Lenny’s Podcast · Podcast episode · Building AI products · Execution, roadmaps & process

The AI-native startup: 5 products, 7-figure revenue, 100% AI-written code

Lenny Rachitskywith Dan ShipperJul 17, 2025128 min▶ 90K
SourceLenny’s Podcast
KindPodcast episode
PublishedJul 17, 2025
Readers▶ 90K
Originalyoutube.com ↗
N:

Dan Shipper, co-founder and CEO of Every, a 15-person company running a daily newsletter, several AI-built products, and an AI consulting arm, discusses how his team operates with AI agents and no hand-written code. The conversation covers his contrarian views on AI and jobs, tools like Claude Code for non-coders, practical team tactics such as an AI operations role, and his thoughts on management skills becoming central in an 'allocation economy.'

01Key learnings

  • Non-programmers can use Claude Code-style command-line agents with local files to process large text sets, like meeting notes or books, by delegating multi-step tasks rather than pasting content into chat.
  • Hire or assign someone to an AI operations role who continuously turns repetitive team tasks into prompts and workflows, and make sure people actually adopt them.
  • Apply 'compounding engineering': spend a bit of effort turning each recurring task (like writing PRDs) into a reusable prompt or command so every later unit of work gets easier.
  • Capture an expert's taste by codifying their feedback into prompts, so writers get early feedback that approximates the editor before they ever need a review.
  • Adoption of AI in a company is best predicted by whether the CEO uses it daily; leaders should share prompts, weekly usage stats, and reward early adopters to spread learnings.
  • Treat working with AI as managing: knowing when to delegate, when to micromanage, how to give feedback, and how to define vision and quality criteria are increasingly core skills.
  • Build small, cheap products by unbundling expensive services (like an assistant or lawyer) that AI makes affordable, starting with general-purpose tools and then spinning out apps once they prove useful internally.
  • Consider fundraising in a lean way (a small seed or flexible 'sip seed' round) so that AI-enabled products can be built with very little capital while keeping optionality.
“whenever I see a kid with ChatGPT, I'm like, "Holy shit, they're going to go so much faster than any other person that I've worked…”Dan Shipper · Lenny’s Podcast · 00:00:20
“the number one predictor is, "Does the CEO use ChatGPT?," or insert your own chatbot.”Dan Shipper · Lenny’s Podcast · 01:12:00
“I think a good definition of AGI is when does it become economically profitable for people to run agents indefinitely?”Dan Shipper · Lenny’s Podcast · 00:17:41

02Frameworks mentioned

Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.