Ethan Mollick: “I suspect that personal AI use will actually be centered on our phone, though not necessarily through apps. Small, local AIs running on your phone’s hardware (something both Microsoft and Apple have demonstrated) can already do much better than Siri at basic assistant tasks, and they can connect to more powerful AIs over the network to handle more difficult requests. For most people, this will be all the AI they need. They can make a request of their local phone AI, and the system will decide how much computing power to put into it. It is a model of ubiquitous AI that does not actually require most users to change habits or devices.”
WSJ: ““We’re talking about creators as the studios of the future,” said Mary Ellen Coe, YouTube’s chief business officer. The platform benefits from a steady stream of fresh content released daily and is particularly popular among people under 24. Today, 150 million people in the U.S.—more than 40% of the population—watch YouTube on connected-TV screens each month, a spokeswoman for the platform said. It has benefited from new features that let viewers shop or chat with one another while watching big events, like a livestream of last month’s Coachella Valley Music and Arts Festival. Unlike traditional media companies, which have to pay hefty sums upfront for programming that they hope will bring in larger amounts of subscription and advertising revenue, YouTube incentivizes the creation of content by sharing 55% of revenue from ads that run in creators’ long-form content, and 45% of revenue from ads on their short-form videos.”
Bloomberg: “Humanoid robots won’t be limited to the factory or warehouse floor. Their advantage over other machines will be in unstructured, fluid work environments, such as a construction site where they may have to step over objects or climb stairs. They will be able to do tasks in buildings and homes because they fit where humans do…Throughout time machines have eased the burden of workers by doing the most physical, repetitive work. This has allowed humans to be more productive and earn more. This new class of robots will repeat this pattern. The difference is that humanoid robots will be working closer with people, putting a priority on safety. The rules on how AI is applied to them must be clearly defined. Still, these robots need to always be seen as just tools that help humans be more productive and always considered just as inanimate objects that can be shut down at any time or recycled when no longer needed.”
Nell Freudenberger: “The essays that have stayed with me over the years don’t follow a pattern. There are some narratives on very predictable topics — living up to the expectations of immigrant parents, or suffering from depression in 2020 — that are moving because of the attention with which the student describes the experience. One girl determined to become an engineer while watching her father build furniture from scraps after work; a boy, grieving for his mother during lockdown, began taking pictures of the sky.If, as Lorrie Moore said, “a short story is a love affair; a novel is a marriage,” what is a college essay? Every once in a while I sit down next to a student and start reading, and I have to suppress my excitement, because there on the Google Doc in front of me is a real writer’s voice. One of the first students I ever worked with wrote about falling in love with another girl in dance class, the absolute magic of watching her move and the terror in the conflict between her feelings and the instruction of her religious middle school. She made me think that college essays are less like love than limerence: one-sided, obsessive, idiosyncratic but profound, the first draft of the most personal story their writers will ever tell.”
FT reviews “How to Become Famous”:”To probe the mystery of why some people become famous and others don’t, Sunstein calls on compelling studies from experimental psychology and behavioural science. Patterns of fame in which “the winners are spectacularly successful, and . . . relatively few” mirror mathematical power law distributions, he says, with surprising predictability. In the fascinating Music Lab Experiment, which asked people to download tracks by unknown artists in an “artificial cultural market”, participants were more likely to enjoy songs they believed more people had downloaded. This was the case even when the information they had been given about a track’s popularity was false.”
