1 00:00:00,000 --> 00:00:01,400 Hey everyone, welcome back. 2 00:00:01,400 --> 00:00:04,960 Ready to dive deep into some seriously cool AI stuff? 3 00:00:04,960 --> 00:00:06,040 Always ready. 4 00:00:06,040 --> 00:00:08,240 So we're diving into this awesome presentation 5 00:00:08,240 --> 00:00:11,120 by Eric Brinjolfsson, a real heavyweight 6 00:00:11,120 --> 00:00:13,240 in the world of digital economics right now. 7 00:00:13,240 --> 00:00:16,240 He's got his finger on the pulse of where AI is 8 00:00:16,240 --> 00:00:19,380 and where it's going, and more importantly, 9 00:00:19,380 --> 00:00:20,640 what it all means for us. 10 00:00:20,640 --> 00:00:22,520 Exactly, the impact is what matters. 11 00:00:22,520 --> 00:00:23,480 Totally. 12 00:00:23,480 --> 00:00:24,840 And what I always find so interesting 13 00:00:24,840 --> 00:00:28,960 about Brinjolfsson's work is how he connects AI 14 00:00:28,960 --> 00:00:32,280 to the real world, the actual economic impact. 15 00:00:32,280 --> 00:00:34,960 Everyone's talking about AI, but he cuts through the hype 16 00:00:34,960 --> 00:00:36,040 and gets to the heart of it. 17 00:00:36,040 --> 00:00:36,840 Yeah, no doubt. 18 00:00:36,840 --> 00:00:38,200 He's all about the practical side. 19 00:00:38,200 --> 00:00:40,780 Right, and he comes out swinging with his bold statement, 20 00:00:40,780 --> 00:00:43,360 claiming that AI is the defining general purpose 21 00:00:43,360 --> 00:00:45,040 technology of our time. 22 00:00:45,040 --> 00:00:48,560 He even compares it to electricity or the steam 23 00:00:48,560 --> 00:00:50,680 engine, those massive game changers in history. 24 00:00:50,680 --> 00:00:52,280 It's a pretty bold comparison. 25 00:00:52,280 --> 00:00:52,960 Yeah. 26 00:00:52,960 --> 00:00:54,660 So I'm curious, what do you think? 27 00:00:54,660 --> 00:00:57,040 Does that comparison hold water? 28 00:00:57,040 --> 00:00:58,560 Well, let's think about it for a sec. 29 00:00:58,560 --> 00:01:01,240 What did electricity really do for us? 30 00:01:01,240 --> 00:01:02,840 I mean, yeah, it lit up our homes, 31 00:01:02,840 --> 00:01:03,960 but it was more than that, right? 32 00:01:03,960 --> 00:01:04,640 Way more. 33 00:01:04,640 --> 00:01:06,600 Right, it completely revolutionized 34 00:01:06,600 --> 00:01:08,320 how we lived and worked. 35 00:01:08,320 --> 00:01:10,320 Brinjolfsson argues that AI could 36 00:01:10,320 --> 00:01:13,480 be even more fundamental, more impactful than even that. 37 00:01:13,480 --> 00:01:15,480 And his reasoning is interesting. 38 00:01:15,480 --> 00:01:20,000 He says it has the potential to actually solve intelligence. 39 00:01:20,000 --> 00:01:24,080 Now, imagine, just imagine, applying that power to things 40 00:01:24,080 --> 00:01:26,840 like, say, designing brand new energy sources 41 00:01:26,840 --> 00:01:30,680 or even personalized medicine that could wipe out diseases 42 00:01:30,680 --> 00:01:32,000 before we even get them. 43 00:01:32,000 --> 00:01:32,640 Are you seeing what I mean? 44 00:01:32,640 --> 00:01:33,160 Oh, totally. 45 00:01:33,160 --> 00:01:34,720 It's like next level stuff. 46 00:01:34,720 --> 00:01:35,600 Right, it's huge. 47 00:01:35,600 --> 00:01:38,600 We're talking about a whole new era of human progress 48 00:01:38,600 --> 00:01:40,360 driven by AI. 49 00:01:40,360 --> 00:01:42,640 It's kind of mind blowing when you really think about it. 50 00:01:42,640 --> 00:01:43,200 It really is. 51 00:01:43,200 --> 00:01:45,040 It actually makes DeepMind's slogan, 52 00:01:45,040 --> 00:01:46,880 that whole solve intelligence, use 53 00:01:46,880 --> 00:01:48,520 it to solve everything else thing, 54 00:01:48,520 --> 00:01:50,160 seem a lot less like, I don't know, 55 00:01:50,160 --> 00:01:52,560 some tech company's pipe dream and a lot more like a real 56 00:01:52,560 --> 00:01:53,320 roadmap. 57 00:01:53,320 --> 00:01:55,440 Yeah, it suddenly feels a lot more tangible 58 00:01:55,440 --> 00:01:57,000 and a lot less like sci-fi. 59 00:01:57,000 --> 00:01:57,520 Exactly. 60 00:01:57,520 --> 00:01:58,020 Yeah. 61 00:01:58,020 --> 00:01:58,960 It's happening now. 62 00:01:58,960 --> 00:02:01,320 But Bryn Dobson, he's not just looking way ahead 63 00:02:01,320 --> 00:02:02,240 into the future. 64 00:02:02,240 --> 00:02:06,680 He also breaks down how AI got to where it is now. 65 00:02:06,680 --> 00:02:08,680 He divides it into these three distinct waves. 66 00:02:08,680 --> 00:02:11,400 First, you've got those early rule-based systems. 67 00:02:11,400 --> 00:02:13,440 Then comes the explosion of machine learning, 68 00:02:13,440 --> 00:02:15,080 what he calls software 2.0. 69 00:02:15,080 --> 00:02:17,040 Which is what everyone's talking about now. 70 00:02:17,040 --> 00:02:21,280 And then now we've got this whole new wave, generative AI, 71 00:02:21,280 --> 00:02:24,760 which is so advanced and moving so fast that even the experts 72 00:02:24,760 --> 00:02:26,320 are like, whoa, hold on a second. 73 00:02:26,320 --> 00:02:27,200 It's true. 74 00:02:27,200 --> 00:02:28,040 It's really wild. 75 00:02:28,040 --> 00:02:31,440 It's like that scene in the Will Smith movie, 76 00:02:31,440 --> 00:02:33,600 where he challenges a robot to write a symphony. 77 00:02:33,600 --> 00:02:35,760 That seemed totally impossible just a few years ago. 78 00:02:35,760 --> 00:02:36,560 But guess what? 79 00:02:36,560 --> 00:02:39,480 Now we have AI composing music, writing screenplays, 80 00:02:39,480 --> 00:02:43,160 even acing those standardized tests like the SAT and GRE. 81 00:02:43,160 --> 00:02:44,840 Science fiction is trying to catch up 82 00:02:44,840 --> 00:02:46,440 to what's actually happening. 83 00:02:46,440 --> 00:02:48,800 And the thing is, it's not just about AI 84 00:02:48,800 --> 00:02:51,060 mimicking human creativity. 85 00:02:51,060 --> 00:02:56,400 It's about showing this capacity for actual problem solving, 86 00:02:56,400 --> 00:02:58,720 for reasoning stuff that people used to think 87 00:02:58,720 --> 00:03:01,000 was totally unique to humans. 88 00:03:01,000 --> 00:03:03,480 And the really cool part, this isn't some far off thing. 89 00:03:03,480 --> 00:03:05,200 It's already happening in the real world, 90 00:03:05,200 --> 00:03:06,880 even in everyday places. 91 00:03:06,880 --> 00:03:07,680 Oh, absolutely. 92 00:03:07,680 --> 00:03:10,400 And Bryn Jolson's research on AI's impact 93 00:03:10,400 --> 00:03:12,640 on productivity in the workplace really 94 00:03:12,640 --> 00:03:13,960 brings it all down to Earth. 95 00:03:13,960 --> 00:03:15,680 And you know what? 96 00:03:15,680 --> 00:03:19,160 He uses a really unexpected example, call centers. 97 00:03:19,160 --> 00:03:20,000 Who would have thought? 98 00:03:20,000 --> 00:03:20,480 Right. 99 00:03:20,480 --> 00:03:22,840 It's not exactly the first place you think of when 100 00:03:22,840 --> 00:03:24,160 you think cutting edge AI. 101 00:03:24,160 --> 00:03:24,660 Right. 102 00:03:24,660 --> 00:03:26,000 So what's the story there? 103 00:03:26,000 --> 00:03:26,960 Think about it, right? 104 00:03:26,960 --> 00:03:28,880 Call center work is often very repetitive. 105 00:03:28,880 --> 00:03:30,560 They're dealing with similar questions, 106 00:03:30,560 --> 00:03:32,480 similar problems all day long. 107 00:03:32,480 --> 00:03:34,180 So Bryn Jolson's team wanted to see 108 00:03:34,180 --> 00:03:36,440 what would happen if they used AI to help out 109 00:03:36,440 --> 00:03:40,400 these human agents in real time, giving them suggestions, 110 00:03:40,400 --> 00:03:42,400 providing information, all while they're 111 00:03:42,400 --> 00:03:43,800 interacting with customers. 112 00:03:43,800 --> 00:03:45,280 OK, so like a little AI assistant 113 00:03:45,280 --> 00:03:46,400 whispering in their ear. 114 00:03:46,400 --> 00:03:47,360 Exactly. 115 00:03:47,360 --> 00:03:50,440 And they went in with this idea that AI would just 116 00:03:50,440 --> 00:03:51,720 make things more efficient. 117 00:03:51,720 --> 00:03:53,280 But the results, they're actually 118 00:03:53,280 --> 00:03:54,960 way more nuanced than that. 119 00:03:54,960 --> 00:03:56,600 It wasn't about replacing the humans. 120 00:03:56,600 --> 00:03:58,840 It was about amplifying their abilities 121 00:03:58,840 --> 00:04:00,600 in a way that helped everyone. 122 00:04:00,600 --> 00:04:01,680 So it was a win-win. 123 00:04:01,680 --> 00:04:02,600 It seems like it. 124 00:04:02,600 --> 00:04:04,280 OK, but what actually happened? 125 00:04:04,280 --> 00:04:07,200 Did the AI make the agents more efficient? 126 00:04:07,200 --> 00:04:10,120 Or were there other unexpected outcomes? 127 00:04:10,120 --> 00:04:10,680 I got to know. 128 00:04:10,680 --> 00:04:12,840 So across the board, they saw big improvements. 129 00:04:12,840 --> 00:04:14,920 The agents were handling more calls per hour. 130 00:04:14,920 --> 00:04:16,920 The time it took to resolve issues went down. 131 00:04:16,920 --> 00:04:20,560 And even, get this, customer satisfaction went up. 132 00:04:20,560 --> 00:04:23,200 But here's where it gets interesting. 133 00:04:23,200 --> 00:04:25,920 The AI seemed to benefit the less experienced workers 134 00:04:25,920 --> 00:04:26,720 the most. 135 00:04:26,720 --> 00:04:28,160 Now, that is unexpected. 136 00:04:28,160 --> 00:04:29,880 You'd think the more experienced agents 137 00:04:29,880 --> 00:04:32,760 would be the ones to really take advantage of this AI helper, 138 00:04:32,760 --> 00:04:33,760 right? 139 00:04:33,760 --> 00:04:34,880 Why do you think that is? 140 00:04:34,880 --> 00:04:37,080 It does seem counterintuitive, right? 141 00:04:37,080 --> 00:04:38,640 Well, Brynjolfsson thinks it might 142 00:04:38,640 --> 00:04:42,640 be because AI can actually capture and share 143 00:04:42,640 --> 00:04:44,720 what he calls tacit knowledge. 144 00:04:44,720 --> 00:04:48,640 That's like the unspoken, those subtle but effective things 145 00:04:48,640 --> 00:04:50,640 that the best performers just do naturally. 146 00:04:50,640 --> 00:04:53,560 It's like having a virtual mentor whispering in your ear, 147 00:04:53,560 --> 00:04:55,000 guiding you to better answers. 148 00:04:55,000 --> 00:04:56,040 Oh, that's so interesting. 149 00:04:56,040 --> 00:04:58,600 So AI is kind of leveling the playing field here, 150 00:04:58,600 --> 00:05:00,760 allowing those less experienced workers 151 00:05:00,760 --> 00:05:03,960 to tap into the expertise of those more seasoned agents. 152 00:05:03,960 --> 00:05:04,920 Yeah, exactly. 153 00:05:04,920 --> 00:05:08,000 It's like AI is democratizing expertise, 154 00:05:08,000 --> 00:05:09,320 which is a really cool thought. 155 00:05:09,320 --> 00:05:09,840 It is. 156 00:05:09,840 --> 00:05:11,880 It challenges all those traditional assumptions 157 00:05:11,880 --> 00:05:15,120 about technology and its impact on jobs. 158 00:05:15,120 --> 00:05:15,760 But you know what? 159 00:05:15,760 --> 00:05:18,640 This whole idea of AI enhancing human abilities rather than 160 00:05:18,640 --> 00:05:22,320 replacing us, it leads to a more maybe a bit 161 00:05:22,320 --> 00:05:25,200 of a cautionary note in Brynjolfsson's presentation 162 00:05:25,200 --> 00:05:27,160 when he starts talking about the Turing trap. 163 00:05:27,160 --> 00:05:29,360 Right, because it's not all sunshine and roses. 164 00:05:29,360 --> 00:05:30,760 Right, so what exactly is that? 165 00:05:30,760 --> 00:05:31,760 What's the Turing trap? 166 00:05:31,760 --> 00:05:33,320 So it's a really interesting concept. 167 00:05:33,320 --> 00:05:36,200 And it's all about shifting how we think about AI development. 168 00:05:36,200 --> 00:05:39,040 Like on the surface, creating an AI that's 169 00:05:39,040 --> 00:05:41,720 basically indistinguishable from a human, 170 00:05:41,720 --> 00:05:44,560 that seems like the ultimate goal, right? 171 00:05:44,560 --> 00:05:47,520 But Brynjolfsson says that this, while super ambitious, 172 00:05:47,520 --> 00:05:50,480 might actually be aiming too low. 173 00:05:50,480 --> 00:05:51,280 Really? 174 00:05:51,280 --> 00:05:52,320 In what way? 175 00:05:52,320 --> 00:05:54,160 It seems like a pretty high bar to clear. 176 00:05:54,160 --> 00:05:56,420 Well, he says that if all we focus on 177 00:05:56,420 --> 00:05:58,860 is replicating human capabilities, 178 00:05:58,860 --> 00:06:02,080 we're going to fall into this trap of just automating tasks 179 00:06:02,080 --> 00:06:04,040 that we already do instead of exploring 180 00:06:04,040 --> 00:06:06,720 what AI is truly capable of, like solving problems 181 00:06:06,720 --> 00:06:09,440 in completely new ways, ways that we can't even imagine. 182 00:06:09,440 --> 00:06:09,920 I see. 183 00:06:09,920 --> 00:06:12,760 Thinking outside the box, not just replicating what's in it. 184 00:06:12,760 --> 00:06:13,720 Exactly. 185 00:06:13,720 --> 00:06:16,000 Brynjolfsson, he uses this really interesting thought 186 00:06:16,000 --> 00:06:17,960 experiment to illustrate his point. 187 00:06:17,960 --> 00:06:22,320 He says, imagine we could send robots back to ancient Greece, 188 00:06:22,320 --> 00:06:24,920 robots that could perfectly copy everything the Greeks were 189 00:06:24,920 --> 00:06:28,240 doing, building temples, making pottery, all of it. 190 00:06:28,240 --> 00:06:31,120 Would that have actually moved human civilization forward? 191 00:06:31,120 --> 00:06:32,680 It's like, would it have even mattered? 192 00:06:32,680 --> 00:06:35,640 Right, or would we be stuck in the same old ways of thinking 193 00:06:35,640 --> 00:06:36,720 and doing things? 194 00:06:36,720 --> 00:06:37,920 Makes you think, huh? 195 00:06:37,920 --> 00:06:39,960 The Turing Trap is basically a reminder 196 00:06:39,960 --> 00:06:43,220 that our goal with AI shouldn't be to simply create 197 00:06:43,220 --> 00:06:45,080 a copycat of ourselves. 198 00:06:45,080 --> 00:06:47,600 It's about using AI's unique abilities 199 00:06:47,600 --> 00:06:50,800 to push the boundaries of what's possible, 200 00:06:50,800 --> 00:06:53,840 to tackle problems that we haven't been able to solve for 201 00:06:53,840 --> 00:06:56,000 centuries, and to create a future that 202 00:06:56,000 --> 00:06:57,400 actually benefits everyone. 203 00:06:57,400 --> 00:06:59,320 So it's about aiming higher and thinking bigger. 204 00:06:59,320 --> 00:06:59,840 Exactly. 205 00:06:59,840 --> 00:07:01,920 It's about looking at the big picture, right? 206 00:07:01,920 --> 00:07:03,800 Not just efficiency for its own sake. 207 00:07:03,800 --> 00:07:07,160 It's about deciding what we want the future to look like, 208 00:07:07,160 --> 00:07:11,560 which brings us to the real meat of Brynjolfsson's argument, 209 00:07:11,560 --> 00:07:12,840 the economic side of things. 210 00:07:12,840 --> 00:07:15,080 Because, yeah, robots in ancient Greece 211 00:07:15,080 --> 00:07:18,160 are super interesting, but it does feel a bit theoretical. 212 00:07:18,160 --> 00:07:20,560 So how does this Turing Trap actually 213 00:07:20,560 --> 00:07:22,200 play out in the real world, especially 214 00:07:22,200 --> 00:07:24,880 when we're talking about the economy and jobs and all that? 215 00:07:24,880 --> 00:07:27,480 OK, so let's imagine, just for a second, 216 00:07:27,480 --> 00:07:30,520 that we go all in on automation. 217 00:07:30,520 --> 00:07:33,440 Those robots in ancient Greece, they're real, 218 00:07:33,440 --> 00:07:35,680 and they're everywhere, in every industry. 219 00:07:35,680 --> 00:07:36,560 What happens? 220 00:07:36,560 --> 00:07:39,200 Well, productivity, it would go through the roof, right? 221 00:07:39,200 --> 00:07:40,640 But then what about wages? 222 00:07:40,640 --> 00:07:42,400 What about people's ability to buy stuff, 223 00:07:42,400 --> 00:07:44,200 to participate in the economy? 224 00:07:44,200 --> 00:07:47,680 Our whole system kind of relies on people having jobs, 225 00:07:47,680 --> 00:07:48,840 earning money, right? 226 00:07:48,840 --> 00:07:51,280 Right, it's like that classic sci-fi problem. 227 00:07:51,280 --> 00:07:53,560 We solve scarcity, but then what? 228 00:07:53,560 --> 00:07:57,080 It seems like this amazing potential of AI 229 00:07:57,080 --> 00:07:58,920 to make everything more productive 230 00:07:58,920 --> 00:08:00,640 could also backfire pretty badly. 231 00:08:00,640 --> 00:08:02,960 Exactly, and that's what Brynjolfsson is getting out 232 00:08:02,960 --> 00:08:03,760 with the Turing Trap. 233 00:08:03,760 --> 00:08:05,260 It's like this warning sign, right? 234 00:08:05,260 --> 00:08:07,760 It's saying, hey, if we're not careful, 235 00:08:07,760 --> 00:08:09,400 we could end up with an economy that's 236 00:08:09,400 --> 00:08:12,720 super efficient on paper, but also incredibly unfair. 237 00:08:12,720 --> 00:08:15,440 So we could end up with a tiny group of people 238 00:08:15,440 --> 00:08:17,240 benefiting from all this automation, 239 00:08:17,240 --> 00:08:20,080 while everyone else gets left behind, which is 240 00:08:20,080 --> 00:08:21,960 the opposite of what we want. 241 00:08:21,960 --> 00:08:23,280 And honestly, it's kind of what everyone's 242 00:08:23,280 --> 00:08:25,120 worried about these days, right? 243 00:08:25,120 --> 00:08:28,920 AI, taking over jobs, widening the gap between the haves 244 00:08:28,920 --> 00:08:32,320 and the have-nots, it feels like we're at this crossroads. 245 00:08:32,320 --> 00:08:34,280 And the decisions we make now, they're 246 00:08:34,280 --> 00:08:35,400 going to determine everything. 247 00:08:35,400 --> 00:08:35,880 You got it. 248 00:08:35,880 --> 00:08:36,880 It's a pivotal moment. 249 00:08:36,880 --> 00:08:40,760 So do we just keep automating everything, hoping for the best, 250 00:08:40,760 --> 00:08:44,200 but risking these potentially terrible outcomes? 251 00:08:44,200 --> 00:08:46,240 Or do we take a different route, one that's 252 00:08:46,240 --> 00:08:49,720 more about using AI as this tool for good, 253 00:08:49,720 --> 00:08:52,680 to work alongside us, to create new opportunities, 254 00:08:52,680 --> 00:08:55,200 and to finally tackle some of those huge problems that 255 00:08:55,200 --> 00:08:56,240 have been around forever? 256 00:08:56,240 --> 00:08:58,800 So like a more optimistic view of the future? 257 00:08:58,800 --> 00:09:00,360 More like a more intentional one, I think. 258 00:09:00,360 --> 00:09:01,680 OK, I like that. 259 00:09:01,680 --> 00:09:03,140 But how do we actually do that? 260 00:09:03,140 --> 00:09:05,360 How do we avoid this Turing trap? 261 00:09:05,360 --> 00:09:08,840 Is it about changing how we develop AI itself? 262 00:09:08,840 --> 00:09:11,920 Or are we talking about bigger societal changes? 263 00:09:11,920 --> 00:09:13,680 It's going to take both, honestly. 264 00:09:13,680 --> 00:09:15,440 It's a multifaceted challenge. 265 00:09:15,440 --> 00:09:18,000 Bryn Jolfsson, he talks about a two-palmed approach, right? 266 00:09:18,000 --> 00:09:20,840 First, we've got to shift our mindset from replacement 267 00:09:20,840 --> 00:09:22,200 to augmentation. 268 00:09:22,200 --> 00:09:25,400 We need to be investing in AI that works with us, that 269 00:09:25,400 --> 00:09:28,200 enhances what we're already good at, not just replaces us. 270 00:09:28,200 --> 00:09:30,000 Remember those call center workers? 271 00:09:30,000 --> 00:09:31,920 The AI wasn't there to take their jobs. 272 00:09:31,920 --> 00:09:34,240 It was there to make them better at their jobs, 273 00:09:34,240 --> 00:09:35,520 to give them new skills. 274 00:09:35,520 --> 00:09:37,720 So it's about finding ways for humans and AI 275 00:09:37,720 --> 00:09:40,720 to be on the same team, each playing to their strengths. 276 00:09:40,720 --> 00:09:41,680 Exactly. 277 00:09:41,680 --> 00:09:45,000 Now, the other piece of this, and it's a big one, is policy. 278 00:09:45,000 --> 00:09:46,580 We need to give companies a reason 279 00:09:46,580 --> 00:09:48,600 to adopt this whole augmentation approach. 280 00:09:48,600 --> 00:09:51,240 Maybe that's tax breaks, grants, something 281 00:09:51,240 --> 00:09:54,120 that encourages them to invest in AI that helps humans, 282 00:09:54,120 --> 00:09:55,440 not just pushes them out. 283 00:09:55,440 --> 00:09:57,760 Right, because if we want companies to do the right thing, 284 00:09:57,760 --> 00:10:00,840 we've got to make it make sense for their bottom line, too. 285 00:10:00,840 --> 00:10:02,800 But even with the right policies and everything, 286 00:10:02,800 --> 00:10:05,560 it still feels like there's so much uncertainty. 287 00:10:05,560 --> 00:10:08,800 I mean, AI is changing so fast, how can we even 288 00:10:08,800 --> 00:10:12,440 begin to predict what the world will look like in 10, 20, 50 289 00:10:12,440 --> 00:10:13,120 years? 290 00:10:13,120 --> 00:10:15,560 It's kind of mind boggling. 291 00:10:15,560 --> 00:10:17,520 And speaking of predictions, Bryn Jolson 292 00:10:17,520 --> 00:10:20,280 mentioned some pretty wild ones about the timeline 293 00:10:20,280 --> 00:10:22,720 for this idea of transformative AI, 294 00:10:22,720 --> 00:10:24,560 like when AI will totally reshape 295 00:10:24,560 --> 00:10:25,960 every part of our lives. 296 00:10:25,960 --> 00:10:27,400 He even brings up some predictions 297 00:10:27,400 --> 00:10:30,360 about artificial general intelligence, or AGI, that 298 00:10:30,360 --> 00:10:32,040 is kind of blowing my mind right now. 299 00:10:32,040 --> 00:10:34,440 Yeah, the pace of a development is really something else. 300 00:10:34,440 --> 00:10:38,040 Even just a couple of years ago, the predictions about AGI, 301 00:10:38,040 --> 00:10:39,840 when AI can do basically anything 302 00:10:39,840 --> 00:10:42,680 a human can do intellectually, we're way, way further out. 303 00:10:42,680 --> 00:10:44,760 Remember that prediction platform we talked about, 304 00:10:44,760 --> 00:10:45,760 Metaculous? 305 00:10:45,760 --> 00:10:49,000 Well, this year, they're saying AGI could be here by 2032. 306 00:10:49,000 --> 00:10:50,860 Whoa, that's right around the corner. 307 00:10:50,860 --> 00:10:52,860 I don't even know what to do with that information. 308 00:10:52,860 --> 00:10:56,680 It makes Bryn Jolson's call for a Project Apollo for AI 309 00:10:56,680 --> 00:10:58,640 seem incredibly important. 310 00:10:58,640 --> 00:11:02,040 Totally, he's talking about this massive coordinated research 311 00:11:02,040 --> 00:11:05,440 effort, something on the scale of, well, the Apollo program 312 00:11:05,440 --> 00:11:07,200 that put people on the moon. 313 00:11:07,200 --> 00:11:11,520 But this time, it's all about understanding and navigating 314 00:11:11,520 --> 00:11:13,800 this AI revolution we're in the middle of. 315 00:11:13,800 --> 00:11:15,720 So it's not just about the technology itself. 316 00:11:15,720 --> 00:11:18,340 It's about figuring out the rules, the social contract, 317 00:11:18,340 --> 00:11:21,560 all of it, to make sure AI benefits everyone. 318 00:11:21,560 --> 00:11:22,360 Exactly. 319 00:11:22,360 --> 00:11:26,800 Bryn Jolson is challenging us to think big, to think long term, 320 00:11:26,800 --> 00:11:29,960 and to start laying the groundwork now for a future 321 00:11:29,960 --> 00:11:33,120 where AI helps us reach our full potential instead 322 00:11:33,120 --> 00:11:34,440 of holding us back. 323 00:11:34,440 --> 00:11:36,400 Yeah, it really makes you think about the future 324 00:11:36,400 --> 00:11:39,080 in a whole new way and all the possibilities, 325 00:11:39,080 --> 00:11:41,160 and maybe a little bit of the uncertainty, too. 326 00:11:41,160 --> 00:11:42,680 It's a lot to take in, for sure. 327 00:11:42,680 --> 00:11:43,180 It is. 328 00:11:43,180 --> 00:11:43,680 It is. 329 00:11:43,680 --> 00:11:45,680 But before we go too far down that rabbit hole, 330 00:11:45,680 --> 00:11:47,880 let's just take a second to recap what we've 331 00:11:47,880 --> 00:11:49,140 learned from Bryn Jolson today. 332 00:11:49,140 --> 00:11:50,400 It's been a wild ride. 333 00:11:50,400 --> 00:11:52,140 He's really changed how I think about AI. 334 00:11:52,140 --> 00:11:53,000 That's for sure. 335 00:11:53,000 --> 00:11:53,600 We do. 336 00:11:53,600 --> 00:11:57,560 So he basically argues that AI is this incredibly powerful 337 00:11:57,560 --> 00:12:01,000 force, maybe even the defining force of our time. 338 00:12:01,000 --> 00:12:06,080 He even says it's potentially a GPT like electricity, maybe even 339 00:12:06,080 --> 00:12:07,920 more so because it can solve intelligence. 340 00:12:07,920 --> 00:12:10,040 And it's not just theoretical either, right? 341 00:12:10,040 --> 00:12:10,920 Right. 342 00:12:10,920 --> 00:12:13,560 He brings it all back to reality with that call center example, 343 00:12:13,560 --> 00:12:15,260 which I thought was super interesting. 344 00:12:15,260 --> 00:12:18,880 Here's this real world example of AI actually 345 00:12:18,880 --> 00:12:21,700 making people's jobs better, even 346 00:12:21,700 --> 00:12:24,080 helping those who might not have as much experience. 347 00:12:24,080 --> 00:12:26,760 Which I think highlights that key point about the Turing 348 00:12:26,760 --> 00:12:27,560 trap, right? 349 00:12:27,560 --> 00:12:29,720 Just automating everything isn't the answer. 350 00:12:29,720 --> 00:12:32,280 It's about finding those win-win situations where 351 00:12:32,280 --> 00:12:34,040 AI and humans work together. 352 00:12:34,040 --> 00:12:34,760 Exactly. 353 00:12:34,760 --> 00:12:35,920 We need both. 354 00:12:35,920 --> 00:12:38,080 And that means we've got to change our thinking about AI, 355 00:12:38,080 --> 00:12:41,240 both in how we develop it and in the policies we put in place 356 00:12:41,240 --> 00:12:44,040 to make sure everyone benefits, not just a select few. 357 00:12:44,040 --> 00:12:46,400 It's about being proactive, shaping the future, 358 00:12:46,400 --> 00:12:48,440 instead of just letting it happen to us. 359 00:12:48,440 --> 00:12:49,780 100%. 360 00:12:49,780 --> 00:12:51,960 And that brings us back to that final question 361 00:12:51,960 --> 00:12:54,240 that Bryn Jolson leaves us with. 362 00:12:54,240 --> 00:12:57,200 If we can use AI to do all these amazing things, 363 00:12:57,200 --> 00:13:00,600 to solve problems, to create new possibilities, what kind 364 00:13:00,600 --> 00:13:01,920 of future could we create? 365 00:13:01,920 --> 00:13:02,880 It's a big question. 366 00:13:02,880 --> 00:13:04,120 It is a really big question. 367 00:13:04,120 --> 00:13:08,080 It goes way beyond the tech, beyond the economics even. 368 00:13:08,080 --> 00:13:11,560 It's about what we value as a society, what we aspire to. 369 00:13:11,560 --> 00:13:14,120 And I love that he doesn't pretend 370 00:13:14,120 --> 00:13:15,480 to have all the answers. 371 00:13:15,480 --> 00:13:16,080 Right. 372 00:13:16,080 --> 00:13:19,040 He's inviting all of us to be part of this conversation, 373 00:13:19,040 --> 00:13:20,880 to grapple with these big questions, 374 00:13:20,880 --> 00:13:22,680 and to make sure that the future of AI 375 00:13:22,680 --> 00:13:24,960 is one that benefits all of humanity. 376 00:13:24,960 --> 00:13:26,000 Perfectly said. 377 00:13:26,000 --> 00:13:27,840 And that's what we've tried to do today, too, right? 378 00:13:27,840 --> 00:13:29,920 Give you the information, the different perspectives, 379 00:13:29,920 --> 00:13:31,560 so you can form your own opinions 380 00:13:31,560 --> 00:13:34,200 about this incredible technology that's changing everything. 381 00:13:34,200 --> 00:13:35,240 So keep learning. 382 00:13:35,240 --> 00:13:37,160 Keep asking those tough questions. 383 00:13:37,160 --> 00:13:40,240 And remember, the future of AI is not set in stone. 384 00:13:40,240 --> 00:13:42,320 It's something we're all creating together. 385 00:13:42,320 --> 00:13:49,320 Until next time, stay curious, everyone.