What can small language models teach us that the largest AI models cannot?

Kelly and Julian are joined by Microsoft Cloud Advocate Gwyneth Peña-Sigüenza to explore why working with small language models (SLMs) may be one of the best ways to understand AI. Rather than relying on increasingly capable models that hide complexity, Gwyneth argues that constraints build stronger fundamentals. From prompt engineering and context management to deployment and security, SLMs force learners to think more carefully about how AI actually works.

The conversation extends beyond AI models into learning itself. Gwyneth shares her self-taught journey from growing up on a remote farm in Ecuador with limited internet access to becoming a Microsoft Cloud Advocate and creator of the Learn to Cloud platform. Along the way, the group discusses productive struggle, mentorship, cloud engineering, Python, security, and what educators should prioritize as AI becomes part of every student's learning experience.

The episode closes with a thoughtful discussion about AI dependency, judgment, and whether we would actually flip the switch and turn AI off if given the choice.

• Gwyneth celebrates the New York Knicks reaching the NBA Finals after more than 50 years.
• Julian shares that he has accepted a new role as a Fractional CTO.
• Kelly reflects on taking her first real vacation in over a year—and how stepping away from work sparked unexpected ideas.

• Why SLMs are valuable teaching tools
• Learning prompt engineering through constraints
• Running models locally on everyday hardware
• When local AI makes sense for classrooms
• Understanding tokens, context windows, and model limitations
• Why bigger models can sometimes hide important lessons

• Learning to drive in an old manual pickup truck as a metaphor for learning AI fundamentals
• Why difficult learning experiences often create lasting understanding
• Building strong habits before relying on more capable tools
• Consistency versus constantly chasing the newest resource

• Growing up without reliable internet in rural Ecuador
• Downloading YouTube playlists to learn programming offline
• Developing discipline through limited access
• The value of repetition and focused practice
• Why mentorship accelerates learning

• Transitioning from cloud engineering to Python advocacy
• Learning Python beyond scripting
• Discovering what "Pythonic" really means
• Wrestling with list comprehensions and other advanced syntax
• Favorite learning resources:
  • Fluent Python
  • Effective Python

• Building an open-source cloud engineering curriculum
• Hands-on labs and automated verification
• AI-assisted assessment
• Supporting self-taught learners around the world
• Creating accessible technical education

• Deploying AI applications to the cloud
• Containers, virtual machines, and serverless deployments
• Why operations and security deserve more classroom attention
• Introducing secure development practices early
• The importance of authentication, secrets management, and responsible deployment

• Helping students understand how AI works instead of simply using it
• Why productive struggle still matters
• The changing role of educators
• Balancing AI assistance with independent thinking
• Preparing students for a future where AI is always available

• AI dependency versus capability
• Judgment as the skill that matters most
• Human connection in an AI-driven world
• Would we actually turn AI off?
• Finding balance between technological progress and intentional learning

In this episode, Python Developer Advocate and author Will Vincent joins the hosts to discuss the lasting appeal of Django, changes in how people learn web development, and the ways AI is reshaping software engineering. While modern AI tools can generate working code in seconds, Django's opinionated design and emphasis on maintainability help developers avoid many of the security and architectural problems that often emerge as projects grow.

Drawing on his background as an educator, author, and Developer Advocate at JetBrains, Will shares his perspective on the challenges facing today's developers and computer science students. The conversation touches on "vibe coding," the misconception that a successful prototype automatically translates into a production-ready application, and the increasing burden AI-generated content is placing on open-source maintainers. Will also discusses the rise of specialized AI models, the importance of human trust in technical communities, and why foundational software engineering skills remain valuable despite rapid advances in AI tooling.

Key Topics Covered

Why Django Still Matters
A look at why Django continues to be a strong choice for building production applications, even if it doesn't receive the same level of attention as newer frameworks.

The Reality Behind "Vibe Coding"
Exploring the gap between generating code with AI and understanding the systems, tradeoffs, and architecture required to build reliable software.

Learning to Program as an Adult
Will reflects on his path from book editing and startup leadership to becoming a self-taught programmer, educator, and author.

AI and Programming Education
A discussion about how AI changes the learning process, why fundamentals still matter, and how concepts like music theory can help explain the value of understanding code beneath the surface.

The Growing Burden on Open Source
How maintainers are dealing with an influx of low-quality AI-generated issues, pull requests, and content, and what that means for community-driven projects.

Local and Specialized AI Models
Why privacy concerns, lower inference costs, and better hardware may drive adoption of smaller, task-focused models rather than ever-larger general systems.

Developer Concerns in the AI Era
How engineers are responding to growing pressure from leadership teams eager to adopt AI, and what trends JetBrains is seeing across the developer ecosystem.

Resources Mentioned
LearnDjango, Will Vincent's platform for learning Django and web development.
Hello World 5 Different Ways, a Django tutorial that introduces key concepts through practical examples.
Django Chat, the podcast Will co-hosts covering the Django ecosystem and web development.
Django News, a weekly newsletter highlighting updates from the Django community.
JetBrains, the software development company behind tools such as PyCharm and IntelliJ IDEA.

Special Guest: Will Vincent.

Kelly talks with Philip Guo, creator of Python Tutor, about how the tool helps students trace code and understand programming basics. They also discuss the challenges AI-generated code creates in the classroom and possible ways to support student learning.

*Wins of the Week
*
Philip: Hiring a second undergraduate student for Python Tutor, including one focused on user experience research with K-12 teachers
Kelly: Finishing a year of in-person teacher trainings and reflecting on how far the teachers have come

*AI, Coding, and Classroom Understanding
*
Much of the conversation focuses on how AI-generated code affects student learning. Kelly describes using AI code with eighth graders and how difficult it can be for them to understand functions, parameters, returns, and other fundamentals when the code is generated all at once. Philip suggests that tools like Python Tutor may be useful for helping students trace code and understand what is happening behind the scenes.

Python Tutor and Possible AI Features

Philip explains that Python Tutor currently visualizes execution and has an AI chat feature that can answer questions about code and errors. They discuss possible future features, including simplified AI-generated examples, alternative execution views that show only the lines actually run, and more guided inline help tied to specific code or variables.

Oral Explanations and Assessment

Kelly describes using a Socratic-style code review with students, where they discuss code aloud in groups. They also talk about using spoken explanations or short oral assessments to check whether students can really explain what code is doing, rather than just copying or prompting AI-generated answers.

Broader Research and “Beyond the Desk”

Philip briefly discusses a new research direction with a PhD student focused on AI support for work beyond the desk, including physical and embodied tasks in science labs and fieldwork. He says this differs from desk-based AI work and involves activities that are harder for current AI systems to support.

**Chapters
**0:25 Python Tutor and AI Learning
1:55 Hiring Help for Python Tutor
4:07 Classroom Wins and AI Reflections
6:11 Teaching Code Through Python Tutor
9:03 AI Code and Student Confusion
14:11 Simplifying Execution Traces
17:19 Functions Are the Hard Part
20:25 Keeping Fundamentals in AI Era
24:25 Socratic Seminars for Code
26:27 Voice-Based Code Thinking
29:27 Learning Beyond Lockdown
36:10 Prompting as a New Skill
36:25 Hardware Troubles and NeoPixels
40:15 Beyond the Code Editor
45:01 New Research on Embodied AI
49:12 PyCon and Community Plans
50:42 Teacher Call to Action

Special Guest: Philip Guo.