[lasercat]

Another exploration of this theme was in an episode of Doctor Who (Smile) which features a colony with nanobots that are optimized for happiness, with disastrous results: https://en.wikipedia.org/wiki/Smile_(Doctor_Who) - "The Vardy were programmed to help construct and operate the colony and make the colonists happy, monitoring the emotional state of the colonists through the emotion badges and avatars. When one of the flight crew died of natural causes, it created grief among the crew, something the avatars were not programmed to register. The Vardy took this as a sign of disease and killed those who displayed unhappiness. This created a "grief tsunami" which rapidly wiped out the flight crew, and will likely wipe out the waking colonists when they discover what happened."

[schoen]

Huh, I was going to wonder if this inspired the AI safety concept about how you shouldn't program a superintelligent AI to optimize the number of things it can see that look like human smiles, since unfortunately there are many ways to maximize that other than making real people genuinely happy.

But it seems like the Doctor Who episode is from 2017 and so it could potentially have been inspired the other way around.