MurmurCast
DiscussionNews

The $1/Hour Worker: Four Robotics CEOs on Humanoids at Home, China's Threat, and the End of Dangerous Jobs

All-In with Chamath, Jason, Sacks & Friedberg1h 8m

Four leading robotics CEOs discuss the current state of humanoid and quadrupedal robots, their real-world deployment across industries, competition from China, and the timeline for achieving general-purpose robotics at scale. They emphasize that recent breakthroughs in AI perception and language models have shifted robotics from research to practical applications, with significant implications for labor, manufacturing, and national security.

Summary

The transcript captures interviews with four major figures in robotics at the Makina conference in Paris: Dr. Peter Funkhauser (ANYbotics), Bernt Borgvik (1X), Amanda McMaster (Boston Dynamics), and Professor Jonathan Hurst (Agility Robotics). Each discusses their company's approach to deploying robots in real-world environments.

Funkhauser explains that ANYbotics' four-legged Anymal robot is deployed for industrial inspection of critical infrastructure like pipelines, power plants, and offshore facilities. The robot's value comes not from labor replacement but from superhuman perception—equipped with thermal cameras, acoustic sensors, and gas detectors—allowing it to detect problems that humans cannot. The company intentionally sources components outside China for data security reasons, viewing Chinese competition as focused on hardware flashiness rather than the full solution stack of autonomy, integration, and data trust that customers require.

Borgvik presents 1X's Neo humanoid as both a consumer product and a platform. Pre-orders sold 10,000 units at $500/month in the first few days, with delivery promised in 2026. He emphasizes that the real breakthrough isn't the hardware but the ability to train on internet video data by making the robot anthropomorphic. 1X has built a World Model Lab that uses a data pyramid: teleoperation data at the top, sensor-enabled human demonstrations, egocentric video, and finally general internet video at the base. This approach leverages the massive scale of YouTube and human video because the Neo's human-like form factor makes that data transferable to the robot. Borgvik believes hard takeoff—where robots become self-sufficient in manufacturing, mining, and data centers—will occur in 3-10 years.

McMaster describes Boston Dynamics' transition from research to commercial deployment. Spot, the quadrupedal robot, has 500+ customers across 46 countries and is the most deployed autonomous robot globally. It performs industrial inspection, acoustic monitoring, and security tasks, with ROI achieved through preventing equipment downtime valued in hundreds of thousands per hour. Atlas, the humanoid, represents the next phase. Boston Dynamics maintains 100% U.S. manufacturing and opposes weaponization, though McMaster acknowledges that if China deploys armed robots, the calculus may change. She stresses focus on industrial applications rather than military use, but hints that government pressure could alter priorities.

Hurst provides historical context, explaining why this moment differs from previous AI winters. The key breakthrough is perception—language models trained on internet data now allow robots to understand their environment without hard-coding. However, robot control remains unsolved; there is no equivalent training dataset for motor commands and physics. Hurst advocates for multiple approaches: learning from demonstration, teleoperation, world models, and sim-to-real transfer. He emphasizes that robots will enable knowledge work by freeing humans from dull, dirty, and dangerous tasks, mirroring how automation improved quality of life in the past century. Agility's Digit V5, releasing later in the year, marks a milestone—the first humanoid that can operate safely without physical barriers between robots and humans in warehouses.

Common themes across all interviews: (1) AI perception breakthroughs, especially from large language models, are the fundamental inflection point; (2) Real-world deployment ROI is measured in avoided downtime, superhuman capabilities, and labor augmentation rather than crude labor replacement; (3) China poses both technological and geopolitical threats through IP theft and cheaper hardware, though full solutions require more than hardware; (4) Humanoid form factors are justified by versatility across multiple tasks and the ability to train on human video; (5) Scaling from hundreds to thousands of deployed robots requires solving safety, reliability, and integration challenges; (6) Business models vary between CapEx sales and robot-as-a-service, with pricing eventually approaching hourly labor rates ($1-3/hour expected long-term); (7) Military applications remain contentious but inevitable if adversaries deploy armed robots; (8) The timeline to general-purpose autonomy ranges from 3 to 10 years depending on the speaker's optimism.

About this episode

<p dir="ltr">(0:00) Intro: Humanoids, Robots, & AI+</p> <p dir="ltr">(0:57) ANYbotics' Dr. Péter Fankhauser: Why ANYbotics Bet the Company on Four-Legged Robot Dogs, Not Humanoids</p> <p dir="ltr">(13:18) Dr. Péter Fankhauser: China's Armed Robot Dogs Are "Stupid and Risky" — Inside the Terminator Debate</p> <p dir="ltr">(15:01) NEO's Bernt Børnich: Neo Ships in 2026 and Becomes an Open Robot Platform</p> <p dir="ltr">(22:18) Bernt Børnich: Robots Building Robots in Just 3 Years — "Hard Takeoff"</p> <p dir="ltr">(33:45) Boston Dynamics' Amanda McMaster: Spot Is Now the Most Deployed Robot on Earth — Atlas is the Most Capable</p> <p dir="ltr">(42:36) Amanda McMaster: Chinese Robotics Security Risks</p> <p dir="ltr">(47:14) Agility Robotics' Jonathan Hurst: Why "This Time Is Different" for Humanoids After 100 Years of False Starts</p> <p dir="ltr">(58:45) Jonathan Hurst: When Do Robots Outnumber Humans on the Factory Floor?</p> <p dir="ltr">Thanks to our partners for making this possible!</p> <p dir="ltr">AppLovin Ads is AppLovin's AI advertising platform reaching over a billion daily active users across mobile games. Full-screen video ads with a 35-second median watch time. Advertisers are profitably spending hundreds of thousands of dollars a day. AppLovin Ads is now open to all advertisers. Sign up at <a href="https://applovin.com/ALLIN">https://applovin.com/ALLIN</a> </p> <p dir="ltr">If your work depends on conversations — meetings, deal flow, interviews, customer calls — Plaud helps you capture and organize everything with highly accurate AI-generated notes that are not just simple summaries, but also highlight pain points, key decisions, next steps, and customizable summary templates. </p> <p dir="ltr">Check out Plaud at <a href="https://www.plaud.ai/allin">https://www.plaud.ai/allin</a> and use code ALLIN for up to 20% off! Which is also available on Amazon: <a href="https://amzn.to/43URLff">https://amzn.to/43URLff</a> (Code: ALLIN20X)</p> <p dir="ltr">Follow the besties:</p> <p dir="ltr"><a href="https://x.com/chamath">https://x.com/chamath</a></p> <p dir="ltr"><a href="https://x.com/Jason">https://x.com/Jason</a></p> <p dir="ltr"><a href="https://x.com/DavidSacks">https://x.com/DavidSacks</a></p> <p dir="ltr"><a href="https://x.com/friedberg">https://x.com/friedberg</a></p> <p dir="ltr">Follow on X:</p> <p dir="ltr"><a href="https://x.com/theallinpod">https://x.com/theallinpod</a></p> <p dir="ltr">Follow on Instagram:</p> <p dir="ltr"><a href="https://www.instagram.com/theallinpod">https://www.instagram.com/theallinpod</a></p> <p dir="ltr">Follow on TikTok:</p> <p dir="ltr"><a href="https://www.tiktok.com/@allin">https://www.tiktok.com/@allin</a><strong> </strong></p> <p dir="ltr">Follow on LinkedIn: </p> <p dir="ltr"><a href="https://www.linkedin.com/company/allinpod">https://www.linkedin.com/company/allinpod</a></p> <p dir="ltr">Intro Music Credit:</p> <p dir="ltr"><a href="https://rb.gy/tppkzl">https://rb.gy/tppkzl</a></p> <p dir="ltr"><a href="https://x.com/yung_spielburg">https://x.com/yung_spielburg</a></p> <p dir="ltr">Intro Video Credit:</p> <p dir="ltr"><a href="https://x.com/TheZachEffect">https://x.com/TheZachEffect</a></p>

Key Insights

  • ANYbotics argues that customer value comes from superhuman perception (thermal, acoustic, gas sensors) enabling detection of problems humans cannot perceive, not from replacing human labor in inspection roles.
  • 1X's core innovation is making the Neo anthropomorphic to enable training on internet video data, which Borgvik claims creates a 'catch-22' that only humanoid form factors can solve due to the massive scale of available human video.
  • Boston Dynamics maintains 100% U.S. manufacturing and component sourcing, explicitly avoiding China, citing both IP theft concerns and customer sensitivities around sending critical infrastructure data to Chinese servers.
  • Spot's ROI is primarily driven by preventing equipment downtime (valued at hundreds of thousands per hour), making the robot's cost economically justified despite being priced at $100,000-$300,000.
  • Four-legged robots became the deployed standard in quadrupedal robotics because wide footprints provide stability in dangerous, slippery, and unpredictable outdoor environments that are the primary use case.
  • The fundamental blocker for robot control—as opposed to perception—remains unsolved; there is no internet-scale training dataset for motor commands, forces, and physics like there is for vision and language.
  • Humanoid robots are justified by versatility across multiple tasks and the ability to train on human video, not because two-legged bipedalism is inherently optimal.
  • 1X's World Model Lab operates on a data pyramid where general internet video is orders of magnitude larger than teleoperation data, making it the only scalable path to training general models.
  • Multiple CEOs predict that if China deploys armed humanoid robots, U.S. companies will face pressure to weaponize their robots or face potential government takeover or contract loss.
  • The breakthrough difference from previous robotics hype cycles is not better hardware but perception: LLMs trained on internet data eliminate the need to hard-code object recognition and semantic understanding.
  • Robot-as-a-service pricing is expected to converge toward $1-3 per hour of operation, creating 90x cost compression compared to current human factory workers ($20-40/hour), justified over 5-year lifecycles with 20+ operating hours daily.
  • Agility's Digit V5 represents a safety milestone—the first humanoid that can operate without physical barriers in human-shared warehouse spaces, enabling broader deployment beyond isolated work cells.

Topics

AI perception and large language models as inflection pointReal-world robot deployment and ROI measurementChina as technological and geopolitical threatHumanoid vs. quadrupedal form factorsData generation through teleoperation and human demonstrationsBusiness models: CapEx vs. robot-as-a-serviceSafety and reliability requirements for industrial deploymentTimeline to general-purpose autonomy and hard takeoffLabor augmentation vs. labor replacementWorld models and sim-to-real transferMilitary applications and weaponization concernsSupply chain and manufacturing sovereignty

Transcript

Hey everybody, it's your boy J.Cal. I'm here in Paris, France at a conference called Makina. Basically AI in the real world. Pardon my robot. Thanks for tuning in and let's get started. App Lovin' started with an $8 domain and no VC funding and became one of the largest ad platforms in the world. Now that same engine powers Applovin ads for e-commerce. Your ads run inside mobile games reaching over a billion people with full screen, distraction-free attention. The platform finds buyers and optimizes for profit. You set the target, it does the rest. One cookware brand went from 4 million to 16 million dollars turned profitable and is on pace for 80 million this year visit applovin.com…

Full transcript available for MurmurCast members

Sign Up to Access
View original source →

More from All-In with Chamath, Jason, Sacks & Friedberg

Get AI summaries like this delivered to your inbox daily

Get AI summaries delivered to your inbox

MurmurCast summarizes your YouTube channels, podcasts, and newsletters into one daily email digest.