Spinout Podcast.
Strategy, AI, and e-commerce. Short, concrete episodes about what actually works, what doesn't, and why.
Career and Recruitment in the AI Era
The U-curve is reshaping who gets hired. A go-kart versus a GT3 race car shows what leverage actually feels like. Two workshops that changed how companies think about talent and career paths.
The Market Is Getting AI Wrong
Stock markets are selling off software companies as if AI makes them obsolete. Meanwhile, personal AI agents are getting the keys to everything. Both stories reveal the same blind spot.
The Bottleneck Has Moved
Your IT processes are optimized for a scarcity that no longer exists. And AI is splitting the market into two races: drag racing and Le Mans. The middle ground is disappearing.
AI Coding: The Reality Beyond the Hype
Disposable software, intelligent spreadsheets, and what AI coding actually looks like for senior engineers. Three perspectives on the same shift: code is getting cheaper, but judgment is getting more expensive.
The Protocol War for Agentic Commerce
The click that proves a human intended to buy something is disappearing. Six layers of infrastructure are fighting over what replaces it. OpenAI and Stripe versus Shopify and Google. Visa versus stablecoins. Your company is caught in the middle.
The Six-Month Window Your Board Needs to Understand
95 percent of enterprise AI initiatives deliver zero ROI. Not because the technology fails, but because companies start with tools instead of processes. Here is what your board needs to decide before Q3.
The bitter lesson: why your AI systems need to get simpler, not smarter
Claude Mythos is a step change, not an upgrade. It reveals the same truth as every major model leap: simpler systems beat complex ones. We walk through four things to audit before the train leaves.
Your E-Commerce Ranks on Google. AI Agents Can't Find You.
Search engines and AI agents discover products in completely different ways. While your SEO is optimised for Google, the AI agents your customers are already using — to ask "which bike tyre should I buy?" — rely on MCP, UCP, and AEO to find and recommend products. This episode explains the three abbreviations that will determine who gets recommended and who gets skipped.
From Vibe Coder to Agent Manager
We see a pattern repeating itself. You build your first AI-generated app, everything works brilliantly, and then it goes off the rails. In this episode we break down what separates those who succeed with AI agents from those who get stuck in loops, and why it is a management challenge, not a technical one.
Why Your AI App Breaks on Thursday
There is a pattern among almost every founder and product manager building their first AI-generated app right now. It starts brilliantly. Then Thursday arrives. Vibe coding gets you 90% of the way — this episode breaks down what happens next.
Claude or ChatGPT? Wrong question.
Most organisations ask the wrong question when evaluating AI tools. Claude and ChatGPT are not variants of the same product — they are optimised for fundamentally different behaviours. It matters.
Why OpenAI Is Spending $600 Billion — and Why Google Is Giving Away Its Best Model
OpenAI is spending $600 billion — not to build a better chatbot, but to own the synthesis layer above your entire organisation. In this episode we break down what that actually means, why Google is giving away its most capable model for free, and how Anthropic has built a massive lead through Claude Code. There is a race underway for your organisation's accumulated understanding. Most people are looking the wrong way.
Everyone Talks AI. Get Your Own Experience.
At D-Congress in Gothenburg, everyone was talking AI. But collective understanding is thinner than ever. The only way to actually get it? Build your own experience, fast.
Your AI memory is trapped in the wrong system
Your AI memory is fragmented and trapped in individual tools. In this episode, we talk about Open Brain – a system we built with Supabase and pgvector to give Claude Code a semantic, persistent memory. Thoughts are sent in via Slack or directly from Claude, stored with vector embeddings and can be searched semantically. A practical walkthrough of how to actually solve AI amnesia.
Lights Out: When the Factory Runs Without People
In manufacturing, 'lights out' means the factory runs 24/7 with no one there. That concept is now moving into e-commerce. In this episode, we explore how AI agents can take over administrative processes like invoicing, procurement, product information, A/B testing and returns — at a fraction of today's cost.
Should You Start Slow and Build Speed?
Most people want to take it easy with AI. Wait until the technology matures. Let others test first. It feels sensible. But it's wrong. When the world changes fast, caution becomes the riskiest strategy. The paradox: trying to learn slowly and safely is like cycling slowly to avoid falling—but cycling doesn't work that way.
All Change Starts at the Top
What happens when leaders don't fully see or understand the AI revolution unfolding around them? In this episode, we explore why all change starts at the top—and what boards and leadership teams need to grasp right now.
The Hidden Costs of Letting AI Write All Your Code
AI code generation is accelerating development—but at what cost? In this episode, we explore the hidden risks: security gaps, technical debt, and the shifting role of senior engineers in an AI-driven world.
Time to Shift Into the Next Gear
Most people using AI today are sitting in a Ferrari but only using it to buy milk. In this episode, we explore the levels of AI automation—from simple chat to orchestrated agent swarms—and why this shift isn't a feature, it's a transformation.
AI Bootcamp
14 episodes from zero to published app. Each episode takes one tool or concept and explains it so you can use it the same day.
Marketeer Bootcamp
8 episodes for marketers who want to use AI for real. From setup and tools to images, video, and publishing.
