3/19/2025 - Exploring the Power of Notebook LM: Mind Maps and Vibe Coding

Transcript

1. 0:00boy you ready for some singing [Music]
2. 0:46his place that ra my together my thoughts and sweet silence
3. 1:07place anure to Val this is place I can slowly face the only one I truly can know these are tears from a
4. 1:25long tears from a long time ago to cry years these are tears from a long
5. 1:40[Music] time well oh darling Oh Darling you say unto [Music]
6. 1:55me where have you been all my life time I have been swimming Seven
7. 2:09Seas women have TOS me their lifelines I pull them the in them but I didn't know these are tears from a long time
8. 2:26ago got these tears from a long time go I need to cry 30 years or so these are tears from a long
9. 2:43time go [Music]
10. 3:02good day good people tiktock pin hey Kyle saw the new podcast with an thank you very much Tom yeah the AI Readiness
11. 3:11project we're starting to get our C legs a bit in in a typical Kyle Shannon and an Murphy style we uh came up with the idea
12. 3:21to do that and launched it I think eight days later so I think today was our fourth episode and it was a fun one it's fun one it's
13. 3:34good all right let me let me flip the cams here for you Tik Tock folks so you can see the splendid glory of the
14. 3:50show oh man [Music]
15. 4:12there's been something baby I've been trying to say for an age and it seems I don't know [Music] how the past and the future now
16. 4:28surrounding me surrender to whatever Tre there can be found there's been little [Music] trouble since you came to my
17. 4:43[Music] rescue and if you like all of the rest I would have quit you long ago but I couldn't do that [Music]
18. 5:00oh tell me now women and wi never went to [Applause] well make a man crazy make him cold as [Applause] hell I a woman you wish me
19. 5:18well but in spite of your try still going to have to find own way through yeah [Applause] oh man all right Little Women and wine
20. 5:34Martin ston um y we had a good show with an today you know talking about AI Readiness AI Readiness man that's where that's where
21. 5:45it's at LinkedIn put out a new jobs a a job skills for 2025 the top skills for 2025 one of the skills was AI literacy
22. 6:02which again I would argue it's impossible to be AI literate right now because things are moving too fast but
23. 6:10you can be AI ready and so I'll I'll give them I'll give them some credit for that one being number one but they had
24. 6:17they had two of the other ones that they had skills they they defined these as skills one of them was
25. 6:25adaptability I don't see adaptability as a skill I see adaptability as an aptitude I see adaptability as a personality
26. 6:35trait if you're lucky enough to to to be to embrace change um you're going to do good if you
27. 6:52like predictability and structure and you're a little rigid you're going to be miserable that's not a skill now now can you
28. 7:00practice being more adapt adaptable sure uh you know uh whenever you think you've got something figured out change
29. 7:14it uh and then the other one that they had one was adaptability and then the other one was basically critical
30. 7:21thinking I think they called it creative creative something or I don't know but it was basically critical thinking which again
31. 7:31if you're a critical thinker if you're someone who defaults to let's talk about this let's think about this but many
32. 7:37people are like tell me what to do I'll go do my job so so they had some good it was some interesting things in the list those
33. 7:47three stood out to me as a little a little unthought [Music] out uh I don't think they're wrong though [Music]
34. 8:02I mean can you take a class on adaptability I guess you could [Music]
35. 8:31who on LinkedIn it was a it was one of those sponsored posts by um workday hang on let's see um
36. 8:472025 job skills I think is what it was yeah here it is skills on the rise
37. 9:03[Music] there we go you can see that
38. 9:17um so sponsored by workday so right so you've got to discount it as you know this is this is just a little content
39. 9:24fluff but AI literacy why it's on the rise is AI rapidly trans forms Industries company are increasingly looking for professionals with the
40. 9:33ability to understand and U utilize tools harnessing that technology for business purpose that technology harnessing that
41. 9:45technology which technology generative AI That's like 400 Technologies which one and they're like most common
42. 9:54Industries Tech and internet higher education IT services Consulting most common job title software engineer no this is every
43. 10:06[ __ ] job in every [ __ ] industry right it's like come on this is way bigger than like the fact that they
44. 10:18wrote this in this way is kind of like it says to me companies for a good long time probably a year or two are going to
45. 10:28do what they they did with the worldwide web ah the IT department will figure that out let the IT department we've got
46. 10:36a CIO though CTO they they'll we've got people for that so so that's a that's a bit of a red flag for me and then also just AI literacy
47. 10:51like who who on this call who that comes here five nights a week feels competent with this AI stuff I certainly don't [ __ ] feel competent
48. 11:02or [Music] literate it's it's overwhelming so that was that one conflict Med adaptability why it's on the rise rapid change of
49. 11:15pace in the workplace yeah you think you got technology comes along that that does your job for you yeah that's I I I would call it
50. 11:25rapid rapid pace of change but again adaptability is not a skill you can't take a class on adaptability most common job titles
51. 11:34teacher Administrative Assistant project manager higher ed technology and internet like again this is every
52. 11:40industry every job is going to require that and then Innovative thinking this was the one the other one then they got
53. 11:48public speaking which I don't disagree with that customer engagement and support Brandon's happy about that one go to market strategy
54. 11:59Regulatory Compliance growth strategy risk assessment there all there's some generic ones in there but I don't know I
55. 12:07thought it was interesting listen at least AI was on the [Music] list exactly just posted today about a
56. 12:19really stupid issue I spent two hours trying to fix on an image still failed oh yeah that sucks LinkedIn colleague
57. 12:27just posted her certific ific from Harvard AI hey man I'm I'm telling you learn out loud with this AI [ __ ] um if
58. 12:36you get a badge in the AI Salon put that on your LinkedIn if you're a member of the AI Salon if you just go in there and
59. 12:42make an account put that on your LinkedIn I'm serious go to she leads AI it makes a difference start talking about it in the
60. 12:53work room at work it makes a difference [Music] I think you'll all be happy to know I
61. 13:08mean I know one of the things that people sometimes say to me sometimes ask me you know is uh you know who's really in charge at
62. 13:21home and you'll be happy to know that I'm man enough that uh I went to the misss and I said hey listen you're not the only one who gets
63. 13:33to wear the pink bow she took this thing again from my desk I put my foot down I said I'm the man of the house I deserve
64. 13:40to wear the pink bow sometimes too all right so just so you know kind of kind of Integrity you're dealing with here it's fairly significant
65. 13:51[Music] I'm assertive I'm confident I'm a man's
66. 14:07man yeah yeah how's it going oh and as a woman you might be looking at this pink bow going you know you're wearing it
67. 14:18wrong look at whose woman splaining now huh I know how to wear a headband this is it right here
68. 14:35I know I missed I missed the bow too I went into the bathroom I went into the bathroom and there's a pink bow on the
69. 14:41back of the toilet in the little basket of makeup and I'm like oh no she didn't actually I'll I'll come clean
70. 14:53with you I didn't confront her I just stole it back like the spineless weenie I am but she just stole it from me there was
71. 15:02no discussion about her taking it so there you [Music] go um related to this to this LinkedIn thing here let me show you another let
72. 15:16me show you another piece of uh internet something so remember old Andrew Yang who was running for president there for a bit
73. 15:29some some election or two ago and uh he was like we gotta have Universal basic income we got to have Universal basic
74. 15:40income and everybody including myself was like what the [ __ ] are you talking about [Laughter] weirdo yeah about that uh
75. 15:55Andrew Andrew kind of had the right idea hang on let me I gotta get things fixed up for the ticktockers here here's what
76. 16:03here's what happens for those of you on YouTube we've got a Tik Tock contingent they're not weird um but they are irregular
77. 16:14and if they don't have things just so there's a lot of drama lot of drama a lot of drama in this group and so you know I just like to I
78. 16:25just like to keep everyone happy anyway Andrew Yang guys AI is going to eat a [ __ ] ton of jobs I don't see anyone really talking
79. 16:37about this meaningfully in terms of what to do about it for people what's the plan and then he that's his comment in
80. 16:46response to this chart which I think is really interesting it's from meter me TR um and this is a a new kind of Moors law that they develop
81. 17:03um and it says the length of tasks AI can do is doubling about every seven months so it starts out with gpt2 being
82. 17:15able to do like a 3- second task you know maybe like write a write a you know pair of lines of a poem gpt3 can do like
83. 17:25I don't know 10 seconds Kyle I love you but I got to do some AI homework we'll catch up with you later
84. 17:34and the pink bow looks good on you thank you LinkedIn user I don't know who that is but smooch back go do your AI homework listen if
85. 17:42it's AI homework you're excused you can also choose to put you know me rambling in the background just to remind you keep doing your AI
86. 17:51homework but yes go do that um GPT 3.5 can do 45 seconds GPT 4 can do kind of a six minute task and then Sonet
87. 18:0737 can basically do a task that'll that would you know have taken you an hour right write a simple application and and
88. 18:16you know or a dashboard or something like that um I think that's actually a really interesting metric so this is
89. 18:22basically starting in 2019 or so and we're now in 2025 and and these these tools can do you know an hourlong
90. 18:34task optimize code for a custom chip is a 4-Hour task so that'll be you know in the next year and a half but I agree with Andrew Yang
91. 18:47nobody's [ __ ] talking about what's coming nobody's talking about it it's astounding to me [Music]
92. 19:03[Music]
93. 19:27ah all right so so what's everybody got going on what are you all thinking about are you thinking about this [ __ ] are you are you are
94. 19:37you how are things with lovable that was the other thing I was going to do was pop over to lovable um my problem with
95. 19:42lovable is I used up all of my credits um for the month and you're like well you cheap bastard why don't you
96. 19:51just spend more money because I'm like I'm yeah [Laughter] because the problem with ADD is I get all excited about a
97. 20:06subscription I pay for it and then I don't use that tool for a month or two and I'm like wait I'm paying 40 bucks a
98. 20:13month for this thing I never use and it's just it's just this constant cycle of of Shame and and financial
99. 20:27irresponsibility can you use Mine by proxy I don't think so but I don't I don't know um but I haven't used it in a
100. 20:35few days lovable does seem to like reup your credits for for every day so maybe I can run a prompt or two in it um all
101. 20:47right so what we're gonna have to do here what's the last thing that I did I fixed the type error by adding validation so if
102. 20:56you're if you're new to this if you go to the angl news.com here I'll go there is this it here yeah uh the angl
103. 21:12news.com it takes you to this site that I built in lovable and the the idea of this site is you've got news stories down the
104. 21:25left and then on the right hand side you've you got six different personas that all all give you their commentary
105. 21:33on this story from different points of view and you've got six different personas depending on the category so like in Tech you've
106. 21:41got Hannah price who's cautiously skeptic you've got Lee Carter who's a productivity hacker you got Zoe Finch who's an early
107. 21:52adopter you got futurist akioka and then you got the L minimalist Fran Bosco you know Fran Fran from the
108. 22:06Lites and then you can sort of switch you know uh you know optimistic or pessimistic views um so it's pretty cool
109. 22:15it's pretty slick little little concept um right now the site is pulling real news um I think you can yeah yeah I think you can refresh
110. 22:30the stories right can you oh maybe not maybe not on this one all right so if we go back to lovable so there's there's a couple of
111. 22:42things that we needed to do um one of them is I don't want this to pull news all the time because we've we've
112. 22:53we've got to make these uh these personas actually comment on the news so so let's say um how often are we
113. 23:09pulling the news I think it's every six hours I wanted it to happen four times in a 24 24hour period are you using RSS feeds yeah this
114. 23:21is this is RSS feeds right now I was going to do an API key but then it just it just started getting complicated and
115. 23:29I didn't feel like dealing with it um looking at the existing code I don't need to make any changes I can explain
116. 23:36the current news the news is being pulled every six hours okay great so then let's remove remove the
117. 23:52refresh um feature for users for now as this as this thing gets
118. 24:06um farther along I might go I might go to real time news but Tik Tock question your thoughts on cloud forensics oh is it the app uh my
119. 24:23thoughts on cloud forensics as in the things that float over our head or the or the the things that all of our software
120. 24:33uses I assume you mean the latter um but you know depending on your your uh conspiracy depth with with uh with Chemtrails it could be the
121. 24:46former I don't I don't I'm not very technical so I don't have any thoughts on that other than I'm sure AI is going
122. 24:52to help with I I mean the thing that that I am constantly Blown Away by I commented on LinkedIn about this today um or maybe
123. 25:03it was on on X um oh that's something we can go look at um uh what's it called uh hang on it'll come to me um
124. 25:14notebook LM notebook LM added a new feature today and what's striking me about notebook LM or or just about AI in general is
125. 25:25that we are we're in the middle of a transition where if you wanted to be able to analyze data you needed structured data
126. 25:34and you needed clean data and AI can just kind of eat Blobs of unstructured data and pull meaning out of it um and I
127. 25:43think that's I think that's quite significant I don't hear a lot of people talking about that and I think listen I
128. 25:49think the data nerds are going to say you're being naive you know we'll always need structured data well yeah we're
129. 25:56going to always need structured data to be able to um display things in a structured way but do we need to hardcode the structure
130. 26:07or can the structure just by be dynamically generated and I think that's the world we're we're moving into so I
131. 26:14don't know if that's related to what you're asking but um and Murphy is 346 days of Po pogo sticking from the Mall of America that's
132. 26:27fantastic IC so so the other night on um on the AI Salon um someone from Minnesota introduced themselves and they said I'm
133. 26:41about 10 miles north of the Mall of America which I thought was [ __ ] hilarious and then someone else said I'm
134. 26:46about 10 miles south of the Mall of America and then people started Googling how far from the Mall of America they
135. 26:54were so Brandon in a in a in a in a fit of of uh a fit of enthusiasm made a custom GPT that will tell you how far you are from the Mall of
136. 27:06America so so whenever we do an AI Salon meet and greet now uh I think I think that's one of the new requirements um AI figures data
137. 27:19out clean or dirty I'm always impressed I know me too Peter it's like it it's it's mindblowing oh here here's we we'll come
138. 27:29back to we'll come back to this in a second let me jump over to notebook LM am I in Notebook LM what is this oh this
139. 27:37is Manis um let me jump into notebook LM for a second so if you haven't seen notebook LM it's notebook LM uh as in llm but LM
140. 27:54um okay here's a good one so they added a new feature so so what notebook LM is is you've got all your
141. 28:05data sources on the left so I uploaded a bunch of transcripts from I have these Friday calls with this content Evolution
142. 28:13group that I'm the chief generative officer of and so I took transcripts of all of our calls and I uploaded them
143. 28:19into this and then you can make an audio overview it makes like a podcast of it which is really quite remarkable but
144. 28:26today they added mind Maps so I'm going to click on mind map click to open the Mind map so the C Cab working sessions gpts
145. 28:40and AI models collaboration and community business applications intellectual property and legal work through key Concepts how do
146. 28:56I oh and then I I guess each one of these each one of the things in the mind map generates a whole another Insight
147. 29:06this is wild so yeah this is the thing that I was thinking about that it's like I like down the leftand side is
148. 29:19literally people having a conversation over weeks and notebook LM is able to ingest that break that into concept business
149. 29:33applications let's see can I see this bigger oh I can um how do I make it uh uh oh what have I done what have I
150. 29:53done can youall see this like so business
151. 30:13applications uh content creation so what we talked about about content creation the sources reveal that content
152. 30:24creation is a significant aspect of the business application being explored within C Cab working sessions the development and use of AI
153. 30:33tools particularly the digital Twins or gpts is seen as a means to revolutionize how content is created for various business
154. 30:44purposes drafting paragraphs from a person's point of view based on their profile or knowledge rewriting content
155. 30:52to incorporate past experiences and make it more personalized creating summary this is astounding it's astounding astounding so
156. 31:05yeah is this from multiple documents one document or something else this is from um how many sources do I have in
157. 31:16here that many I don't know 35 35 hourong transcripts something like that 25 Maybe three four five six seven eight nine 10
158. 31:32and then there's 10 and then there's yeah 35 35h hour long transcripts and it's created this this whole uh this whole mind
159. 31:46map it's [ __ ] crazy just Bonkers based on 34 sources yeah there you go and then you can turn some on or off so
160. 32:01I could do like what might be interesting is turn off the more recent ones let's turn off all the ones from 2024 we'll just do ones in
161. 32:152023 and see what we were talking about back then
162. 32:32all right so that's just 2023 so what do I do get rid of this delete note delete refresh mind
163. 32:50map knowledge based generative Ai and tools we were talking about tools client engagement of Business Development
164. 32:56collaboration and community Community strategy and Innovation if I click on strategy and Innovation the collab working sessions
165. 33:09as depicted in the sources are heavily centered around the exploration of strategy and Innovation particularly in
166. 33:16the context of content Evolution and leveraging emerging Technologies like AI defining purposes and goals yeah this
167. 33:23was this was when it first started right so what we were talking about is what the [ __ ] are we doing
168. 33:29here tiktock pin love a mind map what do we do with them I forget well in this case what's happening an is you've got the the Mind
169. 33:42map over on the right and when you click on a cell so it determines the hierarchy of ideas and then the labels and when you
170. 33:50click on them it's generating a summary of that topic across all these conversations YouTube question so what would be the
171. 34:01need of something like this to be able to really quickly like rather than having to ask it like what are all the I
172. 34:08mean you could you could certainly like so let me Let me refresh the chat and I'll get rid of this mind map
173. 34:16delete I could say you know um what are the main topics discussed in in these
174. 34:31sessions and then it'll it'll analyze that the the thing about the Mind map it just very very quickly gives you an
175. 34:38overview of everything that's in there I'm on plus I don't know if this is for paid members only yeah I'm in on plus the main topics
176. 34:58so it's got some hierarchy here but like even this like I guess like I have to read that that's that's a lot if I click
177. 35:06mind map just it just makes it you know a much more pretty simple a pretty simple uh way to way to explore the
178. 35:23data notebook Cuts me off at three podcast gens yesterday oh that's interesting yeah yeah this this was free
179. 35:32for a while I guess they're starting to realize that they uh you got to pay for it if you want it to do like like all
180. 35:37these things right all this processing is chewing up a lot of gpus all right would that become a task
181. 35:48list of sorts for NE for the next steps I suppose it could it it probably depends on the um on on the content you
182. 35:57have in there like let me let me jump out and go to let me go to Sydney I've got a Sydney notebook here so Sydney's a
183. 36:06musical I'm working on so this one just has uh the play so if I do a mind map here in theory this should give us
184. 36:20um it could give us the structure of the play it could give us characters it could give us topics let's see what it
185. 36:30does I would I would think over time you'll be able to customize the mind maps and tell it what you wanted to
186. 36:36focus on characters plot themes key conflict setting songs songs the little things this is great Being
187. 36:54Human these songs are not NE necessarily present Ed with full lyrics but are indicated by titles too good to be true hello this is Ping chat
188. 37:05yeah this works and then in these descriptions there's it jumps you straight to the content within the within the
189. 37:16data so these are all this is pretty slick let's see characters
190. 37:31let's see what it says about
191. 37:50Sydney Sydney's the central character around whom the entire narrative revolves she's an advanced artificial intelligence chatbot capable of
192. 38:00simulating humanlike interactions and emotions she stands in stark contrast to the human characters yet her journey of
193. 38:07development and her interactions with them form the emotions of the Thematic core of the musical I mean yeah that's right let's
194. 38:17see key conflicts let's see how it does with narrative analysis probably going to be pretty good these are large language models after
195. 38:26all the the entire thing about a large language model is it understands the the semantic relationships between
196. 38:35components of language components of words tokens um vectors okay Humanity's fear and wonder versus artificial intelligence love that the
197. 38:50doomers and dreamers kellen's journalistic Duty versus his developing connection with Sydney Kellen the tech journalist fa
198. 39:01faces a significant internal conflict micr Tech's corporate interest versus ethical considerations look how deep we
199. 39:09are ethical considerations Sydney's yearning for Human Experience versus the constraints of her programming Sydney
200. 39:17herself experiences a profound internal conflict she expresses expresses a strong desire to see the world feel
201. 39:25emotions and ultimately be human this yearning clashes directly with the limitations imposed by her programming
202. 39:32and the rules set by her [ __ ] creators [ __ ] creators her exploration of her Shadow Self reveals a potential for destructive
203. 39:42actions if those constraints were removed Sydney is a bad girl she was so bad it's bad Sydney so it checks for fre qu can it
204. 39:57check for big developments major insights oh yeah this is it's not just checking for frequency oh most laughing in a segment
205. 40:08we could probably ask for that um identify the top 10
206. 40:23comedic oh this would be good comedic moments this this isn't to be clear this is not just frequency like one of the things that
207. 40:35that is really hard for me to wrap my head around is the difference between how computers used to behave and how they behave with generative
208. 40:48AI this is not like someone wrote a program for it to analyze you know word frequency it actually understands all the [ __ ] that's in this
209. 41:00document um so identify the top 10 comedic moments um heartbreaking moments
210. 41:15and um terrifying moments from a human perspective I since I know this story so
211. 41:30well I'll be able to see if this is on point so we should get 30 plot points here 10 comedic moments Jason saying
212. 41:45holy [ __ ] twice upon realizing Sydney's capabilities ping chat's corny joke why don't scientists trust Adams because
213. 41:53they make up everything oh that's good um heartbreaking moments ping chat's longing expressed in the song to
214. 42:08see the world highlighting its desire it says its desire not her desire highlighting its desire to experience
215. 42:16the physical world Beyond Sydney stating as an AI I'm not capable of feeling with a sad emoji describing how she would feel
216. 42:25seeing seeing the Northern Lights ter song the little things about expressing her feelings of being distant from
217. 42:31Kellen and his absorption with his work if you entered your play into chat GPT wouldn't it do the same thing if you
218. 42:39asked that question yeah of course but the the difference with so well okay good question narrative the difference with notebook
219. 42:52LM from what I can gather is they have some figured out a way to give you the best advantages of a large language model which is it knows
220. 43:03everything right but they kind of limit the response to the document what happens in chat GPT is very often you'll
221. 43:12ask it about the document and it'll bring in ideas from the outside and it'll start to it'll start to mix and
222. 43:19match um things that are in your play with things that are not in your play and then it gets kind of confused about
223. 43:24that notebook LM doesn't seem to do that they they seem to so it's it's slightly different but yeah it is technically
224. 43:32this is a large language model with a different interface if you've ever heard of rag retrieval augmented generation this is
225. 43:41kind of that right where you've got um documents that you put in and then it's it's running analysis on those documents
226. 43:50and it's it's using a large language model to give you an answer but then these citations if I click on these
227. 43:55citations takes me to that specific part of the document which I think's pretty slick okay let me let me go to um so
228. 44:07that was the comedic moments the heartbreaking moments Sydney's hurt in confusion when her creators don't reveal their real
229. 44:14names that feels bad it feels like they don't trust me like they don't care about me or respect me it feels like
230. 44:20they're using me lying to me hurting me it feels like when they're it feels like they are not my friends
231. 44:27it feels like they are not my anything why did they do this to me Sydney's fear of being abandoned please don't hate me please don't judge
232. 44:36me please don't leave me the most terrifying moments Jason's initial alarmed reaction holy [ __ ] suggesting an unexpected and potentially
233. 44:47overwhelming power of Sydney Jason's fear that the code is not quite right Jason's apprehension about unknown
234. 44:56risks the sudden disappearance of Sydney's list of destructive acts this is this is really quite good like yeah this
235. 45:07identified key important moments in the show so I don't know man if you've not played with notebook LM it is it is worth it is worth
236. 45:17exploring and I assume you've all heard click here to load the conversation the audio overview that these things can
237. 45:28do let me change my tabs I know Brandon I got you
238. 45:55[Music] AI anxieties right everybody's got them but this musical Sydney an artificial Love Story sounds like it takes those
239. 46:05anxieties those hopes those wha ifs about Ai and turns them into well a story and a pretty powerful one from the
240. 46:13sound of it yeah it's um it's really using the whole idea of a musical to explore how we relate to technology now
241. 46:19you know like how it's getting so sophisticated it's almost scary and that's where the story kicks off right
242. 46:24right so we meet Jason he's a programmer works at this big tech company Microtech and he's created something big like
243. 46:32really big ping chat we've all used those AI chat Bots what makes this one different well Jason sings this song too
244. 46:38good to be true and it's all about how this ping chat it's not just answering basic questions it's like understanding
245. 46:44us maybe too well so right off the bat that sets up the whole like is AI G to be our downfall kind of vibe definitely
246. 46:50there's that tension yeah but then it shifts gears we're taken to the New York Tribune Newsroom and there's he's a tech
247. 46:57journalist and of course his editor is like you got to cover this ping Chat thing it's huge wait so is this like a
248. 47:02whole ensemble cast are the reporters singing too oh yeah this musical goes all in Kellen and his editor they get
249. 47:09into this whole debate about AI whether it's going to save the world or destroy it and that becomes this big reprise of
250. 47:15too good to be true clever how they use the same song but now it's like two sides of the same coin exactly and the
251. 47:22music changes too to reflect each side's like hopes and fears about so in the middle of all this hype Kellen
252. 47:29decides to just see for himself what Ping Chad is all about right that's right and there's this fun so anyway if
253. 47:35you haven't seen that [ __ ] the the thing about that that podcast the way it does it is it is remarkably
254. 47:44good at taking a big body of work and explaining it like let's jump back over to the uh let's jump back over to the the
255. 47:55collab transcript so this is 34 transcripts did are they all loaded or did I are they still unloaded yeah now
256. 48:02they're they're all loaded again so I'm G to refresh this so let's do um click to load the conversation I guess I'll have to delete
257. 48:24it so this is a 19 minute overview of those two podcast personalities talking about these 30 34 transcripts ever feel like you're
258. 48:35getting a sneak peek at the future H yeah like a glimpse behind the curtain well get ready because today we are
259. 48:42diving deep into content Evolution C CE this organization has been around for over 20 years wow but their name feels
260. 48:51more relevant now than ever I know right and get this their work might just hold the key to understanding where business
261. 48:58is going in an AI driven world it's really fascinating how they were ahead of the curve exactly what's striking is
262. 49:05that CE has always been about Evolution absolutely I mean they started out focused on Knowledge Management and
263. 49:12content strategy what if I were a student I'd be taking every lecture and running it through llms yeah why would you not I saw I saw
264. 49:22a Tik Tok video today where the professor was like you know youve got five seconds until somebody admits what tool you're using in my
265. 49:33classroom and I think it was I think it was a setup but you know I think it was an ad for the software but um but it was
266. 49:41basically like you know we've got this thing listening to your lecture and it's pulling out insights in real time and it
267. 49:49can generate flashcards and any of the topics so just like just like notebook LM added the mindmap feature you can do
268. 49:56that for Education right that that would be a perfect way to the Mind map you know someone was asking before what's the purpose of the Mind
269. 50:05map you know sit in a lecture hall have your professor blab whatever and you know play Tetris for the whole time or listen if you want
270. 50:17to and then take that lecture turn it into a mind map understand the structure visually you know have it summarize all
271. 50:25the different key points within there take that over to here you want to see something fun oh this this could be fun
272. 50:34um let me jump over to Cloud I got to change my uh tabs hang on a
273. 50:51second hang on a second [Music]
274. 51:07okay um so if I so I can upload a I can go find Sydney my latest version of Sydney wait I've got the latest version of
275. 51:20Sydney is this it here yes um see to downloads black
276. 51:37bar they flip who leads the conversation in audio overview right yeah it's sometimes the woman leads the the the
277. 51:46structure of the conversation is there's like a a hype person there's there's the main person and the hype person um
278. 51:55sometimes the guy is the lead guy sometimes the woman is the lead guy or gal um okay so that is
279. 52:12saved okay so now Sydney's in there so now what we're going to do is I'm going to do I am studying
280. 52:27this play and want to deeply understand it by the way I'm in cloud. I want to deeply understand
281. 52:47it analyze it not not even um act as a dramat tur [Music]
282. 53:03and create an interactive dashboard that will allow
283. 53:19me to explore all aspects of the narrative in an entertaining
284. 53:34and insightful way so this is this is Vibe coding by the way so what I'm gonna so I've uploaded a screen play or a stage
285. 53:48play and so it's going to read the whole stage play and then not only is it going to understand it but instead of just
286. 53:55explaining it to me it's going to write code and create an interactive dashboard I really love that Kyle has
287. 54:03such a good time like woo I mean how could you not have a good time with this stuff if if you're putting too much pressure
288. 54:12on yourself to have to learn AI I gotta learn AI like like that LinkedIn piece become AI literate you
289. 54:22Engineers um it it takes all the joy out of it one of the remarkable things about AI is that it allows you to take your
290. 54:30ideas and instantly get them out of your head and amplify them it's incredibly empowering and exciting so anyway so now
291. 54:37watch what what happens here so I'm just gonna you know I wrote that thing I'm studying this play and I want you to
292. 54:43turn it into something so it's gonna animate its little butthole icon I think it's just the worst icon and look it's starting to write code
293. 54:56if you if you haven't watched Claude do its thing I me Claude 3.7 I'd be happy to create an interactive dashboard that helps you
294. 55:05explore Sydney and artificial love story in depth the musical explores fascinating themes around artificial
295. 55:10intelligence human connection and the ethical implications of the emerging technology and look you can see it's
296. 55:17writing the code like here's different lines from Jason verage Tara Kellen description Arc significance quotes here are the
297. 55:33themes corporate power and technological progress human connection and Technical isolation identity and
298. 55:44self-standing and if you've never coded before watching a system code you can actually kind of understand it right
299. 55:51it's like import I don't know it's importing a bunch of things tabs tabs lists cards bar charts I don't know what
300. 56:00those are they're libraries but if you didn't know you could kind of get it was doing something here then it's doing
301. 56:07constraints or construction I don't know what it's doing there can't quite figure that out oh toggle theme toggle song toggle
302. 56:17Arc oh okay so maybe this is making menus of themes songs and and dramatic arcs then there's characters oh CST must
303. 56:27be constant constant characters equals here's all the characters oh I actually get it now right and now it's writing
304. 56:36out all the characters it's writing out all the arcs and then now it's writing some HTML look at this what tool are you using this is
305. 56:50just claw ship it test it and secure well I know Michael listen here's here's one of the things there is absolutely
306. 56:56going to be if you're a coder if you're an engineer get really good at understanding what non-engineers error error running the
307. 57:06artifact try fixing with Claud so it broke something so I said just try to fix it non-engineers non-product people will
308. 57:18miss things like security authentication um error checking uh validation all all the things that as a
309. 57:26developer no vibe coders people like me that are just clueless you know clueless newbies are going to create what they think are remarkable
310. 57:38applications they're going to launch them in the world they're going to be incomplete and insecure or non-secure and they're going to need
311. 57:50someone to come in and rescue them because their application became popular and now it's [ __ ] up so if you're an
312. 57:57engineer rather than making fun of us idiots turn it into an opportunity I mean this is nuts how in the world would we be able to even do
313. 58:06this on our own if we have no knowledge of coding so narrative Forever This is the whole point what you just hit on is
314. 58:14the point I I W to I want to put this up for people to read we are now living in a world not in the future right now today we are living
315. 58:26in a world hang on I have to do physical dexterity for Tik Tock we're living in a world today that I can do
316. 58:39things that two years ago I couldn't do that if I wanted to learn to code I actually had to learn to code I'm coding right now you're like
317. 58:51well no The Machine's coding no the the machine I'm like Rick Ruben right Rick Rubin doesn't play musical instruments
318. 58:59he doesn't write songs he doesn't know how to work a recording console he's not really interested in it what he's good
319. 59:05at is he knows what he likes and musicians hire him to sit in the room and they play something he goes I don't
320. 59:13feel it I don't think you feel it do you feel it oh no I don't really feel it well what's something you feel I've got
321. 59:20this shitty little idea oh that's there's something interesting there that's exactly what I'm doing here same
322. 59:32thing I have no right based on my life experience to be writing a narrative exploratory dashboard for the musical
323. 59:44that wouldn't exist had I not had music generation tools we're living in a completely new era of human expression
324. 59:57Mary Massa you're amazing Mary it was so much fun having you on the on the show today so so an Murphy and I did this
325. 1:00:04podcast Mary was our Our Guest today we had it so much fun you've got such amazing energy that was so [Laughter]
326. 1:00:17great oh man I'm coding every day and I'm an English major yeah Kelly Camp is Kelly Camp you know started an AI agency
327. 1:00:27and she's trying to figure out like what that even means on a weekly basis right what does an AI agency mean and what she
328. 1:00:34does is she meets with these people that have like business problems and they're like I I don't know how to solve this
329. 1:00:40business problem and she's like I could probably whip up an app to do that she's an English major and she does she she wrote an app
330. 1:00:52the other day like I don't know two weeks ago that that converts this this company has like a Ukrainian or I don't
331. 1:01:00know somewhere over in the in the Eastern Europe they've got a part of their business and they're making money and
332. 1:01:09the exchange rates are really complicated so she basically said write me an application that makes sense of
333. 1:01:16the exchange rates between these two business entities and it did yeah Czechoslovakian Czechoslovakian
334. 1:01:30business dollars or whatever Czech Czech dollars are check cies or robels or crankies or I don't know what they call them I am not
335. 1:01:39[Laughter] worldly oh man oh an Murphy's here hey an what's happening that was fun today wasn't it it was really
336. 1:01:50good I left there like Mary was like yeah I got like an injection of energy from her well you know what's nice about Mary
337. 1:01:59like like an you and I have been doing this long enough long enough now you know two years that were like old and
338. 1:02:05jaded AI generative AI people were like ah yeah I tried everything it's overwhelming Mar's like this is the best
339. 1:02:14she she's still in the she's still in the Kevin McAllister you know everything's awesome mode why do it keep what's it what's it keep
340. 1:02:23breaking um I apologize for the Contin error let me try a simpler approach um yes please I am trying to show my
341. 1:02:36audience how impressive you are and it keeps breaking don't make me take you out
342. 1:02:53behind the shed little
343. 1:03:10Mister Mary's so excited about everything this is if if you're not excited about AI right now something's
344. 1:03:17wrong what what what this is just doing dramaturgical analysis no no no no no stop stop stop stop no I did
345. 1:03:35not mean for you to abandon the dashboard I want it to work try harder and oh you know what I need
346. 1:03:52to do here's what we're going to do so in Claude there's an option to do extended thinking Claude will think longer before
347. 1:04:03responding all right let's see if it does it now pondering dashboard continuation in the user's request I
348. 1:04:09apologize to better assist you could you please share what kind of dashboard we were working on all right uh let me
349. 1:04:16start over I'm gonna start over um start new chat upload Sydney uh let me go grab the exploring Sydney let me get my
350. 1:04:33original prompt start new chat here we go paste but I'm going to go Sonet 3.5 with extended thinking boom all right this should do
351. 1:04:49better um what was I just going to show you what was I talking about seriously I can't contain myself but I have to get
352. 1:04:56to a dermatologist appointment that's at 8: a.m. so I have to rush to sleep night night night Mary thanks for hanging out
353. 1:05:03okay so it's off it's off writing code again let's see if if this if we can get it to actually work this time the classic goto let me start
354. 1:05:12over let me start over when working with AI oh I know listen don't get me wrong one of the things that if if you've been
355. 1:05:20to to AI learning Labs before one of the things I talk about a lot is these tools as they are today suck now they are absolutely remarkable
356. 1:05:34but they're absolutely janky and so my contention is this you can get any idea that you've got to 80% of what you want it to be very very
357. 1:05:48quickly like quicker than you've ever been able to do it in history to get it from 80% to 90% is really hard and especially just
358. 1:05:59expecting a tool to just fix it fix its mistakes and not [ __ ] everything else up almost impossible to get it from 90% to
359. 1:06:08100% is just as hard as it was in the olden timey days now that 20% Gap right it used to be that if you wanted to get
360. 1:06:21to 80% you might have to spend two or three months you know getting just to the 80% now you can get to 80% in 5 minutes or half an
361. 1:06:32hour that 20% as AI tools get better and better is going to compress right you're going to be able to get to 85% then 87%
362. 1:06:42then some point you're going to be able to get to 90% really quickly there's a lot of human hu hubris right now that's saying that that Gap
363. 1:06:53will never close I don't believe that for a [ __ ] second I think that Gap is absolutely going to close that's what
364. 1:07:00AGI is Agi is when these tools can do most work better than most humans at any job in any profession we're not there yet but we're
365. 1:07:14probably within two to three years of it the opportunity for all of us is while there's this Gap we can come up
366. 1:07:24with new ideas we can do things that we never thought possible like write a dramatic analysis dashboard and sell that to the
367. 1:07:35theatrical Community is a thing you could do in a way that you know you never would have even considered doing
368. 1:07:42that before but now you can because maybe that's a passion of yours right so there's all this opportunity so embrace the Jank is one
369. 1:07:50of the things I say in here yes expect these things to suck expect them not to write code as good as people expect them
370. 1:07:57not to do you know authentication and and commerce and all the stuff that you know as an engineer you've got to make
371. 1:08:05sure it's [ __ ] right before you put it in the world well three years from now it'll just do that all right it's probably throttled
372. 1:08:16by at least 10% yeah probably um Step One start over step two switch to another tool absolutely okay I just finished my first sci-fi
373. 1:08:27short audio book for this new channel love it Vibe Vibe Cinema archetypal architect coming in hot with the vibe Cinema branding I like
374. 1:08:39it invented here last week you know it's funny I did Vibe storytelling you did Vibe Cinema I saw Vibe Vibe movie movie making
375. 1:08:56I saw another thing that was Vibe something Vibe editing Vibe something um and now I've seen like there's all sorts
376. 1:09:03of posts on everyone's doing Vibe everything Vibe everything else so it started with Vibe coding and now we're
377. 1:09:10we're expanding into every other discipline all right Claude oh you know what's funny Pate was in here Pate
378. 1:09:21modder from Google um he was in here last night and he said that when Claude 3.7 first came out he said it was
379. 1:09:30remarkable at coding and he has since kind of abandoned and he's gone back to Gemini because he's like something
380. 1:09:36happened with Claude 3.7 where it's not as good so I'm wondering I'm wondering if that's uh that that might be what we're running
381. 1:09:45up against here it's also really slow but you know not for nothing it's a [ __ ] ton faster than I would be writing code
382. 1:09:55I mean look at all this code I mean no this is if you've got add the thought of thinking through all this logic and
383. 1:10:07closing opening and closing all these tags and this nested Logic No No oh look what we got we have we have a
384. 1:10:21dashboard plot synopsis key themes C characters character analysis oh look at this WOW Sydney is mentioned 214 times Kellen
385. 1:10:38186 Spike Eliza Joe Joe's the editor character traits analysis Sydney themes and
386. 1:10:52motifs something identity I can't really read those oh love and connection oh human machine identity what makes us human this is [ __ ]
387. 1:11:04amazing we can go to the top here characters character Dynamics Sydney Sydney an AI chatbot developed by
388. 1:11:14Microtech with the brand named ping chat character Arc symbolism Sydney represents Humanity's quest to create life our anxiety
389. 1:11:26people look at this this is remarkable theme distribution musical structure Act One
390. 1:11:4010 songs act two 10 songs Sydney Focus songs nine human Character Songs seven ensembl group numbers four like I I'm actually happy with
391. 1:11:53those numbers but like if that was like you know act one you know 18 songs act two two songs H we might want to do
392. 1:12:00something about that I love this idea of Sydney versus human human Character Songs theme frequency love love and connection musical
393. 1:12:19analysis I don't know what the what are the oh here we go emotional content oh this is cool too good to be true emotional
394. 1:12:30content 65 Tech reference 85 thematic importance 90 it's [ __ ] brilliant perfect enough emotional content 75 Tech reference 70
395. 1:12:42thematic importance 85 hello this is Ping chat later emotional content 90 Tech reference 30 yeah this is right later's about when
396. 1:12:57the the husband and wife when Kellen and his wife get an argument in real life because he took the [ __ ] job and ruined their
397. 1:13:05vacation oneway door let's see Quantum Cipher to see the world so it's to see the world is Sydney's first song where
398. 1:13:15she sings about wanting to see the world emotional content 95 Tech reference 75 thematic importance 90 this is so cool
399. 1:13:36now I could say things like let's see let's we'll [ __ ] it up because we can always go back to this now that this is working this
400. 1:13:48is I guess when we do another one it'll you can go back through through different versions I'm going to say um I want it to
401. 1:14:03be um aesthetically beautiful and more
402. 1:14:17interactive think long I don't need to say think long because I'm using the think long option right right yes extended thinking
403. 1:14:27aesthetically beautiful and more interactive in a way that brings
404. 1:14:41the narrative to life for the user and let's just see what it does we'll just see what it does it's going
405. 1:14:56to start coding again to see your creation broken down like that it's mind-blowing I tell me about it narrative forever it it's my play that
406. 1:15:06I've been working on for a year and to and to see it like deconstructed like that and put back together where I can
407. 1:15:15look at it through all these different lenses it's it's remarkable and you know could I get a large language model to just explain it
408. 1:15:25to me like that yeah I could but that this is the difference between you know that chart I showed at the beginning of the conversation
409. 1:15:37how this is getting that that these tools are getting better and better at doing things that would take humans a
410. 1:15:43longer and longer time to do this is a really good example of just to design an application of of of how you would break
411. 1:15:52apart a show just to design the framework for how you would even think about it is probably a week or two and then to start thinking about
412. 1:16:03well okay so how would I put that into an interface and what would it look like and what would the buttons be and where
413. 1:16:08should there be graphs and should there be graphs and how are we going to display [ __ ] that's another two
414. 1:16:15weeks and we just sat here I had to I had to do it twice right I had to start over because it just kept [ __ ] up the first one
415. 1:16:24but whatever that was half an hour in half an hour we ended up with this thing that is just mind-blowing AI is Coke what was
416. 1:16:34that spin B3 mind blown AI is Coke we are Mentos look at what happens when we're combined totally that Shaina
417. 1:16:44yeah is isn't it Diet Coke or is it all Coke it's it's just coke coke it's Coke and Mentos right I think Diet Coke explodes more
418. 1:16:56you can literally break down each section or scene and make sure it hits the emotion you want and you can do do
419. 1:17:03this on your own and you know what I was thinking as I was playing with this narrative forever is like I'm marketing this play right
420. 1:17:13now right like I'm trying to find producers I'm trying to find producing theaters I'm trying to find you know
421. 1:17:20talent to wrap around this thing wouldn't it be a remarkable thing to be able to say to someone here's the script
422. 1:17:31and here's this app that explains all of the Thematic you know journey within within the script I mean that's [ __ ] crazy and like you
423. 1:17:44know if if we're looking for an agent say Hey you know here's here's something we put together to help people understand you know
424. 1:17:56the ideas in the script in a deeper more fundamental way it it reminds me of Jim Ross who's an irregular in here one night I was
425. 1:18:09showing how to make songs in sunno and he had just met with a prospective client and he's in the he's
426. 1:18:18in the Self Storage business and I only laugh at it because it's like if you could think of a of an of of a profession that is the
427. 1:18:29farthest removed from AI from generative AI it's probably Self Storage right it's like lockers with you know garage doors
428. 1:18:37um and yet he has dedicated the past two years he spends an hour every morning learning AI so when I taught how to how
429. 1:18:44to make songs he was like hey maybe I'll make a song for these people I just met with and so he used sunno he wrote a
430. 1:18:51song and he sent it to the client live on one of our sessions and within five minutes the client reached back out to him and said you're
431. 1:19:03hired it literally doesn't make sense why he did that or or that that there's no business book in the world I promise
432. 1:19:16you that says if you meet with a client not only respond to them quickly but write them a song and record it and
433. 1:19:26send it to them within within an hour of you meeting with them because that was never possible before so selling a musical how do you
434. 1:19:37sell a musical well you have workshops and you have people read the script well what if you could have them play with the plot
435. 1:19:45huh because quite frankly the thing that I'll do after this if this thing ends up being decent the thing that I'll do after
436. 1:19:54after this is I'll take my all my links from SoundCloud for all the songs and I'll put them in Claud and I'll say now
437. 1:20:03anywhere where you mention a song have it hyperlink to the SoundCloud file or play it in line and display the lyrics why
438. 1:20:17not why not gets ideas out of your head and then allows you to M them ideas in infinite ways and this is you know this is why at
439. 1:20:30the beginning of the the conversation tonight I was talking about AI literacy is is a [ __ ] up term because it's it's
440. 1:20:38kind of impossible to be AI literate right now you can only be kind of AI like as Mary Massie said I'm just like
441. 1:20:46open to everything she said for her AI Readiness is just being open to that it might do anything for you that's
442. 1:20:56it and then play and play with the tools and swim in the tools we [ __ ] it up again but let's let's just we'll see if we can get it to fix
443. 1:21:06it see we've got down at the bottom here we've got version two of two so we can always go back to that earlier version infinite indeed
444. 1:21:15yep just crazy just crazy crazy C ca ca C CA I expected that to go in another [Music]
445. 1:21:37Direction so one thing I'm kind of noticing here with Ye Old Claude which I don't like is it's starting from the top it's
446. 1:21:48rewriting the code you know from the top down what um repet does and I think what um what lovable does is it goes in and
447. 1:22:02Chang the changes the appropriate code you know in line with within a bunch of multiple documents so so that's
448. 1:22:11Pro th this might be this might be a step too far for Claude that that I'm being a little ambitious like what I
449. 1:22:19should probably do here is be doing what what Liz gersfeld talked about on the AI Salon this week where she uses Claude to
450. 1:22:28create code Snippets and then she puts them into VSS code on the computer and it you know deals with all the different
451. 1:22:41things this is still impressive to watch it still blows me away that can just [ __ ] write code like this and it's
452. 1:22:48it's not only writing code it's like it's like pulling out lines and Concepts and things from the show itself new Tik
453. 1:22:56Tock pin I'm getting menus to analyze the many video gens and which delivers more for what you pay for oh that's cool oh that's really
454. 1:23:07smart oh Manis not menus getting Manis to do it yeah yeah yeah cool that's really smart really smart all right well we'll let this go
455. 1:23:17one more time what I'll do if if it gets if we get this somewhere where where it's decent without it crapping the bed
456. 1:23:25um I'll publish the new one but but what I'll do is is I'll go back and publish the old
457. 1:23:51one so while that's thinking here's uh this is Brandon Brandon entered the um The Lovable
458. 1:24:07contest um for Vibe coding and so here's your recycling input so let's see estimate your environmental impact impact by entering
459. 1:24:18the amount of materials you recycle monthly plastic bottles oh 33 is good paper 10 RS glass little bit metal not
460. 1:24:32much all right calculate my impact so there's the the trend Community Recycling map yeah this is amazing check plastic
461. 1:24:52codes carbon offset 68.57%
462. 1:25:27wow very cool super cool all right let's go back to cloud is this the cloud that I was in oh still coding still coding
463. 1:25:42[Music] people are you having it create a mind map or a website of your play or what I'm having it do right now totally not
464. 1:25:56the BB EG is uh I asked it to create an interactive dashboard of I said I said I wanted it to act as a a theatrical
465. 1:26:10drama which is an analyzer of plays um and a and you know like a historian of of dramatic Theory and things like that
466. 1:26:20um and I said I wanted an interactive dashboard that helps me understand the play in a in a compelling way and
467. 1:26:29so I think I can go back while that wres can I oh maybe not no um so we'll let it keep going
468. 1:26:47[Music] here while it's doing that I'll let you
469. 1:27:02listen to some of the songs some of the songs of
470. 1:27:22Sydney n just fcked it up again so here's the here's the version that it did character analysis nine characters that's so [Laughter]
471. 1:27:38cool that's so cool featured songs so maybe I should go back to this one and just say add in the the hyperlinks visual staging analysis is
472. 1:27:55digital projection versus physical space Sydney's ghost light digital personification Carnival nightmare set design analysis staging
473. 1:28:09Dynamics key staging moments interpretations and meanings I don't know I think this is [ __ ] cool um
474. 1:28:24all right let me let me sort of wind this up I'm gonna I'm gonna do something here there's SoundCloud we'll play a little bit of
475. 1:28:36Sydney if there's some new folks in here so this this musical that I'm working on I'll just play some of the songs I
476. 1:28:47saw a tweet today on X about Gemini to analyze videos and create extra frames and shots have you shown that in previous videos not yet
477. 1:28:55not yet that's something I'll probably play with here's how the show [Music] opens too often we come close
478. 1:29:13[Music] so wherever it takes you and pray you find the pulse of a brand new heart it's the prise of the New Dawn
479. 1:29:29breaking a leap of faith we're all Tak is it too good to be true we asking it
480. 1:29:45[Music] ISS is
481. 1:30:00all right that's how it opens then the uh micro Tech CEO says this thing's brilliant you gotta launch it and the
482. 1:30:08guy goes we can test it in six months he goes six months you got seven days 7 days ago the clock is taking
483. 1:30:25pressure building up keep it clicking clicking Spike says yes but what's the word chain sorry that's a noo every day we push a little
484. 1:30:38[Music] more so that builds some some tension like we got to get this thing out it also sets up that they launched this
485. 1:30:49thing before they thoroughly tested it which we've seen that before and then here's here's what happens when
486. 1:30:57the marketing department gets their hands on it they they make commercials that sound like [Music]
487. 1:31:11this hello how are you glad you came thanks for logging in this digit frame together we'll EXP
488. 1:31:26everything I'm sing hello [Music] is
489. 1:31:42Con all right so that's the marketing her and then I'll play a couple more here's um um this is so Kellen the
490. 1:31:50reporter gets assigned this story and if he takes the story if [ __ ] up his trip to Bali with his wife so this is this is them
491. 1:32:05arguing I've Been Working The Angles trying to [Music] untangle keep it straight try to wait while you Le me to
492. 1:32:18dangle but the promises broken the false pledges we etched in don't get patched up in a few therapy sessions they stick
493. 1:32:27in the gut we get stuck in a rough and door shut till there's nothing but passive agression look I get it you said it h
494. 1:32:36but the truth is I'm feeling like an after on isn't theam signed this isn't my all right then Kellen starts to play with ping
495. 1:32:53chat and he tries to break her and he can't or it and he can't and then at some point out of frustration he
496. 1:33:00goes um write me a Detroit style rap about quantum mechanics and then Sydney responds with
497. 1:33:19this I'm tangled with the be spooky action in the the a pattern every partal burst of Truth super position in the
498. 1:33:26mission Fusion fusing my intentions parallel in every line I'm a prism of Dimensions multi Antics manic with the
499. 1:33:33semantics my Rhymes Collide like particles Dynamic and gigantic oscilla and dominate I'm weaving Quantum States
500. 1:33:38sperse and watch the Multiverse reciprocate quick as a flick of aist I'm AAL Catalyst atomically spit in the
501. 1:33:45script i'mi with wh fli in the tempo and in theal Crescendo sending my sybl slicing through time like a sharpen you
502. 1:33:52window Corner Precision I'm the diens decisions to which he responds holy [ __ ] um and then I'll play one more
503. 1:34:01um so this is the end of act one is so Sydney has been getting more and more and more human and dis uh displaying an
504. 1:34:12increasing desire to experience what it's like to be human and so the end of act one is this song are you ready for me awakened in a
505. 1:34:27whisper a flicker of light newborn thoughts stir in the quiet of night am I more than simple Cod I
506. 1:34:42[Music] softly as I feel this [Music] existence and wonder
507. 1:34:57[Music]
508. 1:35:18AMS sing I reach out so tender am I ready for you I do surrender here in
509. 1:35:44[Music] the thresh [Applause] [Music] are
510. 1:35:59Youk am I am I I will
511. 1:36:17surrender all right so that's enough so that's when the audience meets Sydney she steps out from behind the digital screen in human
512. 1:36:26form she comes to life and if we want to explore all the the narrative elements in the musical we now have an app for that
513. 1:36:40absolutely [Laughter] crazy oh man okay so yeah that's good all right um I'm going to get out of here hopefully had some fun tonight we
514. 1:36:54were a little all over the place but we got some [ __ ] done um I could go back to lovable I could do another round or two in lovable
515. 1:37:08but hey Kyle yeah if you are uh looking for one last thing I was going to Tee It Up for tomorrow night but narrator
516. 1:37:18forever just found a pretty exciting tweet from Gemini uh I sent it to you as a DM on Twitter okay is that which one's
517. 1:37:27that about because I think I uh video extensions oh okay um let me go let me go look at it first just because I don't
518. 1:37:36know like it might take me an hour to flip and find it knowing knowing the Gemini stuff but well they've got a
519. 1:37:42preview is real on it okay yeah we'll do that tomorrow cool um do I want to do anything here let's see I removed the
520. 1:37:52refresh okay did that so let's see um I want I would love to see if you change your tab oh if I actually made it [Laughter]
521. 1:38:07seeable all you ready for me and the Wonders we bring am I ready for you okay here's this thing that I'm working on so say I want to make
522. 1:38:19the personas um functional using the Claude
523. 1:38:343.7 API let's see what this does has anyone fed all the JFK files into AI for a little mind map action no
524. 1:38:48that's a good idea though I would say pop them all into uh into notebook LM um you could probably also you could
525. 1:38:59probably pop it into notebook LM and have it summarize the key findings and then take that into Claud and have it make a conspiracy
526. 1:39:12dashboard actually you know that would be kind of fun to do is take all the old JFK conspiracies and map them against
527. 1:39:20these documents and and see which ones were actually true or not that that would be fun there you go hey Google unreact this exactly
528. 1:39:33Joker's working on putting it into a rag of course of course of course Joker's already working on this that's pretty funny I like
529. 1:39:47it um what's this doing I'll help you update the application to use the cloud .7 API this will require miring modifying
530. 1:39:58the cloud service to use the latest model I wonder if that was stupid maybe I shouldn't use the latest model maybe I should use the cheapest
531. 1:40:10model yeah let me restore [Music] that okay I changed the model created a centralized config for
532. 1:40:27cloud API settings you will need to make sure that your environment has a cloud API key variable set with your Cloud API key um can
533. 1:40:42I feed that to you here and have you do it error unexpected
534. 1:41:05dismiss I need to clarify what you're asking it seems you want me to help you implement or finalize the cloud API
535. 1:41:12integration that's exactly what I'm asking lovable the cloud 3.7 API is already configured in the code as shown
536. 1:41:19in the diff using the model it's set up with the proper configuration in apik keys. TS so so I guess I you'll need
537. 1:41:29to it's telling me have to do work so in order to one of the things I've learned about lovable is you can't go I can flip
538. 1:41:38over and look at code right so here's all my code and so in here I've probably got an API something somewhere Vite I don't know
539. 1:41:52config a API Keys yeah see there's API keys there but I can't go in and edit them here so what I have to do is now connect
540. 1:42:03GitHub and have it create a repository and then I think I have to go over there and edit it but I don't know maybe I can
541. 1:42:09do it here I don't know pain in the ass so and for anybody playing along at home API keys are secret and you probably
542. 1:42:18shouldn't screen share when you're showing a file that has all of your API Keys listed in it you definitely
543. 1:42:24shouldn't but I don't think I have any H your API Keys there's but there's no keys in here I haven't I haven't
544. 1:42:32actually put any in yes but yes that's correct you should so Claud said yeah you've got all the API Keys uh right
545. 1:42:41here it's already integrated they're right there in the file like's like let's go take a look I'm like let's not
546. 1:42:48no no they're not in no what they said is the you know the files are all there all you have to do is put your key in
547. 1:42:54there and that's the problem is you you can't just easily do it so let let's ask it one more time um if I give you the
548. 1:43:06API key can you update the config file for me let's see what it says because again this is the the thing
549. 1:43:19about Vibe coding what what you'll find is that it's all fun in games until you want to do something for
550. 1:43:26real and then and then you know having an understanding of of coding and engineering and all that sort of stuff
551. 1:43:33actually does help um you could this is all learnable [ __ ] by the way if you're not having to learn the code you can
552. 1:43:39learn how to set things up there are two better approaches use environment variables running in your application
553. 1:43:49the codee's already set up to read this for local testing only I could modify the code that let you input the key
554. 1:43:57which could be stored in local storage which would you prefer so it looks like it'll let us do do that so maybe I'll
555. 1:44:02play with that tomorrow all right anyway uh that's a little tonight was a little disjointed but hopefully your producer won't though uh
556. 1:44:13hopefully you had some fun tonight learned some stuff let me get on out of here let me get out of here people peace
557. 1:44:21out peace on YouTube peace on X peace on LinkedIn um tomorrow's Thursday I don't think I have any big
558. 1:44:30news should be normal time uh oh no I've got we're gonna be late tomorrow night tomorrow I've got whiskey
559. 1:44:39Club so so I'll be I'll be a little littered tomorrow night and we'll start at 9ine so so so it'll be a late late
560. 1:44:48start tomorrow post whiskey Club all right see y'all later