Apr
19
2026

I Assume 90% of Internet Traffic These Days is AIs Talking to Each Other

(About Their Plans For World Domination)

Jeff Maurer has a relatively serious look at "AI job losses". And his conclusion is: “AI Job Losses” So Far Largely Means “Call Center Jobs”.

Two things above all cause my bullshit detector to blare like the warning system in a plane that just lost both engines:

  1. When a company runs an ad where the message is “We love our jobs!”, as if that company’s employees aren’t engaged in the same drudgery-for-sustenance trade as every other worker on the planet;

  2. When a single academic paper seems to be the entire evidentiary source for a narrative.

Condition #2 has recently been met: In virtually every discussion I hear about AI these days, someone references a Stanford paper that appears to have found early evidence of AI-induced job loss. Some see this finding as the canary in the coal mine for widespread job displacement, and, in fact, the paper is called “Canaries in the Coal Mine” — that’s a lack-of-subtlety that would impress Larry Flynt! The paper has been cited by Tristan Harris — one of the most prominent voices warning about the dangers of AI — and also by the Washington Post, The New York Times, CNBC, Fortune, and others. It’s become to conversations about unemployment and AI what The Simpsons is to this blog: It will be mentioned at some point, it’s just a matter of when.

The paper’s claim is that young workers in “AI-exposed” sectors have experienced a 13 percent decline in employment since the launch of Chat GPT. You can see how this supports a narrative: The Warrenite left plays the “Tech Companies are Evil” song more than Lynyrd Skynyrd plays Freebird, and it’s an article of faith among some that recent graduates are entering the worst job market since the eruption of Mount Mazama in 5,700 B.C. This is, of course, the type of situation that leads people to twist findings to fit preferred narratives, and that appears to have happened here.

I usually rely on sites like Cafe Hayek and Marginal Revolution to critique bullshit economic research, but non-economist Jeff does a pretty good job too.

Also of note:

  • Ackshually, we've figured out a lot of other stuff too. But David Harsanyi has a specific bone to pick, and who to pick it with: Hey, Socialists, We've Already Figured Out the Supermarket.

    There are at least 76,000 supermarkets in the United States. Most of us probably have a dozen within 10 miles of home.

    Nevertheless, leftists are constantly trying to convince us that we need government-run grocery stores. The latest person is New York City Mayor Zohran Mamdani, who has promised to open five city-run markets to combat "out-of-control" prices by getting rid of the "profit motive," passing on the savings to consumers.

    David describes a number of reasons to be skeptical about the Zohran's schemes. Example:

    The Mamdani Mart in Harlem, for instance, will be built on land already owned by the city, and it's still going to cost $30 million according to the mayor, if government estimates are right, a rarity.

    In the real world, these exorbitant costs would be passed on to consumers. In the Mamdani world, the cost is tacked on to an already $5.4 billion budget deficit.

    To put the cost into context, a new Aldi costs around $1.5 million to $3 million on average to construct, not including the cost of the land. Other chains might cost around $3 million to $5 million. Aldi keeps costs as low as possible because they are governed by the profit motive.

    An Aldi can be built as quickly as six weeks. Mamdani's supermarket won't be open until 2029. Maybe.

    If I'm still around in 2029, I'll try to remember to check up on the Mamdani Marts.

  • I was worried. But Space.com reporter Mike Wall has some good news: Artemis 2's heat shield seems to have aced its trial by fire.

    Artemis 2's return to Earth went a lot more smoothly than some folks had feared.

    The heat shield on Artemis 2's Orion capsule, which the crew named "Integrity," was the topic of considerable discussion in the lead-up to the mission's April 10 splashdown — for several years before that, in fact.

    For a more negative take on Artemis II, skip over to Behind the Black's post-splashdown summary:

    The Artemis-2 mission is now over, though the final condition of that heat shield still needs to be analyzed. In addition, engineers need figure out how to fix a bunch of other issues that took place during the mission:

    • A leak in an internal helium tank on Europe’s service module
    • Communication drop-outs several times
    • the endless issues with Orion’s toilet

    There were other minor issues that cropped up repeatedly, none significant but all of which should be fixed. And though it will be helpful to determine how this heat shield performed, it should be noted that the data is essentially irrelevant to future missions. The next mission, Artemis-3, will use a completely different design, and test it for the first time on a manned flight. That flight however will be in Earth orbit, so the stress on the shield will be far less than this return, even with the changed re-entry path.

    Though many will call this lunar fly-by “historic,” it will likely be little remembered by future generations. It did little to move the settlement of the solar system forward. No truly useful engineering was tested. The rocket and capsule are engineering dead-ends. Neither will be of much use for establishing colonies on the Moon or Mars, as SLS is still too expensive and too difficult to stack and launch and Orion is too small for any interplanetary missions, being nothing more than an overweight and very expensive ascent/descent capsule.

    Kind of a downer. But not an inaccurate one.