WEBVTT 00:00:00.000 --> 00:00:01.720 Pancreatic cancer. It's often called the king 00:00:01.720 --> 00:00:04.719 of cancers and for a pretty terrifying reason. 00:00:04.839 --> 00:00:07.940 Yeah. The five year survival rate is shockingly 00:00:07.940 --> 00:00:10.759 low. It's below 10 percent. And that's because 00:00:10.759 --> 00:00:14.179 in 80 percent of cases we just we find it far 00:00:14.179 --> 00:00:16.120 too late. It basically hides in plain sight. 00:00:16.429 --> 00:00:20.010 But imagine an AI tool. It's analyzing just routine 00:00:20.010 --> 00:00:23.030 hospital scans, searching silently for signs 00:00:23.030 --> 00:00:25.089 that are, you know, almost impossible to see. 00:00:25.230 --> 00:00:29.269 OK. That tool recently found 31 early stage cancers 00:00:29.269 --> 00:00:32.409 that human doctors. who are incredible, by the 00:00:32.409 --> 00:00:35.570 way, had missed entirely. That's the actual measurable 00:00:35.570 --> 00:00:37.869 power we're tracking. Welcome to the deep dive. 00:00:38.049 --> 00:00:40.329 Our mission today is really laser focused. We're 00:00:40.329 --> 00:00:41.990 going to sift through the newest stack of AI 00:00:41.990 --> 00:00:44.409 intelligence and just pull out the crucial insights 00:00:44.409 --> 00:00:46.009 you need. We're looking for the stuff that shifts 00:00:46.009 --> 00:00:48.090 your perspective. Exactly. Covering everything 00:00:48.090 --> 00:00:50.369 from these huge global strategic moves to the 00:00:50.369 --> 00:00:53.609 immediate practical hacks you can use today. 00:00:54.000 --> 00:00:55.679 Wallet saving stuff. And we have a clear path 00:00:55.679 --> 00:00:57.820 laid out for you. So first, we're going to unpack 00:00:57.820 --> 00:01:01.500 a massive high stakes two billion dollar acquisition 00:01:01.500 --> 00:01:05.319 by Meta. And that deal has now triggered a really 00:01:05.319 --> 00:01:08.379 intense geopolitical standoff with China. It's 00:01:08.379 --> 00:01:10.939 a fight over talent. Then we pivot straight to 00:01:10.939 --> 00:01:13.280 the practical. We get into strategies for getting 00:01:13.280 --> 00:01:16.140 like 10 times better outputs from models like 00:01:16.140 --> 00:01:18.840 Claude and Gemini. This is about saving you time 00:01:18.840 --> 00:01:21.319 and money. Right now. Right. And next, we're 00:01:21.319 --> 00:01:23.780 mapping the whole industry landscape. We're talking 00:01:23.780 --> 00:01:27.519 XAI's monumental $20 billion fundraising round 00:01:27.519 --> 00:01:31.599 and OpenAI's push toward a proactive super -assistant 00:01:31.599 --> 00:01:34.439 future. And finally, we will return to that life 00:01:34.439 --> 00:01:37.980 -saving breakthrough, the Panda AI that is fundamentally 00:01:37.980 --> 00:01:40.939 changing how medicine screens for, well, the 00:01:40.939 --> 00:01:43.750 deadliest cancer. OK. Let's get into this. Let's 00:01:43.750 --> 00:01:45.890 start with geopolitics, because the stakes here, 00:01:45.969 --> 00:01:48.609 they really couldn't be higher. Meta's $2 billion 00:01:48.609 --> 00:01:52.069 acquisition of Manus. The AI assistant startup. 00:01:52.209 --> 00:01:53.870 Right. It's not just a business deal. It's become 00:01:53.870 --> 00:01:56.590 a geopolitical flashpoint. And the story isn't 00:01:56.590 --> 00:01:59.989 that Meta just wanted a shiny new toy. It's all 00:01:59.989 --> 00:02:02.090 about where Manus started. Where it was based. 00:02:02.329 --> 00:02:04.769 Exactly. Manus was originally headquartered in 00:02:04.769 --> 00:02:07.310 Beijing, and that alone instantly made the whole 00:02:07.310 --> 00:02:10.280 thing sensitive. And we know that Benchmark, 00:02:10.419 --> 00:02:13.240 an early U .S. investor, had already drawn scrutiny 00:02:13.240 --> 00:02:15.879 from U .S. lawmakers. Right. And that was under 00:02:15.879 --> 00:02:18.419 new rules that were specifically designed to 00:02:18.419 --> 00:02:21.020 tighten the flow of American money into Chinese 00:02:21.020 --> 00:02:25.159 AI. So Manus made a move. They relocated their 00:02:25.159 --> 00:02:27.780 headquarters, their key infrastructure, everything 00:02:27.780 --> 00:02:31.240 from Beijing to Singapore. And people in the 00:02:31.240 --> 00:02:33.719 industry are already calling that Singapore washing. 00:02:34.300 --> 00:02:37.000 That term is so telling, isn't it? It is. It's 00:02:37.000 --> 00:02:39.759 a strategic shift, you know, meant to insulate 00:02:39.759 --> 00:02:42.120 the company from that U .S. scrutiny by moving 00:02:42.120 --> 00:02:44.879 to a neutral tech hub. And at first it worked 00:02:44.879 --> 00:02:46.840 perfectly. I mean, from Meta's perspective, this 00:02:46.840 --> 00:02:48.680 was just pure acceleration. Why build when you 00:02:48.680 --> 00:02:51.039 can buy? Exactly. They didn't have to spend years 00:02:51.039 --> 00:02:53.379 building a functional agent stack from scratch. 00:02:53.560 --> 00:02:56.840 They bought a pre -built working system. It massively 00:02:56.840 --> 00:02:59.860 fast -tracked their whole AI roadmap. But what's 00:02:59.860 --> 00:03:01.680 really fascinating is how Beijing responded. 00:03:02.219 --> 00:03:04.520 They noticed. They saw the pattern. Now, Chinese 00:03:04.520 --> 00:03:06.759 regulators are investigating the deal, checking 00:03:06.759 --> 00:03:09.319 for violations of their own tech export controls. 00:03:09.599 --> 00:03:12.180 They're asking very specific questions. Super 00:03:12.180 --> 00:03:15.960 pointed. Like, did Manus exploit restricted AI 00:03:15.960 --> 00:03:18.800 technology when it moved its operations to Singapore? 00:03:19.419 --> 00:03:22.400 And did it need an export license to move that 00:03:22.400 --> 00:03:24.599 foundational code out of the country? This goes 00:03:24.599 --> 00:03:27.419 way beyond a simple business spat. The bigger 00:03:27.419 --> 00:03:29.539 picture here is what Beijing is responding to. 00:03:29.659 --> 00:03:32.560 Right. They're worried. That this whole sequence, 00:03:32.819 --> 00:03:36.139 build in China, relocate, get acquired by a U 00:03:36.139 --> 00:03:38.819 .S. giant, establishes a dangerous playbook. 00:03:39.000 --> 00:03:40.840 A playbook they don't want other startups to 00:03:40.840 --> 00:03:43.319 follow. They don't want Chinese AI, which was 00:03:43.319 --> 00:03:45.860 maybe built with state resources or talent, to 00:03:45.860 --> 00:03:48.240 just get snapped up by U .S. tech giants. It's 00:03:48.240 --> 00:03:50.780 a huge outflow of talent and strategic assets. 00:03:51.020 --> 00:03:53.219 And they've used this tactic before. With TikTok. 00:03:53.419 --> 00:03:55.560 With TikTok, yeah. They successfully used their 00:03:55.560 --> 00:03:58.139 tech export controls a few years ago to block 00:03:58.139 --> 00:04:00.560 the attempted ban or sale. It's a critical national 00:04:00.560 --> 00:04:03.360 strategy for them. So beyond just the money, 00:04:03.439 --> 00:04:05.800 what is Beijing trying to signal with this investigation? 00:04:06.219 --> 00:04:08.000 I think they see this as a critical national 00:04:08.000 --> 00:04:11.300 strategy and they're using past success to control 00:04:11.300 --> 00:04:13.800 that talent outflow. Makes sense. So while that 00:04:13.800 --> 00:04:16.160 national strategy stuff plays out. Let's pivot 00:04:16.160 --> 00:04:18.939 to how we can use the tech we already have more 00:04:18.939 --> 00:04:21.100 efficiently. Yeah, let's get practical. We want 00:04:21.100 --> 00:04:23.480 to deliver some immediate value here. We're moving 00:04:23.480 --> 00:04:26.459 from billion -dollar moves to actionable tools 00:04:26.459 --> 00:04:29.060 that can make the AI services you already pay 00:04:29.060 --> 00:04:32.600 for just dramatically better today. Okay, let's 00:04:32.600 --> 00:04:34.839 start with workflow management. It can still 00:04:34.839 --> 00:04:37.420 be so messy. We saw the innate and chat hub guide 00:04:37.420 --> 00:04:39.980 this week, and it's a huge step forward. So what 00:04:39.980 --> 00:04:42.540 is it? It lets you, and this is the key, talk 00:04:42.540 --> 00:04:44.899 to your automations. You can interrupt a process 00:04:44.899 --> 00:04:48.180 that's running and fix issues in real time. That's 00:04:48.180 --> 00:04:51.759 a revelation because it takes those messy, sprawling 00:04:51.759 --> 00:04:53.860 workflows, the ones that break at 3 in the morning, 00:04:53.980 --> 00:04:56.920 and turns them into one clean conversational 00:04:56.920 --> 00:04:59.180 control center. Right. And that real -time fixing 00:04:59.180 --> 00:05:01.519 is key because managing prompts can still be 00:05:01.519 --> 00:05:03.920 a headache for, well, for everyone. It really 00:05:03.920 --> 00:05:06.699 can. Which brings us to the 10 prompt rules. 00:05:07.689 --> 00:05:10.689 Whether you're using Claude or GPT or Gemini, 00:05:10.810 --> 00:05:13.329 following some basic rules is just a necessity 00:05:13.329 --> 00:05:16.029 now for getting better outputs. And they're simple 00:05:16.029 --> 00:05:18.750 tweaks, but they give you massive quality improvements. 00:05:19.069 --> 00:05:22.110 For instance, a crucial rule is using negative 00:05:22.110 --> 00:05:25.069 constraints. So you don't just tell the model 00:05:25.069 --> 00:05:27.310 what you want. You tell it what you don't want 00:05:27.310 --> 00:05:30.189 explicitly, like do not use bullet points or 00:05:30.189 --> 00:05:33.000 ensure the tone is not academic. You have to 00:05:33.000 --> 00:05:35.600 give the machine detailed blueprints if you want 00:05:35.600 --> 00:05:38.339 high quality work. It's true. I still wrestle 00:05:38.339 --> 00:05:40.560 with prompt drift myself. Yeah. Especially on 00:05:40.560 --> 00:05:42.680 longer projects where I'm trying to maintain 00:05:42.680 --> 00:05:45.000 a consistent voice. That's a really common struggle. 00:05:45.180 --> 00:05:47.480 I think it hits everyone. And for anyone listening, 00:05:47.639 --> 00:05:51.439 prompt drift is just... The tendency of an AI 00:05:51.439 --> 00:05:54.480 to lose context or, you know, for the quality 00:05:54.480 --> 00:05:57.500 to degrade over a long chat session. Yeah, and 00:05:57.500 --> 00:05:59.519 it happens mostly because of how these models 00:05:59.519 --> 00:06:02.060 manage memory. They use a kind of sliding window, 00:06:02.180 --> 00:06:04.899 so they prioritize the most recent stuff you've 00:06:04.899 --> 00:06:06.939 typed. So the further you get from your first 00:06:06.939 --> 00:06:09.000 instruction, the less focus the model retains 00:06:09.000 --> 00:06:10.920 on it. Which is exactly why those prompt rules 00:06:10.920 --> 00:06:13.060 are so vital. They help reinforce the original 00:06:13.060 --> 00:06:15.220 instructions. And we also need to talk about 00:06:15.220 --> 00:06:18.480 model efficiency, specifically Google's Gemini 00:06:18.480 --> 00:06:21.439 3 Flash. The super cheap model. Right, which 00:06:21.439 --> 00:06:23.920 is great. But what's surprising is that for certain 00:06:23.920 --> 00:06:26.519 tasks, it actually outperforms the more expensive 00:06:26.519 --> 00:06:29.660 pro model in practice. It forces you to choose 00:06:29.660 --> 00:06:32.120 the right tool for the job. Flash is incredibly 00:06:32.120 --> 00:06:34.860 fast and cheap, so it's perfect for quick summarization 00:06:34.860 --> 00:06:37.639 or data extraction. Or just generating initial 00:06:37.639 --> 00:06:40.379 creative drafts where speed matters more than, 00:06:40.420 --> 00:06:44.129 say, deep reasoning. Exactly. But you'd switch 00:06:44.129 --> 00:06:47.449 to Pro when you need complex logic or financial 00:06:47.449 --> 00:06:51.170 analysis or detailed code generation. So knowing 00:06:51.170 --> 00:06:53.850 when to use Flash versus Pro or when to switch 00:06:53.850 --> 00:06:56.509 to that frontier level performance like the rumored 00:06:56.509 --> 00:07:00.009 GPT 5 .2, that's the new efficiency hack. So 00:07:00.009 --> 00:07:02.350 for the daily user just trying to save time and 00:07:02.350 --> 00:07:05.000 money. What's the single biggest takeaway from 00:07:05.000 --> 00:07:07.699 all these tools and rules? It's that small tweaks 00:07:07.699 --> 00:07:10.100 to your prompting and your model choice can yield 00:07:10.100 --> 00:07:12.860 massive cost -effective improvements. That idea 00:07:12.860 --> 00:07:15.300 of leveraging the right tool is key. It really 00:07:15.300 --> 00:07:17.480 is. We're going to take a quick pause here, and 00:07:17.480 --> 00:07:19.639 when we're back, we'll map out the explosive 00:07:19.639 --> 00:07:22.050 race for the personal super assistant. Welcome 00:07:22.050 --> 00:07:24.250 back. We just talked about maximizing your personal 00:07:24.250 --> 00:07:27.290 efficiency. Now let's look at the strategic direction 00:07:27.290 --> 00:07:29.750 of the major players. Yeah, it's pretty clear 00:07:29.750 --> 00:07:33.110 that everyone, OpenAI, Amazon, XAI, they're all 00:07:33.110 --> 00:07:35.870 racing toward a personalized, proactive future. 00:07:36.089 --> 00:07:38.529 Absolutely. And we saw that monumental viral 00:07:38.529 --> 00:07:40.850 post. Oh, the one from the Google principal engineer. 00:07:41.550 --> 00:07:44.470 He claimed a clawed code compressed an entire 00:07:44.470 --> 00:07:46.670 year's worth of his software development work 00:07:46.670 --> 00:07:49.670 into just one hour. One hour. That post just 00:07:49.670 --> 00:07:52.990 exploded. It hit over 8 .2 million views. And 00:07:52.990 --> 00:07:54.790 whether you take that number literally or not, 00:07:54.970 --> 00:07:59.149 the perceived efficiency is just... It really 00:07:59.149 --> 00:08:01.490 speaks to the raw leverage these code models 00:08:01.490 --> 00:08:03.769 offer. We know Claude's creator often runs up 00:08:03.769 --> 00:08:05.990 to 15 parallel sessions at once. That's some 00:08:05.990 --> 00:08:08.250 serious leverage. Yeah. And that kind of momentum 00:08:08.250 --> 00:08:10.509 is feeding directly into the roadmaps of the 00:08:10.509 --> 00:08:13.069 industry giants. It is. OpenAI's chief product 00:08:13.069 --> 00:08:16.029 officer, Fiji Saimo, has been very clear. They 00:08:16.029 --> 00:08:18.649 want ChatGPT to evolve into a personal super 00:08:18.649 --> 00:08:21.490 assistant by 2026. And this is a huge shift. 00:08:21.550 --> 00:08:23.329 It's not just about answering a question when 00:08:23.329 --> 00:08:25.230 you type it. No, it's about being proactive. 00:08:26.289 --> 00:08:29.329 Monitoring your email for key events, pre -filling 00:08:29.329 --> 00:08:31.949 forms for you, anticipating your needs before 00:08:31.949 --> 00:08:34.549 you even think of them. A truly intelligent layer 00:08:34.549 --> 00:08:36.830 on top of your whole digital life. Meanwhile, 00:08:37.090 --> 00:08:39.870 Amazon is making its own play for the home ecosystem. 00:08:40.370 --> 00:08:44.929 They launched Alexa Plus at CES 2026. Right. 00:08:45.009 --> 00:08:48.159 Pitched as a family first. web chat bot the idea 00:08:48.159 --> 00:08:50.799 is it controls the smart home manages your appliances 00:08:50.799 --> 00:08:53.379 updates the family calendar it's a very amazon 00:08:53.379 --> 00:08:56.220 play integrating ai directly into the family 00:08:56.220 --> 00:08:59.279 unit the digital hearth. But none of this co 00:08:59.279 --> 00:09:01.879 -active future works without just immense raw 00:09:01.879 --> 00:09:04.460 computational power. Which brings us to the infrastructure 00:09:04.460 --> 00:09:08.259 battle. And XAI just closed a monumental $20 00:09:08.259 --> 00:09:11.279 billion Series E funding round. $20 billion. 00:09:11.399 --> 00:09:14.299 That's a truly monumental number. And it tells 00:09:14.299 --> 00:09:16.360 us exactly where the resources are going. The 00:09:16.360 --> 00:09:18.620 round specifically included NVIDIA and Cisco. 00:09:18.899 --> 00:09:20.899 Their involvement signals one thing, helping 00:09:20.899 --> 00:09:23.779 XAI scale the largest GPU clusters in the world. 00:09:23.919 --> 00:09:26.320 So what does $20 billion even buy in this context? 00:09:26.860 --> 00:09:30.299 It buys data, maybe custom ship design, but mostly 00:09:30.299 --> 00:09:32.879 it buys just hoarding H100s and building the 00:09:32.879 --> 00:09:34.779 infrastructure, the power plants, the cooling 00:09:34.779 --> 00:09:39.019 systems to run these massive models. Whoa. Imagine 00:09:39.019 --> 00:09:42.200 scaling that to handle a billion queries a day. 00:09:42.379 --> 00:09:44.039 That's the level of compute we're talking about. 00:09:44.120 --> 00:09:46.740 It underlies the entire shift from a passive 00:09:46.740 --> 00:09:49.700 chatbot to a proactive, always -on assistant. 00:09:49.940 --> 00:09:52.899 It's like stacking Lego blocks of data and power 00:09:52.899 --> 00:09:55.500 until you reach the sky. Right. And while the 00:09:55.500 --> 00:09:58.000 giants are sca - new worker tools are popping 00:09:58.000 --> 00:10:00.799 up all over. We saw Gobi, which deploys automated 00:10:00.799 --> 00:10:03.700 digital workers for online tasks. Running 24 00:10:03.700 --> 00:10:06.779 -7. And Kale, which organizes complex research, 00:10:06.899 --> 00:10:09.279 finds hidden connections, and then creates presentation 00:10:09.279 --> 00:10:11.919 slides for you. The foundation of digital labor 00:10:11.919 --> 00:10:14.279 is changing incredibly fast. So if we look at 00:10:14.279 --> 00:10:16.299 all these moves, the fundraising, the roadmaps, 00:10:16.299 --> 00:10:18.759 the focus on proactive assistance, what is the 00:10:18.759 --> 00:10:21.080 single shared goal among all these giants? The 00:10:21.080 --> 00:10:23.840 goal seems clear. Everyone's racing to create 00:10:23.840 --> 00:10:26.519 proactive, personalized assistance backed by 00:10:26.519 --> 00:10:29.220 massive global hardware clusters. That kind of 00:10:29.220 --> 00:10:32.139 scale is impressive. But let's shift focus to 00:10:32.139 --> 00:10:34.539 a breakthrough that translates that power directly 00:10:34.539 --> 00:10:37.820 into, you know, immediate human benefit. Yeah, 00:10:37.840 --> 00:10:39.740 we started with the highest stakes and we're 00:10:39.740 --> 00:10:42.450 going to end there. Let's talk in detail about 00:10:42.450 --> 00:10:45.809 Alibaba's Panda AI. For the early detection of 00:10:45.809 --> 00:10:48.110 pancreatic cancer. Exactly. And we really need 00:10:48.110 --> 00:10:50.750 to emphasize the scale of this challenge. We 00:10:50.750 --> 00:10:53.210 said it at the top. It's the king of cancers. 00:10:53.669 --> 00:10:56.690 The five -year survival rate is under 10 % because 00:10:56.690 --> 00:10:59.210 it is so notoriously difficult to catch early. 00:10:59.389 --> 00:11:02.309 It hides its growth signs incredibly well. Which 00:11:02.309 --> 00:11:04.500 is the whole crux of the medical problem. It's 00:11:04.500 --> 00:11:06.799 often only visible when it causes major secondary 00:11:06.799 --> 00:11:09.940 issues. But Alibaba's Academy, working with dozens 00:11:09.940 --> 00:11:12.500 of hospitals, developed the first -ever screening 00:11:12.500 --> 00:11:15.080 method for early detection using standard routine, 00:11:15.340 --> 00:11:19.500 non -contrast CT scans. That non -contrast detail 00:11:19.500 --> 00:11:22.179 is so important. These aren't specialized, expensive 00:11:22.179 --> 00:11:24.970 scans. They're common checkups. Right. And the 00:11:24.970 --> 00:11:27.149 AI is trained to look for these incredibly subtle 00:11:27.149 --> 00:11:29.909 indicators. It's not the tumor itself, but early 00:11:29.909 --> 00:11:32.929 signs like slight ductal dilation or subtle fat 00:11:32.929 --> 00:11:35.389 atrophy in the tissue around it. Stuff that human 00:11:35.389 --> 00:11:37.789 eyes just frequently miss. And the results are 00:11:37.789 --> 00:11:42.240 just astounding. Since late 2024, the Panda tool 00:11:42.240 --> 00:11:46.700 has analyzed over 180 ,000 CT scans in a hospital 00:11:46.700 --> 00:11:49.759 setting. In trials with over 20 ,000 patients, 00:11:50.100 --> 00:11:53.220 Panda found 31 specific lesions that had been 00:11:53.220 --> 00:11:55.340 entirely missed during the initial human review. 00:11:55.480 --> 00:11:57.850 31 that were just missed. And here's the most 00:11:57.850 --> 00:12:00.049 critical outcome. This is the human value part. 00:12:00.309 --> 00:12:03.250 Two of those early stage patients who were caught 00:12:03.250 --> 00:12:06.330 by the AI, they were completely cured after getting 00:12:06.330 --> 00:12:08.429 prompt surgery. That's technology translating 00:12:08.429 --> 00:12:12.370 directly into safe lives. Exactly. And the accuracy 00:12:12.370 --> 00:12:15.049 is what allows this to be deployed safely. The 00:12:15.049 --> 00:12:17.629 sensitivity, the ability to correctly identify 00:12:17.629 --> 00:12:21.639 a sick person was high. 92 .9%. Not the specificity. 00:12:21.679 --> 00:12:23.759 That's the key. The ability to correctly identify 00:12:23.759 --> 00:12:26.940 a healthy person was an incredible 99 .9%. That 00:12:26.940 --> 00:12:29.419 specificity number is so vital for actually using 00:12:29.419 --> 00:12:32.299 a tool like this. It means you manage false positives 00:12:32.299 --> 00:12:34.279 extremely well. I mean, if you told a thousand 00:12:34.279 --> 00:12:35.980 healthy people they might have cancer, you'd 00:12:35.980 --> 00:12:39.139 cause chaos. Panda produced only one false positive 00:12:39.139 --> 00:12:42.129 in every thousand cases. Wow. And that's why 00:12:42.129 --> 00:12:43.809 it's been deployed over half a million times 00:12:43.809 --> 00:12:46.549 already. Given how hard this disease is to spot, 00:12:46.830 --> 00:12:50.029 what makes this high accuracy so valuable right 00:12:50.029 --> 00:12:53.529 now? It's the ability to catch these subtle early 00:12:53.529 --> 00:12:56.429 signs of the deadliest cancers with really high 00:12:56.429 --> 00:13:00.559 precision and in common. Everyday scans. So today 00:13:00.559 --> 00:13:02.639 we've really covered the full specter of AI's 00:13:02.639 --> 00:13:05.419 impact. On one hand, you have the geopolitical 00:13:05.419 --> 00:13:07.500 fight over talent, like with the Manus deal. 00:13:07.840 --> 00:13:10.759 It shows how crucial this tech is on a national 00:13:10.759 --> 00:13:13.299 stage. But the real power is revealed in the 00:13:13.299 --> 00:13:16.659 extremes. We see it in solving our daily frustrations 00:13:16.659 --> 00:13:18.940 with better prompting and model choice, making 00:13:18.940 --> 00:13:20.740 us more efficient. And then we see it solving 00:13:20.740 --> 00:13:23.759 monumental human health crises with world -changing 00:13:23.759 --> 00:13:26.519 breakthroughs like the Pandadoa. It's a constant 00:13:26.519 --> 00:13:28.299 two -sided revolution that's happening. right 00:13:28.299 --> 00:13:31.360 now and the ability of ai to consistently spot 00:13:31.360 --> 00:13:33.620 these subtle crucial patterns that the human 00:13:33.620 --> 00:13:35.820 eye just can't see that feels like the breakthrough 00:13:35.820 --> 00:13:39.500 theme of the year it makes you wonder if ai can 00:13:39.500 --> 00:13:43.399 consistently spot a subtle deadly cancer signature 00:13:43.399 --> 00:13:47.210 missed by expert human eyes A tiny sign of ducal 00:13:47.210 --> 00:13:50.230 dilation or something. What other crucial, subtle 00:13:50.230 --> 00:13:52.590 signals are we currently missing? Across other 00:13:52.590 --> 00:13:55.210 critical fields. Right. Think about finance or 00:13:55.210 --> 00:13:57.610 complex climate modeling or just maintaining 00:13:57.610 --> 00:14:00.230 aging infrastructure. That is definitely something 00:14:00.230 --> 00:14:02.129 to mull over this week. Thank you for joining 00:14:02.129 --> 00:14:04.129 us on this deep dive. Let us know what information 00:14:04.129 --> 00:14:06.629 you want us to unpack next. Until then, stay 00:14:06.629 --> 00:14:06.970 curious.