How to Use Perplexity AI: The UNDERRATED Search AI

Key Concepts:

• Perplexity AI: An AI-powered search engine that prioritizes credible answers and source citations.
• AI Overview: A feature in Google Search that provides a summarized answer at the top of search results.
• Threads: Individual search queries within Perplexity.
• Sources: The origins of information used by Perplexity, including web pages, academic papers, and social media posts.
• Academic Search: A Perplexity search mode that filters results to only include academic papers.
• Social Search: A Perplexity search mode that filters results to only include social media posts (primarily Reddit).
• Spaces: Customized environments within Perplexity where all threads are governed by the same custom instructions and can include uploaded files.
• Pro Search: A feature in Perplexity Pro that increases the number of sources used for a search.
• AI Model Selection: A feature in Perplexity Pro that allows users to choose the underlying AI model used for searches.

1. Introduction to Perplexity AI

• Perplexity AI is presented as an alternative to Google Search, designed to provide credible answers quickly.
• The video aims to teach viewers how to use Perplexity effectively, including tuning searches and understanding its strengths compared to tools like ChatGPT.
• The Perplexity homepage features a prompt box for asking questions, similar to other AI tools.

2. Core Functionality: Asking Questions and Getting Answers

• Perplexity provides answers with citations to its sources, enhancing credibility. Citations appear as small numbers in gray rounded rectangles within the answer.
• The interface is divided into three panels: the core answer (bottom left), images and videos (right), and a list of sources (top left).
• Clicking on citations leads directly to the original source.

3. Perplexity vs. Google Search

• While Google Search also offers AI overviews with sources, not all search results include this feature.
• Perplexity consistently provides summarized answers with citations, even for recent or niche topics where Google might only show a list of articles.
• Example: Searching "how does AI predict whale movements to reduce ship damage?" yields different results on Google (articles) and Perplexity (summarized answer with citations).

4. Tuning Searches: Academic and Social Sources

• Perplexity allows users to filter search results by source type: web, academic, or social.
• Academic Search: Filters results to only include published scientific papers.
  • Users can review and remove specific sources from the search results.
• Social Search: Filters results to only include social media posts, primarily from Reddit.
  • Useful for gathering the "wisdom of the crowd" on topics where other search methods have failed.
  • Example: Asking "How can I get started making a budget?" and receiving a step-by-step guide based on Reddit threads.

5. Sharing and Saving Threads

• Search results can be shared via a generated link.
• Saving threads and making them private requires signing up for an account.

6. Account Settings and Customization

• Signing up for an account allows users to save their search history and customize their experience.
• Settings include:
  • Appearance (dark mode)
  • AI Data Retention (option to prevent Perplexity from training its models with user data)
  • Profile (custom instructions for the AI)
    • Example: Instructing Perplexity to provide concise, bullet-pointed answers.

7. Spaces: Customized Environments

• Spaces are like dedicated "rooms" within Perplexity where all threads are governed by the same custom instructions.
• Users can create spaces for specific topics or projects.
• Example: Creating a "Business Mentor" space with instructions to provide direct, no-sugar-coating advice.
• Spaces can be enhanced with uploaded files (Pro plan required) and links to provide additional context for the AI.
  • File limits: Up to five files, each less than 25MB.

8. Perplexity Pro Features

• Pro Search: Increases the number of sources used for a search, improving precision and relevance.
  • Free users get three Pro searches per day; Pro users get at least 300.
• AI Model Selection: Allows users to choose the underlying AI model used for searches.
  • Different models may be better suited for different tasks.

9. Perplexity vs. ChatGPT: A Comparative Analysis

• A side-by-side comparison is conducted between Perplexity and ChatGPT (with search enabled) for finding dinner recommendations in Rome.
• While both provide recommendations, Perplexity excels in providing citations.
• For research queries, Perplexity is shown to be more reliable and accurate, especially when citations are crucial.
• Example: A research query about specific tests is more accurately answered by Perplexity with proper sourcing, while ChatGPT initially provides an incorrect answer.
• Perplexity is positioned as a tool for reliability, while ChatGPT is better suited for creativity.

10. Conclusion

• Perplexity AI is presented as a valuable tool for users seeking credible, well-sourced information.
• Its ability to filter search results by source type (web, academic, social) and its focus on citations make it a strong alternative to traditional search engines and other AI tools.
• The video concludes by emphasizing Perplexity's strength in providing reliable answers, contrasting it with ChatGPT's focus on creative tasks.