How to Turn Off AI Overview in Google and Set "Web" as Default

Original Article ↗ Hacker News Discussion ↗

Google’s new AI Overview feature in search results is prompting some users to seek ways to disable it or bypass it entirely by forcing the classic “Web” view or switching to alternative search engines like DuckDuckGo and Kagi. Many commenters criticize the AI summaries as inaccurate, intrusive, and harmful to both user trust and the wider web ecosystem, especially publishers who rely on clicks. Others report finding the overviews useful for quick answers, framing the change as part of a broader tension between “innovation,” product quality, and Google’s financial and organizational incentives.

Ways to Disable or Bypass AI Overview

  • Many prefer to “disable” AI Overview by simply abandoning Google Search or using it only via special modes.
  • Several mention appending udm=14 to the Google URL to force the “Web” view.
  • Some describe browser-specific tricks (e.g., custom search engines in Firefox), though instructions in the article don’t match all platforms (Ubuntu GNOME, Windows, macOS).
  • Others avoid the problem entirely by using non‑Google search engines or command-line tools that scrape SERPs as plain HTML.

Migration to Alternative Search Engines

  • Strong sentiment to switch to other engines: DuckDuckGo, Kagi, Qwant, StartPage, Brave Search, Bing.
  • DuckDuckGo’s “bangs” feature (!g, !w, etc.) is praised as a seamless fallback to Google.
  • Some note DuckDuckGo is largely Bing-backed and ad-supported but still perceived as less intrusive.
  • Kagi receives repeated positive mentions for: clean UI, “quick answers” with explicit citations, domain filters, and non-intrusive, opt‑in AI features.
  • A few say they almost never need Google anymore; others see Google as still necessary for edge cases.

Quality and Reliability of AI Overviews

  • Many examples are shared of AI Overview being factually wrong, incoherent, or internally inconsistent.
  • Some find this dangerous for medical-style queries where users may not fact-check.
  • Others report mostly positive experiences and say it often gives them exactly what they need, faster than clicking links.
  • Non-determinism is criticized: different users or times can yield different answers, making trust and reproducibility harder.
  • A minority likes that AI overviews “burn Google’s cash,” but still sees them as early-stage.

User Experience and UI Concerns

  • Frequent complaints that AI Overview and other modules push organic results far down the page, especially on mobile.
  • Some dislike the “reels-like” scrolling experience and overall visual clutter compared with older, simpler Google.
  • Others appreciate AI for quick summaries of things they already understand (e.g., syntax snippets) to avoid ad-ridden pages.

Impact on the Web and Publishers

  • Concern that AI summaries will reduce clicks to sites that rely on ads or affiliate links.
  • Some speculate Google may eventually bias AI content for commercial reasons, further distorting the ecosystem.
  • One comment notes that if sites themselves start using AI to generate content, the harm may be circular.

Views on Google’s Strategy and Culture

  • Many see AI Overview as driven by hype, shareholder pressure, and internal promotion incentives (“ship new AI” vs. improve existing products).
  • There is nostalgia for older, focused Google products and frustration with a decade of perceived missteps (Reader shutdown, chat fragmentation, Stadia, constant UI churn).
  • Some argue that as long as revenues, profits, and stock price are at records, the strategy is “working” for shareholders, even if users feel product quality is declining.
  • Others blame a fragmented, metrics-driven culture and lack of coherent vision; Google is described as fiefdoms competing for performance metrics.

Broader Reflections on AI and “Innovation”

  • Several commenters are wary of AI being bolted onto everything, calling it half-baked and unsuitable for a mainline, mass product.
  • Others defend experimentation, arguing that integrating AI into core search is a major, necessary change and will improve over time.
  • Some predict an AI investment bubble and eventual commoditization (local LLMs, cheap GPUs); others think it will remain transformative.
  • There is a theme that “innovation” is often a euphemism for user-hostile changes driven by revenue and hype rather than clear user benefit.