Jul 8, 2024 - Unleashing Creativity: How AI is Transforming Artistic Expression

Transcript

1. 0:00you ready to [Music] [Applause] [Music] [Music]
2. 0:21sing I'm just tuning dude he doesn't know the difference between tuning and playing [Music]
3. 0:42[Music]
4. 0:56you ready chy every time I see you now get that look in my eyes every time I see your mouth I hear that
5. 1:10smile in the early Misty morning GL I heard the engine turning the old for outside
6. 1:24[Music] you convince me again today you're leaving this hotel looking
7. 1:39for someone else's Golden Ring all right something's glitching Champ's tuning up champy is tuning up CH Champ's a mess tonight what
8. 1:56do you think champy [Music] [Music]
9. 2:13a freedom came my with that [Music] night just like a jet plane in and out of side I was hauling ass a million miles
10. 2:27an hour war and hard out here when they came into the [Music] station they said I was B Beyond
11. 2:45repairs but I got no quals with my situation say here I am [Music]
12. 3:00say Sheree Shere Shere won't you dare to say Shere Shere Shere won't you dare to say Shere Shere
13. 3:15Shere uhhuh yeah leave a message and your number please take the time to want to satisfy me
14. 3:30take all these old fantasies and send them care of [Music] me good evening good people it's Monday do we get
15. 3:44philosophical Kyle tonight I do not know I don't plan these things it just tends to be that on Monday nights maybe it's
16. 3:52CU I haven't been paying attention to all the news I'm not all full of news and I'm just full of hot wind [Music]
17. 4:12oh yeah good evening good people we get Monday Kyle I have freshly glued my hair so we should be okay on the hair tonight I
18. 4:27know that's always a big concern a lot of lot of people a lot of people come here they're like I come for the hair I
19. 4:34stay for the madness some people quite the opposite I I come for the madness and the hair I just [Laughter]
20. 4:46tolerate it looks like image generated hair AI generated Hair It Is this is a uh a masterful collaboration of four different
21. 4:57tupes cleverly s together to seem like natural hair with an extremely large forehead it's taken me decades to get it like
22. 5:12this the amount of tape and bobby pins it's it's like a Hollywood uh special effects crew in here getting this hair to look like this really
23. 5:23something really really impressive Jenny's a part of that and Bucky Jenny and Bucky are my two primary hair people they're
24. 5:37amazing occasionally Sarah will jump in [Music] fantastic we've got a pool going on the hair
25. 5:54nice so if you're new here my name's Kyle Shannon this the AI learning lab all of a sudden that sounded like a like an NPR kind of [Music]
26. 6:05like hey it's Kyle Shannon little little background guitar like an Off Broadway [Music] show I moved to New York City in
27. 6:231987 my first job was at FAO Schwarz selling yo-yos at the top of the the [Music] escalator one of my favorite memories of
28. 6:39that job was Walter kronite coming up that escalator with his grandson and I was doing some yo-yo tricks and Walter
29. 6:48kronite looked at his grandson and said whoop walking the dog [Music]
30. 7:20that's a true story actually Mr it Kyle Kyle we missed you guys thank you so much I think two days is too much [Music]
31. 7:43did that job take you around the world oh Joker coming in heavy with the dad jokes [Music] tonight you had to be careful with that
32. 7:58cuz the Y y- Yos that they that I was selling were these wooden yo-yos which are fine they were fine they weren't
33. 8:05like Tom [ __ ] you know highly balanced machined aircraft aluminum yo-yos and so I would often like Zing the the yo-yo out like right to
34. 8:17someone's face and then stop it right before it got there there are a couple of times where the spindle in the middle
35. 8:24had worn down and it split the yo-yo in half so that was that was they they had a they had a bigger liability with me selling yo-yos than I
36. 8:34think they realized I met Keith Richards doing that it's pretty [Music]
37. 8:53good yeah those were interesting jobs the FAO Schwarz jobs cuz they mostly hired you know like Broadway actors who are not
38. 9:04working Ana like and share and follow the host thank you so much I agree like share follow the channel share this
39. 9:13thing here's who you should share this Channel with anyone who you think will be pissed off if you share a Channel about
40. 9:22AI speaking a liability let's [ __ ] with your friendships you know you know Jenny listen um look I really appreciate you
41. 9:33sharing that live uh you know the guy was interesting I mean his toupe was fantastic but um all he talked about was
42. 9:40artificial intelligence and you know I am personally affronted by by artificial intelligence so um I would appreciate it
43. 9:49if you do not ever share that person ever again so how long did I watch it well I watched all two and a half hours of it it was fascinating
44. 9:59yeah I like chat BT now but I am I'm personally offended by artificial intelligence but I I do like it a lot
45. 10:05now it's pretty good isn't it you know I can no longer tell if this is a real story or not no the the FAO short stuff that was all real
46. 10:23[Music] yeah you know I was working at fo Schwarz and trying to get an agent trying to get
47. 10:40jobs I auditioned for this thing called the drama [Music] project I had to audition to get in [Music]
48. 11:07I did this monologue from Golden Boy boxing film or boxing [Music] play and I had to audition to get in and
49. 11:20so I decided to pick the prettiest girl in the room and I did the monologue straight to her was about a girl
50. 11:30turns out that pretty lady is now my [Music]
51. 11:44wife turns [Music] out when I did that monologue to her she turned to the girl next to her and she said uh-oh this looks like trouble
52. 11:59[Music] wow so sweet fantastic Love Story also true did I miss a story about Mrs Shannon yes you
53. 12:20did then a group of us broke off from that group started a theater company and we only did original plays which is stupid cuz
54. 12:35no one's ever heard of them and nobody really wants another play in the world but we'd Workshop him in Vermont in a
55. 12:44barn I got a barn let's put on a show in Warren [Music] Vermont and then in the winter we take the plays that didn't suck
56. 13:01[Music] and we perform them on 42nd Street a never before seen picture of Kyle getting his hair done in
57. 13:12Irregulars let's go see this I like the idea I like the idea people uh there's Irregulars new activity
58. 13:28see this is a little behind the scenes I don't know how they got the camera feed from uh getting ready for the evening
59. 13:36but yeah that's that's pretty much how it goes [Music] [Applause]
60. 13:50[Music]
61. 14:14living in New York City is pretty amazing if you've never done it it is it is remarkable it is weirdly the biggest city you've
62. 14:27ever experienced and the most intense you've ever experienced and it's a weirdly small town I remember the first
63. 14:38day the first day I was living in New York City I'd been in a play the summer [Music] before at the mount Greta
64. 14:52[Music] Playhouse we did three Penny Opera and it was was directed by Paul Giovani who was a Broadway [Music] director
65. 15:07[Music] and and then I did good enough there that he offered me an equity job in Washington [Music] DC at the end of the summer but the
66. 15:21problem was I had a full ride of full scholarship to the Alabama Shakespeare Festival so 2 weeks before I was supposed to go to the Alabama
67. 15:29Shakespeare Festival I was like [ __ ] it it's an equity house I got my first paid job I'll get my Equity card so I'll do
68. 15:37the show and I'll move to New York City and so it was the uh crucifer of blood a Sherlock [Music] Holmes murder
69. 15:52mystery and I played a young lap boy and a Pakistani [Music]
70. 16:07conspirator I learned my Pakistani accent I learned how to be a a serving boy for a rich old man and then I moved to New York
71. 16:23City and my first day living in New York city so I moved into an apartment in East Harlem on 114th and 1st Avenue with
72. 16:31some actors that I had worked with the summer before and I decided to take the subway to Time Square and so I'm walking around
73. 16:40Time Square in the middle of rush hour I mean it's [ __ ] just chaos right and I hear hey it's like huh that's weird sounded
74. 16:52like someone said hey Kyle and I heard hey [Music] Kyle and it it was John peel me who's a playright he wrote Agnes of
75. 17:03God and he was in he was in the three Penny Opera at the mount Greta show over the summer and so you know I saw him and
76. 17:14and we ran together and hugged and like in the middle of Time Square with like thousands of people my first day in New
77. 17:21York City I run into someone I had a a deep relationship with that was really cool [Music] and New York's full of [ __ ] like [Music]
78. 17:36that so yes you're going to get a philosophical Kyle [Laughter] tonight all right what do we want to
79. 17:51talk about people we can talk about pickling cucumbers [Music] was in a live this afternoon two young women had a question dilemma they said
80. 18:04let's use chat GPT There's Hope oh there you go I'll tell you what if if you haven't if you haven't sort of crossed the the
81. 18:14Rubicon if you haven't crossed the threshold with chat GPT where whenever you think oh I should use my brain and come up with
82. 18:27something if you haven't cross the threshold where you then correct that and go oh no [ __ ] that let me just try
83. 18:32chat gbt first that's a really fun threshold to cross because you still get to use your brain to come up with the ideas and
84. 18:45curate the ideas and refine the ideas but you can just live at the sort of ideation level that much higher kind of
85. 18:53for me higher fulfillment rather than sort of running down all the Tactical detail rabbit holes that's a that's a fun threshold to
86. 19:01cross would you be on my podcast tell me three things absolutely Tim Lord just uh DM me happy to do I have to come up with three can I use
87. 19:13chbt thing one uh never go on a podcast called three things um thing two uh uh chat GPT told me to delve deeper into
88. 19:27the tapestry of the unfolding Paradigm of podcast speaking thing three uh yeah happy to as you can
89. 19:41tell I'm a scintillating guest I quit using Google in my brain and and my brain thanks to chat GPT exactly exactly I got to tell you if you
90. 19:55haven't um if you're still using Google and you still are under the delusion that it works start using chat
91. 20:02GPT Google gets long in the tooth really [ __ ] fast like I I I I sincerely cannot believe how bad Google search is these
92. 20:14days it's just it's just not usable not usable by the way little coffee with milk in case you're wondering it's it's
93. 20:23not it's not Denver tap water all right I find myself telling people that it was chat GT but I had to rewrite
94. 20:32it five times yeah I got a note from from Allan today um asking about he had heard that I wrote a book using Claude
95. 20:42on the channel here and he goes he goes I heard you put in like a 2hour video and it wrote a book and I was like uh no
96. 20:50I put in a list of quotes and then I took 2 hours 2 and 1/2 hours and about 400 prompts to get it to write a book
97. 21:02so we're not quite at the push a button get a book [Music] yet new new images in Irregulars it
98. 21:16seems new activity will act for I had a little lighter hair back then I was I was more blondish than brownish and it was a wooden yo-yo
99. 21:29like a blonde but that's what it looked like you went up the escalators and I stood right at the top of the escalator right
100. 21:37up here twirling my will act for food that's pretty much it and this is the glamorous life I lead today this is where it
101. 21:49started where it is today oh yeah oh yeah people that's comedy right there I'm finding
102. 22:04the combo of chat PT Gemini and Claud are great for e-commerce yeah so I'm I'm experiencing the same thing where
103. 22:12although I haven't been playing as much with Gemini I it their latest release got me really confused and I like I I
104. 22:19don't know I just sto it pissed me off cuz they're kind of all over the place um so I haven't been using that
105. 22:25but Claude is really really good chat GPT is really really really good and but they do different things better so
106. 22:32bouncing back and forth between them I think is a really good strategy right [Music] now there's also uh Silver Fox Lori um
107. 22:42asked me a question for my AI walk and talks which I haven't gotten to yet but but it was uh it was basically about
108. 22:52um if you want to do something with a particular category of tools should you learn lots of tools thinly or should you
109. 23:02pick one and and go really deep it's a really good question um I think it depends what mode you're in so I I see AI literac literacy being
110. 23:15a journey of kind of three different modes we talk about this on the AI Salon a lot which if you haven't checked out the AI Salon
111. 23:26um there's the address go there and check that out go to the click on the top link there it's the community um
112. 23:40and the three modes are play first mindfully create and generously lead and how I like to think about The Learning
113. 23:50Journey is play first is when you're in sponge mode you're just trying to learn so if you're like oh I want to get into
114. 24:00image generation don't just pick a tool and go deep on it play with as many as you can or if you're doing the same with
115. 24:06video or song generation or even large language models play with as many as you can and you know some of them are going to have
116. 24:14free modes some of them won't you don't you don't want to do all sorts of subscriptions you just want to play you
117. 24:20just want to get a taste for as many as you can and you'll see that oh this one's kind of good at this kind of thing
118. 24:25and that one's good at that kind of thing and you'll start to figure out it's not not like any tool is is better
119. 24:31than another it's that any tool for a given use case and your personality are going to suit you better than
120. 24:42others and the only way to kind of learn that is to try them on it's it's kind of like picking schools for your children
121. 24:49if you've ever had to do that where if you listen to other parents which I made the mistake of doing a couple of times I was like oh
122. 24:58Bob thought this was fantastic this has got to be a fantastic School let's put our kids here and see if it [ __ ] up their
123. 25:05life and it did you you you really have to try this [ __ ] on for size yourself right you can't just say oh someone else
124. 25:12loved this it must be great no no sometimes you'll try a thing someone else said was great you'll hate it and
125. 25:19then you'll try something else that they hated and you'll love it so in that play first mode cast as wide AET as possible
126. 25:28and then once you've got your head around okay I kind of like this tool or or you know like like there's subtleties
127. 25:35like like for art mid Journey if you want high quality images is there's nothing like it unless you want to produce text and
128. 25:48then you need to use um dolly3 within chat jpt or um what you call it uh ideogram those do text really well so
129. 25:59you kind of learn subtleties right then once you decide on your project then you narrow it down so you start really
130. 26:06wide up here trying everything figure out the one or two tools that you really like narrow down and then when you do
131. 26:13your project Run Deep for that project and the generously lead part is share what you've learned so in this
132. 26:23process of sort of learning wide and narrowing down and building something share what you're learning and
133. 26:29share you know what you learned in the play first mode share what you learned as you're building and you will
134. 26:37gain um trust and a following and a reputation for somebody who is thoughtfully building things with AI and a leader saw an article
135. 26:50today that we are most creative don't care what others think at age four rather than seven oh interesting
136. 26:58interesting as we talked about before yeah that makes sense yeah it's generally around age five where something happens where you
137. 27:10realize the world is dangerous and that's when you start to shut your creative brain down yep and what what seren is talking about
138. 27:24what Amelio's wife is talking about there is what people here have consistently experienced is once you sort of break through the what's
139. 27:37this AI stuff all about I don't get it I don't get it it's like gole oh champ champy champ champ
140. 27:51champ champy come sing come on you want to come sing [Music]
141. 28:14were you protecting us from the evil people yeah it's a good dog oh it's a good dog oh good dog um what was I talking
142. 28:26about hey AG five we go to school yeah it could be [Music]
143. 28:41school I wrote a scene in a oneman show about the thing that happened to me when I was five and I compared it to the thing that
144. 28:49happened to my dad when he was five and how we reacted to the same situation [Music]
145. 29:04and that gave me insight as to why I became an actor and he became a pilot life's [Music]
146. 29:23weird all right constantly being taught to color inside the lines yep all right what do you want to talk about people
147. 29:37here what do you want to dig into you want to dig into anything you want me just keep rambling I can keep rambling a
148. 29:43pilot wow amelo has always wanted to be a pilot yeah my dad was a was a pilot in the Air Force he flew um Huey helicopters in
149. 29:55Vietnam he flew uh he was co-pilots on c130s the big cargo planes he trained on t38 t38 training
150. 30:11[Music] Jets I think he might have flown c c 141s see4 141s yeah I think 141s the
151. 30:25big jet cargo planes yep I thought about being a pilot but with my ADD brain well I'd probably been okay being a pilot cuz it's a hyperfocus
152. 30:41thing but the boring [ __ ] in in between in the the checklists at the beginning I mean I know you want the
153. 30:52airplane to stay in the air and you should check all that [ __ ] but I just want to go fly I would not have been a good commercial
154. 31:02pilot I think I'd have been a good fighter [Music]
155. 31:18pilot you are were an actor I was I have a fine arts degree from Penn State University in performance they called it
156. 31:28I don't know why they called it performance it's [Music] acting then I moved to New York City out of school started and ran a theater
157. 31:39company got tired of doing free theater then I wrote seven screen plays in two years started a fake management company
158. 31:48cuz we didn't have agents we got good enough at being fake agents that we got real clients [Music] and then uh in the mid90s when the web
159. 32:01came along I started one of the first online magazines and one of the first digital web agencies and then ended
160. 32:10up building that company from like two people to 2200 people in 5 years so it was pretty [ __ ] surreal to go from
161. 32:18being an outof workor actor to being a chief creative officer of a 2,000 person agency building websites for the biggest
162. 32:26brands in the world we are Penn State exactly worker B worker B knows we are Penn State actually that was my first night
163. 32:40at Penn State my first night at Penn State we went out drinking of course or we might have gone out I think
164. 32:54we went out and got beer and we drank all night we stayed up all night and then we were like hey let's go break
165. 33:00into Beaver Stadium and so so we went and we broke into Beaver Stadium and then we went out and hung
166. 33:08out at the 50 yard line in the middle of Beaver Stadium and watched the sunrise that was pretty cool that was I've had a
167. 33:15couple of really good first first day experiences at places New York City Penn State
168. 33:30[Music] my daughter's an actress she's in Paris does combat video editing dance singing Lexie Baker nice that's awesome
169. 33:45[Applause] Joseph you know it's funny I think that um the acting the the creative arts are about to get
170. 34:02decimated and then I think they emerge a couple of years from now um in a really powerful state but it's going to be completely
171. 34:17different I think what what gets decimated is the current business model I think voiceover actors are that like
172. 34:24that industry is just going to be obliterated but I think special effects like just just a lot of the jobs are going to be [Music]
173. 34:39different but I think if we if if we go more like a 5year Time Horizon here's how I think about [Music]
174. 34:54it the AI stuff is going to automate a lot of the Tactical execution including in the creative arts and so just all the [ __ ] that used
175. 35:07to be done by people is now going to be done by these machines and it's just going to be a handful of like creative
176. 35:12director types that are conducting all these tools and they'll do all that [ __ ] and I think that there's going to
177. 35:20be a real like a dramatic shift toward authentic experiences so theater live music um uh Escape rooms um just just in-person physical events
178. 35:36that that a lot of this online stuff is going to get so saturated with content that people are going to have a
179. 35:43hunger for connection and entertainment and so I think there'll be a rebirth of a lot of Need for talented people to do
180. 35:53interesting [ __ ] um but it's I think it's going to be an ugly 5 years and to be quite honest the the you know it's funny
181. 36:03everyone's talking about oh the you the creative arts are all [ __ ] creative arts have always been [ __ ] I'm a member of Actors Equity the
182. 36:12the union or I was the the union for stage acting in New York and or all over the country but you know the you get a
183. 36:21bro Broadway contract there um they have like a 95% unemployment rate like this is back in the'80s nobody needs another
184. 36:34actor if I'm like well I don't want to do that role you're not paying me enough they'll be like all right there's 10,000
185. 36:40there's literally 10,000 people behind you that that would take you out for a walk- on roll in this play that you're turning down
186. 36:54like it's a brutal industry anywhere where you've got an inequality of amount of jobs for people who want
187. 37:03those jobs it's going to be [ __ ] up and the Arts is that everybody wants a gallery show everybody wants a hit song
188. 37:12everybody wants a stadium tour everybody wants to be on Broadway how many people get to do that six relatively you know 3 mlion million
189. 37:27people a year mve to New York to become an actor 22 make it and most of those 22 have been pound on the boards for 20 [Music]
190. 37:46years I'll tell you what though you want someone who works their [ __ ] ass off hire someone who's been in the theater [Music]
191. 38:05if you hire an actor they can talk for one [Music] thing but they also work their ass off you can't you can't be an artist and not be an
192. 38:17entrepreneur so theatricals as crazy that [ __ ] crazy as they are they'll work their ass off they tend to be smart they
193. 38:25tend to be critical thinkers they tend to be [Music] talented read an article on AI chatbots in interpretation of pathology
194. 38:41reps brought oh brought to mind Story vine oh interesting interpretation of pathology reports interesting yeah you know the whole idea
195. 38:57behind Story vine so Story vine I'm the CEO of this company Story vine that's a an automated video storytelling platform
196. 39:04that um most people assume is AI they have for 12 years it's not it's more Aristotle than AI so the idea is that we figure out
197. 39:15what's the story structure and then we prompt someone with a series of questions that creates that narrative
198. 39:22Arc and then those clips go up to the cloud 5 minutes later you got a fully edited video so for doing things like explaining
199. 39:29things like that it was a Jamma Network article very cool Firefly thank you I'll check it out there should be a want to be a Prof
200. 39:39professional sports player Union yeah exactly same thing Sports is another Arena where you know you can be [ __ ]
201. 39:47massively talented I played one year I worked with this guy that owned a jewelry store and he was a at at some point in
202. 39:57his life was a semi-professional tennis player and he was like in the he was like just below the top
203. 40:051,000 and I played him all summer and he just destroyed me every [ __ ] game except at the very end of the summer I I
204. 40:13won one set from him and it was it was like IID [ __ ] won Wimbleton CU cuz he was so good and I was so bad but I got really good but
205. 40:22um he was saying in sports the the difference between being being out of the top 1,000 and in the top 1,000 has nothing to do with
206. 40:35skill he said once you get to a certain level in tennis and frankly this is any sport once you get to a certain level
207. 40:42but tennis is particularly acute it's no longer about the skill it's all about the head game it's all about the head game he's like why do
208. 40:53the same players end up in the finals year after year after year after year part of its talent but I don't know if you've been
209. 41:06watching Wimbleton like the the level of play is just [ __ ] insane and why is it that the top ranked players somehow find a way to win that
210. 41:20point or Stave off you know a match point three or four times before winning it in five sets it's all
211. 41:34head the Arts is the same way quite frankly cuz the Arts isn't about Talent like you assume at some point at some
212. 41:44level everyone's Got Talent who works the people that go to the parties the people that work the room the people that are always in the
213. 41:55conversation how do you stay in the conversation that's what being an actor is all about and they don't teach that in acting [Music]
214. 42:08school I studied with Bill esper in New York City William esper and it was a 2-year program and the there was some [ __ ] talented
215. 42:23people in my class like [ __ ] CRA you had to audition to get into this class class it's like a master you know
216. 42:29postgraduate Master New York City acting class 2-year program and we had some [ __ ] just crazy talented
217. 42:42actors and they weren't the ones that worked it was the it was the actors that were good but knew how to work the room knew how to be in the
218. 42:54conversation knew how to smoo an an agent knew knew who the producers were at any party that they were at were
219. 43:01always in those conversations those were the ones that worked no one ever taught that at least not anywhere I [Music]
220. 43:16studied and I think a lot of actors get really bitter because of that cuz they're like I'm more talented than them and they you might be
221. 43:27you know it's funny I this is where okay let's get back to AI for a second the reason I am so passionate about the salon the AI
222. 43:42salon and the reason that I think this is the future of work this is the future of how we're all going to work not it
223. 43:48doesn't have to be the AI Salon but something like it if you assume that AI tools are going to get so good that everyone can do
224. 44:00every skill maybe they're not going to be that good okay let's say everyone can do every skill at 80% of a current expert right so okay
225. 44:12maybe they can't do it at 100% but if I could be like 80% as good as a Harvard MBA at business Consulting I might not need to hire that
226. 44:25Harvard MBA who's a [ __ ] I might hire someone else who I know from a group like this where I've got a trusted relationship with them so I
227. 44:38think that skill that actors have to have is going to be a skill that we all have to have over the next 3 to 5 years
228. 44:49and so getting your ass into a community where you're actively learning and exploring and asking questions and
229. 45:00teaching other people and building [ __ ] and throwing pasta at the wall and just all of it it's not about the work that you're
230. 45:11doing in there it's about the relationships that you're building in there because at some point one of those
231. 45:17people's going to be in a position where they have to hire someone and they're going to go well everyone can do this
232. 45:28work who do I want to work with which is different than how we hire today how we hire today is who's got the skills to do the
233. 45:39work what if everyone had the skills that's the world we're moving into then you hire someone for a different reason
234. 45:46you hire them because you trust them you hire them because they're not an [ __ ] you hire them because you know they do
235. 45:51good work you hire them because they make you laugh you know that old adage don't be a dick you don't have to be an [ __ ] in
236. 46:04this world you can make money withing and I think that actually gets really really important in the in the
237. 46:12coming years so all right what are we doing you ready to start we're 50 minutes in I just put the guitar down
238. 46:22Champ's looking at me like what's going on what's going on Dad what are we doing here networking is always important yep it really
239. 46:38is which is I think I think one of the reasons that LinkedIn is feeling so Antiquated and weird lately is that everyone's taken the
240. 46:51course on how to write the cold Outreach you know thing and now everyone's using chat GPT so you know six times a day
241. 47:04it's like hey Kyle really love what you're doing with Story vine there I think it'd be great if we got together
242. 47:09and expanded our networks for Mutual benefit no how about no what's the mutual how about this how about you give me some benefit over a
243. 47:26sustained period of time like three or four or 5 months or a year or two or 10 and then when I need whatever the [ __ ] it is you're
244. 47:38Hawking I'll reach out to you hey Kyle just wondering if you uh were looking for any new clients oh no I've got all that I
245. 47:53need sorry full up on on needing new clients of course I need new [ __ ] clients you think I'm going to reach out to you you pin
246. 48:08head we specialize in helping companies like yours get new leads new hot
247. 48:23leads he seems a little bitter he's he's he was a little aggressive on uh on LinkedIn there didn't you think it was
248. 48:32it was I like LinkedIn I I think it's a professional networking site I mean I don't know I mean sure I use chat
249. 48:40GPT when I reach out to people hey Bob I noticed you uh went to Penn State too wink wink trust relationships
250. 48:57providing value not being a douche get in communities where you're actively
251. 49:11interacting don't just come in here to like what can I who who's going to hire me no one I guarantee you when Kelly Camp gets
252. 49:26a big job for her AI agency and she needs five or seven people cuz it's a big job there's a lot of moving parts to it she won't be going to
253. 49:40LinkedIn she'll be like all right who have I worked with in the past year and a half that I've been on my AI Journey
254. 49:47that have added value to my life that's who she'll hire okay I know that one can do this I know that one can do that I
255. 49:54need someone to know something about politics oh Daisy Thomas really good for that right that's how that's going to
256. 50:02happen this woman works thank you so much for making my night better oh thank you so much I appreciate that I you know
257. 50:12sometimes sometimes just [ __ ] ramble so um I rarely am all that conscious about what I'm saying I'm often barely conscious
258. 50:25period so if this is resonating and and making uh making a difference thank you thank you for saying that cuz I don't
259. 50:32always know yes I will be calling my peeps here says Kelly Camp exactly why does Kelly Camp come here every night oh sure for the hair
260. 50:44yeah comes for the hair stays for the engagement because as I'm rambling I don't know if you noticed this there's a
261. 50:54whole world of activity going on in the comments here here and it also extends over to the AI Salon where this group
262. 51:01the AI learning lab we have our own club over here called the AI learning lab Irregulars and so as I'm staying stupid
263. 51:10[ __ ] and talking about my hair people and my makeup people and people want to make a funny image of that they can
264. 51:16share it and so what's happening there is people are learning people are self-expressing people are leading
265. 51:25people are showing other people oh you can respond to things in real time and they're building relationships that's what this Channel's
266. 51:35about the fact that I talk incessantly with no real destination in mind I've come to learn actually doesn't
267. 51:50matter like sometimes it's brilliant and awesome sometimes it's a complete was of time but it's not it might be a waste of
268. 51:59time if what you came here for was Kyle's going to teach me about artificial intelligence then it might
269. 52:05have been a waste of your time but if you're here and while I'm rambling you kind of lose your train of thought and
270. 52:12you connect with someone else in the chat here or you just go off and start playing with the tool but you're in the conversation this is
271. 52:22valuable You're Building relationships You're Building skills You're Building ways of thinking you may occasionally be [Laughter]
272. 52:37entertained very small rocks rarely occasionally entertained so just
273. 52:51be just be a human be connected be empathetic be generous um be curious be brave do you know how many people in
274. 53:05here over the past year and a half have come in here and be like I don't know anything about it yeah what do I
275. 53:13do and what's what's cool now is that there's enough people here that when someone comes in and says that they kind
276. 53:19of take them under their wing they're like go over to the is Salon do this do that Kyle eventually answer your
277. 53:23question or not just keep coming back and how many of those people have have kept coming back to this thing and three or six months
278. 53:35later having started from I don't know what to do I'm clueless are now like uh I just got asked by like the VP of my
279. 53:45company to give like a talk on AI and I don't really know anything but should I do it [ __ ] yes you know how many people have gone
280. 53:54from like the annoying person in the lunchroom talking about AI that nobody wants to hear about to all of a sudden
281. 54:02running seminars or running like an AI working group in their company a lot that's what this Channel's all about
282. 54:10I mean not not necessarily getting promoted but just like being in the world learning this [ __ ] it will pay dividends as however
283. 54:20behind you think you are you're way ahead if you're willing to sit here and listen to me just [ __ ] Babble
284. 54:31nonsense you're like I think somewhere in here there's something I could glean if you're in this conversation at
285. 54:40all and you're like just curious like well what is this chat GPT thing let me go try it out you're way ahead of where most
286. 54:49people are 47% of people in America have not even heard of chat GPT haven't even heard of it if you do the math of the report this
287. 55:02is from a report that Ethan mik shared like a month ago if you follow the math of the report we're in single digits of anyone who's
288. 55:13actually using chat GPT with any regularity and then probably some like a third of those people actually use it
289. 55:23enough and are on and are Adventure enough to actually know what's coming like know what's coming not like
290. 55:32oh you know these things could get better at some point no no like if you use this stuff a lot you you can sort of
291. 55:39track where it was a year ago to where it is now and if you assume that it's not ever going to get any better than
292. 55:44that for the next 2 or 3 years you haven't been on this planet for the past 50 technology keeps advancing
293. 55:56and AI is accelerating that advancement so so you're early so just be in this conversation don't be a dick Learn out Loud share what you're
294. 56:12learning support other people ask questions ask for help Kyle and Irregulars know more about how to deploy gen AI than 150,000 Folks at
295. 56:27world's largest AI oh agbiz interesting yeah cuno cuno works for a very very large company so
296. 56:41she is she is speaking from experience she's an irregular and what she's saying is people in here are smarter about
297. 56:51AI than most people in her 150,000 person company because there's no curiosity there's fear in fact I'm going I got invited to
298. 57:05speak in Washington DC in fact the AI Salon is hosting a Flyin on July 24th of 12 AI entrepreneurs I guess
299. 57:16we're going to go talk to Congress I don't quite know what it is yet but in in talking about the strategy
300. 57:24you know we we're really AR articulating what the AI salon's about and it's about Gathering Ai optimists and creating a place for those
301. 57:40AI optimists to be able to jam and what what was conveyed to me is that Washington is approaching AI from a place of fear
302. 57:56what's the antidote to fear curiosity optimism Adventure that's what this group is that's what this channel is
303. 58:12about it's funny early on I got a lot of like what are your qualifications Mr Shannon if if I'm in this AI learning
304. 58:19lab uh I don't see any uh certificates uh from Stamford behind you uh it looks like bottles of wh Whi and Pokemon dolls
305. 58:29um just wondering what your qualifications are sir I think it is my lack of qualifications that are my
306. 58:47qualifications DC reminds me Kyle and I have a bunch of stuff to show you soon exactly lunatick we need to connect this
307. 58:54week that's I I am I am the ball dropper so thank you for reminding me yeah we need to connect for sure I'm a
308. 59:00subscriber now I already feel smarter fitty Min thank you so much I appreciate you subscribing so I've got to figure out
309. 59:16what I want to say to some Congressional people that probably don't want to hear from
310. 59:30us are just sitting there thinking like how can we protect the American people from ai ai is going to kill us all what do we do run lock it
311. 59:44down hopefully they don't make that face but I'm going in fully expecting that but I I think I think what I'm I'm going to talk
312. 59:57about is my experience with the people in here and the people at the AI Salon people who have in being
313. 1:00:13willing to have the courage to say [ __ ] it I'll give it a shot are completely Reinventing themselves and rediscovering themselves
314. 1:00:23so I think what I'm going to talk about is as you think about AI legislation think about it from the point of view of
315. 1:00:32what if this could improve things not just how do we protect people from it from the risks yeah yeah yeah we got to deal with the
316. 1:00:41risks understand that but don't just legislate the risks legislate the opportunity and the only way you can legislate the
317. 1:00:50opportunity is to [ __ ] use the tools nobody knows how to use these right now there is a very clear distinction between what ai's been for
318. 1:01:05the past 30 years and what it's been since chat GPT came out generative AI is different generative AI does for AI what
319. 1:01:19the worldwide web did for the internet it made it accessible to everyone you don't don't have to have a Stanford
320. 1:01:27mathematics degree to be good at generative AI um someone just pinned Jim Ross three Three Mile Storage management Jim Ross is the
321. 1:01:40epitome of someone who is just kicking ass and taking names with generative AI in the storage management business Self Storage storage
322. 1:01:52units he's making music he he showed me a a kick video he's starting he's basically becoming like a media
323. 1:02:00company and he's becoming known and trusted for being a thought leader in his entire industry to my knowledge he didn't go to
324. 1:02:10Stamford for theoretical mathematics possible I don't know his background but I'm guessing not he's completely transforming his
325. 1:02:23life and his business because he's willing to have the courage to go hauh let me give it a shot and that's how he talks it's funny
326. 1:02:32he's like he's like hey Kyle how's it going I guess I'll give it a shot he doesn't talk like that though uh
327. 1:02:43an AI Enthusiast AI Enthusiast Jam sounds amazing that's why I love this group so much yeah exactly that's that's
328. 1:02:51what it is by the way whoever posted um Jim Ross's uh comment I actually didn't read the comment I was just commenting
329. 1:02:57on Jim Ross so if you can pop that back up there I'll actually take a look at what he said so yeah and if you've if you've been to
330. 1:03:13that but you're not active in it you should you should go there and you should reconnect and you should
331. 1:03:18ReDiscover people and re reconnect today the Utah Department of Commerce launched the office of artificial intelligence oh fan fascinating
332. 1:03:29fascinating and hopefully Jim I mean listen if you like it it it would probably behoove you to find out you know who's in that
333. 1:03:43and get active in it because you could probably be an incredible contributor to that um I are you I think you're in Utah correct um
334. 1:03:57because if they created the Office of artificial intelligence as a we've got to save the people from The Perils of
335. 1:04:03artificial intelligence if it's that kind of organization it needs help if it's kind of organization that's
336. 1:04:13like we're going to figure this [ __ ] out we're going to figure out how this can help the people of Utah [ __ ] awesome
337. 1:04:18and if that's the case you should be a voice in that conversation it seems most equate AI to scary robots very few know about what
338. 1:04:27generative AI is yeah so so the thing the Breakthrough that I had on this concept Source Camp Kelly is
339. 1:04:36um it it took me it it was a tweet that broke it broke it open for me um I've been experiencing for like the second half of last year in
340. 1:04:50the first half of this year this cognitive battle where in my head I'm like I know these AI tools are probability engines
341. 1:05:05and I know it's a cold heartless probability calculator that is finding a mathematical point in Thousand
342. 1:05:14dimensional mathematical space Oh three M storage I'm working on it I know someone that's two degrees of separation
343. 1:05:21that's a part of it yeah Jim you need to be a part of that like if nothing else you need to go in there just tell them
344. 1:05:26what you're doing if you just regaled them with three or four of the stories that you've shared you know in here um
345. 1:05:34you you need to you need to be in that conversation in the in the Utah legislature that's that's amazing
346. 1:05:43um so for for for the better part of a year I'm like I know cognitively that these AI tools are calculators and I know we call them
347. 1:05:58robots and I know that Hollywood has told us that the robots are going to kill us but every time I use chat GPT it
348. 1:06:05doesn't feel like I'm interacting with a robot it feels like I'm interacting with a virtual employee who by the way is the most
349. 1:06:17educated clueless employee I've ever worked with they know everything but they're clueless about how I do what I
350. 1:06:23do and they're clueless about the world and they make mistakes and they make fuckups but they're incredibly
351. 1:06:31positive and there was just this part of me this nagging part of me that was like it feels human and I know it's not and and it it
352. 1:06:41was even like one of the first things first thing that popped into my head when I when I used chat GPT for the
353. 1:06:47first time I was like oh [ __ ] the Turing test is done this is the the Turing test has been passed like okay check
354. 1:06:57and if you don't know the Turing test is a computer system that is capable enough that when you interact with it you don't know if there's a
355. 1:07:08person on the other side of the screen or not and then I saw this tweet from this woman Anna I forget her last
356. 1:07:18name and I think she was she was inspired by an Ezra Klein podcast that was talking about this and and she said
357. 1:07:26AI is the collective intelligence of humanity and and for me that just went I was like holy [ __ ] that's
358. 1:07:38it the mechanism that is generative AI that currently it's it's mostly about this Transformer architecture the way it
359. 1:07:46works the mechanism of it is a cold unfeeling probability calculator the mechanism is clueless cold
360. 1:08:00technology but what it's been trained on is the output of humanity that's been dumped onto the public
361. 1:08:10internet and as much of it as they could encode over the past four or five or six or seven years and they smashed it into
362. 1:08:21this mathematical thousand dimensional latent space when you interact with chat GPT you're actually collaborating with
363. 1:08:31all the people that came before us it's collective intelligence it's not artificial intelligence this is
364. 1:08:43why the Chief Operating Officer of Black Rock don't know if you know who what black rock is but go to chat GPT and ask
365. 1:08:52about Black Rock they [ __ ] own the world they own everything the Chief Operating Officer of Black Rock said about two months
366. 1:09:00ago he's starting to hire liberal arts analysts in favor of Engineers and mbas Robert Rossy thank you very much
367. 1:09:14it's very generous appreciate it thanks for the lightning bolts liberal arts analysts why liberal arts people tend to be horizontal
368. 1:09:28thinkers they tend to be critical thinkers and they tend to be really good with the human [ __ ] the empathy the how to talk to
369. 1:09:37people the how to understand people those people that are good at chat GPT tend to be people that are good with
370. 1:09:49people if you want to get better at prompting you can take a prompting class but you know what would probably be more
371. 1:09:57beneficial a psychology class or a Sociology class or a philosophy class or just start interacting with chat GPC chat GPT as if it was a
372. 1:10:13philosophy Professor hang on the dog's whining you got to go out what do you want cheese
373. 1:10:49all right we ready to get started I'm trying to figure out how to get a I to work with common sense well okay so wind rider
374. 1:11:0147 that reminds me one of one of my big breakthroughs in life did he just ask jamy if he wants cheese yes I
375. 1:11:10did generally about now is when when the MS gives him a piece of cheese so sometimes he's like I want [Laughter] cheese um you want to get
376. 1:11:25artificial intelligence to work with common sense okay so I was about three or four years into building
377. 1:11:40agency.com in 1997 or 8 and we were growing really fast and [ __ ] was going not good and I said
378. 1:11:54to Aaron Sugarman who was our president of North America I said um I mean what the [ __ ] is with these people it's just common
379. 1:12:03sense and his response to me was he was very calm he was a good balance for me and he said Kyle Common Sense isn't
380. 1:12:17common and it [ __ ] shattered me because I knew he was right I knew in the moment he was right and so I don't know what your
381. 1:12:26definition of common sense is but what I can tell you is if you interact with chat GPT with that
382. 1:12:40intention you will be able to get it to interact with common sense but common sense is very particular to you so it
383. 1:12:48probably means that you want it to behave in a very particular way and so what you should get good at is talking
384. 1:12:55with it as if it's a person that could probably use a lot of coaching to get better at what you consider common sense
385. 1:13:05and then it will get really good so you should look at in chat GPT you should look at a thing called custom
386. 1:13:12instructions so custom instructions allow you to tell it the kind of personality you want it to have and how
387. 1:13:19you want it to behave how you want it to respond when people talk to it those custom instructions you could probably get chat
388. 1:13:28GPT to behave very very close to what you want and if you want to take it a step further build a custom GPT where
389. 1:13:34you actually upload data into its brain and do those custom instructions and then you can create a standalone thing that you can share with
390. 1:13:46other people where it behaves with the common sense as you define it so anyway I just saw that comment that was kind of cool all
391. 1:13:58right that's what my AI personas are for yep AI personas are great for that stuff that's great Janu I'm starting to see a
392. 1:14:06bit more benefits for it evening how's the night treating you very good I'm it is it is philosophy Mondays as as
393. 1:14:16they've come to be come to be known so I started taking weekends off just so I could create some space for myself to you know me
394. 1:14:26Ander and when I come back after not being entrenched in this stuff for 2 days I tend to be very philosophical and
395. 1:14:35it's not an intentional thing it's just something I've been observing and other people have been observing we going to
396. 1:14:41get philosophical Kyle tonight apparently so we're an hour 20 minutes into it and I haven't talked to I've talked very
397. 1:14:51little about artificial intelligence or tools but it doesn't matter because the tools are irrelevant tools are irrelevant they're
398. 1:15:08irrelevant is a digital twin just a custom GPT well no so let me let me unpack what a digital twin is there
399. 1:15:17there's there's lots lots and lots of definitions digital twin has a long history so there's a guy that spoke at
400. 1:15:23the AI salon named Richard Boyd he's got a company called ultis Sim um and they've been doing simulations for 20 years something like that
401. 1:15:37um digital twins can take all sorts of forms digital like Richard Boyd one of the things that he's done over the years
402. 1:15:45he's done a lot of work he sold one of his companies to LED Martin um he worked with the Department of Defense to create a digital twin of a
403. 1:15:54new nuclear class submarine where they built a fully functional um down to the systems like the switches and dials and
404. 1:16:10gauges and software systems as well as the physical representation of the submarine they created a digital twin of that
405. 1:16:22submarine that people in the Navy could train on how to run a nuclear submarine in that's a digital twin um I assume what what you're asking
406. 1:16:37about is more like the digital twin of if I make a digital twin of Kyle what might that look like that could look
407. 1:16:43like a custom GPT so right now I've got three different digital twins of myself I've got one of them for this group
408. 1:16:54called content of ution where we've got these digital personas that are kind of these professional digital personas I've
409. 1:17:00got another one uh that's tied to this book I wrote called AI Futures and Anthology um where the chapter of my
410. 1:17:08book as well as an interview that I did specifically for that um has created this digital twin and and in both of
411. 1:17:16those cases I've got the individual digital twin as well as the group the group of digital twins that
412. 1:17:24also sort of interact together which is fascinating and then I've got a I've got a a standalone one and I've got a couple
413. 1:17:30others that are in the neighborhood of digital twins um it is my deep belief that these kind
414. 1:17:45of representations of our self are going to be it's going to be really important to have the skill to create them and manage
415. 1:17:55them um I'm working with Cindy [ __ ] on a project right now that's all about that and it's it's really exciting
416. 1:18:04um with another guy from the AI Salon um who I also worked with back at agency.com 30 years ago so it all comes
417. 1:18:13around why am I working with him right now cuz he was one of my favorite people back then and I know he can deliver [ __ ]
418. 1:18:26the skills of what he needs to deliver I don't know I don't even know if he's got the skills to do what I'm assuming he's
419. 1:18:33going to do but I figure AI will help out with that [ __ ] so so I think we're going to need to start to
420. 1:18:46understand how we create versions of ourselves that we put in the world right now most of the interactions with these digital twins are
421. 1:18:59um are just text based but you can put chat GPT into voice mode and actually have a conversation with my digital
422. 1:19:09twin probably within 6 months to a year I'll be able to put together a digital twin that has some version of my
423. 1:19:16likeness like it might be a Pixar looking version of me or it might be photographic like this my voice will be
424. 1:19:27sampled and the thing will have knowledge of who I am and you'll be able to have a conversation with some version
425. 1:19:33of me and probably multiple versions of me here's the business Kyle here's the fun zany Kyle from his Tik Tok
426. 1:19:47lives and I'm going to a big part of what I'm going to spend the next two years doing is really understanding what it means to create
427. 1:19:57those what it means to have one of those or to have multiple of those what it means to manage those uh
428. 1:20:06um why anyone would interact with one is really fascinating to me right now like we built all these professional digital
429. 1:20:13personas for this content Evolution thing and we've now put them into this master GPT that's got like we've got
430. 1:20:20like 20 digital twins inside this one GPT and so you can say to this GPT given this business problem who are
431. 1:20:29the top five people in your you you know that are represented here to talk about this and I want you to host a roundt
432. 1:20:39discussion where you ask those people to talk about that topic we can do that right now today and it's fascinating and it's really good
433. 1:20:48it's like surprisingly good and how we made those digital twins was we created a structured interview that captured not only who
434. 1:21:01they were professionally but like how they think how they work who's who their creative Inspirations are who are heroes
435. 1:21:07of theirs um what have been challenges in their life and how they overcame them things like that so every person in
436. 1:21:16there is represented with this same interview that we didn't have them type out we had them talk it answer the
437. 1:21:24questions as if you were human because how we talk is different than how we write and so we wanted these things to
438. 1:21:33have the essence of who they are and and they're [ __ ] surprisingly good but what we're faced with now is
439. 1:21:42why would anyone use this why would we use this what are we using it for and that's what we're trying to figure it
440. 1:21:49out oh my God I had a profound idea why the Nanobots measuring our bodies observe becomes particle I didn't I can't see the rest
441. 1:22:01of it but that sounds amazing you're definitely passionate Canen respect that thank you the cult cult Case
442. 1:22:09Files yeah listen here I you know I am one of the things you should know about this channel even though I'm an AI
443. 1:22:19Optimist I'm also an AI realist but what I am deeply confident in is that the world has already changed it's
444. 1:22:35just that no one knows it yet relatively speaking that these tools are already significantly beyond anything we've ever
445. 1:22:43experienced before what this channel is not about is defending AI or saying if it's good or bad or saying you know uh
446. 1:22:53are are there copyright issues [ __ ] yes there are copyright issues are there issues with how this data was collected of
447. 1:22:58course do you think if any one of these companies had asked anyone permission to use their data anyone would have given it
448. 1:23:07no so that horse is already out of the barn so what we don't talk about here is is it good or bad because it's
449. 1:23:14irrelevant that's like talking about if the Giant Wave that's about to crash across Florida is good or bad it's irrelevant whether
450. 1:23:28it's good or bad it's coming so so this channel is about that why AI not 100% accurate do we not have anyone with the skills or AI lack
451. 1:23:41the data good question um it's just early so what AI doesn't have right now is reasoning reasoning
452. 1:23:55will fix a lot of the issues we have right now so here here's the way it works the way the Transformer model works it's it's a
453. 1:24:07probability it's a token by token probability engine and that doesn't mean [ __ ] but it it kind of works like this you type a
454. 1:24:16prompt and as you're typing that prompt when you hit submit it turns that prompt into a probability of the next token and you can think of a
455. 1:24:27token essentially as a word but tokens also include spaces and paragraphs and [ __ ] like that or uh punctuation so you can think of it as
456. 1:24:35it's it's predicting the next word the next token and so you type a prompt and it goes into this and I know this sounds
457. 1:24:43like [ __ ] but it's really what it is thousand dimensional mathematical space where words are represented in semantic clusters
458. 1:24:55I love talking about this because it sounds like I'm just talking horeshit but this is actually what happens semantic clusters and what do I
459. 1:25:02mean by that so you might have a cluster that has the word dog in it and and that version of the word dog that semantic
460. 1:25:10cluster is like animals and things that live in your house and things that make you happy and then somewhere else in Thousand
461. 1:25:17dimensional latent space is the word dog in a different cluster and that's where like the guy did his girl wrong and he's
462. 1:25:25a dog right both the word dog but very different meanings semantic meanings so the way these things work is you write a
463. 1:25:33prompt and it says the next probable most probable word is blank and then the next most probable word after that is
464. 1:25:40this and this and this so when you see it write a four or five paragraph answer what we think what what we
465. 1:25:48experience is oh it's smart it's it knew the answer to that so it's it's reasoning it's not it's literally going word by word
466. 1:26:00what is the next most probable word and there's some funky math in there to keep it original and interesting and things
467. 1:26:06like that so it's not always repetitive called temperature that's a whole other [ __ ] thing reasoning so so the implications
468. 1:26:18of that are when you ask it to do something and it vomits this stuff out it doesn't know what it said it's not
469. 1:26:28doing any analysis of what it said of whether it's right or wrong so when it gives you some [ __ ] and you're like
470. 1:26:36hey you just gave me some [ __ ] Mister maybe you can fix that up next time and it's like oh I'm sorry I didn't
471. 1:26:41mean to make a mistake let me try that again and then it vomits out the same [ __ ] and you're like uh you're still
472. 1:26:47screwed up you know you know when you've done that 10 times and it and it literally can't fix it that's just the
473. 1:26:53nature the Beast now there's ways that you can prompt it that get it much better but it doesn't know what it
474. 1:27:01created some of the things that are coming this thing called qar that open AI talked about and some other mathematicians have talked
475. 1:27:11about is reasoning and and in that case what it's doing is it's actually paying attention to what it's writing and it's
476. 1:27:18analyzing it and saying is this in the direction of a right answer or away from a right answer and it mathematically
477. 1:27:25sort of keeps trying again to get more and more right so as these systems either have secondary analysis so it writes
478. 1:27:35something it then analyzes it and says is it any good and then it rewrites it and it does that iteration four or five
479. 1:27:42or 10 or 100 times before you see anything then these things will start to get better when they have reasoning
480. 1:27:48systems behind them they're just not there AI is like fire fire it's neither good or bad it's how it's used that's
481. 1:27:53exactly right wind rider yeah it's exactly it's a it's a force right it's a force that like the miraculous part of
482. 1:28:03it is is that these mathematicians have effectively figured out how to compress the knowledge of what's been put on the
483. 1:28:11public internet into this ball into this box they took they took all the knowledge and jammed it in there and
484. 1:28:21then we can make an API call down into this thing and go hey give me the answer to something and it just goes squit
485. 1:28:28here's your answer and it is right enough of the times that it's quite remarkable and the word whimsical is
486. 1:28:36always slipped in Whimsical and um and uh tapestry and delve let's delve into the tapestry oh man AI knowledge is the
487. 1:28:51essence of human knowledge that's exactly right it is collective intelligence delve let's delve together shall we I have a i in my custom
488. 1:28:59instructions in my GPT I have a whole list of words that I want it never to say delve is the first one in the
489. 1:29:10list the best explanation for AI at the moment it's a tool it all comes down to the operator yep it does and it's also
490. 1:29:18so let me explain another thing to you this is um Andre ay um talked about this at a at a conference at MIT about a year
491. 1:29:31ago and people were asking him you know why does it hallucinate and so he he used he used an example of a
492. 1:29:40mathematician which I thought I thought was you know he's a mathematician so it's good um okay so if if gp4 has been trained on how
493. 1:29:54however much of the internet they could smash into it it's got all the stuff on the internet so let's say
494. 1:30:03that there's a particular equation that you want to know something about and you say to chat GPT tell me about this
495. 1:30:15equation and it gives you back an answer and you're like oh that's seems really good and then but you look at it and
496. 1:30:20you're like oh but it got the equation wrong or it got the explanation wrong like no that's not what that equation
497. 1:30:25means at all it it like seemed right but it's wrong why is that the way Andre karpathy described it it it gave a really clear visualization
498. 1:30:38for me so what he said was if you just ask it to tell you about the equation think about what's on the internet on the
499. 1:30:48Internet is the guy that came up with the equation his explanation of it all the other PhD level mathematicians that
500. 1:30:56have talked about that equation over the years and advanced it and all of the undergrad students who learned about it and [ __ ] it
501. 1:31:09up more times than they got it right that's all in the data set so if you just say tell me about this equation CHT
502. 1:31:18is actually mashing together the good [ __ ] the good data with the shitty data and it just vomits out data and the way I thought about this
503. 1:31:28was it's almost like chat GPT is like the Grand Canyon and if you give it a generic shitty prompt you're basically saying
504. 1:31:37we're going to take all of the data that was dumped into the Grand Canyon we're going to we're going to sort of average
505. 1:31:43it out and and give you some sort of answer and it and right but if you go into a sub Canyon if you go into the sub
506. 1:31:51Canyon of mathematicians and and and you say I want you to act like this particular mathematician that
507. 1:31:57specializes in this particular equation and I and now I want you to tell me about that equation you'll get a better
508. 1:32:04answer because you've now gone into a sub Canyon where it's just highlevel mathematicians and it's the data set
509. 1:32:11where that actual mathematicians equation is in there and so that's why prompting matters and that's why things like act
510. 1:32:19like an expert in real estate development makes a difference because you're narrowing the latent space that
511. 1:32:27that probability engine is going to look at so wild right it's crazy an early experiment had AI reviewing Amazon reviews some of the
512. 1:32:39agents self wait some of the agents self to themselves to judge I don't know what that means um can you show us how to
513. 1:32:49tell chat PT not to use certain words absolutely was that giggy mom I like it yes let's go I will show you custom instructions okay so
514. 1:33:02we're going to go to chat J my session has expired see I didn't log into this all weekend
515. 1:33:19uh okay so here we are at chat ja you'll notice we're at chat GPT 40 this works for any of the models you
516. 1:33:28could be in 3.5 if you're not paying for this um you you'll have GPT 40 um but it'll be very limited um I
517. 1:33:38think it may you may still have access to gp4 but I'm not sure but anyway it doesn't matter okay um to access custom instructions you're
518. 1:33:49going to go to the upper right hand corner and you're going to go to well you can go to settings or customize chat GPT so
519. 1:34:00we'll go to customize chat GPT go to customize chat GPT and then it says custom instructions what would you like
520. 1:34:07chat GPT to know about you to better provide responses so I said you're a helpful assistant and are not afraid to
521. 1:34:14criticize as well as support if it's appropriate so I just put that in there cuz one of my pet peeves with chat GPT
522. 1:34:23is also Al one of its greatest strengths is is that it acts like a golden retriever everything's a good idea it's
523. 1:34:29like you know you're like you know I want to smash bugs that's a really interesting idea you know smashing bugs
524. 1:34:36could be a really interesting way to express yourself um so I wanted to sometimes be able to criticize I I don't
525. 1:34:42know that it's good at that but you can tell this anything you could tell it that um you want it to be very concise and not be overly
526. 1:34:56wordy um you could tell it that you want it to act like a standup comedian where it's sarcastic and everything's a
527. 1:35:03joke um so so you can you can give your chat GPT a personality here the second part of custom instructions is how would
528. 1:35:12you like chat GPT to respond so the first thing I have here is for every response give it a sequential response
529. 1:35:19count followed by a response title that summarizes the response you're like what the [ __ ] is that so let me show
530. 1:35:26you so if I go in here and I say um why is the sky blue you'll notice it says response count one response title why the sky
531. 1:35:41appears blue and then it gives me the answer someone in here showed me how to do this I forget who it was anyway um
532. 1:35:54I didn't come up with this someone in here came up with this but I like I I love this thing it gives me the date it
533. 1:35:59gives me um sort of a synopsis of what the answer is going to be and what what count this is where this is valuable is
534. 1:36:06if I've got 40 different responses in a chat thread I can go back and say oh that was where it you know it was
535. 1:36:13response 37 is where it did something interesting and I want to go back from there so anyway that's that's what that thing
536. 1:36:19does and then below that I've also include the date and sometimes this bugs me sometimes I don't like it having that but it's it's I it's
537. 1:36:30decent now I said in custom instructions avoid overused words delve tapestry vibrant landscape realm Embark vital moreover
538. 1:36:42arguably significant ultimately unveil transform robust expansive intriguing facilitate harness and phrases dive into
539. 1:36:52it's important to note important to consider based on the information provided so all this [ __ ] I actually had
540. 1:36:59chat GPT help me come up with this list of things it overuses I don't know if it's right um I think I went out on the
541. 1:37:06internet I Googled I found an article that talked about what are the most overused words and then I had chat GPT
542. 1:37:13expand it and then I edited it and so this is a big ass long list and in fact here's what I'll do I am happy to share this cuz I don't
543. 1:37:23really give a [ __ ] if it's going to help you I think there's probably 38 better ways to do this than I did but I'm going
544. 1:37:31to go to the AI Salon Irregulars Channel I'm going to say here is my anti- delve prompt for custom
545. 1:37:50instructions and then there you go post and that's how you do it so um have I noticed it being better
546. 1:38:04yeah it's a little bit better I'm working on my book with Claude ethical money thanks for the inspiration awesome
547. 1:38:11Joseph that's fantastic it's [ __ ] fantastic yeah like like working with Claude or chat GPT as an as a writing assistant rather than
548. 1:38:22saying he is going to write my book no AI is not going to write your book you're going to write your book but
549. 1:38:28using AI as a writing assistant to block and tackle the outline and block and tackle the main Frameworks of the
550. 1:38:35chapters it's [ __ ] brilliant at that um I don't think it necessarily is going to speed up the editing process I think
551. 1:38:43the editing process Still Still is going to require you to you know look and get frustrated and you know break things and Unbreak things
552. 1:39:01um all right does that help hopefully so 112 people in here it's a party
553. 1:39:15tonight philosophical Kyle's got a got a crowd you could make money withing people are just like what is
554. 1:39:23this channnel what are what am I doing here no yapping supposedly works oh
555. 1:39:38that's funny Joker yeah in custom instructions just tell it not to Yap I screenshotted my comment being
556. 1:39:46pinned thanks that's awesome Joseph I know it's so cool it's it's so cool the mods in here do a really good job of of
557. 1:39:54grabbing grabbing the interesting stuff and and you know supporting people so if you have questions feel free to answer
558. 1:40:01them I can't see everything but the mods do a really really good job of pulling stuff up for me to look at I'm using an
559. 1:40:07old 32-bit laptop running Linux love that word list I have a list of words related yeah yeah just yeah feel free to
560. 1:40:16take that word list that's on the AI Salon in in the Irregulars Channel oh man rookie question okay first of all
561. 1:40:29Brandon rookie questions are the best and by the way it's probably not a rookie question but what does it mean to
562. 1:40:36make your own version of chat GPT and why would I do this okay great so let's go let's go show another thing great question actually okay
563. 1:40:47[Applause] wait uh uh okay we're going to do a little we're going to do a little explanation of what a of what chat GPT
564. 1:41:03is what GPT 4 is and what a custom GPT is also know that these things also exist in other platforms you can do the
565. 1:41:15same thing in po poe.com you can also do it in Claude Claude now has this thing called um the [ __ ] is it called profiles
566. 1:41:26no well claude's got some version of this I forget what it's called Pages no that's that's uh projects thank you AA
567. 1:41:38man projects is it in the boxes okay so here's what we're going to do I'm going to I'm going to explain to
568. 1:41:47you why these research Labs need to hire some [ __ ] branding people because the fact that you're confused
569. 1:41:57about what a custom GPT is is not your fault the the every it's totally confusing the terms are all intermingled
570. 1:42:05right now it's a [ __ ] disaster so it's not on you okay so this box this is gp4 gp4 is literally just like a box
571. 1:42:18full of crap it's a box full of all of the knowledge of mankind that was stuck on the internet in the past 70 years
572. 1:42:25that they could manage to squeeze in the box and then GPT 5 will be a bigger box okay so they took all of what they could
573. 1:42:32find on the internet and they jammed it in here as a developer you can go to the open aai playground and you can get
574. 1:42:41access to this and you can create your own application so um you know some of those marketing applications that do chat GPT
575. 1:42:49like things they're using this but they've got their application that's making it API calls to this this is chat GPT chat GPT is an
576. 1:42:59application that open AI wrote that is the simple little Google interface with like the four examples where you type [ __ ] in and it
577. 1:43:10gives you back cool answers so what happens when you interact with chat GPT is you're interacting with this
578. 1:43:17application that they wrote and when you ask a question it sends an API call into this big model and says hey um Bob had a
579. 1:43:29question and then the GPT model gives back an answer and this thing shows it to you on screen okay that's chat GPT
580. 1:43:37and then if you ever hear people say like well Microsoft um co-pilot is is chat gp4 no it's not Microsoft co-pilot
581. 1:43:49is like chat GPT but it's written by Microsoft they're both calling the same underlying model but they're different applications
582. 1:43:58that's why this one is shittier there's a lot of reasons why this one is shittier we don't need to go
583. 1:44:03into it now but anyway co-pilot chat GPT this one's written by the people that made this so that's why it's better okay now a custom
584. 1:44:19GPT is a particular version of this so I just showed how to do custom instructions where you could have chat GPT act like a
585. 1:44:29standup comedian or talk less or talk more or only respond in I Amic pentameter or whatever it is um so so
586. 1:44:38you can kind of slightly modify this so that it behaves differently but what you can also do is you can
587. 1:44:44create a completely custom version of this that actually sits up here right so I can create a c customized version of chat GPT that just
588. 1:44:55does what I want it to do and your question is a really good one why would I want to do that okay so that's the that's it
589. 1:45:07conceptually I'm going to show you the very first GPT I ever made I made it live on this channel it's somewhere in in our YouTube
590. 1:45:16channel that historical moment was captured okay over in chat GPT they've got this thing it says explore gpts and I'm going to go there
591. 1:45:32and so here are gpts oh and this is the thing that I was going to explain before The Branding nightmare gp4 Omni is that big
592. 1:45:46box chat GPT for Omni is the application that sits on top of it custom gpts are called gpts and are up here and are
593. 1:45:56something different entirely so you've got three different completely different functional objects that are all called
594. 1:46:03gpts it's a [ __ ] disaster okay that's that's one of the reasons it's confusing so again not on you so if you go to explore gpts and you
595. 1:46:13scroll down you see all these different things here's an image generator one here's a scholar GPT here's one that
596. 1:46:20uses canva and creates canva [ __ ] um here's ones that are made by um open AI there's a data analyst there's hot
597. 1:46:31mods you can modify your image into something else here's writing ones AI humanizer copywriter GPT okay so I'm going to go in here and
598. 1:46:42I'm going to search for Sin City sipper and you can go search for Sin City sipper right now too so one of the
599. 1:46:53cool things about gpts is they're sharable so this is Sin City sipper by Kyle Shannon Vegas bartender with a Twist on the
600. 1:47:00classics and so I can come in here and I can play and so here we are this is Sin City sipper and I've got these pre-written
601. 1:47:10questions show me a new fangled oldfashioned make a classic martini classier reinvent a margarita for me or
602. 1:47:17I could I could ask for something I'm going to say um fourth of July just ended I need a cocktail
603. 1:47:33[ __ ] tail for the dog days of summer and what and what this is going to do now this thing has been I prompted
604. 1:47:46it to be a really good bartender and take classic drink recipes and give them a Vegas twist and so it's got kind of a
605. 1:47:54cool personality so it says all right let's whip up something refreshing for those scorching days of summer how about
606. 1:48:00the cucumber melon cooler it's light refreshing and perfect for sipping on a hot hot on a hot day and it's uh let's
607. 1:48:08see it's vodka melon lure like madori lime juice simple syrup cucumber slices soda water and ice nice sounds lovely
608. 1:48:16then it explains how to make it and then it gives you it creates a picture of that drink now could you technically and easily go
609. 1:48:33replicate this yourself just with prompts with chat GPT yes you can go to chat GPT right now and say act like a
610. 1:48:40bartender and and when I ask you for drinks you know start with a classic recipe and then give it a Twist and then
611. 1:48:49after it gives you a recipe you could say okay now make me a photograph of that drink and could you do yes you
612. 1:48:56could do that but you could also just go to Sin City sipper and say make me a drink and it does it and so this is a really simple example
613. 1:49:09of a subset of chat GPT right this it is a very specific version of chat GPT sitting on top of and leveraging all of
614. 1:49:18the Power of chat GPT but doing it in a way that you know is is useful for me someone came into the channel right
615. 1:49:28after the Super Bowl and they said they used Sin City sipper for for this they took a picture of all of the
616. 1:49:36booze that they had they lined up all their booze bottles and they took a picture of it and they said to Sin City
617. 1:49:42sipper I want you to come up with 50 Super Bowl themed drinks and so they created a menu for people at the party
618. 1:49:54and then they created the 50 recipes for these cocktails that it created and so people at the party could ask for you
619. 1:50:01know the the whatever like you know the Carolina Panthers whatever I forget who played this year um and based on the booze and their thing
620. 1:50:14and it it you know and it created you know pictures and things like that so anyway so that's it you could do this for
621. 1:50:24um one of the other cool things you can do with with these gpts is you can upload knowledge to them and then it'll
622. 1:50:31be specialized to that knowledge so we did one for my my company storyvine where we took we have a 50 page sales
623. 1:50:39process document that's this Frankenstein nightmare of a document it's got 50 pages it's got all sorts of
624. 1:50:46scenarios in it about how to sell it how to talk about it what to do in certain situations and so we uploaded that
625. 1:50:53document as the knowledge source to one of these gpts and now our sales guy can use that GPT to ask questions about hey
626. 1:51:01I'm in this situation what do I do and this thing knows it so it they they can be very very powerful or you can have them make drinks for
627. 1:51:10you oh the other thing I did was you know I had to do a little marketing at the bottom this GPT was created by Kyle
628. 1:51:16Shannon with a link to my LinkedIn profile I think oh no that's that's an email co-founder of the AI Salon that's
629. 1:51:24a link to the AI salon and then Creator host of the AI learning lab on Tik Tok so anyone that uses this I get a little ad crazy
630. 1:51:35right so I hope that helps when you go to that explore gpts button in the upper right hand corner you can say create GPT they're
631. 1:51:47really easy really easy to make wow super pull C Tail list from an uploaded photo into chat gbt amazing I know right Firefly like
632. 1:51:57amazing that that's the thing is these custom gpts have all of the Power of chat GPT in them so they've got Vision you can
633. 1:52:07upload photographs to them and say hey Analyze This circuit board and tell me what it is if you had a hardware analysis GPT
634. 1:52:31all right open the AI wait oh open AI reinstated my one with 5K chance but it's stuck at 5K yeah my Sin City sipper it went from
635. 1:52:45200 uses to 300 but I don't know I haven't really been pushing it so have you found a way to make image
636. 1:52:57recognition to be more exact to what it sees not right now Rick Olsen um I would imagine that if you took the
637. 1:53:08time to prompt it right you could probably get it better um I haven't I haven't dug that much into it it's it's
638. 1:53:14pretty remarkable what it can see I mean it can interpret like you can put in a picture of a whiteboard with sketches on it like
639. 1:53:22interface sketches and it can analyze that and turn that into software what's the biggest difference between three and four um there's a lot
640. 1:53:36so so chat GPT 4 and chat GPT 4 Omni are multimodal models and um chat gpt3 .5 is
641. 1:53:53essentially just text chat GPT 4 and 4 Omni um can make images can see they've got a thing called Advanced
642. 1:54:03Data analysis or code interpreter where it can write code and execute code so you can upload a um a CSV file of data and say analyze
643. 1:54:15this data and it'll actually write and execute python code to analyze the data in that data file I mean it's it's
644. 1:54:24staggering um what's coming in the fall is realtime voice where not only will chat GPT be able to understand what you
645. 1:54:34say it'll be it's able to understand the emotional content of what you say so if you're really excited about something
646. 1:54:46it'll know that and it'll be able to respond in kind or if you're really depressed and low it's going to be able to re recognize those
647. 1:54:54differences um it's it's night and day the the the difference between them if you haven't paid um the 20 bucks for chat
648. 1:55:05GPT plus I would recommend it now here's the deal open AI made the free version of cat GPT now uses gp4 Omni but it just uses it in a
649. 1:55:20very limited kind of way um I personally think it's worth paying the 20 bucks a month to chat GPT um that said the new Claude 3.5
650. 1:55:33Sonet which is also free but also very limited um is a is a really interesting competitor um a lot of people like the the Claude
651. 1:55:443.5 Sonet better than they like chat GPT right now so I'm starting to see people saying I'm canceling my chat GPT and I'm
652. 1:55:52moving over to Claude I'm not there yet I'm still paying for both um I think they're both worth the money right now um so for what that's
653. 1:56:10worth all right let's see let me see what else I got going here I think I got to start my 15minute goodbyes cuz I'm a little
654. 1:56:19chatty tonight so I think it's going to be a while can chat act as a writing assistant for a screenwriting project Oh
655. 1:56:25Cassandra absolutely um I've written seven screen plays I created I created a custom GPT where I used the screenplay
656. 1:56:36as the as the input source um so you can ask it questions about a a a screenplay I've written but brilliant writing
657. 1:56:46assistant um but you have to you have to you have to train it to be a better writing assistant so what I would what
658. 1:56:59what I would encourage you to do is um really think about what a good screenwriting assistant would know and and build that into your custom
659. 1:57:13instructions so in your custom instructions tell it it's a screenwriter maybe you tell it um it should use Sid
660. 1:57:22Fields screenwriting you know formula or save the cat or you know some storytelling framework if if you work
661. 1:57:30with a particular storytelling framework put it in the custom instructions um if you write dialogue in a particular
662. 1:57:39way you can describe that you could even put in Snippets of dialogue and say um this is an example of the kind of
663. 1:57:46dialogue I like here's an example of the kind of of dialogue I don't like and why and and like put that all in the custom
664. 1:57:53instructions um or create yourself a custom GPT that has all these instructions in it um because you could
665. 1:58:01you could probably get chat GPT dialed in so that it when it's contributing to your story it's doing it in a in a in a less shitty way
666. 1:58:13um in general how I use chat GPT for for larger writing projects like that is I lean on on it really heavy for
667. 1:58:24outlining less so for filling in the outline and even less so for doing detailed things like dialogue and and
668. 1:58:33you know uh screen directions and things like that sometimes you want those to be very particular so I think the details
669. 1:58:39of the screenplay are still on you but the blocking and tackling of outlining getting the story right having it
670. 1:58:47analyze what you've already written and give you six Alternatives What If This Were a common what if in the middle of act two instead
671. 1:58:54of making this decision they made that decision like I took I've got a fragment of a screenplay right now where I've got the first three
672. 1:59:09scenes I've got like the first half of an outline the first three scenes and then a really shitty incomplete second
673. 1:59:16half of the outline like I know the beginning of it and I know the ending but I'm I'm missing like two and a half
674. 1:59:22to three is just there's nothing there and so I just uploaded that shitty incomplete document into chat GPT and
675. 1:59:30within 20 minutes I had like six different ways I could take it the minute I tried to get it to write scenes
676. 1:59:40after the scenes that I had written they were just shitty so like you got to break it into piece parts and just it's
677. 1:59:46a lot of critical thinking and thinking in modular Lego blocks but it's it's really good at doing the Lego blocks and
678. 1:59:52reorganizing them and things like that so brilliant for that anyway all right um let's see do you see that Sora has done a beta
679. 2:00:04type commercial for toys of Toys R Us yeah Toys R Us did a commercial or the agency did I didn't even know I knew
680. 2:00:11Toys R Us like the brand still existed but I didn't know it was still a functioning entity in any way I'm
681. 2:00:17rewriting the sop at work for an import for transportation sector which platforms should I look at what's the SOP again statement of
682. 2:00:29purpose hang on what is an sop in Biz context standard operating procedure for
683. 2:00:46the transportation sector um you should explore Gemini Claude and chat GPT and probably perplexity probably perplexity so
684. 2:01:03perplexity is kind of like a combination of a Google Search and a large language model where it does research with
685. 2:01:10citations um it's gotten some [ __ ] recently but it's still it's still decent so I'd say play with perplexity
686. 2:01:18um and then I would just explore like with all all of them your critical thinking is the most important piece don't don't put the
687. 2:01:28expectation on any of these tools where you're just going to go like write me a standard operating procedure for the
688. 2:01:34transportation industry and expect it to do anything decent it's going to just be a vomit of generic ideas so what you're
689. 2:01:41going to want to do is put the framework together um with one of these tools and just I I would play with perplexity
690. 2:01:52Claude Gemini and an open Ai and just feel which one feels right to you when you start interacting with it and then
691. 2:02:02just and then just lean on that one this was sort of how the conversation started tonight where if if you've got a project
692. 2:02:09like this and you're just learning this try on as many tools as possible just like playing and then once you find one that
693. 2:02:16feels like ooh that feels right to me that's the one I dig down on the two that I'm feeling like are going to be
694. 2:02:24most valuable for you instinctively are Gemini and um Claude because Gemini's got a million token context window so it's like
695. 2:02:34750,000 words it won't lose its [ __ ] for a long time and claud's 200,000 chat GPT is really short context window
696. 2:02:43perplexity is really short so I would go with probably Claude or Gemini is my guess all right um why should a hospital be
697. 2:02:57nervous cuz we might have another pandemic why should they be nervous about generative AI um I think anything where people's lives
698. 2:03:13are at stake it's going to move slower than than you know general knowledge work I don't know that hospitals should
699. 2:03:19be nervous I think listen the way you frame that question is a fear-based you you framed it in fear why should they be nervous well
700. 2:03:33because maybe with generative AI hackers are going to be able to hack more okay I think a more interesting question
701. 2:03:41is why should they be excited what should they be exploring how could they how could they take the nursing crisis
702. 2:03:48right now and be able to provide better patient care with less people there are all the nervous Hospital there are all sorts of ways
703. 2:04:03that hospitals could aggressively use generative AI in a way that doesn't violate you know patient privacy doesn't you know
704. 2:04:17compromise diagnosis you you don't need to go into the clinical side there are so many things on the operation side
705. 2:04:23where you could dramatically dramatically improve the quality of patient experience um but if you ask the
706. 2:04:30question why should they be nervous all you're looking at is the risk hospitals are not going to lean into something if they only think it's
707. 2:04:41going to harm them they're going to avoid risk so anyway no Kyle you were literally talking about the nervous hos Hospital
708. 2:04:55oh oh we oh you were you were literally talking oh oh I see oh I get it why should the hospital be nervous oh I get it well never mind
709. 2:05:08then carry on so this this is this is why there are absolutely two different conversations going on here whatever the
710. 2:05:15[ __ ] I'm talking about and then the actual conversation I'm very interested in ai's integration into senior living oh I
711. 2:05:22think it's going to be huge Senior Living the digital twin thing for senior living is going to be a [ __ ] Boon
712. 2:05:30before you go can you check out Jace AI maybe but let me go look at it quick if it's if it's just a very particular kind of rapper
713. 2:05:41app probably not let's see meet your new AI assistant focus on meaningful things don't just chat start acting today plan a
714. 2:05:57wedding I'll join the weight list um if this is something you created um I just signed up Kyle story vine.com
715. 2:06:12um I've been weight list weight listed uh you know if this is um if this is agents and it's true agents that can can take action may be
716. 2:06:22interesting um I'll check it out all right we have a Toys R Us at Macy's oh interesting so I guess some PE
717. 2:06:31firm bought up the scraps and they're using the brand like like PE firms are want to do um all right I got to get out
718. 2:06:38of here my son's name is Jay oh that's so cool what a cool name Danielle that's great um all right so a couple of
719. 2:06:48things if you have not thank you Lord digital gos that's very generous of you I really appreciate it thank you thank you thank you
720. 2:06:57um if you haven't followed my channel follow my channel um if you want to directly support it please subscribe to
721. 2:07:05the lives that that supports me directly um but I don't really give a [ __ ] if you do or not what I do care about is keep
722. 2:07:14coming back here why would why would you keep coming back here it's a great question but the people that keep coming back
723. 2:07:23here are Reinventing themselves in incredibly inspiring ways and they're called the Irregulars you you could if
724. 2:07:30you just keep coming back you can call yourself an irregular they're not weird they're irregular and that's why they take the
725. 2:07:43mails it's a poop joke all right um thank you IR regulars you're awesome thank you mods this is you did a really
726. 2:07:54good job tonight cuz I I was able to um be both bombastic and wordy uh but kind of follow a train of thought so thank you
727. 2:08:05for that cuz that that doesn't always happen um and so generally what we do here is we'll do a little Romper Room
728. 2:08:15magic mirror if you remember that if you're a gen xer remember Romper Room and and you're like they never call my
729. 2:08:20name this sucks that's what I do night after night and probably piss off people night after night after night so who's
730. 2:08:28here tonight Mary Mary's here Janu is here against the norm Brandon's here thank you so much Tobias in the house of course
731. 2:08:37Joseph William Baker thank you thank you thank you thanks for hanging out psyllium husk fiber just drank some working on his
732. 2:08:47irregular uh that's awesome gigaom nice to see you thank you very much uh what does subscribe to the live mean um somewhere up
733. 2:08:56here you can subscribe and it's it's a it's a monthly subscription fee I think it's 10 bucks or something like that or
734. 2:09:03four bucks I don't I don't even know um but some percentage of that goes directly to me so it's just supporting
735. 2:09:09just supporting the channel um Dr Jay's in the house tonight Drew learning coming in coming into to us hot from
736. 2:09:15Australia thank you good to see you snoway thank you very much for that I appreciate that wolf Clint good to see
737. 2:09:22you sir Firefly cuno in the house Mike larus in the house um Source Camp is here who else we
738. 2:09:36[Music] got Brandy cjr cuno thank you Kyle thank you so much thanks for coming Lord digital gods
739. 2:09:50in the house tonight Teton ski in the house tonight thank you for the uh lightning bolt really appreciate that
740. 2:09:56Kyle you should ask your Congressional handlers to bring your laptop and do a demo maybe I got to figure out exactly I
741. 2:10:04don't like I have no idea what the what the event is going to be like is it going to be like we're in a big Hall and
742. 2:10:15there's going to be people pretending to give a [ __ ] about what we're saying or is it going to be a bunch of individual
743. 2:10:20things I'll definitely have my laptop there and I'll be prepared that's good I'll I'll definitely have some some uh
744. 2:10:27some parlor trick demos for for the I I'll do some want to see something cool I'll do a lot of that get some sleep
745. 2:10:36Kyle all right thank you all nighty night Joey's in the house uh all right I probably missed some people there's a
746. 2:10:44lot of people in here tonight so um thank you all for your time I will be back tomorrow there's I think there's a
747. 2:10:52after hours event for the salon here in Denver tomorrow but I'm not sure bring your cell phone I want to talk to some
748. 2:10:59of them yeah really I know me too I I gotta I got to learn how to say hey could you get your head out of your [ __ ] ass in a way that you
749. 2:11:09know they can actually hear that um I'm out of here so I might be late tomorrow but I will see you tomorrow night um thanks all it was a
750. 2:11:19fun night tonight I appreciate it have a good night bye