The Phony Comforts of AI Optimism

Source: Where’s Your Ed At Date: 2025-03-24 URL: https://www.wheresyoured.at/optimistic-cowardice/

Summary

Ed argues that Kevin Roose and Casey Newton practice “cautious optimism” about AI as a form of intellectual cowardice. Their claims (“AI models are much better”) are vague, undefined, and rely on executive testimony from parties with financial incentives to promote AI. Neither defines what “better” means, provides specific evidence, or asks about economics and actual capabilities. Both previously hyped the metaverse and NFTs without accountability, and are now repeating the pattern with AI. Meanwhile, the real story — OpenAI and CoreWeave burning billions with no profitability path, the entire AI investment structure depending on interconnected debt — goes unexamined.

Implications

This is Ed’s methodological manifesto for AI journalism: skepticism requires uncomfortable, detailed work; optimism requires nothing except willingness to reprint executive statements. The specific naming of Roose and Newton — with their metaverse/NFT track record — is the accountability thread that connects prior hype cycles to the current one. The “interconnected debt” observation about AI infrastructure is the structural vulnerability underneath the optimism: companies whose valuations depend on each other’s continued spending are fragile in ways that cautious optimists never model.