On a recent night, a woman named Robin was asleep next to her husband, Steve, in their Brooklyn home, when her phone buzzed on the bedside table. Robin is in her mid-thirties with long, dirty-blond hair. She works as an interior designer, specializing in luxury homes. The couple had gone out to a natural-wine bar in Cobble Hill that evening, and had come home a few hours earlier and gone to bed. Their two young children were asleep in bedrooms down the hall.

This might sound like a hot take but it's not: In 50 years, when historians look back on the crazy 2020s, they might point to advances in artificial intelligence as the most important long-term development of our time. We are building machines that can mimic human language, human creativity, and human thought. What will that mean for the future of work, morality, and economics? Bestselling author Steven Johnson joins the podcast to talk about the most exciting and scary ideas in artificial intelligence and an article he wrote for The New York Times Magazine about the frontier of AI.

There was a point for all of us, somewhere near the beginning of the pandemic, where we said to ourselves, oh shit. One of my first oh-shit moments arrived after reading Ed Yong's sobering feature, "How the Pandemic Will End," in the Atlantic in March. The coronavirus, Yong wrote, was "unlikely to disappear entirely," and he explained that it was possible that "COVID-19 may become like the flu is today--a recurring scourge of winter." I repeat: Yong wrote this in March. The day it published, the United States had so far detected around 68,000 cases of COVID-19 and documented less than 1,000 deaths, according to the CDC. In the days ahead of its publication, California became the first state to mandate its citizens stay home; Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases, said that Americans would likely need to socially distance for "at least several weeks"; and President Donald Trump said he wanted to have the country "opened up and just raring to go by Easter."

A deepfake video featuring Chris Evans as Homelander from Amazon's The Boys recently appeared on Reddit. The 13-second clip from Redditor d4danger includes multiple shots from The Boys, but with Chris Evans, who played Captain America in the Marvel Cinematic Universe, as Antony Starr's Homelander. The irony of the MCU's Nazi-fighting Captain America's face on the body of Homelander -- who was in a relationship with an actual Nazi in Season 2 of The Boys -- was not lost on viewers. A deepfake video is a popular trend in which the likeness of one person replaces another with exacting detail. There have been several recent deepfakes involving cinematic superheroes, including one with Jon Krasinski taking Evans' place as Captain America, while another swaps Gal Gadot's Wonder Woman with Lynda Carter's version from the 1970s.

Imagine this: You click on a news clip and see the President of the United States at a press conference with a foreign leader. The news conference is real. You share with a friend. They share with a friend. Soon, everyone has seen it.

As artificial intelligence becomes integrated into more technology in our homes and workplaces, concerns about its ethical implications are growing. It seems like every day there are new tools designed to help workers use their time more efficiently with machine learning. If these trends continue, what will the workplaces of the future look like? What if it goes wrong? Keiichi Matsuda's new short film Merger taps into that uncertainty by imagining a workplace where humans have been proven to be less capable than AI.

That poses a pretty major problem. For example, when a neural network screws up, engineers don't know precisely why, or whether that one screw-up represents a small fluke or a much bigger flaw in the programming. That would be a minor problem with dog pictures, but a much bigger issue when it comes to predicting diseases or driving autonomous vehicles. And sometimes these networks do things we never meant them to. AI trained to recognize objects begins to recognize human faces, for example, or two bots trained to negotiate with each other come up with their own language.

Art depicting Norman, a "psychopath" AI created by researchers at MIT. (Photo: Massachusetts Institute of Technology) Researchers at the Massachusetts Institute of Technology created an artificial intelligence labeled a "psychopath," using disturbing image captions found on Reddit. The AI is named Norman, after the character in the Alfred Hitchcock classic "Psycho." Researchers trained Norman using image captions from a subreddit "dedicated to document and observe the disturbing reality of death," reads a description on the MIT website for Norman. Because of technical and ethical concerns, the team at MIT used captions and not actual images of people dying. "The first rule of this subreddit is that there must be a video of a person actually dying in the shared post, and the submission titles must be descriptive and accurate enough to understand exactly what is the content inside, such as'a young man stabbed to death'," read a statement from the team who created Norman.

These inkblot pictures reveal how a terrifying'psychopathic' machine called Norman was trained to think like a maniac. The AI - named after Hitchcock's Norman Bates in the 1960s film Psycho - was trained on disturbing images of death culled from a group on Reddit. As part of the same study, another'normal' AI was trained on more benign images of cats, birds and people. Both Norman and the regular AI were then shown inkblot drawings used by psychologists to better understand a person's state of mind and asked to interpret them. On one of the drawings, where a'normal' AI saw'a close-up of a wedding cake on a table', Norman's interpretation was much more sinister of a'man killed by speeding driver'.

Mario Klingemann is the Machine Learning Artist in Residence at the Google Cultural Institute in Paris (the same arm of Google behind the new Louvre-busting museum app). But when he's not tinkering with neural networks to figure out how they can be leveraged against Google's database of several million scanned cultural artifacts, Klingemann's training them on human faces... and posting the unsettling results online. Trained on a database of over 20,000 faces organized by rough category (age, gender, pose, etc), Klingemann experiments use a convolutional neural network similar to Deep Dream. Except instead of just recursively reinterpreting the same picture with trippy results, Klingemann's face maker generates ghoulish visages entirely from scratch, according to how computers interpret faces. And how do computers interpret faces?