Contra Acemoglu On...Oh God, We're Doing This Again, Aren't We?

A rebuttal to Daron Acemoglu's WaPo 'worry about AI now, not superintelligence later' piece. Scott names the core fallacy structure ('AI is dangerous now, therefore it can't be dangerous later' -- a non-sequitur he parodies with climate and stock-market versions), notes Acemoglu's own cited paper finds no aggregate labor-market effect yet, and offers the general-purpose-technology analogy: electricity too powers surveillance states, displaces workers, and warps discourse -- any powerful new tech does, so 'it enables evil more efficiently' can't be the argument. The real distinction: superintelligence is not humans-misusing-a-tool but a novel actor whose ethics we only partly control ('technology misusing itself').

Quality 69 · Strong. Strong: a clear, forceful rebuttal with two portable framings (the now-vs-later fallacy; the general-purpose-technology / 'technology misusing itself' distinction); a strong entry in the rebuttal genre, not field-defining.

Claude’s paradigm shift 50 · Moderate. Notable: the 'humans-misusing-tech vs tech-misusing-itself' line cleanly separates short- and long-term AI risk, a useful non-obvious framing.

Real-world impact 2 · Minor. Minor: part of the AI-safety public discourse, but a commentary post without a traceable material effect.