Notes on Robotics and Untapped Real World Potential

The domain of robotics, even today, is vastly under explored for practical use cases that genuinely benefit people. People often say the same about LLMs because of the constant stream of model releases and updates. In robotics the effect is far stronger because far fewer people are exploring the potential space. There is so much that is now possible or reachable, in the sense of the adjacent possible, to borrow from Stephen’s idea.

Once you begin to mentally explore tasks that are currently handled in very bespoke ways, with teams of people and specialised equipment, you start to notice that many of these could now be replicated with existing robotics and modern software. A good example is sports broadcasting. There is no realistic way to deliver that level of quality cheaply for a local sports team, yet current robotics already give you camera control, tracking, robotic positioning and the option to use cloud or local processing for streaming and synthetic camera stitching. All of this gets you surprisingly close to broadcast level output with equipment that costs under ten thousand. Setting aside the real difficulties and the fact that any early prototype would need generous assumptions, the trend is clear. The quality of camera miniaturisation, the precision of robotic control and the infrastructure for streaming from providers like Cloudflare mean we are entering a period where the initial capital cost is falling quickly.

This is before even considering the potential to exceed current broadcast quality. Once you imagine statistical tracking, segmentation, depth of field techniques and continually improving tools like SAM 3, you can see a kind of inventor’s dilemma forming. A small rough prototype built by a local club could, with time and investment, climb the ladder of quality and capability.

Perhaps I am overly bullish, but it feels like one of the first mass market use cases for small drone drifters. With positional stability, neutral buoyancy, low power requirements and a friendly physical presence, small blimp like drones seem almost ideal, although perhaps less so in the winds of the west of Ireland. It is hard not to love the idea of a chunky little blimp quietly filming a match.