When Does Worrying About Things Trade Off Against Worrying About Other Things?

A sharp short essay probing Acemoglu's claim that worrying about long-term AI risk crowds out concern for near-term AI harms. Scott constructs six parallel 'instead of worrying about X, worry about Y' arguments that all sound absurd, and uses them to build a complements/substitutes/neutral toy model of attention: closely related concerns tend to be complements (police brutality + evidence fabrication), unrelated ones are neutral, and genuine substitutes are rare — so it's suspicious that the only claimed substitute-pair in the intellectual universe is near- vs long-term AI. He then lands the real point: long-term AI funding comes from x-risk-motivated funders (LTFF, Musk) who would never redirect to near-term AI, so the crowding-out framing misreads the funding landscape — the recurring fight persists because the thinkpiece class treats defunding weird-sounding AI risk as costless.

Quality 68 · Strong. Strong-floor (68). Original, well-constructed reasoning (the six-arguments technique, the complements/substitutes model, the funding-landscape correction) that genuinely clarifies a recurring discourse; short and tightly bound to one topical dispute, so not a major essay.

Claude’s paradigm shift 48 · Moderate. Moderate (48). A fresh, useful frame on attention/concern tradeoffs and whataboutism, building on standard economic intuitions.

Real-world impact 1 · Negligible. 1 — a discourse-analysis blog post; no concrete real-world effect.