A system that commits to one predicted future becomes fragile in the situations where prediction matters most, whereas carrying multiple weighted futures guards against unusual scenarios at the cost of compute.

A safe autonomous system must account for all valid futures simultaneously — a cyclist might continue straight, turn, or stop — and Waymo's approach of generating many possible paths per road user contrasts with systems that commit to a single prediction. ✦ AI generated

Article Author · ByteByteGo Newsletter · 2026-08-17 · original ↗

Once the system holds a description of its surroundings, the next task is estimating what those objects will do. Several futures are valid at the same moment. For example, a cyclist approaching an intersection might continue straight, turn, or stop. A safe response should account for all possibilities. Waymo describes the system as producing many possible paths for each road user rather than one, drawing on accumulated driving data and accounting for the different ways a car, a cyclist, and a pedestrian move. A system that commits to one predicted future becomes fragile in situations where prediction matters most. However, carrying several weighted futures costs compute on every cycle, but can guard against the possibility of unusual situations.

Read full article ↗excerpt · fair-use quotation