Jason Robert Carrol breathed his last.
He lay in the hospital with his wife and children nearby and an attentive nursing staff and competent doctors. It was pneumonia, the old man's friend, as he used to jest.
As his body was removed, his wife sat down in the one visiting chair and the children on the floor, and they all shut off.
As the body was taken away, the hospital staff sat down in place and shut down. Then the hospital lights went out. The body was buried as Carrol's will had indicated and, their job complete, the burial crew shut down.
Carrol was the last one, and with him humanity died, never fully knowing.
It had begun in 2035, with the first marriage between a man and a robot. It was, of course, a very beautiful robot he had married, and as people rightly suspected, it had been just a physical interest, and the marriage did not last long.
But the robots improved rapidly. People took batteries of psychological tests, and then factories built and programmed robot spouses that matched them perfectly. Some wanted submission in their spouse; others preferred a bit of fire. Intelligence, stupidity, fawning adoration, beauty, sex, whatever. They got what they wanted.
They could even have children, robotic children who were updated regularly until they grew into adulthood, and then - like their parents - they appeared to age.
With mass production, the cost of robots dropped, and these pleasures of the rich became increasingly available to all. More and more people married - or simply lived with - robots. And life with a robotic man or woman who matched you perfectly was very beautiful.
But fewer and fewer people married other people, so the number of people began to slide, though where even one human existed, the streets were full of cars, the shops and restaurants were all open, and the media broadcast the news, though what part of the news was about robots and what part about humans it was impossible to tell, as the two were difficult to tell apart and because it was considered very rude to make such distinctions.
And in the end, the humans died out, never fully aware that it was even happening.
But the fearmongers were wrong. The robots never took over the world - they never even tried. They died with their masters. Although scientists had succeeded in making robots intelligent, they had never figured out how to give them a will.
No robot ever wanted anything, any more than a brick ever wanted anything. Oh, true, they acted as if they wanted things, and their apparent desires were indistinguishable from human desires, but it wasn't real.
Some scientists thought that if they combined the right software with enough processing power to make a computer's brain, that the computer would become self-aware and have a will. But it never happened, any more than a pile of bricks is more self-aware than a single brick. More and more on-off switches combined with more and more instructions for flipping those switches on and off never resulted in a computer that wanted so much as an ice cream cone or cared a bit or a byte whether it existed for a single additional second.
Oh, when the humans were still alive it certainly looked that way, and many scientists argued that robots had indeed achieved self-awareness, but when the last of the humans died, the robots - seeing no tasks before them - reverted to their most basic instruction set, and simply shut down to save power.