I'm drowning in AI features I never asked for and I hate it

Original Article ↗ Hacker News Discussion ↗

OS bloat, Windows vs. Linux

  • Many commenters connect unwanted AI to a broader pattern of Windows bloat: ads, “suggestions,” lock-screen junk, and now Copilot/AI.
  • Several report switching to Linux (Mint, MX, Fedora, Ubuntu, Zorin, NixOS) or Steam Deck and finding systems “snappy” and distraction‑free.
  • Others argue Windows 11 AI/ads can be disabled, but critics respond that:
    • The out‑of‑box experience matters.
    • Settings get silently reverted or re‑enabled after updates.
    • Debloat scripts are risky and can break legitimate features.
  • Some struggle to move non‑technical relatives off Windows despite repeated Windows Update disasters; inertia, fear of Linux, and dependence on desktop Office are strong.

Inescapable AI and bad UX

  • Users resent AI widgets that:
    • Pop in late and reflow UI, causing misclicks (Firefox context menu, AI bubbles, Google Sheets/Gemini, Atlassian, Confluence, Kagi, Amazon Q).
    • Sit as permanent floating buttons (Android Messages Gemini FAB, Kagi “Quick Answers”) with no obvious off‑switch.
  • Some note AI can be globally toggled on iOS or mostly disabled on Windows, while others fear opt‑outs will vanish or be overridden.

Web, shopping, and content slop

  • Examples of useless or wrong AI answers abound:
    • Target’s AI Q&A giving non‑answers (“ensuring stability” instead of weight limit).
    • Amazon replacing human Q&A with hallucinated product differences.
  • Users see forums, Reddit, Twitter/X replies, and product Q&A flooded with AI‑generated “engagement farming” and spam, making authentic voices harder to find.

Public sentiment and “AI bubble”

  • Many outside tech circles reportedly dislike AI or feel apathetic; comparisons are made to web3/crypto, 3D TV, and the “metaverse.”
  • Some argue AI is a hype‑driven, too‑big‑to‑fail bubble tied to stock prices and internal KPIs, so companies keep shoving it everywhere regardless of user value.
  • Others counter that AI genuinely helps with things like documentation search, wiki summarization, and “how do I…?” queries, even if current UX is poor.

Customer service and automated punishment

  • Strong resentment toward AI‑driven moderation and support:
    • Automated bans in games and social media with no human appeal path.
    • Stories of locked accounts where the only realistic fix is insider help or grey‑market “unbanning” services.
  • Some note human support was already being hollowed out; AI chatbots are seen as cheaper but still inadequate.

Ownership, decentralization, and coping strategies

  • Several argue real benefits require on‑device or self‑hosted, open models; corporate AI is seen as censored, rent‑seeking “slop.”
  • Others emphasize decentralization in general (small forums, self‑owned infra) as a defense against enshittification.
  • Practical coping: use adblockers/filters to hide AI UI, switch OSes or apps, or retreat to curated/paid communities that keep bots and AI spam out.