Asking chatgpt to compare two AI tools for literature search and review tasks: a study note

Asking chatgpt (accessed via poe.com) to compare two AI tools for literature search and review tasks: a study note

My question: "I am a University student, currently studying the subject of Research Methods. I come across two Artificial Intelligence (AI) tools; they are perplexity and GPT-3.5. I would like to know what criteria can be used to evaluate the strengths and weaknesses of these two AI tools for performing the literature search and literature review tasks. Could you use these criteria to compare these two AI tools, preferably in the form of a table? Tks."

The reply from chatgpt: "Here’s a comparison table evaluating the strengths and weaknesses of the AI tools, Perplexity and GPT-3.5, for literature search and literature review tasks based on various criteria:

Criteria	Perplexity	GPT-3.5
Purpose	Primarily designed for search and discovery of information	General-purpose language model for various tasks
Search Capability	Strong search capabilities with real-time results	Relies on pre-existing knowledge, not real-time
Source Variety	Aggregates information from multiple online sources	Generates text based on training data, lacks direct sourcing
Ease of Use	User-friendly interface tailored for searches	Versatile but may require prompts for specific tasks
Response Quality	Provides concise and relevant search results	Offers detailed and context-rich responses
Citation Management	Limited citation features	Can generate citations but not directly manage them
Customization	Limited customization options for specific searches	Highly customizable responses based on prompts
Access to Latest Research	Accesses up-to-date online content	Knowledge may be outdated (up to 2021 for GPT-3.5)
User Interaction	Focused on direct query-response	Supports conversational interactions
Learning Curve	Low learning curve	Moderate learning curve, depends on user familiarity
Cost	Free or subscription-based for advanced features	Costs associated with API usage or subscription
Limitations	May not understand complex queries	Can generate misleading or inaccurate information

 

Summary

Perplexity is best suited for real-time literature searches and finding specific information quickly, with a focus on aggregation from diverse sources.

GPT-3.5 excels in generating detailed text and contextual responses, making it useful for summarizing literature and providing insights, but may lack the latest information and direct sourcing capabilities.

This table should help you evaluate the two tools effectively for your literature search and review tasks in your Degree program."

Reference: a collection of blog notes on using chatgpt for research purpose.