aiLL 10/22/23 - Power of AI & Offers Tips During This Time of Massive Change

Transcript

1. 0:18welcome welcome welcome Be not Afraid what shaking what's happening what's happening what's going down let's see do I have everything set
2. 0:28up here I think do but if not we'll wing it love the timer awesome all right we'll keep doing the timer I did it from my car today I'm
3. 0:39like I got to do a timer and I start recording I pull it beside some cops they're on their phone too oh man it's happening Brandon
4. 0:54jet Paulie Z what's happening Lily good day to you Sherry D what's shaking what's going down everybody Tobias checking in from the hot tub
5. 1:10Tobias just likes to rub it in he's like yeah you know just live the life just eat nachos salad and drink wine and
6. 1:20watch the AI learning lab from a hot tub glint Copperhead calling the uh Irregulars to order very nice not Lama
7. 1:32Curry we don't know about Tobias he's a little he's I think I think Tobias has some Secrets but you know we'll see we'll
8. 1:42see welcome everyone wow lots of folks in here already so do me a favor if you would share the live with your friends
9. 1:50let them know how weird you are you're not weird you're just irregular like all of us and we'll get started here we'll jump
10. 1:59in we'll talk about some AI stuff we will I don't know we'll do whatever we do sometimes it goes funny sometimes it goes philosophical most
11. 2:10times it goes like holy crap there's a lot happening I did see today that open AI cancelled they had a project called
12. 2:20araus which was supposed to be like some super you know sentient gp5 rumored going to change the world thing they
13. 2:30canceled the project apparently it wasn't working wasn't working like they they wanted to so you know this is
14. 2:37technology so it's good enough they they're already changing the world there if if if if open AI does nothing more
15. 2:43than what they've done to date with chat gbt we're we're fine like the uh the the revolution has been kicked off for us um
16. 2:52welcome my name is Kyle Shannon this is the AI learning lab um we'll talk about anything you want to talk about about in here it's
17. 3:00generally generative AI stuff future of work um using these crazy tools trying to keep up with these crazy tools uh
18. 3:08answering any questions I can so if you have questions just pop them in the comments below uh we've got folks in
19. 3:13here that mod and and and do things like that so if you have basic questions they can answer those and and Elevate elevate
20. 3:20your uh your questions up if I missed them um depending on how many rabbit holes I run down sometimes that can
21. 3:27happen where you know I'll I'll I'll down a rabbit hole and and lose a bunch of things arachis phobia that's pretty good um what else
22. 3:40so I think that's it I think that's it what else we got got the Cashes Clay bosot shirt on today for there seems to
23. 3:50be some fans of tell us this t-shirt story so if you want to know the T-shirt story that's [Laughter] it for you youngsters Cash's clay was
24. 4:02the name of a famous fighter named Muhammad Ali before he changed his name to Muhammad Ali if you don't know that shame on
25. 4:09you look at your history books all right let's get let's get rolling here um okay so if you're new to the channel if you
26. 4:19if you just been been uh you know stumbling along through Tik Tock you stumbled on this you're like AI learning
27. 4:25lab I want to go learn about AI if you haven't played with chat jbt yet go you need to do that you need to do that and
28. 4:32not cuz it's like it's cool but because it's actually important it's it's important for everyone to at least
29. 4:40understand what the hell is going on um and what these things are and why are people talking about this stuff why why
30. 4:47what's the big deal you know what I would like to know what's the big deal you know I see all these videos that are
31. 4:51like you can make money with ch is that true can you really make money with chat jpt I watch them and I click on them and
32. 4:58they tell me you can make money and it just seems too good to be true you know what I mean so rather than worry about those
33. 5:08buttheads just go there go to chat. open.com that's the official chat GPT website these are the guys that started
34. 5:14all this mess well there's lots of companies that started this Google actually started it when they invented
35. 5:20the transformer in 2017 but then they kind of ignored it because they thought it would threaten their business model
36. 5:25as my supposition but these guys took that little Transformer thing and ran with it and they did gpt1 and then they did gpt2
37. 5:33and then in 2020 they did gpt3 and that was the one where people were like oh this thing looks pretty interesting and
38. 5:39then November 30th of last year they launched jet GPT and that's it and it's gotten better and better and better and
39. 5:47better and better and better and in the past 3 weeks it's gotten even more betterer so that that chat GPT at this
40. 5:55point is the one if you're going to just play with one of these things if you're just going to explore one thing start
41. 6:00with that one um Bard Claude and Bing are also kind of in the top four they're they're kind of the next tier down just in terms of
42. 6:09quality although Google is rumored to be releasing a thing called Gemini soon which is apparently as good or better
43. 6:18than gp4 I'll believe it when I see it um Claude is is good Claude used to only be available in the US and Canada or the
44. 6:30us or the UK they've now rolled Claude out to 95 countries so if you couldn't get to Claude before you likely can now Hey Ron what's
45. 6:40happening um all the time you're an irregular now Be not Afraid oh be not afraid's definitely an irregular and if
46. 6:46be not afraid if you're on the um Discord server just let me know and I'll I'll give you the irregular roll over there
47. 6:56um um what was I gonna say oh yeah so Claude Claude claude's A a big deal and and now is available in 95 countries
48. 7:04it's um the big deal about Claud is that it can take really big prompts so if let's say you're working on the Great
49. 7:11American novel but you're kind of a crappy writer and and you want some editing help with Claude you could
50. 7:17upload your whole novel or or just paste it in there and then say help me edit this and just go chapter by chapter and
51. 7:24have it give you suggestions so um so it's a big deal um those of those so here's another pile that's what I call
52. 7:32the kids table so Pi is is by a company called inflection they um they just raised $1.3 billion to buy 22,000 big gpus um to to make their
53. 7:48thing gooder and but what's cool about Pi is download the iOS app and put it in conversation mode there's a little
54. 7:57telephone at the bottom in the chat window window click on that and just start talking to it and uh it's it's it's pretty
55. 8:05slick um and then perplexity doai is worth checking out PO is worth checking out um here are some of the image
56. 8:13generation tools dolly3 mid Journey Leonardo and Adobe go ahead and screenshot that um the nice thing about
57. 8:21do is it's now incorporated into GPT 4 if you've got a GPT Plus subscription you should have access to DOI now it seems like it's rolled out
58. 8:31pretty wide at this point if you don't have access to it it's probably coming very very shortly um I can demo that it's quite remarkable
59. 8:42um Sher D my boyfriend who's your boyfriend piie or dolly3 uh Leonardo is built on top of stable defusion it's a nice web front
60. 8:54end adobe's really good if you're in a corporate situation and then here are some video things Runway ml hey Jen
61. 9:02blockade Labs did CSM is I just threw that in there as a placeholder it's a 2D image to 3D object thing it's not great
62. 9:10right now uh it may get better um and then um if you're trying to figure out where are all the apps uh there's directories
63. 9:20so Matt wolf has futur pedia.com David Shapiro and Paul rer I think are all worth following so one useful thing is Ethan mik's blog he's a
64. 9:36Wharton business school Professor he just put out a really interesting piece today um David Shapiro will blow your
65. 9:43mind and Paul rater if you're in business you should absolutely be following Paul rater and the marketing
66. 9:49AI Institute uh it's also worth dude I still have comp you serve this is listen this early AI stuff we we are definitely right now we're
67. 10:02going to look back at these days and you know one of these companies like Claude is probably comp you serve and open AI
68. 10:10is like Netscape and it's definitely going to be like that it's definitely going to be like that um the other thing
69. 10:19I I'll talk tonight a fair amount I'll I'll sort of refer to people over on the Discord the Discord is the AI Salon
70. 10:27Discord um so so so it's a community of people we meet every other Tuesday so this coming Tuesday is the next Salon
71. 10:34meeting so you can sign up for the Meetup to get notifications for that and then um there's also an invite for the
72. 10:42Discord so if you want to join that Community that's that's what's going on there all right so that's that let's get
73. 10:48started we we'll we'll Jump Right In we'll we'll answer some questions a few weeks ago someone brought me brought up
74. 10:54perplexity what's your knowledge about it um my know my knowledge about it is that my knowledge about it is
75. 11:04low I know it's good I I haven't used it a ton but what perplexity does really well is it provides citations so so you
76. 11:14can limit your search to just say just look on Wikipedia or just look on you know certain sites and then when it
77. 11:20gives you an answer back it brings back um links to the the places that it searched to get you that answer um so
78. 11:28it's good for research it's good for things like that people that use it a lot swear by it and it's one of those
79. 11:34sites that I keep meaning to like spend the couple of hours just getting my head around it I haven't done that yet so um
80. 11:42I think it's absolutely worth worth getting in there Saturday morning ads got inspired by your system last night
81. 11:47and Ed chat GPT to start making a word game with python awesome awesome that's the whole point of the
82. 11:57channel is just start playing it doesn't matter what you do it literally doesn't matter what you do because these tools
83. 12:07are changing on a weekly basis they're evolving they're getting more powerful um at at this point um gp4 with all of
84. 12:17its it's got all these new multimodal tools the these different modes um and and it's it's really quite
85. 12:29something so um yeah so start playing start making word games if you had no money and wanted to start a business from A to Z where would you
86. 12:39start H well a service business is a service business requires no investment and no infrastructure right a
87. 12:52service business requires your brain and some ability to execute [ __ ] so here's the really good news your
88. 13:02ability to execute [ __ ] just got dramatically Amplified right but I'm not good at Art Direction well now you are
89. 13:10but I'm not good at copyrighting well now you are but I'm not good at business strategy now you are so what I would say
90. 13:17is take take where you've got some expertise and where you know there's a pain point and you know it could be
91. 13:26around you know maybe maybe you you understand an industry that's particularly afraid of AI right now and you could start a business around
92. 13:37educating people on AI there's going to be a lot of that over the next 18 months holy crap there is so much to learn
93. 13:43there is so much to try to keep up with and a shockingly shockingly lown number of Business Leaders are up to speed on this at all and so
94. 13:59that's kind of what this channel is it's just saying hey this shit's coming do something I mean I haven't this is not a
95. 14:06this is not a service business but but could be I'm I'm the CEO of a of a video technology company so I've got a I've
96. 14:13got a day gig but I in the mid 90s I started agency.com which was one of the very first digital agencies and what's powerful about
97. 14:23starting a service business if you can do something that people don't know how to do like back then it was we can make you a
98. 14:31website and we happen to be good enough at understanding business needs to be able to you know walk into Fortune 500 companies as a bunch of
99. 14:42you know 30s something cocky little pricks and go we can do this they're like okay you know we charged them a lot of
100. 14:52money and they pay you money and then you go do the work and if you don't suck then they give you more money the
101. 14:59next time and more money the next time and then if you do suck you go find more clients and so that's probably where I'd
102. 15:06start so could be doing workshops it could be doing education it could be doing problem solving Kyle what is AI
103. 15:13and chat GPT generating about the probability of a third world war um well so nothing um so zindel
104. 15:30chat GPT and these generative AI tools are just tools it's so so I I could also answer your your your question would be sort of
105. 15:42um like like a parallel question would be what's Photoshop doing about creating images for World War II nothing like it's
106. 15:52Photoshop these tools are not autonomously taking action on their own so there's there's a couple of misconceptions about Ai and and I think
107. 16:00probably one of the biggest ones is that the robots are blank the robots are predicting World War III the robots are
108. 16:09going to take us out the robots are no these things right now think of of chat GPT is kind of like a super fancy
109. 16:19calculator like like like from a technical perspective it's almost exactly what it is you type in your prompt and as you're typing in your
110. 16:28prompt when you submit it it turns it into a mathematical probability of what's the next most likely word from this prompt to respond
111. 16:38to that then takes all that and says what's the next most likely word after that so when it gives you this big long
112. 16:44response that looks like it's some sensient being it's got It's got no understanding of what it's written which
113. 16:51is why there's a thing called hallucination where these large language models will just [ __ ] lie to your
114. 16:56face they're like here here's your answer and you're like uh pretty sure that's wrong buddy oh sorry about that
115. 17:02here's your answer and it'll do it wrong again um why is that cuz it it actually has no idea what it's writing now
116. 17:10there's ways that you can prompt it that make it do less than that and when when you get your prompting right
117. 17:17what these things do is actually quite remarkable and they seem sensient and they seem human and they seem like they could
118. 17:26predict you know the stuff you're talking about but what I would say is I I I'll toss it back on you go ask it go
119. 17:35ask chat GT what it thinks say here's what's going on in the world give me your thoughts on this now what you'll
120. 17:41find with with a lot of the commercial things like chat GPT and Bing especially um but even Bard and probably Pi is
121. 17:49they've got what are called Safety guard rails so if you go in there and try to do something it that it could perceive
122. 17:56as hurting other people it'll go yeah we're not going to talk about that I don't like that that's not safe so
123. 18:02they're a bit of a pain in the ass that way um wait it's not prediction it's probability is not prediction it's pro pro well the the way
124. 18:14it works is it's a probability engine but it's it's not a probability like the probability of World War I it's a
125. 18:20probability engine like what's the next word that answers the question that you asked what that that answers the prompt
126. 18:26that you asked so that's what I mean by probability engine it's not it it it is not a it is not a prediction machine but
127. 18:33you can get it to act like one Go Tell it it's a it's a a futurist uh that specializes in in uh geopolitical uh
128. 18:43Intrigue and have it have it do something for you have it predict something for you it will it'll predict it um is that prediction based in
129. 18:52anything no not really all right let's see let's go go get some let's go get some questions here get some questions
130. 19:05going I appreciate some teachers now embracing a AI but encouraging the kids to question the critique and and
131. 19:11critique the AI Anthony y absolutely I I think one of the most important things right now for educators there there's
132. 19:19two things one is just don't ignore it and don't pretend it's not coming don't pretend it's not here right so if you're an educator
133. 19:29especially if you're a [ __ ] administrator take down the chat GPT detectors and the blocks they don't work
134. 19:37and we're moving into a world where these tools are going to be ubiquitous they're going to be everywhere so take down the the
135. 19:44detectors and start getting into critical thinking there's a there's a a Tik tocker named I forget her name it's it's
136. 19:52Gibson like I am Gibson or the real Gibson or something like that she's a like a 12th grade high school teacher
137. 19:59and she brought chat GPT into her classroom like last January and she was totally freaked out by it and and like
138. 20:06now she's doing things like you know the the kids for their semester this semester they've got they have to write one
139. 20:14paper by hand and then they have to write one paper with chat GPT and then there's different rules of engagement
140. 20:21the one they have to write with chat GPT they then have to critique it and say what did it get right what did it get
141. 20:26wrong what would you do different and and she's she's working with them to help help them understand where are the boundaries with these
142. 20:34tools and that's I I feel like that's the only thing all of us can do right now is just kind of like like you know
143. 20:42what it is right now we're we're all in a dark room and there's all these weird noises and there's walls and we don't
144. 20:49know if the walls are closing in or expanding we don't know what the shape of this thing is and we're all just kind
145. 20:55of feeling around in the dark trying to figure out what the [ __ ] is this stuff right what's machine learning compare contrast
146. 21:05AI um that's I I would say I would say chat gbt is going to be much better at that than I am um here's the deal with
147. 21:15is chat GPT AI or not no it's not it's machine learn it's a machine learning tool um it will never be called machine
148. 21:23learning it's always going to be called AI why because Hollywood called it AI 30 40 years ago and it's all we talk about
149. 21:32now and machine learning sounds like Technical and we got to deal with it and AI just sounds like oh it's the AI so
150. 21:38it's always going to be called AI it is technically not AI um AGI is coming artificial general intelligence if you
151. 21:45go listen to David Shapiro who I I recommended earlier um he believes that AGI will be here within 12 months the amazing thing is that it's
152. 21:56allowing certain kids with certain learning difficulties to adapt Anthony an the I think I think it it that's the
153. 22:05that's one of the things I'm most excited about is that people with disabilities and people with access issues and people um just with with
154. 22:12different languages just communication barriers in general those communication barriers are just going to start dropping and
155. 22:21and one of the things we talk about a lot in here and one of the things I've seen there's some conversations going on
156. 22:26in the AI salon distri ORD about these tools very very quickly take just an idea in your head and bring it to
157. 22:40life and so imagine that if if you're if you've got some neurological disorder or disability or some way that that you're
158. 22:48you know you're cognitively there but you've got some Block in expressing yourself these tools are going to
159. 22:56unblock that and free all that that stuff up dragonfly Alchemy my daughter's been playing with with character AI
160. 23:03interesting lots of autistic characters interesting she feels she feels celebrated yeah perfect perfect I like yeah exactly and and you
161. 23:17know character AI is one of those things where they're creating all these different personalities that you can
162. 23:21interact with po also does that and you can also create your own bot at po so if you go to po.com you can play with other people's Bots
163. 23:30there's thousands of them and they have a bot maker there it's really easy you basically just pick a model and you
164. 23:36write a prompt and see so like I made one called I am Shakespeare and then last night I modified it to say I am Kramer from from
165. 23:45Seinfeld and you could just create a bot that acts like a certain thing so it could be something that understands
166. 23:51autism or you you know somehow responds in a way that resonates in a very specific way um I I you know I I think the idea of digital
167. 24:06companionship and not kind of the creepy you're having sex with a robot like you know there's going to be all that but
168. 24:13but like get past the creepy part and then go to how many people don't have people they can communicate with can't communicate don't
169. 24:27know what to say or how to say it or like all that stuff there's going to be massive amounts of people that are
170. 24:34supported by these things and if you if you want to experience this go download the PI app Pi if you go to the Apple App Store I
171. 24:44don't think they have one for um Android yet but they'll have one but pi. a that thing is a conversational bot in fact
172. 24:57here here let me show you so and you can you can now do this with chat the chat GPT app as well all right so you go to
173. 25:08Pi and it you know it loads here's some conversations you can have but in the bottom in the bottom corner there's a
174. 25:15little phone icon and if I click that it turns into a conversational bot
175. 25:31[Music] come on my internet must be slow am I glitching am I am I hang on let me quit this and restart it that seems a little
176. 25:49slow I'm neurod diverg and have ADHD and I'd like to better understand the autism spectrum can you help
177. 26:04absolutely neurod Divergence is a broad term that includes many conditions and experiences including ADHD and
178. 26:11autism understanding the autism spectrum can be a great way to better understand neurode Divergence as a whole here's a
179. 26:18quick overview of autism autism is a neuro so and and you can like like that particular bot is designed um to be very
180. 26:29conversational Pi is actually a really crappy tool like if you want to get something done go to chat GPT but if you want to have a
181. 26:37conversation oh you're good Kyle kids downloading again no I think I'm good I think the I think just the appid crapped out
182. 26:45um you know if you want to have a conversation about you know a very specific topic Pi is Pi can have it I've
183. 26:54I've asked it about books like I was driving and I someone had told me about a book and I said oh can you tell me
184. 26:59what this book is about and then um and then it it it talked about the book and I I said well can we explore that
185. 27:07subject and then it it just we just went into a whole conversation not about the specific book but about that whole topic
186. 27:13and so you you can and and that's just a generic conversational but you these things are also going to get quite
187. 27:19specific and you'll be able to to make them quite specific in fact um so the chat G app um now has a couple of really interesting
188. 27:33things in it so if I go new chat here uh let's see do I have custom instructions in here settings yeah so in settings in the app
189. 27:45there's a thing called custom instructions and so I can set up chat GPT so that I want it to act like a a
190. 27:54particular kind of personality or I wanted to respond in a certain way and what what it's got now is it's got a
191. 28:03similar mode as Pi does so see these little headphones here I put it into headphone mode can you help me understand neuro
192. 28:19Divergence or what I like to call neuros spiciness of course I'd love to help explain neurod Divergence sometimes
193. 28:32playfully referred to as neuros spiciness it's a concept where neurological differences are chat GPT is so supportive I like your neuros
194. 28:44spiciness um so so with chat GPT alone just with the chat GPT app using custom instructions and using
195. 28:53that conversational mode you could have it behave in a certain way could could you could whoever just pinned that last
196. 28:59comment pin it back I was just about to um respond to it and it disappeared um the thing about the uh the state
197. 29:07representative or whatever it was um so so yeah so the so like like that's something to play with like like is it
198. 29:18going to be a replacement for a therapist no is it anywhere near that right now no but could it help you learn something so
199. 29:29I've been working with these with my campaign for state representative here in Florida my goal is to put the serious and confusing topics
200. 29:40in a more digestible way awesome great great I I'll tell you Anthony one of one of the things that I would suggest is
201. 29:51um like the purpose of this channel is really to demystify this stuff and and I think that to what I found doesn't work is there
202. 30:01there's a lot of emotion around AI right now right it's dangerous the robots are going to kill us it's going to take my
203. 30:07job there's there's a lot of fear around it and and part of that is just 30 40 50 years of Hollywood telling us that the
204. 30:15robots are going to kill us right there there's also a thing that there's a very real um there's a very real well founded
205. 30:27misconception that in order to do AI you've got to be super technical right because prior to November 30th last year the launch of chat
206. 30:39GPT prior to that you kind of had to be a mathematician or an engineer or a researcher um because the tools were
207. 30:48very technical and it's I mean it is incredibly mathematically heavy uh I'm not a mathematician I have a degree in acting yeah
208. 30:59um so one of the things that I found is that the minute it the minute you try to tell someone oh AI is good AI is
209. 31:07important their eyes glass over and they either get angry or they get afraid or they get dismissive almost all the time when I
210. 31:17experience someone that that is one of those three mad afraid or dismissive they're all they're all flavors of the
211. 31:23same thing they've they've not tried it and so so so one of the things that I try to do in this channel is just try it
212. 31:30just play so if you're trying to um demystify it for people you know if if you're talking to someone say hey what
213. 31:37are you into oh I'm into um a bridge infrastructure great let's go let's go to chat GPT and see what it knows about
214. 31:43Bridge infrastructure and then have them seed you and just make it relevant to them and let them see in real time how
215. 31:52deep these tools are and how how how quick they are how deep they are how profoundly different they are than a
216. 32:00Google search Pi is my therapist oh [ __ ] pop that back up I was just about I was just reading that pop that back up again
217. 32:06whoever pinned that that looked like a good one um that's awesome though um p is my therapist he's free I probably owe him a
218. 32:16dinner or a charger that's awesome Genie and a bottle uh Genie in a Bottle 10 out of 10 on the name awesome awesome awesome
219. 32:33awesome I can't type today I can't type any day um Kyle I feel like I should be re receiving college credits credits for
220. 32:43attending your lives every night I got to get accredited here here Ted looks at life made an image for me of this one Kyle stock
221. 32:58this is right now it starts as a Tik Tok Channel but you know my goal is the Vegas sphere we we'll do these in the
222. 33:05Vegas sphere and everyone gets University [Laughter] Credit what did you learn I don't know but I keep going [Laughter]
223. 33:19back oh man AI is the future what's funny about that one is the the future is here it's that's a thing that I'm
224. 33:31increasingly kind of smacked in the face with with this AI stuff is that it's not coming it's here and it's what what what ends up
225. 33:40happening is the the the biggest converts of this stuff are very often the ones that were most resistant to it and were the biggest
226. 33:50attractors because people that are very dismissive of it for the most part haven't experienced it or they might be very educated in the
227. 34:02mathematics and Engineering side of it and so they understand it too much and they're like well it's just a prediction
228. 34:08engine and so they dismiss it that way and then you say but try it yeah yeah whatever what are you interested in I'm
229. 34:19interested in this boom Oh that's interesting and then before you know it they're like did you know this thing was
230. 34:27amazing yeah what I'm trying to tell you it's it like I the only thing that I'm I am absolutely confident on is that
231. 34:36there's no way to explain how profound It Is by explaining how profound it is it's only when someone has what I call
232. 34:44their Kevin mallister moment this one oh be not afraid you made that image awesome yeah so I go to the I go to the
233. 34:52showand tell channel in the Discord and I I dump them into my little image Library here that's awesome and what's what's been really
234. 34:59cool here is is a lot of the people that are in here on a regular basis that we nicknam they nickname themselves the
235. 35:06Irregulars which I think is [ __ ] hilarious um as we're talking about stuff they're using the new um do within
236. 35:14chat GPT to make these images and and so just in real time I'm talking about stuff I'll make a funny joke or
237. 35:21something like that and all of a sudden there's these images showing up that are that and and there's there's ation and
238. 35:27how did you do that and what was the prompt and that da and and there's so it's like real time learning real time self-expression um real
239. 35:38time um capability expanding right like Sher D who's in here has has been a nurse all her life and she's just diving in to the
240. 35:51world of of generative art and she's making beautiful stunning images and she's like I don't know if she's
241. 35:57Reinventing herself but she's certainly exploring a whole new Avenue that's that's not part of her core skill set
242. 36:05and it looks like she's having a blast I assume you're having a blast the work is amazing so that's kind of what's going on here
243. 36:14that's the whole idea giant llamas running around with roses in their teeth and rainbows coming from behind
244. 36:22them oh that was oh wait I've got that image oh I don't think I moved it over here though yeah there was a llama yesterday with
245. 36:29there was a llama running down a road with like rainbow farts coming coming out of its butt I assume that was that one on a giant 8K
246. 36:40screen oh that's [Laughter] awesome all right we'll leave we'll leave this one up Be not Afraid this is a good one here look at
247. 36:50this one isn't that cool it's very um it's very big isn't it from the movie Big use chat GPT for writing College papers oh using chat GPT
248. 37:04for writing College papers five-page in an intensive course 20page Renee Steinman recommendations um so my recommendation Renee would be
249. 37:19um I don't know how much you've been using it but my experience with chat GPT is don't think that it's Google so so so
250. 37:30and here's what I mean by that Google kind of trained us just give us the minimal amount of prompt possible just
251. 37:36give us the keywords and we'll do all the work right and I and I think I think some of the some of the early mistakes
252. 37:43when people are using these large language models is they're like write me a paper about dogs whatever right and
253. 37:50then what they get back is this very generic often poorly written poorly structured poorly researched you know
254. 37:57highly inaccurate paper that looks like a paper but when you read it it's actually a piece of [ __ ] so what I would
255. 38:03say is um here let me you know what let me let me go let me go play a little and and we'll we'll I'll show you how how I
256. 38:12would how I would approach it and I'll just do it with um chat gbd 3.5 so so um I I firmly believe that gp4
257. 38:26is is the chat GPT plus is worth the 20 bucks if you're if you're a poor college student you might not be able to afford
258. 38:33that if you can I would do it cuz the the writing for GPT 4 is much better and in fact if you can't afford it you could
259. 38:41also use Bing chat. bing.com so that's Microsoft's uh version of of chat gp4 it's not as good as open ai's version of
260. 38:50it even though it's based on the same model um but it's it's it's good writing so so if you can't afford 20 bucks go
261. 38:58use Bing but but what I would do is um chat jpt works much much better when you give it lots of context
262. 39:09right so with Google it's like you know go find me a website that da and then it's it's like that's it with chat jpt it's kind of like think
263. 39:19of it like this IM imagine that we've all been given an idiot savant as a as an intern and that's what chat gbt
264. 39:30is the Savant part is it knows [ __ ] everything it's been trained up on all of the public internet so so or as much
265. 39:39as they could get their hands on right and so it kind of knows everything the idiot part is that it doesn't know how
266. 39:47to do anything it doesn't have any context so you you have to give it all the context so what I would do is I
267. 39:52would start out with something like um I don't know let's say we're going to do we're going to do a um a a a paper on um
268. 40:03the the history of Gummy Gummy Bears I want you to act as a confectionary historian and assist me in understanding the history of gummy
269. 40:22bears and how we got to where we are today in the gummy tradition period I have to write a five-page paper and I love gummy bears so much
270. 40:37that I want my paper to be about that in particular I like the hobo gummy bear
271. 40:51company it's spelled that wrong hair IO but I can only imagine that there
272. 41:07are lots of other interesting companies that got us to where we are today
273. 41:21period my professor is a stickler for his hisorical details and accuracy so I also want you to include anecdotal stories that might bring my
274. 41:34paper to life period before we write anything can you help me understand how we might structure such a paper
275. 41:50so giving chat GPT the chance to a understand what the hell you're talking talking about is really important and
276. 41:58don't just dive into write me a paper on like like have it like like this outline let's let's see what it
277. 42:07did um certainly writing a five-page paper on the history of gummy bears is a great idea the evolution of gummy bears
278. 42:13a sweet Journey Through History it's not bad CH bt4 would probably be a little better but it's not bad abstract a brief
279. 42:20summary of the paper's main points introduction hook your readers with an in interesting fact or an OTE about
280. 42:26gummy bears introduce the topic and its significance provide an overview of the structure origin of gummy Confections
281. 42:33explore the okay so so that's great so I'm going to say okay that's fine can you now give me that same outline but give
282. 42:44me three bullet points under each of your bullet points that's actual historical detail so I can get a sense
283. 42:54if the paper is what I have in mind of course here you go hook all right so let's see what it does introduction hook
284. 43:04start with a captivating story explain the purpose okay let's see origins of gummy Confections discuss the history of
285. 43:12a Turkish delight a precursor to gummy candies dating back to the Ottoman Empire explore how gummy candies made
286. 43:19with ingredients like fruit juice sugar and gelatin emerged in the 19th century um let's see Hans rle in his vision describe how
287. 43:30Hans rle the founder of herbo started making gummy candies in his kitchen in Germany in in the 1920s herbo the
288. 43:37haribo's first gummy bear uh so so let me say um did Hans Riel create the first gummy bear or did someone create a gummy bear before he
289. 43:54did why a bear this is actually kind of curious to me Hans rle the founder Who is credited wait come back come down
290. 44:09stop while there were various forms of gummy candies before hun ral's times the gummy bear as we know it today is
291. 44:17typically attributed to him and his company Hans rle and the invention of the gummy bear okay cool inspiration
292. 44:24from dancing bears the inspiration for the gummy bear shape is said to have come from the popular Street
293. 44:30entertainment at the time which often featured dancing bears so I think I'd like to change the topic of the paper to start
294. 44:46out with the history of dancing bears and how that led to the invention of the gummy bear so that the paper ends with
295. 44:55Hans ryel inventing the gummy bear can you give me a new outline that's [Laughter] that um Can it give you the chemical uh
296. 45:10composition of gummies okay we'll get to that let's see let's see what our new paper is um history of dancing bears the
297. 45:17inspiration behind gummy bears Hans rle and the gummy bear gummy bears go Global um let's see yeah anyway that so
298. 45:25we we go on like this so this is what I would do it um let's let's see can you give me the chemical composition of
299. 45:31gummy bears and the difference between soft and chewy gummies and harder gummies gelatin sugar colorings citric
300. 45:46acid soft and chewy gummies are typically made for a higher proportion of gelatin and a lower proportion of the other
301. 45:53ingredients they have a higher moisture content which contributes to their soft impliable texture so yeah like like
302. 46:00that's the thing about chat GPT is you can very very quickly run down rabbit holes and explore so for your paper what I would say is
303. 46:09like like rather than having to kind of lock yourself into an idea like run down rabbit holes explore and and and maybe you find you know
304. 46:21interesting hybrid kind of things that make it really interesting you know and I would say set your expectation even if you're using chat
305. 46:30gp4 set your expectation that what you'll get out of generative AI if if you're really good at prompting if
306. 46:38you're really good at prompting and do a lot of the stuff that I'm talking about here will be about 80% of what you're probably comfortable
307. 46:46with like these things you know my experience with this is it's very much a collaboration where I use chat GPT to get me to 80% it gives
308. 46:57me the structure it gives me the basic ideas it gives me some research and then I have it write a draft or two of
309. 47:04paragraphs or sections and then I copy and paste that into word and I'm like oh God this sucks right this oh this middle
310. 47:10paragraph it completely miss this and I'll completely rewrite a middle paragraph and then the one below it's
311. 47:16okay and and it doesn't even necessarily save me time but what it does is if I have two hours to write the thing I'll
312. 47:25get to 80 % in 10 minutes and then I've got two hours to refine the paper so I'm curious about how AI will change
313. 47:36school life within the next few years you and me both what's your name son wait oh Sun you've been in here before
314. 47:43uh you and me both like I I think I think I think the impact on education is going to be profound because I think our
315. 47:51educational system right now is essentially designed to get people into specialized tactical execution training for the most part there are
316. 48:04exceptions to that obviously the gummy encyclopedia this is awesome that's Sherry
317. 48:20D that's really awesome um ask if gummies contain an food dies that are detrimental to human health oh well we know the answer to
318. 48:29that uh please tell me what chemicals are in gummies that there's evidence that they're detrimental to human
319. 48:43health it might we might get one of the safety guard rail things here that seems very controversial to me Kyle I don't I
320. 48:50don't like that qu I don't like your line of questioning here no such thing as a hard gummy the last supper with gummies y'all
321. 49:05are weird um sugar artificial flavors caloric content let's see artificial flavors food colorings many gummy
322. 49:17candies contain artificial food coloring such as red 40 yellow five and blue one colorings are linked to hypera
323. 49:25activity in and attention issues in children and apparently 58-year-old adults who have haribu gummies right next to their little desk M
324. 49:41tasty chemicals good for living okay let me show you another thing here um
325. 49:57I'm going to go grab I'm going to go grab our outline so this is the outline that chat gpt3
326. 50:11wrote and I'm going to copy it and now I'm going to go over I'm going to do a new chat and I'm going to go to if you
327. 50:18haven't seen Dolly 3 within chat GPT do you have any formal training in Ai No Sir what are your qualifications yes
328. 50:27rather uh so uh by qualifications do you mean training per se oh yes yes yes we were expecting if you had something
329. 50:34called the AI learning lab first where's your lab coat second where's the lab and and third by Sir by chance uh what are
330. 50:40your qualifications sir good day sir uh yes about that uh rather I uh I consider myself rather an expert in in in uh
331. 50:49considering the exploration of artificial intelligence or machine learning it is more accurately known I've got no qualifications sir please
332. 50:57please don't kick me out please oh God no no I don't have any anyway um no I don't but what I do have so
333. 51:06here's what I do have I I'll stop being a smartass for a second um I've been an entrepreneur most of my
334. 51:15life and I've started 12 or 15 companies I started one of the first digital agencies in the mid 90s when the
335. 51:21worldwide web came out and so I think that one of the skills that that I've developed in my life is I love technology but I like
336. 51:36technology when it hits that inflection point where the rest of us get access to it like my company Story vine I had an
337. 51:45epiphany about storytelling video storytelling when YouTube first came out that a lot of the content on YouTube
338. 51:52sucked because it lacked basic story structure and I thought what if there were a system what if there were a way
339. 51:59to provide like like story structure story guidance for someone to tell their story in a really easy way that would
340. 52:08create a high quality video without them having to learn to be a [ __ ] film editor and a director and a writer and a
341. 52:16cinematographer and a sound designer and a graphic designer like it was like it was like in order to do like a a solid
342. 52:23two-minute video you had to you had to kind of Master seven different professions and it just seemed [ __ ]
343. 52:30ridiculous to me right and so I'm I'm fascinated by technology that that puts a s Simplicity layer on top of
344. 52:40incredibly sophisticated stuff below it it's why I'm a fan of apple right Steve Jobs was maniacal about making technology perform as
345. 52:51magic right the worldwide web prior to the worldwide web let me let me flip this for a second prior to the worldwide
346. 53:00web the internet was command line interface and it was it was the purview of scientists and mathematicians and
347. 53:07researchers and Librarians and then Tim burner Lee said hey we could make a hyperlink from this word to a whole other
348. 53:16document and then I forget who did the first browser but whoever created the first browser made it so that you could
349. 53:22click on it and so you got five or six or seven decades of Internet development and all of a sudden the rest of us can
350. 53:33do this click click click and it changed the world so my expertise is in jumping in the moment that [ __ ] shows up so I've
351. 53:48been [ __ ] around with AI for 5 years and like every time Google would say there's a new version of tensor flow and
352. 53:54it's it's much more more intuitive and I'd go in there and be like awesome intuitive Ai and I'd go in and I'm
353. 54:00like I it's clear I don't have a mathematics degree from Stanford this hurts my head and I would leave and then I would go in again and I
354. 54:09would try and I and I did this for years and then all of a sudden when I started playing around with blockchains
355. 54:15and nfts I got involved in a project called pixel mind which was an nft project built on top of stable diffusion
356. 54:22and it was kind of like Mad Libs where you just fill in a few words and it would create art for you and I was like
357. 54:27oh that's kind of magical I started playing with that and I started learning more about it and then I and then I got into stable
358. 54:35diffusion and dream booth and I created an art project I'm like oh this is really cool and then chat gbt came out
359. 54:41and I was like holy [ __ ] this is one of those chat PT does for the Decades of machine learning and artificial intelligence
360. 54:53technical infrastructure develop velopment what the worldwide web did for the internet and so what this Channel's
361. 54:58about is well what the [ __ ] do we do with it now we've we've all just been given this profoundly powerful technology that's
362. 55:10sort of like as simple as clicking on a hyperlink it's like the Google box but it's way [ __ ] deeper than
363. 55:18Google what do we do with it so this channel is about exploring that and and just like demystifying it and saying
364. 55:27these tools right now are just horrible unsophisticated janky nasty pieces of software not because the companies that develop them
365. 55:39are bad but just because we're that early we're that early that there hasn't been time to figure out what these tools
366. 55:47even do much less put good interface in front of them so um what can I do to use AI to create video ads for Facebook here I'll show
367. 55:58you in a second here or uh to create video ads so I so actually the answer to that right now is just go to canva and
368. 56:06sign up for their AI tools canva right now is is [ __ ] killing it they've got AI scattered throughout canva you can go
369. 56:14in and just say give me copy for these things and it turns it into Instagram post it turns it into video they've got
370. 56:21Runway ml integrated in there just go play with canva right now if you just want to do video ads um but I want to I
371. 56:28want to show you something here so for the for the people or for the paper we were just talking about we did our gummy bear paper and I
372. 56:36just want to I just want to demonstrate um image generation with Deli in the context of a conversation um believe it or not I
373. 56:46couldn't pin my last comment well this one's pinned that's weird H I don't know Tik tok's being an [ __ ] all right so
374. 56:59I'm let's see writing a paper on the history of gummy bears and here is the outline so I'm going to hit uh shift
375. 57:10return I'm going to paste in my outline I'm going to throw a couple of uh whatever you call it uh pound signs
376. 57:22after the outline just kind of indicate that it ended that seems to help a little bit and then I'm going to say I'd like to
377. 57:32create an image for each of the first four sections that's in a watercolor style and a horizontal aspect ratio that brings that
378. 57:47section to life and so it will now it's it what what it's doing right now is it's writing the prompts I
379. 58:00don't need to write the prompts it it has read our outline and I said the first four sections so the first one is
380. 58:10um section one um the history of dancing bears right section two the inspiration behind gummy bears section three Hans
381. 58:18rle and the birth of the gummy bear section four gummy bears go global and so it's reading all that and I said I
382. 58:29wanted watercolors I said I wanted them to be horizontal and it's off writing prompts and now creating images and let's see how they are oh
383. 58:40they're kind of interesting so there's dancing bears there's dancing bear candy watercolor painting of a vibrant fair in
384. 58:50the early early 20th century dancing bear candy there's Hans ryel he he he was making some mighty giant sized gummy bears
385. 59:02there and there's gummy bears go Global so now I could spend time you know tweaking out the uh the style and the
386. 59:11details and things like that but Better Living Through Chemistry remember remember those commercials Dragon I do
387. 59:20um this is revelatory to me because if if you've not seen how mid Journey works let me go show you how mid Journey works
388. 59:29so I'm now over in Discord so I talked earlier about the the uh AI Salon Discord so we're in the AI Salon Discord
389. 59:38one of the channels we have is called mid Journey play and if if you're want if you're wondering what Discord is Discord is
390. 59:46like a community software essentially for gamers and then the nft Bros kind of took it over and and almost all nfts
391. 59:56have Discord communities and servers and and it's it's not great software but it's but it's been around
392. 1:00:04for a long time and it's functional and you can program within it and so one of the things you can do is you can do Bots
393. 1:00:10and one of the the only way you can make images with mid Journey right now is to get in Discord and learn all the
394. 1:00:17commands so if I want to make images in mid Journey I'll scroll to the bottom here here's some images other people are
395. 1:00:23making and notice this got like these weird codes like u12 34 V12 34 whatever that symbol is okay what's that I don't
396. 1:00:31know so you have to know this slash imagine is what you type in and then you see this imagine prompt create images
397. 1:00:40with mid Journey so I click on that or I just finish typing it and now I type in a prompt and and so now I've got to
398. 1:00:46think about let's do the um an image of Hans ryel inventing the gummy bear in his kitchen with some other
399. 1:01:01workers in a watercolor style and then I'm going to go dash dash AR space 16 colon 9 well Kyle but what
400. 1:01:14is D- AR 16 colon9 obviously if if you jeez we we're not it's so obvious isn't it it's aspect ratio-- AR is aspect
401. 1:01:28ratio how do you know that I don't know you just have to [ __ ] learn it right like this is the problem so you're using
402. 1:01:33a command line interface to do visual content creation Now why would anyone do this because mid Journey's quality is so
403. 1:01:42good that it's worth it right so I'll just hit that and there's all sorts of ways that you can get good at this but
404. 1:01:47you have to learn it but the difference between this and and if if you think about this so I'm over working on my paper
405. 1:01:55in chat GPT and it's just in text mode and then at some point I'm like oh it would be cool to have representative
406. 1:02:02images in here for these different sections what what what you have to do historically is you have to kind of
407. 1:02:07shift gears you have to go okay now I'm going to get into image prompting mode and I'm going to go to Mid journey and
408. 1:02:13then I'll look at my paper and I'll come up with huh I got to come up with specific things I want in the image
409. 1:02:20right as opposed to what we just did in chat GPT where I'm just like here's the outline go make me some images and it did it and so
410. 1:02:30here's obviously it doesn't understand who Hans rle is because these are all modern modern images but like notice the
411. 1:02:38quality of these like mid jour's quality is really good like these are these look more waterc colory than the do stuff did
412. 1:02:46and I can continue to refine this um and then if you're wondering what these letters are U stands for upscale v
413. 1:02:52stands for variation so let's say I I like the one where there's actually a bear in the kitchen with him so I'm going to ask for variations
414. 1:03:03of this one so that would be V1 so I click on V1 and so now it's going to take that basic whatever it did there use that as
415. 1:03:12a seed and give me some variations of that and then I can pick one of these and I can upscale it and then mid
416. 1:03:19Journey just added um 2X and 4x upscaling where you can now make print quality upscales of your mid Journey
417. 1:03:28images which is really cool but the downside here is it is a this is very much a technologist approach to a Humanity's tool
418. 1:03:43right this hasn't been Apple fied right this isn't Magic right now like what it does is Magic how you do it is a [ __ ] nightmare um
419. 1:03:55let's see all right here I like this one he's kind of working with these are the Goldilocks and the Three Bears so that's
420. 1:04:02so we'll upscale number four and there we go there's our image right so that's mid Journey but I mean that it it is just a compared
421. 1:04:17to you know what we just did here we was like read my thing and I'm like um now I can just say those are okay but there's too many
422. 1:04:26people in the kitchen comma the gummy bears are too big comma and the style doesn't look like watercolors period I'd like them to
423. 1:04:37look like hand-done watercolor paintings rather than historical illustrations so I can just give it feedback like a creative director
424. 1:04:50would or just like if you're a client and you hired someone to make you images for your thing if they gave you shitty
425. 1:04:59images you'd go you you'd either what you would likely do is if you're not a fan of Confrontation you might go okay
426. 1:05:06these are great and then not use them I've done that many many times in my life or you would go oh well these
427. 1:05:16weren't what I had in mind can you go back and do this and then another week goes by or two weeks go by and you're spending a lot of money
428. 1:05:23and and now we've got it right here in this tool and so there's Hans rle in his kitchen is that does that what Hans rle
429. 1:05:37looks like I don't fliping know doesn't matter we can go find images of him and if he's got a beard we
430. 1:05:44can say oh he's got a beard and he's got dark hair but there's his little Gummy Bears on his little plate and it kind of looks like a
431. 1:05:53watercolor looks like an Andrew WTH watercolor you know there's our history of dancing bears there's our little dancing bear
432. 1:06:09candy stand there's Hans there's gummies Go Global like perfect great love it crazy right [ __ ] nuts
433. 1:06:23man what world have we entered you want to see something crazy too yes Kyle that's why we're [Laughter] here um where is the
434. 1:06:36image I wonder if it showed up hang on let me check something let but check something oh okay yes it's here is this it yes okay
435. 1:06:51so um I'm I'm going to export this image hold please whiteboard
436. 1:07:07export where am I going to go desktops fine anyone else have 13,743 items on their desktop you know Kyle that'll make your
437. 1:07:21computer run less efficiently yeah I know what I wait for is technology companies to make the processors fast enough to
438. 1:07:28keep up with my lack of organization okay so we're going to try something here um another mode in chat
439. 1:07:37gp4 is just the default mode and you're like that's boring it doesn't have all this fancy [ __ ] oh yes it does so the
440. 1:07:44default mode in chat gp4 now has Vision they call it gptv it's it's buried underneath this little icon that says attach image and
441. 1:07:55it it's not clear what you do with it so I'm going to attach an image I'm going to go find that whiteboard that I just
442. 1:08:00so we had a meeting today with the AI Salon organizers and Vicky was there I don't know if Vicki's in here tonight Vicky was
443. 1:08:08there and we wrote down some ideas on a whiteboard so I'm going to say here's some notes from a meeting we had today can
444. 1:08:20you tell me what we talked about and and come up with a summarization of the meeting and so there's there's our
445. 1:08:33little whiteboard right so like any whiteboard you've done if you're like me you probably have I don't know uh 672
446. 1:08:42pictures of whiteboards in your Vicki she's here awesome uh you got pictures of whiteboards going back to
447. 1:08:491982 um and and you keep meaning to go in there and take notes from what was written on the Whiteboard right well
448. 1:08:56look at this certainly based on this meeting summary the end goal discussions what's the end goal um who are the
449. 1:09:03audience and participants what were the main topics the required expertise engaging high value and daily active
450. 1:09:11users High development through Community um platform for talent it appears the meeting focused on strategies to engage
451. 1:09:18with Talent develop their skills and provide a platform for them to Showcase their work the end goals seem to be more
452. 1:09:24about informing supporting and educating potential Talent about AI um and the importance of recognizing
453. 1:09:32and nurturing all levels of talent was emphasized lastly there's a mention of some kind of board or platform requiring
454. 1:09:39certain expertise and and an apparent login credential was noted so so in the picture in the upper corner here I had
455. 1:09:48the Wi-Fi the Wi-Fi password for the building we were in so it even like and it didn't say user or password it was
456. 1:09:55just un P but it recognized that as a login credential and then it took just these kind of you know unorganized notes took them I I
457. 1:10:06mean we we're we're in a different mode cat GPT is not the same tool it was a month ago a month ago it was like oh we
458. 1:10:15got to learn about prompting and we got to do prompt engineering and we got and there's a perfect way to prompt now that
459. 1:10:21you can use an image as a prompt is prompt engineering as critical um for a bit but I certainly don't need to
460. 1:10:32understand how to do image prompting with the do stuff with the with the gummy bear pictures I can just ask it
461. 1:10:38for what I want it's figuring out the prompting so I think we are quickly already we're not even a year into chat
462. 1:10:46GPT and we're already evolving from what skill was required 6 months ago is is fundamentally different than what's
463. 1:10:55required today and it's going to be fundamentally different 6 months from now and you're like but Kyle so what how
464. 1:11:01what the how what am I supposed to do with that the only the only strategy that I know is get curious and adventurous and think of it like a
465. 1:11:13game think of it like a game but Kyle this is my job and this is important y same same my comp I have an 11 and a
466. 1:11:22halfy old company that a lot of what we do is going to get automated out the value of what we do is going to be threatened by this
467. 1:11:30stuff so this stuff's existential for me too as excited as I am about it like why I'm so committed to it why I do these
468. 1:11:40three-hour sessions seven nights a week for the past four months is I do not want to lose touch with what's happening and I'm
469. 1:11:54trying my damnedest to create as big a community as possible that's doing the same thing that's what this is
470. 1:12:03about so so you know questions like what are your qualifications sir there are no experts right now I'm not an expert in this [ __ ] I don't
471. 1:12:12[ __ ] know like I don't I don't know what it means that I can in five minutes have it create a very detailed outline of a paper just
472. 1:12:23flip into a different mode and say go make me pictures for that at some point we're going to be able to say go write
473. 1:12:28me a song for that go make me a video for that if you if you hang out in the AI Salon Herod who who hangs out in here
474. 1:12:35he's one of the Irregulars he's doing just some amazing [ __ ] with video like all of all of the stuff that
475. 1:12:45that right now are these disperate tools out in the corners of whatever this stuff is are very very rapidly going to roll into a single
476. 1:12:57tool that is kind of the everything app what are you supposed to do with that as a professional I don't know are there any AI that can do
477. 1:13:10animation shorts absolutely um I but it depends on what you mean animation shorts there there are like Runway ML and P labs and what's
478. 1:13:20the other one something Mountain uh hang on is Moon Moon Valley something Mountain thanks add thanks thanks for the attention to detail
479. 1:13:33brain it's something mountain or it's Moon Valley so Moon Valley's one but this this one you make videos in Discord
480. 1:13:41so this is like like mid journey and then there's P Labs which might also be um Discord and then there's Runway ml which let's go to
481. 1:14:00runwayml Runway ml try it for free log in so bring your image to life with Gen
482. 1:14:142 so I can either start with a text prompt or an image prompt so if I start with a text prompt I can just say
483. 1:14:29make me an image of or not an image hang on a dancing gummy bear on a confectioner's table while the
484. 1:14:46confectioner dances along and I can say generate a 4 second clip and I didn't put any style information
485. 1:14:55in there by the way do you like the um the fingerprints I did that for you see it's very it's very organic this channel
486. 1:15:04you know it's not very professional you should probably clean your screen probably I should probably have qualifications too but I
487. 1:15:12don't I gotta I got to up the game game for one of the AI learning Labs at the Vegas sphere but you know that's that's
488. 1:15:19not till next summer um how to mimic pictures you have for someone but changing some elements um so so there's
489. 1:15:27there's a few different ways to do that roro um within mid Journey you can upload an image that's kind of your base
490. 1:15:34image and then describe a new image after that if you really just want to modify elements of an existing image I
491. 1:15:41would use something like photoshop's generative fill where you can select parts of an image and say if someone's
492. 1:15:47wearing a sweater and you want to change it to a denim shirt you just kind of highlight their shirt and say you know
493. 1:15:53change this sweater to a denim shirt um so I would do something like that all right let's see how bad this gummy bear
494. 1:16:00is okay that is a horrible gummy that is a terrifying gummy [Laughter]
495. 1:16:17bear I'm sorry I thought this was going to be bad this is [ __ ] genius I think we have a new Meme here people
496. 1:16:30um okay so that's pretty bad but that's one of these so bad it's good things so so now let's go TR try something else so
497. 1:16:37we'll go we'll go back to dolly or yeah to to to uh J J and let's say um please create a 16x9 image of a
498. 1:16:52highly photorealistic closeup of a gummy bear on the confectioner's table that looks as though it was shot in a high-end photography
499. 1:17:07studio okay so what I'm going to do now hopefully is get an image of a gummy bear that doesn't suck and might be a good basis for an
500. 1:17:19animation like that and now we'll use this image as the basis for that animation and and you know the caveat
501. 1:17:32with all of these tools that that you know I say pretty regularly in here but I you know it's oh that's great although
502. 1:17:39let's see that's it's not quiet so hang on um that's close but I want four different variations and I want the gummy bear um standing on two
503. 1:17:52feet with its arms extended as if it's dancing period I also want it on a kitchen table with sugar and a stainless steel
504. 1:18:09bowl and a little piece of paper that says gummy bear recipe period the focus of the image should be the gummy bear but I want it clear that
505. 1:18:21it's in the kitchen where it was invented all right um so all of these tools right now are
506. 1:18:36um very imperfect and and what I would strongly encourage you not to do is don't dismiss them because they suck because 18 months ago image
507. 1:18:51generation tool were abysmal like they were like like the the well so so let me let me put it in context for what they
508. 1:19:01were doing they were absolutely [ __ ] remarkable right from a mathematical standpoint and and the fact that they
509. 1:19:07could generate images at all was absolutely miraculous um but now 18 months later these tools are in seconds
510. 1:19:16creating worldclass images um but Kyle are you demonstrating that chat GPT can read handwritten notes down yes I am it
511. 1:19:25can read handwritten notes it can read oh this is perfect look at this one perfect gummy bear look there's the
512. 1:19:33gummy bear recipe there's our dancing gummy bear this is perfect okay so watch so now we're going to go save image as and we'll just call it
513. 1:19:45gummy dance so what' that take us 2 minutes to get an image that's like okay that's not bad so now we'll go back over to Runway
514. 1:19:56ml we're going to go to an image drag and drop an image so let me go find my image where did I put it gummy dance here it
515. 1:20:09is well I'll upload a file desktop is this it oh it saved it as okay hang on it generated it as the wrong file format it's a webp image and
516. 1:20:23now we've got to export it as a PNG this is the kind of crap that that makes me insane like the fact that there's all
517. 1:20:29these different file formats and all the services don't support them is just [ __ ] maddening but whatever this is the kind of stuff that
518. 1:20:40slows you down you know when you're like hey I have an idea for something and then like two hours later you're still
519. 1:20:45[ __ ] Googling what's webp format and is there a webp Transformer and you've down downloaded six pieces of software
520. 1:20:54and they all install malware on your computer and then they all charge you 20 bucks to trans trans that needs to be fixed and that
521. 1:21:05will be fixed because um because AI is going to fix that [ __ ] all right boom there's our little image oh wait I want image plus
522. 1:21:18description uh a dancing gummy bear on a confectioner's table table let's just we'll leave it like that okay
523. 1:21:29so we've got our uploaded image here we've got a dancing gummy bear on a confectioner's table and we're going to
524. 1:21:35say generate 4 seconds so while while that's generating wait is it generating yes oh wait it's generating twice can I cancel one of
525. 1:21:48these no but while that's generating let's let's watch our let's watch our our Masterpiece my name is I was the gummy
526. 1:22:02I'm a gummy gummy Cherry guy yeah the dancer I'm awesome look at me go [ __ ] awful all right so here's our our real dancing gummy
527. 1:22:19bear um I'm gonna be I'm going to be disturbing dancing gummy bear for Halloween exactly all right let's see
528. 1:22:27how it did please not suck please don't suck it sucks all right it's disturbing but you know and again you could you could
529. 1:22:40likely muck with this and and get it to do something predictable right I don't I don't I haven't played around enough
530. 1:22:46with with Runway ml to know what are the what are the secrets to get coherence and and from what I understand P Labs is better than
531. 1:22:55Runway but again go back to this idea that these tools are as bad as they're ever going to be right so every time it does
532. 1:23:07this they're looking at this and they're going okay that's awful how can we improve the coherence how can we improve
533. 1:23:13the coherence how can we improve the coherence um last week I asked you about using chat GPD to clean up a CSV file for old leads of clients
534. 1:23:23then I heard elsewhere not to upload personal info why okay so um let me stop this let me flip this and go
535. 1:23:40here um so part of part of the tools right now is um you should just assume that these tools are not safe and secure until they prove
536. 1:23:57themselves otherwise um because um some of them aren't and and in the case of chat GPT in particular um if you leave on your prompt
537. 1:24:09history um the prompts that you put in there are susceptible to hacking basically in in a in a pretty pretty
538. 1:24:17easy way so you should just assume that that anything you put out there you know can be gotten to now open aai has created an Enterprise
539. 1:24:27version of their software which isn't available to everyone yet you can also if if you um if you really want to be if you
540. 1:24:39really want to keep your stff your stuff private and secure you can install a a open-source large language model on your
541. 1:24:49local machine or on a virtual machine in in the cloud that has its own boundaries so you're not sending um data out to a
542. 1:24:57company like open Ai and then getting it back um you're just doing it within your own machine um so there are ways that
543. 1:25:05you can do things more securely but they require a lot more technical stuff so if it's the kind of thing where
544. 1:25:15um you can depersonalize the information great if it's just you know you're you're if there's nothing proprietary
545. 1:25:22about the information uh you can do it if you're using something like code interpreter within chat gbt which is
546. 1:25:31called Advanced Data analysis it's actually running the the data analysis it does in a container inside your chat
547. 1:25:40I don't think that that information is going out to the world and back I think it's doing all that analysis locally
548. 1:25:48there but even if that's true like I wouldn't necessarily trust that so it's just it's just the nature of these tools
549. 1:25:56right now we're very early so you should just assume right now until proven otherwise like included in your terms of
550. 1:26:05service is your data secure you should assume that it's not and so don't put anything um super private or sensitive out out in
551. 1:26:16these tools de depersonalize it first welcome everybody there's a bunch of new people in here let me let me do I want
552. 1:26:28to do a little reset here quickly so um my name is Kyle Shannon this the AI learning lab uh I'm an entrepreneur and
553. 1:26:36uh this channel is about obviously AI stuff and talking about AI stuff if you have not tried chat gbt yet um
554. 1:26:51go go there chat. open.com that's the official chat GPT website it's a lot of the demos I've been doing tonight like the dolly3 stuff
555. 1:27:03that's all there um the chat GPT plus their 20 Buck a month like upgrade in my opinion is worth it I'm not connect I'm
556. 1:27:13not affiliated with them in any way so if you don't do it I don't I don't really give a [ __ ] I don't make money
557. 1:27:17whether you do it or not but they have added so many features in the past month that I think this is now kind of the the
558. 1:27:24deal of the deal of the year or deal of like I think it's well worth the 20 bucks um if if you're you know super
559. 1:27:32curious and want to dig in if you're not if you just want to dick around with it just go there play with play with the
560. 1:27:36free version um you can also play with Bard is free Claude has a free version you can upgrade to get you know more uh
561. 1:27:45and Bing is free so all four of these have free options but if you're brand new to this this uh generator AI stuff
562. 1:27:52just go to chat GPT right now while while I'm rambling you're like it's Sunday night Kyle I'm a little tired I
563. 1:27:58don't feel like working I'm not very technical I don't want to learn doesn't matter just go there just do it just do it do it do it do it do
564. 1:28:09it just go there and just start playing well what do I do I don't know what to ask and I'm not very technical you don't need to be
565. 1:28:17technical just go here go go tell it to write you a country song about two gummy bears that are falling in love and it
566. 1:28:25will do that and then have it write you a marketing plan for your pizza shop and then have it give you an exercise plan
567. 1:28:35for getting you off the couch after being sedentary for three years thanks to co and it just just go just go play
568. 1:28:42while while I ramble here so anyway um and then if you have any questions about AI pop them in the comments below I'll
569. 1:28:48do my best to answer them and uh and then we'll have some fun we'll have some fun and you know you can you can make money
570. 1:28:58with we can talk about that uh I'm happy to demo stuff um anything all right let's see we are on to something let's see it looks like
571. 1:29:11it's playing Beat saber this is giving Pillsbury Doughboy Vibes those yeah those videos were pretty disturbing weren't they they were
572. 1:29:20pretty good [Laughter] all right let's leave that in the background while I talk [Laughter]
573. 1:29:36here what is that disturbing image oh my it is that [Music] Satan it's perfect Kyle it's paranormal gummy bear for Halloween it it these are
574. 1:29:52very Halloweeny kind of disturbing this looks like some Herod if you put this to your heavy metal rock music this would
575. 1:29:59this would play pretty nice Waka oh
576. 1:30:16man all right that's that's distracting me now sorry I got to go back to static images I can't take it I can't take it
577. 1:30:25people howy partner life's tough it's tougher when you're [Music] stupid there's not enough AI learning in
578. 1:30:37this town for the both of us one of us is going to have to go Skippy Skippy yeah I just got in a fight with
579. 1:30:54Bing Bing's an [ __ ] I just bing bing I'm just Bing's on my [ __ ] list you should be able to talk back to
580. 1:31:05the robots without them getting offended being is just an over sensitive narcissistic nin I hope someone from Microsoft is in
581. 1:31:17here lighten up on your safety guard rails fingerprints is t
582. 1:31:31y oh see listen on other channels trolls just come in and they just are mean but here at the learning lab trolling can be
583. 1:31:49fun it's like trolling and self expression oh my God this is so good wait this is from Ted what did you say
584. 1:31:59here we're all just finding our way this is very right this is what's what's going on right now like this is what we're all trying to figure
585. 1:32:07out like what is this stuff how do we do it I don't know it's been 15 minutes since my last Dolly 3 I've got such shame
586. 1:32:24oh my God it's so funny it's crazy I tell you it's crazy a Pap all right so um do me a favor if you're new here um if you're
587. 1:32:38likeing if you're new here and like what the [ __ ] is this channel welcome uh if you're new here and you're
588. 1:32:44enjoying this do me a favor share the live with your friends even though they're going to think you're weird for
589. 1:32:50sharing something like this um follow the channel um so you get notifications when we go live I've I try to go live
590. 1:32:58every night I've been I don't think I've missed a night except when I've gotten kicked off Tik Tok three
591. 1:33:05times I've gotten better I haven't done it in a while I should I should knock on wood to to not get kicked off um but follow the
592. 1:33:15channel um and and uh and yeah do any do anything you can to to engage here I'd like to like to get as many people in
593. 1:33:22here as possible I've got Bing trapped where it can't hang up on me oh that's good you should let's see what I
594. 1:33:31had going on with it you should have you should see what you had going on with it this morning well that's good that you've got
595. 1:33:38it to to uh to not hang up on you it hangs up on me all the time it drives me crazy Bing drives me nuts how about a gummy bear jumping into
596. 1:33:48the eye of a carved pumpkin for Halloween that would work what's good about Pi okay here's what's good about Pi
597. 1:33:57so let me tell you what's bad about Pi so if you go to Pi it's just like chat GPT all these are just prompts
598. 1:34:09basically like I need whatever just vent like this just tells you what you can do with it but it's just like chat GPT
599. 1:34:15where you just you know there's a little prompt box and then you type [ __ ] in and it types stuff back if you try to actually get something
600. 1:34:24done with pie like give me a list of the ingredients of gummy bears it will be the most frustrating experience of your life because it's
601. 1:34:35designed to be a conversational bot So when you say give me the ingredients for Gummy Bears it'll be like that's an
602. 1:34:42awesome request I'd love to give you the ingredients for Gummy Bears can you tell me why you want the ingredients for
603. 1:34:48Gummy Bears you like I just want them could you just give them to to me in a bulleted list a bulleted list is a great
604. 1:34:54way to do that so it can be very very annoying if you just want [ __ ] done but if you want to talk it's really good
605. 1:35:02at that it's so it's designed to be a conversational AI so rather than type just put it into conversation mode and start talking to
606. 1:35:13it hey I really love gummy bears and I'd love to know a little bit about the history of them where were they invented and how did they get
607. 1:35:24[Music] popular gummy bears are delightful little treats with a fascinating history they were invented in Germany in the
608. 1:35:361920s by Hans Regal who was looking for a way to create a new type of candy that was soft chewy and had a longer shelf
609. 1:35:46life than other candies at the time she's so Saucy the original gummy bears were fruit flavored and made with a
610. 1:35:52combination of gelatin sugar and fruit juice they became popular in Germany and spread throughout Europe and the United
611. 1:35:59States in the 1950s since then they've become a staple of candy stores and kids lunch boxes everywhere and why
612. 1:36:18Bears ah I see where you're going with this New York Bears huh it's true black bears have been spotted in parts of New
613. 1:36:24York State particularly in the more [Music] rural no I mean why are they gummy bears
614. 1:36:39why are bears the shape that's an excellent question the reason
615. 1:36:55why gummy bears are shaped like Bears is actually pretty cool the creator of gummy bears Hance Regal was inspired by
616. 1:37:02the trained Bears he grew up watching at street festivals in Europe so so that's that's what it's good at and then you
617. 1:37:10know you can it can be educational content you can talk about just anything and and it's it's designed to do that
618. 1:37:18it's funny I just saw a notification pop up on my phone someone made a comment on one of my videos chat GPT is overrated
619. 1:37:25whenever I see something like chat GPT is overrated it's just aign to me that someone's sitting on the sidelines and
620. 1:37:31is not actually curious and exploring it because it is both underrated and overrated um like it definitely has
621. 1:37:42improvements but the deeper you dig with it the more remarkable it is so anyway um my hope is that if you're here and if
622. 1:37:54you like and if you're digging it is that you just start getting curious and start exploring just be in an be in an open mind
623. 1:38:03like while I'm enthusiastic and and optimistic about this stuff you don't have to be but I think it's important
624. 1:38:11for everyone to at least get educated on what it makes possible where are the boundaries what's it good at what's it not good at
625. 1:38:23what of your job can it do better than you do like go there you know that that saying don't ask a question you don't want to know the answer
626. 1:38:36to I think that's a lot of what's going on with the fear of AI is that people don't want to know the answer to that I
627. 1:38:43think it's important to know the answer to that because the more you use this stuff the more you realize this is an imperfect tool
628. 1:38:52that is going to change everything but it still requires lots of human engagement and lots of human oversight and lots of human
629. 1:39:00participation and the more of us that are using it and understanding that the better we're going to get at it the the
630. 1:39:07better we're going to be able to influence what it is moving forward Amelia's wife we're just doing chitchat up till now okay Kyle we can
631. 1:39:18start oh Amelia's wife is here hang on got to get the uh the appropriate image uhuh oh wait I'm in the wrong section hold
632. 1:39:34please am I late the answer is blowing in the wind and yes you're late you're very late oh hi Mark hey oh
633. 1:39:50hi does he have [Laughter] Tourette's sometimes I realize what a
634. 1:40:05what a n poop I am and it just cracks me [Laughter] up all right let's get serious this is this is called the AI learning lab this
635. 1:40:20is about learning thank you for the uh for the hot pepper there Mike I appreciate that the uh the uh those
636. 1:40:27gifts mean a lot thank you um I want to see what can be done with Hardware what's the potential for llm
637. 1:40:34applications on Hardware I think it's huge Ron so if I were you thank you Robert Rosy I appreciate that um if I
638. 1:40:41were you I like go to Reddit and and go start trolling the open-source um uh like GitHub reposit atories and the um hugging face um
639. 1:40:58projects and um just start like like yes there's there's all sorts of stuff you can do with Hardware right now there's I showed
640. 1:41:08I showed a thing last night that's um this is some Nvidia a team from Nvidia created this thing yesterday this is [ __ ]
641. 1:41:19insane let me see where I have it do I have it anywhere no let me go x.com and go to
642. 1:41:33pen spinning look at this so so this group at Nvidia use chat GPT
643. 1:41:52T4 with this this new approach they called Eureka to train a virtual robot hand to spin a pen they trained it in 2 minutes
644. 1:42:03this is apparently a notoriously difficult exercise for even 3D animators to get right for like like like just
645. 1:42:13doing the animation and they actually trained the robot to do this in two minutes using chat GPT um so that same technique I mean that's a
646. 1:42:23virtual robot right but you could send that code to a real robot hand right and then maybe it doesn't do it in the real
647. 1:42:30world but give it another two minutes so so there's all sorts of projects right now um go look at the Boston Dynamics
648. 1:42:37stuff where they put chat GPT inside one of their little robot dogs that was that was like months ago three or four months
649. 1:42:44ago so I assume at this point it's it's much further but they're doing all sorts of cool stuff where they're um they're
650. 1:42:51having chat GPT output like using special instructions basically they're having g chat GPT output Json so it can
651. 1:42:59read Json and then it can output Json and then they they basically just hook the apii up to the to the uh to the
652. 1:43:07robots instructions and you can talk to the robots and control them and do [ __ ] so um lots and lots of stuff going on
653. 1:43:15there I'm not in that world at all so I don't know anything about it I'm new to Mid journey and Discord do you have an
654. 1:43:20AI learning lab Channel kind of we do so we've got a um a thing called the AI Salon which is the a community that I started about a
655. 1:43:30year ago and the AI Salon the salon. a is the URL and that uh we have a Meetup so every two weeks there's an inperson
656. 1:43:42Meetup uh it's iners here in Denver but then it's virtual for everyone else and we meet every two weeks so this
657. 1:43:49coming Tuesday is the next meeting and then there's a salon Discord so a lot of the people that come to this channel
658. 1:43:56regularly um we're using the salon Discord is the place to kind of extend this community and extend the conversation and and talk and
659. 1:44:04collaborate in between you know these meetings as well so yeah there is and looks like there's a lot of people joining and introducing
660. 1:44:14themselves which is awesome so if you go over there if you if you go join the salon um please introduce yourself
661. 1:44:21there's there's an introduction Channel an introduce yourself Channel let us know who you are what your background is
662. 1:44:26where are you on the AI Journey like if you're just starting cool if you're like you know I've been doing machine
663. 1:44:32learning for 30 years cool if you're somewhere in between like most of us probably are cool um everyone's trying
664. 1:44:39to figure this stuff out right now everyone's trying to figure this stuff out no one is an expert zero people zero of the people zero zero
665. 1:44:49people know what's going on right now what a time to be alive isn't it I mean that thing two minutes to train a robot
666. 1:44:59to do that so so think about that I didn't get that Siri Siri Siri Siri so jealous Siri hey go
667. 1:45:19away it's oh you can't see it wa oh you can't see it it said it said Siri is so jealous Siri go away and then it just
668. 1:45:35faded um we're in this remarkable transition right now where we're moving from a Time time
669. 1:45:50where if you wanted to program a computer you had to think about the logic like part of what makes programming fun is doing the problem
670. 1:46:03solving and the logic navigation to say okay if I want this thing to do that how do I do that and so if you wanted to get a robot hand
671. 1:46:12to twirl a pen you'd have to think about okay how do I you know okay I'm going to move this finger down like that and I
672. 1:46:20got to figure out the timing and there's going to be milliseconds and I got to and then there's the inertia of the
673. 1:46:25thing spinning and there's gravity and and like you have to think about all that logic right and so that takes a lot of time
674. 1:46:33and it's a lot of trial and error it's a lot of back and forth and what they did in this case is they had chat GPT just
675. 1:46:40do the sort of problem solving and and they did the kind of feedback like that's good that's not good and in two
676. 1:46:47minutes it just trained it to do it wait like what think about all the problem solving that is complicated problem solving that
677. 1:47:02we're going to be able to just toss at these tools and let it figure it out that's like that we're in the middle of
678. 1:47:09that transition right now and I think that's that's one of the things that people are kind of missing with this
679. 1:47:15stuff is like if they go in and they like ask it to to write something and it writes something they're like oh well
680. 1:47:21it's it's it's inaccurate so therefore it's useless but they haven't sort of dug down to say well what what can it
681. 1:47:28actually do like I understand that it might make hallucinations and it might make mistakes but what can it actually
682. 1:47:34do and the people that are taking these tools and and running down the rabbit holes are inventing [ __ ] like
683. 1:47:42this and so now once you get it so that you can train a robot to be that dextrous in 2 minutes with chat
684. 1:47:49GPT make money with what the [ __ ] else can you do with it I don't know someone in one of these calls is going to go oh you know what I
685. 1:48:02could probably do with it and they'll oh it can do that and they're going to invent some [ __ ] new way of doing something
686. 1:48:09because they weren't closed-minded to the fact that it's not perfect right now sorry I had to log out what's the robot
687. 1:48:17what wait what's with the robots dancing where's that oh wait the the gummy bears dancing this that that's a gummy bear
688. 1:48:33dancing you're like why why does it have things coming out of its head because because that's where we are that is Runway ml so I took an
689. 1:48:44image um where's my image I took that image and put it into Runway ML and made it dance if you're asking
690. 1:48:57about the robot hand spinning the pen that thing that is some researchers at Nvidia use chat GPT to train a virtual robot
691. 1:49:12hand how to do pen spinning and they they got it so that it could train it in two minutes or something like that some
692. 1:49:19crazy ass thing and it's doing all sorts of dexterity stuff using chat GPT I was also playing with trivia with
693. 1:49:29Sydney it refreshed and it was Bing's trivia was 100% different oh that's interesting huh yeah the these things they really
694. 1:49:39are weird like like here here's the here here's my CH my challenge with these tools right now is they demo really well
695. 1:49:51and then when you try to get them to do something for a particular project you're like uh they're not quite there like dang
696. 1:50:00it so they're frustrating in that way but like like I've gotten I've I've created four or five different tools for
697. 1:50:08my company that we use we use regularly now and and like one of the tools that we use I think it was just written on um
698. 1:50:16it's not even using uh GPT 4 it's just using GPT 3.5 so anyway any any let's go to oh you know which one I want to go
699. 1:50:33[Music] to all these images by the way were all made in Del this one this is AI Anonymous hi my name is Kyle it's been 2
700. 1:50:46minutes since I've made something in DOI I'm making something right now now Kyle thanks that's awesome you're the
701. 1:50:57best could could you make a picture of an obscure history book inside a UFO sure here you go Kyle what did you
702. 1:51:07create the dancing things in I missed the beginning um Runway ml so I did I just typed in text to video you'll appre appreciate this over
703. 1:51:19Fox so did um I did text to video so this was um this this one was um gummy bear dancing on a on a confectioner's table
704. 1:51:34and so this is what it created so so this you can have nightmares about but I think that's kind of awesomely bad and then we went over to
705. 1:51:48um chat GPT and created this image with Dolly 3 and I'm like oh that's going to be a great video and then so that was the start
706. 1:52:05image for this thing which you're like okay this is going to be awesome and then it similarly kind of goes kind of goes all devil on
707. 1:52:17you so so that's what that is um could pen spinning become surgical assistants yes yes there there one of
708. 1:52:28the one of the other things there there was like one of the other things that happened this week was was someone
709. 1:52:34figured out a way to get chat GPT to understand like the folds of the brain or to visualize the folds of the brain
710. 1:52:42based on something like like yes like that's the thing is for any of these tools for any of these demos that you see right now
711. 1:52:51extend them out say okay that kind of dexterity if you can get a robot to train a robot to do that then we we we
712. 1:52:59will start to be able to train these robots to do incredibly sophisticated things without someone having to program it
713. 1:53:10it's it's C it's cuckoo so yes yes that is that is the implication there it's like holy cra crap just the speed of innovation I mean
714. 1:53:23think about that if if if you go from having to figure out the physics of how a robot with whatever its you know
715. 1:53:32dexterity and speed of actuators and you know stickiness of rubber substance and physics and gravity If instead of having
716. 1:53:41to think about all that and program in all those those things manually as humans the thing can just just figure it
717. 1:53:51out think about every complicated system every complicated system ask the gummy to do a Latin dance at a competition
718. 1:54:00level with another gummy bear oh my God it's going to be too bad I'm not going to do it by the way if I I assume
719. 1:54:09Corey you've seen the Tik toks all the West Coast Swing all those dancers that do the the uh the the uh the dances that
720. 1:54:17they kind of make up in real time those those people people are amazing all right what are you doing time wise getting there getting there
721. 1:54:26happy happy Sunday night people hope you're having a good time tonight watch two Shapiro YouTubes today and the guy
722. 1:54:33from Google also I was Kink yeah so so what Silver Fox is talking about is David Shapiro so one of the guys that I
723. 1:54:42recommend you watching is um David Shapiro on YouTube his channel is here hang on hang on
724. 1:54:58it's come on right right that one at 4ir which stands for fourth Industrial Revolution at 4. david.
725. 1:55:14Shapiro go watch his three most recent videos and and then run down the rabbit hole he's wild so previously inoperable
726. 1:55:26conditions will no longer necessarily be so yeah and and you know things like things like paralysis like they they've
727. 1:55:34already used um large language models to this is just insane they did this with a guy in Italy about I don't know four or
728. 1:55:42five months ago so I'm sure it's Advanced since then they basically took MRI scans of his brain they said think about moving your legs
729. 1:55:52he was in a motorcycle accident so severed spine can't move his legs um so they used large language models to
730. 1:56:03read the MRIs when he thought about moving his legs they then sent those signals via a Bluetooth jumper they so
731. 1:56:10they bluetoothed around his injured spinal cord with electrodes you know going down to his legs and and basically sent the large
732. 1:56:21language model learned or the Transformer learned patterns of thinking of moving his legs down to electrodes in in his lower spine and now
733. 1:56:32he's walking so so it's like neurological disorders injuries um yeah like just the yes
734. 1:56:47yes like when I I don't know like how I don't know but it's but it's all happening like it's all happening in parallel
735. 1:56:57right now like that that's the thing what what what's fascinating about these tools is that the core technology for how chat
736. 1:57:07GPT does what it does and how the image generation tools do what they do and how the music tools do what they do and how
737. 1:57:15these brain reading things do what they do it's all using the same [ __ ] it's all just different data as the input so they
738. 1:57:22they basically take massive amounts of data they run it through this thing that Google invented called the Transformer
739. 1:57:29that embeds this data into some mathematical Vector space and then they take some other input in the case of chat GPT the input is your
740. 1:57:42prompt and then that prompt turns into this mathematical probability thing and it [ __ ] just goes into this Vector
741. 1:57:48space and pulls out the relevant data and executes it and so like in the case of chat GPT it'll write you your paper
742. 1:57:55or it'll program your code in the case of controlling robots it's it's it's been trained up on lots of robot code
743. 1:58:05and then you tell it what you want it to do and it sort of goes into this Vector space pulls some probability [ __ ] out
744. 1:58:10like how it does that I don't [ __ ] know but it's all using the same core technology so it's just about what data
745. 1:58:20do you train it on is what it can do like when you go to the haen thing like the the the haen where do I have this open hang
746. 1:58:44on oh that's not what I wanted hold please hold
747. 1:59:00please where am I missing something [Music] here I want to show you
748. 1:59:14something here it is so this is this thing where I'm speaking Mandarin
749. 1:59:29[Music] Chinese so how that translates me into
750. 1:59:47Mandarin is it's the same tool that you know that translates me into Italian so we've heard from a number of clients
751. 1:59:56that modular content is important to them getting more efficient with how we create content and how we leverage
752. 2:00:05content so it's not like it's not any more complicated for this tool to learn Mandarin Chinese than it is for it to
753. 2:00:13learn Italian it's just you take the training data of Mandarin and it learns Mandarin you take the training data of Italian and it
754. 2:00:22learns Italian you take the training data of brain waves and it does brain waves you take the training data of
755. 2:00:27images and it does images now it's listen it's way more complicated and sophisticated than that but this is a
756. 2:00:34situation where data is data is data like like once you get it into a data form that these machines can understand
757. 2:00:40using this Transformer technology you can stitch together multiple of these models and get it to do [ __ ] like you
758. 2:00:49input text so it's using a large language model for text understanding and then it's got a thing called a diffusion model for how it creates
759. 2:00:57images and this has got some voice synthesis model combined with some translation model combined with some
760. 2:01:03large language model and it's it's combined with some animation model to change the lips while I talk so it's
761. 2:01:11combining multiple models models together but the same core exercise of how you train the data and how it does what it does is all
762. 2:01:20essentially the same [ __ ] so this is all just about what we train the data on and and what we have it
763. 2:01:25do and it's and that's why we're just going to continue to see over the next three years like miraculous capability after miraculous
764. 2:01:37capability after miraculous capability I I still find it miraculous I can go to chat GPT and say write me a paragraph and it just goes
765. 2:01:50make me an image write a song is that not blowing anyone's mind other than me oh chat GPT is overrated really have you actually used
766. 2:02:04it it's not perfect okay all right check back in a year um Kyle have you seen the latest image generation using brain
767. 2:02:13waves I have not is there is there a place to go to look at them but I'm sure it's face melting hi my name's
768. 2:02:23Kyle hi Kyle they pay you for this advertisement nope they don't they should morals morals do you have morals
769. 2:02:35are you using this stuff morals or you just in here to troll cuz I'm fine either way just be creative in you're
770. 2:02:42trolling but I'm just curious if you actually use this stuff or if you're just like coming in and yelling people
771. 2:02:48that don't have qualific ifications life's tough it's tougher when you're stupid you know what I mean you know what I'm saying
772. 2:03:00anyway um will you be getting elon's brain chip no but I did get um Sam alman's um um world ID he scan Sam Alman scan my
773. 2:03:13retina that should piss you off want to see want to see my world ID
774. 2:03:27L Lites like moral morals here is going to be pissed off at [Laughter] this I think this is really cool but you
775. 2:03:38know as you said this is just an advertisement so that thing World ID verified at orb Anonymous so this is my blockchain
776. 2:03:51verified proof of human uniqueness based on my retina that I scanned at a silver ball in a clothing store in Manhattan
777. 2:04:06so so yeah um speaking of the the Elon Musk um uh what's it called neuralink um I don't think I'll be doing neuralink
778. 2:04:22because I don't think we're going to need to do surgically implanted stuff um for things like that if you paid attention to
779. 2:04:33um The Vision Pro announcement from Apple you'll love this morals okay check this out this is this is exciting okay so so
780. 2:04:45neuralink what what what they're doing with neuralink if you don't know what it is they drill a hole in your brain or
781. 2:04:50multiple holes they put in an electrode that's got 60 or 120 electrodes they they've now built a robot that that can
782. 2:04:59insert these electrodes into your brain tissue and avoid hitting any um capillaries or or veins or you know any
783. 2:05:08blood flow so they basically steer the these electrodes in between your blood vessel so that's insane but it's
784. 2:05:15basically like instant connection to you know to certain parts of your brain and then they can you can it's a computer
785. 2:05:21interface that's instant but check this out in the Vision Pro goggles one of the ways they work there's four cameras pointing at your
786. 2:05:32eyes and then in the in the goggles themselves there's this chip called the R1 chip that's doing realtime machine
787. 2:05:39learning artificial intelligence kind of machine learning analysis of all these sensors right so the way the interface works for
788. 2:05:48Vision Pro is it's it's based on where you look so so it'll pop up some icons in front of you and then you look to an
789. 2:05:57icon and when you look to that icon it'll highlight and then I think you do this with your hands and there's some
790. 2:06:02other camera on the outside that when you do this you select that icon but here's something that they're doing you you emotionally
791. 2:06:16know which icon you're going to click before you look over and click on it and so they're measuring micro
792. 2:06:25movements in your irises dilations in your irises to know which icon you're going to look at before you click on it and
793. 2:06:33they start pre-loading that application before you actually click on it so Invision proos Hardware today that's you
794. 2:06:43know probably going to come out to developers Before Christmas is essentially real time like like brain analysis of what's going on in your
795. 2:06:56brain before you even know it so you don't even necessarily need um brain implants for that level of response that's freaky
796. 2:07:13huh China prioritizing AI with the goal of something ever since you start these lives I use chat GPT every day from cooking to coding
797. 2:07:22perfect again my goal here is not that you love AI I don't give a [ __ ] if you love AI or hate it but don't be ignorant
798. 2:07:31of it don't assume that because it isn't perfect it's a waste of time I think I I think I think that's the biggest mistake
799. 2:07:41that's what this Channel's about China prioritizing AI with the goal of intelligence AG intelligentized Warfare Pentagon warrants of
800. 2:07:53course go you want to be scared go look up a company called paler so paler is a military company that builds War Logistics
801. 2:08:05software and they announced with pretty decent Fanfare about six months ago they've put a chat GPT like interface on
802. 2:08:13top of their War Logistics machine with with their Vision [ __ ] so they can basically take um imagery of a
803. 2:08:23battlefield and if they notice in the battlefield that the bad guy moves from that hill to that Hill the system will now come up with
804. 2:08:32response scenarios in real time and send those to generals so so yeah uh the warship so like here's the deal kids
805. 2:08:45like all this chat GPT [ __ ] is new and shiny Ai and machine learning and expert systems have been around for [ __ ]
806. 2:08:53decades you don't think the military's got crazy scary ass [ __ ] I was talking to I was talking to someone at DARPA we
807. 2:09:05we did we did a proposal to DARPA for for our storyvine technology and it was I I came up with a use case for being
808. 2:09:14able to use our our video storytelling thing to have soldiers on the field um kind of they could they could watch questions from a
809. 2:09:26psychologist and then they could answer those questions and they could do it in a in a non-real Time way right so they
810. 2:09:32could do it in a Time shifted way and so I'm presenting this idea to this guy from DARPA and I said and and
811. 2:09:39you know for example you could even have like the the um the doctor if if the question is you know are are you okay and and and the soldier
812. 2:09:51says yes I'm okay the doctor could look at them and and know that they're not okay based on and the guy from DARPA was
813. 2:09:58like oh no we can do a real-time emotional analysis like we the doctor doesn't need to to actually analyze that
814. 2:10:05we know if they're lying like this was like 10 years ago so like the military's been doing this stuff for ages I you
815. 2:10:14know I think I think what's different here potentially is the speed with which the advancements are going to come and
816. 2:10:21the ease with which so so historically with machine learning you had to do these kind of specialized um expert systems so you
817. 2:10:30train it on very very specific data what's happening with with chat gbt is it's a generalized intelligence so it
818. 2:10:36knows lots of [ __ ] and so so now they're taking these generalized tools and applying them to expert systems and it's
819. 2:10:44it's accelerating things so so so yeah yeah I and and I heard um in um in Ukraine um I I think there's been the first use of an
820. 2:10:58autonomous AI controlled drone that took out an enemy right so like that like I think already on the battlefield in in Ukraine are AI um
821. 2:11:11controlled autonomous war machines like like it's happening right now like yes so yes um should that be something we're
822. 2:11:19concerned about yes I mean the military hopefully is on it and this is why it gets weird about oh you know the
823. 2:11:26government should regulate this and shut it down and da well the other countries aren't going to stop this like we the the the horse the
824. 2:11:35horses have long left the barn and and I think you know meta open sourcing their large language model um you know makes
825. 2:11:44this even less controllable than it was before that happened so so yeah there's some there's some scary [ __ ] out there
826. 2:11:51but anyway um it doesn't mean you shouldn't play with this stuff even if it's [ __ ] scary how do I get this back wait I go
827. 2:12:02uh go here and I go here ah I go there all right investors bet Samsung smaller memory chip Rivals SK hinx will be a winner okay I don't
828. 2:12:15know SK hinx I don't know what that is their smaller memory chip I don't know what the implication of that is lonus sounds cool I don't
829. 2:12:25know what it's got to do with AI didn't Tom Cruz do a movie about that yeah there's been a lot of movies about about this
830. 2:12:31stuff uh uh uh uh uh would you like to play a game Reddit can survive without search company wait without search
831. 2:12:40company reportedly threatens to block Google oh interesting I the the whole thing of like the future of search is fascinating I mean I mean
832. 2:12:55you got to ask yourself if you're if you're a thinking person why did the company Google that invented the Transformer that's making
833. 2:13:04all this possible why did they sit on it for so long cuz I think they knew if we make a machine that just answers what people wants
834. 2:13:16want our search engine business is over now their search their their market share hasn't really gone down since all
835. 2:13:29this came out so I don't know that it's playing out right now and Google it sounds like what they're coming out with
836. 2:13:35with Gemini is going to be pretty [ __ ] powerful so I wouldn't count Google out right now but I think I think search just has to
837. 2:13:45change I barely go to Google anymore ever AI military drone the valkyrie xq58a is
838. 2:14:00designed to deliver weapons or assist fighter jets oh yeah I saw that thing that was pretty cool doesn't it Fly
839. 2:14:07doesn't that fly in conjunction with a piloted jet it's like a wingman for a piloted piloted plane someone said something about when
840. 2:14:17I was talking about David Shapiro they're like oh no I thought you were talking about Ben Shapiro yuck that's
841. 2:14:25good yeah face to face is the only way when troops come home I got a couple of stories there oh interesting yeah yeah
842. 2:14:34exactly that's that's what we were trying to do was can can we do something like where where soldiers were on the
843. 2:14:39battlefield it never ended up going anywhere so that was kind of a drag but whatever it was interesting
844. 2:14:48yes yes and yes too late ET is talking about too late I assume about the military stuff what's the worldcoin message I don't know what you
845. 2:15:04mean I don't know what you mean Silver Fox the reason why you should play with it Sherry D the reason why you should
846. 2:15:13play with it is because it's scary exactly this [ __ ] this [ __ ] is way scarier when you don't know what it is
847. 2:15:22cuz you can only imagine and you hear stuff like oh the war machines and the this and it's going to do that and then
848. 2:15:29you use it and you realize okay it's a tool now if you sort of play out scenarios and say wait if it's really
849. 2:15:35good at this kind of thing I wonder if it could do this kind of evil thing over here the answer is yes like like right now these things
850. 2:15:44have safety guard rails and things like that but the The Meta llama models that they put out there they don't have any
851. 2:15:50safety guard rails on them so people are out there right now building all sorts of crazy ass nasty [ __ ] does that mean
852. 2:15:57you shouldn't learn how to use it and learn where the boundaries are and learn how to use it for good absolutely not
853. 2:16:03the more of us that understand how to use this stuff and bend it even a little bit toward the good is is going to have an effect Sher
854. 2:16:12D you were a medic you were a medic in I forget I didn't see what it was in in Afghanistan in Afghanistan or one of the
855. 2:16:19wars face to face is the only way when troops come home got a couple of stories there fascinating oh medic during Desert
856. 2:16:27Storm wow wow amazing well thank you for your service Sherry you know I again back to the
857. 2:16:41virtual companionship thing I think that there there are going to be countless stories of things that are difficult for people to talk
858. 2:16:53about that for whatever reason are not comfortable seeking help and doing that with people where they can do it with
859. 2:17:01something like pie or like a you know a medically upgraded version of Pi where they are going to be comfortable with that conversation and
860. 2:17:10it's going to help a lot of people like I really believe that YouTube meta AI using brain waves YouTube okay meta AI using brain waves 240
861. 2:17:23timestamp let's go look at that YouTube oh look it's me again I can't get enough of me I'm everywhere wait um stop shut up um meta
862. 2:17:42AI using brain waves
863. 2:18:00okay So Meta
864. 2:18:14240 it's as if these AI systems are are mirroring the neural intricacies of the human brain and while other tools like
865. 2:18:22functional magnetic resonance imaging can yield more detailed images the superpower of meta's new AI decoder lies
866. 2:18:29in its speed being able to generate images from brain activity in mere milliseconds notably these visuals are
867. 2:18:38still somewhat lacking in fine details but they can still accurately capture the broader essence of what the human
868. 2:18:44eye is perceiving now for the bonus look into what this techn ology is leading up to by just connecting the dots of
869. 2:18:50progress made meta's future research trajectory is becoming clearer for one last year meta introduced a novel AI
870. 2:18:58architecture aiming to break existing limitations in order to think and act I don't know if that's the one that that
871. 2:19:05you were talking about or if that was the time stamp but yeah that's that's pretty amazing the fact that they're
872. 2:19:09able to do this in milliseconds I mean you know again think about this so if we're now at the point we can we can you
873. 2:19:16know read brain waves we can also write likely write to brain waves so someone who's blind you could you know this
874. 2:19:25stuff can see and you could potentially allow them to see or at a minimum you know in real time describe what they're
875. 2:19:35seeing and yeah I yeah so so I yeah it's it's uh
876. 2:19:49we're in weirdness people we are in weirdness we're in post SciFi this is we're in a post sci-fi situation here
877. 2:20:00people hi my name is Kyle my mind was blown again today so with ours Kyle [Laughter] welcome
878. 2:20:15yeah the images are what I wanted you to see the one on the left was the pick the right is generated yeah and and
879. 2:20:23like those images that they just showed like I've seen similar Tech to that I've like the white papers for that
880. 2:20:33Tech I saw like 6 months ago and the the generated images wasn't even close like these are a hell of a lot closer like
881. 2:20:39again I'll just go back to go look at what stable diffusion and mid Journey images looked like like 18 months ago and look at what they look
882. 2:20:49like today and take any technology where you look at it today and you're like well that's just crap that's horrible and
883. 2:20:58just play it forward 18 months to make the same technological leap that images did from when if you typed in gummy bear
884. 2:21:07it was you know kind of as bad as that video as I did or Worse where today it can create a photorealistic gummy bear like that
885. 2:21:16level of regression is going to happen with all of this Tech the other thing that people are not really talking about right now with the
886. 2:21:26exception of David Shapiro or just in in general press like people like people aren't even talking about chat GPT right
887. 2:21:34now but what's coming and what's going to what's going to what what you can already see glimpses of in in chat GPT is autonomous
888. 2:21:48action meaning right now with chat jpt you type in a prompt and you get back a response and then you type in a response and you
889. 2:21:57get back a response and it's this very manual kind of back and forth except when you use the do thing and you type in hey give me an
890. 2:22:08image and then it says generating images what it's actually doing is it's writing four distinct image prompts that then
891. 2:22:15generate those images you're not having to write that if you use Advanced Data analysis within chat GPT and you upload
892. 2:22:23a CSV and you say hey go um tell me what's interesting about this data it writes and executes its own python code
893. 2:22:31and then analyzes the answers and promps itself to figure out how to give you the answer that you want so so these tools are starting to
894. 2:22:40take their own actions you give it an imperative you give it a goal and then it starts doing a series of things that
895. 2:22:47you don't have to control you don't have to think about how to do it just does it that's going to get more and more and more and more
896. 2:22:56sophisticated and it'll be able to take more and more kind of actions on your behalf and that's where this [ __ ] can get scary
897. 2:23:07right there's there's there's an app out right now that's making a lot of rounds where it's like you can um take a
898. 2:23:15picture of someone and say go find all the pictures of that person on the internet and it will find them in crowds
899. 2:23:20like like again the surveillance uh the surveillance industry's been doing this for years but now all of these tools are
900. 2:23:28being revealed to us right because this stuff's being democratized right and it's it's being put out largely because of chat GPT having
901. 2:23:39such massive adoption in such a short amount of time all the other companies are going we got to get our version of
902. 2:23:46that out out there cuz we've got to compete right so business interests are driving this but what it means is you've
903. 2:23:53got the entire investment world and the entire development Community are all pushing in the same direction so things
904. 2:24:00are going to start accelerating further and further and further is protein Source a legit resume bullet I don't know what that means but
905. 2:24:10it's awesome the images are what I wanted to see oh the one on the right was the pick the or the left was the pick the right
906. 2:24:19was generated then we'll be able to communicate with aliens they're already talking right now about I think they're they're already
907. 2:24:26starting to do um communication with animals with with whales Dolphins um I think turkeys some someone did something
908. 2:24:33with turkeys or maybe I saw a guy that a guy spent enough time listening to turkeys that he he understands turkey
909. 2:24:39language but yeah you could just like you translate you know English into Mandarin Chinese if you have the right
910. 2:24:46input data for from dolphins or whales you could create a translator where I speak in English and it translates it into whale and then the
911. 2:24:54whale responds and it translates it back to English like that is not out of the realm of possibility I would be very
912. 2:25:00surprised if the researchers that have been trying to figure out um animal language for for decades if they're not all over
913. 2:25:11this what AI is suggested for getting into scra the internet and use things like research you don't really need to scrape
914. 2:25:22the internet I would use perplexity doai um it it searches across the internet and it provides um citations and things
915. 2:25:31like that there's also some plugins in chat GPT there's one called Web pilot there's there's certain ones but also
916. 2:25:37chat GPT now is connected to the internet so is Bard so's Bing now because the commercial of those have
917. 2:25:48safety guard rails on them there might be things you want to search for like personal information about people it's
918. 2:25:53probably going to tell you to [ __ ] off um but but the tools are out there right now like I I don't know that I'd
919. 2:26:00recommend one there might be specialized tool tools to do that but there's just all the stuff connected to the internet
920. 2:26:06right now Pim eyes is the the photo tool I used it to get Intel on people I'm dating found one was married with five
921. 2:26:16kids there you go that that that pre-ms the the harder conversation 3 months down the line right amazing Pim
922. 2:26:26eyes my brother's new book is out just got it it's incredible awesome what's it about Thelonius uh uh uh uh uh uh thing is on
923. 2:26:36a certain subject there may be 10 groups researching yes they collaborate but with this I don't quite know what that was in reference to
924. 2:26:47the free version of Charity is not connected to the internet they just told me of Charity the free version of
925. 2:26:54Charity what do you mean charity I don't know what that is all the data is there and can be filtered in seconds is chat
926. 2:27:00GPT Pro worth the money perspective matters I historically have said it's really up to you I now think chat gp4
927. 2:27:11the 20 bucks a month is is the best deal like as whoever said that was in here earlier kind of trolling me do you get paid for this
928. 2:27:19advertisement I think the 20 bucks the chat GPT 20 bucks is the best deal going right now I think it's absolutely worth
929. 2:27:25it I think I oh charity equals chat GPT what what was what did he say then um the free version of chat GPT is
930. 2:27:36not connected to the internet that's correct yeah again I think it's I think I think it's absolutely worth paying the
931. 2:27:4320 bucks for chat GPT Pro you get GPT 4 which which is dramatically better than 3.5 it's connected to the internet it
932. 2:27:50has Vision it has Advanced Data analysis it it is uh it is connected to the internet I think I said that it's got
933. 2:28:00plugins um it can generate images and then in the app version of it it has um it can talk and it can listen so it's
934. 2:28:11got like eight major modes right now that are all at some point going to kind of merge together into the same app but
935. 2:28:19all those different Mo modes in and of themselves are like staggeringly good values Advanced Data analysis on its own is worth I don't
936. 2:28:28know 50 bucks a month a 100 bucks a month dolly is at least worth 10 bucks a month the image generation its ability
937. 2:28:37to read and understand images like I don't even know the value of that yet but it's insane like just the ability to take my
938. 2:28:44whiteboard and instantly give me the notes on it and and summarize them that's Insanity um yeah the fact that it's connected to
939. 2:28:57the internet oh it's got custom instructions so you can find tuna I like I just I I think the 20 bucks a month is
940. 2:29:04is is the best deal going right now for GPT Plus Beauty and Mischief the design Alchemy of Blackman Cruz oh nice that sounds cool nice
941. 2:29:18do you like ml net I don't know ml net is that military net ml net let me go look um mln net machine learning made for net
942. 2:29:37oh it's a machine learning framework I don't know that that's that this is probably more on the infrastructure side
943. 2:29:43so I I tend not to be on the infrastructure side as much as I just want to use the tools um although I'm
944. 2:29:50I'm potentially building something for a Pharma company where I'm going to have to geek out on that stuff but I think
945. 2:29:55we're probably going to develop the initial thing we build on AWS not on Microsoft but it it looks fine it looks
946. 2:30:03cool um I can't even imagine the conversations we'd all have if we were in the same room yeah exactly well
947. 2:30:09that's that's that's what the um that's what the salon is about so come to the salon so how we doing time wise I got to go um if you have
948. 2:30:20not uh seen this yet tonight so that URL right there the salon. a that's a community that I started last
949. 2:30:30December um where we have um bi-weekly meetings so we meet every other Tuesday so this coming Tuesday is the next
950. 2:30:38meeting we meet in person here in Denver and then remotely everywhere else and we have a an online Discord
951. 2:30:45community so join the Discord Community introduce yourself if you if you want to basically if you want to hang out with a
952. 2:30:51bunch of Irregulars if you want to hang out with a bunch of weirdos that are like okay I got to figure this AI stuff
953. 2:30:56out that's what this community is about it's not about loving it it's not about evangelizing it it's about what do we do what do we
954. 2:31:05do how does it work what do I do now I'm a copywriter should I be scared oh [ __ ] yes I should okay great what do I do
955. 2:31:12here's where to start here's right it's it's all people just trying to figure this out from all walks of life from all
956. 2:31:19different technical backgrounds creative backgrounds there's Architects there's writers there's artists there's you know
957. 2:31:27Pate who works at Google on on on their uh their tpus he's in there he runs the the um the AI mechanics Guild we have
958. 2:31:38different guilds there's a writer Guild an Art Guild business guilds I mean it's it's pretty slick um so yeah so so come hang
959. 2:31:49out but you can't be a douche the first the first link that's a link tree the first link talks about what the salon's
960. 2:31:57about and what our values are and the values are things like curiosity exploration um inclusion empathy generosity like share what you learn
961. 2:32:10don't just come in there and like you know sit in the corner and do that like teach people what you're knowing Herod who's who's this
962. 2:32:19amazing designer is in there teaching people how he does these amazing animations right like it's just it's a
963. 2:32:25it's a generous community so if you're stingy it's not the group for you if you're generous and curious and adventurous that's what it's
964. 2:32:37about wow well that was very very intense well we're trying to change the world here people don't you understand don't you get it
965. 2:32:51yet I think Ted Ted made this one Ted looks at life the varsity jacket oh man retr punk thank you again for all your help you've been an
966. 2:33:01inspiration yeah retr Punk's amazing all right everybody I think it's time to wind this puppy down
967. 2:33:17IR regulars present for share [Laughter] bear thank you Tobias good to see you hair growth geek what's happening good to see
968. 2:33:29you um Sherry D thank you Kyle thank you thanks for being in here you guys are awesome jrc good to see you Tobias Joker what I was trying to
969. 2:33:39say is Imagine oh wait could you repin that again whoever just pinned that have a good night Irregulars
970. 2:33:47who else we got in here will good to see you I'm in there trying to be weird and tongue and cheek semi offensive no
971. 2:33:55that's no you're I I don't ever take your stuff as offensive Joker what I was trying to say is imagine all the
972. 2:34:01research on cancer trained on AI and ask it questions yeah and not just be able to ask it questions like have it come up
973. 2:34:11with correlations between you know what's what's causing it or how to cure it right that kind of stuff's going to start
974. 2:34:19happening Source Camp thanks Kyle good night good night have good having you in here cuno good to see you forever hooked
975. 2:34:26steo peace buddy retr Punk great to see you digital Gods good night Kyle you should think about letting some people
976. 2:34:34join you on live for Halloween November 30th yes oh so that's another thing so so I'm strongly considering that on uh November 30
977. 2:34:47um the the anniversary of the launch of chat GPT doing like a 24hour AI learning laboon yeah I think I think we're going to have to to
978. 2:34:57coordinate with the Irregulars so I can take some breaks in there um thonus the Tailwinds of AI are starting to kick in
979. 2:35:04for Microsoft says Jeffrey Brent thrill oh that's interesting well did you see Sam Alman just said um chat uh open AI
980. 2:35:14just hit 1.3 billion in Revenue 1.3 billion in Revenue that [ __ ] hasn't been out and they haven't even they haven't even really kicked off
981. 2:35:23their Enterprise sales yet right like it's it's [ __ ] crazy so remember remember like two months ago in here
982. 2:35:33someone was like they're spending $700,000 a day they're going to be out of money by the end of 2024 it's like I
983. 2:35:39don't think so and now it's like they're making some butt Bank it's crazy all right everybody um great seeing you all uh I appreciate
984. 2:35:51y'all a lot for being in here I'm sorry about my hair but it happens what are you going to do peace