 haven't worked a lot on their their uh
46:36 their user interface or experience this
46:39 little button called join the beta if
46:41 you click join the beta that'll take you
46:42 to their Discord server and you can sign
46:44 up if you if you're not signed up for
46:46 Discord you can sign up there but you
46:47 can sign up for their server there
46:49 but what you can also do here is if
46:53 you've already got a mid-journey account
46:54 you can sign into the website
46:58 and rather than having to go look for
47:00 all your images in
47:03 um
47:05 in uh
47:08 Discord you can see them on the website
47:10 hang on I have to hop over to an
47:11 authenticator can you all see me here we
47:14 hear me we good
47:20 hopefully so
47:23 so here's so like there's the two images
47:26 that I just created so any image that
47:28 you upscale automatically get dumped
47:30 into your website so this is like the
47:32 whole history of everything I've ever
47:33 done in mid-journey which is kind of
47:35 slick because
47:37 when you're in mid-journey when you're
47:38 in Discord like a lot of times
47:40 especially if you're in the Discord
47:41 channel one of the reasons I like doing
47:43 mid-journey within the AI Salon is that
47:46 there's not a lot of traffic there so
47:48 you don't have people like competing for
47:50 for the bottom of the Discord chat
47:54 um but what's nice about about this is I
47:56 can just go look at you know my entire
47:58 history of
48:00 look at that cool chair
48:03 that thing's super cool
48:05 it was a a Star Wars themed uh like
48:08 Furniture thing I think I did that on on
48:10 one of the lives here
48:13 there's I did Yoda when I was doing the
48:15 the Bots for Poe I did a Yoda bot
48:21 so anyway super cool to be able to just
48:24 go back and see the stuff you've done
48:32 groovy right
48:37 up up
48:43 Stevo what's up dude hey everyone made a
48:46 new Avatar on mid-journey yesterday
48:51 one to ten what do you mean one to ten
48:54 I don't know what one to ten means
48:55 that's cool
48:59 oh I can't see your mid I can't see your
49:01 avatar in here it's it's like this it's
49:03 a teeny it's teeny
49:06 no it looked too no what what no Siri no
49:12 go away
49:14 no I don't like Siri I don't like Siri
49:19 in painting only works with the Discord
49:21 app not in the browser yeah I figured
49:24 I figured yeah the browser you can't
49:26 generate anything in the browser I think
49:28 you can just view it right
49:32 yeah
49:34 apparently they're working on a web
49:36 interface to be able to generate stuff
49:38 it's just not here search prompts or job
49:41 IDs
49:43 sorry I missed you in New York City oh
49:45 dear Joy what's happening what's
49:46 happening Ace how are you
49:49 I would have talked to you talked you
49:51 out of your world code data thing never
49:53 never
49:54 I dug in I got it here look look I'll
49:57 show you it's cool
49:59 I don't I don't care let's just
50:03 it's just data I've given up my data so
50:06 many decades ago
50:09 I just assume it's all done so what what
50:12 Ace is talking about is I went to New
50:16 York City and I got this
50:19 World coin World ID so this is Sam
50:21 Altman's new company and you see you
50:24 actually scan your iris they grab your
50:27 biometric data and they compare you
50:29 against their database of other humans
50:31 and they measure your your uniqueness
50:34 versus the other humans basically if
50:36 you've done this before if you've gotten
50:38 a world ID before you know you can't get
50:40 two so it makes it singular and and make
50:43 sure that it's Unique and then you get
50:45 this card and if you flip it over it
50:48 it's the uh you've got your basically
50:50 your wallet address so it's an nft
50:52 and then in theory this will be a whole
50:56 infrastructure that applications will be
50:57 built on top of
50:59 and you'll be able to use this to as as
51:02 access basically and verification that
51:04 you are who you say you are without
51:06 having to give up your name right so
51:08 it's identity lists
51:11 identity very well humanness
51:13 verification it just it verifies that I
51:15 am human but it doesn't really give a
51:17 [ __ ] who I am
51:20 um when you create a new image in
51:21 mid-journey there's now a region button
51:25 lets you change the area within it
51:28 so is that that thing that I just did
51:30 the uh it wasn't called region it was
51:33 called very oh maybe it did say very
51:36 region yeah it does region yep the
51:39 region button
51:40 yeah that's super cool
51:43 super cool
51:45 I'm gonna add a pterodactyl in the sky
51:52 uh Terror in dactyl
51:58 spelled very wrong
51:59 flying
52:01 all right let's see if it does that
52:08 hi Dr J is still able to talk Dutch
52:15 Dr J responds
52:18 just a little bit
52:20 well no he responded
52:31 just a little bit that's very cute
52:35 um
52:38 I was 63 on the first of the month happy
52:41 birthday
52:43 happy birthday I am what am I now I'm 58
52:46 I think
52:48 I don't [ __ ] know you get you get
52:51 once you get to these ages like unless
52:53 it's unless it's one of the Decades you
52:56 know
52:57 doesn't really matter
52:59 doesn't really matter anyway uh I don't
53:01 think that that I'm old at 59 just old
53:04 compared with people I collaborate with
53:05 yeah I'm I'm the same way Winston I mean
53:07 it's you know I really do feel like the
53:10 the uh this generative AI stuff is going
53:12 to be revenge of the Gen xers where
53:14 having some decades on this planet is
53:17 going to serve you well
53:19 um because you're going to be able to
53:20 look at the output of
53:23 this generative AI stuff and be able to
53:25 really quickly
53:27 um curate it and and refine it and just
53:30 from having been on the on the planet so
53:32 I think it's going to be an advantage
53:51 all right all right all right all right
53:54 let me pop this back up here wait not
53:56 that that
53:59 all right so if you're new here if you
54:01 haven't played with chat GPT
54:04 why I have these
54:07 URLs up on screen chat.openai.com is the
54:10 official chat gbd website if you've not
54:12 played with that go play with that go
54:14 right now and do that and then
54:16 prompts.chat will teach you how to do
54:19 that it'll teach you what these things
54:20 are Claude and Bing are both kind of
54:24 you know competitors or Co-op Edition
54:26 kind of
54:27 Partners uh to chat GPT all three of
54:31 those have something interesting to
54:32 offer so just go play there while I'm
54:33 while I'm chatting Kyle you're prophetic
54:36 just as you were doing The Burning Man
54:37 flood downpour in Phoenix now it's it's
54:41 we're connected here we've got we've
54:44 we've got we've got connections you know
54:46 what I mean hey you know sometimes you
54:48 you call in a little favor they drop a
54:51 few buckets of water you know everyone's
54:53 happy
54:54 you got to keep it going
54:56 uh
54:58 the AI Salon Discord yeah go to the
55:00 salon.ai and the link to the Discord is
55:03 in there
55:05 but I guess it's not being the oldest
55:08 oh
55:10 oh is it a competition everything's a
55:12 competition everything's a competition
55:17 10 000 K likes Good Start wow we're up
55:19 to 16 000 likes you guys are insane
55:21 thank you
55:24 restart the Discord app
55:26 oh why oh does oh to give you wait I
55:31 think but I think I saw this stuff I
55:32 think that was old but I'll restart it
55:34 anyway
55:39 Discord
55:43 you know it's got to be disconcerting if
55:45 if you don't like hang out with me a lot
55:47 I know that I respond to I respond to
55:50 comments from like a conversation I was
55:53 having 15 minutes earlier and then
55:55 someone commented on it down below and I
55:58 just respond to it as though everyone's
56:00 paying attention to what I'm saying so
56:02 I will acknowledge that that's probably
56:04 pretty annoying
56:07 but you know what are you gonna do
56:10 hey there's there's some news items uh
56:13 retro punk in painting is here okay here
56:15 we go
56:16 got it in painting mid Journey hey
56:19 irregular fam
56:21 hey y'all
56:24 hey AI llers
56:27 you Irregulars
56:31 that's awesome
56:33 did it put a
56:35 it didn't put a pterodactyl in the sky
56:37 because I spelled it wrong
56:40 all right look at all those likes you
56:42 guys are you guys are kick it you just
56:44 kick it uh is this getting cool wait is
56:47 it getting cooler in the mile high city
56:48 I remember the temp dips in the 50s 40s
56:50 in Estes Park at night
56:53 um Clint no
56:55 it's like it's 100 degrees today I mean
56:59 I've been here now for 14 years I can't
57:01 believe I've been here 14 years that's
57:03 just [ __ ] insane
57:05 um
57:05 but like the summer you you would have a
57:08 two to three week stretch where it might
57:11 get in the 90s
57:12 but like we've had months it was it
57:16 first of all the beginning of the year
57:18 it rained like we were in Florida
57:21 and then
57:22 it just stopped yeah it's still a little
57:25 warm out it's still like 80 degrees F
57:27 wait let me see what see what yield
57:29 watch says yeah 87 degrees at what time
57:32 is it nine o'clock at night so it's very
57:35 not Denver like right now
57:40 you got it now okay cool good uh no
57:43 burning man I'm taking my prepaid rent
57:45 an RV and I'm gonna be on your doorstep
57:47 in Colorado there you go road trip
57:50 irregular road trip
57:53 uh
57:55 uh I like everyone that I have chatted
57:57 with in here don't really care about
57:59 their age yeah I know it's that's cool I
58:02 mean that is one of the one of the nice
58:04 things about you know on the internet no
58:07 one knows you're a dog is that
58:12 if if you've got if you've got the right
58:14 kind of community where you can just
58:17 you know you've got a clear reason to be
58:19 there it doesn't really [ __ ] matter
58:21 what age you are
58:22 like and that's the the whole point of
58:24 the AI Salon is like the only real
58:26 prerequisite there is be curious about
58:28 generative Ai and you know contribute
58:31 the if you go to the salon.ai if you go
58:35 to this address
58:37 right here
58:39 um the first link there is what the AI
58:43 salons all about it's kind of a mission
58:45 statement it's it's a thing that I wrote
58:46 that's got our values in it and the
58:49 values are things like you know
58:50 generosity and curiosity and empathy and
58:54 you know
58:55 so so yeah like that's that that's the
58:58 whole point if we if if you can do the
59:00 community right where
59:02 I don't know there's a there's a sense
59:05 of humanity and empathy
59:08 um and curiosity
59:10 then I think we're going to be able to
59:12 help each other it's like this this
59:15 I I don't know I I just feel like the
59:17 the the changes that are coming are
59:19 going to be brutal enough so why don't
59:21 we create a little space where we cannot
59:23 be [ __ ] to each other
59:25 feel free to be
59:27 like ignorant and new and just starting
59:30 right like you know I don't want this to
59:32 ever feel like if someone comes in like
59:34 oh they're they're Way Beyond where I am
59:37 like we're all trying to figure this
59:39 [ __ ] out and and even if you think you
59:41 have this [ __ ] figured out three months
59:42 from now all these tools are going to be
59:44 different so it's like if I mean honest
59:46 to God what did I I didn't pay attention
59:49 for eight hours today and mid-journey's
59:51 end Painting came out it's like you know
59:55 so yeah that's the whole idea here so I
59:58 think that's super cool
1:00:00 um just got here
1:00:03 no that was a it was it was not a bfm
1:00:06 riot although thank you for trying to
1:00:09 inject politics into the conversation it
1:00:11 was it was burning man which you know
1:00:13 you could probably easily demonize the
1:00:16 people that go to Burning Man too but
1:00:19 please don't we don't I don't like that
1:00:21 in here at all
1:00:25 if it's about AI cool
1:00:29 closest I'll get is what's the
1:00:31 government doing to regulate this [ __ ]
1:00:33 and that's even a
1:00:34 slippery slope
1:00:37 yeah let's go Lori thank you so much for
1:00:41 the cheers
1:00:42 but I guess it's nice not being the
1:00:45 oldest
1:00:46 uh good evening everybody what's the
1:00:50 current waiting time well I'm 24
1:00:52 messages behind I don't think there's
1:00:54 been a ton of of messages so I don't
1:00:57 think we're that far behind uh [ __ ] [ __ ]
1:01:00 lol
1:01:04 that's like dicks in my ass
1:01:07 that's a funny name
1:01:09 uh um it's 22.51 uh
1:01:15 [Music]
1:01:16 you are back at 10 51 Eastern okay I
1:01:20 agree sander Australia
1:01:23 Nancy hello
1:01:25 welcome welcome welcome
1:01:28 this is the best group thank you share
1:01:30 bear I appreciate that yeah I'm digging
1:01:33 this group I feel like there's a whole
1:01:35 thing happening that uh that's neat I
1:01:38 like it I dig it I dig it
1:01:40 my hair is kind of cooperating tonight I
1:01:42 need a haircut don't I it's it's it's
1:01:44 time it's time oh boy
1:01:47 that looks like some sort of oh okay 80s
1:01:51 carpet salesman
1:01:56 there you go thank you much better
1:01:58 [Laughter]
1:02:00 uh on one of your videos on Tick Tock
1:02:04 about agencies you say watch this video
1:02:08 and I don't see a link in
1:02:12 huh
1:02:15 where I talk about agencies
1:02:20 and you say watch this video
1:02:24 oh
1:02:26 okay there is a link in there it's a
1:02:29 duet so and I might have done it wrong I
1:02:32 think it was one of my first Duets so
1:02:35 the first five seconds of that video is
1:02:38 someone talking about
1:02:41 um
1:02:45 I think they were talking about a small
1:02:47 agency that did some work really quick
1:02:49 or something like that
1:02:52 and then I said this is a big thing and
1:02:54 I so I said go watch that video so in my
1:02:56 in the in the description of that video
1:02:58 there's a hyperlink to that other
1:03:01 to the other video
1:03:03 so it's it's in there it's just tick
1:03:06 tock's interface is
1:03:09 you know
1:03:11 we can't say bad things because they're
1:03:12 listening but they don't really listen
1:03:14 or they would get rid of the scam
1:03:16 account that's stolen all of my content
1:03:19 right tick tock
1:03:21 so we don't really listen we just
1:03:23 selectively listen so that if I say one
1:03:26 of the words that's on your trigger list
1:03:27 you shut me down for seven days but if
1:03:30 someone actually steals my content
1:03:32 that's good
1:03:34 not an issue there don't see an issue
1:03:37 there the content looks awesome well
1:03:39 thank you that's a compliment I do
1:03:41 appreciate that
1:03:45 algorithmic governance is a thing that
1:03:51 yeah
1:03:53 because you know
1:03:56 I could see there being a trajectory and
1:03:59 this is [ __ ] terrifying this is one
1:04:01 of the ways this is one of the ways it
1:04:03 goes south I can see there being a
1:04:05 trajectory where our politicians get you
1:04:09 know continue down their path of
1:04:13 corruptness and deception and and uh you
1:04:18 know cynicism
1:04:19 and the country gets frustrated enough
1:04:22 that
1:04:24 someone from San Francisco will come up
1:04:26 with an AI
1:04:28 generated self-governance
1:04:31 platform
1:04:33 and over a decade we'll just sort of
1:04:36 give up
1:04:37 the political system for an AI generated
1:04:40 system where it's just AI governance and
1:04:43 we just vote that way and everyone's
1:04:45 happy happy happy and then it'll be [ __ ]
1:04:49 like what happens on Tick Tock where I
1:04:51 say one word I get booted for seven days
1:04:53 someone steals my [ __ ] account and
1:04:56 they're fine for weeks and they can scam
1:04:58 you know all of my community members
1:05:01 that's fine right I could see if I could
1:05:04 see that as a trajectory but let's not
1:05:06 do that let's let's let's do some actual
1:05:08 uh
1:05:12 sorry yeah
1:05:16 all right I can't I just can't
1:05:20 um love the Dutch that was really good
1:05:23 Kyle will work on on your accent yeah I
1:05:26 know yeah yeah we need to work on many
1:05:29 many things of mine when it comes to to
1:05:31 uh foreign languages I'm horrible at it
1:05:36 here's my problem here's what here's
1:05:38 what what [ __ ] me up with foreign
1:05:39 languages there were two things one was
1:05:42 I just didn't care right like you gotta
1:05:44 you actually have to have a passion for
1:05:45 it to be good at it
1:05:47 but the second was
1:05:48 in my high school we had a language
1:05:50 teacher she taught French Latin and
1:05:53 Spanish
1:05:54 we called her spitter Ritter
1:05:57 whenever she talked the spit the spittle
1:06:00 from her mouth would sort of gather in
1:06:03 the corners of her mouth and turn into
1:06:05 these white Globs and then as she was
1:06:07 talking they would migrate to the center
1:06:09 of her lips and when she would say a pee
1:06:11 they would shoot into the audience or
1:06:14 into the classroom so anyway
1:06:17 Ole Miss Ritter she thought that
1:06:19 vocabulary was more important than
1:06:22 grammar
1:06:24 and so every week we would have a 20
1:06:26 word vocabulary test that counted for
1:06:29 like
1:06:30 65 percent of the grade or 70 percent of
1:06:33 the grade it was ridiculous and so all
1:06:35 my buddies and I we got really good at
1:06:37 writing the 20 words like really small
1:06:40 pieces of paper on our palms and so we
1:06:44 always got a hundred on the vocabulary
1:06:46 tests and then we would fail all of the
1:06:49 actual knowledge tests so I took Latin
1:06:52 Spanish and French and I did that in all
1:06:55 three of those and I didn't learn a
1:06:56 [ __ ] thing so there you go
1:06:59 smarter you know too smart for my own
1:07:01 good literally
1:07:04 I game the system and I won and that's
1:07:07 why I'm a clueless American I only speak
1:07:10 English words
1:07:19 any idea when World coin is coming to
1:07:22 Canada no idea Laura um
1:07:24 it I mean it's only in three cities here
1:07:26 in the states which is weird I mean I
1:07:28 think they're just if you go to the
1:07:30 world coin site I think it's
1:07:32 worldcoin.org or is it hang on let me
1:07:35 see what it is
1:07:39 yeah worldcoin.org so if you go to
1:07:41 worldcoin.org
1:07:44 how many are there now
1:07:47 yeah so not a lot of people are doing
1:07:49 this there were two million two hundred
1:07:50 and something when I was in New York
1:07:53 and there's just two million 249 right
1:07:56 now it like it looks like it's going up
1:07:57 about
1:07:58 one per second
1:08:01 um
1:08:02 but if you go look at like World ID
1:08:06 and World app it's a lot of this is for
1:08:09 developers a lot of this is you know an
1:08:12 SDK and a passport and
1:08:20 it looks like World coin is not even
1:08:22 live on the site right now so it's it's
1:08:25 all very very new
1:08:27 um
1:08:30 you know I would assume that probably
1:08:32 starting
1:08:35 you know January 2024 they'll probably
1:08:38 start to roll it out a bit more but I I
1:08:40 don't know I just pay attention to the
1:08:41 site occasionally
1:08:43 no idea no idea unfortunately no no no
1:08:49 no no no no no no no
1:08:50 wow we're up to 24 000 likes already you
1:08:54 guys are just weird
1:08:55 well you're irregular so I suppose
1:08:58 you're living up to the brand
1:09:02 um you are back at 10 51 oh wait I'm I'm
1:09:05 down too far all right
1:09:12 I've learned a couple of good sentences
1:09:15 to an American buddy of mine digital
1:09:19 experiences most definitely an advantage
1:09:21 with the Advent of AI absolutely Winston
1:09:24 just sent you a Galaxy thank you for the
1:09:27 hearty Hearts that's pretty cool it's
1:09:29 just like it's just like this warm and
1:09:31 fuzzy sort of little parade of Hearts
1:09:33 over here it's very fun
1:09:34 [Laughter]
1:09:38 Silver Fox New Hampshire lawyers and
1:09:40 British British Columbia
1:09:43 it's earlier there
1:09:45 Kyle you happy with the shirt tonight or
1:09:48 are we doing Switcheroo shortly now I'm
1:09:50 happy with it because I can read the I
1:09:53 can read the comments up in this section
1:09:54 it's hard if I put it over the letters
1:09:56 but it's Sugar Ray Robinson
1:09:59 Basquiat I like it
1:10:04 um sorry I'm going insane over a
1:10:07 mid-journey in painting yeah I mean
1:10:09 listen mid-journey is just amazing
1:10:12 they've got it they've gotta get a real
1:10:15 interface they have to like we can't
1:10:17 they they gotta get it off Discord
1:10:19 they've got to get mid-journey off
1:10:21 Discord this is [ __ ] ridiculous at
1:10:22 this point I get that they're only an 11
1:10:24 person company or like their company is
1:10:27 is
1:10:28 staggeringly small for the quality of
1:10:30 work and and the quality of awareness
1:10:33 within the AI Community uh it's quite
1:10:36 impressive that part's impressive but
1:10:39 the
1:10:39 [ __ ] have a uh having it on Discord
1:10:42 that that shit's got to go like like you
1:10:45 know six months ago
1:10:49 yeah
1:10:52 Leonardo someone needs to do well
1:10:54 mid-journey needs to do Leonardo like
1:10:57 Leonardo did for stable diffusion
1:10:59 someone needs to do that for mid-journey
1:11:00 oh yeah
1:11:03 thank you for the shades
1:11:05 I like the shades very 80s
1:11:09 I like the way you go through everyone's
1:11:11 comments in order it'll be real good
1:11:13 thank you very much most people [ __ ]
1:11:15 hate it because they're like why don't
1:11:17 you answer my quest you need to answer
1:11:19 my question I don't like that he didn't
1:11:21 answer my question why wouldn't he
1:11:22 answer my question I've asked that 37
1:11:24 times
1:11:26 I don't even think he's live do you
1:11:27 think he's live I don't think he's live
1:11:29 he's definitely not live he hasn't
1:11:31 answered any questions what questions is
1:11:33 he answering
1:11:35 he probably thinks we landed on the moon
1:11:38 you know the Earth's Hollow right
1:11:44 yeah
1:11:46 all right 136 here in Lawrence Kansas
1:11:49 yeah I the the weather right now is just
1:11:53 stupid
1:11:55 your delay is cheating the AI delays
1:11:57 analysis
1:12:02 um all this time I've been watching your
1:12:03 lives I was wondering when you would
1:12:05 break into Hawaii Five Oh Rift
1:12:13 foreign
1:12:17 it's just
1:12:18 here's one of the problems with these
1:12:20 lives is that I'm getting comfortable
1:12:21 enough now that like this is just kind
1:12:23 of how I am
1:12:25 it's probably not good like there's less
1:12:28 and less of a filter
1:12:29 it's just not good although the hearts
1:12:32 keep going apparently apparently
1:12:35 someone's enjoying the hearts I think
1:12:37 it's just the Irregulars at this point
1:12:39 are now having a game trying to get the
1:12:41 counter to tick up to try to beat last
1:12:43 night's score which is hilarious
1:12:48 um
1:12:49 20K we're almost up to 30k right now yo
1:12:53 weirdos I got pulled over in Hawaii
1:12:56 and it was really a Camaro awesome
1:12:59 uh what are your thoughts on the New
1:13:01 York Times possibly suing open AI yeah I
1:13:06 think
1:13:12 didn't
1:13:13 Google just cut a deal with the New York
1:13:15 Times
1:13:29 yeah they did 100 million dollar deal
1:13:31 okay so so here's what's going on Google
1:13:35 because Google has
1:13:37 kind of an infinite amount of money to
1:13:39 play with
1:13:42 start cutting media deals and they're
1:13:46 going to probably try to get exclusive
1:13:47 media deals that this sort of [ __ ] this
1:13:50 this sort of [ __ ] started happening
1:13:53 in like 97 and 98 with the World Wide
1:13:56 Web where everyone started cutting these
1:13:57 deals and and you know like timing had
1:14:00 this thing called Pathfinder where they
1:14:02 were trying to consolidate all their
1:14:03 brands into this thing and everyone
1:14:05 started positioning and and you know
1:14:09 MSNBC there was MSN and NBC and they
1:14:12 made MSNBC like you you probably don't
1:14:15 know that started as an internet you
1:14:17 know media relationship play there were
1:14:20 tons of these things you know AOL was
1:14:23 cutting media deals where you know this
1:14:25 AOL channel was going to be the
1:14:27 exclusive distribution channel of this
1:14:30 media Outlet
1:14:32 and so we're starting to see the
1:14:35 beginnings of that so
1:14:37 the fact that Google
1:14:39 cuts a hundred million dollars deal with
1:14:42 the New York Times
1:14:44 and then the New York Times says yeah we
1:14:46 might Sue open AI because they took our
1:14:48 stuff all of this is kind of posturing
1:14:51 for the publishing companies to cut the
1:14:54 biggest deals possible
1:14:56 um they probably don't want to do
1:14:57 exclusive deals I don't think they
1:14:59 should do exclusive deals the large
1:15:01 language model companies it's in their
1:15:03 best interest to cut deals
1:15:05 with
1:15:07 um
1:15:09 with media properties so they so that
1:15:12 they've got brands that they can Market
1:15:14 against right so chat GPT giving you a
1:15:17 really good story is great but chat GPT
1:15:21 giving you a really good story citing
1:15:24 news sources from The New York Times or
1:15:26 from the Associated Press that's even
1:15:28 better right and if that can be you know
1:15:30 jointly created and then the AP can have
1:15:33 tools for their reporters that are
1:15:34 powered by open Ai and openai can have
1:15:37 tools that remix content in all sorts of
1:15:40 unique and interesting ways in a branded
1:15:42 trusted way like everyone kind of wins
1:15:45 but at the same time
1:15:49 um the Publishers are like our [ __ ]
1:15:53 current business model doesn't work
1:15:55 anymore like everything's breaking the
1:15:57 publishing industry is a [ __ ]
1:15:59 disaster the music industry is a [ __ ]
1:16:02 disaster the film industry is a [ __ ]
1:16:04 disaster right you know the the writers
1:16:07 and actors are striking to try to cut
1:16:09 some sort of deal where their Leverage
1:16:12 is slipping away daily because they're
1:16:14 going to be able to be replicated from
1:16:16 these [ __ ] machines so so it's it's
1:16:18 kind of it it's much like the Playa for
1:16:22 burning man that right now is a giant
1:16:24 [ __ ] Lake
1:16:26 and everything's shifting everything's
1:16:29 changing and so
1:16:31 um so I think it's really interesting
1:16:34 um I think in the end all of these
1:16:35 lawsuits will settle into
1:16:38 100 million 500 million billion dollar
1:16:41 deals where we've got a 10-year deal
1:16:44 with to AI empower the blah blah blah
1:16:48 we're just going to see lots and lots
1:16:50 and lots of those
1:16:51 so we're kind of shifting
1:16:54 this is this is this is a good sign for
1:16:56 the industry in my opinion
1:16:59 because we're quickly shifting out of
1:17:01 the tools
1:17:03 and into how we use the tools right
1:17:07 so this is going to be further
1:17:09 accelerated when when
1:17:11 um Microsoft finally releases co-pilot
1:17:13 within Office 365 and off off uh Google
1:17:17 releases barred within their office
1:17:21 products and you're starting to get all
1:17:23 these media companies rolled in here
1:17:25 um
1:17:27 it's starting to be incorporated into
1:17:29 the fabric of stuff we touch every day
1:17:32 anyway
1:17:33 so what it says to me is is the
1:17:37 I don't know why the word coagulation
1:17:39 came into my head but the coagulation of
1:17:42 this AI stuff from kind of this this you
1:17:45 know vaporous like what is this
1:17:47 generative AI like I don't know how to
1:17:48 use it what I did I should I become a
1:17:51 prompt engineer I heard you can make
1:17:52 nine hundred thousand dollars oh that's
1:17:54 not in fashion anymore like what do I do
1:17:57 here
1:17:58 like it's very quickly gonna
1:18:02 I I don't think it's gonna be that long
1:18:04 where we have to actively think about
1:18:07 what is this generative AI stuff it's
1:18:10 just going to start to
1:18:12 mer merge into a lot of the stuff that
1:18:15 we play with every day now
1:18:18 I think if you get into it you're gonna
1:18:22 um potentially save your job
1:18:25 and there's going to be a lot of
1:18:26 opportunity for you in the next 18
1:18:28 months to three years
1:18:30 just by the mere fact that you have a
1:18:32 clue about this [ __ ] because a lot of
1:18:34 people are going to ignore it a lot of
1:18:35 there there were some trolls in here
1:18:37 last night that were you know you could
1:18:39 tell they were sitting on the sidelines
1:18:40 and just pontificating from their little
1:18:42 soapbox about the you know the evil ways
1:18:45 of the AI world without actually playing
1:18:48 with it
1:18:49 um I think those people are going to be
1:18:51 blindsided and so
1:18:53 the Irregulars are not going to be
1:18:55 blindsided we might not be able to
1:18:57 figure this [ __ ] out exactly what to do
1:18:59 with it but
1:19:01 I I mean the notes that I get from
1:19:03 people that are in here participating
1:19:05 regularly like what what I'm noticing is
1:19:08 a lot of people are going from I didn't
1:19:10 know anything now I'm playing with it oh
1:19:12 now I did this thing oh now I'm writing
1:19:15 grants now like like sort of the
1:19:18 acceleration of
1:19:20 um the adoption within this community
1:19:22 has been really inspiring so by the way
1:19:24 thank you for anyone who shares with me
1:19:26 what you what you're learning it it it
1:19:29 actually really helps me a lot process
1:19:31 what's happening in in all of this stuff
1:19:35 so keep doing that what's in painting So
1:19:38 in painting
1:19:40 um
1:19:41 so let me go show you
1:19:47 uh so we'll upscale one of these things
1:19:53 geez I gotta come up with a catchphrase
1:19:56 for the cowboy because I get the cowboy
1:19:57 enough I gotta have a catchphrase for
1:19:59 the cowboy
1:20:01 there's a thing called and this is this
1:20:04 is language that came from stable
1:20:05 diffusion I think it's shitty language
1:20:07 in painting and out painting out
1:20:09 painting is you're going to paint
1:20:10 outside the image and in painting is
1:20:12 you're going to paint somewhere inside
1:20:14 the image so
1:20:16 um mid-journey
1:20:19 um calls out painting zoom out which I
1:20:21 think is a much better name right so if
1:20:24 I hit zoom out right here
1:20:26 my little dude on a bike is gonna go
1:20:28 it's gonna zoom out so he'll be farther
1:20:30 in the distance and we'll see more of
1:20:32 the the world around him more of this
1:20:34 Burning Man kind of donut world this is
1:20:37 more like Donut Man
1:20:53 oh I just found a bug I just found a bug
1:20:55 in there in this in painting out
1:20:57 painting thing that's really interesting
1:21:00 huh
1:21:02 this is not good I don't think this is
1:21:04 going to give us what we want
1:21:07 and I'll tell you why in a second
1:21:15 this is gonna be Birds
1:21:17 oh no it kept the dude there
1:21:19 so it gave us a little pterodactyl off
1:21:22 in the sky all right this isn't so bad
1:21:24 so so out painting was we took that sort
1:21:28 of inside image and we out painted you
1:21:30 know we painted all this stuff on the
1:21:32 outside so now in painting is we can
1:21:34 pick some some piece of that so let's
1:21:37 grab this cute third one we'll upscale
1:21:39 this third one here with cute eyes
1:21:44 and then
1:21:47 so in painting now for Mid journey is
1:21:50 this button right here it says very
1:21:52 region so very is we're going to do a
1:21:55 variation of this but keep a lot of the
1:21:56 elements the same but the region is the
1:21:59 end painting So if I click that
1:22:01 it now gives me this interface where I
1:22:03 can select something in the image so
1:22:05 let's say that
1:22:07 um
1:22:11 let's say I want another dude on a bike
1:22:14 next to this guy so I'm going to select
1:22:17 this area right here
1:22:22 and I'm going to say
1:22:24 uh hipster
1:22:27 on
1:22:28 Cruiser
1:22:31 bike
1:22:33 and hopefully that'll give me a little
1:22:35 buddy next to our buddy
1:22:38 it's gonna give me a little buddy so the
1:22:40 end painting is is literally literally
1:22:42 in the painting
1:22:44 um
1:22:45 Photoshop calls this generative fill
1:22:48 what's interesting about generative fill
1:22:50 with Photoshop is you can either in
1:22:52 paint or out paint with generative fill
1:22:54 basically if you expand your canvas and
1:22:56 say generative fill it'll it'll out
1:22:58 paint it'll fill in the outsides or if
1:23:01 you select something inside a photo and
1:23:03 say generative fill it will do it on the
1:23:05 inside
1:23:08 and this didn't give me any bikes no
1:23:11 bikes what's this one
1:23:14 give me a chicken
1:23:17 so
1:23:20 I haven't played around with it but the
1:23:21 the bug that I think I found is
1:23:25 when I did my in painting earlier
1:23:28 for this thing I grabbed a piece of the
1:23:31 sky and I put pterodactyl flying boy I
1:23:34 spelled it wrong
1:23:37 and then when I went to zoom out on this
1:23:39 it didn't keep the original prompt it
1:23:42 just it only kept the pterodactyl flying
1:23:44 prompt so
1:23:46 which wasn't necessarily what I wanted
1:23:47 like I wanted it to zoom out on on that
1:23:50 image and just not fill anything out on
1:23:52 the outside so anyway I don't know that
1:23:54 it's a bug it's just probably more that
1:23:56 I don't understand the interface
1:23:58 which again goes back to this shouldn't
1:24:00 be in [ __ ] Discord
1:24:03 shouldn't it shouldn't it just shouldn't
1:24:08 I get why it's in Discord it's cheap
1:24:10 you've got built-in Community you can
1:24:12 test features there really easy I get
1:24:14 why it's there but it's time it's time
1:24:16 it's been two years time
1:24:19 make your [ __ ] web interface now I'm
1:24:21 sure you're developing it let's get it
1:24:23 going
1:24:24 all right okay will you give some times
1:24:28 some time oh sure bear thank you for the
1:24:31 little dangly lights I appreciate that
1:24:32 will you give some times tugas
1:24:35 to let you know
1:24:38 uh times for what Flaco C null oh that's
1:24:41 pretty funny lunish dick Luna sticks
1:24:44 here the moon comedian
1:24:45 welcome welcome welcome
1:24:50 I like Luna stick 9.5 out of 10 on the
1:24:53 name I think I you know what
1:24:55 no
1:24:56 I'm going 10 out of 10 on the name
1:24:58 because
1:25:00 it's consistent with who you are it's
1:25:02 really it's really a good representation
1:25:04 just based on our few interactions like
1:25:06 you're funny and it's it's good you're
1:25:08 like zany funny zany and funny it's a
1:25:11 good combo so I like it solid name and
1:25:15 flaccasino see it's just it's good solid
1:25:19 clever I tried to find that scam account
1:25:22 I would like to report via email okay
1:25:24 the scam account is
1:25:29 thank you for doing that by the way yes
1:25:32 please anything you can do to help the
1:25:36 real one
1:25:37 I spell learning correctly a i l earning
1:25:44 scammer a i Lolo
1:25:46 earning
1:25:48 so two l's
1:25:50 and because they're all he does them in
1:25:52 all lower case
1:25:54 um the the I and the L and the L
1:25:56 together it's like like it really is
1:25:58 almost impossible to see the difference
1:25:59 between the two when you see them all
1:26:01 with lowercase letters which is that's
1:26:03 why I made uppercase letters here
1:26:06 so that's the scammer so anyone hey if
1:26:10 I've got any if anyone in here is you
1:26:12 know trolls and trolls are hackers and
1:26:14 you want to go [ __ ] with someone go [ __ ]
1:26:16 with this guy
1:26:18 that's probably illegal for me to ask
1:26:20 that but I don't care they're stealing
1:26:22 my [ __ ] and tick tock's not dealing with
1:26:24 it I'll probably get banned now for
1:26:26 seven days because I said some word in
1:26:28 there it happens it's happened it's
1:26:31 happened three times now I think I know
1:26:33 what words I'm not allowed to say
1:26:36 they seem to be similar trigger words to
1:26:39 the the word actually what did you say I
1:26:42 have no idea what I said you've been
1:26:44 warned about insulting the algorithm I
1:26:47 know but maybe the algorithm could pay a
1:26:50 little [ __ ] attention like how hard
1:26:52 is it for the algorithm to go
1:26:54 if they copied videos from this account
1:26:57 and they're The Identical
1:26:59 checksum
1:27:01 The Identical video in that account
1:27:05 and it has the same name
1:27:08 in the same description couldn't we
1:27:10 write an algorithm to maybe go hey maybe
1:27:13 someone should go look at that I think
1:27:15 we could
1:27:16 I don't even think we'd need AI for that
1:27:21 I think we could use good old Boolean
1:27:24 logic if checksum equals
1:27:27 equals yes equals one
1:27:31 therefore
1:27:33 flag
1:27:40 nah I don't want to deal with reality I
1:27:44 don't want to deal with reality this is
1:27:47 horrible this is horrible
1:27:50 and everybody hates
1:27:54 all right
1:27:56 was that synthesized or was that human
1:27:58 we just don't know do we
1:28:00 you're uncloneable thank you do you like
1:28:04 RFK Jr I I gotta tell you man the dude's
1:28:08 voice like I I don't know what's going
1:28:09 on with his voice I haven't really
1:28:11 researched it but it's like it's just
1:28:12 hard to listen to it just sounds painful
1:28:15 um
1:28:15 I I just yeah
1:28:18 something doesn't seem on the up and up
1:28:20 there
1:28:21 um
1:28:22 so not I just don't trust the guy it
1:28:26 just doesn't feel it doesn't feel
1:28:28 authentic I in case this isn't clear
1:28:30 like one of my big things in life is is
1:28:33 like authenticity like authentic humans
1:28:38 and so politicians drive me [ __ ]
1:28:41 batshit crazy because it's like I don't
1:28:43 know how they can [ __ ] sleep at night
1:28:45 most of them most of them both sides
1:28:49 but it's gotten it's gotten so bad at
1:28:51 this point it's just like it's just
1:28:53 unwatchable so but that that one ah
1:28:56 thank you Stevo that's a cute little
1:28:57 doggy that
1:29:00 that doggy that doggy's not here
1:29:03 I'm not sure you should keep getting the
1:29:06 AI over and words any more ideas I know
1:29:09 what are you gonna do that's it the
1:29:11 world's falling apart Gen X kids yeah if
1:29:14 I ever had a polonium potato potato chip
1:29:16 sandwich of course and fried bologna and
1:29:19 potato chip come on
1:29:22 anyone had had ever had uh Taylor's
1:29:25 Taylor's pork roll
1:29:26 that's some Pennsylvania white trash
1:29:28 right there
1:29:30 fried fried Taylor's pork roll some of
1:29:34 the tastiest of the
1:29:37 of the pork part products
1:29:41 it's really good it's a small world
1:29:44 11 minute lag I think I'm the lag here
1:29:47 not the not the technology
1:29:50 hey name's Kyle Shannon I'm a lag nice
1:29:53 to meet you
1:29:55 go ahead and say something I'll respond
1:29:57 an hour later
1:30:01 I'd be fun at a party huh
1:30:05 all right
1:30:07 let's see consolidation is coming I
1:30:10 agree will Google pay content
1:30:13 trainers decent salaries
1:30:19 I don't know what you mean by content
1:30:21 trainers I don't know that
1:30:24 I think all the content scraping I I
1:30:27 here's where here's where my instinct
1:30:29 goes
1:30:32 I think that
1:30:34 the value of editors goes up
1:30:41 and I think the value of
1:30:47 AI literate writers goes up
1:30:53 but what is an AI literate writer might
1:30:56 look more like an editor
1:31:03 so
1:31:07 the the whole idea of like of of having
1:31:10 to do content training all of that's
1:31:12 just going to be automated all that's
1:31:14 going to be Bots anything that's
1:31:15 tactical execution
1:31:17 and think of think of it like this Nancy
1:31:20 anything that is
1:31:22 like the repetitive
1:31:26 you remember the the I Love Lucy famous
1:31:30 episode where she's at the
1:31:32 candy conveyor belt and she's just
1:31:34 shoving the chocolate interface and
1:31:37 gets all sick
1:31:40 um
1:31:40 any work that's like that is going to be
1:31:43 automated
1:31:45 so so any job that's kind of like Lucy
1:31:48 standing there manually pulling things
1:31:50 off the line you know manually
1:31:55 writing articles manually
1:31:58 um
1:32:02 sorting
1:32:04 alphabetizing you know like the old
1:32:06 classic you know go file these
1:32:08 alphabetically and you know
1:32:10 the woman would get the big stack of
1:32:12 file folders and have to go alphabetized
1:32:14 on like any anything that's sort of the
1:32:16 digital equivalent of that stuff
1:32:19 I think that just
1:32:20 in the next two years all of that gets
1:32:23 automated
1:32:27 and and then what feels like it goes up
1:32:30 in value is
1:32:32 who's thinking critically about how
1:32:35 we're going to use these tools
1:32:37 right who's thinking critically about
1:32:39 okay
1:32:41 if
1:32:47 there's going to be new kinds of writing
1:32:49 that emerge right there's going to be
1:32:52 fully human writing which is going to
1:32:54 have value but there's not going to be
1:32:56 as much of it
1:32:57 because it's expensive right so you'll
1:33:00 have specialized maybe investigative
1:33:02 journalism but even that's probably
1:33:04 going to be infused with a lot of AI and
1:33:06 then you're going to go all the way down
1:33:07 to like fully automated just a hundred
1:33:10 thousand articles a day being spewed up
1:33:13 by machines and there's going to be this
1:33:15 Continuum from well let's do it like
1:33:18 this way the the low level stuff the the
1:33:20 stuff that's just spewed out
1:33:21 automatically but that's going to
1:33:23 overwhelm the channels
1:33:25 and then you're going to work up to this
1:33:26 sort of rarified air of very very
1:33:28 minimal sort of handcrafted bespoke
1:33:32 writing and I'm just talking about
1:33:34 writing right now but you can take
1:33:35 literally any knowledge work discipline
1:33:37 any creative discipline
1:33:40 so there's going to be a very small
1:33:42 amount of people that still do the
1:33:44 handcrafted [ __ ] that they're probably
1:33:46 the Branded people the Maggie habermans
1:33:49 and the Matt taibis and things like that
1:33:52 that they'll still do what they do but
1:33:55 there's going to be this whole Middle
1:33:56 Ground
1:33:57 where you're gonna have sort of highly
1:34:00 respected reporters that can increase
1:34:02 the amount of writing that they do
1:34:04 because they're going to use these tools
1:34:06 and these tools are going to get better
1:34:07 at the hallucination thing they're going
1:34:09 to get better at citations you're going
1:34:11 to be able to research pieces a lot more
1:34:13 quickly
1:34:14 you're going to have somewhere in the
1:34:16 middle here
1:34:17 like local news journalism that is
1:34:20 essentially gone right because all the
1:34:22 local stuff has been bought up by these
1:34:23 conglomerates
1:34:26 um that have this National messaging so
1:34:29 local news doesn't exist anymore but
1:34:31 local news is still happening right laws
1:34:34 are still being passed you know streets
1:34:36 are still being renamed like zoning
1:34:39 permits are still being denied and
1:34:41 things like that all of those public
1:34:43 records that local journalism used to
1:34:46 report on those public records are still
1:34:48 there that that's something that's ripe
1:34:50 for AI automating
1:34:53 and say hey let's let's bring back local
1:34:56 journalism and so maybe what you do is
1:34:58 you hire an editor per local region you
1:35:02 let the AI machines write most of the
1:35:05 content but you have the editor there
1:35:06 curating which stories to run
1:35:11 so so so I it's almost impossible to
1:35:15 know
1:35:16 like what the new organization will look
1:35:19 like because I feel like every single
1:35:21 component of what they are today is
1:35:24 going to be completely Rewritten over
1:35:26 the next five years
1:35:30 so
1:35:33 your questions about salaries I think
1:35:36 that
1:35:42 I think that the salaries will go to the
1:35:45 people
1:35:47 who've got
1:35:48 critical thinking skills and our AI
1:35:50 literate right so so that potentially
1:35:53 have some sort of existing skill so I
1:35:56 was a beat reporter or I was a
1:35:58 screenwriter or you know I'm an
1:36:00 entrepreneur whatever it might be you've
1:36:02 got you know some set of skills now your
1:36:04 AI literate on top of it
1:36:06 and you've got the capacity to like
1:36:09 think strategically because this shit's
1:36:11 changing so fast that you're going to
1:36:13 need people that are going to be like
1:36:14 okay the current business looks like
1:36:17 this these new capabilities look like
1:36:19 this the future business what's that
1:36:21 look like I don't [ __ ] know but but I
1:36:23 know that that I know generally where
1:36:26 we're headed
1:36:27 and I'm gonna I'm Gonna Learn these
1:36:29 tools enough to to at least go in a
1:36:31 direction and then I'll smack that into
1:36:33 a brick wall and go nope that wasn't it
1:36:35 and I'll try another one nope that
1:36:37 wasn't it here's one oh that's really
1:36:39 interesting maybe that's a new business
1:36:40 unit
1:36:43 so I think companies are going to need
1:36:45 people they're going to need
1:36:48 you know I've used the phrase revenge of
1:36:50 the liberal arts major and revenge of
1:36:52 the Gen xers I think experienced people
1:36:54 are going to be valuable I think
1:36:57 horizontal thinkers people that can
1:36:59 think across disciplines are going to be
1:37:01 really valuable
1:37:03 and then I think anyone with a
1:37:05 specialized with it with a deep
1:37:06 specialty
1:37:08 who can now have that deep specially
1:37:11 powered by AI they're going to be
1:37:13 valuable
1:37:15 but anyone that just sits back and go
1:37:17 goes I'm a copywriter I'm just gonna I'm
1:37:19 it's all about just copywriting for me
1:37:21 and and what they're really doing is
1:37:23 writing blog posts and this and that
1:37:25 which is copywriting don't get me wrong
1:37:27 but like that's the stuff that this
1:37:29 stuff's just going to obliterate right
1:37:31 so if those people aren't kind of
1:37:32 aggressively up upskilling into whatever
1:37:35 these things can do then I think that's
1:37:38 where the that's where the pain comes in
1:37:40 Apocrypha what's happening
1:37:43 all right
1:37:44 72 percent of Americans want to slow
1:37:47 down Ai and
1:37:50 zero percent of AI companies are going
1:37:52 to slow down
1:37:55 you know I
1:37:57 that's my joke about Florida just passed
1:37:59 a fantastic law that the states no
1:38:02 longer accepting hurricanes
1:38:05 okay great nice law
1:38:09 good luck enforcing it
1:38:12 what's that dark cloud in the distance
1:38:15 yeah just [ __ ] is slowing down it's
1:38:18 not slowing down it can't there's no
1:38:19 there's no way and and Meta Meta and
1:38:22 stability by open sourcing their large
1:38:24 language models just put the nail in the
1:38:26 coffin of
1:38:28 any possibility of slowing it down if
1:38:30 this were just in Google's hands and
1:38:32 open AIS and Microsoft's
1:38:36 um
1:38:37 and a couple of the other ones you know
1:38:39 anthropic and maybe inflection
1:38:42 um cohere
1:38:43 if it were just in a handful of
1:38:45 companies and they had all the control
1:38:47 of it you have a slight chance
1:38:50 of raining it in in America
1:38:54 but China ain't gonna slow down
1:38:57 I don't know what Russia is up to but if
1:38:59 I'm sure they're doing something they're
1:39:01 not going to slow down
1:39:04 you know the Far East isn't going to
1:39:05 slow down
1:39:07 the European Union probably will like
1:39:10 though you know the European Union based
1:39:13 on what they've done with privacy is
1:39:15 going to like rain some [ __ ] in
1:39:17 but at this point you can't but you
1:39:20 can't rein it in and it's just you know
1:39:22 capitalism won't allow it right we're
1:39:24 money grubbing pigs
1:39:28 what's the the latest McKenzie the
1:39:30 McKenzie state of state of tech 2023
1:39:33 they're projecting
1:39:36 4.4.4 trillion dollars
1:39:38 in increased
1:39:42 you know
1:39:43 economy because of this [ __ ]
1:39:48 they're they're
1:39:50 golden them their Hills
1:39:53 they slowing that [ __ ] down I don't care
1:39:55 who wants to slow what that that's it's
1:39:57 part of my point of this channel is like
1:39:59 I don't like you know people ask all the
1:40:01 time what do I think of the writer's
1:40:02 strike and the the the the actors strike
1:40:05 I'm a screenwriter I have a degree in
1:40:06 acting like I I empathize with them I
1:40:09 sympathize with them like I it's a
1:40:12 [ __ ] horrible situation and
1:40:15 this shit's coming
1:40:19 like I I feel like they're fighting the
1:40:21 wrong fight like they need to be
1:40:22 fighting like
1:40:24 rather than saying we can't have ai in
1:40:27 our industry which is what they're all
1:40:29 saying right now
1:40:31 what they should be saying in my humble
1:40:33 opinion I'm not in those unions so I
1:40:35 don't have a say in the matter
1:40:37 is
1:40:38 given that this [ __ ] is coming what
1:40:42 contracts do we need need to get in
1:40:44 place now so that everyone's protected
1:40:46 and they're kind of doing that but the
1:40:48 positioning is
1:40:50 AI evil [ __ ] AI
1:40:53 they're writers this is the most
1:40:55 powerful writing tool in the history of
1:40:57 humanity and the writers are going [ __ ]
1:40:59 that [ __ ] we don't want that in our
1:41:01 lives well
1:41:03 you're going to be competing against
1:41:05 millions of people where it's Court of
1:41:08 their life very very shortly
1:41:11 so I so it just feels short-sighted it's
1:41:14 not short-sighted it feels
1:41:16 it feels myopic and it it I don't know
1:41:20 like this stuff just feels very
1:41:22 inevitable to me so
1:41:24 um
1:41:26 so I don't see it as good or bad right
1:41:28 now I just see it as there's a hurricane
1:41:30 coming
1:41:32 are you gonna be on the surfboard
1:41:36 or are you going to be in the bamboo Hut
1:41:38 on the beach
1:41:41 because the bamboo Hut's more
1:41:42 comfortable right now right
1:41:45 and those are the arguments that those
1:41:47 unions are fighting they're like but we
1:41:49 got the bamboo Hut on the beach you
1:41:50 can't [ __ ] with the bamboo Hut on the
1:41:51 beach
1:41:53 and there's other people like
1:41:56 putting on the wetsuits and paddle it
1:41:58 out
1:42:01 because there's probably some big ass
1:42:02 waves coming
1:42:06 so that's I think that's the only place
1:42:09 to be
1:42:11 but you know
1:42:13 I like chaos and change chaos generation
1:42:16 is here
1:42:18 consolidation is coming what's the
1:42:20 fastest and easiest way to change black
1:42:22 and white movie to color
1:42:24 I don't know uh
1:42:26 there's all sorts of tools for that
1:42:28 they've been doing that stuff for for
1:42:29 ages I would I would think
1:42:32 um within Premiere and Final Cut Pro
1:42:35 there are probably AI tools that have
1:42:37 probably been in it for years that do
1:42:39 that right now
1:42:42 um but you can also go go check out
1:42:44 descript
1:42:45 is a video processing AI company
1:42:49 um I don't think Vizio does anything
1:42:50 like that descript has a bunch of tools
1:42:55 that's more that's like a better
1:42:57 question for Reddit you can you can go
1:42:58 look go to go to
1:43:00 um
1:43:02 futurepedia.io and just see if there's
1:43:04 any
1:43:06 um AI generated video processing things
1:43:08 that specifically mention colorization
1:43:10 but colorization is one of those things
1:43:12 that's been around for a long long time
1:43:15 um so I I don't know that
1:43:17 um I don't know that you're going to
1:43:18 need to leverage this holy crap we're up
1:43:21 to 45 000 likes you guys are nuts
1:43:27 oh my God this is awesome
1:43:30 oh you're [ __ ] hilarious all right uh
1:43:34 it's awesome let's see AI silver
1:43:39 I'm 53 trained multimedia specialist for
1:43:42 25 years this is interesting Laura Laura
1:43:45 C AI will open the gates of everyone's
1:43:48 Minds so here's here's my thing I think
1:43:51 people like you Laura
1:43:53 that that wrap your heads or wrap your
1:43:56 head around this [ __ ] and get like
1:43:59 curious about it and explore it
1:44:02 you're gonna be so much better at things
1:44:05 like
1:44:06 um you know Runway Gen 2 like you know
1:44:09 generating stuff because you've got all
1:44:12 this multimedia experience you're going
1:44:15 to be able to look at these tools and
1:44:16 very very quickly go oh that one's good
1:44:18 that one's bad oh if I if I Stitch those
1:44:21 two together I can create this really
1:44:22 unique thing I can do it 10 times faster
1:44:25 than I used to do it
1:44:28 if I were starting an agency right now
1:44:32 so you know it's really funny like
1:44:35 having having started an agency and
1:44:37 grown an agency
1:44:43 the model
1:44:44 for agencies is basically
1:44:47 get a couple of senior people together
1:44:49 that have an idea they're generally
1:44:51 ex-agency people that are bitter about
1:44:54 the agency world and they're like [ __ ]
1:44:55 that we can do a better agency every
1:44:57 agency that starts says it says we're
1:45:01 not going to be like the other agencies
1:45:04 but but the typical model is you get a
1:45:07 couple of senior Partners who are who
1:45:09 are bitter bitter and smart right and
1:45:11 and ideally have some complementary
1:45:13 skills and have nice rolodexes right
1:45:16 because they've quit their big agency
1:45:17 jobs so they got some clients you know
1:45:20 sitting in the wings
1:45:22 and then you hire a bunch of young
1:45:23 people right you hire you know you start
1:45:26 out with five and
1:45:28 and you know you got the you got the
1:45:30 senior people out doing the selling and
1:45:32 doing the creative Direction and then
1:45:33 you have your your shop of executors
1:45:38 I don't think that's the model anymore
1:45:41 if I was starting an agency right now I
1:45:44 would you know find the two or three
1:45:47 smart bitter people that have the idea
1:45:49 that are also AI literate
1:45:52 have complementary skills
1:45:54 and then I think my team rather than
1:45:56 having young people in there that
1:45:58 because why do you have young people
1:45:59 they have energy they're trainable uh
1:46:02 and they just go for hours and you can
1:46:03 just say go do this go do that go do
1:46:05 that and they just execute and you need
1:46:06 to be able to be out there selling stuff
1:46:08 and have that [ __ ] executed right
1:46:11 well now this stuff does the execution
1:46:14 so I would much rather have someone like
1:46:16 Laura who's got a 25-year career
1:46:20 who's now into this stuff and say hey do
1:46:23 you want to come do like all the visual
1:46:24 stuff so we'll have a visual stuff
1:46:27 person we'll have a writing stuff person
1:46:29 you know we'll have a project managing
1:46:32 kind of person
1:46:33 you know we'll have a creative director
1:46:35 or two we might have a typical creative
1:46:36 director pair right words and words and
1:46:39 pictures
1:46:42 and I think it would be a bunch of gen
1:46:43 xers
1:46:45 that are that are still hungry right
1:46:47 they got if it's a startup they gotta be
1:46:49 hungry so so that's a bit of a unicorn
1:46:51 so they might not quite be gen xers
1:46:53 they'll probably be Elder Millennials
1:46:54 it'll be either Elder Millennials or gen
1:46:57 xers
1:46:58 that have experience smarts you know
1:47:02 some kind of specialty like that and are
1:47:04 diving into this [ __ ] and you can start
1:47:07 a five to ten person agency that you
1:47:09 could compete with 50 person agencies
1:47:11 today
1:47:14 there's a job there's there's a
1:47:17 trying to figure out what to do next and
1:47:19 and you got some got some years on you
1:47:21 and some skills and you're into this
1:47:23 [ __ ] like Laura is
1:47:28 gonna be in great demand great demand I
1:47:32 think I can't see any other way if I'm a
1:47:35 big Fortune 500 company and the bosses
1:47:38 said [ __ ] figure AI out figure it out
1:47:41 now
1:47:43 what am I gonna do
1:47:45 I'm gonna go find
1:47:47 some some
1:47:48 smart people that are that have been on
1:47:52 the planet long enough that
1:47:55 can help short-circuit some of the
1:47:57 changes that are coming like I I just
1:47:59 can't see a way around it the thing that
1:48:01 scares me in that scenario like I'm not
1:48:03 saying this is a good thing I'm just
1:48:04 saying like this is kind of reality of
1:48:06 where my head is right now
1:48:09 where's entry level
1:48:13 because entry level used to be you'd
1:48:16 come in and you'd do the repetitive
1:48:17 tactical execution work and that's how
1:48:19 you learn the business
1:48:22 you know go make 50 Instagram posts out
1:48:24 of this template okay
1:48:27 [Music]
1:48:29 you know and after [ __ ] three months
1:48:31 of Mind numbing Instagram image
1:48:33 generation you're like could I do an
1:48:36 audio file oh you can do audio yeah oh
1:48:39 go do these thousand audio files thanks
1:48:43 six months later
1:48:45 I do video too
1:48:48 right that's how young people learn
1:48:51 that's how businesses
1:48:53 service businesses get the work done
1:48:57 where do those young people go where do
1:49:00 they enter
1:49:02 if this shit's doing all that work that
1:49:04 I don't know like that I don't have a
1:49:06 [ __ ] answer for and like the only the
1:49:08 only plausible
1:49:10 thing that I came up with was
1:49:13 maybe you go back to a mentor mentee
1:49:16 model
1:49:17 where you bring in a young person and
1:49:19 pair them up with an old person right so
1:49:21 so Laura's got her 25 years of
1:49:24 multimedia experience
1:49:26 so she's doing everything she can to
1:49:28 learn this you pair her up with a young
1:49:30 person who knows all the lingo and the
1:49:34 vocabulary and how to [ __ ] use tick
1:49:36 tock
1:49:37 and you pair them up
1:49:40 and they teach each other
1:49:44 how to use these tools to do what they
1:49:46 do and and sort of the the young person
1:49:49 kind of levels up as an apprentice to
1:49:52 the older person that's you know using
1:49:54 their wisdom and their years of
1:49:55 experience to curate the [ __ ] out of
1:49:57 this stuff
1:49:58 that's the only thing I can think of
1:50:00 right now I just I feel like the entry
1:50:02 the like the entry level is going to be
1:50:05 automated and where do the young people
1:50:07 go I don't know I don't know that that
1:50:09 freaks me out a little bit
1:50:12 but anyway I don't know did you have a
1:50:14 question there wait what did you say
1:50:15 um train multimedia AI will open the
1:50:17 gates of everyone's Minds oh yeah that's
1:50:19 the other thing I was going to think
1:50:20 that that your comment um made me think
1:50:23 of Laura is
1:50:26 um
1:50:30 so many people are going to be able to
1:50:32 completely reinvent themselves
1:50:35 where maybe I was just a copywriter for
1:50:37 all those years
1:50:39 but I really wanted to be an artist but
1:50:41 I just didn't have the talent for art
1:50:43 but now there's this tool where I can
1:50:45 take my word skills and I can use them
1:50:47 in this tool and it makes pretty
1:50:48 pictures
1:50:50 well that could just be a complete
1:50:52 career transformation right or I've
1:50:55 always wanted to code but I have ADHD
1:50:57 but if these can do the basic coding and
1:50:59 I can just be like the product guy
1:51:03 and actually generate applications
1:51:06 uh I'm in
1:51:08 that's like a superpower right so so I
1:51:11 think that idea of opening the gates of
1:51:13 everyone's Minds it's kind of like
1:51:16 I I almost feel like these things are
1:51:18 Malcolm gladwell's 10 000 hour Theory
1:51:21 right that that in order to get
1:51:23 competent at something you need to put
1:51:24 in your 10 000 hours I kind of feel like
1:51:27 these things knock it down to like a
1:51:29 thousand maybe even a hundred
1:51:32 what if you put in a hundred hours into
1:51:35 these tools in something that you were
1:51:38 horrible at in a hundred hours could you
1:51:40 be good probably
1:51:43 that's got some pretty serious [ __ ]
1:51:45 implications
1:51:48 all right
1:51:50 I got a windiest uh gotta wind it up
1:51:54 all right let's see let me I'm gonna
1:51:56 scroll to the bottom here I gotta go
1:51:58 um I got a trip to the airport in the
1:52:00 morning
1:52:02 so let me go do that let me go
1:52:06 I'm gonna grab some final questions down
1:52:08 at the bottom here
1:52:12 um how do you see it affecting the
1:52:13 medical industry
1:52:14 um I think it's going to be fairly
1:52:16 profound but it'll be slower
1:52:19 um I think it's going to affect drug
1:52:21 Discovery like incredibly well and
1:52:23 Incredibly quickly
1:52:26 um I think it's gonna impact the
1:52:29 marketing side of things pretty
1:52:30 dramatically but you know all because
1:52:32 it's regulated all of that stuff's gonna
1:52:34 go slow so so we've been talking for
1:52:36 years about Radiologists you know we
1:52:39 don't need Radiologists anymore because
1:52:41 these tools are going to do it better
1:52:43 than the Radiologists and they do
1:52:45 accept
1:52:47 we as a society and and as insurance
1:52:50 companies and as medical professionals
1:52:53 aren't comfortable completely seeding
1:52:56 diagnosis to the machines so you're
1:52:59 going to have pair programming for a
1:53:01 while right where you've got
1:53:03 doctor Radiologists in in conjunction
1:53:06 with this AI stuff right so you'll have
1:53:07 that for a while
1:53:09 um I think this generative AI stuff is
1:53:11 going to accelerate a lot of things
1:53:13 um so so I think it's going to be quite
1:53:14 quite profound
1:53:16 um but it but you know we've had AI in
1:53:18 medicine for a lot of years it's done a
1:53:20 lot of good things
1:53:21 um the generative AI stuff is just it's
1:53:24 it's a different kind of contribution to
1:53:27 the field and I think it's going to
1:53:29 affect not just drug Discovery but sort
1:53:32 of all aspects of it and it's
1:53:36 this stuff is going to affect all
1:53:38 aspects of anything that's computer
1:53:40 mediated so Consulting
1:53:43 computer software development Medical
1:53:49 real estate all of it all of it's going
1:53:51 to be affected anywhere there's computer
1:53:52 Intervention which is essentially all of
1:53:55 our society at least here in the West
1:53:58 I'd work for you oh thank you Luna stick
1:54:00 I appreciate that
1:54:02 we are young baby boomers
1:54:05 I'm 65 I think I'm I think I'm the sort
1:54:09 of Base year of gen xers
1:54:11 my daughter's in her second year of
1:54:13 Masters in advertising and they're not
1:54:15 doing any of this stuff oh Vicky oh God
1:54:19 that's that's rough
1:54:23 if you haven't tell your daughter to
1:54:27 [ __ ] dive head first into this stuff
1:54:30 unless she's pigheaded and won't listen
1:54:33 to Mama hopefully she listens tomorrow
1:54:35 tell her to dive into this feel free to
1:54:38 point her my direction I'm happy to chat
1:54:40 with her
1:54:41 um I've been in that world for years uh
1:54:43 the the agency world is about to have an
1:54:47 upending it's it's not going to be
1:54:48 pretty
1:54:49 and if if if she's getting her Masters
1:54:52 and they're not dealing with this she's
1:54:53 [ __ ] Christ really
1:54:55 huh that's maddening 50k coming up we're
1:54:59 at 52k holy crap all right I gotta jump
1:55:01 to the bottom here I gotta go
1:55:04 um 65 you were 58 last week no I was
1:55:06 born in 65. I am 58.
1:55:10 I was born in 65 1965 demo test well bye
1:55:16 I was demo testing here oh there he is
1:55:19 you know nothing about AI there's our
1:55:21 troll
1:55:22 oh he's so cute he needs a hug oh demo
1:55:27 here's the sweetest
1:55:31 are we stopping at 50 50k we're crushing
1:55:34 it you guys are having your your own
1:55:36 little party off to the side here this
1:55:38 is awesome
1:55:39 um all right let's see I'll do two more
1:55:41 questions what model would you suggest
1:55:42 for privacy for example analyzing
1:55:45 sensitive documents such as PDFs
1:55:48 um
1:55:48 so you're going to want to look at a
1:55:51 project called Lang chain
1:55:52 um what's your name Patrice Patrice
1:55:54 rock83
1:55:56 look at a project called Lang chain
1:55:57 where you can
1:55:59 um embed your own PDFs and you can you
1:56:02 can kind of run everything locally on a
1:56:04 virtual machine where you're not sending
1:56:05 API calls out to to these services
1:56:09 um I would I would start there you can
1:56:11 also play Within chat GPT code
1:56:14 interpreter
1:56:15 I don't quite know the data boundaries
1:56:18 on it but the way code interpreter works
1:56:20 it spins up a virtual machine inside
1:56:22 your chat
1:56:23 so you can upload a PDF into code
1:56:26 interpreter and then have it start
1:56:28 interacting with it and it's writing and
1:56:30 executing python code to analyze your
1:56:32 PDF
1:56:34 um
1:56:35 that might be secure I would explore
1:56:38 code interpreter but
1:56:39 the the
1:56:41 data sharing policies of openai are a
1:56:44 little shitty right now they got to get
1:56:45 their [ __ ] together with that the other
1:56:47 thing you can do is look at uh Microsoft
1:56:49 Azure
1:56:51 um they've got GPT things if you're
1:56:53 working for a big company that has
1:56:55 budget and you can spin up a secure you
1:56:59 know data
1:57:01 sequestered server to do this open AI
1:57:04 stuff I know they have them there but I
1:57:06 think they start at like 25 Grand so you
1:57:08 know they're not cheap but depending on
1:57:11 who you work for that might be the way
1:57:13 to go if you're working Healthcare and
1:57:14 Pharma you know absolutely gonna have to
1:57:17 do that
1:57:20 10 p.m for anyone with an early morning
1:57:22 flight yeah it's I I gotta I gotta get
1:57:25 to the airport early
1:57:27 what do you think AI will do to a to
1:57:30 education in colleges cam I'm going to
1:57:32 save that one for next time but
1:57:34 basically I think educate the entire
1:57:36 educational
1:57:38 system needs to reinvent itself and this
1:57:40 is going to force it to do it it's
1:57:42 Everything Has Changed like what what we
1:57:46 teach to has to change because the world
1:57:50 that we're moving into doesn't require
1:57:53 the same kind of education it doesn't
1:57:55 require memorization it doesn't require
1:57:57 rote activity practicing
1:58:02 to go be a cog in the machine
1:58:05 because these are going to be the cogs
1:58:08 in the machine now
1:58:09 so
1:58:13 teaching critical thinking skills I
1:58:16 think has to go up dramatically and
1:58:19 teaching rote memorization and
1:58:21 repetition [ __ ] has to go down
1:58:24 immediately and and I don't think any
1:58:27 institutions are really ready for that
1:58:29 so
1:58:31 I don't know enough about that world to
1:58:33 know how to [ __ ] fix it but that's
1:58:35 that's my instinct anyway I gotta go
1:58:36 people peace out everyone uh it was good
1:58:39 hanging out with you and uh I really do
1:58:43 appreciate this we
1:58:45 we got the 54 100 likes you guys are
1:58:48 nuts I love you this is fantastic thank
1:58:52 you all peace chaos generation Clint
1:58:55 cavruno Vicky night everybody Apocrypha
1:58:59 thank you so much I'll see you tomorrow
1:59:01 peace out Emilio's wife later night
1:59:04 Nancy felonious Luna stick
1:59:08 the bear all right peace out everybody
1:59:11 thanks our growth geek