Episode Summary
Executive Summary: Dan Shipper argues AI is not mainly destroying jobs but reshaping work toward higher leverage, faster skill acquisition, and more “manager-like” behavior. He describes Avery as an AI-native media/software company where a small team uses agents, prompts, and internal AI ops to ship products, improve editorial quality, and train clients. His core thesis: the winners will be people and companies that learn to delegate to AI well.
Main Topics: AI as job reshoring and demand expansion (Priority: 5/5): Shipper’s hottest take is that cheap intelligence will reshore work to the U.S. by making formerly expensive services affordable to smaller firms and individuals, increasing demand and enabling fewer people to serve more customers. AI-native workflows and agentic tools for non-coders (Priority: 5/5): He argues tools like Claude Code and Gemini CLI are underappreciated for non-programmers because they can autonomously process files, work for long periods, and reduce the need for manual interface-driven prompting. Avery’s AI-first operating model (Priority: 5/5): Avery runs as a 15-person company with an AI operations lead, internal prompt libraries, multi-agent workflows, and zero handwritten code on the product team, using AI to scale editorial, engineering, and product work. How AI changes talent development and management (Priority: 5/5): Shipper says the scarce skill in the AI era is managing models and agents: defining tasks, giving feedback, evaluating outputs, and knowing when to intervene. He believes young workers with ChatGPT can accelerate far faster than past generations. Product strategy: incubate expensive services into software (Priority: 4/5): Avery builds products by identifying costly human services—email chief of staff, ghostwriting, file cleanup, content automation—and turning their useful parts into AI tools that can be bundled and launched quickly. AI adoption in companies and consulting (Priority: 4/5): Avery’s consulting arm helps organizations become AI-first through audits, training, and workflow design. Shipper says the biggest predictor of success is whether the CEO uses AI daily and visibly champions it. Funding philosophy and staying creative (Priority: 3/5): He prefers small, flexible financing ('sip seed') to preserve optionality and maintain Avery as a creative playground while still building a serious institution with lasting impact.
Key Arguments: AI can increase, not decrease, demand for skilled labor by making expensive services cheap enough for smaller businesses and individuals to buy. Non-coders can gain major leverage from agentic tools like Claude Code because these tools can read local files, run for long periods, and complete multi-step work autonomously. Avery’s operating model demonstrates that a very small team can ship multiple products when the team is AI-first, multidisciplinary, and supported by a dedicated AI operations lead. Young workers using AI can learn faster than prior cohorts because they can record feedback, encode it into prompts, and avoid repeating mistakes. The key AI-era skill is management: framing tasks, selecting models, providing context, judging outputs, and iterating effectively. Companies that successfully adopt AI usually have CEOs who are personally active users; top-down enthusiasm and visible usage create momentum and realistic expectations. AI products should be built by first solving real internal pain points, then unbundling successful workflows into standalone apps. Avery’s bundle-and-incubate model works because its audience and internal team share a similar AI-forward taste, making them natural first users. Small, flexible capital raises can better fit AI-era experimentation than large venture rounds, because product leverage is much higher and capital needs are lower. Using AI may reduce some traditional skills, but historical precedents like writing replacing memory show that technology often trades one capability for a much larger gain.
Data Points: Team size: 15 people - Avery operates a newsletter, consulting arm, and multiple products with a small team. Subscriber count: About 100,000 subscribers - Avery’s daily newsletter reaches a large AI-focused audience. Products shipped: 4 products - The company has built and shipped multiple apps alongside its newsletter. Public beta users: 2,500 active users - Quora, Avery’s email AI product, had this many active users at launch. Email volume: Millions of emails - Shipper said Quora processes millions of emails through the product. Consulting revenue last year: About $1 million - He said the consulting arm did roughly this amount in the prior year. Consulting growth expectation: Likely to double this year - Shipper expects the consulting business to grow substantially. Quora build cost: About $300K all-in - He estimated the cost to build Quora, including labor. Pre-seed round: $700K - Avery raised a small initial round to preserve optionality. SIP seed committed capital: Up to $2 million - Avery’s recent flexible financing arrangement with Reed Hoffman and Starting Line VC. AI operations cadence: Once a week - Shipper meets weekly with the head of AI operations to identify repetitive work and automate it. Time horizon for agent work: 20–30 minutes - He uses this as a rough current benchmark for how long models can work autonomously. Model comparison scope: 15 Claude Code instances - He described the Quora team as two people plus many AI agents. Management workforce share: 8% of the workforce - Shipper referenced this as the current proportion of managers, arguing management will become more common.
Pivotal Quotes: "I think AI may be one of the biggest force for reshoring American jobs." — Dan Shipper: His hot take on AI’s macroeconomic effect was that it will stimulate domestic demand and make U.S.-based work more competitive. "I think people are truly sleeping on how good Claude code is for non coders." — Dan Shipper: He used this to argue that agentic CLI tools are underused by non-technical people who could benefit from them most. "I hate the headlines that are like entry-level jobs are taken away by AI." — Dan Shipper: He pushed back on fear-based narratives and argued that young workers using AI can accelerate faster than previous generations.
Implications: Listeners should think less about AI replacing all work and more about learning to manage agents, build AI-first workflows, and compound leverage. Companies that adopt early may ship more with smaller teams; individuals who learn AI deeply may advance faster and gain a strong competitive edge.
About Lenny's Podcast
Lenny Rachitsky interviews world-class product leaders and growth experts about building products and growing careers.