aiLL Nov 26, 2023 - Mastering AI Tools: How Anyone Can Unlock Creativity

Description

Chapters

Transcript

0:04 [Music]
0:08 champ are you going to
0:13 [Music]
0:26 [Applause]
0:29 sing
0:32 [Music]
0:45 I think you are first forever hook no no
0:47 you were fifth fourth or fifth you were
0:52 close hey everybody Welcome champy says
0:57 hi champy says hello
1:00 happy Sunday evening Dr J good to see
1:08 you yay I'm on time Clint
1:13 Copperhead present Turtles llas
1:18 hedgehogs do what ye do ye fellow
1:23 Irregulars good day good day good day
1:28 Christopher thank you for the Roes the
1:32 roses are
1:34 fantastic are those your car glasses no
1:37 these are my these are my computer
1:39 glasses these are my car
1:43 glasses like a 60s I look like a like a
1:46 well without the
1:48 hair these are kind of like 60s
1:51 uh gmen
1:58 glasses what's happening to BU
2:02 okay got my phone for Tik Tok iPad for
2:04 Discord trying to get Mighty networks on
2:07 my iPad too I know it's a lot right now
2:10 we'll we'll we'll simplify we have like
2:13 another week or so before we you know
2:15 officially switch over so still trying
2:18 to figure Mighty networks out welcome
2:21 everybody welcome welcome welcome
2:23 welcome to the AI learning lab encourage
2:26 more viewers to share okay hey viewers
2:29 would you share this would you share
2:31 this live and particularly if if you can
2:35 share it with friends or family that you
2:37 know it would
2:38 bug that you're watching a Channel about
2:42 AI that would be good let's get some
2:44 people in here we'll give them their
2:45 first Kevin mallister
2:48 moment
2:50 um I don't I I've got I've got zero
2:53 agenda tonight I mean there's there's a
2:55 lot going on there's a lot of lot of
2:57 things rolling but um but I think it's
3:00 just
3:02 uh I don't know I think it's
3:05 just just cranking along like I feel I
3:07 feel like last week was so dramatic with
3:10 the open AI
3:12 nonsense grock is happening wait I'm
3:15 just not paying for it I hear good
3:16 things yeah I heard that too but I do
3:19 not seem to have access to it let me
3:22 see unless I
3:25 do why is this why do I not
3:28 have there we go
3:32 I got no grock I got no grock on the
3:34 left hand side let me
3:36 reload I'm paying the 16
3:39 bucks verified orgs communities messages
3:43 I got nothing people I got
3:50 nothing so we'll say we'll
3:54 say what was that was oh it was Napoleon
3:58 Dynamite the the brother or the cousin
4:01 he talk like
4:03 this with his lower teeth out I thought
4:06 that was the the funniest acting
4:09 Choice what's up
4:12 Napoleon finally having about finally
4:15 having meeting about an AI pilot project
4:17 tomorrow at work brainstorming ideas oh
4:19 that's a awesome
4:21 kruno man it's it's so cool like you've
4:23 gone from like you know secret cyborg to
4:26 like thought leader in your company in
4:28 in couple of months that's very
4:31 cool as they say in the military no joy
4:34 want to explore Google's
4:36 duet
4:38 um possibly
4:41 except I don't think I have access to it
4:45 because I have Google
4:50 workspace and ironically if you pay for
4:53 Google you don't get this [ __ ] although
4:56 maybe they've added it
4:58 docs let me fire up up a Google doc here
5:01 and see if I got
5:07 it hello
5:13 um I don't think I got
5:16 it
5:18 tools
5:21 no why am I not seeing the
5:28 suggestion
5:32 I don't know what does grock mean okay
5:35 so so welcome everyone my name is Kyle
5:37 Shannon this is the AI learning lab we
5:39 will learn about anything AI anything
5:42 you want to talk about pop it in the
5:44 comments below um I'll do my best to get
5:46 everyone up to speed we'll do some
5:48 little level setting here and then
5:51 um and then we can dig in and and I
5:54 don't know do whatever you want to do we
5:56 can we can do some demos we can I can
5:58 answer questions grock is um the new
6:03 Twitter
6:05 um x.
6:08 a which is
6:11 X's AI arm so X used to be Twitter and
6:16 grock is their new chat GPT
6:19 competitor and it's supposed to be
6:21 launching widely to Twitter premium
6:26 members so basically if you're paying 16
6:29 bucks a month for
6:30 Twitter which I just bit my lower lip
6:33 and and paid it because I want to try it
6:36 um it's supposed to be coming out this
6:37 week the the early uh early responses
6:41 are good what's what's interesting about
6:43 it is it incorp well a it's part of
6:46 Twitter right it's part part of X So
6:49 within within your interface you've got
6:51 this tool and it incorporates X data in
6:55 its answers um so yeah yeah that's
7:01 that that's what it is so so we we
7:04 should see it's also got um it's
7:06 supposed to be not biased it's supposed
7:09 to be there's a funny mode where it'll
7:11 just be a smartass um which I think is
7:13 kind of
7:16 funny what did it say it
7:19 said someone Lex fredman asked it asked
7:23 grock about Elon Musk and said you know
7:26 is Elon Musk wrong about this and it
7:29 said oh the same Elon Musk that was
7:31 wrong about something and it said he's
7:33 wrong more than a wrong Omer at a wrong
7:37 [Laughter]
7:40 Factory oh
7:43 man Sher D CER Rose thank you very
7:48 much unbiased I doubt it I know all
7:51 these things are going to be some level
7:53 of bias depending on what they're
7:54 trained on hey Robert
7:58 Rossy
8:04 [Music]
8:06 trained on tweets it's going to be so
8:08 confused you can get more misinformation
8:15 faster yeah grock where where truth is
8:20 [Laughter]
8:21 relative we'll see we'll
8:25 see I mean like like all of these things
8:28 there's always compromises right he
8:29 wants things to be open and completely
8:32 fair and biased it probably means it's
8:33 going to there's going to be some
8:34 downside to that some unintended
8:36 consequence to that so so we'll see
8:40 we'll see how it goes all right so if
8:42 you've got uh any questions or comments
8:45 Kyle are you real or AI I'm gp5 nice to
8:48 meet
8:51 you uh I think I'm real um I do my best
8:58 I do my best to be
9:00 authentic um I can show you uh I can
9:04 show you the fake version of me here you
9:06 want to see this is funny this is funny
9:10 this is comedy this is
9:13 Haha where is it let's see we'll do it
9:16 here we go there and we'll go there and
9:18 we'll go there and go
9:22 there so I'll show you two versions of
9:25 me one is it's my real voice
9:30 but it's not real video this one hey
9:33 it's Kyle Shannon with the AI learning
9:35 lab so the previous video on my timeline
9:39 is an avatar generated with haen where I
9:42 didn't say the words I just typed in a
9:45 script and it synthesized my voice and
9:48 it was decent but it didn't really sound
9:50 like me this so that's that's my
9:53 recorded voice but but faked video and
9:57 this is not even my recorded voice
10:01 where is it hang
10:02 on hey hey it's Kyle Shannon with the AI
10:05 learning lab you know there are some
10:07 days I just don't want to turn on the
10:09 camera days where I don't even want to
10:10 get out of bed well good news we now
10:14 live in a so there's fake me so so we
10:17 can have fake me we can have all the
10:19 fake me you want um Jim Ross any chat
10:23 GPT birthday plans uh yeah I'm just
10:26 going to I I'm just going to go I just
10:29 I'm going to keep going live and
10:30 whenever I need to take a break I'm
10:32 going to take a break so I think I'm
10:33 going to start in the morning and then
10:35 I'll you know I'll do work [ __ ] then
10:37 I'll jump back into the live then I'll
10:39 jump off then I'll jump back into the
10:41 live so wholesome
10:44 exactly a post sci-fi world and I can
10:48 just type up this script and select my
10:50 avatar I can just type up this script
10:54 and yeah crazy crazy crazy crazy your
10:59 avatar sounds more realistic
11:02 exactly my version of you for $35 is
11:07 just as
11:10 good fake you is more handsome hey now
11:14 hey
11:15 now Kevin
11:18 Vasquez all right uh it's supposed to be
11:23 a leaked paper so qar and and what was
11:26 the other one that you asked about cam
11:29 the
11:31 uh Orca I don't know orca orca 2 let me
11:36 just go quickly do a quick search on it
11:37 see what it
11:39 is uh I can tell you what qar is or what
11:42 I what I understand of
11:44 qar um Orca
11:49 2 Microsoft releases Orca 2 a pair of
11:53 small language models oh yeah
11:57 so and these aren't these didn't the
11:59 these things supposed to didn't they
12:01 show up like better than GPT 3.5 or
12:04 something like that orcu 13
12:07 billion Microsoft's AI model that
12:09 challenges chat GPT yeah satian adala
12:13 talked about this he hinted at
12:16 this a th000 plus pre-built AI apps for
12:19 any use case
12:21 interesting what is Ora 2 at its core
12:24 it's a testament to Microsoft's
12:26 commitment he sort of talked about this
12:28 when they were like wow what's going to
12:29 happen if open AI screws up he's like
12:31 well we're into large language models
12:32 we're into small language models which
12:34 is what this is what what's the big deal
12:37 about a small language
12:39 model is that as they as they optimize
12:43 these as as a combination of chips
12:46 getting faster and these things getting
12:49 smaller and more optimized is you're
12:51 going to be able to
12:52 run GPT 3.5 level large language models
12:57 locally on device right right so on your
13:00 phone on your
13:02 computer and they won't have to go out
13:04 to the Internet so they'll be faster
13:06 more responsive they'll be there'll be
13:08 there won't be L latency issues and
13:10 there won't be privacy issues because
13:11 you'll be running things
13:13 locally so that feels like a pretty big
13:15 deal to me although I didn't this the
13:17 first time I'm really seeing it like I I
13:19 knew I knew it was coming and I saw a
13:20 tweet about it today that this thing
13:22 apparently performed quite well um so I
13:25 don't know about that
13:26 qar um is
13:30 the
13:31 rumored
13:33 um the
13:35 rumored break through from CH from open
13:40 AI where they say well maybe we hit AGI
13:44 um but they've been pretty tight lipped
13:46 about it since all the all the debacle I
13:48 don't think they want more drama right
13:50 now let's see Q
13:57 star see if there's any anything
13:59 interesting
14:03 here a lot of thirst traps trying to
14:06 take advantage of people searching for
14:12 qar but whatever let's see qar open AI
14:16 maybe that'll help
14:17 with oops not
14:28 Zi
14:34 yeah I don't think there's anything new
14:36 so so basically what qar is is it's a
14:39 it's
14:41 a it's a an algorithm let's
14:47 see it sits on top of a large language
14:50 model and helps it basically do
14:52 reasoning and they did reasoning on very
14:56 simple mathematical equations but the
14:59 Breakthrough was apparently that it got
15:01 them 100% right and math is very
15:04 different
15:05 than um than large language than
15:09 language um because it requires
15:11 reasoning and you know running down
15:13 different paths and doing that until you
15:16 get it right and apparently that's what
15:19 they've achieved so that's what qar is
15:21 in in a nutshell what are your thoughts
15:23 on artificial super intelligence well if
15:26 we if we get to AGI so so there's
15:29 artificial general intelligence where um
15:32 it's essentially that
15:35 the the large language model becomes an
15:38 autonomous system that performs better
15:41 than most humans or or better than
15:44 humans at most economically valuable
15:47 tasks that's the open AI Definition of
15:50 artificial general intelligence so
15:53 basically when when when the when when
15:57 chat GPT can do your job better than
15:59 than you can and can do essentially any
16:01 knowledge work job better than humans
16:05 artificial super intelligence is where
16:07 these things are smarter than the all of
16:10 us and you
16:12 know I I don't think we're close to
16:15 artificial super intelligence but the
16:17 minute you get to artificial general
16:19 intelligence and if you have
16:22 the um you use the AGI to improve itself
16:28 then you get to artificial super
16:30 intelligence very
16:32 quickly Kyle um what is Apple doing to
16:35 stay competitive with AI you mentioned
16:37 slm being able to do something that the
16:41 pinned comment just went away uh so if
16:44 you could repin that um so so
16:50 Apple Apple's doing what Apple does wait
16:54 you okay Apple uh hang on what's it say
16:59 slm being able to run small language
17:01 model being able to run on devices Apple
17:04 have anything in the works um so I I
17:08 don't know any more than anyone else but
17:10 just paying attention to the The Vision
17:12 Pro launch and the Apple
17:16 um iPhone 15 launch and the recent M3
17:20 launch of the the M M3
17:23 chip all three of those um
17:27 launches
17:31 the first two didn't mention artificial
17:33 intelligence at all but they did mention
17:35 machine learning and inferences and
17:37 embeddings and things like
17:40 that um and in The Vision Pro goggles
17:43 they've got this chip called the R1 I I
17:46 assume it stands for reality one um but
17:49 it's it's it's doing machine learning
17:52 processing on all the sensors in the
17:55 device in real time um and then the they
17:59 mentioned both the iPhone 15 and the M3
18:02 have advanced machine learning
18:04 capabilities so I think apple is um
18:09 building the foundation right now from a
18:11 hardware perspective um they've
18:14 apparently put they they spent in 2023
18:18 their R&D budget was $22.4 billion and
18:22 apparently a lot of that was on on AI I
18:25 feel like I need to do
18:27 homework I know it should
18:29 that's that's kind of what this is
18:31 feeling like right now because it's I
18:33 I'm right there with you I feel like
18:35 there is so much changing so fast I
18:40 don't I don't even know like I kind of
18:43 know how to use chat GPT right now but I
18:48 don't really like it the fact that it's
18:51 gone fully multimodal where it's got
18:53 vision and it can generate images and it
18:55 can hear and it can talk and it can do
18:57 all this there's so many ways to use it
18:59 right now not even dealing with like the
19:01 developer side of this where there's
19:03 there's a whole other world of what you
19:05 can do with their
19:06 API and so I I I feel like I'm right
19:09 there with you that I I spent the last
19:11 year learning a tool that is so
19:13 fundamentally different now that I don't
19:15 know I don't know I don't know how to
19:17 use it um so I feel like a bit of a
19:21 fraud
19:23 here uh so and it does it kind of feels
19:26 like I got to do my homework I gotta
19:27 learn my [ __ ]
19:29 um knowing that that we're about to you
19:33 know it's going to change again and it's
19:34 going to change again and then Google's
19:37 going to come out with Gemini and you
19:39 know grocka is coming out it's going to
19:40 behave slightly differently so so it is
19:43 this is very much a curiosity hamster
19:47 wheel there's one for you are regulars
19:49 to go make something with the Curiosity
19:51 hamster wheel is Agi hold up based on
19:55 RAM GPU speed board Hardware thickness
20:00 [Laughter]
20:05 space now I think AGI is just based on
20:10 the current current large language
20:12 models are really kind of dumb they're
20:14 just
20:15 dumb um calculators they're probability
20:18 engines that just do it over and over
20:20 again they're not aware of what they
20:22 make it sounds like what they did with
20:24 this qar thing is they said okay you got
20:26 a large language model down here that
20:28 spits out a bunch of [ __ ] it's not aware
20:29 of so why don't we build a system up
20:32 here that can take the output of the
20:36 actions that were taken and score them
20:39 did it do good or did it do bad and if
20:41 it did good give it a higher score and
20:45 if it did bad give it a lower score and
20:47 let's keep track of that score over time
20:49 and so what it allows the large language
20:52 model to do is kind of self-navigate
20:54 toward the right
20:55 solution and that that seems to have
20:58 changed everything it but you know I
21:00 don't know but we don't it's all
21:02 speculation at this point
21:04 there's there's a lot of smart people
21:07 out there talking about it um so I am
21:10 just
21:11 uh this channel is more like chat
21:15 TMZ than news or
21:18 science I'm just sort of paring what
21:20 they're saying at this point we'll see
21:22 can AGI create deliverables or just Aid
21:25 in the creation AGI can create
21:27 deliverables AGI can also take actions
21:30 and AGI is
21:32 autonomous so so right now there's hints
21:36 of autonomous Behavior within chat GPT
21:39 if you watch what it does with code
21:41 interpreter where you give it a goal and
21:43 it writes python code and executes
21:46 python code to hit your goal that's
21:50 that's autonomous Behavior but in a very
21:52 sort of contained little thing what will
21:55 happen with AGI is the system will be
21:58 able to take actions on your behalf in
22:02 ways that it determines that it needs to
22:05 do you give it a goal and it will just
22:07 go off and start doing things and taking
22:09 actions and um do it until it gets there
22:13 Kyle play your Funky Music Play That
22:16 Funky Music W
22:22 boy oh champy Apple's interface chips
22:27 are still a bit of an enigma correct uh
22:30 yeah the or inference
22:33 chips yeah Apple's been fairly tight
22:35 lipped what's what what's what Apple's
22:37 going to do is what Apple does
22:40 is when they're ready to launch AI they
22:43 probably won't call it AI they'll
22:45 probably it'll probably be something
22:47 around assistants or agents you know
22:51 something on your
22:53 behalf
22:56 um and and you know know they'll they'll
22:59 probably come out and try to define a
23:01 new sector that that is kind of consumer
23:04 grade AI right I would argue that chat
23:07 GPT and grock and pi and Po and all of
23:11 them are not consumer grade they're easy
23:15 but they're but they're not consumer
23:16 grade right you still have to do
23:18 prompting and you have to do prompt
23:20 engineering you have to figure out what
23:22 plugins are and then you have to figure
23:23 out what code interpreter is and then
23:25 you you got to go use zapier if you want
23:28 to do automations and it's just
23:30 everything right now is just duck spit
23:32 and duct tape janky ass pieces of
23:35 software apple is not going to come out
23:37 with that so when Apple comes out
23:38 there's is going to be tight and
23:40 buttoned up and it will probably be like
23:42 a lot of Apple products more like a
23:45 Walled Garden where you don't have as
23:47 much control and all the Geeks will go
23:49 oh that's not really interesting cuz
23:50 they're they're you know hiding all the
23:52 controls behind that simple little
23:54 interface that cute little
23:57 Avatar
23:59 and then you know millions of people use
24:00 it because you know they don't have to
24:03 figure out [ __ ]
24:04 [Laughter]
24:10 math
24:15 ah acceleration versus
24:18 incrementalism yeah I think we're I
24:20 think
24:21 2024 I think 2024 we're going to we're
24:24 going to we're going to figure out the
24:26 difference between those two Chris um
24:29 being away from AI for 2 weeks starting
24:32 to feel like a caveman Edgar I've been
24:35 doing this every day I took one day off
24:37 on Thursday for
24:39 Thanksgiving and I still I still feel
24:41 like I'm not caught up I I feel like I'm
24:44 I'm really
24:47 behind but I don't think I mean I mean
24:50 most of the most of what people are
24:53 talking about right now is a lot of
24:54 speculative kind of crap so but I just
24:57 have this this overhang that something
24:59 big is about to change and and so it
25:03 probably won't like it's like it's we're
25:08 probably I don't know I mean it's
25:11 possible that we get the GPT store this
25:13 week hamster wheel Show and Tell
25:17 okay wait which Show and Tell Mighty
25:19 networks or or the
25:24 [Music]
25:27 other
25:31 hamster
25:35 wheel that's
25:41 awesome from B Be not Afraid the
25:44 Curiosity hamster
25:45 [Laughter]
25:54 wheel oh so so where these are thank you
25:58 Drew that's really good
26:00 um these are awesome
26:04 um so these are on the AI Salon Mighty
26:07 networks which is our new home for the
26:09 AI salon I can talk about the AI Salon
26:12 maybe what I'll do here is let let me
26:13 let me do some if you're new to Ai and
26:17 you're you're trying to figure this
26:18 stuff out if you haven't really played
26:19 with chat GPT let me let me do a little
26:21 bit of level level
26:24 [Music]
26:27 setting
26:29 okay so ignore the thing on the right
26:32 but Kyle what's behind the thing on the
26:34 right your head's in the way what are
26:35 you doing your head's in the way Apple
26:37 personal genius yeah it'll be something
26:40 like
26:41 that uh I'm here to let people in okay
26:43 great thanks Vicki that's awesome
26:48 um like playing C ketchup in a hamster
26:51 wheel yeah exactly that's exactly what
26:53 it feels like Robert um if you have not
26:57 tried chat GPT I feel like I'm about to
26:59 sneeze so I
27:03 may maybe I don't know let me drink some
27:06 water hold
27:11 on okay if you've not played with chat
27:14 GPT go there right now go to chat.
27:18 open.com and play with that what what
27:21 you'll get access to is chat GPT 3.5
27:24 which is what
27:25 launched a year ago on November 30th so
27:29 today is the 26th so 4 days from now it
27:32 will have been a
27:33 year and the thing that is at chat.
27:37 open.com if you just use the free
27:39 version is is chat GPT
27:42 3.5 what they launched over the summer
27:46 and then dramatically improved over the
27:48 last month GPT 4 is just a holy
27:52 different thing um they're not allowing
27:55 new people to sign up for it yet but I
27:57 would strong strongly encourage you to
27:59 get on the weit list um it's pretty
28:03 remarkable and I can show some of that
28:04 stuff um so so that's that the the other
28:09 thing that you should do if you want to
28:11 get curious about this stuff is start
28:14 understanding that there's other tools
28:16 that are like chat GPT that are not chat
28:18 GPT um
28:22 Claude is probably their biggest
28:24 competitor it's a company called
28:26 anthropic it's a really good tool um
28:29 co-pilot on the lower right there that's
28:31 Microsoft's version of chat GPT 4 um
28:35 it's free and in the absence of having
28:39 access to chat gp4 from open AI you can
28:41 use co-pilot then you've got Gro which
28:44 is from from X from Elon musk's from
28:47 Twitter um that's coming out this week
28:49 and then you got Bard from Google and
28:52 that's been steadily improving since the
28:54 summer it got kind of usable probably in
28:58 July and it's been getting better ever
28:59 since and it's supposed to get
29:00 dramatically better um so go play with
29:04 those and then I would also encourage
29:06 you to go play with
29:07 these These are four other sort of chat
29:10 GPT like things there's Pi which is very
29:13 conversational get the IOS app and put
29:15 it in conversation mode and talk to it
29:18 it's remarkable it it's connected to the
29:21 internet you can talk to it perplexity
29:23 if you do any kind of research is really
29:25 good for that people are swearing by
29:27 that
29:28 um llama 2 is a whole world of
29:30 open-source large language models um
29:34 that that metap put out and the whole
29:37 developer Community is kind of pouncing
29:39 all over those so that's a whole world
29:40 if you want to geek out on open source
29:43 and running things locally there's
29:45 there's just a whole world there to
29:46 explore and then po lets you play with
29:49 lots and lots of different models you
29:50 can also build these things called Bots
29:54 which are just like uh chat gpt's custom
29:57 gpts that they've launched um very very
30:00 similar to that co-pilot is supposed to
30:02 have co-pilots their version of gpts
30:04 yeah exactly so co-pilot um if you go to
30:09 copilot
30:11 microsoft.com they've apparently
30:13 launched but I don't have access to it
30:15 uh a thing called co-pilot
30:17 studio um where you can make customized
30:20 co-pilots so you know their version of
30:23 gpts just as Chris Ramsey said um so so
30:27 go play with all those
30:29 things chat
30:32 GPT those four and those four and and
30:36 now why
30:39 well because they're all kind of good at
30:41 different things and and they're kind of
30:43 one-upping each other in terms of
30:46 capabilities and what you're going to
30:47 find is that some of them do one thing
30:50 that you want to do but don't do another
30:51 thing so well and so understanding you
30:54 know the strengths and weaknesses of
30:55 them is is is really valuable for the
30:58 next six months or so
31:00 um
31:02 and yeah that's that the other thing if
31:05 you are if you're one of those people
31:08 that is curious about this AI stuff and
31:11 you've kind of gotten obsessed about it
31:13 and you're like the only weirdo in your
31:15 group of friends or family that's into
31:18 this stuff and everyone's looking at you
31:19 like you know you're you're just a freak
31:23 um come join the AI salon so this is a
31:26 group that I started
31:28 a year ago we started this one week
31:30 after chat GPT launched and it's it's
31:36 a community of people that are think of
31:39 it like an AI support
31:42 group hi my name is Kyle I'm into AI hi
31:47 [Laughter]
31:50 Kyle and it's it's a really good group
31:52 of people and we just we just moved over
31:54 to Mighty Network so this is the the
31:56 address for the salon so so AIS salon.
31:59 mn. and go look at about the salon it'll
32:03 talk about the salon um you know we're
32:06 we're a ragtag group of Misfits ndo
32:08 Wells and Mis creant who've gathered to
32:10 explore the Uncharted Wilderness of
32:12 generative AI um and it goes on it talks
32:14 about who we are why we do what we do
32:17 what we're trying to accomplish and it
32:18 talks about our values and our values
32:20 are something I'm really proud of and
32:22 and something that I really want the
32:24 group to maintain as it grows um you
32:27 know one of our values is curiosity we
32:30 want people that are curious and and
32:32 inquisitive and trying to learn thank
32:33 you digital Gods really appreciate that
32:35 um generosity is another one it's a big
32:37 one for me that as you're learning stuff
32:40 share it share what you're learning
32:42 there's no reason to keep any of this
32:44 [ __ ] proprietary because it's changing
32:45 so fast that the thing that you think is
32:48 going to give you the competitive
32:50 Advantage is going to be absolutely
32:52 useless a month from now I promise you
32:54 that so just share what you're learning
32:56 because there are are no experts right
32:58 now everyone's trying to figure it out
33:00 and so generosity is about just
33:02 acknowledging that holy [ __ ] we're in
33:04 the middle of this crazy crazy mess um
33:08 exploration is one of our values we want
33:10 people that are kind of actively
33:11 stepping into this adventure
33:13 collaboration that they're willing to do
33:15 it together inclusion they're willing to
33:17 you know accept people from you know all
33:19 different kinds of work all different
33:22 experience levels things like that and
33:24 then finally empathy is our final value
33:27 and I think it's the most important one
33:30 so let's see Jennifer L thank you I've
33:32 been watching you for over a month and
33:34 something thank you very much I
33:36 appreciate that I talk to Pi for hours
33:38 at a time great sounding board yeah papa
33:41 JJ Pi's Pi's really something I've I've
33:44 used Pi for everything from just like
33:47 wanting to kill some time to I had my
33:50 battery go bad on my car and it helped
33:52 me troubleshoot the bad battery thank
33:54 you been watching you for over a month
33:56 and I'm Blown Away by your dedication oh
33:58 thank you very much Jennifer I
34:00 appreciate that yeah I I mean here's the
34:02 deal like do I do I always want to fire
34:06 up the camera and go live seven nights a
34:08 week I
34:10 don't and and yet I'm doing it why
34:15 because
34:19 um I'm need I'm needing Community right
34:23 I'm needing to connect myself with
34:25 people that are excited about about this
34:27 curious about this trying to figure it
34:29 out trying to reinvent themselves trying
34:31 to reinvent their
34:33 business um and it feels really
34:36 important to me and I here's the other
34:37 thing that I know so I was a little bit
34:39 about my background I'm an entrepreneur
34:41 I was an entrepreneur in the early days
34:42 of the worldwide web and
34:47 um and what's going on with AI right now
34:50 is very very similar to what was going
34:53 on uh in the early days of the web and
34:55 and what I know is that
34:58 the
34:58 beginning of a major shift like what AI
35:03 is about to to do to the world you don't
35:06 get to participate in
35:09 that ever right it's it's it's an
35:12 incredibly rare
35:14 moment to be at the beginning of
35:16 something we are absolutely at the
35:19 beginning of this thing what is this
35:21 thing I don't
35:23 know you know it's AGI it's it's AI it's
35:27 Ma Mach learning it's generative
35:29 AI the values are invaluable thank you
35:32 papa JJ I appreciate that I listen I
35:35 think the values here's the thing about
35:37 values if you if you look at these
35:41 values and they sing to you and you're
35:43 like yeah that's like I want to I want
35:45 to hang out with those kind of people
35:47 great this is probably going to be your
35:49 group of people if you look at those
35:51 values and you're like H that seems
35:53 [ __ ] awful then it's going to be
35:54 [ __ ] awful like like values are not
35:57 there just as like you know successories
36:00 posters right this is how I want people
36:02 to behave and if you don't behave that
36:04 way then it's just it's you're not going
36:06 to fit it's just going to be like a like
36:07 a genetic you know uh you know it'll
36:11 it'll be a it'll the organism will will
36:13 reject it and like you won't enjoy it
36:15 and it won't enjoy you so so those
36:17 things are actually really really
36:19 important I've I've grown a lot of
36:21 things over the years and and values are
36:24 super important and not every one likes
36:28 them because in order to have values you
36:30 actually have to have a point of view
36:32 and we
36:34 do so that's that group we will be
36:36 grateful for what you're doing for years
36:38 to come oh thank you be not afraid I
36:40 appreciate that um I hope so I listen I
36:44 it's not about that right now it's just
36:46 about let's [ __ ] figure this out
36:50 together and and what is this I don't
36:54 know I don't know I mean it's it I
37:00 the more I learn about this the less
37:02 educated I
37:04 feel how do you plan to monetize this
37:06 community You're Building any
37:08 advertisements um well we're not going
37:11 to we're not going to monetize the um
37:14 the community so the the AI Salon what
37:17 we're trying to do right now is just
37:18 cover our costs so so we don't have any
37:20 dues um the mighty networks is free like
37:24 the mighty networks is free and is
37:25 always going to be free
37:28 um right now we're taking um donations
37:31 um and
37:34 um uh we're selling monthly sponsorships
37:37 which which are really cheap right now
37:39 basically just to cover our monthly it's
37:41 like 120 bucks a month for Mighty
37:43 networks we we had to do the business
37:44 version of it for the features we wanted
37:47 um I think over time we might allow
37:49 members within the community to create
37:52 courses and they might be able to sell
37:54 those courses so people can you know mon
37:57 ties things individually I think at some
37:58 point we'll have a merch store that you
38:01 know the proceeds of that will will
38:02 support the group but the whole idea is
38:05 like I I want this to be a like an
38:08 ongoing you know group that anyone can
38:11 join the whole idea is being inclusive
38:14 so um so I'll figure out you know it
38:17 it's we want we want people within the
38:21 community to be able to build their
38:24 brand and monetize that but not monetize
38:27 the community itself and so we're we're
38:30 thinking of all the different ways that
38:32 that's possible what's the huge
38:34 performance limiter on the next year uh
38:37 human
38:41 beings I I think the huge
38:48 limiter is going to be human
38:56 beings
39:03 hm I think we're very very very very
39:07 very unprepared for what's coming for
39:09 what's
39:10 here how much for a Bruce Campell
39:13 impersonation see there's how I'm going
39:14 to monetize things I'm GNA be I'm going
39:17 to do I'm GNA do B celebrity
39:20 impersonations for a fee that's that's
39:23 how we're going to monetize this bad
39:25 boy
39:31 that's pretty funny um huge limiter men
39:34 fighting so we can't have nice things
39:36 yeah don't disagree with that yeah yeah
39:38 men being [ __ ] I think I
39:42 think some limiting factors will
39:46 be centers of
39:51 power actively resisting what's coming
39:54 because they're going to realize
39:58 what's coming is going to [ __ ] with
39:59 their business model get your smart
40:01 pants we're about to take off yeah
40:03 exactly exactly
40:07 bran normally if we work with Bri work
40:10 for British hours we like to wear smart
40:14 puns
40:16 um yes yes
40:19 indeed yeah I think there's going to be
40:21 a lot of resistance to this AI stuff
40:23 they're going to demonize it um they're
40:26 going to dismiss it the problem with
40:29 dismissing it is it's already getting
40:31 powerful enough that I think I think
40:34 companies and people are
40:36 realizing
40:38 um we can't really dismiss this but if
40:41 we accept
40:44 it it's a threat to our business model
40:46 and I can talk more about that how do
40:48 you think the government will react when
40:50 we hit AGI Panic
40:55 um I don't know I don't know I I think
41:00 that the fact that that Biden put out
41:02 the 122 page executive order is
41:05 promising I've got a i' I built a GPT
41:08 that if you want to understand the
41:09 executive order I built a GPT that you
41:11 can you can ask all about it um which I
41:15 shared on the AI Salon
41:18 um I think I did it in the show in tell
41:21 Channel um so if you just if you just
41:23 look for that there um
41:28 so I so I think the fact that
41:32 um that that there's awareness at the
41:35 highest level that this is important is
41:37 is
41:39 good
41:41 um the government doesn't understand
41:45 WiFi so so
41:49 um and the government's not good at
41:51 moving
41:53 fast so so it's absolutely possible that
41:56 there's going to be some Panic you know
41:59 if we hit AGI or if it's declared or if
42:02 some version of it is rolled
42:05 out
42:07 um I don't know I like
42:10 I we're in such a like we're living in
42:14 an episode of black mirror right now
42:16 Obama is currently in charge of AI yeah
42:19 he's he's he's one of the senior
42:21 advisers on it I don't know if he's in
42:22 charge of it but he's one of the senior
42:24 advisers for for what to do with it you
42:26 know he's a thoughtful guy he's really
42:28 smart he's internet Savvy like that's
42:31 good like you're going to need some some
42:32 smart people right so if it's Obama
42:35 he'll build a network of people that
42:37 have their [ __ ] together some artists
42:39 and writers are seething still listen
42:41 people are going to be seething it
42:45 I I've written seven screenplays like I
42:48 know that world well I have a degree in
42:50 acting right I moved to New York City
42:52 out of school to pursue a life in the
42:53 Arts most most of my friends from back
42:56 then
42:57 are in that
42:58 business um I have students take moral
43:01 vocal stands against AI in my class
43:04 Christopher Ramsey the younger Jen
43:06 hasn't bought in listen I I agree I've
43:09 got 23-year-old twin boys who have not
43:11 bought in they they could give two shits
43:13 about it they're they're not actively
43:15 against it they're very cynical about it
43:20 um you can take a moral stand all you
43:23 want this channel why I do these lives
43:27 is whether you take a moral stand or not
43:31 whether you I wrote a play over the
43:33 summer Kyle very cool lunar stick
43:37 um whether you um agree with it or not
43:41 whether you love it or hate it whether
43:45 um you're completely ignorant of it it's
43:48 here and it's not going away and it's
43:51 getting stronger and better dramatically
43:55 fast and
43:57 we ain't seen nothing yet
44:00 apparently if the if the if the rumors
44:03 of of sort of what's been put out there
44:05 or are
44:07 true
44:09 that
44:12 the
44:15 um they kind of slow
44:18 [Music]
44:19 rolled the qar stuff they actually
44:22 stopped it before it started improving
44:25 itself
44:27 because if if AGI comes and they start
44:33 having the system improve the system
44:37 then things get really powerful really
44:39 fast now there's pieces of that that are
44:41 remarkable because we start doing things
44:43 like curing diseases really fast and
44:46 curing scientific problems really fast
44:48 and curing economic you know um
44:52 disparity models really fast or societal
44:54 models really fast
44:57 but that's going to break a bunch of
44:59 [ __ ] and so I don't think anyone's
45:01 prepared for that so I don't know I'm an
45:03 artist and I think it's amazing so so
45:05 Otto here's here's the thing that that
45:07 I've seen
45:09 um almost all of the artist friends that
45:13 I know who have tried it are finding new
45:17 ways to express themselves with it and
45:20 and they find it remarkable and then
45:22 some put boundaries in place right so
45:24 I've got one of the one of the women
45:26 who's an organizer for the AI Salon Jane
45:29 she's a writer and she's she she writes
45:31 professionally and she uses AI for
45:33 writing you know all the time and and
45:36 you know it's remarkable and she's like
45:37 I can't imagine doing my job without it
45:39 now but she draws a line she won't do
45:42 her personal creative writing with AI
45:45 That's like a Sacred Space for her so I
45:46 think some artists are finding
45:48 boundaries like what's in what's out
45:51 where can I use it where can I not use
45:52 it other artists are just going [ __ ] it
45:55 I will never use it I'm taking a stand
45:58 and that's where the where the wga
46:00 started was [ __ ] AI
46:02 right they then came around and the new
46:05 Writers Guild contract says we can use
46:08 AI but you can't use AI to replace us is
46:10 effectively what it says it makes you
46:12 think as an artist in a different way
46:14 it's hard to describe I totally agree
46:16 with that like thankful Jones that's
46:18 that's exactly that's exactly what it is
46:20 for me is and and it's funny when I when
46:24 I first started this channel PE a lot of
46:26 the questions I got were well if it's so
46:28 creative doesn't it make you less
46:30 creative and the reality is is no it
46:32 actually makes you more creative because
46:35 you can get ideas out of your head
46:37 faster and in some form of um
46:42 articulation really really quickly and
46:45 what I found is that chat GPT can get me
46:48 to like 80% like that and then the last
46:52 20% like if I get really good at
46:55 prompting and get it really dialed in it
46:57 can get me to like 85% or maybe 90% it
47:00 it almost never gets me to the Finish
47:02 Line um but I get 10 ideas out of my
47:08 head that never would have seen the
47:10 light of day historically and so what
47:12 it's doing is it's just it's it's like
47:14 writer block as a thing of the past
47:18 um my my ability to like my confidence
47:23 the the whole sort of crazy straw of
47:27 creativity and and being neuros spicy is
47:31 is that there's a lot of Shame and
47:34 imposter syndrome and and that sort of
47:35 [ __ ] right and what I found with the
47:38 AI stuff is that
47:41 it it kind of erases that like it's just
47:46 like like you don't have to really
47:50 confront well you know I'm not I'm not
47:52 really good at drawing so that idea for
47:54 that card game would wouldn't be good
47:56 I'd have to go get someone that was good
47:58 at drawing and now I can just have a
48:00 card game and now I can put it out there
48:02 and you know I can't do creative writing
48:06 with it but man do I love its editing
48:09 capabilities and ability to argue yep
48:12 exactly like again that's what what my
48:15 what my strong strong recommendation
48:17 here is start playing with these tools
48:19 and figure out where the boundaries of
48:21 the technology are like what's it good
48:23 at what's it not good at what right and
48:25 then figure out where the boundaries of
48:28 your personal work is right there might
48:32 be some things that you're like oh I'll
48:34 absolutely use it for project management
48:36 but I would never use it to generate art
48:38 because I love illustrating great then
48:40 carve that out and don't do that but use
48:42 it for all this other stuff yep I use it
48:45 for storyboarding for the video process
48:47 yeah
48:48 exactly
48:50 um yeah and and you can take like I I've
48:53 done this now a couple of times where
48:54 I've taken projects that were mothballed
48:58 for
48:59 decades and I was like oh let me pull
49:02 out that half a screenplay I started and
49:05 then and then like within 10 minutes I
49:07 had an outline finishing the screenplay
49:10 and I was like oh that's pretty good I
49:12 could actually finish
49:14 that amazing I want to be my own artist
49:16 why does my creativity have to stop
49:19 where artists Define
49:22 it absolutely it's we're seeing that so
49:25 so Broody you've you've been coming here
49:27 a fair amount now like so you're you're
49:29 uh you are quickly becoming an irregular
49:31 so the people that show up regularly um
49:34 have named themselves the Irregulars and
49:36 that's that's the the real criteria for
49:38 being an irregular is just [ __ ] show
49:40 up but what's happening is a lot of the
49:43 Irregulars are are completely
49:45 Reinventing themselves and what they
49:47 enjoy and what they're good at and what
49:49 they where they're finding they've got
49:51 skills where they never thought they'd
49:53 have
49:54 skills and that's absolutely remarkable
49:58 right and and to your point like why why
50:03 have boundaries we now have a tool that
50:07 can let us level up in the areas that we
50:09 sucked at
50:11 before and then what we can do is just
50:13 think creatively
50:15 horizontally knowing that the tool can
50:18 do some sort of execution on all the
50:19 vertical Specialties just me in the
50:22 trees that's why it's so important for
50:23 teachers to start learning and how to
50:24 teach students to use it
50:27 absolutely absolutely just me and the
50:29 trees this is it's so funny at the
50:31 beginning of the year I was kind of
50:33 agnostic
50:35 about education and I'm like well if if
50:38 Educators don't want to do it that's
50:39 their prerogative and if they want to
50:41 block
50:41 it and and now I'm just like
50:45 no no Educators need to get their [ __ ]
50:49 together
50:52 and understand that
50:54 that
50:56 I mean even if you're just dealing with
50:58 college the world that freshmen today
51:02 are going to enter in four years is not
51:05 going to resemble the world that we're
51:06 in right
51:07 now so freshmen today are being taught
51:11 things that are going to be Irrelevant
51:14 in four years a lot a lot of the things
51:17 but there's going to be new skills that
51:20 are going to be absolutely in
51:22 demand and so I I think it's imper ative
51:27 for educational institutions to learn
51:30 this stuff to get curious about it and
51:34 to start like for me the big skill
51:37 moving forward is critical thinking and
51:39 so to the extent that you're not
51:41 actively teaching and not even teaching
51:44 critical thinking that you're not
51:47 actively exploring critical thinking
51:49 with these tools you're doing our future
51:52 a disservice I'm in training to be a
51:54 clinical psychologist and trying to
51:56 build tools to expand extend my presence
51:59 so sunshower I think that um so if you
52:01 haven't played with it yet go go play
52:05 with so there's an app called Pi this is
52:09 my
52:09 iPad and so go to
52:13 Pi and then you'll see you know it's
52:16 just like a normal chat gbt like thing
52:19 and it'll look like that when you log in
52:21 and then you can you can talk with it
52:25 but in the lower right hand corner you
52:27 see where it says Pi if you click on
52:30 that there's two modes for it now one of
52:34 them is pi just PI right the old one and
52:39 one of them is support
52:42 pi and support
52:46 Pi is like a
52:49 therapy fine-tuned version of Pi and
52:53 then the other thing you can do with pi
52:55 is you can put it into into conversation
52:56 mode so see that little thing right
53:03 there can you describe support pi to me
53:06 and how it's different than normal
53:14 Pi haha you're really hitting me with
53:17 some great questions support p is a new
53:19 in Creative term you just coined support
53:22 p is not a real thing but let's role
53:24 play like it
53:31 no I mean the pi model Pi that you're
53:34 based on there's two different modes
53:36 there's just pi and there's support Pi
53:39 can you talk about what support Pi
53:45 [Music]
53:48 is ah now I understand what you're
53:50 asking you're correct that there are two
53:53 different modes of the P model there's
53:55 the general information mode or just P
53:57 mode and there's the support mode which
53:59 you've been referring to as support P
54:02 the general information mode is designed
54:04 to give accurate comprehensive
54:05 information on any topic you can think
54:08 of whereas the support mode is more
54:10 focused on providing emotional support
54:13 empathy and companionship the support
54:16 mode so so if if you're trying to Fig so
54:19 there's two two things I would do so
54:21 that's Pi I would explore that and then
54:25 the gpts So within chat GPT you can make
54:29 a GPT that can take your philosophy of
54:33 how you want to do therapy and let's say
54:35 you've written a paper on it and you
54:37 want to do your therapy based on your
54:39 you know your dissertation or whatever
54:41 it might be you can use that as
54:43 knowledge to build your own custom GPT
54:45 where someone could interact with a
54:48 version of you and then chat GPT also
54:52 has conversation mode so on the right
54:54 hand side you can also put chat
54:57 GPT into a similar conversation
55:05 mode absolutely let's continue our
55:08 conversation how did you feel during
55:10 these different moments in your dream so
55:13 for example were you amused scared so
55:16 this was a GPT that I made called Dream
55:18 Weaver that's a dream analysis tool and
55:20 so I can have a conversation with
55:23 Dreamweaver and tell it my dream and
55:25 it'll do a whole analysis and even
55:27 visualize it Otto got to got to go eat
55:30 later everyone later Otto thank you for
55:32 coming thank you for hanging
55:33 out
55:36 um all right let's see are there any
55:39 copyright issues to worry about if GPT
55:41 helps you write your screenplay I don't
55:44 think so I mean if if
55:48 listen it's it's on you to not be a
55:52 douche right if if you just say you know
55:56 give me the give me the plot for Star
55:58 Wars and you know make it a western and
56:01 change all the character names then you
56:04 know if you just you know do something
56:06 like that then you're setting yourself
56:08 up for uh a problem but just using it in
56:11 general based on the training I don't
56:12 think so in fact a number of the
56:15 lawsuits where where people are doing
56:17 these class action lawsuits you know
56:19 suing based on content that was used in
56:22 training a lot of those are being um
56:26 tossed out by judges so I I I wouldn't
56:29 worry about it right now I I think that
56:32 IP attorneys are going to be suing each
56:34 other both directions for the next 20
56:36 years so at some point some precedent is
56:39 going to be set that'll that'll do for
56:42 AI what happened with hip hop and
56:44 sampling music in the 80s and 90s where
56:47 where we sort of landed on some version
56:48 of fair use and how people get
56:50 compensated and attribution and things
56:53 like that so some some version of that
56:55 will happen and and there will be tools
56:57 that come out that support it in fact um
57:02 YouTube I just saw a new tool that
57:04 YouTube's got that
57:07 um if you're doing a YouTube short you
57:10 can create custom background music for
57:12 your YouTube short in the style of
57:15 Charlie poth or Tay right so they've cut
57:18 deals with Charlie poo and P T where um
57:23 it'll generate the song that sounds like
57:24 Charlie po wrote it and and it's singing
57:27 and it sounds like him he's probably
57:29 getting paid every time someone
57:31 generates a song um so there'll be all
57:34 sorts of different new new kinds of
57:36 models that come out that'll deal with
57:38 it so but I wouldn't worry about it
57:39 right now just don't go steal you know
57:41 don't go blatantly steal [ __ ] because
57:43 then you'll just set yourself up for a
57:45 for you know trouble um I wonder how
57:48 this can be used to benefit the trades
57:50 I.E construction so anywhere Jerry G
57:52 anywhere where you've got um
57:59 computers
58:02 mediating a
58:05 process is is is going to be set up for
58:08 you you can do all sorts of interesting
58:10 things um so it could be
58:15 um finding ways to do construction
58:18 project planning more efficient more
58:20 efficiently you could you could upload a
58:23 spreadsheet of the last job
58:26 and say you know do some analysis on say
58:29 how can AI be used in construction what
58:32 I would say is this what are problems
58:34 that you're trying to
58:36 solve start with that and say okay if I
58:38 could solve this problem that that would
58:42 that's a major pain point for us and if
58:45 that problem in any way can be used as a
58:48 computer or could use a computer um then
58:51 you could say okay how can I use AI to
58:53 solve that problem and and then you'll
58:55 likely find ways to do it very very
58:58 quickly cuz you can just go to chat GPT
59:01 or any of these other things and just
59:03 say here's the problem I'm facing here's
59:05 how I currently deal with
59:07 it help me come up with a strategy to
59:11 make that better and it will and then
59:13 you'll get inspired you'll be like oh
59:15 and then you'll ask it some more
59:16 questions just think of it kind of like
59:18 you've got a business
59:20 consultant at the ready you have to give
59:22 it lots of instruction you have to give
59:24 it lots of Direction
59:25 but it knows
59:28 everything you can ask for specific NE
59:31 NEC code sections for electrical it's
59:33 pretty cool yeah we uh so my my
59:36 co-founder of my current business
59:38 storyvine she's got a shitty
59:41 construction site next to her house and
59:44 this construction company is just
59:46 they're just violating codes like
59:48 multiple times a day they're just
59:50 they're just shitty and they don't care
59:53 and and so we buil bu her we took the
59:56 Denver building codes website and we
1:00:00 built a GPT so she now has a GPT that
1:00:03 she can use to basically say here's what
1:00:07 they did today is is that um is that
1:00:11 against
1:00:13 code so Kyle come in please come in
1:00:17 where to he gone oh froze come
1:00:22 back the model is being tweaked can you
1:00:24 hear
1:00:25 am I good now tap the tap the hearty
1:00:28 Hearts if you can hear
1:00:31 me okay I'm seeing
1:00:34 aart you're back okay I'm back
1:00:38 cool I don't know what I was saying
1:00:42 um yeah yeah you can do code stuff you
1:00:45 could
1:00:47 do um one of the guys at the building I
1:00:51 work work for created a thing called
1:00:53 maintenance mentor
1:00:56 that um would would do like you know if
1:00:59 you have questions about the a AC or the
1:01:02 heating or Plumbing or electrical it
1:01:05 will help you troubleshoot problems
1:01:06 within the building um so you can it's
1:01:09 just it's kind of endless so again where
1:01:11 I would start is start with the problem
1:01:13 you want to solve first and then go see
1:01:17 how much chat GPT knows or or any of
1:01:19 these tools how much they know um they
1:01:22 will likely surprise you at how good
1:01:24 they are
1:01:25 especially things like construction
1:01:27 there's so much been written on it
1:01:28 there's so much that's been on the
1:01:30 internet about it that it just kind of
1:01:34 assimilates all that information and it
1:01:36 kind of knows everything it's pretty
1:01:37 remarkable how many gpts have you made
1:01:40 I've probably
1:01:42 made I don't know 25 or 30 most of them
1:01:45 I've made live on on on this channel um
1:01:48 we can make one now if you
1:01:50 want um if someone has a specific thing
1:01:53 they want to do I'm happy to do that
1:01:57 hope you have a good holiday Kyle with
1:01:59 your gpts thank you your company
1:02:03 co-founder is my hero oh that's so cool
1:02:06 my
1:02:07 Shiro Monique's
1:02:10 amazing yeah no Monique and I we we knew
1:02:13 each
1:02:14 other I don't know five or six years
1:02:17 before we started storyvine and I was
1:02:19 Consulting with her she had started a
1:02:21 social media
1:02:23 agency and I was Consulting with her on
1:02:25 that and that ended up kind of falling
1:02:27 apart and there were kind of you know
1:02:29 founder issues and things like that and
1:02:31 so I knew she was available and I knew I
1:02:33 wanted to start story Vin and so I
1:02:34 called her up and said hey I've got this
1:02:36 idea you know do you want to start it
1:02:39 cuz I knew she was good at the [ __ ] I
1:02:41 wasn't as we describe our relationship
1:02:43 Becky I'm the hand wavy guy and she
1:02:45 keeps the trains running on time um and
1:02:49 and you know it's been it's been a
1:02:51 really really really good partnership
1:02:53 and if you started companies before
1:02:56 finding good partners is almost
1:02:58 impossible teaching autistic and add
1:03:01 math please do it oh teaching autistic
1:03:04 and add math for for for a GPT we can do
1:03:08 that that would be a good one
1:03:12 um yeah yeah we can do that I'm trying
1:03:15 to think are there
1:03:17 any if you've got any
1:03:19 sources we we'll we'll we'll make one
1:03:21 without sources right now and then I'll
1:03:23 show you how you could improve with
1:03:25 sources so let's go make a GPT he
1:03:27 creates so right now cam on on Mighty
1:03:31 networks um if you go to show and
1:03:35 tell okay hang on a
1:03:38 sec oh wait hold please
1:03:43 hold so I'm in the showand tell channel
1:03:47 on Mighty networks
1:03:49 right and in the lower right that's how
1:03:51 you add a post to show and tell
1:03:58 and then it actually just prompted that
1:03:59 see where it says the hashtag thing so
1:04:03 when you create a
1:04:07 post you bring up a list of
1:04:13 hashtags can you see it oh come on there
1:04:16 you go well they're down there see that
1:04:18 one hashtag is gpts so if you tag your G
1:04:22 you you tag your post in show and tell
1:04:24 with GPT you can then go um sort
1:04:29 by all the hashtags so so there's like a
1:04:32 pull down menu where you can say show me
1:04:33 all the ones that are tagged with GPT
1:04:35 it's not a perfect way so we we might
1:04:38 improve that but right now how we're
1:04:39 handling it is with hashtags so share it
1:04:42 in Show and Tell and just tag it just
1:04:44 tweak my GPT if you want Kyle
1:04:49 oh that's right
1:04:52 Pate so pate's got I we don't you know
1:04:55 what I think what I'd rather do is point
1:04:57 them to your so whoever just said to
1:05:00 create a
1:05:01 GPT um for autistic uh and add
1:05:07 students um on the channel Pate is a is
1:05:11 a a he's a swell guy works at Google he
1:05:16 heads up the AI mechanics Guild on the
1:05:20 AI salon and he
1:05:23 created a
1:05:25 GPT for for you know in that
1:05:29 neighborhood and so that could be
1:05:31 modified so let me go to chat GPT here
1:05:34 hang
1:05:35 on now I think did I look at it I'm sure
1:05:39 I looked at it here
1:05:43 recently explore hang
1:05:53 on recently
1:05:59 used where is it P what's it called
1:06:02 again top of the hourish Irregulars
1:06:11 present ask
1:06:18 Mary
1:06:20 H I don't know how to
1:06:23 see
1:06:26 unambiguous Ally or Aly Ally hang
1:06:31 [Music]
1:06:33 on delete
1:06:48 post hold
1:06:53 please
1:07:01 search and M big
1:07:10 us all right if you go to go to go to
1:07:14 Mighty networks and find it yeah if you
1:07:17 go to the AI Salon Mighty networks and
1:07:19 search for unambiguous
1:07:21 Ally this is the GPT that Pate wrote
1:07:25 for
1:07:27 empowering neuros
1:07:30 spices all right um so that's a good one
1:07:35 so let's let's do something else since
1:07:37 since that one's already built like
1:07:38 that's a good one and I actually want to
1:07:39 support that one because I feel like
1:07:42 you've got a vision for that pain so USA
1:07:45 loves Tik Tock okay all right no picture
1:07:49 must
1:07:50 pass what do you mean no picture must
1:07:53 pass I don't know what that means
1:07:55 what do that
1:07:56 [Laughter]
1:08:03 man who did this
1:08:06 one who did this darling
1:08:11 Rober that's just rude it's true but
1:08:14 [Laughter]
1:08:18 rude oh
1:08:21 man the Curiosity hamster wheel that's
1:08:24 pretty
1:08:26 fun all right so let's go let's go make
1:08:29 another one um what do we want to make
1:08:31 for a GPT uh uh give me another give me
1:08:34 another topic we'll we'll do one on the
1:08:36 Fly here feel free to try it out and
1:08:39 provide feedback it's already been
1:08:41 improved thanks to the community that's
1:08:43 awesome Pate that's really
1:08:46 cool um which is completely different
1:08:49 than the
1:08:50 androgynous
1:08:53 Ally
1:08:55 uh wait crit faced can you show us how
1:09:01 you use it to
1:09:03 storyboard um I haven't used it to
1:09:06 storyboard but we can certainly do that
1:09:07 let me go that that might be a fun thing
1:09:09 to do all right so let's go do that so
1:09:12 we're going to go to um here's chat
1:09:17 GPT
1:09:20 um I'm in chat gp4 mode so one one of
1:09:24 the things if if you haven't used chat
1:09:27 GPT a lot um chat GPT is now multimodal
1:09:31 meaning you can upload images you can
1:09:38 um you can generate images you can
1:09:41 upload
1:09:42 um spreadsheets and CSV files and it can
1:09:46 analyze them it can it can write things
1:09:48 so one of the things I'm going to do
1:09:50 here is I'm going to
1:09:53 go I'm going to upload a screen
1:10:03 [Music]
1:10:15 play that's Microsoft Word I'll just
1:10:18 I'll upload the text version of it all
1:10:20 right so here's a this is a screenplay I
1:10:22 wrote in 1990 one or two called salmon
1:10:26 flying
1:10:28 East and
1:10:30 so um so one of the things I'm going to
1:10:32 do is I'm going to
1:10:39 say please
1:10:42 write a
1:10:44 compelling Hollywood log line that is
1:10:48 the elevator pitch for this
1:10:51 screenplay period
1:10:55 act as if you're a high-powered
1:10:58 Hollywood agent and make sure
1:11:04 that anyone that hears this will want to
1:11:07 make this
1:11:10 movie so so so log lines are especially
1:11:14 if you if if you write the screenplay
1:11:15 writing log lines about it are almost
1:11:17 impossible cuz you can't separate what
1:11:20 you know is wrong with the screenplay
1:11:22 from from the marketing piece of it so
1:11:25 let's see what it comes up with salmon
1:11:27 flying East in the misty docks of 1991
1:11:30 Gig Harbor Jack a rugged fisherman
1:11:33 haunted by his Vietnam War past grapples
1:11:36 with shadows of memory and longing of a
1:11:38 life he can't reclaim when his path
1:11:41 crosses with ping a striking Vietnamese
1:11:43 woman entwined in his wartime memories
1:11:45 Jack's quest for Redemption takes a
1:11:47 dramatic turn as Jack as Jack navigates
1:11:51 the treacherous Waters of his past and
1:11:53 present he finds him himself Torn
1:11:54 Between the ghost of his love lost in
1:11:57 Vietnam and the reality of his solitary
1:11:59 life blah blah blah okay this is so the
1:12:02 content may violate our content policy
1:12:04 because probably because it talks about
1:12:05 war um that's great can you make it much
1:12:10 shorter and
1:12:14 punchier I'll just leave it that see
1:12:16 what it
1:12:22 does
1:12:25 all right a haunted Vietnam vet and
1:12:27 rugged fisherman in 19901 Gig Harbor
1:12:30 confrences w war torn past when he
1:12:34 reencounters ping a figure from his
1:12:35 Vietnam days and past and present
1:12:38 Collide Jack search for Redemption and
1:12:41 peace propels him on a thrilling journey
1:12:43 through love loss and the Quest for
1:12:44 healing I'm going to say
1:12:48 um give me a log line that
1:12:51 compares this screenplay to two
1:12:55 different movies meeting one
1:12:58 another I don't know if it'll understand
1:13:00 what that means like Jaws meets
1:13:04 Rocky we'll see what it comes up with
1:13:07 what I what I find about chat GPT is it
1:13:09 tends to be um way better at
1:13:11 understanding me than I am like I don't
1:13:15 understand what I say half the time
1:13:17 salmon salmon flying East is where the
1:13:19 deer hunter meets a river runs through
1:13:21 it oh that's kind of cool a poignant of
1:13:24 a Vietnam veteran fishing uh veteran
1:13:27 fisherman wrestling with his past
1:13:28 entangled in a haunting romance that
1:13:30 Bridges the turbulent Waters of War and
1:13:33 the Serene struggles of life back home
1:13:36 that's cool that that sounds that that's
1:13:38 in the neighborhood so now I'm going to
1:13:40 say
1:13:41 okay okay great now I'd like you to
1:13:43 create a list
1:13:52 of visual ual scenes from the
1:13:58 opening of the
1:14:01 movie period describe them as if they
1:14:04 were a
1:14:09 storyboard period in fact create a
1:14:14 table where one column
1:14:17 describes the
1:14:19 action and the next column describes
1:14:23 what's going on
1:14:31 visually all right scene one Gig Harbor
1:14:35 Docks pre-dawn a foggy Serene Gig Harbor
1:14:38 the camera slowly pans across the glassy
1:14:40 water of the Puget
1:14:42 Sound fishing boat interior Jack's
1:14:45 morning
1:14:47 routine Jack steps out of the boat onto
1:14:49 the dock all right so great so this is
1:14:53 so this is now this is now reading my
1:14:56 screenplay and and pulling the visuals
1:14:59 from
1:15:01 it flashback to a crowded bar in
1:15:15 Vietnam so let me say
1:15:22 so great I'd like to start storyboarding
1:15:26 this so I want you to
1:15:29 create
1:15:31 16x9 wide
1:15:34 orientation
1:15:38 images one per
1:15:42 scene
1:15:45 period I'd like the style to
1:15:47 be loose
1:15:52 watercolor in
1:15:55 style but that clearly
1:16:00 indicate what the camera
1:16:03 sees
1:16:04 and what
1:16:09 lens should be
1:16:12 used period why don't we start
1:16:19 with scene one to get the style dialed
1:16:27 in and and so notice what I'm doing here
1:16:32 is while I'm technically prompting this
1:16:35 isn't really prompt engineering right
1:16:37 you keep hearing this term you got to be
1:16:39 a prompt engineer and prompt engineering
1:16:41 is going to be big and prompt
1:16:43 engineering prompt engineering all that
1:16:45 does in my opinion is intimidate people
1:16:48 to not want to deal with AI at all I
1:16:50 don't want to be an engineer I'm not a
1:16:52 PR engine I don't even know what a
1:16:54 prompt is and now I got to engineer it
1:16:57 no no what you actually have to do is
1:16:58 just
1:16:59 talk you just you just talk your words
1:17:02 now could I use prompt engineering to
1:17:05 get things better sure I could
1:17:07 absolutely do
1:17:08 that um so there's my thing okay that's
1:17:19 great okay that's terrific but I don't
1:17:23 want the splattery edges period I want
1:17:27 the image to go all the way to the edge
1:17:30 of the frame
1:17:36 period I'd also like to create two
1:17:45 distinct color
1:17:48 grades one for the present which is more
1:17:52 saturated
1:17:54 and vibrant and one for the
1:17:58 past that captures the fog of
1:18:03 War
1:18:07 period let's
1:18:11 create two images one for the
1:18:15 present and one from a scene from the
1:18:20 past so I can compare them
1:18:24 all right so it like it already hit me
1:18:27 that this looks pretty cool but it would
1:18:30 it would also be cool to have you know
1:18:31 the past look like the past you know how
1:18:34 films
1:18:36 do but but like but what I'm doing here
1:18:39 is just I I am acting kind of as a
1:18:42 director kind of as a writer kind of as
1:18:45 an art director kind of as a creative
1:18:47 director that's way too colorful so we
1:18:50 hate that and we will tell it them we
1:18:53 hate
1:18:57 that but that does look like Gig
1:19:00 Harbor which is amazing
1:19:05 actually and so now it's creating
1:19:07 there's the one from the past right so
1:19:10 listen at a minimum it kind of got that
1:19:13 you
1:19:13 know there's your present there's your
1:19:16 past
1:19:19 right so now I'm just going to say
1:19:22 you've got the right idea but the images
1:19:25 still
1:19:27 have blotchy edges rather
1:19:31 than the image going all the way to the
1:19:34 edge of the frame period
1:19:37 also the saturated image is way too
1:19:41 vibrant period it should still look like
1:19:43 a
1:19:46 film but be more saturated than the
1:19:49 images from the past
1:19:51 period why don't we put
1:19:54 an olive green Hue on the color grading
1:19:58 from the images from the
1:20:00 past and
1:20:04 a soothing blue hue for the images from
1:20:08 the
1:20:10 present let's see what that
1:20:17 does make it as an elevator pitch for a
1:20:19 David swimmer single dad
1:20:21 sitcom exactly we could do
1:20:24 that well that's I mean that's the thing
1:20:27 I could take I could absolutely take
1:20:29 this
1:20:30 screenplay and have it rewrited as a
1:20:32 comedy like give me a new outline with
1:20:34 this is a comedy and then use the
1:20:37 writing style of how this is written and
1:20:40 let's let's turn this into a comedy you
1:20:43 could absolutely do
1:20:46 that all right there's our blue hue
1:20:49 again that looks like Gig Harbor in fact
1:20:53 that actually looks like where my dad
1:20:56 had a boat that looks like where where
1:20:58 we exited that looks like Point Defiance
1:21:01 and there's the there's a Narrows Bridge
1:21:03 right there like this actually looks
1:21:06 familiar to me that's
1:21:09 crazy there you
1:21:12 go so there's your two different styles
1:21:15 it still doesn't quite get that I don't
1:21:16 want these blotchy edges but that's all
1:21:18 right so so we'll say fine so
1:21:22 say
1:21:30 great now can you make an image for each
1:21:35 scene using our new color
1:21:44 grading
1:21:46 framework based
1:21:48 on whether the scene is in the present
1:21:51 or in the past
1:21:54 and it should just start doing it should
1:21:56 just start going through and
1:21:58 doing why are those lines in that
1:22:01 orientation the I what it did Pate these
1:22:04 these lines you mean the these are film
1:22:07 strip
1:22:10 sockets it it it realized it was a film
1:22:13 so it made it like like this is what the
1:22:16 frame of film's going to look like but
1:22:17 the the sprocket holes would be down the
1:22:19 left and right not across the top it's
1:22:21 listen it's trying
1:22:26 this like I said I gave it you know I
1:22:29 gave it some shitty prompts you know I
1:22:31 talked before about don't be a prompt
1:22:33 engineer but obviously I would need to
1:22:35 be a bit more specific about the
1:22:39 style all right let's see is that scene
1:22:44 one what's scene
1:22:47 one scene one a foggy Serene Dawn at Gig
1:22:51 Harbor inside the r fishing boat cabin
1:22:54 okay so that's scene two so let's see so
1:22:58 that is scene one which is
1:23:00 right that's scene
1:23:03 two there's Scene Three Jack's
1:23:12 up this is
1:23:14 good who suggested this this was a good
1:23:17 idea no this is really good that is
1:23:19 totally what the docks looked like
1:23:21 although these are more like Easter in
1:23:23 fishing boats the the Ducks where where
1:23:26 my dad had a fishing boat was in uh they
1:23:28 were more modern boats but but like that
1:23:32 looks very familiar although the dock
1:23:33 kind of ends right there it's pretty
1:23:36 funny there he is at the what was scene
1:23:39 four was so wait what was scene
1:23:45 four Jack emerges on the docks the baate
1:23:48 shop inside de baate shop Jack greets
1:23:51 Martha the owner's wife
1:23:53 Jack prepares for
1:23:54 fishing this is
1:23:57 good I've never I've never done this on
1:23:59 my own screenplay before but I mean holy
1:24:02 crap like look at this he's obviously
1:24:05 not preparing that that's after they
1:24:07 caught something although I guess that
1:24:09 could be bait but he's he's salmon
1:24:11 fishing those should be
1:24:13 smaller all right there we go into the
1:24:15 past so he's cutting bait and
1:24:18 remembering the
1:24:20 past like I I mean you can just you can
1:24:23 almost hear it right
1:24:25 like this is
1:24:28 cool oh I'm going to do one one other
1:24:30 one final thing
1:24:33 here I think it can do
1:24:35 this let's let it keep
1:24:38 going so wait that was preparing bait
1:24:41 what was that
1:24:43 one that was flashback to the Vietnam
1:24:46 War abrupt transition to a helicopter
1:24:48 flying
1:24:49 low Jack back in the present flashback
1:24:52 to a crowded
1:24:54 bar so so we have some back and forths
1:24:56 now between the past and the present all
1:24:59 right so he's he's at the bait shop now
1:25:01 he's preparing to go out now he has that
1:25:04 flashback now he's getting ready to go
1:25:06 out now he has that
1:25:09 flashback look at that I mean this is
1:25:11 amazing there was an
1:25:14 error oh generating response I don't
1:25:16 want to regenerate all
1:25:18 this
1:25:21 h is it going to start
1:25:26 over yeah it's going to start over we
1:25:29 lost all that oh that sucks anyway what
1:25:33 I what I would then do what I would then
1:25:36 do is you could tell it to take that
1:25:40 table and all those images and put it in
1:25:43 an Excel file and so it can actually
1:25:46 generate a a
1:25:51 uh let's see
1:25:56 it could actually generate a um a
1:26:00 storyboard for you in an Excel
1:26:03 file this is wild isn't it it's a movie
1:26:06 yeah oh and actually you know what we
1:26:08 could do now holy [ __ ] I should do
1:26:13 this because I could take all those
1:26:15 scenes into
1:26:17 um I could take all those scenes into
1:26:20 into Runway ml and and turn them into to
1:26:24 action and just do it with voice over
1:26:26 and stuff like
1:26:29 that hit the
1:26:36 2-2 to get the previous reply oh
1:26:40 really oh oh I see hang
1:26:43 on oh that's cool I didn't know you
1:26:45 could do that wait oh it's still
1:26:49 creating
1:26:50 images oh okay cool
1:26:53 so I didn't lose everything okay that's
1:26:55 something new I learned thank you
1:26:57 whoever said that who is that thankful
1:26:59 Jones okay so see see right here where
1:27:01 it's got one of three now it says three
1:27:04 of three so these are the new images
1:27:06 it's
1:27:08 creating there's Jack getting up I
1:27:10 actually like that shot better than what
1:27:11 it gave us before there he is on the
1:27:14 dock that that looks closer that's
1:27:16 that's in the neighborhood of of what it
1:27:17 looked like for
1:27:21 real this is just
1:27:25 trippy there he is meeting what's her
1:27:34 name any way to capture a receipt of a
1:27:37 user accessing my created GPT not right
1:27:41 now brother 52 so
1:27:44 um if you if you go to your um your
1:27:50 account and you go to my GP s when you
1:27:54 go to my gpts it will show you to the
1:27:56 right of the GPT how many people have
1:27:58 used it you can't click through and get
1:28:00 any more metadata on that I would assume
1:28:03 that when we get to when there's the GPT
1:28:06 store they're going to give you stats
1:28:09 and things like that um but but right
1:28:11 now it's just you can see how many
1:28:13 people have used it but that's about
1:28:17 it this is much much more Eastern tler
1:28:20 than Pacific salmon fisherman but that's
1:28:24 okay that's all stuff this is all stuff
1:28:27 like like that being a big tler is is is
1:28:31 is not right but like that's stuff that
1:28:32 you could fix you could say give me this
1:28:34 he should be cutting smaller bait than
1:28:36 this it should be on a modern boat there
1:28:38 shouldn't be a troller in the background
1:28:40 it should be you know Pacific Northwest
1:28:43 you know Puget Sound hills and mountains
1:28:46 or like the The Narrows Bridge or
1:28:48 whatever the hell they call that thing
1:28:50 how do you change the aspect ratio of
1:28:52 Imes you just ask for it you can say so
1:28:54 when I do when I make my videos for Tik
1:28:57 Tock reminders uh for the lives I just
1:28:59 say give me a 9 by6 image with this
1:29:03 subject that says AI learning Lab at the
1:29:06 top and then and then it makes it
1:29:07 vertical in this case I said make them
1:29:10 16
1:29:11 by9 wide you know wide
1:29:14 orientation that is so cool that it just
1:29:17 goes it it just it's it's remembering
1:29:19 all this
1:29:21 stuff all right each image captures it
1:29:24 um so so let me okay so now I want to
1:29:26 try something I want to
1:29:29 go okay great can
1:29:33 you create an Excel
1:29:42 file that has the
1:29:50 first seven scenes
1:29:56 from the table
1:30:00 above
1:30:02 and add a third
1:30:05 column where you include the
1:30:09 image you created
1:30:12 [Music]
1:30:14 here embed the actual image in the Excel
1:30:21 file now this
1:30:24 should so see where it says analyzing
1:30:27 here that's that's using a thing called
1:30:29 code interpreter and if you open this
1:30:32 what it's actually doing is writing
1:30:34 python code to to build this file that I
1:30:37 just asked
1:30:39 for so it's putting data into the into
1:30:42 the
1:30:43 things it looks like that's just that's
1:30:46 cell that's column
1:30:50 one there's column
1:30:52 two the
1:30:56 descriptions and then I I'm going to I'm
1:30:58 going to close this cuz I want to be
1:31:01 surprised to see if it actually does it
1:31:04 right I missed the first part of this
1:31:07 did he give it a story screenplay or
1:31:09 something and now
1:31:12 it's generating storyboards that's
1:31:15 exactly that's exactly what you're
1:31:18 witnessing so so let me go back to the
1:31:20 beginning while it's while it's making
1:31:21 that file so I uploaded a screenplay
1:31:25 that I wrote In 1991 or 2 I had to give
1:31:28 me a log line which was okay then I had
1:31:30 it make it shorter which was okay and
1:31:32 then I had it compare two movies it said
1:31:35 salmon flying East is where Deer Hunter
1:31:37 meets a river runs through it which is
1:31:39 that's a really good like someone would
1:31:40 go oh yeah that sounds good um then I
1:31:43 had it describe the first 10 scenes in
1:31:47 the in the movie like what what happened
1:31:50 and what the visual thing was
1:31:52 then I did a couple of passes of uh sort
1:31:56 of look and feel and then realized that
1:31:59 I wanted to do a look and feel where the
1:32:01 present was was kind of in color and the
1:32:03 past kind of had this funky Hue to it
1:32:07 and then I said now just uh give me the
1:32:11 scenes you know give me the images from
1:32:13 the scenes in that
1:32:15 spreadsheet and so it did the first
1:32:18 seven and let's see I've created an
1:32:21 Excel file with the first seven scenes
1:32:24 described including the action and
1:32:25 visual descriptions however embedding
1:32:27 the actual images into the Excel file is
1:32:29 not possible with my current environment
1:32:32 you can download the Excel file and man
1:32:34 manually insert them as
1:32:37 needed um let me try
1:32:40 something is there a way for
1:32:44 you to use the image command in
1:32:48 Excel and put the URL to each image in
1:32:51 there so that the Excel file itself
1:32:54 displays the image but it's pulling that
1:32:56 image from the
1:32:59 internet let's see if it's let's see if
1:33:01 it can think its way out of this
1:33:07 one and we can while it's doing that we
1:33:09 can go look at what it gave
1:33:15 [Music]
1:33:21 us
1:33:27 um okay yeah it didn't it didn't give us
1:33:30 even placeholders for the images
1:33:33 so
1:33:36 fail fail fail fail and you know what I
1:33:39 what I'm doing here is you know while
1:33:43 I'm in in theory teaching you like I'm
1:33:45 doing this for the first time I've never
1:33:47 done this before
1:33:50 so this is what I I would encourage all
1:33:53 of you to do is is don't think that
1:33:55 someone has the answers because the fact
1:33:58 is that everyone's trying to figure this
1:34:01 out and and this dear God why would you
1:34:04 bring Excel into this that's
1:34:12 great however you can manually only
1:34:14 embed the images yeah I know I can
1:34:16 manually embed them but I don't want to
1:34:18 let's see
1:34:20 um can you give me a list of the URLs
1:34:24 for each of those
1:34:29 images certainly here's the URLs okay
1:34:33 great oh but they're
1:34:46 broken those URLs do not
1:34:50 work wait
1:34:54 those URLs do not work can you please
1:34:57 provide URLs that do
1:35:05 work NOP it can't placeholder URL all
1:35:08 right well they're broken
1:35:11 anyway so so yeah there's your things
1:35:13 and then this this new thing where you
1:35:15 can take the you can go back in time is
1:35:19 that's new to me so let me show you that
1:35:21 again if you were not here so this
1:35:24 is this is the second batch of images
1:35:27 some of these I like okay but some of
1:35:29 the original ones it did were really
1:35:30 good so see where it says three of three
1:35:33 here that means three of three answers
1:35:36 for this particular prompt so I can go
1:35:38 back to the second version of that where
1:35:40 I stopped it and the first version and
1:35:43 here are the images that it created like
1:35:45 I love that one although I did like the
1:35:48 one with the the alarm clock in the
1:35:49 foreground so maybe do you'd sort of
1:35:52 combine those two then that's really
1:35:54 good he's smoking a cigarette in his
1:35:58 boat he's going out he's at the bait
1:36:01 shop there's supposed to be a woman here
1:36:03 so we that's something we'd fix there
1:36:06 he's cutting bait that bait's too big
1:36:07 but that's okay we'd fix that there's
1:36:09 his
1:36:10 flashback now he's back in the present
1:36:13 looking tortured right cuz he just
1:36:15 thought of that looking
1:36:17 tortured now he's back in the in the
1:36:20 bar and then that's where it
1:36:23 stopped but um holy crap that's really
1:36:27 good that's really good really good nuts
1:36:33 nuts it's nuts I tell you nuts I tell
1:36:39 you you know and we could even do you
1:36:41 know what would be kind of fun is
1:36:46 say as an experiment can you
1:36:50 create
1:36:53 two images one from the past and one
1:36:56 from the
1:36:57 present based on the scenes
1:37:01 above but do them
1:37:03 in a
1:37:07 photographic cinematic style so they
1:37:11 actually look
1:37:13 like stills from the
1:37:19 film period
1:37:24 the
1:37:26 lighting should be
1:37:30 professional and
1:37:32 Moody and
1:37:35 the set
1:37:38 decoration and locations should be
1:37:43 impeccable see if we can get it to see
1:37:46 if we can get it to give us some
1:37:48 Hollywood
1:37:49 action ask for a list of code block or
1:37:53 markdown oh yeah that's a good idea retr
1:37:55 Punk let me try
1:38:07 that oh that's pretty cool that looks
1:38:12 nice that very much looks like Gig
1:38:15 Harbor very much looks like
1:38:19 it very much in
1:38:27 fact didn't quite capture the the color
1:38:38 grading the image from the past doesn't
1:38:41 quite have the color grading that we
1:38:43 discussed
1:38:44 earlier can you make it more
1:38:47 dramatic from
1:38:49 a stylistic point of view
1:38:56 view can it include sound seagulls maybe
1:38:59 it cannot um sounds sounds and video
1:39:04 will probably be within within six
1:39:07 months like I'm sure I'm sure chat GPT
1:39:11 very shortly here becomes true
1:39:16 multimodal
1:39:18 um and at that point it's probably
1:39:22 that doesn't look photorealistic
1:39:25 anymore the revised
1:39:33 image okay that's better stylistically
1:39:36 but it doesn't look photographic and
1:39:38 cinematic anymore can you give me a
1:39:40 final version from the
1:39:45 past that is much more
1:39:49 photographic that retains that
1:39:54 [Music]
1:39:56 style oh by the way you could you could
1:40:01 absolutely turn this into a
1:40:03 GPT so you could you could come up with
1:40:06 a way to
1:40:09 um feed it feed it a piece of a
1:40:12 screenplay have it break that screenplay
1:40:16 down and oh look this is so cool look
1:40:19 what it
1:40:20 did th this is like all the from the
1:40:23 film you know this is like the
1:40:24 cameraman's view the cinematographers
1:40:27 view that's
1:40:30 hilarious not exactly what we wanted for
1:40:32 a storyboard but that's kind of
1:40:40 cool wow uh so let me try something here
1:40:45 um could you
1:40:50 create a
1:40:52 table using
1:40:56 markdown that
1:40:59 includes the columns for the scene the
1:41:03 visual
1:41:05 description and the embedded image so
1:41:08 that I can view it here in the browser
1:41:09 as a
1:41:11 table so markdown is a is a sort of
1:41:16 local rendering it's it's kind of like
1:41:18 mini HTML let's see if it gets the in
1:41:22 there did it oh it broke darn
1:41:46 it this is great but the images are
1:41:49 breaking can you
1:41:52 use whatever coding you used above
1:41:54 because above I can see the images but
1:41:57 in this table I
1:42:00 can't let's see what it does here I
1:42:03 understand the issue let's try embedding
1:42:04 the images using a different method in
1:42:06 the markdown
1:42:08 table oh it's putting in base 64
1:42:11 truncated truncated for brevity well
1:42:13 that means that the image won't show you
1:42:15 dumb
1:42:16 Dum oh well oh well what are you going
1:42:19 to do all right so we can sort of get it
1:42:21 working if if I if I messed around with
1:42:23 this for a while you you could probably
1:42:24 get this thing to be pretty
1:42:26 sophisticated anyway this would be a
1:42:28 good GPT to create is is create a
1:42:30 storyboarding GPT that invites someone
1:42:34 to upload a scene or write a scene or
1:42:37 have it write a scene and then it will
1:42:39 create a visualization like a a scene by
1:42:42 a shotby shot table based on the
1:42:46 screenplay and then prompt them for
1:42:49 visual style and then generate all this
1:42:51 crap and then output something and or
1:42:53 save it to a something or save it to a
1:42:55 Google
1:42:58 Drive crazy crazy crazy and you could
1:43:02 probably knock that out on a Saturday
1:43:05 [Laughter]
1:43:12 afternoon holy goodness
1:43:15 people
1:43:17 crazy it's cuckoo I tell you it's cuckoo
1:43:25 cachu hello
1:43:28 [Music]
1:43:29 hello my name's
1:43:34 Kyle I like
1:43:36 ai ai is good AI is tasty have you had
1:43:40 any taste the AI taste it you can make
1:43:45 money
1:43:47 with it's so
1:43:50 exciting
1:43:54 I don't know what I'm doing I do not
1:43:56 know what I'm doing like whatever we
1:43:58 just did
1:44:00 there that was chat add is it at its
1:44:03 finest I hope you got something out of
1:44:06 that there's there's enough people in
1:44:08 here hopefully you got something out of
1:44:09 that kind of blew my
1:44:13 mind I mean it it's kind of making me
1:44:17 want to go back I've got seven feature
1:44:20 length screen
1:44:22 plays how do you upload your script
1:44:24 script to chat GPT so you just there's a
1:44:28 little paperclip so you you have to be
1:44:30 using GPT 4 for one
1:44:35 thing GPT 4 and then that little
1:44:38 paperclip you can upload anything so you
1:44:41 want to see something fun yes Kyle of
1:44:44 course we do um let me go back to the
1:44:47 chat that I just had let me go find an
1:44:50 image that I
1:44:58 like I really like that first one where
1:45:01 he's laying in
1:45:04 bed or the one where he's smoking the
1:45:06 cigarette yeah this one all right is
1:45:09 that the download button yes all right
1:45:12 so that just downloaded that so I can go
1:45:14 to chat
1:45:15 GPT if you're in chat
1:45:17 gp4 you can upload anything so I
1:45:20 uploaded a text file you could upload a
1:45:22 PDF you could upload a Word document you
1:45:25 could upload a
1:45:26 CSV
1:45:28 um I'm going to upload one of those
1:45:31 images from the
1:45:34 storyboard this
1:45:37 one and I'm going to
1:45:44 say this is an image from a storyboard
1:45:48 for a film I am working on
1:45:52 period can you
1:45:57 please describe this in as much detail
1:46:00 as possible from a cinematography and
1:46:03 set decoration point of
1:46:07 view followed
1:46:12 by a
1:46:15 table
1:46:17 of props and their descriptions
1:46:22 [Music]
1:46:33 image showcases a moment of quiet
1:46:34 introspection in the small boat cabin
1:46:36 filled with natural and artificial light
1:46:39 sources blend to create a Serene blue
1:46:41 Ambiance evoking a sense of morning
1:46:43 Tranquility the scene is primarily is
1:46:46 lit primarily by Soft natural light
1:46:48 coming from a circular window suggesting
1:46:50 early Dawn which is complemented by the
1:46:53 Warm Glow of an oil lantern indicating
1:46:55 the cabin's lack of modern amenities the
1:46:58 subject the man with unkempt beard blah
1:46:59 blah blah the set decoration is
1:47:04 meticulous Furniture adds a touch of
1:47:06 authenticity the color palette is
1:47:08 predominantly shades of blue okay here's
1:47:10 a table of props and descriptions oil
1:47:12 lantern an old-fashioned brass oil
1:47:15 lantern alarm clock ropes and Tackle
1:47:18 ship's wheel wooden suit this is this is
1:47:21 what you have assistants on Films spend
1:47:26 hours
1:47:28 doing it just knocked that [ __ ]
1:47:32 out port hole wicker basket bed with
1:47:36 pillows unmade and utilitarian
1:47:38 suggesting a character's transient
1:47:40 lifestyle like capturing the
1:47:44 the like the creative Essence as well as
1:47:49 what it is right you you need bed you
1:47:51 need a bed with pillows but we're not
1:47:53 getting you know something from from uh
1:47:57 from a fancy store here right this is
1:47:59 utilitarian this is this is well-used
1:48:01 [ __ ] transient
1:48:03 lifestyle this is well-worn
1:48:08 bedwar I'm going to
1:48:18 say could you create
1:48:23 a production
1:48:30 schedule for what needs to be done in
1:48:33 what order to achieve this
1:48:47 shot multiple departments including set
1:48:49 design lighting camera Crews and others
1:48:52 belows a simplified version
1:48:54 pre-production phase script breakdown
1:48:55 and storyboarding one to two weeks
1:48:57 before the shoot analyze the script
1:49:00 create storyboards
1:49:02 scouting set design prop sourcing
1:49:05 casting and wardrobe production phase
1:49:08 set construction and and dressing set
1:49:10 the dress set set build the set and
1:49:13 dress it with necessary props make
1:49:15 adjustments technical walkth through
1:49:17 lighting setup day of shoot morning
1:49:21 rig up natural light sources adjust for
1:49:24 desired Shadows camera setup and
1:49:26 rehearsal principal photography
1:49:30 midday review the Daily's day of shoot
1:49:36 evening
1:49:40 amazing
1:49:42 amazing oh mying it's so cool it's so
1:49:46 cool I don't know if you've seen this
1:49:47 have you ever seen AI it's so cool it's
1:49:50 really cool you can make money with
1:50:02 ch oh man yall are cool all
1:50:07 right you remind me of Peter sers
1:50:10 inspector cluso when you say I don't
1:50:13 know I don't know what you
1:50:19 mean
1:50:26 oh ask get to use the same image at a
1:50:29 different time of day we can do
1:50:35 [Music]
1:50:40 that let's
1:50:49 see
1:51:04 okay I'd like to
1:51:07 create a
1:51:14 still
1:51:17 from the photographic image
1:51:22 of the marina at
1:51:25 dawn and keep
1:51:29 the image composition
1:51:34 identical but change the time of day to
1:51:37 be high
1:51:40 noon there should
1:51:43 be some clouds in the sky but the sun is
1:51:47 shining directly on the
1:51:49 marina
1:51:50 [Music]
1:51:52 not High Moon High
1:51:56 Noon high
1:51:59 noon and this is Dawn not
1:52:02 Dan all right let's see what we
1:52:06 get Vicky's gone now she said you are in
1:52:09 the door now okay all right great thank
1:52:19 you
1:52:22 you're all caught up
1:52:24 okay fix on I
1:52:26 did let's see what it does here we'll
1:52:30 see if it it it might not remember what
1:52:32 scene I was talking about it might but
1:52:34 we'll
1:52:38 see nope it
1:52:41 didn't and that's not really high noon
1:52:46 but look it says High
1:52:49 Noon
1:52:50 High Noon present
1:52:54 day that's more like 4 in the afternoon
1:52:57 but that's okay where it's like it's
1:53:00 like 10 o00 in the morning and still
1:53:02 burning off the still burning off the uh
1:53:05 the low Ling clouds and
1:53:08 [Laughter]
1:53:13 Mist now but what I could possibly do so
1:53:16 here's another way to do
1:53:19 this
1:53:21 we
1:53:22 grab
1:53:25 this save image
1:53:29 as Gig
1:53:31 Harbor
1:53:39 Dawn save it to the desktop we'll do a
1:53:44 new a new chat jie
1:53:47 Tai and we'll upload that
1:53:54 not
1:53:56 that that okay I'm going to upload
1:54:03 that describe this scene in as much
1:54:06 detail as possible including the
1:54:09 photographic
1:54:12 style all right so we'll get it to
1:54:15 describe it watching a masterpiece being
1:54:19 born yeah this is actually I mean you
1:54:22 know jrc it it's like
1:54:25 what's what's amazing about this is you
1:54:29 know I'm just you know this is as much
1:54:33 just entertainment just me kind of
1:54:35 [ __ ] around here but I mean imagine
1:54:38 if if you really put in the work right
1:54:43 to say hey I want to I want to I want to
1:54:46 take this screenplay that I wrote I I
1:54:48 really want to get it to the next level
1:54:49 I want to get this thing
1:54:51 fundable and you put in a
1:54:54 solid
1:54:55 week can you imagine like just me
1:54:58 [ __ ] around in in the last 45 minutes
1:55:01 or so whatever it's
1:55:03 been imagine putting in a week of really
1:55:06 focus time on this what youd come out
1:55:07 with on the other side you'd come out
1:55:10 with it's just
1:55:13 amazing it's just amazing okay so now
1:55:16 I'm going to
1:55:19 say perfect now I'd like to recreate
1:55:22 that exact same scene at a different
1:55:25 time of day
1:55:35 period how about late
1:55:37 afternoon where the sun is behind the
1:55:40 camera and lighting up the
1:55:45 marina with magic
1:55:48 hour sunlight
1:55:55 is
1:55:56 reasoning able to come up with new ideas
1:55:59 the only ability missing for True
1:56:01 AGI I would argue that the current
1:56:05 technology absolutely comes up with new
1:56:08 ideas
1:56:09 um it seems there was an issue
1:56:12 generating
1:56:19 this
1:56:20 um but the me the reasoning is missing
1:56:23 it's its ability to take autonomous
1:56:25 action and its ability to reason to to
1:56:28 basically figure out if is what it's
1:56:30 doing as it's taking these actions
1:56:33 working or not and the qar thing that
1:56:37 everyone's talking about right now is
1:56:39 apparently a layer that sits on top of
1:56:41 the current large language model GPT 4
1:56:45 or five if they're working on that and
1:56:48 it adds that reasoning component
1:56:49 component to it and that autonomous
1:56:51 component to it and if that's the case
1:56:53 then I I that seems to me to
1:56:57 be a hurdle for at least some stage of
1:57:02 AGI um try again with making the
1:57:18 image thank you JC that's very
1:57:22 sweet oh my God my phone typos oh I
1:57:26 typing on a phone is anyone else in
1:57:29 instead of hitting the space bar you
1:57:31 always hit the N key and so you just
1:57:33 have one long word with all these n's in
1:57:35 the middle of the words it's
1:57:38 like like AI should learn that I'm how I
1:57:43 [ __ ] up my typing and fix it for me
1:57:46 shouldn't
1:57:47 it yes it should
1:57:52 why yes it
1:58:02 should come on come on create that image
1:58:05 oh you you
1:58:07 stinker any recommendations now that the
1:58:10 upgrade to GPT 4 is not available yeah
1:58:13 Vic so if you go to co-pilot
1:58:17 microsoft.com so so a couple of things
1:58:20 um
1:58:21 Claude now has a 200,000 token context
1:58:25 window and it's got a it's got a free
1:58:28 option um perplexity doai has a free
1:58:31 option and apparently you can upload
1:58:33 300,000 Words which is like a 400,000
1:58:37 token context window which I don't
1:58:40 believe but maybe that's true someone
1:58:42 told me that yesterday on the live then
1:58:44 you can also go to oh this is the AI
1:58:47 Salon you can go to the AI salon and
1:58:49 join that that I'll tell you how to do
1:58:51 that in a second but you can also go
1:58:55 to um
1:58:58 co-pilot
1:59:01 dot
1:59:05 microsoft.com and so this is oh I have
1:59:08 to sign in to Microsoft I think this
1:59:11 screwed up for me the other day okay now
1:59:13 I'm in okay so let's let's try
1:59:16 this can I upload
1:59:21 can I upload here add an image oh I
1:59:25 can't huh oh and it's only got a 2,000
1:59:28 token context window all right but I
1:59:31 should be able to let's
1:59:34 see let me
1:59:37 upload from this
1:59:41 device all right we'll do that we'll
1:59:43 upload that
1:59:48 image
1:59:51 describe this still
1:59:54 from a film's storyboard in as much
1:59:57 detail as possible including the
1:59:59 photographic
2:00:03 style so this
2:00:05 should acts relatively like what we just
2:00:08 did over in chat GPT but
2:00:13 Microsoft is sometimes a little
2:00:16 weird and by a little weird I mean
2:00:18 shitty
2:00:23 all right this is a still from the film
2:00:25 storyboard the image is a wide shot of a
2:00:27 harbor with boats
2:00:30 and
2:00:32 great recreate this image at
2:00:39 a a different time of day question mark
2:00:46 or no no question
2:00:48 mark let's see what it does I'm sorry
2:00:51 but I'm not be able to recreate the
2:00:53 image however based on the
2:00:55 description I can suggest
2:00:59 some all
2:01:01 right create an image based on the
2:01:07 description it should be able to do that
2:01:10 see this is what it just did there this
2:01:12 is what I experienced with Microsoft all
2:01:13 the
2:01:16 time you could use an AI image generator
2:01:19 this is an AI image
2:01:22 generator although maybe they didn't
2:01:24 bring Dolly over to this I don't know
2:01:32 see let's go try
2:01:35 Bing chat.
2:01:37 [Music]
2:01:39 bing.com ah this is maddening no this
2:01:42 looks like it's the
2:01:44 same let's try
2:01:48 it
2:01:50 upload from
2:01:53 device please describe this in as
2:01:59 much detail as
2:02:03 possible all
2:02:05 right chat GPT made a joke last week
2:02:08 unprompted during a conversation I had
2:02:10 with it
2:02:11 wow that's wild well it looks like grock
2:02:14 you can put grock into funny mode and it
2:02:16 looks like it's going to just be a real
2:02:18 cutup
2:02:20 this is a photo okay good beautiful
2:02:23 terrific um create a new version of this
2:02:29 image at a different time of day spelled
2:02:35 wrong but that's
2:02:39 okay no it
2:02:42 doesn't I don't understand what
2:02:45 Microsoft is doing this used to have
2:02:46 Dolly built into it and now it seems
2:02:48 like it doesn't
2:02:51 I don't get it and if if I go to
2:02:54 bing.com
2:02:56 create that should be um dolly3 SLC
2:03:03 create jump start your
2:03:08 creativity but I can't upload an image
2:03:10 here I don't know so I'm totally
2:03:14 confused right now so we are discovering
2:03:17 together that it looks like have gutted
2:03:21 co-pilot to remove all the multimodal
2:03:24 [ __ ] from it except it can describe an
2:03:27 image it can see it can see images but
2:03:30 it's not able to create them which is
2:03:32 weird so that might just be a temporary
2:03:35 thing but that's very strange anyway
2:03:39 um I was going to say you could use
2:03:42 co-pilot but now you can't so
2:03:48 um
2:03:50 I got nothing I got nothing for
2:04:00 [Music]
2:04:03 you sometimes this shit's really
2:04:06 frustrating if if if by the
2:04:10 way if when you're using this AI [ __ ]
2:04:12 you're like it seemed so good when Kyle
2:04:15 did it now I can't get it to do
2:04:18 [ __ ] I have happens to me multiple times
2:04:21 a day where there are sometimes where I
2:04:23 just get so pissed off with it I just
2:04:25 like bail on the project and then there
2:04:28 are other times where it's like so
2:04:30 magical I'm like it can do
2:04:33 anything and that's just the nature of
2:04:35 where we are right now so so don't let
2:04:38 its lack of consistency dissuade
2:04:42 you just keep playing and keep finding
2:04:45 the boundaries and then what what what
2:04:48 you'll find is that
2:04:50 that you'll try to do something and
2:04:52 it'll suck at it and then you'll go try
2:04:54 other tools and you'll get something
2:04:56 close and then you'll just realize okay
2:04:58 this stuff isn't quite there yet and
2:05:01 then you'll come back like 3 months
2:05:02 later and you try the same thing and
2:05:04 you're like oh holy
2:05:06 [ __ ] oh it does it now it's
2:05:11 like it's it's um it's been progressing
2:05:15 per
2:05:16 fast per fast
2:05:20 yeah mhm I like them taters mhm Tater
2:05:24 taters tasty
2:05:26 mhm they call it a sling
2:05:28 [Music]
2:05:38 blade I think I'm I think I'm crisped
2:05:41 out I think I might be
2:05:43 done let me see if we got any
2:05:46 other stuff here let me you know what I
2:05:49 want to do um since there's a bunch of
2:05:51 new people in here let me give you get
2:05:53 ready to do some screenshotting they
2:05:55 call me tater salad that's it Mr
2:06:00 K um okay if you're brand new to this
2:06:05 here AI [ __ ] if you're like Ma what you
2:06:09 were doing look cool and everything I
2:06:10 don't know how to do any of that crap
2:06:12 I'm going to I'm going to give you some
2:06:14 some stuff to play with I'll give you
2:06:15 some homework we got Mr K here he's an
2:06:18 educator he will give you homework and
2:06:20 this is this is my son tater tot
2:06:26 exactly um go there right now if you've
2:06:29 not played with chat gbt I am wagging my
2:06:32 finger at you and you know what that
2:06:34 means I don't either but go there right
2:06:38 now and just start playing with it ask
2:06:40 it to write you a poem ask it to write
2:06:42 you a project plan ask it to to invent a
2:06:45 new kind of pizza for you what like
2:06:47 whatever you're interested in just ask
2:06:49 ask it to do that try to stump it if you
2:06:51 think you're super clever on a
2:06:52 particular kind of trivia ask it about
2:06:55 that see how it does and just go start
2:06:58 playing with it and just play with it
2:07:00 until you have what I call your Kevin
2:07:03 McAllister moment which
2:07:05 is it can do
2:07:07 that um maybe you've seen some of that
2:07:10 tonight some of that storyboarding stuff
2:07:12 I was just doing like kind of is melting
2:07:15 my face like
2:07:17 I I wrote that screenplay 20 odd years
2:07:21 ago and watching
2:07:24 it synopsize my
2:07:27 screenplay break the scenes down into
2:07:30 visuals and then create visuals like it
2:07:33 looked like what I had in my head it's
2:07:36 like are you kidding
2:07:38 me so so this stuff's just wild so so
2:07:42 anyway I made a chat GPT for your button
2:07:45 oh for this button you make money with
2:07:47 chat GPT
2:07:49 we need a you can make money with chat
2:07:51 GPT button I think I might have I might
2:07:53 have made one called that you can make
2:07:55 money with chat
2:07:58 GPT um but that's awesome that's awesome
2:08:00 share it on the on the salon I'm gonna
2:08:02 I'm GNA talk about the salon here in a
2:08:04 second okay so so that's chat GPT chat
2:08:07 GPT in my opinion right now there's
2:08:09 nothing even close to it GPT 4 okay the
2:08:13 problem is if you don't have access to
2:08:15 chat GPT 4 right now they're not taking
2:08:17 new uh
2:08:19 accounts so get on the wait list um if
2:08:22 you do have access to it and you haven't
2:08:24 really dug into it start digging start
2:08:27 digging because there's all sorts of
2:08:29 [ __ ] it does including these making
2:08:30 these gpts okay
2:08:34 but I would encourage you all to get
2:08:37 very very curious about this space in
2:08:39 general I'm now going to show you eight
2:08:41 different alternatives to chat GPT that
2:08:44 are all worth at least exploring
2:08:46 okay um Claude grock Bard and co-pilot
2:08:51 are kind of your big four um grock isn't
2:08:54 out yet that's going to be the one
2:08:55 that's inside Twitter x x.com and
2:08:59 x. um Claude is anthropic that's a big
2:09:04 competitor to chat GPT Bard is from
2:09:07 Google co-pilots from Microsoft that's
2:09:10 ostensibly chat GPT 4 but whatever I
2:09:13 looked at tonight was the gutted version
2:09:15 of it so that was very disappointing so
2:09:17 I'm hope hopefully that's just a
2:09:20 blip um these are another four that I
2:09:23 think are worth exploring Pi perplexity
2:09:25 and po are all three different very
2:09:28 specific kind of flavors of these chat
2:09:30 gpts and then llama 2 is kind of a whole
2:09:33 world in and of itself uh that meta put
2:09:36 out these open-source uh large language
2:09:39 models called llama uh that's why we
2:09:41 have llamas in here a lot because every
2:09:43 time I mention llama someone pops a
2:09:45 llama on
2:09:47 screen um and then here's kind of four
2:09:50 different image generation tools to go
2:09:52 play with and I'm not going to go beyond
2:09:54 this so screenshot
2:09:57 that screenshot that if you
2:10:01 haven't and then screenshot that and
2:10:06 then if if you are curious about this
2:10:09 and you're like you don't have you
2:10:11 haven't found your peeps yet like if all
2:10:13 the people in your life are like that AI
2:10:15 stuff's stupid it's evil it's really
2:10:17 dumb why are you wasting your life
2:10:19 if you got a lot of that going on and
2:10:21 you want to hang out with people that
2:10:22 are up to the same thing um join the AI
2:10:25 salon so this is a group that I started
2:10:27 a year ago next Tuesday is it next
2:10:30 Tuesday no a week from Tuesday is our
2:10:34 one-year anniversary so I started the
2:10:36 salon the week after chat gbt came out
2:10:39 cu when it came out I lost my [ __ ]
2:10:42 mind and so go to AIS salon.
2:10:46 mn. that's the that's the the address
2:10:50 please go to the about the salon
2:10:56 page that will tell you what we're all
2:10:59 about community of extraordinary
2:11:01 adventurers exploring the AI unknown who
2:11:04 we are why we're here how we explore
2:11:07 what we seek and what our values are and
2:11:10 please please please please it is my
2:11:12 deep request please go look at the
2:11:14 values of this community I want this
2:11:17 community to grow but I don't want it to
2:11:20 lose what we have right now which is
2:11:21 pretty [ __ ] magical so the values are
2:11:24 curiosity generosity like share what
2:11:27 you're
2:11:28 learning
2:11:29 exploration
2:11:31 collaboration inclusion and
2:11:35 empathy and and you know go read what
2:11:37 the values are really about and what the
2:11:39 community is about if you resonate with
2:11:41 these jump in start exploring start
2:11:44 being generous start being curious um if
2:11:46 if they don't resonate with you that's
2:11:48 totally cool but it's probably not your
2:11:50 group of people you probably will not
2:11:52 like
2:11:55 us uh if you see things about hair
2:11:58 you're in the wrong
2:12:02 place oh yeah yeah exactly if you go to
2:12:06 uh I did this the other day if you go to
2:12:08 the salon. mn. that's a hair place I
2:12:13 gave that uh URL out for for a while on
2:12:15 the first day when we moved over to this
2:12:17 new platform and that was that was a
2:12:19 mistake so anyway so go there join us
2:12:23 and um do me a favor follow this channel
2:12:26 I'm trying to trying to grow this
2:12:27 channel so follow this if you like what
2:12:29 was
2:12:30 here I want to say thanks for suggesting
2:12:32 pie she has been super helpful oh that's
2:12:35 that's terrific n Nash
2:12:38 n Asha or Asia NAA NAA I think I like
2:12:43 that
2:12:44 name um good afternoon for me and good
2:12:47 night to all night good afternoon drew
2:12:50 the values are the best part of this
2:12:52 group I'm stealing these values for my
2:12:54 classroom rules oh awesome I know I'm
2:12:56 really what's cool about the values is
2:12:59 we didn't just like make up a pile of
2:13:02 values we sat down and we said okay we
2:13:05 want to grow the salon but how do we not
2:13:06 lose what it is and so those values were
2:13:09 derived from how people were
2:13:12 acting and so those are just a
2:13:14 reflection of what's already there and
2:13:16 so that's when when when values are
2:13:20 right that's what they are they're just
2:13:21 a reflection of what's already there but
2:13:23 but they're super important because as
2:13:25 things grow you can lose that so that's
2:13:27 why I push push hard on on the values so
2:13:32 all right everybody I'm going to get my
2:13:34 ass out of here into bed multimod modal
2:13:37 Works in co-pilot try create an image of
2:13:39 a Philadelphia Eagle defeating a buffalo
2:13:42 yeah it might be just I sometimes when I
2:13:45 log in and it says I've got a 2,000
2:13:47 token cont context window it's non-
2:13:50 multimodal gp4 and then there's other
2:13:53 times I log in and I've got a 4,000
2:13:55 token context window and then it does it
2:13:58 so I don't know what's going on there
2:13:59 some something's weird on my end tonight
2:14:02 so thanks Kyle Kyle is C thank you very
2:14:07 much Irregulars mods present Turtles
2:14:10 llamas or
2:14:12 hedgehogs thank you Clint Copperhead
2:14:14 thank you mods um thank you everyone who
2:14:18 uh supports this and shows up
2:14:20 nightly let's see Clint
2:14:23 Copperhead who else we got jrc good to
2:14:25 see you Tobias Mary Mary as always good
2:14:28 to see you Mr K you're
2:14:30 amazing good night cor Corey Sandler
2:14:34 good to see
2:14:35 you
2:14:38 uh quick take Jake good to see you
2:14:42 Amelio's wife you were
2:14:45 late random thoughts by TK good to see
2:14:48 you all right
2:14:50 everybody um I'm Crispy drink your
2:14:53 fluids I got to go drink mine peace out
2:14:55 steo um thanks for your help you're
2:14:57 welcome thanks everybody for hanging out
2:14:59 and hope you had a good time and go join
2:15:01 the salon
2:15:04 peace