Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

The man who turned 225 million dollars into 5.5 billion dollars explained on camera exactly why he made his biggest bet. This is Leopold Aschenbrenner, the same person whose Bloom Energy position is now worth close to 2 billion dollars after Oracle's 2.8 gigawatt fuel cell deal laying out...

MilkRoadAI's profile picture

Milk Road AI

47,008 subscribers

4,641,938 görüntüleme • 4 ay önce •via X (Twitter)

Haberler & PolitikaBilim & TeknolojiFinans
LIVE

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

Leopold Aschenbrenner laid it out in 2024 and nobody listened But it's not too late to make money and here's why: • GPT-4 training cluster: $500M, 10 megawatts • 2024 cluster: billions, 100 megawatts • 2026 cluster: tens of billions, 1 gigawatt • 2030 cluster: $1 trillion, 100 gigawatts, 20% of all US electricity He said energy was the constraint the entire time Here's his full thesis: • GPUs were the trade of 2023-24 • Energy and power infrastructure is the trade now • AGI arrives by 2027 and markets haven't priced it yet • Electricity was the trade all along
1:26

Leopold Aschenbrenner laid it out in 2024 and nobody listened But it's not too late to make money and here's why: • GPT-4 training cluster: $500M, 10 megawatts • 2024 cluster: billions, 100 megawatts • 2026 cluster: tens of billions, 1 gigawatt • 2030 cluster: $1 trillion, 100 gigawatts, 20% of all US electricity He said energy was the constraint the entire time Here's his full thesis: • GPUs were the trade of 2023-24 • Energy and power infrastructure is the trade now • AGI arrives by 2027 and markets haven't priced it yet • Electricity was the trade all along

Leopold Stock Tracker

80,725 görüntüleme • 1 ay önce

The man who turned 225 million dollars into 5.5 billion dollars just laid out on camera exactly when he believes the world changes permanently with specific dates. Leopold Aschenbrenner's argument follows a single trend line that has held for over a decade without breaking. Right now in 2025 and 2026, the models being built are already smarter than most college graduates across the board. By 2027 and 2028, AI hits expert level as capable as the best professionals in any field operating not as a chatbot but as what he calls a drop-in remote worker. You assign it a project, It goes off, writes drafts, runs tests, iterates, and comes back with finished work fully autonomously, for hours at a time. The key unlock he describes is what he calls unhobbling, today's models are already more capable than most people realize, but artificially constrained by how they are deployed. Once agents can use computers freely and run long-horizon tasks without human checkpoints, the economic value unlocks almost overnight. His best guess for true AGI is the 10 gigawatt cluster range, a single data center drawing more electricity than most US states produce in total. By 2030, the trillion-dollar training cluster consumes over 20 percent of all US electricity production for a single training run. This is the direct line between that prediction and his 875 million dollar Bloom Energy position. He did not buy a power company because he liked the chart but rather bought a power company because he ran the math on what AGI physically requires to exist, and concluded that electricity is the asset class of the decade. The position is already worth close to 2 billion dollars, and his own timeline says the demand that drove it is just getting started.
1:11

The man who turned 225 million dollars into 5.5 billion dollars just laid out on camera exactly when he believes the world changes permanently with specific dates. Leopold Aschenbrenner's argument follows a single trend line that has held for over a decade without breaking. Right now in 2025 and 2026, the models being built are already smarter than most college graduates across the board. By 2027 and 2028, AI hits expert level as capable as the best professionals in any field operating not as a chatbot but as what he calls a drop-in remote worker. You assign it a project, It goes off, writes drafts, runs tests, iterates, and comes back with finished work fully autonomously, for hours at a time. The key unlock he describes is what he calls unhobbling, today's models are already more capable than most people realize, but artificially constrained by how they are deployed. Once agents can use computers freely and run long-horizon tasks without human checkpoints, the economic value unlocks almost overnight. His best guess for true AGI is the 10 gigawatt cluster range, a single data center drawing more electricity than most US states produce in total. By 2030, the trillion-dollar training cluster consumes over 20 percent of all US electricity production for a single training run. This is the direct line between that prediction and his 875 million dollar Bloom Energy position. He did not buy a power company because he liked the chart but rather bought a power company because he ran the math on what AGI physically requires to exist, and concluded that electricity is the asset class of the decade. The position is already worth close to 2 billion dollars, and his own timeline says the demand that drove it is just getting started.

Milk Road AI

375,886 görüntüleme • 4 ay önce

Two years ago, a former OpenAI researcher sat down on a podcast and calmly mapped out everything happening in markets right now. Then he ended on one question. This week the market answered it, with his own fund. In 2024, Leopold Aschenbrenner told Dwarkesh Patel that AI was an industrial process, and walked through the math: A gigawatt data center by 2026, "the power of the Hoover Dam," costing tens of billions. 10 gigawatts by 2028. More power than most US states. A trillion-dollar cluster by 2030, drawing over 20% of US electricity. Roughly $1 trillion of total AI investment by 2027. It sounded like science fiction. Then Google guided capex to $200B. Tesla started building a chip fab. Nuclear startups began raising nine figures to power data centers. The map held. But the clip ends with a sentence that aged differently than the rest: "These are forward-looking investments. The question is, do they pay off?" He raised a hedge fund on his own answer. It hit $20 billion in two years and was up 439% through June, per the FT. In July the market repriced that question. The fund fell 67%, and Ken Griffin's Citadel bought its book as it raced to meet margin calls, per the WSJ. He asked the question two years early. The answer came back with a margin call attached.
4:09

Two years ago, a former OpenAI researcher sat down on a podcast and calmly mapped out everything happening in markets right now. Then he ended on one question. This week the market answered it, with his own fund. In 2024, Leopold Aschenbrenner told Dwarkesh Patel that AI was an industrial process, and walked through the math: A gigawatt data center by 2026, "the power of the Hoover Dam," costing tens of billions. 10 gigawatts by 2028. More power than most US states. A trillion-dollar cluster by 2030, drawing over 20% of US electricity. Roughly $1 trillion of total AI investment by 2027. It sounded like science fiction. Then Google guided capex to $200B. Tesla started building a chip fab. Nuclear startups began raising nine figures to power data centers. The map held. But the clip ends with a sentence that aged differently than the rest: "These are forward-looking investments. The question is, do they pay off?" He raised a hedge fund on his own answer. It hit $20 billion in two years and was up 439% through June, per the FT. In July the market repriced that question. The fund fell 67%, and Ken Griffin's Citadel bought its book as it raced to meet margin calls, per the WSJ. He asked the question two years early. The answer came back with a margin call attached.

WOLF

64,879 görüntüleme • 20 gün önce

This is Amazon's new $11 billion AI data center campus in St. Joseph County, Indiana. It will draw 2.2 gigawatts of electricity - the output of a large power plant, enough to supply 1 million homes. For years the public was told energy had to be rationed: use less, dismantle reliable sources, accept scarcity - all to save the planet. That story has now been abandoned. AI requires massive continuous electricity to power the data centers, and AI is now seen as the superior control system. So energy is no longer a liability. It is a strategic necessity. Power will now come from wherever. Whatever works. Whatever scales. The climate narrative was tossed out the moment AI dominance became the new priority. Power was to be rationed. Now it is required. Power was the problem. Now it is the solution.
0:47

This is Amazon's new $11 billion AI data center campus in St. Joseph County, Indiana. It will draw 2.2 gigawatts of electricity - the output of a large power plant, enough to supply 1 million homes. For years the public was told energy had to be rationed: use less, dismantle reliable sources, accept scarcity - all to save the planet. That story has now been abandoned. AI requires massive continuous electricity to power the data centers, and AI is now seen as the superior control system. So energy is no longer a liability. It is a strategic necessity. Power will now come from wherever. Whatever works. Whatever scales. The climate narrative was tossed out the moment AI dominance became the new priority. Power was to be rationed. Now it is required. Power was the problem. Now it is the solution.

Electroverse

41,103 görüntüleme • 6 ay önce

Elon Musk is attempting something extraordinary. Not just building another company. He is building the largest AI training cluster in the world. A million GPUs wired together. A brute-force attempt to scale intelligence itself. The capital required to do this is staggering. Forty to sixty billion dollars for one cluster. And yet this is just one experiment. This is the scale of ambition in AI right now. It is not about incremental software. It is about rewiring the fundamental limits of what machines can learn and know.
1:02

Elon Musk is attempting something extraordinary. Not just building another company. He is building the largest AI training cluster in the world. A million GPUs wired together. A brute-force attempt to scale intelligence itself. The capital required to do this is staggering. Forty to sixty billion dollars for one cluster. And yet this is just one experiment. This is the scale of ambition in AI right now. It is not about incremental software. It is about rewiring the fundamental limits of what machines can learn and know.

Nina Schick

517,586 görüntüleme • 9 ay önce

Mark Zuckerberg says he is not an AI researcher -- so here is the one number he obsesses over to win the superintelligence race: "I'm not an AI researcher. So it's not like I'm in there telling them what research ideas to do." "The things that I can do as CEO of the company are: I can make sure that we have the very best people and talent density. I can make sure that we have by far the highest compute per researcher, and that we do whatever it takes to go build out all that capacity." "We're building this Prometheus cluster, which is going to be the first gigawatt-plus single cluster, a contiguous cluster for training, coming online next year." "We're building a cluster in Louisiana that is going to be 5 gigawatt, and we're building several more clusters like that that are just going to be multiple gigawatts." "We're talking about many hundreds of billions of dollars of capital. So you both need to have a good business model that can support it, and you need to believe in it." The tell: the richest lab in the world is not optimizing headcount. It is optimizing compute-per-researcher, and betting hundreds of billions that whoever owns the biggest contiguous cluster owns the frontier. Talent gets you into the room. Watts decide who leaves with the model.
1:06

Mark Zuckerberg says he is not an AI researcher -- so here is the one number he obsesses over to win the superintelligence race: "I'm not an AI researcher. So it's not like I'm in there telling them what research ideas to do." "The things that I can do as CEO of the company are: I can make sure that we have the very best people and talent density. I can make sure that we have by far the highest compute per researcher, and that we do whatever it takes to go build out all that capacity." "We're building this Prometheus cluster, which is going to be the first gigawatt-plus single cluster, a contiguous cluster for training, coming online next year." "We're building a cluster in Louisiana that is going to be 5 gigawatt, and we're building several more clusters like that that are just going to be multiple gigawatts." "We're talking about many hundreds of billions of dollars of capital. So you both need to have a good business model that can support it, and you need to believe in it." The tell: the richest lab in the world is not optimizing headcount. It is optimizing compute-per-researcher, and betting hundreds of billions that whoever owns the biggest contiguous cluster owns the frontier. Talent gets you into the room. Watts decide who leaves with the model.

Karl Mehta

255,587 görüntüleme • 1 ay önce

Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology right now. The United States currently consumes roughly 500 gigawatts of electricity on average. To double that capacity which is what continued AI expansion on the current terrestrial trajectory would eventually require would mean building as many power plants as currently exist in the entire country. He is not arguing that this is technically impossible, just that communities are not willing to accept it, that permitting timelines make it unrealistic, and that the hard ceiling on Earth based power generation means the expansion of AI compute will eventually hit a wall that no amount of capital can overcome on the ground. His observation is that in space, that wall does not exist. A solar panel in orbit produces roughly five times more power than the same panel on Earth, operates in continuous sunlight uninterrupted by weather or nighttime, and benefits from the vacuum of space as a completely passive cooling system meaning the two largest operating costs of any terrestrial data center, energy and cooling, are effectively eliminated. He then said that you could theoretically increase harnessed energy by a factor of one million and still be using less than a millionth of the sun's total energy output. This is the underlying physics of why SpaceX filed with the FCC to launch up to one million solar powered AI satellites, and why they described that constellation in their own filing as a first step toward becoming a Kardashev Type II civilization capable of harnessing the full power of the sun. To understand what makes this credible rather than visionary, you need to understand what SpaceX already controls that no other company on earth possesses. Starship, once operating at full cadence, can deliver 100 to 150 tons of payload to orbit per launch, at a target cost per kilogram that is an order of magnitude lower than any existing vehicle. Musk's stated ambition is to scale Starship to 10,000 to 30,000 launches per year, a frequency that would allow the deployment of orbital compute infrastructure at a pace that is currently unimaginable with any existing rocket. He told xAI staff earlier this year that achieving space-based AI at scale will eventually require manufacturing facilities on the moon, building solar panels and heat dissipation structures from lunar silicon and aluminum, and launching them into orbit from there rather than from Earth's surface because the moon's lower gravity makes the economics of launch dramatically more favorable. SpaceX's S-1 filing explicitly states that its launch capabilities could enable massive AI compute satellite constellations with the potential for millions of satellites for orbital data centers, with the first launch potentially occurring as soon as 2028. Google and Alphabet are already in advanced talks with SpaceX about deploying space-based data centers. Starcloud, a startup running Nvidia H100 GPUs in orbit, has already validated that high-performance AI inference workloads can operate in space, with plans to scale to five gigawatts of orbital compute power by 2035. This is why Musk believes the cost crossover happens in two to three years because SpaceX's launch cost trajectory intersects with the accelerating energy constraint on the ground in a way that makes space genuinely cheaper, faster, and less regulated at exactly the moment AI demand is hitting its hardest physical limits.
1:01

Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology right now. The United States currently consumes roughly 500 gigawatts of electricity on average. To double that capacity which is what continued AI expansion on the current terrestrial trajectory would eventually require would mean building as many power plants as currently exist in the entire country. He is not arguing that this is technically impossible, just that communities are not willing to accept it, that permitting timelines make it unrealistic, and that the hard ceiling on Earth based power generation means the expansion of AI compute will eventually hit a wall that no amount of capital can overcome on the ground. His observation is that in space, that wall does not exist. A solar panel in orbit produces roughly five times more power than the same panel on Earth, operates in continuous sunlight uninterrupted by weather or nighttime, and benefits from the vacuum of space as a completely passive cooling system meaning the two largest operating costs of any terrestrial data center, energy and cooling, are effectively eliminated. He then said that you could theoretically increase harnessed energy by a factor of one million and still be using less than a millionth of the sun's total energy output. This is the underlying physics of why SpaceX filed with the FCC to launch up to one million solar powered AI satellites, and why they described that constellation in their own filing as a first step toward becoming a Kardashev Type II civilization capable of harnessing the full power of the sun. To understand what makes this credible rather than visionary, you need to understand what SpaceX already controls that no other company on earth possesses. Starship, once operating at full cadence, can deliver 100 to 150 tons of payload to orbit per launch, at a target cost per kilogram that is an order of magnitude lower than any existing vehicle. Musk's stated ambition is to scale Starship to 10,000 to 30,000 launches per year, a frequency that would allow the deployment of orbital compute infrastructure at a pace that is currently unimaginable with any existing rocket. He told xAI staff earlier this year that achieving space-based AI at scale will eventually require manufacturing facilities on the moon, building solar panels and heat dissipation structures from lunar silicon and aluminum, and launching them into orbit from there rather than from Earth's surface because the moon's lower gravity makes the economics of launch dramatically more favorable. SpaceX's S-1 filing explicitly states that its launch capabilities could enable massive AI compute satellite constellations with the potential for millions of satellites for orbital data centers, with the first launch potentially occurring as soon as 2028. Google and Alphabet are already in advanced talks with SpaceX about deploying space-based data centers. Starcloud, a startup running Nvidia H100 GPUs in orbit, has already validated that high-performance AI inference workloads can operate in space, with plans to scale to five gigawatts of orbital compute power by 2035. This is why Musk believes the cost crossover happens in two to three years because SpaceX's launch cost trajectory intersects with the accelerating energy constraint on the ground in a way that makes space genuinely cheaper, faster, and less regulated at exactly the moment AI demand is hitting its hardest physical limits.

Milk Road AI

12,738 görüntüleme • 2 ay önce

Jensen Huang just told you exactly which AI bottleneck never goes away and it points directly to one of the best trades in the market right now. He did not say chip bottlenecks are permanent. He said the opposite, more chip capacity is a two to three year problem, more CoWoS packaging capacity is a two to three year problem, and none of the manufacturing constraints currently limiting Nvidia's ability to ship are structural barriers that cannot be solved. What he said is permanent or at least, far harder to solve is energy. You cannot build AI factories, reindustrialize the United States or build robots and next-generation compute without energy and energy does not respond to large purchase orders the way foundry capacity does. It involves regulatory timelines, grid interconnection queues, permitting cycles, and national policy decisions that no single company can accelerate regardless of how much capital they deploy. The US is staring down a 19 gigawatt power gap by 2028, and PJM launched an emergency integration plan earlier this year just to handle current data center load, not future load, current load. Nearly half of the data centers planned for 2026 are already delayed or canceled not because of chip shortages but because of transformer shortages, switchgear backlogs and grid capacity constraints that have nothing to do with silicon. This is the exact environment that makes Nebius a structurally differentiated position. While every other AI cloud buildout is fighting the same power bottleneck Jensen described, Nebius has already secured over 2 gigawatts of contracted power capacity with a 1.2 gigawatt campus in Missouri, another 310 megawatts in Finland and a Pennsylvania site adding another 1.2 gigawatts to the pipeline. Power is the constraint Jensen says no one can shortcut and Nebius has already locked in more of it than almost any independent AI cloud operator on the planet. That power moat sits underneath a $27 billion contracted revenue deal with Meta, a Microsoft partnership ramping to full run rate in 2027, and a 684% year over year revenue growth number that just printed in Q1. The Nebius thesis was always about infrastructure scarcity in a world of accelerating demand, Jensen just confirmed on camera which scarcity actually matters long-term, and it is exactly the one Nebius spent the last two years solving before anyone else was paying attention. Milk Road Pro called Nebius early, has been sitting on a massive gain on the position, and continues to track the infrastructure plays that matter before they become obvious to the rest of the market, come join us at the link in bio/below!
0:32

Jensen Huang just told you exactly which AI bottleneck never goes away and it points directly to one of the best trades in the market right now. He did not say chip bottlenecks are permanent. He said the opposite, more chip capacity is a two to three year problem, more CoWoS packaging capacity is a two to three year problem, and none of the manufacturing constraints currently limiting Nvidia's ability to ship are structural barriers that cannot be solved. What he said is permanent or at least, far harder to solve is energy. You cannot build AI factories, reindustrialize the United States or build robots and next-generation compute without energy and energy does not respond to large purchase orders the way foundry capacity does. It involves regulatory timelines, grid interconnection queues, permitting cycles, and national policy decisions that no single company can accelerate regardless of how much capital they deploy. The US is staring down a 19 gigawatt power gap by 2028, and PJM launched an emergency integration plan earlier this year just to handle current data center load, not future load, current load. Nearly half of the data centers planned for 2026 are already delayed or canceled not because of chip shortages but because of transformer shortages, switchgear backlogs and grid capacity constraints that have nothing to do with silicon. This is the exact environment that makes Nebius a structurally differentiated position. While every other AI cloud buildout is fighting the same power bottleneck Jensen described, Nebius has already secured over 2 gigawatts of contracted power capacity with a 1.2 gigawatt campus in Missouri, another 310 megawatts in Finland and a Pennsylvania site adding another 1.2 gigawatts to the pipeline. Power is the constraint Jensen says no one can shortcut and Nebius has already locked in more of it than almost any independent AI cloud operator on the planet. That power moat sits underneath a $27 billion contracted revenue deal with Meta, a Microsoft partnership ramping to full run rate in 2027, and a 684% year over year revenue growth number that just printed in Q1. The Nebius thesis was always about infrastructure scarcity in a world of accelerating demand, Jensen just confirmed on camera which scarcity actually matters long-term, and it is exactly the one Nebius spent the last two years solving before anyone else was paying attention. Milk Road Pro called Nebius early, has been sitting on a massive gain on the position, and continues to track the infrastructure plays that matter before they become obvious to the rest of the market, come join us at the link in bio/below!

Milk Road AI

21,316 görüntüleme • 3 ay önce

"The limiting factor for AI deployment is fundamentally electrical power." That was Elon Musk in a conversation with Larry Fink during his first-ever appearance at Davos. And it's the most honest thing anyone in Big Tech has said in a while. Forget the hype about superintelligence and robots. Forget the promises about productivity gains that CFOs still can't measure. The bottleneck is power. And that bottleneck is very real. Here's why & how to position yourself to make the most out of this: We're producing more chips than we can turn on. AI chip production is increasing exponentially. US electricity generation is growing 3-4% annually. The math doesn't work. US data center power demand is expected to hit 75.8 gigawatts in 2026, up from 61.8 gigawatts in 2025. By 2030, it could reach 134 gigawatts. The largest US grid operator, PJM, expects to add 31 gigawatts of data center load over the next five years. But only 28 gigawatts of new generation capacity is planned. The deficit is already here. Residential electricity rates near data centers have jumped as much as 267% compared to five years ago. Regular Americans are subsidizing Silicon Valley's power consumption. Whether AI delivers on its big promises or not, the electricity bills are coming due right now. Elon made another point that deserves attention: China is solving the energy problem while America talks about AI miracles. He noted China's solar deployment is "tremendous." And that's an understatement. I verified the numbers. They're insane: China installed 275 gigawatts of solar capacity in the first 11 months of 2025. That's more than the ENTIRE installed solar capacity of the United States. In May 2025 alone, China added 93 gigawatts. Roughly 100 panels every second. For the first time in history, a single country surpassed 1,000 gigawatts of total solar capacity. China added more solar in one month than America has built in its entire history... But this ISN'T about climate policy. It's about strategic positioning. China understands something Wall Street is ignoring: whoever controls the energy infrastructure controls the next era of computing. While American investors chase the latest AI stock, China is building the power grid that will actually run the technology. My take: The data center buildout will fall short of expectations. Big Tech wants trillions in infrastructure. The grid can't deliver it. Not at this pace. Not with 3-4% annual electricity growth against exponential demand projections. Something has to give. Either the buildout slows dramatically, or electricity prices spike to levels that destroy the economics of the entire AI story. The AI trade has been built on the assumption that infrastructure will materialize to meet demand. That assumption is looking increasingly shaky. BUT the companies selling power to Big Tech win either way. Scarcity means pricing power. Utilities don't need AI to cure cancer. They just need hyperscalers to keep signing power agreements. And they are. Microsoft, Amazon, Meta, and Alphabet spent ~$350B in 2025 on data centers. That money flows to utilities and grid infrastructure whether AI changes the world or not. US utilities are forecasting a ~6% jump in capital expenditures to $228B in 2026. Cumulative utility capex is expected to surpass $1.1T through 2029. Dominion Energy is investing $50B through 2029, projecting 5-7% annual earnings growth. Entergy plans $41B between 2026 and 2029, targeting more than 8% compound annual earnings growth. These aren't speculative bets on AI changing everything... They're regulated utilities with contracted demand and predictable cash flows. So beware the AI story. The productivity miracle remains unproven. The physical constraints are becoming impossible to ignore. But the picks-and-shovels play? THAT'S where the smart money is looking. The hyperscalers have to pay their electricity bills. I'd rather own the companies collecting them.
1:39

"The limiting factor for AI deployment is fundamentally electrical power." That was Elon Musk in a conversation with Larry Fink during his first-ever appearance at Davos. And it's the most honest thing anyone in Big Tech has said in a while. Forget the hype about superintelligence and robots. Forget the promises about productivity gains that CFOs still can't measure. The bottleneck is power. And that bottleneck is very real. Here's why & how to position yourself to make the most out of this: We're producing more chips than we can turn on. AI chip production is increasing exponentially. US electricity generation is growing 3-4% annually. The math doesn't work. US data center power demand is expected to hit 75.8 gigawatts in 2026, up from 61.8 gigawatts in 2025. By 2030, it could reach 134 gigawatts. The largest US grid operator, PJM, expects to add 31 gigawatts of data center load over the next five years. But only 28 gigawatts of new generation capacity is planned. The deficit is already here. Residential electricity rates near data centers have jumped as much as 267% compared to five years ago. Regular Americans are subsidizing Silicon Valley's power consumption. Whether AI delivers on its big promises or not, the electricity bills are coming due right now. Elon made another point that deserves attention: China is solving the energy problem while America talks about AI miracles. He noted China's solar deployment is "tremendous." And that's an understatement. I verified the numbers. They're insane: China installed 275 gigawatts of solar capacity in the first 11 months of 2025. That's more than the ENTIRE installed solar capacity of the United States. In May 2025 alone, China added 93 gigawatts. Roughly 100 panels every second. For the first time in history, a single country surpassed 1,000 gigawatts of total solar capacity. China added more solar in one month than America has built in its entire history... But this ISN'T about climate policy. It's about strategic positioning. China understands something Wall Street is ignoring: whoever controls the energy infrastructure controls the next era of computing. While American investors chase the latest AI stock, China is building the power grid that will actually run the technology. My take: The data center buildout will fall short of expectations. Big Tech wants trillions in infrastructure. The grid can't deliver it. Not at this pace. Not with 3-4% annual electricity growth against exponential demand projections. Something has to give. Either the buildout slows dramatically, or electricity prices spike to levels that destroy the economics of the entire AI story. The AI trade has been built on the assumption that infrastructure will materialize to meet demand. That assumption is looking increasingly shaky. BUT the companies selling power to Big Tech win either way. Scarcity means pricing power. Utilities don't need AI to cure cancer. They just need hyperscalers to keep signing power agreements. And they are. Microsoft, Amazon, Meta, and Alphabet spent ~$350B in 2025 on data centers. That money flows to utilities and grid infrastructure whether AI changes the world or not. US utilities are forecasting a ~6% jump in capital expenditures to $228B in 2026. Cumulative utility capex is expected to surpass $1.1T through 2029. Dominion Energy is investing $50B through 2029, projecting 5-7% annual earnings growth. Entergy plans $41B between 2026 and 2029, targeting more than 8% compound annual earnings growth. These aren't speculative bets on AI changing everything... They're regulated utilities with contracted demand and predictable cash flows. So beware the AI story. The productivity miracle remains unproven. The physical constraints are becoming impossible to ignore. But the picks-and-shovels play? THAT'S where the smart money is looking. The hyperscalers have to pay their electricity bills. I'd rather own the companies collecting them.

George Noble

23,931 görüntüleme • 6 ay önce

BlackRock CEO Larry Fink says the biggest job boom coming to America has nothing to do with AI software or Wall Street, it's in infrastructure. Specifically, the construction, power grids, and data centers. That is where he believes the massive job growth is coming from. And his warning is that we are not ready for it. Companies are committing hundreds of billions to build out AI infrastructure across the US, Microsoft, Amazon, Meta, and Alphabet are projected to spend $650 billion on it this year alone. But that buildout requires physical workers. Electricians, pipefitters, construction crews, grid engineers. Fink told the Trump administration directly, the US doesn't have enough skilled workers to meet the demand these projects will create. Microsoft's Brad Smith called the skilled labor shortage the "single biggest challenge for data center expansion in the U.S." In some regions, Microsoft had to relocate workers or ask them to commute 75 miles just to fill critical roles. The electrical work alone accounts for 45–70% of total data center construction costs. Over 200,000 electricians are expected to retire in the next decade. The pipeline to replace them doesn't exist at the scale needed. Energy resilience is the other piece Fink flagged. Data centers don't just need to be built, they need consistent, reliable power. The US grid hasn't been meaningfully expanded in decades. China, by contrast, generates roughly twice the electricity the US does and is actively building more. Fink's point is direct because if the US doesn't win the buildout of its physical infrastructure, China will. That is why BlackRock just committed $100 million to fund training programs for 50,000 skilled trade workers over the next five years. The largest asset manager in the world is putting money into electricians and plumbers because it sees the bottleneck clearly.
0:29

BlackRock CEO Larry Fink says the biggest job boom coming to America has nothing to do with AI software or Wall Street, it's in infrastructure. Specifically, the construction, power grids, and data centers. That is where he believes the massive job growth is coming from. And his warning is that we are not ready for it. Companies are committing hundreds of billions to build out AI infrastructure across the US, Microsoft, Amazon, Meta, and Alphabet are projected to spend $650 billion on it this year alone. But that buildout requires physical workers. Electricians, pipefitters, construction crews, grid engineers. Fink told the Trump administration directly, the US doesn't have enough skilled workers to meet the demand these projects will create. Microsoft's Brad Smith called the skilled labor shortage the "single biggest challenge for data center expansion in the U.S." In some regions, Microsoft had to relocate workers or ask them to commute 75 miles just to fill critical roles. The electrical work alone accounts for 45–70% of total data center construction costs. Over 200,000 electricians are expected to retire in the next decade. The pipeline to replace them doesn't exist at the scale needed. Energy resilience is the other piece Fink flagged. Data centers don't just need to be built, they need consistent, reliable power. The US grid hasn't been meaningfully expanded in decades. China, by contrast, generates roughly twice the electricity the US does and is actively building more. Fink's point is direct because if the US doesn't win the buildout of its physical infrastructure, China will. That is why BlackRock just committed $100 million to fund training programs for 50,000 skilled trade workers over the next five years. The largest asset manager in the world is putting money into electricians and plumbers because it sees the bottleneck clearly.

StockMarket.News

46,903 görüntüleme • 5 ay önce

This is Amazon’s new $11 billion dollar massive Data Center Campus in St. Joseph County, Indiana It will be primarily dedicated to training and running AI models It will use 2.2 gigawatts of power, equivalent to the electricity needed to power roughly 1 million homes and approximately 300 million gallons of water per year
2:09

This is Amazon’s new $11 billion dollar massive Data Center Campus in St. Joseph County, Indiana It will be primarily dedicated to training and running AI models It will use 2.2 gigawatts of power, equivalent to the electricity needed to power roughly 1 million homes and approximately 300 million gallons of water per year

Wall Street Apes

3,026,162 görüntüleme • 8 ay önce

Elon Musk explains the “series of miracles” xAI had to achieve to bring one gigawatt of power online for its supercomputer cluster in Memphis The xAI team coordinated dozens of gas turbines, faced permit challenges in Tennessee, and ultimately moved the power generation just across the border into Mississippi. From there, they had to run high-voltage transmission lines and build an entire power plant on the Mississippi side and it was very difficult to build that But the deeper point is even more important Most people calculate AI power needs by looking only at the GPUs. That is a serious mistake. A real data center also requires substantial power for networking, CPUs, storage, and most critically cooling In Memphis, where summer temperatures are extreme, peak cooling demand can add roughly 40% more power. On top of that, operators must maintain extra capacity so generators can be taken offline for servicing without shutting down the entire system. This redundancy typically adds another 20 to 25 percent When all of these factors are included, Elon stated that approximately 110,000 GB300 GPUs require roughly 330 megawatts of actual power generation This is why building and powering frontier AI data centers is far more difficult and far more capital intensive than it appears from the outside
2:43

Elon Musk explains the “series of miracles” xAI had to achieve to bring one gigawatt of power online for its supercomputer cluster in Memphis The xAI team coordinated dozens of gas turbines, faced permit challenges in Tennessee, and ultimately moved the power generation just across the border into Mississippi. From there, they had to run high-voltage transmission lines and build an entire power plant on the Mississippi side and it was very difficult to build that But the deeper point is even more important Most people calculate AI power needs by looking only at the GPUs. That is a serious mistake. A real data center also requires substantial power for networking, CPUs, storage, and most critically cooling In Memphis, where summer temperatures are extreme, peak cooling demand can add roughly 40% more power. On top of that, operators must maintain extra capacity so generators can be taken offline for servicing without shutting down the entire system. This redundancy typically adds another 20 to 25 percent When all of these factors are included, Elon stated that approximately 110,000 GB300 GPUs require roughly 330 megawatts of actual power generation This is why building and powering frontier AI data centers is far more difficult and far more capital intensive than it appears from the outside

X Freeze

140,377 görüntüleme • 3 ay önce

Dr. Eric Schmidt says AI needs “energy in all forms: renewable, non-renewable, whatever.” “It needs to be there, and it needs to be quickly.” “People are planning 10 gigawatt data centers.” “An average nuclear power plant in the United States is one gigawatt.” “Data centers will require an additional 29 gigawatts of power by 2027 and 67 more gigawatts by 2030.” “These things are industrial at a scale I have never seen in my life.” This is exactly why I believe the market is still underestimating the AI power and infrastructure opportunity in companies like $TE, $BE, $PUMP, $PSIX, $SMR, $FCEL, $MARA, $KEEL, and $OKLO. The biggest constraint to scaling AI remains electricity. The companies building power generation, grid infrastructure, electrical equipment, cooling systems, and AI data centers could be among the biggest beneficiaries over the next decade. Not financial advice. Do your own research.
0:51

Dr. Eric Schmidt says AI needs “energy in all forms: renewable, non-renewable, whatever.” “It needs to be there, and it needs to be quickly.” “People are planning 10 gigawatt data centers.” “An average nuclear power plant in the United States is one gigawatt.” “Data centers will require an additional 29 gigawatts of power by 2027 and 67 more gigawatts by 2030.” “These things are industrial at a scale I have never seen in my life.” This is exactly why I believe the market is still underestimating the AI power and infrastructure opportunity in companies like $TE, $BE, $PUMP, $PSIX, $SMR, $FCEL, $MARA, $KEEL, and $OKLO. The biggest constraint to scaling AI remains electricity. The companies building power generation, grid infrastructure, electrical equipment, cooling systems, and AI data centers could be among the biggest beneficiaries over the next decade. Not financial advice. Do your own research.

Leopold Stock Tracker

20,643 görüntüleme • 1 ay önce

OpenAI is building a $20 billion dollar project in GA, Residents just found out The project will consume 3.2 gigawatts of electricity, enough to power more than 2 million homes - almost half of all homes in Georgia. The project will use massive amounts of water in an area that is already facing shortages. At nearly $30 billion dollars total cost, this will be the most expensive project in Georgia's history. Open AI won't pay a dime in sales tax. No vote was held, the people of Georgia were not asked if they wanted this.
2:30

OpenAI is building a $20 billion dollar project in GA, Residents just found out The project will consume 3.2 gigawatts of electricity, enough to power more than 2 million homes - almost half of all homes in Georgia. The project will use massive amounts of water in an area that is already facing shortages. At nearly $30 billion dollars total cost, this will be the most expensive project in Georgia's history. Open AI won't pay a dime in sales tax. No vote was held, the people of Georgia were not asked if they wanted this.

Orwell Day

23,295 görüntüleme • 24 gün önce

This is Amazon’s new $11 billion dollar massive Data Center Campus in St. Joseph County, Indiana It will be primarily dedicated to training and running AI models It will use 2.2 gigawatts of power, equivalent to the electricity needed to power roughly 1 million homes and approximately 300 million gallons of water per year Edward Snowden Telegram Channel
2:09

This is Amazon’s new $11 billion dollar massive Data Center Campus in St. Joseph County, Indiana It will be primarily dedicated to training and running AI models It will use 2.2 gigawatts of power, equivalent to the electricity needed to power roughly 1 million homes and approximately 300 million gallons of water per year Edward Snowden Telegram Channel

Michael

562,461 görüntüleme • 7 ay önce

Elon is a genius and there's no other way to explain what he built. The bears biggest argument against SpaceX's $1.75–2 trillion IPO valuation has always been the orbital data center play. Space based compute is unproven, technically complex, and expensive at a scale that defies easy modeling. SpaceX's own S-1 acknowledged it directly that these data centers may not achieve commercial viability. Critics were right to flag it as speculative but what they missed was that Elon was already building the hedge. While the debate about orbital compute was still theoretical, he was assembling the largest terrestrial AI infrastructure footprint on Earth. Colossus 1, now leased to Anthropic houses 230,000+ GPUs at 500 MW. Colossus 2 (Macro-Hard) runs 550,000 Blackwell GPUs targeting over 1 gigawatt enough to power roughly 750,000 American homes. The third facility, Macro-Harder, adds another ~500 MW of capacity across 810,000 square feet in Southaven, Mississippi, pushing the total campus toward 2 gigawatts and over 1 million GPUs. He secured all of this, the land, the power contracts, the chips before the rest of the market understood that power and compute would become the single most constrained resource in the AI economy. Now that constraint is everywhere and he owns the supply. Starlink is already generating $11.4 billion in annual revenue with 63% EBITDA margins and 10 million+ subscribers. The company is projecting $20 billion in total revenue across the combined entity in 2026 and that existing cash engine now funds the entire AI buildout and the Anthropic compute deal alone is estimated to generate an incremental $4–5 billion in revenue this year. The orbital data center bet may still pay off on top of all of this. But it doesn't have to, the terrestrial capacity alone, at current utilization rates, subsidizes Grok training, generates hyperscaler revenue, and floors the SpaceX valuation story against every bear case scenario. Elon didn't just build rockets but rather he built the infrastructure layer that the entire AI industry now depends on and then started selling access to the people who need it most. Go PRO at Milk Road to see how our analysts are positioning for what could become the trade of a lifetime, the SpaceX IPO and the AI infrastructure supercycle. Link below!
1:05

Elon is a genius and there's no other way to explain what he built. The bears biggest argument against SpaceX's $1.75–2 trillion IPO valuation has always been the orbital data center play. Space based compute is unproven, technically complex, and expensive at a scale that defies easy modeling. SpaceX's own S-1 acknowledged it directly that these data centers may not achieve commercial viability. Critics were right to flag it as speculative but what they missed was that Elon was already building the hedge. While the debate about orbital compute was still theoretical, he was assembling the largest terrestrial AI infrastructure footprint on Earth. Colossus 1, now leased to Anthropic houses 230,000+ GPUs at 500 MW. Colossus 2 (Macro-Hard) runs 550,000 Blackwell GPUs targeting over 1 gigawatt enough to power roughly 750,000 American homes. The third facility, Macro-Harder, adds another ~500 MW of capacity across 810,000 square feet in Southaven, Mississippi, pushing the total campus toward 2 gigawatts and over 1 million GPUs. He secured all of this, the land, the power contracts, the chips before the rest of the market understood that power and compute would become the single most constrained resource in the AI economy. Now that constraint is everywhere and he owns the supply. Starlink is already generating $11.4 billion in annual revenue with 63% EBITDA margins and 10 million+ subscribers. The company is projecting $20 billion in total revenue across the combined entity in 2026 and that existing cash engine now funds the entire AI buildout and the Anthropic compute deal alone is estimated to generate an incremental $4–5 billion in revenue this year. The orbital data center bet may still pay off on top of all of this. But it doesn't have to, the terrestrial capacity alone, at current utilization rates, subsidizes Grok training, generates hyperscaler revenue, and floors the SpaceX valuation story against every bear case scenario. Elon didn't just build rockets but rather he built the infrastructure layer that the entire AI industry now depends on and then started selling access to the people who need it most. Go PRO at Milk Road to see how our analysts are positioning for what could become the trade of a lifetime, the SpaceX IPO and the AI infrastructure supercycle. Link below!

Milk Road AI

63,920 görüntüleme • 3 ay önce

Nebius is one of the most undervalued AI infrastructure companies in the public markets right now (Save this). Leopold Aschenbrenner, the former OpenAI researcher who wrote the 165-page essay predicting AGI within this decade and then launched the $13.7 billion Situational Awareness Fund around that thesis just filed a 13G disclosing a 5.6% stake in Nebius, representing 12.41 million Class A shares. This is the man whose entire investment framework is built on one core conviction, AI will advance faster than anyone expects, and the binding constraint will not be algorithms or model architectures, it will be physical computing infrastructure, data center capacity, and energy. Now look at what Nebius actually is and why this conviction is justified by the numbers alone. Nebius is a GPU native AI cloud platform, a neocloud built from the ground up specifically for AI training and inference workloads, founded by Arkady Volozh, the former CEO of Yandex who divested all non-Russian assets and left Russia in direct opposition to Putin before relisting the company on Nasdaq. In Q1 2026, Nebius reported $399 million in revenue, a 684% increase year over year from just $50.9 million while also delivering EBITDA and adjusted EPS that beat consensus estimates by 43% and 50% respectively, in a quarter where analysts had already built in aggressive assumptions. The scale of the infrastructure buildout is what makes the valuation argument so compelling. Nebius has raised its contracted power capacity guidance to over 4 gigawatts for 2026, with a target of 5 gigawatts of AI computing capacity deployed by 2030, including multiple gigawatt-scale AI factories across the United States and Europe. The Finland campus coming soon to Lappeenranta will be 310 megawatts powered by low-carbon energy, making it one of the largest AI data centers in Europe, specifically located in a cold-climate, energy-stable region that dramatically reduces cooling costs and carbon intensity. The 2026 capacity is already effectively sold out according to management disclosures, which means every megawatt Nebius brings online has a revenue contract attached to it before the facility opens. The strategic backing validates the thesis at every level. NVIDIA committed a $2 billion strategic investment in Nebius by 2030, with the two companies co-developing an inference stack, implementing NVIDIA's GPU health monitoring systems, and deploying next-generation architectures including Rubin GPUs, Vera CPUs, and Bluefield storage systems meaning Nebius gets preferential access to the hardware that every other AI company is begging Jensen Huang for. Meta signed a $27 billion agreement with Nebius, with $12 billion in dedicated computing resources confirmed and up to $15 billion in additional capacity over the coming years. And Nebius just partnered with Bloom Energy on a $2.6 billion deal guaranteeing 328 megawatts of installed capacity through modular fuel cell systems behind the meter power that eliminates grid dependency and accelerates deployment timelines. The forward valuation math is where the undervaluation case becomes undeniable. Nebius is pricing in $3.5 billion in revenue for 2026 and $11 billion for 2027, which puts the forward price-to-sales ratio at 16.6 times for this year and just 5.3 times for next year for a company growing revenue at 684% year over year with sold out capacity, NVIDIA backing, a $27 billion Meta contract, and a path to 4+ gigawatts of contracted power. Milk Road has been positioned in Nebius and we believe the convergence of Leopold's conviction stake, NVIDIA's $2 billion endorsement, Meta's $27 billion commitment, and a physical infrastructure buildout that is sold out before it opens represents one of the highest-quality risk-reward setups in AI infrastructure today. Come join Milk Road Pro and get our full Nebius thesis including the exact framework we use to think about neocloud valuation, the power capacity math that determines when revenue accelerates, and every catalyst we are watching through 2027. Link in bio/below.
0:20

Nebius is one of the most undervalued AI infrastructure companies in the public markets right now (Save this). Leopold Aschenbrenner, the former OpenAI researcher who wrote the 165-page essay predicting AGI within this decade and then launched the $13.7 billion Situational Awareness Fund around that thesis just filed a 13G disclosing a 5.6% stake in Nebius, representing 12.41 million Class A shares. This is the man whose entire investment framework is built on one core conviction, AI will advance faster than anyone expects, and the binding constraint will not be algorithms or model architectures, it will be physical computing infrastructure, data center capacity, and energy. Now look at what Nebius actually is and why this conviction is justified by the numbers alone. Nebius is a GPU native AI cloud platform, a neocloud built from the ground up specifically for AI training and inference workloads, founded by Arkady Volozh, the former CEO of Yandex who divested all non-Russian assets and left Russia in direct opposition to Putin before relisting the company on Nasdaq. In Q1 2026, Nebius reported $399 million in revenue, a 684% increase year over year from just $50.9 million while also delivering EBITDA and adjusted EPS that beat consensus estimates by 43% and 50% respectively, in a quarter where analysts had already built in aggressive assumptions. The scale of the infrastructure buildout is what makes the valuation argument so compelling. Nebius has raised its contracted power capacity guidance to over 4 gigawatts for 2026, with a target of 5 gigawatts of AI computing capacity deployed by 2030, including multiple gigawatt-scale AI factories across the United States and Europe. The Finland campus coming soon to Lappeenranta will be 310 megawatts powered by low-carbon energy, making it one of the largest AI data centers in Europe, specifically located in a cold-climate, energy-stable region that dramatically reduces cooling costs and carbon intensity. The 2026 capacity is already effectively sold out according to management disclosures, which means every megawatt Nebius brings online has a revenue contract attached to it before the facility opens. The strategic backing validates the thesis at every level. NVIDIA committed a $2 billion strategic investment in Nebius by 2030, with the two companies co-developing an inference stack, implementing NVIDIA's GPU health monitoring systems, and deploying next-generation architectures including Rubin GPUs, Vera CPUs, and Bluefield storage systems meaning Nebius gets preferential access to the hardware that every other AI company is begging Jensen Huang for. Meta signed a $27 billion agreement with Nebius, with $12 billion in dedicated computing resources confirmed and up to $15 billion in additional capacity over the coming years. And Nebius just partnered with Bloom Energy on a $2.6 billion deal guaranteeing 328 megawatts of installed capacity through modular fuel cell systems behind the meter power that eliminates grid dependency and accelerates deployment timelines. The forward valuation math is where the undervaluation case becomes undeniable. Nebius is pricing in $3.5 billion in revenue for 2026 and $11 billion for 2027, which puts the forward price-to-sales ratio at 16.6 times for this year and just 5.3 times for next year for a company growing revenue at 684% year over year with sold out capacity, NVIDIA backing, a $27 billion Meta contract, and a path to 4+ gigawatts of contracted power. Milk Road has been positioned in Nebius and we believe the convergence of Leopold's conviction stake, NVIDIA's $2 billion endorsement, Meta's $27 billion commitment, and a physical infrastructure buildout that is sold out before it opens represents one of the highest-quality risk-reward setups in AI infrastructure today. Come join Milk Road Pro and get our full Nebius thesis including the exact framework we use to think about neocloud valuation, the power capacity math that determines when revenue accelerates, and every catalyst we are watching through 2027. Link in bio/below.

Milk Road AI

61,932 görüntüleme • 2 ay önce

The man who won the Nobel Prize just told the world that AI is not the energy crisis, it is the cure for it. Everyone has been screaming about how much electricity AI consumes, the data centers, the training runs, the billions of queries every single day. Hassabis just said AI will extract 30 to 40 percent more efficiency out of national power grids, grids that, right now, operate at only 30 percent of their total capacity according to Stanford researchers. That means the grids we already built are massively underused, and AI is the key to unlocking what is already there. But that is the smallest part of what he is saying. He is saying AI will crack nuclear fusion, the energy source that has been 30 years away for the last 60 years and DeepMind is already working with Commonwealth Fusion in the US to help AI contain plasma inside fusion reactors. His personal mission is to use AI to discover a room-temperature superconductor, a material that would allow electricity to travel with zero loss, something physics has never been able to deliver. The implications of that alone would reshape the entire global economy overnight. He is also saying AI will design next-generation batteries and build the best climate modeling systems humanity has ever had, using them to figure out exactly where the planet is breaking down. Think about what this means, the thing everyone is blaming for the energy crisis is the same thing being bet on to end it forever.
0:51

The man who won the Nobel Prize just told the world that AI is not the energy crisis, it is the cure for it. Everyone has been screaming about how much electricity AI consumes, the data centers, the training runs, the billions of queries every single day. Hassabis just said AI will extract 30 to 40 percent more efficiency out of national power grids, grids that, right now, operate at only 30 percent of their total capacity according to Stanford researchers. That means the grids we already built are massively underused, and AI is the key to unlocking what is already there. But that is the smallest part of what he is saying. He is saying AI will crack nuclear fusion, the energy source that has been 30 years away for the last 60 years and DeepMind is already working with Commonwealth Fusion in the US to help AI contain plasma inside fusion reactors. His personal mission is to use AI to discover a room-temperature superconductor, a material that would allow electricity to travel with zero loss, something physics has never been able to deliver. The implications of that alone would reshape the entire global economy overnight. He is also saying AI will design next-generation batteries and build the best climate modeling systems humanity has ever had, using them to figure out exactly where the planet is breaking down. Think about what this means, the thing everyone is blaming for the energy crisis is the same thing being bet on to end it forever.

Milk Road AI

41,417 görüntüleme • 4 ay önce

Elon Musk announced a chip factory 10 times the size of Tesla's Gigafactory. The goal is to produce enough AI compute to equal twice the entire electricity consumption of the United States. He called it the Terafab. Here is the number that stopped me cold. The entire global AI chip industry right now is on track to hit around 100 gigawatts per year of compute. Every Nvidia GPU, every Google TPU, every chip from every company on earth combined. 100 gigawatts. Musk wants one factory to produce a terawatt per year. A terawatt is 1,000 gigawatts. Ten times the output of the entire global industry. From a single building. To put the scale in physical terms, the Terafab would need to be around 100 million square feet. You would need Starship point to point transport just to get from one end to the other. But the reason for the scale is not ambition for its own sake. To launch meaningful AI compute into space, you need a billion chips per year running at a kilowatt each. That is not a number the current industry can produce. The Terafab is the only way to get there. The timeline he put out: a gigawatt of space AI compute annualized by end of next year. Then 10x per year from there. 10 gigawatts by year two. 100 gigawatts by year three. A terawatt beyond that. Most people think orbital data centers are a decade away. Musk is building the factory to make them possible by next year.
3:26

Elon Musk announced a chip factory 10 times the size of Tesla's Gigafactory. The goal is to produce enough AI compute to equal twice the entire electricity consumption of the United States. He called it the Terafab. Here is the number that stopped me cold. The entire global AI chip industry right now is on track to hit around 100 gigawatts per year of compute. Every Nvidia GPU, every Google TPU, every chip from every company on earth combined. 100 gigawatts. Musk wants one factory to produce a terawatt per year. A terawatt is 1,000 gigawatts. Ten times the output of the entire global industry. From a single building. To put the scale in physical terms, the Terafab would need to be around 100 million square feet. You would need Starship point to point transport just to get from one end to the other. But the reason for the scale is not ambition for its own sake. To launch meaningful AI compute into space, you need a billion chips per year running at a kilowatt each. That is not a number the current industry can produce. The Terafab is the only way to get there. The timeline he put out: a gigawatt of space AI compute annualized by end of next year. Then 10x per year from there. 10 gigawatts by year two. 100 gigawatts by year three. A terawatt beyond that. Most people think orbital data centers are a decade away. Musk is building the factory to make them possible by next year.

Ihtesham Ali

317,824 görüntüleme • 2 ay önce

The cheapest place to put AI will be space. Elon Musk says solar panels in space produce 5 times more power than on the ground. No atmosphere, no clouds, no day-night cycle. No batteries needed. And critically, no permits. Scaling energy on Earth means fighting land use regulations, environmental reviews, and grid interconnection queues that take years. Space has none of that. The math gets wild fast. SpaceX could launch hundreds of gigawatts of AI compute into orbit annually. That's more compute than currently exists on Earth. Ten thousand Starship launches per year. One launch every hour. And you only need 20 to 30 ships to sustain that cadence. Earth gets about half a billionth of the Sun's total energy output. A millionth of the Sun's energy is 100,000 times all of civilization's current electricity production. The AI race is not a software race. It is a power plant race. And software people are about to get a very hard lesson in hardware.
2:19

The cheapest place to put AI will be space. Elon Musk says solar panels in space produce 5 times more power than on the ground. No atmosphere, no clouds, no day-night cycle. No batteries needed. And critically, no permits. Scaling energy on Earth means fighting land use regulations, environmental reviews, and grid interconnection queues that take years. Space has none of that. The math gets wild fast. SpaceX could launch hundreds of gigawatts of AI compute into orbit annually. That's more compute than currently exists on Earth. Ten thousand Starship launches per year. One launch every hour. And you only need 20 to 30 ships to sustain that cadence. Earth gets about half a billionth of the Sun's total energy output. A millionth of the Sun's energy is 100,000 times all of civilization's current electricity production. The AI race is not a software race. It is a power plant race. And software people are about to get a very hard lesson in hardware.

Shruti

11,700 görüntüleme • 1 ay önce