Jan 24, 2024 - Create Your Own GPTs and Transform Content Creation

Transcript

1. 0:00some good singing oh yeah you're ready I can tell oh you're ready you ready we do a little Jesse Winchester [Music]
2. 0:16[Laughter] [Music] huh the boys was singing shingling the summer that we met
3. 0:31you were 10 and 17 how could I forget every Star from near and far was watching from
4. 0:47above watching two teenagers fall in [Music] love the way we danced was not a dance but more a long
5. 1:03Embrace held on to each other and Flo there in space and I was shy to kiss you while the whole wide world could
6. 1:19see so shinging said everything for [Music] me and all the poor old old folks they thought we' lost our
7. 1:34minds they could not make heads or tals of the Young Folks funny [Music] rhymes but you and I knew all the words and we always sang
8. 1:50along to Old Cham ding sh dang [Music]
9. 2:08[Laughter] so after years and after tears and after Summer's passed the old folks tried to warn us wait the old folks Tred to want us our
10. 2:25love would never last but all we yet was soaking wet from walking in the rain and singing shammer shingling
11. 2:42again and all the poor old old folks they smile and walk [Music] away but I bet they did some shamama
12. 2:57Lama a ding dong in the [Music] day I know I should learn this song better right that's a beautiful Jesse Winchester
13. 3:12[Music] song the boys were singing Shing the summer that we
14. 3:28met I don't even deserve to play [Music]
15. 3:43[Music] [Laughter] [Music]
16. 3:58today [Music] you and I here all alone Sunday morning here at
17. 4:13home sky blue and a cough is strong it's true thank you Robert Rossy good Lord very generous thank you so
18. 4:23much good day good day Cindy [ __ ] sitting in SFO she boards for as in a couple hours and this is how she spends
19. 4:32her time good Lord a singing dog an attitude [Music] [Laughter] [Music]
20. 4:57guitar you and I all long Sunday morning hair at home sky blue and the CU is strong it's
21. 5:13true then I open my eyes to dream realiz in front of me and I hav't got a clue what in the world is happen to
22. 5:28me think I think I'm happy like first day summer vacation happy got to get a little rest and relaxation
23. 5:40happy like to quiet on Sunday morning singing [Music]
24. 5:58true [Music] who you good that
25. 6:16[Music] enough I think he's done I think he's digging this singing I I think he's digging
26. 6:30it he comes he comes up here now he lays right by me he knows what's coming he knows he's singing for his
27. 6:39supper this is this is all for Pate this is for chicken sticks this is all for p subscription pays for your chicken
28. 6:49sticks so you got to do it you can't not sing welcome everybody happy Wednesday had kind of a nothing day today just messed
29. 7:04around messed around all day didn't see anything too remarkable out there like I said I think we are be being lulled into a false sense of
30. 7:15security I think some shit's about to hit the fan with this new with some new AI stuff but uh but we'll have to see
31. 7:23we'll have to see maybe nothing's going to happen in 2024 maybe 2023 is not the last normal year I think maybe 2023 maybe was just a
32. 7:32year 2024 is going to be non- remarkable very possible and then we'll just sing [Laughter]
33. 7:47songs all right people let's get this show started for the chicken champy singing for the chicken only music itions don't follow the
34. 8:03singer what are you smoking did you wait did you not see mid Journey replacement for free no I didn't wait what's that what's mid
35. 8:15Journey replacement for free what the what the [Music] heck
36. 8:30oh that was wrong right mid journey
37. 8:49journey mid Journey just pushed out some updates to the web Alpha can you you make can can non wait pix art-
38. 9:07Alpha but is it is it mid Journey pix art-
39. 9:25Alpha Transformer based text image diffusion system hm no I don't know this I do not know
40. 9:40this it looks like it's a paper with a little but it looks like you can play with it yes is this the one you're talking about
41. 9:54Sherry D oh you're not looking there but so it's a it's it's a hugging face spaces right now yeah I mean it like it quality-wise
42. 10:04it might be there but what I'm looking forward to is mid Journey like mid journey I actually meant to show this the other night mid
43. 10:13Journey um they're improving their their website so if I sign in well let's get started we'll get started here we'll come back to
44. 10:27this oh I got to do multiactor authentication what's the latest on Gemini I don't think anything but maybe I missed some [ __ ] today
45. 10:40Gemini uh Ultra news about Gemini nothing honey nothing
46. 10:57honey nope nothing we ain't got we ain't got no news all right everybody so welcome welcome welcome this is the uh this is the AI learning
47. 11:10lab I'm stumbling around in the dark here bunch of Google texted video papers came out yeah I saw a couple of those
48. 11:17those were okay you need to be using Suzanne's GPT for AI news yeah I should do that although the last time I tried to use
49. 11:24it she did not work what's it called again Source camp [Music] let's see
50. 11:38suzan um Welker is it Welker is that right no w e MK [Applause]
51. 11:53no AI daily rundown AI daily run down s Welker Jurgens okay today's AI
52. 12:11news today's AI news let's see if anything comes up yeah Google put out a couple of papers today on some video
53. 12:20stuff that looked pretty cool the the problem with the with the Google papers coming out as they do the it's it's a
54. 12:27preview of what's to come but it's not really what's there like they've got some really exciting papers they put out
55. 12:33that you can get to if you build using their API stuff Suzanne I don't know if you're on here oh now it's working error creating
56. 12:46image this this GPT breaks for me for some reason it's weird Google back's National AI research pilot okay cool Google Google's gradient
57. 13:00invests in send AI for data extraction for complex documents all right so they invested in a startup service now rise in cyber
58. 13:14attacks yeah I don't think there's anything super massive that I saw anyway so welcome uh my name is Kyle Shannon
59. 13:23this is the AI learning lab so what we'll do here is we'll get started we're not really started yet we're just kind
60. 13:27of just kind of of babbling just kind of babbling back and forth um and then we'll show some stuff we'll talk about some
61. 13:38stuffs 30 minute hellos exactly Danielle Danielle's got it figured out she understands how this works let me get my
62. 13:46little mod Channel up in case there's any panicked Kyle you got to go look at something SpaceTime text to video
63. 13:56diffusion model by Google research all right so oh Lumi Lumiere didn't we look at this last night Lumiere Dov video.
64. 14:08GitHub is this new hasn't this been out did they update it oh they're
65. 14:25using oh looks like they rolled a bunch of things together this looks like they rolled rolled style drop into some video
66. 14:36stuff so let's this is actually probably worth looking at H let's see Loom
67. 14:52Year Dot video. GitHub Doo that's spelled right oh Dash
68. 15:10video lumere DV video. github.io let's watch this video so so it looks like they're they're rolling some tech together I don't know if I
69. 15:20don't know if uh P's in here um but um but let's let's just watch the video it's text to video so they got text to video image to
70. 15:35video stylized generation so this is that this is that style drop thing although this doesn't look nearly as
71. 15:46good as the when it was just images on their own and then there was video stylization so you can take a source video and just
72. 15:54describe different styles and it'll recreate it then there's cinemagraphs which are just animated gifts and
73. 16:02there's video in painting which is cool like change her outfit huh let's see what they say about [Music]
74. 16:27it [Music] is it all [Music]
75. 16:48[Applause] [Music] square [Music]
76. 17:27w [Music] [Applause] [Music] [Applause]
77. 17:41[Music] [Applause] [Music] all right that's interesting I mean but if you if you remember there's this
78. 17:57thing called style drop style drop Google that they're using they're using some of the same Tech in it it looks
79. 18:10like this style transfer technology but this looks really slick where where is it oh this was just an article oh
80. 18:24God are you going to give me a hyperlink to the actual site you dumb DS anyone no I'm screw you I can't I can't
81. 18:34deal with I can't deal with um what are they called um media sites that don't link you out of their site they're like well we don't
82. 18:45want them to leaving to site or they'll never come back well guess what if you don't link me out of your site I'm not
83. 18:49going to stay there in the first place you big fat Dum Dums okay so this is this is a paper this is a paper that this came out this
84. 19:03was ages ago this came out like June maybe June or July and then they've got this you know you put one image in and then it makes
85. 19:15these static images that are identical to that input image and they have a bunch of different styles and these
86. 19:21things look great and then what they just showed in the video like it was this same input image and it sort of L
87. 19:28looked like this stuff but it looked like [Music] [ __ ]
88. 19:44[Applause] [Music] [Applause] [Music] here
89. 19:59[Music] see like
90. 20:15the I don't know the there's something about the video it just it looks it looks really janky but the static images look [ __ ] perfect
91. 20:27like there's the how do I go left here can I go back where's the left Arrow there like look how good those so
92. 20:43there's the input like sticker and look how good those stickers look and then if we go to this this here VI there's the
93. 20:56sticker see how Jang they look they just don't look they don't look good so anyway all right whatever and it only
94. 21:04looks like they're making Square video the research stuff is rarely able to go into production immediately that's why
95. 21:10it's research that's from Pate from Google thank you Pate yeah I know and and that's like as I'm looking at this
96. 21:16this isn't ready to go anyway like it's it's Square only video but but it what what's really cool here is you can see
97. 21:23where it's going right you can see where it's going to it's going to end up being this tool that is you know you can
98. 21:31generate it you can start with video you can start with nothing you can start with an image you can start with a style
99. 21:36you can in paint you can out paint you can right so so you can see where this is headed but what we don't know is when
100. 21:44are mere mortals going to get their hands on this [ __ ] right and that's that's again that's more what this
101. 21:48channel is now if you're a developer you should definitely be if it was open AI you'd be using it well if it was open AI
102. 21:58have a I'd have a consumer interface I could use it on not a development environment like like Pate I don't I you'll notice I
103. 22:09don't hang out in in open ai's playground I I hang out in chat GPT right I could hang out in Google's
104. 22:18development environment but but it's not hey I'm not a developer and you you did a Freudian sleep slip if it was open AI you'd be
105. 22:30losing oh I'd be losing it no I wouldn't I no I would still be bitching about this the fact that it can't do 16x9
106. 22:37video it's only showing Square video and the fact that the the degradation in image quality when you animate these
107. 22:44things is so dramatic no I'd be [ __ ] on open AI I dude I [ __ ] all over the open AI store the open AI store is a
108. 22:52piece of steam and poo it's not even a store have you looked at V v. yeah I looked at I looked at that one that was
109. 23:02pretty good it's okay there's a I mean here's the problem with all these individual little tools is like there
110. 23:07are too many of them right there's just there's thousands and thousands and thousands look this is the the GPT store is is [ __ ]
111. 23:17horrific it's it's not have they have they added more than seven SE sections and is this more than just a an anchor tagged web page no it's
112. 23:26not and do any of these go more than 12 featured things see more 12 no no this is a steaming pile of poo
113. 23:38like I think the the plugin store was better than this so I I uh but you know chat GPT is
114. 23:53still better than Bard so until until that that changes Pate that's where we're starting we're starting with this is
115. 24:03state-of-the-art and when they put out piles of [ __ ] I'll say it they're they this store is really [Music]
116. 24:13bad what's the name of the website that that Google website was lumere l u m i e r e lumere vid or- video. github.io
117. 24:30and this is chat. open.com yeah there you go thank you Brandon thank you thank you thank you did you see the GPT to
118. 24:40search the GPT store someone created it no of course they did cuz someone had to create something that sucked less than
119. 24:49the the [ __ ] site that they built I can't believe we had to wait three months for that Steam and P laoo Sam Alman announced that at the
120. 25:01beginning of November and then they didn't launch it in November because they decided to fire Sam Alman and then they waited an
121. 25:11additional two months to launch something that like that this must have been what they coded in November and then they just got busy
122. 25:23trying to you know lick their wounds after F firing Sam Alman or deal with AGI or whatever the [ __ ] was
123. 25:30going on over there but they they have they they're like the they're clearly not taking the the GPT store very
124. 25:37seriously right now and it might be listen who knows why right it might be that um they know what's coming and they
125. 25:47don't want to put time into this CU they're doing something different I like I don't know but but it's it's it's a steam and pile of poo all
126. 25:55right let's see let's get going going I haven't even gotten started yet we're all we're all running down all these rabbit holes
127. 26:02already welcome everybody if you're new to AI um we will go look at [ __ ] um I'll give my my two cents on it um I tend to
128. 26:14be uh an AI Optimist but I try to be decently realistic um other than the fact that I do refer to what it does as
129. 26:24magic more than Pate is comfortable with Pate works at Google he like it's math this is a probability engine dear
130. 26:32sir we shall not refer to mathematics as magic unless it works like [Music] magic yeah I you know it's it's close
131. 26:48enough it's close enough for for me check out GPT finder in the GPT store okay I will check check that out all right
132. 27:00fantastic fantastic Bob Hey Hey listen what are you doing tonight what are you guys doing tonight are you going to go
133. 27:07look on the internet for some of that AI stuff huh are you go hey hey Kyle you going down to the mall today you going
134. 27:16to the orange julius hey Kyle you going down to the mall today I grew up in central Pennsylvania that's what I listen to for most of
135. 27:26my schooling years was that accent I had to get out I had to get out of there you going down to the mall
136. 27:38today Kyle hey Kyle you going down to the mall got to change oral in the car I've only ever seen in my lifetime
137. 27:51I've only seen one thing on TV that had that accent and it was the Tracy Olman show and she did a skit called the
138. 28:02Baltimore steps and it was about she played a like a girl leaving home and going to college and everyone was crying
139. 28:10and she was packing her stuff and trying to leave and they wouldn't let her leave and that was the accent they all had and
140. 28:16it was it was [ __ ] perfect I was like holy [ __ ] somebody got it you going down in a mall today d
141. 28:31[Music] you can take the magic out and it'll still work fine take the math out and it's useless that's a very good point Pate very good
142. 28:49[Laughter] point letter Kenny hi AI people hello ISU what's happening welcome welcome
143. 29:03welcome I used to sound just like Blake station he's from my hometown I fixed it my Uncle Earl lived in
144. 29:17Pittsburgh Uncle Earl Earl would change change my car but magic magic makes it pretty well I I'll tell you what makes
145. 29:26the the AI stuff magic is what it was trained on right it's the it's the p in GPT pre-trained what what makes chat GPT so
146. 29:37remarkable the math makes it possible but it's the data that's in there that it's what has been trained on right and
147. 29:44so that's why it you know it responds to you in ways that are like seemingly human and you know you you're kind of in
148. 29:53this weird have we passed the Turing test I'm pretty sure we're passed the Turing test right we're kind of in that
149. 29:58land um and that's like it's funny like why I refer to it as magic is it's just like it it's just um I've spent my whole [ __ ] life
150. 30:10looking at sci-fi where it's like that's not going to happen in my lifetime and then I'm sitting here playing with the
151. 30:16exact same computer I've been using you know for 30 years well not same computer but same basic interaction and all of a sudden
152. 30:25it's acting different and it's it's it's wild I'm so ashamed of the GPT store that I don't want to make a follow-up
153. 30:34demo I promised my audience so Cena here's the deal keep building keep showing stuff just because the the GPT store is not
154. 30:46what it it needs to be um don't stop sharing [ __ ] talk about the store and you can talk about hey this has got a long way
155. 30:56to go right now it's basically you can get stuff up there and people can find it are people going to come to that
156. 31:01store because of the store no but if you build really good gpts that people can use put them out there as a calling card
157. 31:09build calls to action for your services into the gpts just keep building keep building keep building keep talking
158. 31:15about it if you've got an audience keep going and if you're [ __ ] bored with chat GPT show them image stuff make videos do a suo song
159. 31:28and I think that's what what I'm going to do I think what I'm going to do for the next next few days um is is I'm going to take the the list
160. 31:39of of recommended sites that I tell people to go check out and I'm just going to pick a random one or two and
161. 31:46we'll go look at them and see how they are um because these things are changing really really fast and so we'll we'll
162. 31:52we'll just sort of putter around to to different things um if you got questions about AI pop them in the comments below
163. 31:59the mods do a really good job of of you know sort of floating floating questions up so I can see them because as as I
164. 32:07start demoing stuff I get I just you know I I kind of lose my mind and can't look at everything so so just be patient
165. 32:15and your your your questions will will be floated to the top oh that's really nice who did that Kevin Vasquez I'm graduating
166. 32:28someone's got their AI learning lab certificate if you're really nice if you're really nice uh I'll give you an Irregulars I
167. 32:41think we have to make some Irregulars uh diplomas I think I think it's important question I get wait question I
168. 32:51get is Will these gpts and info be available for those not on a play paid plan not until further notice Lori I
169. 33:01mean as far as I can tell um gpts right now are what did Sam mman say I wrote it on my
170. 33:21whiteboard he said something about chat GPT and he said or cat gp4 and he and he he basically said chat gp4 is just for practice like
171. 33:35he he kind of disparaged chat gp4 as that's just to get people ready for what's coming and so I think that gpts kind of
172. 33:47live in that space so what I take from gpts is we have now we have now moved into a world where non-programmers are
173. 33:57going to be able to generate sophisticated applications like anyone will be able to gen generate sophisticated applications
174. 34:10um the form of gpts right now my guess is that they won't be anything like that 6 months from
175. 34:24now and in in my Olden timey mentality I'd have been like well [ __ ] it why would I build it but in my in in you know knowing what
176. 34:41I know now or you know have having been having been through this before with the worldwide web what I know what I know is
177. 34:48that any time that you spend right now building stuff building prompts building gpts figuring out how to do do images
178. 34:57figuring out how to make videos that have music behind them and all that sort of stuff is training you to be ready when
179. 35:09the cool [ __ ] comes and you're like but Kyle it's already so cool it's already so amazing yeah we you don't I don't
180. 35:18know if you follow Matt wolf Matt Wolf's the one that put together futurepedia doio he said um he said in a tweet like two days ago he said
181. 35:31um like he was he was thinking that the ability to do like a 30 or 60 Minute video from a prompt was probably two or three years
182. 35:46away and and he said I was wrong wait till you see what's coming in two or three months so I just think I think on all
183. 35:57levels I think when when Gemini Ultra from Google comes out I think when GPT 5 comes out which I mean that could be
184. 36:04later in the year but we don't know when whatever the next generation of these video tools come out um we're going to have to relearn
185. 36:18everything and you should be building today because here's what you're doing you're establishing yourself as an expert in generative
186. 36:28AI You're Building skills and and more than building skills you're building an awareness that something new is
187. 36:39possible and not an awareness like oh something new is possible yeah ai's going to do cool stuff no like internalizing like oh [ __ ] I'm a
188. 36:49horrible illustrator but that no longer matters if I want to make art oh [ __ ] I can't sing or write music super good but
189. 36:58now I can oh [ __ ] I wasn't a cinematographer but now I am oh [ __ ] I wasn't an application developer but I've
190. 37:05made if your name's Daisy Thomas more than 80 gpts 80 applications she's put out in the world Suzanne's done 40 an's probably done 40 or
191. 37:2150 right so so that's why to do this now but I wouldn't get too caught up in yeah but no one's going to use my GPT
192. 37:35because they don't have access to GPT 4 at some point open AI will probably do that just understand the game you're
193. 37:42playing and do the best you can within the game you're playing and so open AI puts out a shitty store
194. 37:50okay the fact that non-programmers still can still make thank you very much the fact that fact that non-programmers can
195. 37:58still put together these incredibly sophisticated gp4 custom gpts is miraculous people are discovering whole
196. 38:11new ways to express themselves because of this and so what those people are going to have an advantage on the rest of the
197. 38:20people is when the next version of these things comes along they're going to be like I know EX exactly what I want to do now it can do
198. 38:28that now holy [ __ ] let me take what I learned in my gpts and jump over to whatever the next version of them becomes and blow the
199. 38:38doors off the thing Source Camp I hope what I'm learning gives me a bit of foundation for the next oh it absolutely will
200. 38:46Corey see see I think Corey's doing it right Cory a lot of people in here are doing it right what a lot of people in here are doing is is
201. 38:58they're taking who they are as human beings and they're finding new ways to manifest their talents with these new
202. 39:07tools and it's really [ __ ] exciting and it's really empowering and it's really inspiring and so yeah Corey I think you
203. 39:18I think I think all of what you're learning is transferable who should I follow to learn prompting um
204. 39:33here's why it's a complicated question it depends what you want to do with your life if if you're just talking about
205. 39:39learning prompting as you know using chat gbt type tools um you can learn structured prompting structured prompting is not a bad thing
206. 39:50to learn um you can you can follow re Rachel Woods she's one of the people I recommend you follow um there are all sorts of videos on
207. 40:02prompting the the challenge with it is like like this this is one of those I feel like it's it's important right now to
208. 40:12play um so you can so what I would say is follow some people to learn some fundamentals but don't get to locked into there's a way to prompt
209. 40:27because if you look at what's coming with Gemini Ultra where you've got images are are as important in the prompt as
210. 40:38words or if you even in chat gp4 right now if you use GPT Vision you can upload a picture of a whiteboard and just say
211. 40:47what is this and it analyzes what was on that whiteboard that's that's not really a prompt but that is like You' you've got to understand
212. 40:57that prompting is just an input to the computer system and there's going to be lots and lots of ways to prompt so so
213. 41:04learn the concept of prompting more than the Tactical details because these systems as they get better at
214. 41:13reasoning and as they get better at here's a good here's a good word for you get your get your notepads out people as
215. 41:20they get better at a gentic behavior where they're taking action on your behalf they're going to prompt
216. 41:28themselves all you have to do is give them a goal so if you want to be a developer and work on the back end of these
217. 41:38systems where you're putting the prompts in that make these things do their stuff then it probably makes sense to
218. 41:44learn a more sort of technical approach to prompting but if you're just a human being trying to figure out how to do AI
219. 41:52just start playing follow Rachel Woods she's got a very she she sort of attacks this stuffff in a much more structured
220. 42:00way than I do right I'm this is Chad add this Channel's Chad add like my prompting style is fast and loose like can I get it can I get it
221. 42:13there it's 80% awesome I'm done other people are like I'm going to sort of prompt the [ __ ] out of this
222. 42:20thing and I'm going to do details and I'm going to she's more like that there's no real white right way
223. 42:27right now and it's changing like it's changing dramatically fast in fact a thing that would be really
224. 42:38interesting actually let's go to chat so here we are chat GPT right and so you're like well I don't know anything about
225. 42:44prompting okay fine let's go over to the internets and look up uh uh Civil War
226. 42:59letters images that's a cool
227. 43:24one okay I'm going copy that image I'm going to go back to my chat GPT I'm going to paste that image and say what is
228. 43:43this appears to be an old postcard with both an illustration and handwritten text it depicts a Native American scene
229. 43:50with an individual individuals in traditional clothing some on Horseback with the American flag present the handwriting seems to be a
230. 43:59personal message from someone named Henry who is a recipient name isn't entirely clear because it overlaps with the
231. 44:06illustration uh so you can say can you read it best wishes Henry Sneed dear sister I
232. 44:21received your something let me say try again try again you can do it apparently um giving uh chat GPT
233. 44:38encouragement look dear sister I received your loving letter on the 26 and was glad to hear from you once more
234. 44:45that you getting along so well I'm in the best of health thank God I'm not all Idol but have the sweet
235. 44:52feeling of success and cheerful thought of honest days toiling toiling Throne at large something my only earning and saving I'm
236. 45:04doing fairly well and and am quite well known here in somewhere and have a good many friends the message continues but it
237. 45:13comes in increasingly difficult all right let's go find another one because that wasn't really Civil War let's see if we can get a Civil War
238. 45:24One that's got a date oops see if I can read this one copy image and you're you're like what are you what are you talking about I ask you
239. 45:39about prompting what is
240. 45:54this Dear Mother thinking of your whereabouts look for me home concluded to write you a few lines to inform you I shall go to Winstead today
241. 46:06it's Glaston Berry from 1863 and I shake my account settled letter appears to be a son informing his mother of his travel plans and
242. 46:18whereabouts it also me mentions settling accounts the letter provides a glimpse into personal Communications from from
243. 46:25the Civil War era um what was happening at this time and where might this person
244. 46:54be give the locations mentioned in the letter Glastonberry Winstead Hartford and Middletown likely a person was in
245. 47:02Connecticut a state that was part of the Union in the Civil War in October of 1863 significant several event significant events were
246. 47:13happening and then it talks about the aftermath of Gettysburg the Battle of chicka malga which I've never heard of and the siege of
247. 47:23Chattanooga Connecticut played a significant so so all of a sudden here with so what what prompts did I write I said what is
248. 47:32this and then I said what was happening at the time I don't need to use prompting and then so so my idea was
249. 47:39well so maybe let's go back out here let's go to this thing go whiteboard
250. 47:54images whiteboard
251. 48:16notes let's get a crappy whiteboard thing going here whiteboard displaying notes oh wait that's good this is good okay copy copy
252. 48:31image chat JP I'll do a new chat I'll do a new chat paste what is
253. 48:47this so why I'm doing this is if you're enjoying this conversation join us at the salon. thank you thank you thank you thank you um
254. 48:56so this is figuring this thing out so you could take a course to learn prompting but what if you spent a
255. 49:05week doing nothing but non-traditional prompting right use images as prompts use um take stuff from chat
256. 49:20GPT the chat jpt comes up with go over toso and just say write me a song about whatever the topic was that chat GPT
257. 49:30came up with based on an image um take some image that was made from chat GPT um over to Runway ml turn it into a
258. 49:45movie like I like I wouldn't get too wrapped up impr prompting I made a GPT yesterday fantastic you can upload an image and
259. 49:55find the font yeah oh yeah I know it's so amazing isn't it I know I know we are getting to play with actual
260. 50:05machine learning that's what that's what see the the data scientists have been excited about this [ __ ] for a long
261. 50:12time but they could do all the geeky math stuff to make it do its magic and that's why it remained narrow
262. 50:21right chat GPT you know marks the new era of General general knowledge tools that the rest of us can use um let me say um make a
263. 50:39table organizing the
264. 50:54notes so let's say you did this brainstorming session here and you were assigned with creating notes the table is simplified it looks like
265. 51:10looks like you truncated do it all I something's up with something's up
266. 51:24with chat Che it's it's being weird I saw I saw some notes on on Twitter today that chat GPT is just being weird
267. 51:36and like people's gpts aren't functioning right and [ __ ] like that I'm always afraid to go over my GPT
268. 51:46prompt word limits so I don't prompt it how does how does it work wait I'm always afraid to go over my GPT prompts words
269. 52:00limits um I don't know apparently if you go to teams if you go to GPT team your context window goes up to 32,000
270. 52:14tokens um and normal chat GPT 4 is 8,000 tokens so it looks like it looks like when I told it to do it for real let's do it it for
271. 52:31real all right so that's good so just so now I'm going to say um can you group the notes into
272. 52:48categories in each of the three sections [Laughter]
273. 53:05[Music] lunar stick what's happening you must your ears must have been burning I was talking about all the all the dang gpts you've been
274. 53:17making I also sent you a note um I had time today but uh just send me some availabilities can chat empowered pixels I'm inspired going
275. 53:29to start building some marketing strategy gpts tomorrow no no no no empowered pixels no you can make money with you can make
276. 53:42money all right I want you to go make a GPT right now right now right now thank you for the dog butt I dog
277. 53:52butts always make me smile um I want you to do it right now so here's what we're going to do why is this oh because I
278. 53:58asked it to to categorize these things oh wow this is oh by the way if if you've been watching the screen while
279. 54:06I'm talking yeah Emilio's wife empowered pixels today so we're gonna We we'll go build one here um notice what's notice what's happening
280. 54:16here um so I threw in that image I said what is this and then I said put the notes in a table and it screwed that up and then
281. 54:27I said no do it do it completely don't get lazy and then it did it and then I said can you group the notes into
282. 54:34categories in each of the three sections and so it's it's just it went off and did that so it's kind of prompting itself
283. 54:43here I mean I gave it this this little goal and then it broke the content up into sections and created categories and
284. 54:53made Separate Tables for each each of the three areas each suggestion represents a community provided idea I mean
285. 55:05dang let me see this I'll let me we'll try this um is there any idea Within These
286. 55:20Concepts that is visually interesting the answer will probably be yes heated seats napping pods box
287. 55:36Fort hot Community cocoa check-in system at the front desk so this is Library stuff I
288. 55:50guess great I'd like you to make a 16x9 wide image of of a library concept that features these seven visuals that you've
289. 56:08identified and again what's my prompting here am am I learning how to like have I learned prompt strategy from Rachel
290. 56:16Woods for how to do high-end fancy structured prompting nope I'm talking to it like I have an intern and I'm giving it really crappy
291. 56:31directions and it's responding with better than I deserve so don't get hung up on prompts can you incorporate a YouTube
292. 56:44video in a GPT you could certainly incorporate links in there I don't think you could embed one but you could
293. 56:51certainly you could certainly take a CSV file of uh it made a bunch of different things all right let's do um I'd like a single
294. 57:07room in the library that has these elements in it period as opposed to each element has its own room which is what you created please fix
295. 57:23this [Music] um what was the question the question was oh YouTube videos you could you could do a uh a CSV file with a bunch of
296. 57:38YouTube descriptions and links to them and use that as your source and then God dang it this new dictation thing from OSX it just keeps
297. 57:51typing um so you could certainly build something that jumps you out to specific YouTube videos but you could also just go to
298. 58:01Bard and Bard does that pretty good so so yeah I mean this is the other thing start learning your tools all right
299. 58:10there's our there's our fancy uh there's our fancy library with our napping pods and whatever the hell else was in there
300. 58:15our hot cocoa here's our hot cocoa um but I mean I I so for for me I could go learn prompting strategy
301. 58:29and prompting structure or I could learn what this tool can actually do and just play with it and just learn holy [ __ ] it
302. 58:38can do all this without me having to learn [ __ ] well I had to learn that it could do this right basic question can you explain tokens
303. 58:48absolutely um the good news is we've got paid in here from Google so if I screw this up he will chew me out and he will fix it okay
304. 58:59so to a computer there's no well there's a difference there's no difference between a letter and a space and a comma
305. 59:08and a space so tokens are either a word a space a punctuation mark or some fragment of a word right so so when when
306. 59:19um when you hear this has got an 8,000 token here let me flip let me flip here here when you hear chaty BT has an 8,000 token context
307. 59:35window you shouldn't have to think about that that is for people like Pate to think about so tokens are basically just the
308. 59:44way these machine learning models handle information there are about it's about there are about 70 5% words available for the number of
309. 59:59tokens that you're given so if you're given 8,000 words you or if you're given 8,000 tokens you have about a 6,000 word
310. 1:00:08context window now what's a [ __ ] context [Laughter] window so your phone is actually a really good example if you if you're
311. 1:00:19viewing a text document on your phone and you make the text really small whatever you can see on your screen
312. 1:00:28that's kind of your context window right it it's it's it's a similar metaphor to chat gpt's context window so the the
313. 1:00:36text starts at the bottom and then it Scrolls up and it Scrolls up and it Scrolls up and so so let's say the first
314. 1:00:42thing that you put there is um uh I want you to act like a silly goose right and so the the phrase I want
315. 1:00:51you to act like a silly goose is there and then you start typ typing and it's acting like a silly goose and you're
316. 1:00:56having silly goose interactions and you're having silly goose interactions and then all of a sudden chat GPT stops
317. 1:01:03acting like a silly goose and you're like what the [ __ ] just happened well silly goose went out of the context window and if you didn't
318. 1:01:11remind it somewhere in this conversation that you wanted it to act like a silly goose it essentially just
319. 1:01:18forgets um so so in chat gp4 you have about 6,000 words how do you voice chat with it so how you voice chat with you
320. 1:01:27can do it right now with pi and with chat GPT you do it with chat GPT with their app only and same with same with pi for
321. 1:01:39for whatever reason I got I got Pi's website to do it once and then the next time I went back I couldn't get it to do
322. 1:01:46it I think they took the feature away but anyway uh Mr K day so essentially if you use most of your token it will start to
323. 1:01:55forget the conversation correct so and this is why Mr K um why a lot of people use Claude instead of chat gbt is that Claude has
324. 1:02:09120,000 token context window so what's the what's the famous book um I don't know like like famous novels are you
325. 1:02:20know could be in like the 50,000 word range so if you have 120,000 token context window that means you've got about 80
326. 1:02:3085,000 words okay so with Claude you can literally paste an entire novel or an entire screenplay or an entire research
327. 1:02:42paper into the prompt and what what's cool what TD what what Claude does which is really cool so if I do that let me see do I
328. 1:02:53still have my screenplay up yes I do so here's a screenplay I have so it's 120 pages so if I go to chat JP so here's
329. 1:03:04fancy chat GPT that can do everything and P claims that I all I am is a is a sick aant for Sam Alman it's not true
330. 1:03:12Pate because watch I paste in my screenplay and I say read this and it goes oops the image you did was so long finally understood why it
331. 1:03:25forgets what we're talking about yeah exactly so now watch this I go over to cloud cloud claw. Claude doai I'll continue with my little
332. 1:03:40Google my little Google account that's Google single sign on you know why I use Google single sign on because I want Google to know
333. 1:03:51everything about me so I just pasted it and look what cloud does it it turns it into a pasted document it says 145k
334. 1:04:064,833 extracted lines and then I can say um please summarize this and it's going to go now read my
335. 1:04:21screenplay now you can do this with chat GPT but you you what you do with chat GPT is you upload it as a file
336. 1:04:32right but that's different than having a larger chat context window Claude has a larger chat context window so one of the
337. 1:04:41ways that I use Claude is when I take um these videos I've got a really long prompt that I wrote to turn the YouTube
338. 1:04:52transcript from my two and half three- hour long videos into a YouTube description in chapters and so I go to
339. 1:05:00YouTube I copy the transcription I come to Claude paste it in there and then paste in my prompt that knows
340. 1:05:09how to you know I've figured out ways to make it do that and then it makes me a YouTube description title description
341. 1:05:17and and chapters it's really cool this is really great info thank you you're welcome here's a summary the script
342. 1:05:24centers around Jack a Vietnam War veteran who owns a fishing boat in Washington state in a flashback Jack is
343. 1:05:29s shown serving as a helicopter pilot in Vietnam along with his close friend and partner snake after snake apparently die
344. 1:05:37apparently dies in a helicopter crash and another friend is killed Jack returns home with money from snake to buy a fishing
345. 1:05:51boat so I read my whole [ __ ] screenplay I wrote a little synopsis of it how cool is that so how did I know
346. 1:05:59that you just got to get curious about this [ __ ] you've got to stumble on to it if you go to something like perplexity I don't know what
347. 1:06:09perplexities context window is I assume it's not that big um please Su Ariz this paste
348. 1:06:34huh I guess it read it yeah Reddit all right so perplexity
349. 1:06:48has got a big context window want more AI resources join us at the AI the salon AI yep I'll I'll talk about that in a
350. 1:06:55second here Brandon that's awesome thank you um so perplexity surprised me there and actually ate that screenplay and you
351. 1:07:05know chewed it up and spit it out it did it really fast too um there's a lot of people in this channel that are huge
352. 1:07:12fans of perplexity um I was just given access to the pro version of it for a year because I bought a rabbit R1
353. 1:07:20because I know how to live and make good life choices but one of the really cool things about perplexity it's got this thing called
354. 1:07:29Focus so if I just want to search if I want it to answer me questions but I only want it to use academic research as
355. 1:07:37the The Source material I could say um how can I chapiz video using um AI for the
356. 1:07:52visual elements understanding the question let's see scene changes object
357. 1:08:09recognition look at this so so it says what specific visual elements do so this okay not for nothing people this is now
358. 1:08:18I'm getting excited I I I know I've known for a while I need to do more with perplexity because people seem excited
359. 1:08:24about it um and I didn't know why so perplexity now has this thing in it called co-pilot which is actually
360. 1:08:31unfortunate because that's Microsoft's whole [ __ ] product name is co-pilot so they shouldn't have named a feature
361. 1:08:38within their competitor of co-pilot the name of their competitor anyway these are the people making this
362. 1:08:49technology are research Labs they're not marketers people we've got to give them some Grace but that's really dumb
363. 1:08:56anyway they knew it was coming they knew it was called GitHub co-pilot they knew it was coming anyway so so I asked it
364. 1:09:06how can I Chapter IE video and then co-pilot says what specifical V uh what specific visual elements do you want to
365. 1:09:12use for chapter zation scene changes object recognition emotional detection so I just want to do the visual things
366. 1:09:21right now I might want to do object detection later but I can go I can go back and try that and then if yes um you know throw
367. 1:09:29it in here what's exciting about this to me is that this essentially just wrote a little mini application for us so it it
368. 1:09:40took my question it thought about it and then it said Hey in order for me to give a better question here let me put some
369. 1:09:48interface together so it put together you know multi-choice um you know checkboxes asked me that and now it's now it's going to go respond
370. 1:10:09right it'll be curious to see if it if it brings up any papers that's what I was hoping it would do we'll see we'll
371. 1:10:18see related um so it didn't give us any of that stuff how accurate what benefit so I can say um please show me any papers
372. 1:10:33that uh talk about tech that does this searching scholar found eight sources show five more so here's eight
373. 1:10:52papers robust and realtime handling of something real-time hand detection and tracking of monocular video
374. 1:11:03so so the the thing to understand here folks why does he sing like that that's really annoying isn't it yeah that's
375. 1:11:15what he does it's annoying is he trying to chase people out sometimes he does yeah sometimes people away
376. 1:11:24scared but they come back he's harmless he's got a nice dog [Music]
377. 1:11:47dude um okay so at the AI Salon which I'm going to talk about in a second we have this this page called AI salon tool
378. 1:11:58recommendations and in this list the the thing at the top is chat GPT because chat GPT is still the damn
379. 1:12:05best um when that changes it will no longer be at the top if that changes no one's caught them yet um but then right
380. 1:12:14below chat GPT I've got what I consider the four biggies of large language models Claude Bard perplexity and
381. 1:12:23co-pilot from Microsoft see here's a whole product called co-pilot perplexity you might want to rechange your feature
382. 1:12:30name unless that's because Microsoft invested money in you and then it's on purpose which maybe it is who knows um
383. 1:12:39and then you've got specialty large language models like Pi Gro llama 2 which llama 2 is actually a whole world
384. 1:12:47of open-source large language models so if you are frustrated with the safe safety guard rails of big commercial tools like chat gbt and
385. 1:13:00perplexity then you can go build your own [ __ ] without any safety guard rails on it um and there's there's po so I've
386. 1:13:06got right now and this list I I personally maintain this list this is my curated list in this is in my opinion [ __ ] you should know
387. 1:13:22about and what I can promise you is if it's on this list it won't be a waste of your time with the sole exception of
388. 1:13:29grock which which which is just stupid right now but it's got a sense of humor so if you want to see a chat GPT that
389. 1:13:36can make jokes grock is kind of funny uh but it's just not useful and it cost you 16 bucks a month to be able to watch it
390. 1:13:44make stupid jokes um but it's going to get better right Elon Musk is going to make sure that that thing turns real so
391. 1:13:51so it's it's it it's there it's it's there just in anticipation of it not sucking at some point in the future
392. 1:14:00but at some point I'm going to go play with grock and it's going to be really good right at some point I'm going to
393. 1:14:07like like what I just discovered with perplexity that it's kind of writing this this little mini computer
394. 1:14:14application for me to be able to you know refine my search um that's pretty slick every time I go use one of these
395. 1:14:22tools some new thing shows up like Microsoft co-pilot um yesterday I discovered that it's got co-pilot
396. 1:14:36gpts so it's got a vacation planner a cooking assistant a fitness trainer so Microsoft invested $1 billion in open
397. 1:14:46AI open ai's got these things called gpts Microsoft has this thing called co-pilot which is like chat GPT and now it has these things called
398. 1:14:55co-pilot gpts and we knew this was coming but like this wasn't part of the site two weeks ago and now it is and
399. 1:15:08so if you want to be AI literate for the next 18 months you just need to be curious and you need to be
400. 1:15:24adventurous and you need to be in community because there's so much happening so fast and it's changing just
401. 1:15:35dramatically that if you're not in a community of people that is figuring this [ __ ] out with you it it gets I what I can promise
402. 1:15:46you is it gets very overwhelming very stressful very um like you just feel like you can't ever possibly learn enough and you're like oh
403. 1:15:58my God everybody knows more than I do help what to do if you can find a community that you can connect with so you can learn these
404. 1:16:11tools you're going to be okay that's what this group is the salon. so I want you to go to that website now this
405. 1:16:19doesn't cost any money so I'm not trying to sell you anything like you can make money with that's not this
406. 1:16:27if I turn this channel channel into a sleazy Commerce Channel I'll let you know right I won't I won't make fun of
407. 1:16:33it with a gold button I'll actually just go you can make money with chat jbt just send me 1995 a month we we might go there I might turn
408. 1:16:41this into an NPC NPC Channel where I just say hot dog hot dog rabbit rabbit and then make lots of money I might
409. 1:16:53twerk right if I need money I I might twerk we could turn this into a twerking channel do you see all those Minnesota
410. 1:17:00dance dancers doing their Twirls I could turn it into a dance competition Channel where all I do is Just Dance for
411. 1:17:08hours but it's not there right now and neither is the salon so the salon this is a group that I started the week after
412. 1:17:15chat gbt came out because I've been in this movie before I was an entrepreneur in the early days of the worldwide web in the mid
413. 1:17:2790s in New York City and started a similar group called the worldwide web artist Consortium and um what I learned back then is that when
414. 1:17:40you're this early when you're at the beginning of the thing that is going to change the world and you recognize that before
415. 1:17:52before other people do one it's really exciting and two you will feel like a clueless idiot for a lot of years
416. 1:18:04why because you can't get a degree in this right now well but you can go to Stanford and get a degree in machine learning yep but that's not
417. 1:18:13this chat GPT November 30th 2022 so just over a year ago marks the beginning of a new era just just like the worldwide web
418. 1:18:25marked the beginning of a new era for the internet the internet had been around for decades and Tim burner Lee comes up with
419. 1:18:34this hypertext transfer protocol thingy that allows you to underline a word and do some magical prompt behind
420. 1:18:45the scenes that when you click on that word it takes you to some other server Halfway Around the World and so the worldwide web went from a
421. 1:18:57command line interface that only researchers and scientists and Engineers used because they love command line
422. 1:19:07interfaces to click click click click click and the rest of us got to use it and so being there at the beginning of
423. 1:19:16that was absolutely remarkable but nobody knew how to build websites nobody knew what websites were like like
424. 1:19:26browsers didn't exist like the browsers that that had images in them did not exist I I I created one of the first
425. 1:19:35online magazines a magazine called Urban desires and and the first couple of issues all of the images are left
426. 1:19:42Justified on the page because the center tag didn't apply to images it was that early right that's where we are with AI right now
427. 1:19:54the difference with AI is the tools are significantly more sophisticated and there's an entire infrastructure in
428. 1:19:59place to be able to deliver them right so that's what allowed when when chat GPT was launched it went from zero users to 100
429. 1:20:10million users in six weeks Word of Mouth only open AI didn't think Chachi was a big deal they thought maybe it would get
430. 1:20:25a blog post or two in 5,000 comments that's what they said the night before they launched it six weeks later they had a 100 million
431. 1:20:34users Word of Mouth only holy [ __ ] right so what I know is nobody knows what they're doing right now so the only
432. 1:20:48way to to navigate this space is do it with others right so like why is there a crew here heading off into the promised
433. 1:20:54land through all the Adventure and toil of of the adventure cuz it's it's the it's that that's it this this I know like this I
434. 1:21:04know there's a lot of stuff on this channel I don't know right there's a lot of stuff on this channel I'm just
435. 1:21:09bullshitting my way through because nobody knows anything including me but what I know is that we're so early in
436. 1:21:16this thing and it is so profound and so different and we and we so don't know what what these tools are how they work
437. 1:21:24what the implications of them are what work is going to look like what our jobs are going to look like if we're going to
438. 1:21:30have jobs how we're going to have jobs how money gets made all of that shit's going to change but that's going to take 10 or 15 or 20
439. 1:21:40years but the next three years are the primordial soup of that future and so all you have to do is be in the conversation show up to
440. 1:21:53my [ __ ] lives Cindy [ __ ] is at the airport right now getting ready to fly to Asia she's got some hours to kill and
441. 1:21:59she's here I think I think I think that says a little bit more about her than than us right no but she's be she's inspired by
442. 1:22:09this oh what what's Cindy Coon's job by the way she's a futurist she's been a futurist for decades she's a futurist
443. 1:22:24she is lit on fire about what's happening with generative AI she is lit on fire about being part of this
444. 1:22:33community as are a lot of the Irregulars the people that show up here regularly we call them the Irregulars they come
445. 1:22:40here night after night after night after night and you're like but Kyle you're supermodel beautiful of course they come
446. 1:22:47night after night I I get it I get it thank you for the compliment it's it's super sweet of you I'm like an old fat guy on Tik Tok
447. 1:22:58why are people coming here night after night after night because on a regular basis we'll stumble onto something where it's like
448. 1:23:07oh holy [ __ ] that changes things and then we'll say o the GPT stores coming out it's going to change
449. 1:23:18everything everyone's going to be great maybe you won't need gp4 to use G gpts we're going to make money cuz he said
450. 1:23:24there'd be Revenue sharing and then they launched the store and we all kind of collectively discover huh [ __ ] well that was a
451. 1:23:32dud most people haven't even heard of chat GPT much less have a point of view on the chat GPT store it's impossible to learn at all
452. 1:23:47just pick things that are of interest that interest you the most that's exactly it Source Camp is just find the
453. 1:23:53[ __ ] that resonates with you and start playing and then why you do it with other people is that someone with a completely
454. 1:24:00different skill set you know just like when you go on an adventure like if we were going to write this as a screenplay
455. 1:24:05right you're going to have the the person that does the magic you're going to have the muscle you're going to have
456. 1:24:12um you know the the Mage the you're going to have the wizard you're going to have the really smart lock
457. 1:24:20picker everyone's got different skills right so no one's got all the skills right now even to learn this
458. 1:24:30stuff because you keep coming back night after night Kyle and we have fomo exactly well you know I mean there's a
459. 1:24:38selfish reason that I keep coming back night after night here and that's that it keeps me in the conversation I just I just know about
460. 1:24:50myself I am easily distracted distracted it's it's one of the one of the joys and tyrannies of my brain right it's neuros spiciness at its
461. 1:25:04best and I know that this is too important not to be in the conversation it's too important I was looking at
462. 1:25:12autogen seems really great yeah the the autonomous agent stuff is is really exciting um I'm sitting on the sidelines
463. 1:25:19with that stuff right now just because I don't I don't have time to dig around in the open source the more technical stuff
464. 1:25:26but yeah it's if you're if you're technical at all playing around with the open source projects like autogen and
465. 1:25:32Lang chain and things like that it's actually really important all right so you go there you go to the salon.
466. 1:25:40aai and it's going to take you to a link tree
467. 1:25:54and the first link is to the AI Salon uh Community the second link is actually really interesting too if you click on
468. 1:26:02that it takes you to a thing called the GPT for good directory this was something that I teamed up with Ann Murphy from empowered
469. 1:26:11fundraiser and then Vicky Baptist built this website for us in real time while we were doing this event so in 7 days we
470. 1:26:18put together this event where we were going to teach as many people as we could how to make gpts for nonprofits
471. 1:26:24this was in in anticipation of the store coming out was was the idea we wanted to teach as many people
472. 1:26:31as possible to build gpts for nonprofits and in 24 hours we we did eight sessions and did four I did four this community built and shared
473. 1:26:43more than 150 gpts for nonprofits and so you can just go see like it's they're super inspiring and the site's beautiful and then
474. 1:26:52what what Vicki did was she took the instructions that we gave people for the event and she built a page here so if
475. 1:26:59you go to this page and you want to make your own GPT for a nonprofit or frankly for anything it it's eight steps they're
476. 1:27:07really easy like building a GPT is copying and pasting that pair of sentences and replacing the bracketed
477. 1:27:15things with your org name and website of the nonprofit that that little thing will build your GP It's [ __ ] amazing anyway that's that
478. 1:27:27but if you go to the AI Salon community and go to the welcome to the salon section you'll see that that groovy picture that was made in uh
479. 1:27:41Dolly not that it matters but it was we're a community of extraordinary adventurers exploring the AI unknown and
480. 1:27:53you know extraordinary adventurers why why adventurers like we're choosing to be on this adventure and you know it's
481. 1:28:02funny like like you know it's it's good and nerdy to have a good you know kind of D action you know fantasy Adventure kind of theme
482. 1:28:15um but if you think about the fear going on right now with AI the fear and the anger and people are afraid of it it's
483. 1:28:22going to kill what if the robots kill us and it's stealing and it's ruining artist's lives and all that sort of stuff going
484. 1:28:31on it's kind of like in one of those fantasy novels or those fantasy movies where we just keep hearing that people
485. 1:28:40go out into the woods and they don't come back and then we keep hearing large Roars in the distance and and there's there's all
486. 1:28:48this fear of what's out there and it becomes clear that something needs to be done but nobody wants to go [ __ ] out [ __ ] out in
487. 1:28:58those woods because they know they're not coming back right and like whatever that thing is that's making all that noise it sounds scary as
488. 1:29:07[ __ ] and then what happens someone in the community goes we got to go I'm going to go I'll be the muscle and then you know someone else
489. 1:29:20goes you know well I can do m spells and someone else goes and right and so the the five five or seven merry adventurers
490. 1:29:27put their wacky band together three of them don't want to be there one of them's an [ __ ] two of them are super
491. 1:29:33nice and are going to fall in love like we know the whole story but they choose to go on that Adventure they choose to do
492. 1:29:46it and it ends up saving the town and it ends up transforming them right it's Joseph Campbell it's the hero's journey
493. 1:29:54so this is a community of people that are choosing to be on this adventure this is a community of people that in
494. 1:30:02the face of their friends and family going you know what Sally I don't want to hear about AI anymore AI is stupid
495. 1:30:09why are you talking about AI all the time you used to be so cool and now it's just like Ai Ai and and did you ever go
496. 1:30:15to the AI learning lab on Tik Tok I don't care about Tik Tok I don't care about your stupid AI learning lab
497. 1:30:23and they keep going yeah I know you're scared but you want to see something cool you want to see something
498. 1:30:28cool I don't even know what a GPT is or does I'll show you in a second here June Mary so anyway so come here and just
499. 1:30:36know that that's who this community is and so we talk about that and then we also talk about our values and the
500. 1:30:42values are really important to me please read them if you come here if this is sounding exciting to you I'm talking it
501. 1:30:48up it doesn't cost anything by the way you're like oh sure here it comes it's going to be $49.95 a month for the next
502. 1:30:56five months no doesn't cost you anything it's free but it does have a cost I want you to read the values my request is read
503. 1:31:06the values and like take them in curiosity we want people that are curious generosity this is a big one I
504. 1:31:14don't want people in here that learn some [ __ ] and hoard it I want people in here that are like oh I figured out something o oh how'd
505. 1:31:22you do that well I took this and I duct taped it together with that and then it made this and then this is what came out
506. 1:31:28and people like holy [ __ ] that's amazing that's generosity when you're when you're this early in a new
507. 1:31:39era you can afford to be generous in fact it's a requirement because nobody knows what the [ __ ] they're doing and
508. 1:31:46the thing that you learn that you think is the proprietary gold that's going to make you rich is going to be completely
509. 1:31:53useless 2 months from now cuz the technology will have moved on so you might as well share it with the community and everyone gets to
510. 1:32:01learn generosity exploration how do I how do I know about perplexity how do I know about Claude how I I because I'm like I got to go
511. 1:32:12figure this out I got to let's go into the woods let's start mapping it collaboration do it with other people
512. 1:32:19connect with other people be generous connect with other people thank you Empower and I know I'll show you how to
513. 1:32:24make a GPT cuz we've got we've got to explain what they are and how to make them but we're going to make one
514. 1:32:30tonight inclusion people come from all over there are artists in here there are technologist in here they are PhD
515. 1:32:41mathematicians this guy Kyle's the real deal thank you rock appreciate that there's blind people in here
516. 1:32:52Rudy's in here from South Africa there's people from all over the world every diff every imaginable profession every imaginable level of
517. 1:33:06understanding of this stuff including I was just scrolling through Tik Tok and I thought it was fat Alec Baldwin and now
518. 1:33:14I'm just what's he talking about like that level of experience is absolutely fine here and that person that is that clueless
519. 1:33:25about AI right now will probably be the one that gives you some remarkable breakthrough 6 months from now because they're bringing some life
520. 1:33:36experience or some way of looking at the world that you never would have considered and they'll use tools differently than you do so that's
521. 1:33:45why you include people and the final one is empathy this [ __ ] scary it is scary and as excited as I am
522. 1:33:56about this and as Evangelistic as I am and as much as this seems like Jesus this is this is he's up on a soap
523. 1:34:05box this shit's also really scary right this is existential if you're an artist if you've been an illustrator for 30
524. 1:34:13years and you go look at what mid Journey can do it's a little how do they say [ __ ] terrifying Ling if you've been a
525. 1:34:27copywriter for 20 years in the ad business and all of a sudden there's a thing that can write 150 ads in 38 seconds and they're
526. 1:34:37good because it also wrote The Creative brief that's [ __ ] terrifying so this is a community that
527. 1:34:53understands that sometimes people are going to come in here and they're going to be pissed off they're going to be
528. 1:34:56scared and they're going to be like this is evil your people are evil yep we understand I understand how
529. 1:35:04it's that feels like that want to see something cool so anyway go join and then and then my request is if you like all that in
530. 1:35:13the community section of the site the very first thing we have is introduce yourself so my request is use your real
531. 1:35:19name ideally put your picture there if you're comfortable because we want to get to know one another and then introduce
532. 1:35:26yourself tell us who you are tell us what you do tell us where you are on this journey my daughter-in-law in the
533. 1:35:32Philippines says you can make money with ch gbt you can make money with that's awesome um and then also what you should
534. 1:35:43do so after you introduce yourself go to the show int tell channel so the show int tell channel is just a [ __ ] blast
535. 1:35:48so every night when I'm doing this because you know I'm all over the place right so as people that come here
536. 1:35:54regularly are like oh here goes Kyle off on one of his tangents they'll start collaborating and jamming and making images and sharing
537. 1:36:02them here and and they take you know little stupid things I say and they turn it into there you go little memes right
538. 1:36:09and it's it's happening in real time and like some of the shit's beautiful some of the shit's hilarious a lot of the
539. 1:36:15stuff is trolling me in the best possible way these are amazing oh it's Valentine's coming enough right
540. 1:36:23beautiful a that's beautiful nice nice Papa JJ oh that's cool wow that's pretty gorgeous the hand Shadows with the the
541. 1:36:35people as the object wow that's super cool Cory Sandler everything is moving so fast them haves wait for the oh for them haves but
542. 1:36:48the have knots need this technology too well that's that's the other thing about this technology is it is it it it
543. 1:36:55provides superpowers to everyone that really uses it oh thank you so much Jen okay let's go talk about gpts so
544. 1:37:04anyway check out showand tell um also oh the other thing that you should check out here is the AI tools and resources
545. 1:37:11so that's where this list is that I personally curate and so in this in this list is if you're brand brand new to this stuff just start with
546. 1:37:22GPT if you can afford the 20 bucks a month pay the 20 bucks a month and what I have listed here is what are the
547. 1:37:29features that you get with GPT 4 that you don't get with gpt3 here's what you get with gpt3 here's all the additional
548. 1:37:36stuff you get with chat gp4 every one of these bullets is a whole separate Rabbit Hole so if you're just getting started
549. 1:37:46just start here knowledge is power something something I didn't see the rest of that um and then I've got eight other large
550. 1:37:55language models we we'll add more if more come and they probably will um I've got a bunch of image generation tools
551. 1:38:03Dolly 3 mid Journey Kaa Deco here which is really interesting and fascinating um Leonardo which started as a just kind of
552. 1:38:12like a stable diffusion web front end for for image generation the as stable diffusion is launching more and more
553. 1:38:21open- Source models like things that make video and do real time they're adding it all to Leonardo so this is
554. 1:38:27turning into an incredibly sophisticated tool then you got Firefly from Adobe then we've got video creation bunch of tools there then music
555. 1:38:36and audio then 3D worlds and objects and and to be to be fair 3D worlds and objects right now this is very thin
556. 1:38:45there's probably a hell of a lot more than this but like this is a whole separate Universe right like when you
557. 1:38:50get into into 3D stuff it it gets much more technical I'm kind of waiting for this world to simplify and get better um
558. 1:38:59kind of like videos done now like like you know you can go to um Runway ml right now or P labs and generate some
559. 1:39:06pretty sexy [ __ ] video without knowing much technically a lot of people in Show and Tell are sharing there
560. 1:39:14um and then I've got people to follow Ethan mullik David Shapiro Paul rer and Rachel Woods are the the the biggies
561. 1:39:21that I think are important voices right now all right so that's that all right now you're talking about these gpts what are
562. 1:39:32gpts um if you don't know what gpts are that is perfectly normal if you don't know what chat GPT is that's
563. 1:39:41perfectly normal in fact McKenzie did a report I don't know why this isn't reloading McKenzie did a report um two three months ago
564. 1:39:52where they tracked the adoption of chat GPT I assume they just did this in the US but it might be Global I don't I
565. 1:39:58don't I have ADD I don't pay attention to details like this but here here are the big broad Strokes of the study at the beginning of 2023
566. 1:40:1088% of people had not used chat GPT in I think it was November of 2023 82% of people had not used chat
567. 1:40:27GPT so we made six we made a Sixpoint increase in 2023 as much of a big deal as it seemed 18% of human beings in in
568. 1:40:39whatever the story was I think the US have used chat GPT more than 80% have not used it now of that 18% that have used it
569. 1:40:52that probably includes people that have used it once right which is probably a lot of them oh yeah it's just like
570. 1:40:58Google I get it and then they you know they use it once and they've never come back so so some let let's say a third of that
571. 1:41:0718% actually uses chat GPT with any level of regularity that's less than 10% right six right 3 * 6 is 18 Pate I need your PhD math thing
572. 1:41:23let's say 6% of people use chat GPT with any regularity so if you're like I don't know what these things are I don't know
573. 1:41:30what the [ __ ] you're talking about you're in the majority we're early and here's the good news about that if you step on if you step into
574. 1:41:44this community if you join the adventure there's a lot of opportunity here for anyone that joins right now the next 18
575. 1:41:52months you just need to be AI literate okay so chat GPT itself what is chat GPT perfect question thank you for asking it Mr Hammer
576. 1:42:03okay chat GPT is what's called a large language model and and I'm going to unpack okay I'm going to calling an
577. 1:42:16audible here we we're going to props so what people in this channel know is that even though this looks like just a normal Tik
578. 1:42:32Tok Channel and I look like just like an old dude with t-shirts this is a $20 million video production
579. 1:42:42facility so the production values here are much higher than you think so I'm going to show you I'm using some props to explain some
580. 1:42:56[ __ ] this might just look like a box of a laptop stand that was sent to me along with a bottle of Blandon but no
581. 1:43:04no this is a prop from our studio from our fancy Studio okay so this box represents
582. 1:43:19gp4 you're like well that's chat GPT no it's not gp4 is this box what's that box well if you open the box look what's in it
583. 1:43:30nothing so so what did they put in it so chat GPT the GPT in in GPT stands for generative pre-trained Transformer the T
584. 1:43:42the Transformer part is just the technology in 2017 Google released this paper called attention is all you need and they
585. 1:43:51invented this thing called the Transformer they didn't nobody thought it was a big deal at the time it turned
586. 1:43:55out it was the seminal paper of all of the [ __ ] that we're playing with today today what the Transformer did was let
587. 1:44:04these machine learning models be trained on massive amounts of data with massive amounts of compute power in the form of
588. 1:44:12gpus graphic processing units and what it turns out is the more data you throw at these things and the
589. 1:44:20more P power you throw at these things the better they get and so Sam Alman and the open AI team recognized this very
590. 1:44:26early on and and what they did starting with gpt1 then gpt2 then gpt3 which was 2020 and then GPT 4 was they they went
591. 1:44:37and they got as much of the public internet as they could and they they embedded it they they they did this special treatment of
592. 1:44:50all that that data that turned it into points it doesn't matter these things called vectors so they took they
593. 1:44:59took as much of the public internet as they they could get they put it in this special form and they dumped it into
594. 1:45:04this box so so so when you're playing with chat GPT you're actually sort of tapping into 40 50 years of humanity whatever Humanity
595. 1:45:19dumped onto the public internet as much of that as opena I could get is what's in this box which is one of the reasons it's so
596. 1:45:28remarkable and and like the fact that this box can hold all that [ __ ] is remarkable okay chat GPT is an application that's built on
597. 1:45:40top of this thing right so this is just a um a thing that makes it simple um like the little search bar that looks
598. 1:45:48like the Google search bar that's part of chat GPD and when you type something in it takes what you typed in it goes
599. 1:45:55down into this thing and pulls it up um I'm going I'm going to give you something just cuz it's funny so Microsoft has this thing called
600. 1:46:04co-pilot which is essentially chat GPT 4 for free so there's your 20 Buck a month chat GPT and there's co-pilot which is
601. 1:46:15free and you're like but Kyle why would I not use Microsoft's CH for f because if I want to you can make money with I don't want
602. 1:46:29to spend all my money here when I could have it for free hair okay why these are both kind of similar is they're both calling the same
603. 1:46:42underlying data they're they're both using the same model but they're actually different applications what
604. 1:46:47happened what happened with co-pilot it was called Bing chat at the time when it first came out it was really
605. 1:46:55good and it was kind of decently unrestricted you could you could have really long talks with it you could ask it
606. 1:47:02anything and so Kevin Roose who is the um the tech reporter for the New York Times he said I'm going to go talk to
607. 1:47:13Bing chat and he had a 4-Hour conversation with it turns out its name was Sydney and she kind of took a lik into
608. 1:47:24old Kevin and she told him that she loved him and that he should really leave his wife for her Bing chat kind of lost her mind
609. 1:47:34Sydney lost her mind so Kevin Roose being a New York Times Reporter screenshotted that turned in the the the
610. 1:47:42piece to his editors and they put it on the cover of the New York Times so here's what Microsoft did to to Bing chat
611. 1:47:51they lobotomized it this is Sydney's brain they put that off to the side and she is but a hollow shell of herself so
612. 1:47:58this is chat GPT this is co-pilot and that's why this is worth 20 bucks a month okay that's what it is or I mean that's sort of how it
613. 1:48:11works so this application this blue box on top of that big box it's sitting on top of this um this giant thing and the the other two
614. 1:48:23letters so I talked about Transformers the technology the GP in GPT stands for generative meaning it's generating
615. 1:48:32content and the P stands for pre-trained that's that it's been pre-trained on all of the internet so you don't need to
616. 1:48:38bring any data it kind of knows all the [ __ ] right and it it it knows a lot and it's getting better and better and
617. 1:48:44better and GPT 5 is even going to be better and llama 3 from meta is being trained up on just massive amounts of
618. 1:48:50compute power it's going to be awesome so these things are going to get better and better but they're they're very very
619. 1:48:55powerful right now so you can kind of ask them anything so I'm going to flip over to gpt3 and I'm going to do what I
620. 1:49:01call parlor tricks and just parlor tricks are just if you don't know what these things can do these things can
621. 1:49:07kind of melt your face a bit and what we what we call them in here is you'll have your a Kevin McAllister
622. 1:49:14moment right where when you see what these things can do you you you you kind of go oh shit's different now and and
623. 1:49:23and that by the way once you have that and get into this world those things just keep happening okay so if I go in
624. 1:49:29here and I say um let's see explain Quantum Mechanics for me in an Eminem RP oh it it did M&M's like the
625. 1:49:44candy so no I meant the rapper um please
626. 1:49:59write a WAP in the style of Eminem that explains quantum mechanics there we go all right so it'll just do that now I can say
627. 1:50:14um explain the difference between quantum mechanics and gummy bears but format it in a table sure here's a table comparing
628. 1:50:31quantum mechanics and gummy bears so the G and GPT stands for generative it's generating this stuff right it's it's not copying and pasting
629. 1:50:41this this didn't exist anywhere and it wrote it that quick the nature of study a branch of physics dealing with the behavior of
630. 1:50:49particles at a Quantum level gummy bears a tasty and chewy candy treat now now wrap it with your hat sideways exactly um then I could do
631. 1:51:04um make me a python application that asks someone five questions and determines if they're thinking more about quantum mechanics or
632. 1:51:19more about gummy [Laughter] bears and it will write you this python code and this code works like you can
633. 1:51:30take this over to repet and make [ __ ] if you're like well I'm a coder is this I'm an engineer my son's going to school for
634. 1:51:39computer science should he not be learn yeah this is why this stuff's scary and exciting the engineers that that are embracing
635. 1:51:51AI um are dramatically increasing their um their productivity and these tools are getting incredibly powerful
636. 1:52:04okay so then you go to gp4 and gp4 does even more right so so gp4 you can do things like uh make me a wide image of a
637. 1:52:2070s muscle car in an abandoned Factory and it will just start making images so so gp4 it's got image generation you can talk to it it can
638. 1:52:36talk to you it can see images and and understand them it can do Advanced Data analysis it's connected to the internet
639. 1:52:46um you can give it a personality so so just just this stuff's crazy crazy crazy okay and and when I say when when I talk about um
640. 1:53:04if you're brand new to this stuff where you should start is with chat GPT it's because all of these tools are so
641. 1:53:14sophisticated that you don't really need to leave it to do remarkable stuff now if you want to do video or if you want
642. 1:53:21to do songs there's specialty stuff you can go do all that but it's really easy to get distracted when you're just getting
643. 1:53:30started cuz there's so [ __ ] much so if you just stick with this one tool and just understand what this one tool can
644. 1:53:36do it is a window into what's possible and it's it's it's kind of a glimpse into what's coming okay um it's janky it
645. 1:53:48lies to you it gives it it it makes mistakes it confidently answers you even if it's wrong like I like to call it mansplaining as a
646. 1:53:58service um like all the good men in your life confidently answer you even if they're wrong um okay so now so what's a GPT so
647. 1:54:09a GPT so so what open AI did two months ago November 6th 2 and months ago is they launched these things called
648. 1:54:23gpts and what gpts are is they are customized personalized versions of chat GPT that you can share that you can
649. 1:54:34author you can create them and you can share with the world okay so 70s muscle car in an abandoned Factory is was just it's just like one
650. 1:54:45of those parlor trick prompts that I like '70s muscle cars and one of the ways when I when I first go to an image
651. 1:54:52generation tool I'll just type in that prompt and just see what it does like mid journey is amazing and it mid
652. 1:54:59Journeys always look good other tools like when I first used um Adobe Firefly I said 70s muscle car in an abandoned
653. 1:55:08Factory and it made muscle cars that were no later than 1962 and what that told me about Adobe Firefly because they trained on public
654. 1:55:21domain images and non- copywritten images is that it didn't have any 7s muscle cars in its training data like so
655. 1:55:29so you know just do using the same prompt across Services can can sometimes teach you stuff but I like muscle cars so I created a
656. 1:55:43GPT called if I type in 70s called the '70s muscle car machine so so design a dream 1970s Resto Resto
657. 1:55:56mod in three simple steps and and what I remember being a kid is that muscle cars like you could get these posters that
658. 1:56:05you ordered from your school that you could hang in your room of these cool muscle cars and I thought wouldn't it be
659. 1:56:10cool for people to be able to just you know go do that again make make a cool poster for their room do do the little
660. 1:56:17Nostalgia thing now is everyone going to be into muscle cars no most people like just wouldn't care to do this like they wouldn't even think
661. 1:56:27to run chat GPT down this rabbit hole but I'm into that and so what it allows is anyone to take anything they're
662. 1:56:35interested in and turn it into something that's super cool so just watch what this does so I've got this pre-written
663. 1:56:41question here I'm ready to design my dream car so if you click on that by the way if you have access to gp4 um chat
664. 1:56:48GPT go to explore gpts and type in70s muscle car machine you can make one along with me um so I say I'm ready to
665. 1:56:57design my dream car first thing it's going to do is ask me uh for the model so let's go with a uh let's go let's go uh
666. 1:57:1267 Mustang fast back you like but it's it's a 70s muscle car machine 67 Mustang Fastback what an iconic Choice let's move to the second
667. 1:57:25question what color and finish let's go silver with black stripes and so the GPT is an application
668. 1:57:41right it's walking me through this process that's that's fantastic what do you want let's see let's do um um a sexy hood scoop and racing
669. 1:58:03tires now it's creating an image didn't you make a song that was an homage to Sydney I did I'll I'll play it
670. 1:58:13for you it's actually pretty good I really like it boom Stang like if you were if you were a kid in the 70s I mean this feels familiar
671. 1:58:29right like and I and so I coded this thing in such a way or coded is a is a generous word I wrote the prompts in such a
672. 1:58:40way that it asks you these questions it asks you three questions and then upon those three questions it comes up with a
673. 1:58:48nickname for the car and puts that in big bold letters above it and makes you a poster now you could go learn to prompt
674. 1:58:56all that yourself you could do this all yourself but like you can just go find these gpts like like uh um Corey Cory
675. 1:59:05Sandlers this is a crazy one how does what's it called the pottery is it going to come up here pottery wheel Pottery Prodigy is that it
676. 1:59:23no crap I got to remember the name of this one Pottery glaze I didn't realize how powerful it was regarding coding oh I yeah so so so
677. 1:59:38capture card company Pottery Whisperer Pro okay oh she's still in here hi Corey poing whisper this this one this one just
678. 1:59:48blows me away ER Pro because like similar to similar to my muscle car one like Cory's an amazing artist an amazing
679. 1:59:58ceramicist and and so what this is is this will help you um design glaze recipes and so this is my favorite one create a random glaze
680. 2:00:10for for cone 6 Pottery and I didn't know what cone 6 was do you know what cone 6 is no well well dum dum cone 6 is a temperature
681. 2:00:22range I had no [ __ ] idea what cone 6 was but this thing told it to me but now look what it's doing so so it's saying
682. 2:00:30the glaze type mystical Emerald sea Rich emerald green the surface type is glossy with subtle crystalline effects cone
683. 2:00:38cone temperature cone 6 firing atmosphere oxidation I don't know what any of this [ __ ] means specific gravity
684. 2:00:461.45 you got to know your specific gravity gravity as well as your alumina to silica ratio everybody knows that
685. 2:00:53you're not a Potter if you're not thinking about your alumina to silica ratio and then it gives you the recipe
686. 2:01:00but then then watch what you can do let let me see what this would look like and it [ __ ]
687. 2:01:16does and this is the [ __ ] where where I I um there it's creating the image this this is the [ __ ] where this is where
688. 2:01:28pate and I and and I differ where where he says oh error creating the image let's see let me do try again for the image
689. 2:01:39um it's all just mathematics that's making this possible but this is [ __ ] magic I'm sorry this is Magic and it's
690. 2:01:49and it's like you know she's got a lifetime of experience and she understands the joy of if oh it didn't
691. 2:02:00it didn't do the glaze right um let me see um show me an example of the glaze you just um
692. 2:02:16created we'll try it again but this takes you know like a lifetime of her experience and it makes that knowledge accessible to
693. 2:02:27people that are just getting started in this and that's what these gpts are all right it broke again and also
694. 2:02:38also the other thing about these gpts uh try again the the other thing about all all all 100% of all of the AI tools that we're playing
695. 2:02:52with today are janky pieces of [ __ ] just know that I designed one called CT chulu Vision which transforms your
696. 2:03:02surroundings into Lovecraft photos super cool that's great anyway this is broken it's not making images but keep oh wait
697. 2:03:12did it make one oh no it just made that image uh but try this it my my gp4 is janking out right now um but but go play with go play with
698. 2:03:22this it's like it's [ __ ] amazing and and you can just go explore the store and find a bunch of those things okay
699. 2:03:28now back to I don't know if if anyone's still in here before that was like how do you cthulu yeah exactly uh how do you
700. 2:03:37make one of these things it is trivially simple to make one now it's not triv trivially simple to make them good but
701. 2:03:48if you join the AI Salon there's a bunch of people in there that participated in the GPT for good challenge that have
702. 2:03:56made dozens and dozens and dozens of these now so anyway you click on create on the explore page wow it could refine redefine
703. 2:04:07Transportation altogether home surgeon this is why this channel exists because when people have their Kevin
704. 2:04:16McAllister moment the I didn't know it could do that it could do that these these systems I don't even know what to [ __ ] call them these
705. 2:04:35platforms like like chat gbt 4 and and the other tools the other large language models are General general knowledge
706. 2:04:46systems and they're incredibly capable and as you know Sam Alman describes it chat jp4 is really just practice that's
707. 2:04:55that's how he defined it at Davos last week this is really just allowing us to practice with what's coming so if you can start to get your
708. 2:05:08head around what chat G pt4 makes possible yeah it's like every single thing is going to be revolutionized
709. 2:05:16every single thing from science to Art to math to just all of it all of it I asked for a squirrel riding a shark
710. 2:05:29the other day and it was beautiful awesome cookie yeah it's good at that okay so how I got here so this is the
711. 2:05:36chat this is the GPT Builder within chat GPT so on the left hand side this is where you make it this is where you do
712. 2:05:44your this is where you will it into existence and then over here here's where it shows up and it's instant and so on the left hand side
713. 2:05:53there's two modes there's the create mode and there's the configure mode the configure mode you can upload an image
714. 2:05:58name it describe it write a prompt um write some questions upload files but you got to think like this is
715. 2:06:07this is like coding if you're not a coder this is intimidating so what they did was they put chat GPT as the
716. 2:06:15interface to build these things so the hardest thing to do is to come up with something to build so so here's what I
717. 2:06:22would strongly suggest take something you're passionate about the GPT Builder the most exciting change in software
718. 2:06:29since the mouse I tired Runner I totally agree I actually wrote a piece if you go to go to LinkedIn I wrote a piece on this because
719. 2:06:43I I think this is the biggest deal the this is uh check your LinkedIn app oh God hang on a sec H resend I got to go I got to go 17 got
720. 2:07:00to go 17 Factor authentic all right I'm back um go to LinkedIn my my handle on LinkedIn is Kyle Shannon not not super uh
721. 2:07:14creative but where oh I go to my profile I'm just trying to get the yeah the dawn of the everyday developer so if you look
722. 2:07:23for look up dawn of the everyday developer so this is a piece I wrote and I and lunatick Daisy Thomas has also written some pieces on this
723. 2:07:33um I talk about the fact that the fact that you've got um the fact that you can with zero coding knowledge create an incredibly
724. 2:07:46sophisticated application in minutes is a is like like the impact of it is is hard to it's hard to State um and even though gpts can only
725. 2:08:01be used by chat GPT 4 members and they're janky as [ __ ] and they don't always work and nobody knows what to do
726. 2:08:07with them none of that is relevant what's relevant is you get to take the power of this [ __ ] insanely powerful
727. 2:08:16engine and you get to it to your will and you get to to self-express as a human inside this thing it's it's it's
728. 2:08:26remarkable so the toughest thing to do is to come up with what are you going to do with it I see an Murphy's in here
729. 2:08:32hello an an an Murphy um she and I teamed up to do this GPT for good Challen challenge she is remarkable so
730. 2:08:40when you go in the AI Salon one of the clubs toward the bottom is called the weekend GPT Club challenge I don't know whatever
731. 2:08:48ever it is um but it's her Club every every Saturday she's got a 2hour GPT building Community um that that's a
732. 2:08:56really cool intersection of lots of other communities but you can join that in the AI Salon okay so we're going to do um ah I got
733. 2:09:17one I'd like to create a GPT that makes it very easy for someone who's not sophisticated with wine to
734. 2:09:29pick the right wine for the meal they're about to have what's what's going on
735. 2:09:43here why did that image gets so big and weird anyway whatever um this is the entire prompt so watch I'm going to hit
736. 2:09:58submit updating GPT it is building this for me right now and you'll say oh boom there it is it's over here done uh great
737. 2:10:08choice a GPT that helps selecting wines for meals sounds both useful and enjoyable let's give it a name that reflects purpose how about Veno
738. 2:10:16Matchmaker does that name work for you sure that's actually not bad so sure now it's generating a profile
739. 2:10:26picture so it named it for us Kyle is breaking chat GPT I'm totally breaking it so there's the there's the little image video
740. 2:10:40Matchmaker or Veno Matchmaker and let's see what wine pairs well with salmon I'm having pasta tonight any wine
741. 2:10:46suggestions could you recommend wine for a vegan dish so so this is done so I can test it now right um what wine would you
742. 2:10:55recommend for a vegan dish absolutely I'd be happy to have it for you great okay so then what we're going
743. 2:11:03to do is this I'm going to go back and we're going to refine this now and actually I I'll do it over in configure so you can see that so in
744. 2:11:11general what I recommend is um you use the create mode to create it and then do any modifications over in
745. 2:11:22the configure mode it just works better because if you do it all in the create mode it can break [ __ ] um okay so I'm going to add I'm
746. 2:11:32going to add a a modification here so after you make a wine recommendation create a 16x9 wide
747. 2:11:50photograph of the glass of wine next to the meal that it was paired with period make the photo high quality as if it came from a
748. 2:12:05glossy magazine all right so now oops
749. 2:12:21um it should just do that like I think it updates it in real time so let's do what P what wine pairs well with
750. 2:12:30salmon so actually it's going to create the image first which is not what I told it to do but you can go fix
751. 2:12:39that and by the way for those of you that said you were going to build gpts tomorrow if you're still here you should
752. 2:12:44be building along with me I hope you are so there's our little wine with our salmon gave us a Chardonnay now now here here so hang
753. 2:13:00with me here let's say you've got a Vineyard and you've got your wines and you describe them you could do this for
754. 2:13:07your specific wines or you could do this for a wine store or you could do this for a restaurant I mean take your like like
755. 2:13:18one of the things that you can do in the configure mode is you can upload files you could upload knowledge right so let's say
756. 2:13:27you've got a wine store you could put a database of your inventory of your wines with descriptions of the wines of the brands with
757. 2:13:35prices and then on a weekly basis you could say here's our wine specials as a separate knowledge that you put in here
758. 2:13:43and then you could give out this tool for someone to pair wines from your store with the meals that they're having that
759. 2:13:52weekend and you could put a call to action in here that says if you want to buy this wine click here and order it
760. 2:14:00and we'll deliver it or whatever it is right so so these things can they can get very sophisticated very fast and
761. 2:14:09then so in the upper right hand corner now I'm going to save this and so I can save it for just myself anyone with the
762. 2:14:16link or I can publish it to the to the store and we're going to put it in lifestyle category and I'm going to confirm
763. 2:14:26that boom so if you go to the GPT store if you go to explore gpts and you search for Veno Matchmaker you'll see this very GPT so
764. 2:14:41that's what a GPT is all right um I got to get out of here that was a that was a long [ __ ] rabbit [Laughter]
765. 2:14:54hole all right I don't know is like Giant on mine um tired Runner what input any code or image you produce by using it is that
766. 2:15:06legally their property wait what about any code or image you produce using it is that legally there if someone uses
767. 2:15:15your GPT yeah I'm pretty sure they can do whatever they want with it thank you Lord digital Gods thank you for the
768. 2:15:22roses that very nice very much appreciated um say you create something spectacular and try to sell it I you can
769. 2:15:31absolutely try to do that there there's lots of different ways you can you can try to monetize these things one way is
770. 2:15:37you could just say um what what Suzanne uh Welker Jurgens does is she puts a call call to action in her gpts that
771. 2:15:47says this GPT was created by her at this company and you can see a list of all the ones I've built here and if you want
772. 2:15:55to hire me click here other people are doing calls to action to go to their website to buy stuff um one of the women
773. 2:16:03from the um Ann Murphy's GPT weekend thing um her mother did one that had affiliate links in it and apparently got enough people to use it
774. 2:16:15that she made $30,000 in affiliate commissions uh since they launched the GPT so like these things these things
775. 2:16:23are real here's the other thing not a lot of people have access to them right like this is only available for gp4
776. 2:16:31users how's it protected wouldn't the AI app also own it since it created it I in terms of the I do not don't worry a ton
777. 2:16:44about the IP [ __ ] right now just assume everything is public domain everyone will steal anything you put in your GPT
778. 2:16:53there are ways that you can discourage people from stealing the [ __ ] within your GPT but these are highly non-secure
779. 2:17:01systems so don't put anything proprietary in there that you don't want out in the world and I wouldn't worry a
780. 2:17:07ton about monetizing it right now the infrastructure is is way too janky right now to do it I would think about it I would play and
781. 2:17:18experiment and put in affiliate links and like I would play up one side and down the other but just don't put too
782. 2:17:24many expectations that this is an infrastructure that is ready for prime time because it's not Kyle the question
783. 2:17:31is owning images Etc the the images that are created out of this you you own they are original images that are generated now they
784. 2:17:42are they are derived from a data set where there might be um copywritten things in
785. 2:17:59there um but but you you own them so how would you implement affiliate links You' just build it into the prompts you'd say when
786. 2:18:10when appropriate or when you answer a question you know or when you recommend a product let's say you put in a a whole
787. 2:18:16bunch of product recommendation when you recommend a product include the link and that link has your affiliate
788. 2:18:22Link in it you could upload a CSV file so anyway all right 32 minute goodbyes all right I know I gotta get
789. 2:18:31out of here people I gotta get out of here all right um so thank you to everyone that's been here there's a lot of people in here tonight
790. 2:18:45um tonight was a little weird a little odd we sort of went deep in a lot of places it wasn't nearly as soap boxy as
791. 2:18:51the other night so I don't know so cool dude okay good awesome thank you as always great poly z um thank you to my
792. 2:18:58mods thank you to my subscribers thank you to the Irregulars that show up here night after night after night after
793. 2:19:04night Cindy [ __ ] if you're still here travel safe um show and tell let me let me go quickly to show and tell then I got to
794. 2:19:13get out of here I got to get out of here people yeah people another thing Johnson code Creator hey I learning nice
795. 2:19:26mechanics you see you see what we do here the rabbit R1 everyone's laughing at him yeah that's [Laughter] me very
796. 2:19:45cute beautiful all right all right everybody are you uploading the videos to YouTube I didn't do it today but I will do it tomorrow
797. 2:19:56please post this Live on YouTube I want to learn how to do this keep coming back here keep hanging out so everybody okay
798. 2:20:04last little thing I'm going to do one last little ad here go to the AI the salon. go there that's a community of people who
799. 2:20:15if you're curious about this stuff you want to be a part of this community go there click on the first link check it
800. 2:20:22out check out about the salon check out our values make sure that you resonate with the values if you do this is your
801. 2:20:30people if you don't probably not your people but take them seriously they're important um that's it all right I got
802. 2:20:39to get out of here thank you all so much let me do a little uh a little Romper rooming thank you Danielle be not
803. 2:20:46afraid Clarence Hopkins good to see you Tobias thank you Amelio's wife Clint Copperhead cookie Brandon thank you so much digital
804. 2:20:56Gods that's Lord digital Gods to you Angelique thank you Kyle and mods good night to you all exactly the mods
805. 2:21:06make this this manageable somehow paully Z thank you Joker good to see you Mary Mary oh that's very beautiful Robert
806. 2:21:15Rossy thank you very much that was super cool thank you Silver Fox thank you for the Roses very much appreciated quick take
807. 2:21:22Jake Mary Mary all right all right fantastic Bob tell him what he's won you've won a new car you can make money
808. 2:21:33all right everybody peace out thank you EMP power I I appreciate a lot quick take Jake Big K thank you you're welcome
809. 2:21:39Robert Rossy thank you thank you thank you JC good to see you all right everybody that's it I'm out of here I'm
810. 2:21:45going to get rid of this giant thing that's all over my face peace out Aloha from Hawaii wow awesome Aloha see you guys tomorrow