Loading video player...
Can AI make game dev truly “prompt to play”? Jack Wakem (Founder, Tempest AI) breaks down how his team is building “Lovable for games” — a consumer platform where anyone can prompt, iterate, and ship, including a timeline that lets you scrub your build like a YouTube video. Tempest started as an AI-native RPG engine (RAG, context engineering, custom schemas) and evolved into a fast, consumer game-creation workflow. We dig into the hard parts (state, memory, assets), product pivots, and what Jack learned dropping out, working night shifts, and building in public with Build Club (Annie Liao). What you’ll learn Why “AI-native games” are hard (context tracking, long-running state) and the pragmatic path that worked. How Tempest’s timeline lets creators jump to any checkpoint, edit, and keep playing, slashing iteration time. Where AI fits today in game development vs. live asset generation (and what’s still rough). “Lovable for games”: lowering the barrier so non-devs can prompt changes directly in-game. Market context: gaming + AI is ~USD $1.5–2B today with projections to “tens of billions” by 2029. Founder mindset: give yourself a focused year, keep lights on with a job, compress learning through reps. Timestamps: 00:00 – Gaming is huge; meet Jack (Tempest AI) 03:02 – Farm kid to game dev: early life 04:19 – First console memories (Nintendo DS, strict screen time) 07:00 – Designing engaging games; making content to learn 08:05 – Indie projects, clients, and the post-school fork in the road 12:00 – De-risking vs passion; finance path vs gaming; GPT-3 moment 16:03 – Dropping out; night jobs while building Tempest 17:26 – Founder psychology: knowledge, ego, experience 20:20 – Community fuel: Annie Liao & Build Club 24:00 – Tempest focus + the gaming-AI market snapshot 25:18 – Vision: AI-native “never-ending” games (and the hard bits) 28:04 – Context engineering; gen-AI in the dev pipeline 30:00 – What Tempest is: “Lovable for games” (consumer prompt UX) 31:18 – The slow iteration problem in game dev 32:07 – Coding agent + timeline scrubber (edit your game like a video) 33:28 – Building with users: sit-with testing to dashboards 35:10 – Will AI reduce or amplify creativity? 38:05 – The next 5 years: more personal, more experimental games 39:31 – Fundraising: where the round is at 40:08 – How they raised: two years of public building & updates 42:00 – Wrap and where to follow Tempest Guest Jack Wakem — Founder, Tempest AI. Rural NSW origin story → Uni (Finance + CS) → drops out, works nights (pest control/industrial cleaning) while building Tempest. About Tempest AI Started by prototyping endless, AI-driven RPGs with custom context schemas and an in-house engine to give LLMs reliable world state. Pivoted ~3 months ago into a consumer game-creation platform with a friction-free UX and a unique timeline editor. Tempest website: https://alpha.tempestengine.ai/ Jack Wakem: https://www.linkedin.com/in/jack-wakem-ab7170230/ Quote to remember “You only lose when you quit… keep working at one thing long enough and compounding kicks in.” Join the conversation What’s the first workflow you’d hand to an AI co-dev, level layout, quest logic, balancing, or playtesting feedback? If you enjoyed this episode Subscribe for weekly founder-led AI product stories Drop a comment with your favourite takeaway Share it with a teammate who’s building with LLMs #AIGameDev #TempestAI #GenerativeAI #GameDevelopment #IndieDev #BuildClub #DigitalNexusPodcast