“What’s your goal?,” R asked.
My starting premise is simple: robots will be adopted. I’ve seen the look on workers’ faces when a robot first folds a shirt or packs a box of insects. Curiousity, excitement, and perhaps even fear. Robots will replace certain types of work, but not everything. That’s why it matters how we build them. I’ve previously written a little about this here and here.
Ezer’s goal is simple: robots adopted safely and affordably. Robots that small businesses can afford. Robots that workers can operate after hours of training. Robots that shift people into safer, more skilled roles.
Most humans can scrub a shirt clean quite quickly, or wash dishes faster than a dishwasher. But I don’t know many who would choose to wash clothes or dishes for many hours a day. Technology has always shifted labour.
That said, the adoption of technology is not neutral. When the washing machine arrived, it replaced hours of manual labour. It didn’t eliminate households; it changed how time was spent.
AI robots are more consequential than most technologies before them. As they interpret, discern and act in the physical world, the widespread adoption of robots will shape safety, labour, money and power. We must design robots with humans in mind.
Ezer deploys modular robots into businesses including small businesses folding laundry in London and handling insects in Ely. We start with teleoperation and gradually build autonomy. Throughout this process, humans stay in the loop, and robots learn in the real world, not in the lab. Where research labs focus on compute and models, hardware companies focus on building cool robots, Ezer focuses on real world and real work.
The truth is that there has been an overpromise of robots. In 2013, Elon Musk said autonomous cars were three years away. It’s 2026 and they are not everywhere. I spoke with a co-founder of a autonomous driving unicorn and he mentions he thinks widespread adoption is still 5-10 years away. It’s not that straightforward to adopt autonomous cars in the real world. I’ve built these systems in the AI robot space and it is arguably even more difficult. Getting a robot to reliably grasp a slippery garment in poor lighting, or navigate a cramped backroom in a homeless shelter, is far harder than a demo suggests. Real deployment is messy.
Robots will come, slowly but surely. The question is how they are deployed and who they serve. Ezer exists to make sure they serve people, not just capital.
DNMZ