Ithaka S+R has been tracking an impressive list of AI tools aimed at higher education.
The Bellack Librarians have also been examining and evaluating generative AI tools. You can see what we've gathered about them below.
ChatGPT (Generative Pre-training Transformer) is a large language model chatbot from OpenAI that lets you interact using natural language to generate answers to questions and create content. It falls into the general category of AI text generator.
The GPT technology is what quite a number of AI text generators use (Jasper, Anyword, Rytr, etc.)
Some things ChatGPT (and GPT technology, in general) can do
Here are some things ChatGPT can't do (for examples)
ChatGPT 3.5 vs. GPT 4.0
ChatGPT 3.5 is the free version of this popular generative AI tool, available from the Open AI website (https://openai.com/blog/chatgpt). GPT 4.0 is the version available through Open AI's subscription-based tool called ChatGPT+. You can also get to a free, customized version of 4.0 through the Bing search engine.
So what does ChatGPT+ have over ChatGPT 3.5?
Price
The free version of ChatGPT is available to everyone. Upgrading to Plus, Team, or Enterprise offers a more powerful experience through additional features and access to GPT-4.
Like ChatGPT, Google's Gemini is a natural language processor, but instead of being based on a large language model, it is based on the Language Model for Dialogue Applications (LaMBDA). That said, it produces similar results, but here are some key differences (as of June 2023).
Gemini 2: Deep Research
Gemini 2: Deep Research, introduced by Google DeepMind in December 2024, is designed to conduct web-based research on behalf of the users. It is part of the Gemini Advanced subscription, which requires a paid plan. By following a structured, multi-step process, it gathers, organizes and refines information.
Here’s how it works:
A user writes a question.
Deep Research creates a “multi-step research plan” for the user to either revise or approve.
Once the user approves, Deep Research refines its analysis over the course of a few minutes — searching, saving potentially interesting pieces of information, and then starting a new search based on what it’s learned.
The process repeats multiple times, and once it’s finished, Deep Research generates a report of the key findings.
Price
It is part of Gemini Advanced, which should be subscribed for 1 month free and $19.99 monthly.
A Quick Video Tutorial
Gamma
Humata AI
Generate summaries of academic papers. Humata will answer your submitted questions about the paper(s) and highlight the relevant sections of the papers for you to double-check accuracy. Registration is free. The first 60 pages of PDFs are free. After that, there are tiered pricing plans.
Third party learning platform with a higher education arm. Automates some of the teaching and learning experience for both students and faculty. Examples below, as listed on Cognii's website.
Elicit calls itself a "research assistant using language models like GPT-3 to automate parts of researchers’ workflows". Currently, there are two main workflows in Elicit.
Other features include:
Show the sources for each answer
Integration with Reference Management Tools like Zotero
Orient with a quick summary of 4 top papers
Limitations
Created by instructors to "provide formative feedback during the revision process. By focusing on the revision process instead of the final grade, [there is] an incentive for students to embrace the process of self-reflection and analysis, while at the same time, minimizing the impulse of students to go directly to ChatGPT and ask it to write the essay for them." Less expensive than ChatGPT, with an emphasis on accessibility for all students writing papers and looking for instant feedback on the quality of those papers before turning them in.
Research Rabbit is a literature mapping tool, helping you discover connections between papers. As you build your collection of papers, it uses AI to help improve it's recommendations for you. It is also a collaboration tool where you can invite team members into your collections.
Research Rabbit also has a data extraction tool, but there are two potential issues with it.
To start using ResearchRabbit:
Publications can be added either by uploading a RIS or BibTeX file or by using ResearchRabbit’s search, powered by PubMed, if users are searching the medical sciences, or Semantic Scholar, for any other subject area.
While ResearchRabbit uses PubMed’s and Semantic Scholar’s search engines, the company claims its unique database of “100s of millions of academic articles” is second in size only to Google Scholar.
Once publications are in a collection, ResearchRabbit’s algorithm will begin generating recommendations. These recommendations can be explored through two modes:
by Papers that are Similar work, Earlier work, or Later work
by People that provide additional publications that These authors or Suggested authors have published. (Cole & Boutet, 2023)
Price
Originally created for K-12 teachers, Magic School has expanded to include features for higher ed faculty as well. After you create a free account, you can use it's many tools
As always, the results are variable and will undoubtedly need editing.
Here's an example from the Multiple Explanations Generator.
Perplexity
Perplexity is an advanced AI-powered search assistant designed to provide answers by searching the internet in real-time. Its core purpose is to deliver source-cited information across a wide range of topics.
Key Features
Chatbot style queries
It can quickly retrieve and synthesize information from multiple online sources in real-time
It cites sources within the text response
It lets you choose from various LLMs like the Default Model, GPT-4 Omni, Claude 3.5 Sonnet, Claude 3 Opus, etc.
The “Focus” feature lets you specify which sites to search, including Reddit, Wolfram Alpha, academic writing, YouTube, or the entire web.
It lets you ask follow-up questions based on its previous response.
Price
It has a free standard plan and a professional version for $20 per month. The Pro version gets you more thorough answers with additional sources of evidence. You can have access to a limited number of "Pro" searches in the free version.
Litmaps
LitMaps is a visual literature discovery tool that enables users to explore and uncover relevant articles through citation connections up to two degrees away. By leveraging data from OpenAlex, Semantic Scholar, and Crossref, LitMaps allows users to search for papers using keywords, titles, authors, or DOIs, and organize their research using tags and workspaces.
Key Features
Search for articles in a catalog of 270+ million papers.
With advanced version: Sync with Zotero
Visualizing and annotating the articles
The "Litmaps Monitor" feature helps users stay updated on new research within their field by alerting them to recent publications.
Price
It has a free version with limited options and a pro version for $10 per month
ASReview
ASReview is a free open-access program created in 2021 that allows users to screen through multiple articles and rank the selected articles in order of most relevant to least relevant. (Chan et al., 2024)
It employs active learning and multiple machine learning models to assist researchers in making inclusion and exclusion judgments. It does not replace expert decisions but improves the efficiency and precision of the screening process. (Quan et al., 2024)
ASReview LAB
is one of the products of the ASReview research project and is a free (Libre) open-source machine learning tool for screening and systematically labeling a large collection of textual data. It’s sometimes referred to as a tool for title and abstract screening in systematic reviews or meta-analyses, but it can handle any type of textual data that must be screened systematically
Key Features
ASReview LAB implements three different options:
Oracle: Screen textual data in interaction with the active learning model. The reviewer is the ‘oracle’, making the labeling decisions.
Simulation: Evaluate the performance of active learning models on fully labeled data.
Validation: Validate labels provided by another screener or derived from an LLM or AI, and explore benchmark datasets without being an oracle.
Complete Instruction can be found here.
Price
Free and open access
