After his mother got dementia, Bari became forgetful. It was little things, like hanging up the wet laundry on time so it wouldn't stink; spraying pesticide on their patch of sea wall against the adventures of crabs and mutant fish; checking the AQI meter before leading his mother out for her evening walk along New Karachi's polluted shoreline. Did something break in your brain, too, when you took care of people who once held you on their lap, helped you count the last straggling trees in the mohalla courtyard? Overwhelmed by their needs and your grief, perhaps you were split into two halves, each perpetually being run into the ground. It wasn't like he had a sibling or a spouse to lean on.

Why does everyone assume that the AI revolution will either lead to a fiery apocalypse or a glorious utopia, and not something in between? Of course, part of this is down to the fact that you get more attention by saying "The end is nigh!" or "Utopia is coming!" But part of it is down to how humans think about change, especially unprecedented change. Millenarianism doesn't have anything to do with being a "millennial," being born in the 90s and remembering Buffy the Vampire Slayer. It is a way of thinking about the future that involves a deeply ingrained sense of destiny. A definition might be: "Millenarianism is the expectation that the world as it is will be destroyed and replaced with a perfect world, that a redeemer will come to cast down the evil and raise up the righteous."

We are here today,I take it, to appraise what has been done, and to discern the future, if we may. I notice that a man's worth these times is in the words he speaks and writes. The understanding that may lead to a publishable paper is much to be preferred to the understanding that leads to a useful machine. "But I say unto you, that every idle word that men shall speak, they shall give account thereof in the day of judgment. For by thy words thou shalt be justified, and by thy words thou shalt be condemned." Can we say anything true and useful in generalization about Pattern Recognition? Are there any broad statements that have any chance of being helpful to someone building a better machine?