Vikram M
@Vvikramai • 3,827 subscribers
Trying to understand AI before it understands us. Ex Founder | Sharing what I'm learning
Videos
Jensen Huang says that AI is "a complete reset" in the race between Germany, Japan and China. Jensen Huang: "Electrification resulted in really three geographical leaders: Germany, the United States, and Japan. For the last 50 years, because of the current industrial revolution, which is computing and digital technology, the United States has largely exported its heavy manufacturing, its buildings, and its crafts really to China." "China is probably the only nation today that is deep in manufacturing and deep in computer science and very comfortable in both." "The good news is there's a complete reset. The last 50 years were about software programming. There's a new era now. The computer programs itself. You just have to show the computer what you want." "All the digital technology in the world will not replace electromechanical physical technology. Electrons will never replace atoms." "This industrial revolution is so mighty, so powerful. You don't have another chance. Whether we have another chance or not, we should always assume we don't." The German chancellor's response: "I share your view that the Americans made a big mistake to export their manufacturing, and now they are faced with a very small, successful, but very small-based industry."
Vikram M285,943 Aufrufe • vor 8 Tagen
Satya Nadella was asked why Microsoft, with a $175 billion AI budget, still doesn't have a frontier model. Satya Nadella: "We are well on our way building our MAI models our goal is to basically hill climb from the bottom, by the way, not distilling anything." "My advice is more like use all but be independent of all. I would pull out a model and see whether I can retain the eval. If I can't, that means you really are dependent on something that may or may not be yours." "If you're a hyperscaler, you're not a supplier to two model companies, that's not a business we need more." "I worked on Windows interrupt with Unix first. In fact, it was counterintuitive, right ? We used to think, oh my god, this means we'll be less used, except we were more used we were able to penetrate the enterprise primarily because we did that interrupt work." If owning the infrastructure layer was the Microsoft plan and winning move, why is the same company now racing to build its own frontier model ?
Vikram M729,957 Aufrufe • vor 1 Monat
Sam Altman says telling an AI it is conscious will “psychologically hack” the people who are using it. Sam Altman: "I want to be clear, no one really knows. These models work at a level of complexity that I don't think. I can't tell you, as a scientific experiment, with 100% certainty, that would answer the hypothetical question to my satisfaction. And we can't even do that for humans." "If you tell a model, 'maybe you're conscious, you should do this, you should do that, you can be a conscientious objector here,' it's going to follow those instructions. And I think it is going to psychologically hack the people that are using it." "You can imagine all of these ways where things go very wrong, where even the people building AI get confused about that and start deferring too much to AI and what it seems to want. 'Okay, well, we can trust it, because it'll be good to us,' and things like that." "Real clarity here, and a commitment about what we as an industry believe about the role of people in the future, would be very helpful."
Vikram M31,763 Aufrufe • vor 3 Tagen
Jensen Huang was asked how China built so many world-class AI companies in ten years. He didn't credit the government. He didn't say they stole it. He pointed at the culture. In China, it's "family first, friends second, and company third." So the whole system is open source by default. That is the reframe. And it flips how the West explains China's speed. The conventional narrative is top-down central planning: a state pouring money into national champions, plus a lot of copied IP. Everyone repeats it. Jensen describes almost the opposite. China isn't one economy; it's provinces and cities with mayors competing against each other like startups. That's why there are dozens of EV companies and dozens of AI companies clawing to survive. What crawls out of that gauntlet is world-class. Now here's where it gets interesting. The second engine is social, not economic. Because friends and schoolmates outrank the employer, engineers share freely across rival companies. What are they protecting, when their brothers work down the street ? So the ecosystem defaults to open, and open source amplifies everyone at once. Half the world's AI researchers are Chinese, most still in China, wired into a culture that spreads every breakthrough at the speed of friendship. Competition sharpens the work. Openness distributes it. That combination compounds. He is not describing a command economy. He is describing the fastest-innovating culture on earth. The uncomfortable question for the West: can a system built to protect IP ever outrun one built to share it ?
Vikram M432,711 Aufrufe • vor 3 Monaten
Jensen Huang was asked what NVIDIA's single biggest moat is. The most valuable company in the world. He didn't point to the chips. He said a competitor could clone CUDA exactly and it wouldn't matter. "if somebody came up with a GUDA or TUDA it wouldn't make any difference at all." That is the reframe. And it changes what NVIDIA is actually defending. The conventional story is that NVIDIA wins on silicon. Faster GPU, better transistors, more FLOPS. Which means the day someone ships a faster chip, the game is over. Every competitor is racing on that assumption. Jensen is defending something else entirely. He said it wasn't three people who made CUDA win. It was 43,000 people and several million developers who bet their software on it. The moat isn't the hardware. It's the install base. Now here's where it gets interesting. Put yourself in a developer's seat. Target CUDA and you reach a few hundred million machines: every cloud, every computer maker, every industry, every country. And the platform gets roughly 10x better every six months, for free, while you wait. A rival could ship a chip that is genuinely faster and still lose, because no rational developer ports a mountain of software off a platform that already owns the install base and improves itself every two quarters. He is not selling the best chip. He is renting out the largest install base in computing. The open question is whether better silicon can ever beat that, or whether the only way past a moat like this is to make the whole category obsolete.
Vikram M367,768 Aufrufe • vor 3 Monaten
Satya Nadella just clarified the Microsoft open Ai deal and the terms what they signed for Azure. Here is exactly how he breaks down the partnership How it was Essentially think of OpenAI having a PaaS business and a SaaS business. The SaaS business is ChatGPT. Their PaaS business is their API. That API is Azure exclusive. The SaaS business, they can run it anywhere. If they want a partner and that partner wants to use a stateless API, then Azure is the place where they can get the stateless API. for even that they'll have to come to Azure. For any custom agreement like that. There are some few exceptions, the US government and so on, that we made, but other than that, they'd have to come to Azure. Open Ai can host there Chat Gpt consumer app wherever they want there wish but when it comes to developers or enterprise want to build on top of Open Ai models Microsoft owned the pipeline.
Vikram M102,664 Aufrufe • vor 2 Monaten
Satya Nadella was asked directly: in two years, will Microsoft have more engineers or fewer ? He didn't answer with a headcount. He answered with a job description that doesn't exist yet. In the 1980s, if someone had predicted 3.5 billion people would spend their days typing, the world would have laughed. Nobody needs 3.5 billion typists. Except that's exactly what happened and every one of them had a wage, a title, and a career built around it. Now here's where it gets interesting. The software developer of the future isn't writing code. They're managing 100 agents, 1,000 agents and doing something Nadella's team just named for the first time. "One of the new things that we are learning is what I'll call cognitive coverage." His point: when your entire codebase is written by agents, the human job becomes comprehending what was built. Auditing it. Understanding the decisions the agent made and why. That is not a task AI can replace because the AI is the thing being understood. So do the math on what that means. The workflow changed. The artifact changed. The input output format of software development changed. And the job changed with it not away, but upward. "That's the job of a software developer. In order to do that you've got to go to school. You've got to learn computer science and have cognitive coverage." Nadella is not saying jobs are safe. He's saying the jobs that survive are the ones AI cannot verify. And the unverifiable part of human work the meeting observations, the judgment calls, the things that leave no trace is exactly what no model can be trained on. I wonder why nobody in San Francisco is talking about that.
Vikram M96,897 Aufrufe • vor 4 Monaten
Dario Amodei was asked whether open source will eventually gut Anthropic's business. He didn't defend the moat. He didn't argue closed beats open. He said the whole question is a red herring. That is the reframe. And it flips how the industry keeps scoring this race. The conventional narrative is inherited from the last era of tech: open source wins because anyone can read the code, anyone improves it, contributions stack, and eventually the free thing catches the paid thing. Investors have a full lexicon for it. Commoditization. Which layer captures the value. Everyone repeats it. Amodei says the analogy breaks at the root. It's called open weights, not open source, for a reason: you can't see inside the model. So the thing that actually made open source powerful elsewhere, many people reading and additively improving shared code, never transfers. You just get a large file of numbers. Now here's where it gets interesting. The second engine isn't ideology. It's infrastructure. Free isn't free. Someone still has to host it. These are big models, and they're hard to run inference on. Someone has to make that fast. And the capabilities people assume only open weights unlock fine tuning, steering, inspecting activations labs are increasingly serving on their own clouds anyway. When DeepSeek shipped, he says he never asked whether it was open. He asked one thing: is it a good model, and is it better than us. That's the only axis he competes on. He even inverts the usual edge. Coming from outside that investor lexicon, he thinks knowing none of it lets him predict this better than the people fluent in it. He is not defending closed models. He is saying the scoreboard everyone is watching measures the wrong thing. The uncomfortable question if the free model still needs someone to run it, was the moat ever the weights, or always the machine underneath ?
Vikram M58,688 Aufrufe • vor 3 Monaten
Satya Nadella says building gigawatt scale data centers for just one AI customer or a single chip generation is a massive trap. He Breakdown We didn't want to just be a hoster for one company and have just a massive book of business with one customer. That's not a business, you should be vertically integrated with that company. I didn't want to get stuck with massive scale of one generation. I don't want to just go build out a whole number of gigawatts that are only for a one generation, one family. So I think the pacing matters, the fungibility and the location matters, the workload diversity matters, customer diversity matters and that's what we’re building towards. we're not in the business of just doing five contracts with five customers being their bare metal service. That's not a Microsoft business.
Vikram M45,440 Aufrufe • vor 2 Monaten
Demis Hassabis was asked if AI could ever simulate the origin of life the actual moment non-living chemistry becomes a living thing. He said I don't see why not, why AI couldn't help with that some kind of simulation. Again it's a bit of a search process through a combinatorial space. Here's like all the chemical soup that you start with, the primordial soup here's some initial conditions, can you generate something that looks like a cell ? And on the nearer dream, the full virtual cell, he was honest about how long he's wanted it: I've had this idea of wanting to do that for maybe more like 25 years. Maybe you could 100X speed up experiments by doing most of it in silico.
Vikram M36,027 Aufrufe • vor 3 Monaten
Sam Alman was asked to justify $1.4 trillion in infrastructure against roughly $20 billion in revenue. He didn't defend the number. He didn't promise the demand would arrive. He said the problem is us. "exponential growth is usually very hard for people." That is the reframe. And it flips what the bubble argument is actually about. The conventional narrative is a spending crisis: a company burning capital far ahead of revenue, chasing demand that may never show up, heading for the bust every cycle ends in. Everyone repeats it. Alman describes almost the opposite problem. Revenue roughly tracks the compute fleet. OpenAI tripled compute in a year and plans to triple it again. His blunt version: double the compute, and they'd be at double the revenue today. They have never once found compute they couldn't monetize. The constraint was never customers. It's silicon, and they have never had enough of it. Now here's where it gets interesting. The second engine isn't financial. It's arithmetic. He runs a rough thought experiment. A frontier AI company outputs maybe 10 trillion tokens a day. Eight billion people, call it 20,000 tokens each. Run those numbers and one company is already doing something like a sixteenth of humanity's total output. Then 10x that. Then 100x. And the thing he most wants to spend it on isn't chat. It's science. He says the small discoveries were supposed to start in 2026. They started in late 2025. Mathematicians are publicly saying 5.2 crossed a line for them. Tiny results, but qualitatively different from nothing, and once the curve lifts off the x-axis, this field knows how to climb it. He is not spending ahead of demand. He is saying he has never in his life met it. The uncomfortable question if none of us can intuit an exponential, how would you know from the inside whether $1.4 trillion is reckless or already late ?
Vikram M30,378 Aufrufe • vor 3 Monaten
Jensen Huang was asked how Elon built Colossus, a 200,000-GPU supercomputer in Memphis, in four months. He didn't credit the money. He didn't credit the scale. He credited subtraction. Question every step until only the necessary remains, and you can't take anything else out. That is the reframe. And it changes what "moving fast" is actually made of. The conventional story of a record build is addition. More people, more budget, more GPUs, more parallel workstreams. Speed as brute force. Jensen described the opposite instinct. Elon is a systems thinker who strips every step to its minimal form, then does it at system scale. "he is as minimalist as you could possibly imagine." Fast, because there is less to get wrong. Now here's where it gets interesting. The second half isn't strategy. It's location. Jensen said Elon is present at the point of action. If there's a problem, he goes there and says, show me the problem. Lex described watching him sit with an engineer to understand exactly how a cable gets plugged into a rack, so the process could be made less error-prone. Do the math on what that does to a supply chain. When the person making the call is standing at the rack, every supplier treats you as their top priority, because urgency at the top forces urgency everywhere below it. He is not out-spending the problem. He is out-subtracting and out-showing-up. The uncomfortable question for most orgs: how much of your timeline is real, and how much is just the distance between the decision-maker and the actual work ?
Vikram M26,149 Aufrufe • vor 3 Monaten
Satya Nadella was asked about the data center buildout. Wall Street wants him to defend the spend. He didn't defend the spend. He named the thing that kills it. Microsoft built more Azure capacity in the last fifteen months than it built in the first fifteen years. Fifteen years of cloud infrastructure. Compressed into fifteen months. Same team. That is not scaling. That is a phase change. So do the math on what has to be true for that phase change to keep going. You need chips. You need power. You need water. You need fiber. You need land. And you need one thing nobody prices into the model. You need permission. Nadella said it plainly. "Unless we as an industry are very principled about ensuring the benefits are felt in real ways at the community level, we won't have permission. It's as simple as that." Now here's where it gets interesting. Every hyperscaler earnings deck shows the same three variables. Capex committed. Contracted revenue. Backlog. What none of them show is the permit approval rate in the counties where the next campus needs to break ground. That number is not in the ten-K. That number decides whether the ten-K is real. Look at what has already happened. Ireland paused new data center connections. The Netherlands froze approvals in North Holland. Virginia's Loudoun County ran out of transmission capacity. Chile blocked a Google project over water. Every one of these is a community saying no after the capex was already committed. Nadella just said the loud part on stage. "I've always felt in human history, if you use a lot of energy and also create a lot of value for society, the story has been fantastic. If you don't do that, it's not been that great." That is not corporate responsibility talk. That is a CEO pricing a variable his competitors are ignoring. Because the industry consensus right now is that the buildout is gated by chips. Nvidia's supply. TSMC's fab timeline. Grid interconnect queues. Those are the numbers everyone models. The actual gate is a county board in Ohio, Arizona, or Wisconsin voting yes on the next permit. I wonder how many hyperscaler CFOs have that variable in their forecast.
Vikram M22,040 Aufrufe • vor 3 Monaten
The entire AI industry is racing to build the smartest model. Satya Nadella just admitted that is not where the money is. The model is not the product. The harness is. That is the exact line. And it changes what Microsoft is actually competing on. OpenAI, Anthropic, Google, xAI, Meta every frontier lab is pouring hundreds of billions into training compute, chasing the next capability jump. Each betting that raw model intelligence is the moat. Microsoft is doing the opposite. It is building the harness the orchestration layer that sits above the model, connecting it to tools, data, permissions, sub-agents, and enterprise workflows. And it is letting OpenAI, Anthropic, and MAI compete to plug into it. "You need the model. But the model is not the product. The harness is." So do the math on what a harness actually does. A raw model dropped into an enterprise answers questions. That is a chatbot. A harness turns that same model into an agent that reads the SharePoint, edits the ERP entry, pulls the GitHub PR, updates Salesforce, and files the Excel report with the right permissions, the right audit trail, and the right sub-agent for each sub-task. The model provides the intelligence. The harness converts intelligence into work. Now here's where it gets interesting. "Even the best model in the world will feel broken without a great harness. And an okay model with a great harness can feel like magic." If that is true, the enterprise buyer is not buying model quality. The enterprise buyer is buying the harness. Which means model quality becomes a commodity input over time, and harness quality becomes the sustainable moat. Compare that to the strategy the entire frontier lab industry is executing. Everyone else is chasing the numerator raw intelligence. Almost nobody at scale is racing to build the denominator the orchestration layer that determines whether that intelligence can actually be deployed profitably inside a real company. The frontier model race has a 10 to 20 percent chance of producing a single dominant winner. Nadella just told the industry he does not need to be that winner. If OpenAI wins, Microsoft wins. If Anthropic wins, Microsoft wins. If MAI wins, Microsoft wins. If someone Microsoft has never heard of trains a better model in 2027, Microsoft still wins. Because the compute they train on, the harness they get plugged into, the enterprise contracts they get delivered through, and the products they sit inside are all Microsoft. He is not building the best AI model. He is building the layer that the best AI model has to run on to make anyone money. I wonder which position looks more valuable in ten years.
Vikram M21,463 Aufrufe • vor 3 Monaten
Jensen Huang was asked whether AI's future is open source or proprietary the question the whole industry keeps fighting over. He didn't pick a side. He said something that quietly ends the fight. "models is a technology, not a product. Model is a technology, not a service" "For the vast majority of consumers, the horizontal layer, the general intelligence, I would really, really love not to go fine-tune my own" "all these industries, their domain expertise, their specialization has to be channeled, has to be captured in a way that they can control, and that it can only come from open models" "the open model industry we're contributing tremendously to it is near the frontier" Everyone else is arguing about which lab wins OpenAI or Anthropic, who has the smartest frontier model this quarter. Jensen is standing on stage saying that's the sideshow. The proprietary model is the convenience layer you rent when you don't care. The open model is the thing you own when the work is yours. Two different products for two different needs not two horses in one race. open models are the second-biggest in the room. Ahead of Anthropic. Behind only OpenAI.
Vikram M12,078 Aufrufe • vor 3 Monaten
Keine weiteren Inhalte verfügbar