Generative AI Tools and Resources : AI tools for Academics

Unlike many of the popular generative tools that are still developing, many of the tools below have been developed over periods of several years. Four of these tools are developed by non-profit research/programming organizations with the goals of accelerating research methods for things like systematic reviews, visualization, and even generative summaries. Each of these four tools below is free to use and sets a precedent for being some of the most transparent developments in research-focused AI's.

• AS Review Lab: A free-to-use, open-source systematic review assistant that leverages state-of-the-art active learning techniques in order to help review large amounts of articles/content for specific key words and qualifiers. Originally developed by a team of researchers from Utrecht University, the software is now part of a non-profit initiative driven by multidisciplinary researchers and system engineers from across the Netherlands. ASReview is still coordinated at Utrecht University. You can install the software locally on your device and keep full control over your data. You need a reference management system, such as Endnote, as well,  Tutorial: How to use AS Review Lab.
   
• Elicit: This online platform generates brand new summaries and findings that are created using an AI reviewer and the contents of each respective paper. Developed by the Ought non-profit learning research lab, the system helps researchers find papers to cite and defined research directions. It currently provides both summary services in addition to an in-depth literature review toolset, all sourced through the over 200 million academic papers available through Semantic Scholar.  Tutorial: dig into details or Tutorial: How to use Elicit for empirical research
   
• Semantic Scholar: Pulling from publisher partnerships, data providers, and web crawls, this AI-driven search and discovery toolset provides a streamlined search platform for over 200 million academic papers. Developed within the Allen Institute for AI, this open project has been worked on since 2015 with the goal of supporting high-quality research. Free to use. Partnerships with PubMed, Springer Nature, Taylor & Francis, SAGE, Wiley, ACM, IEEE, arXiv, and Unpaywall. Connects to Endnote.
   
• Research Rabbit: Using a complex AI-backed methodology, this platform is a literature mapping tool. Using a “seed” article, the platform finds relevant articles and content to link. Visualized by a bubble map of connected topics, this toolset can be invaluable for finding additional connections across a variety of disciplines as well as staying up to date with up and coming publications. This initiative is supported by a development collective, which has pledged to keep the core discovery functionalities open for all to use. Sign up required to access at researchrabbit.ai.  (University of South Florida Libraries, 2023).
   
• Litmaps: Can be used for literature mapping with the creation of a visual network of academic papers. Just like Research Rabbit it helps users to start with a seed article and explore related research. Next to that it provides recommendations based on citations, references, and interconnectedness. Uses advanced algorithms to find relevant papers beyond a direct citation. 
   
• Paper Digest: Paper Digest is an AI-powered scholarly assistant designed to help navigate and analyze academic research more efficiently. It helps with discovery and summarization of academic papers. It only uses verifiable resources and promises "100% no hallucination". You can create alerts for (conference) papers, track topics, helps with academic reading & writing. 

The following are examples of other AI tools for literature searching and mapping currated by Cambridge Libraries (University of Cambridge, 2023):
Audemic.io, Andi, ChatPDF, Claude 2, 
Connected Papers, Consensus search engine, 
Dimensions, Inciteful, Iris AI, 
Keenious, Perplexity AI, Phind, Scite Assistant, SciSpace and Talk to Books 

NotebookLM is developed by Google. It is an AI-powered research assistant designed to streamline the process of gaining insights from a personal selection of content. It allows users to ground the AI in their own documents, creating a personalized assistant capable of summarizing, explaining, and generating new ideas based on the user’s provided content.

Here are the key features and capabilities of NotebookLM:
NotebookLM allows users to consolidate and analyze information from multiple sources, acting as a virtual research assistant. Its main functions include:

• Summarizing uploaded documents
• Answering questions about the content
• Generating insights and new ideas based on the source material
• Creating study aids like quizzes, FAQs, and outlines

NotebookLM is particularly useful for:

• Students and researchers synthesizing information from multiple sources
• Content creators organizing ideas and generating scripts
• Professionals preparing presentations or reports
• Anyone looking to gain insights from complex or lengthy documents.

Further reading: LibGuide on NotebookLM. 

ChatGPT is a chatbot that provides human-like answers to questions from it's users by engaging in human-like conversations. It can carry out a multitude of smaller and even more complex tasks.

Within research, ChatGPT can support every stage of the research process. It can help with:

• developing and refining a research question
• explaining complex concepts
• overview of existing literature and provide other angles
• translating literature
• summarizing literature and extract key points
• organize notes
• drafting documents
• writing code for data analysis

Within education, ChatGPT can support staff to enhance their teaching, for instance with:

• creating learning assessments
• acting as a personal tutor
• developing learning activities

Further reading: Benefits and risks of ChatGPT for Education.