Adventures in AI

Adventures in AI

Download on the App Store

Adventures in AI episodes

  • Deep dive: Claude Design

    Host Ollie is joined by Design Lead Hugo to take a deep-dive into Claude Design and review how AI is changing the world of Design.

    Claude Design was in early preview at time of recording and is now available to all Claude users. Hugo Campanella, Rightbrain's lead designer, has been testing it since day one — and this episode is the map of where it earns its keep in the software development lifecycle, and where it doesn't.

    We walk the five SDLC stages — ideation, discovery, build, launch, delivery — and place Claude Design, Figma, Cursor, Lovable and FigJam against each one. Short version: Claude Design wins at ideation and early visual direction. Cursor still wins at low-fidelity wireframing — Hugo's method is ASCII wireframes grounded in the actual codebase, back in under a minute. Figma still does the production work. None of them replace each other yet.

    Also in this one: why Figma's stock fell 7% on Claude Design's launch day, and the first entry in a new segment — Top tips for Claude users.

    46 min
  • The map of the AI

    You chat with AI every day. But is there more to it? This episode is the map.

    Ollie left the clever stuff for later and went back to basics: what Claude actually is, how it's built, and the one thing that turns a rubbish answer into a great one. No jargon, nothing assumed — the ground floor for getting properly good with AI, whether you're on Claude, Copilot, Gemini or ChatGPT.

    In this episode:

    • Where work happens has shifted. Excel → Jira → Notion → Claude. The tool you reach for first is changing for every role — analyst, recruiter, finance lead.
    • The three places to work in Claude: Chat to ask, Cowork to do, Code to build — what each is good at, where each falls down, and the path most people travel between them.
    • How an AI interface is actually put together — prompts, custom instructions, skills and threads, explained plainly.
    • The big idea: the best AI is the smallest AI. Targeted context beats more context every time. Ask a librarian for "a good book" and who knows what you'll get; ask for the new Dan Brown and you're sorted. AI's no different.
    • The four levels, and how to set them up — raw chat → custom instructions → skills → your context. By level two you've more than doubled the value of every answer. By level four it feels like an assistant with ten years' experience and a perfect memory.

    Next time: The Temple of Claude — a love story. How Ollie fell for it, and the times they nearly broke up.

    Useful links:

    • Watch this episode on YouTube
    • More from Rightbrain: rightbrain.ai
    • Follow Ollie on LinkedIn

    Subscribe so you don't miss the next one — we go from the map to the adventure.

    34 min
  • Raiders of the automation

    In this episode, Ollie shares his journey from ChatGPT skeptic to product lead and AI podcast host at Rightbrain. He covers the tools that didn't work, the ones that did, why context is everything, and the exact moment AI stopped being a novelty and started being indispensable. Plus: what Rightbrain actually does , why he joined the team and what to expect from the podcast.

    14 min

About Adventures in AI

From the publisher's feed

Adventures in AI is a podcast from the team at Rightbrain. Follow Ollie Happy's journey learning, breaking, and building with AI tools. The goal: help you become more efficient, more organised, and…