Beyond Books and Screens

Introduce to Use AI as a Tool in Penn Library

In the dynamic realm of our ever-evolving digital landscape, the utilization of Artificial Intelligence (AI) has become imperative for libraries seeking not just to survive but to flourish. This blog serves as a gateway to introduce several AI tools, empowering librarians to seamlessly integrate AI into various facets of library operations, programming, and services. Delving into the transformative potential of AI, we will explore how these technologies can elevate libraries into more dynamic and user-centric spaces. Armed with the necessary insights, this exploration aims to equip Penn Library with the tools required to harness AI as a transformative force, effectively meeting the evolving needs of library users and advancing the institution's mission of knowledge dissemination and innovation.

Machine learning

Macmillan Learning & Packback

Macmillan Learning has announced a collaboration with the education technology company Packback. They are combining artificial intelligence technology with hundreds of course materials on a digital learning platform, with the aim of assisting educators in fostering students' critical thinking abilities and enhancing their learning experiences.

 
Achieve

Achieve is Macmillan's digital learning platform. The company has conducted replicable research with IRB approval, demonstrating its effectiveness in helping students learn and improve their exam scores. Achieve's user interface is designed to incorporate features that support active learning, metacognition, and evidence-based teaching practices. Some of these features include:

  • Goal Setting and Reflective Surveys: Supporting students' metacognition through a series of actionable surveys, helping them set goals for themselves and reflect on their learning performance throughout the semester.
  • Adaptive Quizzes: Providing each student with a personalized learning experience that becomes more challenging as they progress.
  • Rich Interactive Content: This includes charts, videos, simulations, and animations, creating engaging learning experiences for students.
Packback

Packback Questions is an inquiry-driven learning discussion platform that encourages students to explore questions both inside and outside the classroom, discovering the real-world applications of course materials and elevating their critical thinking abilities. Students can also use this platform to express their thoughts, engage with peers, and practice writing, fostering their self-directed learning skills. What sets Packback apart is its use of artificial intelligence, known as the "digital assistant," which aids in managing discussion communities, assessing the quality of student contributions, providing timely feedback, and offering proactive guidance to ensure that discussion topics remain focused, relevant, and productive. Furthermore, the platform's proprietary AI model combination can also:

  • Provide feedback on students' grammar.
  • Assign "curiosity scores" to students.
  • Prompt students about potential violations of academic guidelines in their posts before submission.

 

Khan Academy (Khanmigo)

Back in March of this year, Khan Academy created Khanmigo, a brand-new tool based on GPT-4. This innovative tool takes on the role of a virtual tutor, engaging with students through a series of prompts and even conversations (without allowing them to directly cheat for answers). Its purpose is to guide students in mastering knowledge and methods during their learning journey, thereby enhancing their critical thinking and problem-solving skills. Discussing artificial intelligence, it is crucial not to focus on the technology initially but instead to consider the problems it can address. The emergence of Khanmigo does not imply replacing the role of teachers in the classroom but rather "can do certain things to alleviate unnecessary burdens for teachers, allowing them to engage in more one-on-one conversations with students, assign more projects, provide support, and do all those things that we know humans can do and technology cannot." For example, this AI tool functions as a personal tutor, guiding students in completing assignments in various subjects rather than directly providing answers. Teachers use it for tasks such as lesson planning, task assignments, and providing feedback to students, while students use it to assist their learning in mathematics, science, and humanities.

Chegg (ChaggMate)

Chegg announced the launch of its in-house AI learning tool, CheggMate. This tool, based on GPT-4, combines advanced artificial intelligence systems with a robust repository of verified, specialized knowledge content on its platform. It promises to provide students with a more efficient, personalized, accurate, and helpful learning experience. 

Serving as a virtual tutor/subject expert, CheggMate engages in a series of dialogues with students. It tailors its instructional guidance, tests, exercises, and feedback based on the information provided by students, such as their chosen subjects, learning objectives, questions they wish to address, and their current learning pace. Without directly providing answers, it guides students to pose new questions, offers answers in various forms, and breaks down steps to help students gain a deeper understanding of concepts and master knowledge, ultimately enhancing learning efficiency and outcomes. 

Chegg has collaborated with AI data infrastructure provider Scale AI to develop proprietary large language models (LLMs). These LLMs will power Chegg's personalized learning assistant, offering a generative experience. This experience is set to be rolled out over the next two semesters and will combine the best techniques in generative artificial intelligence with Chegg's high-quality content, showcasing the ability to improve student performance. Students can expect to see a more user-friendly conversational interface, personalized learning, deeper content, and the capability to automatically transform their content into innovative learning tools, such as practice tests, study guides, and flashcards. This collaboration allows Chegg to develop unique LLMs that will enhance the quality of data annotation and classification, enabling them to offer new product experiences to customers starting from the fall while retaining their own intellectual property rights.

Author

Date

December 14, 2023

Categories

Share