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.

Rockets are hard. Flying robots are harder. Sensing, compute, and software stack that parallels a self-driving car, but three orders of magnitude smaller in size and weight. Plus the full gamut of aerospace complexity spanning aerodynamics, thermals, vibration. Very high flight volumes, low and close to the ground, in... show more

Adam Bry

7,655 subscribers

17,924 views • 5 months ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

7 Comments

Adam Bry5 months ago

Full @sourceryy conversation with @MollySOShea, guest appearance by @skydiohq's Ben Thompson

Chuck Petras5 months ago

@GrandpaRoy2

Sugardaddy5 months ago

@hadip Dm me

Steven Cheng5 months ago

Fighting for every gram hits hard. I battle tight limits on my delta bots, but adding aerodynamics and thermals to a vision stack? Brutal optimization loop. Huge respect.

James Ryan Taylor5 months ago

Its like this

Lukas w/ K5 months ago

nice rage bait lmao. Kids can design and build drones. Yea, they are simple and not as performant but they fly. No person on this earth can build a space ship or rocket by themselves. Big teams with millions of funding fail..

Mike Carrieri5 months ago

9/30/2019

Related Videos

🇮🇱 Ben-Gurion University in South Israel just built a robot that switches between driving and crawling depending on the terrain. Nobody asked why Israeli engineers are suddenly very motivated to solve the problem of robots that can navigate rough ground and climb obstacles.

🇮🇱 Ben-Gurion University in South Israel just built a robot that switches between driving and crawling depending on the terrain. Nobody asked why Israeli engineers are suddenly very motivated to solve the problem of robots that can navigate rough ground and climb obstacles.

Mario Nawfal

316,658 views • 6 months ago

Zero-shot navigation of complex terrain. The true advantages of humanoid robots are their ability to work in the true difficulty of human environments. All using onboard sensing and compute. Take a look ->

Zero-shot navigation of complex terrain. The true advantages of humanoid robots are their ability to work in the true difficulty of human environments. All using onboard sensing and compute. Take a look ->

Chris Paxton

10,945 views • 1 month ago

In Florida, a woman driving a Chevrolet Silverado ran over a low-slung Lamborghini Huracán. She missed the car due to its low profile and the truck's blind spot. The sports car driver remains unharmed, but the repair costs are expected to be very high.

Visegrád 24

89,032 views • 6 months ago

#WATCH | Sriharikota, Andhra Pradesh: Pawan Kumar Chandan, CEO and Founder of Skyroot Aerospace, says, "Today, we are here at the Spaceport of India, Sriharikota, where we are going to launch India's first private orbital rocket... The first time in India that a private company has developed an orbital rocket and taken it to the launch site and is going to take off soon. And it's a proud moment for India... very few companies in the world have actually launched a rocket to orbit, and very few companies in the world are operating regular orbital rockets. This is a very unique capability which the world needs. And as an Indian company, we are very proud that we are going to launch our first test flight of Vikram-1 soon..."

#WATCH | Sriharikota, Andhra Pradesh: Pawan Kumar Chandan, CEO and Founder of Skyroot Aerospace, says, "Today, we are here at the Spaceport of India, Sriharikota, where we are going to launch India's first private orbital rocket... The first time in India that a private company has developed an orbital rocket and taken it to the launch site and is going to take off soon. And it's a proud moment for India... very few companies in the world have actually launched a rocket to orbit, and very few companies in the world are operating regular orbital rockets. This is a very unique capability which the world needs. And as an Indian company, we are very proud that we are going to launch our first test flight of Vikram-1 soon..."

ANI

97,210 views • 3 months ago

Tesla's "Full Self-Driving" software is a joke. Here FSD encounters a blocked left turn, so decides to do a three-point turn and then drive the wrong way down a one-way street into oncoming traffic! Tesla's defective software is orders of magnitude away from safe operation.

Tesla's "Full Self-Driving" software is a joke. Here FSD encounters a blocked left turn, so decides to do a three-point turn and then drive the wrong way down a one-way street into oncoming traffic! Tesla's defective software is orders of magnitude away from safe operation.

Dan O'Dowd

28,351 views • 1 year ago

Legacy automakers are in a very precarious position right now. They haven't yet figured out how to make compelling EVs profitably, so now many of them are starting to fall back on the gas products that do make them money, but this will put them even further behind in the future, which also means they will likely underinvest in self-driving capabilities and the best product for that (EVs). Ford is working on its "Universal EV Platform" designed to reduce costs, with the first vehicle with that aiming for a 2027 release, but it remains to be seen whether that plan will live up to its promise. I think we’re going to see a lot of consolidation across the legacy auto industry over the next few years. Many of these companies will struggle to catch up in self-driving because they spent years dragging their feet. At this point, self-driving is largely a software problem, and software is where most legacy automakers are weakest. Many still struggle with basic in-car UI and software updates, let alone building and iterating on full self-driving systems at scale (pretty much non-existent for them).

Legacy automakers are in a very precarious position right now. They haven't yet figured out how to make compelling EVs profitably, so now many of them are starting to fall back on the gas products that do make them money, but this will put them even further behind in the future, which also means they will likely underinvest in self-driving capabilities and the best product for that (EVs). Ford is working on its "Universal EV Platform" designed to reduce costs, with the first vehicle with that aiming for a 2027 release, but it remains to be seen whether that plan will live up to its promise. I think we’re going to see a lot of consolidation across the legacy auto industry over the next few years. Many of these companies will struggle to catch up in self-driving because they spent years dragging their feet. At this point, self-driving is largely a software problem, and software is where most legacy automakers are weakest. Many still struggle with basic in-car UI and software updates, let alone building and iterating on full self-driving systems at scale (pretty much non-existent for them).

Sawyer Merritt

201,715 views • 10 months ago

Elon Musk in new interview on how Tesla’s Optimus robot will learn more quickly: "What we're going to need to do is build a lot of robots and put them in a kind of like a "Optimus Academy" so they can do self-play in reality. We're actually building that out. We're gonna have at least 10,000 Optimus robots, maybe 20,000 or 30,000, that are doing self-play and testing different tasks. Tesla has quite a good reality generator. A physics-accurate reality generator that we made for the cars. We'll do the same things for the robots, and actually have done that for the robots. You can do millions of simulated robots in a simulated world, and you use the tens of thousands in the real world to close the simulation to reality gap.”

Elon Musk in new interview on how Tesla’s Optimus robot will learn more quickly: "What we're going to need to do is build a lot of robots and put them in a kind of like a "Optimus Academy" so they can do self-play in reality. We're actually building that out. We're gonna have at least 10,000 Optimus robots, maybe 20,000 or 30,000, that are doing self-play and testing different tasks. Tesla has quite a good reality generator. A physics-accurate reality generator that we made for the cars. We'll do the same things for the robots, and actually have done that for the robots. You can do millions of simulated robots in a simulated world, and you use the tens of thousands in the real world to close the simulation to reality gap.”

Sawyer Merritt

250,956 views • 9 months ago

ELON: TESLA’S ROBOTS ARE MOVING FROM SIMULATIONS TO REALITY “For the robot, we’re going to need to build a lot of them and put them in an Optimus Academy so they can do self-play in reality and test different tasks. Tesla has quite a good, physics accurate, reality-generator, and we’re doing that for the robots. So you can do millions of robots in a simulated world, and then 10k in the real world to close the simulation to reality gap. Grok would orchestrate the behavior of the Optimus robots.” Source: Cheeky Pint Podcast, Elon Musk

ELON: TESLA’S ROBOTS ARE MOVING FROM SIMULATIONS TO REALITY “For the robot, we’re going to need to build a lot of them and put them in an Optimus Academy so they can do self-play in reality and test different tasks. Tesla has quite a good, physics accurate, reality-generator, and we’re doing that for the robots. So you can do millions of robots in a simulated world, and then 10k in the real world to close the simulation to reality gap. Grok would orchestrate the behavior of the Optimus robots.” Source: Cheeky Pint Podcast, Elon Musk

Mario Nawfal

19,289 views • 9 months ago

Cathie Wood says humanoid robots are roughly 200,000 times more complex than robotaxis, with the hands being the hardest part to solve. But she argues Tesla has a major advantage because humanoid robots rely on the same core technologies it is already developing for autonomous vehicles: robotics, batteries, and AI. “Obviously the hands being the most complicated part.” “Elon says maybe late ’28 into ’29 for scaling. We would put that a couple of years later.” “The reason Tesla is furthest ahead on this is it’s the same three technology platforms that are converging in humanoid robots as it is in robotaxis.” “Robots, battery-operated electric, and they are powered by AI, just like robotaxis.” Wood also believes Tesla is now very close to solving autonomous driving. “We’re very close to completely solving the robotaxi problem.” “If any of you are driving with Tesla’s FSD with the latest software update, it’s unbelievable.” “It’s magic. I would never drive another car.” Peter H. Diamandis, MD Cathie Wood

Cathie Wood says humanoid robots are roughly 200,000 times more complex than robotaxis, with the hands being the hardest part to solve. But she argues Tesla has a major advantage because humanoid robots rely on the same core technologies it is already developing for autonomous vehicles: robotics, batteries, and AI. “Obviously the hands being the most complicated part.” “Elon says maybe late ’28 into ’29 for scaling. We would put that a couple of years later.” “The reason Tesla is furthest ahead on this is it’s the same three technology platforms that are converging in humanoid robots as it is in robotaxis.” “Robots, battery-operated electric, and they are powered by AI, just like robotaxis.” Wood also believes Tesla is now very close to solving autonomous driving. “We’re very close to completely solving the robotaxi problem.” “If any of you are driving with Tesla’s FSD with the latest software update, it’s unbelievable.” “It’s magic. I would never drive another car.” Peter H. Diamandis, MD Cathie Wood

KanekoaTheGreat

37,717 views • 1 month ago

Today marks the beginning of a new era. Introducing: Cowboy Space Corporation. We are building orbital infrastructure for the AI era: a fully integrated system of rockets and satellites designed to deliver high-performance compute and optical data transmission directly from Low Earth Orbit.

Today marks the beginning of a new era. Introducing: Cowboy Space Corporation. We are building orbital infrastructure for the AI era: a fully integrated system of rockets and satellites designed to deliver high-performance compute and optical data transmission directly from Low Earth Orbit.

Cowboy Space Corp.

809,706 views • 6 months ago

Ashok Elluswamy, Tesla's AI lead, during a GTC discussion, highlighting the fundamental similarity in AI approaches for self-driving cars and humanoid robots: - Hierarchical decision making is useful, but it has to be done as part of the same decision-making process as lower-level controls. - We haven't seen the long tail of humanoid robotics, but Tesla has seen the long tail of self-driving, where high and low-level decisions have to be jointly made at a pretty high framerate. - Optimus's architecture is designed in a similar way, where there's a hierarchy but it's all running as part of the same model and the latencies involved in decision making are well modeled. - This architecture will scale quite well with humanoid robots. - The distinction of the decision-making levels is only in the developer's mind. For the model, it's a continuous space of decision making, where there are dials available to make them more fine or coarse. - Humanoids have more sensor modalities and higher degrees of freedom compared to self-driving, but the fundamental constraints remain the same: you need to make real-time decisions. There's obviously a hierarchy to these control signal outputs, but the lowest frequency cannot be too low, because the safety of the robot cannot depend upon things running at very low frequencies.

Ashok Elluswamy, Tesla's AI lead, during a GTC discussion, highlighting the fundamental similarity in AI approaches for self-driving cars and humanoid robots: - Hierarchical decision making is useful, but it has to be done as part of the same decision-making process as lower-level controls. - We haven't seen the long tail of humanoid robotics, but Tesla has seen the long tail of self-driving, where high and low-level decisions have to be jointly made at a pretty high framerate. - Optimus's architecture is designed in a similar way, where there's a hierarchy but it's all running as part of the same model and the latencies involved in decision making are well modeled. - This architecture will scale quite well with humanoid robots. - The distinction of the decision-making levels is only in the developer's mind. For the model, it's a continuous space of decision making, where there are dials available to make them more fine or coarse. - Humanoids have more sensor modalities and higher degrees of freedom compared to self-driving, but the fundamental constraints remain the same: you need to make real-time decisions. There's obviously a hierarchy to these control signal outputs, but the lowest frequency cannot be too low, because the safety of the robot cannot depend upon things running at very low frequencies.

The Humanoid Hub

670,485 views • 7 months ago

Let’s talk about Zipline’s test sites and who we’re hiring for 🧵 They are where elite talent meets 24/7 large scale high volume testing. They’re the engine that helped scale Zipline to the largest autonomous delivery service on earth, 5,000+ autonomous trips around the world and we’re just getting started. Zipline's testing in 2025: 315,000+ test flights 🚀 35,000+ flight hours (480+ straight days of nonstop flying) 🙂‍ 3,000+ flights per day 📈 Our test sites are built to push our system to the max so that we de-risk tomorrow. Each one tackles a different brutal edgecase to make sure the system’s reliability is bulletproof. Our 'engineering test site' in the video I posted is reconfigured every few weeks: new obstacles are added, new flight apps reviewed, new edge cases are tested in any weather condition. Every new software build deploys here first, into live airspace. Our other sites are placed around the U.S. and are focused on testing in severe conditions that ground most if not all other forms of transport. They operate in scorching heat of up to 125 degrees, high-altitudes, intense rainstorms, 60+ mph winds, hail, sleet, and extreme cold-weather, down to -20F. Heavy ice and snow accumulation on propulsion and sensors is the norm. We aggressively chase these conditions in test so we dominate when it really matters. We are now developing 5+ new test sites, each one dialed in to push even more extreme weather and edge conditions. What I am especially proud of is that our safety has kept improving even as flight volume, complexity and environmental hostility increases. We’ve been able to test and develop at a scale that’s unprecedented in aviation history because our teams own a 100% fully vertical tech stack, built hand in hand with our flight and application software teams. We have to have the most robust airspace and fleet management tools on earth because Zipline will soon operate the largest fleet of aircraft on earth. Now to the fun part, we're hiring! DM me or send me an email to marcusZipline.com if you want to join Zipline Test Operations! We're hiring Flight Test Operators, Test Site Operators, Flight Test Electricians, Flight Test Construction Staff, Flight Test Project and Program Managers, Flight Test Engineers, Flight Test Safety Managers, Flight Test Security and, most importantly, Flight and Fleet Application Software Engineers. Please, cut to the chase. We have zero requirements on degrees, formal education or tenure. What counts is what you can do, merit, and what you’ve shown you’re capable of. Highlight that. Let's go!

Let’s talk about Zipline’s test sites and who we’re hiring for 🧵 They are where elite talent meets 24/7 large scale high volume testing. They’re the engine that helped scale Zipline to the largest autonomous delivery service on earth, 5,000+ autonomous trips around the world and we’re just getting started. Zipline's testing in 2025: 315,000+ test flights 🚀 35,000+ flight hours (480+ straight days of nonstop flying) 🙂‍ 3,000+ flights per day 📈 Our test sites are built to push our system to the max so that we de-risk tomorrow. Each one tackles a different brutal edgecase to make sure the system’s reliability is bulletproof. Our 'engineering test site' in the video I posted is reconfigured every few weeks: new obstacles are added, new flight apps reviewed, new edge cases are tested in any weather condition. Every new software build deploys here first, into live airspace. Our other sites are placed around the U.S. and are focused on testing in severe conditions that ground most if not all other forms of transport. They operate in scorching heat of up to 125 degrees, high-altitudes, intense rainstorms, 60+ mph winds, hail, sleet, and extreme cold-weather, down to -20F. Heavy ice and snow accumulation on propulsion and sensors is the norm. We aggressively chase these conditions in test so we dominate when it really matters. We are now developing 5+ new test sites, each one dialed in to push even more extreme weather and edge conditions. What I am especially proud of is that our safety has kept improving even as flight volume, complexity and environmental hostility increases. We’ve been able to test and develop at a scale that’s unprecedented in aviation history because our teams own a 100% fully vertical tech stack, built hand in hand with our flight and application software teams. We have to have the most robust airspace and fleet management tools on earth because Zipline will soon operate the largest fleet of aircraft on earth. Now to the fun part, we're hiring! DM me or send me an email to marcusZipline.com if you want to join Zipline Test Operations! We're hiring Flight Test Operators, Test Site Operators, Flight Test Electricians, Flight Test Construction Staff, Flight Test Project and Program Managers, Flight Test Engineers, Flight Test Safety Managers, Flight Test Security and, most importantly, Flight and Fleet Application Software Engineers. Please, cut to the chase. We have zero requirements on degrees, formal education or tenure. What counts is what you can do, merit, and what you’ve shown you’re capable of. Highlight that. Let's go!

Marcus Mueller

312,708 views • 8 months ago

Lindsey could have you laughing so hard your ribs would hurt. He had a self-deprecating wit, a twinkle in his eye, and the rare ability to fight fiercely while earning the respect and affection of every senator around him.

Lindsey could have you laughing so hard your ribs would hurt. He had a self-deprecating wit, a twinkle in his eye, and the rare ability to fight fiercely while earning the respect and affection of every senator around him.

Ted Cruz

154,986 views • 4 months ago

Detection. Something very important in a game like this. There are several ways to approach this mechanic and we started basic. The NPCs have a “cone” of vision in front of them, and depending on how closer you are to the middle of that cone and also how close you are to the front of the NPC, the faster you are detected. It isn’t perfect, by any stretch of the imagination, but it is working and serves it’s purpose. As we advance in development this will both evolve and be refined from its current “primitive” state. Surprisingly hard to get something working that actually feels both fair and fun! I’d love to eventually have them “look” from their eyesight, so height matters. Ignore the ground flashing, unity editor is weird like that.

Detection. Something very important in a game like this. There are several ways to approach this mechanic and we started basic. The NPCs have a “cone” of vision in front of them, and depending on how closer you are to the middle of that cone and also how close you are to the front of the NPC, the faster you are detected. It isn’t perfect, by any stretch of the imagination, but it is working and serves it’s purpose. As we advance in development this will both evolve and be refined from its current “primitive” state. Surprisingly hard to get something working that actually feels both fair and fun! I’d love to eventually have them “look” from their eyesight, so height matters. Ignore the ground flashing, unity editor is weird like that.

ShyGirl Team 🌸🎮

39,998 views • 6 months ago

#WATCH | Delhi | On Drones and sixth-generation fighter aircraft programmes, Defence Secretary Rajesh Kumar Singh says, "On the sixth-generation fighter aircraft programmes, there are multi-country programmes ongoing, one involving the UK, Japan, and Italy, and another in which France and Germany were together, although Germany is now no longer part of it. We had expressed interest in both of those groups of countries because all of them are friendly foreign countries for us. We are veering around to the idea that, now that Germany has left, perhaps the French programme, since we have a large inventory of French aircraft, is something we should examine. We are pursuing discussions with the French, and we'll see how that goes, but we have expressed our interest. On drones, the number and size of the orders are impossible for me to give because a lot of those orders are placed under delegated powers at the command and Vice Chief levels. At the centralised level, there are the larger drone orders. Of course, you are aware of the 31 MQ-9B drones. That's about $3.5 billion, and that order was placed about two years ago. Those are the HALE-class drones. A lot of smaller orders have been placed by all three services under delegated powers. After Operation Sindoor, under emergency procurement as well, significant orders for both drones and counter-drone systems have been placed by the services. We are in the process of trying to ensure that, since drones have a very short technological cycle, there is no point in undertaking multi-year procurement for them. We will try to procure these mostly through the fast-track process..."

#WATCH | Delhi | On Drones and sixth-generation fighter aircraft programmes, Defence Secretary Rajesh Kumar Singh says, "On the sixth-generation fighter aircraft programmes, there are multi-country programmes ongoing, one involving the UK, Japan, and Italy, and another in which France and Germany were together, although Germany is now no longer part of it. We had expressed interest in both of those groups of countries because all of them are friendly foreign countries for us. We are veering around to the idea that, now that Germany has left, perhaps the French programme, since we have a large inventory of French aircraft, is something we should examine. We are pursuing discussions with the French, and we'll see how that goes, but we have expressed our interest. On drones, the number and size of the orders are impossible for me to give because a lot of those orders are placed under delegated powers at the command and Vice Chief levels. At the centralised level, there are the larger drone orders. Of course, you are aware of the 31 MQ-9B drones. That's about $3.5 billion, and that order was placed about two years ago. Those are the HALE-class drones. A lot of smaller orders have been placed by all three services under delegated powers. After Operation Sindoor, under emergency procurement as well, significant orders for both drones and counter-drone systems have been placed by the services. We are in the process of trying to ensure that, since drones have a very short technological cycle, there is no point in undertaking multi-year procurement for them. We will try to procure these mostly through the fast-track process..."

ANI

22,036 views • 1 month ago

Sam Altman: “By the end of this year, for $100–$1,000 of inference and a good idea, you’ll be able to create software that would have taken teams of people a year to do. That magnitude of economic change is very hard to wrap your head around.”

Sam Altman: “By the end of this year, for $100–$1,000 of inference and a good idea, you’ll be able to create software that would have taken teams of people a year to do. That magnitude of economic change is very hard to wrap your head around.”

vitrupo

579,330 views • 9 months ago