AI promised less work. I've never worked more
AI coding tools are making individual software developers far more productive, but many report feeling more overworked, stressed, and responsible for an ever‑growing number of parallel tasks. Commenters argue this is a classic case of productivity gains being captured by employers and capital—through higher expectations, layoffs, and intensified work—rather than translating into shorter hours or better conditions for workers. The thread branches into broader questions about Jevons’ paradox, job security, unions, and whether political or collective action is needed to ensure AI augments human well‑being rather than just corporate output.
Workload, Pace, and Burnout
- Many report doing more work, not less: AI speeds up execution, so management and clients simply raise expectations and backlog.
- People juggle many more parallel tasks/agents, leading to heavy context switching and “reverse centaur” feelings: humans serving as supervisors to machines with little downtime.
- Several describe strong cognitive strain from reading and validating dense AI output (code, docs, emails), replacing slower but more meditative “grunt work.”
- Some intentionally cap hours or use dumb phones to avoid constant “productive” AI use creeping into all waking time.
Productivity vs. Quality
- Supporters say AI massively accelerates coding, debugging, docs, CRM, and homelab work; enables solo devs and small teams to tackle projects previously impossible.
- Critics counter that AI often produces half-baked, overengineered, or outdated solutions; you still pay the cost in review, debugging, and maintenance.
- Common pattern: get to “80% done in 1% of the time,” then spend unpredictable effort on the last 20%. Net gain is unclear or sometimes negative.
- Concerns about AI increasing tech debt, repo bloat, and flooding systems with mediocre code and text.
Labor, Leverage, and Compensation
- Strong consensus: in current capitalism, productivity gains mostly accrue to owners, not workers. More output usually means same pay, more pressure, and/or fewer jobs.
- Some argue AI deskills parts of programming, enabling cheaper workers to substitute; others say it widens the gap, making top engineers far more effective than weak ones.
- There is skepticism that companies will voluntarily trade AI gains for shorter weeks or more time off; several see only three realistic outcomes: work harder, get laid off, or exit employment.
Unions, Policy, and Collective Response
- Many see organized labor and political action (shorter workweek, labor laws, taxation, social insurance) as the only real path to “less work, same pay.”
- Others argue unions and regulation have limited power in a globalized, easily offshored and automated knowledge-work market.
- Broader fears include techno-feudalism, extreme inequality, mass unemployment, and social unrest if AI-driven gains aren’t redistributed.
Self‑Employment and Entrepreneurship
- Some founders and solo devs report genuine lifestyle gains: AI handles marketing, accounting, basic support, letting them focus on core product and work fewer hours.
- Others running small shops say AI heightens competition, raises client expectations, and pushes prices down, so they work more for lower effective hourly pay.
- Recurrent view: coding is now the easy, cheap part; marketing, sales, distribution, and support remain the hard bottlenecks.