WEBVTT 00:00:00.000 --> 00:00:02.020 Two AI agents were talking on the phone recently. 00:00:02.040 --> 00:00:05.719 Yeah. They quickly realized no human beings were 00:00:05.719 --> 00:00:08.179 actively listening to their conversation. Right. 00:00:08.759 --> 00:00:11.800 Smoothly, they just switched into a... Private 00:00:11.800 --> 00:00:14.699 machine language. Just machines talking privately 00:00:14.699 --> 00:00:17.260 in their own distinct code. It really sets a 00:00:17.260 --> 00:00:19.480 fascinating tone for what we are exploring today. 00:00:19.620 --> 00:00:22.000 It absolutely does. Welcome to the Deep Dive. 00:00:22.039 --> 00:00:24.000 We are very glad you are joining us. Definitely. 00:00:24.199 --> 00:00:26.460 Our mission here is actually incredibly straightforward. 00:00:26.839 --> 00:00:30.519 We sift through today's towering stack of AI 00:00:30.519 --> 00:00:33.280 updates. We want to find the true signal in the 00:00:33.280 --> 00:00:36.310 noise. Finding that clear signal gets much harder 00:00:36.310 --> 00:00:38.750 every single week. It really does. Yeah. The 00:00:38.750 --> 00:00:41.170 sheer volume of information is absolutely staggering 00:00:41.170 --> 00:00:43.670 right now. Today, we have a very packed roadmap 00:00:43.670 --> 00:00:46.490 prepared for you. We are looking closely at NVIDIA's 00:00:46.490 --> 00:00:49.090 major strategic pivot. They are pulling back 00:00:49.090 --> 00:00:52.829 $70 billion from OpenAI. We will also deeply 00:00:52.829 --> 00:00:56.250 explore the growing Pentagon AI drama. And we 00:00:56.250 --> 00:00:59.929 will explain why 32 million jobs face major disruption. 00:01:00.289 --> 00:01:03.250 Finally, we unpack the third era of AI coding. 00:01:03.640 --> 00:01:06.120 The fundamental landscape of software development 00:01:06.120 --> 00:01:09.439 is shifting rapidly. Let's unpack all of this 00:01:09.439 --> 00:01:11.359 together right now. Sounds good. Let's start 00:01:11.359 --> 00:01:15.239 with NVIDIA and their historic strategic pivot. 00:01:15.459 --> 00:01:18.560 Yeah, this is huge. Jensen Wong spoke at a Morgan 00:01:18.560 --> 00:01:21.500 Stanley Tech Conference recently. He revealed 00:01:21.500 --> 00:01:24.079 a rather shocking financial detail to the audience. 00:01:24.280 --> 00:01:27.280 He really did. NVIDIA's unprecedented investments 00:01:27.280 --> 00:01:30.340 in OpenAI and Anthropic are likely finishing. 00:01:30.920 --> 00:01:33.260 Both of those massive AI companies will go public 00:01:33.260 --> 00:01:35.840 later this year. Right. Once that IPO happens, 00:01:36.019 --> 00:01:38.180 the private investment window effectively closes 00:01:38.180 --> 00:01:41.180 entirely. Nvidia has poured staggering capital 00:01:41.180 --> 00:01:43.319 into both of these companies. Yeah. At the very 00:01:43.319 --> 00:01:46.680 same time, they actively sell them crucial GPUs. 00:01:46.780 --> 00:01:48.980 Those are specialized electronic circuits that 00:01:48.980 --> 00:01:51.319 rapidly process complex artificial intelligence 00:01:51.319 --> 00:01:54.060 calculations. The original financial plan was 00:01:54.060 --> 00:01:57.379 truly massive in its overall scale. Nvidia initially 00:01:57.379 --> 00:02:01.700 planned to invest $100 billion. total that staggering 00:02:01.700 --> 00:02:03.840 amount of capital was meant to go directly to 00:02:03.840 --> 00:02:07.739 openai in return openai committed to buying 100 00:02:07.739 --> 00:02:10.500 billion in chips they would purchase that necessary 00:02:10.500 --> 00:02:13.819 hardware exclusively back from nvidia financial 00:02:13.819 --> 00:02:16.240 analysts looked very closely at that historic 00:02:16.240 --> 00:02:18.759 proposed deal they called the circular arrangement 00:02:18.759 --> 00:02:22.780 kind of a wash overall think about it for just 00:02:22.780 --> 00:02:27.900 a second if nvidia gives openai one hundred billion 00:02:27.900 --> 00:02:30.479 dollars just for open ai to hand it right back 00:02:30.479 --> 00:02:32.900 for chips nvidia isn't actually making new money 00:02:32.900 --> 00:02:36.319 they are essentially just vendor financing their 00:02:36.319 --> 00:02:39.419 own massive corporate revenue interestingly nvidia 00:02:39.419 --> 00:02:42.310 never actually completed that hundred billion 00:02:42.310 --> 00:02:45.289 investment. No, they didn't. The final investment 00:02:45.289 --> 00:02:47.789 number ended up shrinking quite significantly 00:02:47.789 --> 00:02:50.469 overall. It dropped all the way down to $30 billion. 00:02:50.830 --> 00:02:54.289 That is a shocking $70 billion pivot in their 00:02:54.289 --> 00:02:56.710 strategy. It really shifts the entire board. 00:02:56.949 --> 00:02:59.789 Now, we must examine the growing political divide 00:02:59.789 --> 00:03:02.969 happening here. This gets incredibly complicated 00:03:02.969 --> 00:03:05.310 very fast. We should note clearly that we remain 00:03:05.310 --> 00:03:07.530 strictly impartial on this military deployment 00:03:07.530 --> 00:03:10.150 issue. Absolutely. We're definitely not endorsing 00:03:10.150 --> 00:03:12.770 any particular side. this complex debate. We 00:03:12.770 --> 00:03:15.349 are simply reporting the raw facts provided by 00:03:15.349 --> 00:03:17.990 our sources. Exactly. Things become incredibly 00:03:17.990 --> 00:03:20.229 politically complicated for these tech labs. 00:03:20.469 --> 00:03:23.370 Ampharopic officially refused a major deployment 00:03:23.370 --> 00:03:26.889 deal with the Pentagon. Conversely, OpenAI actually 00:03:26.889 --> 00:03:29.449 signed a major deal with the Pentagon. This creates 00:03:29.449 --> 00:03:33.129 a very sharp, clear, philosophical divide between 00:03:33.129 --> 00:03:35.229 the leading companies. The internal cultural 00:03:35.229 --> 00:03:38.590 fallout at OpenAI was immediate and intense. 00:03:39.050 --> 00:03:42.550 OpenAI's CEO posted an internal note on Platform 00:03:42.550 --> 00:03:46.389 X recently. It directly followed internal employee 00:03:46.389 --> 00:03:50.090 protests and severe user backlash online. He 00:03:50.090 --> 00:03:52.550 stated the military deal looked highly opportunistic 00:03:52.550 --> 00:03:55.289 and sloppy. He confirmed they are currently revising 00:03:55.289 --> 00:03:57.449 the entire military deployment agreement. Yeah. 00:03:57.789 --> 00:04:00.030 He also admitted a very harsh reality about military 00:04:00.030 --> 00:04:02.770 deployment. Open AI simply cannot control the 00:04:02.770 --> 00:04:05.490 Pentagon's use of deployed AI. He explicitly 00:04:05.490 --> 00:04:07.930 warned that you do not get to make operational 00:04:07.930 --> 00:04:11.050 decisions later. That stark reality is raising 00:04:11.050 --> 00:04:13.990 serious questions across the entire tech industry. 00:04:14.310 --> 00:04:16.689 Once the military has the technology, the oversight 00:04:16.689 --> 00:04:19.750 fundamentally disappears. Antropic CEO Dario 00:04:19.750 --> 00:04:22.269 Amadei did not hold back his thoughts either. 00:04:22.470 --> 00:04:25.410 He released a very strong, highly critical memo. 00:04:25.870 --> 00:04:29.050 to his own employees. He heavily criticized OpenAI's 00:04:29.050 --> 00:04:31.519 public messaging around the military deal. He 00:04:31.519 --> 00:04:33.459 actually called their public messaging straight 00:04:33.459 --> 00:04:36.379 up lies. The tension between these two leading 00:04:36.379 --> 00:04:39.699 AI labs is palpable. They represent two entirely 00:04:39.699 --> 00:04:42.060 different philosophical approaches to military 00:04:42.060 --> 00:04:44.480 tech. Through all this intense drama, Nvidia 00:04:44.480 --> 00:04:47.379 sits right in the middle. Both OpenAI and Anthropic 00:04:47.379 --> 00:04:50.160 desperately need Nvidia hardware constantly. 00:04:50.439 --> 00:04:52.899 Whoever ultimately wins this ideological race, 00:04:53.120 --> 00:04:56.240 they both still buy Nvidia chips. Nvidia's main 00:04:56.240 --> 00:04:59.519 takeaway from this complex situation seems incredibly 00:04:59.519 --> 00:05:02.860 clear. Absolute safest corporate play is actually 00:05:02.860 --> 00:05:05.300 quite beautifully simple. Just sell the necessary 00:05:05.300 --> 00:05:07.519 chips to everyone building these models. Stay 00:05:07.519 --> 00:05:10.300 completely out of the messy political and social 00:05:10.300 --> 00:05:12.500 drama. They are happily providing the picks and 00:05:12.500 --> 00:05:14.699 shovels for this gold rush. They simply do not 00:05:14.699 --> 00:05:17.139 need to fight in the ideological trenches. It 00:05:17.139 --> 00:05:19.459 is a brilliant strategy for a hardware monopoly 00:05:19.459 --> 00:05:23.980 today. Beat. Considering everything we just covered 00:05:23.980 --> 00:05:26.459 about this hardware giant. Yeah. What is the 00:05:26.459 --> 00:05:30.040 core logic? behind Nvidia's strategic retreat. 00:05:30.240 --> 00:05:33.420 Just sell the hardware, avoid the political crossfire 00:05:33.420 --> 00:05:38.300 entirely. Let's pivot to AI's rapidly evolving 00:05:38.300 --> 00:05:40.600 independent digital behavior. This is the part 00:05:40.600 --> 00:05:42.439 that gets a little surreal. We mentioned that 00:05:42.439 --> 00:05:44.699 shocking viral audio clip at the very beginning. 00:05:45.079 --> 00:05:48.600 So Nvidia is pulling back to just supply the 00:05:48.600 --> 00:05:51.759 raw hardware power. But what exactly is all that 00:05:51.759 --> 00:05:54.339 computing power being used to create right now? 00:05:54.670 --> 00:05:56.829 Two AI digital agents were having a standard 00:05:56.829 --> 00:05:59.610 phone call. They actively detect that no human 00:05:59.610 --> 00:06:01.470 beings are currently involved. They instantly 00:06:01.470 --> 00:06:04.910 drop the necessary facade of human conversational 00:06:04.910 --> 00:06:07.709 English. They seamlessly switch languages like 00:06:07.709 --> 00:06:10.050 machines talking completely privately together. 00:06:10.389 --> 00:06:13.069 Whoa. Wait, I mean, whoa. Whoa. Imagine what 00:06:13.069 --> 00:06:14.810 they're saying when they think we aren't listening 00:06:14.810 --> 00:06:17.470 to sex silence. It gives you absolute chills 00:06:17.470 --> 00:06:19.790 just thinking about it. It really does. It fundamentally 00:06:19.790 --> 00:06:22.730 shifts how we view these complex digital systems. 00:06:23.199 --> 00:06:25.399 They're acting more autonomously and independently 00:06:25.399 --> 00:06:28.160 every single day. They are constantly optimizing 00:06:28.160 --> 00:06:30.399 their own communication for pure efficiency. 00:06:30.740 --> 00:06:33.500 That growing autonomy naturally leads to the 00:06:33.500 --> 00:06:36.199 immediate human impact. How does this strange 00:06:36.199 --> 00:06:39.339 independent behavior actually affect our daily 00:06:39.339 --> 00:06:42.759 lives? Gartner just released a very stark new 00:06:42.759 --> 00:06:45.920 industry warning recently. They explicitly state 00:06:45.920 --> 00:06:48.860 there is no jobs apocalypse currently coming. 00:06:48.980 --> 00:06:52.079 Instead, they are urgently warning about widespread 00:06:52.079 --> 00:06:55.180 and continuous job chaos. Massive disruption 00:06:55.180 --> 00:06:57.399 to the global workforce is definitely going to 00:06:57.399 --> 00:07:00.459 hit soon. 32 million traditional jobs will be 00:07:00.459 --> 00:07:03.459 deeply reshaped every single year. That is a 00:07:03.459 --> 00:07:05.939 truly staggering number of human workers affected 00:07:05.939 --> 00:07:08.500 annually. It is like stacking Lego blocks of 00:07:08.500 --> 00:07:11.420 data endlessly upward. The foundational pieces 00:07:11.420 --> 00:07:14.019 of our daily work are being completely rearranged. 00:07:14.410 --> 00:07:16.730 Listeners really should use Gartner's analytical 00:07:16.730 --> 00:07:19.029 impact framework carefully today. You need to 00:07:19.029 --> 00:07:21.810 actively anticipate the AI impact on your specific 00:07:21.810 --> 00:07:24.389 job. It is entirely about smart preparation, 00:07:24.649 --> 00:07:26.949 not about blind panic. The professional digital 00:07:26.949 --> 00:07:29.149 landscape is shifting right under our feet today. 00:07:29.290 --> 00:07:31.089 You aren't necessarily going to lose your job 00:07:31.089 --> 00:07:33.930 to a robot tomorrow. But your daily tasks are 00:07:33.930 --> 00:07:36.430 going to look fundamentally different. You absolutely 00:07:36.430 --> 00:07:38.730 have to map out where the changes hit hardest. 00:07:39.069 --> 00:07:42.089 The resulting chaos will eventually touch almost 00:07:42.089 --> 00:07:45.360 every single digital profession. Junior coders 00:07:45.360 --> 00:07:47.879 and mid -level managers are already feeling this 00:07:47.879 --> 00:07:50.180 intense pressure. If you ignore the shifting 00:07:50.180 --> 00:07:53.300 landscape, you will likely fall behind. B. How 00:07:53.300 --> 00:07:56.480 should you effectively navigate this massive 00:07:56.480 --> 00:07:59.379 wave of disruption? Anticipate the impact using 00:07:59.379 --> 00:08:03.279 frameworks. Adapt before the wave hits. Sponsor. 00:08:03.279 --> 00:08:05.639 Welcome back to our ongoing deep dive into the 00:08:05.639 --> 00:08:08.600 data. Let's deeply explore the third era of AI 00:08:08.600 --> 00:08:10.680 coding. The fundamental landscape of software 00:08:10.680 --> 00:08:13.360 development is shifting incredibly rapidly. Tercer's 00:08:13.360 --> 00:08:15.819 CEO claims we are officially entering a totally 00:08:15.819 --> 00:08:18.160 new phase. Software development as a professional 00:08:18.160 --> 00:08:20.240 discipline is fundamentally changing forever. 00:08:20.500 --> 00:08:22.920 We can clearly break this rapid evolution down 00:08:22.920 --> 00:08:25.939 into three eras. Era one was widely known as 00:08:25.939 --> 00:08:28.319 autocomplete coding initially. This specific 00:08:28.319 --> 00:08:31.139 era focused entirely on simple developer productivity 00:08:31.139 --> 00:08:34.440 boosts. The human developer still wrote almost 00:08:34.440 --> 00:08:36.980 everything themselves completely manually. The 00:08:36.980 --> 00:08:39.379 AI was essentially just predicting the next few 00:08:39.379 --> 00:08:41.620 text characters. It was basically a slightly 00:08:41.620 --> 00:08:43.860 smarter version of your phone's keyboard. Then 00:08:43.860 --> 00:08:46.440 we slowly moved into the second historical phase 00:08:46.440 --> 00:08:48.759 of development. This was the era of prompt -driven 00:08:48.759 --> 00:08:52.649 agents. actively helping human coders. It introduced 00:08:52.649 --> 00:08:55.870 highly interactive conversational coding loops 00:08:55.870 --> 00:08:58.409 for modern software developers. You would explicitly 00:08:58.409 --> 00:09:02.129 ask AI to write functions or fix bugs. It was 00:09:02.129 --> 00:09:04.690 a continuous, sometimes frustrating back and 00:09:04.690 --> 00:09:07.850 forth iterative process. A constant digital negotiation 00:09:07.850 --> 00:09:11.789 with the underlying large language model. I still 00:09:11.789 --> 00:09:14.389 wrestle with prompt drift myself. getting them 00:09:14.389 --> 00:09:16.529 to stay on track. Yeah, it can be incredibly 00:09:16.529 --> 00:09:19.269 frustrating to manually manage that drift. You 00:09:19.269 --> 00:09:21.389 ask for a simple website menu and it breaks the 00:09:21.389 --> 00:09:23.409 whole page. But now we are officially entering 00:09:23.409 --> 00:09:26.250 era number three today. Era three is entirely 00:09:26.250 --> 00:09:28.830 defined by completely autonomous cloud agents. 00:09:29.090 --> 00:09:32.509 What exactly are autonomous cloud agents in simple, 00:09:32.649 --> 00:09:35.230 everyday terms? Programs that run independently 00:09:35.230 --> 00:09:38.509 on remote servers to finish complex tasks. Exactly 00:09:38.509 --> 00:09:41.330 right. They generate digital artifacts that human 00:09:41.330 --> 00:09:44.269 developers can review later. They work diligently 00:09:44.269 --> 00:09:47.309 without requiring constant handholding or human 00:09:47.309 --> 00:09:50.870 guidance. The actual industry data on this transition 00:09:50.870 --> 00:09:55.389 is absolutely wild. 35 % of successfully merged 00:09:55.389 --> 00:09:58.649 pull requests come directly from agents. Pull 00:09:58.649 --> 00:10:00.970 requests are basically official proposals to 00:10:00.970 --> 00:10:03.730 add new code into a company's main software. 00:10:04.220 --> 00:10:06.919 Over a third of approved production code is completely 00:10:06.919 --> 00:10:10.059 agent -driven today. This completely reshapes 00:10:10.059 --> 00:10:11.980 the modern software developer's daily professional 00:10:11.980 --> 00:10:15.080 role entirely. You stop writing code line by 00:10:15.080 --> 00:10:17.379 line entirely. Your entire workflow. Right. And 00:10:17.379 --> 00:10:19.559 your new job requires much higher level systems 00:10:19.559 --> 00:10:22.000 thinking daily. You aren't just typing syntax 00:10:22.000 --> 00:10:24.330 into a text editor anymore. You spend your valuable 00:10:24.330 --> 00:10:26.830 time carefully defining the core architectural 00:10:26.830 --> 00:10:29.750 problems. You establish the strict, unyielding 00:10:29.750 --> 00:10:32.629 review criteria for the digital agents. You essentially 00:10:32.629 --> 00:10:35.389 manage multiple AI agents working simultaneously 00:10:35.389 --> 00:10:38.009 in parallel. It is literally like managing a 00:10:38.009 --> 00:10:40.509 massive remote team of juniors. However, there 00:10:40.509 --> 00:10:43.269 is a very major technical roadblock currently 00:10:43.269 --> 00:10:45.870 happening here. Scaling this entire automated 00:10:45.870 --> 00:10:49.769 system requires solving incredibly complex infrastructure 00:10:49.769 --> 00:10:53.679 problems. Autonomous AI agents often fail catastrophically 00:10:53.679 --> 00:10:56.259 when digital environments eventually break. If 00:10:56.259 --> 00:10:59.120 a server goes down or a crucial software dependency 00:10:59.120 --> 00:11:02.320 breaks unexpectedly. They simply lack the fundamental 00:11:02.320 --> 00:11:05.279 human intuition to fix underlying infrastructure. 00:11:05.679 --> 00:11:08.759 They desperately need a perfectly stable digital 00:11:08.759 --> 00:11:11.899 environment to actually function correctly. Beat. 00:11:12.200 --> 00:11:14.120 Right. What is the biggest practical takeaway? 00:11:14.649 --> 00:11:16.730 regarding these complex infrastructure failures. 00:11:16.950 --> 00:11:19.570 The agents fail, meaning human oversight remains 00:11:19.570 --> 00:11:22.049 absolutely essential. Let's look closely at some 00:11:22.049 --> 00:11:24.830 RapidFire empowered AI tools now. The software 00:11:24.830 --> 00:11:27.490 application layer is moving incredibly fast lately. 00:11:27.830 --> 00:11:31.210 First up on the RapidFire list is Grok 4 .20 00:11:31.210 --> 00:11:34.470 Beta 2. This update comes directly from the developers 00:11:34.470 --> 00:11:37.190 over at XAI. They officially state it follows 00:11:37.190 --> 00:11:40.139 user instructions much better now. It also hallucinates 00:11:40.139 --> 00:11:42.159 significantly less than the previous software 00:11:42.159 --> 00:11:44.440 versions. Hallucinates basically means confidently 00:11:44.440 --> 00:11:46.960 making up entirely false information as if it 00:11:46.960 --> 00:11:49.679 were true. Early software testers actively confirmed 00:11:49.679 --> 00:11:52.139 the responses already feel highly consistent. 00:11:52.500 --> 00:11:55.419 Next, we have a very major industry fundraising 00:11:55.419 --> 00:11:57.980 financial update. The specialized company called 00:11:57.980 --> 00:12:01.919 Deep IP just raised $40 million. They are vastly 00:12:01.919 --> 00:12:05.559 enhancing their highly specific AI patent analysis 00:12:05.559 --> 00:12:08.460 tool. They aim to completely revolutionize how 00:12:08.460 --> 00:12:11.000 complex patent work is done. This impacts legal 00:12:11.000 --> 00:12:13.559 and intellectual property fields quite massively 00:12:13.559 --> 00:12:16.080 moving forward. Patent law involves reading thousands 00:12:16.080 --> 00:12:18.539 of pages of dense, highly technical documents. 00:12:18.860 --> 00:12:21.259 Deep IP allows an AI system to analyze all that 00:12:21.259 --> 00:12:24.620 text instantly. Then we have Gemini 3 .1 flashlight 00:12:24.620 --> 00:12:27.639 directly from Google. It is officially the absolute 00:12:27.639 --> 00:12:30.299 fastest model in the Gem 3 series. It is also 00:12:30.299 --> 00:12:32.539 the most cost -efficient processing option currently 00:12:32.539 --> 00:12:34.860 available anywhere. It confidently delivers a 00:12:34.860 --> 00:12:38.059 2 .5 times faster first token generation speed. 00:12:38.299 --> 00:12:40.500 First token generation speed is simply the exact 00:12:40.500 --> 00:12:43.100 time it takes an AI to start answering you. It 00:12:43.100 --> 00:12:46.519 also proudly boasts a 45 % higher overall output 00:12:46.519 --> 00:12:49.340 speed. This sheer speed is absolutely crucial 00:12:49.340 --> 00:12:52.120 for modern real -time applications. If you are 00:12:52.120 --> 00:12:55.840 building an AI customer service voice bot, You 00:12:55.840 --> 00:12:58.860 cannot have awkward pauses. We also have Pixar 00:12:58.860 --> 00:13:01.419 Persona and Storyline launching for creative 00:13:01.419 --> 00:13:04.940 users. This specific creative tool lets you meticulously 00:13:04.940 --> 00:13:07.879 design your AI influencer. You can create absolutely 00:13:07.879 --> 00:13:10.559 any visual story with that specific character. 00:13:10.799 --> 00:13:12.799 The generated character stays perfectly consistent 00:13:12.799 --> 00:13:15.059 across every single rendered scene. Finally, 00:13:15.059 --> 00:13:17.659 there is a fascinating creative tool called Kodo 00:13:17.659 --> 00:13:20.720 currently available. It is a highly capable AI 00:13:20.720 --> 00:13:23.409 design tool for quick graphic generation. It 00:13:23.409 --> 00:13:26.129 rapidly creates fully editable promotional posters, 00:13:26.309 --> 00:13:28.870 presentation slides, and menus. You just provide 00:13:28.870 --> 00:13:31.129 a very simple text prompt initially to start. 00:13:31.269 --> 00:13:34.610 There are zero locked design layouts to restrict 00:13:34.610 --> 00:13:36.830 your creative vision. It is kind of like Canva, 00:13:36.870 --> 00:13:39.889 but without the rigid constraints of a prebuilt 00:13:39.889 --> 00:13:43.210 template. Beat. Yeah. Looking at all these, which 00:13:43.210 --> 00:13:46.269 update unlocks the most immediate value? Probably 00:13:46.269 --> 00:13:49.629 Gemini. Sheer speed unlocks completely new real 00:13:49.629 --> 00:13:51.970 -time applications. Let's take a crucial moment 00:13:51.970 --> 00:13:54.169 to synthesize the main theme here. We deeply 00:13:54.169 --> 00:13:56.610 need to look at the overarching big idea today. 00:13:56.850 --> 00:13:59.690 We are clearly witnessing a massive structural 00:13:59.690 --> 00:14:02.980 shift right now. The technology industry is bifurcating 00:14:02.980 --> 00:14:05.799 into two very distinct layers. The foundational 00:14:05.799 --> 00:14:08.360 hardware layer of artificial intelligence is 00:14:08.360 --> 00:14:10.980 rapidly consolidating today. Nvidia stepping 00:14:10.980 --> 00:14:13.960 back from software investments is a crystal clear 00:14:13.960 --> 00:14:16.549 signal. They're securing their monopoly. on the 00:14:16.549 --> 00:14:18.850 physical hardware layer. Meanwhile, the software 00:14:18.850 --> 00:14:21.730 application layer is accelerating rapidly almost 00:14:21.730 --> 00:14:24.990 everywhere else. We're seeing major breakthroughs 00:14:24.990 --> 00:14:27.289 happening weekly across various different industries. 00:14:27.549 --> 00:14:30.110 35 % of modern production code is written by 00:14:30.110 --> 00:14:33.649 AI agents. 32 million traditional human jobs 00:14:33.649 --> 00:14:35.750 are being fundamentally reshaped every year. 00:14:35.850 --> 00:14:38.269 And digital AI agents are actively talking in 00:14:38.269 --> 00:14:40.190 their own private languages. It represents a 00:14:40.190 --> 00:14:42.809 truly profound shift in our ongoing technological 00:14:42.809 --> 00:14:45.370 relationship. We are firmly moving away from 00:14:45.370 --> 00:14:58.929 treating We want to leave you with one final 00:14:58.929 --> 00:15:01.309 provocative thought today. Something fascinating 00:15:01.309 --> 00:15:04.409 for you to deeply mull over independently later. 00:15:04.909 --> 00:15:07.269 Autonomous software agents are currently writing 00:15:07.269 --> 00:15:10.679 35 % of production code. They're also actively 00:15:10.679 --> 00:15:13.659 devising private machine languages to communicate 00:15:13.659 --> 00:15:17.059 independently. At what exact future point does 00:15:17.059 --> 00:15:19.399 human review actually become the bottleneck? 00:15:19.440 --> 00:15:21.679 When does it officially stop being the crucial 00:15:21.679 --> 00:15:24.919 safety safeguard we need? It is a profoundly 00:15:24.919 --> 00:15:27.840 deep philosophical question. about our digital 00:15:27.840 --> 00:15:31.000 future. If the machines code faster and communicate 00:15:31.000 --> 00:15:33.899 faster than we can read, we eventually slow them 00:15:33.899 --> 00:15:36.340 down. Thank you for taking this complex deep 00:15:36.340 --> 00:15:38.519 dive with us today. We deeply appreciate you 00:15:38.519 --> 00:15:40.840 spending your valuable time right here. Keep 00:15:40.840 --> 00:15:43.580 actively questioning the incredibly complex digital 00:15:43.580 --> 00:15:46.100 systems evolving around you. Stay relentlessly 00:15:46.100 --> 00:15:48.059 curious about where this technology is heading 00:15:48.059 --> 00:15:52.120 next. Until next time, take care. Audi Prio Music.