In episode 4, Nathan and Tony are at it again. Why do so many AI models exist in each of these LLMs, and why should I care? Don't worry, you're not the only one asking. Step into the 'Model Locker' and try some on. Did Nathan and Tony just compare AI models to the shoes in your closet. Yep, you have to hear this.

Takeaways

• Choose the right AI model for the right task
• Avoid unnecessary upgrades and token burn
• Don't fall for model FOMO

Chapters

• 00:00 Introduction and LLM Definition
• 01:15 Organic Conversations and Model Analogy
• 03:41 AI Models and the 'Model Locker'
• 06:27 Token Burn and FOMO
• 08:49 Choosing the Right Model
• 10:04 AGI and Hype Cycle

Prompting an AI is basically bossing around a very literal six-year-old — and Tony proves it by playing one, one dirty dish at a time. Nathan asks him to "wash the dishes," and chaos (and grandma's heritage china) ensues.

Tony and Nathan break down why AI needs short, specific instructions with clear end states — and what to do when it doesn't know what to do next. If grandma understood a word of that, they did their job — though frankly, grandma's been giving flawless one-step, no-ambiguity prompts since 1962. She never once had to ask twice about washing the dishes... properly.

Governance sounds like a corporate buzzword — until you think of it as a healthy habit. Tony and Nathan use a burning toaster, a Waymo, and a knitting hallucination to explain how to use AI without setting the house on fire. If Grandma understood what they said, they did their job, but let's be honest, Grandma already has a governance framework, it's that look. And somehow it's more effective than most corporate compliance training.

Referenced in this episode: Electric Toaster in Downton Abbey https://youtu.be/F_o0_Qel9QU?si=vk4FFrcSJD71Om-e

Two friends bootstrapped a podcast to demystify AI for real people. In this pilot episode, Tony and Nathan walk through their first recording setup (Zencaster + Otter), compare LLMs with VHS-era analogies, share quirky origin stories (including owls and knitting), and show how everyday tools like you.com and Claude can make AI practical and approachable — all with humour, chemistry, and the goal that even Grandma could follow along.