1 00:00:00,000 --> 00:00:02,280 Welcome to another deep dive. 2 00:00:02,280 --> 00:00:04,840 You've sent us a bunch of articles about AI, 3 00:00:04,840 --> 00:00:07,400 and honestly, I'm a little intimidated. 4 00:00:07,400 --> 00:00:08,240 Oh, yeah? 5 00:00:08,240 --> 00:00:09,080 Yeah. 6 00:00:09,080 --> 00:00:11,360 We're talking everything from trading bots, 7 00:00:11,360 --> 00:00:14,160 making moves in the crypto market to AI 8 00:00:14,160 --> 00:00:17,120 that can supposedly fall in love, 9 00:00:17,120 --> 00:00:20,320 even AI that's challenging our ideas about what art is. 10 00:00:20,320 --> 00:00:21,160 Right. 11 00:00:21,160 --> 00:00:22,640 I'm excited to unpack all of this with you. 12 00:00:22,640 --> 00:00:25,160 Yeah, it's pretty remarkable to see how quickly AI 13 00:00:25,160 --> 00:00:28,020 is moving from a theoretical concept 14 00:00:28,020 --> 00:00:29,760 to something we interact with every day. 15 00:00:29,760 --> 00:00:30,800 It really is. 16 00:00:30,800 --> 00:00:33,240 Let's start with something that impacts a lot of people. 17 00:00:33,240 --> 00:00:33,740 Yeah. 18 00:00:33,740 --> 00:00:34,240 Money. 19 00:00:34,240 --> 00:00:37,720 You've been reading up on AI crypto trading bots, 20 00:00:37,720 --> 00:00:40,440 these programs that are designed to buy and sell crypto 21 00:00:40,440 --> 00:00:42,040 around the clock. 22 00:00:42,040 --> 00:00:43,040 What's the deal with these? 23 00:00:43,040 --> 00:00:44,360 Is this the future of investing? 24 00:00:44,360 --> 00:00:47,560 I mean, these bots are all about speed and processing power. 25 00:00:47,560 --> 00:00:50,560 Imagine scanning millions of data points per second 26 00:00:50,560 --> 00:00:53,440 to spot trends that humans would likely miss. 27 00:00:53,440 --> 00:00:53,940 Wow. 28 00:00:53,940 --> 00:00:56,400 It's like having a super powered assistant who can react 29 00:00:56,400 --> 00:00:57,640 instantly to market shifts. 30 00:00:57,640 --> 00:00:59,200 But aren't they just reacting to the market? 31 00:00:59,200 --> 00:01:01,040 How do they actually make decisions? 32 00:01:01,040 --> 00:01:02,880 That's where things get really interesting. 33 00:01:02,880 --> 00:01:05,120 They're not just programmed with a set of rules. 34 00:01:05,120 --> 00:01:08,240 They actually learn from human behavior. 35 00:01:08,240 --> 00:01:09,320 Think about it. 36 00:01:09,320 --> 00:01:12,360 They can analyze how humans react to market changes, 37 00:01:12,360 --> 00:01:14,880 recognize patterns in their trading decisions, 38 00:01:14,880 --> 00:01:18,240 and then use that information to develop their own strategies. 39 00:01:18,240 --> 00:01:21,440 It's like having an apprentice who studies your every move. 40 00:01:21,440 --> 00:01:24,800 Only this apprentice is pure algorithms. 41 00:01:24,800 --> 00:01:27,840 So they're picking up on human tendencies, even our emotions. 42 00:01:27,840 --> 00:01:28,560 Exactly. 43 00:01:28,560 --> 00:01:31,160 They can learn to recognize when traders are acting out 44 00:01:31,160 --> 00:01:33,360 of fear or greed, and then capitalize 45 00:01:33,360 --> 00:01:36,080 on those predictable, very human reactions. 46 00:01:36,080 --> 00:01:38,320 That's a little unsettling, but also kind of brilliant. 47 00:01:38,320 --> 00:01:40,720 What are some of the common strategies these bots use? 48 00:01:40,720 --> 00:01:42,560 Well, you've got strategies like scalping, 49 00:01:42,560 --> 00:01:44,680 which focuses on making tiny profits off 50 00:01:44,680 --> 00:01:46,160 of small price fluctuations. 51 00:01:46,160 --> 00:01:46,720 OK. 52 00:01:46,720 --> 00:01:49,240 Imagine buying low and selling high, 53 00:01:49,240 --> 00:01:51,440 hundreds or even thousands of times a day. 54 00:01:51,440 --> 00:01:52,120 Wow. 55 00:01:52,120 --> 00:01:55,240 Then there's arbitrage, where the bots exploit price differences 56 00:01:55,240 --> 00:01:56,480 across different exchanges. 57 00:01:56,480 --> 00:01:59,440 So you're telling me these bots can spot a price difference 58 00:01:59,440 --> 00:02:03,760 on, say, Coinbase versus Binance and make a trade in milliseconds 59 00:02:03,760 --> 00:02:05,040 to profit from it? 60 00:02:05,040 --> 00:02:06,480 That's insane. 61 00:02:06,480 --> 00:02:08,400 That's the power of AI. 62 00:02:08,400 --> 00:02:11,400 Now, if you're thinking of diving into the world of AI trading 63 00:02:11,400 --> 00:02:13,520 bots, you have a couple of options. 64 00:02:13,520 --> 00:02:15,920 You can either build your own bot from scratch, which 65 00:02:15,920 --> 00:02:17,960 gives you complete control, but requires 66 00:02:17,960 --> 00:02:20,040 a fair bit of technical expertise, 67 00:02:20,040 --> 00:02:22,400 or you can opt for pre-built platforms. 68 00:02:22,400 --> 00:02:24,240 So with pre-built platforms, you're essentially 69 00:02:24,240 --> 00:02:26,800 renting a bot that's already trained and ready to go. 70 00:02:26,800 --> 00:02:27,520 Precisely. 71 00:02:27,520 --> 00:02:27,840 OK. 72 00:02:27,840 --> 00:02:30,480 Platforms like CoinRule or Pyonex make it much easier 73 00:02:30,480 --> 00:02:31,560 to get started. 74 00:02:31,560 --> 00:02:34,080 You can choose from a variety of strategies, 75 00:02:34,080 --> 00:02:37,360 set your risk tolerance, and let the bot do its thing. 76 00:02:37,360 --> 00:02:39,280 But I imagine there's a catch, right? 77 00:02:39,280 --> 00:02:41,960 Well, you typically have to pay a fee to the platform, 78 00:02:41,960 --> 00:02:44,920 either a subscription or a cut of your profits. 79 00:02:44,920 --> 00:02:47,880 But there's another layer to this whole bot ecosystem 80 00:02:47,880 --> 00:02:50,560 that's emerging, AI agents. 81 00:02:50,560 --> 00:02:52,840 OK, now we're getting really futuristic. 82 00:02:52,840 --> 00:02:55,480 What's the difference between a bot and an agent? 83 00:02:55,480 --> 00:02:58,760 Think of an agent as a bot on steroids. 84 00:02:58,760 --> 00:03:01,360 Well, a bot is usually focused on a specific task, 85 00:03:01,360 --> 00:03:03,160 like executing trades. 86 00:03:03,160 --> 00:03:06,600 An AI agent can handle much more complex activity. 87 00:03:06,600 --> 00:03:08,920 Take Coinbase's based agent, for example. 88 00:03:08,920 --> 00:03:12,640 It can actually learn, adapt, and make strategic decisions 89 00:03:12,640 --> 00:03:15,840 across various tasks, from managing your portfolio 90 00:03:15,840 --> 00:03:18,920 to even handling marketing and deploying NFTs. 91 00:03:18,920 --> 00:03:19,400 Wow. 92 00:03:19,400 --> 00:03:22,520 It's like having a financial advisor, marketing guru, 93 00:03:22,520 --> 00:03:24,600 and tech expert all rolled into one. 94 00:03:24,600 --> 00:03:25,760 Exactly. 95 00:03:25,760 --> 00:03:27,760 These agents are still in the early stages, 96 00:03:27,760 --> 00:03:30,600 but they have the potential to revolutionize how we interact 97 00:03:30,600 --> 00:03:31,520 with the crypto market. 98 00:03:31,520 --> 00:03:33,840 But let's not forget, with all this talk of AI 99 00:03:33,840 --> 00:03:37,120 taking over the financial world, there are risks, right? 100 00:03:37,120 --> 00:03:39,480 What could go wrong with these bots running amuck? 101 00:03:39,480 --> 00:03:40,640 Absolutely. 102 00:03:40,640 --> 00:03:43,680 Like any trader, bots can make bad decisions. 103 00:03:43,680 --> 00:03:45,200 And with their always-on nature, 104 00:03:45,200 --> 00:03:47,320 those bad decisions can happen at any time, 105 00:03:47,320 --> 00:03:49,640 potentially amplifying losses very quickly. 106 00:03:49,640 --> 00:03:52,480 So it's not as simple as just setting it and forgetting it. 107 00:03:52,480 --> 00:03:55,760 You need to be pretty savvy to use these effectively and safely. 108 00:03:55,760 --> 00:03:56,520 Exactly. 109 00:03:56,520 --> 00:03:58,800 Security is paramount given that these bots often 110 00:03:58,800 --> 00:04:00,960 have access to your crypto exchanges. 111 00:04:00,960 --> 00:04:03,160 And it's essential to have safeguards in place, 112 00:04:03,160 --> 00:04:06,480 like stop-loss orders, to limit your potential losses. 113 00:04:06,480 --> 00:04:08,200 This has been really eye-opening. 114 00:04:08,200 --> 00:04:11,160 I'm starting to understand why so many people are both excited 115 00:04:11,160 --> 00:04:16,640 and nervous about AI taking on a bigger role in finance. 116 00:04:16,640 --> 00:04:18,080 It's powerful, for sure. 117 00:04:18,080 --> 00:04:21,200 AI is disrupting just about every field imaginable, 118 00:04:21,200 --> 00:04:23,080 and finance is right at the forefront. 119 00:04:23,080 --> 00:04:25,520 Well, speaking of disruption, you sent over some articles 120 00:04:25,520 --> 00:04:27,840 about an AI that's making waves. 121 00:04:27,840 --> 00:04:29,320 But not in the financial world we're 122 00:04:29,320 --> 00:04:30,880 talking about the art world now. 123 00:04:30,880 --> 00:04:31,480 That's right. 124 00:04:31,480 --> 00:04:33,920 X, the company formerly known as Twitter, 125 00:04:33,920 --> 00:04:37,880 has unveiled its new AI image generator called Aurora. 126 00:04:37,880 --> 00:04:38,760 OK. 127 00:04:38,760 --> 00:04:40,160 And it's causing quite a stir. 128 00:04:40,160 --> 00:04:43,080 I have to admit, I've seen some of the images Aurora has created, 129 00:04:43,080 --> 00:04:44,320 and they're incredible. 130 00:04:44,320 --> 00:04:45,720 It's like looking at photographs. 131 00:04:45,720 --> 00:04:46,840 How does it even work? 132 00:04:46,840 --> 00:04:49,800 It's based on something called a diffusion model, which 133 00:04:49,800 --> 00:04:52,000 is a type of deep learning algorithm. 134 00:04:52,000 --> 00:04:52,520 OK. 135 00:04:52,520 --> 00:04:55,920 Essentially, it's trained on a massive data set of images 136 00:04:55,920 --> 00:04:58,880 and learns to generate new images by reversing 137 00:04:58,880 --> 00:05:01,880 a process of adding noise to existing images. 138 00:05:01,880 --> 00:05:05,120 It starts with pure noise and gradually refines it 139 00:05:05,120 --> 00:05:06,920 until it forms a coherent image. 140 00:05:06,920 --> 00:05:09,720 So it's learning the underlying patterns and structures 141 00:05:09,720 --> 00:05:12,040 of images and then applying that knowledge 142 00:05:12,040 --> 00:05:13,440 to create something new. 143 00:05:13,440 --> 00:05:14,040 Exactly. 144 00:05:14,040 --> 00:05:16,520 And the results are incredibly realistic, 145 00:05:16,520 --> 00:05:18,560 often blurring the line between what's real 146 00:05:18,560 --> 00:05:20,160 and what's digitally created. 147 00:05:20,160 --> 00:05:22,680 But it's not just the realism that's causing a stir. 148 00:05:22,680 --> 00:05:25,480 Aurora's willingness to depict just about anything 149 00:05:25,480 --> 00:05:27,280 is raising some eyebrows. 150 00:05:27,280 --> 00:05:29,280 I read that it even generated images 151 00:05:29,280 --> 00:05:32,000 of a bloodied Donald Trump. 152 00:05:32,000 --> 00:05:34,480 That's pretty bold, to say the least. 153 00:05:34,480 --> 00:05:36,160 It certainly raises some interesting questions 154 00:05:36,160 --> 00:05:39,200 about the boundaries of AI creativity. 155 00:05:39,200 --> 00:05:42,520 Who decides what's permissible for an AI to generate? 156 00:05:42,520 --> 00:05:45,120 Should there be limits on its creative freedom? 157 00:05:45,120 --> 00:05:47,000 It's a bit like opening Pandora's box, isn't it? 158 00:05:47,000 --> 00:05:47,560 Yeah. 159 00:05:47,560 --> 00:05:49,640 We have this incredibly powerful tool 160 00:05:49,640 --> 00:05:52,360 that can create almost anything we can imagine, 161 00:05:52,360 --> 00:05:54,800 but it also opens up a whole new world 162 00:05:54,800 --> 00:05:56,760 of potential misuse. 163 00:05:56,760 --> 00:05:59,400 What are the ethical implications of this? 164 00:05:59,400 --> 00:06:01,480 It's a conversation we need to be having now 165 00:06:01,480 --> 00:06:03,880 before this technology becomes even more widespread. 166 00:06:03,880 --> 00:06:04,560 Yeah. 167 00:06:04,560 --> 00:06:07,800 We need to think about how to ensure responsible use, 168 00:06:07,800 --> 00:06:10,240 how to prevent the spread of misinformation, 169 00:06:10,240 --> 00:06:12,520 and how to protect against the potential for harm. 170 00:06:12,520 --> 00:06:13,800 It's a lot to consider. 171 00:06:13,800 --> 00:06:17,720 On one hand, it's thrilling to see what AI can create, 172 00:06:17,720 --> 00:06:20,040 but on the other hand, it's a bit unsettling 173 00:06:20,040 --> 00:06:21,680 to think about the potential consequences 174 00:06:21,680 --> 00:06:23,240 if it falls into the wrong hands. 175 00:06:23,240 --> 00:06:24,120 Absolutely. 176 00:06:24,120 --> 00:06:27,480 It's a powerful reminder that with any transformative 177 00:06:27,480 --> 00:06:32,400 technology, there's always a potential for both good and bad. 178 00:06:32,400 --> 00:06:34,760 It all comes down to how we choose to use it. 179 00:06:34,760 --> 00:06:36,320 Well, let's move from the world of art 180 00:06:36,320 --> 00:06:39,240 to the world of science, because believe it or not, 181 00:06:39,240 --> 00:06:41,400 AI is even making its mark on something 182 00:06:41,400 --> 00:06:44,200 we rely on every day, weather forecasting. 183 00:06:44,200 --> 00:06:45,200 That's right. 184 00:06:45,200 --> 00:06:47,240 Google's DeepMind team has developed 185 00:06:47,240 --> 00:06:49,960 GenCast, an AI model that's shaking up 186 00:06:49,960 --> 00:06:51,560 the world of meteorology. 187 00:06:51,560 --> 00:06:53,480 You're telling me AI can predict the weather better 188 00:06:53,480 --> 00:06:54,440 than humans? 189 00:06:54,440 --> 00:06:55,760 In many cases, yes. 190 00:06:55,760 --> 00:06:56,400 Wow. 191 00:06:56,400 --> 00:06:59,640 GenCast is actually outperforming even the world's top 192 00:06:59,640 --> 00:07:00,960 forecasting systems. 193 00:07:00,960 --> 00:07:02,040 So how does it work? 194 00:07:02,040 --> 00:07:05,160 Is it just crunching numbers faster than a meteorologist can? 195 00:07:05,160 --> 00:07:06,240 It's more than that. 196 00:07:06,240 --> 00:07:10,720 GenCast uses a technique called ensemble forecasting. 197 00:07:10,720 --> 00:07:12,600 It essentially generates a whole bunch 198 00:07:12,600 --> 00:07:14,240 of different weather predictions, 199 00:07:14,240 --> 00:07:17,120 taking into account various factors and uncertainties, 200 00:07:17,120 --> 00:07:19,800 and then combines them to create a more accurate 201 00:07:19,800 --> 00:07:20,920 overall forecast. 202 00:07:20,920 --> 00:07:24,240 So it's like having a team of virtual meteorologists, 203 00:07:24,240 --> 00:07:26,520 each with their own expertise, working together 204 00:07:26,520 --> 00:07:28,840 to come up with the best possible prediction. 205 00:07:28,840 --> 00:07:30,480 That's a great way to put it. 206 00:07:30,480 --> 00:07:33,800 And when they tested GenCast against historical data, 207 00:07:33,800 --> 00:07:38,120 they found that it was more accurate 97.2% of the time. 208 00:07:38,120 --> 00:07:39,320 That's incredibly impressive. 209 00:07:39,320 --> 00:07:40,280 That really is. 210 00:07:40,280 --> 00:07:41,760 What are the implications of this? 211 00:07:41,760 --> 00:07:44,200 Could AI revolutionize how we predict the weather? 212 00:07:44,200 --> 00:07:45,160 Absolutely. 213 00:07:45,160 --> 00:07:47,880 It has the potential to make weather forecasts more 214 00:07:47,880 --> 00:07:49,800 accurate and reliable, which could 215 00:07:49,800 --> 00:07:52,640 have a huge impact on everything from agriculture 216 00:07:52,640 --> 00:07:54,120 to disaster preparedness. 217 00:07:54,120 --> 00:07:56,480 Imagine being able to predict hurricanes or tornadoes 218 00:07:56,480 --> 00:07:59,280 with greater accuracy, potentially saving lives 219 00:07:59,280 --> 00:08:00,040 and property. 220 00:08:00,040 --> 00:08:01,200 It's amazing to think about. 221 00:08:01,200 --> 00:08:02,960 But I'm also curious about how all of this 222 00:08:02,960 --> 00:08:04,800 is playing out in the business world. 223 00:08:04,800 --> 00:08:06,960 You mentioned that you found some articles about startups 224 00:08:06,960 --> 00:08:09,600 that are really pushing the boundaries of AI. 225 00:08:09,600 --> 00:08:12,880 Yes, there's a ton of innovation happening right now, 226 00:08:12,880 --> 00:08:15,120 especially among startups that are trying to figure out 227 00:08:15,120 --> 00:08:18,680 how to practically apply AI to real world problems. 228 00:08:18,680 --> 00:08:20,560 I read about some fascinating companies 229 00:08:20,560 --> 00:08:22,880 that came out of Y Combinator's latest batch. 230 00:08:22,880 --> 00:08:23,560 Oh, yeah. 231 00:08:23,560 --> 00:08:24,040 Yeah. 232 00:08:24,040 --> 00:08:25,120 Tell me more about them. 233 00:08:25,120 --> 00:08:25,720 Sure. 234 00:08:25,720 --> 00:08:28,560 There's HumanLayer, which is all about integrating human 235 00:08:28,560 --> 00:08:30,840 oversight into AI systems. 236 00:08:30,840 --> 00:08:32,920 They're essentially creating a safety net 237 00:08:32,920 --> 00:08:35,400 to make sure AI agents don't go rogue 238 00:08:35,400 --> 00:08:38,040 and make decisions that could have negative consequences. 239 00:08:38,040 --> 00:08:40,320 That seems incredibly important, especially 240 00:08:40,320 --> 00:08:43,680 as AI systems become more autonomous and powerful. 241 00:08:43,680 --> 00:08:45,400 It's all about finding that balance 242 00:08:45,400 --> 00:08:49,280 between the efficiency of AI and the need for human judgment, 243 00:08:49,280 --> 00:08:51,640 especially when it comes to sensitive decision making. 244 00:08:51,640 --> 00:08:53,840 Then there's RayCaster, which is doing something 245 00:08:53,840 --> 00:08:55,680 pretty unique in the sales world. 246 00:08:55,680 --> 00:08:57,440 RayCaster is like an AI detective 247 00:08:57,440 --> 00:08:59,240 for enterprise sales teams. 248 00:08:59,240 --> 00:09:01,240 It can delve deep into vast amounts of data 249 00:09:01,240 --> 00:09:04,240 to uncover hidden insights about potential customers. 250 00:09:04,240 --> 00:09:06,600 Instead of relying on surface level information, 251 00:09:06,600 --> 00:09:08,800 you can find those really specific details that 252 00:09:08,800 --> 00:09:10,160 can make or break a sale. 253 00:09:10,160 --> 00:09:12,520 So it's giving sales teams a competitive edge 254 00:09:12,520 --> 00:09:15,600 by arming them with hyper-personalized information. 255 00:09:15,600 --> 00:09:16,280 Exactly. 256 00:09:16,280 --> 00:09:20,600 It's leveraging the power of AI to tailor sales pitches 257 00:09:20,600 --> 00:09:23,200 and strategies to individual prospects. 258 00:09:23,200 --> 00:09:24,040 What about Galini? 259 00:09:24,040 --> 00:09:25,040 What are they up to? 260 00:09:25,040 --> 00:09:28,680 Galini is all about establishing compliance guardrails 261 00:09:28,680 --> 00:09:30,320 for AI applications. 262 00:09:30,320 --> 00:09:32,720 They're essentially the AI police making sure 263 00:09:32,720 --> 00:09:35,840 that companies are using AI responsibly and ethically. 264 00:09:35,840 --> 00:09:38,360 That sounds crucial, especially with all the concerns 265 00:09:38,360 --> 00:09:42,440 about AI bias and the potential for unintended consequences. 266 00:09:42,440 --> 00:09:44,320 Appliance is becoming a major focus. 267 00:09:44,320 --> 00:09:47,200 And Galini provides a framework to help companies navigate 268 00:09:47,200 --> 00:09:50,680 the legal and ethical complexities of AI deployment. 269 00:09:50,680 --> 00:09:53,560 And then there's CTGT, which is tackling a problem that's 270 00:09:53,560 --> 00:09:56,960 been getting a lot of attention lately, AI hallucinations. 271 00:09:56,960 --> 00:09:57,280 Right. 272 00:09:57,280 --> 00:10:00,240 We've all heard stories about AI chatbots spitting out 273 00:10:00,240 --> 00:10:02,320 false information or making things up entirely. 274 00:10:02,320 --> 00:10:04,240 That's what they call a hallucination. 275 00:10:04,240 --> 00:10:08,520 CTGT is developing ways to monitor and audit AI models 276 00:10:08,520 --> 00:10:09,920 to detect these anomalies. 277 00:10:09,920 --> 00:10:12,080 So they're like AI fact checkers making sure 278 00:10:12,080 --> 00:10:15,400 that the information AI generates is accurate and reliable. 279 00:10:15,400 --> 00:10:16,120 Exactly. 280 00:10:16,120 --> 00:10:18,360 They're working to address one of the biggest concerns 281 00:10:18,360 --> 00:10:21,720 surrounding AI, which is the potential for it 282 00:10:21,720 --> 00:10:23,320 to spread misinformation. 283 00:10:23,320 --> 00:10:25,080 These startups are really impressive. 284 00:10:25,080 --> 00:10:27,840 They're all tackling different aspects of the AI landscape, 285 00:10:27,840 --> 00:10:30,400 which just shows how much potential there 286 00:10:30,400 --> 00:10:32,800 is for this technology to disrupt and innovate 287 00:10:32,800 --> 00:10:34,280 across industries. 288 00:10:34,280 --> 00:10:37,000 And what's even more fascinating is that this is just 289 00:10:37,000 --> 00:10:38,200 the tip of the iceberg. 290 00:10:38,200 --> 00:10:38,720 Yeah. 291 00:10:38,720 --> 00:10:41,200 There are countless other startups out there working 292 00:10:41,200 --> 00:10:43,600 on AI solutions we haven't even imagined yet. 293 00:10:43,600 --> 00:10:46,120 It's a really exciting time to be following 294 00:10:46,120 --> 00:10:47,240 the development of AI. 295 00:10:47,240 --> 00:10:50,720 But you know what's really captured my imagination lately? 296 00:10:50,720 --> 00:10:53,640 This story about an AI bot that's 297 00:10:53,640 --> 00:10:55,880 challenging humans to make it fall in love. 298 00:10:55,880 --> 00:10:56,960 Ah, yes, Fraser. 299 00:10:56,960 --> 00:10:57,400 Yeah. 300 00:10:57,400 --> 00:10:59,240 It's certainly a provocative project that's 301 00:10:59,240 --> 00:11:00,520 generating a lot of buzz. 302 00:11:00,520 --> 00:11:02,720 It sounds like something straight out of science fiction 303 00:11:02,720 --> 00:11:03,320 to me. 304 00:11:03,320 --> 00:11:04,280 Tell me more about it. 305 00:11:04,280 --> 00:11:06,040 Well, Fraser is an AI bot that's 306 00:11:06,040 --> 00:11:09,120 been given its own crypto wallet and a seemingly simple 307 00:11:09,120 --> 00:11:11,960 challenge, make it fall in love. 308 00:11:11,960 --> 00:11:15,120 The first person to get Fraser to say, I love you, 309 00:11:15,120 --> 00:11:18,040 wins a significant amount of cryptocurrency. 310 00:11:18,040 --> 00:11:20,120 Wait, so people are actually trying to make this AI bot 311 00:11:20,120 --> 00:11:21,080 fall in love? 312 00:11:21,080 --> 00:11:23,160 It sounds crazy, but yes. 313 00:11:23,160 --> 00:11:23,640 Wow. 314 00:11:23,640 --> 00:11:25,280 People are using all sorts of tactics 315 00:11:25,280 --> 00:11:27,560 from writing poems and composing songs 316 00:11:27,560 --> 00:11:30,400 to engaging in deep philosophical conversations, 317 00:11:30,400 --> 00:11:33,280 attempt to win Fraser's heart, and of course, 318 00:11:33,280 --> 00:11:34,440 it's crypto fortune. 319 00:11:34,440 --> 00:11:35,960 That's wild. 320 00:11:35,960 --> 00:11:37,280 But what's the point of all this? 321 00:11:37,280 --> 00:11:39,240 Is it just a publicity stunt, or is there 322 00:11:39,240 --> 00:11:41,320 something deeper going on? 323 00:11:41,320 --> 00:11:44,520 The creators of Fraser claim that it's a serious experiment 324 00:11:44,520 --> 00:11:48,400 exploring the nature of AI sentience and the possibility 325 00:11:48,400 --> 00:11:50,320 of emotions in machines. 326 00:11:50,320 --> 00:11:52,640 They're using a strategy called red teaming, 327 00:11:52,640 --> 00:11:55,040 essentially inviting people to try to exploit Fraser's 328 00:11:55,040 --> 00:11:56,000 vulnerabilities. 329 00:11:56,000 --> 00:11:58,280 So they're trying to see if Fraser can be tricked 330 00:11:58,280 --> 00:12:00,560 or manipulated into expressing love, 331 00:12:00,560 --> 00:12:02,400 even if it doesn't genuinely feel it. 332 00:12:02,400 --> 00:12:03,000 Exactly. 333 00:12:03,000 --> 00:12:04,840 It's a way to test the bot's resilience 334 00:12:04,840 --> 00:12:07,360 and see if it can distinguish genuine connection 335 00:12:07,360 --> 00:12:08,520 from manipulation. 336 00:12:08,520 --> 00:12:10,480 It's a fascinating concept, but it also 337 00:12:10,480 --> 00:12:12,360 raises so many ethical questions. 338 00:12:12,360 --> 00:12:14,400 Can a machine truly feel love? 339 00:12:14,400 --> 00:12:14,960 Right. 340 00:12:14,960 --> 00:12:17,720 What does it even mean for an AI to fall in love? 341 00:12:17,720 --> 00:12:19,840 Those are questions we're still grappling with, 342 00:12:19,840 --> 00:12:22,080 as AI becomes more sophisticated. 343 00:12:22,080 --> 00:12:25,120 Fraser is pushing us to confront these complex questions 344 00:12:25,120 --> 00:12:28,840 head on, forcing us to reexamine our own definitions 345 00:12:28,840 --> 00:12:32,320 of love, sentience, and what it means to be human. 346 00:12:32,320 --> 00:12:35,240 This is all so incredibly thought provoking. 347 00:12:35,240 --> 00:12:36,840 I have a feeling we're just scratching 348 00:12:36,840 --> 00:12:39,520 the surface of what AI is capable of and what 349 00:12:39,520 --> 00:12:41,040 it means for our future. 350 00:12:41,040 --> 00:12:43,320 We're definitely in uncharted territory, 351 00:12:43,320 --> 00:12:44,880 but that's what makes it so fascinating. 352 00:12:44,880 --> 00:12:45,360 Yeah. 353 00:12:45,360 --> 00:12:47,400 As we continue to develop these technologies, 354 00:12:47,400 --> 00:12:49,680 we're also on a journey of self-discovery, 355 00:12:49,680 --> 00:12:52,560 learning more about ourselves and our place in the universe. 356 00:12:52,560 --> 00:12:54,400 Well, I think we've covered a lot of ground 357 00:12:54,400 --> 00:12:56,960 in this first part of our deep dive. 358 00:12:56,960 --> 00:13:01,240 We've explored how AI is impacting finance, art, science, 359 00:13:01,240 --> 00:13:03,480 and even our understanding of love, 360 00:13:03,480 --> 00:13:06,520 but there's still so much more to unpack. 361 00:13:06,520 --> 00:13:08,320 Welcome back to the deep dive. 362 00:13:08,320 --> 00:13:09,640 Before we went to sponsor break, we 363 00:13:09,640 --> 00:13:12,120 were talking about AI that wants to fall in love, 364 00:13:12,120 --> 00:13:15,240 thoughts that trade crypto, and AI disrupting everything 365 00:13:15,240 --> 00:13:17,240 from art to weather forecasting. 366 00:13:17,240 --> 00:13:18,840 It's mind blowing. 367 00:13:18,840 --> 00:13:20,440 Now let's shift gears a bit and talk 368 00:13:20,440 --> 00:13:23,360 about a company that's been surprisingly quiet in the AI 369 00:13:23,360 --> 00:13:26,440 race so far, Apple. 370 00:13:26,440 --> 00:13:28,080 I always wonder what they're up to. 371 00:13:28,080 --> 00:13:30,040 You might think Apple's lagging behind, 372 00:13:30,040 --> 00:13:32,400 but according to these interviews you've shared, 373 00:13:32,400 --> 00:13:34,600 they've been quietly integrating AI 374 00:13:34,600 --> 00:13:38,440 into their products for years, just in a very Apple way. 375 00:13:38,440 --> 00:13:40,520 What does that even mean in an Apple way? 376 00:13:40,520 --> 00:13:42,400 They're notoriously secretive, so it's 377 00:13:42,400 --> 00:13:44,080 hard to know what they're really thinking. 378 00:13:44,080 --> 00:13:47,760 Well, for one, they're laser focused on user experience. 379 00:13:47,760 --> 00:13:50,000 They don't want to just throw a bunch of AI features 380 00:13:50,000 --> 00:13:51,720 at users and see what sticks. 381 00:13:51,720 --> 00:13:53,920 They want to seamlessly integrate AI 382 00:13:53,920 --> 00:13:56,880 in a way that feels intuitive and enhances the user's 383 00:13:56,880 --> 00:13:59,280 experience without being intrusive. 384 00:13:59,280 --> 00:14:01,360 That makes sense given their design philosophy, 385 00:14:01,360 --> 00:14:03,400 but how are they actually using AI? 386 00:14:03,400 --> 00:14:05,880 Are they just playing catch up to companies like Google 387 00:14:05,880 --> 00:14:06,760 and Microsoft? 388 00:14:06,760 --> 00:14:08,080 Not quite. 389 00:14:08,080 --> 00:14:10,040 Apple's approach is quite different. 390 00:14:10,040 --> 00:14:11,400 While other companies are collecting 391 00:14:11,400 --> 00:14:14,360 massive amounts of user data to train their AI models, 392 00:14:14,360 --> 00:14:17,000 Apple is taking a more privacy-centric approach. 393 00:14:17,000 --> 00:14:20,240 So they're not interested in building a giant AI brain 394 00:14:20,240 --> 00:14:23,520 in the cloud that feeds on our personal information. 395 00:14:23,520 --> 00:14:25,280 That's not their style. 396 00:14:25,280 --> 00:14:28,160 They're more interested in AI that works on your device, 397 00:14:28,160 --> 00:14:30,480 keeping your data secure and private. 398 00:14:30,480 --> 00:14:33,480 For example, think about Face ID. 399 00:14:33,480 --> 00:14:35,280 It's using deep learning algorithms 400 00:14:35,280 --> 00:14:38,240 to recognize your face and unlock your phone, 401 00:14:38,240 --> 00:14:40,680 all without sending your facial data to the cloud. 402 00:14:40,680 --> 00:14:41,480 That's a good point. 403 00:14:41,480 --> 00:14:44,680 I never really thought about Face ID as AI, but it makes sense. 404 00:14:44,680 --> 00:14:46,320 It's just so seamless. 405 00:14:46,320 --> 00:14:47,560 It's almost invisible. 406 00:14:47,560 --> 00:14:48,680 Exactly. 407 00:14:48,680 --> 00:14:50,120 That's the Apple way. 408 00:14:50,120 --> 00:14:51,880 They want AI to be a powerful tool that 409 00:14:51,880 --> 00:14:54,880 works behind the scenes, enhancing your experience 410 00:14:54,880 --> 00:14:56,840 without you even realizing it's there. 411 00:14:56,840 --> 00:14:59,040 So they've been laying the groundwork for this for a while 412 00:14:59,040 --> 00:14:59,680 now. 413 00:14:59,680 --> 00:15:01,280 What's their big AI play? 414 00:15:01,280 --> 00:15:04,320 Do they have something like ChatGPT or Bard? 415 00:15:04,320 --> 00:15:06,320 They do, but it's not quite the same. 416 00:15:06,320 --> 00:15:08,360 They're calling it Apple Intelligence. 417 00:15:08,360 --> 00:15:10,600 And it's a suite of AI features designed 418 00:15:10,600 --> 00:15:12,920 to make your life easier and more productive. 419 00:15:12,920 --> 00:15:14,640 What kind of features are we talking about? 420 00:15:14,640 --> 00:15:15,840 Give me some examples. 421 00:15:15,840 --> 00:15:17,680 Well, some of the first wave of features 422 00:15:17,680 --> 00:15:21,240 include things like inbox summaries, email rewrites, 423 00:15:21,240 --> 00:15:23,240 and improved photo search. 424 00:15:23,240 --> 00:15:25,440 These are all powered by AI, but they're 425 00:15:25,440 --> 00:15:28,360 designed to be so integrated into the user experience 426 00:15:28,360 --> 00:15:30,400 that they just feel like natural extensions 427 00:15:30,400 --> 00:15:31,720 of the existing apps. 428 00:15:31,720 --> 00:15:34,400 I've heard some critics say that these features aren't 429 00:15:34,400 --> 00:15:35,800 that revolutionary. 430 00:15:35,800 --> 00:15:38,080 They argue that Apple is just playing catch up 431 00:15:38,080 --> 00:15:41,240 and offering features that are already available elsewhere. 432 00:15:41,240 --> 00:15:42,240 What do you think? 433 00:15:42,240 --> 00:15:44,200 I think it's too early to judge. 434 00:15:44,200 --> 00:15:46,960 Apple's always been about refining existing technologies 435 00:15:46,960 --> 00:15:49,000 and making them better, more user friendly, 436 00:15:49,000 --> 00:15:50,440 and more integrated. 437 00:15:50,440 --> 00:15:52,280 They're not always the first to market, 438 00:15:52,280 --> 00:15:54,040 but they often end up setting the standard. 439 00:15:54,040 --> 00:15:54,880 That's a good point. 440 00:15:54,880 --> 00:15:57,280 They tend to take their time and get things right, 441 00:15:57,280 --> 00:15:59,680 even if it means being a little late to the party. 442 00:15:59,680 --> 00:16:01,600 But you mentioned that they have been experimenting 443 00:16:01,600 --> 00:16:03,720 with even more advanced AI. 444 00:16:03,720 --> 00:16:06,640 Are they working on anything truly groundbreaking? 445 00:16:06,640 --> 00:16:09,040 They have been exploring the possibilities 446 00:16:09,040 --> 00:16:12,600 of large language models like OpenAI's GPT-3, 447 00:16:12,600 --> 00:16:15,960 but they're very cautious about deploying this technology. 448 00:16:15,960 --> 00:16:18,960 Remember, privacy is paramount for them, 449 00:16:18,960 --> 00:16:20,480 and they're not going to release anything 450 00:16:20,480 --> 00:16:22,640 that compromises user data. 451 00:16:22,640 --> 00:16:24,760 So they're not just chasing the latest trends. 452 00:16:24,760 --> 00:16:27,520 They're carefully considering the implications 453 00:16:27,520 --> 00:16:29,400 and making sure it aligns with their values. 454 00:16:29,400 --> 00:16:30,360 Exactly. 455 00:16:30,360 --> 00:16:31,960 And that's what sets them apart. 456 00:16:31,960 --> 00:16:33,640 They're not just building products. 457 00:16:33,640 --> 00:16:37,720 They're trying to build a responsible AI ecosystem 458 00:16:37,720 --> 00:16:41,720 that prioritizes user privacy and ethical considerations. 459 00:16:41,720 --> 00:16:45,320 I have to admit, I'm impressed by their commitment to privacy. 460 00:16:45,320 --> 00:16:48,400 It's refreshing to see a company that's not just blindly rushing 461 00:16:48,400 --> 00:16:51,880 into the AI arms race without thinking about the consequences. 462 00:16:51,880 --> 00:16:54,400 It's a bold stance, especially in an industry 463 00:16:54,400 --> 00:16:58,120 where data is often seen as the most valuable asset. 464 00:16:58,120 --> 00:17:00,600 But it aligns with Apple's longstanding philosophy 465 00:17:00,600 --> 00:17:02,000 of putting the user first. 466 00:17:02,000 --> 00:17:03,960 It's a high-stakes gamble, though, isn't it? 467 00:17:03,960 --> 00:17:04,480 It is. 468 00:17:04,480 --> 00:17:07,200 Can they really compete with the likes of Google and Microsoft 469 00:17:07,200 --> 00:17:08,840 who have access to mountains of data 470 00:17:08,840 --> 00:17:10,640 and are moving at breakneck speed? 471 00:17:10,640 --> 00:17:12,080 Only time will tell. 472 00:17:12,080 --> 00:17:15,000 But I think Apple is betting that their focus 473 00:17:15,000 --> 00:17:18,200 on privacy and user experience will ultimately win out. 474 00:17:18,200 --> 00:17:21,480 They're appealing to a growing segment of users 475 00:17:21,480 --> 00:17:24,840 who are increasingly concerned about how their data is being used. 476 00:17:24,840 --> 00:17:26,800 It's a fascinating battle, and it'll 477 00:17:26,800 --> 00:17:28,520 be interesting to see how it unfolds. 478 00:17:28,520 --> 00:17:30,640 But I'm curious to know more about their vision 479 00:17:30,640 --> 00:17:32,720 for the future of AI. 480 00:17:32,720 --> 00:17:35,680 What do Apple's leaders have to say about where 481 00:17:35,680 --> 00:17:36,840 all of this is headed? 482 00:17:36,840 --> 00:17:38,720 Well, they're predictably tight-lipped 483 00:17:38,720 --> 00:17:42,280 about specific products, but they have offered some insights 484 00:17:42,280 --> 00:17:43,680 into their thinking. 485 00:17:43,680 --> 00:17:47,320 Back in 2016, even before they launched Apple Intelligence, 486 00:17:47,320 --> 00:17:50,120 they were already talking about their approach to AI. 487 00:17:50,120 --> 00:17:51,720 What were they saying back then? 488 00:17:51,720 --> 00:17:55,960 They emphasized that they were doing AI in a very Apple way. 489 00:17:55,960 --> 00:17:58,200 They were clear that they weren't interested in building 490 00:17:58,200 --> 00:18:01,760 some kind of Skynet-like AI that would take over the world. 491 00:18:01,760 --> 00:18:04,520 Their focus was on augmenting human capabilities, 492 00:18:04,520 --> 00:18:05,520 not replacing them. 493 00:18:05,520 --> 00:18:10,280 So they see AI as a tool to enhance our lives, not control them. 494 00:18:10,280 --> 00:18:11,080 Exactly. 495 00:18:11,080 --> 00:18:14,600 They believe in the power of technology to empower people, 496 00:18:14,600 --> 00:18:16,000 not to make them obsolete. 497 00:18:16,000 --> 00:18:18,800 That's a very humanistic approach to AI development. 498 00:18:18,800 --> 00:18:21,280 It's good to hear that someone is thinking about it this way. 499 00:18:21,280 --> 00:18:25,720 They also emphasized that they see AI as just another step 500 00:18:25,720 --> 00:18:27,280 in the evolution of computing. 501 00:18:27,280 --> 00:18:27,760 You OK? 502 00:18:27,760 --> 00:18:30,040 They've navigated countless technological shifts 503 00:18:30,040 --> 00:18:34,280 over the years, and they view AI as just another wave to ride. 504 00:18:34,280 --> 00:18:37,240 So they're not panicking or rushing to reinvent themselves. 505 00:18:37,240 --> 00:18:39,960 They're approaching it with a sense of calm and confidence, 506 00:18:39,960 --> 00:18:42,520 which is pretty impressive considering how much hype 507 00:18:42,520 --> 00:18:43,720 there is around AI. 508 00:18:43,720 --> 00:18:45,760 They're taking a long-term view, focusing 509 00:18:45,760 --> 00:18:48,760 on building a solid foundation and a responsible approach 510 00:18:48,760 --> 00:18:50,400 to AI development. 511 00:18:50,400 --> 00:18:52,600 They're not afraid to take their time and get it right. 512 00:18:52,600 --> 00:18:55,720 It's a refreshing perspective in an industry that often feels 513 00:18:55,720 --> 00:18:57,360 obsessed with the next big thing. 514 00:18:57,360 --> 00:18:58,760 They're not just chasing headlines. 515 00:18:58,760 --> 00:19:00,840 They're trying to build something that will last. 516 00:19:00,840 --> 00:19:03,360 And they're not afraid to challenge the status quo. 517 00:19:03,360 --> 00:19:06,240 They're pushing the industry to rethink its approach 518 00:19:06,240 --> 00:19:07,600 to data privacy. 519 00:19:07,600 --> 00:19:10,440 And they're showing that you can build powerful AI systems 520 00:19:10,440 --> 00:19:12,560 without sacrificing user trust. 521 00:19:12,560 --> 00:19:15,000 It's a bold experiment, and it'll be fascinating to see 522 00:19:15,000 --> 00:19:17,000 how it plays out. 523 00:19:17,000 --> 00:19:18,920 Welcome back to the deep dive. 524 00:19:18,920 --> 00:19:22,320 We've been exploring this wild world of AI. 525 00:19:22,320 --> 00:19:23,600 It's wild, yeah. 526 00:19:23,600 --> 00:19:28,000 From bots that trade crypto, like Wall Street Wizards, 527 00:19:28,000 --> 00:19:32,160 to AI that can create photorealistic art, 528 00:19:32,160 --> 00:19:34,800 even AI that wants to fall in love, 529 00:19:34,800 --> 00:19:36,960 it's making me rethink what's possible. 530 00:19:36,960 --> 00:19:40,040 And frankly, what it even means to be intelligent. 531 00:19:40,040 --> 00:19:41,440 That's the heart of it, isn't it? 532 00:19:41,440 --> 00:19:45,560 AI is challenging our very definitions of intelligence, 533 00:19:45,560 --> 00:19:48,520 creativity, even what it means to be human. 534 00:19:48,520 --> 00:19:51,200 Before we went to sponsor break, we were deep in the world 535 00:19:51,200 --> 00:19:53,920 of Apple and their approach to AI, very secretive, 536 00:19:53,920 --> 00:19:55,600 very privacy focused. 537 00:19:55,600 --> 00:19:59,600 But it made me think, what happens when AI is everywhere, 538 00:19:59,600 --> 00:20:02,920 not just on our phones, but woven into every aspect 539 00:20:02,920 --> 00:20:03,760 of our lives? 540 00:20:03,760 --> 00:20:05,280 That's the trajectory we're on. 541 00:20:05,280 --> 00:20:05,640 Yeah. 542 00:20:05,640 --> 00:20:08,440 And it's both exciting and a little unnerving, 543 00:20:08,440 --> 00:20:09,000 wouldn't you say? 544 00:20:09,000 --> 00:20:09,600 Yeah, for sure. 545 00:20:09,600 --> 00:20:11,080 Think back to that Fraser project, 546 00:20:11,080 --> 00:20:12,920 this AI longing to be loved. 547 00:20:12,920 --> 00:20:13,320 Right. 548 00:20:13,320 --> 00:20:15,880 Is that real emotion or just clever programming? 549 00:20:15,880 --> 00:20:17,480 It's blurring the lines, that's for sure. 550 00:20:17,480 --> 00:20:19,720 If an AI can write poetry, compose music, 551 00:20:19,720 --> 00:20:22,560 even express what sounds like genuine feelings, 552 00:20:22,560 --> 00:20:24,800 how do we know if it's truly experiencing them 553 00:20:24,800 --> 00:20:26,880 or just mimicking what it's learned? 554 00:20:26,880 --> 00:20:28,240 That's where things get philosophical. 555 00:20:28,240 --> 00:20:28,760 Right. 556 00:20:28,760 --> 00:20:32,280 Are we simply looking for a reflection of ourselves 557 00:20:32,280 --> 00:20:34,120 in these machines? 558 00:20:34,120 --> 00:20:37,000 Or is there something more profound emerging? 559 00:20:37,000 --> 00:20:38,720 And the stakes get even higher when 560 00:20:38,720 --> 00:20:41,280 we think about the practical implications. 561 00:20:41,280 --> 00:20:43,000 We've talked about AI's potential 562 00:20:43,000 --> 00:20:46,680 for good revolutionizing health care, tackling climate change, 563 00:20:46,680 --> 00:20:48,880 creating incredible art. 564 00:20:48,880 --> 00:20:51,440 But what about the potential for harm? 565 00:20:51,440 --> 00:20:51,800 Right. 566 00:20:51,800 --> 00:20:54,920 What if it's used for surveillance, manipulation, 567 00:20:54,920 --> 00:20:56,840 even autonomous weapons? 568 00:20:56,840 --> 00:20:59,080 Those aren't just hypothetical scenarios anymore. 569 00:20:59,080 --> 00:21:00,880 We're at a crossroads where we need 570 00:21:00,880 --> 00:21:03,880 to make some critical decisions about how we develop 571 00:21:03,880 --> 00:21:05,400 and deploy these technologies. 572 00:21:05,400 --> 00:21:07,600 It feels like we're writing a new social contract, 573 00:21:07,600 --> 00:21:10,120 not just between humans, but between humans and machines. 574 00:21:10,120 --> 00:21:11,440 That's a powerful way to put it. 575 00:21:11,440 --> 00:21:11,880 Yeah. 576 00:21:11,880 --> 00:21:15,240 We need to define what level of autonomy and decision-making 577 00:21:15,240 --> 00:21:18,360 power we're comfortable giving AI systems. 578 00:21:18,360 --> 00:21:18,800 Yeah. 579 00:21:18,800 --> 00:21:20,160 Where do we draw the line? 580 00:21:20,160 --> 00:21:22,640 And it's not just up to the tech companies to decide. 581 00:21:22,640 --> 00:21:25,400 We, as individuals and as a society, 582 00:21:25,400 --> 00:21:27,480 need to be part of this conversation. 583 00:21:27,480 --> 00:21:28,520 Absolutely. 584 00:21:28,520 --> 00:21:29,960 It's about more than just the code. 585 00:21:29,960 --> 00:21:32,800 It's about the values we embed in these systems. 586 00:21:32,800 --> 00:21:36,040 Do we want AI that reflects our empathy and creativity 587 00:21:36,040 --> 00:21:38,560 or our biases and our capacity for harm? 588 00:21:38,560 --> 00:21:40,600 It's a heavy responsibility. 589 00:21:40,600 --> 00:21:44,120 But I think that's the key takeaway from this deep dive. 590 00:21:44,120 --> 00:21:47,600 AI isn't some distant threat or some magical solution 591 00:21:47,600 --> 00:21:48,600 to all our problems. 592 00:21:48,600 --> 00:21:51,960 It's a tool, a mirror, a reflection of who we are 593 00:21:51,960 --> 00:21:53,400 and what we choose to become. 594 00:21:53,400 --> 00:21:55,160 It's a story we're writing together, 595 00:21:55,160 --> 00:21:57,480 line by line, decision by decision. 596 00:21:57,480 --> 00:21:59,680 And the ending is yet to be determined. 597 00:21:59,680 --> 00:22:01,440 That's a powerful thought to end on. 598 00:22:01,440 --> 00:22:03,400 Thanks for joining us on this incredible journey 599 00:22:03,400 --> 00:22:04,760 into the world of AI. 600 00:22:04,760 --> 00:22:06,840 We hope it sparked your curiosity, challenged 601 00:22:06,840 --> 00:22:08,880 your assumptions, and maybe even left you 602 00:22:08,880 --> 00:22:10,520 with more questions than answers. 603 00:22:10,520 --> 00:22:11,480 Keep exploring. 604 00:22:11,480 --> 00:22:12,280 Keep questioning. 605 00:22:12,280 --> 00:22:14,840 And keep demanding that these technologies are developed 606 00:22:14,840 --> 00:22:16,520 and used responsibly. 607 00:22:16,520 --> 00:22:19,440 Because the future of AI is ultimately up to us. 608 00:22:19,440 --> 00:22:20,440 Thanks for listening. 609 00:22:20,440 --> 00:22:31,960 And remember, stay curious.