Google will start showing AI-powered search results for users who didn't opt-in
Google’s move to roll out AI-generated “overview” answers directly in search results for users who never opted in is prompting concern over accuracy, transparency, and the company’s ad-driven incentives. Commenters worry that LLM summaries will both mislead users and siphon traffic and revenue from the sites that supply the underlying information, accelerating the “dead internet” effect. Others welcome faster, synthesized answers or have already shifted to alternative engines like Kagi and SearxNG, highlighting a broader shift in how people expect to find and consume information online.
Reactions to Google’s AI Overviews Becoming Default
- Many see this as a tipping point to seek alternative search engines (Kagi, SearxNG, Bing, LLMs directly).
- Others view it as a normal product change: users rarely “opt in” to new features anyway.
- Some users already had SGE enabled and report good experiences, especially when summaries show multiple “stanzas” with source carousels.
Trust, Quality, and UX
- Strong skepticism about LLM reliability: hallucinations, mixing up facts, and “distilled blogspam.”
- Others say AI summaries are often “good enough,” especially for programming or quick factual questions.
- A recurring preference: search as a “librarian” that surfaces sources vs. a “black box” that answers directly.
- Kagi’s approach (optional AI, clear citations, minimal nagging) is praised as a better UX model.
Incentives and Business Model
- Concern that Google’s goals (maximize ad impressions, engagement) conflict with users’ goal (accurate, useful information).
- Some argue Google has long tried to keep users from clicking through (info boxes, rich snippets); AI Overviews are the next step.
- Debate on whether sources deserve revenue sharing when their content is summarized; disagreement over “who owns” facts vs. value-added synthesis.
Impact on the Web Ecosystem
- Fear that AI answers will starve content creators of traffic, killing remaining high-effort, organic content.
- Counterpoint: much of what dies will be SEO spam; high-quality sources (research, docs, forums, GitHub) will persist due to external incentives.
- Worry about a “dead internet” flooded with AI-generated junk, creating a feedback loop of low-quality training data.
Broader Social and Ethical Concerns
- Anxiety about platform power: AI layers let Google/Amazon further mediate and shape what people see (ads, political or corporate spin).
- Debate around bias and RLHF: some see “DEI” shaping as ideological filtering; others say tuning is mostly to avoid PR disasters and regulation.
- General unease that building a “good product” is increasingly secondary to monetization.