Read the first part of the series: An AI for Academia
Science is in a precarious state. A recent Vox article reveals that the once highly productive field is now experiencing a decline in truly groundbreaking research. Although funding has increased by over ten times since 1955, many feel that the golden age of science has come to an end.
In the early 20th century, discoveries such as DNA and nitrogen fixation had a major impact on society, but these types of discoveries have been notably absent in recent years. The problem is highlighted in a new Nature paper called Papers and patents are becoming less disruptive over time, which found that although there are more scientific papers and patents than ever before, fewer are considered to be revolutionary or transformative.
The authors of the paper used a new metric called the “CD index” to determine if papers were mostly consolidating knowledge or pointing towards fresh avenues of research. According to the Nature paper, “disrupting” papers, or those that point in new research directions, are on the decline. For instance, in the social sciences, the average CD5 decreased 92% from 0.52 in 1945 to 0.04 in 2010. In the physical sciences, the average CD5 decreased 100% from 0.36 in 1945 to 0 in 2010.
The question that arises is: “Why is science slowing down?” There are many possible and commonly cited answers, including issues with the current system, a lack of incentives to innovate, or simply that low-hanging fruit has already been picked. One, however, that has not been explored is to what extent expanding our idea of what the web is—from a flat structure infested with extractive AI to a multi-layered structure with symbiotic AI—can enable the unprecedented levels of collaboration, knowledge building, and capacity for collective intelligence and wisdom that will enable science and society to address our planetary challenges.
AI is certainly a field where papers, patents, and tools are becoming increasingly disruptive with great potential to point towards fresh avenues of research and change society. This article explores how AI and the emergent Academic Overweb could help revive science, leading to new breakthroughs and expanding the adjacent possibles we may want to live in. As described in Part 1 of this series, the Academic Overweb is an emergent AI-enabled overlay on the Web that enables scientists, researchers, and academics to collaborate on accelerating the pace of innovation and breakthroughs.
Academic research has come a long way since the days of dusty library stacks and handwritten notes. Today, the internet and digital technologies have revolutionized the way researchers collect, analyze, and share information. And now, with the emergence of AI and an Academic overlay on the Web, the research process is set to undergo yet another transformation. In this article, we’ll explore the different stages of the research process and examine how an AI-enabled Academic Overweb is poised to change each of them.
Finding Relevant Information and Potential Collaborators
The first step in any research project is to find relevant information and collaborators. AI-powered search tools will certainly help with the former. With the AI-enabled Academic Overweb, we can find additional information sources as well as collaborators. Using on-page presence, researchers can easily find others who are focusing or have recently focused on webpage content directly related to their study. By simply going visible on relevant webpages of related studies and articles, they can review people on the visible list. They can also search the page-level chat for messages related to their study, reply to related messages, and/or post their own message. AI can help researchers decide what webpages to go visible on, find relevant content, and figure out whom to engage.
Bridging the Information Ecosystem
Once researchers have found the relevant information, they need to organize it into a coherent and useful format. This is where bridges come into play. Bridges are an essential component of the AI-enabled Academic Overweb, allowing researchers to connect all obviously and non-obviously related online content. When an academic information ecosystem is mature with all relevant studies bridged, adding AI to bridges could automate literature reviews. The AI can also help identify knowledge gaps based on the existing constellation of relevant bridges. AI-powered tools can suggest topics for research papers or academic essays based on a writer's interests, their previous works, and the relevant information ecosystem.
AI-Assisted Writing
Writing is one of the most critical stages of the research process, and the AI-enabled Academic Overweb can help researchers in many ways. For example, AI can potentially automate certain parts of the research writing process that involve routine or standardized information, or AI can assist with generating or suggesting text for certain sections of a research paper based on established practices or data. AI can assist with ideation in the writing process and identifying logical fallacies in arguments. Additionally, AI can help with language and grammar checking, ensuring that their writing is free of errors. Citation and reference management is also made easy with AI-powered tools that help writers to manage their references and citations, ensuring that they are formatted correctly and helping to save time.
AI can also analyze study drafts and a writer's style and provide feedback on areas for improvement, such as overall structure, argumentation flaws, sentence structure, or word choice.
Facilitating the Peer Review Process
Peer review is an essential part of the research process, and AI can play a crucial role in streamlining and optimizing it. AI can be used to facilitate the peer review process by analyzing submitted manuscripts and identifying potential reviewers based on their expertise and previous work. This can help to speed up the peer review process and ensure that reviewers are well-suited to assess the quality and relevance of the submitted work.
AI can also be used to detect plagiarism and other forms of academic misconduct, making the review process more transparent and reliable. Autonomous AI reviewers can be deployed to review specific aspects of articles. Bridges, notes, and conversations within specific meta-communities can be used by the peer review team to document, discuss, and debate their findings.
Automating the Formatting and Trend Analysis
Once a manuscript has been accepted for publication, the AI-enabled Academic Overweb can help to automate the formatting of submitted manuscripts, ensuring that they are consistent with the journal's guidelines and reducing the workload for journal editors.
Trend analysis is another area where AI can play a significant role in the academic research process. By analyzing trends in submitted manuscripts, AI can identify emerging topics and areas of interest in a particular field. This can help journals to stay up-to-date with the latest research and ensure that they are publishing work that is relevant and impactful.
Furthermore, AI can also be used for data analysis, allowing researchers to quickly and accurately analyze data from published studies. This can help to identify patterns and relationships that may not be immediately apparent, advancing the understanding of a particular field and identifying areas for further research.
Post-Publishing
Once the research is complete and the paper is published, the AI-enabled Academic Overweb can continue to provide benefits. Bridges can be used to connect the study to the information ecology, ensuring that it is easily discoverable by other researchers. Relevant meta-communities can have conversations about the study with smart tags, allowing for ongoing discussion and debate. As relevant data surfaces, more bridges can be built, creating a more interconnected and robust academic ecosystem.
One day, a meta-community within the Academic Overweb may operate as a DAO consisting of autonomous AI agents and subject matter experts who initiate research, collaborate with researchers, facilitate peer-review processes, and publish journals.
In conclusion, it is essential to revitalize science if we wish to continue benefiting from transformative and disruptive discoveries. The decline in truly groundbreaking research could have a devastating impact on society if it persists, given the challenges we face on our planet. The potential for AI and the Academic Overweb to revolutionize the research process is vast, but it requires us to expand our understanding of the Web and embrace AI. By leveraging AI-powered tools and Overweb features, researchers can streamline their workflows, automate time-consuming tasks, collaborate with subject matter experts and AI agents, stay up-to-date with the latest research in their field, and even conduct truly disruptive research. The emergence of the Academic Overweb could lead to a new era of scientific research, one that is collaborative, innovative, and impactful, enabling us to solve some of the most pressing problems of our time.