Video · 2026-07-29 · 1h 9m · 6 moments

The $1/Hour Robot Is Coming: Four Industry Leaders Explain What’s Next

✦ AI generated

GUEST Dr. Peter Funkhouser GUEST Bernt Børnich GUEST Amanda McMaster GUEST Professor Jonathan Hurst

timeline · colored by role

Claim◆

Industrial inspection robots are not about labor replacement — they achieve superhuman perception that prevents catastrophic downtime, paying for themselves in hours saved.

Peter Funkhouser argues that four-legged inspection robots deliver value not by replacing workers but by detecting gas leaks, overheating, and other hazards humans can't sense, preventing hundreds of thousands in hourly downtime costs.

transcript

Dr. Peter Funkhouser: For us, it's not about labor replacement, right? It's what can we do better? What can we do? Super human. Inspection is a great example. Our eyes and ears don't perceive all the signals. Micro gas leakages, temperature equipment overheating with the cameras on the robot, thermal cameras, acoustic, you know, microphones, gas concentrations and all of that. We pack it full of sensors and AI and you can go way beyond what a human can do. So, the monetary benefit is avoiding downtime. These assets, if they stop, they lose revenues in hundreds of thousands per hour. So every minute, every hour we can save them essentially pays for the robots.

gives example · 1

Mechanism◆

The only path to true embodied intelligence is to build a robot so physically similar to a human that it can be pre-trained on all internet video data, which is orders of magnitude larger than any robotics-specific data set.

Bernt Børnich explains 1X's decade-long bet: by making Neo closely resemble a human in hands, form, and movement, the robot can leverage the entire internet's video of humans as training data — far more than any competitor can collect through teleoperation or sensor suits alone.

transcript

Bernt Børnich: the big bet that we made which is this decade long bet in 1x is if you get the robot to be similar enough to a human then you can train on all of the available video data out there of humans. And we're starting to see some very good proof that this is actually working incredibly well. And that's the reason we started the WX World Model Lab because we now finally have the scaling loss on that. And we're seeing that this — So I think ultimately it's very simple right the model is going to be as good as the data. And if you think about the data pyramid then on the top you have like tele operation data very high quality small fine tuned data set... then you have the data which is what you're talking about with like put the sensors on the human go and gather data. you have more of that and it's very close to the robot but it's not the robot... Then you have egocentric video from humans point of view... And then you have general video data... And because Neo is so similar to a human, we can actually utilize all of that data. Now, the bottom layer in the pyramid, which is this video data, general video data is absolutely ludicrously immense compared to anything else. it's YouTube, it's everything. So if you look at what is needed to actually achieve true intelligence, you need multiple orders of magnitude more data than anyone is even close to collecting over the next few years with egocentric data or with this sensor data. And all of the major breakthroughs that we've seen as far as I'm aware of in AI have been because someone figured out how to use a huge new data source that previously we were not able to use. you unlock some new set of data and now your model capability greatly improves.

supports · 1

Prediction

Hard takeoff — where robots build robots, data centers, and chip fabs autonomously — is less than 10 years away, and my current bet is 3 years.

Bernt Børnich predicts that within 3 to 10 years, robots will be building other robots, running data centers, and operating chip fabrication plants — creating a self-sufficient system of physical intelligence that he calls 'hard takeoff.'

transcript

Bernt Børnich: 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 system that is just — Under 10 years. my current bet would be 3 years. But like if it takes 10 like in the in the history of humanity, right? It's still like a blip. It doesn't really matter.

extends · 1

Data

Spot is the most deployed mobile autonomous robot on the planet, with over 500 customers across 46 countries delivering real industrial ROI in under two years.

Amanda McMaster details Spot's real-world traction: 500+ customers in 46 countries performing industrial inspection, gauge reading, vibration detection, and security patrol — and the inflection point was proving customer ROI in under two years.

transcript

Amanda McMaster: We started with our spot robot, which many people know. That's our mobile quadroed um in industrial — It has been deployed in real customer sites. It's pre providing really customer value. Um at this point we have over 500 customers over 46 countries. Wow. Um it is the um it is the mobile um autonomous robot that's used more than any other on the planet right now. ... the customers are finding a lot of value in industrial inspection. So, they're using it for both, you know, acoustic um, gauge reading, vibration detection. So, assets that, you know, if they have expensive assets in their facility and they want to monitor them, this allows for them to do that. Now, it can do that during the day and then it can do security perimeter work at night. Um, so the answer is yes, we do all of that. And, um, and the real inflection point was customer ROI, right? We want customers to find value in this to do really useful work. It's not just about yes, it's cute and it dances, but it's long past dancing at this point. It's now doing real work. And um and customers need to see your ROI in under two years.

supports · 1

Claim◆

The United States must not allow Chinese humanoid robots into the country and needs a coordinated national robotics strategy to protect IP and prevent data leakage back to China.

Amanda McMaster takes a firm stance: Chinese quadrupeds already leaking data back to China, the US cannot repeat the semiconductor mistake with robotics, and must bring manufacturing and IP protection into allied countries through a national robotics strategy.

transcript

Amanda McMaster: No. It's not safe. Right. We've already heard um about leaks that are happening with some of the quadripeds that you're seeing um in the United States and it's being back channelled back to China. Listen, we we have seen what happens if we let China win in the semiconductor space. You know, we can't do that with robotics. So, we need to have a concerted effort to protect our IP to um make sure that we are bringing manufacturing of of this ecosystem into the United States or into um our allied countries. And that means that we need to take our national robotic strategy. We're lucky enough that we get to sit at the table in some of these discussions. Um I'm hoping that more companies in the US join us um in in taking taking up this mission.

Claim◆

This time is different for humanoid robotics because large language models have essentially solved perception, and the combination of safety-certified hardware and sub-$1-per-hour operating costs makes widespread deployment inevitable.

Jonathan Hurst explains that LLMs have made robot perception 'all but solved' — robots can now understand their environment semantically. Combined with Digit V5's safety certification (no physical barrier needed between robot and human) and economics approaching $1/hour vs. $20-40/hour human labor, the path to mass deployment is clear.

transcript

Professor Jonathan Hurst: It is very easy to make a robot that looks like a person. It's very hard to make a robot that can do useful things in human spaces. And we're starting to see that today and that's the difference. So even if it doesn't look exactly like a human, but maybe a little bit humanoid, but it's doing useful work, that's where the impact matters. And because of large language models, a lot of things have now become free. When these robots look at a table here and you say, 'What's on the table?' It knows that's a phone. It knows this is paper, tea, water. It probably knows how many ounces are in each. If it were sitting here 3 or 4 years ago, it wouldn't actually know what was in the world. You would have to program it in a very narrow way. Perception was incredibly difficult. And the fact that perception is all but solved at this point is a really, really huge inflection point. ... Digit V5 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. ... we get 20 hours 365 days a year, you know, now you're in that 78,000 hours a year. Let's put it at 8,000 5 years 40,000 hours of work. ... People tend to think these things are going to cost 20, 30, $40,000. They will at some point. ... So that's a dollar an hour. These people are being paid in factories currently $40 an hour.

supports · 3

Highlight slides

Superhuman Perception, Not Labor Replacement✦ from: Industrial inspection robots are not about labor replacement — they achieve superhuman perception that prevents catastrophic downtime, paying for themselves in hours saved.Avoiding Catastrophic Downtime✦ from: Industrial inspection robots are not about labor replacement — they achieve superhuman perception that prevents catastrophic downtime, paying for themselves in hours saved.The Data Pyramid for Robot Training✦ from: The only path to true embodied intelligence is to build a robot so physically similar to a human that it can be pre-trained on all internet video data, which is orders of magnitude larger than any robotics-specific data set.The 1X Bet: Human-Like Form Unlocks Internet Data✦ from: The only path to true embodied intelligence is to build a robot so physically similar to a human that it can be pre-trained on all internet video data, which is orders of magnitude larger than any robotics-specific data set.Why Scale of Data Matters for Embodied Intelligence✦ from: The only path to true embodied intelligence is to build a robot so physically similar to a human that it can be pre-trained on all internet video data, which is orders of magnitude larger than any robotics-specific data set.Chinese quadrupeds already leaking data to China✦ from: The United States must not allow Chinese humanoid robots into the country and needs a coordinated national robotics strategy to protect IP and prevent data leakage back to China.National robotics strategy as a security imperative✦ from: The United States must not allow Chinese humanoid robots into the country and needs a coordinated national robotics strategy to protect IP and prevent data leakage back to China.LLMs Solved Robot Perception✦ from: This time is different for humanoid robotics because large language models have essentially solved perception, and the combination of safety-certified hardware and sub-$1-per-hour operating costs makes widespread deployment inevitable.Safety + Economics Unlock Deployment✦ from: This time is different for humanoid robotics because large language models have essentially solved perception, and the combination of safety-certified hardware and sub-$1-per-hour operating costs makes widespread deployment inevitable.

Related episodes

Video · The TWIML AI Podcast

AI Factories Explained: What’s Actually New?

→

Video · The Cognitive Revolution

Intelligence on the Edge: Liquid AI's Ramin Hasani on the Search for Device-Native Foundation Models

→

Video · The Cognitive Revolution

Fable's Back, AI Engineer Recap, & SambaNova

→