A sustainable web career, for when all this blows over

AI-driven coding tools and agents are transforming web development, raising doubts about advice that suggests technologies like React are “over” and that the industry will simply revert to a pre-AI status quo. Commenters broadly agree that large language models are here to stay and already good enough to automate many low-end websites, but differ on whether this represents a temporary hype bubble or a lasting structural shift that will deskill large parts of the profession. Many argue that future-proof careers will depend less on hand-coding and specific frameworks and more on system design, fundamentals, and knowing how to direct AI effectively.

React, Frontend, and the Article’s Core Claim

  • Many disagree with the idea that “React is out”; they say job listings are still overwhelmingly React-focused.
  • Some argue the article is nostalgic, assuming we’ll “go back to 2019” dev practices, which others see as unrealistic.
  • A few defend the article’s emphasis on fundamentals, UX, and accessibility, but think its framing (“when this blows over”) is misleading or wishful.

AI/LLMs: Hype vs. Permanence

  • Broad consensus: LLMs are already useful and not going away, even if current vendors or VC-backed models crash.
  • Some compare this to the dot-com boom: a bubble may burst, but the underlying tech (like the internet then, LLMs now) will remain and grow.
  • Others warn of hype, likening AI enthusiasm to previous tech manias, but still acknowledge ML’s real utility.

Changing Nature of Web and Software Work

  • Several say they now rely heavily on AI or agents for coding, especially frontend; some claim they rarely hand-write code anymore.
  • Strong view from some that “frontend will be first to go”: agentic tools can already produce multi-platform UIs faster and often better than typical devs.
  • Others see AI as a powerful tool rather than total replacement: devs shift toward architecture, system understanding, UX, and accountability for running systems.

Low-End Web Dev and Small Sites

  • Multiple anecdotes show AI can already build small-business sites and simple SaaS-style apps that previously cost thousands.
  • Consensus that the “$0–$10k brochure site” segment is under severe pressure from site builders and LLM tools.
  • Similar erosion happened earlier when small businesses shifted from bespoke sites to social media pages.

Skills and “AI-Resistant” Areas

  • Suggested durable skills: system design, distributed systems, security, “agentic” automation, clear specs, testing, and cross-platform thinking.
  • Some argue deep programming fundamentals still matter; others claim code syntax itself is becoming niche.

Broader Anxiety, Politics, and Social Dynamics

  • Many express fear or despair about deskilling and job loss; some are excited and feel they’re at “the heart” of a historic shift.
  • There’s a contentious side thread about racism and political takes in the article, with debate over calling out bigotry vs. empty name-calling.
  • Several note that even if an AI crash happens, society is unlikely to “go back” technologically unless there’s broader catastrophe.