Rehearsing Reality.
I watched the Nvidia GTC keynotes a few months ago, and for the first time in a while I felt both excited and overwhelmed by the possibilities.
What landed most strongly was the idea of using simulation to train robots. Not abstract environments, but high-fidelity worlds that closely resemble the real one.
At Nearmap, I see how much accuracy and realism matter, and how they shape better decisions. But until recently, I hadn’t fully connected those ideas to robotics. That changed when I read World Labs’ case study: robots learning and being evaluated inside worlds that look and behave like our own.
Once you see it, the implications are hard to ignore. Training accelerates. Risk drops. Reality no longer has to be the first place you make a mistake.
A quiet inflection point. When simulation stops standing in for the world and starts extending it.
