The last time my family was replaced by technology

Retraining, displacement, and job quality

• Several commenters argue “retraining” after technological shocks is often a euphemism for downward mobility: skilled workers become low‑status, low‑pay service workers (e.g., typesetters → prepress at minimum wage, factory workers → retail/gig work).
• Others push back: some workers do retrain successfully; adaptability and willingness to learn matter, but going “back to college” is unrealistic for mid‑career people with mortgages and families.
• Even when retraining works, the new job often has worse conditions, pay, or respect, so “successful adaptation” may still feel like loss.

Horse analogy and “this time is different”

• The horse→car story is used heavily:
  • Pessimists: humans are now the horses; once machines can do all economically useful tasks, most people become redundant.
  • Optimists: humans are not horses; we’re adaptable, we’re consumers, and economies will invent new roles.
• Debate over whether many jobs will never be worth automating (e.g., physical trades, niche instructors) vs. view that cost curves and inequality will eventually push automation into almost everything, including luxury “human touch” services.

Scope and pace of AI vs prior tech revolutions

• Some say AI is “just another” disruption like railroads, cars, computers, or spreadsheets; humanity has always predicted doom and then adapted.
• Others stress differences:
  • Previous waves mostly automated muscle; AI targets reasoning and knowledge work.
  • Past changes played out over decades and often across generations; AI diffusion (and cloud centralization) feels simultaneous and extremely fast.
  • Unlike cars, AI can potentially improve and work on itself, shrinking the residual space for humans.

Meaning, mastery, and the nature of work

• Many programmers distinguish two camps:
  • People who primarily value outcomes (solving problems, shipping products) tend to embrace AI assistance.
  • People who love the process (architecture, debugging, deep understanding, craftsmanship) feel their source of meaning is being hollowed out.
• Some propose new forms of mastery: orchestrating agents, understanding models deeply, focusing on system‑level design and maintenance, or building personal tooling.
• Others are unconvinced that “prompting” or supervising opaque systems is a satisfying or respectable craft; they describe an “identity crisis.”

Social, economic, and political implications

• Concerns go well beyond individual careers:
  • If AI + robotics can do nearly all labor, what happens when capital no longer “needs” most people? Fears include a surplus “useless class,” intensified inequality, and weakened bargaining power (strikes don’t work if no human labor is needed).
  • Comparisons to NAFTA and deindustrialization: headline GDP can rise while large regions are permanently hollowed out.
  • Some foresee states preferring make‑work jobs for social control rather than unconditional welfare.
• Others argue that massive unemployment historically leads to unrest and political change; they expect government intervention long before total displacement.

Current impact and near‑term uncertainty

• Some claim fears are overblown: today’s LLMs still require humans in the loop; many industries and physical jobs remain largely unchanged.
• Others counter that for math and coding they’re already highly capable, and that benchmark progress over ~5 years has been extraordinary; they expect continuing rapid improvement.
• Many note that even partial automation changes hiring: fewer juniors, more pressure on mid‑career workers, and more “developer‑adjacent” roles with AI doing much of the heavy lifting.
• There is broad uncertainty—and disagreement—about what concrete, “good” jobs displaced knowledge workers should retrain into, and whether those paths will remain safe for long.