Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Prismatic Arc Warlock Is lowkey so fun. This Build has -Near Unlimited Jolts -Amplified, Devour, Radiant, Blinding, Weaken, Unravel, Slow/Suppress -Comically fast Super & Prismatic meter charging -Enough orbs to feed a small village when in Prismatic #BuildoftheWeek Destiny 2 Team 𝙰𝚗𝚍𝚢 𝚂𝚊𝚕𝚒𝚜𝚋𝚞𝚛𝚢 Br1 dmg04

Sneak

388,224 subscribers

38,514 views • 2 years ago •via X (Twitter)

#BuildoftheWeek

Anya Rossi• Live Now

Private livecam show

9 Comments

Sneak2 years ago

build here btw

Ubertini2 years ago

What exactly are you using?

Sofia2 years ago

I keep wanting to run that build, but the melee requiring sprint -> slide -> melee screws me up so bad on m&k. I can't figure out good bindings to pull it off without misfires.

fiery tepig2 years ago

"I need more [orbs of] power" -devil from the devil may cry series

whirlydirly2 years ago

Crown of Tempest is way better for this build. And Inmost Light + Synthoceps is the final form.

Underplayed Creations2 years ago

Why not use spirit of vesper with spirit of inmost light? Or crown of tempests?

DominantCub722 years ago

As a hunter main, I have to say out of the 2 prismatic classes I've played so far, warlock is by far superior and way more fun

Dario2 years ago

People still playing this game? Crazy

Dougdot2 years ago

Tried it a bit, fun playstyle! I swapped vesper out for felwinters instead so I can get 30% weaken + suppression instead of arc blind. Swapped to Strand super and put on facet of defiance. Worked out much better in my favor with end game content.

Related Videos

Destiny 2 left a permanent mark on my life. The memories I’ve made with this game, both as a player and as someone lucky enough to work on it, will forever stay with me. Not just because of late nights running King’s Fall with friends, or grinding out the Crucible Glorious Seal in solo queue like a complete maniac, but because Destiny 2 challenged me, shaped me, and pushed me toward becoming the creative I am today. My journey into digital art and photography started back in 2007, when I began taking screenshots in Halo 3 and entering Bungie community art contests. The relationships I built during that time eventually opened the door to an opportunity with Bungie’s Gameplay Capture team in 2014, helping the team capture footage for a new game called Destiny. I had no idea then that a commendably short two-week contract would turn into an incredible 12-year journey with this franchise. I originally came to Bungie hoping to pursue environmental concept art, but along the way I had the opportunity to work on marketing art and quickly fell in love with it. Creating marketing art for a franchise like Destiny challenged me in so many different ways. It forced me to expand my technical, creative, and communication skill sets, constantly adapt, and keep learning new tools and workflows just to keep pace with the live-service beast that was Destiny 2. It means a lot to be able to look back and say with confidence that I had a visual impact on Destiny 2; but the most meaningful part of it all was getting to work alongside and learn from so many incredible artists at Bungie, and seeing firsthand just how much care, effort, and humanity it takes to make work like this possible. It’s hard to fully capture how much energy, skill, and collaboration goes into every visual part of a game like Destiny 2. That work is shaped by artists from different backgrounds, experiences, disciplines, and perspectives. Each of them, including me, left a small part of themselves in what they created. To me, that’s what makes a game like Destiny 2 feel truly meaningful and memorable. Especially now, in a world increasingly saturated with content and driven by instant output and gratification through AI, I keep coming back to the value of process. For me, and for so many of my peers, it was never only about arriving at the final image. It was about the journey it took to get there. The late nights. The iteration. The problem-solving. The trust. The shared pursuit of trying to make something special. To my peers, I’m deeply proud of what we built together, but even more grateful for how we built it. We challenged each other, inspired each other, and kept showing up for one another through every high and low, as Destiny 2 has had many. That kind of shared effort leaves a lasting mark. What we made mattered. What we gave mattered. And the impact of what we built together will stay with us, and this community, for a long time. Shoutout to the current and former members of Creative Studios, the VizD team, and the many Bungie developers and marketers who helped shape this chapter of my life. I’m especially grateful to the teammates who believed in me, encouraged me to embrace failure and new beginnings as essential parts of artistic growth, and pushed me to take on challenges even when they felt beyond my reach. You showed me that the strength of a team will always surpass that of any individual hero. As my work on #Destiny2 comes to a close and I look toward the future, I plan to spend the next few weeks sharing some of the pieces I had the chance to create or art direct that mean the most to me. For now, I’ll leave you all with a collage of some of my personal favorite pieces to work on across Destiny 2’s lifetime. These projects mean so much to me because many of them started as personal passion projects or late-night concept sketches, inspired by playing early builds of Destiny 2 and by the incredible work of our development team.

Destiny 2 left a permanent mark on my life. The memories I’ve made with this game, both as a player and as someone lucky enough to work on it, will forever stay with me. Not just because of late nights running King’s Fall with friends, or grinding out the Crucible Glorious Seal in solo queue like a complete maniac, but because Destiny 2 challenged me, shaped me, and pushed me toward becoming the creative I am today. My journey into digital art and photography started back in 2007, when I began taking screenshots in Halo 3 and entering Bungie community art contests. The relationships I built during that time eventually opened the door to an opportunity with Bungie’s Gameplay Capture team in 2014, helping the team capture footage for a new game called Destiny. I had no idea then that a commendably short two-week contract would turn into an incredible 12-year journey with this franchise. I originally came to Bungie hoping to pursue environmental concept art, but along the way I had the opportunity to work on marketing art and quickly fell in love with it. Creating marketing art for a franchise like Destiny challenged me in so many different ways. It forced me to expand my technical, creative, and communication skill sets, constantly adapt, and keep learning new tools and workflows just to keep pace with the live-service beast that was Destiny 2. It means a lot to be able to look back and say with confidence that I had a visual impact on Destiny 2; but the most meaningful part of it all was getting to work alongside and learn from so many incredible artists at Bungie, and seeing firsthand just how much care, effort, and humanity it takes to make work like this possible. It’s hard to fully capture how much energy, skill, and collaboration goes into every visual part of a game like Destiny 2. That work is shaped by artists from different backgrounds, experiences, disciplines, and perspectives. Each of them, including me, left a small part of themselves in what they created. To me, that’s what makes a game like Destiny 2 feel truly meaningful and memorable. Especially now, in a world increasingly saturated with content and driven by instant output and gratification through AI, I keep coming back to the value of process. For me, and for so many of my peers, it was never only about arriving at the final image. It was about the journey it took to get there. The late nights. The iteration. The problem-solving. The trust. The shared pursuit of trying to make something special. To my peers, I’m deeply proud of what we built together, but even more grateful for how we built it. We challenged each other, inspired each other, and kept showing up for one another through every high and low, as Destiny 2 has had many. That kind of shared effort leaves a lasting mark. What we made mattered. What we gave mattered. And the impact of what we built together will stay with us, and this community, for a long time. Shoutout to the current and former members of Creative Studios, the VizD team, and the many Bungie developers and marketers who helped shape this chapter of my life. I’m especially grateful to the teammates who believed in me, encouraged me to embrace failure and new beginnings as essential parts of artistic growth, and pushed me to take on challenges even when they felt beyond my reach. You showed me that the strength of a team will always surpass that of any individual hero. As my work on #Destiny2 comes to a close and I look toward the future, I plan to spend the next few weeks sharing some of the pieces I had the chance to create or art direct that mean the most to me. For now, I’ll leave you all with a collage of some of my personal favorite pieces to work on across Destiny 2’s lifetime. These projects mean so much to me because many of them started as personal passion projects or late-night concept sketches, inspired by playing early builds of Destiny 2 and by the incredible work of our development team.

Biwald

73,183 views • 2 months ago

I've been spending a lot of my time digging into the combat of Arknights: Endfield to figure out who I want to have the most fun investing in. So far... the Physical Team has been my favorite (disregarding Limited). I believe Physical teams really help teach a player how the combo & battle skill system work, while giving a decent bit of skill expression. I'm not finished with my true first impressions of the game, but here's a short clip of this Pogranichnik combo I'm working on. I've designed it so that I stack Vulnerability as fast as possible, but delay 2 inputs long enough to build a final SP for Pog. This lets me immediately convert the stacks, activate his combo skill, and if my ULT is active he can launch 2-3x soldiers between each hit. Speaking of his Ultimate, this dood is basically "Aftershock the character" for anyone familiar with Zenless Zone Zero. I didn't understand it at first but now that I'm getting the hang of it, he might be my favorite of the standard 6-stars. He's got sauce!

I've been spending a lot of my time digging into the combat of Arknights: Endfield to figure out who I want to have the most fun investing in. So far... the Physical Team has been my favorite (disregarding Limited). I believe Physical teams really help teach a player how the combo & battle skill system work, while giving a decent bit of skill expression. I'm not finished with my true first impressions of the game, but here's a short clip of this Pogranichnik combo I'm working on. I've designed it so that I stack Vulnerability as fast as possible, but delay 2 inputs long enough to build a final SP for Pog. This lets me immediately convert the stacks, activate his combo skill, and if my ULT is active he can launch 2-3x soldiers between each hit. Speaking of his Ultimate, this dood is basically "Aftershock the character" for anyone familiar with Zenless Zone Zero. I didn't understand it at first but now that I'm getting the hang of it, he might be my favorite of the standard 6-stars. He's got sauce!

Iyo

184,560 views • 6 months ago

"What you do is you prioritize the top few benchmarks and then you probably bullshit the rest." We talk a lot about how AI is making coding easier for non-technical folks, but don't hear much about how the most elite engineers are delegating their most technically complex work. Ankur Goyal (CEO Braintrust) and team has to deliver huge volumes of data at lightning fast speeds, which means they're always looking at ways to optimize their DB stack. The problem? Technical hypotheses are expensive and slow to test when your human team has to manually review, so you look at 1-2 things and then move on with your best guess. With AI, they're able to bring much more rigor to their most technical work. In this episode, he shows us: - how a coding agent can run a full matrix of benchmarks against ever open source column store; something no eng had time to manually do before - how to build evals the same way you'd write a PRD - how he uses his best engineer's "vibe check" as the gold standard, and scales it across the product as scoring functions This is an episode by engineers for engineers, where Ankur and I geek out on CEOs who code, brute force coding agents, and why you should be using AI to solve your hardest engineering problems. Watch now:

"What you do is you prioritize the top few benchmarks and then you probably bullshit the rest." We talk a lot about how AI is making coding easier for non-technical folks, but don't hear much about how the most elite engineers are delegating their most technically complex work. Ankur Goyal (CEO Braintrust) and team has to deliver huge volumes of data at lightning fast speeds, which means they're always looking at ways to optimize their DB stack. The problem? Technical hypotheses are expensive and slow to test when your human team has to manually review, so you look at 1-2 things and then move on with your best guess. With AI, they're able to bring much more rigor to their most technical work. In this episode, he shows us: - how a coding agent can run a full matrix of benchmarks against ever open source column store; something no eng had time to manually do before - how to build evals the same way you'd write a PRD - how he uses his best engineer's "vibe check" as the gold standard, and scales it across the product as scoring functions This is an episode by engineers for engineers, where Ankur and I geek out on CEOs who code, brute force coding agents, and why you should be using AI to solve your hardest engineering problems. Watch now:

claire vo 🖤

17,764 views • 2 months ago

👀 Life update: Today I’m announcing Rehearsals! Rehearsals watches every user and tells you why people pay, why they leave and what exactly to build next. So why now? • Big companies are getting smaller and small teams are doing more. • Multimodal models (especially Gemini’s) have reached a cost to performance inflection. • The cost of good code has plummeted and in abundance, focus is more important than ever. Focus is everything in a world when sonnet 4.5 and GPT5 codex can do most software engineering tasks. So what is Rehearsals? It’s three revolutionary products: • User Simulator: You can instantly test ideas and pricing on AI clones of your real customers. • Vibe Analytics: We spot what’s blocking growth and tell you what to build next. • Generative Backlog: UX mocks and tickets appear, already privatized by impact, ready for any AI engineer. We focused our craft on making this the most useful product for these three types of people at big and small companies: 👩‍💻 Founders: You need to make great decisions fast and with as little resources as possible. Rehearsals keeps your team small and scrappy. 🧑‍🎨 Marketers: You’re trying to stretch every marketing dollar to its maximum. Rehearsals connects the impossible dots from creatives to the confirmation page. 🥷 Product: You’re tired of working with more data science and user research teams to get a simple answer. Rehearsals gets you in-depth research and decisions in minutes for your next PRD. This is the platform I always wanted at Google and when founding companies in the past, and it’s so exciting bringing tomorrows technology to today. 💸 In collaboration with one of our investors (a16z speedrun 🧊), we’re giving 2 months of our pro plan for free (worth $3,000) only during SF and LA tech week #techweek Show your interest in the comments and I’ll send over the details! More updates coming tomorrow 😇

👀 Life update: Today I’m announcing Rehearsals! Rehearsals watches every user and tells you why people pay, why they leave and what exactly to build next. So why now? • Big companies are getting smaller and small teams are doing more. • Multimodal models (especially Gemini’s) have reached a cost to performance inflection. • The cost of good code has plummeted and in abundance, focus is more important than ever. Focus is everything in a world when sonnet 4.5 and GPT5 codex can do most software engineering tasks. So what is Rehearsals? It’s three revolutionary products: • User Simulator: You can instantly test ideas and pricing on AI clones of your real customers. • Vibe Analytics: We spot what’s blocking growth and tell you what to build next. • Generative Backlog: UX mocks and tickets appear, already privatized by impact, ready for any AI engineer. We focused our craft on making this the most useful product for these three types of people at big and small companies: 👩‍💻 Founders: You need to make great decisions fast and with as little resources as possible. Rehearsals keeps your team small and scrappy. 🧑‍🎨 Marketers: You’re trying to stretch every marketing dollar to its maximum. Rehearsals connects the impossible dots from creatives to the confirmation page. 🥷 Product: You’re tired of working with more data science and user research teams to get a simple answer. Rehearsals gets you in-depth research and decisions in minutes for your next PRD. This is the platform I always wanted at Google and when founding companies in the past, and it’s so exciting bringing tomorrows technology to today. 💸 In collaboration with one of our investors (a16z speedrun 🧊), we’re giving 2 months of our pro plan for free (worth $3,000) only during SF and LA tech week #techweek Show your interest in the comments and I’ll send over the details! More updates coming tomorrow 😇

Liam Bolling

10,442 views • 10 months ago

Q: How do you build a great company? In the clip below, Sam Altman walks through 9 things he has seen the best founders do: #1 Get to know your users really well “The best founders do customer support themselves. They go visit their users—in the case of Airbnb they go live with them. You want to get to know your users really really well.” #2 Have a short cycle time & understand compound growth “The cycle here is basically: talk to customer to understand pain point → build product to address that → get product in front of user → see what they do → repeat cycle. This cycle is how you iterate and improve. The law of compound growth being what it is: if you can get 2% better every iteration cycle, your iteration cycle is every four hours rather than every four weeks, and you compound that over the course of a few years, you’ll be in a very very different place. Make it one of your top goals to build one of the fastest iterating companies the world has ever seen.” #3 Make a long-term commitment “Most companies have a 2-3 year time horizon. But companies are almost always a 10 year project if they work. If you think about it that way from the very beginning, you will make very different and much better decisions. I think this is the only arbitrage opportunity left in the market. Almost no one makes a fairly long-term commitment to a new project. But if you do that, you will think in a different way, you will hire different people, and it will work very well.” #4 Stay lean until everything is working really well “In the early days, when you’re experimenting and zig zagging, you’re like a fast little speed boat and want to be able to turn the whole company on a dime. You can’t do that if you’re a big company—cash burn aside, which is another problem. The flexibility of the company basically decreases with the square of the number of employees, so you want to stay really small until you’re sure things are working. Once things are working, then you can get really big.” #5 Resist the urge to hire; especially resist the urge to hire mediocre people “Vinod Khosla has a saying that I love: ‘the team you build is the company you build.’ This is really true and I never appreciated how true this was for a long time. If you build a team of great people and you have a product that people love, you’ll have a 90%+ chance of success. Those are both really hard to do, and they’re independent variables. But don’t ignore the team component. The best CEOs I know spend huge amounts of their time recruiting and retaining good talent.” #6 Relentless execution “You have to keep going, and do things perfectly, and get all of the details right. You have to care too much about every experience that a customer has with your company.” #7 Startups are about not giving up “One of the very best companies in the last YC batch applied 7 times before they got in. This is just a version of what happens in startups all of the time: you get beat down, again, and again, and again. And that last time when you get pushed down and don’t think you have enough energy to get back up—that’s the time it actually works. This is what you sign up for if you’re going to start a startup.” #8 Fiduciary duty to take care of yourself “This is a 10 year marathon and you have a fiduciary duty to your shareholders to take care of yourself. Some people treat startups like an all-nighter: they don’t take care of their health, they don’t sleep, they don’t maintain their personal relationships. It is true that startups are a bad choice for work-life balance. But you have a duty to yourself, your team, and your investors to take care of yourself.” #9 Clear mission “You don’t have to figure this out on Day 1, but all of the most successful startups I’ve been fortunate enough to be a part of pretty quickly—in the first one to two years—figure out a really important mission. It’s this mission that gets people to join them. It drives the founders. It gets the media to write about them. And even if you start off building a project that’s just interesting to you and solves a problem in your life—which is how you should start—remember that you should have a clear mission at some point… That is what will convince people to come help you, and that is how you will build this idea into a huge company with a ton of people that really love your product.”

Q: How do you build a great company? In the clip below, Sam Altman walks through 9 things he has seen the best founders do: #1 Get to know your users really well “The best founders do customer support themselves. They go visit their users—in the case of Airbnb they go live with them. You want to get to know your users really really well.” #2 Have a short cycle time & understand compound growth “The cycle here is basically: talk to customer to understand pain point → build product to address that → get product in front of user → see what they do → repeat cycle. This cycle is how you iterate and improve. The law of compound growth being what it is: if you can get 2% better every iteration cycle, your iteration cycle is every four hours rather than every four weeks, and you compound that over the course of a few years, you’ll be in a very very different place. Make it one of your top goals to build one of the fastest iterating companies the world has ever seen.” #3 Make a long-term commitment “Most companies have a 2-3 year time horizon. But companies are almost always a 10 year project if they work. If you think about it that way from the very beginning, you will make very different and much better decisions. I think this is the only arbitrage opportunity left in the market. Almost no one makes a fairly long-term commitment to a new project. But if you do that, you will think in a different way, you will hire different people, and it will work very well.” #4 Stay lean until everything is working really well “In the early days, when you’re experimenting and zig zagging, you’re like a fast little speed boat and want to be able to turn the whole company on a dime. You can’t do that if you’re a big company—cash burn aside, which is another problem. The flexibility of the company basically decreases with the square of the number of employees, so you want to stay really small until you’re sure things are working. Once things are working, then you can get really big.” #5 Resist the urge to hire; especially resist the urge to hire mediocre people “Vinod Khosla has a saying that I love: ‘the team you build is the company you build.’ This is really true and I never appreciated how true this was for a long time. If you build a team of great people and you have a product that people love, you’ll have a 90%+ chance of success. Those are both really hard to do, and they’re independent variables. But don’t ignore the team component. The best CEOs I know spend huge amounts of their time recruiting and retaining good talent.” #6 Relentless execution “You have to keep going, and do things perfectly, and get all of the details right. You have to care too much about every experience that a customer has with your company.” #7 Startups are about not giving up “One of the very best companies in the last YC batch applied 7 times before they got in. This is just a version of what happens in startups all of the time: you get beat down, again, and again, and again. And that last time when you get pushed down and don’t think you have enough energy to get back up—that’s the time it actually works. This is what you sign up for if you’re going to start a startup.” #8 Fiduciary duty to take care of yourself “This is a 10 year marathon and you have a fiduciary duty to your shareholders to take care of yourself. Some people treat startups like an all-nighter: they don’t take care of their health, they don’t sleep, they don’t maintain their personal relationships. It is true that startups are a bad choice for work-life balance. But you have a duty to yourself, your team, and your investors to take care of yourself.” #9 Clear mission “You don’t have to figure this out on Day 1, but all of the most successful startups I’ve been fortunate enough to be a part of pretty quickly—in the first one to two years—figure out a really important mission. It’s this mission that gets people to join them. It drives the founders. It gets the media to write about them. And even if you start off building a project that’s just interesting to you and solves a problem in your life—which is how you should start—remember that you should have a clear mission at some point… That is what will convince people to come help you, and that is how you will build this idea into a huge company with a ton of people that really love your product.”

Michael McGuiness

500,098 views • 2 years ago

Q: How do you build a great company? In the clip below, Sam Altman walks through 9 things he has seen the best founders do: #1 Get to know your users really well “The best founders do customer support themselves. They go visit their users—in the case of Airbnb they go live with them. You want to get to know your users really really well.” #2 Have a short cycle time & understand compound growth “The cycle here is basically: talk to customer to understand pain point → build product to address that → get product in front of user → see what they do → repeat cycle. This cycle is how you iterate and improve. The law of compound growth being what it is: if you can get 2% better every iteration cycle, your iteration cycle is every four hours rather than every four weeks, and you compound that over the course of a few years, you’ll be in a very very different place. Make it one of your top goals to build one of the fastest iterating companies the world has ever seen.” #3 Make a long-term commitment “Most companies have a 2-3 year time horizon. But companies are almost always a 10 year project if they work. If you think about it that way from the very beginning, you will make very different and much better decisions. I think this is the only arbitrage opportunity left in the market. Almost no one makes a fairly long-term commitment to a new project. But if you do that, you will think in a different way, you will hire different people, and it will work very well.” #4 Stay lean until everything is working really well “In the early days, when you’re experimenting and zig zagging, you’re like a fast little speed boat and want to be able to turn the whole company on a dime. You can’t do that if you’re a big company—cash burn aside, which is another problem. The flexibility of the company basically decreases with the square of the number of employees, so you want to stay really small until you’re sure things are working. Once things are working, then you can get really big.” #5 Resist the urge to hire; especially resist the urge to hire mediocre people “Vinod Khosla has a saying that I love: ‘the team you build is the company you build.’ This is really true and I never appreciated how true this was for a long time. If you build a team of great people and you have a product that people love, you’ll have a 90%+ chance of success. Those are both really hard to do, and they’re independent variables. But don’t ignore the team component. The best CEOs I know spend huge amounts of their time recruiting and retaining good talent.” #6 Relentless execution “You have to keep going, and do things perfectly, and get all of the details right. You have to care too much about every experience that a customer has with your company.” #7 Startups are about not giving up “One of the very best companies in the last YC batch applied 7 times before they got in. This is just a version of what happens in startups all of the time: you get beat down, again, and again, and again. And that last time when you get pushed down and don’t think you have enough energy to get back up—that’s the time it actually works. This is what you sign up for if you’re going to start a startup.” #8 Fiduciary duty to take care of yourself “This is a 10-year marathon and you have a fiduciary duty to your shareholders to take care of yourself. Some people treat startups like an all-nighter: they don’t take care of their health, they don’t sleep, they don’t maintain their personal relationships. It is true that startups are a bad choice for work-life balance. But you have a duty to yourself, your team, and your investors to take care of yourself.” #9 Clear mission “You don’t have to figure this out on Day 1, but all of the most successful startups I’ve been fortunate enough to be a part of pretty quickly—in the first one to two years—figure out a really important mission. It’s this mission that gets people to join them. It drives the founders. It gets the media to write about them. And even if you start off building a project that’s just interesting to you and solves a problem in your life—which is how you should start—remember that you should have a clear mission at some point… That is what will convince people to come help you, and that is how you will build this idea into a huge company with a ton of people that really love your product.” Follow Startup Archive for more tactical startup advice!

Q: How do you build a great company? In the clip below, Sam Altman walks through 9 things he has seen the best founders do: #1 Get to know your users really well “The best founders do customer support themselves. They go visit their users—in the case of Airbnb they go live with them. You want to get to know your users really really well.” #2 Have a short cycle time & understand compound growth “The cycle here is basically: talk to customer to understand pain point → build product to address that → get product in front of user → see what they do → repeat cycle. This cycle is how you iterate and improve. The law of compound growth being what it is: if you can get 2% better every iteration cycle, your iteration cycle is every four hours rather than every four weeks, and you compound that over the course of a few years, you’ll be in a very very different place. Make it one of your top goals to build one of the fastest iterating companies the world has ever seen.” #3 Make a long-term commitment “Most companies have a 2-3 year time horizon. But companies are almost always a 10 year project if they work. If you think about it that way from the very beginning, you will make very different and much better decisions. I think this is the only arbitrage opportunity left in the market. Almost no one makes a fairly long-term commitment to a new project. But if you do that, you will think in a different way, you will hire different people, and it will work very well.” #4 Stay lean until everything is working really well “In the early days, when you’re experimenting and zig zagging, you’re like a fast little speed boat and want to be able to turn the whole company on a dime. You can’t do that if you’re a big company—cash burn aside, which is another problem. The flexibility of the company basically decreases with the square of the number of employees, so you want to stay really small until you’re sure things are working. Once things are working, then you can get really big.” #5 Resist the urge to hire; especially resist the urge to hire mediocre people “Vinod Khosla has a saying that I love: ‘the team you build is the company you build.’ This is really true and I never appreciated how true this was for a long time. If you build a team of great people and you have a product that people love, you’ll have a 90%+ chance of success. Those are both really hard to do, and they’re independent variables. But don’t ignore the team component. The best CEOs I know spend huge amounts of their time recruiting and retaining good talent.” #6 Relentless execution “You have to keep going, and do things perfectly, and get all of the details right. You have to care too much about every experience that a customer has with your company.” #7 Startups are about not giving up “One of the very best companies in the last YC batch applied 7 times before they got in. This is just a version of what happens in startups all of the time: you get beat down, again, and again, and again. And that last time when you get pushed down and don’t think you have enough energy to get back up—that’s the time it actually works. This is what you sign up for if you’re going to start a startup.” #8 Fiduciary duty to take care of yourself “This is a 10-year marathon and you have a fiduciary duty to your shareholders to take care of yourself. Some people treat startups like an all-nighter: they don’t take care of their health, they don’t sleep, they don’t maintain their personal relationships. It is true that startups are a bad choice for work-life balance. But you have a duty to yourself, your team, and your investors to take care of yourself.” #9 Clear mission “You don’t have to figure this out on Day 1, but all of the most successful startups I’ve been fortunate enough to be a part of pretty quickly—in the first one to two years—figure out a really important mission. It’s this mission that gets people to join them. It drives the founders. It gets the media to write about them. And even if you start off building a project that’s just interesting to you and solves a problem in your life—which is how you should start—remember that you should have a clear mission at some point… That is what will convince people to come help you, and that is how you will build this idea into a huge company with a ton of people that really love your product.” Follow Startup Archive for more tactical startup advice!

Startup Archive

407,717 views • 2 years ago

people always say, "why is your syndra build so weird"? (mandate + stormsurge + shadowflame + cryptbloom) it's so that I can go around 1shotting people like the lobby admin IN CHALLENGER while still having haste/speed for fights, without EVER having to rely on my teammates to make plays. with tier 3 sorcs buff next patch, my build will be even more OP. don't buy disgusting rocketbelt / blackfire for "more rotations" or "mobility" when you can get the job done in 1 rotation with so much mpen that you deal true damage. and if your enemies manage to survive, your 2nd rotation is still up super fast with mandate and they can't play the game. FULL BUILD ORDER: your first 2 items are ALWAYS imperial mandate and stormsurge, with sorcerer's shoes. The purchase order depends on matchup. can 1shot: go stormsurge first, with a faerie charm to proc JoaT can't 1shot: go mandate first for more E's. prioritize 3 amp tombs and faerie charm for mana / jack. 3rd item: always shadowflame. 4th item: cryptbloom default. You deal true damage to every1 in the game with this purchase, or VERY close to it. 4th item alternative: seeker's armguard -> cryptbloom, or upgrade to zhonya's if needed (vs zed,fizz,qiyana,etc.) 5th item: deathcap/zhonyas/banshees. Final Item Insight: if you bought zhonya's or banshees, chances are you don't have Deathcap in your build. Honestly, i make it this far in like 1% of games, because my build is EXTREMELY good at closing games out. But if this happens to you, just sell shadowflame or stormsurge for deathcap when you can afford it. I prefer to keep stormsurge if i need movespeed, otherwise if I have flash/ghost then I sell stormsurge and keep shadowflame. Alternatively, if you are GIGA rich, you can sell your boots for deathcap (you need summs or else u are too slow). Another possibility is selling mandate for Cosmic Drive for speed/durability, so your build could look like: DCAP, Shadowflame, Stormsurge, Cosmic Drive, Zhonyas, Cryptbloom (no boots) RUNES: In 90% of games, go Comet, Manaflow, Transcendence, Scorch, with Cashback, Jack of all Trades. Your first 2 items are cheap, getting full utilization of Cashback ASAP. In 10% of games, vs unkillable matchups (yasuo,ekko,galio,any tank), go First Strike, Cashback, Triple Tonic, Jack of All Trades, Manaflow, Transcendence. EXPERT RUNES: Swap out manaflow band for Axiom Arcanist. You need to be VERY good at mana management to take this greedy option. When you deal true damage, 12% increased R damage is A LOT. And I have noticed this rune singlehandedly get me kills and win me games I had no business winning where my team was 5-10k gold down, just because you can transcend Syndra's damage limits and 1v9. And no, I NEVER go oom after mandate is purchased since that's when your passive starts refunding a ton of mana. But before mandate, you might go oom if you use too many abilities without getting a Splinter stack out of it. Syndra is uniquely able to go this build because she has mana sustain, whereas every other mage doesn't. Abuse for free LP.

people always say, "why is your syndra build so weird"? (mandate + stormsurge + shadowflame + cryptbloom) it's so that I can go around 1shotting people like the lobby admin IN CHALLENGER while still having haste/speed for fights, without EVER having to rely on my teammates to make plays. with tier 3 sorcs buff next patch, my build will be even more OP. don't buy disgusting rocketbelt / blackfire for "more rotations" or "mobility" when you can get the job done in 1 rotation with so much mpen that you deal true damage. and if your enemies manage to survive, your 2nd rotation is still up super fast with mandate and they can't play the game. FULL BUILD ORDER: your first 2 items are ALWAYS imperial mandate and stormsurge, with sorcerer's shoes. The purchase order depends on matchup. can 1shot: go stormsurge first, with a faerie charm to proc JoaT can't 1shot: go mandate first for more E's. prioritize 3 amp tombs and faerie charm for mana / jack. 3rd item: always shadowflame. 4th item: cryptbloom default. You deal true damage to every1 in the game with this purchase, or VERY close to it. 4th item alternative: seeker's armguard -> cryptbloom, or upgrade to zhonya's if needed (vs zed,fizz,qiyana,etc.) 5th item: deathcap/zhonyas/banshees. Final Item Insight: if you bought zhonya's or banshees, chances are you don't have Deathcap in your build. Honestly, i make it this far in like 1% of games, because my build is EXTREMELY good at closing games out. But if this happens to you, just sell shadowflame or stormsurge for deathcap when you can afford it. I prefer to keep stormsurge if i need movespeed, otherwise if I have flash/ghost then I sell stormsurge and keep shadowflame. Alternatively, if you are GIGA rich, you can sell your boots for deathcap (you need summs or else u are too slow). Another possibility is selling mandate for Cosmic Drive for speed/durability, so your build could look like: DCAP, Shadowflame, Stormsurge, Cosmic Drive, Zhonyas, Cryptbloom (no boots) RUNES: In 90% of games, go Comet, Manaflow, Transcendence, Scorch, with Cashback, Jack of all Trades. Your first 2 items are cheap, getting full utilization of Cashback ASAP. In 10% of games, vs unkillable matchups (yasuo,ekko,galio,any tank), go First Strike, Cashback, Triple Tonic, Jack of All Trades, Manaflow, Transcendence. EXPERT RUNES: Swap out manaflow band for Axiom Arcanist. You need to be VERY good at mana management to take this greedy option. When you deal true damage, 12% increased R damage is A LOT. And I have noticed this rune singlehandedly get me kills and win me games I had no business winning where my team was 5-10k gold down, just because you can transcend Syndra's damage limits and 1v9. And no, I NEVER go oom after mandate is purchased since that's when your passive starts refunding a ton of mana. But before mandate, you might go oom if you use too many abilities without getting a Splinter stack out of it. Syndra is uniquely able to go this build because she has mana sustain, whereas every other mage doesn't. Abuse for free LP.

Strompest

435,202 views • 13 days ago

Today Meesho became a public company. Congratulations to Vidit Vidit Aatrey, Sanjeev Sanjeev Barnwal and the entire Meesho team on building one of the most mission-driven companies India has produced in the last decade. We at Elevation Capital Elevation Capital feel privileged to have been partners since 2017. Back then, it took us 1.5 years and multiple meetings before we finally partnered, and watching what's unfolded since then has been nothing short of remarkable. Last week, Vidit and I went back to Cafe Noir @ UB City where, in June 2017, we had decided to partner. Same place, same table. We hadn’t returned together in eight years. And sitting there, it felt like time folded in on itself. That day in 2017, Vidit cancelled another investor meeting to meet me. I’ve always remembered that as a small but decisive moment. What I learned only last week was that he thought the other fund was more likely to convert. And still, he chose to meet me. I can’t thank him enough for that decision - because it led to one of the most defining partnerships of my career and one of the most fulfilling relationships of my life! Sitting there, we ended up reminiscing about the whole arc: the early days with very little runway; the brutal months in 2016-17 that forged the culture; the courage to pivot from reseller-first to consumer-first; and the conviction to go zero commission when many people thought it was impossible. From those early conversations to today: 200 million+ customers, 500k+ sellers, and 100k+ livelihoods created through Valmo. Meesho has truly democratized e-commerce in India. And beyond the numbers, what has stood out for me the most is the character of the journey: deep customer obsession, the courage to make non-consensus decisions, and stellar execution to back them - again and again. Watching Vidit and Sanjeev has been a masterclass. Vidit's rare combination of super long term thinking and operational depth, seeing clearly where the world is heading, and having the conviction to act; Sanjeev’s quiet excellence and the stellar engineering team and culture he’s built - one that so many teams in India now look up to. And through it all, both staying humble and grounded. As they remind us, this is just chapter one. The e-commerce market in India is still in its early days and the mission is just getting started! I couldn't be more excited about what comes next. Congratulations once again to the exceptional team that has delivered on super auditious goals time and time again - Dhiresh, Roopa, Debdoot, Milan, Prasanna, Megha, Siddharth, Ashish, Sourabh, Steffie, Harshit, Prateek, Kirti, Jatin, UK. And thank you for the partnership. It has been a privilege! Penned down my reflections on the journey here:

Today Meesho became a public company. Congratulations to Vidit Vidit Aatrey, Sanjeev Sanjeev Barnwal and the entire Meesho team on building one of the most mission-driven companies India has produced in the last decade. We at Elevation Capital Elevation Capital feel privileged to have been partners since 2017. Back then, it took us 1.5 years and multiple meetings before we finally partnered, and watching what's unfolded since then has been nothing short of remarkable. Last week, Vidit and I went back to Cafe Noir @ UB City where, in June 2017, we had decided to partner. Same place, same table. We hadn’t returned together in eight years. And sitting there, it felt like time folded in on itself. That day in 2017, Vidit cancelled another investor meeting to meet me. I’ve always remembered that as a small but decisive moment. What I learned only last week was that he thought the other fund was more likely to convert. And still, he chose to meet me. I can’t thank him enough for that decision - because it led to one of the most defining partnerships of my career and one of the most fulfilling relationships of my life! Sitting there, we ended up reminiscing about the whole arc: the early days with very little runway; the brutal months in 2016-17 that forged the culture; the courage to pivot from reseller-first to consumer-first; and the conviction to go zero commission when many people thought it was impossible. From those early conversations to today: 200 million+ customers, 500k+ sellers, and 100k+ livelihoods created through Valmo. Meesho has truly democratized e-commerce in India. And beyond the numbers, what has stood out for me the most is the character of the journey: deep customer obsession, the courage to make non-consensus decisions, and stellar execution to back them - again and again. Watching Vidit and Sanjeev has been a masterclass. Vidit's rare combination of super long term thinking and operational depth, seeing clearly where the world is heading, and having the conviction to act; Sanjeev’s quiet excellence and the stellar engineering team and culture he’s built - one that so many teams in India now look up to. And through it all, both staying humble and grounded. As they remind us, this is just chapter one. The e-commerce market in India is still in its early days and the mission is just getting started! I couldn't be more excited about what comes next. Congratulations once again to the exceptional team that has delivered on super auditious goals time and time again - Dhiresh, Roopa, Debdoot, Milan, Prasanna, Megha, Siddharth, Ashish, Sourabh, Steffie, Harshit, Prateek, Kirti, Jatin, UK. And thank you for the partnership. It has been a privilege! Penned down my reflections on the journey here:

Mukul Arora

38,917 views • 8 months ago

A surveillance regime is being assembled in front of our very eyes. Most of you do not understand the tyrannical nightmare directly ahead. People still talk like this is a future problem. It's not. The battle against dystopia is now. 1984 is here. It arrives as "efficiency," procurement, integrations, device adoption, and a slow widening of what the state can do to you without asking, without noticing you, without needing you to consent. The attached video shows a federal agent in tactical gear wearing Meta Ray-Ban smart glasses on his face while carrying out state work. A camera at eye level. A microphone. A network connection. Constant recording that becomes a file, then a feed, then a searchable object that can be attached to a case, cross-referenced, stored, shared, and re-used. Raw footage is noise until it can be fused, searched, and operationalized. That is where AI enters. ICE is paying Palantir to build "ImmigrationOS," described as a platform for near real-time visibility into enforcement workflows, including tracking people and helping decide who gets targeted. Immigration is a perfect testing ground because the public has been trained to tolerate exceptional measures when the target group is already politically disposable. The capability does not stay there. It never stays there. The nightmare is the stack itself. Capture at the edge, aggregation in the middle, AI scoring and targeting at the top. Doorbell cameras feed the state. License plate readers feed the state. Data brokers feed the state. Local sharing agreements feed the state. Wearables feed the state. AI compresses the labor cost of suspicion, so the state can watch more people, more often, with less human effort. That is the mechanical shift most people still do not understand. AI does not need to be sentient to be oppressive. It needs to be cheap enough to scale the state’s attention and fast enough to outrun your ability to contest what it thinks it knows. Let me paint you the picture: an agent walks up the block wearing smart glasses, and your face, your voice, your license plate, and the people you are standing with get captured immediately. That clip hits an internal system where AI transcribes, tags, and links it to whatever identifiers already exist, your name, your address, your contacts, your past crossings, your employer, your car, your social graph. The platform fuses that with location trails and camera networks, then assigns a "priority" score that quietly moves you up a queue nobody outside the system can see. A caseworker opens a dashboard and clicks through prebuilt options that generate a task list, knock, detain, transfer, pressure, repeat, with the paperwork already half-written by AI. This is tyranny by procedure. The real chokehold is anticipatory control. Once people know they can be indexed, scored, and surfaced for enforcement, they start trimming their speech, their associations, their routes, their friends. Dissent becomes a risk factor, organizing becomes exposure, and the state does not need to ban protest when it can make participation costly enough to thin the crowd. It inevitably expands to include anyone who interrupts the smooth operation of power, and this kind of surveillance gives power the ability to punish quietly, repeatedly, and selectively, with plausible deniability stapled to every click. This is the formula for tyranny. The state gains capability, then finds incentives to justify using it. Agencies protect budgets by "demonstrating output." Contractors protect revenue by "delivering results." Politicians protect narratives by demanding visible "enforcement." Restraint gets treated as "inefficiency." Efficiency gets treated as "virtue." Your rights get treated as "friction." Any system that allows institutions to assemble your biography without your consent, then act on it without due process, is an engine of domination. Some people tolerate it because they think they will never be the target. That belief is childish. Power does not remain polite. It expands to fill the permissions you give it. A surveillance regime always needs new enemies to justify itself, because a machine built for pursuit must pursue. The regime being assembled is not subtle, it is simply normalized. It runs on boredom, exhaustion, and the assumption that someone else will handle it. It's all being wired now, and once it is wired, it will not ask your permission to be used.

A surveillance regime is being assembled in front of our very eyes. Most of you do not understand the tyrannical nightmare directly ahead. People still talk like this is a future problem. It's not. The battle against dystopia is now. 1984 is here. It arrives as "efficiency," procurement, integrations, device adoption, and a slow widening of what the state can do to you without asking, without noticing you, without needing you to consent. The attached video shows a federal agent in tactical gear wearing Meta Ray-Ban smart glasses on his face while carrying out state work. A camera at eye level. A microphone. A network connection. Constant recording that becomes a file, then a feed, then a searchable object that can be attached to a case, cross-referenced, stored, shared, and re-used. Raw footage is noise until it can be fused, searched, and operationalized. That is where AI enters. ICE is paying Palantir to build "ImmigrationOS," described as a platform for near real-time visibility into enforcement workflows, including tracking people and helping decide who gets targeted. Immigration is a perfect testing ground because the public has been trained to tolerate exceptional measures when the target group is already politically disposable. The capability does not stay there. It never stays there. The nightmare is the stack itself. Capture at the edge, aggregation in the middle, AI scoring and targeting at the top. Doorbell cameras feed the state. License plate readers feed the state. Data brokers feed the state. Local sharing agreements feed the state. Wearables feed the state. AI compresses the labor cost of suspicion, so the state can watch more people, more often, with less human effort. That is the mechanical shift most people still do not understand. AI does not need to be sentient to be oppressive. It needs to be cheap enough to scale the state’s attention and fast enough to outrun your ability to contest what it thinks it knows. Let me paint you the picture: an agent walks up the block wearing smart glasses, and your face, your voice, your license plate, and the people you are standing with get captured immediately. That clip hits an internal system where AI transcribes, tags, and links it to whatever identifiers already exist, your name, your address, your contacts, your past crossings, your employer, your car, your social graph. The platform fuses that with location trails and camera networks, then assigns a "priority" score that quietly moves you up a queue nobody outside the system can see. A caseworker opens a dashboard and clicks through prebuilt options that generate a task list, knock, detain, transfer, pressure, repeat, with the paperwork already half-written by AI. This is tyranny by procedure. The real chokehold is anticipatory control. Once people know they can be indexed, scored, and surfaced for enforcement, they start trimming their speech, their associations, their routes, their friends. Dissent becomes a risk factor, organizing becomes exposure, and the state does not need to ban protest when it can make participation costly enough to thin the crowd. It inevitably expands to include anyone who interrupts the smooth operation of power, and this kind of surveillance gives power the ability to punish quietly, repeatedly, and selectively, with plausible deniability stapled to every click. This is the formula for tyranny. The state gains capability, then finds incentives to justify using it. Agencies protect budgets by "demonstrating output." Contractors protect revenue by "delivering results." Politicians protect narratives by demanding visible "enforcement." Restraint gets treated as "inefficiency." Efficiency gets treated as "virtue." Your rights get treated as "friction." Any system that allows institutions to assemble your biography without your consent, then act on it without due process, is an engine of domination. Some people tolerate it because they think they will never be the target. That belief is childish. Power does not remain polite. It expands to fill the permissions you give it. A surveillance regime always needs new enemies to justify itself, because a machine built for pursuit must pursue. The regime being assembled is not subtle, it is simply normalized. It runs on boredom, exhaustion, and the assumption that someone else will handle it. It's all being wired now, and once it is wired, it will not ask your permission to be used.

Dylan Allman

327,629 views • 6 months ago

This Chinese developer linked two $2,999 NVIDIA DGX Sparks into one box and runs the full Qwen3-235B at home, after dropping his $1,999-a-month cloud bill to zero. He wired 2 small boxes into a single computer, split a giant 235-billion-parameter model in half between them, and serves it across his own network at about 10 tokens a second, with no internet, no cloud, right there on the desk. No data center, no thousand-dollar graphics cards, no monthly cloud bill. Just him, 2 gold boxes the size of a sandwich, one cable between them, and 1 power strip. And here is the whole payoff. He used to pay the cloud $1,999 a month for the same model, and the meter ticked on every request. Now he paid $5,998 once for 2 boxes, they covered their cost in 3 months, and after that he sends as many requests as he wants for free, only electricity. The two Sparks talk over one fast cable, each holds 128GB of memory, and together they carry the whole model, about 73GB loaded per box, with the chip inside pinned near the limit at 96%. Both boxes work as one and keep trading data over the cable, with no cloud in the loop and no single word leaking out. The ready model sits on one local address, and any app on his network calls it as easily as ChatGPT. And here is how he described, in plain words, what this pair of boxes does: "this is a pair of boxes that holds the huge Qwen3-235B model and serves it to one network. the model is split in half, and each box owns its half. parts: // Box 1 (holds the first half of the model and starts the answer fast, the first word appears in under a second) // Box 2 (holds the second half and writes out the rest, about 10 tokens a second) // Cable (connects the 2 boxes and moves data between them on every step, with no lag) // Address (one local address where any app sends its request, like to a cloud model) // Test (a script that runs big prompts through and measures speed and delays) // Monitor (checks temperature, power draw, and load on both boxes every 2 seconds). the model never goes to the cloud. he only steps in when a box runs hotter than 80 degrees or the cable between them starts dropping data." So the system knows exactly what it is, what it is for, and where its limits are. It knows it has to hold the whole huge model across 2 boxes on its own. It knows it has to answer every request locally, with no meter, no limits, and no internet. It knows the human is only needed when a box overheats or the link between them stalls. → The setup runs around the clock on 2 boxes, each pulling under 60 watts → However many requests he sends, the monthly bill is $0, only electricity → The first box starts the answer in under a second → The second writes text at about 10 tokens a second → One request at a time: 838 tokens in 85 seconds, first word in 0.8s → Two requests at once: 697 tokens in 108 seconds, first word in 0.7s → Both boxes sit at 96% load and warm up to 76-78 degrees And only when a chip in a box runs hotter than 80 degrees or the cable between the 2 Sparks drops data does the system call the owner. And when he himself is out on a run or in a coffee shop, he still reaches his own model at home from his phone: sends a big prompt to the local Qwen3-235B, gets the full answer back in under a minute and a half, with no token meter ticking and no limit to hit. Here is what the test shows on his screen during one of the night runs: "one request at a time: 838 tokens in 84.9 seconds, first word in 0.8s, then 0.1s per token." "two requests at once: 697 tokens in 107.6 seconds, first word in 0.7s, then 0.15s per token." "Box 1: chip at 96% load, 76 degrees, 56 watts, 73GB used in memory." "Box 2: chip at 96% load, 78 degrees, 56 watts, the Qwen3-235B model fully loaded." And while everyone around is paying for AI by the month and bumping into limits, his top-tier model just sits on the desk and works as much as he wants: his own little power plant instead of a forever meter. He has no server rack of his own and no cloud account behind it. Just 2 DGX Spark boxes on a desk, one model split in half between them, one local address, and a folder of prompts next to it. Out of everything I have seen this year, this is the cleanest way to stop paying for AI: $5,998 of hardware on the desk once, $0 a month to the cloud, unlimited forever, and between them 2 gold boxes, 1 cable, and the full Qwen3-235B answering at home with no internet.

This Chinese developer linked two $2,999 NVIDIA DGX Sparks into one box and runs the full Qwen3-235B at home, after dropping his $1,999-a-month cloud bill to zero. He wired 2 small boxes into a single computer, split a giant 235-billion-parameter model in half between them, and serves it across his own network at about 10 tokens a second, with no internet, no cloud, right there on the desk. No data center, no thousand-dollar graphics cards, no monthly cloud bill. Just him, 2 gold boxes the size of a sandwich, one cable between them, and 1 power strip. And here is the whole payoff. He used to pay the cloud $1,999 a month for the same model, and the meter ticked on every request. Now he paid $5,998 once for 2 boxes, they covered their cost in 3 months, and after that he sends as many requests as he wants for free, only electricity. The two Sparks talk over one fast cable, each holds 128GB of memory, and together they carry the whole model, about 73GB loaded per box, with the chip inside pinned near the limit at 96%. Both boxes work as one and keep trading data over the cable, with no cloud in the loop and no single word leaking out. The ready model sits on one local address, and any app on his network calls it as easily as ChatGPT. And here is how he described, in plain words, what this pair of boxes does: "this is a pair of boxes that holds the huge Qwen3-235B model and serves it to one network. the model is split in half, and each box owns its half. parts: // Box 1 (holds the first half of the model and starts the answer fast, the first word appears in under a second) // Box 2 (holds the second half and writes out the rest, about 10 tokens a second) // Cable (connects the 2 boxes and moves data between them on every step, with no lag) // Address (one local address where any app sends its request, like to a cloud model) // Test (a script that runs big prompts through and measures speed and delays) // Monitor (checks temperature, power draw, and load on both boxes every 2 seconds). the model never goes to the cloud. he only steps in when a box runs hotter than 80 degrees or the cable between them starts dropping data." So the system knows exactly what it is, what it is for, and where its limits are. It knows it has to hold the whole huge model across 2 boxes on its own. It knows it has to answer every request locally, with no meter, no limits, and no internet. It knows the human is only needed when a box overheats or the link between them stalls. → The setup runs around the clock on 2 boxes, each pulling under 60 watts → However many requests he sends, the monthly bill is $0, only electricity → The first box starts the answer in under a second → The second writes text at about 10 tokens a second → One request at a time: 838 tokens in 85 seconds, first word in 0.8s → Two requests at once: 697 tokens in 108 seconds, first word in 0.7s → Both boxes sit at 96% load and warm up to 76-78 degrees And only when a chip in a box runs hotter than 80 degrees or the cable between the 2 Sparks drops data does the system call the owner. And when he himself is out on a run or in a coffee shop, he still reaches his own model at home from his phone: sends a big prompt to the local Qwen3-235B, gets the full answer back in under a minute and a half, with no token meter ticking and no limit to hit. Here is what the test shows on his screen during one of the night runs: "one request at a time: 838 tokens in 84.9 seconds, first word in 0.8s, then 0.1s per token." "two requests at once: 697 tokens in 107.6 seconds, first word in 0.7s, then 0.15s per token." "Box 1: chip at 96% load, 76 degrees, 56 watts, 73GB used in memory." "Box 2: chip at 96% load, 78 degrees, 56 watts, the Qwen3-235B model fully loaded." And while everyone around is paying for AI by the month and bumping into limits, his top-tier model just sits on the desk and works as much as he wants: his own little power plant instead of a forever meter. He has no server rack of his own and no cloud account behind it. Just 2 DGX Spark boxes on a desk, one model split in half between them, one local address, and a folder of prompts next to it. Out of everything I have seen this year, this is the cleanest way to stop paying for AI: $5,998 of hardware on the desk once, $0 a month to the cloud, unlimited forever, and between them 2 gold boxes, 1 cable, and the full Qwen3-235B answering at home with no internet.

Blaze

93,871 views • 2 months ago

Check out one of our Alpha Testers Mobile Demo showcasing the ATM’s AI Driven capabilities to automatically detect tokens, make purchases on-chain on your behalf, and selling off for the profit level that the user sets for their strategy. The team has been building the ATMs Brain for almost a year now integrating a WIDE variety of custom built indicators built into the ATMs Brain (ie: Bubblemaps, paid Dex updates, paid trending, Alpha wallets, etc etc.. there are countless indicators built into the AI component of how it works). The @ATMdottrade will be versatile. If you are advanced enough and want to build out your own custom automated strategies you will be able to. There will also be preset strategies for the more beginner users. But what’s even more powerful on how we can really build a large market share in the trading platform space is how the advanced trader can use this as an all in one platform for alpha signals and lightning fast manual trades as well. DEGEN MODE IS COMING. If you signed up for BETA We have announced all of our acceptances into various phases. BETA is coming very soon. You can check here if you applied! We had over 1,800 applications for our BETA program: Right now the platform is in pure alpha and everyday we continue to enhance and add integrations so we can truly build the ultimate trading platform for every trader on Solana. Also after seeing the revenue generating from just a small amount of users, we are starting to see how valuable the $MARS ecosystem assets will become. Photon, BullX, Trojan, they cannot compete when it comes to this component. We are giving a Lion’s Share of the revenue back to our asset holders and believe we will be first to market to actually deliver a serious stream of passive income to our holders (50% to $MARS token Holders, 20% to Martians NFT (Powered by $MARS) Holders, 10% Community Treasury/DAO, 20% Back to Team, Marketing, Server costs, etc.) Think Bigger, we do. $MARS

Check out one of our Alpha Testers Mobile Demo showcasing the ATM’s AI Driven capabilities to automatically detect tokens, make purchases on-chain on your behalf, and selling off for the profit level that the user sets for their strategy. The team has been building the ATMs Brain for almost a year now integrating a WIDE variety of custom built indicators built into the ATMs Brain (ie: Bubblemaps, paid Dex updates, paid trending, Alpha wallets, etc etc.. there are countless indicators built into the AI component of how it works). The @ATMdottrade will be versatile. If you are advanced enough and want to build out your own custom automated strategies you will be able to. There will also be preset strategies for the more beginner users. But what’s even more powerful on how we can really build a large market share in the trading platform space is how the advanced trader can use this as an all in one platform for alpha signals and lightning fast manual trades as well. DEGEN MODE IS COMING. If you signed up for BETA We have announced all of our acceptances into various phases. BETA is coming very soon. You can check here if you applied! We had over 1,800 applications for our BETA program: Right now the platform is in pure alpha and everyday we continue to enhance and add integrations so we can truly build the ultimate trading platform for every trader on Solana. Also after seeing the revenue generating from just a small amount of users, we are starting to see how valuable the $MARS ecosystem assets will become. Photon, BullX, Trojan, they cannot compete when it comes to this component. We are giving a Lion’s Share of the revenue back to our asset holders and believe we will be first to market to actually deliver a serious stream of passive income to our holders (50% to $MARS token Holders, 20% to Martians NFT (Powered by $MARS) Holders, 10% Community Treasury/DAO, 20% Back to Team, Marketing, Server costs, etc.) Think Bigger, we do. $MARS

trenches.com

77,182 views • 1 year ago

After 2 years in “stealth mode”, we’re finally sharing about Payam Music, starting with my role: I met Payam when he was running a small local piano school, and I saw its transformative impact on students, starting with my own son, Darius. The Payam Method works wonders, as we hear over and over from parents and in online reviews. It has just the right mix of elements to make students fall in LOVE with playing piano and writing their own music. Once I saw firsthand how much better learning can be with the Payam Method, I felt compelled to help spread it, leveraging my experience scaling to 100 million students. I joined as CEO two years ago, and we began hiring a team and recruiting support to help us scale. With the recent endorsement of world-famous composer Hans Zimmer, Payam Music is available in cities around the US and expanding rapidly. Our schools teach 1-on-1 lessons, in person and even online. We have limited spots, so if you or your child want to learn piano, sign up now! And if we don’t have a school near you, join our wait list, we’re growing fast. Fall in Love with Piano P.S. This week we’re announcing the amazing business leaders who are investors in Payam Music, to help us bring the joy of learning to more students everywhere, and also to provide children an alternative to screentime. If you’re worried about kids’ obsession with screens and social media, the solution is to give them a new obsession: piano. ❤️🎹❤️

After 2 years in “stealth mode”, we’re finally sharing about Payam Music, starting with my role: I met Payam when he was running a small local piano school, and I saw its transformative impact on students, starting with my own son, Darius. The Payam Method works wonders, as we hear over and over from parents and in online reviews. It has just the right mix of elements to make students fall in LOVE with playing piano and writing their own music. Once I saw firsthand how much better learning can be with the Payam Method, I felt compelled to help spread it, leveraging my experience scaling to 100 million students. I joined as CEO two years ago, and we began hiring a team and recruiting support to help us scale. With the recent endorsement of world-famous composer Hans Zimmer, Payam Music is available in cities around the US and expanding rapidly. Our schools teach 1-on-1 lessons, in person and even online. We have limited spots, so if you or your child want to learn piano, sign up now! And if we don’t have a school near you, join our wait list, we’re growing fast. Fall in Love with Piano P.S. This week we’re announcing the amazing business leaders who are investors in Payam Music, to help us bring the joy of learning to more students everywhere, and also to provide children an alternative to screentime. If you’re worried about kids’ obsession with screens and social media, the solution is to give them a new obsession: piano. ❤️🎹❤️

Hadi Partovi

57,930 views • 2 months ago

📢 $BVM MARCH UPDATES 📢 GM. In Feb, we added these products to the BVM ecosystem: ✅BVM Staking, ✅Bitcoin Stamps, ✅SRC-20 Bridge, ✅Avail, ✅Jackal, ✅Arweave & ✅ IPFS. We're excited to share with the community what we’re shipping next in March. 1. FOUR MORE BITCOIN L2 BLOCKCHAINS BVM’s mission is to scale Bitcoin by powering thousands of Bitcoin L2s. In March, we’ll work with 4 different builders to help them launch their own Bitcoin L2 blockchains and ultimately scale Bitcoin with unlimited throughput and rich user cases. ⛓️ Swamps: DeFi Bitcoin L2 for SRC-20 @swamps_src20 is launching the first-ever Stamps-based Bitcoin L2 blockchain with BVM, designed for SRC-20 DEX. We’re super excited to be working with @0xnovae5Db RMZ 🟨 and the Swamps team. ⛓️ Naka: DeFi Bitcoin L2 for BRC-20/BRC-404 Naka Chain is an Ordinals-based Bitcoin L2 blockchain designed for BRC-20. They just added support for BRC-404. We're excited to work with DeBit | NakaChain.xyz and the Naka team. ⛓️ Satoshi Sync: Bitcoin L2 for Inscription Markets SatoshiSync is building the first chain-agnostic Inscription Protocol with BVM, enabling everyone to launch their own Inscription Market. It’s been a pleasure working with @okubalinska and the Satoshi Sync team. ⛓️ ███████: Bitcoin L2 for ████ This is a big one. We'll share it when it's ready. Super excited to be working on this with ███████ and the ███████ team. Note: If you're thinking about launching your own Bitcoin L2, please DM us. Would love to help! 2. BITCOIN MODULE STORE BVM is a modular Bitcoin L2 infrastructure. The more building blocks, the more developers can build on Bitcoin. This month, we’ll be adding these new modules to the Bitcoin Modular Store. 🧱 Filecoin Filecoin is one of the best decentralized storage networks. Developers can build their Bitcoin L2 and choose Filecoin to store their data. 🧱 Syscoin zkDA Syscoin is a Bitcoin-backed Data Availability layer. It is secured by Bitcoin’s own PoW plus Syscoin’s finality. We think it's a super interesting DA solution. 🧱 BVM-ETH Bridge Ethereum bridge will enable us to flow assets between BVM and Ethereum. You can trade $BVM on Uniswap or bridge ETH to various Bitcoin L2 blockchains powered by BVM. Note: please do make suggestions on which modules we should integrate next! 3. BRC-404 Starting out as a fun weekend experiment, BRC-404 shows us that it could become a promising protocol. $3.5M trading volume so far! 🤯🤯🤯 🎨 Fractionalize Ordinals with BRC-404 The idea is to take an existing Ordinal Inscription and turn it into BRC-404. Then, Ordinals would be much more accessible to many smaller buyers, and at the same time, the collectors would make a lot more on trading fees. That's really the beauty of semi-fungible tokens. You can move between the fungible worlds and the non-fungible worlds. 🔒 ███████ for BRC-404 This product is exciting. We can't quite talk about it yet. But we'll do it when it's ready later this month. 4. COMMUNITY While our product team is busy crunching out code, our community team is building the foundation for our community to thrive in the coming months. 🚀 Setup a home for BVM holders to hang out 🚀 Work on airdrop deals for BVM holders 🚀 List BVM on Coin Market Cap 🚀 List BVM on Gecko Terminal 🚀 List BVM on DEX 🚀 List BVM on CEX This is our plan for March. We hope to get most of it done for the community. See you on #Bitcoin

📢 $BVM MARCH UPDATES 📢 GM. In Feb, we added these products to the BVM ecosystem: ✅BVM Staking, ✅Bitcoin Stamps, ✅SRC-20 Bridge, ✅Avail, ✅Jackal, ✅Arweave & ✅ IPFS. We're excited to share with the community what we’re shipping next in March. 1. FOUR MORE BITCOIN L2 BLOCKCHAINS BVM’s mission is to scale Bitcoin by powering thousands of Bitcoin L2s. In March, we’ll work with 4 different builders to help them launch their own Bitcoin L2 blockchains and ultimately scale Bitcoin with unlimited throughput and rich user cases. ⛓️ Swamps: DeFi Bitcoin L2 for SRC-20 @swamps_src20 is launching the first-ever Stamps-based Bitcoin L2 blockchain with BVM, designed for SRC-20 DEX. We’re super excited to be working with @0xnovae5Db RMZ 🟨 and the Swamps team. ⛓️ Naka: DeFi Bitcoin L2 for BRC-20/BRC-404 Naka Chain is an Ordinals-based Bitcoin L2 blockchain designed for BRC-20. They just added support for BRC-404. We're excited to work with DeBit | NakaChain.xyz and the Naka team. ⛓️ Satoshi Sync: Bitcoin L2 for Inscription Markets SatoshiSync is building the first chain-agnostic Inscription Protocol with BVM, enabling everyone to launch their own Inscription Market. It’s been a pleasure working with @okubalinska and the Satoshi Sync team. ⛓️ ███████: Bitcoin L2 for ████ This is a big one. We'll share it when it's ready. Super excited to be working on this with ███████ and the ███████ team. Note: If you're thinking about launching your own Bitcoin L2, please DM us. Would love to help! 2. BITCOIN MODULE STORE BVM is a modular Bitcoin L2 infrastructure. The more building blocks, the more developers can build on Bitcoin. This month, we’ll be adding these new modules to the Bitcoin Modular Store. 🧱 Filecoin Filecoin is one of the best decentralized storage networks. Developers can build their Bitcoin L2 and choose Filecoin to store their data. 🧱 Syscoin zkDA Syscoin is a Bitcoin-backed Data Availability layer. It is secured by Bitcoin’s own PoW plus Syscoin’s finality. We think it's a super interesting DA solution. 🧱 BVM-ETH Bridge Ethereum bridge will enable us to flow assets between BVM and Ethereum. You can trade $BVM on Uniswap or bridge ETH to various Bitcoin L2 blockchains powered by BVM. Note: please do make suggestions on which modules we should integrate next! 3. BRC-404 Starting out as a fun weekend experiment, BRC-404 shows us that it could become a promising protocol. $3.5M trading volume so far! 🤯🤯🤯 🎨 Fractionalize Ordinals with BRC-404 The idea is to take an existing Ordinal Inscription and turn it into BRC-404. Then, Ordinals would be much more accessible to many smaller buyers, and at the same time, the collectors would make a lot more on trading fees. That's really the beauty of semi-fungible tokens. You can move between the fungible worlds and the non-fungible worlds. 🔒 ███████ for BRC-404 This product is exciting. We can't quite talk about it yet. But we'll do it when it's ready later this month. 4. COMMUNITY While our product team is busy crunching out code, our community team is building the foundation for our community to thrive in the coming months. 🚀 Setup a home for BVM holders to hang out 🚀 Work on airdrop deals for BVM holders 🚀 List BVM on Coin Market Cap 🚀 List BVM on Gecko Terminal 🚀 List BVM on DEX 🚀 List BVM on CEX This is our plan for March. We hope to get most of it done for the community. See you on #Bitcoin

Bitcoin Virtual Machine

32,165 views • 2 years ago

Thermodynamic computing is here There is a new computing paradigm emerging from the noise, and its arrival may be as significant as the dawn of deep learning or the advent of cloud virtualization. A new company, Extropic, has just launched its first thermodynamic computer, a device they call a TSU, or Thermal Sampling Unit. While the web is already filling with deep technical dives, what’s more important for most of us is building a clear intuition for what this technology is, how it’s fundamentally different from anything that’s come before, and why it’s generating so much excitement. This isn’t just another chip; it’s a new way to think about computation itself. Seeing is Believing: Solving Puzzles in One Shot To understand what a TSU does, let’s look at two classic, notoriously difficult computer science problems: Sudoku and the Eight Queens problem. When you or I solve a Sudoku, we use a process of sequential logic, guess-and-check, and backtracking. We make an assumption, follow its logical conclusion, and if we hit a dead end, we erase and try again. A classical computer does the same, just much faster. A TSU, however, approaches this in a completely different way. Using a TSU simulator, one can “program” the problem by first clamping the known values—the clues already on the board. Then, you program in the constraints: no duplicate numbers in any row, column, or 3x3 square. With the problem thus defined, the TSU doesn’t “search” for a solution; it anneals one. In a single computational step, the solution simply emerges, backfilling all the empty squares correctly. The same principle applies to the Eight Queens problem, a challenge to place eight queens on a chessboard so that none can attack any other. This is a complex combinatorial problem with 92 distinct solutions. A classical computer would have to iteratively search for these. A TSU, by contrast, can be programmed with the constraints (the “anti-affinity” between queens on the same row, column, or diagonal) and then set to sample the “solution space.” In this context, a valid solution is one with a “problem energy” of zero. The TSU’s physical nature allows it to naturally find these zero-energy states. A simulation of this process shows the TSU discovering all 92 unique solutions, demonstrating its ability to not just find an answer, but to explore the entire landscape of all correct answers. This is a fundamentally new approach, one that bypasses the brute-force, iterative methods we’ve relied on for decades. The Physics of Computation: Using Noise, Not Fighting It This new power comes from a radical design philosophy. For the last 70 years, computing has been about one thing: order. We build chips that are deterministic, logical, and precise. The great enemy has always been noise, heat, and randomness. We spend billions on cooling and error correction to eliminate these very things. Quantum computing, in many ways, is the ultimate expression of this, requiring temperatures near absolute zero to eliminate all thermal noise and achieve quantum coherence. Thermodynamic computing is the polar opposite. It doesn’t fight the noise; it uses it. The TSU is built on the understanding that the natural, stochastic noise from “leaky” transistors—the very randomness we’ve tried to engineer out of existence—is itself a powerful computational resource. Think of it this way: a GPU, which is central to today’s AI, has to simulate noise. When a generative AI model creates a new image or sentence, it’s using complex algorithms to fake randomness. The TSU doesn’t need to fake it; it harnesses the actual physical randomness of thermodynamics. It is a piece of hardware that directly computes with probability. This makes it a hybrid, sitting somewhere between a purely analog computer (which might use light or sound waves to compute) and a digital GPU. It’s a physical device that leverages the laws of physics itself to find solutions, rather than just using logic gates to simulate them. From a Lost Hiker to a Million Bouncy Balls Perhaps the best way to build intuition is with a metaphor. Imagine that solving a complex optimization problem is like trying to find the lowest point of altitude in a 100-square-mile mountainous landscape. Classical computing, using an algorithm like gradient descent, is like being a single hiker dropped into this landscape at night. You have no map or satellite view. All you have is an altimeter and the sensation of the slope under your feet. You can only take one step at a time, always walking downhill, hoping you don’t get stuck in a small local valley when the true, lowest canyon is miles away. Thermodynamic computing is a completely different approach. It’s like having a million bouncy balls and a helicopter. You drop all million balls simultaneously across the entire 100-square-mile landscape. Then, you “turn on an earthquake,” shaking the entire system. The balls bounce and jostle, but as the shaking (the “annealing”) subsides, where do they all end up? They naturally settle into the lowest points. The balls that collect in the deepest valley represent the optimal solution. The TSU is, in essence, a physical device for dropping those million balls at once and letting the laws of thermodynamics find the lowest “energy” state for you, all at the same time. Beyond Puzzles: The Real-World Impact This is far more than just a clever way to solve brain teasers. This ability to instantly find the lowest energy state for a complex, constrained system has staggering real-world applications. One of the most immediate is protein folding. Companies like Google’s DeepMind have made incredible progress with AI like AlphaFold, which predicts protein structures. But this is still a predictive model trained on existing data. A TSU could potentially solve the folding problem directly, treating the protein as a system of atomic affinities and repulsions and finding its most stable, lowest-energy configuration almost instantaneously. This could revolutionize drug discovery and materials science. An even more profound possibility lies in nuclear fusion. One of the greatest engineering challenges in history is controlling the superheated plasma within a tokamak reactor. This requires shaping unimaginably complex magnetic containment fields in real-time to prevent the plasma from touching the reactor walls. This is a real-time optimization problem so complex it’s currently beyond our capabilities. A TSU, however, could be fast enough. Its ability to compute with electricity itself, rather than abstracting the problem through layers of software, might allow it to update the magnetic fields fast enough to stabilize the fusion reaction. One could even imagine a future where thermodynamic computing elements are built directly into the tokamak’s walls, allowing the reactor to physically and intelligently react to the plasma’s state in real time. A ‘GPT-2 Moment’ for a New Era It’s easy to become numb to hype, but what we are witnessing with the TSU feels different. This is what you might call a “GPT-2 moment.” For those who were there, GPT-2 was the first generative AI model that wasn’t just a toy; it was the first time you could play with it at home and see the spark of true generative intelligence. It was the precursor that pointed directly to the GPT-3 and ChatGPT revolution that has since changed the world. This TSU has that same feel. It’s the “SDK” for a new computing paradigm. This technology is as different from classical computing as quantum computing is, but with a critical difference: a team of 15 built this in two years, and it runs at room temperature on your desk. Quantum computing has seen decades of work and billions in funding, and it still hasn’t produced a commercially viable, scalable machine. The TSU is here now. Based on a two-decade-long career at the cutting edge of technology—from seeing the obvious future of virtualization in 2007 to an early conviction in deep learning and GPT—this has all the same hallmarks of a fundamental, world-changing shift. We are not just building faster calculators; we are learning to compute with the universe itself. Pay close attention to this. This is the next big thing.

Thermodynamic computing is here There is a new computing paradigm emerging from the noise, and its arrival may be as significant as the dawn of deep learning or the advent of cloud virtualization. A new company, Extropic, has just launched its first thermodynamic computer, a device they call a TSU, or Thermal Sampling Unit. While the web is already filling with deep technical dives, what’s more important for most of us is building a clear intuition for what this technology is, how it’s fundamentally different from anything that’s come before, and why it’s generating so much excitement. This isn’t just another chip; it’s a new way to think about computation itself. Seeing is Believing: Solving Puzzles in One Shot To understand what a TSU does, let’s look at two classic, notoriously difficult computer science problems: Sudoku and the Eight Queens problem. When you or I solve a Sudoku, we use a process of sequential logic, guess-and-check, and backtracking. We make an assumption, follow its logical conclusion, and if we hit a dead end, we erase and try again. A classical computer does the same, just much faster. A TSU, however, approaches this in a completely different way. Using a TSU simulator, one can “program” the problem by first clamping the known values—the clues already on the board. Then, you program in the constraints: no duplicate numbers in any row, column, or 3x3 square. With the problem thus defined, the TSU doesn’t “search” for a solution; it anneals one. In a single computational step, the solution simply emerges, backfilling all the empty squares correctly. The same principle applies to the Eight Queens problem, a challenge to place eight queens on a chessboard so that none can attack any other. This is a complex combinatorial problem with 92 distinct solutions. A classical computer would have to iteratively search for these. A TSU, by contrast, can be programmed with the constraints (the “anti-affinity” between queens on the same row, column, or diagonal) and then set to sample the “solution space.” In this context, a valid solution is one with a “problem energy” of zero. The TSU’s physical nature allows it to naturally find these zero-energy states. A simulation of this process shows the TSU discovering all 92 unique solutions, demonstrating its ability to not just find an answer, but to explore the entire landscape of all correct answers. This is a fundamentally new approach, one that bypasses the brute-force, iterative methods we’ve relied on for decades. The Physics of Computation: Using Noise, Not Fighting It This new power comes from a radical design philosophy. For the last 70 years, computing has been about one thing: order. We build chips that are deterministic, logical, and precise. The great enemy has always been noise, heat, and randomness. We spend billions on cooling and error correction to eliminate these very things. Quantum computing, in many ways, is the ultimate expression of this, requiring temperatures near absolute zero to eliminate all thermal noise and achieve quantum coherence. Thermodynamic computing is the polar opposite. It doesn’t fight the noise; it uses it. The TSU is built on the understanding that the natural, stochastic noise from “leaky” transistors—the very randomness we’ve tried to engineer out of existence—is itself a powerful computational resource. Think of it this way: a GPU, which is central to today’s AI, has to simulate noise. When a generative AI model creates a new image or sentence, it’s using complex algorithms to fake randomness. The TSU doesn’t need to fake it; it harnesses the actual physical randomness of thermodynamics. It is a piece of hardware that directly computes with probability. This makes it a hybrid, sitting somewhere between a purely analog computer (which might use light or sound waves to compute) and a digital GPU. It’s a physical device that leverages the laws of physics itself to find solutions, rather than just using logic gates to simulate them. From a Lost Hiker to a Million Bouncy Balls Perhaps the best way to build intuition is with a metaphor. Imagine that solving a complex optimization problem is like trying to find the lowest point of altitude in a 100-square-mile mountainous landscape. Classical computing, using an algorithm like gradient descent, is like being a single hiker dropped into this landscape at night. You have no map or satellite view. All you have is an altimeter and the sensation of the slope under your feet. You can only take one step at a time, always walking downhill, hoping you don’t get stuck in a small local valley when the true, lowest canyon is miles away. Thermodynamic computing is a completely different approach. It’s like having a million bouncy balls and a helicopter. You drop all million balls simultaneously across the entire 100-square-mile landscape. Then, you “turn on an earthquake,” shaking the entire system. The balls bounce and jostle, but as the shaking (the “annealing”) subsides, where do they all end up? They naturally settle into the lowest points. The balls that collect in the deepest valley represent the optimal solution. The TSU is, in essence, a physical device for dropping those million balls at once and letting the laws of thermodynamics find the lowest “energy” state for you, all at the same time. Beyond Puzzles: The Real-World Impact This is far more than just a clever way to solve brain teasers. This ability to instantly find the lowest energy state for a complex, constrained system has staggering real-world applications. One of the most immediate is protein folding. Companies like Google’s DeepMind have made incredible progress with AI like AlphaFold, which predicts protein structures. But this is still a predictive model trained on existing data. A TSU could potentially solve the folding problem directly, treating the protein as a system of atomic affinities and repulsions and finding its most stable, lowest-energy configuration almost instantaneously. This could revolutionize drug discovery and materials science. An even more profound possibility lies in nuclear fusion. One of the greatest engineering challenges in history is controlling the superheated plasma within a tokamak reactor. This requires shaping unimaginably complex magnetic containment fields in real-time to prevent the plasma from touching the reactor walls. This is a real-time optimization problem so complex it’s currently beyond our capabilities. A TSU, however, could be fast enough. Its ability to compute with electricity itself, rather than abstracting the problem through layers of software, might allow it to update the magnetic fields fast enough to stabilize the fusion reaction. One could even imagine a future where thermodynamic computing elements are built directly into the tokamak’s walls, allowing the reactor to physically and intelligently react to the plasma’s state in real time. A ‘GPT-2 Moment’ for a New Era It’s easy to become numb to hype, but what we are witnessing with the TSU feels different. This is what you might call a “GPT-2 moment.” For those who were there, GPT-2 was the first generative AI model that wasn’t just a toy; it was the first time you could play with it at home and see the spark of true generative intelligence. It was the precursor that pointed directly to the GPT-3 and ChatGPT revolution that has since changed the world. This TSU has that same feel. It’s the “SDK” for a new computing paradigm. This technology is as different from classical computing as quantum computing is, but with a critical difference: a team of 15 built this in two years, and it runs at room temperature on your desk. Quantum computing has seen decades of work and billions in funding, and it still hasn’t produced a commercially viable, scalable machine. The TSU is here now. Based on a two-decade-long career at the cutting edge of technology—from seeing the obvious future of virtualization in 2007 to an early conviction in deep learning and GPT—this has all the same hallmarks of a fundamental, world-changing shift. We are not just building faster calculators; we are learning to compute with the universe itself. Pay close attention to this. This is the next big thing.

David Shapiro (L/0)

83,649 views • 9 months ago

Over the past two years, AI video models have been competing on realism, resolution, and duration. But no matter how impressive the results look, we remain passive viewers: we press play, watch the clip, and it ends. AlayaWorld Alaya Lab is attempting something fundamentally different. Instead of generating a fixed video, it generates a world that continues to unfold as you move through it. These three demos show the same journey toward a green village rendered in three distinct styles: photorealistic, oil painting, and line art. As the camera moves forward, the model continues generating the road, fences, trees, and distant village. This is not simply an existing video with different filters applied. The environment is generated continuously along the camera trajectory, allowing the scene to develop as the user explores it. AlayaWorld streams video at 720p and 24 FPS while supporting camera movement and viewpoint control. The real breakthrough is not just image quality. Once generation becomes fast enough to respond within an interactive loop, the user is no longer merely watching a video. They become a participant inside the generated world. The world can also respond to new instructions. During generation, users can introduce prompts that trigger spells, summon characters, create explosions, or transform the environment. Most video models follow an initial prompt and produce a predetermined clip. AlayaWorld can respond to changing intent while the world is still running, allowing subsequent events to evolve according to the user’s commands. Generating an attractive frame is relatively easy. Maintaining a coherent world over time is much harder. As a video model repeatedly predicts the next frame, small errors can accumulate until roads, buildings, and objects begin to distort or disappear. AlayaWorld combines spatial memory with compressed historical context, helping the model remember both where things are and what has already happened. This enables stable generation lasting more than one minute while improving consistency when the camera leaves an area and later returns. This may be the next step for AI video: not simply generating a longer movie, but generating a world that can be explored, changed, and interacted with. AlayaWorld is developed by Alaya Lab. The team is progressively releasing its inference code, training code, and datasets, with an online experience expected to launch near the end of the month. Project page:

Over the past two years, AI video models have been competing on realism, resolution, and duration. But no matter how impressive the results look, we remain passive viewers: we press play, watch the clip, and it ends. AlayaWorld Alaya Lab is attempting something fundamentally different. Instead of generating a fixed video, it generates a world that continues to unfold as you move through it. These three demos show the same journey toward a green village rendered in three distinct styles: photorealistic, oil painting, and line art. As the camera moves forward, the model continues generating the road, fences, trees, and distant village. This is not simply an existing video with different filters applied. The environment is generated continuously along the camera trajectory, allowing the scene to develop as the user explores it. AlayaWorld streams video at 720p and 24 FPS while supporting camera movement and viewpoint control. The real breakthrough is not just image quality. Once generation becomes fast enough to respond within an interactive loop, the user is no longer merely watching a video. They become a participant inside the generated world. The world can also respond to new instructions. During generation, users can introduce prompts that trigger spells, summon characters, create explosions, or transform the environment. Most video models follow an initial prompt and produce a predetermined clip. AlayaWorld can respond to changing intent while the world is still running, allowing subsequent events to evolve according to the user’s commands. Generating an attractive frame is relatively easy. Maintaining a coherent world over time is much harder. As a video model repeatedly predicts the next frame, small errors can accumulate until roads, buildings, and objects begin to distort or disappear. AlayaWorld combines spatial memory with compressed historical context, helping the model remember both where things are and what has already happened. This enables stable generation lasting more than one minute while improving consistency when the camera leaves an area and later returns. This may be the next step for AI video: not simply generating a longer movie, but generating a world that can be explored, changed, and interacted with. AlayaWorld is developed by Alaya Lab. The team is progressively releasing its inference code, training code, and datasets, with an online experience expected to launch near the end of the month. Project page:

Rachel🥥

78,241 views • 1 month ago

Xiaomi SU7 Ultra, with 1548 horsepower, as a true production vehicle, recorded a 7:04.957 Nürburgring lap time in June 2025, breaking the previous records held by Rimac Nevera (7:05.298) and Porsche Taycan Turbo GT (7:07.55). In a run against our BMW i4 M50, the SU7 made it feel as if we were driving a broken B58 in an M2. But stepping back before the final driving impression — our inspection of the vehicle brutally impressed us. Unlike mainstream media that focuses on panel gaps, leather on the gear selector, and other superficial “cup holder review” topics, we went a bit deeper under the surface. The first – and essentially the only serious downside – is that the car has SGW (Secure Gateway), meaning the OBD diagnostic port is locked. By searching for the powertrain CAN bus somewhere along the wiring harness, we were able to manually access the PT network, bypass the SGW, sniff the network and capture packets. The battery system is Qilin 2, with a CATL-derived BMS stack with only minor modifications (similar approach as seen in Xpeng and NIO). On the powertrain side, the CAN network looks very similar to a UAES/Bosch architecture (0x1A0–0x1AF). The official diagnostic tool costs around €10,000, but we are already developing our own tool so we can support this rocket within our service ecosystem. The BMS is an 800V system with 214 NMC prismatic cells in series, forming a 93.7 kWh battery pack tailored for high performance. The car offers up to 630 km (391 miles) CLTC range. It supports 800V fast charging, capable of 10–80% in just 11 minutes (5.2C rate). The pack supports 16C discharge, while the battery mass is 638 kg. The vehicle comes fully in track-focused configuration: massive ceramic brake discs, pads, calipers, wheels, and tires. Unlike the Model S Plaid, this car can take repeated track abuse for 6–7 hours continuously. The rear motor is a dual-unit setup with torque vectoring, which works brutally well but requires proper understanding of the system and tuning in Track Mode. The operating system is HyperOS, very similar to Tesla’s system — perhaps even more detailed in some aspects. Interior quality is excellent, driving characteristics are outstanding, and the overall ride comfort is surprisingly well balanced. The seats are actually more comfortable than those in the BMW i4. Now for the most interesting part. When disassembling interior plastics and cosmetic components, we discovered a surprisingly strong similarity to Tesla Model 3 / Model S next-generation chassis design. The front shock tower, shotgun rails, front HVAC carrier, even the rear seat clips — underneath, the battery penthouse layout, wiring routing, reinforcements, and cable positioning look extremely familiar. The vehicle weighs 2360 kg, while the full-carbon prototype series reportedly weighs around 1900 kg. impression? The overall quality of the vehicle is impressive for a company building its first electric car, especially considering it already broke every major performance record. A similar precedent happened once before — when Tesla introduced the Model S. The SU7 Ultra “democratizes” 1550 horsepower for under €100,000 — performance that is now literally accessible to everyone. Some street racing veterans spend over €250,000 on forged internals, turbo systems, and tuning just to reach similar numbers. Conclusion? This car doesn’t just challenge the German automotive industry — in this segment, it completely dominates it and practically euthanizes it. Every major component on this vehicle has been internally developed and manufactured in China, and pushed very close to perfection. The only thing we still don’t know is how repairable and serviceable the vehicle will be long-term — and only mileage and time will answer that question. Would we have one? Yes! We have already ordered one and started building the import and homologation network. Among our previous favorites — Tesla and BMW — this one is now clearly #1.

Xiaomi SU7 Ultra, with 1548 horsepower, as a true production vehicle, recorded a 7:04.957 Nürburgring lap time in June 2025, breaking the previous records held by Rimac Nevera (7:05.298) and Porsche Taycan Turbo GT (7:07.55). In a run against our BMW i4 M50, the SU7 made it feel as if we were driving a broken B58 in an M2. But stepping back before the final driving impression — our inspection of the vehicle brutally impressed us. Unlike mainstream media that focuses on panel gaps, leather on the gear selector, and other superficial “cup holder review” topics, we went a bit deeper under the surface. The first – and essentially the only serious downside – is that the car has SGW (Secure Gateway), meaning the OBD diagnostic port is locked. By searching for the powertrain CAN bus somewhere along the wiring harness, we were able to manually access the PT network, bypass the SGW, sniff the network and capture packets. The battery system is Qilin 2, with a CATL-derived BMS stack with only minor modifications (similar approach as seen in Xpeng and NIO). On the powertrain side, the CAN network looks very similar to a UAES/Bosch architecture (0x1A0–0x1AF). The official diagnostic tool costs around €10,000, but we are already developing our own tool so we can support this rocket within our service ecosystem. The BMS is an 800V system with 214 NMC prismatic cells in series, forming a 93.7 kWh battery pack tailored for high performance. The car offers up to 630 km (391 miles) CLTC range. It supports 800V fast charging, capable of 10–80% in just 11 minutes (5.2C rate). The pack supports 16C discharge, while the battery mass is 638 kg. The vehicle comes fully in track-focused configuration: massive ceramic brake discs, pads, calipers, wheels, and tires. Unlike the Model S Plaid, this car can take repeated track abuse for 6–7 hours continuously. The rear motor is a dual-unit setup with torque vectoring, which works brutally well but requires proper understanding of the system and tuning in Track Mode. The operating system is HyperOS, very similar to Tesla’s system — perhaps even more detailed in some aspects. Interior quality is excellent, driving characteristics are outstanding, and the overall ride comfort is surprisingly well balanced. The seats are actually more comfortable than those in the BMW i4. Now for the most interesting part. When disassembling interior plastics and cosmetic components, we discovered a surprisingly strong similarity to Tesla Model 3 / Model S next-generation chassis design. The front shock tower, shotgun rails, front HVAC carrier, even the rear seat clips — underneath, the battery penthouse layout, wiring routing, reinforcements, and cable positioning look extremely familiar. The vehicle weighs 2360 kg, while the full-carbon prototype series reportedly weighs around 1900 kg. impression? The overall quality of the vehicle is impressive for a company building its first electric car, especially considering it already broke every major performance record. A similar precedent happened once before — when Tesla introduced the Model S. The SU7 Ultra “democratizes” 1550 horsepower for under €100,000 — performance that is now literally accessible to everyone. Some street racing veterans spend over €250,000 on forged internals, turbo systems, and tuning just to reach similar numbers. Conclusion? This car doesn’t just challenge the German automotive industry — in this segment, it completely dominates it and practically euthanizes it. Every major component on this vehicle has been internally developed and manufactured in China, and pushed very close to perfection. The only thing we still don’t know is how repairable and serviceable the vehicle will be long-term — and only mileage and time will answer that question. Would we have one? Yes! We have already ordered one and started building the import and homologation network. Among our previous favorites — Tesla and BMW — this one is now clearly #1.

EV Clinic

22,172 views • 5 months ago

Super Bowl MVP Why do the odds on Polymarket get it wrong, while the statistics point us to one winner? Super Bowl MVP is always 50% statistics and 50% storyline. We have 4 finalists, but in my opinion there are only two real contenders. - Sam Darnold (The Redemption Arc) He has the strongest media narrative so to speak, a shot-down pilot who returned from hell. His strengths: if he delivers a clean game (250+ yards, 2 TD, 0 INT), journalists will give him the vote simply for the beautiful comeback story. His weaknesses: under pressure in finals, Darnold is still prone to hero ball, risky throws that turn into interceptions. MVPs are not given for mistakes. - Jaxon Smith-Njigba (The Volume King) for me he is Cooper Kupp 2.0 His strengths: in games where the pass rush eats the quarterback alive, JSN becomes the life ring, 10–12 catches per game is his norm. His weaknesses: receivers become MVP only when they completely overshadow their QB. He needs 150+ yards and at least 2 touchdowns to take the statuette. - Kenneth Walker III (The Home Run Hitter) The most explosive running back in the league. His strengths: he is the only one here who can change the course of the game with one touch (70-yard TD run). His weaknesses: the modern NFL, in my opinion, is a passing league. For an RB to become MVP, he needs to gain 150+ rushing yards, which against elite defenses in the final is almost unrealistic. - Drake Maye (The Chosen One) His strengths: he plays like a veteran in the body of an athlete. His ability to extend plays with his legs (scramble) and throw into tight windows is exactly what creates TV “highlights”. His weaknesses: inexperience can lead to a slow start in the first quarter. I bet that the MVP will be taken by DRAKE MAYE Why exactly will he take the award? Three iron arguments: - Quarterback Bias The math is simple: over the last 15 years, quarterbacks have become MVP in 70% of cases. If Maye’s team wins, the award is automatically given to him. To change this, someone else needs to deliver a historic performance (hat trick), which in a final is almost unrealistic. - Clutch Factor Maye is the king of decisive moments. He showed the best numbers in the entire league in the metric “EPA per play” specifically in the 4th quarter and on third downs. And Super Bowls are won precisely under the pressure of the final minutes. - Versatility (The Dual Threat) Look at the difference in styles: •Darnold is only passing (Pocket Passer). •Walker is only running. •Maye is a dual threat. He breaks defensive schemes both with his arm and with his legs. The opponent’s defense simply physically will not be able to fully contain him. I bet here :

Super Bowl MVP Why do the odds on Polymarket get it wrong, while the statistics point us to one winner? Super Bowl MVP is always 50% statistics and 50% storyline. We have 4 finalists, but in my opinion there are only two real contenders. - Sam Darnold (The Redemption Arc) He has the strongest media narrative so to speak, a shot-down pilot who returned from hell. His strengths: if he delivers a clean game (250+ yards, 2 TD, 0 INT), journalists will give him the vote simply for the beautiful comeback story. His weaknesses: under pressure in finals, Darnold is still prone to hero ball, risky throws that turn into interceptions. MVPs are not given for mistakes. - Jaxon Smith-Njigba (The Volume King) for me he is Cooper Kupp 2.0 His strengths: in games where the pass rush eats the quarterback alive, JSN becomes the life ring, 10–12 catches per game is his norm. His weaknesses: receivers become MVP only when they completely overshadow their QB. He needs 150+ yards and at least 2 touchdowns to take the statuette. - Kenneth Walker III (The Home Run Hitter) The most explosive running back in the league. His strengths: he is the only one here who can change the course of the game with one touch (70-yard TD run). His weaknesses: the modern NFL, in my opinion, is a passing league. For an RB to become MVP, he needs to gain 150+ rushing yards, which against elite defenses in the final is almost unrealistic. - Drake Maye (The Chosen One) His strengths: he plays like a veteran in the body of an athlete. His ability to extend plays with his legs (scramble) and throw into tight windows is exactly what creates TV “highlights”. His weaknesses: inexperience can lead to a slow start in the first quarter. I bet that the MVP will be taken by DRAKE MAYE Why exactly will he take the award? Three iron arguments: - Quarterback Bias The math is simple: over the last 15 years, quarterbacks have become MVP in 70% of cases. If Maye’s team wins, the award is automatically given to him. To change this, someone else needs to deliver a historic performance (hat trick), which in a final is almost unrealistic. - Clutch Factor Maye is the king of decisive moments. He showed the best numbers in the entire league in the metric “EPA per play” specifically in the 4th quarter and on third downs. And Super Bowls are won precisely under the pressure of the final minutes. - Versatility (The Dual Threat) Look at the difference in styles: •Darnold is only passing (Pocket Passer). •Walker is only running. •Maye is a dual threat. He breaks defensive schemes both with his arm and with his legs. The opponent’s defense simply physically will not be able to fully contain him. I bet here :

AdiiX

22,658 views • 6 months ago

✨New demo: what if vibe coding felt more visual? Brian Lovin Mary Rose Cook and I did a game jam using Notion as our "IDE": launching Cursor agents from a task board, and making a custom image for each task 😎 The demo shows 3 ideas for the future of agents: 1) Agents should collaborate across apps. Each app has its focus--Notion AI is good at drafting specs and organizing tasks; Cursor is good at coding. So let them specialize! Today we're launching a new integration where Notion AI can kick off Cursor Cloud Agents to do coding tasks. The Cursor API accepts natural language prompts, so I think of this as "cross-app sub-agents" -- it's kinda cute how it resembles humans hiring outside contractors 😊 BTW: the parallelism of cloud agents is incredibly freeing for creativity, but it also creates a new problem: sooo much work to keep track of! Which brings us to the next idea... 2) Agent orchestration is a data visualization problem. A powerful frame for designing agent UIs is to think of the chat transcripts as the "raw data" and ask: what visual projections might help people make sense of this data at scale? We need to engage our human GPUs -- our visual processing -- to understand what the computer GPUs are doing for us! One thing we can do is use AI to populate traditional UIs like progress bars and status updates. But there are also new possibilities now... For example: when you have a lot going on, it can be hard to identify tasks just by text titles. So we tried generating an AI image for each task -- turns out this helps a lot by giving it a unique visual identity! And of course, it also just makes it super fun to build with friends 😃 Speaking of friends... 3) The future of coding is collaborative. Sometimes it feels like IC engineers are being reduced to middle managers: shuffling information between the team's context and the coding agents that they individually manage. The solution: bring all the people and agents into one shared space, with shared context and visibility! In the video you can get a glimpse of how this feels. Mary, Brian and I record ourselves chatting about ideas, and then we use AI to turn that conversation into a list of tasks on a shared board. As the ideas get built in parallel, we can all monitor progress and review the work together, nothing is siloed. My main takeaway from this game jam was: damn, creativity with friends, at the speed of conversation, is incredibly fun. --- Our goal here is to let anyone use Notion as a fun and creative "software factory" to build software together with your team. Give the Cursor integration a shot and let us know what you think! (AI Image gen in Notion isn't GA yet, but coming soon and already out to some users) And let me know if you'd want a template or more detailed instructions on the setup we showed in this demo...

✨New demo: what if vibe coding felt more visual? Brian Lovin Mary Rose Cook and I did a game jam using Notion as our "IDE": launching Cursor agents from a task board, and making a custom image for each task 😎 The demo shows 3 ideas for the future of agents: 1) Agents should collaborate across apps. Each app has its focus--Notion AI is good at drafting specs and organizing tasks; Cursor is good at coding. So let them specialize! Today we're launching a new integration where Notion AI can kick off Cursor Cloud Agents to do coding tasks. The Cursor API accepts natural language prompts, so I think of this as "cross-app sub-agents" -- it's kinda cute how it resembles humans hiring outside contractors 😊 BTW: the parallelism of cloud agents is incredibly freeing for creativity, but it also creates a new problem: sooo much work to keep track of! Which brings us to the next idea... 2) Agent orchestration is a data visualization problem. A powerful frame for designing agent UIs is to think of the chat transcripts as the "raw data" and ask: what visual projections might help people make sense of this data at scale? We need to engage our human GPUs -- our visual processing -- to understand what the computer GPUs are doing for us! One thing we can do is use AI to populate traditional UIs like progress bars and status updates. But there are also new possibilities now... For example: when you have a lot going on, it can be hard to identify tasks just by text titles. So we tried generating an AI image for each task -- turns out this helps a lot by giving it a unique visual identity! And of course, it also just makes it super fun to build with friends 😃 Speaking of friends... 3) The future of coding is collaborative. Sometimes it feels like IC engineers are being reduced to middle managers: shuffling information between the team's context and the coding agents that they individually manage. The solution: bring all the people and agents into one shared space, with shared context and visibility! In the video you can get a glimpse of how this feels. Mary, Brian and I record ourselves chatting about ideas, and then we use AI to turn that conversation into a list of tasks on a shared board. As the ideas get built in parallel, we can all monitor progress and review the work together, nothing is siloed. My main takeaway from this game jam was: damn, creativity with friends, at the speed of conversation, is incredibly fun. --- Our goal here is to let anyone use Notion as a fun and creative "software factory" to build software together with your team. Give the Cursor integration a shot and let us know what you think! (AI Image gen in Notion isn't GA yet, but coming soon and already out to some users) And let me know if you'd want a template or more detailed instructions on the setup we showed in this demo...

Geoffrey Litt

88,919 views • 5 months ago

Anthropic CEO Dario Amodei just gave THE MOST accelerated talk on how scaling will continue to make models exponentially more powerful for a long time to come — and that there is "NO WALL." 🔥 - From 'Alex Kantrowitz' YT Channel (Full Video link in comment) --- "The thing I think is real that I've said over and over again is the exponential. The idea that every few months we get an AI model that is better than the AI model we got before. And we get that by investing more compute in AI models, more data, more new types of training models. Initially, this was done by what's called pre-training, which is when you just feed a bunch of data from the internet into the model. Now we have a second stage that's reinforcement learning or test time compute or reasoning or whatever you want to call it. I think of it as a second stage that involves reinforcement learning. Now both of those things are scaling up together, as we've seen with our models and as we've seen with models from other companies. I don't see anything blocking the further scaling of that. There's some stuff about how do we broaden the tasks on the RL side of it. We've seen more progress on, say, math and code, where the models are getting pretty close to a high professional level, and less on more subjective tasks, but I think that is very much a temporary obstacle. So when I look at it, I see this exponential and I say, look, people aren't very good at making sense of exponentials, right? Like, if something is doubling every 6 months, then 2 years before it happens, it looks like it's only 1/16th of the way there. And so we are sitting here in the middle of 2025, and the models are really starting to explode in terms of the economy. If you look at the capabilities of the model, they're starting to saturate all the benchmarks. If you look at revenue, and you know, Anthropic's revenue every year has grown 10x. Every year we’re kind of conservative and we say, it can’t grow 10x this time. I never assume anything and actually always am very conservative in saying I think it's going to slow down on the business side. But we went from zero to $100 million in 2023, we went from $100 million to $1 billion in 2024, and this year, in the first half of the year, we've gone from $1 billion to, I think as of speaking today, it's well above $4 billion, it might be $4.5 billion. And so if you think about it, suppose that exponential continued for 2 years. I'm not saying it will, but suppose it continued for 2 years. You're well into the $100 billions. I'm not saying that'll happen. I'm saying the situation is that when you're on an exponential, you can really get fooled by it. 2 years away from when the exponential goes totally crazy, it looks like it's just starting to be a thing. And so that's the fundamental dynamic. We saw that with the internet in the '90s, right? Where it was like networking speeds and the underlying speed of the computers were getting fast, and over a few years it became possible to have to basically build a digital global communications network on top of all this when it wasn't possible just a few years ago and and almost no one except for a few people really saw the implications of that and how fast it." - Anthropic CEO Dario Amodei

Anthropic CEO Dario Amodei just gave THE MOST accelerated talk on how scaling will continue to make models exponentially more powerful for a long time to come — and that there is "NO WALL." 🔥 - From 'Alex Kantrowitz' YT Channel (Full Video link in comment) --- "The thing I think is real that I've said over and over again is the exponential. The idea that every few months we get an AI model that is better than the AI model we got before. And we get that by investing more compute in AI models, more data, more new types of training models. Initially, this was done by what's called pre-training, which is when you just feed a bunch of data from the internet into the model. Now we have a second stage that's reinforcement learning or test time compute or reasoning or whatever you want to call it. I think of it as a second stage that involves reinforcement learning. Now both of those things are scaling up together, as we've seen with our models and as we've seen with models from other companies. I don't see anything blocking the further scaling of that. There's some stuff about how do we broaden the tasks on the RL side of it. We've seen more progress on, say, math and code, where the models are getting pretty close to a high professional level, and less on more subjective tasks, but I think that is very much a temporary obstacle. So when I look at it, I see this exponential and I say, look, people aren't very good at making sense of exponentials, right? Like, if something is doubling every 6 months, then 2 years before it happens, it looks like it's only 1/16th of the way there. And so we are sitting here in the middle of 2025, and the models are really starting to explode in terms of the economy. If you look at the capabilities of the model, they're starting to saturate all the benchmarks. If you look at revenue, and you know, Anthropic's revenue every year has grown 10x. Every year we’re kind of conservative and we say, it can’t grow 10x this time. I never assume anything and actually always am very conservative in saying I think it's going to slow down on the business side. But we went from zero to $100 million in 2023, we went from $100 million to $1 billion in 2024, and this year, in the first half of the year, we've gone from $1 billion to, I think as of speaking today, it's well above $4 billion, it might be $4.5 billion. And so if you think about it, suppose that exponential continued for 2 years. I'm not saying it will, but suppose it continued for 2 years. You're well into the $100 billions. I'm not saying that'll happen. I'm saying the situation is that when you're on an exponential, you can really get fooled by it. 2 years away from when the exponential goes totally crazy, it looks like it's just starting to be a thing. And so that's the fundamental dynamic. We saw that with the internet in the '90s, right? Where it was like networking speeds and the underlying speed of the computers were getting fast, and over a few years it became possible to have to basically build a digital global communications network on top of all this when it wasn't possible just a few years ago and and almost no one except for a few people really saw the implications of that and how fast it." - Anthropic CEO Dario Amodei

Rohan Paul

74,406 views • 1 year ago