
Hosted by Zach Justus and Nik Janos · EN
Unfixed is a podcast for anyone interested in how generative AI is transforming higher education, from teachers and administrators to students and curious learners. Hosted by the faculty behind the blog Melts into Air, it offers candid reflections on a university system in flux, where even the mission of higher ed is up for debate.

Asynchronous online courses have become a cornerstone—and a cash engine—of modern higher education, but in the age of AI, the model is starting to crack. In this episode, we unpack how large-scale, low-touch online classes create fragile conditions for trust, especially as generative AI makes it easier than ever for students to outsource their thinking. We explore the growing tension between institutional incentives to scale enrollment and the reality of eroding learning quality, faculty burnout, and widespread uncertainty about what student work actually represents. Drawing on firsthand teaching experience, we examine how instructors are adapting and where those efforts fall short. The result is a stark question: can asynchronous education be rebuilt for the AI era, or are we sustaining a system we no longer fully believe in?Thoughts and suggestions? Email us at unfixed@meltsintoair.orgYou can find the full show notes and our writing at meltsintoair.orgSubscribe to the Unfixed Newsletter—The AI Higher Ed Breakdown our bi-weekly newsletter unpacking the top news stories at the intersection of AI and higher education.

Generative AI is transforming how students write and how professors teach writing. In this episode of Unfixed, we’re joined by Stephanie Tran (Cypress College) and Tamara Tate (UC Irvine Digital Learning Lab) to explore what AI means for composition, academic writing, and digital literacy in higher education. Together, we examine the scale of the challenge facing writing instructors as AI models become increasingly proficient at generating polished text. What does “writing” mean across disciplines? And how can educators integrate AI into writing instruction without offloading the essential cognitive work that writing is meant to develop? Stephanie shares her work leading district-wide faculty development on equitable and human-centered AI practices. Tamara discusses her leadership at UC Irvine’s Digital Learning Lab and her experience developing AI-based writing tools designed to support — not replace — student skill building.Tamara TateUC Irvine Digital Learning LabPapyrusAI Resources, “Starting to use PapyrusAI or generative AI for writing?”Thoughts and suggestions? Email us at unfixed@meltsintoair.orgYou can find the full show notes and our writing at meltsintoair.orgSubscribe to the Unfixed Newsletter—The AI Higher Ed Breakdown our bi-weekly newsletter unpacking the top news stories at the intersection of AI and higher education.

Higher education is facing two converging pressures: generative AI disrupting early-career labor markets and new federal “earnings test” accountability rules tying program viability to graduate income. In this episode of Unfixed, we examine how AI is thinning entry-level work, why mid-tier public universities may be especially exposed, and how federal accountability metrics could amplify wage suppression caused by automation. Drawing on reporting from Jeffrey Selingo, PBS NewsHour, The Chronicle of Higher Education, and U.S. Department of Education materials, we explore why GenAI and federal regulation are no longer separate stories. We close by asking what university leaders, faculty, and systems—especially in California—should do before signaling effects and enrollment shocks reshape the sector.Jeffrey Selingo, New York Magazine / Intelligencer, What Is College For in the Age of AI?PBS NewsHour, How AI may be robbing new college graduates of traditional entry-level jobsU.S. Department of Education, U.S. Department of Education Reaches Consensus on Historic New Accountability FrameworkBrian O’Leary, The Chronicle of Higher Education, Colleges Face a New Test on Their Grads’ Earnings. These Programs Fail It. (Note: Paywalled)California State University, Strategic Planning for the CSU (CSU Forward)Thoughts and suggestions? Email us at unfixed@meltsintoair.orgYou can find the full show notes and our writing at meltsintoair.org

Ep. 22 Building AI Capacity in Public Higher Ed with James FrazeeIn this episode of Unfixed, we’re joined by Dr. James Frazee, Vice President for Information Technology and Chief Information Officer at San Diego State University, to unpack how large public universities are navigating AI strategy, governance, and equity at scale. As a nationally recognized leader in AI and data-informed decision-making, Frazee shares insights from spearheading SDSU’s AI and digital transformation survey—an effort that has since been adopted systemwide across the California State University system.James FrazeeSDSU Statement on Generative AI (GenAI) PrivacySDSU AI SurveyThoughts and suggestions? Email us at unfixed@meltsintoair.orgYou can find the full show notes and our writing at meltsintoair.org

In this episode of Unfixed Newswire, Nik and Zach break down the most consequential—and occasionally absurd—AI news. The conversation spans OpenAI’s push toward ads and potential NSFW content; the relentless churn of new AI models like Gemini 3 and GPT-5.x; and why tools like Claude Code and Claude Cowork may signal a real shift toward AI agents as co-workers. They also interrogate the current AI investment bubble, unpack why enterprise tools like Google Gemini’s Canvas integration may matter more than headline-grabbing launches, and close with some deliberately irresponsible speculation about humanoid robots.OpenAI begins piloting advertisements AI Daily Brief, January 19, 2026, How to Make ChatGPT Ads Not SuckThe Verge,ChatGPT’s ‘adult mode’ is expected to debut in Q1 2026, https://www.theverge.com/news/842657/openai-chatgpt-adult-mode-debut-q1-2026The Crimson, 1994, Porn on the internet!Humai, GPT-5.2 vs Claude Opus 4.5 vs Gemini 3 Pro: The Complete ComparisonThe AI Daily Brief, January 13 2026, Claude Cowork is Claude Code for everyone else Wired, Anthropic’s Claude Cowork Is an AI Agent That Actually Works Instructure, Unleash the Power of AI in Canvas with Gemini LTI™Neo, your home robotMarques Bownlee (MKBHD) on the problem with NEO the robot Thoughts and suggestions? Email us at unfixed@meltsintoair.orgYou can find the full show notes and our writing at meltsintoair.org

In this episode of Unfixed, Nik and Zach take on a difficult but necessary conversation about nostalgia in higher education. Using real media headlines, faculty experiences, and their own teaching histories, they explore how AI panic taps into a deeper longing for a pre-pandemic, pre-ChatGPT past that may never have been as stable as we remember. The discussion unpacks what nostalgia is, why it’s so powerful right now, and how scapegoating AI can obscure long-standing challenges around reading, writing, and motivation. Rather than a call to “go back,” this episode walks the line between insight and wake-up—arguing that AI often reveals what’s already broken and, in some cases, opens space to fix it.Zach Justus and Nik Janos, Inside Higher Ed, No, the Pre-AI Era Was Not That GreatHua Hsu, The New Yorker, What Happens after A.I. Destroys College WritingNew York Magazine, Everyone Is Cheating Their Way Through CollegeLindsay McKenzie, 2018, Inside Higher Ed, Learning Tool or Cheating Aid?Raphael, The School of AthensThoughts and suggestions? Email us at unfixed@meltsintoair.orgYou can find the full show notes and our writing at meltsintoair.org

Nik and Zach review the best and worst of AI in higher ed for 2025. AI and Higher Education: 2025 We unpack the collapse of AI detection, the rise of system-wide AI partnerships like CSU and OpenAI, uneven assessment redesign, growing faculty stratification, and mounting concern that generative AI is hollowing out entry-level knowledge-work jobs. The episode closes by looking ahead to 2026, including model updates, political backlash, environmental impacts, and what higher education faces if the AI bubble bursts.Nik and Zach’s top five of the year:Zach Justus, That terrible “Everyone is cheating their way through college” essayNik Janos, The Assistant for the Rest of UsNik Janos, Builders: Designing the Post-AI UniversityZach Justus and Nik Janos, No, The Pre-AI Era Wasn’t that Great (Inside Higher Ed)Zach Justus and Nik Janos, Why professors are more important than ever in the AI era (EdSource)Mentioned:Rampant AI Cheating Is Ruining Education Alarmingly Fast (Intelligencer / New York Magazine, May 7, 2025)Experts Weigh In on “Everyone” Cheating in College (Inside Higher Ed, May 20, 2025)Unfixed episode 16: Inside the CSU-OpenAI PartnershipOpenAI and the California State University system bring AI to 500,000 students and faculty (OpenAI, Feb 2025)CSU, OpenAI roll out ChatGPT Edu to California college students (Axios, Feb 4, 2025)OpenAI targets higher education with ChatGPT rollout at CSU (Reuters, Feb 4, 2025)2025 EDUCAUSE AI Landscape Study: Into the Digital AI Divide (EDUCAUSE, Feb 2025)AI is Destroying the University and Learning Itself (Current Affairs, Dec 2025)These Students Use AI a Lot — but Not to Cheat (The Chronicle of Higher Education, 2025)How AI Is Changing—Not ‘Killing’—College (Inside Higher Ed, Aug 29, 2025)2025 AI Index Report (Stanford HAI, 2025)How to Think of AI in Education (MIT Open Learning, July 2025)Unfixed episode 11: AGI and the Future of Higher Ed: talking with Ray SchoederUniversities risk becoming passive arms of Silicon Valley if they don’t question AI (Business Insider, Dec 2025)Thoughts and suggestions? Email us at unfixed@meltsintoair.orgYou can find the full show notes and our writing at meltsintoair.org

Dr. Shelley Hart joins Nik and Zach to explore how children and adolescents are forming relationships with generative AI. We discuss what “too young” means in an AI-saturated world, and how parents, teachers, and caregivers can responsibly shape norms, boundaries, and ethics around AI use. Drawing on her expertise in child development, Dr. Hart offers practical insights for guiding kids through emerging technologies.Shelley HartNik Janos, Typing, Talking, Googling: Seeing the AI-first generation Todd Feathers, Wired, Parents Fell in Love with Alpha School’s Promise then They Wanted OutFirst 1000 DaysServe and returnUC Irvine study, California Parents Report their Fears and Hopes for AIMichael Connelly’s novel The Proving Ground Thoughts and suggestions? Email us at unfixed@meltsintoair.orgYou can find the full show notes and our writing at meltsintoair.org