A study note on MS copilot and perplexity (AI tools) evaluation

A study note on MS copilot and perplexity (AI tools) evaluation (information from poe.com)

My question: "I am a Research Methods student, currently learning to do dissertation project; could you kindly tell me the strengths and weaknesses of the two AI tools (i.e. MS copilot and perplexity) for supporting the research tasks of (1) literature search and review, and (2) desk research on topics with public domain sources?"

The reply: "When considering the use of Microsoft Copilot and Perplexity AI for literature searches, reviews, and desk research on public domain sources, each tool has its own strengths and weaknesses. Here’s a breakdown of how they perform in these specific research tasks:

1. Literature Search and Review

Microsoft Copilot:

• Strengths:

  • Natural Language Processing: Copilot allows users to ask questions in natural language, making it user-friendly for those unfamiliar with traditional search queries. It provides concise answers and can summarize complex topics effectively [2].
  • Integration with Microsoft Products: Being integrated into Microsoft’s ecosystem (like Edge and Office), it offers seamless access for users already using these tools, enhancing productivity [2].
  • Contextual Understanding: Copilot can handle follow-up questions well, maintaining context from previous interactions, which is beneficial for iterative literature reviews [2].

• Weaknesses:

  • Limited Source Transparency: While Copilot provides answers, it may not always clearly cite sources, which can be a drawback for academic rigor in literature reviews [2].
  • Potential for Errors: Like other AI tools, it can generate incorrect or biased information, necessitating careful verification of facts [2].

Perplexity AI:

• Strengths:

  • Source Citation: Perplexity AI emphasizes transparency by citing sources for the information it provides, which is crucial for academic literature reviews [1].
  • Focused Search Options: Users can specify the type of sources (web, academic, or social media), allowing for tailored searches that can yield more relevant results for literature reviews [1].
  • Deep Research Feature: This feature generates in-depth reports, which can be particularly useful for comprehensive literature reviews [1].

• Weaknesses:

  • Limited Contextual Awareness: Perplexity AI has a more limited context window compared to Copilot, which may affect its ability to provide nuanced answers in ongoing discussions [2].
  • User Experience: While it has a clean interface, some users may find it less intuitive than Copilot, especially if they are accustomed to conversational interfaces [2].

2. Desk Research on Public Domain Sources

Microsoft Copilot:

• Strengths:

  • Real-Time Information Retrieval: Copilot can search the web in real-time, which is beneficial for obtaining the most current information on public domain topics [2].
  • Conversational Interface: The chat-like experience allows users to refine their queries and explore topics in depth, making it suitable for desk research that requires iterative questioning [2].

• Weaknesses:

  • Subscription Cost: As a subscription-based service, it may not be accessible to all students or researchers, limiting its use for those on a budget [2].
  • Dependence on Microsoft Ecosystem: Its integration primarily with Microsoft products may restrict usability for those who prefer other platforms [2].

Perplexity AI:

• Strengths:

  • Open Source and Free Access: Being a free tool, Perplexity AI is accessible to all users, making it a great option for students and researchers [2].
  • Diverse Source Access: It can pull information from various types of sources, including academic databases, which is beneficial for thorough desk research [1].

• Weaknesses:

  • Depth of Analysis: While it provides good summaries, the depth of analysis may not match that of more specialized tools, potentially limiting its effectiveness for complex desk research tasks [2].
  • Variable Quality of Sources: The quality of information retrieved can vary, and users may need to sift through results to find credible public domain sources [1].

Conclusion

Both Microsoft Copilot and Perplexity AI offer valuable features for literature searches and desk research, but they cater to different needs and preferences. Copilot excels in user experience and integration within Microsoft products, while Perplexity AI stands out for its transparency and free access. Depending on your specific research requirements and resources, one tool may be more suitable than the other."

Reference: 

A collection of blog notes on using chatgpt for research purpose.