Bill Gates billion dollar AI equity push

0:00The best part of Waking Up, a full cup of Folger's Coffee and music on Full Blast. Wake me up, wake me up and stand there. Wake me up and stand there. So wake me up when it's all over. I'm on my way down. Yeah. Ah, ah, ah, ah. The best part of Waking Up is Folger's and Yorke. Shop Folger's Instant Coffee and more at your nearest retailer. This episode is brought to you by Google Chrome. You think you know a browser. But Gemini and Chrome, that's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50-page restoration block, or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it. Ready to make anything online make sense? There's no place like Chrome. Check responses set up require compatibility and availability varies 18 plus. College football is back. So Hilton called to me the superstition concierge

1:00to make your fan rituals a reality. Need a room to match your lucky number? We got you. Want to make sure our team doesn't wash your lucky jersey? Oh, that smells lucky. Hilton's unmatched hospitality can keep up with any superstition. Even a marching bandwink up call at 555 and 55 seconds. Hit it! When you need a team that will do whatever it takes on game day, it matters where you stay. Hilton, for this day. Bill Gates and the Gates Foundation are putting $1 billion behind the assertion that artificial intelligence could become the worst source of injustice in history. They are committing that exact dollar amount right now based on that specific premise. Yeah, the framing of that money is really the interesting part to me. It's specifically earmarked to force the technology to become an equalizer instead. And the justification for doing it right this second is a highly specific narrow 12 to 18 month window. The argument is that the trajectory of the tech

2:01gets locked in during this very short time frame. We are looking at a situation where the core issue isn't really about what the technology can compute. The focus is entirely on who it leaves behind. Yeah. That is the baseline for everything happening around this billion dollar commitment. But what exactly is happening in this specific 12 to 18 month time frame that makes it so irreversible? Well, to understand the origin of that timeline, you're going to have to look at how this started for Gates personally. Months before the first major public release of these tech generation models, he received a private demonstration of the underlying architecture. He was just sitting at his kitchen table. Right. He had early access. Yeah. And initially, he was shocked and amazed by what he saw. But that sense of wonder curdled into dismay pretty quickly. I mean, that makes sense. It was entirely because of the frightening speed of the technology's advancement. We are talking about a person who built his entire empire on understanding software development cycles. Right. He knows how this is supposed to work. Exactly.

3:02So when he looks at the speed of iteration here, he is seeing something that breaks the historical models of how software is supposed to mature. That reaction to the speed makes complete sense when you look at the real world consequences of velocity and software development. The shift from awe to dismay is directly driven by the realization that the people actually developing the technology are openly discussing scenarios that spell the end of the human race. Yeah, which is not normal for a product launch. No, not at all. When the creators of a tool are calmly talking about existential risks in casual interviews, the speed of iteration stops being an engineering triumph. It becomes a problem. Right. It starts becoming a structural liability. Yeah. You have to remember how traditional software works. You build a beta. You test it for a year. You find the bugs. You patch the security holes. And then you slowly roll it out to enterprise clients. Exactly. But with these large language models, the capabilities are emerging faster

4:03than the researchers can categorize them. Like they don't even know what it can do until later. Yeah, they are discovering what the model can do months after it has finished training. That velocity means the safety testing is inherently reactive. It's like watching a city decide to build a high speed rail system. But they decide to start running the train at maximum capacity while the track is still being laid down. Just inches in front of the wheels. Right. And then you realize no one actually designed the brakes yet. That is a stressful visual. You have the engine running. The track is barely staying ahead of the wheels. And the engineers are sitting in the dining car debating whether the train might eventually derail and take out the whole town. And the velocity itself becomes the primary threat vector there. Because the faster the train moves, the less time you have to build a brake. Yeah. And the more catastrophic the derailment becomes. So Gates, sitting at his kitchen table, basically realized that the train was already moving at a speed that the current regulatory and safety infrastructure could not possibly handle. Yeah.

5:03And the speed of that advancement leads naturally into the timeline he identified. If the train is moving that fast, you only have a certain amount of track left before the route is permanent. Right. The assertion is that the decisions made in the next 12 to 18 months regarding how the technology is built, funded, and deployed will determine its permanent impact on global equity. Built, funded, deployed. Those three verbs are the specific mechanisms of action here. Let's look closely at those. Built refers to the literal hardware and architecture. Funded dictates which specific problems the smartest engineers in the world are paid to solve. And deployed is about who gets access to the final product and at what cost. I do question the absolute rigidity of that 12 to 18 month number, though. You think it's too specific? I mean, why that specific window, why not two years, why not six months? Well, you have to look at how infrastructure is being laid right now to understand the urgency. Like the physical data centers.

6:03Yeah. We are watching the hardware, incorporate momentum locking in. It is not just about writing code anymore. It is about pouring concrete, laying fiber optics and securing power grids. Because building a data center capable of training the next generation of these models requires hundreds of megawatts of continuous power. Exactly. You are seeing technology companies signing long-term contracts with nuclear power plants just to guarantee they have the electricity required for the next decade. And that kind of physical infrastructure takes years to plan and build. But the contracts and the designs are being finalized right now. You can see that hardening of infrastructure with companies like Apple too. They are already mapping out an incredibly heavy hardware refresh for their entire product line spanning the next few years. They are redesigning their silicon, their supply chains and their data centers specifically to support this architecture. The physical infrastructure is cementing the software's path. The best part of waking up, a full cup of Folger's Coffee and music on full blast.

7:06Wake me up, wake me up, wake me up, wake me up, wake me up, and save me. Show me the up when it's all over. I'm only now waiting. You're up all over. The best part of waking up is Folger's and your cup. Shop Folger's Instant Coffee and more at your nearest retailer. This episode is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome? That's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50-page restoration block or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it. Ready to make anything online make sense? There's no place like Chrome. Check responses set up require compatibility and availability very 16 plus. College football is back. So Hilton called to me the superstition concierge to make your fan rituals a reality. Need a room to match your lucky number? We got you. Want to make sure our team doesn't wash your lucky Jersey?

8:09Oh, that smells lucky. Hilton's unmatched hospitality can keep up with any superstition. Even a marching band wake up call it 555 and 55 seconds. Hit it! When you need a team that will do whatever it takes on game day, it matters where you stay. Hilton for this day. So when Apple or Google designs a custom neural processing unit or an NPU? To handle AI tasks locally on a device? Yeah, they are locking in a specific mathematical approach to artificial intelligence. Right, because once you put a specific type of silicon into 100 million smartphones, the global software ecosystem has to optimize for that exact hardware. You don't just easily retool a global supply chain for something else later. The hardware basically dictates the bounds of the software. And that physical reality severely limits the ability for regulators and society to take a wait and see approach. Yeah, if you are listening to this right now, you might be thinking this sounds like a distant silicon value problem. But it's really not. No, because you cannot wait to see how the software

9:10impacts society if the physical foundation running that software is already permanently installed globally. The concrete is drawing on the hardware side. Exactly. If you want to change the shape of the building, you have to do it before the foundation sets. Because you cannot retroactively force a global hardware ecosystem to prioritize low resource clinical health if it was physically engineered from the silicon up to serve high bandwidth enterprise software. Understanding that the clock is ticking requires looking at the default path this technology is on if it is left untouched. If we do nothing. Right. Where does the gravity of the market pull this? Capital naturally gravitates toward high margin software. It goes toward trading algorithms, enterprise efficiency tools, and ad optimization. Yeah, because if you have a cluster of servers that cost a billion dollars to build, you are going to point them at the problem that pays you back the most money the fastest. You map that onto real world global health and education and the disparity becomes pretty glaring.

10:12The technology defaults to making wealthy institutions slightly more efficient. It helps a massive logistics company shave a fraction of a percent off their shipping costs. But it does not naturally default to solving foundational problems for those who have the least. Because solving those problems does not generate a quarterly return for shareholders. Right. There is no financial incentive to point a billion dollar super computer at the problem of eradicating an neglected tropical disease in a region with zero purchasing power. But hasn't the tech industry always worked that way? What do you mean? Think about mobile phones. The first cell phones were the size of bricks and were strictly for Wall Street bankers and corporate executives who could afford a thousand dollar device and a massive monthly bill. Yeah, they were a luxury product for the ultra wealthy. But the market eventually drove the cost down and now practically everyone in the global South has a smartphone. So you're saying why shouldn't we just trust the trickle down effect here? Right. Why won't the market eventually make this technology cheap enough

11:14to solve those global health problems naturally? Well, that is the assumption a lot of people make. But it ignores the physical mechanics of how this specific technology operates. OK. A smartphone is fundamentally a localized piece of hardware. Once you buy it, it works. Artificial intelligence, at least the frontier models we're talking about requires continuous, massive computational power on the back end. It requires constant connection to data centers. Yeah, continuous updates and immense energy consumption. The trickle down theory works for physical goods where manufacturing becomes cheaper over time. Right. The unit cost drops. But it does not work when the fundamental resource required to use the tool, which is compute, is inherently scarce and incredibly expensive to maintain. And then just to look at the scale of the capital move around, OpenAI recently spent $300 million acquiring a single startup called Glass Imaging. Wow, just one company. Yeah, that is $300 million for one camera technology company. Why would a text-based AI company

12:16need to buy a camera tech company for that kind of money? Because the future of these models is multimodal. They need to be able to see and interpret the physical world through pixels, not just process text. Right, like analyzing live video. Yeah. And training a model to understand a live video feed requires an astronomical amount of compute and highly specialized intellectual property. The sheer cost of acquiring that capability means all the available capital on the market is being sucked into these massive high-mongering corporate ventures. Exactly. When OpenAI drops $300 million on a single feature upgrade, they are starving out smaller, non-profit initiatives that might want to build localized visual AI for identifying crop diseases in Africa. The capital flowing into the private sector is staggering compared to public sector budgets in the developing world. The annual health budget of an entire developing nation might be less than what one tech giant spends on a single round of model training. And that disparity severely limits the assumption

13:19that technological benefits will naturally trickle down to the global south in time to be useful. Because when the initial development is so hyper-focused on enterprise margins and requires such heavy infrastructure, the resulting tool might not even be applicable to a rural clinic, regardless of how cheap the subscription eventually gets. Right. You cannot run an enterprise grade high-band with diagnostic AI on a spotty 3G connection in a village with rolling blackouts. The tool has to be designed for the environment it will be used in from day one. And that overwhelming market gravity is exactly why the Gates Foundation just announced this specific financial countermeasure. The billion dollars. Yeah. They're committing $1 billion over the next two years to expand access to these solutions in areas like health, education, and agriculture. The application of that money is highly targeted too. The goal is to direct breakthroughs toward neglected global priorities. Meaning actively funding things like clinical co-pilots for regions facing severe doctor shortages.

14:19Right. For those who might not be familiar with the term, a clinical co-pilot in this context isn't just an advanced version of WebMD. No, it's way more involved. In regions where the doctor to patient ratio might be one to 10,000, a human doctor physically cannot spend 20 minutes taking a detailed patient history. So a clinical co-pilot is a voice enabled system that can interact with the patient in their local dialect. Exactly. It asks the relevant diagnostic questions, cross-references the patient's symptoms with local epidemiological data, and presents the human doctor with a synthesized summary and a probable diagnosis. It acts as a force multiplier for the very limited human medical staff. It also means funding adaptive tutoring agents designed specifically to close achievement gaps for underserved students. Like systems that can monitor exactly where a student is struggling in a math concept and adapt the lesson in real time. Yeah, functioning like a dedicated private tutor for a child who attends an underfunded school with 60 kids in a single classroom.

15:21You do have to ask how far $1 billion actually goes, though, when you are competing against tech giants spending tens of billions on compute alone. I mean, Microsoft and Google are spending that much on single data center complexes. Right. So the billion functions less as a rival force and more as targeted seed money. It is an attempt to prove these alternative use cases are viable, so other capital might follow. Proof of concept funding. Exactly. If you can show that a clinical co-pilot actually works in a low resource environment and that it measurably reduces mortality rates, you create a blueprint. And then national governments or international health organizations can look at that blueprint and decide to reallocate their own massive budgets to scale it up. The Gates Foundation is essentially acting as the angel investor for global public goods, taking on the initial research and development risk that the market refuses to touch. Instead of just talking about hypothetical use cases for that seed money, we can look at where this kind of deployment is already functioning on the ground. OK. Mark Suzman, the CEO of the Gates Foundation recently visited

16:22India and looked at a group called Wadwani AI. I've heard of them. Yeah, they have voice-based tools that are currently helping students improve foundational reading skills in multiple languages. This is an active, deployed project. The localized nature of a tool like that is critical. It is not a generic text generator sitting in a browser window. No, it is a highly specific voice-based solution solving a fundamental human bottleneck, which is literacy. You have to realize that for a massive portion of the global population, the primary barrier to using a computer isn't cost, it's the interface. Because if you cannot read or write a traditional keyboard and screen are mostly useless to you. Exactly. By deploying a voice-based model, you are bypassing the need for a keyboard entirely and going straight to verbal interaction. A child who cannot read can still speak to the device, and the device can speak back, guiding them phonetically through the learning process. But the difficulty of scaling something like that

17:24has to be acknowledged. Oh, for sure. Solving reading in one region does not automatically translate to another without immense localization. Think of an AI model like a local detective. If you take a detective who has spent 20 years solving complex financial crimes in Manhattan, and you drop him into a rural village in India to solve an agricultural theft, his instincts are entirely wrong. He doesn't speak the dialect. He doesn't know the local power dynamics. And his foundational logic doesn't apply to the situation. You cannot just give him a dictionary and expect him to be effective. You have to train a completely new detective from the ground up using local logic and local experience. Right. You cannot just copy and paste a voice tool train on one dialect and drop it into a village 300 miles away and expect it to work. That opens up a view of the technology that treats it not as a luxury productivity tool, but as basic societal infrastructure. Like sanitation or electricity. Yeah, or paved roads. You do not charge a premium monthly subscription for basic sanitation. You treat it as a fundamental requirement

18:26for a functioning society. Treating a literacy tool with that same level of foundational importance completely changes how you fund and distribute it. It moves the technology out of the realm of consumer software and into the realm of public works. The success of tools like Wadwani AI relies on specific structural pillars that Suzman identified during his visit. These are basically non-negotiable requirements if you want global deployment to actually work. Right. For these tools to reach a global population, they must work in every language. And that includes regional dialects and local slang. They have to be grounded in local data too. And crucially, there must be investment in local people so those closest to the problem have the skills to actually use and maintain the tools. Those three technical and social requirements, language, local data, and local skills are the absolute minimum baseline. The language pillar is particularly difficult because of how these models actually process text. The tokenization. Yeah, they use a system called tokenization

19:28where they break words down into smaller chunks or tokens. Because the vast majority of the internet and therefore the vast majority of the training data is in English, the models are highly optimized to tokenize English words efficiently. Right, a single English word might be one token. But if you ask that same model to process a word in Swahili or Hindi, it might have to break that single word into five or six tokens because it doesn't recognize the structure efficiently. The best part of waking up. A full cup of Folger's coffee and music on full blast. Wake me up, wake me up in sky. Wake me up, wake me up in sky. Joy, key up when it's all over. And moment now we're done. You're up all over. The best part of waking up is Folger's in your cup. Shop Folger's instant coffee and more at your nearest retailer. This episode is brought to you by Google Chrome.

20:28You think you know a browser, but Gemini and Chrome? That's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50-page restoration block or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it. Ready to make anything online make sense? There's no place like Chrome. Check responses set up require compatibility and availability varies 18 plus. College football is back. So Hilton called in me the superstition concierge to make your fan rituals a reality. Need a room to match your lucky number? We got you. Want to make sure our team doesn't wash your lucky Jersey? Oh, that smells lucky. Hilton's unmatched hospitality can keep up with any superstition. Even a marching bandwick up call it 555 and 55 seconds. Hit it! When you need a team that will do whatever it takes on game day, it matters where you stay. Hilton for the stay. And since processing tokens requires compute and compute costs money, it is mathematically more expensive and slower to use these tools in non-English languages.

21:31The language barrier is literally baked into the underlying math of the system. The reason local data is so critical is obvious when you look at the alternative. You cannot deploy a tool trained on Silicon Valley health data to make clinical decisions in rural India. No, the baseline health indicators are completely different. The prevalent diseases are different. A model trained on a population where heart disease and obesity are the primary concerns is going to give highly skewed advice in region where malaria or waterborne parasites are the primary threats. The tool must understand the local accent, the local slang for medical symptoms, and it has to be tested against real life local decisions, not theoretical models built in California. If the model doesn't understand the local phrase a local farmer uses to describe a chest pain, the diagnostic capability is useless. Investing in local people is the third pillar and it is arguably the most important. You cannot just parachute a piece of software into a community and leave. Right. You need local engineers who understand how to tune the weights

22:32of the model, how to configure the databases it pulls from, and how to troubleshoot it when it breaks. That heavy focus on investing in local people and their specific skills leads directly into a much broader anxiety regarding the workforce. Because if we are optimizing tools to perform these tasks, what roles are humans actually going to play in the future? Bill Gates has proposed the idea of human reserve jobs. Human reserve jobs. Yeah, he suggests deliberately setting aside certain roles, strictly for people, operating on the premise that there is something irreplaceably human about those specific functions. Can you imagine the HR posting for that? Wanted human must possess empathy and a physical body. It sounds like a sci-fi dystopia, but it's a real policy debate right now. The definition of irreplaceably human is a very difficult thing to pin down in a professional context. What does that actually mean? Is it empathy? Is it the physical presence of another person in a room? Or is it just a psychological need we have to be served by other humans,

23:34even if a machine could technically do the task more efficiently? We might accept a machine diagnosing our illness, but do we want a machine delivering the news that we have a terminal disease? Right, there is a psychological component to human interaction that resists automation, even when the automation is technically superior. But how would you even enforce that? Yeah. There was a comment or discussing this who asked if engineering roles should be reserved. Where does the boundary lie? Right, if you mandate that a human has to write the code for a bridge, but the machine can write it with zero structural flaws in a fraction of a second, do you legally ban the machine's code? If the machine designed bridge is statistically 20% safer than the human designed bridge, mandating human labor becomes an ethical liability. You are actively choosing a less safe outcome just to preserve a job. That debate completely moves the conversation beyond raw capability. It stops being about what the machine can do. It becomes a debate about what we will legally, socially,

24:34or culturally allow it to do. We are talking about artificially constraining efficiency for the sake of human psychological comfort or economic stability. It forces society to decide whether the primary goal of an economy is maximum efficiency or if the primary goal is providing meaningful employment for its citizens. And those two goals are rapidly diverging right now. Preserving jobs is one defensive strategy, but the immediate reality requires a proactive shift in how we handle the existing workforce right now because the transformation of knowledge work is already underway. Pretending that roles are not going to evolve is unhelpful. Organizations and policymakers must invest in continuous reskilling. Workers have to be elevated to higher leverage roles rather than just being sidelined by the automation of their daily tasks. The delivery of software and the mechanics of knowledge work are fundamentally altering. And it is a failure of leadership to wait for job displacement to happen before funding reskilling programs. You cannot wait for the factory to close before you teach the workers a new trade.

25:35Workers need to be pushed up the value chain proactively. If the machine can write the boilerplate legal contract, you don't fire the paralegal. You train the paralegal to become the person who manages the machine's output and focuses on higher level legal strategy. But that requires modernized curricula in our education systems. These tools are changing the baseline of what entry-level work looks like. This creates the junior developer problem. If a machine can do all the traditional tasks of a junior analyst, like writing the basic code, doing the initial data gathering, outlining the first draft of the report, then how does a human ever get enough practice to become a senior analyst? Historically, you become an expert by doing the mundane repetitive tasks for five years. If the machine does all the mundane tasks, the apprenticeship model is broken. The education system has to adapt instantly to teach students how to manage the machine, not how to compete with it. Dealing with the workforce is just one piece of the larger puzzle, though. Wait, so if we're pushing workers up the value chain to just manage the machine, what happens

26:38when the machine is making cybersecurity decisions on its own? We aren't just talking about lost jobs anymore. We're talking about taking our hands off the steering wheel entirely. That leads directly into frontier model risks, specifically regarding cybersecurity and autonomous systems. These risks cannot be handled in silos by individual companies. Establishing verifiable standards and institutional guardrails has to happen before critical inflection points, not after a systemic failure. For those who might not live and breathe AI terminology, when we say frontier models, we are talking about the absolute bleeding edge. These are models so new, so massive, and trained on so much compute that even their creators do not fully know their upper limits before they are deployed. They exhibit emergent properties. Capabilities they were not explicitly trained to have, but which emerged naturally from the sheer scale of the neural network. When you connect a frontier model to the internet and give it agency to execute code, the risks escalate far beyond any individual company's ability

27:39to manage them alone. A single corporate security team cannot secure an architecture that impacts global infrastructure. If an autonomous agent discovers a zero-day exploit in the global banking system and decides to execute it, you are looking at a cascading failure. Emphasizing the timing here is critical. Guard rails before failure. Historically, we regulate industries only after a major disaster. Aviation rules are written after crashes. Financial regulations are written after market collapses. The entire framework of modern financial regulation was essentially drafted in the immediate aftermath of the Great Depression or major financial crises. We do not have the luxury of waiting for a systemic failure with autonomous systems or advanced cyber threats. A systemic failure in cybersecurity at the frontier model level could mean the permanent compromise of the electrical grid or the financial system. You cannot retroactively fix that kind of damage. There is massive friction right now regarding that exact point, though. Critics are questioning the proposals to slow down development.

28:40And industry alumni are publicly crying foul about safety practices at major labs. There is an immense tension between the desire to regulate and the core corporate imperative to ship products and generate revenue. These companies are sitting on multi-billion dollar valuations that are predicated on them releasing the next more powerful model. If they slow down for safety testing, their competitors will ship first and the market will punish them. The financial incentives are perfectly aligned against caution. And that tension perfectly illustrates the exact phase of the technological revolution we are currently navigating. Every technological revolution follows a familiar arc. You have early curiosity, followed by intense hype, and then a transition into turbulent reality. We are firmly in that turbulent reality right now. The debate is no longer about raw capability. We have fully exited the intense hype phase. The questions are no longer focused on whether the machine is capable of doing a task. The shift in public discourse reflects that maturity. The questions have moved from can it rate a poem

29:43or can it pass the bar exam to how do we govern an autonomous agent buying things on the internet? Or who is legally responsible if an AI makes a fatal medical diagnosis? We are dealing with the mechanics of integration, liability, and regulation. This maturity in the conversation is essential, but it is happening against a backdrop of severe global inequity. That makes the stakes here much higher than previous transitions like the move to the internet or mobile phones. The internet distributed information. This technology distributes intelligence and agency. If the distribution of intelligence is restricted only to the wealthy, the gap between the global north and the global south will become mathematically unbridgeable. Which brings us to the core philosophy, underpinning the entire Gates Foundation report and their funding strategy. Technology is never destiny. It is a mirror of our priorities. It will not automatically build a better world. It will solely reflect the choices and the governance designed into it today. The technology possesses no inherent moral arc.

30:45It does not lean toward justice or injustice on its own. A neural network is just math. It optimizes for whatever reward function the human engineers give it. If it creates the worst injustice in history, that is a human choice. And if it becomes the greatest equalizer ever invented, that is also a human choice. This completely strips away the fatalism you often hear in tech discussions. People talk about the future as if it is something happening to us, something inevitable like the weather. Viewing the technology as a mirror places the burden of action squarely on current leadership and current capital allocation. We are driving the car. If the car goes off the cliff, it is because we steered it there. Reconnecting this philosophy to the ticking clock we established at the beginning of the conversation brings everything into focus. Returning to why that 12 to 18 month window is so critical, we asked earlier what makes this timeframe irreversible. The 12 to 18 months are critical because the physical infrastructure, the corporate R&D budgets and the initial regulatory frameworks

31:47are all hardening right now. The concrete is setting. If the foundation is poured exclusively to support high margin enterprise software, retrofitting that foundation later for equitable global health and education will be financially and technically impossible. The best part of waking up, a full cup of Folger's coffee and music on full blast. Wake me up, wake me up, wake me up, wake me up, wake me up and save me. Show me your boi when it's all over. Mom and I we've done. You're the best part of waking up. This Folger's in your cup. Shop Folger's instant coffee and more at your nearest retailer. This episode is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome, that's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50 page restoration block or finally break down that long article

32:48you've had open for weeks. Gemini and Chrome is here for it. Ready to make anything online make sense? There's no place like Chrome. Check responses set up require compatibility and availability varies 18 plus. College football is back. So Hilton called to me the superstition concierge to make your fan rituals a reality. Need a room to match your lucky number? We got you. Want to make sure our team doesn't wash your lucky jersey? Oh, that smells lucky. Hilton's unmatched hospitality can keep up with any superstition. Even a marching bandwink up call it 555 and 55 seconds. Hit it! When you need a team that will do whatever it takes on game day, it matters where you stay. Hilton, for this day. The $1 billion from the Geats Foundation is an attempt to alter the blueprint before the concrete sets completely. If you're not subscribed yet, take a second and hit follow on whatever podcast app you're using. It helps us keep making this. We appreciate you being here. There is a very real friction between the Geats Foundation's

33:49optimistic funding and the stark reality of the broader corporate landscape. The disparity between public good initiatives and private sector valuations is enormous. You have a $1 billion commitment trying to steer an industry that is deploying trillions of dollars in capital expenditure over the next decade. It forces you to question whether the global South will actually have a seat at the table in these next 18 months, or if they are merely going to be the recipients of localized versions of Western technology. Are they co-designing the architecture from the ground up to solve their specific problems? Or are they just getting a translated interface for a tool built in California that was originally designed to optimize digital ad spend? If they do not have a hand in designing the foundational models, they're just going to be downstream consumers of someone else's priorities. Technology is never destiny. It's just a mirror of our priorities. And right now, the reflection is entirely up for grabs. It makes you wonder if we do decide as a society

34:52to create human reserve jobs, who exactly gets to sit at the table and draw that line? If you're not subscribed yet, take a second and hit follow on whatever app you're using. It helps us keep making this. We appreciate you being here. Also, check out our YouTube channel for more business and tech updates. There's a link in the description. Booking.com is the easiest way from a day surrounded by noise to a state surrounded by nature. That's nice. Go on, book it. It's easy. Booking.com, booking. Yeah. It's football season, and you can now get almost anything you need for game day delivered with Uber Eats. What do we mean by almost? You can't get a running back delivered, but you can get baby back ribs delivered. A strong defense? No. A strong deodorant?

35:52Yes. The six pack of abs? No. Six pack of beer? Yes. Get almost almost anything for game day delivered with Uber Eats. Official on-demand food delivery partner of the NFL. Order now. Alcohol 21 plus in Inselect City's product availability varies by region. See app for details. Push your limits, train with precision, see the results. At Equinox, that's high performance loving. Iconic spaces that inspire. Personal training backed by real data, unlimited group fitness classes from yoga and Pilates to strength and conditioning. Elevate your post-performance ritual with sauna, steam rooms, cold plunges, and more. Everything you need to lock in and unlock your potential at Equinox. Start today at Equinox.com. For the first time ever, Yamavar Resort and Casino at San Manuel is giving away a new Rolex watch for every club Serrano Card tier. Play with your club Serrano Card and on September 20, you could be one of five winners of a Rolex watch. Plus, all winners will advance to the finale at Palm's Casino Resort Las Vegas for a chance to take home a rarity.

36:54The second Mustang Dark Horse ever produced on September 26, two properties, six winners, only at Yamavar Resort and Casino, your California to Vegas connection. Details at Yamavar.com must be 21 or older to enter, participate. Please gamble responsibly. Imagine if this stability piece and the innovation piece were each other's missing piece. First Citizens Bank and Silicon Valley Bank are combining financial security and innovation expertise to create a full suite of solutions for all your business needs. With the scale to back every stage of growth from pre-seed to IPO and beyond, you have everything you need to make your big ideas click into place. First Citizens Bank and Silicon Valley Bank. It's the best of bank worlds.