I am done with this shit
Engineers describe increasingly “soul-sucking” workplaces where large language models generate specs, code, and tickets, while humans are pushed to ship faster without truly understanding or reviewing what’s produced. Many fear a hollowing-out of software craftsmanship, job quality, and team culture as AI accelerates existing corporate pressures to prioritize output over comprehension, quality, and mentoring. Others counter that the role is simply evolving toward higher-level design and problem-solving, arguing that AI can be a powerful tool if used with discipline, but acknowledge the risk of fragile, incoherent systems and a painful jobs correction along the way.
State of AI‑Driven Corporate Engineering
- Many describe big-company and agency environments where nearly all specs, tickets, and code are produced by LLMs under extreme “ship fast” pressure.
- Engineers feel reduced to “pressing enter” on AI output, with little time to understand systems, review code, or fix bugs.
- Some note this culture predates AI (“code grinding,” Scrum hellholes); LLMs simply amplify existing dysfunction.
Work Quality, Maintainability, and Comprehension
- Reports of “vibe‑coded” repos: over‑engineered, inconsistent with surrounding code, hard to follow, and brittle.
- Concern that AI-generated logic is arbitrary, poorly defined, and accumulates cognitive debt faster than humans can pay it down.
- Others argue LLMs can help refactor legacy messes, identify bottlenecks, and at least enforce minimal review in places that previously had none.
- Strong theme: you can’t ship faster than you can comprehend without sacrificing quality.
Jobs, Careers, and Market Dynamics
- Anxiety about overemployment in software and an impending correction; glue‑code and framework configuration roles seen as especially vulnerable.
- Some expect fewer but more demanding, better‑paid roles in security, reliability, and “real” engineering; others think traditional software craftsmanship is becoming niche.
- Freelancers feel squeezed: competitors using LLMs can churn out flashy but low‑quality work faster and cheaper, making ethical, maintainable work harder to sell.
Psychological and Social Impact
- Several express burnout, grief, and loss of identity: the craft and camaraderie of building software together is eroding as everyone delegates to LLMs.
- Juniors ask fewer questions; seniors are isolated; product becomes a proxy between management and AI.
- Enthusiasts report the opposite: unprecedented productivity on side projects and solo products, describing LLMs as a “virtual dev team.”
Governance, Culture, and Ironic Details
- Management often treats AI as a path to radical headcount cuts and unrealistic timelines, dismissing pushback as anti‑AI or not being a “team player.”
- Multiple commenters note the irony of an AI bot auto‑summarizing a thread full of complaints about AI hollowing out the work.