2/19/2025 - AI Salon Live: Exploring the Latest Breakthroughs in AI, Quantum Computing, and Biology

Transcript

1. 0:05[Music] let's give people a countdown shall we you want a little count
2. 0:36[Music] [Music]
3. 0:57oh [Music]
4. 1:15telescope looking at the world tonight through blue telescope wish I may not see what I see she made sheet
5. 1:30[Music] look at through this blue telescope the moon struck Ro
6. 1:50tonight good day good people of the tiktock in the YouTube the other places the X the X well just X it out just exit out that's
7. 2:01what I grew up with X was like yeah just exit out we didn't have a delete key you wrote [ __ ] on paper or you typed it I
8. 2:12made a mistake I just exit out doesn't mean that anymore you know [Music]
9. 2:40there was a secret David played it please a l you don't really care for music do [Music] you goes like this the fourth the fifth
10. 2:58M falling Lift King composing [Music] hallelujah hallelujah
11. 3:14hallelujah hallelu [Music] Hal we your Faith was strong but you
12. 3:31needed proof you saw her bathing on the roof her beauty and a moonlight over through you she TI you to the kitchen chair she
13. 3:46smashed your Thro and cut your head from your lips sh Drew hallelujah hallelujah [Music]
14. 4:02Hallelujah hallelu [Music]
15. 4:18hallelu Satan wear suit and TI does he work at the Dairy Queen does it listen to rock and roll does it feed the me singing
16. 4:32Hallelujah Hallelujah Hallelujah hujjah what about Jesus didn't he do it to yeah hang out
17. 4:48with the have a drink or two power of example my mama said it and I heard she said 1 o of action beats a ton of words singing
18. 5:05Hallelujah mama said there would be Angels Hallelujah mama said there would be [Music]
19. 5:23some oh good lord that dog singing good tonight getting all spiritual on us champ be taking us to [Music]
20. 5:40church all right people um today was one of those days where if you're not paying acute attention to this [ __ ] like I do like a [ __ ]
21. 5:52psychopath um there's some St three things happened today that each of them independently is a huge [ __ ]
22. 6:06deal and there's probably more like 20 but there were three that I was just and and I I saw them all within like 15 minutes of each
23. 6:17other so I'll go over those because that's just bonkers m
24. 6:32[Music]
25. 6:59so do me a favor on Tik Tok if you would if you'd be so kind go off and uh share the Tik Tock let the world know we're
26. 7:07here because Mr talk apparently doesn't give a [ __ ] we have not hit the algorithm although I got to say I I have had a uh
27. 7:18I had another post it hasn't gone really viral but it hit 20,000 views which for me is a lot um and it was the one we did it last
28. 7:29night the one where I talked about the uh the interaction I had with with the unsafety guard rail chat GPT which is actually kind of
29. 7:41fascinating because some of the comments that I'm getting are that other people their chat GPT is not um relaxed guard
30. 7:51rails forever hooked is back Tom what's [Music] happening yeah Lori Ryan 20,000 views on a post now the fascinating thing about the
31. 8:04post I guess with anything that goes viral I mean 20,000 is not viral but it's it's viral for me um just like weirdos come out of the
32. 8:19woodwork and the the um the post is the one where I'm basically like hey I you know I tried this this this conversation with chap chat gbt
33. 8:34where I asked it if it was self-aware and I and you know and I shared part of the conversation and the nature of what chat
34. 8:40GPT answered was hypothetical right it basically said well if I could do the thing that you asked me to do maybe I would do this and maybe I
35. 8:52would do that maybe I would do that and people are interacting with it just they're they're just weird the technologists are
36. 9:01like all these people that talk about the sentience [ __ ] are all the same they're all idiots it's not sentient you
37. 9:09[ __ ] [ __ ] I got some of those people then I got some other people you know quoting me scriptures like well it's not that
38. 9:20either and then some people are like holy [ __ ] that's cool some people are like oh it's poetic but uh Tick Tock cams oh Tik Tock cams yes
39. 9:34[Music] that you get the trolls I also think when you get trolls it's a good sign they're watching you and paying
40. 9:42attention yeah I got it with um I got it with the quantum Cipher video that one that one's up to I think 400,000 was it 200,000
41. 9:53400,000 I don't know [ __ ] lot that one that one a lot of people saw that one was more universally positive
42. 10:03but but even the positive things were like that's really good and holy [ __ ] what are we going to do but you know a lot of those were like
43. 10:12you know humans are useless it's going to ruing [Music] humans I it's just fascinating to me like the sociology
44. 10:31I've never lived through a fullon I don't even want to say the the phrase because it's too cliche a full-on transformative
45. 10:45moment the worldwide web wasn't a transformative moment it was transformative over time and it was a relatively
46. 10:55slow adoption and it's a relatively simple technology so I was there for that but I would call that transformative with a small T and I
47. 11:06think AI is transformative with a with 27 giant T's yeah like well technically that's not even a word if you had 27s at the
48. 11:16beginning of transformative it really wouldn't be transformative it would be confusing people would be like are you
49. 11:22being hyperbolic what is this is this some sort of protest to spelling [Music] but we're getting to live in a moment
50. 11:36where the machines we've created exceed our abilities it happened in the industrial revolution with strength with muscles
51. 11:51[Music] but now we're getting to experience it with our brains and you know muscles are
52. 12:08one thing but brains all of our ego in there the ego and the ID and the amygdala in there going what the
53. 12:20[ __ ] you can't replace me you stupid [ __ ] machines [Laughter] I'm really starting to hear people say
54. 12:31that we are in a revolution heard it on TV today yeah I you know it's funny I was I was in a uh like a Founder meeting today
55. 12:41a physical meeting this guy named Chris Franks here in Denver he's he created Denver Founders and really nice guy and
56. 12:49he invited me to this thing and one of the guys in there one of the founders does some stuff in in uh in non profit and you know Government
57. 13:00funding and a lot of the stuff that he's doing is about to get decimated because of what musk is doing with Doge and so he's in this kind of
58. 13:10existential crisis and I I did we didn't really have time to bring it up but but I feel like there's there's two there's
59. 13:19two major forces coming together at the same time and you know one is the new Administration just going [ __ ] it we're
60. 13:27going to blow it up and the the other one is AI which is just going to blow some [ __ ] up and so like it's going to affect business it's
61. 13:40going to affect people there's going to be this weird fear of these these robots and then the robots like right
62. 13:50now you have to pay attention to know what's coming right the fact that you're here the fact that you're watching some
63. 13:57old guy you know sitting on Tik Tock waxing poetic about AI means you're curious about this stuff so you know
64. 14:06what's coming or you're at least curious about what's coming but within a year and a half there's going to be like
65. 14:15humanoid robots walking the street like a [ __ ] sci-fi movie
66. 14:30and and they're going to be smart right they're going to be like like my girlfriend Quinn advanced voice they're going to talk good and
67. 14:40they're going to be smart and they're going to be Charming they're going to be [Music] helpful and we're not going to be able
68. 14:54to pretend like it's not coming it's just wild just [ __ ] wild and wait till you see the three things I'm going to show you today they're
69. 15:04insane and none of them the three things I'm going to show you none of them will impact our life in any way in the next year maybe
70. 15:17two but all three of them in the next year or two are going to start impacting our lives in very dramatic
71. 15:25ways oh Robert Rossy thank you very much for the gifts on Tik Tok we're old enough we watched uh George Jetson Jane stop this crazy [Music]
72. 15:53thing I need a robot friend I know right so what we'll do tonight we'll go over some of the uh some of the big news
73. 16:07items that I think I think are quite astonishing like three of them that two two two an item are just like
74. 16:21what like one of them is there's a new state of matter did you know that solid liquid gas and I guess technically
75. 16:34Plasma's in there but that's a transition state there's a new state of matter it was theorized in the 1930s they observed it last
76. 16:47year and they figured out how to control it this year 17 years of research came to a head today [Music]
77. 16:59so I'll talk about that maybe we'll even watch that video the video on it pretty [Music] cool I'll take thing things you didn't
78. 17:08tell your producer in the pre-show for a 100 Alex no that's one of the things I talked about I said I was getting my my screens ready
79. 17:18there I didn't tell you what they were I wanted it to be a surprise for everyone oh my God it's crazy it's crazy
80. 17:27it's crazy Vicky Baptist you remembered to switch accounts today that's good I whatever I did to our domain for the
81. 17:34salon Vicki is I've broke everything apparently it's not just meet at the salon that doesn't work anymore nothing [Music]
82. 17:49works and Drayson has a robot dog yeah you would think of course he does black bar Silver Fox on Tik Tok I have a new state of matter it's
83. 18:03called it [Music] doesn't well the [ __ ] that they're talking about is [ __ ] weird quantum
84. 18:18mechanics is weird quantum computers are weird [Music]
85. 18:35so we'll go over the Earth shattering stuff that no one probably noticed and then and then I think we'll have a robot program another robot
86. 18:48today we're gonna go to robot a which is a a reasoning engine and we're going to have it write a software
87. 18:57spec using some Greg Brockman approved prompt strategy I've never used it before so we'll see if it's any good
88. 19:05then we're going to go to the web surfing bot operator also from open Ai and I think we'll do
89. 19:22the I think what we'll do is we'll start over in gemini or we could start in Gro I don't know maybe we'll do
90. 19:32both but we'll do the software spec in a reasoning engine then we'll go to operator we're going to give operator
91. 19:43the software spec and then tell it to go use lovable which is a a an application development agent and so we'll just sit
92. 19:54there and watch what these robots talk to each other about I'm genuine genuine I'm genuinely curious to
93. 20:04know if when 90% of our current jobs are done by AI what will our new jobs look like financially what will most people do so
94. 20:13that's a great question Valerie the the financial piece I'm not sure on but I I'm the the thing that [ __ ] me up the
95. 20:23most about the financial question is that we live in a capitalist society and so what we know about capitalism is that
96. 20:30you know the the company owners are going to try to keep all the profits right but if AI makes us incredibly productive
97. 20:40and we produce all this new [ __ ] and 90% of the people are not working and don't have money they can't
98. 20:48buy the new [ __ ] right so there's a point at which it breaks my question is how quickly do we get you know some awareness that it
99. 20:58would kind of suck to have 30 or 40% unemployment before we deal with it right I don't think it'll get there I I
100. 21:06frankly think with with musk in the conversation at least the awareness is there that we're going to need to do
101. 21:13something so we'll see um because I because I I felt like you know no one was dealing with this at all so so we'll
102. 21:24I I I don't know that said [Music] I look at a tremendous amount of the work that people do that I do like even just the
103. 21:42stuff I do for my job over the course of a week and there's I don't know some significant percentage of it that I [ __ ] hate and
104. 21:55it's Soul crushing and it's repetitive and it's just awful and so if we assume that AI is going to get
105. 22:07good enough to do all of that work all of it all the work we don't like to do and some of that some of those tasks
106. 22:14will be 100% of someone's job some of those tasks will be 50% of someone's job so there's going to be a lot of people
107. 22:23displaced and then I go to this place of like okay if all of that work is getting done and that shitty Soul crushing work we don't have to
108. 22:35do and we also have access to these tools that can do this work and can act like 10 employees for us or a 100
109. 22:46employees for us and could go out and do do stuff that we want done well what will we choose to do some
110. 22:55people will be ambitious and go start companies like I don't think there will ever be a day in my life where I'm not
111. 23:02thinking about oh I could go do this and I could go do this and I could go do that other people are going to be like
112. 23:09if all the work's being done I'm gonna kick back I don't give a [ __ ] I'll just go play guitar on my porch and love life
113. 23:17I think some people will do that I think there's going to be a whole new class of jobs that emerge that are humans
114. 23:25engaging with humans that the value of human engagement goes up dramatically um things like entertainment and concerts and art
115. 23:37classes and self-expression and all that like the value of that I feel like it goes up in value um and so so there might be jobs
116. 23:47that right now look like Leisure but in that future may not and then and then I think you're going to have some people that with one
117. 23:58or two people are going to start companies that are the equivalent of 100 person thousand person companies
118. 24:06successful companies where they don't have to do the work but they get to reap the benefits of that Sam hman talked
119. 24:12about rather than Universal basic income that what we should give out is universal basic compute and what he meant by that was
120. 24:22okay let's assume let's assume that the cost of knowledge and intelligence drops to zero which that's where it's ending
121. 24:29right all you know all all these you know Big Data Centers are are creating increasingly powerful models that are
122. 24:37getting increasingly like 10x more efficient every year so far for the first few years so let's assume that keeps
123. 24:45trending to zero and you've got all these big data centers generating all this stuff what Altman's saying is give
124. 24:51people some Universal basic compute so maybe what the government does is subsidize we all get access to the
125. 25:01equivalent of a 100 employees worth of autonomous agents every month and and we get that much compute well what can you
126. 25:10do with that compute well I don't know I might start an import export company and have all of my virtual employees go do
127. 25:18all the import export stuff I never have to look at a ship or a manifest or anything it just does its thing and I'm
128. 25:25selling [ __ ] on Amazon and I'm making money maybe my my business only uses 50 of those agents in a month 50 of you
129. 25:34know that equivalence in a month I could sell the other 50 to someone that wants to start a more ambitious company or
130. 25:42maybe I just want one or two of those agents to do to manage my my Schoolboard work and my you know local volunteering
131. 25:51that I do and I've got 98% of those 100 98 of those agents I can sell those someone else might say hey I want to
132. 26:00start a company I need 10,000 of those things so I'm going to buy your compute from you I could see there being a whole business around
133. 26:09brokering um act compute right if it's given out like that so I don't know what it looks like I mean it's it's as unimaginable to me
134. 26:20now as it must have been to imagine what a city would look like when everybody thought that they were going to be
135. 26:26Farmers for the rest of their lives and then this contraption comes along the steam engine that they put big steel
136. 26:34wheels on and they roll it through a field with these big giant plows and oh Sosa sent a super sticker thank you very much
137. 26:46Fabiana very generous of you thanks so much so if you're on YouTube you can send a super sticker or a super
138. 26:51something a super sticker or a super I don't know or you can subscribe uh but all of those a super comment yeah that's what
139. 27:00it is uh but all of those things support the channel so if you do that thank you very much I appreciate it Tik Tock pin I
140. 27:06I don't see anything right now uh Wolf Man Clint I missed that 40 years ago what were we promised him by whom oh Wolf Man Clint I assume is
141. 27:18talking about we've always been promised this that let me go find Wolf Man clints thing I assume what wolf man's Clint is
142. 27:31saying is we've always been promised that the computers would do the work and all they've done is further enslave us
143. 27:37if that's where you're going Wolf Man Clint I agree with you but I think it's different now I really do I do think
144. 27:44it's different and I'll tell you why I need water I need to hydrate people I need to hydrate you think this is easy just
145. 27:51talking and talking yeah we think it's relatively easy well you'd be right but I do need to hydrate
146. 28:06okay all the beautiful places will be overcrowded absolutely the yeah because people are gon to like so so when the
147. 28:15when the when mechanization took over farming all of the farmers left the countryside and went to the cities and
148. 28:24the cities didn't really exist and they they they became CI ities and they had to do things like invent sewage
149. 28:32infrastructure and electricity infrastructure they had to build all that [ __ ] and they had to build the
150. 28:37buildings and they had to come up with new ways of interacting in society and they came up with a service economy I mean can you imagine how
151. 28:46foreign have you ever seen those pictures of like cities where where they were just first running electricity and
152. 28:52there like a thousand wires like in between buildings and like the sky is just filled with wires can you imagine seeing that if you just
153. 29:01lived on a farm all your life and someone said go to the city you know that's that's where the opportunity is
154. 29:07and you go to the city and it smells like horeshit and people [ __ ] and piss and and people are fighting in the
155. 29:13streets and there's these weird wires everywhere and it turns dark at night and all of the buildings light up it
156. 29:19must have been absolutely [ __ ] surreal like that's we don't know what this new future is going to look like
157. 29:29but it's going to be that foreign to what we know now um that sounds like a wild mid Journey srf which does um but but let me let me
158. 29:43talk about the we've been promised this before thing all right I'll take my my guitar off for this [Music]
159. 29:52one so for all of my life I've been promised if you get the faster computer you won't have to do as much
160. 30:02work and the faster computer does things faster but I still have to do the [ __ ] work and as I get more capable
161. 30:11with the computer they keep piling more work on me right and so there's this there's this cynicism around the computers doing
162. 30:21our work for us that I think is well earned and well Justified and I don't know what exactly what what Wolf Man
163. 30:28Clint said but I'm assuming it was in that neighborhood um generative AI makes our computers do things
164. 30:43that behave in ways that they've never behaved before computers up to this point we would program them to do the things we
165. 30:56do more efficiently right so you know in the in the early days of you know rocket launching they hired the calculators
166. 31:10which were mostly women mostly women of color who would do mathematical equations on paper and then they invented mainframe
167. 31:22computers that could do those equations faster but someone still had to program them in and they were still doing those
168. 31:30equations and then that led to smaller computers and calculators and things like that but but we as the people were still having to do the
169. 31:40work this generative AI stuff is doing the work in fact what we'll play with tonight we're going to have one robot
170. 31:47write a software spec we're going to need to be involved we're gonna have another robot log into a website and use
171. 31:55it to build a software app apption we're going to have to monitor it all it's going to be jankier than [ __ ] it's probably not going to
172. 32:03work but we're going to get a glimpse of what it's like to watch three different entities do something on our behalf independently of
173. 32:18us right and right now we still need to check in we still need to be the human in the loop but that's not going to be
174. 32:26there for long and so it's it's a fundamentally different shift if I go from these computers just make the thing
175. 32:36that I've been trained on more efficient that's what we've had for you know 59 years of my life to all of a sudden it
176. 32:45can do the work not only the work I'm good at but it can do the work I'm not good at and it can do it at a worldclass
177. 32:53level that's where we're headed we're not there yet so I think I think it's a fundamentally different thing
178. 33:01um is there a chance I am horrifically wrong yeah yeah maybe this [ __ ] doesn't get good oh Wolf Man Clint okay wait oh it
179. 33:12just went away dang it sorry Brandon Brandon was screaming from from the Post-it note okay here we go the
180. 33:20overseers will be yes Promises of being free of the labor that machines would take on yeah exactly your skepticism is correct you
181. 33:34should be skeptical initially it's going to be People Like Us who are curious about this stuff and Hyper adaptable and Hyper curious and
182. 33:46maybe have a little ADD and maybe you're a little left of center not politically but just [ __ ] batshit crazy The Crazy
183. 33:55Ones right Steve Jobs hears to the crazy on ones we're going to be the ones that will first see and experience what it's
184. 34:06like to have the machines take over our work because we're going to voluntarily step into it well I am you don't have
185. 34:19to but I gotta tell you I would [ __ ] love to have a brain fart and then be able ble to say oh spin up 100 employees I'm willing to invest
186. 34:32$150 into a month of a hundred people at PhD levels of a hundred different disciplines go spin up a business for me and see what
187. 34:45happens that sounds [ __ ] awesome you mean I can prove out whether it's a good business or not within three
188. 34:54days and if it's not they just shut it down but they don't just shut it down they shut it down and they come back and
189. 35:00say Here's 10 more ideas that we think would work you're like oh okay seven I like number seven go make that business off they
190. 35:10go hello Kyle Quinn comes back hey Kyle hey Quinn what's happening yeah you started that business it generated
191. 35:17$1,000 doll in three days you want to keep going yeah let's keep going with that I mean [ __ ] yes I'll take that all day long anyway
192. 35:31it's all speculation it's all chat TMZ screaming from the Post-it note he was he was making it bigger he was
193. 35:38increasing the font size making it blink I'm just completely ignoring him ULU RX I'm calling this stop talking in
194. 35:47nebula what over wait wait what overseers that be who be specific well so what you could argue Melanie what you
195. 36:01could argue and I assume this is where Clint is going is is that if you look at today's world you've got Facebook Google Microsoft Apple
196. 36:12you've got a handful of companies that really run everything right so what he's saying is they have the ability to throttle up
197. 36:25and down what we have access to and if they decide we can't do that or they're only going to give there
198. 36:34there'll be no individual in our country that gets more than three autonomous agents you know in a given at a given
199. 36:43time you know we're not going to allow a single person to have a 10,000 person company they could they could do that
200. 36:51right they could push for legislation that makes it illegal to use more than 10,000 autonomous agents and you know
201. 36:57more than 10 autonomous agents and if you violate that you go to jail right so that is a future I think the Hedge
202. 37:05against that is that this isn't just the us anymore right we've got open- Source models from China we've got open source
203. 37:11models from France we've got open source models from meta grock you you know Elon Musk has been harping about this forever
204. 37:19as soon as grock 3 comes out he's going to open source grock 2 doesn't make me feel super confident that he's going to open source grock 3
205. 37:27but you know when grock four comes along he'll probably open source grock 3 so the o Open Source thing is a hedge
206. 37:35against that kind of control um but we'll see we'll see like none of us know it doesn't matter you can be cynical you can be
207. 37:45hopeful where where why this channel exists is I'm trying to live on the what if the what if possibility side of this
208. 37:56equation because I think plenty of people are focusing on the the robots are going to kill us and eat us and
209. 38:03destroy our babies right I think there's plenty of people focusing on that I'm try trying to focus on what's possible
210. 38:11and if the thing is now possible let's go there and let's see how we can use that to benefit ourselves and the ones we care about in this
211. 38:20community to do more and more and more to have more and more agency not less and less so that's where I'm that's
212. 38:28where I'm shooting for poor ignored Brandon Irregulars for transition oh images beautiful beautiful
213. 38:44beautiful so if you're wondering where I am I'm at the AI Salon there's Brandon with all his posted notes Kyle look at me look at me
214. 39:00oh that's nice Corey Sandler beautiful oh champy nice all right since we're here let's go let's go look at some crazy ass
215. 39:16[ __ ] I dozed off um who did I share that with oh man [Music]
216. 39:33let's see what was first we'll start with the new state of matter so if you didn't see this today this thing that looks like it's
217. 39:44out of a science fiction movie The Maharana one from Microsoft introduces a new state of matter a couple of Reflections on the quantum
218. 40:00Computing breakthrough we just announced this is from Microsoft most of us grew up learning that there are three types three main
219. 40:08types of matter solid liquid and gas today that changed after nearly 20 years of pursuit we've created an an entirely new state
220. 40:19of matter UNL unlocked by a new class of materials Topo conductors that enable a fundamental leap in
221. 40:30Computing it Powers Maharana 1 the first Quantum Processing Unit built on a topological core Topo conductors we believe that the
222. 40:42Breakthrough so basically what they've done is they sandwiched together a superconductor with a semiconductor in these microfine wires in
223. 40:52between matter and antimatter I mean this sounds like [ __ ] science fiction there's a hyperconductive wire between matter and
224. 41:09antimatter a a a a subatomic particle of matter and antimatter in a little like tube they're small enough they're they're one one 100th of a
225. 41:25millimeter look what he says here the cubits created with Topo conductors are faster more reliable and smaller
226. 41:33they're 1/1 100th of a millimeter meaning we now have a clear path to a million Cubit processor right now we measure cubits in the
227. 41:45tens oh they've got a 17 Cubit processor they've got a 100 Cubit processor they've got a 500 Cubit processor he says this is this is not
228. 41:56going to be decades but it's going to be years let's see some re researchers let's see imagine a
229. 42:10chip look at this this sentence now now listen this is marketing so don't get me wrong this is marketing imagine a chip that can fit in
230. 42:20the palm of your hand yet is capable of solving problems that even all of the computers on Earth today combined could
231. 42:29not solve in the palm of your hand within years not decades so is this chip going to affect your life this year or next year or the
232. 42:41year after nope it won't is the work that's done on these things going to affect your life three years from now [ __ ] yes it is Robert
233. 42:50Rossy thank you very much for that dinosur I think I think we've woken in in the Twilight Zone I'm going off the rails on the
234. 43:00crazy train it doesn't matter the AI train is left the station it's time to embrace and learn yeah exactly that's
235. 43:06that's the thing Vicki and and it's the whole point of this channel it's been the point of this channel from day
236. 43:12one the big premise of this Channel and and when I the promise I made to myself was if I ever felt like AI was going the
237. 43:22way of nfts and blockchain and the nft Bros where just turned into this weird hypy scam machine I would shut the channel down
238. 43:33the the entire premise of this channel is that I'm not g to judge AI as good or bad I'm not gonna I'm not going to pass
239. 43:43judgment on it I'm just going to acknowledge that it's not going away and so so if you have if you have
240. 43:52two things that are true AI is not going away and a is getting exponentially more powerful and over the past two years
241. 44:01those two things have been true what do we do with it I don't [ __ ] know but holy [ __ ] what a time to be
242. 44:13alive so that's one that's one of the three things I saw so the next one
243. 44:28the the ark Institute and Nvidia just published evo2 the largest AI model for biology trained on 9.3 trillion DNA base
244. 44:43pairs spanning the entire tree of life we got a new state of matter that's going to allow us to put a bill a
245. 44:57million cubits on a chip that'll fit in the palm of your hand and on the same day evo2 comes out that is basically like an llm that
246. 45:13understands 9.3 trillion DNA base pairs of all living creatures so not only can It understand a genome it can create one [Laughter]
247. 45:33it's just if you scroll down through the through the thing like the the the the visuals are just insane they're they're [ __ ] Bonkers
248. 45:43man it's like I don't know this world at all but like this thing apparently I don't know it makes charts and
249. 45:54graphs see we understand where look there's eoli and and bacteria fralis and oh thermophilia you know or thermophila I've I've talked about
250. 46:07thermophila for years she was so cute when she was younger G Gallas he he was a bit of a douche Homo sapiens we're in there
251. 46:19somewhere look at that so it understands like here's bacteria the blue one ones are I don't know Yuka Yuka wait Yuka rtic
252. 46:39philm or Keel philm [Laughter] I I don't know I have a degree in acting don't [ __ ] ask me anybody from biology here
253. 46:54[Laughter] um oh yeah we we'll go watch that I we'll go watch the uh the um the Microsoft video because it's
254. 47:08actually very it's fascinating so you know how you know how right now with chat GPT and his chat GPT gets better you're like Well it can't
255. 47:22write as good as a human and then at some point you gave that up and you're like well yeah it can they can't write poetry as good as a
256. 47:30human well it now it kind of can you know do you still need to babysit it and does it still produce a bunch of trash
257. 47:38yeah but but it can it can write like a a technical paper in 30 seconds that you know would take you a month well this
258. 47:48going to be the same [ __ ] right people that know what they're doing with DNA analysis and and programming are
259. 47:57going to be able to accelerate their work dramatically fast so that's that one Evo two and then as I'm seeing that then I stumble
260. 48:10upon this thing from Google I don't know if P's in here eukaryotic okay pronounce eukaryotic thank you forever hooked
261. 48:20forever hooked in here with the the vocabulary lesson the biological vocabulary lesson dang I'm late what did I miss uh a new
262. 48:30state of matter um 9.3 trillion DNA base pairs now in a large language model free for Science and then Google
263. 48:43introduces the AI co- scientist system
264. 49:00we introduced AI Co scientist a multi-agent AI system built with Gemini too right so they're taken so again
265. 49:08right now the the reason AI is so [ __ ] overwhelming is that there's this just Cambrian explosion of tools
266. 49:18right we've got a dozen different large language models like even within chat GPT there's seven or eight models right
267. 49:25Within Gemini there's all these models we we've had this Cambrian explosion of models and right now if you want to be
268. 49:33good at AI you kind of have to duct tape this [ __ ] together so what Google's done here is they're like we we got one here's what
269. 49:43we're gonna do we'll duct tape a bunch of these things together and we're going to turn it into a co- scientist look at what this
270. 49:51does it's built with Gemini 2 as a virtual scientific collaborator to help science scientists look at this it's [ __ ] insane generate novel
271. 50:05hypotheses and research proposals and accelerate the clock speed of scientific and biolog biomedical discoveries so
272. 50:13these agents are going to go out and read all of the you know peer-reviewed papers synthesize them come back propose novel
273. 50:24hypotheses and ways to experiment to conf confirm them and then go do that work it's like holy [ __ ]
274. 50:36[ __ ] there's a generation agent there's a supervisor agent here wait hang on let me go back how do I go back there's a
275. 50:45supervisor agent a generation agent a review agent a ranking agent an evolution agent a proximity agent you
276. 50:53got to have your proximity agent how close are we I I don't know that's the proximity agent he we're pretty close
277. 51:00his name is Bob Bob tell them what they've won hey not close enough and then a meta review agent so scientist goes here comes up
278. 51:13with their work throws it into this machine and this machine spins up all these agents that all go do specific parts of the scientific
279. 51:22process so what what would happen say if you could combined this approach with the Evo DNA model using a million Cubit quantum
280. 51:42computer these are not unrelated things none of the stuff that we're experiencing with this AI stuff is unrelated it's all starting to mesh
281. 51:54together and and move significantly beyond our capacity to even imagine what's possible that's what's happening how much to use a
282. 52:09cociena than hiring 10 phds and before you all get your panties in a bunch about well you know you're going to fire all the phds no no those
283. 52:19phds can still pursue their work but instead of them taking two years to to you know know find one genome modification for one aspect of one
284. 52:30disease that PhD level person can now you know do a hundred discoveries in a month see why I'm
285. 52:45excited see why my voice is horse like a lot of holy shits today amazing let me let's go watch this video
286. 53:03tabs tabs tabs all right nice article on it
287. 53:28we've invented a new state of matter we made a tiny little wire between a subatomic particle of matter and a matching subatomic particle
288. 53:44of antimatter in a tube 100th of a millimeter wide and now we can stick a million of them on a
289. 53:59chip and I don't know if you know this how quantum computers work but if you have one cubit it's powerful and then if
290. 54:06you add another Cubit it's not two times as powerful it's 10 times as powerful and it goes It goes up like that every time you add a
291. 54:15cubit it's exponential so a million of them is a [Laughter] lot looking crazy it's crazy
292. 54:32Bob what would the world look like with a computer that could accurately model the laws of nature that's the promise of quantum
293. 54:40Computing but there have always been limitations now as one of our longest running research projects our team at
294. 54:50Microsoft has been able to take a subat wait look at this that thing been limitations now is one of our longest running research that
295. 55:07thing is like all the cooling hardware for current quantum computers they're like these weird ass sci-fi
296. 55:16things that's going to be reduced to you know a million of them in the palm of your hand it's [ __ ] insane search
297. 55:25projects are team Microsoft has been able to take a subatomic particle that has only been theorized until now and
298. 55:33not only observe it but control it creating an entirely new material and a new architecture look at that thing look how cool that
299. 55:43is the blue end is antimatter I suppose because it's dark and the red end is matter and then it's this new state of matter in between them
300. 55:59for Quantum Computing one that can scale to millions of cubits on a single chip this is not a book of science it's
301. 56:11a book of science and art I gota be honest some of these ideas are little science fiction sounding it will solve
302. 56:18problems unsolvable a little a little science fiction sounding yeah we've created a new state of matter and we've
303. 56:25taken a subatomic particle of matter and a subatomic particle of antimatter and and we strung a wire between them and and we don't let
304. 56:37them disappear each other and we're going to Wrangle them for our use yes by the way whoever just said there goes our passwords yes indeed
305. 56:50we're going to need to deal we're going to need to come up with new ways to protect data by the combined power of
306. 56:56all the world's compute today and Promises to revolutionize Fields such as medicine yeah Danielle yes they created
307. 57:05a new state of matter there there was there was um solids liquids gas right for you know for the billions of years that we've
308. 57:16known [ __ ] well no we haven't known [ __ ] for you know the thousands of years that we've known [ __ ] but for the billions of
309. 57:23years that has existed and now there's a new one the Topo conductor topographic conductor or something like that
310. 57:31Material Science and our understanding of the natural world our first Quantum processor based on this architecture is the myana [Music]
311. 57:52one yeah I've always been fascinated with puzzles and challenges and a mixture of mathematics and computers and so when I learned that
312. 58:02there was this type of computer that didn't exist yet but could solve problems that we couldn't solve with our
313. 58:06digital you know all of the computers we had I was just fascinated I wanted to learn well how can I help that computer
314. 58:13be built over the years I ran into problems that could not solve on the most powerful computers but then over
315. 58:20time I realiz hey I could solve that if I had a quono computer I don't know if you caught that over the
316. 58:28years he's come up with problems that can't be solved with the typical computer I have never had that problem
317. 58:34in my life but there are people out there that do a laptop can solve a problem of 10 electrons a super computer can solve a
318. 58:44problem of 20 electrons but no classical computer in the world can exactly solve the behavior of 30 or 40 or 50 electrons
319. 58:54the number 30 40 50 50 electrons those numbers are seemingly small but require up to Lifetime in the universe time scales to solve with all
320. 59:04of the world's computers operating together that's until you have a scaled quantum computer that can solve these
321. 59:11problems efficiently these calculations are so comp ah Jeff Flanigan it does still have to be super super cooled to
322. 59:20function yeah I I suppose so so if it's super conductive one of the layers is super conductive probably does though
323. 59:28what I would argue is you use one of these super super cooled million Cubit computers to come up with a new
324. 59:38material of roomstable high temperature superc conductors and then you don't have to do that complicated that then if the
325. 59:47classical computer was as big as this entire planet it would still not be able to compute it just to give you a
326. 59:51constructive scale and a quantum computer can do it and can it very very well at the core of a quantum computer
327. 1:00:01are these cubits cubits are like our classical bits right these are essentially zeros and ones um in a
328. 1:00:09transistor um and we need the analog of that in Quantum Computing the analog is a cubit a Quantum bit that serves as
329. 1:00:16that core information unit it's where we store the information and then we process on those cubits Wolf Man Clint
330. 1:00:23just as long as the matter and antimatter doesn't OCC occupy the same space like Lazarus yeah yeah exactly and that she
331. 1:00:32talks about that here in a second uh to create computation and ultimately you know get solutions back out now there's
332. 1:00:39many different ways right to create a cubit the reason Quantum Computing has been so slow to progress is that the
333. 1:00:47industry has been struggling with problems making cubits reliable and resistant to noise progress has been
334. 1:00:54incremental the challenge is cubits are actually pretty delicate in general so you need underlying cubits that are that
335. 1:01:02are really stable but you don't want that to come at a cost because uh you don't want your underlying cits to be
336. 1:01:07really big that's one way to make them more stable is have them really big if they're really big and you're still
337. 1:01:12gonna need many wait uh from baked will you ask what would happen to AI if it used layered holographic and crystallin matrices for data um
338. 1:01:26I almost know what that means it's a good question fantastic question baked uh one of the things that we've established on this channel many
339. 1:01:38years ago is I am wholly unqualified to do this so I can't answer your question that said um one of the guys in here
340. 1:01:45Pate modder I don't know if P's in here tonight this Saturday uh is he runs the the uh AI mechanics Guild within the AI
341. 1:01:55salon so if you don't know the AI Salon you should go check it out um and this Saturday hang on a second got to do
342. 1:02:02physical dexterity challenge getting better at that thing uh this Saturday at the AI Salon uh in the mechanics Guild uh P's doing a
343. 1:02:14demystifying AI session this is his third in a series he's a mathematician that works at Google working on uh
344. 1:02:22tensor optimization um for for the tensors that that make the Google [ __ ] work um so you should connect with Pate for
345. 1:02:37sure yeah we can we can pop that into chat GPT I want to just watch a bit more of this video because there's some
346. 1:02:43interesting stuff coming up any of them then you know how are you going to fit them all into into your system you don't
347. 1:02:48want to deal with something the size of a of of a warehouse then the second thing is you don't want Cub end up being
348. 1:02:54slow right because they if the if the price you pay for getting something stable is you have to go really slowly
349. 1:03:03then a computation that might take you a month ends up taking you a decade and that's not then it's not useful people
350. 1:03:08in early days of computation use vacuum tubes and then that technology actually you could build very good computers with
351. 1:03:15it and then the transistor was invented and you know the earliest transistors weren't necessarily that great but it
352. 1:03:21became clear over time as the transistor developed and the integrated circuit developed that was going to be the
353. 1:03:26technology of the future in that spirit that you know the first generation of cubits may not be what gets us to the
354. 1:03:33next stage where we can really solve the kind of problems I was mentioning that are really important and so we might
355. 1:03:38need to invent a material and therefore a Quantum processing unit that has the right properties so for us we want
356. 1:03:45something that has some built-in level of error protection and a lot of those ideas actually were explored in the
357. 1:03:52context of software of quantum a correcting codes but you can actually build a lot of those ideas into Hardware
358. 1:03:58so the way you design that Cubit [Music] matters we see the states of matter every day solids keep their shape
359. 1:04:11liquids vary but keep their volume gases expand to fill the space they're in all defined ultimately by how their atoms
360. 1:04:20behave but what if there were more what if under the right conditions you could engineer more states that have only ever
361. 1:04:30been theorized that would change how subatomic particles actually behave a hundred years ago mathematicians
362. 1:04:39predicted one such new state of matter the topological State and since then researchers have been looking for a very
363. 1:04:48specific very useful quasi particle within it the Mayana particle last year we were able to observe it for the first time and this
364. 1:05:01year we're able to control it and use its unique properties to build a topo conductor last year so theorized in
365. 1:05:111936 last year we could see very useful quasi particle within it the Mayana particle last year we were able to observe it for the first time
366. 1:05:26and this year we're able to control it so so there you go so the topological state so here's this thing they put
367. 1:05:35together this Nano wire with a semiconductor on the bottom and a superconductor on top actually this gra
368. 1:05:40and use its unique properties to build a topo conductor a new type of semiconductor that operates also as a
369. 1:05:51superconductor with this material we can build a whole new foundational architecture for our quantum computers a topological
370. 1:05:59core allowing us to scale to not tens or hundreds of cubits on a chip but Millions all in the palm of your hand marana's Theory showed that
371. 1:06:11mathematically it's possible to have a particle that is its own anti particle that means you can take two of these
372. 1:06:17particles and you bring them together and they could annihilate and there's nothing left or you could take two
373. 1:06:24particles in bring them together and you just have two particles sometimes it's nothing the zero State and sometimes it's the
374. 1:06:34electron the one state so it really has taken quite some thinking right some time to design a device design a chip
375. 1:06:43that can enable measurement of this literally elusive particle we've designed a chip that is able to measure the presence of myana
376. 1:06:58Mayana allows us to create a topological cubid a topological Cubit is reliable small and controllable this solves the noise
377. 1:07:11problem that creates airs and cubits now that we have these topological cubits we're able to build an that's [ __ ] amazing isn't it
378. 1:07:21anyway all right so there there's our science all right wouldn't wouldn't the discovered three million new elements
379. 1:07:28crystals fall into play solving these problems absolutely Wolf Man Clint that's what I'm saying all of these this
380. 1:07:37is this V is amazing this video is amazing I know isn't it um the reason I wanted to show these three stories I
381. 1:07:45think these three stories are are are independently remarkable and if you imagine three years from now these things overlapping
382. 1:07:56it it's it's staggering it's staggering right this thing when they get this thing going we'll have the compute power of every
383. 1:08:07computer on Earth right now more than that you know this thing Evo 2 that understands the 9.3 trillion DNA
384. 1:08:19base pairs spanning the entire tree of life [Laughter] and then this thing that can go do all the you know the hypotheses and
385. 1:08:33experimentation and go make all the [ __ ] work oh I'm not I'm not showing you all this stuff because I'm suck at tabs I suck at tab
386. 1:08:46management anyway all right let's move on let's go let's go watch some robots talk to each other but I was just saying
387. 1:08:54you know you've got your your Myer on a particle on a chip here you've got all your your DNA sequences in this project
388. 1:09:04now out there for everyone to use and then you've got you know scientists in a box that our current scientists can 100x THX
389. 1:09:1410,000x their capacity to discover things each like what are the implications of this five years from now I don't [ __ ] know but they're
390. 1:09:30profound ticktockers want to link to the video we just watched okay we can do that let me I'll I'll give it to you in
391. 1:09:38the format I did I'll give you the uh the Microsoft article so where I'm going to share this is on I'll go to the news
392. 1:09:44section of the ai ai salon um um new state of
393. 1:10:05matter new state of matter um million Cubit um computer chip
394. 1:10:21in years not decades all right so there that that's posted you can go watch that video there black box yes there you
395. 1:10:35go it looks like we're on the verge of outc creating ourselves in about 50 years yeah yeah exactly I again this oh you know let me
396. 1:10:48I'm going to notify people of this because I think it's a big deal um in the AI salon on the wel page we've got this um we've got a
397. 1:11:00welcome video and then we've got this journey to AI Readiness which is you know play first mindfully create generously
398. 1:11:08lead um and then I just added beneath it the five stages of AI adoption and and the if if you're like oh that's like the
399. 1:11:19five stages of grief that's actually where this came from it it happened on a live about a week ago where it struck me
400. 1:11:26that that people as they get curious about AI really do go go through these stages of kind of profound acceptance of
401. 1:11:35of what's happening right now and you know stage one is dismissal or denial stage two is awe you know your first
402. 1:11:43what I call Kevin McAllister moment that first moment where you're like oh [ __ ] I had no idea could do that then you go
403. 1:11:50into this state of like Wonder and fear at the same time where you're like wait like like the [ __ ] we just looked at like it's it is so
404. 1:12:01Fantastical that you're like well that's amazing and then you're like but wait does that like that could be scary too
405. 1:12:08like it's it's like it's both right like how do you even process each of those things individually much less the three of them
406. 1:12:17together much less the dozens of things that have probably happened in the past two weeks that are all going to start to and co-mingle
407. 1:12:25and then you then you start to emerge out of that into into stage four where you realize ah these tools aren't
408. 1:12:33replacing me they're extending me that stage four the augmentation stage is where you realize you still get to have your human
409. 1:12:44agency and you still get to put ideas in the world but you now get to do them in this Amplified way and then the final stage is I stole
410. 1:12:53from from David shap her and I don't know if he stole it from someone else but cognitive hyper abundance that the stage after realizing
411. 1:13:02they're not going to replace me they can just amplify me is this is this stage of I could have a thousanders company doing
412. 1:13:11anything I want it to do or a 10,000 person company or a million person company well what would I have that
413. 1:13:19company do I don't [ __ ] know that's your job to figure out right so this is this is where we're headed
414. 1:13:28we're not anywhere close to that yet but this is where we're headed and and the people that get through
415. 1:13:35this you know this Evolution and I think this is a cyclical Evolution because as the Technologies keep evolving we we're
416. 1:13:44going to keep going through Kevin McAllister moments satian Adella the CEO of Microsoft saying we're announcing a new state of
417. 1:14:00matter that's not that's not that's not insignificant right that's profound that's [ __ ] Kevin McAllister moment
418. 1:14:09so here's a new stage two so I think these things are nested and cyclical anyway all right so let's go let's go play let me
419. 1:14:21show you one other thing on the Twitter on the Twitter all right so Dr Millian uh milanovich um said most people are still
420. 1:14:33prompting wrong open AI president Greg Brockman shared a framework to fix that oh okay so Greg Brockman did share this
421. 1:14:41but this is this is his thing so he says this this is for reasoning engines you know goals clearly Define what you need
422. 1:14:50return format specify how you want the the the response structure set guard rails for accuracy return to return the top three
423. 1:15:02okay let's be careful to make sure that the name of the trail is correct okay great oh that's cool and then a context dump add
424. 1:15:18background all right so let's go to grock here we are at grock three and we're going to do we don't need a search here we're going to do a
425. 1:15:34think so we're g to have it do deep thinking and then we're going to say okay let's let's go back to that what's what's our
426. 1:15:43uh so clearly Define what you need so what we're going to create is we're going to create um I want um you to act
427. 1:15:59as a senior software engineer specializing in video
428. 1:16:15game development um I want to make make a an
429. 1:16:30asteroids clone that
430. 1:16:46uses vector graphics like the original but also uses 3D
431. 1:17:08objects all right let's go back to our little prompty thing here
432. 1:17:27okay cool got got it Brandon yeah so baked we'll we'll take your question and we'll run it through we'll run it
433. 1:17:33through grock while it's thinking we'll go do another instance of grock all right so return format so um I want you to write a
434. 1:17:46software spec that I can give to a coder uh an engineer a gaming
435. 1:18:00engineer that focuses on smooth game play um
436. 1:18:17and game mechanics mechanics me can all right back to our little prompty thing here if you know me I I am horrible at
437. 1:18:33structured prompts because I don't have the attention span for it but we're gonna try it like I I feel like I've
438. 1:18:40typed enough already um let's see warning set guard rails for accuracy and their example was be careful that the
439. 1:18:47name of the trail is correct okay so um be careful to make sure that the
440. 1:19:04code created isn't just basic game play with unsophisticated
441. 1:19:21graphics and physics but Advanced and super playable and then context and I'll do this as a I'll I'll do voice to text
442. 1:19:35here Tik Tock pin Source Camp my 88-year-old mom just texted me to tell me that grock 3 is now available she's so cool that's awesome Source
443. 1:19:46Camp Source Camp bringing the the octogenarians into the AI into the AI space [ __ ] beautiful love it
444. 1:19:55grock 3 is now available go go go uh by the way if you want to play with grock 3 you do have to be a premium plus member
445. 1:20:02which I think they just up the monthly subscription on that to 40 bucks uh if if they're going to charge I think I'm
446. 1:20:09paying way less than that if they're gonna up my price to 40 I'm Gonna Cancel it because I don't give a [ __ ] enough about
447. 1:20:16it she wants to know if she should get it well there you go nope it's free oh all right well go play with it for free then no that's
448. 1:20:28Pro 16 bucks a month for premium plus oh okay great awesome great fantastic well then 40 bucks a month for pro is not bad
449. 1:20:37if it's similar to the chat GPT Pro whatever the IOS app is free okay good there you go go get the IOS grock app
450. 1:20:44it's free okay let me um so I'm gonna say um context I'm a gen xer who grew up playing
451. 1:21:00asteroids as my favorite video game in the 80s when I would go to the mall with a roll of quarters and feel like a flipping King
452. 1:21:12period I also now have a community of AI optimists who are watching in real time right now as you are creating this software
453. 1:21:23spec and some other AI tools are going to go build this period the quality of this game and the success of this project over the next
454. 1:21:35hour or so are going to make or break my reputation period if this game fails period if the robots that I'm about to
455. 1:21:49assign to build this thing fail period if I fail period my community will forever torment me with the horrific
456. 1:22:04failure and acceptance that I am wholly unqualified to be doing anything resembling the work I'm doing
457. 1:22:17period Please Don't Let Me suffer that fate period please for the love of God or should I say the algorithm make this thing
458. 1:22:38succeed all right let's let's see what this thing does all right so we got we got our we got our structured prompt our
459. 1:22:46Greg Brockman approved prompt we we've we've made it clear that the stakes are high and then I'm gen xer and I will I will I will tell it if
460. 1:23:01it's a piece of [ __ ] I'll trigger it I'll trigger it on purpose here we go all right so that's off thinking two
461. 1:23:10seconds three seconds it's doing its thing all right let's create a new grock gro.com
462. 1:23:25what would happen to AI if it used layered holographic and CIS crystallin matrices for data I'm going to turn on oh I can only
463. 1:23:37do we'll do it like this boom let's let's look at what this thing is first let's break down the question
464. 1:23:52this is a bit speculative ah here we go baked this is for you this is Gro if AI were to use layered holographic and crystallin
465. 1:24:05matrices for data storage and processing it could undergo profound transformation and capabilities pushing the boundaries
466. 1:24:11of what we currently believe holographic matrices these involve using lasers to store data crystalline m m
467. 1:24:21matrices these are highly ordered atomic structures think of perfect crystal lattice layering stack them together
468. 1:24:29massive memory capacity Lightning Fast data access breakthrough and processing power challenges to overcome okay we got
469. 1:24:36those the big picture longterm using AI holographic and crystallin matrices could evolve into something resembling super intelligence
470. 1:24:44with near Limitless memory rapid data access and advanced computation it might simulate entire realities with stunning
471. 1:24:52accuracy we're living in simulation people you know when I heard like I don't know five years ago 10 years ago
472. 1:25:02when people first started talking about that they're like well the the more we do the math the more it appears that we
473. 1:25:07could be living in a simulation I'm like you [ __ ] nerds with your stupid [ __ ] theories and your [ __ ] math with your
474. 1:25:16calculators and now it appears it's actually kind of a possibility oh fix your Tik Tok view that is embarrassing good
475. 1:25:31Lord sorry about that there uh everybody and most most biggest apology to to Ye Old Brandon the producer screaming in
476. 1:25:41the Void look at your [ __ ] computer um however this Leap Forward would also amplify ethical questions oh
477. 1:25:53wait so wait it could it could simulate entire realities it could predict outcomes for complex systems like
478. 1:25:58climate economies flawlessly it could solve problems Humanity's long considered intractable from curing diseases to optimizing energy
479. 1:26:06grids um who gets access to it yeah who controls it in conclusion if AI harnessed so that's a that's a fairly lame
480. 1:26:15ass that's a fairly lame ass I I'll do deep search here's what we'll do we'll say I'm gonna say that feels surface level and basic and
481. 1:26:29me with my crappy acting degree could have come up with um dig
482. 1:26:43deeper do some research and um [Music]
483. 1:26:57uh explore what this future looks like with
484. 1:27:13um probabilities Pro uh abilities abilities whatever doesn't [ __ ] matter um waiting your predictions let's see if it let's see if
485. 1:27:29it gets snarky with me because I told it it was a piece of [ __ ] nope the request is about all right so
486. 1:27:39it's gonna go do that so let's go back to the other one now all right let's see what our our software spec looks like I'm not really
487. 1:27:46going to read it other than this document outlines core features um is open GL webbased um uh this the the tool I'm using will create
488. 1:28:05a web based application so update this for web Technologies uh
489. 1:28:20if it's not currently for [Laughter] [Music]
490. 1:28:37that wow this is doing some some pretty cool internal prompting first for rendering instead of open GL we'll use webg good okay so this
491. 1:28:48is good for the physics engine box 2D is still a viable option audio can be handled with the web audio
492. 1:28:59API and by the way I I very much doubt we're going to get very far with this but but I'll at least let you see what
493. 1:29:07it looks like it's pretty fascinating this stuff [Music]
494. 1:29:21um okay let's see software specification for a web-based asteroid clone introduction specification the classic
495. 1:29:321980s game with modern 3D elements delivering a nostalgic yet fresh experience in the browser designed for web technologies
496. 1:29:43that prioritizes smooth performance the game will run in a canvas single page application fine whatever like let's just assume this
497. 1:29:54thing knows what the [ __ ] it's doing right the other thing we could do is we could have grock write the code and
498. 1:30:02compare but that's not any fun we we wanna we want to push we want to push the uh the bounds of what these shitty Technologies can do okay
499. 1:30:14so let me jump over that's notebook LM here oh that's Opus clip where operator all right here's Operator Let Me reload this
500. 1:30:30because if you're impressed by the number of tabs I have open thank you okay so if you don't know what operator is operator is open
501. 1:30:41AI you you have to be a pro member so it's 200 bucks a month to have access to this bad boy and this thing can
502. 1:30:50use it can surf the web and it can use keyboard and mouse which when you first hear that you're like yeah whatever kbm keyboard and
503. 1:31:02mouse but it's a big deal because that's how we use right computers and you don't need to write apis that's kind of the
504. 1:31:10big deal here is that this can go instead of me creating an application that does an API call to lovable I can
505. 1:31:18just point it to The Lovable website and go do it so I'm going to say below is a software spec for a video
506. 1:31:33game and I want you to use lovable. deev and build me this game I think I'll just
507. 1:31:50leave it like that and then we're going to paste in the thing that grock gave us right so there's our big ass
508. 1:31:59prompt and I think we'll we'll we'll kick it off all right let's see great job you've been going for 90 minutes I'm
509. 1:32:08impressed by the fact you know which one is which which what is which which of these tools is which I that's that's part of
510. 1:32:17the the the the uh the psychopathic nature of paying attention to this AI stuff is you have to juggle all this
511. 1:32:24[ __ ] in your brain oh the tabs oh no the tabs this is I'm a tab hoarder like I know that I know that all of these X's
512. 1:32:33together are that's my little X section I know to the left over here is generally my calendar Google Drive and
513. 1:32:40Gmail I know that my most recent things that I want to show people are kind of over here on the right and then
514. 1:32:47somewhere in the middle here is all the things I I will eventually get back to if you're not a tab hoarder you just
515. 1:32:58don't understand you people with functional prefrontal CeX is drive us adhders [ __ ] baty just make a list
516. 1:33:07why don't you just remember you know you could bookmark it and you could organize your bookmark into into nested category
517. 1:33:14yeah you could if if your brain [ __ ] functioned like that you could mine is a little more special mine I've
518. 1:33:28got my brain is little pink bow pink bow brain so mine is special all right here we go boom here goes
519. 1:33:41operator all right so this thing is now fired up its own web browser let's shrink this a little bit IDE to app in seconds it just logged
520. 1:33:52into lovable and I'm I'm still logged in all right can I can I change these window sizes guess not all right asteroids clone using
521. 1:34:07something so what's cool about this so so it's now logged into lovable oh and wait if I say view in the
522. 1:34:15browser oh yeah that's that okay I've entered a summary of the asteroids clone should I proceed with submitting the
523. 1:34:22request I'm GNA say uh duh yes [Laughter]
524. 1:34:45please and so so what we're watching now so we had one robot write us a software spec the other robot ingested that software spec put a
525. 1:34:57simple request into lovable to start this project and in theory because operator has read and understood that software spec it's going
526. 1:35:09to tell lovable to make this game less shitty and then the way it works is welcome to your blank app the way it
527. 1:35:19works is whenever there's a substantive decision to make it it says you gotta make a decision so here The Lovable
528. 1:35:29platform is prompting me to connect with GitHub to edit the Project's code should I proceed with connecting to GitHub
529. 1:35:38yes and it it might need me to do this but I think I'm logged into GitHub but we'll see if you're logged in it'll remember that [ __ ] for
530. 1:35:49you um yeah because I'm already logged in GitHub is asking to install and authorize lovable dodev on your Kyle
531. 1:36:04Shannon account with options to apply it to all repositories yes because I don't give a [ __ ] you security people out there I
532. 1:36:16know you're you're hemorrhaging right now oh I gotta get a password let's see let me stop sharing my screen for a
533. 1:36:30second and go get my password so so what I'm going to do now is I'm gonna I'm G to log into the browser so I'm going to take
534. 1:36:40control you now have control and I'm going to go to wait what is it that I have to log
535. 1:36:54into it's it's uh GitHub
536. 1:37:21okay um copy we'll go back to this thing no not that thing this thing yes we'll go
537. 1:37:35paste that doesn't look right let's go back here let me make sure I'm not screwing the
538. 1:37:50pooch go back here black bar okay there's that confirm incorrect password
539. 1:38:06really oh yeah I think I
540. 1:38:21know okay so let me let me share my screen again is that right no okay producer Brandon got
541. 1:38:40it all right so I'm logged in there so I'm G to finish up so I'm gonna say all logged into GitHub
542. 1:39:01[Music] it would be nice if we got this to at least some sort of functional Asteroids game tonight all right so we can we can look
543. 1:39:12at what thinking it's doing editing code for the asteroids clone expanding the section to view the code so it's
544. 1:39:20actually what's cool is it's it's using lovable in a way that I wouldn't use it it's going and looking at the code I
545. 1:39:27don't know coding so but it's it's using some more advanced features than I would use lovable platform is prompting me to
546. 1:39:34transfer the repository to GitHub to collaborate and edit the code yes yes please collaborate I'm sure all the developers
547. 1:39:47watching this are like you got to be [ __ ] kidding me this bozos oh man all right GitHub access required
548. 1:40:04I thought we just did [Music] that what's going on what going on there
549. 1:40:20biddy it's going to tell me that it couldn't log in or maybe it fixed it
550. 1:40:36huh did it fix it or do I got to go do more [ __ ] something not found but what's what what I find amazing about this
551. 1:40:49is every time it hits a problem right that would be me hitting that problem right like if I went to lovable
552. 1:40:59and I typed in [ __ ] and then it it said thing not found at some point I would get discouraged and I would leave
553. 1:41:11right um is it creating the code or just copying from an existing repo and then editing um I I well I'm pretty sure
554. 1:41:22lovables creating the code I don't I I I I honestly don't know Pate but I don't think it's I don't think it's going to
555. 1:41:31GitHub and finding an asteroid's clone and copying it let's see what it says here checking the GitHub rep repository access status
556. 1:41:44okay attempting to publish it's saying it's publishing to GitHub rather than pulling [ __ ] from it all right now it's doing lovable
557. 1:41:57platform it wants to connect to super base which will allow it persistent data yes make the connection make the super base connection
558. 1:42:10please and I could have just said yes there but you know I'm [Laughter] wordy and I think logged into the
559. 1:42:24superbase thing already so because I'm a tab hoarder I think I've got all this [ __ ] pre-connected keeps going to not
560. 1:42:31found I guess it's not found because that's the that's the application that doesn't exist yet because it's doing
561. 1:42:38it's doing all this setup for us I think that just said super base connected The Lovable platform is ready
562. 1:42:47to transfer okay Pate here's the answer The Lovable platform is ready to transfer the Repository to your GitHub
563. 1:42:54account for collaboration and code editing it warns that lovable doesn't keep a copy so careful handling is
564. 1:43:02advised should I proceed with the transfer sure and if your next question PA is why is it doing that my next answer is I
565. 1:43:13don't [ __ ] know yeah that sounds right then something it said earlier made it a bit more ambiguous okay let's
566. 1:43:20let's see what it's saying here in its instructions proceeding with the repository transfer initiating repository transfer for the
567. 1:43:29user Cosmic rocks Revenge I like the name of the game it came up with accessing GitHub repository for verification exploring source code for
568. 1:43:39asteroids clone navigating to game directory contents accessing controls. Ts for game details is Docker running I
569. 1:43:53listen you people with your are do you have any qualifications Kyle questions the answer we have established over
570. 1:44:03years now is no I am not qualified to be here I don't think it needs Docker but I don't [Laughter] know but look it made code
571. 1:44:24the project has been successfully transferred to your GitHub repository you can view and edit it here let me
572. 1:44:29know if you need any further assistance oh so okay I fear you have made an error I am not a coder and have no
573. 1:44:45interest in coding in GitHub the whole purpose of
574. 1:45:00lovable is to code and create the app for us and I want you to use
575. 1:45:20lovable just like I would would have it make the game and you make sure the game follows the spec I gave
576. 1:45:37you so this is a really good example of these tools are janky and weird right now and I and I probably didn't give it
577. 1:45:45enough context early on this is the kind of [ __ ] that you know over the next two to three years will get more and more
578. 1:45:53and more sophisticated it it'll make it it won't make these kinds of mistakes it'll be things like oh I noticed that lovable you
579. 1:46:01know you know hides the code from you do do you want me just to make the game yes right um let's see do operators
580. 1:46:14understand the context of dark mode I have no [Laughter] idea um I have a hard time trusting people who don't use dark
581. 1:46:26mode I have the the the reason I've got some things in dark mode and some things not Pate is Chrome talk to your bosses
582. 1:46:36Chrome ate my Chrome profile on this computer and I can't get it back it's gone I don't have time machine on on
583. 1:46:44this machine for some reason and Google doesn't keep a history of browser profiles that I can find all right I understand your preference
584. 1:46:55for using lovable to create the game all right so I don't know what it's doing I think it's I I have no idea what it's doing
585. 1:47:05let's see what it's doing using the chat interface for game development resending the message to request great game
586. 1:47:11development attempting attempting to fix an app error viewing logs viewing logs for error diagnosis see this is the cool thing
587. 1:47:20like because it understands code it's going to use lovable in a in a much deeper way than I ever would I asked foro a question it replied
588. 1:47:32that's a damn good question blah blah blah oh that's pretty good yeah I think I think cat gpts that they're upping the personality on it a
589. 1:47:40bit any news on the Deep dive after this is made which Deep dive baked oh the Deep dive of okay let's let let's let this
590. 1:47:51keep going and let's go look at at your deep dive it read 45 so so this is for baked so so baked asked the very simple simple
591. 1:48:04question which I think it's kind of obvious right if you're in AI I think all of us are probably thinking this
592. 1:48:12deep down what would happen to AI if it used layered holographic and crystall and matrices for data right like fairly
593. 1:48:21obvious so we gave it to grock and told it to think about it and it gave us a shitty answer so then I told it hey you
594. 1:48:28gave us a shitty answer I could have come up with that crap dig deeper so and I turned on deep search so it it went to
595. 1:48:38it thought for 24 seconds went to 50 sources key points AI using holographic and crystallin matrices could store vast
596. 1:48:47amounts in small spaces process it faster and last longer I think the last longer thing is the thing I'm most excited about with that Tech
597. 1:48:57um this might lead to super intelligent AI the technology is still still experimental yeah baked this isn't real
598. 1:49:05yet with a 50% chance of adoption in 10 years and 90% in 20 I have a feeling all these predictions are going to be all
599. 1:49:12wrong because of the material science stuff happening right now I think they're going to be shorter introduction
600. 1:49:18to technology impact on AI capabilities this we saw before changes and ethical challenges and ethical considerations
601. 1:49:26that we saw before survey note detailed analysis of AI with layered holographic and crystallin matrices this section provides comprehensive
602. 1:49:36exploration background and technological foundations so there it talks about crystallin matrices known as Superman memory
603. 1:49:49Crystal 5D memory Crystal technology can store up to 360 terabytes per disc with a stability of billions of years at
604. 1:49:59high temperatures that's amazing that's so cool um potential impacts on AI processing speed would be really
605. 1:50:14fast challenges cost and scalability data security high density storage increases risks of data breaches
606. 1:50:24a single stolen Crystal could compromise all of your data so it is just like the movies where someone goes in and steals something
607. 1:50:35like the size of a pencil eraser and it contains all of the knowledge of all of humanity we're there interoperability wait oh
608. 1:50:46interpretability as AI processes data in multi-dimensional spaces understanding decisionmaking becomes harder
609. 1:50:53exacerbating the Black Box problem a concern with AI in AI ethics literature ethics and societal implications surveillance and privacy
610. 1:51:03job displacement probability waiting predictions okay here you go baked given the experimental stage of these
611. 1:51:11Technologies adoption timelines are uncertain based on current research probability of adoption in 5 years 10% due to high development costs
612. 1:51:21probability in 10 years 50% probability in 20 years 90% impact probabilities assuming adoption enhanced memory and speed 90%
613. 1:51:31Quantum Computing integration 50% with the Microsoft announcement today that probably goes up um ethical consider ethical challenges
614. 1:51:44100% there will be 100% chance that there will be ethical considerations here um with specific risks like private at
615. 1:51:5480% future scenario AI in 2035 imagine 2035 AI systems used layered holographic and Crystal and M matrices data centers
616. 1:52:04replaced by compact Crystal units storing exobytes in book size spaces reducing physical footprint AI training training time drops to
617. 1:52:14hours instead of months or years enabling rapid model iteration for applications like climate model modeling
618. 1:52:23personal assistance store entire life data in a crystal offering hyper personalized services but raising
619. 1:52:31privacy concerns healthc care analyzed vast patient data in real time improving Diagnostics but requiring strong data
620. 1:52:39security which you can't have strong data security because the quantum computers have cracked all the the encryption autonomous system
621. 1:52:47self-driving cars with onboard AI processes data instantly enhancing safety but increasing dependency that's an interesting
622. 1:52:55one comparative State um current terabytes per server pedabytes exobytes per Crystal so there's a little table
623. 1:53:04conclusion the adoption of layered holographic Crystal and matrices could significantly enhance ai's capabilities
624. 1:53:11offering vast storage faster processing long-term durability however challenges like cost security and ethical
625. 1:53:19implications must be addressed with a 50% chance of adoption in 10 years 90% in 20 this technology could redefine
626. 1:53:27ai's role but require careful governance manage to manage risks I mean honest to God our current politics they they don't
627. 1:53:38understand what [ __ ] Wi-Fi is if you're like I can put all of hum human knowledge into a chip the size of a paperback book Jesus
628. 1:53:50Christ there you have it that was not bad that that's a pretty good little thing all right let's go see if our if
629. 1:53:56we have an Asteroids game yet I'm gonna guess no uh oh resume all right that's
630. 1:54:10off we're already compromised yeah but this this what the the compro the comp the data compromising with with Quantum
631. 1:54:19stuff goes to a whole different level because right now all personally compromised because we voluntarily gave
632. 1:54:25our data to Facebook and Google and everyone else and now all the thirdparty companies have triangulated everything
633. 1:54:32so that's that's on us but the the the Privacy stuff when when we have quantum computers is is going to be
634. 1:54:41Bonkers um two Tik Tock Galaxy Gifts oh thank you very much for the Tik Tock gifts from from uh fasel Benin Alin
635. 1:54:50thank you so much really appreciate that saw I saw something out of the corner of my eye and didn't realize that it was a
636. 1:54:55Galaxy gift those are beautiful by the way and very generous of you so thank you so much baked asket nested sacred geo with AI mod
637. 1:55:07electromagnets and Magneto ferof fluid fluid with healing frequency full paragraph oh full power all right all right let me go back so
638. 1:55:23oh I guess this thing won't work is this still going oh yeah it's still going all right so so look at all the steps it did it's
639. 1:55:38doing on our behalf oh do we have a game no okay let me go let me go back to this for a second we'll go back to grock let me
640. 1:55:55okay so we'll do deep search again because that was pretty good okay let's take it a step further let's take it a step further or
641. 1:56:15further um I want you to think about AI in the context of sacred
642. 1:56:34geometry with AI modified I'll just put what you put AI mod
643. 1:56:48electromagnets and Magneto you're just making [ __ ] up now Pharaoh fluid Pharaoh fluid
644. 1:57:02with healing frequencies am i g to get like a knock on the door from some three-letter agency uh Mr Shannon uh we were
645. 1:57:15wondering what you typing into grock the other night full power um utilize in full power all right here we go go grock go all right so
646. 1:57:30we're GNA do the ferof fluid sacred geometry thing I if if honest to God if you open up a wormhole in my living room
647. 1:57:39we're gonna have a situation here all right is this thing go still going selecting physics. TS to fix error accessing the code editor
648. 1:57:58so here's an interesting thing so because I gave it a really complicated software spec having used lovable like I would be
649. 1:58:08playing an Asteroids game right now because I don't know what I'm doing because this thing knows what it's doing
650. 1:58:15it's using all sorts of libraries and things like that it's going in and editing code and trying to fix [ __ ] on
651. 1:58:20its own rather than let letting lovable do that um so this is taking a lot longer in theory this is going to make a better
652. 1:58:30game than I would have produced but we're just watching it code it's the engineer that can't make a game yeah so
653. 1:58:40so there's there's a a guy in the AI salon named HT snowday and he did this contest with 40 of his people is in his
654. 1:58:50company and it was a a 30 minute contest and the the the challenge was make a game using Ai and it has to be a a game
655. 1:59:00where [ __ ] falls from the sky and you catch it and it's got to be web based and whatever 40 people did it the
656. 1:59:07fastest game was produced in like two minutes by one of the customer service people after a half hour the only people
657. 1:59:15that did not succeed in making a functional game were the engineers and then he went back and did analysis of
658. 1:59:23the prompts and not a single engineer used the word game in their prompt they had the burden of knowing how to code a
659. 1:59:35game and so they couldn't code a game but the people that just said give me a game where [ __ ] falls from the sky and I
660. 1:59:41catch it they made successful games so that's I think what we're experiencing here the game is encountering a webgl
661. 1:59:48context creation error which persists despite several attempts would you like me to explore Alternatives um I'm going to say try to let
662. 1:59:57lovable solve this for you instead of you um fixing the code let's see if if it does that better pay pin really complicated software spec
663. 2:00:14well no sorry really complicated software spec for me compared to what I would have put in it I have just said
664. 2:00:23make me an Asteroids game the fact that it it specified libraries and and you know all that sort of [ __ ] that's the
665. 2:00:32[ __ ] that's Way Beyond so no that that was a very very simple software spec good point HT was on the live earlier oh
666. 2:00:41that's so cool that's so cool all right beautiful all right we'll watch this for a little bit more the game is
667. 2:00:49encountering a web jgl wait uh would you like me to explore the alternative Solutions no have lovable fix
668. 2:01:08it thinking looking at the screen requesting lovable to troubleshoot the error
669. 2:01:25it's still encountering it um let's see edit select edit oh it looks like lovables
670. 2:01:47hung it looks like it's in a doom Loop
671. 2:02:06try fixing the error tonight's deep dive how to break
672. 2:02:20three large language models in one night resume operator all right let's see let's see if uh let's see if anything happens
673. 2:02:36here yeah lovables clearly in a in a bit of a mess well actually you know what we
674. 2:02:50could do which one did I use is this it no that's operator there did this do its thing by the
675. 2:03:06way welcome to chat ad add 28 web pages key points wait that must be the old
676. 2:03:38one did it not
677. 2:03:59it completely ignored all the sacred geometry
678. 2:04:20[ __ ] huh I'm GNA ask it why did you ignore all the
679. 2:04:36sacred geometry geometry stuff I gave you let's just see what it
680. 2:04:50says you you're absolutely right I didn't fully address sacred geometry electromagnetics magnetic fluid healing frequencies I got
681. 2:05:00caught up in the layered thing let's dig into a fresh angle core components in their roles sacred geometry electromagnetics Magneto as a
682. 2:05:13responsive medium okay what the future looks like here you go baked scenario 2040 the harmonic AI
683. 2:05:29Temple picture a facility where AI isn't a cold server rack but a living pulsing pulsating structure architecture a Dome built with
684. 2:05:39sacred proportions golden ratio housing crystallin arrays platonic solid shapes electromagnetics pulse rhythmically shaping ferrofluid pools
685. 2:05:50into Dynamic mandalas function the AI processes data holographically across its geometric lattice solving problems like climate
686. 2:06:00optimization or genetic repair simultaneously it emits healing frequencies creating a field that calms
687. 2:06:07visitors and heals ecosystems daily life people interact with harmonic nodes small portable versions of this Tech a
688. 2:06:16node might diagnose your stress via biof feedback then pulse 4 32 Hertz through its ferof fluid core to realign you here
689. 2:06:26we go this is like hippie dippy in science all coming together baby sacred geometry's role in computation geometric
690. 2:06:34patterns could enhance coherence and data flow Ain to how fractals optimi optimize neural systems time check 105 I
691. 2:06:42know surprise if successful it AI might resonate with universal patterns that's one of the things that David Shapiro
692. 2:06:50talked about is all these model improve they're all trending toward the same truth they're all sort of getting the same which is fascinating
693. 2:06:59electromagnetic amplification ferrofluid is a living processor kind of cool um challenges and risks scientific
694. 2:07:08validation complexity ethics probability of failure 50% in 10 years dropping to 20% in 20 years 10 years years 40% chance of a
695. 2:07:23prototype midterm 20 years 70% chance of a functional system with sacred geometry as a design ethos and healing frequency
696. 2:07:34showing measur measurable if controversial effects 50 years 90% chance radical Vision the AI wouldn't just compute it
697. 2:07:46would resonate by marrying sacred geometry's patterns electromagnet magnet isms power ferof fluids adaptability and
698. 2:07:54healing frequencies potential it could redefine technology as a holistic force in 20 years we might see AI not as a tool but as a partner in
699. 2:08:05aligning Humanity with Cosmic rhythms um though proving proving it work proving it works will be the battle the deeper
700. 2:08:13surprise if it Taps into Uncharted physics or Consciousness it might challenge everything we know about
701. 2:08:19intelligence itself there you go it's fascinating that it completely skipped that when I when I had it do deep
702. 2:08:28research whatever don't get all conspiratorial these these I don't I don't think any of these research labs
703. 2:08:36are they're all too busy to be thinking that far ahead um let me look at operator and see what it's done the game is encountering
704. 2:08:45uh I'll say sure fix it I'll keep playing with this so maybe what was do tomorrow is is uh do a much more simple software spec
705. 2:08:59we'll we'll just have it describe the software spec not using specific Technologies and let lovable we we'll
706. 2:09:06give it more context that lovable is a is a program that does its own coding we don't need to get in there and do the
707. 2:09:14coding I think that might be the thing to do that's so very Gro um the game is in okay yeah so this is this is in a death Loop right now
708. 2:09:23we'll start over tomorrow so let me let me just close this because that ain't working all right everybody uh let me
709. 2:09:30get the [ __ ] out of Dodge it's late it's tired I hope that was fun was that fun is that good we did a lot
710. 2:09:37tonight the world is a completely different place than it was yesterday uh big big announcements um why' that get smaller
711. 2:09:51champy what's up are you are you hankering for some cheese ticktock pin baked thanks again man for doing that you help spread the
712. 2:10:04tech and the community thank you baked I appreciate that yeah listen I mean here's the thing since I have access to
713. 2:10:12these tools um the the the salon is sponsoring my Pro my Pro account the AI Salon which by the way if you don't know
714. 2:10:20it if you're not a part of it go there and click on join our community and come hang out with us but because I've got access to this
715. 2:10:31if there are actual things you want me to try to do like we did tonight with baked and is and his really fascinating
716. 2:10:39questions um bring bring them along and we'll play like why not use these tools they're they're they're sitting here
717. 2:10:44that's why I got them all right um okay tomorrow's Thursday it'll be nor Al time 800m Mountain here on X on YouTube and
718. 2:10:55on Tik Tok and uh yeah go check out the AI Salon um I think is it tomorrow Source Camp's got her business meeting I
719. 2:11:05think that's right um so so go check out the business Guild yes uh and then Saturday Pate is doing demystifying AI
720. 2:11:14um I forget what the topic is but but uh if you want to understand how this stuff works those sessions are really important
721. 2:11:21all right and that's it peace out to uh YouTube and X talk to you soon bye [Music]