
Humanoid Robots: CTO Jarad Cannon Reveals the Big Breakthroughs
The sudden momentum in humanoid robotics is no coincidence. Numerous companies, from tech giants to agile startups, are pushing the development forward. They all aim to bring human-like robots with versatile functions into the workplace and, eventually, into our everyday lives.
What seemed like wild speculation a few years ago — that we would actually get robots in our lives — now appears within reach. Cutting-edge technology, fresh billion-dollar investments, and the increasing need for automation are potentially giving birth to a trillion-dollar industry: humanoids.
On this note, Jarad Cannon, Chief Technology Officer at Humanoid, dives into the multitrillion-dollar future of humanoid robots and how they could transform industry and society.
By discussing key issues such as safety, compliance, and human-robot interaction, he breaks down what this means for the future workforce, business, and our lives.
Q. To begin with, can you tell us something about your role at Humanoid? Any key highlights that you want to mention?
Cannon: I joined Humanoid as CTO not long ago, and I can already say it’s one of the most exciting work environments I’ve ever been part of. One thing that stands out here is how fast everything’s moving. We’re scaling rapidly — new people, new functions, new product developments. It requires a lot of coordination across engineering, product, and business. It requires alignment of many cross-functional teams.
There’re no instructions or manuals for building the robotic future, no one’s done it before at commercial scale. But that’s exactly what I like about it. In simple words, my job is to make sure all the ambitious tech we’re developing actually becomes real — on schedule, at scale, and ready for the real world.
Humanoid is a commercial-first company, not just a research lab, and that perfectly aligns with my background as well. Before joining Humanoid, at Brain Corp, I helped scale from 5 to over 40,000 deployed robots and led the launch of 13 different robots into the market. We’re entering a crowded space, a so-called “humanoid race”. But I’m sure that we’re not late, but exactly on time. Humanoid actually has what we call a second-mover advantage: we’ve learned by example, studied what works and what doesn’t, and now we’re coming in with a team full of veterans who built the first generation of humanoid systems.
As I said at the Humanoids Summit in London this May, we’re a hardcore humanoid robotics company. For us, building humanoids isn’t a side project or hype. It’s our entire mission. Every decision, from system architecture to business model, serves that goal. That is our superpower.
We have the rare combination of timing, world-class talent, and technology to make humanoid robotics real, and we’re on track to launch both wheeled and bipedal alpha models later this year. It’s a big moment.
Q. Do you wish to say something about your latest humanoid robot, HMND 01? How is it unique from the other robots?
Cannon: To start, I love competition, it makes everyone better and drives the whole industry forward. But what sets HMND 01 apart is our focus on real-world deployment. It’s not a research platform. It’s a reliable, commercially viable and affordable product — so it can work and scale in industries that need it most.
We’re starting with logistics, retail, and warehouse environments because that’s where the demand is real. We are developing market-ready solutions, starting with PoC really early. For example, recently, we completed our first POC with a pre-alpha wheeled prototype. And all this in just a year! That pace is incredible for this industry, most of our competitors have been there for years.
Also, our robot is modular by design: HMND 01 robots feature modular lower bodies and end-effectors. This approach allows customers to adapt robots to different tasks and reduce cost across deployments.
Another game-changer is AI-led design. We’re using simulation not just to train the robot but to design it as well: sensor placement, walking policies — everything runs through environments like Isaac Sim. We’ve integrated reinforcement learning from the very beginning and we’re also built around hardware-enabled AI.
Our goals are very ambitious. To achieve them we’ve built a team of world-class talent, more than 160 experts across three hubs: in London, Boston and Vancouver. I believe in this team, and I believe in this moment.
Q. With robots capable of doing a range of tasks, will they replace humans in the future?
Cannon: I believe “Humans vs. Humanoids” is a completely wrong frame. It’s not “Humans vs. Humanoids”, it’s “Humans with Humanoids.” At Humanoid, we believe robots should amplify human potential, not displace it.
We’re entering a moment where entire industries, from manufacturing to retail, are struggling to fill essential jobs. We are facing an unprecedented labor crisis across developed nations. In the U.S. alone, there’s a shortfall of over two million workers, and Europe isn’t far behind. Meanwhile, the population is aging, and birth rates are declining. So who’s going to do the work?
Humanoid robots are developed to take on dull, repetitive, and physically demanding jobs that few people want to do. This shift will free humans and let them focus on more creative, meaningful, or strategic tasks.
Also, let’s not forget that the robotics industry creates thousands of new jobs in design, AI, hardware, support, ethics, and operations, among others. It’s somewhat similar to what AI is doing right now, reshaping entire economies.
Q. Ensuring that robots interact safely with humans in various environments is crucial. Major points to note to address these safety concerns in human-robot interaction?
Cannon: Safety is essential for humanoid robotics. Robots will soon work and live side-by-side with people, often in dynamic environments. We can’t afford to get it wrong.
At a technical level, it’s incredibly complex. Take voice commands, for example — we believe they will become a natural way of interacting with robots. But we can’t allow just anyone to give instructions. There are big questions around command authorization, data privacy, and even intent detection. And once these robots move into households, the bar gets even higher. That’s why safety and compliance should be embedded in design from day one.
Our headquarters is in London, so we’re building our robots, aligning with the high safety standards of the UK and EU. We believe it will also give us a competitive edge against companies operating in looser regulatory environments.
Right now, we’re focused on industrial use cases: logistics, warehousing, retail, and manufacturing. They are more structured, more predictable environments, which allow us to test and refine our safety systems, and only then consider more complex household interactions.
Also, we’re targeting a wheeled platform first for industrial use cases. It’s inherently safer, easier to control, and already capable of tackling a broad range of tasks. We’ll learn, iterate, and then apply that knowledge as we transition into bipedal models.
Q. How can one ensure that AI-powered humanoid robots are developed and used responsibly?
Cannon: Before letting humanoids into warehouses or households, developers have to conduct extensive testing in diverse real-world scenarios to identify unexpected risks. Early PoCs give us real-world data on how robots interact with people, systems, and environments. That lets us identify edge cases early and train our robots more safely and intelligently.
I believe that companies should also implement robust safety mechanisms: include fail-safes, emergency stop functions, and behavioral constraints to prevent harm.
Q. As robots become more autonomous, questions arise about their moral status and whether they should be held accountable for their actions. What is your thought process on this?
Cannon: It’s one of the most fascinating discussions in robotics. When people talk about accountability, they’re often imagining humanoids like creatures with thoughts and feelings. But in reality, we’re talking more about automation tools.
In technical words, this is the question of autonomy versus control. There are three levels of control in humanoid robotics: full human control (safe, but loses the point), shared autonomy (robots perform tasks, humans monitor and step in if needed), and full autonomy (robots act independently).
Even at level three, there’s a key principle: robots don’t make moral decisions; they act within predefined goals, based on what they’ve been trained to do. They don’t have intent. That’s why, for me, accountability lies with the people who design, deploy, and manage these systems.
It’s no different than the accountability you’d expect with autonomous vehicles, medical tech, or financial algorithms.
At the end of the day, a humanoid robot is more like a smartphone than a pet. It’s an intelligent tool, and we’re the ones responsible for how that tool is used.
Q. How can the public/humans be educated about the potential benefits and risks of humanoid robots?
Cannon: Right now we’re standing at a turning point. For the first time in history, we’re truly close to living with robots. It’s a moment humanity has imagined for decades, just look at how many books and movies have explored this. To be honest, most of those images are dystopian. So, the skepticism we see now, especially on social media, isn’t surprising. It has deep roots in pop culture.
I think acceptance will come gradually, and organically. That’s why I support the strategy most companies, including us, are taking: starting with industrial use cases. These predictable environments let people get used to robots in a more relaxed way. You’ll hear the stories: someone worked with their first robot colleague in a warehouse or saw one moving goods in a supermarket. That’s how perception starts to shift — through lived experience.
Then we’ll see robots move into the service sector, and eventually into homes. Each step will bring new benefits and new questions. The shift won’t happen overnight, it’s an evolution.
I also think the media has a major role to play here too. We need more serious, thoughtful reporting and fewer clickbait headlines that multiply fear.
Q. Future trends in Humanoid Robotics: What to expect in the next decade?
Cannon: As I’ve already said, we’re at the most exciting moment the field has ever seen. For years, we’ve had pilots and demos, but no real deployed humanoid robot. That’s finally about to change. Thanks to advances in AI, simulation, and hardware, we’re on the edge of actual use cases scaling into real industries. The biggest trend is that humanoid robots will actually come from research labs into the real world.
At Humanoid, we see the rollout in three waves. By 2027, humanoids will be transforming manual work in logistics, warehousing, manufacturing, and retail. By 2029, we expect them to play critical roles in the service sector. From 2031, robots will enter households with the potential to impact over 3.5 billion homes worldwide. They’ll manage daily chores, support care needs, and maybe even help fight loneliness.
A decade feels like a lifetime in robotics. Remember that just ten years ago, AI was still mostly stuck in academic papers and sci-fi scripts, and now it helps millions of people every day. We believe humanoids will follow a similar trajectory.