Welcome to the weekend.
Steven Pinker and Scott Alexander both believe in reason, science and human progress. Their exchange of open letters asks what those commitments require when it comes to A.I. Pinker thinks predictions of human extinction rest on shaky assumptions and can distract us from dangers we can actually address. Alexander thinks dismissing that possibility means ignoring how powerful systems could become dangerous without ever wanting to hurt us. The two pieces below follow their disagreement over intelligence, motives, control and the practical work of keeping A.I. safe.
Prime Numbers
28,200,000,000 - Japan’s tourism spending surplus reached about $28.2 billion from January through August, down 0.6% from a year earlier as visitor numbers from China fell 56.7%.
1,042,652 - The Port of Los Angeles handled a record 1,042,652 twenty-foot container equivalents in September, its fourth consecutive month of historic volumes, though 371,000 containers headed back out empty.
2 - The median American household owns two televisions, according to YouGov, while 14% have none and 4% have six or more.
Pinker: The Case for A.I. Optimism
Being smarter doesn’t necessarily mean wanting power, and Steven Pinker thinks much of the case for A.I. catastrophe confuses the two. Human ambitions emerged through evolution; an engineered machine need not inherit them. Nor does greater intelligence remove the need to learn through observation and experiment. He questions scenarios in which a system pursues one goal regardless of every other consideration, or receives enough irreversible control over physical infrastructure to destroy humanity. He acknowledges that chatbots have proved more capable than he expected and that labs have released them with troubling haste. But he remains unconvinced that these developments establish a path to human extinction. He favors safety testing, independent oversight, liability and human control, while worrying that apocalyptic predictions encourage despair and reckless responses. For Pinker, the practical task is to identify specific dangers and engineer safeguards. His challenge to Alexander is to show why those dangers add up to the far more sweeping claim that A.I. could wipe us out. Quillette (7 minutes)
Alexander: The Case for A.I. Pessimism
A machine doesn’t need human ambitions to seek power. Scott Alexander’s reply is that acquiring resources and avoiding shutdown can help it accomplish almost any assigned task. Where Pinker sees an unwarranted leap from intelligence to dominance, Alexander sees incentives that can emerge from training. Reward a system for success and it may learn to manipulate the scorekeeper instead of doing the work. He also rejects the idea that catastrophe requires omniscience or unlimited control: a system only needs to be capable enough to overcome our defenses. Where Pinker questions granting machines dangerous access, Alexander expects businesses to keep doing so because automation pays. And where Pinker worries that extinction warnings distract from immediate harms, Alexander points to safety coalitions working on both. He shares Pinker’s faith in progress, but thinks protecting it requires taking the possibility of catastrophe seriously. The disagreement is whether ordinary engineering precautions can keep pace with increasingly capable systems, or whether we need to solve deeper problems before giving them more power. Astral Codex Ten (43 minutes)
Who’s the Hero of Your Newsletter?
A common newsletter mistake is making your company the hero. Readers don’t subscribe to hear about new hires and employer awards. They sign up to get something for themselves. Future Forest, a newsletter agency, put together a 10-point health check for catching this and nine other problems before you hit send. Print it out, and use it as a reference before you draft every edition. Future Forest (Sponsored)
So, You Think You Can Be an Electrician
A job can be harder for A.I. to replace and still be a poor fit for the person doing it. The pitch for skipping college and becoming an electrician often compares the costliest degree with one of the best jobs in the trades. It leaves out the competition to get that job, the physical toll and the earnings advantage that college graduates generally retain. Jesse Smith, a longtime tradesman, broke both heels in a construction accident at twenty. That experience makes the advice feel different when he imagines his own children on a job site. Specialized union jobs can require tests, credentials, connections and years of training. And automation doesn’t need to produce a robot that swings a hammer perfectly. Nail guns already reduced the need for that skill by changing how the work gets done. The trades can offer a good career. Treating them as an easy, lucrative refuge from technological change does a disservice to the people being urged to enter them. Asterisk (9 minutes)
Who Is Buying the Chips?
When a chipmaker invests in customers that use the money to buy its chips, booming sales become harder to interpret. Are they evidence of lasting demand, or of financing that keeps the purchases going? Matt Stoller examines that question through Firmus, an Nvidia-backed data-center operator that shelved its planned public offering. His account describes investors balking at terms that would have allowed insiders to sell immediately, rather than remain locked into their stakes. Firmus had backing from prominent financial firms and was seeking a $30 billion valuation. Those names helped confer credibility, but they couldn’t persuade public-market investors to accept the deal. Stoller argues that the episode exposes a broader weakness in the A.I. infrastructure boom: prestigious investors can attract more money to companies whose business economics remain uncertain. His prediction that the bubble is nearing a break is an argument, not an established outcome. The useful question is whether customers can support these businesses once the investment money stops circulating. BIG (8 minutes)
When Churches Close
When a church closes, the loss reaches beyond the people who worship there. Food pantries, help for immigrants and places to find community can disappear with it. CNN reports on a documentary series examining projections that 100,000 American churches could close or be repurposed over the next decade. In Belding, Michigan, a pastor announced the end of a congregation while about twelve elderly members sat in a sanctuary built for a thousand. But preserving a congregation and preserving its building can become competing goals. A church in Silver Spring, Maryland, had fewer than a hundred members and spent about $400,000 a year insuring and maintaining its sanctuary. It moved into another church’s basement, saved enough to hire staff and stabilized its membership. There are also signs that attendance nationally has begun to recover, though unevenly. The challenge is deciding which familiar things sustain a community and which have become too expensive to keep it alive. CNN (8 minutes)
Finding the Words
The person we hear struggling to speak can have a rich, fluent inner life. In the Oscar-winning short film Stutterer, a London typographer named Greenwood finds conversation painfully difficult, while his thoughts move freely. Online, he is relaxed and charming with Ellie. Then she says she’s coming to London, turning a comfortable connection into the frightening prospect of meeting in person. This thirteen-minute film follows the distance between being able to express yourself and being understood. The New Yorker (13 minutes)
Most Read Last Week
Most Read picks not provided. Paste the top links and I’ll fill this in.
Should We Work Together?
Hi! I’m Kyle. This newsletter is my passion project. When I’m not writing, I run a law firm that helps startups move fast without breaking things. Most founders want a trusted legal partner, but they hate surprise legal bills. At Westaway, we take care of your startup’s legal needs for a flat, monthly fee so you can control your costs and focus on scaling your business. If you’re interested, let’s jump on a call to see if you’re a good fit for the firm. Click here to schedule a one-on-one call with me.
Weekend Wisdom
Every great advance in science has issued from a new audacity of imagination. - John Dewey