AI search engines review: they are worse than Google at navigational queries, mixed on information queries, and offer helpful citations in exploration queries
The AI search tools are getting better — but they don't yet understand what a search engine really is and how we really use them. X: @banafaahmed , @gaganghotra_ , and @catcheronthesly X: Prof. Ahmed Banafa / @banafaahmed : The AI search tools are getting better — but they don't yet understand what a search engine really is and how we really use them. https://www.theverge.com/... Gagan Ghotra / @gaganghotra_ : Of course friends of SEO community from Verge 😏 did little bit of digging and found that of course we SEOs have NOT ruined the internet 😏 https://www.theverge.com/... [image] Ved Nayak / @catcheronthesly : 'The AI search tools are getting better — but they don't yet understand what a search engine really is and how we really use them.' Here's why AI search engines really can't kill Google https://www.theverge.com/...
Context & Ripple Effects
This review distinguishes among navigational, informational, and exploratory search rather than treating AI search as a single substitute for Google. That framing fits later coverage finding that AI search was more of a user-interface revamp than a replacement for blue-link search.
The citation benefit in exploratory queries is consequential because it makes source attribution part of the product experience. But later reporting on incorrect news citations across AI search engines shows that visible citations do not by themselves establish reliability.
First-order effects
- Google retains a practical advantage for users trying to reach a known site or destination, while AI search tools are more useful for open-ended research that benefits from surfaced sources.
- AI-search users must choose the interface by task: concise synthesis and citations can help exploration, but informational answers remain uneven and navigational requests are a weak fit.
Second-order effects
- AI-search providers face pressure to improve query-intent recognition and source selection, rather than optimizing only for fluent answer generation; failures on short queries were also evident in early ChatGPT Search testing.
- Publishers and SEO teams have an incentive to track whether AI citations create meaningful referral paths, not merely brand exposure, as answer interfaces mediate discovery.
Third-order effects
- Search is likely to evolve into a task-segmented market: conventional ranked results remain important for navigation and verification, while AI layers compete for research and synthesis workflows.
- If citation quality remains inconsistent, trust and publisher attribution will become a central constraint on AI-search adoption, strengthening demand for clearer controls and accountability around use of web content.
The trend: AI search is developing as a complementary answer-and-exploration layer, with its long-term position determined by task reliability and trustworthy source attribution rather than conversational UX alone.