The GenAI models are now the new search engines. Previously, of course, search engines were built to find page links. Generative AI models are built to answer questions. And the questions can include requests to find links, if they are desired.
That functional shift changes a lot of things. Instead of sending you to ten blue links, GenAI reads across sources, thinks, and returns a direct, contextual response often with follow-up help (summaries, drafts, tables, emails) in the same conversation.
Why the shift is happening
Answers over links: GenAIs generate concise, human-readable explanations rather than a ranked list of pages to open and scan.
Context sticks: They remember what you asked earlier in the session and adapt. That’s hard to replicate with traditional search.
Multi-step help: beyond facts, GenAIs can structure plans, compare options, and transform data (e.g., turn a PDF into bullet points, or distil Excel file into insights).
Fewer dead-ends: When information is fragmented or jargon-heavy, GenAIs bridge gaps and provide plain-English summaries.
A few key facts
GenAIs are generative, not merely retrieval tools: they produce a composed answer. Search engines primarily retrieve and rank documents. GenAIs can work with your own material: many allow you to upload files (documents, spreadsheets, images) so the answer reflects your context. And they’re interactive: the quality of the output improves as you iterate. Search engines don’t improve their result list just because you asked a better second question.
Tips for getting much better results than Google
Ask for sources: “Cite sources with links and dates” or “show me the top three reputable references”.
Set the brief by building your own constraints: country (e.g. UK), time window (e.g., “since 2023”), your actual objective in requesting the search, and output format (bullets, email, section of a document).
If applicable, provide your files: Upload the spreadsheet/report and say what you want from it (summary, any errors found, chart, comparison with...).
Force structure: “Give a table with columns: option, cost, risk, next step.”
Direct deliberately: Follow up with “shorter please”, “in plain English”, or “challenge the assumptions”.
Cross-check facts: ask the GenAI to provide references and then open them yourself.
Localise: Specify UK spellings, regulations, metrics, and your particular context.
The Bottom Line
GenAIs offer fast thinking, drafts, and/or help with your own files paired with quick source checks for cited facts. GenAIs are a much more productive starting point than a traditional search engine.
