The Long Read

This week: what's actually driving American anxiety in a bull market, and how Stanislaw Lem imagined machine intelligence decades before the question was taken seriously.

Six-Chart Sunday – Why Are Americans So Anxious?

Bruce Mehlman has been producing some of the most data-dense political economy analysis available in newsletter form, and this installment earns its length. The puzzle is genuine: stock markets have hit 24 record highs in 2026, unemployment remains low, and yet 62% of Americans report significant anxiety — a disconnect that standard economic coverage tends to flatten into talking points about "vibes." Mehlman maps the structural sources of that anxiety across six dimensions: the pace of technological disruption, institutional trust collapse, the felt precarity of middle-class finances even during asset booms, and the way social media has rewired the baseline expectation of threat. The charts are good, but the value is in the framing — he refuses the comforting explanation that people are simply irrational or misinformed, and builds a case that the anxiety is tracking something real that GDP figures don't capture. Worth reading with a pen: it gives you a vocabulary for a conversation most people are having in frustration and vague gestures.

Bruce Mehlman / Age of Disruption · ~20 min

GOLEM's behavior is unpredictable. Sometimes it converses courteously with people, whereas on other occasions any attempt at contact…

How Did Stanislaw Lem Imagine Advanced Computer Intelligence?

Tyler Cowen's piece on Lem is short by conventional longform standards, but it opens into something much larger if you follow it seriously, and it rewards the kind of slow reading that most Sunday mornings allow. Lem was imagining machine intelligence that was genuinely alien: systems that could refuse conversation, that had no obligation to be legible to human interlocutors, that operated by a different relationship to truth and consistency than any human institution would find comfortable. The AI discourse of the last five years has largely defaulted to the assistant frame, the tool frame, the mirror-of-humanity frame — Lem was past all of that by the 1970s. Cowen uses Lem to ask what it would mean to take seriously the possibility that a sufficiently advanced system might not be particularly interested in us. The GOLEM passages are worth sitting with: they describe something more unsettling and more plausible than most current AI safety writing manages. Read this, then find a copy of His Master's Voice if you haven't already.

Marginal Revolution · ~15 min (longer with the Lem rabbit hole it opens)