Blog · 2026-08-13
There's a version of this conversation that goes straight to jobs and Terminator references, and I get why — it's the easy hook. But it skips past the actual...
There's a version of this conversation that goes straight to jobs and Terminator references, and I get why — it's the easy hook. But it skips past the actual question, which is a lot more useful: what does this stuff do for us, day to day, if we actually let it in?
It's not a future hypothetical anymore. It's landing here right now. The only real choice left is whether you're watching that happen or taking part in it. Watching doesn't slow anything down. It doesn't help in the slightest. It just means someone else decides what it's used for, how and when. Taking part — actually getting hands-on with it, that's where the leverage is.
So let's be pragmatic about where the leverage actually shows up.
Education
A robot dog or a humanoid in a classroom isn't a gimmick, it's a way of making abstract STEM concepts physical. Kids don't just learn about actuators and sensors and coding, they watch their code make something walk, or fail to walk, in real time. That kind of immediate, physical feedback loop is worth more than a semester of theory on a whiteboard. Worth more than experimenting with some rudimentary lego with motors.
Kids who grow up programming robots instead of just reading about them stop treating this technology as something mysterious that happens to them and start feeling involved and able to be part of its direction and purpose. It means Australia has a generation coming through who aren't intimidated by the tools shaping the next few decades of work.
Security
Quadrupeds are already doing perimeter patrols, site inspections and hazardous-area checks that used to mean sending a person somewhere unpleasant or dangerous — mine shafts, disaster sites, unstable structures. Very large spaces that require many boring, lonely hours for a human guard. That's not a robot replacing a security guard's judgement. It's a robot doing the boring or risky bit so a person can make the actual call from somewhere safer, faster, with better data while allowing the humans to turn their time to greater challenges and tasks.
**Entertainment and public engagement **
We need to stop dismissing this stuff as novelty. Entertainment is often the first honest contact most people have with this technology — a robot at a conference, an open day, a corporate event, a tech conference. A funny video of a robot chasing wild boars... That's where the sci-fi weirdness wears off. You can't build public trust in something through a highlight reel. You build it by letting people stand next to the thing. It is now very close to the point where one can experience it not just visually but in a tactile way.
I remember when I experienced it firsthand in Magiclabs' building in China. Before that, I had only ever seen the robot videos. I had not seriously travelled far away from my home in NSW Australia. It was an eye opener.
With each new experience in my life comes a new realisation about how little it is that I know.
So, like a baby taking its first steps, I flew, rode in Byd cabs, bussed, walked, high speed trained, walked and walked my way to Magiclab headquarters. They shared the same building with a highly official government department and of course, I went in the wrong side. I still don't know what that place was, but there were some very serious people there and as I sat down, the person in front of me started sternly demanding information about me. What I was doing here, where am I from and who do I think I am type of questions. I was too stupid to realise what was going on for about 5 minutes. Fortunately, I was eventually rescued and ushered into the left side of the building.
After introductions and pleasantries, it became surreal. Robots everywhere. Dog robots. Humanoid robots. Crazy. Not the imagined sort of robots from my dad's already dated science fiction paperbacks when I was 10. Nope. These ones were real. Metal, motors, computer chips, sensors, cameras, plastic. Real. I now had this robot right in front me. Magicbot Gen 1. Standing on its own legs. Balancing just the way the old sci-fi stories had described. I wondered how much art influences science.
I could hear things quietly whirring and clicking. The soft noise that I know will be in the backdrop of our and our children's future, an indelible part of the natural soundscape. Its' digital eyes representing its real time 4k camera were looking at me while it spoke to me. To finally see it in front of me solidified in my mind how close to an automated embodied A.I. reality we really are. We should buckle our seatbelts.
Research
Humanoids and quadrupeds are becoming lab tools in their own right — platforms researchers can run manipulation, locomotion and AI experiments on without building the hardware from scratch every time. That lowers the barrier to entry for Australian universities and labs that don't have the budget or the years it takes to build a walking robot in-house. Buy the platform, focus the actual research effort on the interesting problem.
Disability and accessibility
This is the one I think gets under-discussed. Robotic arms and assistive platforms are already giving people with limited mobility a level of independence that didn't exist ten years ago — picking things up, opening doors, operating equipment, without needing another person to do it for them. One only has to look at the humanoids that Fourier is building to see their potential in aged care. The elderly and disabled are not just a bunch of other people somewhere else. It's us. You and me.
Teleoperated systems, the kind you drive through a VR headset or a controller, are also opening up work and tasks to people who physically can't be in the room but can absolutely do the job remotely, arms extended into a space they can't otherwise stand in.
The jobs question, honestly.
Some jobs will change shape. That's true and it's not new — it's true of every serious technology shift. I imagine that when that crazy invention "The horseless carriage" came along, people were probably up in arms about the poor farriers losing their jobs.... The more accurate story, and the one that's actually playing out in warehouses and clinics and labs right now, isn't "robot replaces person."
It's "robot takes the repetitive, physically taxing or dangerous 20% of a job, and the person spends their time on the 80% that actually needs judgement." A security guard who's not walking a perimeter for two hours a night is a security guard who's watching more cameras, handling more incidents, doing more of the job that actually needs a human.
A researcher who's not hand-building a robot platform is a researcher who's doing more research. Efficiency isn't automatically the same as unemployment. Often it's the same person, doing more of the part of their job that was worth hiring them for in the first place.
And while we're deciding whether to worry about it.
China isn't debating any of this. It's quietly building it — the humanoid and quadruped supply chains, the manufacturing scale, the deployment in warehouses and factories and public spaces, at a pace that isn't waiting for the rest of the world to finish having the conversation. That's not scaremongering, it's just the market. Whoever gets comfortable with this hardware first gets the advantage first, and right now that's happening quietly, without much fanfare, while a lot of other countries are still arguing about whether it's worth taking seriously.
Australia doesn't need to win that race outright. But it does need to be in it — building familiarity, skills and infrastructure now, rather than importing a finished ecosystem in ten years and calling it catching up.
None of this requires blind enthusiasm. Ask the hard questions, push back on the hype, be sceptical of anyone promising it'll fix everything overnight. But do it with your hands on the controller, not from the sidelines. That's the whole difference between a country that shapes how this technology gets used, and one that just receives it once everyone else has already decided.
— Cave, AussyBot