History warns: AI as normal tech may cause generational harm

Davidmanheim's LessWrong post challenges the comforting narrative that technology always improves human welfare. He points to the agricultural revolution: while it enabled population growth and surplus, it also introduced widespread malnutrition, increased disease burdens, and larger-scale warfare for millennia. The industrial revolution similarly harmed those immediately affected—factory workers faced grueling conditions, child labor, and pollution, with mortality spiking in the mid-1800s (the 'urban penalty') before sanitation improvements came decades later.

Using ChatGPT's list of five major technological transformations, Davidmanheim examines writing and metallurgy as well. Writing initially empowered elites for taxation and control, not literacy. Metallurgy brought better tools but also fueled conquest and inequality. His conclusion: if AI is a 'normal' technology, we should expect severe transitional harms—perhaps even existential risks—before any net benefit arrives. The piece urges caution against assuming AI will be an exception.