Dentistry

Perplexity AI provides the ability to search portions of the web as well as a body of academic literature. It's useful when finding current grey literature is a priority (e.g., think tanks, institute reports, white papers). It can also help you track down some peer reviewed articles as a starting point for exploration. In addition to the standard and "pro" search options, there is also a "deep research" feature to try.
 

Below are some tools to explore, keeping the strengths and challenges and recommendations below in mind.  If you would like help matching a tool to your information need at hand, please feel free to contact us: judsmith@umich.edu

For brief video introductions to a few of these tools, please see Wake Forest University's Speed Dating with [AI] Databases and Tools guide.

Generative artificial intelligence (AI) tools like ChatGPT or Gemini offer rapidly evolving ways to interact with information, and may be helpful in the literature search process. While each tool is unique, it is important to consider broad strengths and weaknesses when using AI in the literature search process:

Strengths

• Might help brainstorm research topics or generate alternate search terms for your literature search
• Can offer more approachable summaries of complex research texts
• May help translate passages of text into other languages

Challenges

• AI tools can generate citations to scientific papers that are actually nonexistent, but appear credible
  (often referred to as hallucination)
• Generative AI tools are not search engines, they are trained on massive data sets to generate text that predicts speech
• Baseless responses may sound authoritative
• Content may not be current - it's important to fully understand the underlying the data
• Many tools have not been trained using paywalled articles, omitting a vast body of scientific and biomedical research

Note: Some content in this guide has been adapted from: AI, ChatGPT, and the Library by Amy Scheelke from Salt Lake Community College, licensed CC BY-NC 4.0, except where otherwise noted.

As you use AI tools in the literature search process:

• Consider using AI to explore ideas and research questions, or to generate additional search terms for your literature search

• Avoid using article citations produced by generative AI without careful validation in credible sources (e.g., PubMed, the publisher or journal's website, Google Scholar, or other scientific databases), and read cited papers to verify claims

• Review generated text critically and always verify across multiple credible resources 

• Plan for transparency: many publishers require authors to acknowledge how AI tools were used in the research process

• Consider the privacy, security, and copyright implications of any text pasted into generative AI platforms, and any content produced from these tools

The articles below can be used as a starting point for exploring the intersection of AI and health policy.   To explore the topic further, you could click here for additional PubMed results, or you could citation track in Google Scholar or Scopus to see who has cited an article since it was published.