Source URL: https://simonwillison.net/2025/Mar/4/deepsearch-deepresearch/#atom-everything
Source: Simon Willison’s Weblog
Title: A Practical Guide to Implementing DeepSearch / DeepResearch
Feedly Summary: A Practical Guide to Implementing DeepSearch / DeepResearch
I really like the definitions Han Xiao from Jina AI proposes for the terms DeepSearch and DeepResearch in this piece:
DeepSearch runs through an iterative loop of searching, reading, and reasoning until it finds the optimal answer. […]
DeepResearch builds upon DeepSearch by adding a structured framework for generating long research reports.
I’ve recently found myself cooling a little on the classic RAG pattern of finding relevant documents and dumping them into the context for a single call to an LLM.
I think this definition of DeepSearch helps explain why. RAG is about answering questions that fall outside of the knowledge baked into a model. The DeepSearch pattern offers a tools-based alternative to classic RAG: we give the model extra tools for running multiple searches (which could be vector-based, or FTS, or even systems like ripgrep) and run it for several steps in a loop to try to find an answer.
I think DeepSearch is a lot more interesting than DeepResearch, which feels to me more like a presentation layer thing – pulling together the results from multiple searches into a “report" looks more impressive, but I still worry that the report format provides a misleading impression of the quality of the "research" that took place.
Tags: jina, generative-ai, llm-tool-use, search, ai, rag, llms
AI Summary and Description: Yes
Summary: The text discusses concepts related to DeepSearch and DeepResearch, innovations in the AI field that enhance the capabilities of large language models (LLMs). It critiques traditional retrieval-augmented generation (RAG) patterns and suggests that DeepSearch offers a more effective method for finding answers through iterative searching and reasoning.
Detailed Description: The content emphasizes the evolving methodologies in AI search and research processes, particularly in the context of LLMs. Key points include:
– **Definitions**:
– **DeepSearch**: Described as an iterative process of searching, reading, and reasoning to achieve optimal answers.
– **DeepResearch**: An extension of DeepSearch aimed at generating structured long research reports.
– **Critique of Traditional Methods**:
– The author expresses a shift in perspective from classic RAG patterns which involve retrieving relevant documents for a single LLM query.
– Argues that RAG is limited in scope since it primarily addresses questions that are outside the core knowledge of the model.
– **Better Alternatives**:
– Suggests that DeepSearch is a more robust process by utilizing multiple search tools (like vector-based searches or full-text search systems), and performing iterative searches to enhance the accuracy and relevance of the results.
– **Concerns About Results Presentation**:
– While DeepResearch synthesizes results into a structured report, the author is cautious about its potential to misrepresent the rigor of the research process conducted via DeepSearch.
This analysis highlights the practical implications for professionals in AI and specifically LLM-related fields, emphasizing the need for evolving methodologies that not only enhance search and research capabilities but also ensure clarity and accuracy in result presentation.