WEBVTT - Nuclear Space Junk Cleaners, Inc. - Week in Tech 0:00:11.600 --> 0:00:14.800 Taylor, Rie, Kyle, I've been obsessed with heists in the 0:00:15.000 --> 0:00:17.730 AI era. And heists gone wrong are, of course, some 0:00:17.750 --> 0:00:20.150 of the great comic gold. Has anyone been following the 0:00:20.190 --> 0:00:21.370 stolen NVIDIA trucks? 0:00:21.870 --> 0:00:22.410 No. 0:00:22.829 --> 0:00:23.680 What's going on? 0:00:23.880 --> 0:00:25.420 It wasn't me, if that's what you're asking. 0:00:26.150 --> 0:00:28.190 So there are a couple of trucks with NVIDIA branding 0:00:28.350 --> 0:00:33.270 on them that some thieves thought were full of NVIDIA chips, which. 0:00:33.150 --> 0:00:34.950 Of course are very, very valuable. 0:00:36.130 --> 0:00:39.750 And so the trucks were stolen, but it happened they 0:00:39.790 --> 0:00:44.530 were just full of sand. And not the sand that's 0:00:44.610 --> 0:00:47.790 used to make chips. The trucks belong to Plus AI, 0:00:47.890 --> 0:00:51.390 who makes self-driving trucks. And in order to make self-driving 0:00:51.430 --> 0:00:54.270 trucks good, you have to simulate cargo loads. So they're 0:00:54.290 --> 0:00:58.190 doing that with tons of sand, 20,000 tons per truck. 0:00:58.370 --> 0:01:01.570 They also happen to have a partnership with NVIDIA. So 0:01:01.590 --> 0:01:04.330 the NVIDIA logos were just on the truck to kind 0:01:04.370 --> 0:01:08.110 of flex about the NVIDIA partnership, like the NASA logo 0:01:08.330 --> 0:01:09.029 on my computer. 0:01:09.190 --> 0:01:10.209 Good advertising. 0:01:10.310 --> 0:01:13.209 If I see someone selling sand on eBay, I will 0:01:13.250 --> 0:01:14.069 know where it came from. 0:01:14.090 --> 0:01:18.399 Well, actually, Plus AI managed to recover the trailers and 0:01:18.540 --> 0:01:20.380 all 40,000 tons of sand. 0:01:20.400 --> 0:01:21.060 So happy ending. 0:01:21.140 --> 0:01:21.740 Happy ending. 0:01:26.950 --> 0:01:27.729 Welcome to Tech Stuff. 0:01:27.790 --> 0:01:30.050 I'm Oz Veloshian, and this is The Week in Tech, 0:01:30.150 --> 0:01:32.850 where I'm joined by the world's most plugged-in reporters to. 0:01:32.810 --> 0:01:35.229 Break down what's really happening in tech right now. 0:01:35.860 --> 0:01:38.940 Today, we're joined by Reid Abregotti, tech editor for Semaphore 0:01:39.140 --> 0:01:41.480 and host of the new podcast, as of this week, 0:01:41.560 --> 0:01:45.420 Future Proof, Taylor Lorenz of UserMag, and back from leave, 0:01:45.660 --> 0:01:48.300 Kyle Chaker, who writes the Infinite Scroll column for The 0:01:48.360 --> 0:01:48.740 New Yorker. 0:01:48.780 --> 0:01:49.260 Welcome, all. 0:01:49.480 --> 0:01:50.340 Thanks for having us. 0:01:50.620 --> 0:01:52.610 Reid, welcome to the podcasting ranks. 0:01:53.570 --> 0:01:54.090 Can't hear you, Reid. 0:01:54.110 --> 0:01:54.760 Oh, we can't hear you? 0:01:57.000 --> 0:01:58.440 First you have to turn your mic on. 0:01:59.500 --> 0:02:01.020 I'm learning all the rules now. 0:02:01.040 --> 0:02:04.840 Don't mute your mic. 0:02:05.020 --> 0:02:07.520 Well, starting off with you, I mean, first, just give 0:02:07.560 --> 0:02:09.040 us a brief plug for the new podcast, but then 0:02:09.080 --> 0:02:10.400 I want to ask you what's going on in space. 0:02:10.700 --> 0:02:13.610 Yeah, well, it's called Future Proof. And, you know, we're 0:02:13.630 --> 0:02:16.790 going to interview people from all areas of the tech 0:02:16.830 --> 0:02:19.840 industry and go deep on on topics and try to 0:02:19.980 --> 0:02:23.960 teach people something. It'll be newsy, journalistic, but also really 0:02:24.000 --> 0:02:26.780 go deep on stuff and help listeners understand. 0:02:26.800 --> 0:02:28.160 I'm excited to check it out. 0:02:29.000 --> 0:02:30.900 This week's story for you feels a bit like the 0:02:31.000 --> 0:02:35.400 plot of a 90s Bond movie involving the US competing 0:02:35.440 --> 0:02:40.200 with Russia and China to build nuclear powered permanent bases. 0:02:39.880 --> 0:02:42.359 On the moon. What's going on? I mean, is this 0:02:42.380 --> 0:02:43.209 like a fantasy or? 0:02:44.410 --> 0:02:47.169 It depends on who you ask. I personally like, you know, 0:02:47.230 --> 0:02:50.730 I think everything in the space industry feels like fantasy 0:02:50.810 --> 0:02:53.310 until it actually happens. I remember when I first came 0:02:53.370 --> 0:02:57.169 out to the tech industry, SpaceX was sort of just 0:02:57.270 --> 0:03:01.650 becoming a thing. And people were talking about satellites, you know, 0:03:01.690 --> 0:03:06.450 in space providing internet. And really smart people told me 0:03:06.490 --> 0:03:10.030 that was completely impossible. It would never happen. Now I own, 0:03:10.210 --> 0:03:13.010 you know, two Starlinks. Um, So I just take the 0:03:13.070 --> 0:03:15.750 timelines for these things and I just extend them in 0:03:15.790 --> 0:03:18.750 my mind, but I don't dismiss anything outright. I mean, 0:03:18.870 --> 0:03:22.050 nothing here is totally impossible, right? But, you know, does 0:03:22.110 --> 0:03:25.460 it make economic sense? Will it make economic sense in 0:03:25.480 --> 0:03:27.820 the near future? That I don't, I just don't know. 0:03:28.180 --> 0:03:29.880 I think the US has tried it once before and 0:03:29.940 --> 0:03:34.320 Russia has been sending nuclear reactors into space since the 60s. 0:03:34.360 --> 0:03:37.560 But somehow this kind of new space race thing is 0:03:37.600 --> 0:03:38.470 particularly fascinating. 0:03:38.510 --> 0:03:39.450 Kyle, have you been following it? 0:03:40.050 --> 0:03:42.850 Just as far as the data centers in space. I mean, 0:03:42.880 --> 0:03:46.920 I think that whole concept is really appealing to me. 0:03:47.340 --> 0:03:49.740 And it kind of seems like the only answer. If 0:03:50.340 --> 0:03:52.780 we're kind of having all these fights over data centers 0:03:53.060 --> 0:03:55.780 on Earth, you know, maybe it would be better to 0:03:55.820 --> 0:03:58.800 put them in space and have them powered in some 0:03:58.840 --> 0:03:59.260 other way. 0:04:00.490 --> 0:04:01.090 So yeah, I don't know. 0:04:01.190 --> 0:04:04.090 I think it reads right about the timelines extending longer 0:04:04.150 --> 0:04:04.950 than you might think. 0:04:05.270 --> 0:04:07.720 But I mean, Kyle, do you know what a space... 0:04:07.750 --> 0:04:10.630 I've actually looked into this recently. A space data center 0:04:10.750 --> 0:04:14.550 is not like a floating data center. It is basically, 0:04:15.210 --> 0:04:18.529 you have a bunch of satellites that have a GPU 0:04:18.589 --> 0:04:22.460 or two or a rack, and they're spread out very 0:04:22.560 --> 0:04:25.839 far apart, and they're connected by lasers. So they communicate 0:04:25.880 --> 0:04:28.980 with each other by lasers instead of on Earth through 0:04:29.040 --> 0:04:32.390 fiber optic cable or copper, right? And because. 0:04:32.900 --> 0:04:35.420 It's not a Death Star. It's not like one floating 0:04:35.500 --> 0:04:36.779 massive data center. 0:04:36.920 --> 0:04:37.140 Right. 0:04:37.200 --> 0:04:40.500 But because, you know, there's no air in space, you 0:04:40.520 --> 0:04:42.420 can't cool these things in the same way you cool 0:04:42.440 --> 0:04:45.980 them on Earth. So the idea is that you'll somehow 0:04:47.040 --> 0:04:50.550 pull the heat away from these GPUs and dissipate it 0:04:50.750 --> 0:04:54.030 on a massive... They're going to essentially like a massive 0:04:54.130 --> 0:04:56.429 heat sink that's about the size of a tennis court. 0:04:56.529 --> 0:04:59.790 I mean, it is... like really kind of just the 0:04:59.850 --> 0:05:03.990 craziest concept you can possibly imagine. But it's not impossible. 0:05:04.110 --> 0:05:06.490 I mean, but I don't know if I agree that 0:05:06.550 --> 0:05:08.450 that's going to be the thing that happens before we 0:05:08.470 --> 0:05:10.550 just get our act together on Earth and build some 0:05:10.570 --> 0:05:11.410 data centers already. 0:05:11.790 --> 0:05:14.510 Yeah, I mean, the concept of space junk is kind 0:05:14.560 --> 0:05:16.710 of interesting. I mean, sending up this tennis court, or 0:05:16.740 --> 0:05:20.640 I think in some cases even football pitch or field size, 0:05:20.720 --> 0:05:24.360 kind of non-convective heat dispersing panels into space. I mean, 0:05:24.740 --> 0:05:27.020 you have to wonder... what the implications of all that 0:05:27.060 --> 0:05:29.820 stuff floating around in orbit are. And I think some 0:05:29.860 --> 0:05:31.720 of this, when I first saw the story, the headline 0:05:31.740 --> 0:05:34.320 about the nuclear reactors in space, I kind of assumed 0:05:34.380 --> 0:05:37.320 somehow it had to do with power generation for AI. 0:05:37.360 --> 0:05:39.320 But to my great surprise, it wasn't about that. It 0:05:39.360 --> 0:05:41.059 was about, go ahead, Taylor. 0:05:41.420 --> 0:05:44.279 I'm going to start a business that is a space 0:05:44.420 --> 0:05:47.960 junk removal cleanup business, because I think this is going 0:05:47.990 --> 0:05:51.160 to be huge business Going forward, I mean, we're about 0:05:51.180 --> 0:05:53.940 to have something like a million satellites or Elon wants 0:05:53.960 --> 0:05:57.260 to scale to a million Starlink satellites. And I think 0:05:57.380 --> 0:05:59.980 if we put data centers up there too, yeah, there's 0:06:00.020 --> 0:06:02.700 just going to be a lot of debris. And I 0:06:02.720 --> 0:06:04.400 think that poses risks. 0:06:05.000 --> 0:06:06.060 Especially when it's nuclear. 0:06:07.490 --> 0:06:08.010 I mean, right. 0:06:08.029 --> 0:06:10.290 It's kind of, well, nuclear doesn't have that much risk 0:06:10.330 --> 0:06:12.830 in space. I mean, you know, nuclear meltdown in space isn't... 0:06:12.850 --> 0:06:16.080 Unless it falls to Earth. Well, yeah, it'll burn up 0:06:16.160 --> 0:06:17.980 before it falls to earth. But I mean, what happens 0:06:18.000 --> 0:06:21.180 to these giant, you know, heat dissipating football field or 0:06:21.220 --> 0:06:23.910 tennis court size things if like a little, a little 0:06:24.029 --> 0:06:26.310 speck runs into it and puts a hole in it? Like, 0:06:26.350 --> 0:06:28.950 does it still work? And then if something breaks, they 0:06:28.970 --> 0:06:31.360 just drop it down and let it burn in the atmosphere, 0:06:31.380 --> 0:06:34.650 which is I'm sure horrifying to people who are here 0:06:34.710 --> 0:06:37.220 trying to get their hands on like a single GPU 0:06:37.279 --> 0:06:40.760 right now. But I thought the nuclear reactor thing is like, 0:06:40.800 --> 0:06:42.980 you got to build those on the moon, right? I mean, 0:06:43.060 --> 0:06:45.100 that was what you were talking about. Because if you 0:06:45.140 --> 0:06:48.350 want to send Mars missions... Staging post. 0:06:48.760 --> 0:06:49.110 Yeah. 0:06:49.170 --> 0:06:50.930 Is that what... That's what you were talking about, right? 0:06:50.950 --> 0:06:51.750 Yeah, exactly. 0:06:51.810 --> 0:06:55.210 And so the current like solar power cannot generate enough 0:06:55.290 --> 0:07:00.010 power on the moon to have actually warm habitable bases. 0:07:01.100 --> 0:07:01.250 Right. 0:07:01.270 --> 0:07:01.359 So. 0:07:01.990 --> 0:07:04.550 And China and Russia are collaborating. Russia have the best 0:07:04.730 --> 0:07:08.869 space nuclear technology. China, I guess, have good rocketry. So 0:07:08.970 --> 0:07:10.310 the new Cold War in space is on. 0:07:10.330 --> 0:07:13.530 Do you guys watch For All Mankind? Just out of curiosity, 0:07:14.230 --> 0:07:17.550 did you watch that show? No, I missed it. I mean, 0:07:17.570 --> 0:07:20.270 I think if you're interested in this topic, you've got 0:07:20.290 --> 0:07:24.150 to watch For All Mankind. Just go. It's on Apple TV. 0:07:24.190 --> 0:07:28.580 Go watch it. It's a sort of alternative history show 0:07:28.760 --> 0:07:31.720 about what would have happened if the if the Cold 0:07:31.760 --> 0:07:34.940 War sort of evolved a little bit differently and space 0:07:35.020 --> 0:07:37.300 just continued to be the thing that we put all 0:07:37.340 --> 0:07:40.740 of our, you know, human effort and ingenuity into. And like, 0:07:40.840 --> 0:07:42.820 I don't want to spoil the show for you guys, 0:07:42.880 --> 0:07:45.970 but you know, a lot of interesting stuff happens in 0:07:46.050 --> 0:07:48.350 space on a timeline that is not, that is a 0:07:48.370 --> 0:07:51.170 lot more accelerated than we have today. It's a really 0:07:51.230 --> 0:07:54.300 fun show. I think it's, you know, I've enjoyed it. 0:07:54.870 --> 0:07:56.650 Do we need more space propaganda? 0:07:56.680 --> 0:08:01.900 Like pro-space propaganda? pro-space content to encourage us all. 0:08:02.360 --> 0:08:05.140 It beats the propaganda we have right now. I mean, 0:08:05.160 --> 0:08:09.640 I'll take that over all the current propaganda. 0:08:09.800 --> 0:08:12.500 I've got a NASA sticker on my computer. What's the 0:08:12.540 --> 0:08:14.520 take on Jared Eisenman? I mean, the head of NASA, 0:08:14.560 --> 0:08:18.620 Trump had initially rejected him during the Trump-Elon Fodor, but 0:08:18.660 --> 0:08:22.860 then changed his mind and Eisenman was confirmed, obviously, a 0:08:22.960 --> 0:08:26.240 private sector billionaire. What's he doing at NASA? And what's 0:08:26.260 --> 0:08:29.000 the public-private relationship in space in the US today, Reid? 0:08:29.680 --> 0:08:32.940 Well, I mean, he's certainly helping Elon, right? There was 0:08:32.960 --> 0:08:36.040 a lot of tension with Elon and the Biden administration. 0:08:36.120 --> 0:08:40.660 And I think Jared's helping Elon. There's less tension between 0:08:40.720 --> 0:08:43.500 SpaceX and NASA. But, you know, I haven't heard any 0:08:43.559 --> 0:08:46.420 sort of controversies about him. He seems to be, you know, 0:08:46.520 --> 0:08:49.260 doing a fine job. We'll see what the effects of 0:08:49.360 --> 0:08:53.000 his term are. But I mean, the general trend, I think, 0:08:53.080 --> 0:08:55.200 in NASA is like more and more of this is 0:08:55.240 --> 0:08:59.420 becoming privatized. You know, there's good things and bad things 0:08:59.460 --> 0:09:01.900 about that, right? I think the public interest mission of 0:09:02.000 --> 0:09:03.980 space that I think we're all used to from our 0:09:04.020 --> 0:09:06.360 childhood of, you know, space shuttle going up and doing 0:09:06.380 --> 0:09:09.520 all these science experiments. I mean, some of that is 0:09:09.580 --> 0:09:11.800 for sure going to be going to be lost when 0:09:11.830 --> 0:09:13.689 the private sector sort of takes over. And I think 0:09:13.710 --> 0:09:16.710 that's kind of sad. But at the same time, it's incredible, 0:09:16.830 --> 0:09:20.110 the stuff that has been accomplished. I think it's amazing, 0:09:20.170 --> 0:09:23.010 not to mention just having a Starlink on my camper 0:09:23.030 --> 0:09:25.250 van in California. That's like the ultimate life. 0:09:25.370 --> 0:09:29.220 So I do want to just note that on the 0:09:29.340 --> 0:09:32.840 kind of safety issue with the nuclear reactors in space, 0:09:32.880 --> 0:09:37.040 there was a Russian nuclear reactor thing that came down 0:09:37.100 --> 0:09:40.280 to Earth. And US intelligence say that five people were killed. 0:09:40.300 --> 0:09:43.220 The Russians deny this. But just before we go to 0:09:43.240 --> 0:09:45.239 the break, Taylor, I mean, this idea of kind of 0:09:45.380 --> 0:09:49.260 cleaning space junk and including nuclear space junk, have you 0:09:49.280 --> 0:09:50.679 met any companies in this space? 0:09:51.620 --> 0:09:54.250 Yeah, I talked to Orbital, which is, you know, seeking 0:09:54.280 --> 0:09:58.350 to put the data centers in space recently and visited 0:09:58.450 --> 0:10:01.940 their offices in LA. So I do think that there's 0:10:01.980 --> 0:10:03.480 sort of like a space race with a lot of 0:10:03.520 --> 0:10:07.240 different startups right now trying to build out infrastructure in 0:10:07.260 --> 0:10:07.700 that realm. 0:10:07.720 --> 0:10:09.370 What did you learn at Orbital? 0:10:09.800 --> 0:10:11.810 What they're doing is cool. I mean, I think it's 0:10:11.870 --> 0:10:13.830 just like they're up against a lot of competition, as 0:10:13.890 --> 0:10:16.849 Reid said, too. It's like everything seems impossible until it 0:10:17.300 --> 0:10:19.720 somebody does it. And I personally think that the data 0:10:19.740 --> 0:10:22.100 center fight is not going away anytime soon, and we're 0:10:22.120 --> 0:10:23.860 going to need more compute. So yeah, if they can 0:10:23.880 --> 0:10:24.559 put them in space. 0:10:25.160 --> 0:10:25.520 Why not? 0:10:26.059 --> 0:10:28.800 Well, what's interesting about Orbital is the founder was the 0:10:28.870 --> 0:10:32.830 scooter guy, right? So he had a scooter startup in 0:10:32.870 --> 0:10:36.250 San Francisco. And then he kind of worked for Ford, 0:10:36.290 --> 0:10:36.650 I think. 0:10:36.710 --> 0:10:37.330 And then he. 0:10:37.470 --> 0:10:38.170 He sold it. 0:10:38.250 --> 0:10:41.170 And now he's like, I'm going to, I'm going to 0:10:41.210 --> 0:10:43.460 just make a space data center company because that's the 0:10:43.540 --> 0:10:44.540 obvious next move. 0:10:44.940 --> 0:10:48.080 Well, the reason why is because he has a lot 0:10:48.100 --> 0:10:51.069 of connection in manufacturing and in China. So I do 0:10:51.110 --> 0:10:55.030 think that like. building a scooter company ironically kind of 0:10:55.070 --> 0:10:58.550 prepared him to just do some sort of hardware startup generally. 0:10:58.570 --> 0:11:01.429 Yeah, I was being tongue-in-cheek, but actually I think that 0:11:01.530 --> 0:11:03.710 there is a point to be made, which is that 0:11:03.950 --> 0:11:08.050 the whole idea of SpaceX and Blue Origin and these 0:11:08.090 --> 0:11:12.250 companies to get infrastructure up into space is that you 0:11:12.270 --> 0:11:16.230 sort of enable a new crop of entrepreneur to work 0:11:16.309 --> 0:11:20.699 on space-related ideas and companies and So it's like, if 0:11:20.740 --> 0:11:22.819 you think of it as like all this infrastructure as 0:11:22.860 --> 0:11:27.250 a platform, it lets this guy who was a scooter entrepreneur, 0:11:27.640 --> 0:11:29.810 all of a sudden be a space entrepreneur, which never 0:11:29.850 --> 0:11:32.069 would have been possible if you didn't have the, you know, 0:11:32.110 --> 0:11:35.570 SpaceX rockets to get his equipment up into space, right? 0:11:35.630 --> 0:11:38.150 That he was never going to start a rocket company, 0:11:38.610 --> 0:11:40.420 but now he can start a space company. And I 0:11:40.460 --> 0:11:43.640 think that's, I think that's actually kind of a profound point. He, 0:11:43.740 --> 0:11:45.219 I talked to him too. He doesn't even want to 0:11:45.260 --> 0:11:46.679 go to space. I was like, do you want to 0:11:46.700 --> 0:11:48.099 go to, he's like, no, I have no interest in 0:11:48.120 --> 0:11:52.390 going to space. I'm like, that's, I think that's pretty profound. 0:11:52.990 --> 0:11:55.610 I mean, being in space kind of sucks. It's cold 0:11:55.710 --> 0:11:57.830 up there. It's like living in a submarine. 0:11:58.050 --> 0:12:00.449 Until you get the nuclear reactors on the moon. 0:12:00.470 --> 0:12:02.030 Right, right. 0:12:02.070 --> 0:12:04.250 To warm your hands by the warmth of a nuclear 0:12:04.270 --> 0:12:06.250 reactor on the moon. I have to say, there's always 0:12:06.270 --> 0:12:09.390 that classic question, what would you like the first line 0:12:09.410 --> 0:12:11.590 of your obituary to be? And I guess he was 0:12:11.630 --> 0:12:14.110 never going to build rockets is one of them. When 0:12:14.630 --> 0:12:18.930 we come back, who pays when chatbots lead you astray? 0:12:31.490 --> 0:12:31.890 Welcome back. 0:12:32.420 --> 0:12:34.510 Taylor, there was a piece in Futurism this week that 0:12:34.890 --> 0:12:37.429 there's a kind of golden oldie quality to it, a 0:12:37.470 --> 0:12:39.970 kind of tech made me do it story. I remember 0:12:39.990 --> 0:12:43.410 there was that episode of The Office where Michael Scott 0:12:43.710 --> 0:12:46.870 drives his car into the lake on purpose because GPS 0:12:46.929 --> 0:12:51.150 told him to. But nowadays, it's not GPS, but AI 0:12:51.210 --> 0:12:55.460 and cloud in particular that's leading not drivers, but hikers astray. 0:12:56.679 --> 0:12:58.980 Yeah, I actually read about this in the New York Times. 0:12:59.520 --> 0:13:02.510 They did a good story on it. But basically, this 0:13:02.590 --> 0:13:06.229 guy was hiking in British Columbia and he went off 0:13:06.290 --> 0:13:10.570 the trail. He used Claude as part of his planning 0:13:10.610 --> 0:13:13.580 process and ultimately had to be rescued off the side 0:13:13.600 --> 0:13:15.780 of a cliff because he was not a qualified enough 0:13:15.820 --> 0:13:18.400 hiker to, you know, to get himself out of the 0:13:18.440 --> 0:13:21.840 situation he had encountered. used. I don't think this is 0:13:21.900 --> 0:13:26.340 an AI problem necessarily. I actually ended up literally last 0:13:26.380 --> 0:13:28.620 week or two weekends ago when I was back in LA, 0:13:28.640 --> 0:13:30.740 I went to go on a hike and ended up 0:13:31.800 --> 0:13:35.600 in this like burned out section of Altadena where there 0:13:35.640 --> 0:13:39.240 was like no water, the trail was a mess. And 0:13:39.300 --> 0:13:42.400 I had read about it online and And even like 0:13:42.460 --> 0:13:44.880 looked it up, you know, ahead of time. And it's 0:13:44.900 --> 0:13:46.920 just that that route after the fire had changed a 0:13:47.100 --> 0:13:49.830 year ago and all this stuff. Of course, my friends 0:13:49.870 --> 0:13:52.650 were like, why wouldn't you use all trails? I think 0:13:52.790 --> 0:13:55.630 all trails is sort of the standard for this. And 0:13:55.670 --> 0:13:58.689 if you go anywhere else outside the internet, you're sort 0:13:58.730 --> 0:13:59.650 of risking it. 0:14:00.450 --> 0:14:03.189 Kyle, you nodded sagely about all trails. You seem like 0:14:03.250 --> 0:14:04.050 a trail guy. 0:14:04.929 --> 0:14:09.770 I'm not, which is the whole problem. I am not 0:14:09.809 --> 0:14:11.900 a hiking guy. I think I would end up in 0:14:11.940 --> 0:14:14.800 a bad situation. But I think if you have ever 0:14:14.940 --> 0:14:19.160 explored all trails, you know why relying on AI for 0:14:19.180 --> 0:14:23.450 your hiking knowledge is a terrible idea. All trails is 0:14:23.510 --> 0:14:28.810 the most human... intricate detailed set of commentary and data 0:14:28.910 --> 0:14:31.460 that you can possibly have i think in an app 0:14:31.860 --> 0:14:34.900 because it'll be like okay a hundred feet to your 0:14:34.960 --> 0:14:37.720 right like don't take this trail look for the rock 0:14:37.780 --> 0:14:40.479 and go the other way or if you aren't an 0:14:40.540 --> 0:14:44.940 experienced hiker maybe don't even bother uh so like all 0:14:44.960 --> 0:14:48.610 trails i don't know i want a person directly to 0:14:48.670 --> 0:14:51.300 tell me in their own words what I should and 0:14:51.320 --> 0:14:53.060 shouldn't do on one of these hiking trails. 0:14:54.420 --> 0:14:55.800 And a person recently. 0:14:55.820 --> 0:14:57.980 Yeah, yeah, it changes all the time. 0:14:58.460 --> 0:15:00.450 Like the hike that I went on a couple weeks ago, 0:15:00.490 --> 0:15:03.489 like it was rated really, people had blogged this, you know, 0:15:03.510 --> 0:15:08.090 like it was just not up-to-date information. And so, yeah, 0:15:08.150 --> 0:15:11.120 I just think that like, There's a lot of people 0:15:11.170 --> 0:15:14.620 interested in outdoors, but you need to rely on, this 0:15:14.680 --> 0:15:17.040 is like an all trails ad. Like, I think like, 0:15:17.060 --> 0:15:19.890 you know, he could have not used Claude and I 0:15:19.930 --> 0:15:21.750 don't know that he wouldn't have ended up in the 0:15:21.770 --> 0:15:24.420 same situation, especially when you're going off the trail, you know? 0:15:25.110 --> 0:15:27.250 He got rescued, sorry to interrupt, and the rescue manager said, 0:15:27.310 --> 0:15:29.410 Claude has never been in that specific area, has never 0:15:29.430 --> 0:15:31.610 been on that route, and has never been in the outdoors. 0:15:32.700 --> 0:15:36.200 Is this like a sticking rocks to the pizza with 0:15:36.300 --> 0:15:37.380 glue type story? 0:15:37.440 --> 0:15:39.600 Or is it like, what's the, not so much, Taylor? 0:15:39.820 --> 0:15:42.440 Yeah, and that was so silly, too. Like, I mean, 0:15:42.480 --> 0:15:45.140 I guess, like, especially Futurism is a website that, like, 0:15:45.200 --> 0:15:48.260 loves to write up this stuff as, like, it's just 0:15:48.390 --> 0:15:50.870 rage bait content to kind of be like, ha ha, 0:15:50.910 --> 0:15:53.930 you idiot, you know, look at you. And to me, 0:15:53.970 --> 0:15:56.410 it's like, look, it's a real problem with accuracy. I 0:15:56.450 --> 0:16:00.430 think this shows the value actually of these human curated, 0:16:00.510 --> 0:16:02.930 like very human updated apps. Like that's what it shows 0:16:02.980 --> 0:16:05.380 the value of. It's not that AI is inherently unreliable. 0:16:05.440 --> 0:16:08.280 I would say 99.999% of things that you ask AI 0:16:08.320 --> 0:16:10.840 and including that, like Google, you know, Gemini, I guess, 0:16:10.900 --> 0:16:13.440 review that had suggested you put glue on your pizza. 0:16:13.480 --> 0:16:15.700 And that was from a Reddit thread, you know, Early 0:16:15.720 --> 0:16:19.420 iterations of AI were certainly less reliable. They're getting smarter 0:16:19.440 --> 0:16:22.540 and smarter. Again, it just has to do with the 0:16:22.590 --> 0:16:24.880 fact that when you're going out into the physical world, 0:16:25.020 --> 0:16:29.390 hiking specifically, until we get drones, you know, going out 0:16:29.470 --> 0:16:33.090 and scouting these trails for us, I think, like, something 0:16:33.110 --> 0:16:34.630 like all trails is just the way to go. 0:16:34.650 --> 0:16:36.550 Or the humanoid robots. 0:16:37.030 --> 0:16:41.030 Right, we send the humanoid robots out to go chart 0:16:41.070 --> 0:16:41.510 the trails. 0:16:41.750 --> 0:16:44.430 But it was striking to me, like, Hiking is a 0:16:44.510 --> 0:16:50.330 subjective thing. Difficulty is subjective. How much your body can 0:16:50.370 --> 0:16:54.050 take is subjective. AI, if you're not an experienced hiker, 0:16:54.150 --> 0:16:55.830 AI is not going to be able to tell you 0:16:56.130 --> 0:16:59.200 what you can handle in this situation. So maybe in 0:16:59.240 --> 0:17:02.740 this case, don't ask Claude, ask your friends. 0:17:03.480 --> 0:17:05.520 In my experience, I don't know if you guys go 0:17:05.560 --> 0:17:09.119 backpacking and stuff, but if something really bad doesn't happen 0:17:09.160 --> 0:17:11.220 where you think you might die, it's kind of a 0:17:11.280 --> 0:17:13.720 boring trip. Those are the things that you always remember, 0:17:13.960 --> 0:17:14.240 you know? 0:17:14.260 --> 0:17:15.080 I think. 0:17:15.119 --> 0:17:16.939 Okay, I am not trying to go on the hikes 0:17:16.960 --> 0:17:20.800 that Reid is going on. I love hiking a leisurely 0:17:21.140 --> 0:17:24.119 trail in the woods. I think also, you know, this 0:17:24.160 --> 0:17:26.280 is a teenager. I think he was 16 years old. 0:17:26.800 --> 0:17:30.040 So I think there's also just teenagers don't traditionally have 0:17:30.080 --> 0:17:34.399 a great track record of judging risk and... Potential for 0:17:34.440 --> 0:17:35.100 bodily harm. 0:17:35.119 --> 0:17:37.400 Yeah, people have always been doing dumb stuff. 0:17:37.840 --> 0:17:41.179 There is something in the automation bias, like AI psychosis 0:17:41.240 --> 0:17:45.830 adjacent thing here. I got my green fingers burnt earlier 0:17:45.869 --> 0:17:48.090 this year when I really went to town on pruning 0:17:48.130 --> 0:17:50.609 my roses with the encouragement of Gemini. And actually, I 0:17:50.630 --> 0:17:52.129 think I ended up killing them and I had to 0:17:52.170 --> 0:17:55.169 confess to the landlord. So I feel like you can 0:17:55.210 --> 0:17:58.129 lean in too hard to the reassuring advice. 0:17:57.869 --> 0:17:58.610 Of AI, certainly. 0:17:59.490 --> 0:18:01.649 The story that kind of sits behind this story, for me, though, 0:18:01.670 --> 0:18:06.879 is the kind of AI... responsibility and liability story. Taylor, 0:18:06.920 --> 0:18:08.520 I think this has been also in the news this 0:18:08.560 --> 0:18:13.380 week in terms of how insurance payments work out when 0:18:13.560 --> 0:18:17.740 AI does harm and who's held liable, like AI hugging 0:18:17.880 --> 0:18:20.780 face episode, for example, like in that case, hugging face 0:18:20.800 --> 0:18:23.220 didn't sue OpenAI, but if they had, who would have 0:18:23.240 --> 0:18:24.600 had the liability in that situation? 0:18:25.160 --> 0:18:29.399 Well, it's a company hacking another company. So the company 0:18:29.440 --> 0:18:32.600 has a liability. It intentionally hacked the other company. 0:18:33.450 --> 0:18:36.419 Yeah, exactly. I think the question though, there is an, 0:18:37.030 --> 0:18:39.170 I think there is a really interesting legal question that's 0:18:39.190 --> 0:18:41.600 going to have to be answered, which is like, let's 0:18:41.640 --> 0:18:44.899 say you're not open AI and there's got to be 0:18:44.920 --> 0:18:48.720 some cases where you could end up accidentally doing harm, right? 0:18:48.740 --> 0:18:50.959 And then I think- Here's the question, I think, or 0:18:50.980 --> 0:18:52.860 not to interrupt, but just to finish, or I think 0:18:52.880 --> 0:18:54.340 this was like the second part maybe that you wanted 0:18:54.359 --> 0:18:56.379 to get at, but like, here's the question that I 0:18:56.400 --> 0:18:58.300 think we're going to have to answer is when we 0:18:58.340 --> 0:19:00.900 move into this more agentic era of the internet, who 0:19:00.980 --> 0:19:04.580 is responsible for the agent's actions? And with OpenAI, it 0:19:04.630 --> 0:19:06.619 is literally their model. Of course, they're responsible. This is 0:19:06.680 --> 0:19:09.560 a test that they were doing. But if I use Grokbot, 0:19:09.600 --> 0:19:12.709 for instance, to go out into the world and do 0:19:12.730 --> 0:19:14.929 a lot of things for me, automate my lifestyle, and 0:19:14.970 --> 0:19:18.470