Month: August 2025

George Siemens on Agentic AI and the Future of Education

On August 22, 2025, Dr. George Siemens, a former colleague and learning theorist at UTA, returned to campus to discuss a topic at the forefront of educational technology: agentic AI. Introduced by Pete Smith and Peggy Semingson, Dr. Siemens’ presentation highlighted the transformative potential of agentic AI and criticized the slow response of higher education institutions to this new frontier. The talk was a deep dive into the practical applications, challenges, and future of AI in learning design. *Note: This post was co-authored using AI to summarize ideas and revised by CRTLE.

Dr. George Siemens

---

The Disappointing Response from Higher Education

Dr. Siemens expressed his “woeful disappointment” with the higher education sector’s response to generative AI. He argued that university leadership, both in the U.S. and globally, has failed to take advantage of AI’s potential and has not meaningfully influenced its development. As a result, educational institutions are simply accepting tools that big tech companies think they want, rather than actively shaping the deployment of AI to serve their specific needs.

---

Key Research on AI in Education

Dr. Siemens shared findings from a 2023 paper on AI in education, which was published just before the public awareness of ChatGPT. The research revealed that discussions around AI at the time focused on personalization, profiling, assessment, and tutoring. While these applications were seen as beneficial for personalized learning and reducing administrative costs, significant challenges were also identified, particularly the lack of ethical considerations.

He highlighted the need for new methodologies to study how learners interact with large language models (LLMs). He mentioned the use of “evals,” where researchers analyze the “trace state of cognitive exchanges” to understand learner confusion and thought processes at a conversational level, something previously only possible with “think-aloud protocols”.

---

Agentic AI in Practice

Dr. Siemens explained that an LLM’s potential is only fully realized through the “support infrastructure that universities create around it”. He provided an example of a multi-agent system used in a medical setting, where different AI agents handle specific tasks like validating medications or providing diagnoses. He emphasized that a similar “virtual team of experts” could be created in an educational context, with agents for questioning, provocation, and deeper inquiry.

---

The Future of Learning Design

According to Dr. Siemens, the future of learning design will be “us designing for AI”. Instead of creating content solely for human consumption, educators will need to design a curriculum that AI can process and deliver to students in novel ways.

“The future of learning design is to create curriculum content for AI (not humans).” -George Siemens

He was particularly impressed by AI’s ability to transmute information and reconfigure it for different audiences. For example, an LLM can explain a complex concept like the Model Context Protocol (MCP) using an analogy of a chef or a farmer, something that was historically cost-prohibitive to produce at scale.

---

Memory Management in AI

Dr. Siemens also touched on the importance of memory management in AI, explaining that while the core LLM has a knowledge cutoff, tools like Retrieval Augmented Generation (RAG) and tool use (like a web search) allow it to access current information. He noted that memory limitations have been a big issue, but now models like Claude and Gemini are actively incorporating memory to recall past conversations. The final personality of an LLM is given during the post-training “fine-tuning” phase, which is much less expensive and time-consuming than the initial training.

---

A Practical Demonstration of Crew AI

The presentation concluded with a practical demonstration of Crew AI, a tool for creating and managing AI agents. Dr. Siemens explained that the process is “super simple” and provides a visual editor for users to create and adjust agents. He suggested that individuals, particularly graduate students, should get into the “agentic space” because most universities “often don’t have the capability to conceive and meaningfully develop these kinds of technologies”.

Full Workshop Recording

Key Ideas from the recording (generated by Microsoft CoPilot):

• Introduction of George Siemens: Peggy introduced George Siemens, highlighting his extensive career in learning theory and his contributions to online learning. George Siemens was a colleague at UTA and is now working at Southern New Hampshire University. 5:02

• Agentic AI and Learning: George Siemens discussed agentic AI and its applications in education. He emphasized the importance of tools like cloud code for experiencing agentic AI and mentioned the challenges and benefits of using such tools in educational settings. 8:13

• Higher Education’s Response to AI: George Siemens expressed disappointment in the higher education sector’s response to generative AI. He criticized the lack of meaningful engagement with AI technologies and the failure to shape the deployment of AI in education. 9:45

• Research on AI in Education: George Siemens shared insights from a 2023 paper on AI in education, highlighting the focus on personalization, profiling, assessment, and tutoring. He emphasized the need for new methodologies and the importance of understanding cognitive exchanges in learning. 12:14

• Challenges and Benefits of AI in Education: George Siemens discussed the challenges and benefits of AI in education, including ethical considerations, curriculum development, and the potential for AI to generate content. He mentioned the importance of understanding the limitations and capabilities of AI. 13:31

• Agentic AI in Practice: George Siemens provided examples of agentic AI in practice, including the use of agents in medical and educational settings. He discussed the importance of coordination and orchestration in multi-agent systems. 41:32

• Future of Learning Design: George Siemens suggested that the future of learning design will involve designing curriculum for AI rather than for humans. He emphasized the need for educators to create content that AI can process and deliver to students. 1:43:31

• Memory Management in AI: George Siemens explained the importance of memory management in AI, including the use of long-term and short-term memory to enhance the learning experience. He discussed the challenges of maintaining context and coherence in multi-agent systems. 1:53:19

• Practical Demonstration of Crew AI: George Siemens provided a practical demonstration of Crew AI, showing how to create and manage agents for various tasks. He highlighted the importance of understanding the tools and processes involved in developing agentic AI. 1:55:53

Join the Conversation

We’d love to hear your thoughts on Dr. Siemens’ presentation and the future of agentic AI in education! How are you currently using AI tools in your teaching or research? What opportunities or challenges do you see with multi-agent systems in education? Have you experimented with tools like Crew AI, Cloud Code, or similar platforms? What support would you need to implement agentic workflows in your courses? Share your experiences, questions, and insights in the comments below, or reach out to us at CRTLE@uta.edu.

How can we be both champions and critics of AI in education? What does it mean to redesign our courses thoughtfully in an era of rapid technological change? In our recent AI Course Redesign Institute, faculty from across UTA explored these essential questions while gaining practical tools and strategies for integrating AI into their teaching practice.

What We Covered

The institute brought together faculty for a full day of learning, discussion, and hands-on work. Pete Smith from Modern Languages opened with a thought-provoking keynote about being both “crusaders” and “critics” of AI—embracing the technology’s potential while maintaining the critical lens that academia demands. The CRTLE team then guided participants through policy development, instructional strategies, and practical applications, with special sessions on Khanmigo, AI chatbots, and effective prompting techniques.

Key Takeaways

• Live in the Complexity. Pete Smith challenged us to sit with the contradiction of being both AI advocates and critics. As he noted, we can champion these tools while still asking hard questions about bias, environmental impact, hidden labor, and the concentration of power in “big seven” tech companies.
• Policy as Foundation. Dr. Peggy Semingson emphasized that clear AI policies aren’t restrictions—they’re roadmaps. The institute provided copy-and-paste policy templates that faculty can adapt immediately for their syllabi, establishing transparency and expectations from day one.
• Beyond the Hype, Into the Classroom. Karen Magruder and Heather Philip showed practical ways to redesign traditional assignments into AI-inclusive learning experiences. Think: students using AI as a thought partner rather than a replacement for thinking, with alternative assessments that showcase genuine learning.
• Tools That Scale. The team introduced Khanmigo’s Canvas integration—an AI teaching assistant that helps with everything from lesson planning to rubric creation. As Melissa Roach and Joseph Rutledge demonstrated in the July webinar, these tools are instructor-facing only and have been vetted for security and accessibility.
• Work Time = Real Progress. The institute wasn’t just talk—participants spent dedicated time redesigning actual course materials and creating personal action plans for AI integration. Some redesigned traditional essays into AI-collaborative projects; others developed comprehensive AI literacy modules.

Part One of the Panelists

The initial panel focused on redesigning an AI course, emphasizing the dual role of supporting AI advancement while critically examining its societal impacts, biases, and ethical concerns.

Part One Big Ideas (summarized by Microsoft CoPilot):

• Course and policy introduction: The session begins with an overview of free AI-related resources and outlines the agenda, including policy crafting and instruction discussions. 1
• Balancing enthusiasm and critique: The speaker highlights the challenge of being both advocates for AI and critical evaluators, especially in academia, encouraging reflection on this dual stance. 2
• Historical context of language models: Large language models have been used in modern languages and translation since the early 2000s, with tools like Google Translate as early examples, evolving significantly since 2017. 2
• Focus on multilingual performance: The course track emphasizes how language models perform across various languages beyond English, such as Spanish, Swahili, Icelandic, and Indonesian. 2
• Critical voices and scholars: Key critics like Gary Marcus, Emily Bender, and Noam Chomsky are introduced, raising questions about model understanding, biases, fairness, and the ethical implications of AI development. 2
• Ethical and labor concerns: The discussion addresses issues such as gender bias, hidden labor exploitation in AI training data, corporate control of AI technologies, and intellectual property disputes. 2
• Environmental impact awareness: The environmental costs of AI, including data center construction and energy use by companies like Meta, are acknowledged as a growing concern in the AI community. 

Part Two of the Panelists

Summary of Key Ideas from Part Two (summary from Microsoft Co-Pilot):

The main topics discussed include the environmental impact of large language models and data centers, the use of AI in education, and the development of AI literacy among students.

1. Environmental Impact of AI:
   • Pete Smith highlighted concerns about the environmental impact of large language models and data centers, including their power and water consumption. Examples include data centers being built by Meta, Tesla, and AIX, and the potential political and ecological issues they cause
   • The importance of considering the ecological impact of AI and data centers is emphasized, with references to experts like Satya Lucione and Henry Spike-sentra
2. AI in Education:
   • Ideas about the integration of AI in educational settings, including the use of AI tools for generating ideas, brainstorming, revising, and editing were shared.
   • Various AI tools like Adobe Firefly, ChatGPT, Midjourney, and Canva are mentioned for creating images, graphics, and videos
   • The importance of accessibility in AI-generated content is highlighted, including the need for alt text and color contrast
3. AI Literacy and Ethical Use:
   • Ideas about stressing the need for developing AI literacy among students, including understanding the ethical use of AI and the potential risks and drawbacks were shared.
   • The importance of transparency and context in using AI is discussed, with examples of how AI can be used appropriately in different contexts
4. Practical Applications and Teaching Strategies:
   • Several practical examples of using AI in the classroom are provided, such as generating case studies, refining assignment directions, creating rubrics, and developing discussion topics
   • Ideas were shared about the use of AI for role plays, critiquing AI responses, and making assignments more engaging and innovative
5. Resources and Further Reading:
   • Check out the OER book titled “AI Powered Education: Innovative Teaching Strategies to Elevate Student Learning,” which provides step-by-step guidelines for implementing AI in education.

Overall, part two provided a comprehensive overview of the current state of AI in education, the environmental impact of AI, and the importance of developing AI literacy and ethical use among students.

Want to Get Started?

Here are the resources shared at the institute to help you dive in:

• PowerPoint Slides: All slides compiled
• Policy Templates: UTA AI Guidelines
• Khanmigo Access: Guides available on the Pedagogy Next blog
• Book Collection: CRTLE’s lending library provides a variety of books on AI in teaching and learning.

Join the Conversation

We’d love to hear from you! How are you balancing enthusiasm and skepticism about AI in your courses? What redesigns have you tried? Share your thoughts, questions, and experiences in the comments below. Let’s keep the conversation going!

How can we transform the chaotic first weeks of the semester into a launchpad for student success while reducing our own cognitive load?

This energizing CRTLE hybrid session brought together faculty from across UTA to explore six research-backed strategies for starting strong and maintaining momentum throughout the semester. The workshop delivered practical tools and innovative approaches that help instructors create clarity, build community, and establish productive learning patterns from day one.

Watch the full recording here:

Managing Cognitive Load: The Foundation

The session opened with a crucial principle: managing cognitive load for both instructors and students. When we streamline and make things predictable, we reduce overwhelm while increasing engagement. Visual aids emerged as a powerful tool—think visual syllabi or roadmaps that help students grasp the big picture at a glance. Instead of wading through 17 pages of text, students can quickly understand course structure through icons, graphics, and clear visual organization. These elements pair perfectly with UTA’s new syllabus templates, which include visual, graphic, and digital format options.

Communication That Works: Intentional and Automated

The first major strategy focused on intentional communication while leveraging automation. Weekly emails on consistent days, course trailer videos, and Canvas’s pre-scheduled announcements let you set up the semester once while technology handles the routine. The standout tool was Microsoft Bookings—UTA’s scheduling solution that integrates with Outlook, allows student self-scheduling, auto-populates Teams meetings, and eliminates endless emails. After an hour of initial setup (request through Service Now), the time savings are substantial, with students able to join meetings via mobile app from anywhere on campus.

Building Community from Day One

Creating genuine classroom connections requires intentional design. The Common Activity has groups of three to six students discovering non-obvious connections—not visible traits but deeper commonalities like birth order or coffee preferences that create lasting semester-long alliances. Creative icebreakers ranged from discussing binged shows to answering “If you had someone’s undivided attention for 10 minutes, what would you talk about?” Other innovations include creating class playlists from students’ favorite songs and using advice cards from previous students to guide current learners.

Low Stakes, High Impact

Low-stakes assignments solve multiple challenges simultaneously: providing early data for progress reports, building confidence gradually, and creating feedback opportunities without high pressure. The key insight focuses on process over product—when students receive credit for struggling, messy drafts, and working through problems, they’re less likely to default to AI shortcuts. Canvas’s recording feature for verbal reflections, in-class activities with immediate artifacts, real-time exit tickets, and process-based assignments all provide alternatives to traditional discussion boards while combating AI misuse and benefiting all students, especially first-generation graduate students.

Creating Space for Productive Struggle

Learning requires struggle within a psychologically safe environment where mistakes become learning opportunities rather than shortcuts to AI. The solution involves deliberately building time for grappling with concepts, followed by reflection where “aha moments” crystallize. Breaking complex concepts into digestible pieces, providing strategic milestone reminders, and rotating classroom participation (“Today I want to hear from the left side of the room”) all ensure diverse engagement without spotlighting individual students while maintaining the productive challenge essential for deep learning.

Motivation Through Connection

Pre-course surveys using Question Pro or Microsoft Forms reveal cohort-specific needs, knowledge levels, and technology access, enabling tailored instruction. Making content relevant means connecting to careers when possible, incorporating current events, or sharing instructor passion through videos and stories. Visual roadmaps showing the semester journey help students understand not just where they’re going but why each step matters—transparency about instructor challenges gives students permission to struggle productively themselves.

Khanmigo: Your AI Teaching Assistant

The session’s showstopper was Khanmigo, Khan Academy’s teacher-facing AI tool now integrated into Canvas at UTA. This tool generates discussion prompts, exit tickets, hook activities, learning objectives, rubrics, lesson plans, and question banks with answer keys—all while keeping content within UTA’s system for privacy and safety. To enable it, navigate to Canvas course settings, find “Khanmigo Teacher Tools” in navigation, drag to visible items, and save. After one-time setup, you’ll access tiles for various teaching tasks that reduce prep time while maintaining pedagogical quality through customizable, iterative outputs.

Addressing the AI Challenge Head-On

The session confronted AI’s impact on traditional assignments by reimagining assessment rather than fighting technology. Solutions include moving to in-class completion, using multimodal formats (audio/video), focusing on process over product, and requiring personal connections that resist automation. The paradox is clear: using AI tools like Khanmigo to reduce instructor workload frees time to create more engaging, AI-resistant assignments that maintain authentic student engagement.

Resources

• Khanmigo Tutorial: Available on the Pedagogy Next blog
• Visual Syllabus Templates: Available on the Provost’s website
• Faculty Technology Support: Trinity Hall Technology Bar/Service Desk
• CRTLE Team: Available for one-on-one consulting

Join the Conversation

Which strategies will you implement first this semester? Have you tried Microsoft Bookings or Khanmigo yet? What creative solutions have you found for maintaining authentic student engagement in the age of AI? Share your experiences and questions in the comments below. Remember: don’t just start the semester—define it!

Post written by Dr. Sarah Shelton and Dr. Peggy Semingson

At UT Arlington, we believe early alerts (feedback in courses by the end of week three or four) are a critical prerequisite to offering meaningful and productive academic support to students, providing them with a sense for how well they’re doing as soon as possible and you with data on how well your course design is supporting course goals. In general, the first month of class is a critical time and letting students know how they’re doing sooner rather than later can make a big difference in whether or not they feel engaged, find a sense of belonging, and believe they can succeed. By giving students constructively critical feedback during the first weeks of a course, you:

• Help Students Make Sound Educational Decisions: Early feedback helps students adjust study habits, seek support, or decide whether to drop a course.
• Help the University Identify and Support At-risk Students: Timely interim grades enable the university to provide targeted support to at-risk students.
• Inform Your Teaching: Early insight into student performance allows you to adjust your teaching for better learning outcomes.
• Foster a Culture that Values Inquiry and Dialogue: Feedback shows you care about students’ growth and encourages academic dialogue.

But giving students early feedback does involve some additional preparation time and effort as you rethink how you evaluate your students. To do this successfully try to:

• Develop short but meaningful evaluation tasks.
• Schedule short periods of class time to allow students to respond.
• Review, record, and return students’ work in a timely fashion.
• Take time to reflect upon what you’re learning about your students.
• Adjust as you see fit.

Begin by designing low-stakes assignments graded quickly (24-48 hours, ideally) to promote student engagement and build confidence early in the semester. This will also help with gathering grades for the four-week progress report. Below are some ideas that can help you gather early data on student performance, build rapport, and engage learners without overwhelming your grading load. Many can be adapted for large classes and graded on a simple completion scale.

Syllabus & Course Familiarity

Syllabus Scavenger Hunt
Ask students to find key information in the syllabus (e.g., policies, due dates) and share it. Ensures they’ve read it and know where to locate important details.

“About Me” Pre-Course Survey
Use a short Canvas or Microsoft Forms survey to learn about students’ backgrounds, interests, and prior experience with the subject. Helps you tailor examples and build rapport.

Early Understanding, Reflection, and Content Checks

Class Summary Statements
At the end of class, have students write a two-minute summary of the day’s main points, or summarize the previous meeting’s main points at the start. For large classes, use:

• Canvas Discussion Board/Quiz: Students submit summaries online.
• SpeedGrader Assignment: Upload summaries with a 3–4 point rubric.
• Survey/Poll: Ungraded Canvas survey with one open-ended summary question.

“Muddiest Point” Cards
Students jot down the concept they found most confusing. Review responses to guide next class; grade for submission only.

Concept Mapping
Students create a simple visual map linking key terms from the first lecture or reading. Submit a paper or digital version; grade for completion.

Quick Poll + Debrief
Pose one or two conceptual questions via Canvas or polling app. Discuss results live; award participation credit.

Sorting or Categorizing Task
Give examples, problems, or images for students to group into categories or sequences. Submit as a quick poll or photo.

Minute Demonstration or Problem-Solving
Have students solve one problem, label a diagram, or interpret a chart in class. Collect for completion credit.

The Evolving Study Guide
Have students submit two potential exam questions after each class to build a collaborative course study guide. Rate each for quality (0–3 scale) and consider using strong questions on actual exams. (Collect via Canvas discussion board, shared doc, or wiki.)

The Five-Minute “Microtheme”
Ask students to write a short in-class essay (one paragraph on a 5×8 card or single page) in response to a prompt. Grade quickly with a simple rubric or point scale. (For larger classes, use SpeedGrader in Canvas or collect only from a rotating sample of students each time.)

Group & Community-Building Activities

Peer Interview & Introduction (Pairs/Small Groups)
Students interview a partner or group members, then introduce them to the class (or post in a discussion board).

Group Brainstorm Board
Small groups brainstorm examples or applications of a concept on a shared online whiteboard or large paper. Submit a link or photo for completion credit.

More Multimodal Options

Video Reflection or Verbal Processing
Have students use Canvas’s media/record tool (or their phones) to submit a short 1–2 minute video or audio reflection on what they learned, how they approached an assignment, or how they’d explain a concept to someone else. Grade for completion or use a simple 3–4 point scale. (In larger classes, rotate who submits each week.)

Photo-Based Process Log
Ask students to take a photo of a physical or digital work-in-progress—such as notes, a diagram, a whiteboard sketch—and write 1–2 sentences explaining their process. Submit as an image + caption in Canvas.

Concept or Process Infographic
Students create a simple visual (using Canva, Google Slides, or even hand-drawn and photographed) that explains a key idea or process from the week. Grade for clarity and completeness.

Audio “Mini-Podcast”
Have students record a 1–2 minute audio segment where they summarize a reading, explain a concept, or interview a peer. They can submit as an audio file through Canvas.

Slide Deck Snapshot
Students create 2–3 slides illustrating the main takeaways from a reading or class discussion (images + key words). Upload as a PDF or PowerPoint file in Canvas.

Caption This!
Provide an image, chart, or figure from the week’s materials and have students create a one-sentence “caption” that accurately describes or interprets it. Collect via discussion board for quick viewing.

Collaborative Video Explainer
In small groups, students create a 1–3 minute video explaining a concept from class, using simple visuals, demonstrations, or examples. They can record on phones or via Zoom and upload to Canvas. Grade for clarity and teamwork, not production quality.

Returning Feedback Quickly

If the point is to get students feedback in the first weeks, what we design has to be sustainable for our unique workloads and class sizes. How to provide early feedback is completely up to you. The key is to keep the costs low while still providing the attendant benefits. Here are some things to consider to help maximize the return for students while keeping the labor sustainable for you:

• If you want a fast way to measure performance, develop a quick grading scale like 0–3, where 0 indicates no submission, 1 reflects minimal or incomplete work, 2 shows basic competence with some errors, and 3 demonstrates strong understanding and complete work.
• If the activity is more about engaging and going through the process, use the Complete/Incomplete option in Canvas to track that participation.
• If you’re worried that smaller assignments could throw off your grade weights, check the “Do not count this assignment towards the final grade” box in Canvas. You can still mark them complete or incomplete, use a rubric or scale to indicate feedback, or leave a few quick comments for feedback without it being calculated in the final grade.
• If you’re more concerned about giving specific feedback for revision or growth but don’t have the time to give in-depth comments for every assignment or student, use AI like Copilot to help you write rubrics with more detailed feedback built into them.

Digital Resources for Low-Stakes Assignments

You don’t have to start from scratch or design these low-stakes assignments without support. Try out the below digital tools, all integrated into Canvas, to help you spark ideas and create the assignments themselves.

Perusall is a social annotation tool already available to us in Canvas. It includes the option to allow Perusall to grade students’ annotation work based on factors like quality of comments, time in the document, keystrokes, etc. For a low-stakes option, consider having students annotate a reading, video, or podcast before a class where they need that content for the lecture or activity. See our Vendor-led demo and our interview with Dr. Katie Welch for more info and ideas.

Khanmigo is an AI-powered learning companion developed by Khan Academy that supports student learning and enhances classroom instruction. It can help you quickly create low-stakes assignments like short practice quizzes, guided problem-solving sets, writing prompts with automated feedback, scaffolded discussion questions, exit tickets and more. See our demo and info session with CDE’s Melissa Roach and Joseph Rutledge for more info and ideas.

DesignPlus is a suite of tools in Canvas that helps instructors create visually engaging, well-organized course content with minimal effort. Part of that design can integrate low-stakes assignments (like short self-check quizzes with immediate feedback, simple reflection journals, or visually guided problem sets that walk students through a process step-by-step) directly into content. See our session with CDE’s Jess Kahlow for more info and ideas.

Inspire for Faculty (click here) allows you to see a heat map of student engagement as well as tips for email “nudges” to help students get on track.

Join the Conversation  

We’d love to hear from you! Let’s share ideas here and keep the conversation going to inspire and support each other. How are you planning to get feedback to your students early in the semester and/or incorporate low-stakes assignments/activities into your course?

How can we transform solitary reading into a collaborative learning experience that keeps students engaged and accountable in today’s educational landscape?

That’s the question Nolan Quella, Engagement Manager at Perusall, addressed in this comprehensive CRTLE training session. This hands-on webinar walked participants through the complete setup and implementation of Perusall in Canvas courses, demonstrating how this social annotation platform can revolutionize the way students interact with course materials.

Watch the full recording here:

Quick Start: Canvas Integration Made Simple

One of the session’s key takeaways was just how straightforward it is to add Perusall to Canvas courses at UTA. Since the university has already upgraded to LTI 1.3 and Perusall is globally installed, integration takes just three clicks:

• Automatic grade sync – grades flow seamlessly back to Canvas
• Roster sync – student information transfers automatically
• Group sync – Canvas groups can be imported directly
• Single sign-on – students access Perusall through Canvas without creating separate accounts

Everything is created in Perusall first, then syncs automatically to Canvas, eliminating duplicate work and technical issues.

Building Your Perusall Course

Course Setup Options

The platform offers two pathways for course creation:

• Empty course – Build from scratch with full customization
• Copy existing course – Save time by duplicating previous courses, colleague templates, or instructional designer frameworks (copies only the shell, no student data)

Starter Assignments

Perusall provides two pre-built assignments to help students learn the platform:

• “Making the Most of Learning with Perusall” – A 30-minute scaffolded tutorial teaching students how to write quality comments and use all platform features
• Syllabus Review Assignment – Upload your syllabus and track whether students actually read it while gathering their questions and feedback

Content Selection

The platform supports multiple content types:

• Over 1.4 million textbook titles (with bookstore integration for financial aid)
• PDFs, Word documents, PowerPoint presentations
• YouTube and Vimeo videos
• Podcasts and audio files
• Google Docs and Slides
• Images and EPUBs
• Even quizzes imported from Canvas

The Power of Social Learning: Groups and Engagement

One of Perusall’s standout features is its intelligent grouping system. Rather than having 30+ students create 300+ comments on a single assignment, the platform automatically:

• Creates groups of 4-12 students (customizable based on class size)
• Shuffles groups between different content types for fresh perspectives
• Maintains consistent groups within each content piece for continuity
• Allows manual group creation or Canvas group import when needed

This structure creates more manageable conversations where students feel comfortable asking questions and engaging with peers.

Analytics That Matter: Beyond Simple Metrics

The analytics dashboard provides insights instructors typically never see:

The Confusion Report – Automatically identifies areas where students struggled most, solving the age-old problem of “Any questions about the reading?” met with silence.

Engagement Tracking includes:

• Viewing time – How long the assignment was open
• Active engagement time – Actual interaction with content (2-minute windows tracking clicks, scrolls, comments)
• Page view heat maps – Shows which pages students spend the most time on
• Comment submission patterns – Reveals when students complete work

Student Activity Reports help identify when students claim they “did all the work” but actually spent only 7 minutes on a 45-minute assignment.

This approach ensures students who struggle with one skill (like comment quality) can still succeed by excelling in others (like consistent engagement or peer interaction).

Resources

• Canvas Instructor Quick Setup: https://support.perusall.com/hc/en-us/articles/14047817114263-Canvas-Instructor-Quick-Setup
• Canvas Integration Features: https://support.perusall.com/hc/en-us/articles/22854094490775-Canvas-Integration-Features
• Dr. Katie Welch’s Setup Screencast: https://websites.uta.edu/pedagogynext/icymi-recap-of-july-23-session-reading-in-the-age-of-ai-introduction-to-perusall-with-katie-welch-and-peggy-semingson/
• Contact Nolan Quella: engagement@perusall.com

Join the Conversation

We’d love to hear from you! Are you ready to transform your course readings into collaborative learning experiences? What aspects of Perusall excite you most – the analytics, the social learning environment, or the automatic grading? Have you tried social annotation tools before? Share your thoughts, questions, and experiences in the comments below. Let’s keep the conversation going!

We’ve assembled a collection of books to support faculty interested in AI and education. These resources cover pedagogy, curriculum design, and practical AI applications in teaching. All books are available for checkout.

Available Books

Co-Intelligence: Living and Working with AI by Ethan Mollick
How to work alongside AI as a collaborative partner in education. Mollick, a Wharton professor, provides practical strategies for using AI to enhance rather than replace human creativity and teaching. Includes real classroom examples and immediately applicable techniques.

AI Snake Oil by Arvind Narayanan and Sayash Kapoor
A critical look at what AI can actually do versus marketing hype. The book helps educators identify which AI tools deliver real value and which are empty promises. Essential reading for making informed decisions about AI adoption.

More than Words by John Warner
Rethinking writing instruction in the age of AI. Warner addresses the fundamental question of what writing means when machines can generate text, arguing that writing is about thinking, not just producing words. Offers practical approaches to maintaining academic integrity while embracing useful AI tools.

AI Optimism by Becky Keene
A guide to using AI effectively in educational settings. Keene focuses on equity and accessibility, showing how AI can help level the playing field for diverse learners. Provides frameworks for ethical implementation across different educational contexts.

AI with Intention by Tony Frontier
Principles and action steps for teachers and school leaders. This book moves beyond random AI experiments to strategic, purposeful implementation. Ideal for department chairs and coordinators leading AI initiatives.

The AI Workshop by Milo Foster
A beginner’s guide to understanding AI basics. Written specifically for educators without technical backgrounds, Foster breaks down complex concepts into clear, manageable chapters. Perfect starting point for faculty new to AI.

Teaching with AI by Jose Antonio Bowen and C. Edward Watson
Practical guide for integrating AI into teaching practice. The authors combine pedagogical expertise with AI insights, offering specific examples across various disciplines. Includes case studies from the humanities, sciences, and professional programs.

The Power of AI for Educators by Victoria R. Summers
Tools and techniques for classroom use. Summers focuses on time-saving applications and student engagement strategies that work today. Features step-by-step guides for popular AI education tools.

The AI Con by Emily M. Bender and Alex Hanna
Critical perspectives on AI development and implementation. The authors examine the dangers of unchecked AI adoption and advocate for technology that serves educational rather than corporate goals. Important counterbalance to tech industry narratives.

Empire of AI by Karen Hao
Global perspectives on AI’s impact and implications. Through investigative journalism, Hao reveals the human stories behind AI development worldwide. Helps faculty understand the broader context students will face in an AI-driven world.

How to Borrow

1. Fill out the checkout form here or at the QR code below.
2. Pick up the book from the CRTLE office.

Checkout Form: Here
Location: Trinity Hall, Room 106
Contact: crtle@uta.edu

Join the Conversation

We’re also accepting suggestions for additional books on teaching, learning, and AI. Let us know in the comments below what titles you’d like to see in our collection.

CRTLE Faculty Facilitators – Learning Analytics Fellowship (Two faculty members) 

Position Title: University Faculty Facilitator: Learning Analytics Fellows

*Use the following subject line in your application email.: 2025-26 Analytics-Focused CRTLE Facilitator Application. 

Hosted by: CRTLE (Center for Research on Teaching and Learning Excellence) in conjunction/collaboration with University Analytics. 

Duration: 2 Years (Fall and Spring Semesters)  

Weekly Time Commitment: 3-4 hours per week.

Stipend: $4,000 per year [paid $2000 at end of fall semester and $2000 at end of spring semester] 

Overview: This would be two faculty members to join the CRTLE Faculty Facilitator program. There will already be eight funded faculty facilitators (one per academic unit). The Learning Analytics Fellowship program will be part of this broader program and would aim to enhance our understanding and application of learning analytics to improve student outcomes and institutional effectiveness. These two positions are funded by University Analytics and the position is in CRTLE but those in these roles will collaborate with University Analytics.

Expectations: 

• Collaborate with faculty, CRTLE, and University Analytics to identify key areas for learning analytics application to teaching and learning (e.g., DFW Courses or large courses, gainful employment). 

• Design and implement data-driven strategies to enhance teaching and learning practices with accessible deliverables for faculty (e.g., videos, handouts, guides, blog posts, infographics). 

• Conduct virtual-only or blended workshops and training sessions for faculty and staff on the use of learning analytics tools and methodologies with clear applications to teaching and learning. 

• Work with University Analytics to analyze and interpret data to provide actionable insights and recommendations for CRTLE. 

• Participate in regular CRTLE Faculty Facilitator meetings and contribute to the community through knowledge sharing and collaboration. 

• Work collaboratively with other units as needed (e.g., career center) to look at issues related to teaching and learning.  

Qualifications: 

• Current full-time faculty member with a strong interest and background in learning analytics and educational technology. 

• Demonstrated experience in data analysis, research, and applications to teaching is preferred. 

• Background in applying analytics or quantitative approaches to pedagogy preferred.  

• Solid communication and presentation skills. 

• Ability to work collaboratively. 

• Commitment to improving student learning outcomes through innovative approaches and analytics. 

Application:

To Apply, submit the following by or before Friday, August 15, 2025 (5 pm) to CRTLE@uta.edu. Use the following subject line in your email: 2025-26 Analytics-Focused CRTLE Facilitator Application. Please be sure to note this is for the Analytics facilitator role in addition to it being noted in your materials, as well.

1. Detailed cover letter with your background related to teaching, why you want to be a facilitator, experience with CRTLE or faculty professional development related to teaching, and ideas you would bring to the role. 
2. Include a formal letter of support from your department chair or indicating that they are aware of the time demands of being a facilitator.  
3. Include a second letter from another colleague who can support your qualifications to be a facilitator, including the criteria listed here.  
4. Copy of your Curriculum vita

Before 2020, online learning was growing steadily but was not as ubiquitous as it is today. Then, in the spring semester of 2020, everything changed due to the COVID-19 pandemic, requiring rapid, widespread adoption of remote instruction. Millions of faculty and students, many of whom had never experienced digital learning environments, experienced online education up close and personal. This period undeniably accelerated the integration of technology into how we approach education and, ever since, faculty have been learning what works and what doesn’t when the physical environment is removed.

Many of us for the first time had to take a crash course in learning the distinction between delivering content online and facilitating learning online. This is where Regular and Substantive Interaction (RSI) moves from a compliance check box to an essential pedagogical strategy. RSI has been a requirement in some form or fashion since 1992, when the federal government distinguished between correspondence courses and online learning, noting (of correspondence courses) that “Interaction between the instructor and student is limited, is not regular and substantive, and is primarily initiated by the student.” In 2005 the phrase “regular and substantive interaction” officially entered the lexicon in the Higher Education Reconciliation Act of 2005 and was further clarified and codified in 2021 in response to the COVID-19 pandemic (1). Quality distance learning programs, regardless of the technologies used, must provide a certain amount of interaction and mentoring with faculty.

In asynchronous classes, the absence of the physical environment and in-person connection with peers can lead to feelings of isolation and disengagement if not carefully counteracted, according to Gasell, et al. (2). They report that students in asynchronous classes may report a lack of engagement, feeling impatient, bored, and a lack of instructor involvement. This is important, as instructor-to-student interaction significantly influences student perception, satisfaction, and achievement in online courses. The work of Gasell et al also found that 23% of analyzed online courses had no instructor posts in discussion boards. This method of “set and forget” is clearly not meeting the needs of our students.

Fortunately, regular and substantive interaction doesn’t have to be a daunting task. Here are some steps faculty can take to ensure that students feel engaged in asynchronous online classes:

1. For larger online classes that use discussion boards, put students into discussion groups. Instructions on how to do this can be found here. When grading the discussion boards, you can interact with the students by focusing on themes and specific interactions within the group.
2. Use Inspire for Faculty. Inspire for Faculty allows instructors to send individual correspondence to groups of students based on certain criteria. For example, Inspire has the functionality to sort students by performance on a certain assessment or assignment (<70%, for example). A  neat feature of Inspire is that you can then send an email to the group who meets the specified criteria and it will automatically personalize it by addressing the email to the individual student (so instead of “dear student” it will say “Dear Tyler”). I’ve used this strategy myself and I have received overwhelmingly positive feedback, with some students saying that was the first time a professor had ever emailed them personally before.
3. Hold virtual office hours or scheduled one-on-one meetings. Tools like Microsoft Teams can be used to set up regular times when students can drop in to ask questions, discuss course material, or just check in. For larger classes, you could offer sign-up slots for brief 10-15 minute individual meetings. This provides a clear, predictable opportunity for direct, live interaction. The personal connection can be especially valuable for students who might be struggling or feel isolated in a fully asynchronous environment. With online asynchronous classes it’s easy to default to “by appointment” but by holding regular “drop-in hours” at a set time (and actively encouraging students to attend!), students feel more comfortable and welcome knowing that you’re going to be there regardless of whether they have an appointment or not.

Ultimately, the journey since 2020 has challenged us to reframe our understanding of online teaching. The distinction between simply delivering content and facilitating learning has become clearer than ever, with Regular and Substantive Interaction (RSI) standing as the bridge between the two. The strategies outlined here—using small discussion groups, leveraging technology for personalized communication, and offering virtual meetings—demonstrate that meeting the criteria for RSI is not a burden, but an opportunity. It is a chance to intentionally design courses that prioritize student engagement, combat feelings of isolation, and create a supportive educational community, no matter the physical distance. In doing so, we not only meet a regulatory standard but also fulfill our primary mission as educators.

Join the Conversation

How has your approach to online teaching changed since 2020? What strategies have you found most effective in fostering regular and substantive interaction with your students—especially in asynchronous environments? Share your reflections, challenges, and success stories in the comments below. Let’s continue learning from one another as we move from compliance to connection in digital teaching.