Jul 29, 2026 · 1h 8m · allin
The $1/Hour Robot Is Coming: Four Industry Leaders Explain What’s Next
Jason Calacanis · 27m spoken Bernt Børnich · 10m spoken Professor Jonathan Hurst · 10m spoken Dr. Peter Fankhauser · 6m spoken Amanda McMaster · 5m spoken
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Hosted by Jason Calacanis live from the Machina conference in Paris, this episode of the All-In Podcast features interviews with leading robotics executives from ANYbotics, 1X, Boston Dynamics, and Agility Robotics to explore the rapid commercialization, economic models, AI architectures, and societal impact of real-world physical automation.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 44.9% of the talking time here. How this is scored →
The hosts as informed peer 4.7 Guest teaching 3.6 Guest disagreement 1.6 The hosts pushing back 3.2
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 45:24 Resistance to Military Weaponization Mandate
McMaster firmly resists Jason's insistence that Boston Dynamics will inevitably be forced by government mandates to build armed military robots, asserting their deliberate anti-weaponization focus.
Hardest push from the hosts ▶ 45:40 Insisting Military Deployment is InevitableJason refuses the guest's peaceful framing and aggressively insists that state competition with China and presidential mandates will force Boston Dynamics to arm their platforms.
Biggest teaching moment ▶ 51:13 LLM Data Gap in Motor ControlProfessor Hurst reframes Jason's assumption that internet LLM data solves robot control for free, explaining that text datasets completely lack motor torque commands and physical dynamics.
The host holds their own ▶ 58:02 Amortized $1/Hour Unit Economics CalculationJason demonstrates sharp financial analysis by breaking down 20-hour daily duty cycles over a 5-year lifespan to prove a $1/hour amortized cost against $20-40/hour human labor.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| ANYbotics Interview: Four-Legged Inspection Robots vs. Humanoids | 3 | 3 | 1 | 3 | Jason probes why the quadrupeds became the industry standard form factor and playfully challenges Fankhauser on why coffee shops shouldn't use centaur formats instead of humanoids. Fankhauser grounds the discussion in physical reality, explaining how four legs offer stability on slippery or uneven industrial surfaces. | |
| ANYbotics Interview: Industrial Inspection, Economics, and Environment Durability | 4 | 3 | 1 | 2 | Jason demonstrates understanding of high-power compute requirements and battery duty cycles in field robotics. Fankhauser educates on the ROI metrics of preventing million-dollar facility downtimes and operating in explosive ATEX environments. | |
| ANYbotics Interview: Supply Chain Sourcing and Chinese Competition | 5 | 4 | 1 | 3 | Jason raises sharp questions about IP theft and undercutting from Chinese manufacturers. Fankhauser reframes the threat, noting that while Chinese hardware does impressive backflips, enterprise clients buy full-stack trust, ISO cybersecurity, and software integration. | |
| ANYbotics Interview: Defense Policy, NATO, and Anti-Weaponization Stance | 4 | 4 | 3 | 4 | Jason pushes Fankhauser on whether ANYbotics could retool its inspection software for NATO military applications. Fankhauser pushes back on the premise, explaining that defense applications require completely different sub-millisecond autonomy stack architecture. | |
| 1X Interview: The NEO Humanoid Robot and Consumer Platform | 4 | 3 | 1 | 2 | Jason draws parallels between early humanoid pre-orders and tech launches like Apple Vision Pro or Tesla deposits. Børnich outlines 1X's deployment timeline and subscription pricing model while positioning NEO as an open developer platform. | |
| 1X Interview: Teleoperation, Data Collection, and Human Video Training | 5 | 5 | 2 | 3 | Jason cites a real-world example of remote cashiers in Manila to explore teleoperation business models. Børnich clarifies that teleop fails to leverage full hardware capabilities due to lack of haptic fidelity, explaining why human-conforming hardware unlocks training on raw video data. | |
| 1X Interview: Embodied AGI, Hard Takeoff, and Labor Abundance | 5 | 4 | 2 | 4 | Jason connects AI agent loops and physical wingspan concepts to embodied robotics. Børnich delivers a bold prediction of a hard takeoff within 3 to 10 years, which Jason gently presses on as ultra-optimistic. | |
| Sponsor Message: PLAUD AI Audio Intelligence | 4 | 1 | 0 | 1 | Jason opens with a sponsor message for Plaud AI before expertly recapping Boston Dynamics' full corporate acquisition history from Google to SoftBank to Hyundai, which interim CEO Amanda McMaster validates. | |
| Spot Use Cases, Hardware Pricing, and Business Models | 4 | 3 | 1 | 2 | Jason asks detailed questions about hardware pricing and deployment scale across industrial markets. McMaster explains why industrial clients prefer CapEx models for Spot while Atlas humanoids will likely utilize Robot-as-a-Service. | |
| Robot Reliability, Continuous Operation, and Battery Swapping | 5 | 4 | 2 | 5 | Jason brings up fully baked union labor rates ($40-60/hr) and directly challenges McMaster on whether avoiding 'labor replacement' terminology is just corporate diplomacy. McMaster reframes the narrative toward risk mitigation and preventing catastrophic air leaks. | |
| Technical Architecture: Onboard Compute vs. Cloud Reasoning | 5 | 4 | 2 | 3 | Jason probes the division between onboard edge processing and cloud reasoning. McMaster introduces a 'two-brains' architectural model, distinguishing real-time physical balance on-board from high-level semantic reasoning in the cloud. | |
| Defense Policy, Anti-Weaponization Stance, and Non-Lethal Defense Use | 6 | 3 | 4 | 7 | Jason forcefully asserts that Boston Dynamics will inevitably be compelled by the U.S. government to build armed defense systems as China deploys weaponized quadrupeds. McMaster holds her ground on their anti-weaponization policy, insisting on focusing on high-ROI industrial applications. | |
| Boston Dynamics' Growth, Recruitment, and Interview Conclusion | 4 | 3 | 1 | 1 | Jason jokes about secret military deployments before transitioning to Professor Jonathan Hurst. Hurst provides historical context on how robotics shifted over 20 years from academic research labs into commercial deployment. | |
| Perception as the Inflection Point: AI and LLM Integration | 5 | 4 | 1 | 3 | Jason highlights how multimodal vision-language models give robots semantic understanding of everyday objects for free. Hurst agrees, identifying solved perception as the key inflection point moving robotics toward general utility. | |
| Training Motor Control: Teleoperation, World Models, and Sim-to-Real | 5 | 5 | 3 | 4 | Jason draws an analogy between robotic motor control and recursive learning loops like AlphaGo. Hurst corrects the premise by pointing out that internet data lacks motor torque commands, making sim-to-real transfer and physical practice essential. | |
| Fleet Learning, Digit Hardware, and Removing Safety Cages | 5 | 4 | 1 | 3 | Jason highlights the network effects of fleet learning via Wi-Fi. Hurst reveals Digit V5's milestone capability to balance and step out of safety cages directly alongside human workers without physical barriers. | |
| Industrial Safety Engineering and Cageless Human Collaboration | 7 | 4 | 2 | 4 | Jason demonstrates strong financial domain expertise by calculating the amortized cost of a robot working 20 hours a day over 5 years (~40,000 hours) down to ~$1/hour against $20-40/hour human labor. Hurst validates the economic potential while politely demurring on exact client pricing. | |
| Macroeconomic Drivers: Labor Shortages and Progressive Automation | 5 | 4 | 1 | 4 | Jason frames the necessity of robotics through macroeconomic drivers such as stagnant population growth and labor shortages. Hurst clarifies that while lights-out specialized automation works for fixed warehouses, humanoids excel in dynamic environments built for people. | |
| Doorstep Delivery Roadmaps and Eliminating Dull, Dirty, Dangerous Work | 5 | 4 | 1 | 3 | Jason asks why humanoids shouldn't have four or six arms like General Grievous. Hurst explains engineering first principles, demonstrating why two arms hit the optimal trade-off between task utility and system complexity. |
Statements from this episode (49)
Fankhauser: Four-legged robots are the ideal format for industrial facilities
“In the facilities that we work, four leg stability, there's enough space to go around. It's the perfect format.”
Fankhauser: Stopped industrial assets lose hundreds of thousands in revenue hourly
“These assets, if they stop, they lose revenues in hundreds of thousands per hour.”
Fankhauser: Some ANYbotics customers run robot inspection missions 40 times daily
“Some of our customers run these missions 40 times a day.”
Fankhauser: Industrial robotics customers want data and insights, not robots
“They don't care about the robot. Actually, they don't even want the robot. They want the data. They want the insights. The robot is a means to an end to collect the data precisely.”
Fankhauser: ANYbotics built a spark-proof robot certified for explosive environments
“Most importantly, we have a robot now that goes into explosive atmospheres, which is, you know, in oil and gas and chemicals, methane in the air, you're not allowed to, you know, create a spark. So we built a special robot that's guaranteed not to create a spa…”
Fankhauser: Chinese robots offer impressive hardware but lack complete solutions
“If you look at the robot from China today, that device is a piece of hardware that can walk. Beautifully, great engineering. Love it. Do backflips. But they're not solving the problem. Our customers don't compare a platform to the full solution that we have.”
Fankhauser: ANYbotics does not build and has no plans to build military products
“Today we're not doing it, not intend to do it, right?”
Fankhauser: Commercial inspection robotics autonomy cannot be easily repurposed for military use
“No, autonomy is very different, right? So for example, we do autonomy, you have time to set up a robot, and it does inspections, all of that. In military, it's about millisecond being in, right? Remote controlled human in the loop, different communications, di…”
Fankhauser: Videos of weaponized quadruped robots are early demonstrations, not active deployments
“I think these are all very early tests. If I look at those videos, these are demonstrations. I've not seen these types of robot active drones.”
Fankhauser: ANYbotics signed an open letter with Boston Dynamics opposing robot weaponization
“Four years ago, we wrote a letter together with our friends at Boston Dynamics and others, right, who condemned the weaponization Of robots for exactly that reason, that as engineers, we don't want to see it being used, and we think it's just dangerous and ris…”
Børnich: 1X will ship initial NEO humanoid robots in 2026
“So we will ship in twenty-twenty-six. Now, expectation managing here, it'll be slow in the beginning. We want to do it right. But there will be a handful of customers that get their NIO in twenty-twenty-six”
Børnich: 1X will launch the NEO humanoid robot as a developer platform
“We are going to allow a lot of people to build on NIO. So we are also launching NIO as a platform.”
Børnich: 1X sold out 10,000 NEO pre-orders in first few days
“We sold out the first 10 K in the first few days.”
Børnich: 1X expects to ship a fully autonomous NEO experience
“I am very happy to say that I think we will actually be able to ship something that's very close to full autonomy, which we did not want to promise when we launched this because it was too early, but, and I'm not going to fully promise it yet, but the way it's…”
Børnich: No general AI model solves everything for robotics yet
“Currently, if you look at where this, where the field is, there is no one general model that solves everything for robotics. It's not there yet.”
Børnich: Humanoid physical form enables robot AI training on internet video
“If you get the robot to be similar enough to a human, Then you can train on all the available video data out there of humans.”
Børnich: Robotics 'hard takeoff' with autonomous self-replication is under a decade away
“I am extremely sure that we're less than a decade away from hard takeoff. And when I say hard takeoff, I mean robots building the robots, the data centers, the chip fabs, Doing the mining and refining, actually a true abundance of labor, a self-sufficient syst…”
McMaster: Boston Dynamics is focused on real-world deployment, not R&D
“Yeah, we are big on deploying robots, so it's no longer an AI lab. It's not a research and development company anymore. We're now focused on real world deployment.”
McMaster: Boston Dynamics has over 500 customers across 46 countries
“At this point we have over 500 customers, over 46 countries.”
McMaster: Spot is the most deployed mobile autonomous robot on Earth
“It is the it is the mobile autonomous robot that's used more than any other on the planet right now.”
McMaster: Industrial robotics customers require an ROI under two years
“And customers need to see your ROI in under two years.”
Boston Dynamics will likely offer Atlas as robot-as-a-service
“We'll be doing a, probably a robot as a service model, likely with Atlas.”
McMaster: Boston Dynamics' Spot robot costs $100k to $300k fully loaded
“Ranges anywhere between, you know, a 100,000 dollars for the base robot, all the way up to 300,000 when we're fully loaded with services, integration deployed.”
Boston Dynamics: Spot robot achieves over 3,000 hours between human interventions
“We think about in terms of mean time between intervention, and we're at over 3000 hours.”
Boston Dynamics: Atlas humanoid robot autonomously swaps its own batteries
“No Atlas does it itself. So Atlas will have two batteries, so it turns its torso around, and you replace one and put it with the other one, it always has a backup.”
McMaster: Boston Dynamics focuses on labor augmentation over replacement for Spot
“We haven't necessarily looked at labor replacement for a spot. While, while that is a metric you might look at, we thought about, you know, how do we bring spots in there to augment human labor?”
McMaster: Human workers often skip assigned routine inspection tasks
“Humans weren't doing the task, even if they were tasked with it, they weren't actually doing it.”
McMaster: Labor replacement will inevitably be used as a robotics ROI metric
“There's going to be an element of labor replacement for this as a metric, because it's easy.”
McMaster: Boston Dynamics splits robot compute between onboard physical control and cloud reasoning
“We think about two brains, right? It's my simplified version of telling the stories. There's two brains. There's the brain that controls the physicality of the robot, which is what Boston Dynamics is known for. You know, the dynamic movement, reliability, the …”
McMaster: Chinese quadruped robots in US are leaking data to China
“We've already heard about leaks that are happening with some of the quadrupeds that you're seeing in the United States and being back channeled back to China.”
McMaster: Weaponizing robots is a distraction from Boston Dynamics' core focus
“For what we're trying to do right now in industrial use cases, it's a distraction for our business.”
Calacanis: Humanoid robots could reach 1:1 ratio with humans in 20-30 years
“That I think we can both agree will be, maybe in the next 2030 years, one-to-one with humans on the planet.”
Hurst: AI perception for robotics is all but solved
“Perception is incredibly difficult, and the fact that perception is all but solved at this point is a really, really huge inflection point.”
Hurst predicts robots will soon perform general, useful tasks
“AI is really enabling that much more broad you know, context awareness for these robots, so people can see that this is going to be useful, generally doing many useful things very soon.”
Hurst: Internet data does not exist for physical robot control
“These language models are trained off of the entire data on the internet. And that data does not exist for robot control.”
Hurst: Human teleoperation has a hard limit for optimal robot training
“You can teleoperate the robot and start to train from that data. You can give it animation input or motion capture input or any number of different things, but that's also got a real hard limit. Because a person controlling a robot is not really getting to wha…”
Hurst: Simulation gaps make physical real-world robot practice indispensable
“But there's always a massive sim to real gap. Things aren't simulated perfectly. And then, you know, as you pick up something in the real world and there's wave dynamics and there's condensation on the glass and the dynamics of the robot are not perfectly mode…”
Hurst: Agility Robotics will release barrier-free Digit V5 this year
“And Digit V-Five, which is coming out later this year, is the first time that a humanoid robot, a robot which is balancing, can step out of a work cell and does not need a physical barrier between the robot and the person to maintain safety in this warehouse.”
Hurst: Agility's Digit v5 robot operates 20 hours per day
“20 out of 24 hours for our Digi V-Five robot.”
Hurst: Humanoid robots will reach $20k-$40k at 100k unit scale
“They will at some point. It's gonna need to go through the scaling and have 100,000 robots out there before that actually is real.”
Agility Robotics offers both CapEx and Robot-as-a-Service pricing
“We do both. We do a capex for customers that prefer that. We also do robot as a service for customers that prefer that.”
Hurst: Increased automation is the only way to grow US GDP
“How do we, in the United States anyway, how do we build our GDP? It's not a growing population. It's increased efficiency and capability, and the only way we could do that is more and more automation.”
Hurst: Humanoid robots belong in human environments, not lights-out factories
“At that point, it's probably specialty automation doing those things, because it's just twenty-four-seven doing it, and a humanoid doesn't make sense. It's not the most efficient thing for that specific task. A humanoid is useful for walking into human environ…”
Calacanis: Autonomous vehicles and robots will deliver packages to doorsteps
“Most people are incapable at this point of imagining a Waymo robo-taxi, an Uber self-driving car, and a robot getting out and bringing the packages to your doorstep. That's gonna happen.”
Jonathan Hurst: Agility Robotics demonstrated doorstep package delivery seven years ago
“You know what, that was one of our very first use cases that we explored with Ford. And there's a nice video online of our very first digit robot getting out of a vehicle, walking up to someone's front porch, and dropping a package there. Stairs and everything…”