Can chatbots really tutor language learners—or are they just good at sounding like they can? In this episode, the hosts explore what it takes for AI to move beyond conversation and become an effective language tutor. Drawing on decades of research in computer-assisted language learning (CALL), they connect lessons from traditional intelligent tutoring systems to today’s AI chatbots. Along the way, they look at tools like Google Translate’s practice mode and consider how chatbots deliver feedback, correct errors, and guide learners through language practice. But fluent conversation doesn’t necessarily mean effective teaching. Join the hosts as they unpack the promise and limitations of AI tutoring and ask what it really takes for chatbots to support language learning.

In this episode:

Hubbard P, Siskin CB. Another look at tutorial CALL. ReCALL. 2004;16(2):448-461. doi:10.1017/S0958344004001326

Levy, Michael, Computer-Assisted Language Learning: Context and Conceptualization (Oxford, 1997; online edn, Oxford Academic, 31 Oct. 2023), https://doi.org/10.1093/oso/9780198236320.001.0001, accessed 22 Oct. 2026.

Heift T, Schulze M. Tutorial computer-assisted language learning. Language Teaching. 2015;48(4):471-490. doi:10.1017/S0261444815000245

Schulze, M. (2025). ICALL and AI: Seven lessons from seventy years. In Y. Wang, A. Alm, & G. Dizon (Eds.), Insights into AI and language teaching and learning (pp. 11–31). Castledown. https://doi.org/10.29140/9781763711600-02

In this episode, Mat and Phil continue to explore Sam Illingworth’s Slow AI, this time turning their attention to AI detectors. Beyond the question of whether AI detectors work, the hosts discuss the technical, ethical and pedagogical concerns surrounding their use in education. They address the risk of false positives, the ways in which second-language writers may be disproportionately affected, and examples of students altering their writing or using AI-humanizer tools to avoid AI detection. They also consider the potential harms of AI detectors to teaching and learning, including its use of energy and water, the proliferation of AI slop, and the possibility of addictive use and brain rot.

In this episode:

Sam Illingworth’s Slow AI: Knowing When to Use AI and When to Leave It Alone, https://samillingworth.com/book

Sam Illingworth’s substack: https://substack.com/@samillingworth

Another of Illingworth’s books (open access): Illingworth, S., & Forsyth, R. (2026). GenAI in higher education: Redefining teaching and learning. Bloomsbury Publishing. https://www.bloomsburycollections.com/monograph?docid=b-9781350535824

When should you use AI, and when should you leave it alone? In today’s episode, we dig into the central question of Sam Illingworth’s book Slow AI: Know When to Use AI and When to Leave it Alone, exploring what it means to move beyond AI tools and techniques and toward using AI in ways that are genuinely worthwhile.

Considering a variety of language teaching practices and processes—from adapting texts to different reading levels to creating lesson plans or grading student work—Mat and Phil discuss the importance of judging when a task can truly benefit from AI and when AI removes something valuable, whether it’s human judgment or processes that are themselves essential to learning.

In this episode:

Sam Illingworth’s Slow AI: Knowing When to Use AI and When to Leave It Alone https://samillingworth.com/book