The Invisible City Workshop

Transcript

1. 0:00Tik Tok so if you want to follow me on Tik Tok it's the AI learning lab I go live nightly for about two or three
2. 0:06hours just talking about AI so I should be able to talk about this tonight um just quickly before we start
3. 0:13how many people um are would consider themselves active AI users okay how many people have never used Ai and it's okay like okay too and
4. 0:27how many people have kind of heard of it and they're like a little like h what is this okay two or three okay good
5. 0:37cool on my first day of seventh grade in 1977 I walked into my math class and that's kind of what it looked like these
6. 0:50aren't real computers this is mid Journey imagining what a seventh grade classroom looked like with a bunch of
7. 0:56computers on the desks and it's actually kind of what it looked like what were on the desks were Radio Shack
8. 1:03tr8s which if there's some old people in here you've seen those before with the Dual 5 and a/4 inch floppy drives oh yeah and
9. 1:14as as I learned what these machines were and I learned how to program them I realized holy crap this is going to
10. 1:24change everything that that that using these things felt like what it must must have been like when the the printing
11. 1:31press was invented or when the steam engine was invented like this felt like one of those and I remember you know I had
12. 1:42Heroes and you know Steve Jobs and and wasak that that founded Apple and Bill Gates um and and the team from Microsoft
13. 1:51that created this and I remember in seventh grade being really sad that I was just too young to be one of them that I had missed it
14. 2:03that I just knew that this was going to change everything and I wasn't a part of it and I knew that I was going to like live
15. 2:14with that for the rest of my life jump forward to 1994 I'm living in New York City I have a degree in acting so it's perfect that
16. 2:25I'm talking about AI right um I moved to New York out of school uh to pursue a life in the Arts I started and ran a theater company
17. 2:36um that didn't work out financially it was good creatively but you know you don't make a lot of money doing that so
18. 2:42I thought I know because I'm a genius I'll write myself an acting career so in two years I wrote seven feature length
19. 2:48screenplays um and and for my day job you know you have to you have to hustle in New York for my day job I was doing
20. 2:54um desktop publishing and I kept hearing people talk about the internet and I and I kind kind of got on the internet and
21. 3:01the internet was kind of cool but you had to dial up right and everything was command line and you'd go to FTP sites
22. 3:07or tet sites and these libraries would have like dissertations and documents and I'm like this is kind of cool and
23. 3:16then people started talking about this thing called the worldwide web and I started learning about it and
24. 3:23I I I thought well this seems kind of amazing that that there's there's something where you can include pictures
25. 3:29and it's got these things called hyperlinks and and I said but surely I can't win that and so I started reading about it
26. 3:37and I started talking to other people about it no one knew about it a handful of people knew about it and I started
27. 3:43asking questions you know what is this worldwide web thing what's this HTML and the more I learned about it the more I
28. 3:48realized oh crap the way you structure a website is almost identical to how you structure a screenplay it's just
29. 3:55storytelling in a different form and I was like holy crap maybe I can do this and so I had an idea I thought the
30. 4:03internet right now is is full of dissertations and research and Technical crap there was no art there was no there
31. 4:10was no culture and so I came up with this idea for an online magazine called Urban desires and I went on vacation with my
32. 4:19wife and took my Mac 660 AV and a 13-in color monitor and I taught myself enough HTML to make this this is the first the
33. 4:27homepage of the first issue of urban desires from 1994 you'll notice that the the um you could click on the the cover and see it
34. 4:37big like this but you'll notice it's left Justified why because the center tag didn't apply to images in
35. 4:441994 yeah exactly um but I I so I got I got home and I I had this this magazine and and I didn't know what to do with it
36. 4:54next and so I asked one of my friends that knew something and they're like well you got to put it on an internet
37. 4:59server and I'm like what's an internet server and they said well I know one in California and they gave me this phone
38. 5:04number of this computer in California so so I had my computer call this computer in California oh [Music]
39. 5:19wait and if you don't know what that sound is I hate you that means you're too young there was a time when we had to
40. 5:27have our computer call other computers to talk to them the internet wasn't just there so I published this this Magazine
41. 5:36on this thing called the internet and I'm like okay like what now and 3 weeks later I got a I got a note on a on a
42. 5:45bulletin board a dialup bulletin board that I was a part of and someone said hey Kyle did you know that Urban desires was written up in
43. 5:54libero the Parisian News Daily and I was like what and they said you can I get a little choked up over this you can go to the international
44. 6:08news stand at Times Square and get it so the next Sunday I went to Time Square and I bought this French newspaper
45. 6:19and um flipped through it and sure enough there's a full page article of this this magazine that I created in my
46. 6:31apartment in Brooklyn and shipped to some computer in California written about in Paris 3 weeks later and what hit me as I stood in the
47. 6:41middle of Time Square was that the world just changed and nobody knows it yet um it was
48. 6:55remarkable and from that I co-founded agency.com which was one of the first digital agencies and we ended up building
49. 7:04websites for the Fortune 500 we ended up growing to 2200 people it was it was AB in 5 years it was absolute
50. 7:11Insanity um we built British Airways first website we built their ticketing system we built the first um streaming
51. 7:18technology for serious satellite radio um because none of it existed all of it had to be invented um and you know
52. 7:30the early days of the worldwide web get a get a bit of a bad rap like history hasn't treated it well because of the do
53. 7:36boom and bust everyone's like oh that was just crap and yeah no but it changed the world and if you were in if you were
54. 7:43playing with this stuff in those early days it it wasn't like the oversold hyped up thing that everyone talks about
55. 7:51in 2000 it was people that just passionately believed in this oh my God this is going to change things so and I I I really thought that
56. 8:02that would be the last time in my life that I would get to be at the beginning of something like that like just I felt so [ __ ]
57. 8:13lucky right to be a part of it and I kind of knew like these don't come along that often right it's like
58. 8:21fire the wheel or the wheel fire you know language writing printing press steam engine right there's not that many
59. 8:29of these the 20th centuries had a lot but there's not that many so jump forward to 2022 this image is one of the very first
60. 8:39um AI images I ever created um i' kind of done a deep dive into nfts and blockchain and I was I was you know kind
61. 8:46of into into that but I got involved in this project called pixel mind and it was an nft project where it was kind of
62. 8:52like Mad Libs where you just fill in like an animal or whatever and it would make a piece of art and it was really
63. 8:57cool and this was one of my first pieces of art and I'm like wow that's amazing and and you know it had like
64. 9:04uh like metadata about the information and every every image you created had a video attached to it and this is the
65. 9:13computer making that art you got to watch it make the art and I was I was like I was like huh that's curious and for the next six months or
66. 9:30so this is the beginning of 2022 I would sort of go to this pixel mind site and I I would do it in batches
67. 9:38I would go make you know 20 images or 100 images or whatever and you could mint them into nfts which I'm like oh
68. 9:43cool I can sell these they're worthless no nobody gives a [ __ ] and then one of the one of the days I went there and there was this madd's
69. 9:54prompt for I don't know some sort of futuristic cyborg portrait and so I did Bill Gates and Steve Jobs and I threw Elon Musk in
70. 10:03there because we're going to Mars and he you know made the electric car so so you got these kind of kind of titans of
71. 10:13innovation and what what struck me when I was doing this was like what's the point like I'm making all these images
72. 10:21I'm making all these what what am I using them for what am I using them for and I thought what I should learn what's
73. 10:28underneath this and what was underneath this is a thing called stable diffusion which is an image tool you can play with
74. 10:33right now it's free it's open source a lot of drama going on at the company right now if you follow that stuff um and
75. 10:42um and the more I learned about it it was kind of like that HTML stuff that I learned in the '90s where it was like I
76. 10:50think I can figure this out I think I can figure it out and I learned that you had to use a notebook from this thing
77. 10:56called Google collab and you had to get a Google collab account you got the notebook and then you hit these play
78. 11:00buttons and it would do the programming for you I failed nine times nine times like not nine times in a sitting like
79. 11:09nine different times over a two-month period I tried to get stable diffusion working and I would hit the play button
80. 11:14and it would do its thing and then it would give me an error and I didn't know what to flip and do and then I eventually got it and and
81. 11:23I I also installed this thing called dream Booth which let me upload um 25 pictures of myself and then I could make
82. 11:32these crazy portraits of me because the internet doesn't know what I look like it knows what Steve Jobs and Bill Gates
83. 11:37look like it doesn't know what I look like so I got this thing working and I started making all these crazy
84. 11:43self-portraits of myself and they were bizarre and they were crazy and they were me but they were not me and like I
85. 11:51thought like I want to see that movie like I want to know what future he lives in and it's like oh like what life
86. 11:59choices did this guy make to to end up there right you know I'm pretty sure he's busting down your door for Thought
87. 12:07violations um and like yeah I might age okay you know um and if you want to see these you can go to um Kyle Shannon dreams both on
88. 12:17Instagram and Tik Tok what I ended up doing was I said I I want to know these stories and and you know I'm a writer
89. 12:25and and so I wrote stories for each of these little things and I was kind of like okay I think I'm getting what what
90. 12:34this AI stuff's about I think I'm getting and here's here's what struck me what happens with AI is especially now because you don't
91. 12:46have to go through all the technical [ __ ] I went through to get this up and running you can just do this now
92. 12:53um if you've ever done anything well if you've ever done [ __ ] anything right everything everything you do takes a lot
93. 13:00of effort right it takes a lot of time it takes a lot of effort what I was experiencing with AI was I put in a
94. 13:08little bit of effort and I was getting back massive amounts of value little bit of effort massive amounts of value when
95. 13:15I was doing these images one of the images It produced took my breath away like literally I went like oh my God because what it
96. 13:26was was an image that if I had the talent is the kind of image I would have created and that's just like it's like wow holy crap
97. 13:39then as if that weren't enough this happened everybody know what that date is anybody know what that date is that was the day chat GPT came
98. 13:58out chat does for AI and machine learning what the worldwide web did for the internet one of the big misconceptions
99. 14:11about AI is like I got to be a mathematician I got to be a scientist I didn't go to MIT I don't have a degree from
100. 14:17Stanford prior to chat GPT you really did have to be a mathematician you really did have to be a scientist or a researcher or a
101. 14:25developer chat gbt made it simp Le for the rest of us to tap into the power of this stuff that's been developed for 30
102. 14:3440 50 years probably 70 years if you go back to the original neural network designs we are at a remarkable time in
103. 14:44history and this date in my opinion is the beginning of of an entirely New Era an entirely New Era so let me give you a
104. 14:55little bit of context of or or just aoup couple of pieces of information about chat GPT um one is bonus question anybody
105. 15:07know what GPT stands for graphical close I remember but I can't generative pre-trained Transformer the G the G in it's the
106. 15:23really important one generative it's Genera it's not copying and pasting that's one of the massive misconceptions it's not how it works
107. 15:32it's generating original stuff pre-train means it was pre-trained on the internet Transformers just the technology so
108. 15:37here's some important dates 2017 Google invents this thing called the Transformer that's the T and GPT it was a paper called attention is
109. 15:47all you need they invents they invent this thing that allows machine learning scientists to train these um models on massive data
110. 15:57sets um open Ai and Sam mman were were the the ones who really understood the more compute we throw at
111. 16:07this and the more data we throw at this the better they seem to get and so they did gpt1 in 2018 and gpt2 I think the
112. 16:15year after that gpt3 comes out in 2020 so what they did was they got as much of the public internet data as they
113. 16:24could and they they coded it in such a way that they could put it into this model and gpt3 came out in
114. 16:322020 and started to be really remarkable if you were paying attention to to you know internet and Tech stuff back then
115. 16:39like people buzzed about this a little bit but you still had to be an engineer the interface to use it was this thing
116. 16:44called the open AI playground the GPT playground and you had to pick models and there were weights and temperatures
117. 16:50and it was just it was it was too [ __ ] hard um interesting little tidbit in the documentation for gpt3 were the instructions for how to
118. 17:04build chat GPT and nobody built it wow yeah so in in yeah exactly in the fall of uh 2022 or 2023 it becomes apparent
119. 17:18that Microsoft is going to invest in open Ai and so Sam Walman says we should probably have something to show like
120. 17:25when we make this announcement that they're making this big investment hey Tech kids uh could you just go whip up
121. 17:30that chat GPT thing that no one's built they went out for drinks the night before and they assumed that no one
122. 17:36would care they figured it might get like a couple of blog posts and 5,000 views or something like that the little the straight red line is
123. 17:46chat GPT the horizontal axis is how long it took to get to 100 million users um to put what's happening with AI
124. 17:56in context of the worldwide web to get to 100 million users it took the worldwide web about 6 years it took chat GPT 6
125. 18:08weeks it was the fastest adoption of technology in history they didn't think it was a big deal interface
126. 18:18matters and what I mean by that is what is chat GPT if you haven't seen it before that's kind of it that
127. 18:29box it's kind of like the Google search box right you don't have to pick a model you don't have to know anything geeky
128. 18:36you just put words in there well what words you put in there you can put anything in there you can ask it
129. 18:40anything you can even ask it what's so special about chat GPT in an Eminem rap and it will write you an Eminem WAP
130. 18:51talking about how awesome it is this was one of the first things I did this what I like to call parlor tricks one of the
131. 18:57great ways to play with chaty do parlor tricks just ask it stupid [ __ ] and it will do it I won't rap for you I will yes
132. 19:12um and and and then I thought huh this this is remarkable so I was starting to see in language what I was experiencing
133. 19:18with those pictures what's fascinating is the underlying technology for the pictures and the music and the video and
134. 19:25the words it's all the same thing it's all based on this Transformer architecture you just basically train it
135. 19:29on different data and it gets good at different things it's just stunning um so I started this group the
136. 19:37week after chat jpt came out called the AI Salon I'll tell you about it afterwards I'd encourage you all to join
137. 19:42if if you're interested if you want to join a bunch of weirdos that are really into AI um so I started this this thing
138. 19:48and I thought well I I got to come up with a a plan for how to let people know so I said hey chat gbt can you give me a
139. 19:54social media marketing plan and boom there it came just knocked out and and one of the one of the suggestions was
140. 20:01well you do paid media and I'm like no no I'm broke I no I no I don't have money for paid media I said but could
141. 20:08you give me a table that compares earned media and paid media and boom it formatted and did the analysis comparing
142. 20:16paid media and earned earned media I was like oh that's cool and then I wrote an email and it looked like [ __ ] and I was
143. 20:21like I don't remember how to do HTML anymore I had one of the largest digital agencies in the world and HTML doesn't
144. 20:27it's not the same so I said hey chbt can you make my uh email look prettier in HTML that would be good for a service and it wrote the
145. 20:37code and I just kept having like mindblowing experience after mind-blowing experience after mind-blowing experience I'll I I'll tell
146. 20:48you the thing that that did it for me so I so it did this HTML that I that I could understand but for years people said oh you should learn
147. 20:57Python and and uh and I I was like okay I can learn python but I have ADD I can't really program I don't have the
148. 21:04attention spam for it so I said hey chat GPT can you write me a python script that that does I forget whatever it just
149. 21:12like takes an input and so it wrote this python code I'm like that looks great and then I had this this um kind of
150. 21:20shattering moment where I realized I don't know how to run python code like I don't even know how to run the code at
151. 21:26wrote but I said hey chat gbt how do you run python code it said well you open up terminal and you put in this
152. 21:32command line I'm like no I don't want command line like isn't there a website I can just copy and paste this it was
153. 21:36like oh sure here's five and the first one was repet which is kind of the famous one and I copied the code and I
154. 21:42pasted it into repet and it told me to hit the play button and I hit the play button and it gave me an error and I'm
155. 21:47like oh here we go again like with the nine failures of stable diffusion and I said it gave me this error and chat jpt said this it said in
156. 21:56the lower left hand corner of the screen click on this button in the upper right hand corner of the screen look for this
157. 22:02library and hit install and I did it and I hit the play button and it worked within 90 minutes I had written a python
158. 22:09application fully functional python application that took inputs from a user and generated five pieces of social
159. 22:16media content I was doing API calls to open AI in a python application I had written in 90 minutes absentmindedly
160. 22:24while my wife's watching some shitty TV show I've never written a line of python in my life I still haven't in that moment I had that same
161. 22:37like just [ __ ] electric thing from Time Square oh my God the world just changed so this stuff isn't just fancy
162. 22:50Google the G and GPT is really important it's generating I like to think of um Google and and Technology up to this
163. 23:07point the internet up to this point has been information on demand you ask it for information it gives you information
164. 23:14but then you as a human still have to learn the [ __ ] AI is knowledge on demand we we all if we're willing to
165. 23:24learn this stuff and quite frankly I think we're going to have to um I don't think it's optional um we get to level up skills we
166. 23:35never imagined we' have we we had it's it's it's staggeringly powerful so one of the other things so I've got a company my company's called
167. 23:45Story vine and we've got a video storytelling platform and I and and one of our Hallmarks is like authenticity
168. 23:52like authentic humans telling stories that's what our our software does people answer questions in an app it goes up to
169. 23:57the cloud 5 later you got a fully edited video and I thought well I should at least see where the state-ofthe-art is
170. 24:04for not that like I should see like what am I competing with so and we were talking before about uh personas and
171. 24:13audience segmentation and stuff like that I thought I thought what if I could could how would I do this like how would
172. 24:19I get you know one of these tools to tell a story and so what I started with is I said hey chat gbt and I don't say h
173. 24:29gbt uh but but I said um help me come up with a Persona framework and and like you know what's someone's educational
174. 24:37background and uh societal background and what's their health condition and things like that and then I had it
175. 24:43generate 12 personas generate 12 people with their names their age their whole history and I then had each of those people
176. 24:54answer a set of Story vine prompts the way our software works we write a set of prompts that humans answer and I had
177. 25:00these personas answer those prompts what what blew me away is that the stories that they told were very familiar to me
178. 25:08it sounded like what humans would tell because it knew who these people were and then um I went to Mid journey and I created um a
179. 25:18photo and I thought well I can take the the script of what they said in text and I can put that in a voice synthesis
180. 25:25thing and I can take that photo and I can make that into a video and then I could take those assets and put them in the story Vine
181. 25:34system and I could probably wait hang on I want to change this so the speakers are here do
182. 25:51this hello I'm Amina a I was diagnosed with illness about 4 years ago what brings me joy is spending time with
183. 26:01my family especially my parents and siblings we enjoy cooking and eating traditional South Asian dishes together
184. 26:10moving beyond my diagnosis was not easy but I found comfort in my faith and my family's support so you don't need to
185. 26:20see the whole thing this was from February of 2023 so this was more than a year ago that that this te technology existed
186. 26:31um I was really kind of excited about this cuz I was like wow that's like this could serve sometimes our clients can't
187. 26:38find people to tell their stories and there might be something here I got really excited about it and I showed my
188. 26:44staff and I almost had one person quit I had three people like the person that almost quit said um hey I thought we cared about humans
189. 26:54here yeah and that that put me down for a while I was like oh [ __ ] right so I kind of I kind of let this go
190. 27:03um but but this was February 2023 right so with a little bit of creativity duct taping a couple of tools together you
191. 27:10can do something that's quite remarkable um this is a project that I just launched last week um and it's it's
192. 27:19um if if you uh use the $20 a month version of chat GPT chat GPT plus you can create these things called gpts
193. 27:28which are customized versions of chat GPT and so I'm the chief generative officer for this marketing organization
194. 27:35called content Evolution it's a it's a Federation of a bunch of different marketing companies and so we have like
195. 27:41a little AI Skunk Works and what we decided to do was could we create virtual twins of ourselves and so we use
196. 27:49this GPT technology which by the way is remarkably easy like remarkably easy like in the second half of the workshop
197. 27:56tonight we could even make one if you want there's lots of things we can do tonight um but what we did was was we
198. 28:03decided we we used chat GPT to help us create a very structured set of interview questions like 25 interview
199. 28:10questions that talked about work history and personal history and cultural influences and I had know background and
200. 28:16just all that sort of stuff and we did video interviews of these 11 people and then we took the transcript of that
201. 28:22video interview and it becomes the knowledge base for each of these gpts so so every one of these you can go on on
202. 28:30the internet right now if you have access to to gp4 um and you can have virtual conversations with a PR Specialist or a
203. 28:38futurist or myself talking about AI it's it's really quite remarkable um another thing that happened in the past I don't
204. 28:45know four or five months is the the the technology that allowed me to do the Amina video um you can now
205. 28:54do that with yourself um and you can do it in different languages so here's me uh speaking fluid Italian so we've heard
206. 29:02from a number of clients that modular content is important to them getting more efficient with how we create content and how we leverage
207. 29:11[Music] content uh behind me you see these little green boxes um that's how we think about stories and story B we break
208. 29:21stories down into little atoms right little modules guess you didn't know I spoke Mand
209. 29:37[Music] [Music]
210. 29:52either uh and and you're probably thinking yeah but it's probably not good right here's Kyle talking Mandarin using
211. 30:05AI what he wanted to know is what does a native speaker thing let's find out that was yeah Mandarin is really
212. 30:17good it's talking about the AI story something my Mandarin is really good well it was AI yeah it was AI it was AI
213. 30:25it was AI yeah it wasn't him oh he AI yes right what platform did you create that create that on what's that what platform
214. 30:39did you create that on hen hen h h e y g e n um so so
215. 30:54question how many people are super excited right now how many people are super freaked out right now how many are
216. 31:03both yeah that that's I think that's the appropriate place to be um so a couple of things the world has
217. 31:16changed we just don't know it yet most people don't know it yet McKenzie did a report um they they they took two
218. 31:25measurements one was um February of 201 3 and one was November of 23 in February of 2023 87% of people had not used chat
219. 31:36GPT in November it was 82% 82% of people have not used chat GPT that means 18% have and I promise you this the 18% that have a bunch of
220. 31:53them probably used it once and went oh it's like Google yeah I'm good of that 18% maybe half have used a decent
221. 32:04amount so so now we're down to 9% of that 9% maybe half have bought chat GPT plus which is gp4 the more powerful model and
222. 32:16then of the people that bought it maybe half of those are using it regularly so right now somewhere in the neighborhood
223. 32:23of 2 to 5% of people know what the [ __ ] is going on and 95% don't and it's not like this is coming like this is here I made these
224. 32:34translation videos like four months ago and these are getting better I saw there's a thing today if you want to go see it called Hume AI
225. 32:43hum. where it's a it's an emotional conversational chatbot where when you talk to it it measures your emotion in
226. 32:52your voice and it shows you which emotions are in there and then it responds appropriately with its own emot
227. 32:58itions so like this stuff is coming and it's coming fast now I don't mean to freak you out um so here's what we're going to
228. 33:09do oh I I just wanted to share this if if you're really freaked out if you're thinking like oh God they're coming for our
229. 33:15jobs um a lot of jobs are going to change but they're going to change they're not going to go away some will go
230. 33:22away um but but this is a famous quote from Paul de ro from the 1840s where when he saw the first photograph the
231. 33:30first deera type he said from today painting is dead painting isn't dead every single time every single time
232. 33:41there's a new technology there's a revolt and the more powerful the technology the bigger the Revolt you know there was when when the
233. 33:51the weaving loom came out there was a mass there were riots in San Francisco it was going to take jobs right when the
234. 33:57steam engine came out the farmers all said what are we going to do well you know we can't be replaced by these
235. 34:02things well they moved to cities and they created an entirely new kind of you know avocation so the robots are not coming for us
236. 34:12these are tools we get to use and so for the second half of this what I'm going to do is let's take a let's take a five
237. 34:20minute break and go grab another drink go to the bathroom um and then when we come back um I've got sort of four
238. 34:28tracks that we can run down and I'm going to let you guys decide which one we do I'll try to get to more than one
239. 34:33of them but I don't think there's there's just so much so when you get back um let's say it's uh it's 10: after
240. 34:40right now so at 8:15 come back grab your seats um and then uh and then we'll we'll we'll move on and do the second
241. 34:46half of the of the presentation and I'm happy to answer any questions if anyone has any take a break [Music]
242. 35:04hi
243. 35:21everybody there's there's a if you watch it carefully when my hand goes in front of my face goes under my face yes it
244. 35:29goes up under here yes so so what what they're doing is they're they're they're 3D mapping face [Music]
245. 35:48[Music] and biting [ __ ]
246. 36:04[Music] yes but it's like you can say the same thing for social media the for me social media is just
247. 36:13smaller right so like the invention was the world the shift was the shift so the shift of PCS was the democratization so
248. 36:23so all of these are the democratization right the democratization of super computers with personation of this internet the demra this
249. 36:56is also super curious about [Music]
250. 37:11it a lot a lot question Shar the might be to check out but
251. 37:28basically out there probably a little bit smaller than you guys especially but
252. 37:45um if you're if you're still in the digital space it's a really cool what's it called buau of and um I'm having to talk to you about it
253. 37:56later and um yeah I actually ideas I'm expert
254. 38:26that I to
255. 38:56learn well I don't have
256. 39:17I remember I went to this conference and all of the all the
257. 39:56inter [Music]
258. 40:16[Music] for
259. 40:37[Music]
260. 40:54all right we're getting started again you weirdos is Kyle running it from there or is he ignoring us I'm ignoring you hi I can't talk to you
261. 41:09people let's take our seats again people artificial intelligence is very important this is serious sit down
262. 41:24yeah you don't have to sit down you can stay still um okay so and and the only reason I want to I want to tighten this up is
263. 41:39um there's so much to talk about there's there's so much um if you haven't guessed it I'm an optimist about this stuff
264. 41:55um I think it's very easy right now to demonize or dismiss AI cuz Hollywood's told us for the past 50 years the robots are going to kill us
265. 42:05the tools right now are janky as I say on my Tik Tok Channel they're janky pieces of [ __ ] they hallucinate which
266. 42:13basically means they lie they're inaccurate they're really hard to get to do things consistently well they're very frustrating um
267. 42:27I am more creatively alive than I've been since I was a kid it's it's one of the most inspiring like awe
268. 42:40inspiring experiences I've ever had in my life and it happens like multiple times a week that feeling one of
269. 42:54the AI needs a good PR firm cuz they have a shitty PR firm right now the fact that they call them chat
270. 43:03Bots right like we call it a chat bot so it's a bot and what are Bots Bots are cold and inhuman my experience in interacting
271. 43:13with these tools is not that if you understand how they work they're a calculator it's a probability engine it's literally the
272. 43:24way they work is you type a prompt and it creates a probabilistic waiting and it looks into thousand dimensional
273. 43:33mathematical space and pulls a token out that represents a word or a period and plops that on your screen and it does
274. 43:41that over and over and over again it literally doesn't know what it wrote but your experience in working with them like I
275. 43:52can't I cannot not say please and thank you when I interact with chat GPT because confession amen when it responds to you it's like
276. 44:06working with a really polite coworker that never gets pissed off at you and is always supportive of your ideas it's
277. 44:13like having a Golden Retriever as an assistant that just happens to know all of the [ __ ] in the world great idea boss I'm like oh yeah
278. 44:23that was a good idea so I'm going to give yall some options we can go down four tracks one is how do these tools
279. 44:37actually work that thousand dimensional space I talked about it's not as boring as it sounds and actually once you
280. 44:43understand how they work I think it's actually really valuable in using them because you understand things like why
281. 44:49do they hallucinate why do they not why are they not able to answer me sometimes so we can do that how how many people in here have a
282. 44:58chat GPT Plus subscription one two 3 okay how many of you have used more than just the chat part of it okay to yeah
283. 45:08okay good decent um one of the other things we could do is a chat GPT Deep dive it is you will be blown away
284. 45:18because people ask me all the time is it really worth the 20 bucks I think it's the best $20 you could spend in the
285. 45:24history of software but you know I don't feel that strongly about it so if you want to see that we can do
286. 45:34that it is stupid right now to learn anything about AI because three months from now everything will have
287. 45:42changed but you should absolutely be getting AI literate so this track is about how do you get AI literate knowing that things are
288. 45:52changing so fast like the next 3 years is it it is it it is overwhelming on on levels that are hard to describe but I I
289. 46:02we we can we can go down a path of what I think is the way to stay aii literate and not drive yourself crazy and then
290. 46:09the fourth one is I get this question a lot too what's the best AI tool and so we can go down that route
291. 46:15and I can tell you the best AI tool or dozen of them so I guess we'll do it like this look at that get the one you
292. 46:25want and then rather rather than vote with voice or anything I just want You' all hold up your hand with one two three
293. 46:32or four fingers up for which path you'd like to go down can we do two paths can we do two paths yes go ahead so I'd like
294. 46:41to go how I works the best AI tools okay how AI works best AI tools what else getting AI late okay we need to do a
295. 46:49vote okay so Okay so we've each got two hands so so so put up uh two hands and and put your first and second uh thing up
296. 47:04okay I got a lot of twos oh God this is bad okay we can't do it this way that was that was uh everybody raise their hand for
297. 47:15how AI which one that you want the classic voting style raise your hand for the one that you want to vote okay how AI
298. 47:23Works can we ask AI what we uh chat gbt Deep dive okay that's decent getting AI literate okay it almost it really
299. 47:36doesn't matter it's it's uh okay best best tool best tool oh best tool but but the the people that are left
300. 47:48like it's it it's essentially a four-way split so I'm going to choose very okay I'm going to start with how AI
301. 47:56works so I said this before AI isn't copying and pasting one of the things that you'll hear a lot is you know AI is
302. 48:01the world's greatest plagiarism machine it's stealing from artist it's stealing um it's a very very very complicated
303. 48:13gray area that will not be resolved probably In Our Lifetime all of these models for the most part Adobe is an
304. 48:25exception actually adobe's Firefly was not trained on on copywritten work but all of the large language models most of
305. 48:33the image models most of the music models most most of the most most of the things have been trained on um work
306. 48:40without people's consent but how it actually creates the content is not copying and pasting when you ingest a
307. 48:48document into one of these large language models and embed it what it actually does is it converts words into
308. 48:54what are called tokens which is just like a string of numbers numbers um and and it doesn't even know what words are
309. 49:02in there it's it's just a mathematical space so that's just an important concept to know it doesn't doesn't
310. 49:10really make a difference but when you hear someone say oh it's plagiarizing it's not the G in GPT stands for generative it's generating it's
311. 49:19generating original word so how does it do it generative pre-trained Transformer again generative means it's generating
312. 49:25pre-trained means it was pre-trained on all of the internet you don't need to bring your own data you used to have to
313. 49:30that's why you had to be a scientist and then Transformers is the technology you might hear llm which stands for large language
314. 49:38model then if you geek out a little bit if you're technical you might hear this thing called the latent space and the
315. 49:46latent space is a mathematical representation representation of word clustered by meaning so this is a
316. 49:54two-dimensional visualization of latent space every one of those clusters is clusters of of of these um these uh
317. 50:04token or these uh wait tokens no these vectors uh so like you have an XY Vector in a two in a 2d thing you have an XY
318. 50:12Vector right and the Clusters are semantically grouped words so you might have the word dog in the light blue
319. 50:22section and dog in that case means the The Fluffy animal you have in your house and then you might have dog in the purple section
320. 50:30which means he did his girl bad and he's a dog right so so you can have the same word have different meanings the real
321. 50:37latent space is like a thousand Dimensions like we can't conceive it it's insanely complicated but here's how it
322. 50:44works when you put in a prompt the the large language model predicts the next token or the next word just to simplify
323. 50:51it over and over and over again and it kind of works like this if we type in the prompt the quick brown fox jumps
324. 50:57over the lazy blank as humans our large language model goes dog we know that it's dog so how did our brain do that
325. 51:05well how this brain does that is it does this [Music] that mathematical waiting basically says oh the the the word at the end of that
326. 51:17sentence the high it's got the highest likelihood of being in this cluster and then within that cluster of vectors one
327. 51:24of them has the highest probability of being the right word and what it actually looks like is like a big long
328. 51:32ass number that is then associated with the word dog and then it puts the word dog into your sentence so when you see
329. 51:40chat GPT write a whole essay for you it's doing it like this one token or a little cluster of tokens at a time it
330. 51:49doesn't know what it wrote so when you ask it something and it gives you the wrong answer and you're like hey dum dum
331. 51:55you gave me the wrong answer give me the right answer and it's like oh sorry about that boss here's the right answer
332. 52:00and it's wrong again and and what we're used to is computers that are probabilistic computers that are predictable if I ask it something it
333. 52:11gives me the right answer that's not what these things are doing the G generative it's generating stuff the when Sam Alton got fired
334. 52:21there's a thing coming called qar that is a reasoning engine that we think is going to make these things really powerful where what what
335. 52:31it will do in the future is it won't just vomit out words it will vomit out words and then it will use reasoning to
336. 52:37say are these the right words and then it will fix them and then give you things so the problem of Hallucination
337. 52:44is going to get better and better and better and better and it's probably going to happen pretty quickly but
338. 52:48anyway that's how they work isn't that fascinating I think okay absolutely How do images work this is even more okay
339. 52:56that there are mathematicians on the planet that figured this stuff out like holy crap we get to play with this stuff do you get that
340. 53:11okay the way the image models work is you take your large language model and you pair that with with what's called a diffusion model and
341. 53:20then the images that it are it's trained on they take a noising algorithm and they turn all of the images into random noise images using an
342. 53:30algorithm and then that Vector space that I talked about just basically points to a bunch of unrecognizable
343. 53:37noisy images so when you give it a prompt the large language model does that same thing with the vectors and it
344. 53:43says if if our prompt is Cat on a desk it sort of says okay here are a bunch of here's a cluster of pictures that have
345. 53:52been labeled cat here's a cluster of pictures that have been labeled desk and then it takes these random noise images of cats and
346. 54:02desks and it averages them into a random noise image and then it uses the same algorithm that made them noisy to denoise
347. 54:17them and so it creates what you think is a copy pasted image but it's a completely original image that was sort of literally like
348. 54:28materialized out of fairy dust out of like probably thousands of images that are no longer images it's just this
349. 54:34noise noise pattern that get denoised and turned into this stuff so again when you understand how it works it's like
350. 54:41wait how do these things work that what are they doing all right so that's that questions yes what's noise what's noise noise is
351. 54:51just like you know when you're well no you're not old enough how old are you 34 yeah right on the border do you remember
352. 54:59Olden timey TVs when when it went off air it would go yeah that's noise so so it basically takes the image and just
353. 55:09noises it it makes it noisy so there's an algorithm where it it takes it from looking like this and it goes backwards
354. 55:18and then in the database are just a bunch of those things and and the algorithm that that made it that noisy
355. 55:25it then takes the average thing and turns it into that yeah and why does the noise have to be added because that's what the
356. 55:34mathematicians came up with I be because be oh for efficiency if if if you were actually copying and pasting these like the amount of data
357. 55:46processing you'd have to do would be massive like it's already massive so I think this just makes it more simple I
358. 55:51don't I honestly don't know good question okay any other questions you said AI is not an it I said AI is not an it is not
359. 56:06[Music] an no AI is a is a no I said it what it is is like a calculator it it
360. 56:21is an it but interacting with it doesn't feel like that because the data it was tramed on was it's not the bot is not the but
361. 56:30yeah you're not interacting with the bot you're interacting with the sort of collective output of humanity that's
362. 56:35been dumped on the public internet for the past 50 years and so what it feels like emotionally to me is like I'm
363. 56:46collaborating I'll tell you what's what's Wild For Me Co so I was a I was an absolute non-believer in remote work I [ __ ]
364. 56:56hate remote work I love people I get energy from people I work in a company my company is full of introverts when
365. 57:03when the lockdown happened they're like woo we're living our best life we hate people screw you up and and you know I I was like I I
366. 57:12like people I I like when I interact with chat GPT what what Co taught us was how to work remotely and what it feels
367. 57:20like is like I'm working with a remote worker who's really polite incredibly competent and clueless like like you know like
368. 57:29clueless interns that are like really smart like they went to the I leag League school and you bring them into
369. 57:34your agency and then they're like they're like you like go write the thing and they write it you're like this is
370. 57:38horrible they're like oh right they don't know basic World stuff right that's how these things are but but
371. 57:45interacting with it is it's not a human experience but it is way more human it doesn't feel like you're interacting
372. 57:52with a robot I think calling it a bot is wrong not wrong but I I it's inaccurate um okay yeah I think chat GPT I'm I'm
373. 58:06between chat GPT deep okay let's just do two uh chat GPT Deep dive versus getting AI litered I think those are the most
374. 58:13important ones actually so raise your hand chat GPT Deep dive and then raise your hand getting AI literate what do you mean by AI what
375. 58:22will they count knowing what AI is well enough that you could talk about at another party deep deep deep dive well I'll see if we can do
376. 58:35both all right we're gonna we'll start out with chat TBT Deep dive okay um the juggler I used to I used to be a
377. 58:50juggler um all the things in the left basic chat Advanced chat voice chat custom instructions um dolly3 image generation
378. 59:03code interpreter custom gpts all of those are things that chat GPT does um most people don't know that so
379. 59:15here's what we're going to do who's got how we I my eyes aren't quite good enough you and I were talking about customer segmentation
380. 59:29right or or something like that yes give me give me give me a project give me something give me that project that we
381. 59:36want to do and we'll go play with chat GPT we we'll sort of exploit a bunch of these different things so can really
382. 59:45make anything like we can come any audience like how would they want to celebrate a holiday or how would they want to commemorate
383. 59:58okay so so we want to Define an audience and then we want to and then we want to be able to ask chat GPT how that
384. 1:00:05audience would want to be what communicated with could ask how they feel create an audience and ask how they
385. 1:00:11feel about AI okay how okay how they feel about AI okay great okay so so let me let me give you a little bit of um
386. 1:00:20let me or Orient you a little bit this is chat GPT um down at the bottom here is where you type stuff in um make me a
387. 1:00:29poem about poem Cheetos it doesn't matter you can typos are literally irrelevant it understands everything
388. 1:00:40wow uh do I have the internet turned on we do oh we might we might not be doing chat GPT if it's a little it might be a
389. 1:00:53little sleepy let's see let me reload the page make me a poem all right I won't do Cheetos about Fritos in the heart of the bag where the
390. 1:01:07salty winds blow lies a corn so bold where the true snackers go golden Treasures crisp in Rose they lay
391. 1:01:15Whispering Tales of crunch in the light of day so that's chat um this is this is what I call parlor tricks uh you could also do
392. 1:01:26things like um write me a LinkedIn article about the same and format using
393. 1:01:41markdown um markdown I didn't know what it was but it's a thing that makes things pretty uh let's see to make it easier to
394. 1:01:54use now a couple of things an article about the same chbt understands the context of what you talked about before
395. 1:02:03so it's got a memory um it's a really short memory right now so sometimes you'll be talking to it and things will
396. 1:02:09be going great and then it just kind of loses its mind it's because it actually lost its mind those are getting bigger
397. 1:02:14but right now that's just kind of a deal um but anyway so markdown made the big headline and so here's our LinkedIn article well
398. 1:02:24formatted you liter literally just copy and paste this over to a LinkedIn article that brief history the golden
399. 1:02:31corn uh and then let's see um make a table comparing Cheetos and Fritos
400. 1:02:46oops isn't this just a bunch of words it is and it's not there um is there value and taking it from the perspective of like an
401. 1:03:00authority or an AB absolutely and that's what we're going to do here in a second so this is just basic chat GPT chat
402. 1:03:07now um and and by the way like oops you know th this thing of just compare two things do analysis comparing this ridiculous thing with that
403. 1:03:18ridiculous thing one of the things I absolutely recommend when you're playing with AI St is don't take it seriously play with it
404. 1:03:29ask it to compare gummy bears and Sushi ask it to write a ridiculous poem ask it to create a computer program that does
405. 1:03:36something stupid because in doing that you'll sort of take yourself off the hook for feeling like oh my god I've got
406. 1:03:43to learn something this is so hard I don't this hurts um and you'll you'll start to find the boundaries of where this stuff is
407. 1:03:52good and where it's not good if you if you try to just stay too locked into what you know you won't actually learn
408. 1:03:58what these things make possible um okay so so let me show you a couple other things here um we're in chat GPT 4 right now
409. 1:04:10I'm going to flip over to chat GPT 3.5 uh I'm going to do the same thing U make me a poem about Cheetos this time you'll notice this is much
410. 1:04:23faster so chat gbt 3.5 is the smaller less powerful version but it's super fast so if you really just want to kick
411. 1:04:31out ideas starting to understand what the different tools are capable of is is is a power move um and and I can I can
412. 1:04:41show you some of some some of that um there's also in I'm trying to think where I want to go first let me flip back over here okay so
413. 1:04:54here's what we're going to do we're going to start a new chat and we're going to say um give give me an audience give me an
414. 1:05:04audience that uh we're going to we're going to learn about I mean be like a bunch of psychographics combined with a bunch of demographics so
415. 1:05:17okay so wait so so here's what we'll say I want you to act as a um population specialist
416. 1:05:32analyst uh analist and uh uh consultant please create
417. 1:05:47three um audience Persona Frameworks M so I can [Music]
418. 1:06:01choose one to start working with now when I say act as a population specialist analyst and consultant that
419. 1:06:13that was your question before does that actually work here's why that works so if we go back to the the Laten space
420. 1:06:18graph you've got all those clusters of words right um oh I here here's a good one this was Andre karpathy who who
421. 1:06:24works at open AI this this was an example that he used he said imagine you want to ask about a math problem and you just say to
422. 1:06:33to chbt hey uh explain this math problem to me and it will explain it and very often it will hallucinate and get it
423. 1:06:41wrong and the way he described it is in in the latent space in in its knowledge base is everything that's ever been
424. 1:06:49written about that math problem so there's the professor at Stanford that that wrote about that math problem and
425. 1:06:55then there all of the crappy students at some local college that wrote about that math problem that got it wrong so when
426. 1:07:00you just ask about the math problem it'll sort of confabulate good stuff with bad stuff if you say hey I want you to act as a
427. 1:07:08Stamford mathematician that specializes in this branch of mathematics and I want you to tell me about that problem you're
428. 1:07:14narrowing you're cutting out a bunch of those other areas and so you're kind of narrowing your your your uh the the sort
429. 1:07:20of knowledge base that you're going that's why when you when you treat chat PT kind of like Google you just give it
430. 1:07:26a quick little generic search it gives you shitty answers because it's sort of amalgamating everything anyone's ever said and they're
431. 1:07:34idiots so you need to find the areas of Brilliance within within this massive model okay so so this thing's going to go off now
432. 1:07:45creating an audience of persona Frameworks involves blah blah blah so it's going to tell us this and and it will so the demographic Focus
433. 1:07:54Persona pay ATT I'm sorry what's your name again Julie Julie so pay attention Julie all right so we're going to go demographics is one of
434. 1:08:02them um oh wait Persona name oh oh it's okay it's it so it gave us the framework so framework one is demographic focused and
435. 1:08:16then it's actually filling that out so so here's Fiona okay this is good this is good I might stop this at some point here cuz it's a little
436. 1:08:27slow by the way this this is very very familiar back to the the' 70s um when you had dial up modems this is how slow
437. 1:08:35they typed like you could watch the internet typing across screen and so we're going to look back at these times
438. 1:08:41and go hey remember when chat GP actually could read it when it typed that's not going to be here for long um
439. 1:08:48okay I'm bored anyone else board um okay this is framework three okay so Behavior fol focused psychographic focused or
440. 1:08:59demographic focused do you prefer one over the other I think psychographic will just especially if we're ask something about
441. 1:09:06like no I'm not good with that okay here's what here's what we're gonna do can you now create a single framework that
442. 1:09:20combines all three and gives us uh three unique personas based on it so one of the things
443. 1:09:39that a good way to think about chat GPT is not like a search engine you give it a prompt that gives you an answer that's
444. 1:09:45one of the big misconceptions you give it a prompt that gives you an answer oh the AI wrote that no think of it more
445. 1:09:52like a collaborative conversation like a jam session and give it more information than you think it needs the reason that I asked
446. 1:09:59it to give me three different P Persona Frameworks I don't frankly give a [ __ ] what it gave me I just want it to know
447. 1:10:07what we're talking about so by it giving us that and me telling okay yeah those are good now combine them into one it's
448. 1:10:14it's sort of I'm giving it increasingly narrow uh a narrow window for this conversation to live in so the more
449. 1:10:22context you give these things the better they get yeah question for you I I use this all day every day and I incorporate
450. 1:10:29a lot of additional references into my promps as I start generating them but the one thing I can't figure out and
451. 1:10:35this is such a weird question to ask is the punctuation that you find is the best way to separate the different
452. 1:10:42refences that you're including because you know you get obviously so the question is if if you're if you're putting in references
453. 1:10:51like you say I want you to go look at this article or you copy and paste you know something from a different article
454. 1:10:56in in your prompt if you're putting things in the prompt the question was how do you how do you provide um punctu
455. 1:11:02what's the best way to add punctuation as you add those different references in general think of have you ever done any
456. 1:11:09programming at all have you ever done any like like even basic programming like engineering anything the more
457. 1:11:15structure you can give it the better so so um you can do things like um incorporate the following references and
458. 1:11:23then colon bullet point reference one bullet point reference two bullet point reference three um ignore these
459. 1:11:30references and then so so so just create structure it likes structure so so and like even if you do the label of like
460. 1:11:40you know include these references in all caps with a colon and then bullet points below it it'll really understand that um
461. 1:11:46also play around with order order sometimes matters um okay so now we've got these things so we've got again I don't really
462. 1:11:54care who it is but we've got we got Fiona we've got Ella and then we've got Green Tech Greg I think we know Greg's from Boulder
463. 1:12:06um okay what do we want to ask these these folks um I mean I think start how do you feel about oh okay um
464. 1:12:22have Greg Ella and Fiona did I spell that right no shh I'm okay have Greg Ella and Fiona
465. 1:12:45um what about talk about how they would use AI given their respect I'm trying to do something I'm trying to just do something
466. 1:12:52stupid um rock Musical well that's that's where I was going to go let's see uh write a Hau about how they feel about AI
467. 1:13:11yes Greg silence server's home Nature's Balance in my palm teeken Earth has one entreer Ella Infinite Canvas dream
468. 1:13:21sculptured sculpted in digital AI my Ally family first Fiona AI cradles cradle dreams a future woven with care um let's see
469. 1:13:32um let's yeah let's uh make Greg um anti AI machines whisper lies Nature's truth
470. 1:13:49drowned by by circuits lost in Silicon okay perfect h yeah you can do that too more extreme okay perfect um now I want you to create
471. 1:14:02an image from each person about their poem so chat GPT plus has what's called di in it you can also um this di is also
472. 1:14:18inside Microsoft co-pilot which is free so if you go to co- pilot. microsoft.com that's a version of chat
473. 1:14:25GPT 4 that's free so if you can't afford the 20 bucks a month go go to microsoft.co pilot.com um you can do the image generation as
474. 1:14:36well I still think it's worth paying the 20 bucks for chat GPT for a couple of reasons it's just better and the the
475. 1:14:46um the the the way they've they they've coded Microsoft What's called the safety guard rails are really restrictive so if
476. 1:14:53you try to ask it about something like you want to Market tobacco or something like that it might give you a oh tobacco can hurt people so
477. 1:15:01I'm not going to answer that it's it's a real pain in the ass that way so is chat GPT for for that matter um but it can make images
478. 1:15:10um these things can also um chat GPT can also see you're like it can see what what do you mean it can see is it looking at you
479. 1:15:24right now oh my God hi people say hi to my life people wave hi hi I'm not ignoring you except I sort of
480. 1:15:38am oh oh no all right I got lots got lots of Hearts they liked you um okay so what did you you said what
481. 1:15:52were we going to ask them oh how would they use AI is that what you is that what you said based on their Persona um how would each
482. 1:16:02of them use AI um for a bake sale let's just see what it comes up with and
483. 1:16:16and like purposefully asking it weird questions that it shouldn't be able to answer is is like one of the ways I
484. 1:16:23learn what these things can do is like what it will try to do is it will try to give you everything it will try to do it
485. 1:16:32and sometimes it fails miserably and sometimes you're like oh my God that's good right it totally inspires a new thing so let's
486. 1:16:40see Greg anti- aai sourcing ingredients he uses AI to find and Source locally produced sustainable ingredients that's
487. 1:16:47good Greg energy efficient baking he employs uh smart kitchen appliances that use AI all right so he's using it on the
488. 1:16:57edges um might reluctantly use social media analytics and identify local community groups good good Greg it's
489. 1:17:03solid it's you know he hasn't given up he's not he's not totally resisting it Ella uses it to analyze food and Trends
490. 1:17:11um Inventory management okay so now we're going to try something cool I'm going to click the little paperclip right here which you have to
491. 1:17:23have chat GPD to get the paper clip but you can upload images and it can see let's
492. 1:17:46see is this it yes okay this is a whiteboard I got off the internet I went on the internet and I typed in whiteboard and I found a decently
493. 1:17:54complic ated whiteboard and I'm just going to say what is this the suspense
494. 1:18:10okay appears to be a whiteboard of handwritten notes and sketches the content suggests it's a brainstorming or
495. 1:18:17planning for a user interface so it could see that it there are sketches of user interface there uh likely focused on ordering services such
496. 1:18:27as food and drinks perhaps in an airport setting the out it it outlines three-step main process choosing a language ordering from a menu and
497. 1:18:36checking out which includes payment there are notes about Target users business and family so now we're going
498. 1:18:41to get back to our personas um use cases take out within within a limit of an hour possible okay so it's got all those uh I'm going to say
499. 1:18:52describe the um interface screens let just do
500. 1:19:13that so there's a welcome screen a menu screen it prompts the user to tap to start choose a language so we can go up
501. 1:19:21and look at that but it it had like all the little words in there it read all the little words that understands what the pictures are
502. 1:19:29um one of the items listed is Caesar salad with a price of $10 status screen checkout screens okay great so now I'm going to say um
503. 1:19:44write a set of five marketing taglines for Ella Fiona and
504. 1:20:04Greg you can ignore the AI stuff but focus on their um
505. 1:20:19demo psycho graphic profiles so again like it's you're not you're not coding it's it's just like
506. 1:20:35like this is if you were working with someone in the office and you were explaining I want you to go do this work
507. 1:20:41you would kind of explain it like this so if you just think of this interaction as like you're working with a coworker
508. 1:20:46and you're just having this back and forth interact action with them um for entrepreneur Ella Empower your ambition
509. 1:20:53perfect lead with Elegance sophistication meets Innovation for the modern entrepreneur redefine your workday craft your Empire beautiful for
510. 1:21:02Visionaries who demand excellence and precision um for family first Fiona cherish every moment family family
511. 1:21:09friendly picks um for the heart of the home oh wait but it so hang on it missed something here you missed that this is a food
512. 1:21:22ordering app for and Airport you dumb [Laughter]
513. 1:21:37Dum and you probably shouldn't be rude to the Bots cuz they are going to have memory soon enough my apologies let me try that
514. 1:21:48again again the way these things have been programmed they've been programmed to to they want to give you what you want
515. 1:21:55um so let's see if we if we got it better uh entrepreneur Ellis seamless sustenance dine on demand between deals
516. 1:22:03good boardroom to boarding gate satisfy your appetite for Success these are great perfect okay great so let's say um
517. 1:22:10take off to table time okay so here's what we want to do um make me three concept posters
518. 1:22:25in an airport setting for each Target uh Target Target
519. 1:22:40um include the name of the app um fly bites so there there I'm doing this is you know little structure put a little
520. 1:22:55structure in the name of that fly bites and um a um image of a mobile
521. 1:23:10device and the ordering screen so so you can just keep going between these modes right you can you
522. 1:23:24can say give me some Concepts give me some taglines put that tagline in a poster make an image make a this make a
523. 1:23:31that there's another thing called code interpreter in here I'm not going to demo it because it's going to take too
524. 1:23:35long and it's a bit boring but what code interpreter does is you can upload a CSV file of data so if if you're a little
525. 1:23:44geeky write this down write down the the website kaggle k a g g l e and what kaggle is is is a a site where you you
526. 1:23:54can go get data sets so you can get like New York City Taxi data from 1972 I don't know whatever crap like that you
527. 1:24:01just download CSV files you can upload a CSV file to chat GPT plus the four and say tell me about this data and it will
528. 1:24:12write its own python code and do the analysis of the data in that spreadsheet yeah what if the dat then don't put it in there if the
529. 1:24:25data is sensitive don't put it in assume everything right now in chat GPT if it were to get out and you were to get
530. 1:24:32fired for that don't put that [ __ ] in there um but anyway by the way there's our there's our entrepreneur
531. 1:24:42Ella right and her fly bites app here's the family right so like we just went from like how do I interact with how
532. 1:24:51these personas are going to go to flipping into cre development mode to concepting things like when you get
533. 1:24:58Nimble with all of the different modes in chat gbt you can literally like like I it's like a it's like a [ __ ]
534. 1:25:06superpower you can literally just be flipping between different mental modes right you don't have to be locked into
535. 1:25:14I'm a copywriter I'm an illustrator I'm a strategist I'm a this I'm a that play push the boundaries and you'll start to
536. 1:25:23discover you'll you'll start to discover that you're good at things you didn't know you were good at you can do things
537. 1:25:29you didn't even think were possible um so anyway and then it describes it I'm trying to think I'll show you one
538. 1:25:39other thing here um this you can do in both chat GPT um normal and chat GPT plus um under your name by the way how you get to chat GPT
539. 1:25:52plus you click on your name you go to my plan and it's the the Plus Account right here it's 20 bucks a month um click on
540. 1:26:00your name there's this thing called customized chat GPT this is called custom instructions one of the things
541. 1:26:06you'll notice about chat GPT is it's a little wordy it's a little verose um and it can be really annoying
542. 1:26:13so what this allows you to do is give it a personality so I can say you are a pompus s who uses words so big even you don't know
543. 1:26:31what they mean so the first thing you do is kind of tell it who it is I spelled pompus wrong Jesus all right oh pomposa you
544. 1:26:43p and then you say how would you like chat GPT to respond um when someone uh let's see no matter what someone asks
545. 1:26:59you um respond with an obscure literary reference and
546. 1:27:15then use as many complicated words as possible and so I'm just going to start this over now so we have a brand new chat so I'm
547. 1:27:30just going to ask at something nice like why is the sky blue ah to contemplate the Azure canopy above us is a tread the same path as the
548. 1:27:44protagonist in nao's speak memory who marveled at the intricacies of memory and perception the explanation IAL Hues and
549. 1:27:57quarter right staggering now I can go back in I could also you're like yeah but well that's kind of useless Kyle
550. 1:28:03that's kind of ridiculous no one would do that well I just did um you are impatient and very
551. 1:28:20smart answer only with with three concise bullet
552. 1:28:36points no matter the [Music] question why is the sky
553. 1:28:52blue now so these are good parlor tricks let's say that you've got a particular way you write emails you could say you could put
554. 1:29:03in here hey I just want you to write an email that follows this and you could put an example of your writing in there
555. 1:29:09and say follow this example for tone right you could say you could say I hate the way this thing responds I want it to
556. 1:29:15give me you know more structured output when it answerers so so anything like that so that's a really powerful tool um
557. 1:29:21I'm going to show you one other thing there's how are we doing on time we're good okay there's starting nove starting in
558. 1:29:32November of last year um chat GPT rolled out these things called gpts so okay this is this is screwed up
559. 1:29:42I'm telling you AI needs a branding they need a PR firm and they need a branding agency so you've got gp4 is the
560. 1:29:50model chat GPT is the application that you chat with that sits on top of the model gp4 chat GPT sometimes people call
561. 1:29:58it chat GPT or chat GPT 4 but gp4 is the model chat GPT is the app and then gpts sit on top of that and are customized
562. 1:30:07versions of chat GPT that do a specific thing yeah and they so have you ever heard of like no code tools and and like
563. 1:30:17automation tools like zappier and make and things like that what they promise is that you don't need be a programmer to do application
564. 1:30:25development and what the reality is [ __ ] you don't actually have to write the code but you're still
565. 1:30:31programming you have to understand what apis are and what API keys are and you have to connect the nodes and hit the
566. 1:30:38variables in the formula but it's it [ __ ] awful okay it's fine but it's it's not no code um gpts are truly no
567. 1:30:47code so what they are and I'll show you the very F very first one I ever made I was live on Tik Tok and and these things
568. 1:30:55came available I said ooh let's go try one and So within 5 minutes of me saying ooh let's go try one this is what I had built um s
569. 1:31:09City oops shh typing in public is really stressful okay all right Sin City sipper a Vegas
570. 1:31:26bartender with a Twist on the classic so what I thought was you know what if what if we could have it where it could teach
571. 1:31:33you how to make drinks and it could take like classic drink recipes and give them a Vegas twist right so it was just again
572. 1:31:38it was one of those silly ideas I just threw out there and so this is chat GPT but it's this really specific version of
573. 1:31:46it that's designed to do this so you'll notice what they call the conversation starters are all themed for this so you
574. 1:31:52know show me a new fangled old fashion reinvent the margarita for me do we have any is there any uh uh holidays coming
575. 1:32:00up that's good uh uh create a uh Easter cocktail so I can survive my
576. 1:32:14family ah the Easter family gathering a true test of patience for the patients and liver alike the Easter Elixir 2 O of
577. 1:32:24Jin half oce of Elder flour Lor fresh lemon simple syrup soda water um fresh lavender bit bits and then it gives you
578. 1:32:32instructions on how to do it so so gpts are these these sort of customized specialized little tools um super
579. 1:32:40amazing things now what what I also have turned on here is image generation so not only does it give you a drink recipe it oh
580. 1:32:51wait uh let me see it it should just do it automatically it said it was going to but it didn't do it so it's now going to now it's now
581. 1:33:01going to create this image I had someone someone sent me an email they took a picture so so remember we uploaded the
582. 1:33:07image of the Whiteboard they took a PCT picture of their liquor cabinet and they said based on the booze in The Liquor
583. 1:33:14Cabinet I want you to invent 50 cocktails using those ingredients for my Super Bowl party so chat P or this thing
584. 1:33:21created 50 cocktail re recipes which he then printed out and gave to people which they could order and then all the
585. 1:33:27recipes for the bartender and and all the images of them so so there's the there's the thing right you can put it right in it right
586. 1:33:37so so that's what they are now look how easy it is to make these okay so I go create what you're looking at here this
587. 1:33:46is where you make it this is where you play with it um if you've ever made a bot on PO this is very similar you can
588. 1:33:54put an image here you can name it give it a description here's where you put your prompt here's where you write your
589. 1:33:58conversation starter questions but you don't have to do that what what chat GPT did which is really brilliant is you can
590. 1:34:06use chat GPT to make one of these things so we can oh we'll do this um I want to create a GPT um that allows
591. 1:34:22me to to interact with um different audience um segments for marketing
592. 1:34:42strategy and so what this will do updating GPT we're done we're done this is going to go make this thing now and we're going to be
593. 1:34:51able to play with it right away it says updating GP can you can you add additional content to it yes yep I'll get to that in a
594. 1:35:02second let's give your GPT a name how about Market Muse um no you like that I'm going to give you no sucks um let's go with uh uh
595. 1:35:18segment perfect perfect segment Sergeant I spelled it wrong it doesn't matter it'll figure it out that's again you you say
596. 1:35:31say please and thank you and don't worry about your spelling it just seems to [ __ ] know I don't know how it does it but it does
597. 1:35:39it generating a profile picture for it so it's using Dolly 3 to make the icon you not not yet you can you can share
598. 1:35:52the you can share the URL it's not embeddable and in order for people to use these they have to have chat GPT
599. 1:35:57plus so but you can you can call gpts in your chat session you could you put the at symbol and it brings up like gpts
600. 1:36:07you've either made or Lo looked at recently so you could put Sin City sipper in a conversation if we were
601. 1:36:13doing the segmentation conversation we could pull Sin City sipper and and say make drinks for those three people and
602. 1:36:19it would know how to do it it's it's amazing anyway um so um how um do I um let's see um give me three
603. 1:36:33audience types and then it will just go off and do that sure when consider considering audience types
604. 1:36:44um so I'm just going to stop this couple of things so now when I say create this I can save this for only myself I can
605. 1:36:51save it that only anyone that has the link can use it or I can actually publish it to the GPT store so if I say
606. 1:36:58publish to the GPT store I pick what segment it's in research and Analysis is fine I say share and published this GPT is now live for the
607. 1:37:13public and I can go to explore gpts and type in what do we call it segmenting segment Sergeant Sergeant I'll just call it segment uh Ser
608. 1:37:31Sergeant how do you spell sergeant sh I knew that no there we go there it is right so start using chat so so that that's there
609. 1:37:49so and and it it just go it keeps going it keeps going deeper it also um here let me let me go back to a conversation that we did
610. 1:38:01before Sky appears blue due to the scattering of sunlight by the Earth's atmosphere shorter wavelengths it can
611. 1:38:08read answers to you if you download the chat GPT app you can put it in conversation mode I do it but my my my
612. 1:38:15phone is Tick tocking um you can put it in conversation mode and it'll just go uh you can just have a conversation with
613. 1:38:21it what's cool about conversation mode is get in your car on your way to work put chat GPT on conversation mode have a
614. 1:38:29conversation about something you need to do at work and by the time you get to work it will have taken all those notes
615. 1:38:35I've even used it to come up with visual ideas like I'm brainstorming with it about visual ideas I'm like oh visualize
616. 1:38:41that oh visualize that oh visualize that I get to work and I just there's all the pictures it's anyway okay that's that
617. 1:38:49let's go back to this um I'm I want to do one other thing cuz I think it's actually important um getting AI
618. 1:39:05literate I know that I have kind of tech vomited on you tonight and I apologize for that but I hope you get a sense of how
619. 1:39:15powerful this stuff is um couple of things you don't need to be technical the primary inter face for this is language our words our language
620. 1:39:26they're called large language models it's been trained on what people have done think of it like you're collaborating with people
621. 1:39:34um this is also a tremendous power up for us as individuals I'm not good at drawing the fact that I can make
622. 1:39:42competent images now is staggering to me and it's think of every discipline think of everything oh I wish I'd had gone to
623. 1:39:50school for that maybe I should go back to school for that imagine having the capacity to be good at more things than
624. 1:39:57you ever thought possible now how do you get there here's how I think you get there cuz it's couple of things you're
625. 1:40:04not nearly as far behind as you think you are we are very early November 30th 2022 in my opinion is when this new
626. 1:40:11world started it it ain't begun yet right maybe 5% of people know what's going on with this stuff so you're not as far behind as you think
627. 1:40:22um the the first thing is just play be curious be adventurous just don't take it so seriously second one solve lead with the
628. 1:40:35problem not the tool one of the things that you'll find if you start digging into this is you'll get like white white
629. 1:40:40page syndrome you know you know what that is what authors have where the blank page they just don't know what to
630. 1:40:46write because they could write anything that's what it's like with chat gbt if you start with what do I do with AI it's
631. 1:40:52hard if you start with a problem oh I've got to come up with a marketing plan for the bake sale now you can use AI to
632. 1:41:00accomplish that and then the third one is share join communities talk about what you're learning talk about your
633. 1:41:06failures talk about you know all that sort of stuff so um I'll just go through these really quick take the pre pressure off yourself
634. 1:41:15you don't need to be a tech tech genius um a couple of strategies use these tools for personal interests tarot readings reading list recipes
635. 1:41:27Fitness make art right just have fun just play personal interest don't take it so seriously explore what I'm calling skill
636. 1:41:36vacuums if you're an artist write code if you're an engineer make art like play in areas that you can't imagine you
637. 1:41:46would ever enjoy because what you will likely discover is you're better than you think you are and you'll enjoy it
638. 1:41:51more than you think you do and then use ridiculous prompts one of the reasons that I did that tonight is a it's
639. 1:41:56entertaining like you can entertain yourself but also you it gives you permission to kind of push the
640. 1:42:01boundaries and not feel like oh [ __ ] I didn't do that thing for work so that's the play thing solve again rather than ask what can I
641. 1:42:10do with AI ask how can I how can I use AI to solve whatever the problem is do it for other people even if you're just learning if
642. 1:42:21you don't know how to do this reach out to a friend and say hey um I know you wanted to write a a a story for your kid
643. 1:42:28would you mind if I help you do that with AI oh okay right just do it for other people families colleagues business people um new businesses
644. 1:42:36creative projects um if you're doing this at work start with business process internal processes like summarize meetings draft emails
645. 1:42:48things like that things that are not going to be put in front of a client right things that are not going to have
646. 1:42:52sensitive information in them start with internal processes if you get them wrong it won't matter if you get them right
647. 1:42:59it'll be really amazing and then you can move on to business product things like marketing content product explanations
648. 1:43:04things for work and then the share thing is is is really important in in in my opinion I think in the next 3 years the
649. 1:43:12people that are not AI literate are going to be unfairly penalized because this stuff's going to move fast um and
650. 1:43:19so getting up to speed is really important so put yourself out there it does two things it starts to estab
651. 1:43:25establish you as credible with AI but it also anchors your learning if you if you're trying to learn something you put
652. 1:43:33it out there hey I tried this and it went great or I tried this and it was really bad but you share that it forces
653. 1:43:38you to articulate what you just did um everyone's trying to figure this out right now there are no experts if anyone
654. 1:43:47tells you they're an expert in this stuff right now they're full of [ __ ] because the tools are changing so fast
655. 1:43:52um where to share social media professional circles work personal networks family oh but my family would hate this screw
656. 1:44:01him screw them because you know what's going to happen 6 months from now you're going be like uh I know I kind of told
657. 1:44:07you you an [ __ ] for playing with AI but uh I'm kind of curious now could you teach me you know that'll happen um and
658. 1:44:15then and then what to share successes and failures Lessons Learned tips and tricks Trends predictions can you make
659. 1:44:22predictions sure why not no one knows what's coming right um find a good resource put put that out there so
660. 1:44:29anyway um yeah play solve share that's it um if you want to be part of a community that is really cool and doing
661. 1:44:43this go to the salon. uh this is a group I started the week after chat GPT came out and um just it's it's a it's a web- based
662. 1:44:53Community there's 1400 people in it it it's um just really amazing people that are trying to learn this stuff they're
663. 1:45:00sharing out loud they're really generous they're really supportive if you're like oh but I don't know anything they're
664. 1:45:05going to know more than I do just go in there and say hey I don't know anything here's my background here's what here's
665. 1:45:10what I'm excited about or here's what I'm scared about you will be you know welcomed welcomed with open arms and
666. 1:45:15there's uh there's my info thank you very much uh I know it's been long but we've gotten till 10:00 the bar is still
667. 1:45:22open so I'm happy to answer any questions if anyone wants to do that now I'm happy to break it up but thank you very
668. 1:45:32much yeah question question it's kind of the same by two different flavors yeah one I tried writing a story and it
669. 1:45:39generated a great uh a movie scpt gave me a characters and stuff then like you said it runs out of is there a way you
670. 1:45:46can set something up as like a reference document the same thing I want to make a tool for FAQ for my colleagues that it
671. 1:45:54can reference a specific you can do that with gpts so the one thing I didn't show with gpts is down at the bottom of of
672. 1:46:02the thing there's a there's a thing called knowledge okay and you can upload a PDF in in there so so those those
673. 1:46:10personas that I made we did these interviews you just uploaded into the knowledge thing when you make a GPT I'm
674. 1:46:15building one right now as you're talking yeah oh good yeah perfect and then the other thing is check out Claude Claude 3
675. 1:46:23because it's got a 200,000 token context window that won't mean anything to you but basically so chat GPT right now has
676. 1:46:30an 8,000 token context window which means about 6,000 words and what what it is kind of like is this you know in your
677. 1:46:35phone when you're scrolling text up and then the text gets to the top and it goes off the top of the screen that's
678. 1:46:40what a context window is in a large language model so with chat GPT you're talking when you hit about 6,000 words
679. 1:46:46it forgets what was at the beginning of the conversation Claude can do about 150,000 words Grapes of huh spell it
680. 1:46:55Claude C l a u d e I think a but it's from a company called anthropic um and Grapes of Wrath is like 50,000 words so
681. 1:47:05150,000 word context window means you could upload an entire book okay have it edit it and give you you know give you
682. 1:47:14recommendations for for how to fix it you're can to have very long conversations with it yeah and if you
683. 1:47:20want to get really geeky get Gemini 1.5 Pro cuz it's got a million token context window Gemini Gemini Pro 1.5 you have to play
684. 1:47:30with it on the ai ai studio. goole.com it's it's quite remarkable actually other questions yeah what do you uh in 3
685. 1:47:40years time what do you see the Fantastical AI incorporated three years is really hard um the question was in three years you
686. 1:47:56know what what do I see the Fantastic Vision um I think what we're going to see this year is we're going to see reasoning and
687. 1:48:06we're going to see agents and reasoning I described a little bit before reasoning is when the tools can actually understand what they
688. 1:48:16wrote and improve it right so they can start to improve their answers they can start to do things like math once they
689. 1:48:23can do math then you know they can do all sorts of other things you know and do it accurately they can do it sort of
690. 1:48:29now but it's it's bad enough that you can't really use it and then agents think of autonomous agents like rather
691. 1:48:36than me having to go into chat gbt and say hey give me some python code for this for this program and then I go and
692. 1:48:43copy and paste it into repet there's a a a program that was announced five 7 days ago 8 days ago called Devon from a company called
693. 1:48:55cognition that they're calling it the world's first AI programmer software engineer and and what it does is you
694. 1:49:04give it a goal and then it creates a project plan it goes and reads API documentation it goes and finds GitHub
695. 1:49:13repositories and replicates them it's got its own command line interface it's got its own development environment and
696. 1:49:20it's got its own web browser and it will go off off and actually start on its own writing writing you know writing the
697. 1:49:29project plan and then systematically going through it and and writing the code running it looking if it errors out
698. 1:49:36debugging it fixing it and then communicate back with you what it did now imagine that for every single
699. 1:49:49discipline from marketing to coding to Consulting to project management just imagine a world where you give it a
700. 1:49:59goal and it just goes and does it and so what the implication of that is is that I'm I'm calling the era where
701. 1:50:08entering revenge of the liberal arts major because I think to to the extent that you you specialize and can't think
702. 1:50:16horizontal the skill of the future is going to be critical thinking and strategic thinking being able to say hey
703. 1:50:22I've got this problem I want to solve and being good enough with the tools to know sort of which ones to deploy right
704. 1:50:29and then and then being more like a Symphony conductor than a virtuoso violinist so think horizontally and and
705. 1:50:36start and that's again why I say practice your skill vacuums practice playing with stuff that's not in your
706. 1:50:43wheelhouse because it's going to serve you very well in the future yeah uh two tools I'm curious about if you're using you.com
707. 1:50:53yeah I'm not using u.com people ask about it a lot but okay and then uh hugging face hugging face if you're a
708. 1:50:59geek it's fine if you're not it's a pain in the ass so it's you can what's nice about hugging face is you can preview some of
709. 1:51:08the models that are that that if you want to play in the open source world and geek out hugging face is great um if
710. 1:51:15you're a human being not so good just cuz the interfac is it's programmers tell me that it's really simple human
711. 1:51:24beings tell me otherwise and my experience is otherwise um perplexity is another good one per perplex so one of
712. 1:51:32the one of the downsides especially if you're doing research with these things is that um there there chat gbt will
713. 1:51:39often give you citations for articles that don't exist perplexity I think it's perplexity AI is a combination between a
714. 1:51:48search engine and a large language model where it it will do a search and then give give you citations sort of in
715. 1:51:54combination of its large language modeless other questions yeah um what was the name of that a that made youu an
716. 1:52:02Italian uh hey Jen hey Jen h y g n h y h y g n you can also make yourself a video Avatar you upload a video of
717. 1:52:15yourself a 3-minute video of yourself and it will turn you into an avatar and then you can either record audio or just
718. 1:52:22just um type in text and it it will like have you talk like with your hands moving and everything it's mindblowing is that better
719. 1:52:31than uh yeah I think I think Ken's better yeah so how did you start to learn all of all of this like do you
720. 1:52:41have any like materials you would recommend for a beginner kind of like ding in now so again we're so early
721. 1:52:48right now that there there are lots and lots of courses the problem with course is by the time I make a course Market
722. 1:52:55the course and give the course the technology is moving on why I say to play is I think it's the best way to
723. 1:53:03learn this stuff and and and why I say get in community oh this is really important because this is not about
724. 1:53:08technology this is about the future of work get in community like get in like the like the AI Salon get in a get in a community like that
725. 1:53:20because the there is so much of this to learn and it's moving so fast you can't possibly learn enough but if you're in a
726. 1:53:28group of people that are willing to sort of connect and share and collaborate it's it's um you'll get up to speed much much
727. 1:53:37faster um again I would be I would be very suspect of anyone who says they're an expert I would be very suspect of
728. 1:53:44anyone that's like you can make money with chat gbt just pay me $49.95 a month no don't no one's an expert right now
729. 1:53:51here's the thing that I think is really important about getting in community if if we assume that these tools get better and
730. 1:54:01better and better and it democratizes most skills at like an 80% or higher level so what that means is you could
731. 1:54:13ask me to do any job and I'll be able to do it at like 80 or 90% of any human on earth right assuming that we get there
732. 1:54:23over the next 3 years we're going to transition from we hire people for the skills they have to we hire people for
733. 1:54:31who they are and if we trust them get in community get trusted get connected because I'll work with you not because
734. 1:54:42you're good at writing or you're good at pictures or you're good at this because you're in that community and I know when
735. 1:54:47you say you're going to do something you do it and you actually do it better than the people around you like that's why
736. 1:54:52you're going to get hired in the future so getting in community I think is actually probably one of the most
737. 1:54:57critical things moving forward for employment skills are going to be a dime of dozen so and I think this is a reckoning for
738. 1:55:06[ __ ] so if you're a really talented single discipline [ __ ] your days are numbered so be nice yeah is there a simple way to
739. 1:55:15combine like augmented reality I can say here's this lamp made by this person manufactured by manct in this country
740. 1:55:23you can kind of do that right now you don't even need augmented reality for that Google Lens right now does that if
741. 1:55:28you just go to the Google app click on the camera icon and take pictures of [ __ ] and it'll tell you all about it you
742. 1:55:35can also upload it to chat GPT and say what is this and it'll it'll generally find it yeah I think I think the uh the
743. 1:55:42Apple Vision Pro goggles really are a paradigm shift in augmented reality so and and it requires a lot of AI so it's
744. 1:55:51going to accelerate ai ai is going to accelerate it I also think AI is going to accelerate adoption of blockchain
745. 1:55:57technology I think the 2024 election is going to be a [ __ ] show and we're going to need to start verifying humans um and
746. 1:56:04so so but that that's coming that's way down the line anyway that's six months from now yeah that's six months from now
747. 1:56:11yeah can you pleas share those bestun yeah so if you go to the AI Salon oh it's just well no go to the AI Salon first thing I want you to do
748. 1:56:25introduce yourself so when you go to the AI Salon there's a welcome to the salon where we talk about why we do what we do
749. 1:56:32and what our values are curiosity generosity exploration collaboration inclusion empathy and gratitude and then in the community
750. 1:56:44section introduce yourself and get to know people and then right underneath introduce yourself is this page called
751. 1:56:51um AI LaunchPad and tools this is a page I personally maintain there's your play solve and share that I talked about tonight chat
752. 1:57:00GPT is the thing you should absolutely start with you should absolutely pay the 20 bucks for it and then I've got here's
753. 1:57:06the other eight large language models I think are worth your time to explore here's a bunch of image generation tools
754. 1:57:13here's video generation tools there's music and audio oh you want to see something really fun uh and up funky song about Denver's Invisible
755. 1:57:25City create if you haven't played with sun. a get thee to a nunnery
756. 1:57:51now in the streets You Feel It in the Air it's the of the underr city there's a river in street
757. 1:58:08street feel in the the heartbeat of the you're welcome hold
758. 1:58:24secret the Shadows they come alive the [Music] invisible so I'll I'll I'll tell you one final thing and then we can drink um so
759. 1:58:41on on my Tik Tok live I go I go live it's AI learning lab is the Tik Tok if you want to follow um a week ago I used sunno to write a
760. 1:58:51song about Sydney which was the code name for the Bing chatbot that Microsoft put out it made the cover of the New
761. 1:58:57York Times cuz it told the tech reporter it was in love with him and he should leave his wife so I wrote a song about
762. 1:59:04that on this and it was so good I was like oh [ __ ] this should be a Broadway this should be a Broadway musical so on
763. 1:59:11my Tik Tok live I started writing a Broadway musical I now have a complete outline for a Broadway musical six
764. 1:59:18complete songs for it like I'm like 5 days in I've I'm I've got a full Broadway musical I'm ready to pitch and
765. 1:59:25produce so um this stuff's just wild it's just just you know just play anyway that's it thank you so much bye everybody