A traditional search engine gives you a list of the pages most relevant to your query; opening, reading and drawing conclusions is up to you. AI-powered search tries to do that step for you: it searches the web, reads the pages it finds, combines the information into a short answer and shows the sources it used. Perplexity, founded in 2022, was one of the pioneers of this approach, and Google, Microsoft and OpenAI have since added similar features to their products.
How does it work?
These systems usually use a method called “retrieval-augmented generation” (RAG). First, current pages relevant to your question are found; then a language model writes an answer using the information on those pages. Numbered links in the answer show which sentence came from which source. This makes it possible to produce answers based on current web content rather than only on the model’s training data.
Pros
- For complex questions it gives a quick summary without opening several sources one by one.
- The topic can be explored further with follow-up questions: “So what are the drawbacks?”
- Source links show where the information came from.
Cons
- Faulty summaries: The model may misread information in a source or attribute to a source something it does not say.
- Source quality: Unreliable pages or promotional content can creep into answers.
- Freshness: On very recent events, information may be incomplete or contradictory.
- Impact on publishers: Because users get answers without clicking, sites that produce content may lose traffic, which has led to copyright and licensing debates.
Tips for using it well
- For important information, open at least one source and read the relevant part yourself.
- Prioritise sources from official bodies, academic publications and established news organisations.
- Compare dates, figures and quotes in the answer with the source.
- On health, law and finance, treat the AI summary as a starting point and consult an expert for the final decision.
AI tools’ features, plans and prices change often. Always check the tool’s official page for current information.
Frequently asked questions
Will AI search replace classic search?
There is no definite answer. AI summaries are popular for quick information needs, while classic search is still widely used for detailed research, shopping and reaching a specific website.
Sources and further reading
- Patrick Lewis et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks”, NeurIPS, 2020.
- Google, “AI Overviews” help pages.
- Pew Research Center, research on how AI summaries affect clicking behaviour (2025).
Leave a comment
Your email address will not be published. Required fields are marked *