Applied
Wizardry
The engineering behind the magic
A podcast for engineers and technical founders about how we build with AI.
"When art critics get together they talk about form and structure and meaning. When artists get together they talk about where you can buy cheap turpentine."
We talk about the turpentine tokens.
Applied Wizardry is conversations with the founders and engineers building with AI: what they use, what works, what breaks, and how it changes their teams and businesses. It's for people who already build with AI and want to do it better.
For software engineers, engineering leaders and technical founders who build with AI every day.
Guests are practitioners, and we talk specifics: tools, costs and results. Claims are tested against real experience. Tools and platforms are covered on their merits, and any partners are disclosed.
Season 1
Twelve episodes
These are working themes; episode titles are still to come.
- 01
Cut your token burn
Where AI spend goes, and how teams bring it down without slowing down. With Dave Slutzkin.
- 02
Coding harnesses roundup
Claude Code, Codex, Cursor and the rest: what each is good at, and how builders choose.
- 03
Is code review dead?
When agents write most of the code, what review is for and who does it.
- 04
WTF is a software factory?
What the term means, what it looks like in practice, and where the humans fit.
- 05
Domain-specific harnesses
Building products and businesses around a harness made for one industry or job.
- 06
The agent-native stack
The infrastructure and platforms that agent-built software runs on.
- 07
Do programming languages matter any more?
Whether language choice still matters when agents write the code.
- 08
Why is eval important?
Measuring whether AI output is good enough, and how that differs from testing.
- 09
Cutting cyber risk
The security risks that come with agents, and how teams reduce them.
- 10
The MCPs that matter
Which MCP servers and integrations are worth using, and which aren't.
- 11
Managing your own context
How builders manage their own attention while running many agents at once.
- 12
How the sharpest startups are built to build
Team structure, roles and hiring at startups built around AI.
What we cover
Six pillars
Frontier
New models and tools, and what they change for builders.
Factory Floor
How software gets made now: workflows, tools and agents.
Trust
Evals, code review, security and quality.
Org & Economics
Team structure, hiring, and the cost of building.
Arbitrage
Opportunities that exist now and may not last.
Platforms
Infrastructure and platforms for building with AI.
Your host
Yaniv Bernstein
Engineer · founder · podcaster · Sydney
- PhD
- Computer science
- 10 yrs
- Engineering at Google
- VP Eng
- and COO at Airtasker
- CTO
- and co-founder of Vera
- 300+
- episodes as co-host of The Startup Podcast
Multi-time founder and engineering leader. PhD in computer science, ten years of engineering at Google, VP Engineering and COO at Airtasker, and CTO and co-founder of Vera.
Yaniv builds his own startup hands-on with AI coding tools, on Cloudflare Workers, and co-hosted more than 300 episodes of The Startup Podcast.
Founding sponsor
Reach the people building with AI
Founding sponsorship for Season 1 is open. Sponsor reads are disclosed, and episode content stays independent.