{"course":{"slug":"choosing-your-tools","title":"Choosing and Using the Tools","url":"https://addaly.com/learn/choosing-your-tools","level":"beginner","track":"work"},"lesson":{"slug":"research-and-search-tools","title":"Research and search tools","url":"https://addaly.com/learn/choosing-your-tools/research-and-search-tools","position":61,"of":76,"minutes":9,"answer":"Search-grounded answers are bounded by what the search returned, so click two citations to check they support the claim and read the source list before the report; and where you can supply the sources yourself.","keyIdea":"Search-grounded answers are bounded by what the search returned, so click two citations to check they support the claim and read the source list before the report; and where you can supply the sources yourself, the failure mode becomes a safe one."},"body":{"format":"markdown","text":"## Three products that look similar and are not\n\n**Search-grounded assistants.** Perplexity and the search modes inside the major assistants. They run a search, read a few results, and answer with citations. Fast, good for current information, bounded by what the search returned.\n\n**Deep research modes.** The assistant runs many searches over several minutes, reads dozens of pages, and produces a long structured report with sources. Genuinely useful for a survey of an unfamiliar area. Expensive in tokens and in waiting.\n\n**Grounded-on-your-own-documents tools.** NotebookLM and similar. You supply the sources; the tool answers only from them, with citations back to the passage. This is a different and often better proposition, because the failure mode changes: instead of finding the wrong source, it tells you your sources do not say.\n\n## The failures, and how to catch them\n\n**The citation that does not support the claim.** The most common failure and the easiest to miss. The link is real, the page exists, and it does not say what the answer says. Click two citations on anything you will rely on. Not all of them: two is enough to tell you whether this answer is the careful kind.\n\n> Figure: Two research tools that look alike and fail in opposite directions. Search-grounded answer: Bounded by what the search engine returned; Marketing pages rank well on commercial questions; A citation can be real and still not support the claim; Paywalled, printed and other-language sources are invisible; Fails by finding the wrong source and sounding certain. Grounded on documents you supplied: Bounded by the material you chose; Citations point at a passage you can open; Questions outside the material get a refusal; Retrieval can miss a passage worded differently; Fails by saying your sources are silent when they are not. Click two citations on anything you will act on, not all of them; two is enough to tell you whether this answer is the careful kind. Where you can supply the sources yourself, the failure mode becomes the safe one.\n\n**Search-engine junk as a source.** The tool reads what the search engine ranked, and for commercial questions that is often marketing copy or an article written to rank rather than to inform. An answer built from three such pages is fluent and worthless.\n\n**Recency confusion.** The model's training gives it an out-of-date belief; the search gives it a current page; the answer sometimes blends them. Watch for answers that mix a current fact with an obsolete one, particularly on prices, versions and job titles.\n\n**The one-source answer.** Some answers rest entirely on a single page dressed up with several citations to the same domain. Check the source list for variety, not just for length.\n\n**Missing what is not on the web.** Anything behind a paywall, in a book, in a database, or in a language the search did not query, is invisible. A survey of the literature built only from what is freely readable is a survey of what is freely readable.\n\n## What deep research is for\n\nIt is at its best when you know nothing about an area and need a map: who the main players are, what the standard positions are, what vocabulary to use, what to read next. As an orientation exercise it can save a day.\n\nIt is at its worst as a final answer on anything contested, because the length and structure create an impression of thoroughness that the underlying source selection may not deserve. Read the source list before the report. If the sources are weak, the report is weak however well it reads.\n\n## The grounded-on-your-own-documents option\n\nFor anyone with a body of material (a syllabus, a regulation, a set of contracts, a year of meeting notes) this is the most reliable configuration available to a non-programmer. The answers come from documents you chose, the citations point to a passage you can check in one click, and questions outside the material get \"the sources do not cover this\" rather than an invention.\n\nIt is not perfect. Retrieval can miss a passage that is relevant but worded differently, and an answer that says the sources are silent is sometimes wrong about that. But the direction of the failure is far safer than the alternative.\n\n## Choosing between them for a given question\n\nThe rule is short. If the answer must be current (a price, a rule, a fixture, who holds an office) use a search-grounded tool and click the citation. If the answer must come from a specific body of material you hold, use the grounded-on-your-own-documents route. If you need to understand an unfamiliar field, use deep research and treat its output as a reading list.\n\nAnd if the answer is stable general knowledge, a plain assistant with no search is often better, because searching introduces the possibility of a bad page where the model's own knowledge was adequate. More retrieval is not uniformly safer.\n\n## The habit\n\nFor anything you will act on: read the sources, not the summary. The summary is a convenience for finding which sources to read, and treating it as the finished product is where people get caught. A researcher who uses these tools well spends the saved time on reading the three papers that matter rather than on the forty they no longer had to skim."},"check":{"question":"A deep research report reads thoroughly and cites twenty sources, eighteen of them from one commercial domain. What does this indicate?","options":[{"text":"The topic is narrow enough that one authoritative source legitimately dominates the available material.","correct":false,"tempting":"Grants authority on the basis of volume, when concentration on a single commercial domain is what needed explaining."},{"text":"Deep research modes weight recent pages heavily, so an actively updated site will naturally dominate the sources.","correct":false,"tempting":"Explains why one site might rank often without addressing that a single commercial voice now carries the report."},{"text":"The search tool failed to deduplicate results, so the same source was counted many times.","correct":false,"tempting":"Blames a counting artefact rather than the substantive problem that one interested party supplied nearly all the content."},{"text":"It rests on one perspective, and the citation count creates an impression of breadth it lacks.","correct":true}]},"prev":{"slug":"video-tools","title":"Video tools, generated and edited","url":"https://addaly.com/learn/choosing-your-tools/video-tools"},"next":{"slug":"document-and-ocr-tools","title":"Documents, PDFs and OCR","url":"https://addaly.com/learn/choosing-your-tools/document-and-ocr-tools"},"authorship":"Founder and series editor: Dr Mustafa Mun. Written by AI models to a written standard. Exam keys checked blind by a second model: 3,665 of 3,666 agreed. Academic review invited.","review":{"status":"Draft, open for academic review"},"cite":"Addaly (2026) 'Research and search tools', in Choosing and Using the Tools. Edited by M. Mun. Addaly. Available at: https://addaly.com/learn/choosing-your-tools/research-and-search-tools (Accessed: 13 October 2026).","licence":{"name":"CC BY-NC-SA 4.0","url":"https://creativecommons.org/licenses/by-nc-sa/4.0/"},"reader":{"schema":1,"title":"Research and search tools","courseTitle":"Choosing and Using the Tools","module":{"index":7,"slug":"beyond-chat","title":"Beyond the chat box"},"answer":null,"keyIdea":"Search-grounded answers are bounded by what the search returned, so click two citations to check they support the claim and read the source list before the report; and where you can supply the sources yourself, the failure mode becomes a safe one.","hero":{"kind":"figure","figure":{"t":"figure","label":"Figure","figure":{"left":{"title":"Search-grounded answer","items":["Bounded by what the search engine returned","Marketing pages rank well on commercial questions","A citation can be real and still not support the claim","Paywalled, printed and other-language sources are invisible","Fails by finding the wrong source and sounding certain"]},"type":"compare","right":{"title":"Grounded on documents you supplied","items":["Bounded by the material you chose","Citations point at a passage you can open","Questions outside the material get a refusal","Retrieval can miss a passage worded differently","Fails by saying your sources are silent when they are not"]},"title":"Two research tools that look alike and fail in opposite directions","caption":"Click two citations on anything you will act on, not all of them; two is enough to tell you whether this answer is the careful kind. Where you can supply the sources yourself, the failure mode becomes the safe one."},"alt":"Figure: Two research tools that look alike and fail in opposite directions. Search-grounded answer: Bounded by what the search engine returned; Marketing pages rank well on commercial questions; A citation can be real and still not support the claim; Paywalled, printed and other-language sources are invisible; Fails by finding the wrong source and sounding certain. Grounded on documents you supplied: Bounded by the material you chose; Citations point at a passage you can open; Questions outside the material get a refusal; Retrieval can miss a passage worded differently; Fails by saying your sources are silent when they are not. Click two citations on anything you will act on, not all of them; two is enough to tell you whether this answer is the careful kind. Where you can supply the sources yourself, the failure mode becomes the safe one."}},"blocks":[{"t":"h2","id":"three-products-that-look-similar-and-are-not","runs":[{"text":"Three products that look similar and are not"}]},{"t":"p","runs":[{"text":"Search-grounded assistants.","b":true},{"text":" Perplexity and the search modes inside the major assistants. They run a search, read a few results, and answer with citations. Fast, good for current information, bounded by what the search returned."}]},{"t":"p","runs":[{"text":"Deep research modes.","b":true},{"text":" The assistant runs many searches over several minutes, reads dozens of pages, and produces a long structured report with sources. Genuinely useful for a survey of an unfamiliar area. Expensive in tokens and in waiting."}]},{"t":"p","runs":[{"text":"Grounded-on-your-own-documents tools.","b":true},{"text":" NotebookLM and similar. You supply the sources; the tool answers only from them, with citations back to the passage. This is a different and often better proposition, because the failure mode changes: instead of finding the wrong source, it tells you your sources do not say."}]},{"t":"h2","id":"the-failures-and-how-to-catch-them","runs":[{"text":"The failures, and how to catch them"}]},{"t":"p","runs":[{"text":"The citation that does not support the claim.","b":true},{"text":" The most common failure and the easiest to miss. The link is real, the page exists, and it does not say what the answer says. Click two citations on anything you will rely on. Not all of them: two is enough to tell you whether this answer is the careful kind."}]},{"t":"p","runs":[{"text":"Search-engine junk as a source.","b":true},{"text":" The tool reads what the search engine ranked, and for commercial questions that is often marketing copy or an article written to rank rather than to inform. An answer built from three such pages is fluent and worthless."}]},{"t":"p","runs":[{"text":"Recency confusion.","b":true},{"text":" The model’s training gives it an out-of-date belief; the search gives it a current page; the answer sometimes blends them. Watch for answers that mix a current fact with an obsolete one, particularly on prices, versions and job titles."}]},{"t":"p","runs":[{"text":"The one-source answer.","b":true},{"text":" Some answers rest entirely on a single page dressed up with several citations to the same domain. Check the source list for variety, not just for length."}]},{"t":"p","runs":[{"text":"Missing what is not on the web.","b":true},{"text":" Anything behind a paywall, in a book, in a database, or in a language the search did not query, is invisible. A survey of the literature built only from what is freely readable is a survey of what is freely readable."}]},{"t":"h2","id":"what-deep-research-is-for","runs":[{"text":"What deep research is for"}]},{"t":"p","runs":[{"text":"It is at its best when you know nothing about an area and need a map: who the main players are, what the standard positions are, what vocabulary to use, what to read next. As an orientation exercise it can save a day."}]},{"t":"p","runs":[{"text":"It is at its worst as a final answer on anything contested, because the length and structure create an impression of thoroughness that the underlying source selection may not deserve. Read the source list before the report. If the sources are weak, the report is weak however well it reads."}]},{"t":"h2","id":"the-grounded-on-your-own-documents-option","runs":[{"text":"The grounded-on-your-own-documents option"}]},{"t":"p","runs":[{"text":"For anyone with a body of material (a syllabus, a regulation, a set of contracts, a year of meeting notes) this is the most reliable configuration available to a non-programmer. The answers come from documents you chose, the citations point to a passage you can check in one click, and questions outside the material get “the sources do not cover this” rather than an invention."}]},{"t":"p","runs":[{"text":"It is not perfect. Retrieval can miss a passage that is relevant but worded differently, and an answer that says the sources are silent is sometimes wrong about that. But the direction of the failure is far safer than the alternative."}]},{"t":"h2","id":"choosing-between-them-for-a-given-question","runs":[{"text":"Choosing between them for a given question"}]},{"t":"p","runs":[{"text":"The rule is short. If the answer must be current (a price, a rule, a fixture, who holds an office) use a search-grounded tool and click the citation. If the answer must come from a specific body of material you hold, use the grounded-on-your-own-documents route. If you need to understand an unfamiliar field, use deep research and treat its output as a reading list."}]},{"t":"p","runs":[{"text":"And if the answer is stable general knowledge, a plain assistant with no search is often better, because searching introduces the possibility of a bad page where the model’s own knowledge was adequate. More retrieval is not uniformly safer."}]},{"t":"h2","id":"the-habit","runs":[{"text":"The habit"}]},{"t":"p","runs":[{"text":"For anything you will act on: read the sources, not the summary. The summary is a convenience for finding which sources to read, and treating it as the finished product is where people get caught. A researcher who uses these tools well spends the saved time on reading the three papers that matter rather than on the forty they no longer had to skim."}]}],"check":{"question":[{"text":"A deep research report reads thoroughly and cites twenty sources, eighteen of them from one commercial domain. What does this indicate?"}],"options":[{"text":[{"text":"The topic is narrow enough that one authoritative source legitimately dominates the available material."}],"correct":false,"tempting":[{"text":"Grants authority on the basis of volume, when concentration on a single commercial domain is what needed explaining."}]},{"text":[{"text":"Deep research modes weight recent pages heavily, so an actively updated site will naturally dominate the sources."}],"correct":false,"tempting":[{"text":"Explains why one site might rank often without addressing that a single commercial voice now carries the report."}]},{"text":[{"text":"The search tool failed to deduplicate results, so the same source was counted many times."}],"correct":false,"tempting":[{"text":"Blames a counting artefact rather than the substantive problem that one interested party supplied nearly all the content."}]},{"text":[{"text":"It rests on one perspective, and the citation count creates an impression of breadth it lacks."}],"correct":true,"tempting":null}],"order":[2,3,1,0],"why":null},"prev":{"slug":"video-tools","title":"Video tools, generated and edited","minutes":8},"next":{"slug":"document-and-ocr-tools","title":"Documents, PDFs and OCR","minutes":8}}}