aiLL 8/10/23 - Surviving the AI Revolution While Getting Stuff Done

Transcript

1. 0:00foreign [Music]
2. 0:22welcome to the lab the AI learning lab what's going down with shaking What's Happening welcome welcome welcome
3. 0:34all right let's get this party started we're gonna get this water started all right let people get rolling on in here
4. 0:50I think I got to do something about these eyebrows starting to get all Santa clausy I can't have Santa 's eyebrows
5. 1:01it's not good it's not good for anyone what do you do for Dallas I am a founder and CEO of a company called storyvine
6. 1:14and uh we've been around for 11 years we do a lot of work in healthcare and Pharma it's an automated uh video
7. 1:21storytelling platform and uh I do this here uh AI stuff for for Fun and Profit you can make money with jbt not a lot of profit right now but
8. 1:32spending a lot of time on it um rolling a lot of AI into my current company the current product Vision things like that so that's that's the
9. 1:41day job and then at night time I'll do do these things do a little bit of sim racing things like that
10. 1:48um yeah this is the the AI learning lab and uh I'll talk about these things here in a in a second we'll let some people roll in here
11. 1:57Rick Olson how are you sir Jeff Clark welcome welcome
12. 2:12breathing water not smart how did he go I waterboarded himself [Laughter] do you believe in palantir well
13. 2:27I believe that they're strapping guns to large language models and that's [ __ ] terrifying do I think they'll be effective Maybe
14. 2:43we did a piece on palantir in uh everyday AI in the newsletter I did for a while and uh yeah it's interesting
15. 2:52stuff it's interesting stuff what they're doing like it's potentially beneficial to Warfare but I don't know these large language models
16. 3:00and weaponry right now feel like they shouldn't be anywhere near each other because if I can't get this thing to
17. 3:06write a [ __ ] email predictably I'm not sure I want it targeting the enemy just yet why don't we give it a year or two and
18. 3:15let it let it settle out and figure out if it actually works or not anyway welcome everybody my name is Kyle
19. 3:21Shannon this AI learning lab a little bit about me I'm an entrepreneur many time entrepreneur started I don't know
20. 3:2712 15 companies something like that number of which should have never been started number of which have been really
21. 3:33successful and lots of stuff in between um I was an entrepreneur back in the early days of the World Wide Web so I have
22. 3:42seen I've been in one of these things before and one of these things meaning uh one of these technology changes that's gonna
23. 3:52transform things and uh it's wild it's wild I can tell you that from from the from the early days of the web kind
24. 4:00of writing writing the web rocket up and being part of the.com bubble and then.com burst and all of that sort of [ __ ] it's pretty
25. 4:11intense so I didn't think I'd see another one in my lifetime but I have and its name is Chad gbt generative AI
26. 4:21um and yeah I think we're in a pretty amazing thing so anyway I love that shirt thank you very much appreciate that crypto Joe
27. 4:28um all right let me let me talk about what I got going on here what are these things so if you're if you're brand new to
28. 4:35generative AI if you're like all these people keep talking about Ai and I don't know what the hell they're talking about
29. 4:40I don't like it I just don't like it one bit if you're one of those people and haven't tried this my strong strong strong recommendation is
30. 4:49go to this top URL chat.openai.com that's the official chat GPT website if you've never been there before go there right now
31. 4:56fire up your laptop or just leave me and go there um and start playing with that now all these other things down below it are
32. 5:06also they're either variations of chat GPT or competitors of chat gbt so if you have played with this and you haven't
33. 5:13explored these things below these are the seven right now that I think are worth you at least understanding what they do now there's there's
34. 5:23they're roughly in order of what I think is you know most to least useful um but the bottom two are useful in kind
35. 5:32of different ways the bottom two are kind of more think of the bottom two more as kind of a bit more consumery pie
36. 5:37is very chatty Poe you you have Bots there you can play with a lot of different models um the top five are are the real players so
37. 5:47you've got chat GPT you got Bing which is Microsoft's version of chat GPT you got Claude which is a competitor of chat
38. 5:55gbt has a really big context window so you can you can put really big documents in Claude and that's amazing then you
39. 6:03got Bard which is not embarrassingly bad anymore and it can read images and so it's worth understanding what it's good at
40. 6:12um you got lab stop perplexity that's uh perplexity hosting um metas llama 2 models llama 7B 13B and 70b
41. 6:23um so that's worth playing with so all these are worth playing with but if you're new if you're brand new to this
42. 6:27[ __ ] and you're like but Kyle I don't know what to do I don't know where to start oops I just changed my I just
43. 6:34changed the chart into a one um go to that top one the top top one chat.openai.com that's gonna be the one
44. 6:42you're gonna look at you're gonna look at that one if you if you're down in the Bible you're trying to figure out what
45. 6:48we're doing with this yo chat Jupiter and you're gonna go to that one you're gonna go to chat.openai.com
46. 6:55take it right out in the swamp with you all right now if you're like but Kyle I don't know why you'd say it like that but if you do
47. 7:07you're like I don't know what to do then I got I got that for you too so prompts.chat that'll teach you how to
48. 7:14prompt these things it's like a little primer um that that uh will tell you what the large language models are how they work
49. 7:22how to talk to them and it's a whole big long list of prompts that you can literally copy and paste in the chat
50. 7:27will give you lots of ideas uh this this other thing here uh prompt book is for images and down here the three little
51. 7:35image things I would recommend playing with mid journey is the best fireflies the most commercially safe and Leonardo
52. 7:42is a really nice web front end sitting on top of stable diffusion which stable diffusion is good it's catching up to
53. 7:49Mid Journey but it's not quite there yet all right so that's what those things are all right so if you've got questions about AI pop
54. 7:59them in the comments below do my best to answer them uh I try to answer them all that causes issues because
55. 8:07I don't get I just it causes issues because I'm answering all the questions from the past and you're down in the
56. 8:14present going why the [ __ ] isn't the answer in my questions oh and I swear a little bit too so if you don't like that
57. 8:19I'd get out get out get out he's gonna be swearing he's gonna do it he'll do it he'll do it right up he'll swear right
58. 8:26up he'll do it I've seen him do it I come to these lives every night and sometimes he's just nice as a church
59. 8:33church mouse he's just super nice and then other times he's just got a filth mouth he's just got a filth mouth and then so
60. 8:40I'll tell you what sometimes he's so bad they'll kick him right off the tick tock they kick him right off the tick tock
61. 8:47they'll do it too I've seen it I've seen him mid-sentence just Boop he's gone we don't see him for weeks all right
62. 9:03all right I don't know what I'm gonna say tonight you know what I'm gonna do I'm gonna answer some questions there's
63. 9:07Emilio's wife what's happening 8 14 in Colorado it is hello the bear how are you Lori 711 let's shake him
64. 9:19all right let me go grab a question or two here and oh it's gonna be hard to read on this shirt sorry about sorry in
65. 9:26advance I might need to go change the shirt out hi Kyle I told a bunch of people about you in a librarian
66. 9:34webinar yesterday so you might have some Librarians I hope they don't mind cussing oh well well he's very I
67. 9:44he's very boisterous isn't he certainly seems enthusiastic all right and listen I'm not saying that Librarians that's how they all are
68. 9:53definitely not just the ones that yelled at me when I was in the library Mr Shannon Mr Shannon Mr that is completely inappropriate
69. 10:07we do not toss peanut butter at the other patrons we don't do that here Mr Shannon thank you thank you all right beautiful
70. 10:20so you've got a video portal I have a video technology company so it's a uh storyvine is a platform that
71. 10:31we have an app that guides an end user through telling their story so we do a lot of work in healthcare and Pharma so for example if um
72. 10:39if Pfizer wants you know one of their patients for a certain drug to share their diagnosis Journey our app will
73. 10:48guide that patient you know introduce yourself here talk about what it felt like when you were diagnosed it walks
74. 10:53them through the process those clips go up to the cloud five minutes later you got a fully edited video
75. 10:58and then I'll hold back in management system to deal with all that so that's what Story vines all about
76. 11:05Rick Olson fantastic thank you sir do I believe in palantir answered that one I believe in it I believe it's [ __ ] scary but you know
77. 11:13I was so I had a choice to make in life when I was a kid my my dad was a pilot in the Air Force and I thought I think I
78. 11:20might want to be a pilot and then at some point it was like do I want to be a pilot or do I want to be an actor I went
79. 11:26the uh I went the actor route hang on one second
80. 12:09all righty all right all right all right all right all right retail consultant by day oh that's nice retail retail is going to be really
81. 12:20interesting with all this AI stuff all right hi what's the best for help with python code I don't know if you I don't
82. 12:27know what you mean with help for python code if you mean um You Want To Learn Python coding that's different than if you want to do python
83. 12:38coding for you which is different than if you want to kind of pair program with AI if you want to pair a program if you're like uh
84. 12:47you know putting projects together I would say uh github's Copilot seems to be um seems to be popular amongst the kids and uh
85. 13:00if you just want to use um this stuff to write python code for you chat GPT is really good I hear that
86. 13:08Bard's good but I would say chat gpt4 is of all of these things gpt4 is still the best I'm holding stock with them who are
87. 13:16you holding stock with oh palantir oh uh yeah I would then then as an investment yeah I I'd say they're it's probably worth holding
88. 13:29stock in palantir like whoever whoever sort of figures out the [ __ ] they're trying to do is not bad
89. 13:35I I mean it's not it's not it's not bad or good uh whoever wins that that race is gonna be they're gonna do good
90. 13:43financially I would think uh this will definitely be more transformative than the worldwide but yeah exactly the World
91. 13:49Wide Web super transformative uh but if you think about what the technology was it was actually quite simple right a
92. 13:55page with hyperlinks in it that connect to other pages that transformed a lot of things right it made Communications immediate it it
93. 14:04you know reduced the need for travel um it's the foundation of all the stuff we do today it's the foundation of a lot
94. 14:10of this stuff the ability to use this stuff um that said the the the basic product that the World Wide Web was was quite simple this stuff
95. 14:20is quite sophisticated and it's moving faster a lot faster um so the the work product of these things is significant
96. 14:31love your shirt thank you two shirt compliments tonight very nice it's hard to read the text against but Pi dot AI
97. 14:39is very empathetic to talk to but asks too many questions if you want to get [ __ ] done totally agree with that what's your name I am Aryan
98. 14:47all right wait I can't I can't read against my shirt I might need to go change it um yeah Pi dot AI if you get the PI app if you search on
99. 14:55Apple for pi personal assistant you'll get that app you put it into speech mode and you can talk to it it's really good
100. 15:02to talk to but if you want to actually get [ __ ] done it's a pain in the ass so so you want to get [ __ ] done connected
101. 15:10to the internet that would be barred or Bing you want to get [ __ ] done not connected to the internet gpt4 gpt4 with
102. 15:17code interpreter and Claude um if you want to experiment and build your own [ __ ] use the Llama 2 models
103. 15:25from from uh meta and if you want to make Bots the PO the post site is pretty good for making Bots that you can share
104. 15:33if we are familiar with promoting is it worthwhile to find a job remotely doing promoting with AI wait if oh if
105. 15:43you're familiar with promoting is it worthwhile to find a job remotely doing promoting with AI yeah I don't
106. 15:51know a lot about that world what I would say is if you know a lot about promoting get AI literate and start figuring out
107. 15:57how you can do what you've done historically better faster and cheaper right and so it's not necessarily can you get
108. 16:06a job with AI what I would say is if you're a good promoter become a way better promoter leveraging these tools
109. 16:12right so you might use mid-journey to make images for posters and you or you know social Graphics you might these
110. 16:19things use these things to write the copy you might use code interpreter to generate you know CSV files full of copy
111. 16:27that you then take over to uh canva and do bulk create and just create you know tons of promotional content and put that out in the world um
112. 16:40you know you could use you could use the fact that Bing or Pi or Bard are connected to the internet to find
113. 16:47Opportunities I don't know I mean the the short answer to any question of you know I've got I'm in this kind of
114. 16:55profession should I be using AI the short answer to it is yes and but I wouldn't even necessarily limit yourself
115. 17:02to a given profession what I would say is if you're a promoter if you're good at doing things in that world you have a
116. 17:12lot of skills right you can't be a promoter and kind of be you know singular in your focus you you got to do a lot of [ __ ] and do it well
117. 17:19learning how to use AI to do those things really well you may discover something that this stuff amplifies what
118. 17:27you do in a way that you couldn't even imagine I'll share I I just wrote a if you go to my LinkedIn profile I just uh wrote an
119. 17:35article today about I had a fairly profound um interaction with chat GPT today so so this guy I'm volunteering for a
120. 17:47United Way thing for a small business thing and I'm going to go tell them about storytelling and so the guy running the program it's
121. 17:55a cohort for small businesses for first-time Founders and he he wanted me to put together like a little mad libs
122. 18:03thing for people to do their elevator pitch and so we talked and you know he said the the people are very nervous they're
123. 18:12first time Founders like the idea of pitching makes them really nervous they're resisting it they don't want to
124. 18:18pitch their business they don't want to feel sleazy and so I thought sure I can you know I've done a ton of those in my day so I
125. 18:26can whip up a little mad libs kind of thing like here's what my business is about here's here's the problem and
126. 18:31here's the solution and you know here's what my business is about um and then I went to chat gbt and I wanted to get
127. 18:42there's uh a guy named Simon sinek who's got a famous Ted Talk called um um what is it uh lead lead with why
128. 18:54so so instead of talking about what you do talk about why you do it so I went there just to get that quote and then I started
129. 19:04kind of brainstorming with chat EBT I was going back and forth about how I could incorporate this quote into this little mad libs thing
130. 19:11and before I knew it an hour had passed and I completely like redefined I got rid of the term elevator pitch and I
131. 19:22came up with the term your belief brief and it's a it's a whole framework for how someone can tell the why of their
132. 19:29story here's why I started it and and things like that and it you know it was just this amazing sort of I went into a
133. 19:36hyper focused tunnel with my ADD and you know I I went into it wanting to do just a very simple kind of search and an
134. 19:46hour later I had created this really sophisticated framework to give to these first-time entrepreneurs and so I shared that on
135. 19:55LinkedIn and and you know it wasn't so much what I wrote that was the remarkable thing it was the speed with which I went from
136. 20:04you know one one intention and then you know I got I gained a little bit of insight into that and then it just
137. 20:11opened up this whole new world and and what would have taken me a week or two to really think about and really
138. 20:18write and really put together and iterate and back and forth I did all that back and forth within within about an hour
139. 20:25and the piece that I wrote is it's like a three-page document with a worksheet and with tips and tricks on how to do it
140. 20:32and the actual framework itself and examples and opening paragraphs that describe what it is um so just you know remarkable so so
141. 20:42back to the promoting thing you likely have skills that are worth exploring how you do those different
142. 20:51things with these tools right and with the image generation tools and in the case of music maybe with the music tools the animation tools
143. 21:01um go to um futurepedia.io and start geeking out on what are some of the tools that are out there you might have some fun doing
144. 21:09that but if you're new to this stuff just start with chat GPT and uh but yes I think I think it's probably you know Wise to get yourself AI
145. 21:19literate all right um oh good you're here I am here The Tick Tock why do older people say
146. 21:34hmm say that before any technical application I do it ironically as a joke um because I'm a bitter cynical Gen X are you a little [ __ ]
147. 21:48but but we used to say back when I started agency.com in the mid 90s um we used to have a thing like like uh
148. 21:59I need to get the internet installed on my computer oh yeah did you just start I did just start I keep hearing about AAA AI
149. 22:12automation agencies is that a big opportunity new here hey John how's it going um I think AI automation agencies I the
150. 22:25fact that there's an acronym for it already just blows my mind um basically if you can use zapier you can
151. 22:34get people to pay you to do [ __ ] right now um I think AI automation agencies are a short-term opportunity yes
152. 22:46I like here's here's my deal this this shit's moving so fast that whether it's an agents like I wouldn't I if I were
153. 22:55going to start an agency right now I'd just call it an agency and and or I wouldn't even call it that I'd call it
154. 23:02like the like I've got the AI learning lab I I also put together a project called the AI leverage lab which I was
155. 23:09just going to turn into an agency um but you know I've got too much too much other [ __ ] going on including the company I'm CEO of um
156. 23:18and and I think right now if if you want to get work right now you can do it in a couple of ways you can teach people how to use
157. 23:28AI you could do workshops um you could build stuff for people if you're not super technical you could
158. 23:36build stuff for people using automation tools like zappy excuse me zapier but I wouldn't lock yourself into any
159. 23:45particular flavor of how you're building [ __ ] for people or advising them because it's just moving too fast right so if you say
160. 23:54we're in the AI automation agency and all of a sudden autonomous agents get their [ __ ] together and now instead of
161. 24:00having to hand program automations you just go in and you say I want you to do [ __ ] for me and the Machine just goes off and does that
162. 24:11um your [ __ ] out of business right but at that point then you might need to understand how to use those agents and
163. 24:18how to string them together and that might not be an automation agency that might be an agent agency so so now I'm
164. 24:25an AI agent agency not an AI automation agency still Triple A so so again I think that I think that the
165. 24:36the opportunity over the next 18 months is so vast that that the only real requirement right now is don't sit on
166. 24:45the sidelines like spend hours I'm spending two to three hours every night just you know rambling my [ __ ] mouth here
167. 24:57just to keep my head engaged in this stuff um and and to get other people up to speed on it and and start playing with
168. 25:04it so so if you haven't go here and start playing if you're thinking about doing that go go build an Automation in uh
169. 25:12in in zapier like if you've got an agency background a good way to start like the thing things that I've built
170. 25:18over the past six months have been content generation engines where I'll create a Google form with some inputs I'll take those inputs
171. 25:27incorporate them into a prompt send that off to open Ai and then either post that to a slack Channel send it in an email
172. 25:35post it to a Google spreadsheet you can knock one of those out in an hour and and it's really good practice to
173. 25:46work on your prompting and work and understand the concept and then yeah go tell people you can automate [ __ ] for them or just go find
174. 25:55out what problems people have and say hey can I help you solve that problem if you're just getting up to speed just go
175. 26:00offer to help people do a couple for free help a neighbor she's having a bake sale she wants to organize [ __ ] she's overwhelmed hey can
176. 26:09I help you with that I'm learning AI okay I don't know what you're talking about but sure and then use these things
177. 26:15to make her project plans and spreadsheets and all sorts of happy horse [ __ ] so so start at whatever level you're at
178. 26:22do it for other people let other people know that you're doing this stuff and then yeah I think it's a good idea
179. 26:28dude have you not learned about the shirt issue yet yes I have yes I've learned about it and I also call it the tick tock I'm just old what
180. 26:37do you want from me people we're not like that anymore oh librarian we're not like that anymore oh spinstery spinstery kind of oh Mr Shannon
181. 26:49yeah I know I I had I had just too many of those Librarians in I had them in high school didn't have any in College college is fun
182. 26:59high school I had him um all right woohoo so happy to be back on your channel I was in Castle Rock at
183. 27:06my friend's house oh that's cool Winston welcome welcome welcome back um I think I'm gonna go change my shirt
184. 27:10I'll be back do me a favor share the live tap the like screen maybe get some people here and then when they come in
185. 27:16and like why am I looking at shelves you can just tell them that's what this channel is all about
186. 27:48foreign all right is that better who's bitching at me
187. 28:02about the shirt uh um someone was no costume change Mr Shannon just joking I still can't read oh I can still read I
188. 28:16know I could sort of read but this is better all right we're good okay uh the answer to the question coffee again
189. 28:26I haven't had whiskey in a while agua and a little bit of crack cocaine we don't joke about drugs we don't do that around here
190. 28:42all right Chachi PT is great for debugging code yeah hey retro Punk what's happening yeah chat GPT amazing for debugging code amazing for
191. 28:50documenting code amazing for explaining code I've done that before I took a sequel of a fairly long sequel query and
192. 28:59I popped it into chat GPT and I said explain what this is doing I don't know SQL and then it explained it and then I said
193. 29:06oh could you fix some [ __ ] and it did and it fixed it wrong first and I only found that out the hard way and look
194. 29:13here's the report it's done and they're like these numbers don't look right it was like right and I definitely checked that
195. 29:23um one of my new friendships I made while I was in Colorado was a Top Gun pilot inspector uh instructor at the USA
196. 29:29AF that's pretty cool yeah those guys they'll [ __ ] buzz Denver every now and again you'll just feel like
197. 29:36like four jets flying in formation Blast by and it's pretty cool hey you got a backup account I do not have a backup account
198. 29:46um I should have a backup count when they for when they kick me off but I'm only off for seven days they give me
199. 29:51like the slap on the wrist one the the problem with it is like like I'll just be doing my live I'll be talking about
200. 29:58whatever going back and forth and all of a sudden just go you've been banned and I'm like you know and then I uh whatever I I uh um
201. 30:08resist it what's that called appeal it and then they say no [ __ ] you you're still you're still an evil person but
202. 30:14they don't tell me exactly they tell me what rule was broken but not what I actually said to break the rule so I
203. 30:21don't know how I as a responsible adult am supposed to learn if I don't know what the [ __ ] I did um I trade Futures what AI program
204. 30:33would you recommend um I would recommend not taking any sort of investment advice from a guy with an acting degree that's what I would
205. 30:42definitely recommend I'll tell you the I'll tell you the things that I that I think are interesting right now
206. 30:53and it the the here's the deal it's so impossible to know this stuff because things are moving so fast and I think that the I think that the
207. 31:05the big step change Innovations are likely going to come from unexpected places so I can tell you what I think is going
208. 31:12on right now so so Chachi BT um they have their [ __ ] together Sam Altman's really thoughtful I feel like I feel like they are
209. 31:22steps ahead of the rest of the industry both in um how they're training their models how they're thinking about alignment how
210. 31:30they're thinking about scale um how they're thinking about um managing through the disruption that these tools are going to cause
211. 31:44so if you look at Sam Altman's World coin I think that's a really interesting project um and if you don't know what that is
212. 31:51it's basically is cryptocurrency and identity blockchain project that and payment system um that is about verifying human beings
213. 32:05with Iris scans with a thing called the orb which is [ __ ] terrifying but it's about it's about redistributing some of the insane wealth
214. 32:17that Altman is predicting that these AI the prosperity that these AI tools are going to generate he's basically
215. 32:23committing to take a big chunk of that from open Ai and and give it back to the people that he's going to do that
216. 32:28through World coin so I think they're really interesting Claude is anthropic um a couple of things about anthropic
217. 32:37their ex open AI people so they're smarter than [ __ ] I just watched the the an interview with the CEO last night
218. 32:45um he he's wicked smart and um but they've got a thing called the Constitutional learning model which openai uses humans to do their
219. 32:57um training Claude is using this thing called the Constitutional learning model feels like it has future to me
220. 33:04um Google is way on their heels they're trying to catch up but I wouldn't count them out but I wouldn't I wouldn't lean
221. 33:09into them right now at all meta just open sourced llama two pie is super interesting because of uh they just raised 1.3
222. 33:20billion dollars uh and it's a deep mind founder is the one that created inflection so PI right now looks like a
223. 33:27little toy but I think that they're going to be a very significant player um I think Runway ml is interesting I
224. 33:34think mid-journey is by far the best of the image generation tools but their CEO is a bit of a clown and he's like ah
225. 33:41[ __ ] copyright he's gonna get his ass sued into Oblivion so I'd sit on the sidelines with that one
226. 33:47um I think stable diffusion stability AI is really interesting um they're open sourcing all their Tech so I think they're going to get their
227. 33:54asses sued into the ground too but because they're open sourcing things that might be a hedge against lawsuits uh um
228. 34:03I don't know Microsoft is playing like a startup again so they're interesting Nvidia seems to have a [ __ ] Monopoly
229. 34:13um amd's trying to break into the into the GPU Market but whatever the [ __ ] barrier is for scaling up chip production
230. 34:24um like like Nvidia seems to have the entire developer Community locked and you know but you know they got a
231. 34:31trillion dollar market cap so I don't know what the upside is there but they're [ __ ] killing it um Apple
232. 34:40so they're apparently so a couple of things built into iOS 17 is voice cloning it's in there right now in the beta that you can play with um
233. 34:51they've got they're rumored to have created a a language or a framework called Aja I think it's called Ajax uh
234. 34:59for large language model creation and they've got something called Apple GPT that they're testing internally and um
235. 35:10they're spending the bulk of their 24 billion dollar r d budget 24 billion dollar r d budget which is bigger than the market cap of open AI
236. 35:24Apple's spending like their couch change all on AI on generative AI so I have a sneak in suspicion that Apple's gonna
237. 35:33come out [ __ ] swinging and you know they've got this spatial Computing thing that's going to be heavy in AI I think they're just going to
238. 35:42bleed out all that AI technology apples Apple's interesting because they're going to be able to they've got credibility with privacy and
239. 35:50they've got the technology for privacy and they've got the technology to run machine learning [ __ ] locally you're not
240. 35:57gonna have to send it to Apple so Apple's going to be able to legitimately say you can GPT the [ __ ] out of all of your
241. 36:05data and we're not going to steal it or give it to the government and I think they're the only ones that can say that
242. 36:12um they've they've taken a pretty hard line on that [ __ ] so um yeah I don't know I wouldn't go invest in any of that [ __ ]
243. 36:22on my advice but those are the ones I'm looking at all right uh he was in the Air Force for 20 years as a fighter trainer now just
244. 36:29flying as a commercial pilot yeah a lot of the a lot of the Air Force dudes end up going commercial it's funny I was
245. 36:34talking to a commercial pilot once and and we saw we we hit some some weather and he he like he flew through it fine and
246. 36:45and I was talking to him and he told me he was in an Air Force fighter pilot and uh I said what's it like flying one
247. 36:53of these things he said it's like driving a bus he was like really disappointed and I said well we hit all
248. 36:59that weather there like did that get your adrenaline going at all and he was like no so you know I there's probably not much else short of
249. 37:12driving an F1 car that's like flying a you know nf-16 or an F-18 so so yeah all those guys gotta go commercial and
250. 37:23you know I suppose the the ones that are still Adrenaline Junkies go do the air shows and things like that but it's
251. 37:30crazy that's really funny all right are there any decent local hostable large language models you can train yourself
252. 37:40um yes but I'm I haven't played deep in that world but go so where I would start is go to Reddit and just you know search for open
253. 37:51source large language models I'm sure there are two or three subreddits that are just dedicated to that [ __ ]
254. 37:57the developer Community right now has kind of lost their mind on llama 2 so you can go play with the Llama 2 models
255. 38:04at labs.perplexity but if you want to go train your own [ __ ] um you know just just go hang out on Reddit look at the projects that they're
256. 38:13doing um there's a project called GPT for all that lets you um use lots of different large language models if you want to start training
257. 38:23large language models on your own data then I would go look at a project called Lang chain which which can chain different
258. 38:31components together including the embeddings and Vector databases that you search against um Lang chain is quite powerful
259. 38:40um most of the developers that I know are using it that are doing any any any of this sort of stuff um I'm I've got an instance of it in a
260. 38:50like a Sandbox for my company storyvine I've got another project that I'm doing and the developer there is using it
261. 38:56pretty extensively he's building some really cool stuff with it so um but there are lots there are lots and
262. 39:04lots and lots of Open Source large language models that you can train or that other people have trained that you can use
263. 39:13some of them you can install on non-optimized computers like M1 Max and up and uh you know like a simple GPU on a PC and up
264. 39:25I think within a year we'll be able to install GPT 3.5 um equivalent models like just locally on our phones locally on our laptops so
265. 39:37it's it's gonna be it's gonna be a something it's going to be quiet something and again pay attention to things like uh
266. 39:46like what Apple's doing apples Apple's Gonna Come Out Swinging and and there they're gonna they're gonna be
267. 39:54I wouldn't be surprised if Apple finds a way to reinvent how we think about these large language models
268. 40:01either with autonomous agents that you know like personal assistants that do you know imagine if Siri didn't suck and don't listen to me
269. 40:12she keeps turning herself on every time I say her name Simon is great Simon cynic yeah he's amazing um anyway I was very busy with about 30
270. 40:20people at my friends and I wished I could have met with you too oh yeah if you get to Denver drop me a line you can
271. 40:26come to we I do an AI Salon that's the other thing if you're if you're into uh into this stuff um I started a group called the AI Salon
272. 40:35you can go to the salon.ai and that's it's a link tree that's got our Meetup our Discord server and our LinkedIn group
273. 40:43and like our mission statement and [ __ ] like that and we meet every other Tuesday we have a meeting coming up this
274. 40:50coming Tuesday the whatever it is the 15th and John Knack from adobe's AI group is going to be speaking there he
275. 40:57spoke with us back in January and they hadn't done Firefly yet they were just they were like we're not sure
276. 41:05if we want to do this AI thing like we're not sure if we want to like lead into this generative AI thing because at
277. 41:11the time it was very very like the artists were [ __ ] pissed off like way more than they are now like like I knew
278. 41:18some artists that were were doing some AI generated art projects that were getting [ __ ] death threats
279. 41:25um so Adobe early on was like we're not sure if we're gonna do this we might dip our toes in the water and then like two
280. 41:31months later they came out with firefly and they've just been Off to the Races they got themselves a Discord server
281. 41:37they got Firefly up and running they rolled that [ __ ] straight into generative fill within Photoshop and uh and the the other thing about
282. 41:45Adobe right now is all of their models are trained on public domain and Adobe stock licenses that they own
283. 41:56um so they haven't trained anything on you know active copywritten artists and they're they're creating a system to
284. 42:03allow artists to opt out of being trained up into being trained attribution models compensation models so I think adobe's doing it right so
285. 42:11anyway John's gonna come to the salon next Tuesday and he's going to talk about whatever he's allowed to talk about but
286. 42:19he'll probably give us a little insight into like what I want to hear from John is what have they learned
287. 42:25thank you David gold appreciate that Gould probably not gold there's a u in there so and I'm an old guy the David Gould The Tick Tock um
288. 42:40was I gonna say yeah yeah I wanna I wanna hear from John like what have they learned because like they just sort of you know
289. 42:47hit the ground running and they didn't seem to catch the Wrath of the art community because they did it right like
290. 42:52they talked to all of the major issues um so I'm just curious what they've learned what they're excited about
291. 42:58any little sneak peeks they can give us about what's coming and John's great he's he's a guy I used to work with him at agency.com he was
292. 43:06actually an intern at agency.com and uh if anyone often became like a big [ __ ] deal he was a product guy at
293. 43:14photoshopped for years then he went to Google for a bunch of years now he's back at Adobe doing AI stuff
294. 43:21so that should be interesting so yeah so come to the salon and yeah if you're in Denver ping me let's go grab a coffee or something
295. 43:29all right let's see moving on Tick Tock jumped me to the bottom I'm trying to figure okay there's Simon is great all right
296. 43:39um hello from Maria Lupita from the Laguna Beach area welcome welcome Maria glad to have you here nice Emoji work
297. 43:49there I'd say 9.5 out of 10 on the Emoji work it's solid the work acceleration from oh the work acceleration from AI is so
298. 44:00exciting I know Aaron it's [Music] well it's funny it's funny like the story I don't know if you heard earlier
299. 44:06when I talked about I just had this little 10 minute project to do today and I went to Chachi BT to do a little
300. 44:11research and then it just ballooned into this sort of hour and a half long um brainstorming session with chat GPT and
301. 44:21I created something really remarkable so in that case it actually slowed the speed of the work but it accelerated the
302. 44:28the quality of the work dramatically like like the guy that asked me to do something was just expecting like a
303. 44:36little mad libs kind of paragraph and I sent him a three-page [ __ ] document that redefined what a what an elevator
304. 44:42pitch was for a first time founder I'm like that's pretty cool um so so yeah it's I I just I I swear to
305. 44:52God man every day I'm just like really really you can do that I don't I don't know when that's gonna go away I'm trying to think back in the
306. 45:02early days of the web I mean in the early days of the web the technology was so basic that the things we got excited about
307. 45:11tended to be like incremental technological improvements like so when I first started building websites you
308. 45:20could center text but you couldn't Center images the center tag didn't apply to images so so all images were left Justified so
309. 45:29that drove us [ __ ] crazy so we invented a whole little series of pixel High invisible images to scoochar
310. 45:36GE our images to the right so we could center our images with the text um and then when you know you know when
311. 45:43the center tag came out we were like and I remember when when um flash came out and we could do animations because we had been we had
312. 45:54been writing uh Pearl scripts to to do animations where the the animation was actually a series of calls back to the
313. 46:03web server and that was the animation and then animated gifs came out and Flash came out and we were doing games
314. 46:09within banners it was it was crazy but all of that stuff were like incremental improvements that we got excited about here it's kind of like
315. 46:19I I mean they're they're exciting technological things happening but it's it's almost more like the more people and companies start to
316. 46:29get their head around this stuff they're like well here's how we're using it and then you look at it you're like oh [ __ ]
317. 46:33I've never even thought about that so yeah it's it's just crazy times right now it really is I find that AI is helping me bring my
318. 46:44vision out of the ethers and into the tangible Denise that that's I that to me feel that sorry I I agree with you so
319. 46:53much I literally can't talk um I think that's the biggest thing for me that chat GPT does is I'll have a whisper of an idea and within three
320. 47:08minutes five minutes that Whisper of an idea is at 80 percent execution now it still takes me very often two hours to get the last 20 done
321. 47:19but like that's the reality anyway we're creative people right we're constantly you know pushing the boundaries of deadlines but but it
322. 47:29that ability to get something out of your head quickly and articulated quickly and then and then and to have
323. 47:37that idea Spur new ideas that you can bring those to life quickly like I find myself knocking out five or six major
324. 47:44Concepts in 30 to 45 minute periods and then I kind of curate those and some will park some will toss oh that was
325. 47:54just a stepping stone that's I totally that like that's it for me that's it for me is it's it's yeah it's it's a it's a
326. 48:06it's an idea amplifier an ability amplifier Ian my wife is sleeping so I can watch you what do you mean you're watching Kyle again
327. 48:18you don't want to watch TV with me yeah my wife was like oh you're going on Tick Tock again so I get it I totally get it Claude and
328. 48:28Poe gave me the McAllister moment the other day oh Nicholas tell me what happened what so Claude info so I'm guessing Claude it
329. 48:37was about the hundred thousand token context limit um and uh Poe I'm not sure maybe it was the the Bots
330. 48:45um the the Kevin McAllister moment if you don't know so so I I say on the channel a lot that you should use these
331. 48:51tools long enough to have your Kevin McAllister moment which is the Home Alone moment um another handsome shirt thank you
332. 49:00belief belief brief that's F and Brilliant so the bear thank you I I'm I'm really excited about the belief brief thank you very much um
333. 49:11yeah it started out as an elevator pitch and then it I went to Simon cynics y thing and then I was saying is there a
334. 49:18framework for how to talk about the why and then it just evolved and evolved and evolved and so I had this whole
335. 49:23framework and then it said here's your elevator pitch uh framework or whatever it was recipe and I'm like elevator pitch still kind
336. 49:33of sucks like that still feels like this intimidating thing and so I did a naming session with chat GPT and I probably came up with
337. 49:4260 or 80 names I was doing them in batches of 10. and it it was just like shitty batch of names shitty batch of names but then
338. 49:50there'd be like one little phrase in there I'm like oh that phrase is good and I sort of pull that out and I do iterations on that
339. 49:57and it was I don't know it was like idea 84 was belief brief and uh I was like that's good so yeah I know isn't that fun
340. 50:08um zapier versus make versus python um I would say the bear are you stalking me at the AI learning lab the kids you kids you kids
341. 50:22no fraternizing amongst the amongst the uh the audience um I don't give a [ __ ] Fred nice all you want um zapier versus make versus python
342. 50:35zapier's been around for longer which is good and bad it's good because and I don't know why I'm talking to you with a
343. 50:41pen but listen little mister listen little Mr Ian um zapier's been around longer so it's got more um connections more API hooks to
344. 50:52different Services it's got like more than five thousand make is newer and and has a sexier interface but they're
345. 51:00they're both kind of exactly the same thing right you've you've got you know it's kind of a linear flow that can branch um
346. 51:10and then python is if you know python like here's the thing with no code the the problem with no code is that
347. 51:19you have to learn their [ __ ] platform right you have to learn all the intricacies of the platform and what
348. 51:25works and what doesn't and like every little module has its own [ __ ] rules and it's probably if someone is really
349. 51:34competent with python they could probably do what you do in zapier faster than I could do it in zapier but if I
350. 51:43don't want to [ __ ] Learn Python I'm going to use zapier right and I just I don't know make I've looked at make I
351. 51:50like how it looks but I went to build something and it just gave me one of those weird [ __ ] automation errors
352. 51:57like I whatever the [ __ ] it was it's just some stupid [ __ ] error that if I had read the manual um I wouldn't have had the issue but but
353. 52:06um uh but I don't read manuals because that's just not how I roll I've never read manuals I have a son that reads
354. 52:12manuals he drives me [ __ ] crazy I'm like there's three screws in a bolt what are you reading the manual for what
355. 52:19could we possibly [ __ ] up and then of course I [ __ ] some things up with that attitude but there's other
356. 52:26times when I just build [ __ ] and it was fine and I can use my brain I don't need a [ __ ] manual I can figure it out anyway
357. 52:35thank you for the heart um so I'd say whichever one you're more comfortable in they're probably all in the neighborhood of the same
358. 52:44morgie calling you all in thank you I assume that means you shared this thank you very much well it's been blessed with good eh
359. 52:53oh I see you're talking to each other cool wait oh Morgan in a few hours is the all-in podcast John interesting
360. 53:02crypto Joe lady that went viral on the plane saying the guy wasn't real is said to be a marketing exec yeah those
361. 53:09marketing exists execs dear little cuckoo advertising is full of special people they're special the big Egos and big crazies
362. 53:22but they're smart and fun yeah ad people are nuts marketing people are crazy grades and crazy but they're storytellers what the [ __ ]
363. 53:34do you want your storytellers are gonna have a little scroll boost every once in a while how else could they tell the story
364. 53:42they're The Observers of society you know what I mean we're all out there doing the storytellers are sitting back
365. 53:49going huh very interesting very interesting well that Line's going into play I use I'll tell you what if you've never
366. 53:59been to New York or if you haven't been there in a while I [ __ ] miss New York I haven't I haven't been there in way too long
367. 54:07um one of my favorite things in New York was just go to go [ __ ] anywhere go to a diner go to the subway and just
368. 54:15have a notebook open and just listen oh my God oh the lines the lines all right I went to block and bottle
369. 54:29in Castle Rock they went to a pretty large potty party at Casa Bonita how was the food at Casa Bonita I haven't been
370. 54:35yet I hear it's good although although all the servers went on ahead and sued uh the South Park guys because I guess
371. 54:42they weren't paying them or they weren't doing what they promised or they weren't opening or something
372. 54:46this guy is cool I don't know who you're talking about ah it's not me I say The Tick Tock that's not cool any decent
373. 54:54locally hostable large language models we already talked about that oh by the way if I'm not answering your
374. 55:02questions it's because I'm reading all the questions so I'm I'm now 38 messages or more behind so don't go hating on me and don't go
375. 55:11asking the same question 47 [ __ ] times because as I scroll through it I'm gonna see him we had a situation last
376. 55:17night with a guy with an unpronounceable name ask me questions 462 [ __ ] times it got ugly I do not suffer fools lightly especially
377. 55:27when I can't read their name I don't know what's going on with my voice tonight it ain't it ain't something something is awry up here
378. 55:40and something is a pumpernickel huh it's comedy all right we have fun we have fun we like to have fun I like that shirt but this record is
379. 55:55also good I know it's good right albums the vinyl spin those Tunes yes much better you were bitching about the shirt Kyle
380. 56:07oh you were bitching about the shortcut I know Emilio's wife all right much better shirt for reading thank you yo yo which AI can take a zip
381. 56:15file and read through it that would be claude.ai and um code interpreter not Claude chat.openai.com has code interpreter
382. 56:27that can read a zip file I think Claude can also read a zip file let me go check I know Claude can do PDFs
383. 56:42where's an archive come on I got I now got one in here somewhere [Music] wait any archive no archive
384. 56:56name contains where I don't get it I just don't get it people what's happening
385. 57:14oh maybe it can't um oh it can't okay so
386. 57:30um okay chat gbt using Code interpreter you can upload a zip file to that and it can be full of a bunch of other files PDFs and csvs um
387. 57:44code interpreter also just you can now do multiple multiple files up to code interpreter and I'm happy to show that here if
388. 57:53anyone wants to see it have I seen hey Jen Avatar pre-release today I tagged you on LinkedIn oh cool retro Punk yeah
389. 57:59I saw that I've seen hey Jen is is clearly getting ready to release something because I've seen them mentioned like 10 times in the past two
390. 58:07days so their PR firm is is earning their stripes um so so something's up with hey Jen um I saw a tick tock today where a guy was
391. 58:18talking about hey Jen and canva um I've seen a couple of people talking about it on X so yeah how can we ever
392. 58:28prove AI can be sentient if we can't solve the hard problems of consciousness well user 588-59-3919 um by the way 0.5 on the name come on
393. 58:40seriously like like Bob at least right anyway um Bob five eight eight five nine three nine one nine um
394. 58:56I don't know what you mean about if we can't solve the hard problem of being conscious we are I I would say certainly
395. 59:03in America we are largely unconscious um Society right now we're all living on our amygdalas just [ __ ] triggered every minute of every day
396. 59:15um so it would be nice if we got back to a little more um consciousness I Here's my thought on sentience and AGI
397. 59:28and is it smarter than us or not all that sort of [ __ ] from where I sit is [ __ ] irrelevant like who cares
398. 59:38okay at the point at which we go okay AI sentient now what changes nothing right these things are already more
399. 59:47capable like these things already know more than any of us individually will ever be able to learn ever in our lives
400. 59:55if we dedicated the rest of our lives to learning all of the [ __ ] we could not keep up with where these things are right now
401. 1:00:02we can't code as fast we can't write as fast we can't write as comprehensively and these things are in their shittiest
402. 1:00:10form that they will ever be so at the point at which we say okay now it's sentient now it's conscious now it's AGI
403. 1:00:21it's just it like I feel like at this point it's just a Continuum right we're we're already at the point and it's even
404. 1:00:28funny like there's people that have got there are all these different um like definitions of when these things
405. 1:00:37pass the Turing test it's like okay well it passes the Turing test for five minutes but I want to talk to it for
406. 1:00:44four hours and if it if it can convince me that it's not a computer for four hours that then it's past the touring
407. 1:00:51well okay like whatever like these things are no longer Alexa in Siri right these things are Way Beyond anything that we have historically
408. 1:01:02called a chat bot Way Beyond right these things are transforming work and jobs and all that stuff right now in
409. 1:01:15their shitty janky duct tape ass form um so I agree with you we can't get our [ __ ] together as humans and I don't
410. 1:01:25think it's relevant when we label that these things are more human than human or not like it's going to be good for a cover
411. 1:01:32of Time Magazine but how about IG account Instagram what I don't know what you're talking about Ian don't take any advice from Jim Cramer
412. 1:01:45you should get a backup account because I think that subsequent bands can be longer oh well they can be if well if
413. 1:01:53they if if they do it again I I'm trying to be I'm trying to be a little careful now I I think I know the areas that I'm
414. 1:02:00not allowed to dance around it basically yeah I'm not even going to talk about it because I can't even talk about it or
415. 1:02:07let's shut this thing down any any I any AI can read zip files and interpret the code okay answers yes
416. 1:02:16and what we'll do so here we are at chat gbt um in order to use code interpreter you need access to GPT plus when you go to
417. 1:02:27the gpt4 area you'll see code interpreter and plugins if you pay for chat GPT plus and you don't see these
418. 1:02:33that's because they're hiding it from you because they're evil little bastards so to turn that [ __ ] on go to your settings
419. 1:02:43and go to Beta beta features and then you turn them on why are they doing that because they want to protect you from yourselves I
420. 1:02:52don't know they're pretending like that's safer than not turning that [ __ ] on okay so now we're in code interpreter
421. 1:03:00next to your little doodaddy here is a plus button and I can now upload uh anything videos images uh uh zip files csvs PDFs
422. 1:03:16so let's go grab some new data so I'm going to run over to a site called kaggle kagg Le wait k a
423. 1:03:32g g l e okay and I'm going to go to data sets and then we're gonna go find us a data set oh we got one here oh it's only 7K
424. 1:03:44we we have a data set for McDonald's nutrition that would be a fun one let's see top 200 Spotify songs database Magnus Carlson's
425. 1:03:55games you know what we're gonna do we're gonna do a filter here we're gonna we're gonna get some decent sized ones we'll go 10 megabytes
426. 1:04:07to a hundred megabytes and let's grab csvs all right things on Reddit
427. 1:04:21S P 500 Global education statistics all indicators that sounds interesting um TV and movie metadata with genres and ratings uh
428. 1:04:36um Metro train data set let's do Global education okay so I'm gonna download this 18 megabyte file looks like it's down is it down
429. 1:04:51did this just come in today at 9 10 yes okay so we're gonna go back to your yield chat GPT you can make money
430. 1:05:06we're gonna grab archive two I'm gonna open it okay so that's gonna now upload that file so it's taking a little little moment here and
431. 1:05:18we're gonna type in here [Music] please open this archive and examine the files within it and tell
432. 1:05:33me what might be interesting about them so you can just do that you can just say go do [ __ ] without being very articulate and then
433. 1:05:45it goes and does it so so what it's actually doing here is spun up a virtual machine inside this chat session
434. 1:05:55and it's writing and executing its own python code and something is up with the focus again this new phone's been better on the
435. 1:06:06focus but this is oh it's because hang on my hand is in there let me see if I can get it so my hand's not in the shot
436. 1:06:14and not really let's see can we do it there get rid of that oh hey hang on please just uh I'll be there okay okay maybe that'll work
437. 1:06:30so it's it's writing and executing its own python code um it says this archive contains 15 CSV files their names seem to be mostly uuid
438. 1:06:41based except for one that mentions education statistics um one what do you want me to do I can check the first few rows of each file
439. 1:06:55investigate any unique patterns or outliers would you like me to start by examining or is it uh yeah just uh
440. 1:07:06you decide that's what I'm gonna say you decide all right let's start by examining the first few rows of a selection
441. 1:07:17of these files and so again it's going off so it's it's starting with the one that it thinks is the most interesting
442. 1:07:26it's it's not starting with the uuid files and again what it's doing there is writing its own python code so it says here's a quick overview
443. 1:07:39unfortunately I encountered an error there's literacy proficiencies of adults
444. 1:07:53language proficiency of second grade from the initial examination it seems these files might contain educational
445. 1:08:01statistics possibly for various countries and years the data is potentially related to proficiency levels blah blah to further understand
446. 1:08:08this data we could check the number of unique countries examine any trends or examine files to the C so let's say um trends
447. 1:08:19we want to look at Trends and so let's see what it does so so now what it's going to do is it's going to go like what's
448. 1:08:26why this feels like the future to me is my input here is very minimal including my my knowledge of data analysis and how
449. 1:08:37to even deal with data sets right I don't know how to do that but I can come in here and with a handful of things
450. 1:08:44just turn it loose to start working so this is an example of an autonomous agent um it's not really going out and you
451. 1:08:53know using other tools and things like that but it's running this you know virtual computer system within my chat
452. 1:09:01session and it's got access to python libraries and it understands python code and it understands how to problem solve
453. 1:09:10and so it is remarkable um I'm in the middle of a project right now where I've got an old Legacy Excel spreadsheet that just isn't
454. 1:09:22working I can't get it to work and so I'm doing my data analysis for this report that this spreadsheet was supposed to do I'm doing
455. 1:09:30it all in code interpreter and you know it takes me like 10 minutes I go in I grab raw data I turn it into tables I turn it into
456. 1:09:38charts and graphs I give it to the account people and they go deal with it and and I don't have to do any coding I
457. 1:09:43don't have to learn SQL SQL queries all that sort of stuff so let's see what it's doing it seems to be there's an error so they're just made
458. 1:09:52a chart with nothing on it and it observed that there's no there's no data available so it made an empty chart uh I'm gonna say find enter
459. 1:10:07esting data in one CSV let's see what that does I might have picked a a janky data set here
460. 1:10:24again if I knew more what I was doing with data and Analysis that's so so one of the other things that I think is is important to recognize
461. 1:10:34me being clueless about data analysis I'm going to be able to use code interpreter like just a little bit like
462. 1:10:40a wee little bit but if someone's actually good at data analysis they're going to be able to use this thing in a remarkable way
463. 1:10:49um let's see the file contains 270 countries 140 unique series descriptive statistics given the area a few interesting paths
464. 1:10:59we can take country analysis series analysis or Trends over time let's do trends over time so if you're good at a particular thing
465. 1:11:14you're going to use chat gbt or code interpreter or Claude or mid-journey you're going to use it way different
466. 1:11:21than I use it and that's a good thing right and that's another reason I like this channel is the more people that
467. 1:11:28come in here with different points of view and perspectives like teach us all different ways of looking at these
468. 1:11:37things and it's that's what we need right now right the only way we can really learn what's possible with stuff
469. 1:11:43that's this different from how we've done things historically requires lots of lenses being looked through
470. 1:11:52all right and this is getting boring because it's like it hasn't quite figured out how to pull the data out of these things
471. 1:12:00but anyway that's that's code interpreter that's how it works there just made Another Empty chart okay we're
472. 1:12:05gonna we're gonna bail you on you on you stop generating stop just stop just stop we have no use for you code interpreter
473. 1:12:15code interpreter is useless tonight all right I'll get a different data set wait you could rip the transcript of your offering
474. 1:12:30oh of your offending video give it to chatgpt along with Tick Tock rules oh that's a good idea River
475. 1:12:36thank you David Gould what is that thing oh it's a tick tock music thing it's a 3D neon tick tick tock tick tickety tock
476. 1:12:48um River that's a really good idea take take so down download the offending video transcribe it put it into chat GPT with
477. 1:12:59the tick tock rules yeah in the middle find it yeah I could do that I don't really care thank you Kay appreciate that
478. 1:13:09what about IBM's Watson I know so so they did they did um announce like six weeks ago Watson Watson AI was it Watson AI
479. 1:13:24or Watson oh Watson X yeah they announced Watson X yeah so they had Watson and then they had Watson health
480. 1:13:30and that was going to be that was gonna change the world right thanks David Gould and um um I really do appreciate that
481. 1:13:39um and then it just flamed out they screwed it up somehow I don't I don't know what they did I mean but but you
482. 1:13:45know Watson Health was pre-transformers right all these tools that we're playing with now this is all possible because of
483. 1:13:52the Transformer that was invented in 2017. I think Watson Health was kind of on its you know on its downswing at that point wow David
484. 1:14:01thank you so much and and TM Tim good name um yeah so all these things are are leveraging Google's Transformer
485. 1:14:15technology and so it's mid-journey and stable diffusion and so are the image generation or the the video generation
486. 1:14:23tools and the voice synthesis tools and the animation tools they're all using the same [ __ ] right now which the other
487. 1:14:29good Lord David that's very that's awesome thank you so much um um what was I gonna say the uh oh so so
488. 1:14:40right now all of these things are using um you know the Transformer model but there's all these other different models
489. 1:14:46and there's models that have that are in research right now that we don't know anything about that all of a sudden are
490. 1:14:52going to spring up on the industry and it's going to change everything again so so I think we'll see um something like the transformer comes
491. 1:15:02along that creates this step change in capability and then all the developers sort of rush to to sort of catch up with
492. 1:15:09it integrate it and then go beyond it and then there'll be another one that does something different better faster
493. 1:15:15more whatever and then same thing will happen and we are definitely in sort of the hockey stick of innovation
494. 1:15:25um and we're just starting like we're just this is nuts and and it really is Chachi PT that kicked off this whole frenzy because
495. 1:15:36that came out you know machine learning had been around for a few decades and Chachi PT came out and all of a
496. 1:15:43sudden 100 million people in six weeks said oh [ __ ] this is cool and those were not 100 million machine
497. 1:15:50learning scientists and researchers and data people those were like human beings Like Yours Truly here that went this is pretty swell
498. 1:15:59I think I could even use the AI now and all the scientists are like oh [ __ ] really them oh that's a cool one the hell is that
499. 1:16:11never seen that before David Gould absolutely killing it with the action I'm gonna turn this into a goddamn MPC Channel yet
500. 1:16:23I'll do it I'll twerk and I'll go oh cowboy hat oh cowboy hat oh cowboy hat oh Tick Tock hat Tick Tock hat
501. 1:16:37I can't tempt David it's not it's not good David it's not good it's not good don't do it don't do it it serves no one it serves no one
502. 1:16:53oh good lord Bagels Bagels um how you doing okay okay holy crud on Pi dot Ai and Nvidia does have a good lock on GPS gpus and
503. 1:17:10likely on newer tpus yeah I just I like it it baffles my mind that all the chip well you know I assume
504. 1:17:22there was a lot of consolidation in chip manufacturers but like how did Intel blow it like how did Intel not get in
505. 1:17:29the GPU game and and how did AMD blow it and how did like there are like it was just like they seated they
506. 1:17:40said okay [ __ ] it video games can go to Nvidia and then the Bitcoin [ __ ] came out they're like well that's fine that's just crypto
507. 1:17:49and now it's like all of AI and like like they've had the market cornered long enough now that like all of the drivers
508. 1:17:59and all of the software infrastructure and all of the development infrastructure is all made for NVIDIA so
509. 1:18:05it's not even can you make the chips like AMD can make the chips but they're soft their their software is all [ __ ]
510. 1:18:11up if you listen to the there's an interview with uh with uh Lex Friedman and George hotz and George is talking
511. 1:18:18about wanting to use someone other than Nvidia and he was trying to use AMD chips and their [ __ ] drivers crashed
512. 1:18:26when he tried to do machine learning [ __ ] on him the things that were designed for machine learning and they
513. 1:18:32just [ __ ] crashed and he's like seriously so that's just amazing to me just amazing I want to drive an F1 car I
514. 1:18:43would love to drive an F1 well I don't honest to God I don't know if I would want to drive an F1 car I I would love
515. 1:18:49to at least start with an F4 car like I feel like the amount of stress just getting in an F1 car and and what was I
516. 1:18:58don't know if you ever saw that Top Gear episode where um Hammond gets in gets into an F1 car and he's and he's in the pits and they
517. 1:19:07train him for like two hours you know how to do this thing and they get him in the pit they start the car and he
518. 1:19:12[ __ ] stalls it like if I stalled an F1 car I just I would just sort of melt into the seat so I think if I started with like an F4 F3
519. 1:19:23maybe go to an F2 car those F1 cars are just they're just [ __ ] crazy [ __ ] nuts be fun to drive a NASCAR
520. 1:19:31too because they're just [ __ ] Beastie Wild just [ __ ] blast that thing around the two mile track left turns only
521. 1:19:43um Lang chain and meta's llm are on top use within EV uh within Enterprises yeah I'm probably gonna set up we we've got
522. 1:19:52I'm going to New York Sunday to talk to a big Pharma client I think we're gonna do three pilot projects um using llms um quite frankly
523. 1:20:02some some projects that are beyond what I you know know exactly how to do it's kind of exciting it's good
524. 1:20:13um but I'm gonna probably end up going with llama uh just for data data safety um asked yesterday but I didn't hear you
525. 1:20:23get to it sorry if I missed it what did you ask oh if Steve Jobs were alive would he partner with or compete against Elon hmm
526. 1:20:37that's an interesting one um I certainly think he would respect him I you know I think that you know jobs in Elon
527. 1:20:51he you know are both both struggle on the you know capacity to be Human Side although job jobs was pretty pretty
528. 1:21:04good on it publicly but not good on it privately um but but they're both relentlessly committed to what they believe in right
529. 1:21:14and so I think I think at a minimum he would respect Elon um I don't know that's a really good question
530. 1:21:29I don't know I don't know there was a someone did a uh was it with Elon that was it was Joe Rogan and Steve
531. 1:21:40Jobs someone did a deep fake with Joe Rogan interviewing Steve Jobs it'd be it'd be fun to do one with Elon Musk and
532. 1:21:46Steve Jobs and just you know what would they say to each other that'd be something to be Pro decently easy to put
533. 1:21:52together that's a really good question I don't know I think they I think that they would like musk isn't going after Apple at all
534. 1:22:05there's nothing there's nothing that Apple's doing that threatens what musk is about now if so so one of the things musk wants to do
535. 1:22:16with X and why he called it X is he wants to make it the everything app right he wants he wants to make it like
536. 1:22:21WhatsApp and and your payment system and your communication system and your publishing system and your money making
537. 1:22:28system and like he wants to make it the everything app Apple if they start getting into autonomous agents they've
538. 1:22:35already got the payment system they've already got credit cards they've already got the entertainment piece they don't
539. 1:22:40really have a social network but once they once they start getting good with the AI stuff I could see X and apple
540. 1:22:53um competing in the same space but I don't see musk getting into Hardware right and that's primarily what Apple's about so
541. 1:23:02I don't know I don't know it's a fascinating one damn it I scrolled at the wrong time heard you go play with llama and was
542. 1:23:10pumped there was a spot what Oh no just did it just jump me to the bottom oh Tick Tock why do you do that
543. 1:23:18Tick Tock stop here and you made me sad you made me very sad when I just start watching with the misses we just started watching um or no
544. 1:23:30she she was watching it for a while the idol it's about all these bratty musicians in La it's just very good acting but like
545. 1:23:40I just I couldn't relate I couldn't relate there was this very there's this very spoiled they're also spoiled
546. 1:23:47standing around in their skimpy little negligee with their little skinny little bodies someone needs a burger
547. 1:24:01I almost moved to LA I was gonna when I was living in New York we were living north of the city and I decided I couldn't take
548. 1:24:09Westchester that area north of the city is just weird couldn't take it anymore and I was gonna move back to New York City
549. 1:24:18come out to Boulder or go to La and so I went to LA and I went to a bunch of startup meetings and man they were just
550. 1:24:27[ __ ] weird La is a weird [ __ ] town everything's about positioning and Angling I'd share an idea with someone like oh
551. 1:24:37man I had an idea just like that man we should hang out man everyone wanted to hang out and what it really was was no one had
552. 1:24:44any [ __ ] ideas and they just wanted to hang out with people they thought had ideas or they thought had power
553. 1:24:51it was a weird town man she's weird man you know man you know what I'm saying man I don't know you ever been to L.A man it's just weird man
554. 1:25:02I'm just trying to buy time to figure out where I was in the comments here so just hang with me I'll get there
555. 1:25:10Ian I think IG is Kyle Shannon huh oh what's my Instagram I have at Kyle Shannon and I also have at Kyle Shannon
556. 1:25:20dreams which is my art project with my stable diffusion self-portraits okay I think I'm almost back to my
557. 1:25:31questions here all right oh there's the Steve Jobs one I scrolled on the wrong time heard you go play with llama and was pumped there
558. 1:25:42was a spot oh this is what I was trying to figure out before where's the salon I'm local and would love to join you for the Adobe speaker
559. 1:25:51uh altitude Tactical Good Name by the way I'll go 8.5 out of 10 on the name um we meet at the Catalyst building in Rhino so it's uh 35
560. 1:26:0213 3513 Brighton and we meet from five to seven and we're on the seventh floor boardroom and all that informations if you go to
561. 1:26:13the Meetup so if you go to the salon.ai go to the go to the uh the link tree here and sign up just become a subscriber to the Meetup we
562. 1:26:23send out the information on the day of the meeting as well so yeah come by say hi um John Knox not gonna be in person he's
563. 1:26:34going to be virtual because he's off in Adobe land I will make time to get on the salon.ai good name thank you appreciate that
564. 1:26:45which AI to make pictures animated bring them to life thank you all right let's go do that we'll go make an image and then we'll go
565. 1:26:52bring it to Life we'll do both we'll do both you're getting a bonus here this is bonus normally I charge double for this but tonight just
566. 1:27:021X not 2x just 1X you only pay the single amount of zero all right so we are gonna go here oh this was this was the the images that
567. 1:27:18I made for the for the uh article that I wrote today for LinkedIn and it was the title of the article was um
568. 1:27:27chat gbt gave me superpowers is that what it was called hang on let's have an 80d moment shall we I'll get back to the other thing calm
569. 1:27:37down you people um LinkedIn um
570. 1:27:52yeah Chachi BT just give me superpowers so that was the image I put in my article here's the article Isn't that cool
571. 1:28:02cool old image and so I did that over here in Discord so if you've never used mid Journey you do it in Discord I'm in the AI Salon Channel
572. 1:28:11and there's a mid-journey play Channel and so that's where I made this thing is that one that's good to animate
573. 1:28:19actually that one's not bad to animate um but I'll uh yeah let's let's do that one I have a feeling it's going to make these things
574. 1:28:27in the sky Fly so let's let's grab this one all right so I'm gonna open in browser I'm Gonna Save as
575. 1:28:39um all right we'll save as there okay so now I'm gonna go to runwayml.com and by the way this is goblin.tools which is a
576. 1:28:51I looked at this quickly last night I played with it a little more today this is a to-do maker a to-do list maker
577. 1:28:57where you give it a single to do and then you hit the little magic wand button and it'll break that to do down
578. 1:29:03into all of the requisite steps it sort of is a project manager in a box goblin.tools pretty cool okay um we're gonna go to runwayml uh
579. 1:29:20M I don't know what's going on with the old throat tonight I might be getting a little tickle in my throat
580. 1:29:26you might be getting it so I went to Runway ml I logged in and then I went to um what the hell was it text to video but
581. 1:29:39oh they added new interface here this is cool so you can either use a text prompt or an image prompt so I'm going to use an image prompt
582. 1:29:47and I can upload a file so I'm going to go grab that file that I just saved is this it yep right
583. 1:30:04all right so there's our picture and now we're just going to say generate and I I I saw right before I came on Live tonight
584. 1:30:16I think Gen 2 you can now extend a video up to 16 seconds so you generate four seconds and then you can add another
585. 1:30:25four seconds three three more times so you can do it videos a total of 16 seconds in length it's exciting isn't it it's so exciting
586. 1:30:38oh the other thing oh Manny so I'm showing you this this might not be what you're looking for you might be looking for how do I turn like
587. 1:30:48a talking head into an actual talking head I'll show you that as well but let's let's see how this thing does
588. 1:30:55with our with our Leo pitcher we'll see how it does our little picture and then I'll go show you how to make an animated dude or dudette
589. 1:31:03dudes doodetts skinny people fat people hairy people bald people you can do it all you can do it all people with no teeth people with
590. 1:31:16too many teeth people with 17 fingers in one hand all right here it is let's see if it got anything interesting are we ready are we
591. 1:31:26excited all right yeah that was pretty cool all right I like it you know what I'm
592. 1:31:41gonna let's extend it four seconds 20 credits per generation I don't know how many credits I have
593. 1:31:55so now I should have an eight second video this is pretty slick um yeah welcome to save the day
594. 1:32:11[Music] welcome everybody if you're new here my name is Kyle Shannon this is the AI learning lab I just ramble on for hours
595. 1:32:23night after night after night I'm insufferable welcome [Laughter] thank you we also talk about AI stuff occasionally
596. 1:32:39I'll sometimes make fun of people I'll sometimes be even a little mean a little Saucy a little Saucy sometimes
597. 1:32:48I'm nice other times I'm saucy I'm Saucy if you don't like the sauce get out of the sauce kitchen you know what I'm saying
598. 1:32:57you understand oh I think we have video I think we have eight seconds of video Let's see if there's any good
599. 1:33:06yeah I kind of died at the end there oh wait sorry my hand's in the way it's cool at the beginning but it kind
600. 1:33:16of died at the end okay anyway so that's how you do runway ml image to video which you know again if you think about that
601. 1:33:24you know imagine if you've got an idea for a film and you go to Mid journey and you do your storyboarding well you can come
602. 1:33:32here and turn those still images into Cuts you could do a rough cut of a movie um probably over a weekend and people are
603. 1:33:42doing that you can see that stuff on the YouTubes okay now we're gonna go somewhere else the other tool to to animate still images
604. 1:33:51that I use there's a number of these that you can use is D Dash ID there's Synthesia there's hey Jen I think is one
605. 1:34:00that like I said they're they're getting their money's worth out of their uh PR firm these days I've been hearing
606. 1:34:08more and more about them okay so let me log in here we'll do it with the Google since they have all my data anyway we
607. 1:34:20might as well just log into everything with them all right so if I say create video what happens here is you can you can either
608. 1:34:30upload an image or you can pick from one of their avatars so they've got some avatars here um and then I've got some that I did
609. 1:34:39like here's a dude that I did that's like a Pixar version of me um and then you can either upload audio that you recorded audio that you
610. 1:34:51synthesized in some other program like 11 labs or you can just you know type in copy so let's let's hop over to chatgpt and I'll say um
611. 1:35:08write me a humorous script in the style of Pixar that's a short paragraph talking about how fun it is living with a chubby dog
612. 1:35:21that barks too much all right that's too big foreign can you make it shorter
613. 1:35:40right there you go I'm not even going to read it we're just gonna go with whatever the hell it wrote that's what that's what we get
614. 1:35:46that's what we get you get what you pay for you ain't paid [ __ ] honey you got nothing on me nothing
615. 1:35:57you know what I'm saying where is my oh there it is okay so now so we come over here pop in that script and then let's
616. 1:36:06go grab a character so there's the ladies and here are the gents all right guy Davis Brandon I guess I could look
617. 1:36:15like a Brandon Eric Jacob Roger Stefan definitely not a Stefan Tony Tony we're gonna go with Brandon oh Brandon
618. 1:36:28okay so so we just typed in the script that Chachi PT gave us we've got an image that we made in mid-journey and modified
619. 1:36:37in Photoshop with generative Phil I did that blurred background using generative fill I can show you how to do that too
620. 1:36:43that's pretty [ __ ] slick and then we're gonna say generate video I got 34 credits left I'm spending three
621. 1:36:53of my credits I'm spending 10 of my credits here to show you how to do this come on tell me that's not something that's something
622. 1:37:03it's crazy I think it's just [ __ ] nuts all right it's done we have a 31 second video in a world where barks and belly rubs
623. 1:37:16Collide thank you the pudgy perpetually excited pup who treats doorbells like rock concerts and breakfast like a five-star feast
624. 1:37:25living with biscuit is a symphony of surprises an unpredictable dance of Joy that'll have you laughing till your sides ache
625. 1:37:32who needs an alarm clock when you've got biscuits booming good morning Serenade sure they might be a little extra in the
626. 1:37:39noise department but every woof and wiggle fills life with more love and laughter than you ever thought possible
627. 1:37:47yeah that was cringy but you know what did it take us 10 seconds no you know 20 seconds to generate the script
628. 1:37:55and then we had to generate the thing so here here's my thought on these things um okay my thought on these things is while they
629. 1:38:10are not ready for prime time now if you go back a year and look at what image generation looked like a year ago and compare it to what
630. 1:38:20it looks like today it's not in the same neighborhood um actually with images it was probably close to a year and a half ago but like
631. 1:38:28a year and a half ago like the images were laughably bad um all this stuff is gonna get dramatically better dramatically quick so
632. 1:38:39um yeah I mean and that's just using like off-the-shelf low level [ __ ] um you know some of the uses for this stuff
633. 1:38:49that are super practical is um like I do a lot of work in healthcare and we're going to be able to take videos of patients speaking English
634. 1:39:01and have them speak any language that our client wants in their voice in a different language you know
635. 1:39:12um with with the right inflections and all that sort of stuff uh grammar all that sort of stuff in in relatively short order and while it's
636. 1:39:21kind of weird now it's not going to be weird two years from now um so so that kind of stuff is coming
637. 1:39:28wild huh isn't that crazy and thank you for whoever ever dumped a pile of coins down the screen that was cool appreciate
638. 1:39:35that yep the old days it's crazy times exciting times it is I so agree with that too what AI rule wait what AI rule will it be in digital
639. 1:39:47currency Wasim I don't know quite what you mean what AI rule will it be in digital if you can clarify your question I'm not
640. 1:40:01quite sure what you're asking I like llamas I like totals I'm gonna go show that living on the edge
641. 1:40:19back here live at the waterfront village with my friend the zombie Jonathan you're looking back Jonathan just got an
642. 1:40:24awesome face paint job what do you think I like turtles the Waterfront Village open for the next 11 days so here are three
643. 1:40:37oh it cracks me up every [ __ ] time good night all good night Denise uh who is this Kyle guy anyway what are your qualifications sir sir
644. 1:40:54I am demanding to know your qualifications what makes you think you can speak upon artificial intelligence with any credibility whatsoever
645. 1:41:04none none good day to you sir good day all right carrying on you mentioned Ai and Retail and how it will be interesting can you elaborate
646. 1:41:17how it will affect retail um I think in lots and lots and lots and lots and lots of ways um
647. 1:41:32certainly on the e-commerce side I think I think things like you know virtually trying things on in in really seamless ways
648. 1:41:44dynamically generating sites and recommendations for products so I could just be in some sort of conversation and
649. 1:41:53say oh that's kind of an interesting thing or I kind of like that look and it just goes and finds the products and and identifies
650. 1:42:01all of the different sources where I might get product from and and sort of custom builds me a website from multiple sources
651. 1:42:10um so I think it's going to be amazing on that front um in physical retail you know you're starting to see things like fast food
652. 1:42:21chains experimenting with virtual you know synthesized people at the drive up window um you know imagine imagine going into
653. 1:42:35a store and having questions about something and instead of having some Clueless you know temporary employee who doesn't
654. 1:42:43really want to be there you've got some virtual employee that no can answer any question about any product from any time in history
655. 1:42:52um I think you could start to see interesting interactions that way I think if you if you look at um if you kind of play Apple's spatial
656. 1:43:03Computing you know their Vision Pro goggles if you play that out five years and assume that the form factor of that
657. 1:43:09gets smaller and actually wearable out in public and things like that the augmented reality stuff you could do in
658. 1:43:16retail is going to be insane that's that's all predicated on machine learning and AI um I don't know dynamic pricing
659. 1:43:26um personalization I I I think we've we're just scratching the surface of personalization um I'm working on a project right now
660. 1:43:36doing personalized images and print on demand stuff uh which is really interesting so lots and lots and lots I
661. 1:43:43don't I don't know that world super well um so it'll be interesting to see [Music] it's easier to learn a platform than learning python
662. 1:43:59and you don't even really need to Learn Python now just have it write it for you now people that know python are like well it
663. 1:44:07doesn't write it as good as I do it writes it in a very sloppy way and that may be true right if you want something bulletproof hire
664. 1:44:15an engineer if you want to get something done you can have to do this this will get it done it might not do it as elegantly but it will
665. 1:44:25again you know I give these machines Grace your son reads manuals Gen X doesn't need no sticking manuals exactly I don't
666. 1:44:34know how he got that like [ __ ] Lego manuals reads them every page Ikea manuals Ikea manuals you have to read those things are crazy
667. 1:44:45but like I'm trying to think what it was it was something it literally had like two bolts and a nut just like what are we read in the manual
668. 1:44:54for uh I need you to record my outgoing message from my phone you can make money watch out every day we could do that here just everyone
669. 1:45:06we'll we'll do a coordinated screen recording and then I'll do you can make money with chat gbt and you can have it
670. 1:45:13for your for your outgoing for your outgoing voice call that's really funny I paid a guy for a year to create e-learning I built a
671. 1:45:28better version in a day with Chachi VT crazy that's hilarious so so that's actually a really good point TM
672. 1:45:37um TMM wait TMM keep Rising I want to know what that means I think that's a cool name but I don't know what it means um
673. 1:45:46so everyone [ __ ] about these things well it doesn't write as good as a human well it might not write as good as a good human but like
674. 1:45:55the amount of marketing professionals that I've hired in my day that have been [ __ ] incompetent is staggering
675. 1:46:07what's really interesting about that is so you did it for for an e-learning thing um it's gonna be everything
676. 1:46:16it's going to be every SAS platform it's going to be every um you know learning course right all of the educational
677. 1:46:26courses that people have done you're going to be able to just make your own learning course dynamically on the Fly
678. 1:46:32customized for you at the level you're at right now that is going to be better than any of the [ __ ] out there and and then
679. 1:46:40where I go with that is okay what are the implications of that I don't know like who gets an area right is it all just
680. 1:46:50everything goes to 20 Buck a month chat GPT like services that end up doing everything or are there Specialists out there
681. 1:47:00that's you know have specialized data sets where you can make an e-learning on very specialized data and they're and
682. 1:47:07they're charging you to the data set rather than the mechanism to make the course but I just I just feel like the the the
683. 1:47:16economics and the customer dynamics of and this goes back to retail as well like every interaction is going to be
684. 1:47:25redefined by these tools because we're all kind of assuming that the things that we do today are the things that
685. 1:47:31we'll be doing in the future what's personalization on a streaming service well they kind of listen to what you like and you can tell them
686. 1:47:41explicitly what you like and then they give you recommendations for playlists of music that other people wrote okay well
687. 1:47:52why does it have to be music that other people wrote what if it's a playlist that just plays music written for you
688. 1:47:59in the mood you're in right now for what you've got going on today and it's all original music and they'll
689. 1:48:09sprinkle in you know 10 of historically recorded songs songs from the Before Time for a little nostalgia and then
690. 1:48:22everything else is personalized and they can tell because your GLA glasses micro movements in your irises and and your
691. 1:48:32earbuds are reading the micro temperature fluctuations in your in your ear canal it understands what music you
692. 1:48:39like and what music you don't like before you even hear it so all of the music you listen to gives you exactly the feeling you were hoping
693. 1:48:50to have I mean what's weird about talking about that is we're not that far away from it it's cuckoo it's cuckoo don't you think
694. 1:49:09all right how are you oh how are you doing I like your conversation thanks for your life thank you James James come out I appreciate
695. 1:49:19that cool name come out you have lovely nostrils wow thank you very much I I was my earlier lives I had my camera too low I've now got
696. 1:49:30everything set up on on large boxes and things like that so my camera was lower so it was shooting up so
697. 1:49:37um I I didn't get any you'll have lovely nostrils most of what I got was you should shave your nose hair foreign
698. 1:49:49it did get me to shave my nose here but thank you for the nostril comment I don't know that I've ever had a lovely nostril comment
699. 1:49:59this is how I spend my break from class thank you Thelonious oh why can't I screenshot your lives security policy no I think that's a
700. 1:50:09local setting someone else asked that the other night I think that's a local setting on your side um I'm not blocking it thank you for the
701. 1:50:16for the hearty Hearts um I'm not blocking anyone from from screenshotting this I should put I should put a little
702. 1:50:24background back here screenshot my lives one dollar a little little coin slot where you can drop in nfts [Laughter]
703. 1:50:34oh good lordy lordy lordy oh my God you've switched cams right about the time you said they'll shut me down and I
704. 1:50:43thought yep oh I know what are you gonna do gazuntite thank you wait zai yeah I think Gesundheit yes
705. 1:50:55I'm sure wait it's sure been glitchy for me the past several days yeah my internet the technical term is poo poo it'd be in
706. 1:51:06a poo poo my internet being a poo poo poo poo we can't watch the TV we can't surf on the internet the boys
707. 1:51:14can't play the video games and I can't do my Ticky talks you know what I'm saying you know you know and do you know what I'm saying
708. 1:51:27how do you hey anyone know how to turn this thing off huh how do you do this NC all right
709. 1:51:44cool you are on the kaggle data sets this live is the best awesome thank you Winston thank you for the tip on the chat gbd
710. 1:51:53plugins you're welcome not sure which tip I gave but cool if I want to learn how to code using chat GPT where should I start
711. 1:52:02um I would start with chat you PT like I so so I guess I mean I would I would ask a couple of things if
712. 1:52:18I would start with what do you want to accomplish and then why do you want to learn to code because these tools are going to get
713. 1:52:29more and more sophisticated so you might not if you want to learn to code to understand conceptually how coding works
714. 1:52:37and how you think about coding and you know Frameworks and languages and syntax and all that sort of stuff just so you
715. 1:52:44can get your head around it so you can understand what's actually happening um that's one thing if you want to learn
716. 1:52:55the code to get a coding job I think coding's jobs moving forward are likely going to be coders that are also AI literate right
717. 1:53:05they're going to want coders that are increasingly efficient and understand how to use these tools so so I'd say step one would be get
718. 1:53:15really clear on what your goals are and then I mean seriously so if I were you if you've got the 20 bucks a month pay for gpt4 at open AI
719. 1:53:29and and just start working in chat gbt4 if you can't afford that go to Bing so Bing is gpt4 and you can you can do this the same
720. 1:53:38[ __ ] there but but the open AI one is just it's it's just less restrictive and it's better um until further notice and I'll at the
721. 1:53:49point at which one of these things gets better than gpt4 I'll flop the order but that's it right now and then what I would do is I would go
722. 1:53:57there and I would say you know sort of here's a little bit about me here's what I'm trying to learn here's my goals um
723. 1:54:07what's the best way for me to learn this skill set and like what do I need to learn what are the things that I don't
724. 1:54:13know that I need to learn and have that thing put together a learning plan for you to do it and you could even say
725. 1:54:21I want you to teach me how to in fact let's let's go do this because this could be this could be interesting this
726. 1:54:28is something I haven't done before that I think could be kind of fun all right so you're gonna go to church
727. 1:54:34EBT you're going to go to Just default um chat gbt and you're gonna say Okay um I'm just getting started with coding and I want
728. 1:54:49to learn how to program python but I literally don't know what step to take first period could you tell me the broad categories
729. 1:55:05of things that I should learn so that I can become a good python programmer period
730. 1:55:19it can be fun and rewarding especially with python basics of python introdu introduction what is it and why use it
731. 1:55:27installing it and setting up the environment running a simple hello world program data types and variables numbers
732. 1:55:34integers floats booleans basic operators blah blah intermediate python functions okay so this is this is like a
733. 1:55:44programming manual right so for me because I don't like reading [ __ ] manuals this doesn't work for me this is driving
734. 1:55:51me [ __ ] crazy so number seven is projects and practical application so what I'm going to say is I'm going to
735. 1:55:59tell it I don't like its approach I have to say I don't like your approach period this looks too much like a manual period
736. 1:56:11can you teach me Python program with Hands-On use case driven projects rather than functions let's see what it does to that it should
737. 1:56:24apologize certainly learning through Hands-On projects is a great approach number one calculator build a simple calculator okay this is good
738. 1:56:40build assume for calculator that can perform addition subtraction blah blah blah okay so I like this I like this approach it's
739. 1:56:49telling me what skills I'm going to learn on these projects I can get my head around it this feels more like when
740. 1:56:54you get an Arduino the the the books that come with arduinos are like little projects make the Blinky lights make the
741. 1:57:00this make the you know TV interrupter thingy so I like this all right this is a good approach this is making me happy
742. 1:57:10all right so now I'm gonna say oops sorry happy coding okay let's start with number one period
743. 1:57:24I want to get right into the programming but I want you to teach me step by step period I don't want you to move on to
744. 1:57:35the next step until I've said okay I understand the previous one do you understand absolutely I understand your approach
745. 1:57:49let's jump in silver simple calculator step one setting up set up python if you haven't already installed python you can download it
746. 1:57:58from the official website here two your first python script okay so this is this is not listening to what I said
747. 1:58:08so I'm going to stop generating I'm going to say I don't want you to give me all the steps at once I want you to give them to me one by one
748. 1:58:19period do not move on to step two until I've completed step one period I apologize for the oversight step one
749. 1:58:31setting up python check and see if it's installed open your terminal or command prompt type python version or python3 so I'm going to say um
750. 1:58:41I'm on a Mac running OS X period can you customize these instructions for my specific setup of course
751. 1:58:57ensure Python's installed on your Mac open your terminal so so anyway um that's how I that's how I would do it right just think of
752. 1:59:13think of these tools as the most capable smartest intern you've ever had and if you're like I'm too young to have an intern
753. 1:59:26well you have one now they will do anything you ask them to do it's what they're programmed to do they
754. 1:59:33want to give you what you need it's it's what they're all about um so I would just do some version of what I
755. 1:59:40just did there but but before you just jump in and start learning [ __ ] I would get clear on what what you want to learn
756. 1:59:48and why you want to learn it because learning to code might not be the best option but it might be so get
757. 1:59:56clear on that first and then you could even use chat gbt as kind of a career coach to say what should I be doing
758. 2:00:03given that all of this shit's going to be automated does it make sense for me to learn coding right now maybe it does
759. 2:00:10maybe it doesn't and it would probably have some pretty good Insight on that all right wow the spreadsheet example shared
760. 2:00:21the uh oh the code interpreter thing yeah it's crazy all right this is the future you say you don't know about data analysis but
761. 2:00:30you're teaching others AI Eda oh yeah you know I'm fairly deep into the AI Eda if you know what I mean
762. 2:00:39if I knew what AI Eda was I would tell you I mean I definitely know but I don't you know what it is I don't want to just
763. 2:00:46give away the answers I want I want you to understand them on your own that's how I roll welcome everybody it looks like we got
764. 2:00:57some new folks here my name is Kyle Shannon this is the AI learning lab uh we have fun here we have fun we just
765. 2:01:05hang out we talk we talk we have talks we we go back and forth we have we have witty banter we have witty banter we
766. 2:01:15talk about all these here tools that one's that's chat GPT have you used Chachi BT I don't know if you knew this but you can make money
767. 2:01:25right there is chat gbt and then you got Claude this is a bunch of pesky nerdo Wells that used to work here and they
768. 2:01:32left and said I don't like how you do [ __ ] we're gonna go do our own and so they did and they're pretty [ __ ] good and then you got Bing
769. 2:01:41Bing was little Microsoft remember Microsoft remember Bing how cute it was that they were trying to compete with Google for years
770. 2:01:50and then all of a sudden they teamed up with old Sam Altman up here and kick Google's ass and Google's like wait we're losing to these guys
771. 2:02:03I think you got barred that's Google Google's starting to get their [ __ ] together after nine months nine months of like oh [ __ ]
772. 2:02:12they released the Kraken what do we do I guess we better use that technology we invented huh and then oh Mark Zuckerberg was like uh
773. 2:02:25We're Not Gonna compete on this [ __ ] why don't we just give it away here you go so that's that's that one that's llama
774. 2:02:33and then you got inflection and you got quora doing their thing with question and answers that Poe just a wild time it's wild
775. 2:02:44it's a wildfire of innovation why do I have a generically Southern accent why do I talk like this no one knows no one knows
776. 2:03:05All Right Moving On is this 3.5 or 4 I can't hear you hmm this is the exploratory data analysis of the future it is it's insane that's code
777. 2:03:19interpreter that was code interpreters GPT 4 Forward 3.5 is free and it's super fast but it doesn't have any of the fancy [ __ ]
778. 2:03:30four is the really good one it can code really well it can write really really well yeah why really well it can wait really well and uh
779. 2:03:40it can do code interpreter and you can do plugins and plugins are interesting and I think both of them can do custom
780. 2:03:47instructions but maybe not that might just be gpt4 are you a paid actor well I'm an actor and I've been paid have I been paid for my acting
781. 2:04:03uh I have not in a while not a good long while although I act like the CEO and I get paid for that that counts doesn't it
782. 2:04:21which is the best for generating animation I showed that earlier so I I don't know I don't know um I'm not I'm not digging deep deep into
783. 2:04:35the into the animation world I I think if if you're talking about video like image to video um then runwayml but but there's there's
784. 2:04:43a bunch of them out there and um if you're talking about like the synthesized voice to animation thing um
785. 2:04:52I think did but it's expensive and so is Synthesia and so is there's another one out there oh hey Jen h-h-e-y-g-e-n
786. 2:05:06um I don't know how much that one is though but I I would imagine it's expensive what I'm waiting for is sort
787. 2:05:12of the equivalent of stable diffusion and I know stable diffusion does some animation but it would be nice if
788. 2:05:17someone came out with something like runwayml or whatever that wasn't so expensive but I the the reason they're
789. 2:05:24expensive is likely that it's burning massive amounts of GPU cycles and it's probably expensive for them to run it so until that [ __ ] gets
790. 2:05:35um optimized and efficientized and cheap then it ain't going to be cheap we use all these GPU cycles and pass the
791. 2:05:48costs on to you it's the American way welcome to capitalism [Laughter] shouldn't the government be giving us
792. 2:05:58access to this oh that's cute that's you you're a card it's a government oh good lord he set the government and
793. 2:06:10competence in this same sentence Oh that's hilarious oh oh good lord all right quit trying to make Watson
794. 2:06:20happen I know it's so true it's so true like why resurrect that brand why not just call it like IBM X or IBM does AI
795. 2:06:30or I'm an AI or at IBM or anything other than Watson X Watson X is like trying to say I'm geriatric hip
796. 2:06:45but whatever it's they they try oh my God the early days of the web the uh the the web group at IBM they started
797. 2:06:55a Consulting service in a in a web group and oh my God they were so bad for so long they got their [ __ ] together
798. 2:07:01eventually but the early days the early days that was special I suggest paying for gpt4 I think it's great I I do too I
799. 2:07:10think it's worth it GPT is worth the 20 bucks in my opinion what's that little little wait a little Raven shark nest
800. 2:07:23little Raven's heart Nest I like the name I don't know what it means um yeah so I don't I'm not affiliated with these guys in any way so I'm
801. 2:07:34when I tell you I think it's worth you paying for chat EBT like why it's on my lives for the entire time
802. 2:07:41um is is because these are the ones that I think are worth um paying attention to right now um I'll remove I I'm close to removing
803. 2:07:50Poe and I I just but I but the the fact that it does Bots and the fact that you can play with multiple models on it
804. 2:07:58makes it kind of worth having there and Pi is is specialized enough right now that it's it's almost not worth putting on there but
805. 2:08:08they just raised 1.3 billion dollars so I have a feeling they're going to be a major player so I think all these are
806. 2:08:14worth playing with right now but I do I do agree that and and I pay for uh open AI chat GPT four if it feels worth it to me
807. 2:08:25yeah is there an AI that does video editing there's um d-script is one so so the problem with all the AI video tools right now is
808. 2:08:35they're all kind of like dscript where you have it's like a page of discrete tools that all do like specialized things
809. 2:08:43they're starting to roll some of those features into Premiere and things like that but not that I know but what I would tell you is go to
810. 2:08:56futurepedia.io once once we start getting into specialized tools like video AI assisted video editing editing tools there are
811. 2:09:06probably a lot out there that I don't know about but if you go to futurepedia.io that's just a big old
812. 2:09:12directory of AI tools they have a video section just go start clicking on logos I would sort by popularity for sure because
813. 2:09:23I would be willing to wager that there are 50 video tools or more and ninety percent of them are going to be steaming piles of poo
814. 2:09:33and you might not even find out how big a steam and pile of poo they are until you pay your you know sign up for our
815. 2:09:40free seven day trial and then you sign up and like oh [ __ ] this is a piece of [ __ ] God yeah rigm frigm Rock them suck
816. 2:09:49em and then you forget to cancel your subscription and three months later you have to sell the children it's just not good so just be
817. 2:10:00careful there be careful out there people there are people asking for 20 bucks a month everywhere you look
818. 2:10:08by the way you can subscribe to AI learning lab for 20 a month welcome no you can't it's a joke it's comedy it's comedy comedy
819. 2:10:19All Right Moving On Moving on something is happening with my voice and I don't know what it is it's not God it's bad
820. 2:10:29maybe I should start a cartoon wouldn't it you know if I had if I had if I were clever I would actually like well no if I were
821. 2:10:37clever and didn't have ADD I could write down all the voices I do and then match them to little AI heads I stick on
822. 2:10:44myself and just quickly switch between them so when I have a hairbrained voice I want to do I like pop on the character
823. 2:10:50it'd be like a whole little show my own little version of The Simpsons without the comedy or the good voice acting all right moving on
824. 2:11:06working in that cowboy hat would have dire consequences I agree I think I would get banned for life it's like twerking on Elder talk
825. 2:11:21gear is a hoot yeah it is it's good uh you have exactly right on the history and what is going on in designing of
826. 2:11:28gpus oh that's interesting yeah I like it I just like how corner of the market it's just mind-blowing to me
827. 2:11:44like like was it a conscious choice that Intel went well you know it's all about CPUs and gpus gpus are the garbage of
828. 2:11:55the CPU business am I right Bob until gpus are the thing we need like how do you [ __ ] that up
829. 2:12:09we're not in the GPU business that's the trailer trash hey what time what's our tee time let's go sorry I'm sorry they're they're uh their
830. 2:12:22market cap is what a trillion dollars just I when's the board meeting I can't breathe
831. 2:12:37that was that was my Intel executive finding about out about nvidia's trillion dollar market cap in June right before the golf game oh
832. 2:12:52Tick Tock just jumped me to the bottom oh uh and gift fundraising uh can we talk automations yes I will do an automation
833. 2:13:01for you I promised I would do that um oh my God the turtle guy made a new video um let's see uh by the way my name is
834. 2:13:11normally stevo42 I remember you I'll change it back in five days and counting uh have you seen any accounting
835. 2:13:19bookkeeping AI products uh I have not seen anything on bookkeeping I would assume that quickened everyone is gonna have uh
836. 2:13:29AI layers sitting on top of their [ __ ] you know what's going to be amazing when the tax preparation folks put an AI
837. 2:13:36layer on top and the AI has just read all the documentation and you just it just looks at your [ __ ] and goes let me
838. 2:13:43prepare that for you blimp move this slider what's your risk tolerance for deductions um so for bookkeeping I'd go to
839. 2:13:54futurepedia.io and see if there's anything in financial services what do you do for work I'm a founder and CEO of a company called storyvine
840. 2:14:05um all right I'm gonna I'm gonna jump down to the automation one I lost my place in the um Amelia's wife as a consumer I'd prefer
841. 2:14:19to just to talk to AI with a generated video rather than just a text chat yeah and I think we'll get there soon again I
842. 2:14:26have a sneaking suspicion Apple's gonna come out with that sooner than sooner than later if you looked at the if you looked at the um
843. 2:14:36if you watch The Vision Pro announcement pay attention to the section where they talk about being on a zoom call it's wild
844. 2:14:51so when you first get the goggles you flip them around backwards and you face them it yourself and it scans your head
845. 2:14:57and you know you scan it at all angles and so it makes a model of your head and then you put on these giant stupid ass goggles
846. 2:15:05that have screens on the outside so people can see your eyes it like projects it takes the cameras and projects them onto the screen on the
847. 2:15:12outside of these things they're [ __ ] wild Apple's just insane and then because they've got your synthesized
848. 2:15:19face and because they've got cameras pointing down at your mouth and they've got cameras pointing at your eyes when you get on a zoom call
849. 2:15:26you see like giant screens of the other people in your living room and then they see you they see your
850. 2:15:34whole face talking and your eyes moving and your eyebrows moving it's all synthesized and they're sort of in real
851. 2:15:41time composing all of these different camera views and reconstructing your face with your voice with your lip synced and everything so
852. 2:15:53they've built that [ __ ] so that's the complicated solution so now they just need to take a scan of your face with their little lidar
853. 2:16:01cameras that are all over these phones now and I think I think we'll be able to be be able to chat with personas of people
854. 2:16:11probably soon I think Apple's gonna be big in that game I think they're going to do it well all right um
855. 2:16:18you wanted to look at automations let's go let's go do an automation let me let me get my head around what we want to do
856. 2:16:27because I've done this before but on alive this could be boring as [ __ ] so all right so I want this not to be
857. 2:16:33boring okay how are we gonna do this hold all right let me get let me get a couple of tools set up here
858. 2:16:52[Music]
859. 2:17:07foreign [Music] all right [Music]
860. 2:17:22who's hitting Hearts this is not good enough singing for hearts this this this deserves no Hearts he deserves no hearts
861. 2:17:31for that kind of singing a trash incompetent off-key off temba good Lord you people are simpletons he's not entertaining there's nothing
862. 2:17:44entertaining about this he says the tick tock who says that he does okay
863. 2:18:00I'll flip the screen over in a second let me let me find my peace Parts first that was the challenge engine wait where is it
864. 2:18:13challenge engine no chat gbt lesson plan maker where's the where's my example I lost my example
865. 2:18:33that's it I think this is it yes okay now visualize your zaps zoom out and use that okay whatever stop
866. 2:18:49giving me new features I know how to use it now don't give me new [ __ ] oh no what have they done oh visual editor
867. 2:18:59classic editor okay we're gonna go to the classic editor just because I don't know how to use a visualizer okay okay
868. 2:19:19so when you come to zapier okay so wait let me let me let me start at the beginning let me let me go uh
869. 2:19:31I'm just trying to figure out where to start where where it's gonna make sense and it's gonna be somewhat remotely
870. 2:19:37interesting so so bear with me as I try to get my head around what I want to do okay uh here so the way you can think about
871. 2:19:46automation tools like zapier and make make is another one that's like this that's got a slightly different
872. 2:19:51interface but it's the same basic idea is all of them function effectively the same you've got what they call a trigger
873. 2:20:00so that's something that starts the automation and then you've got some series of actions um and what I have on screen right here is
874. 2:20:11is kind of the most simple automation you can do where you've got an input some sort of processing and then an output
875. 2:20:19so the input I have is a Google form the processing I have is send a prompt with the data from that Google form to to
876. 2:20:32um chat gbt to gpt3 in this case and then take the result of that and pop it in an email and send that email so
877. 2:20:43the inputs here could be you add a new row to a Google form you upload a file into a directory on Google Drive you create a new Google doc
878. 2:20:55there's all sorts of things that can trigger this it can be someone posts something in a slack channel right so the trigger
879. 2:21:05is something happens somewhere that's connected into your thing and then that that starts this Cascade
880. 2:21:13and so how how you add another step is there's this little plus button here and then you you add it and and so so let me
881. 2:21:20go through this so so what I did was I created a really simple Google form the most simple Google form that basically just said what's your
882. 2:21:31favorite color and then your answer and submit so so that was the form that I created then over in zapier when you create one of these um
883. 2:21:44one of these zaps you could you kind of build them think think of these as Lego blocks you build it a Lego block at a
884. 2:21:49time and you kind of start from the input and work your way down so I said I want to create a Google form so
885. 2:21:57if I change this so so you come into this little thing which says you know now it says change the trigger but if I
886. 2:22:04was creating this from scratch it would say create a new trigger and then I would say Google form right there's Google forms right there
887. 2:22:14um and then I select well what do you want to do on this Google form well I want a new form response so so
888. 2:22:22depending on the module you choose you've got different options that that API allows right so I can either say
889. 2:22:29anytime someone put something new in the form or if they put something new or update their form I'm going to take this
890. 2:22:35action and then I hook my account up and then I say what what form do I want So based on my account I got lots of forms
891. 2:22:44I can choose from so I choose the I chose the uh the one that said favorite color so that's the one with our single color
892. 2:22:52there and then I test it and um I have a form response from when I created the form I did a little test and
893. 2:23:02I put in the color green and you'll notice over here when I when I tested it it pulled in the color green so pull in
894. 2:23:09a bunch of other [ __ ] about the form but I'm just interested in can it see the answer that I typed into that form it can
895. 2:23:17so now what happens is now that I've got a trigger and I've got some data input from that trigger now I can use that in
896. 2:23:24anything that follows it and that's one of the cool things about the automations is the more modules you add you can use
897. 2:23:32elements from anything above it okay all right so now we're going to say continue with that that record so now
898. 2:23:42it's moved us down to our open AI prompt our openai module right and so an important thing to know is there's an
899. 2:23:51open AI module and there's a chat GPT module um I'm a little cranky about the chat GPT one because the the the fields are not
900. 2:24:01the same as they are over in in opening eyes playground but that's neither here nor there okay so so what's happening here is it's
901. 2:24:09it's a very similar thing um to what we just did so we're doing an open AI module we've got a list of things we can do here so we can check
902. 2:24:20moderation to see if someone cusses at us we can create a transcription we can create a translation we can generate an
903. 2:24:27image with Dali we can send a prompt so that's the one I selected so we're going to send a prompt then I hook up my account then I
904. 2:24:37um now this is where I do my prompt so here's the model that I'm using I'm using text DaVinci 3 and here's my
905. 2:24:45prompt I want you to use the following variables color and so let me get rid of this so so over in in the um
906. 2:24:57GPT playground I sort of figured out the prompt I wanted to do and this one's just a stupid stupid example prompt but
907. 2:25:04I want you to use the following variables color so because I've got that form above that I've I've now got access to I can now take any of
908. 2:25:18the input fields from that form and put them into my prompt so I'm going to find the one that says green
909. 2:25:24which is the bottom one here says what's your favorite color I click green and so that's now in my prompt so so this is a
910. 2:25:31variable within my prompt so what that means is anytime someone fills out a form it's going to take that color and
911. 2:25:37it's going to stick it in my prompt and it's going to send that to open AI okay and then like here's the temperature
912. 2:25:43here's the maximum length so this is all stuff that you play with at playground the temperature if I move it to zero
913. 2:25:50it'll be more dry if I move it to two it'll be um more creative one is somewhere in the middle um I ignore a lot of the other [ __ ] just
914. 2:26:02because I'm I'm generally just doing fairly straightforward things and then I test that and so so basically what the
915. 2:26:09prompt was write a write a poem about a blank colored animal right and so it gave me back this is the test response a
916. 2:26:18silly frog of green croaks and hops and you know it so it's a green frog poem so that worked and now I say continue and
917. 2:26:26my next module was Gmail so I pick Gmail I hook my account up I say we're gonna we're gonna send an email
918. 2:26:36so you fill in all the fields of send the email and then I'm going to say here's your poem about a and so I pulled
919. 2:26:45the the green variable from the form right so here's your poem about a green animal that's the subject and then the
920. 2:26:54body of the email is hello there you asked for a poem about a green animal a blank animal here it is and then this
921. 2:27:03other variable is the response from open AI So within this email it's it's using both inputs from the form and from open AI
922. 2:27:13and then when I go test that retest action did it do it it should send me an email so I'll go over to Gmail
923. 2:27:35[Laughter] and so here we go me from me here's your poem here's your poem about a green animal hello there you asked for a poem about a
924. 2:27:51green animal and I don't have a space there so I see there's a typo in my form so I could go fix that and then there's
925. 2:27:57my poem right and so now anytime anyone goes to that form and fills in a color within 20 seconds they'll get an email
926. 2:28:05with a poem about something colored animal so that's automation that help was that boring we lost 20 people
927. 2:28:20we lost 20 people doing the automation we lost them they're like I can't take it he's just talking about forms and variables
928. 2:28:30this is like program I didn't want homework I came here for entertainment now he's talking about zapier zaps and variables
929. 2:28:45I can't take it Sally again I'm done I'm out AI learning lab my ass never again oh I'll be back tomorrow or I'll
930. 2:29:00see you tomorrow all right bye all right I stay up to see you heart heart but gotta go to bed good night Dawn thank you
931. 2:29:13um hey guys by the way my name is normally Stevo yep we saw that one earlier you must read some manuals that's how the Challenger exploded
932. 2:29:21no the Challenger exploded well they didn't read the manual but didn't the dude come in and bang the thing on the
933. 2:29:27table and said this is a problem and they're like sure get out pee on with your engineering skills yeah I know I I'm not I'm just not gonna
934. 2:29:40do it I ain't never read a manual ain't never gonna read a manual I ain't gonna do it and now haha wait who said I needed to do that
935. 2:29:51Lynette Lynette ha these things will read the manual for me do you know you can take full-on documentation and pop them in these
936. 2:30:01things and just have it give you instructions on how to do [ __ ] you don't ever have to read a manual
937. 2:30:07again we got machines for that now I got people for that I don't really knows thinking manual never I will never do it never
938. 2:30:18all right best hour of my day just spent here on this chat thank you TM thank you peace love it oh my God the turtle guy
939. 2:30:28made a new video I know I know I got messaged by someone impersonating you be careful there are fraudsters lurking I got really
940. 2:30:40that's terrifying um Lomas camper if if you would DM me and tell me specifically what happened there I'm I'm curious well first of all yeah
941. 2:30:54all right I got I got nothing going on here Commerce wise but I suppose someone could say bye my class
942. 2:31:03if I buy my class on you can make money watch out oh my God I didn't think oh that's annoying all right well we'll
943. 2:31:15have to deal with that if I find them I'll tell him to knock it the [ __ ] off sorry about that that sucks
944. 2:31:21Kyle can you do automation so I hope that helped Dan I hope that was uh that gave you what you needed to learn there
945. 2:31:27um it's it's kind of a it's kind of a Deep River to run in but once you get your header like the basic idea with
946. 2:31:35automations like the biggest thing to get is you've got inputs you process the inputs and then you output them
947. 2:31:41somewhere like once you get your head around that then it's all just variations on the theme but some of those modules are really janky and
948. 2:31:48complicated and just overly whatever this is crazy I'm adding I I am now adding monetization to my platform using
949. 2:31:57stripe oh that's great that's great with chat GPT chat GPT is writing it all yeah it's isn't it nuts it's nuts so so this is it so what tm's
950. 2:32:07doing is really good one of the toughest things to sort of flip your head around that the thing that took me a while is I would just
951. 2:32:15kind of sit and stare at these things like they were gonna do something I'm like all right what do I do with it I
952. 2:32:22should I I guess I could make another poem uh um what do I is there a class because they can do anything they can
953. 2:32:34get very overwhelming and and the best way to get your head around what they can do is is don't think about what can
954. 2:32:42I do with AI flip it around and say what's something in my life that I'm trying to do that it would be interesting to see if I
955. 2:32:51could do with these things and then you've got a then you have an imperative you're like okay all right so
956. 2:32:57I wanna I wanna write a Mad Lib style thing for first-time Founders okay that that I can let me go here and see if I
957. 2:33:05can do that and then you'll start doing it and then at some point it will give you some sort of output where you'll go oh God
958. 2:33:13I didn't know I could do that when you have your Kevin McAllister moment and then your life will never be
959. 2:33:23the same and then you just show up on these [ __ ] tick tocks night after night going why am I here what have I done with my life
960. 2:33:29oh my God more chat ubt oh it's so good it's so good I used to like pudding but now it's Chad gbt I could have pudding and chat EBT oh
961. 2:33:42no this is so good and then three hours later you're like where's that pudding oh do you own a cat I am not a cat person
962. 2:33:55I'm a dog person I don't like the kitty cats sorry sorry kitty cats are fickle I like I like unbridled love doggies are
963. 2:34:08just doggies are like you walk in and like oh my God you're here this is the best thing ever I don't like to like please pet me
964. 2:34:19please pet me [ __ ] you that doesn't do it for me cats make me crazy like you want the love you get the love
965. 2:34:28you want the love and then you scratch the [ __ ] out of me um Homie don't play that all right I know coder's using chat gbt
966. 2:34:38to help them with bug fixes and I use it for quick templates yeah that's a good idea that's cool yeah you can oh one of the things these
967. 2:34:46things are really good at if you haven't done it it's really good at mimicking stuff so you can give it like for Story vine we
968. 2:34:54write these things called video guides which are a series of prompts that show up in our app that walks someone step by
969. 2:35:00step through telling their story they're written in a very particular way that you know because of the nature of how
970. 2:35:05the app works and and we've sort of figured out over the years what works and what doesn't and you start with an
971. 2:35:11introduction and then it's got a story arc and just all sorts of [ __ ] in it and so I was trying for a long time to get it
972. 2:35:19to do our prompts correctly and I was just you know writing these increasingly complicated prompts and then it hit me
973. 2:35:25oh [ __ ] why don't I just give it an example and then I said I I explained you know I have a video company and we
974. 2:35:32write these story Frameworks called video guides and here's what they look like and so now write me a story that's
975. 2:35:39got five questions in it you know that follows that format that tells this story and then out it comes and it was [ __ ] perfect
976. 2:35:50like Bang from day one so if you've got something you have a particular kind of writing style or you've got something
977. 2:35:57that's formatted in a particular way or you've got in your case a code template pop that template in there and tell it
978. 2:36:05to solve a problem within that template or follow the tone of the email it's [ __ ] brilliant for that
979. 2:36:11do you believe humans will use AI also in weapons in the future someone was asking about palantir earlier so yes we
980. 2:36:17already are one of the dangers in my opinion totally agree um Boston Dynamics has come out publicly and said you know we probably shouldn't
981. 2:36:27put weapons on our robo dogs thanks Boston Dynamics I'm sure that's what they tell the public and then with their defense contractors
982. 2:36:39they're like wait can is there any way we can stick on six more guns paint pointed in all directions
983. 2:36:46so you know they're doing that [ __ ] but you know long as it's with the military um yes I do think that uh I do think
984. 2:36:55that AI will be used in weapons and you know bad guys will get their hands on AI and good guys will get their hands on AI and
985. 2:37:02well well yeah yeah and will there be big news stories where something went awry absolutely we are we are predisposed right now
986. 2:37:13to be afraid of these things for a couple of reasons one is any new technology especially ones that are going to be disruptive
987. 2:37:22scare the [ __ ] out of people and it's happened throughout history happened with the tractor it happened with the camera
988. 2:37:32with the daguerreotype it happened with the printing press it happened with the World Wide Web and so so humans are you know resistant to
989. 2:37:47change they don't like change they're afraid of new technology we've also got 50 years of history of Hollywood telling
990. 2:37:54us the robots are going to kill us so we've seen all the movies so the minute that we have a story that's like oh that's like that scene in
991. 2:38:04Terminator here we are you know it'll make the cover of the New York Times and it'll be on all the news channels for three days
992. 2:38:13and it'll probably knock this [ __ ] back six months so yeah that's gonna happen and is it gonna suck yeah it's gonna suck but you know
993. 2:38:25they ain't gonna be uninvented they ain't going away I'm sorry about the hair I don't know what just happened there
994. 2:38:31it was normal and now it's not that's the way she rolls it's just my lot in life all right um it will be used a lot by experienced
995. 2:38:44coders it will be used a lot what I've heard Winston from uh the the all-in podcast Jason kalkanis and those guys is that what what what
996. 2:38:53they were saying is that um Tech shops in San Francisco that are using um GitHub co-pilot as a as a normal
997. 2:39:02course of business they're seeing 30 to 40 percent increases across the board in those companies that's that's [ __ ]
998. 2:39:10significant and this is and this is like the new raw not really working version of this technology like imagine two
999. 2:39:17years out when you can you can just kind of Imagine a sophisticated you know subroutine and it just writes it for you
1000. 2:39:27and integrates it and updates all your API calls imagine never having to update a library ever again because the AI is just
1001. 2:39:35sitting there waiting for Library changes it just scans your codes looks at all your libraries and dependencies goes out finds any updates
1002. 2:39:45finds replacements for things that when things have been deprecated automatically update your code what's gonna happen to all of the Ops
1003. 2:39:55people that that's all they do with their time I don't know they can go right the Bots that go do that [ __ ]
1004. 2:40:03and then they can sit back and smoke a big fat cigar going yeah my machines are coding themselves [Laughter] bring me another cocoa
1005. 2:40:19hey I will take software engineering jobs accountants lawyers too so Jorge Ramirez yes AI is gonna so any any all
1006. 2:40:31all jobs that are intermediated with computers so every knowledge work job every creative job every
1007. 2:40:46anything Logistics manufacturing anything with computer mediation the jobs every single job is going to be
1008. 2:41:00redefined by these tools every single job now does that mean every single person gets replaced no what it means so so there's a there's a
1009. 2:41:11guy I talk about a lot on here named Paul rates or he's got the marketing AI Institute and he he said
1010. 2:41:17AI will not replace your job but someone using AI will and so one of the things that this channel is about is just encouraging
1011. 2:41:24people like why I have these things up on screen for the entire [ __ ] thing is if you're not playing with these
1012. 2:41:30things go play with them go right now to chat GPT if you haven't done it yet and start playing start getting AI literate because
1013. 2:41:41when those disruptions start to happen in every business the powers that be at some point are going to go oh [ __ ] we got to get our sh
1014. 2:41:50our [ __ ] together with AI who knows anything about AI uh Jorge was talking about Chachi PT in the lunchroom
1015. 2:41:58the other day uh bring Jorge to my office Jorge the [ __ ] going on with this AI stuff I don't know it's you know
1016. 2:42:08would you like me to write you a poem about quantum mechanics here you go um so so get literate get literate start
1017. 2:42:19figuring out what these things do what they can't do what they're good at what they're bad at um even if you don't have a goal in mind
1018. 2:42:25it's gonna be it's it's gonna serve you well um so yeah now here's the good news so here's what this is just my gut right
1019. 2:42:36this is all this is all speculation right now we're just starting to see impact on the workforce you know we're
1020. 2:42:44starting to hear oh so and so laid off 4 000 people and they're replacing them with AI and this company is not going to
1021. 2:42:50hire any more people because of AI and we got the SAG strike and the writer strike right now this stuff is kind of at the periphery
1022. 2:42:58of the conversation it's sneaking in a little bit but it's at the periphery of the conversation over the next two to three years
1023. 2:43:07big swaths of jobs are going to be impacted and it's they're just going to get bowled over and I don't think we're
1024. 2:43:13frankly prepared for it so I think there's going to be a lot of pain and a lot of disruption what we don't know is
1025. 2:43:21on the other side of that pain and disruption all of these Technologies as they get more sophisticated are going to
1026. 2:43:28generate lots of prosperity and they're going to generate lots of new jobs and lots of new job demand that
1027. 2:43:36the workforce won't be trained for because it's going to be entirely new things so who's going to be first to fill those
1028. 2:43:43jobs people that at least understand something about AI right and so as it does this and the workforce kind of
1029. 2:43:51diminishes and does this it's the people that sort of you know hop out of whatever they were doing before start
1030. 2:43:58learning this AI stuff and they're going to be able to ride that that wave up um so I think five years out we start to
1031. 2:44:06see all sorts of New Opportunities and jobs but the two to three year period you know sort of probably starting 2024
1032. 2:44:17I think it's going to be painful I've used AI to help program an idea I have but it doesn't do certain languages that I guess I need um
1033. 2:44:29it's possible um I would just keep asking it questions because there's with programming there's always many ways to to solve the problem
1034. 2:44:43um so I would keep uh keep uh I would keep on trucking I would I would tell it that you're frustrated and you
1035. 2:44:52know if it says it doesn't do certain languages say well what languages do you do like I would just keep pushing on the
1036. 2:44:59edges of it just keep having it educate you about what you need so here's here's an example when I first started using
1037. 2:45:07chat EBT I thought oh I'm gonna have it write me some python code I don't know python and so it wrote python code I was like cool
1038. 2:45:17and then I realized I don't know how to run python code so I told it I don't know how to run python code and it said oh sure you just
1039. 2:45:24go to your terminal and you type in this command line and and I was like oh no I don't want to type in command lines I
1040. 2:45:30don't know I'm not yeah I don't know it's that feels like work to me I said isn't there a website that I can use
1041. 2:45:35that'll just do this for me and it gave me five websites and I picked one and then it taught me how to use the
1042. 2:45:41website to run python code so there there there may be ways for you to to get around that the other thing to do is um
1043. 2:45:51talk to people in your life that you're doing this that you're playing with AI even if they think you're weirdos
1044. 2:45:57someone in there might be playing with this stuff also and maybe they've got coding skills and could support you on
1045. 2:46:04that so I would sort of get generous about talking about how you're exploring it and by the way 9.5 out of 10 on the name verified mom
1046. 2:46:15is a solid name very solid like it my internet is bad today [ __ ] Xfinity Chris you too Xfinity it's so bad
1047. 2:46:27you got another account I did not I just came back they only booted me for seven days can we talk automations we just did um
1048. 2:46:37prompt engineering oh by the way there's 90 folks here I am more than 99 messages behind I because I'm an idiot I like to read
1049. 2:46:50every comment from every person but it means as the I've been doing this now for like two hours two and a half something like
1050. 2:46:59that I don't know way too long what time is it oh I gotta go to bed it's [ __ ] 11. um I'm just I'm behind oh I really I gotta
1051. 2:47:07go I've got a workshop I Gotta Give tomorrow and I I am behind I haven't done my homework for my little workshop yeah it's an add thing
1052. 2:47:19as long as it's an emergency I can get it done um anyway I'm really behind on the on the comments so I'm gonna you know what
1053. 2:47:31I'm gonna do I'm gonna jump down to the bottom because I need to go soon so I'm just gonna skip a bunch of questions so if I
1054. 2:47:37missed your question sorry about that so I've now jumped down to the bottom and I'm gonna just go back a few and just
1055. 2:47:43grab a couple more questions Kyle's hair is tired now says Kyle um I'm a software engineer use it every
1056. 2:47:50day to help write code that's cool Sam the name wait and the name is a letter or so off that's what you were saying AI learning lab 2ls huh
1057. 2:48:06did I misspell something somewhere Maybe gotta pivot to architecture or cyber security architecture is going to get
1058. 2:48:16really cool in the next couple of years what's gonna happen when planned obsolescence is thwarted by self-updating programs
1059. 2:48:25I think everything's gonna get [ __ ] up I you know I think it I think it's going to be really fascinating to watch
1060. 2:48:40the power infrastructure not not electric power but the powerful corporations as these tools start [ __ ] with the power centers
1061. 2:48:54the death throws of those power centers are going to be fascinating they're going to try to shut this [ __ ] down and
1062. 2:49:01they're going to try to they're going to try to box it in and I think Meta Meta releasing the Llama 2 models
1063. 2:49:07commercially is a really interesting hedge against that kind of control um it's going to be fascinating like I have
1064. 2:49:17no idea what it's going to look like but but chaos Generation by the way 10 out of 10 on the name I like chaos generation is a
1065. 2:49:26really good name um yeah I just just keep paying attention just like like here's the deal if if you look
1066. 2:49:38at this [ __ ] from the outside it it looks a certain way and it's like oh it's over hyped and it's Geeks and it's this and that
1067. 2:49:45when you get on the inside of this stuff and actually start playing with it and see what it makes possible and start to
1068. 2:49:50just imagine even two years in the future if this if this thing I'm doing right now got just a little bit better who would that threaten
1069. 2:50:00start paying attention to that [ __ ] and if you start seeing the who would that threaten people starting to make noise
1070. 2:50:08you're starting to see the death throws of a of a change in power air traffic controllers let's replace them with AI
1071. 2:50:15my big my biggie right now is uh customer support call centers they need to go they need to go and be replaced with competent AI
1072. 2:50:26and phone answerers that talk to me and can actually solve my [ __ ] problem without putting me on hold for 45
1073. 2:50:32minutes and then hanging up with me or hanging up on me after my fourth person I'm talking to where I've had to re-explain my problem
1074. 2:50:42to each person every single time that [ __ ] needs to go so you take your air traffic controllers I'll take the
1075. 2:50:50customer support people let's get these things moving on what's it going to do to digital marketing completely totally [ __ ]
1076. 2:50:59obliterated and upend the entire agency structure time and materials is a relic of the past but no one knows it yet
1077. 2:51:13the whole ad industry is but what if I like turtles whole ad Industries what I mean what's cool is as as the ad
1078. 2:51:23industry starts to make things more efficient and let go of people as they find efficiencies they're still going to
1079. 2:51:30struggle to hold on to their time and materials model and what's going to happen is the 10 of the people that they
1080. 2:51:36cut due to efficiencies or twenty percent the Smart Ones of those are already learning this [ __ ] if they
1081. 2:51:43haven't already learned it and they're going to go start 10 person agencies that can come kick the [ __ ] out of 200 person agencies
1082. 2:51:51the holding companies are in trouble and they're being [ __ ] arrogant about it right now I'm talking to the clients like we've
1083. 2:52:00got big ass clients storyline's a small company but we have big ass clients and we have some pretty trusted relationships right now
1084. 2:52:08and they're telling us what they think of their holding companies and it ain't pretty they're in trouble they don't know it yet
1085. 2:52:18because they're you know full of themselves because they're big powerful people that have been big and powerful for a lot of decades
1086. 2:52:27but they got some [ __ ] coming we're incorporating touch EBT into our six-figure blueprint digital business that's super cool I like that
1087. 2:52:39I guess I'll be staying oh styling hair forever if you're dealing with me we got we got some issues to deal with
1088. 2:52:46I hope you go work for Trump huh okay I don't even know what that means yep I'm the weirdo in my circles that's weirdos gotta stick together
1089. 2:53:01that's what that's what this channel is all about this is this is a weirdo collection Bowl [Laughter] oh my God somebody is impersonator You
1090. 2:53:13by a similarly named account Kyle I'll message you oh okay got it all right we'll get them whoa get them uh whatever you call it
1091. 2:53:24we'll get them we'll get them tick-tocked out like what why why oh are they like stealing my content and doing like I don't what
1092. 2:53:38all right I'll have to go check it out there's no there's somebody impersonating you and scamming people it's AI learning lab with two L's
1093. 2:53:51oh ail learning lab oh I see okay all right I'll go check it out and how are they scamming people oh that's exhausting okay can we stop it
1094. 2:54:05with the [ __ ] scamming like really isn't it just as easy to start a [ __ ] business as it is to like scam people
1095. 2:54:12and just try to [ __ ] sleep at night never understood that can you try the Vizsla plug-in the link is in response where the video outcome
1096. 2:54:22is I tried it the other night deadly and it didn't work it was broken so I'll try it at some point I got to go to bed now
1097. 2:54:29but I actually got to go do some work now um I think they were telling you about the impersonator with a misspelled
1098. 2:54:35screen name I will find them uh Richard feyman came in and dropped an O-ring into ice water and showed it became brittle
1099. 2:54:44oh that yeah that's what it was yeah exactly exactly that was the The Challenger thing yeah they said this is bad
1100. 2:54:53all right this this stream makes me anxious why all right oh my God I'm so ready for AI to replace customer call centers I know
1101. 2:55:01right exactly all right people I gotta go it's time good night Kyle do you have homework to do I do I have homework to
1102. 2:55:12do all right peace everybody like honestly AI makes it easier to scam than start a business that's a big fear yeah
1103. 2:55:19I know we'll get after him thanks Eric growth geek appreciate it peace out everybody peace out the bear Dr J chaos generation
1104. 2:55:30Rebel Alex peace out everyone thank you talk to you tomorrow peace