The description of events is slightly suspicious. For all the predictive smarts of the car and its impressive LIDAR, keeping from bumping into things seems like it would be the highest priority possible second only to loss of life. The corners of the car have some kind of distance sensor so the car would have to have known it was about to collide with something. In CPU time there was plenty of time to react and its always looking at all four corners.
The article make it sound like the AI blindly changed lanes into the bus. It seems most likely that the AI knew about the impending collision and decided colliding was safer than any other option. It'd be great to know what the other options were, but I imagine we'll probably never get more detail.
It didn't change lanes, which is why the description didn't say it did. The entire event took place in a single extra-wide right lane, which are pretty common. The extra width is to accommodate on-street parking as well as right hand turns.
Notice that it doesn't say it changed lanes; rather, it shifted to the leftmost-side of its lane, later to return to the center of the lane (which is when the bus hit it).
I get your point, though, and I imagine it could have gone something like this: Car predicts bus will slow down (as reported), moves back to center of lane. Bus doesn't, and the car's sensors inform of it. During the seconds (or the fraction of a second) before the collision, the car came up with the safest avoidance maneuver, and started executing it. Crash.
The article make it sound like the AI blindly changed lanes into the bus. It seems most likely that the AI knew about the impending collision and decided colliding was safer than any other option. It'd be great to know what the other options were, but I imagine we'll probably never get more detail.