AI Tools for Researchers: (Hopefully Helpful) Suggestions
ChatGPT and its growing family of generative artificial intelligence (AI) chatbot counterparts are just the start! Read on to learn more (with lots of links) .....
Apart from chatbots, there is an avalanche of other AI tools, partially AI-informed tools, and tools that can integrate with AI functions. Many of these tools have been evolving since long before AI chatbots were a thing. We’ve talked about some of the issues (and how you can embrace AI for organizational success) in an earlier blog.
For many researchers, grasping the sheer mass of AI tools, what they do, and assessing their trustworthiness and validity is a collective burden.
This post surveys the AI tool landscape for researchers, highlighting tools for tasks like literature review, data analysis, and manuscript preparation.
Generally understanding generative AI and large language models (LLMs) is a good place to start. Then, you can situate ChatGPT and the like in the bigger AI picture.
From there, the AI tools make more sense and can become part of your workday toolbelt!
Generative AI tools and LLMs in research
Generative AI tools, including LLMs, use artificial intelligence to create text, images, and other content based on user input. They act like virtual helpers, processing vast amounts of information, with abilities limited only by the user’s ability to instruct them.
A 2024 McKinsey Global survey found nearly 65% of organizations regularly use AI— way up from just 10 months earlier. Research by Slack found about 80% of workers reported greater productivity in tasks like writing assistance, workflow automation, and content summarization. For researchers, they can scour data and reply in human language, for basically every aspect of every research stage.
Yet skepticism remains regarding many AI tools, and rightfully so. The “crisis” of ChatGPT- and LLM-generated text, comprising equal parts unethical and absurd, is sure to segue into AI output that’s harder to detect because it’s much higher quality. Until then, these tools will be in progress. However, many other AI tools are only partially AI- and LLM-informed. They focus on specific tasks and their domains (e.g., literature searches and identifying suitable journals for publication) are less subject to gray areas and misinterpretation.
ChatGPT et al.
ChatGPT is categorized as an AI chatbot. You can interact with it in conversational text and in many languages, even across languages (e.g., put in Chinese and Arabic text and ask it to create an English summary). These functionalities allow researchers and librarians to automate various tasks such as finding and summarizing information, editing text, translating, crunching data, and now, scanning and reporting on documents and websites. Its ability to generate seemingly coherent text in a simple user interface rocketed it to widespread use in early 2023.
Emerging competitors include all the above functions to some degree and are continually being updated. The race is still just past the starting line.
Gemini (Google) is notably compatible with Google Scholar.
Copilot (Microsoft) provides AI help in Microsoft Office apps with real-time suggestions.
Claude (Anthropic) is quite similar to ChatGPT, but trained on different data and showing different proficiencies.
Apple Intelligence may reshape the playing field again, as it incorporates personal data with the capabilities of the above chatbots and platforms.
Limitations such as made-up information (including academic references), refusal to work on certain themes, and inability to handle certain file types exist in spades. However, the developers are almost certain to iron these out and introduce new features, so one can’t really dwell on them for too long.
Now, on to the more specified tools. Note that the list is subject to change and many more options are currently available or will no doubt become available. The inclusion/exclusion, order, and descriptions are as of this writing and do not imply endorsement. Use all of these at your own risk and responsibility.
AI tools for the early stages of research
The early stages include identifying gaps, formulating ideas and approaches, preliminary literature searches, summarizing key findings, and ensuring the accuracy and reliability of the data sources. The following AI tools can help. A few notable ones are detailed for each section and a host of others.
Assistant by scite
This AI-powered tool aims to make literature retrieval more reliable and efficient. Unique among LLMs, it validates its outputs against an internal database to give accurate and trustworthy references. Assistant leverages full-text articles and deep learning models to demonstrate how scientific publications cite each other, providing both quantitative and qualitative insights.
Connected Papers
This tool streamlines the literature review process by visually mapping connections between papers based on citation patterns. Search by keyword, title, DOI, PubMed URL, or any other identifier and click to generate a cobweb-like visual representation. Papers frequently cited together are likely to be related and they’ll appear in the Document Tree. Each publication is shown as a circle, with proximity and line darkness indicating similarity, larger circles reflecting higher citation frequency, and darker colors indicating more recent publications.
IRIS.AI
IRIS.AI is a platform that assists researchers by finding relevant papers, summarizing findings, identifying connections between different studies, and extracting data. This product of an EU-funded project can solve the problem of digging through thousands of publications (including lesser-known ones) by using a neural network algorithm to understand context and document similarity. Apart from literature searches, its tools can filter, extract, and systemize data, analyze sets of documents and help with inclusion/exclusion, and let you “interact” with your data via an LLM-based chat tool.
ChatPDF – For extracting specific information from PDF-based articles and understanding complex information through research-related queries.
Consensus – Aggregates and synthesizes scientific findings, providing summaries of the consensus on specific research questions.
Elicit – A tool that assists researchers by finding relevant papers by research topic, summarizing the findings, and answering specific research questions.
Lateral – Suggests relevant papers and visualizes connections between different studies.
LitMaps – Maps the connections between papers, uncovers new papers of interest, visualizes academic landscapes, aids collaboration with colleagues, and gives updates on new papers.
Perplexity – Takes complex scientific documents and produces accessible summaries. Reduces reading time, makes studies easier to comprehend, and gives insights to researchers.
Research Rabbit – Explores a topic of interest by suggesting related papers, tracking associated research trends, and organizing potentially interesting findings.
Scholarcy – Reference management tool that aids in finding, managing, and sharing academic papers. Provides recommendations for other papers of interest based on user input and research trends.
SciSpace – Formerly Typeset, this is a platform that enhances academic writing and collaboration with tools for managing references, generating citations, and improving the overall research workflow.
SciSummary – Summarizes scientific papers, providing concise overviews of research findings and key points of complex studies.
Semantic Scholar – Finds and helps users understand relevant articles, extracts key information, and provides recommendations for other papers based on the research topic.
AI tools for the middle stages of research
The middle stages of research focus on organizing and analyzing the data gathered. This includes project management, extracting and interpreting data, and drafting manuscripts. Especially for the latter task, many tools have existed and been evolving since before ChatGPT.
Mapify
Mapify (formerly Chatmind) is an AI-powered tool for transforming various types of content into mind maps, so complex information becomes clearer and more manageable. It can easily convert text, images, audio, and even long documents into structured mind maps. Mapify also has built-in AI templates, a chat interface for interactive mind map editing, and features like real-time web access and text-to-image generation.
Jenni
This AI-powered writing assistant is for enhancing academic and content writing. Its features include AI autocomplete, in-text citations, paraphrasing, and custom writing tones for drafting, refining, and citing. Jenni.ai supports multiple citation formats and styles and includes tools like an outline builder and a plagiarism checker. It can also read and summarize research PDFs.
AudioPen – A transcription tool that converts spoken research notes into written text.
Cohesive – For creating, refining, editing, and publishing documents. While not specifically for research writing, it can certainly be applied to it.
Dante – Assists with data analysis and visualization, providing recommendations based on complex datasets.
Fable Fiesta – Think outside the box with this one – a tool for storytelling and narrative creation that can be used for by crafting compelling narratives around research findings.
Grammarly – A widely used writing assistant that checks for grammar, punctuation, clarity, and readability. Has many plug-ins and implementations to work it into your browsing and other software.
Quillbot – A writing assistant for paraphrasing, summarizing, and improving writing. Quillbot provides suggestions for improving clarity, coherence, and style.
Tableau – A powerful business-oriented data visualization tool for analyzing and presenting data in interactive and visually appealing charts, graphs, and dashboards.
AI tools for the latter stages of research
The final stages of research work are to ensure quality and integrity, find and submit to a suitable publication (though some experts contend this should happen at an earlier stage), and steps such as peer review and revision. The following tools provide support.
Penelope.ai
This tool helps academic authors make sure their manuscripts meet journal requirements. It integrates with manuscript-handling software to perform more than 30 configurable checks, covering areas such as ethics, declarations, title page, abstract, structure, figures and tables, and referencing, and gives feedback and highlights sections that need attention. The tool can help automate the review process and speed up the publication process, and it improves the experience for authors, making it easier to prepare manuscripts for successful submission.
Journal finders/selectors
As we’ve discussed in earlier posts, selecting a good, effective and career-enhancing journal is key. Many academic publishers, companies, and academic service providers (especially editing companies) have created journal selection tools. Most are algorithm-driven, with the data and inside works the only difference, including the use of AI. The UI on most of these is quite similar. Drop in your abstract or keywords and get a list of suggestions. Find one that works for you. Just a few are:
Jane (journal author name selector), Journal Selection dashboard (Dimensions with Altmetric data), and MY journal selector (Edanz).
Copyscape – Plagiarism detection tool that scans texts for duplicate content and provides detailed reports on potential matches.
Explore the tools, put them to work, stay updated
No prescription can be offered on the right/wrong way to use AI tools, which to trust, and which to recommend. These are choices you’ll find on your own. Use ChatGPT or another chatbot as a gateway (if you’re not already) to generative AI. Try to compare the lit search tools and writing assistance tools. Decide for yourself and revisit what’s on offer because it’ll change quickly.
“AI tools have made it easier to conduct research. Researchers can search and find the research topics and relevant literature on the topic very quickly and efficiently”.
Here are some sites for keeping up to date on the latest goings on in AI tools and techniques:
AI news website: https://www.artificialintelligence-news.com/
Developer Tech: https://www.developer-tech.com/categories/developer-ai/
MIT News: https://news.mit.edu/topic/artificial-intelligence2
OpenAI News: https://openai.com/news/
Great compilation. I will use it for future reference.
A very helpful article
thanks a lot