I was going to make a comment about the hood not being soft enough for pedestrians hitting it, but maybe the computer would prevent such a collision.
Right? If your bio isn't a joke, you'd better darned well have that attitude. Otherwise I don't see how you'd get through a typical workday.
The SL-1 nuclear reactor accident (possibly murder-suicide) happened because a human was actively actuating a control rod by hand and pulled it out too fast. Passive systems that limit rod withdrawal rate are better.
The Three Mile Island accident happened because a sensor mislead the human operators, who then did the wrong thing and ended up dropping the coolant level below the core, which subsequently partially melted. Passive safety systems like a pool of low-pressure coolant preclude this entire class of accidents.
Chernobyl happened because humans could and did manually disable all the automatic safety systems that told them the reactor was in an unstable configuration. Passively safe reactors can't physically get into unstable configurations.
Fukushima had active cooling systems powered by diesel generators. After the earthquake, they started up and worked fine. But when the tsunami came, it flooded the basement. The operators for god knows what reason put the fuel supply and electric switches in the basement, which flooded. The active safety systems failed, the coolant boiled, and the cores melted. As with TMI, passively safe reactors with low-pressure coolant and/or natural-circulation driven decay heat removal (i.e. no diesel backup power needed) would preclude this condition.
The nuclear industry is very into passive safety features, from experience. The first true passive safety demos happened in Idaho in April 1986 (weeks before Chernobyl) at a reactor called the EBR-II.
In summary, making a system safer with active systems is one approach. It's often both more elegant, more reliable, and cheaper to improve a system passively via design ingenuity. Thus, passive safety has a place here at Hacker News.