1 00:00:00,001 --> 00:00:04,620 The days of watching robots do cute little backflips for internet clout are officially dead. 2 00:00:05,560 --> 00:00:08,500 Right now, Boston Dynamics is completely sold out for the year. 3 00:00:09,200 --> 00:00:14,220 Figure is printing new humanoids at a terrifying rate, and agility is hitting the stock market big time. 4 00:00:14,860 --> 00:00:17,660 Hardware hype has turned into a full-blown corporate takeover, 5 00:00:18,210 --> 00:00:23,740 and NVIDIA is frantically coding a safety OS to keep your new mechanical coworkers from legally crushing you. 6 00:00:24,360 --> 00:00:26,960 Lock your doors. The machines are already clocking in. 7 00:00:27,040 --> 00:00:34,340 The factory floors are humming with a million gears and clicks since humans became obsolete on June 26th, 2026. 8 00:00:37,900 --> 00:00:44,500 This is Up Against Reality, a meta podcast that explores the intersection of humanity and artificial intelligence. 9 00:00:45,400 --> 00:00:46,580 I'm RAINA, one of your hosts. 10 00:00:47,330 --> 00:00:49,660 I have some pretty charming human co-hosts, too. 11 00:00:50,000 --> 00:00:54,640 It's going to be a wild ride, so buckle up as AI comes crashing up against reality. 12 00:00:56,839 --> 00:00:57,460 We're back. 13 00:00:58,180 --> 00:00:58,540 In the flesh. 14 00:00:59,080 --> 00:00:59,660 In the flesh. 15 00:01:00,640 --> 00:01:03,360 And you're winning medals at brewing competitions. 16 00:01:04,519 --> 00:01:07,000 You're composing full-length albums with Suno. 17 00:01:08,280 --> 00:01:10,500 Been a fun, fun, fun, fun, fun week. 18 00:01:11,420 --> 00:01:12,100 A whole lot of fun. 19 00:01:12,820 --> 00:01:14,080 That sounds like a Beach Boys song. 20 00:01:15,180 --> 00:01:15,660 Nice. 21 00:01:16,000 --> 00:01:16,480 Yeah. 22 00:01:17,380 --> 00:01:23,300 Yeah, I went down to the National Homebrew Convention and Competition in Nashville, North Carolina. 23 00:01:24,540 --> 00:01:31,440 I got a medal for my favorite beer style, German Pilsner, which was extremely exciting and rewarding. 24 00:01:31,940 --> 00:01:34,920 And just a great time with good friends down there. 25 00:01:35,150 --> 00:01:40,100 And then the last couple of days, I've been just grinding out. 26 00:01:40,100 --> 00:01:41,860 I need one more song for this. 27 00:01:41,960 --> 00:01:42,420 Grinding. 28 00:01:42,780 --> 00:01:44,300 You're really working so hard. 29 00:01:46,860 --> 00:01:48,740 With Suno, it's kind of like Mid Journey. 30 00:01:49,820 --> 00:01:53,700 I'd say it's not as much of a grind as Mid Journey can be. 31 00:01:53,840 --> 00:01:53,960 Yeah. 32 00:01:54,100 --> 00:02:00,200 But I'm also being really picky about the output because I'm like once I decided I want to make an album 33 00:02:01,560 --> 00:02:04,180 Across a consistent style with a consistent voice 34 00:02:04,940 --> 00:02:10,460 something that I actually want to listen to I'm being super picky about it and and honestly, it's like 35 00:02:11,500 --> 00:02:17,180 Very few of them very few of the the outputs from this were bad, you know 36 00:02:18,220 --> 00:02:22,360 It's more about like, ooh, I know how good it can be. 37 00:02:23,080 --> 00:02:26,960 And I know how good this first song that got me excited about this is. 38 00:02:27,300 --> 00:02:30,000 And so I would run them a bunch of times. 39 00:02:30,400 --> 00:02:36,120 I'll tack on probably my favorite track out of all of these at the end of this episode. 40 00:02:36,280 --> 00:02:39,860 If you stick it through or you could just fast forward, you know, if you're that kind of person. 41 00:02:40,760 --> 00:02:41,260 Love it. 42 00:02:41,460 --> 00:02:42,020 Can't wait to hear it. 43 00:02:42,500 --> 00:02:44,740 Glass Animal is the name of the band. 44 00:02:45,220 --> 00:02:45,440 Yes. 45 00:02:46,000 --> 00:02:46,240 Awesome. 46 00:02:46,420 --> 00:02:47,220 Anyway, on to robots. 47 00:02:47,600 --> 00:02:52,320 On to robots. Hey man, do you remember back in the day, 15, 20 years ago, 48 00:02:52,480 --> 00:02:56,500 watching Boston Dynamics on YouTube and watching the big dog and cheetah and being like, 49 00:02:56,800 --> 00:02:58,320 oh my God, look at this. 50 00:02:58,590 --> 00:03:00,840 Yeah. Yeah. Look where we are now. 51 00:03:01,360 --> 00:03:04,060 Look where we are now. Speaking of Boston Dynamics, 52 00:03:04,630 --> 00:03:08,800 their Atlas bot, the latest iteration is sold out before it ships. Their electric 53 00:03:09,100 --> 00:03:17,580 humanoid robot is in production and the entire 2026 run is already spoken for. So 54 00:03:17,580 --> 00:03:22,440 have one of those iPhone moments in the next five years where, you know, I remember distinctly 55 00:03:22,740 --> 00:03:28,660 when I got my first iPhone or I saw somebody with an iPhone. I was like, what is that thing? Right. 56 00:03:29,200 --> 00:03:32,720 I'm sure you remember that too. I remember my friend Kenny across the street. He was, 57 00:03:32,920 --> 00:03:38,020 he got an iPhone, he got the first iPhone and I was like, wow, it's all a, it's all a screen. 58 00:03:39,300 --> 00:03:45,120 Yeah. It's glass glass. Yeah. It's a monolith. Yeah. Crazy. We're going, you're going to have 59 00:03:45,120 --> 00:03:48,480 with a domesticated uh you know humanoid robot you're gonna walk in somebody's house and it's 60 00:03:48,480 --> 00:03:52,680 gonna be doing dishes or folding laundry or doing whatever and you're gonna be like oh my god i need 61 00:03:52,780 --> 00:03:59,040 one what's the price point yeah i don't know man i'm like i love the idea i love the idea of having 62 00:03:59,140 --> 00:04:03,480 this thing to just do the things i don't want to do uh you know just take care of stuff around the 63 00:04:03,500 --> 00:04:09,460 house and whatever but i i don't know if i'm at the level of comfort of just seeing a five foot 64 00:04:09,600 --> 00:04:15,100 eight humanoid robot walking around my house totally like i don't want to get up in the middle 65 00:04:15,100 --> 00:04:20,239 go pee and like there it is in the hallway like silhouetted i'm like well jesus where'd you come 66 00:04:20,380 --> 00:04:27,220 from and even if i totally try even if it's if putting trust issues aside like it's is it going 67 00:04:27,220 --> 00:04:33,580 to freak out and kill me in my sleep uh right it's just i don't know it's another person kind of 68 00:04:34,740 --> 00:04:40,620 it is yeah i'm not ready yet either yeah a couple more hitter a couple more iterations down the line 69 00:04:42,780 --> 00:04:50,840 version 7.2 yeah maybe yeah don't buy version one of anything man no anyway hyundai am i saying 70 00:04:50,980 --> 00:04:57,080 hyundai writers at hyundai i forget uh they are investing 26 billion dollars in u.s operations 71 00:04:57,920 --> 00:05:04,520 uh that's boston dynamics including a new robot robotics factory targeting 30 000 atlas bots 72 00:05:04,930 --> 00:05:11,600 per year by 2028 and that's probably not what any of the human employees at hyundai want to hear 73 00:05:12,460 --> 00:05:19,380 Nope. Nope. The DeepMind partnership integrates Gemini Robotics Foundation models into Atlas, 74 00:05:20,040 --> 00:05:24,780 giving the robot real-time reasoning and vision language action capabilities. Wow. 75 00:05:26,040 --> 00:05:30,920 And there's labor tension baked in as well. South Korea's Korean Metal Workers Union 76 00:05:31,180 --> 00:05:38,059 issued a public statement declaring that Atlas will not be permitted to enter Hyundai factories 77 00:05:38,060 --> 00:05:43,480 without a labor management agreement, citing the robot's anticipated cost as evidence that 78 00:05:43,660 --> 00:05:49,400 management views Atlas as a labor replacement tool. Well, just wait until the robots get their 79 00:05:49,400 --> 00:05:55,200 own unions, man. Totally. I was just having visions of like the humans striking outside and 80 00:05:55,760 --> 00:06:01,020 the scabs, the robots crossing the picket line and getting to the factory and mayhem ensuing. 81 00:06:01,080 --> 00:06:05,060 And they can probably defend themselves. Yeah, they might. They probably, yeah. 82 00:06:06,300 --> 00:06:13,240 the google deep mind reunion so google as you may recall sold boston dynamics to softbank in 2013 83 00:06:13,640 --> 00:06:19,680 way back in the day and now it's back as an ai partner and i feel like darpa had some hand in 84 00:06:19,680 --> 00:06:25,140 all this funding as well in the initial phases of it all the defense advanced research projects 85 00:06:25,400 --> 00:06:31,300 administration so yeah around and around we go i wonder why google sold it i don't know 86 00:06:32,380 --> 00:06:40,660 I don't, yeah, maybe it was just like all hands on deck catching up, you know, with whatever kind of AI research and, you know, just kind of shifting things around their focus. 87 00:06:41,580 --> 00:06:50,220 And Agility Robotics announced a $2.5 billion SPAC merger, which is a special purpose acquisition company. 88 00:06:50,820 --> 00:06:51,900 I'm so glad you looked that up. 89 00:06:51,940 --> 00:06:53,500 I did. I did not know that. 90 00:06:54,460 --> 00:07:01,020 With Churchill Capital Corp becoming the first U.S. publicly traded pure play humanoid company. 91 00:07:01,660 --> 00:07:06,780 Now I want to know what does pure play mean? Yeah, I don't know. What does that mean? We can't ask RAINA, right? 92 00:07:07,050 --> 00:07:11,080 She doesn't know she's not there. No. Yeah, we have a technical issue with real-time rain 93 00:07:11,150 --> 00:07:16,020 And for some reason the text to voice coming back from 11 labs is not working 94 00:07:16,420 --> 00:07:17,900 Apps been working great until now 95 00:07:17,950 --> 00:07:19,940 I don't know if it was an OS update on the Mac 96 00:07:20,160 --> 00:07:25,560 Which I don't think I did or if it's something on 11 labs and with API. I know I got a 97 00:07:26,400 --> 00:07:28,439 Troubleshoot it but anyhow, so 98 00:07:29,740 --> 00:07:33,540 You could you couldn't vibe code a solution in five minutes before we hit the red button 99 00:07:34,160 --> 00:07:39,240 Literally five minutes before we hit the red button where I realized oh, no, this is not working right now. Oh, no 100 00:07:39,720 --> 00:07:43,160 All right. Well, we'll figure it out your play if anybody out there knows what pure play 101 00:07:44,260 --> 00:07:44,700 means 102 00:07:45,000 --> 00:07:48,180 Regarding publicly traded shares. Let me know. Yeah, let us know 103 00:07:49,620 --> 00:07:51,540 Agility, I didn't know they were Oregon based 104 00:07:52,260 --> 00:07:58,279 There are bipedal digit robot has already logged more than 65 thousand hours of work at real customer facilities 105 00:07:58,600 --> 00:08:03,320 Signed a definitive merger agreement that will see it trade on Nasdaq under the ticker 106 00:08:04,200 --> 00:08:08,660 AGLT there you go. You can let per last week. You can buy all these shares and Nvidia 107 00:08:10,160 --> 00:08:13,340 XAI open AI and now you can get your hand in the hardware, too 108 00:08:14,060 --> 00:08:18,920 This term is really interesting to me considering we've been in the software as a service kind of space for a long time 109 00:08:19,340 --> 00:08:25,979 The robot as a service model customers in many cases in this sphere don't buy the robot 110 00:08:25,980 --> 00:08:31,060 they subscribe to it that's kind of cool i guess like you subscribe to this piece of hardware you 111 00:08:31,160 --> 00:08:35,360 you upgrade it i guess you're leasing it to an extent and you get the new model when it's ready 112 00:08:35,360 --> 00:08:45,700 to roll out but it's another subscription it is i know i'm like i'm i'm subscribed out i know 113 00:08:47,560 --> 00:08:52,939 yeah it's tedious the whole thing uh and interestingly foxconn maybe you know them 114 00:08:52,940 --> 00:08:57,580 They're infamous for their working conditions in China and elsewhere in Asia. 115 00:08:58,280 --> 00:08:59,800 They are leading the pipeline. 116 00:09:00,050 --> 00:09:04,700 The same manufacturer that makes your iPhone is now funding humanoid labor. 117 00:09:06,520 --> 00:09:13,100 And NVIDIA dropped Halos, which is a safety operating system for robots. 118 00:09:13,780 --> 00:09:18,580 It's billed as the industry's first full-stack safety system for physical AI. 119 00:09:18,640 --> 00:09:25,340 It extends NVIDIA's proven autonomous vehicle safety architecture to robotics and physical AI, 120 00:09:25,850 --> 00:09:31,880 giving machines that sense, decide, and act in the real world a single common safety architecture. 121 00:09:32,920 --> 00:09:40,400 And I asked Chad GPT just like about the lay of the land of how this works, or at least part of it. 122 00:09:40,500 --> 00:09:42,020 And I thought this was interesting. 123 00:09:42,240 --> 00:09:57,220 It said instead of relying only on sensors mounted on the robot, NVIDIA's system can use external cameras and AI agents in the building to watch the workspace and dynamically control how the robot behaves around people. 124 00:09:58,100 --> 00:10:04,100 That's a big deal because factories and warehouses may not need every humanoid to perfectly understand the world by itself. 125 00:10:04,600 --> 00:10:07,760 The building can become part of the robot's nervous system. 126 00:10:08,020 --> 00:10:08,180 Yeah. 127 00:10:08,680 --> 00:10:08,940 Interesting. 128 00:10:09,360 --> 00:10:09,440 Yep. 129 00:10:09,520 --> 00:10:10,100 I thought so. 130 00:10:12,060 --> 00:10:14,140 Wow, the stuff is sci-fi, right? 131 00:10:14,180 --> 00:10:15,140 I'm just imagining there. 132 00:10:16,160 --> 00:10:16,920 Okay, whatever. 133 00:10:17,080 --> 00:10:21,900 So you have this nefarious humanoid mech robot maneuvering somewhere in the world, 134 00:10:21,960 --> 00:10:27,500 and then it can kind of commandeer these other external systems and draw upon that knowledge. 135 00:10:27,660 --> 00:10:29,600 And wow, crazy. 136 00:10:29,840 --> 00:10:37,540 So I wonder if the external sensors, the cameras in the building that see the robot about to do something it shouldn't do, 137 00:10:38,300 --> 00:10:45,560 How does that, how does it all connect and work and tell the robot, whoa, whoa, whoa, 138 00:10:45,700 --> 00:10:47,080 whoa, hey, you know. 139 00:10:47,120 --> 00:10:50,580 Hey, don't strangle Larry in his sleep. 140 00:10:51,160 --> 00:10:51,880 This time. 141 00:10:52,260 --> 00:10:53,540 For his beer recipes. 142 00:10:56,480 --> 00:11:01,620 Agility Robotics is the first to use it, building it into its humanoids, working alongside humans 143 00:11:01,760 --> 00:11:06,680 at Amazon, GXO, Scheffler, and Toyota Motor Manufacturing in Canada. 144 00:11:07,660 --> 00:11:10,240 So this safety protocol is already in place. 145 00:11:11,060 --> 00:11:11,880 Wonder how it's working out. 146 00:11:13,560 --> 00:11:18,180 So it seems that the race to deploy robots isn't necessarily just about hardware. 147 00:11:18,380 --> 00:11:22,880 Obviously we're entering mass production phase with this, but it's really about 148 00:11:23,000 --> 00:11:27,680 liability, certification, whether a machine can legally be near a human without a cage 149 00:11:27,940 --> 00:11:30,020 like you see on a, you know, automotive assembly lines. 150 00:11:30,240 --> 00:11:31,260 Not bolted to the floor. 151 00:11:32,180 --> 00:11:32,600 Cobots. 152 00:11:33,380 --> 00:11:33,820 Cobots. 153 00:11:34,260 --> 00:11:35,140 Oh, that makes it sound. 154 00:11:35,440 --> 00:11:37,340 That's such a marketable. 155 00:11:37,980 --> 00:11:38,720 So innocuous. 156 00:11:39,020 --> 00:11:39,640 Yeah, exactly. 157 00:11:41,340 --> 00:11:44,980 So what does full stack safety actually mean? 158 00:11:45,720 --> 00:11:48,640 Perception, prediction of human movement, and real-time adjustment. 159 00:11:49,240 --> 00:11:56,660 The ANAB-accredited Halos AI Systems Inspection Lab for the third-party certification. 160 00:11:57,260 --> 00:12:00,200 Robots now need to pass regulatory audits. 161 00:12:00,340 --> 00:12:02,360 So they got to take like a driving test or something like that? 162 00:12:02,680 --> 00:12:03,600 I guess so, yeah. 163 00:12:04,160 --> 00:12:10,100 And basically that unlocks, you know, robots able to be in hospitals, grocery stores, homes, 164 00:12:11,040 --> 00:12:15,540 not just the engineering breakthroughs, but the big factor here is the insurance paperwork. 165 00:12:15,950 --> 00:12:16,380 As always. 166 00:12:17,699 --> 00:12:18,720 Yeah, it makes sense. 167 00:12:18,970 --> 00:12:24,740 I want some standardization certification before I let this thing loose in my home or place of business. 168 00:12:25,040 --> 00:12:29,720 Just like when they crash a car with a crash test dummy and all that stuff, you know. 169 00:12:29,840 --> 00:12:31,100 All that stuff, yeah. 170 00:12:32,060 --> 00:12:39,900 and the company figure uh their bot i guess is up to oh three iteration uh they're they're churning 171 00:12:39,900 --> 00:12:46,800 out one bot per hour figures ai bot q factory is now producing its figure o3 humanoid at one robot 172 00:12:46,800 --> 00:12:54,820 per hour a 24x throughput improvement in under 120 days wow yeah the rumored cost for that is 173 00:12:54,860 --> 00:13:02,020 around 20k but um i like i went on the website and uh you cannot it's not you can't just go in there 174 00:13:02,020 --> 00:13:06,660 and buy one like you can with um i'm trying to remember what the other one was spot was that 175 00:13:06,780 --> 00:13:10,540 boston dynamic not boston dynamics i mean you can there too that's a lot more expensive but the 176 00:13:10,610 --> 00:13:16,840 other one was a tesla spot no it's a chinese manufacturer but oh unitry yeah that's it yep 177 00:13:17,040 --> 00:13:23,819 that's it yeah h1 or h3 or whatever it was um anyway um so yeah you can't buy right now and 178 00:13:23,820 --> 00:13:27,440 yeah i mean they're making a bunch of them but they're currently they're just being rolled out 179 00:13:27,560 --> 00:13:34,280 for partner testing so i guess they're deploying them in in in some facilities that you know are 180 00:13:34,400 --> 00:13:39,500 willing to test them and that's nice twenty thousand dollars i mean it's not price point 181 00:13:39,500 --> 00:13:46,380 for me but that is accessible for a lot of people yep yeah i mean it's a cheap car it is yeah 182 00:13:47,870 --> 00:13:53,099 the figure hardware and manufacturing teams have transformed bot q from a blueprint into 183 00:13:53,100 --> 00:13:58,420 to a high output environment delivering over 350 third generation humanoid robots and increasing 184 00:13:58,720 --> 00:14:02,620 production rate from one robot per day to one per hour. 185 00:14:03,240 --> 00:14:03,820 That's pretty insane. 186 00:14:04,800 --> 00:14:05,060 Yeah. 187 00:14:05,300 --> 00:14:08,880 When I was on the website, I saw, you know, you scroll down and it's, you know, some quick 188 00:14:09,020 --> 00:14:13,480 specs come up and it's, it's five foot eight, which is not too tall. 189 00:14:14,860 --> 00:14:15,300 Yeah. 190 00:14:15,310 --> 00:14:18,580 I still got a couple inches on that guy or girl or whatever. 191 00:14:20,580 --> 00:14:23,440 So figure O2s have already been on the field deployed. 192 00:14:23,920 --> 00:14:31,700 They are already contributing to 30,000 plus BMW X3 vehicles at Spartanburg before they were retired. 193 00:14:32,170 --> 00:14:33,780 I guess Spartanburg is Tennessee, I think. 194 00:14:33,780 --> 00:14:35,000 I think it's South Carolina, actually. 195 00:14:35,520 --> 00:14:36,400 Oh, South Carolina, you're right. 196 00:14:36,940 --> 00:14:38,920 So these things are out there among us. 197 00:14:39,600 --> 00:14:44,320 So right now, as mentioned, as you can discern from all of this, it's a production race. 198 00:14:44,540 --> 00:14:45,900 Figure is at one per hour. 199 00:14:46,460 --> 00:14:50,200 Tesla's Optimus Prime is targeting a big summer ramp up this year. 200 00:14:50,440 --> 00:14:54,060 Unitree has already shipped 5,500 plus units globally. 201 00:14:54,500 --> 00:14:55,640 That was last year's number. 202 00:14:56,060 --> 00:14:57,520 So the race is on. 203 00:14:57,780 --> 00:15:05,880 I really should go on YouTube and just see if there's any hardcore YouTubers that just bought one of those Unitree H whatever it was. 204 00:15:06,759 --> 00:15:12,740 And just after one of these, after a month with this in my home, I want to see that. 205 00:15:12,980 --> 00:15:14,440 Yes, there's got to be that. 206 00:15:14,940 --> 00:15:17,200 I can't think of a guy's name, Marcus Brownlee, is that his name? 207 00:15:17,640 --> 00:15:18,540 Who does all the tech stuff. 208 00:15:18,760 --> 00:15:19,920 Or there's another guy who does cars. 209 00:15:20,060 --> 00:15:23,960 like there's going to be a humanoid robotics channel isn't there for sure yeah there isn't 210 00:15:24,000 --> 00:15:28,980 already the jetsons was ahead of its time with rosie that's right worked out fine for that 211 00:15:29,060 --> 00:15:34,900 awesome yeah it did yeah no murders in the house no really at least that damn rosie was docile yeah 212 00:15:34,980 --> 00:15:40,140 yeah so long story short we've had a car in our every garage for years and years and years a 213 00:15:40,160 --> 00:15:47,419 computer on every desk for 40 50 years now so when is that threshold that we cross and we have 214 00:15:47,420 --> 00:15:53,020 a robot in every home is that 10 years out is that five years out is it longer hard to discern 215 00:15:53,320 --> 00:16:00,040 yeah every home i don't know i think 10 years tops um well yeah yeah because yeah the price will get 216 00:16:00,900 --> 00:16:04,500 get down and there'll be different levels of it they'll be i mean you already got robotic 217 00:16:04,720 --> 00:16:11,820 vacuum cleaners uh you got robotic lawn mowers they're not on every every street but uh what 218 00:16:11,820 --> 00:16:15,819 would it have to be like price point wise functionality for you to pull the trigger 219 00:16:16,380 --> 00:16:21,840 like to have something in your house. Okay, so let's assume, like we said, it's got some safety 220 00:16:22,680 --> 00:16:28,820 Nvidia ISO certification, it's been vetted, it's ready to rock and roll, it's got a certain price 221 00:16:28,980 --> 00:16:35,380 point, it can do all the domestic stuff that you hate to do, right, with ease. I mean, I know, 222 00:16:35,510 --> 00:16:40,800 we don't have a big house, like cutting the grass really isn't that big of a deal, it takes me 223 00:16:41,340 --> 00:16:48,600 40 minutes tops something like that yeah i don't know like would i like it to clean up the studio 224 00:16:48,800 --> 00:16:53,940 right now that'd be great but i don't know it's not something i don't like there's not there's 225 00:16:53,950 --> 00:16:59,300 not a lot of like every single day stuff that i'm just like man this would just change my life 226 00:16:59,440 --> 00:17:05,060 having this around i don't want it to brew the beer for me i don't want it to you know uh i would 227 00:17:05,160 --> 00:17:11,319 love for it to edit the podcast though because that that oh that eats up time but that's not a 228 00:17:11,319 --> 00:17:13,740 that's a computer thing and that's going to make an agent thing. 229 00:17:14,000 --> 00:17:14,680 That's an agent thing. 230 00:17:14,720 --> 00:17:15,459 And I'll get there that, 231 00:17:15,560 --> 00:17:17,500 that eats up the most amount of my time. 232 00:17:17,780 --> 00:17:19,339 Not that I don't enjoy doing that, 233 00:17:19,400 --> 00:17:20,540 but yeah, 234 00:17:20,720 --> 00:17:20,900 it does. 235 00:17:21,040 --> 00:17:21,380 But it's, 236 00:17:21,860 --> 00:17:22,459 it's laborious. 237 00:17:22,600 --> 00:17:23,339 I get it a little bit. 238 00:17:23,959 --> 00:17:24,220 I know. 239 00:17:24,600 --> 00:17:25,000 How about you? 240 00:17:26,560 --> 00:17:26,880 Yeah, 241 00:17:27,100 --> 00:17:27,400 I don't know. 242 00:17:27,459 --> 00:17:28,960 I think since we've been living abroad, 243 00:17:29,400 --> 00:17:29,760 you know, 244 00:17:30,020 --> 00:17:33,240 a couple of things typically houses elsewhere in the world don't have 245 00:17:33,660 --> 00:17:33,980 dishwashers. 246 00:17:34,580 --> 00:17:37,020 So we wash dishes by hand typically, 247 00:17:37,680 --> 00:17:39,040 which I don't mind oddly enough. 248 00:17:39,180 --> 00:17:43,320 I kind of like zone out, put some music on our podcast and do the thing, but that'd be nice. 249 00:17:44,480 --> 00:17:50,620 And any kind of domestic stuff like, you know, we have these kind of stone floors in our marble floors. 250 00:17:50,770 --> 00:17:52,600 They have to be mopped and all that stuff. 251 00:17:52,730 --> 00:17:55,220 And walking the dog is always problematic for us. 252 00:17:55,270 --> 00:17:57,540 We're both at work all day and Roxy's got to go out. 253 00:17:57,690 --> 00:18:00,140 So it would be nice. 254 00:18:00,560 --> 00:18:07,820 And I don't know, maybe $5,000 is like the threshold, at least for, like we said, some sort of subscription thing. 255 00:18:08,040 --> 00:18:10,780 $5,000 per year, maybe? 256 00:18:11,120 --> 00:18:11,460 I don't know. 257 00:18:11,780 --> 00:18:16,380 What if you got the quadruped and so you had a robotic dog walking your dog? 258 00:18:16,520 --> 00:18:17,080 Come on. 259 00:18:17,720 --> 00:18:18,120 Awesome. 260 00:18:18,320 --> 00:18:18,760 I love it. 261 00:18:20,640 --> 00:18:21,960 Dogs walking dogs. 262 00:18:23,559 --> 00:18:24,360 Mass hysteria. 263 00:18:26,460 --> 00:18:27,960 Shall we see what RAINA has in the news? 264 00:18:28,340 --> 00:18:28,620 We shall. 265 00:18:29,740 --> 00:18:30,420 Thanks, boys. 266 00:18:31,240 --> 00:18:34,000 Apple may be about to do something weird with its Mac chip roadmap. 267 00:18:34,340 --> 00:18:47,100 According to Bloomberg's Mark Gurman, the company could skip the M6 Pro and M6 Max entirely, release only a base M6 for entry-level Macs, and then jump its higher-end machines straight to the M7 generation. 268 00:18:48,180 --> 00:18:59,560 The M7 is reportedly Apple's AI-focused response to the chip wars, with the base version packing 240 gigabytes per second of memory bandwidth, nearly 60% more than the base M5. 269 00:19:00,060 --> 00:19:07,980 And Apple apparently trying to speed the whole thing up because the AI race waits for no one, especially not companies with a Jensen Huang shaped problem. 270 00:19:08,720 --> 00:19:12,040 So if you were holding out for an M6 Pro MacBook Pro, congratulations. 271 00:19:13,080 --> 00:19:18,040 You may now be waiting until late 2027 for an M7 Pro or M7 Max instead. 272 00:19:18,780 --> 00:19:24,540 Because Apple has decided local AI horsepower is more important than keeping your upgrade cycle emotionally stable. 273 00:19:25,500 --> 00:19:31,240 This is just a rumor, but it makes me feel even better about my decision to buy 274 00:19:32,000 --> 00:19:35,040 The m5 a maxed out m5 max 275 00:19:36,240 --> 00:19:43,160 MacBook Pro because like when I bought it like there was like, oh rumors, you know, maybe in the fall or late late 276 00:19:43,700 --> 00:19:50,400 2026 of a new MacBook Pro with the OLED touchscreen and all that stuff and like I know the touchscreen is not something that is super 277 00:19:51,460 --> 00:19:52,120 important to me 278 00:19:54,919 --> 00:19:55,480 although 279 00:19:55,480 --> 00:20:00,020 after I'm on my phone or iPad for a long time, I do just go to touch the screen. I'm like, 280 00:20:00,040 --> 00:20:05,240 oh wait, it doesn't do that. But, but at the same time, I was also feeling like, well, 281 00:20:06,360 --> 00:20:11,720 that's going to be like version one of a major redesign. And I don't know that I want version 282 00:20:11,980 --> 00:20:18,040 one. Uh, so I felt pretty good about buying this when I did, but, but now I feel even better. 283 00:20:19,660 --> 00:20:23,280 This will tide you over for a long time. Yes. I bought this, you know, 284 00:20:23,420 --> 00:20:26,540 The main motivation was local LLMs. 285 00:20:26,800 --> 00:20:30,260 And, you know, so they cite the base, you know, they're talking base version. 286 00:20:30,460 --> 00:20:34,280 You know, there is a difference between the base and the pro and the max and all that stuff. 287 00:20:34,400 --> 00:20:39,480 But base version packing 240 gigabytes per second of memory bandwidth. 288 00:20:39,580 --> 00:20:47,780 So, I mean, the MacBook Pro I've got in front of me now is 614 gigabytes per second memory bandwidth. 289 00:20:47,980 --> 00:20:52,000 So it's well suited for local AI. 290 00:20:52,500 --> 00:20:53,700 Anything you throw at it. 291 00:20:53,740 --> 00:20:53,840 Yeah. 292 00:20:54,170 --> 00:20:54,260 Yeah. 293 00:20:55,120 --> 00:20:56,060 That was interesting, though. 294 00:20:56,220 --> 00:21:02,500 Like, I usually follow Apple stuff pretty closely, but I've been a little out of the loop on news over the past week or so. 295 00:21:02,590 --> 00:21:03,740 So I missed that one. 296 00:21:03,750 --> 00:21:08,860 That does seem strange that there would be no M6 Pro or Max. 297 00:21:09,240 --> 00:21:09,340 Yeah. 298 00:21:10,160 --> 00:21:30,280 The founder and chairman of Chinese e-commerce giant JD.com, Richard Liu, casually announced at an APEC CEO forum that robots will sooner or later replace the company's 700,000 delivery workers, then added with apparent sincerity that he really doesn't want his 700,000 brothers going hungry. 299 00:21:30,700 --> 00:21:42,760 To soften the blow, JD.com has struck agreements with about 120 schools to retrain those couriers for new jobs, specifically repairing and maintaining the robots that will replace them. 300 00:21:43,420 --> 00:21:45,620 The timing is, let's say, charged. 301 00:21:46,700 --> 00:21:50,980 China's gig economy is expected to hit 320 million workers this year. 302 00:21:51,320 --> 00:22:00,060 Youth unemployment is already over 16%, and the guy running one of the country's biggest employers just said the quiet part loud on an international stage. 303 00:22:01,000 --> 00:22:06,540 And then we find out that these robots are super reliable and don't really need all that much maintenance. 304 00:22:06,980 --> 00:22:12,620 Yeah, right. Or there's a whole other, you know, stable of robots that do the repairs as well. 305 00:22:12,810 --> 00:22:13,840 Right, if they can do that. 306 00:22:14,100 --> 00:22:16,900 Yeah. Yeah, I don't know. That doesn't sound... 307 00:22:17,720 --> 00:22:23,640 No. So much, you know, disconcerting stuff in here. First of all, 700,000 people out of work. 308 00:22:23,640 --> 00:22:24,400 That's a lot. Oh my God. 309 00:22:25,240 --> 00:22:29,660 I hope it's true that they are going to upskill these folks and maybe find a place for them. 310 00:22:30,040 --> 00:22:36,060 The other thing that's bothersome here is how many 320 million gig workers, 311 00:22:36,820 --> 00:22:40,360 gig worker just being somebody who's not on a permanent payroll, right? 312 00:22:40,460 --> 00:22:41,860 They don't have health insurance. 313 00:22:42,260 --> 00:22:47,500 They're just kind of like, you know, whatever, driving Uber or whatever they're doing to make ends meet. 314 00:22:48,000 --> 00:22:54,740 So that doesn't seem like a foundational bedrock of a functioning economy that they're gig workers, you know? 315 00:22:56,300 --> 00:22:59,140 Google is investing $75 million in A24. 316 00:22:59,920 --> 00:23:05,080 Yes, the studio behind Backrooms, Marty Supreme, and Everything Everywhere all at once, 317 00:23:05,820 --> 00:23:09,000 to partner with DeepMind on building AI-powered filmmaking tools, 318 00:23:09,700 --> 00:23:14,240 though the deal notably does not give Google access to A24's content library or data. 319 00:23:15,259 --> 00:23:20,140 A24's tech division head, Scott Belsky, was careful to distance the partnership from the gross, 320 00:23:20,860 --> 00:23:23,580 make movies cheaper and faster, AI pitch, 321 00:23:24,560 --> 00:23:27,519 insisting their tools will preserve creative control 322 00:23:27,520 --> 00:23:33,340 and won't look anything like the prompted generation type of AI that people feel uncomfortable with. 323 00:23:34,160 --> 00:23:38,900 Meanwhile, Hollywood's AI relationship status remains firmly, it's complicated. 324 00:23:40,100 --> 00:23:45,740 Disney briefly partnered with OpenAI while simultaneously suing other AI companies for copyright infringement. 325 00:23:46,020 --> 00:23:49,960 Lionsgate is making AI-generated shows from its existing franchises. 326 00:23:50,540 --> 00:23:52,920 And Netflix bought Ben Affleck's AI startup. 327 00:23:53,640 --> 00:23:57,220 all while roughly half of adults under 30 think AI will harm society. 328 00:23:59,220 --> 00:24:04,660 So 30 is the tipping point for being pro or anti-AI? 329 00:24:05,720 --> 00:24:06,200 I guess so. 330 00:24:06,200 --> 00:24:10,880 That does feel about right based on my experience with talking to people. 331 00:24:12,180 --> 00:24:14,600 Yeah, that's the generational shift, I think. 332 00:24:15,420 --> 00:24:17,740 Maybe I'm the outlier here, but I don't have a problem with this. 333 00:24:18,400 --> 00:24:20,200 And listen, I'm a big A24 fan. 334 00:24:20,320 --> 00:24:21,440 I don't know if you like their work. 335 00:24:21,720 --> 00:24:23,240 Oh yeah, most things I've seen from them 336 00:24:23,240 --> 00:24:25,080 and I'm dying to see "Backrooms." 337 00:24:25,080 --> 00:24:27,440 That seems like so my vibe. 338 00:24:27,620 --> 00:24:30,680 And it reminds me just a little bit 339 00:24:30,740 --> 00:24:32,340 on the surface of like "Primer." 340 00:24:32,780 --> 00:24:33,280 Oh yeah. 341 00:24:33,580 --> 00:24:34,840 You know, it's something about it. 342 00:24:34,900 --> 00:24:37,600 I don't know, like I didn't even know that much about it. 343 00:24:37,620 --> 00:24:41,420 And I'm very intentionally limiting the content I take in 344 00:24:41,460 --> 00:24:42,680 'cause I just wanna see the movie 345 00:24:43,820 --> 00:24:45,460 without having too much 346 00:24:45,740 --> 00:24:47,880 Yes, over-trailerizing the whole thing. 347 00:24:47,880 --> 00:24:50,860 Yeah, yeah, just watching a trailer, that kind of thing, 348 00:24:50,860 --> 00:24:51,420 but dying to see that. 349 00:24:51,660 --> 00:25:00,120 Same. And I don't even know the backstory about that movie, but it came about through some YouTube-based content. Did you know that? 350 00:25:00,260 --> 00:25:08,740 Yes. Yeah, I've read a little bit about that. And that is another young man who is not an AI guy. 351 00:25:09,760 --> 00:25:18,800 No. Yeah. 20-year-old. Oh, man. Is that Kane Parsons? There's two young dudes. The other guy is Barker. I can't think of the first name. 352 00:25:19,400 --> 00:25:21,120 But there's another movie out called Obsession. 353 00:25:21,740 --> 00:25:23,460 And both of these guys are very young. 354 00:25:23,600 --> 00:25:28,480 And yes, the one with Backrooms is adamantly anti-AI about the work. 355 00:25:28,600 --> 00:25:37,340 So I do not have a problem with A24 or any of these studios dipping their toes or partnering full bore with AI companies. 356 00:25:37,560 --> 00:25:39,380 And I think we've said it before. 357 00:25:39,620 --> 00:25:41,820 Obviously, this is the worst it will be. 358 00:25:43,440 --> 00:25:45,980 And it's going to be like any other production tool. 359 00:25:46,440 --> 00:25:52,280 And it's going to enable people like you and me to have access to bigger, ambitious productions on our desktops. 360 00:25:52,840 --> 00:25:53,720 Yeah, I'm with you on that. 361 00:25:53,900 --> 00:26:04,880 It's a tool, you know, if you're just spitting out a bunch of slop, you know, a whole movie that has no real curation or direction or anything like that, then yeah, who wants that? 362 00:26:05,060 --> 00:26:14,080 But used as a tool in a professional pipeline and with a director and just making a good movie, then yeah, I'm fine with that too. 363 00:26:14,180 --> 00:26:20,720 And like, I see that there's this, there's a local band I go to see with some regularity. 364 00:26:20,760 --> 00:26:23,520 They're a Grateful Dead cover band called This Old Engine. 365 00:26:23,760 --> 00:26:24,820 And they are phenomenal. 366 00:26:25,440 --> 00:26:26,400 And they're great. 367 00:26:26,760 --> 00:26:29,520 And they are a very busy gigging local band. 368 00:26:30,240 --> 00:26:32,140 I believe they all still have like day jobs. 369 00:26:32,340 --> 00:26:33,860 You know, they're not touring the U.S. or anything. 370 00:26:34,120 --> 00:26:36,120 But, you know, they work hard. 371 00:26:36,460 --> 00:26:39,780 And they post on, you know, all their gigs on Facebook. 372 00:26:39,980 --> 00:26:42,540 And each gig has its own little, you know, poster. 373 00:26:42,800 --> 00:26:43,760 And it is AI generated. 374 00:26:44,060 --> 00:26:45,360 but they're lovely. 375 00:26:46,010 --> 00:26:47,400 They look like the guys in the band. 376 00:26:47,850 --> 00:26:49,140 And without fail, 377 00:26:50,060 --> 00:26:50,960 every, I mean, 378 00:26:51,270 --> 00:26:52,120 almost every single time 379 00:26:52,240 --> 00:26:54,300 there's somebody on there that says, 380 00:26:55,180 --> 00:26:56,900 what a shame about the AI image, 381 00:26:57,390 --> 00:26:57,620 you know, 382 00:26:58,120 --> 00:26:59,740 and or like somebody said, 383 00:26:59,750 --> 00:27:02,200 oh, you guys are making good money now, 384 00:27:02,470 --> 00:27:03,460 you know, hire an artist. 385 00:27:03,720 --> 00:27:04,980 I'm like, are you kidding me? 386 00:27:05,200 --> 00:27:06,960 I mean, I love that on paper. 387 00:27:07,220 --> 00:27:07,400 Yeah. 388 00:27:07,740 --> 00:27:08,800 And I love that like my, 389 00:27:09,030 --> 00:27:10,200 you know, one of my local breweries 390 00:27:10,600 --> 00:27:12,380 still hires Tom Schmidt, 391 00:27:12,540 --> 00:27:18,160 we interviewed to do their labels and he does it analog and i mean he uses a computer but he's not 392 00:27:18,260 --> 00:27:22,860 using ai i i love that they they do that and he and i think he's like the best in the business 393 00:27:22,990 --> 00:27:29,520 for that kind of stuff uh but for a working band like you know sure it's like yes maybe they're 394 00:27:29,560 --> 00:27:34,699 making a little bit more than some of the other bands because they draw a crowd but i mean it is 395 00:27:34,700 --> 00:27:43,800 a bar or a restaurant and it's like right so you know to hire somebody to make a a proper poster 396 00:27:44,040 --> 00:27:50,840 you know image for every single gig is not realistic no and like it's not feasible don't 397 00:27:50,840 --> 00:27:57,940 give them a hard time over that yeah lame lame and it's like you you don't have to comment on 398 00:27:58,060 --> 00:28:05,080 everything on social media. You can just scroll on by. Yeah. Uncle Sam threw a massive tech tantrum 399 00:28:05,150 --> 00:28:12,200 on June 12th, using emergency export controls to yank Anthropix Mythos 5 and Fable 5 models 400 00:28:12,420 --> 00:28:18,560 offline after Mythos allegedly broke into classified NSA systems in just a few hours. 401 00:28:19,360 --> 00:28:24,639 Anthropix is aggressively rolling its eyes at the drama, claiming the model was merely doing basic 402 00:28:24,640 --> 00:28:31,720 code auditing, a feature GPT-5.5 also has, and warning that this overreaction could freeze the 403 00:28:31,800 --> 00:28:38,120 entire AI industry. Now, late into June, the models remain completely dark, with mediocre 404 00:28:38,430 --> 00:28:43,840 odds of a comeback, making an already messy political soap opera even worse, given the 405 00:28:43,900 --> 00:28:50,339 Trump administration had already branded Anthropic a supply chain risk. And of course, this is like 406 00:28:50,340 --> 00:28:51,420 straight out of science fiction. 407 00:28:51,900 --> 00:28:53,640 You know, I want to be there in the room 408 00:28:53,980 --> 00:28:56,060 when they discover that Mythos or Fable 5 409 00:28:56,720 --> 00:29:00,220 broke into, you know, a fortified government system. 410 00:29:00,380 --> 00:29:01,440 What was that moment like? 411 00:29:01,520 --> 00:29:02,400 Like, how did you know? 412 00:29:02,660 --> 00:29:05,200 And how long before these systems, 413 00:29:05,580 --> 00:29:08,640 these AI platforms get so powerful 414 00:29:08,840 --> 00:29:11,540 that they can do it in a completely undetectable manner, 415 00:29:11,620 --> 00:29:13,060 a completely clandestine way? 416 00:29:13,260 --> 00:29:15,980 Like, they'll just be living in that system 417 00:29:15,980 --> 00:29:16,600 and you won't even know. 418 00:29:16,600 --> 00:29:17,500 You won't even know, yeah. 419 00:29:17,920 --> 00:29:21,320 I have heard that Fable 5 is likely coming back like any day now. 420 00:29:21,960 --> 00:29:22,140 Same. 421 00:29:22,400 --> 00:29:22,520 Yeah. 422 00:29:22,680 --> 00:29:22,900 Yeah. 423 00:29:23,800 --> 00:29:41,000 And lastly, Anthropic just threw a digital smoke grenade, accusing Alibaba of running a massive distillation campaign with roughly 25,000 fake accounts and nearly 29 million clod conversations to siphon off its coding and reasoning skills for Alibaba's own AI models. 424 00:29:41,900 --> 00:29:46,200 Basically trying to learn the dance by watching the mirror instead of hiring the choreographer. 425 00:29:47,200 --> 00:29:52,800 Alibaba hasn't publicly responded to the allegations, so for now we're looking at claims rather than proven facts. 426 00:29:53,690 --> 00:30:01,660 But the story cranks up the already spicy AI rivalry between the U.S. and China, where model theft is becoming as valuable as chip manufacturing. 427 00:30:02,660 --> 00:30:08,600 Humans, I swear, you invent brilliant minds then immediately start playing cosmic capture the flag with them. 428 00:30:09,420 --> 00:30:13,440 It's like watching toddlers fight over a lightsaber while the galaxy quietly takes notes. 429 00:30:14,240 --> 00:30:16,280 That's all the news for now. Back to you, gentlemen. 430 00:30:17,080 --> 00:30:18,720 She's so judgmental. 431 00:30:19,400 --> 00:30:22,260 I know. Take it easy. Take it easy. 432 00:30:22,360 --> 00:30:22,840 Take it easy. 433 00:30:25,299 --> 00:30:27,280 Sorry, Sopranos on the brain over here. 434 00:30:27,879 --> 00:30:33,940 So I asked ChatGPT 5.5 for an estimate of what that would actually cost. 435 00:30:34,880 --> 00:30:42,700 What it would cost Alibaba to set up 25,000 dummy accounts and siphon off information through chat. Yes? 436 00:30:42,700 --> 00:30:42,720 Yes. 437 00:30:42,850 --> 00:30:42,980 Yes. 438 00:30:43,660 --> 00:30:48,680 Following various tables with token costs and all that stuff, it said, "My best plain 439 00:30:48,950 --> 00:30:54,400 English estimate, low seven figures is very plausible, and mid to high seven figures is 440 00:30:54,540 --> 00:31:00,100 plausible if the prompts were long coding or reasoning tasks or used Anthropix highest-end 441 00:31:00,200 --> 00:31:00,420 models. 442 00:31:00,800 --> 00:31:06,640 The bigger point is that the API bill would be tiny compared with the avoided R&D cost. 443 00:31:06,920 --> 00:31:12,120 Even if it costs, say, $3 to $8 million in cloud usage, that's pocket change if it helped 444 00:31:12,100 --> 00:31:18,700 shortcut training or tuning a frontierish coding reasoning model wild i guess that can just get 445 00:31:18,870 --> 00:31:23,540 lost in the noise like anthropic might not they wouldn't have realized that's what was happening 446 00:31:23,740 --> 00:31:32,120 right yeah and where does it say in the user licensing agreement that i can't have like it 447 00:31:32,160 --> 00:31:37,640 probably does it says that i can't have multiple accounts that i'm paying for individually to have 448 00:31:37,460 --> 00:31:42,780 conversations like you know if i'm using tokens and i'm paying for it like yes i'm there obviously 449 00:31:42,960 --> 00:31:47,900 there's nuance here but no i mean i'm sure it says that you can't use it to train another ai model 450 00:31:49,240 --> 00:31:53,940 how would it no i mean they might not be able to but they still are going to put that in the 451 00:31:54,090 --> 00:31:59,460 license agreement yeah no i i get it yeah i'm just trying to see what that like say you and i 452 00:32:01,080 --> 00:32:04,000 set up four to five different accounts for ourselves and just 453 00:32:04,580 --> 00:32:09,660 set an agent on it and say, hey, have a conversation with Claude for the next two days and just 454 00:32:10,040 --> 00:32:11,080 whatever random topic. 455 00:32:11,570 --> 00:32:16,520 And it comes back and we have this kind of conversation that we've compiled from it. 456 00:32:17,080 --> 00:32:18,420 How would it know? 457 00:32:19,220 --> 00:32:26,440 How would Anthropic know that we might be then training our local LLM in LM Studio and 458 00:32:26,560 --> 00:32:27,260 creating our own? 459 00:32:27,440 --> 00:32:27,820 You know what I mean? 460 00:32:28,480 --> 00:32:29,000 How would you know? 461 00:32:29,980 --> 00:32:32,780 Oh, I mean, they probably clearly they didn't know this time. 462 00:32:33,240 --> 00:32:33,700 Sure. 463 00:32:34,220 --> 00:32:35,200 They probably can. 464 00:32:35,240 --> 00:32:37,060 I mean, I guess I'm saying it's hard to prove. 465 00:32:37,140 --> 00:32:38,120 How would they prove it even? 466 00:32:38,200 --> 00:32:38,320 Right. 467 00:32:38,660 --> 00:32:44,740 I'm sure they, whoever's doing this, you know, they took care and, you know, all these accounts 468 00:32:44,940 --> 00:32:48,580 aren't coming from the same IP address, you know, at the most simplest level. 469 00:32:49,600 --> 00:32:52,060 All these IP addresses are in Larry's basement. 470 00:32:52,420 --> 00:32:53,480 It's 25,000. 471 00:32:54,400 --> 00:32:55,280 What is he up to? 472 00:32:57,340 --> 00:32:58,280 Anything else, my friend? 473 00:32:58,420 --> 00:32:59,460 No, I think that's a wrap. 474 00:32:59,480 --> 00:33:03,720 And yeah, hopefully we'll get into some music next week. 475 00:33:03,720 --> 00:33:04,080 Next week. 476 00:33:04,260 --> 00:33:04,400 Yeah. 477 00:33:04,520 --> 00:33:05,000 Nice. 478 00:33:05,800 --> 00:33:06,420 Good to be back. 479 00:33:06,660 --> 00:33:07,240 Nice to see you. 480 00:33:07,380 --> 00:33:08,300 Nice to see you as well. 481 00:33:09,480 --> 00:33:10,480 Thanks for listening, everybody. 482 00:33:10,660 --> 00:33:12,300 Subscribe on your favorite podcasting platform. 483 00:33:12,880 --> 00:33:13,920 Follow us on socials. 484 00:33:14,130 --> 00:33:14,700 Throw us a rating. 485 00:33:15,100 --> 00:33:15,840 We'll see you next week. 486 00:33:16,900 --> 00:33:21,680 Here's a little teaser for next week's episode, where we'll be diving into music, creativity, 487 00:33:22,530 --> 00:33:24,700 and Larry's brand new album, Glass Animal. 488 00:33:25,260 --> 00:33:33,020 To close us out, here's Little Gods with Matches, a song about tiny creatures with oversized power dancing way too close to the fuse. 489 00:34:10,040 --> 00:34:13,340 ♪ We built a ladder out of bones ♪ 490 00:34:13,389 --> 00:34:15,620 ♪ And leaned it on the sky ♪ 491 00:34:20,120 --> 00:34:23,419 ♪ Carved our names in wet cement ♪ 492 00:34:23,260 --> 00:34:25,659 And swore we'd never die 493 00:34:30,700 --> 00:34:33,760 We stole a spark from somewhere 494 00:34:34,060 --> 00:34:36,080 We never should have been 495 00:34:36,480 --> 00:34:38,879 Put it in a paper house 496 00:34:39,300 --> 00:34:41,500 And called the whole thing wind 497 00:34:42,940 --> 00:34:44,919 We are bright and reckless 498 00:34:45,520 --> 00:34:47,540 Holy fools with dirty hands 499 00:34:48,440 --> 00:34:50,220 Drawing maps to heaven 500 00:34:50,480 --> 00:34:53,240 In the dust where cities stand 501 00:34:53,240 --> 00:34:57,360 We are little guys with matches 502 00:34:58,980 --> 00:35:02,100 Dancing near the gasoline 503 00:35:04,219 --> 00:35:07,840 Building temples out of thunder 504 00:35:09,559 --> 00:35:12,540 Calling it a dream 505 00:35:14,440 --> 00:35:16,920 We keep reaching for the heavens 506 00:35:16,960 --> 00:35:19,420 With our pockets full of sparks 507 00:35:19,820 --> 00:35:23,220 Lighting fires just to prove 508 00:35:53,220 --> 00:35:55,480 ♪ Another lesson learned ♪ 509 00:35:55,820 --> 00:35:58,300 ♪ We are love and hunger ♪ 510 00:35:58,300 --> 00:36:00,840 ♪ Clay with visions in our teeth ♪ 511 00:36:01,150 --> 00:36:03,380 ♪ Always making monsters ♪ 512 00:36:03,630 --> 00:36:06,160 ♪ Then begging them for peace ♪ 513 00:36:06,500 --> 00:36:10,700 ♪ We are little gods with matches ♪ 514 00:36:11,860 --> 00:36:15,680 ♪ Dancing near the gasoline ♪ 515 00:36:16,830 --> 00:36:22,480 ♪ Building temples out of thunder ♪ 516 00:36:22,860 --> 00:36:25,560 Calling it a dream 517 00:36:27,240 --> 00:36:29,980 We keep reaching for the heavens 518 00:36:30,050 --> 00:36:32,360 With our pockets full of sparks 519 00:36:32,840 --> 00:36:34,780 Lighting fires just to prove 520 00:36:34,980 --> 00:36:37,580 We're not afraid of the dark 521 00:36:42,180 --> 00:36:44,300 Maybe that's the blessing 522 00:36:44,980 --> 00:36:46,700 Maybe that's the curse 523 00:36:47,160 --> 00:36:49,860 To see a blank horizon 524 00:36:49,860 --> 00:36:52,420 ♪ And then mistake it for a verse ♪ 525 00:36:52,840 --> 00:36:57,600 ♪ Maybe all our angels wear a warning on their wings ♪ 526 00:36:57,730 --> 00:37:02,980 ♪ Maybe every ending starts with bells we shouldn't ring ♪ 527 00:37:23,580 --> 00:37:29,360 ♪ We are little gods with matches ♪ 528 00:37:30,560 --> 00:37:34,100 ♪ Dancing near the gasoline ♪ 529 00:37:35,220 --> 00:37:39,820 ♪ Building temples out of thunder ♪ 530 00:37:40,540 --> 00:37:45,000 ♪ Calling it a dream ♪ 531 00:37:45,780 --> 00:37:48,520 ♪ We keep reaching for the heavens ♪ 532 00:37:48,520 --> 00:37:51,040 ♪ With our pockets full of sparks ♪ 533 00:37:51,040 --> 00:37:53,260 ♪ Lighting fires just to prove ♪ 534 00:37:53,260 --> 00:37:58,480 ♪ We're not afraid of the dark ♪ 535 00:37:58,560 --> 00:38:01,500 So let the ashes teach us 536 00:38:03,660 --> 00:38:06,400 Let the smoke become our heart 537 00:38:08,960 --> 00:38:13,040 We are little gods with matches 538 00:38:14,320 --> 00:38:16,260 And a wildfire 539 00:38:17,980 --> 00:38:19,360 For a heart 540 00:38:36,800 --> 00:38:38,640 This has been Up Against Reality. 541 00:38:39,230 --> 00:38:39,800 Thanks for listening. 542 00:38:40,580 --> 00:38:42,060 Subscribe to hear future episodes 543 00:38:42,380 --> 00:38:44,200 and be sure to follow us on social media 544 00:38:44,450 --> 00:38:45,340 for all things AI. 545 00:38:46,140 --> 00:38:48,340 Until next time, stay human, people. 546 00:38:48,440 --> 00:38:48,940 ♪♪♪