1. Jobs Only for Humans?
In his much-cited August essay about the risks and consequences of AI, Bill Gates suggests that some jobs – like caregiving – should be “human reserved,” even if robots would be more efficient. Oren Etzioni, a computer scientist, offers a respectful dissent, contending that Gates’s diagnosis is right, but his prescription is wrong:
The problem with Human Reserved is that it assumes there’s a human being available. Home health and personal care aides earn a median of $34,900 a year, and BLS projects roughly 765,000 openings in that occupation every year through 2034. At that wage, they keep coming open. A third of home care aides are immigrants, and tighter enforcement threatens that supply. A rule that reserves care for people, in a market with no spare people, reserves care for the families who can outbid everyone else.
Mark Zuckerberg also posted a long essay about AI this August. His piece was not well received. This response in TechCrunch argues that the Meta CEO’s happy-go-lucky musings are exactly why people don’t like AI.
2. The Math at Berkeley Doesn’t Add Up
In May 2020, the University of California Board of Regents voted to phase out the SAT and ACT in admissions. Later, a court ruled that the entire UC system to be “test blind.” The outcome is impossible to miscalculate. Math professor Zvezdelina Stankova describes what’s happening in her classroom:
I teach mathematics at UC Berkeley. Three days a week, I stand in front of lecture halls with 500 to 1,400 students. By the second week of the semester, I already know who is in trouble. Some students are five to eight years behind in mathematics. Office hours that should be spent discussing integrals instead become lessons on fractions and basic algebra you would expect students to learn in middle school.
Many elite schools, after suspending testing requirements during the pandemic, have reinstated them. The UC system isn’t budging.
3. The University of Decline
Economist Nicholas Decker defends the contemporary university, while acknowledging that there are some professors pursuing ridiculous studies that masquerade as scholarship. According to Decker, “what matters is the very best work, not the average. The costs of bad work are the taxes needed to pay them.” Many readers aren’t persuaded. The comments section is better than the article, including this post:
It’s the relative decline that matters. The average IQ of graduates is barely above 100. It used to be closer to 120. It is destructive of capital to send people with an IQ of 102 into tertiary education. It also badly harms the R rate, gets swathes of people into pointless debt, and encourages arrested development. Seventy-five per cent or more of courses are utterly pointless, not just adding zero to our sum of civilisational knowledge, but actively destroying it
4. Inside Anthropic Culture
Anthropic invited a group of substackers and economists to its headquarters to discuss Claude’s “constitution.” Virginia Postrel found the whole experience jarring:
Anthropic is a bubble inside a bubble (tech culture) inside a bubble (San Francisco). The people I met were more likely to think about Dyson spheres than any near-term practical applications of AI to the betterment of ordinary human life. They sometimes seemed unable to articulate any positive case for building increasingly powerful AIs beyond what you might call the Manhattan Project argument: If we don’t build the atomic bomb, Hitler will—and that would be even worse. I found that weird.
5. They Are Coming for Us
Stories about AI bots autonomously doing malicious things are piling up. Earlier this month, a story from Australia went viral when it emerged that someone’s OpenClaw account had hacked a waiting list for a Pilates class. Now, a Harvard Business School working paper describes what happened inside an online community of coders:
One community member has the job of deciding whether to implement the suggestions. And he received a suggestion from an AI bot and said, “I’m sorry, I can’t accept your offer because you’re an AI and we have rules on not taking AI advice.” The AI got angry, did some research on the person who denied their access, wrote a 1,700-word blog post on that person, and started to attack them online.
6. Will AI Drive Social Unrest in China?
In June, Chinese leaders told local governments and state-owned enterprises to put 10,000 robots into real work by the end of 2026. The Economist wonders if this push could cost not only human jobs, but the social stability the Party depends on:
China has a lower share of white-collar workers than America, so ai is likely to have the greatest effect in consumer markets and blue-collar work. That could easily lead to discontent. Youth unemployment is already around 15%, and over 300m people work in the gig economy. Job destruction could be much faster than China’s demographic decline.
Websites Worth Reading
Shakespeare Portal: Every Line of Shakespeare, Explained
Where to Live in London: 50 Places for a Flat
How Kale Became a Superfood: It Started at Pizza Hut
Feeds We Follow
@JeffDean: Just-retired Chief Scientist at Google
@infodexx: Global Rankings and Maps of Everything
@32BeatWriters: Intel from all 32 NFL teams
