AI and Open Science: The Joined Forces Leading Us into the Future of Education and Research
Editor’s Note: Today’s post was written by Prof. Dina Darwish, an OSEI Expert and Vice Dean of Faculty of Computer Science and IT at the Ahram Canadian University, Egypt.
With open access now widely embedded in academic research and publishing, it is well understood that open science encourages the free and open exchange of information, facilitating rapid discovery and global participation from researchers, libraries, publishers, and beyond. It is creating a stronger, more integrated research ecosystem through the use of open infrastructure, collaborative research initiatives, open data sharing, and open access publication. An evolution of traditional publishing, open access improves the credibility of scientific findings, makes said research easier to replicate, and fosters public confidence in science by opening up the scientific process to more people. Even more rapidly advancing, artificial intelligence is opening new pathways that can reshape open education and deepen and expand the principles of open science across research and learning environments.
By encouraging openness across the entire research continuum, open science represents a seismic shift in the way we will carry out science, and not just the disciplinary sciences.
A Major Step in the Right Direction
Open science is fundamentally changing the traditional method of scientific publishing. New organizations as well as non-profit educational and research sectors are openly traversing and adopting open access publishing models. To name a few, the Public Library of Science (PLOS) provides a repository of open access periodicals and peer-reviewed scientific literature; the Aggregator of Global Open Science and Research (AGOSR) platforms millions of open access content to the scholarly community; arXiv provides electronic preprints across many disciplines (e.g., Computer Science, Quantitative Biology); F1000Research provides a platform for rapid publishing of life sciences papers and associated research outputs, while bioRxiv operates as an open access online archive and distribution service for preprints in the biological sciences.
Open access to research data, or open data, is arguably a major step forward in fostering the openness and transparency of science. Because research utilizes data more than ever, developments in scientific understanding are inherently tied to access to data. Open data allows scholars to incorporate existing resources in new and complementary ways.
Additionally, other components of research, such as methodology and code, have similar importance and follow a similar approach. Due to the public funding of these works, they need to be available under a Creative Commons (CC) license to protects users from copyright infringement.
“Public Funding, Public Knowledge”
Maintaining the theme of “Public Funding, Public Knowledge,” a category of computer resources (open research computation) is also important to the operating system, as researchers need software to process data and information. These spaces include resources like the Open Science Framework, a cloud-based management system that supports and encourages open collaboration in science while housing project information and archiving data; Ibercivis, a computing platform that is applied to run scientific simulations using CPU time that is not already being utilized; and Experiment, a web-based platform that crowdsources funding for scientific research projects.
Open-source software is associated with open-source licensing for various software packages to eliminate barriers, in keeping with open collaboration philosophy. While open data, infrastructure, and computational resources form the backbone of open science, emerging technologies are now further expanding the possibilities of where open science can take us.
Adding AI to the Mix
Understanding the current landscape of open science, let’s take a step further and look at what is also rapidly advancing and becoming the face of future research: artificial intelligence (AI). AI is moving quickly, and it’s changing many areas, including education (particularly in high schools and colleges). It gives students a mix of chances and challenges, which affects how well they do in school. AI could largely exceed the boundaries of education by being able to meet the needs of every student, such as intelligent tutoring systems, which can personalize learning by tailoring help, support, and feedback to adapt to the learning styles of each student. Other applications that could greatly improve teaching and learning are instructional robots, learning analytics dashboards, adaptive learning platforms, and human-computer interfaces.
For example, research has shown that intelligent tutoring systems can help students do better by giving them feedback. Adaptive learning systems also use AI to change course materials to fit each student’s needs. AI tools are being utilized in education, giving rise to numerous advantages, including:
- Personalized Learning: AI analyzes data on students to personalize lessons based on their pace and styles, ensuring improved learning and retention.
- Operational Automation: AI can automate educators’ tasks to be more efficient, including grading, scheduling, and preparing reports.
- Creative Development: AI can support educators in developing themes, designing lessons, creating interactive materials, and providing visual aids.
- Student Support: The resources provide support to students on demand, which may resolve questions, clarify complex subjects, and provide study assistance.
Adapting Despite the Risks
Educators are beginning to use a variety of AI-enabled services to facilitate their daily routines, such as home voice assistants, grammar checkers, sentence completion applications, and essay writing aids. As AI systems become more widely available, more teachers are putting effort into evaluating and implementing these systems. Educators understand the opportunity to use AI-enabled capabilities like speech recognition and translation to better support students with disabilities, multilingual students, and others who could benefit from more options and personalized digital learning materials. Recognizing the benefits, educators are testing if AI will help with writing, revising lessons, or developing course materials—at the same time being aware of the rising risks to data privacy and security.
As AI use in scientific research continues, it is likely that advanced techniques like fine-tuning, convolutional neural networks (CNNs), semi-supervised learning, and reinforcement learning can be used. The future of AI in scientific research is a generative, multimodal, multifunctional tool for both lecture instruction and research. Scientific educators will benefit greatly from massive multimodal models (LMMs) such as OpenAI’s GPT-4o and Sora, as well as Google’s Gemini 1.5 Pro, which can analyze and generate text, images, audio, and video. These capabilities allow instructors to integrate various approaches to assessment and learning within their standard lecture formats.
The Convergence of AI and Open Science
The articulation of AI with open science principles sets the stage for their intersection to provide a potent relationship based on transparency, the sharing of data, and collaborative research— where AI technologies facilitate scientific discovery, accessibility, and trust in research. While AI is capable of analyzing very large datasets and automating tedious processes, open science complements it by ensuring that AI data, methods, and code are publicly accessible and easy to verify, replicate, and expand on. The convergence of these two makes up a supportive environment that can include funding regulations around open data, journal publisher policies around open access, and open research infrastructure.
There are many opportunities for AI and open science to overlap:
- Open Access and Open-Source AI: Open frameworks allow academics and the public alike to access research and conduct research using AI tools.
- Data Sharing: AI can support collaboration in open science through sharing data and collaborating across disciplines.
- Research Integrity: Open science reduces bias and mistakes in research findings through the appropriate use of AI algorithms and datasets to produce trustworthy scientific evidence.
Given the opportunities, the blending of the open scientific method with AI has the potential to expand the traditional limitations of research in a truly collaborative and modern way.
Embracing the Future of Open Education
While AI offers powerful tools to enhance education and research, its integration must remain grounded in the core values of open science to ensure transparency, replicability, credibility, and equitable access to knowledge. Aligning AI development with open science principles presents both opportunities and challenges, particularly given the nature of many AI systems and the growing competition between private industry and academia. Nevertheless, evolving incentive structures within the research ecosystem may encourage broader adoption of open practices that embrace both AI development and scientific openness. By thoughtfully combining these approaches, the research and education communities can foster a more collaborative, trustworthy, and inclusive future for scientific discovery and learning.
Discover open access articles about AI in education on the AGOSR database:
Ai in Education a Systematic Literature Review
Ai-Driven Transformation of Vocational Education Opportunities, Challenges, and Future Paths
A Systematic Review of Ai Ethics in Education Challenges, Policy Gaps, and Future Directions
