1 00:00:00,520 --> 00:00:03,240 Speaker 1: Already and this is this is the Daily OS. 2 00:00:03,400 --> 00:00:04,360 Speaker 2: This is the Daily OS. 3 00:00:05,160 --> 00:00:15,880 Speaker 1: Oh, now it makes sense. Good morning and welcome to 4 00:00:15,960 --> 00:00:18,600 Speaker 1: the Daily OS. It's Monday, the twenty third of March. 5 00:00:18,680 --> 00:00:22,720 Speaker 1: I'm Lucy Tassel, I'm Tam Kazlowski. The developer of POKEMONO 6 00:00:23,040 --> 00:00:26,160 Speaker 1: has revealed a new use for the billions of images 7 00:00:26,200 --> 00:00:29,480 Speaker 1: captured by its users over the years, and that's AI 8 00:00:29,640 --> 00:00:33,520 Speaker 1: mapping technology. Earlier this month, the company announced the tech 9 00:00:33,640 --> 00:00:37,400 Speaker 1: will be used to help delivery robots navigate city streets. 10 00:00:37,720 --> 00:00:41,840 Speaker 1: The announcement has raised questions about the ethics of how 11 00:00:41,880 --> 00:00:46,040 Speaker 1: the data was collected, including the nature of player consent. 12 00:00:46,600 --> 00:00:49,360 Speaker 1: We're going to get into all of that in today's episode. 13 00:00:49,560 --> 00:00:52,840 Speaker 1: But first, Sam, what do you remember about the release 14 00:00:52,880 --> 00:00:55,480 Speaker 1: of Pokemon Go way back in twenty sixteen? 15 00:00:55,960 --> 00:00:59,080 Speaker 3: Yeah? Wow, twenty sixteen. It was this era of watching 16 00:00:59,120 --> 00:01:02,320 Speaker 3: people look down at their phones. The streets of Sydney 17 00:01:02,360 --> 00:01:05,839 Speaker 3: were filled with people trying to find Pokemon on the street, 18 00:01:06,120 --> 00:01:08,880 Speaker 3: and it was quite remarkable. I'd never played it. Not 19 00:01:08,959 --> 00:01:11,880 Speaker 3: a big video game person, I reckon you would have 20 00:01:11,880 --> 00:01:13,040 Speaker 3: played it, am I right. 21 00:01:13,000 --> 00:01:15,640 Speaker 1: I did, but not because I have become a video 22 00:01:15,720 --> 00:01:18,200 Speaker 1: game person as I get older, but back then I 23 00:01:18,360 --> 00:01:20,960 Speaker 1: genuinely was just interested in whatever the big trend was. 24 00:01:21,400 --> 00:01:23,640 Speaker 1: And I also I was in first year UNI doing 25 00:01:23,640 --> 00:01:26,080 Speaker 1: a lot of kind of general media courses and it 26 00:01:26,120 --> 00:01:29,039 Speaker 1: was coming up a lot in my classes. So I 27 00:01:29,400 --> 00:01:33,200 Speaker 1: was one of the five hundred million people worldwide who 28 00:01:33,280 --> 00:01:36,880 Speaker 1: downloaded Pokemon Go in the first sixty days after it 29 00:01:36,959 --> 00:01:38,720 Speaker 1: was released in the middle of twenty sixteen. 30 00:01:38,920 --> 00:01:42,160 Speaker 3: I remember just the number of injuries that were happening 31 00:01:42,200 --> 00:01:44,840 Speaker 3: because of Pokemon Go. I've just found actually from twenty 32 00:01:44,880 --> 00:01:47,960 Speaker 3: sixteen a statement from the National Safety Council in the 33 00:01:48,040 --> 00:01:52,240 Speaker 3: US talking about distracted driving distracted walking eleven thousand injuries 34 00:01:52,240 --> 00:01:55,160 Speaker 3: in the US alone in six months from Pokemon Go. 35 00:01:55,480 --> 00:01:56,919 Speaker 3: That was a brutal sport. 36 00:01:57,160 --> 00:01:59,760 Speaker 1: Yeah. Yeah, there was a setting where I used to 37 00:01:59,800 --> 00:02:01,640 Speaker 1: play on the bus on my way home from UNI, 38 00:02:01,680 --> 00:02:03,080 Speaker 1: and there was a setting where I would sort of 39 00:02:03,160 --> 00:02:05,600 Speaker 1: say like, are you driving, Please, don't be driving while 40 00:02:05,600 --> 00:02:08,200 Speaker 1: you're playing this. But I wasn't driving, so I could 41 00:02:08,240 --> 00:02:10,440 Speaker 1: catch all the Pokemon along the side of Paramatta Road. 42 00:02:10,360 --> 00:02:12,440 Speaker 3: And just remind me of the mechanics of the game. 43 00:02:12,639 --> 00:02:15,680 Speaker 1: Yeah. So basically the key aspect that made it different 44 00:02:15,680 --> 00:02:18,560 Speaker 1: from other phone games of the era. It's an element 45 00:02:18,600 --> 00:02:22,160 Speaker 1: called augmented reality or a R. So I'm going to 46 00:02:22,200 --> 00:02:25,079 Speaker 1: try and pronounce that carefully because we're also talking about AI. 47 00:02:25,840 --> 00:02:29,840 Speaker 1: A R games use your device's camera and screen to 48 00:02:29,960 --> 00:02:33,680 Speaker 1: layer graphics over the real world, and basically the game 49 00:02:33,800 --> 00:02:37,119 Speaker 1: asks users to grant access to their location, and once 50 00:02:37,160 --> 00:02:40,480 Speaker 1: it has that location, it'll show you a graphic overlay 51 00:02:40,480 --> 00:02:44,800 Speaker 1: map of where you currently are and different Pokemon those 52 00:02:44,919 --> 00:02:48,360 Speaker 1: you know, the little the pocket monsters developed by Nintendo 53 00:02:49,240 --> 00:02:54,120 Speaker 1: around your location, as well as something called Poke Stop. 54 00:02:54,120 --> 00:02:56,720 Speaker 1: We'll get into that later, but the key thing is 55 00:02:57,080 --> 00:02:59,960 Speaker 1: it's all location based. So when I opened it up 56 00:03:00,080 --> 00:03:02,760 Speaker 1: again on my phone this morning, it showed me like, 57 00:03:02,840 --> 00:03:05,640 Speaker 1: there's a squirddle around the corner. If I go further afield, 58 00:03:05,720 --> 00:03:08,600 Speaker 1: I can find different Pokemon. And it was developed by 59 00:03:08,600 --> 00:03:13,000 Speaker 1: a company called Neantik in collaboration with Nintendo, which obviously 60 00:03:13,040 --> 00:03:13,760 Speaker 1: owns Pokemon. 61 00:03:14,000 --> 00:03:16,920 Speaker 3: And so just to clarify that is a live game today, 62 00:03:17,160 --> 00:03:19,560 Speaker 3: there's still players, there's still communities around the world playing 63 00:03:19,560 --> 00:03:20,080 Speaker 3: Pokemon Go. 64 00:03:20,360 --> 00:03:24,360 Speaker 1: Yeah, as recently as twenty twenty four, there were about 65 00:03:24,360 --> 00:03:27,240 Speaker 1: one hundred million active users of the game, so a 66 00:03:27,280 --> 00:03:30,360 Speaker 1: fifth of the original downloaders. But still a huge number 67 00:03:30,360 --> 00:03:31,359 Speaker 1: of people, and so. 68 00:03:31,320 --> 00:03:35,840 Speaker 3: It's collecting you said location data. It's also collecting some images. 69 00:03:35,880 --> 00:03:38,360 Speaker 3: If you're taking photos of Pokemon that you find on 70 00:03:38,400 --> 00:03:42,640 Speaker 3: the ground, is that info all feeding back to Neantik. 71 00:03:43,040 --> 00:03:46,800 Speaker 1: Yeah, so Neantik says, quote, merely walking around playing our 72 00:03:46,880 --> 00:03:50,200 Speaker 1: games does not train an AI model. The key thing 73 00:03:50,440 --> 00:03:54,680 Speaker 1: that Pokemon Go involves is images. So I mentioned those 74 00:03:54,680 --> 00:03:57,880 Speaker 1: poke stops earlier. This is basically it's a real world 75 00:03:58,040 --> 00:04:01,800 Speaker 1: location that within the game takes on significance because it 76 00:04:01,880 --> 00:04:04,640 Speaker 1: might be in real life a statue. There's a couple 77 00:04:04,680 --> 00:04:08,320 Speaker 1: near our office, so a statue, a park, a playground, 78 00:04:08,640 --> 00:04:10,480 Speaker 1: so it exists in the real world, but then in 79 00:04:10,520 --> 00:04:13,080 Speaker 1: the game it is a place where you can get 80 00:04:13,120 --> 00:04:17,559 Speaker 1: special items. And once you reach level twenty in the game, 81 00:04:17,600 --> 00:04:21,599 Speaker 1: you can scan the real world location with your phone 82 00:04:21,680 --> 00:04:24,640 Speaker 1: and access even more items. So you have to have 83 00:04:24,680 --> 00:04:26,880 Speaker 1: played for kind of a while to get to that stage. 84 00:04:26,960 --> 00:04:30,920 Speaker 1: But basically you then scan a real world location with 85 00:04:31,040 --> 00:04:33,640 Speaker 1: your phone for something that happens to you in the game. 86 00:04:33,920 --> 00:04:36,080 Speaker 1: I can only explain it so much because I only 87 00:04:36,120 --> 00:04:39,880 Speaker 1: know so much about AR technology. But I have got 88 00:04:39,880 --> 00:04:42,680 Speaker 1: an expert, so we are going to hear now a 89 00:04:42,680 --> 00:04:45,560 Speaker 1: little bit of an interview with an expert on this subject, 90 00:04:45,600 --> 00:04:50,159 Speaker 1: and that's Bond University's Associate professor, doctor James Bert. He 91 00:04:50,360 --> 00:04:54,479 Speaker 1: has studied AR technology such as Pokemon Go, and here's 92 00:04:54,520 --> 00:04:56,839 Speaker 1: a little bit of how he described the game in 93 00:04:56,880 --> 00:05:00,280 Speaker 1: an interview with TDA journalist Elliott Laurie last week. 94 00:05:00,640 --> 00:05:07,080 Speaker 2: Pokemon Go was really just a gamification layer for allowing Neantic, 95 00:05:07,240 --> 00:05:12,280 Speaker 2: the company behind it, to essentially gather data of real 96 00:05:12,279 --> 00:05:15,440 Speaker 2: world locations. So they had about one hundred and fifty 97 00:05:15,520 --> 00:05:24,599 Speaker 2: million people worldwide taking photographs of landmarks, locations, streets, etc. 98 00:05:25,040 --> 00:05:29,719 Speaker 2: All around different cities, parks, suburbs all around the world. 99 00:05:30,160 --> 00:05:34,400 Speaker 2: And because they set up gyms and little stops and 100 00:05:34,440 --> 00:05:38,080 Speaker 2: stuff in the places that they wanted to capture, essentially 101 00:05:38,120 --> 00:05:42,160 Speaker 2: they were able to target the various data points that 102 00:05:42,200 --> 00:05:48,560 Speaker 2: they were able to generate multiple views images, pictures in 103 00:05:48,640 --> 00:05:52,680 Speaker 2: different lighting conditions, weather conditions, et cetera. And so they 104 00:05:52,680 --> 00:05:58,239 Speaker 2: were able to build out about thirty billion images from 105 00:05:58,320 --> 00:05:59,680 Speaker 2: their platform itself. 106 00:06:00,240 --> 00:06:03,159 Speaker 3: That's a really interesting application of user collected images. The 107 00:06:03,360 --> 00:06:06,640 Speaker 3: idea that you can almost build replicate that site because 108 00:06:06,680 --> 00:06:08,920 Speaker 3: of just how much rich data there is from one 109 00:06:08,960 --> 00:06:13,680 Speaker 3: particular location. How is Neantik or at least, how are 110 00:06:13,720 --> 00:06:16,880 Speaker 3: they telling us that they're using the images and the 111 00:06:16,960 --> 00:06:18,120 Speaker 3: data that they've collected. 112 00:06:18,480 --> 00:06:22,160 Speaker 1: Yeah, So, last year Neantic sold off its video game 113 00:06:22,279 --> 00:06:26,120 Speaker 1: arm including Pokemongo okay, but it kept all of the 114 00:06:26,279 --> 00:06:31,039 Speaker 1: mapping data and technology and its database of images for 115 00:06:31,240 --> 00:06:35,920 Speaker 1: an AI spinoff called Neantic Spatial. Now, Neantik says it's 116 00:06:36,000 --> 00:06:40,280 Speaker 1: used what it's called quote player contributed scans to build 117 00:06:40,440 --> 00:06:43,800 Speaker 1: a model of the world from this database of images. 118 00:06:44,160 --> 00:06:48,719 Speaker 1: It's called a large geospatial model, which is part of 119 00:06:48,839 --> 00:06:53,080 Speaker 1: its visual positioning system. So there's some big keywords here, 120 00:06:53,160 --> 00:06:57,680 Speaker 1: sort of jargon. Basically, the visual precisioning system is also 121 00:06:57,720 --> 00:07:01,000 Speaker 1: called a VPS, and that is similar to a GPS 122 00:07:01,480 --> 00:07:05,080 Speaker 1: or a global positioning system, but it is distinctly different. 123 00:07:05,720 --> 00:07:08,800 Speaker 3: Right, So the GPS that we have in phones and 124 00:07:08,880 --> 00:07:11,560 Speaker 3: cars and tells us how to get wherever we want 125 00:07:11,560 --> 00:07:15,720 Speaker 3: to go, similar principles, slightly different here with this VPS model. 126 00:07:16,400 --> 00:07:19,760 Speaker 3: Explain though, the technology that sits behind the GPS, because 127 00:07:19,760 --> 00:07:21,920 Speaker 3: I think that's probably important to first understand. 128 00:07:22,160 --> 00:07:25,160 Speaker 1: Yeah, so I learned this today, and I apologize if 129 00:07:25,160 --> 00:07:28,320 Speaker 1: everyone knows this except for me, But GPS is actually 130 00:07:28,360 --> 00:07:29,800 Speaker 1: owned by the US government. 131 00:07:30,120 --> 00:07:34,080 Speaker 3: So the technology that sits behind maps for example, Wow. 132 00:07:33,880 --> 00:07:36,680 Speaker 1: That is a US government property. Yeah, they developed it 133 00:07:36,680 --> 00:07:40,000 Speaker 1: in the nineteen seventies. It's currently operated by the department 134 00:07:40,080 --> 00:07:43,120 Speaker 1: known as the Space Force. It is a network of 135 00:07:43,160 --> 00:07:47,240 Speaker 1: satellites in space around the Earth that communicate with stations 136 00:07:47,280 --> 00:07:50,160 Speaker 1: on the ground and also with tech in your phone 137 00:07:50,160 --> 00:07:52,840 Speaker 1: and in your car that's called a receiver. So the 138 00:07:52,880 --> 00:07:56,760 Speaker 1: satellites in space can see the entire world from basically 139 00:07:56,800 --> 00:08:00,240 Speaker 1: a bird's eye view. That system works really well out 140 00:08:00,280 --> 00:08:03,680 Speaker 1: in the open, but it can struggle in dense environments 141 00:08:03,720 --> 00:08:07,200 Speaker 1: like cities with really tall buildings that can interfere with 142 00:08:07,280 --> 00:08:08,200 Speaker 1: the signal. 143 00:08:07,880 --> 00:08:10,720 Speaker 3: It drops out in tunnels, all that kind of stuff exactly. 144 00:08:10,800 --> 00:08:14,560 Speaker 1: So that's where VPS comes in. So using the images 145 00:08:14,680 --> 00:08:17,720 Speaker 1: taken by pokemon Go users of cities around the world 146 00:08:17,760 --> 00:08:20,160 Speaker 1: and even beyond cities. I mean, as I said, five 147 00:08:20,200 --> 00:08:23,240 Speaker 1: hundred million people anywhere that someone has been with a 148 00:08:23,240 --> 00:08:26,720 Speaker 1: smartphone and taken a photo of an environment for pokemon 149 00:08:26,760 --> 00:08:30,200 Speaker 1: Go that has been used to train, as I said, 150 00:08:30,240 --> 00:08:34,800 Speaker 1: this model that then powers the VPS. So Neantic says 151 00:08:34,880 --> 00:08:38,560 Speaker 1: its model is based on training of fifty million neural 152 00:08:38,920 --> 00:08:43,559 Speaker 1: networks and that it's trained those neural networks on those scans. 153 00:08:44,040 --> 00:08:46,520 Speaker 1: Sam as our resident AI expert, maybe you can explain 154 00:08:46,559 --> 00:08:47,720 Speaker 1: what a neural network means. 155 00:08:47,760 --> 00:08:49,400 Speaker 3: I'm just going to look it up in real time 156 00:08:49,440 --> 00:08:51,679 Speaker 3: to make sure that I am not completely wrong. No, 157 00:08:52,080 --> 00:08:54,600 Speaker 3: I'm on the right track. So basically, a neural network 158 00:08:54,920 --> 00:08:58,959 Speaker 3: is meant to copy the way our brain works. So 159 00:08:59,160 --> 00:09:02,240 Speaker 3: the brain has neural pathways. If you do something once, 160 00:09:02,360 --> 00:09:04,360 Speaker 3: it might be an unfamiliar skill. If you do it 161 00:09:04,400 --> 00:09:06,840 Speaker 3: over and over and over again, the neural pathway gets 162 00:09:06,840 --> 00:09:10,040 Speaker 3: stronger and that behavior gets super entrenched. It's my understanding 163 00:09:10,240 --> 00:09:14,319 Speaker 3: that a neural network does a similar thing. So if 164 00:09:14,360 --> 00:09:17,200 Speaker 3: a whole bunch of users kind of exhibit the same 165 00:09:17,240 --> 00:09:20,040 Speaker 3: behavior and the technology, it will really get good and 166 00:09:20,120 --> 00:09:21,440 Speaker 3: how to do that exact skill. 167 00:09:21,679 --> 00:09:24,560 Speaker 1: Yeah, So, with those networks that have then been trained 168 00:09:24,600 --> 00:09:28,360 Speaker 1: on player images, Neantik has built this model which is 169 00:09:28,400 --> 00:09:31,800 Speaker 1: able to mimic human thinking by filling in the gaps 170 00:09:31,800 --> 00:09:34,520 Speaker 1: between images to imagine what the back of a statue 171 00:09:34,600 --> 00:09:37,840 Speaker 1: or a building looks like without explicitly seeing it. Well, 172 00:09:37,920 --> 00:09:40,480 Speaker 1: so apparently AI has been really bad at this in 173 00:09:40,520 --> 00:09:44,760 Speaker 1: the past, but given this huge database of images, Neantik 174 00:09:44,800 --> 00:09:47,559 Speaker 1: has used the images trained the neural networks. The neural 175 00:09:47,600 --> 00:09:52,640 Speaker 1: networks form the model, the model powers the visual positioning system. 176 00:09:52,800 --> 00:09:56,679 Speaker 3: It all sounds really cool, It sounds quite creative in 177 00:09:56,800 --> 00:09:59,400 Speaker 3: terms of its application of technology. But how is the 178 00:09:59,440 --> 00:10:02,560 Speaker 3: company a actually using all of this? Because do we 179 00:10:02,600 --> 00:10:04,319 Speaker 3: really need to see the back of statues. 180 00:10:04,600 --> 00:10:06,880 Speaker 1: So the very first application that we've heard about is 181 00:10:06,920 --> 00:10:10,680 Speaker 1: that Nantik is partnering with a delivery robots startup called 182 00:10:10,760 --> 00:10:15,400 Speaker 1: Coco Robotics. So these delivery robots are at street level 183 00:10:15,640 --> 00:10:19,080 Speaker 1: just like people are, and using Niantic's tech, they're able 184 00:10:19,120 --> 00:10:22,280 Speaker 1: to identify locations figure out where they need to go 185 00:10:22,640 --> 00:10:27,640 Speaker 1: by matching their camera footage against the large geospatial model, 186 00:10:27,640 --> 00:10:30,679 Speaker 1: which is built on people walking around taking photos. 187 00:10:31,080 --> 00:10:32,920 Speaker 3: It kind of sounds like some of those little pizza 188 00:10:33,080 --> 00:10:36,480 Speaker 3: delivery robots that we've seen in the streets of the US. 189 00:10:36,600 --> 00:10:39,320 Speaker 3: I want to unpack though, the ethics about that sort 190 00:10:39,320 --> 00:10:44,120 Speaker 3: of offering being powered by images from Pokemon Go. But 191 00:10:44,200 --> 00:10:49,640 Speaker 3: first let's hear a quick word from our sponsor, so Lucy. 192 00:10:49,800 --> 00:10:53,720 Speaker 3: To recap twenty sixteen, hundreds of millions of people pick 193 00:10:53,800 --> 00:10:56,400 Speaker 3: up their phones to play Pokemon Go. In the real world, 194 00:10:56,679 --> 00:11:00,360 Speaker 3: lots of those players are scanning real world locations time 195 00:11:00,400 --> 00:11:03,119 Speaker 3: of the day. Some of those locations are inside buildings, 196 00:11:03,360 --> 00:11:05,480 Speaker 3: and the company that developed the game has used all 197 00:11:05,520 --> 00:11:08,920 Speaker 3: of these photos to develop this new piece of AI technology, 198 00:11:09,120 --> 00:11:12,200 Speaker 3: an AI model of the world that it's selling to 199 00:11:12,280 --> 00:11:17,280 Speaker 3: businesses that need highly detailed navigation. Did users or even 200 00:11:17,360 --> 00:11:19,800 Speaker 3: you know, users that play the game today, are they 201 00:11:19,800 --> 00:11:20,600 Speaker 3: consenting to that? 202 00:11:21,000 --> 00:11:25,120 Speaker 1: Well in a sense, yes, maybe not knowingly explicitly, but 203 00:11:25,480 --> 00:11:28,640 Speaker 1: all Pokemon Go users approved to the collection and third 204 00:11:28,679 --> 00:11:32,040 Speaker 1: party use of their data by agreeing to the terms 205 00:11:32,080 --> 00:11:32,680 Speaker 1: of service. 206 00:11:33,120 --> 00:11:35,479 Speaker 3: And so scroll scroll, scroll, tick exactly. 207 00:11:35,920 --> 00:11:38,400 Speaker 1: And so when you scan a poke stop you take 208 00:11:38,400 --> 00:11:41,000 Speaker 1: a photo of it, you have already given your consent 209 00:11:41,280 --> 00:11:45,240 Speaker 1: for Nantik to use that photo. And I'll note as well, 210 00:11:45,280 --> 00:11:47,640 Speaker 1: in order to scan a poke stop you need to 211 00:11:47,640 --> 00:11:50,760 Speaker 1: be level twenty, so you have to be a little 212 00:11:50,840 --> 00:11:54,840 Speaker 1: more than a casual player. So I picked up Pokemon 213 00:11:54,880 --> 00:11:57,319 Speaker 1: Go for the first time since twenty sixteen. This week. 214 00:11:57,440 --> 00:12:00,880 Speaker 1: I played for about fifteen minutes across two days. I 215 00:12:01,000 --> 00:12:04,160 Speaker 1: rose to level five, and then the leveling up slowed 216 00:12:04,400 --> 00:12:06,880 Speaker 1: way down. I think it would probably take me a 217 00:12:06,960 --> 00:12:10,920 Speaker 1: few hours to reach level twenty, let alone level thirty five, 218 00:12:10,920 --> 00:12:12,600 Speaker 1: which is what you need to be in order to 219 00:12:12,920 --> 00:12:16,480 Speaker 1: submit your own pokey stop, got it? But obviously there 220 00:12:16,520 --> 00:12:18,520 Speaker 1: is still kind of this aspect of what are you 221 00:12:18,559 --> 00:12:21,960 Speaker 1: actually agreeing to. Here's doctor Burt again with a little 222 00:12:21,960 --> 00:12:22,960 Speaker 1: bit more on this topic. 223 00:12:23,559 --> 00:12:27,680 Speaker 2: I would argue that most players would not have thought 224 00:12:28,120 --> 00:12:33,000 Speaker 2: that their hunting of Pope, Pikachus and other things would 225 00:12:33,040 --> 00:12:37,920 Speaker 2: have actually led to the training of a artificial intelligence system. 226 00:12:38,320 --> 00:12:41,280 Speaker 2: Is it legal, well, it seems to be legal. Is 227 00:12:41,320 --> 00:12:45,600 Speaker 2: it ethical That might be a gray area. And is 228 00:12:45,640 --> 00:12:49,800 Speaker 2: it deceptive? Well, in terms of the legalities, it's not. 229 00:12:50,080 --> 00:12:55,040 Speaker 2: But in terms of maybe how players were i guess 230 00:12:55,200 --> 00:12:59,679 Speaker 2: encouraged to use a platform for a different purpose possibly. 231 00:13:00,200 --> 00:13:03,440 Speaker 3: I think what's interesting about this topic is if you 232 00:13:03,440 --> 00:13:05,960 Speaker 3: think about the way that Google Maps and street View 233 00:13:06,040 --> 00:13:08,840 Speaker 3: has been developed, and I'm sure you've seen those cars 234 00:13:09,200 --> 00:13:12,200 Speaker 3: driving around with the big machines on top of them. 235 00:13:12,640 --> 00:13:15,880 Speaker 3: In some ways, the data that this company now has 236 00:13:16,120 --> 00:13:19,400 Speaker 3: isn't confidential in nature. I mean, this is kind of 237 00:13:19,600 --> 00:13:23,800 Speaker 3: real world environments that exist, and if they had the resources, 238 00:13:23,840 --> 00:13:26,199 Speaker 3: they could probably take a camera there themselves and document 239 00:13:26,240 --> 00:13:29,280 Speaker 3: it themselves. But what's interesting is the unexpected ways in 240 00:13:29,320 --> 00:13:33,400 Speaker 3: which data collected through a game is being used to 241 00:13:33,520 --> 00:13:37,000 Speaker 3: inform something that has all different ways that it can 242 00:13:37,080 --> 00:13:40,160 Speaker 3: be used. I mean, I'm sure delivering pizzas could be 243 00:13:40,160 --> 00:13:42,679 Speaker 3: the first stop, but there's so many different ways that 244 00:13:42,679 --> 00:13:45,640 Speaker 3: that data can be applied. Do you think that we 245 00:13:45,720 --> 00:13:48,679 Speaker 3: need to have a broader conversation about terms and conditions 246 00:13:48,720 --> 00:13:51,719 Speaker 3: with AI and think about the data that we all 247 00:13:51,720 --> 00:13:54,000 Speaker 3: put inside CHATWT and Claude for example. 248 00:13:54,480 --> 00:13:56,600 Speaker 1: Yeah, I mean this has certainly been a topic of 249 00:13:56,640 --> 00:14:00,680 Speaker 1: discussion over the years. I know that the A Triple 250 00:14:00,760 --> 00:14:03,760 Speaker 1: C has looked into non specifically terms and conditions, but 251 00:14:03,960 --> 00:14:06,680 Speaker 1: just sort of how a website's designed to have you 252 00:14:06,760 --> 00:14:10,480 Speaker 1: interact with them, and what sort of things are highlighted 253 00:14:10,520 --> 00:14:12,280 Speaker 1: for you to click on and not click on in 254 00:14:12,360 --> 00:14:15,800 Speaker 1: terms of like clicking on yes I agree to these terms, 255 00:14:15,880 --> 00:14:18,800 Speaker 1: and then a separate box it says, yes, I want 256 00:14:18,840 --> 00:14:23,120 Speaker 1: you to unsubscribe me from communications. So, yeah, I think 257 00:14:23,160 --> 00:14:25,960 Speaker 1: it's definitely an evolving area as we think about, yeah, 258 00:14:25,960 --> 00:14:28,400 Speaker 1: the different ways that our tech can be used. And 259 00:14:28,440 --> 00:14:31,840 Speaker 1: certainly there's a sort of a widespread saying on the 260 00:14:31,880 --> 00:14:35,600 Speaker 1: Internet that if the product is free, you're the product, 261 00:14:35,800 --> 00:14:38,440 Speaker 1: meaning like you're giving them your data in exchange for 262 00:14:38,560 --> 00:14:40,240 Speaker 1: use of an app. So yeah, we're going to have 263 00:14:40,320 --> 00:14:42,640 Speaker 1: to keep close tabs on it. 264 00:14:42,760 --> 00:14:46,440 Speaker 3: As these things evolve, Lucy, as data gets more valuable, 265 00:14:46,480 --> 00:14:47,960 Speaker 3: I think this is a topic that's going to come 266 00:14:48,040 --> 00:14:49,600 Speaker 3: up more and more. Thank you so much for taking 267 00:14:49,640 --> 00:14:51,680 Speaker 3: us through it. Thanks Sam, and thank you for joining 268 00:14:51,800 --> 00:14:54,200 Speaker 3: us on the Daily os this morning. We really appreciate 269 00:14:54,240 --> 00:14:56,320 Speaker 3: your company here. We're going to be back with the 270 00:14:56,480 --> 00:14:59,320 Speaker 3: evening headlines this afternoon, but until then, have a beautiful 271 00:14:59,360 --> 00:15:00,920 Speaker 3: start to your week. 272 00:15:04,120 --> 00:15:06,480 Speaker 1: My name is Lily Maddon and I'm a proud Arunda 273 00:15:06,680 --> 00:15:11,479 Speaker 1: Bunjelung Kalkadoon woman from Gadighl Country. The Daily oz acknowledges 274 00:15:11,560 --> 00:15:13,720 Speaker 1: that this podcast is recorded on the lands of the 275 00:15:13,760 --> 00:15:17,360 Speaker 1: Gadighl people and pays respect to all Aboriginal and torrest 276 00:15:17,400 --> 00:15:20,240 Speaker 1: Rate island and nations. We pay our respects to the 277 00:15:20,240 --> 00:15:23,040 Speaker 1: first peoples of these countries, both past and present.