1 00:00:00,160 --> 00:00:01,880 Speaker 1: Yea, thanks for listening to the show. This is Better 2 00:00:01,880 --> 00:00:04,760 Speaker 1: than Yesterday. Useful tools and useful conversations to help make 3 00:00:04,800 --> 00:00:07,480 Speaker 1: your day to day better than yesterday. Every episode, every 4 00:00:07,520 --> 00:00:10,480 Speaker 1: week since twenty thirteen Minams Lotter Ginsburg and I'm very 5 00:00:10,480 --> 00:00:13,200 Speaker 1: glad you're here. I hope you had a wonderful week. 6 00:00:13,240 --> 00:00:16,360 Speaker 1: I went to south By Southwest. I went there on 7 00:00:16,400 --> 00:00:19,280 Speaker 1: Wednesday with my friend Campbell, who illustrated the book So 8 00:00:19,320 --> 00:00:21,520 Speaker 1: What Now What, which you can get in the show notes. 9 00:00:22,840 --> 00:00:25,880 Speaker 1: The day I went there, I went to work there 10 00:00:26,360 --> 00:00:30,080 Speaker 1: working for a big international marketing company, to talk about 11 00:00:30,800 --> 00:00:34,919 Speaker 1: artificial intelligence and what the creative industry looks like in 12 00:00:34,960 --> 00:00:39,519 Speaker 1: a time when saer iiO exists. Well, VO three, you 13 00:00:39,560 --> 00:00:42,680 Speaker 1: can just type in a complete ad campaign, it'll come 14 00:00:42,720 --> 00:00:45,720 Speaker 1: out the other end of you. Machane. And I also 15 00:00:45,720 --> 00:00:49,640 Speaker 1: saw there's really brilliant, brilliant conversation with an AI safety 16 00:00:49,800 --> 00:00:55,160 Speaker 1: researcher who didn't mince about when he was basically saying, look, 17 00:00:55,960 --> 00:01:00,279 Speaker 1: one of the big AI models, Clawed, uses their own 18 00:01:00,360 --> 00:01:03,640 Speaker 1: model to type seventy percent of all the code that 19 00:01:03,680 --> 00:01:07,760 Speaker 1: goes into their product. And at the World Benchmark rankings 20 00:01:07,760 --> 00:01:11,480 Speaker 1: of the best coder in the world. AI has been 21 00:01:11,520 --> 00:01:13,039 Speaker 1: the best chess player in the world for a long time. 22 00:01:13,760 --> 00:01:16,080 Speaker 1: Less than a year ago, AI was the one hundred 23 00:01:16,080 --> 00:01:18,680 Speaker 1: and seventy fifth best coder in the world, So there's 24 00:01:18,680 --> 00:01:22,440 Speaker 1: one hundred and seventy four humans better than AI. AI 25 00:01:22,600 --> 00:01:26,760 Speaker 1: is now the sixth best coder in the world, and 26 00:01:26,760 --> 00:01:28,720 Speaker 1: that's in less than six months that it got to 27 00:01:28,800 --> 00:01:31,800 Speaker 1: do that and be that good. So pretty soon a 28 00:01:31,920 --> 00:01:34,199 Speaker 1: large language model will be able to write code better 29 00:01:34,240 --> 00:01:37,920 Speaker 1: than any human. Therefore, like any human might not even 30 00:01:37,959 --> 00:01:40,040 Speaker 1: be able to know what else is going on in there. 31 00:01:40,720 --> 00:01:42,640 Speaker 2: So the idea that we're. 32 00:01:44,000 --> 00:01:47,560 Speaker 1: Old mate at the very end of Castaway losing the 33 00:01:47,640 --> 00:01:50,680 Speaker 1: rope and the raft getting away from us, that is happening. 34 00:01:52,080 --> 00:01:55,040 Speaker 1: It's scary shit, but it's okay because we also need 35 00:01:55,040 --> 00:01:56,440 Speaker 1: to keep our eyes on the ball. I wanted to 36 00:01:56,520 --> 00:01:59,120 Speaker 1: kind of talk a bit about AI today, so we're 37 00:01:59,160 --> 00:02:00,440 Speaker 1: going to go and listen to a little bit of 38 00:02:00,440 --> 00:02:03,960 Speaker 1: Matt Agnew, doctor Matt Agnew, who has a master's degree 39 00:02:04,040 --> 00:02:08,320 Speaker 1: in AI research. He's an incredibly talented man. He's an astrophysicist, 40 00:02:08,360 --> 00:02:12,360 Speaker 1: he's a PhD, and he once a bachelor. 41 00:02:12,720 --> 00:02:13,480 Speaker 2: Which is really lovely. 42 00:02:13,480 --> 00:02:15,120 Speaker 1: So we're going to talk a bit about AI today, 43 00:02:15,280 --> 00:02:17,399 Speaker 1: and also I want to tell you about the most 44 00:02:17,440 --> 00:02:20,880 Speaker 1: valuable thing I learned that we can do ourselves as 45 00:02:20,880 --> 00:02:24,960 Speaker 1: these machines show up and do all of our jobs 46 00:02:24,960 --> 00:02:27,240 Speaker 1: better than us. So here's just a moment from Matt 47 00:02:27,280 --> 00:02:29,480 Speaker 1: from our conversation from episode five, one hundred and fifty. 48 00:02:29,480 --> 00:02:32,880 Speaker 1: The whole episode's fantastic, but in this bit, Matt talks 49 00:02:32,880 --> 00:02:36,800 Speaker 1: about how artificial intelligence is fundamentally different from other innovation 50 00:02:36,840 --> 00:02:40,119 Speaker 1: seeing technology, and he also kind of explores about AI's 51 00:02:40,200 --> 00:02:44,560 Speaker 1: ability to replace mental labor and how because it's doing that, 52 00:02:44,600 --> 00:02:47,280 Speaker 1: it could lead to a paradigm shift and how we've 53 00:02:47,639 --> 00:02:49,840 Speaker 1: even view work and view productivity. 54 00:02:51,760 --> 00:02:56,200 Speaker 3: We've had machines replace labor many many times over in 55 00:02:56,240 --> 00:03:00,919 Speaker 3: the past, we've never really had machines replace mental labor 56 00:03:01,280 --> 00:03:04,640 Speaker 3: in the same way that AI does, And so I 57 00:03:04,680 --> 00:03:08,640 Speaker 3: think there's this potential for us to really change the 58 00:03:08,680 --> 00:03:12,840 Speaker 3: paradigm of what making the living means. And that's kind 59 00:03:12,840 --> 00:03:16,840 Speaker 3: of been a fairly constant idea. I say, for a while, 60 00:03:16,919 --> 00:03:19,919 Speaker 3: you know, you work your forty hour week, you sleep 61 00:03:20,000 --> 00:03:21,840 Speaker 3: for however many hours, and then you've got a bit 62 00:03:21,840 --> 00:03:24,280 Speaker 3: of time to yourself, and I think there's a way 63 00:03:24,280 --> 00:03:25,000 Speaker 3: that we could change that. 64 00:03:25,040 --> 00:03:26,400 Speaker 2: I think work is always important. 65 00:03:26,440 --> 00:03:29,480 Speaker 3: I think that sort of purpose is fairly critical to humans, 66 00:03:30,160 --> 00:03:32,280 Speaker 3: but we could kind of, you know, let's can we 67 00:03:32,360 --> 00:03:34,800 Speaker 3: dial that down a bit and can we re amplify 68 00:03:34,920 --> 00:03:36,720 Speaker 3: some of the things. 69 00:03:36,400 --> 00:03:38,560 Speaker 2: That humanity or humans. 70 00:03:38,240 --> 00:03:41,560 Speaker 3: Really get a boost from, which which largely I would 71 00:03:41,560 --> 00:03:45,480 Speaker 3: say is hobbies, passions, and relationships. And so I think 72 00:03:45,840 --> 00:03:49,600 Speaker 3: the idea of okay, well let's change rather than the 73 00:03:49,640 --> 00:03:51,960 Speaker 3: five day working week and forty hours a week and 74 00:03:52,280 --> 00:03:56,000 Speaker 3: work life balance, it's like, what if making a living? 75 00:03:56,040 --> 00:03:59,280 Speaker 3: What if life? We can change that in quite a 76 00:03:59,520 --> 00:04:02,680 Speaker 3: significant way, a bit of an overhaul and you know, 77 00:04:03,040 --> 00:04:04,040 Speaker 3: let's have a bit of a reset. 78 00:04:04,680 --> 00:04:04,800 Speaker 1: All. 79 00:04:04,880 --> 00:04:07,360 Speaker 3: A lot of this is kind of rollover legacy stuff 80 00:04:07,360 --> 00:04:10,160 Speaker 3: that we've implemented over the years as we've had more 81 00:04:10,200 --> 00:04:14,040 Speaker 3: and more developments, and this potentially is an opportunity to 82 00:04:14,080 --> 00:04:17,159 Speaker 3: reset that and go, let's start again, because at some 83 00:04:17,200 --> 00:04:20,680 Speaker 3: point someone just made stuff up, right, Someone just made 84 00:04:20,760 --> 00:04:22,800 Speaker 3: up the forty hour work week, someone just made. 85 00:04:22,640 --> 00:04:24,040 Speaker 2: Up how you spend your time. 86 00:04:24,080 --> 00:04:26,080 Speaker 3: All these things were made up, and it's just kind 87 00:04:26,080 --> 00:04:30,760 Speaker 3: of rolled over and become the norm. But we can renormalize, 88 00:04:30,920 --> 00:04:34,560 Speaker 3: and so I think that's probably the biggest upside AI 89 00:04:34,640 --> 00:04:36,320 Speaker 3: is the ability to do that, and I think that 90 00:04:36,360 --> 00:04:37,960 Speaker 3: could be really interesting. 91 00:04:38,000 --> 00:04:39,120 Speaker 2: It could be done really. 92 00:04:38,920 --> 00:04:43,279 Speaker 3: Wrong if it's not careful and considered. But I think 93 00:04:43,760 --> 00:04:48,560 Speaker 3: I'd like to have faith in human ingenuity in that 94 00:04:50,080 --> 00:04:52,840 Speaker 3: we will tackle these problems and we can solve them 95 00:04:52,880 --> 00:04:58,040 Speaker 3: in a clever way that ultimately is beneficial for everyone. 96 00:04:58,800 --> 00:05:01,520 Speaker 1: So it's very clear that this technology has the potential 97 00:05:01,600 --> 00:05:05,440 Speaker 1: to redefine the way we live and the way we work. 98 00:05:05,520 --> 00:05:09,200 Speaker 1: When I was at south By Southwest, I asked the 99 00:05:09,240 --> 00:05:11,919 Speaker 1: two people I was on stage with, the Global CEO 100 00:05:12,080 --> 00:05:15,839 Speaker 1: and the Australia Pacific Asia APACK or whatever APAC mate, 101 00:05:16,400 --> 00:05:18,400 Speaker 1: what are the skills that we should probably have up 102 00:05:18,400 --> 00:05:21,240 Speaker 1: our sleeves as we move forward, And they said, the 103 00:05:21,320 --> 00:05:24,520 Speaker 1: ability to learn, unlearn and relearn something. It's probably the 104 00:05:24,520 --> 00:05:27,840 Speaker 1: most valuable thing you can have. For example, I don't 105 00:05:27,880 --> 00:05:30,520 Speaker 1: know before pedncillin, if if you showed up to the 106 00:05:30,520 --> 00:05:33,200 Speaker 1: doctor with syphilis, they would go, oh, geez, tough weekend, 107 00:05:33,400 --> 00:05:36,080 Speaker 1: here's a leech because that's the best thing we had 108 00:05:36,240 --> 00:05:38,960 Speaker 1: was leeches. A week later, you show up with your 109 00:05:38,960 --> 00:05:41,680 Speaker 1: syphilis and you're worried about the leaches and no, no, no, no, 110 00:05:41,720 --> 00:05:43,880 Speaker 1: it's fine. We have penicillin now, so from one week 111 00:05:43,920 --> 00:05:45,719 Speaker 1: to the next, the way we did the job is 112 00:05:45,720 --> 00:05:48,480 Speaker 1: completely different. We're still trying to heal the person, but 113 00:05:48,520 --> 00:05:51,600 Speaker 1: the way we do it is completely different. The job 114 00:05:51,600 --> 00:05:53,720 Speaker 1: we're doing is still the job we're doing, but the 115 00:05:53,720 --> 00:05:55,880 Speaker 1: way we do it we might need to relearn how 116 00:05:55,880 --> 00:05:58,440 Speaker 1: to do it very very quickly, and to unlearn an old. 117 00:05:58,240 --> 00:05:59,040 Speaker 2: Way of doing it. 118 00:06:00,120 --> 00:06:02,000 Speaker 1: Another thing I spoke to Matt Agnew about this is 119 00:06:02,040 --> 00:06:03,480 Speaker 1: episode five point fifty. By the way, if you want 120 00:06:03,480 --> 00:06:05,360 Speaker 1: to go back and listen to it, just scroll on back. 121 00:06:05,720 --> 00:06:09,640 Speaker 1: Is exploring how AI is working in education, it's already 122 00:06:09,640 --> 00:06:12,520 Speaker 1: influencing education, and how kids are often way more at 123 00:06:12,520 --> 00:06:16,040 Speaker 1: depth of understanding tech than grown ups. He also does 124 00:06:16,080 --> 00:06:19,920 Speaker 1: explore a bit of the ethical considerations of AI and 125 00:06:19,960 --> 00:06:22,280 Speaker 1: the implications that has on coming generations. 126 00:06:23,480 --> 00:06:27,360 Speaker 3: Children are moving into an increasingly AI driven world and 127 00:06:28,000 --> 00:06:33,080 Speaker 3: education is power and being adequately prepared is I guess 128 00:06:33,080 --> 00:06:36,520 Speaker 3: it's the responsibility of our generation to make sure the 129 00:06:36,600 --> 00:06:39,359 Speaker 3: children are are ready. And I think tied in with 130 00:06:39,480 --> 00:06:42,240 Speaker 3: that as well is I think we've spoken about it before, 131 00:06:42,320 --> 00:06:44,920 Speaker 3: the kind of decline in STEM in boys and girls 132 00:06:45,000 --> 00:06:48,960 Speaker 3: disproportionately more in girls in boys and girls, both in performance, 133 00:06:49,000 --> 00:06:53,280 Speaker 3: but more concerningly participation. And so I think again, creating 134 00:06:54,640 --> 00:06:59,240 Speaker 3: science and scientific literature and discourse that children engage with 135 00:06:59,839 --> 00:07:02,760 Speaker 3: is is really important. So yeah, there's kind of a 136 00:07:02,800 --> 00:07:06,840 Speaker 3: piece of let's keep fostering a love of science and children, 137 00:07:07,360 --> 00:07:11,440 Speaker 3: but then also let's adequately prepare children for this AI 138 00:07:11,520 --> 00:07:14,040 Speaker 3: driven world that can be scary if they're not prepared, 139 00:07:14,560 --> 00:07:16,200 Speaker 3: and the kind of I guess the secondary part of 140 00:07:16,200 --> 00:07:19,200 Speaker 3: that is that there's a lot of parents who or 141 00:07:19,280 --> 00:07:22,840 Speaker 3: an adults who aren't across it, and I think some 142 00:07:22,920 --> 00:07:24,640 Speaker 3: of them are finding it quite challenging to talk to 143 00:07:24,640 --> 00:07:27,400 Speaker 3: their children about AI or even to know do I 144 00:07:27,480 --> 00:07:29,160 Speaker 3: need to be do I need to be concerned? 145 00:07:29,560 --> 00:07:31,640 Speaker 2: Are they being exposed to AI in dangerous ways? 146 00:07:31,840 --> 00:07:35,000 Speaker 1: Kings could handle this kind of stuff. There's a misconception 147 00:07:35,280 --> 00:07:37,880 Speaker 1: that I was too complicated. We can't tell kids. This 148 00:07:37,960 --> 00:07:40,400 Speaker 1: is Reuben's big I bring up again, but it's Reuben's 149 00:07:40,400 --> 00:07:42,560 Speaker 1: big thing. He has all these models made of molecules, right, 150 00:07:43,040 --> 00:07:45,000 Speaker 1: and his thing is are we teaching the ABC's. We 151 00:07:45,120 --> 00:07:48,800 Speaker 1: teach them numbers, but we don't teach them the numbers 152 00:07:48,840 --> 00:07:51,240 Speaker 1: and the ABC's are the actual physical world that they inhabit. 153 00:07:51,880 --> 00:07:55,280 Speaker 1: Gravity pressure sets them. I don't have spasic shit. And 154 00:07:56,120 --> 00:07:59,920 Speaker 1: he gives these kids these molecules and they build they 155 00:08:00,520 --> 00:08:03,080 Speaker 1: they build things in front of him. And he sent 156 00:08:03,120 --> 00:08:04,480 Speaker 1: me a photo one day and he said, this is 157 00:08:04,480 --> 00:08:07,760 Speaker 1: a Grade eleven physics exam. It was completed by seven 158 00:08:07,840 --> 00:08:08,200 Speaker 1: year old. 159 00:08:08,360 --> 00:08:08,560 Speaker 2: Yeah. 160 00:08:08,640 --> 00:08:10,600 Speaker 1: Right, and this kid's not some sort of genius at 161 00:08:10,640 --> 00:08:12,720 Speaker 1: making up to a movie about there's just someone who's 162 00:08:12,760 --> 00:08:14,640 Speaker 1: had to explain to them and they can figure it out. 163 00:08:15,080 --> 00:08:16,840 Speaker 1: But for some reason, we don't tell these kids there's 164 00:08:16,840 --> 00:08:19,560 Speaker 1: stuff until a're sixteen, which is ridiculous. 165 00:08:20,200 --> 00:08:21,920 Speaker 2: It's too hard for you. You can't deal with that, That's it. 166 00:08:22,040 --> 00:08:25,080 Speaker 3: Yeah, I think I think the current education system needs 167 00:08:25,360 --> 00:08:27,280 Speaker 3: a bit of an overhaul. And that's not kind of 168 00:08:27,320 --> 00:08:30,440 Speaker 3: a dig at teachers. I think they're tasked with these 169 00:08:31,400 --> 00:08:36,160 Speaker 3: nearly impossible tasks of you need to teach thirty children 170 00:08:36,800 --> 00:08:39,440 Speaker 3: this particular topic where they all have very unique learning 171 00:08:39,440 --> 00:08:42,480 Speaker 3: styles in a limited period of time. Yeah, I think 172 00:08:42,480 --> 00:08:46,040 Speaker 3: that there's a lot of relics of an outdated education 173 00:08:46,120 --> 00:08:48,360 Speaker 3: system in place, and yeah, this is one of tho. 174 00:08:48,400 --> 00:08:52,200 Speaker 3: It's like it's the pacing, right, there's certainly concepts that 175 00:08:52,280 --> 00:08:54,120 Speaker 3: children can grasp much more quickly than that. 176 00:08:54,240 --> 00:08:55,280 Speaker 2: Probably a few, like. 177 00:08:55,240 --> 00:08:57,599 Speaker 1: It's common knowledge if you speak another language to a 178 00:08:57,679 --> 00:08:59,920 Speaker 1: kid before the age of twelve, they will learn it 179 00:09:00,360 --> 00:09:03,560 Speaker 1: like that, Like, tell me it's any different, It's no different. 180 00:09:03,640 --> 00:09:07,360 Speaker 3: Yeah, they're very, very robust and adapt they're sponges, right, 181 00:09:07,960 --> 00:09:12,680 Speaker 3: So yeah, it's and kind of almost counterintuitively, the book 182 00:09:12,720 --> 00:09:16,240 Speaker 3: I've written for AI, the age group of targeted are 183 00:09:16,240 --> 00:09:19,920 Speaker 3: probably already much more across it than the parents. And 184 00:09:20,000 --> 00:09:21,599 Speaker 3: so it's funny that, you know, I've written it for 185 00:09:21,640 --> 00:09:24,240 Speaker 3: eight to twelve, and certainly there'll be a group within 186 00:09:24,280 --> 00:09:25,400 Speaker 3: that cohort that find it. 187 00:09:26,040 --> 00:09:27,480 Speaker 2: You know, it's all new, but a lot of them 188 00:09:27,480 --> 00:09:27,959 Speaker 2: are already. 189 00:09:27,960 --> 00:09:33,040 Speaker 3: They're doing coding, they're doing it adjacent subjects, and so 190 00:09:33,120 --> 00:09:34,959 Speaker 3: a lot of the stuff they're probably, oh, yeah, I 191 00:09:35,040 --> 00:09:37,199 Speaker 3: kind already knew about this, but this is a little 192 00:09:37,200 --> 00:09:40,719 Speaker 3: more detailed for me to understand what the implications are 193 00:09:40,920 --> 00:09:42,640 Speaker 3: and what are the ethical considerations. 194 00:09:43,880 --> 00:09:48,040 Speaker 1: So yeah, it's very important that we prepare both children 195 00:09:48,120 --> 00:09:51,440 Speaker 1: and adults for the future, which is literally a month 196 00:09:51,440 --> 00:09:53,600 Speaker 1: from now. So this is not ten years from now, 197 00:09:53,600 --> 00:09:55,440 Speaker 1: this is not beyond two thousand this is not twenty 198 00:09:55,640 --> 00:09:58,360 Speaker 1: thirty two, this is not Olympic year. This is before 199 00:09:58,400 --> 00:10:02,640 Speaker 1: my next birthday. This ship is happening, Okay, if not already, 200 00:10:02,880 --> 00:10:06,240 Speaker 1: it's so so, so so important. The ethics around this 201 00:10:06,240 --> 00:10:09,040 Speaker 1: stuff is already bolted out of the gate. I did 202 00:10:09,080 --> 00:10:11,040 Speaker 1: actually make on mons stage with these two marketers. I 203 00:10:11,040 --> 00:10:13,280 Speaker 1: think there's one thing the advertising industry is known for, 204 00:10:13,640 --> 00:10:14,439 Speaker 1: it's ethics. 205 00:10:15,280 --> 00:10:15,959 Speaker 2: I got a good laugh. 206 00:10:16,000 --> 00:10:18,040 Speaker 1: I was pretty happy with that line because my book's 207 00:10:18,040 --> 00:10:20,959 Speaker 1: been stolen by the libgen hack, my first book back 208 00:10:21,000 --> 00:10:23,920 Speaker 1: after the Break, That and many many other books were 209 00:10:24,000 --> 00:10:26,320 Speaker 1: used to train all these AI models which are now 210 00:10:26,360 --> 00:10:30,760 Speaker 1: making schoollions of bucks based off stolen data. And no 211 00:10:30,760 --> 00:10:32,960 Speaker 1: one's going to stop it. I can't stop it. No 212 00:10:32,960 --> 00:10:35,559 Speaker 1: one can stop it. And so the ethics around this 213 00:10:35,559 --> 00:10:39,280 Speaker 1: stuff is really really full on. There was one more 214 00:10:39,280 --> 00:10:41,760 Speaker 1: thing that we kind of discovered about the best thing 215 00:10:41,800 --> 00:10:44,440 Speaker 1: we can do when it comes to making sure we 216 00:10:44,520 --> 00:10:46,920 Speaker 1: keep our humanity in the time when we're using all 217 00:10:46,920 --> 00:10:49,959 Speaker 1: these machines. But I do have to take a commercial break. 218 00:10:50,440 --> 00:10:52,000 Speaker 1: We'll be back with that and a little more from 219 00:10:52,000 --> 00:11:02,160 Speaker 1: that agnew in a moment. Thanks for listening to the show. 220 00:11:02,679 --> 00:11:04,800 Speaker 1: We're just kind of just recapping a little bit with 221 00:11:04,880 --> 00:11:08,240 Speaker 1: Matt Agnew and our conversation about AI after I went 222 00:11:08,280 --> 00:11:10,640 Speaker 1: to south By Southwest this week and my head kind 223 00:11:10,640 --> 00:11:15,480 Speaker 1: of got exploded by how terrifying and groundbreaking the next 224 00:11:15,720 --> 00:11:18,760 Speaker 1: six months of the future are. It's coming so fast. 225 00:11:18,760 --> 00:11:20,960 Speaker 1: I don't know if people are really aware of how 226 00:11:21,040 --> 00:11:24,320 Speaker 1: quickly things are changing. One of the lectures I saw 227 00:11:24,360 --> 00:11:29,640 Speaker 1: that day there was a quadrupedal robotic flamethrowing dog that 228 00:11:29,800 --> 00:11:32,600 Speaker 1: is autonomous and runs on AI and you can buy 229 00:11:32,679 --> 00:11:36,680 Speaker 1: it for nine grand to defend your home in America, obviously, 230 00:11:37,080 --> 00:11:40,400 Speaker 1: But do I want a flamethrowing dog based on AI 231 00:11:40,600 --> 00:11:42,360 Speaker 1: that I don't know if it's going to be able 232 00:11:42,360 --> 00:11:44,160 Speaker 1: to tell the difference between my neighbor and a batty 233 00:11:44,400 --> 00:11:46,320 Speaker 1: And even that, do I want to barbecue a battye 234 00:11:46,320 --> 00:11:46,840 Speaker 1: on my property? 235 00:11:46,880 --> 00:11:47,200 Speaker 2: Probably? 236 00:11:47,240 --> 00:11:53,600 Speaker 1: Not? So so weird. I don't want AI driven death robots. 237 00:11:53,640 --> 00:11:56,360 Speaker 1: I don't want that, but they're happening. But never mind. 238 00:11:58,679 --> 00:12:00,280 Speaker 1: There was one more thing that Matt Agnew and I 239 00:12:00,320 --> 00:12:03,560 Speaker 1: talked about, and it goes beyond what's happening at the moment, 240 00:12:03,640 --> 00:12:05,920 Speaker 1: but it's coming way quicker than you or I can 241 00:12:05,960 --> 00:12:11,160 Speaker 1: possibly know, and that is artificial general intelligence, and we 242 00:12:11,240 --> 00:12:14,760 Speaker 1: talk about what that actually is. So talking about how 243 00:12:14,800 --> 00:12:19,560 Speaker 1: does artificial general intelligence AGI differ from current AI and 244 00:12:19,600 --> 00:12:22,560 Speaker 1: why it's called the holy grail of AI research. We 245 00:12:22,600 --> 00:12:25,160 Speaker 1: also talk a little bit about the implications of AGI 246 00:12:25,720 --> 00:12:31,239 Speaker 1: achieving sentience and surpassing human intelligence. It's a lovely listen enjoy. 247 00:12:31,880 --> 00:12:34,360 Speaker 3: I would say that for most big tech companies playing 248 00:12:34,360 --> 00:12:38,440 Speaker 3: in the AI space as sort of AGI, a artificial 249 00:12:38,480 --> 00:12:42,240 Speaker 3: general intelligence is seen as the sort of holy grail. 250 00:12:43,360 --> 00:12:45,600 Speaker 3: And what I mean by general intelligence is. 251 00:12:47,120 --> 00:12:47,760 Speaker 2: Chat GBT. 252 00:12:48,120 --> 00:12:51,680 Speaker 3: While it seems human like in its conversation, is it's 253 00:12:51,920 --> 00:12:55,600 Speaker 3: the illusion of conversation, right, It's the illusion of thinking 254 00:12:55,640 --> 00:12:57,040 Speaker 3: and the illusion of intelligence. 255 00:12:57,080 --> 00:13:00,520 Speaker 2: It's it is algorithmic, algorithmic driven. 256 00:13:01,080 --> 00:13:04,040 Speaker 3: But that also means that it's got a very narrow 257 00:13:04,280 --> 00:13:08,120 Speaker 3: kind of application or use case, and that use case 258 00:13:08,240 --> 00:13:11,600 Speaker 3: is creating human like conversation. If you asked to do 259 00:13:11,640 --> 00:13:14,679 Speaker 3: something more general, like drive my car from A to B, 260 00:13:15,440 --> 00:13:18,679 Speaker 3: it can't do that. It's got nothing programmed in at 261 00:13:18,720 --> 00:13:21,480 Speaker 3: all that relates to being able to navigate from A 262 00:13:21,559 --> 00:13:24,240 Speaker 3: to B, just like you can't ask Google Maps to 263 00:13:24,240 --> 00:13:26,920 Speaker 3: tell you you know what's two times too. It's these 264 00:13:26,920 --> 00:13:30,040 Speaker 3: things have very specific use cases and they're quite narrow 265 00:13:30,120 --> 00:13:35,120 Speaker 3: in function. General intelligence is the idea like us. You 266 00:13:35,160 --> 00:13:37,160 Speaker 3: can give it any sort of problem and it can 267 00:13:37,200 --> 00:13:39,640 Speaker 3: solve it the same way we solve it. You if 268 00:13:39,679 --> 00:13:41,079 Speaker 3: I say how do you get from A to B? 269 00:13:41,280 --> 00:13:43,360 Speaker 3: You can tell me in some way you can solve 270 00:13:43,360 --> 00:13:46,920 Speaker 3: that problem. I I give you some algebraic problem I 271 00:13:46,960 --> 00:13:50,920 Speaker 3: want you solve again, you could solve that. AGI I 272 00:13:50,920 --> 00:13:53,440 Speaker 3: think is seen as the holy grail because it's saying, 273 00:13:53,480 --> 00:13:55,439 Speaker 3: all right, not only do we have something that can 274 00:13:55,559 --> 00:13:58,120 Speaker 3: you can interact with in a human like fashion like 275 00:13:58,200 --> 00:14:01,000 Speaker 3: chat GBT, but it can solve things really like humans. 276 00:14:01,440 --> 00:14:05,600 Speaker 3: And I think that's probably arguably the biggest milestone. And 277 00:14:05,640 --> 00:14:09,520 Speaker 3: I think that sort of scene as the first step 278 00:14:09,520 --> 00:14:12,600 Speaker 3: on the way to what's called the singularity, which is 279 00:14:13,120 --> 00:14:16,240 Speaker 3: a whole other kettle of fish, being as that general 280 00:14:16,280 --> 00:14:20,680 Speaker 3: intelligence approaches the point of human like intelligence and ultimately blows. 281 00:14:20,400 --> 00:14:24,160 Speaker 1: Past it and it gets that's where get identified as sentient. 282 00:14:24,960 --> 00:14:29,240 Speaker 3: Yes, yes, so, I mean sentience is still very high. 283 00:14:29,320 --> 00:14:32,000 Speaker 3: You know, you ask ten different signs hard to define 284 00:14:32,160 --> 00:14:34,960 Speaker 3: sentience and you get ten different answers. But yeah, if 285 00:14:34,960 --> 00:14:37,360 Speaker 3: you had something that kind of is mimicking human intelligence. 286 00:14:37,400 --> 00:14:40,280 Speaker 3: It's very hard to argue against sentience at that point. 287 00:14:40,680 --> 00:14:43,960 Speaker 3: I think anything before that, okay, yeah, it's you know, 288 00:14:44,040 --> 00:14:46,120 Speaker 3: how do you It's a bit more nebulous, right, But 289 00:14:46,200 --> 00:14:49,120 Speaker 3: as soon as something is as smart as you, very 290 00:14:49,240 --> 00:14:50,480 Speaker 3: very difficult to say I'm sentient. 291 00:14:50,480 --> 00:14:51,000 Speaker 2: But it's not. 292 00:14:52,440 --> 00:14:55,800 Speaker 1: So I don't know. The singularity is both super exciting 293 00:14:55,840 --> 00:15:00,200 Speaker 1: and terrifying, isn't it. I mean, good gosh. If if 294 00:15:00,520 --> 00:15:03,640 Speaker 1: you're not already having ethical conversations with your family about 295 00:15:03,720 --> 00:15:06,160 Speaker 1: using this stuff, it's important that you do. And if 296 00:15:06,200 --> 00:15:08,800 Speaker 1: you're not already speaking to your MP about the ethical 297 00:15:08,880 --> 00:15:11,200 Speaker 1: use of these tools, it's important that you do, because 298 00:15:11,200 --> 00:15:15,080 Speaker 1: they're happening so so fast, And like I said, it's 299 00:15:15,080 --> 00:15:17,920 Speaker 1: not five years, ten years, twenty years, it's like within 300 00:15:18,000 --> 00:15:21,040 Speaker 1: a year. Shit is changing so fast we're going to 301 00:15:21,080 --> 00:15:22,200 Speaker 1: have to get on top of it. But there is 302 00:15:22,240 --> 00:15:24,520 Speaker 1: one thing that I did we did talk about on stage, 303 00:15:25,600 --> 00:15:30,400 Speaker 1: because if art is a conversation and all creativity is 304 00:15:31,320 --> 00:15:34,400 Speaker 1: just an idea, and another idea would have never been 305 00:15:34,440 --> 00:15:38,280 Speaker 1: combined yet combining, and that is a new idea. If 306 00:15:38,480 --> 00:15:41,520 Speaker 1: the ideas we are being inspired by are being generated 307 00:15:41,760 --> 00:15:45,200 Speaker 1: by AI, and then we type into another AI. Hey, 308 00:15:45,240 --> 00:15:50,560 Speaker 1: I want to do this thing. We are unwillingly just 309 00:15:50,600 --> 00:15:53,480 Speaker 1: a warm meat bag in a part of the AI 310 00:15:53,560 --> 00:15:57,360 Speaker 1: slop loop. And that's no good if the only inspiration 311 00:15:57,440 --> 00:16:01,680 Speaker 1: we're getting is just slop, you know, and we are 312 00:16:01,720 --> 00:16:04,720 Speaker 1: a part of that. And so we were talking when 313 00:16:04,720 --> 00:16:07,040 Speaker 1: we're on stage. We were talking about how important it 314 00:16:07,080 --> 00:16:09,200 Speaker 1: is to put your fucking phone down and go out 315 00:16:09,240 --> 00:16:13,000 Speaker 1: and touch the world, feel the world, feel tactile things 316 00:16:13,000 --> 00:16:18,000 Speaker 1: that a machine can't experience. Something human, the emotion, the 317 00:16:18,040 --> 00:16:22,120 Speaker 1: connection that what connects you another person, particularly when it 318 00:16:22,120 --> 00:16:24,760 Speaker 1: comes to like and that is my thoughts now when 319 00:16:24,800 --> 00:16:27,000 Speaker 1: it comes to connecting with your friends. I mean, even 320 00:16:27,280 --> 00:16:28,920 Speaker 1: since Saw two came out in the last like the 321 00:16:28,960 --> 00:16:31,120 Speaker 1: last two weeks, I scroll through my feet and I 322 00:16:31,320 --> 00:16:32,920 Speaker 1: don't even know it's real or not. And I'm so 323 00:16:33,040 --> 00:16:35,760 Speaker 1: reluctant to even look at my freaking phone. But if 324 00:16:35,760 --> 00:16:38,080 Speaker 1: I'm not looking at my phone, am I connecting with 325 00:16:38,120 --> 00:16:41,640 Speaker 1: people that I can't see with my own eyes? So 326 00:16:42,280 --> 00:16:44,960 Speaker 1: I'm thinking more and more it's really important to get 327 00:16:45,040 --> 00:16:47,240 Speaker 1: out there and actually look at another human being and 328 00:16:47,280 --> 00:16:49,040 Speaker 1: see what the light does to their eyeballs when they 329 00:16:49,080 --> 00:16:50,920 Speaker 1: look at you, and see how the wind flicks through 330 00:16:50,960 --> 00:16:52,480 Speaker 1: their hair, and listen to the tone of their voice, 331 00:16:52,480 --> 00:16:54,480 Speaker 1: so they tell you about the sandwich they are yesterday 332 00:16:54,560 --> 00:16:57,680 Speaker 1: and actually connect with an the human being, because it's 333 00:16:57,720 --> 00:17:02,440 Speaker 1: that what makes us human well be and it is 334 00:17:02,480 --> 00:17:04,640 Speaker 1: the thing that we cannot have taken away from us, 335 00:17:05,040 --> 00:17:06,879 Speaker 1: and it will be the thing that allows us to 336 00:17:07,000 --> 00:17:10,600 Speaker 1: use these tools to create things and do stuff that 337 00:17:10,640 --> 00:17:13,680 Speaker 1: other people are not doing if we maintain our humanity 338 00:17:13,720 --> 00:17:18,080 Speaker 1: and maintain our perspective on things. Thank you so much 339 00:17:18,080 --> 00:17:20,040 Speaker 1: for listening. I really appreciate it. I haven't done one 340 00:17:20,040 --> 00:17:22,439 Speaker 1: of these little episodes in a while. Thank you very 341 00:17:22,520 --> 00:17:25,000 Speaker 1: much to add a bunch of for chopping it together. 342 00:17:25,240 --> 00:17:27,080 Speaker 1: If you want to hear more of Dodr Matt Agnew's 343 00:17:27,400 --> 00:17:30,960 Speaker 1: insights into AI, it's brilliant stuff. Look, he's a super 344 00:17:30,960 --> 00:17:33,719 Speaker 1: brilliant man. He has a PhD in astrophysics and master's 345 00:17:33,760 --> 00:17:36,879 Speaker 1: in AI and he goes also once a bachelor, and 346 00:17:36,920 --> 00:17:40,600 Speaker 1: he's a delight he's a delightful man. Episode five, one 347 00:17:40,640 --> 00:17:43,440 Speaker 1: hundred and fifty is the show. Just scroll on back 348 00:17:43,520 --> 00:17:45,879 Speaker 1: in your feed and you'll grab it there. Thanks so 349 00:17:45,960 --> 00:17:48,240 Speaker 1: much for listening, And if you're still listening from now 350 00:17:48,600 --> 00:17:50,400 Speaker 1: you want to come to story Club on the ninth 351 00:17:50,440 --> 00:17:54,639 Speaker 1: of November. Mark Humphrey's, Harley Breen, Nina Ayama, Beck Wilson, 352 00:17:55,320 --> 00:17:57,920 Speaker 1: and myself and a new guest. We are yet to announce, 353 00:17:58,080 --> 00:18:00,320 Speaker 1: but it's probably maybe the life I just want to 354 00:18:00,320 --> 00:18:02,240 Speaker 1: get to do for the year. It'll definitely sell out. 355 00:18:02,560 --> 00:18:05,560 Speaker 1: Get your tickets. The linkers in the show notes, thank 356 00:18:05,560 --> 00:18:15,600 Speaker 1: you so much for listening. I'll see you Wednesday.