AI PM — Source Registry
Master index of all source material. Sortable by status, type, author, and date.
Source Status Key
| Status | Meaning |
|---|---|
unread |
Captured, not yet read |
reading |
Currently working through |
read |
Finished, notes added |
processed |
Ideas extracted into knowledge entries |
All Sources
By Status
Unread
- 2023-02-05 Marily Nika AI and Product Management — Marily Nika (AI Product Leader, Meta Reality Labs; formerly Google 8 years; PhD in Machine Learning; Executive Fellow Ha
- 2024-07-09 How Close AI Replacing Product Managers — Mike Taylor (professional prompt engineer, co-author O’Reilly’s Prompt Engineering for Generative AI) guest post on Lenn
- 2024-08-13 Summary AI Product Management Marily Nika — Human-edited written summary (by Gaurav Chandrashekar) of the Feb 2023 Marily Nika podcast episode. Detailed breakdown o
- 2024-09-08 LinkedIn Transformation Tomer Cohen — Tomer Cohen (CPO LinkedIn) on Lenny’s Podcast. The inside story of how LinkedIn transformed its feed into an engaging co
- 2025-01-07 Guide AI Prototyping Product Managers — Colin Matthews (PM @ Datavant, AI prototyping course instructor) guest post on Lenny’s Newsletter. Comprehensive guide t
- 2025-04-22 Sahil Lavingia Gumroad AI Native CEO — Sahil Lavingia (CEO, Gumroad) on How I AI. Demonstrates exact workflow for 40x productivity increase
- 2025-04-28 Farhad Manjoo ChatGPT Writing Workflow — Farhad Manjoo (fmr NYT columnist) on How I AI. Three AI-enhanced writing workflows
- 2025-05-05 Cody De Arkland Vibe Coding 3D Game — Cody De Arkland (Sr. Dir DevEx, Sentry) on How I AI. Two workflows demonstrated
- 2025-05-12 Joel Unger Designer Guide Cursor Prototyping — Joel Unger (Design Director, Atlassian) on How I AI. Three AI workflows for designers
- 2025-05-13 State Product Job Market 2025 — Lenny Rachitsky’s annual PM job market analysis using TrueUp dataset. Over 6,000 open PM roles globally
- 2025-05-16 Vibe Check Codex OpenAI Coding Agent — OpenAI incorporated the taste of a senior software engineer into how Codex writes code. Built for senior engineers. Comp
- 2025-06-02 Hiten Shah ChatGPT Workplace Tool — Hiten Shah (serial founder, Crazy Egg, KISSmetrics) on How I AI. Three advanced ChatGPT Project workflows
- 2025-06-05 Anthropic CPO Mike Krieger — Mike Krieger (CPO Anthropic, co-founder Instagram) on Lenny’s Podcast. 90-95% of code for some Anthropic products now wr
- 2025-06-09 Luke Harries ElevenLabs AI Playbook — Luke Harries (Head of Growth, ElevenLabs) on How I AI. Three workflows for AI-native marketing
- 2025-06-10 Get Entire Team Prototyping AI — Colin Matthews (guest post on Lenny’s Newsletter) presents a comprehensive guide to scaling AI prototyping across produc
- 2025-06-11 Every Two Engineers Ship Like Team of 15 — Two Every engineers shipped 6 features, 5 bug fixes, 3 infrastructure updates in one week by designing AI agent workflow
- 2025-06-16 Dennis Yang Cursor PM Workflow — Dennis Yang (Principal PM for Generative AI, Chime) on How I AI. Two workflows using Cursor as a central PM hub
- 2025-06-23 John Blackman 91yo Vibe Coding Replit — John Blackman (91yo retired electrical engineer) on How I AI. Two-part workflow to build a complex church event manageme
- 2025-06-30 Elizabeth Lin Designer Cursor Prototypes — Elizabeth Lin (independent design educator, Design is a Party) on How I AI. Three workflows showing designers how to use
- 2025-07-06 Base44 Bootstrapped Startup Maor Shlomo — Maor Shlomo (founder, Base44) on Lenny’s Podcast. Solo founder bootstrapped AI app builder to $1M ARR in 3 weeks, 400K+
- 2025-07-07 Priya Badger Yelp AI Product Design — Priya Badger (PM, Yelp) presents three workflows for designing conversational AI products
- 2025-07-08 What People Are Vibe Coding — Lenny Rachitsky crowdsourced 1,000+ replies on what people are vibe coding that they actually use regularly. 50+ example
- 2025-07-14 Colin Matthews High Fidelity AI Prototypes — Colin Matthews (PM, founder, Maven instructor) on How I AI. Three workflows for on-brand AI prototyping
- 2025-07-28 Terry Lin Vibe Coding Fitness App Cursor — Terry Lin (PM, vibe coder) on How I AI. Two workflows for building Cooper’s Corner, a voice-powered fitness app
- 2025-07-31 Vibe Check Claude New Agents Confusing — Each agent has separate conversation history and memory. Can’t see what others are doing unless you explicitly pass inf
- 2025-08-04 Hamel Husain Debugging AI Writing Evals — AI consultant Hamel Husain presents a 6-step systematic error analysis framework for debugging AI products
- 2025-08-07 GPT-5 Early Review Coding Product Managers — Early hands-on GPT-5 review. In Cursor, feels like working with a productive staff engineer
- 2025-08-18 Zach Leach Gamma AI Feedback Synthesis — Zach Leach (Head of Design, Gamma) on How I AI. Three workflows for a lean 30-person team serving a 60% international us
- 2025-08-25 Tomasz Tunguz Parakeet Podcast Processor — Tomasz Tunguz (Theory Ventures) presents two main workflows including ‘Parakeet Podcast Processor’
- 2025-08-28 Claude Code Camp Workflows — Subagent patterns for Claude Code: on-the-fly agent creation, log investigator agents, when to use subagents vs. direct
- 2025-09-01 Anjan Panneer Selvam Meeting Notes Prototypes — Anjan Panneer Selvam (CPTO, Acolyte Health) on How I AI. Three workflows rethinking B2B product management through trans
- 2025-09-02 I Started Talking to My Computer Instead of Typing. It Changed How I Think. — Katie Parrott: switching to voice dictation (Monologue + ChatGPT projects) removes friction between thought and output; shifts focus from mechanics to substance; shadow side: frictionless workflow creates entrainment cycle and work bleeds into weekends
- 2025-09-05 Silicon Valley’s Top Coaches Want You to Stop Fearing AI — Three Silicon Valley coaches: create a ‘codex vitae’ for personalized AI coaching sessions; weekly 45-min AI coaching; Jungian dreamwork via ChatGPT; fear of AI replacement is grief from identity collapse; ‘centaur model’ — client + human coach + AI working together
- 2025-09-07 AI Reshaping Product Role Udezue — Oji and Ezinne Udezue (50+ years combined PM leadership at Microsoft, Twitter, Atlassian, Calendly, Typeform) on Lenny’s
- 2025-09-08 How I Run Three AI Models in Parallel Without Losing My Mind — Katie Parrott coins ‘model manager’ role and ‘allocation economy’: ChatGPT extracts quotes, Monologue handles voice revisions, Claude analyzes style — cycling between tasks; writing time down 40%, weekly output tripled; four boundaries: compartmentalize, leave breadcrumbs, limit to 3 simultaneous, capture immediately
- 2025-09-08 Prerna Kaul FDA Submissions Claude Streamlit — Prerna Kaul (product/platform leader, Amazon Alexa to Moderna to Panasonic Well) on How I AI. Two workflows
- 2025-09-09 Why 95 Percent of AI Pilots Fail—And How to Avoid It — Marc Malott: MIT study 95% of corporate AI pilots fail; the failure mode is success — when AI unlocks 40,000 human hours, leadership raises targets and prices to capture ROI immediately, destroying slack for innovation and paradoxically impeding further progress
- 2025-09-10 How to Use Claude Code as a Second Brain — Noah Brier (Alephic) uses Claude Code as thinking partner by pointing it at his 1,500-note Obsidian vault; frame as ‘thinking mode’ not ‘writing mode’; instruct AI to ask clarifying questions before generating; mobile via Termius + Tailscale
- 2025-09-11 Build Places, Not Products — Lucas Crespo: AI generates median statistical interfaces accelerating toward ‘a single, hyper-optimized, bloodless template’; key distinction: design (solves problems) vs. art direction (shapes atmosphere, visual north star); Cora email assistant chose oil-painted sky over sterile dashboard
- 2025-09-15 Zach Davis LaunchDarkly Enterprise AI Coding — Zach Davis (Director of Engineering, LaunchDarkly) on How I AI. Three workflows for weaving AI into enterprise engineeri
- 2025-09-17 He Got Thousands of Users Before His AI App Even Launched — Naveen Naidu (Monologue GM): never build in darkness (Discord from day one); single power user as north star; dogfooding as feature development; fast-ship muscle: ship four products, expect most to fail; technical stack: Windsurf + Claude Code + Codex; 1M words transcribed weekly
- 2025-09-18 Launch Day Lies—Day Two Tells the Truth — Naveen Naidu: real validation appears on day two when novelty fades; five rules from four rapid-iteration products; Monologue breakout: internal user used it 100x/day by day two; ‘when everything is easy to make, nothing is easy to love’
- 2025-09-22 Amir Custom GPT Workflows — Amir Klein (AI PM, monday.com) on How I AI. Three workflows for offloading PM m
- 2025-09-22 I Fed My Essays to ChatGPT Until It Learned My Voice — Katie Parrott builds a ‘writing brain’ — customized ChatGPT project trained on published essays with four components: overview, core directives (signature structure), red flags (writing weaknesses), performance playbooks; iterative AI as sparring partner across all writing stages
- 2025-09-23 You’re Probably Using AI Wrong — Rhea Purohit: efficiency framing produces flat output; reframe — use AI for meaning-making not productivity; Eiger analogy: humans should pursue activities aligned with personal meaning; use Claude as brainstorming partner (paste drafts with blanks, request multiple options, select yourself)
- 2025-09-24 Cognition’s CEO on What Comes After Code — Scott Wu (CEO, Cognition/Devin): CS fundamentals still matter — you’re ‘at the helm making decisions’; vision: programming by description; AI-first workflow: map plans conversationally before any code; Devin vs. Claude Code distinction; AGI has already arrived by decade-old benchmarks
- 2025-09-25 I’ve Stopped Writing Prompts—DSPy Does It Better — Mike Taylor (co-author O’Reilly prompt engineering book): DSPy automated optimization supersedes manual prompt writing for most use cases; ‘severely underhyped’ (Shopify CEO); democratizes prompt quality, making expertise less individually dependent
- 2025-09-29 Ravi Mehta Data Driven Prototyping Midjourney — Ravi Mehta (fmr CPO, Tinder) presents two workflows for data-driven prototyping and structured Midjourney prompts
- 2025-09-29 Vibe Check Claude Sonnet 4.5 — Sonnet 4.5 collapsed a two-chat writing/editing workflow into one model. Retains natural voice from dictation instead of
- 2025-09-30 How to Make AI Write Less Like AI — Chris Silvestri distinguishes prompt engineering (narrow instructions) from context engineering (rich foundational materials: customer interviews, brand voice guides, competitive analysis); context creates a ‘data-informed world’ for AI; scaling production without scaling human judgment is a recipe for disaster
- 2025-10-06 Andrew Mason Nabil Hyatt AI Cofounder — Andrew Mason (founder of Groupon, CEO of Descript) and Nabil Hyatt (partner, Spark Capital) on How I AI. Used AI as a th
- 2025-10-08 How Box Is Building an AI-first Company—Without Cutting Jobs — Aaron Levie (CEO, Box): AI expands work via Jevons Paradox; AI automates ‘tasks’ not ‘jobs’; ICs increasingly resemble managers directing agents; three enterprise adoption lessons: frame as growth accelerant, institutionalize learning via public demos, foster first-principles thinking about why processes take current form
- 2025-10-09 How to Use Claude Code for Everyday Tasks—No Programming Required — Katie Parrott: ‘apartment vs hotel’ metaphor — cloud AI is a hotel (fresh start), Claude Code is your apartment (persistent, customized); best for: files exceeding cloud limits, extended processing time, repeated workflow automation; use cases: expense tracking, content analysis, customer support via GitHub
- 2025-10-10 Seeing Science Like a Language Model — Dan Shipper: reductionist science has reached its limits — the replication crisis shows human behavior is nonlinear and contextual; LLMs reveal that valid understanding can emerge through pattern recognition rather than explicit rules — a challenge to reductionism itself
- 2025-10-13 Tim McAleer Ken Burns Documentary AI — Tim McAleer (producer, Ken Burns’ Florentine Films) on How I AI. Three custom AI tools for documentary post-production
- 2025-10-17 Seeing Business Like a Language Model — Dan Shipper: business ‘principles’ are context-dependent patterns that fail unpredictably, not universal laws — five reasons including equifinality, impossible experimentation, sensitive dependence; path forward: develop gut-level pattern recognition (how LLMs work) rather than memorizing rules
- 2025-10-20 Michal Peled HoneyBook LinkedIn Recruiting NotebookLM — Michal Peled (Technical Operations Engineer, HoneyBook) on How I AI. Three ‘little helper’ workflows
- 2025-10-22 He Built an AI Ghostwriter With Taste — Danny Aziz (GM of Spiral): design philosophy — deliberately slows writing by asking clarifying questions before drafting; multi-agent architecture (interviewer + writer in parallel); presents three directional options rather than one ‘best’ answer; craft is about judgment, not authorship
- 2025-10-24 Seeing Creativity Like a Language Model — Dan Shipper: AI shifts creative work from knowledge economy to ‘allocation economy’ — value flows to those who direct intelligence (vision, taste, quality control) rather than execute; AI amplifies creativity via finding inspiration, articulating taste, and producing more
- 2025-10-27 Lucas Werthein AI Wellness Coach — Lucas Werthein (Head of Technology, Cactus) on How I AI. Three workf
- 2025-10-28 How My Company Turned a Data Crisis Into an AI-native Rebirth — Stella Garber (CEO of Hoop): pivoted after Slack killed data access; rebuilt as AI-native — prototype, website, and ad campaign in one week using Claude Code plus 10 parallel agents; key insight: senior teams face greatest challenge accepting AI outperforms humans; ‘teams stop doing the work and start directing it’
- 2025-10-29 How to Use Claude Code Like the People Who Built It — Cat Wu and Boris Cherny (Anthropic founding engineers): use plan mode for complex tasks (doubles/triples success rate); shared settings.json for team defaults; stop hooks to run test suites automatically; competitive subagent review — multiple specialized subagents challenge each other’s findings
- 2025-10-29 Vibe Check Cursor 2 Composer 1 Alpha — Cursor 2.0 agent view prioritizes delegating to and managing agents over reading/writing code. Claude Code terminal simp
- 2025-10-31 How Tools Shape How We See the World — Dan Shipper: tools shape cognition — each era conceptualizes the mind through its most advanced technology; traditional computers reinforced reductionism (binary logic); language models represent a paradigm shift — distributed neural networks reveal aspects of reality that defy reductionism
- 2025-10-31 Marco Casalaina Microsoft GitHub Spark Halloween — Marco Casalaina (VP Core AI, Microsoft) live-codes a Halloween fortune teller mobile app using GitHub Spark
- 2025-11-03 Tim Trueman Alexa Cerf Faire Vibe Analysis — Tim Trueman and Alexa Cerf from Faire’s data team demonstrate how MCPs are revolutionizing data analysis workflows
- 2025-11-03 Vibe Check Claude Skills Share Button — Skills work across Claude web/Code/API but lock you into Anthropic’s ecosystem. Gap between building a skill and deployi
- 2025-11-04 The Tool That Lets You Switch Models Without Losing Your Place — Katie Parrott profiles Droid (Factory): multi-model agentic coding tool that preserves conversational context via history compression; GPT for research/planning, Claude for implementation; multi-agent orchestration with specialized subagents; non-technical automation without coding
- 2025-11-05 What Jason Fried Learned From 26 Years of Building Great Products — Jason Fried (37signals CEO): ‘wholeness’ as core principle — Frank Lloyd Wright architecture metaphor; on AI: 37signals engineers rarely use AI for core coding to avoid losing creative skills; ‘the age of undifferent’ — software all looks like wireframes; only way not to be the same is to be yourself
- 2025-11-09 Good Start Claude Skills Research Strategies — Taught AI to analyze Figma designs and produce implementation plans, then it built 5 pixel-perfect screens autonomously.
- 2025-11-10 Claire Vo Beginners Guide Coding with AI — Claire Vo solo episode: complete zero-to-one walkthrough for non-coders. Four workflows
- 2025-11-10 Matt Britton Suzy Sales Calls GTM Engine — Matt Britton (founder/CEO, Suzy) on How I AI. One mega-workflow in Zapier
- 2025-11-12 He Built AI Agents to Launch a Million Businesses — Henrik Werdelin (Prehype/Audos): build around specific customers, not technology — Barkbox expanded to Bark Air because of relationship depth; three dimensions of relationship capital: depth (feeling seen), density (brand belonging), durability (permission to expand); emerging pattern: constellations of specialized agents serving identical customer segments
- 2025-11-13 AI Solved the Problem I Couldn’t Explain to Managers — Katie Parrott: AI tools (Cora, Monologue, ChatGPT) provided essential infrastructure for managing bipolar disorder at work — not by increasing productivity metrics, but by enabling consistent function; neurodiverse workers report significantly higher AI satisfaction; ‘technologies that change the very texture of work might be the most valuable AI applications we build’
- 2025-11-14 AI Ran Out of Internet. Now It’s Learning by Playing Games Again. — Alex Duffy: internet-trained AI models have exhausted public data and exhibit a ‘jagged frontier’; games solve this by providing synthetic training environments with clear success metrics; games are ideal because scenarios are controllable, outcomes scoreable, data proprietary and dynamic
- 2025-11-17 Claire Vo Thanksgiving Party Hub Lovable — Claire Vo walks through two personal AI workflows: vibe coding a custom Thanksgiving party hub using Lovable
- 2025-11-18 When AI Can Do Your Job, Who Else Are You? — Danny Aziz (GM of Spiral): identity crisis when AI automated his technical work; two responses: coasting vs. going deeper; he chose deeper — journaling revealed underlying interest in UX and product design; professional evolution through AI-enabled introspection: from ‘how does data move?’ to ‘how does this feature make me feel?’
- 2025-11-19 How Two Engineers Ship Like a Team of 15 With AI Agents — Kieran Klaassen and Nityesh Agarwal: shipped 6 features, 5 bug fixes, 3 infra updates in one week; meta-prompt using Anthropic’s Prompt Improver transforms rough ideas into detailed GitHub issues; mental model from Andy Grove’s High Output Management — catch issues in planning phase; Claude Code ranked top tier
- 2025-11-24 Claire Vo Clawdbot Moltbot Autonomous Agent — Claire Vo solo episode: 24-hour hands-on test of Clawdbot (open-source autonomous AI agent) named ‘Polly’. Three workflo
- 2025-11-25 Inside The Browser Company: Why They Killed Arc to Build Dia — Josh Miller and Hursh Agrawal (Browser Company): abandoned successful Arc browser to build Dia, an AI-native browser; pivot started as casual experiment that went viral via Arc Search; AI ‘memory’ feature nearly didn’t ship after 9 months of failure, then revived when ‘fundamental building blocks changed’; mission-first framing over profitable growth
- 2025-12-01 Claude Code: The Most Common Questions Beginners Ask — Nityesh Agarwal compiles 23 FAQs from 200 Claude Code beginners; Claude Code requires Pro or Max (not free tier); MCP optional extensions; bypass permissions mode for autonomous work; ‘apartment vs hotel’ metaphor (Claude Code = your own apartment with AI); trust and mindset matter as much as technical setup
- 2025-12-02 Think First, AI Second — Ines Lee (MIT research): sequencing matters — students who thought first then used AI maintained cognitive engagement comparable to AI-avoiders; two modes: passive AI use (outsourcing cognition) vs. active AI use (AI as challenge to reasoning); as critical thinking becomes scarce, its value rises
- 2025-12-03 Claire Vo Gemini3 vs Opus45 vs Codex51 Design Battle — Head-to-head model comparison for front-end design: same prompt across Gemini 3, Claude Opus 4.5, and Codex 5.1
- 2025-12-04 LinkedIn Replacing PMs Full Stack Builders — Tomer Cohen (CPO LinkedIn) on Lenny’s Podcast. LinkedIn scrapped traditional APM program, replaced with Associate Produc
- 2025-12-04 Opus 4.5 Collapsed Six Months of Development Work Into One Week — Dan Shipper built complete iOS reading companion app in one week using Opus 4.5, writing no code himself; ‘features are written in English, not code’; ‘prompt-native will be where many features start, and they will be hardened into code over time as they stabilize’
- 2025-12-05 This Prompt Optimizer Learns From Its Mistakes Like DNA — Mike Taylor explains GEPA (Genetic Pareto): evolves prompts via genetic mutation and natural selection — 25% better performance than other methods, 35x fewer trials; prompt accuracy improved from 26% to 71% in 30 minutes; no-code access via LangWatch and Opik; prompt engineering is being automated
- 2025-12-08 Two Ways to Win in the Post-software Era — Sumeet Singh applies Sutton’s ‘bitter lesson’ to argue most AI SaaS wrappers will be absorbed — task-length models can handle doubles every 7 months; two viable paths: The Model Economy (infrastructure) or Post-skeuomorphic apps (workflows impossible without AI)
- 2025-12-11 Compound Engineering How Every Codes Agents — Nobody at Every writes code manually anymore. Four-step engineering process for teams where 100% of code is agent-writte
- 2025-12-12 How This Venture Capitalist Sees Into the Post-software Future — Sumeet Singh (Worldbuild VC): 3-step research framework for post-software opportunities; investment thesis: win via infrastructure models need OR via truly novel apps impossible without AI — not SaaS wrappers; case study: Sardine (fraud in crypto → fintech compliance, 10x return)
- 2025-12-14 Humans Biggest Bottleneck Alexander Embiricos — Alexander Embiricos (Product Lead, Codex, OpenAI) on Lenny’s Podcast. Codex grown 20x since August, trillions of tokens/
- 2025-12-15 Cortney Hickey Zapier EA AI Systems — Cortney Hickey (EA to CEO at Zapier) walks through four AI workflows that transform traditional EA work into scalable or
- 2025-12-16 CJ Hess Flowy Claude Code Codex QA — Software engineer CJ Hess reveals Flowy, a custom-built tool that transforms Claude’s ASCII diagrams into interactive vi
- 2025-12-16 I Talked to More Than 100 Companies About AI—Here’s What’s Actually Working — Natalia Quintero (Every’s head of consulting): core problem is clarity not technology — ‘we have tools, we have power users, we don’t know where to go’; three blocking personas: skeptics, overwhelmed, tool-jumpers; success pattern: 3-5 AI champions, peer-built tools that spread organically, leadership that models personal AI usage
- 2025-12-17 How I 10x Engineering with AI — 100% of PRs opened by AI tools. Three modes of LLM collaboration. Git worktree-based parallel agents
- 2025-12-17 Vibe Check Claude 3.7 Sonnet Claude Code — Claude Code launch review: agentic coding in terminal, no more copy-pasting code
- 2025-12-17 Vibe Check Gemini 2.5 Pro Flash — Gemini 2.5 Pro became default in Cursor and Windsurf. Claude still preferred for style coherence and UI in code
- 2025-12-17 Vibe Check OpenAI o3 GPT-4.1 o4-mini — Three OpenAI models dropped simultaneously. Model selection confusion even for AI-savvy teams
- 2025-12-18 AI Growth Playbook 2026 Elena Verna Lovable — Elena Verna (Head of Growth, Lovable) on Lenny’s Podcast. Lovable hit $200M ARR in under a year with 100 employees
- 2025-12-18 How AI Can Cut Your Planning Cycle From Two Weeks to Two Days — Austin Tedesco (Every’s head of growth): 7-step AI-assisted planning workflow using shared knowledge hubs, AI notetakers, and speech-to-text; key instruction: always have AI ask probing clarifying questions before large tasks; gains: 2-4 weeks earlier execution time, sharper strategic thinking
- 2025-12-19 OpenAI Gave Us a Glimpse Into Their AI Coding Playbook — Four OpenAI engineers built Android Sora app in 28 days using Codex; practices: treat AI as new hire (every session is onboarding), don’t overload context (skills > monolithic files), specialize agents (compose police + tech lead + bug hunter), filter information ruthlessly (only surface failures)
- 2025-12-21 Vibe Check Opus 4.5 Coding Model — Opus 4.5 excels at both planning and coding simultaneously. Makes vibe coding g
- 2025-12-22 Brian Greenbaum Pendo Team AI Adoption — Brian Greenbaum (product designer at Pendo) shares a 3-step playbook for driving company-wide AI adoption
- 2025-12-23 Four Predictions for How AI Will Change Software in 2026 — Dan Shipper and Brandon Gell: (1) agent-native architecture reshapes software into three access levels; (2) designers become builders; (3) three engineering archetypes emerge (AI-accelerated, vibe coder, agentic engineer); (4) AI training shifts toward independence; both humans and agents will be ‘first-class citizens’ in next-gen software
- 2025-12-29 Rachel Wolan Webflow AI Chief of Staff — Rachel Wolan (CPO, Webflow) walks through three workflows including a custom AI Chief of Staff app
- 2025 Every Inside AI Workflows Six Engineers — Each of Every’s six engineers reveals their personalized AI stack. Claude Code vs. Codex side-by-side testing
- 2026-01-05 Wade Foster Zapier AI Recruiting Culture — Wade Foster (CEO, Zapier) presents three interconnected AI workflows for recruiting and culture
- 2026-01-05 How AI Made Pricing Hard Again — Anh-Tho Chuong: AI products destroy SaaS economics — where SaaS enjoyed 70-85% margins, AI pays LLMs per token making growth expensive; explores five pricing models for the AI era; pricing decisions now require engineering involvement since variable AI costs represent both liability and competitive advantage. Partial paywall
- 2026-01-06 I Asked Claude the Question I Could Never Ask My Boss — Katie Parrott used Claude and ChatGPT to analyze her own performance data — discovered she was producing 15% of content output but 25-27% of subscription trials; AI made the self-assessment possible with no social cost; key insight: anyone can build a performance case with available metrics and AI
- 2026-01-07 Reid Hoffman Makes Five Predictions About AI In 2026 — Reid Hoffman: 2026 will be ‘the year we move from agentic coding to agents in everything else’; enterprise meetings will be routinely recorded and summarized by agents; practical AGI = individuals directing agent teams; biology as the next AI language
- 2026-01-08 The Heyday of the Writing-first Practitioner — Eleanor Warnock argues that AI floods make writing-first practitioners more valuable, not less; authentic voice, cultivated networks, and re
- 2026-01-09 Agent-native Architectures: How to Build Apps After the End of Code — Dan Shipper’s foundational essay: traditional apps are like skyscrapers (precise blueprints); agent-native apps are like cities (emergent
- 2026-01-12 Alexander Embiricos OpenAI Codex Workflows — Alexander Embiricos (Product Lead, Codex, OpenAI) presents four workflows progressing from beginner to advanced
- 2026-01-12 The Boring Businesses That Will Dominate the AI Era — Tina He identifies 5 business archetypes irreplaceable when AI agents are primary users: knowledge compounders, workflow commons, reality’
- 2026-01-13 Vibe Check Claude Cowork — Claude Cowork: third tab in Claude desktop alongside Chat and Code. Brings agentic async ‘hand off and come back’ workfl
- 2026-01-14 Why Your AI Learning Projects Keep Fizzling Out — Nir Zicherman argues LLMs make it easy to go deep but don’t solve retention; real learning requires spaced repetition and mental model
- 2026-01-15 AI Can Build Anything. Social Dandelions Decide What Spreads. — Lewis Kallow uses 1930s Iowa corn adoption research to argue distribution advantage comes from understanding how trust travels through comm
- 2026-01-20 What AI Is Teaching Us About Management — Mike Taylor argues AI management and human management are converging: same techniques (clear context, decomposed tasks, feedback loops) appl
- 2026-01-21 Opus 4.5 Changed How Andrew Wilkinson Works and Lives — Andrew Wilkinson (Tiny cofounder) describes using AI to build personal tools and rethinks investing thesis: AI commoditizes code, so mo
- 2026-01-22 What the Team Behind Cursor Knows About the Future of Code — Coverage of Every’s Cursor Camp: IDEs are dead (replaced by AI-native editors); model selection is ‘vibes-based’; retrieval beats compre
- 2026-01-23 I Stopped Reading Code. My Code Reviews Got Better. — Kieran Klaassen stops reading code line-by-line; 50/50 rule: half review time on business logic, half on architectural decisions; fix is
- 2026-01-25 Every 2026 Predictions — Claude Code named most-impactful tool of 2025 across Every team. Dan predicts ‘agentic engineers’ as third kind of engin
- 2026-01-25 Four Predictions AI Change Software 2026 — 2025 was the year coding agents became reliable. Predictions for how software will be built, who will build it, and what
- 2026-01-26 How I Use Claude Code to Ship Like a Team of Five — Kieran Klaassen’s foundational piece on compound engineering: Claude Code feels like delegating to a colleague; orchestra conductor mental m
- 2026-01-26 John Lindquist Advanced Claude Code Techniques — John Lindquist (co-founder egghead.io, AI DX at Vercel) presents three advanced AI engineering workflows for senior engi
- 2026-01-27 My AI Had Already Fixed the Code Before I Saw It — Kieran Klaassen finds Claude Code self-reviewed its own PR unprompted, citing 3 prior PRs — defines ‘compounding engineering’: every fix
- 2026-01-28 Stop Coding and Start Planning — Kieran Klaassen: 10 minutes of AI-assisted planning prevents 3 hours of debugging; plans teach AI how you think; three planning fidelities
- 2026-01-29 Teach Your AI to Think Like a Senior Engineer — Kieran Klaassen’s 8 planning strategies for AI-assisted engineering; running specialized research agents in parallel before coding prevents
- 2026-01 Agent Native Architectures After Code Ends — Skyscraper metaphor: traditional software as towering facade. Claude Code demonstrated that coding agent = general-purpo
- 2026-02-03 We Trained an AI on a Board Game. It Became a Better Customer Support Agent. — Good Start Labs fine-tuned Qwen3-235B on Diplomacy and found game-based RL training produces generalizable skills that transfer across do
- 2026-02-04 Every’s Head of Consulting Just Automated Her Job — Natalia Quintero shares three patterns in AI-forward companies; built ‘Claudie’ AI project manager that cut weekly PM workload from 15 ho
- 2026-02-05 Teresa Torres Claude Code Task Management — Teresa Torres’s Claude Code system for task management, research automation, and context libraries (Claire Vo / ChatPRD)
- 2026-02-06 Agent Native Engineering — Restructuring engineering orgs around agents as ICs: task levels, code review, token spend (Andrew Pignanelli)
- 2026-02-06 Brex Agent Human Ops — How Brex is redesigning operations with AI: 3-level ops model, agent platform, hiring for generalists (First Round)
- 2026-02-06 Build AI Product Sense — Using Cursor for non-technical work to build AI product sense — RAG, memory, context engineering, tool calling (Lenny’s Newsletter)
- 2026-02-06 Teresa Torres Claude Code Guide — Claude Code for non-technical people: comparison with browser/desktop, competitive analysis example, data portability (Product Talk)
- 2026-02-07 Iterative Workflow Refinement With AI Agents — Organic draft: iterative refinement of AI agent skill files through three rounds of feedback (Dudgeon & Claude)
- 2026-02-07 Simon Scrapes Every Level Claude Code — Seven-level progression framework for Claude Code mastery: intentional prompting, CLAUDE.md personalization
- 2026-02-07 Zapier Workflows CRM Automation Meeting Prep — Zapier MCP workflows for CRM automation, meeting prep, and self-improving customer feedback systems (ChatPRD)
- 2026-02-08 AI Augmented Product Development Cycle — Maps AI tools across full product dev cycle: ideation, research, MVP, dev, launch, post-launch optimization (Product-Led Alliance)
- 2026-02-08 Beyond the Pillars AI Digital Products — Academic research synthesis on agentic systems, full-lifecycle orchestration, AI-first PM frameworks (Medium)
- 2026-02-08 Claire Vo How I AI Podcast Launch — Announcement post for How I AI podcast on Lenny’s network; thin on extractable knowledge (ChatPRD Blog)
- 2026-02-08 Claire Vo Idea to Product 30 Min — PRD, Devin (AI coding agent) async delegation; replaced Cursor with Devin for async workflow; future-proof PM competencies (Creator Economy)
- 2026-02-08 Make PM Fun Again AI Agents — Agentic spectrum definition, five agent design principles, Zapier Agent tutorial for meeting prep, meta-prompt for platform selection (Lenny’s Newsletter)
- 2026-02-08 Marily Nika AI Tool Stack — Six-tool AI PM stack: AI Studio, Opal, NotebookLM, Perplexity, ChatGPT, Fireflies; “be like a crab” career advice (Aakash Gupta)
- 2026-02-08 Marily Nika PM AI Toolkit — Google AI PM demos end-to-end workflow: Perplexity, custom GPT PRD, v0.dev prototype, Flow/Sora video in 20 min; workflow inversion — prototype before PRD (How I AI / Lenny’s)
- 2026-02-08 Reid Robinson Navigating AI Integrations — Tie AI to existing “skeleton-in-the-closet” problems; Zapier’s ChatGPT integration as fastest-growing ever (BetaKit)
- 2026-02-08 Teach AI Think Like Senior Engineer — 8 strategies to teach AI your codebase, patterns, and preferences. Running multiple specialized research agents in paral
- 2026-02-08 USIDO Framework AI Product Management — U.S.I.D.O. Framework (Understand, Strategize, Ideate, Define, Optimize) mapping AI across the product lifecycle, with companion Custom GPTs (Medium)
- 2026-02-09 CJ Hess Build AI Developer Tools Claude Code — Software engineer CJ Hess (Tenex) demonstrates two workflows: Flowy visual planning tool and model-vs-model code reviews
- 2026-02-09 Compound Engineering: The Definitive Guide — Kieran Klaassen introduces ‘compound engineering’ — each unit of AI-assisted work makes subsequent work easier; launch post for comprehen
- 2026-02-09 Managing Agents New Horizontal Domain — Identifies “Managing Agents” as a nascent horizontal skill domain — agent selection, onboarding, performance retros, context management (note)
- 2026-02-09 Scaling People to Scaling Agents — Maps CHJ’s Scaling People management framework to AI agent management: mechanism layer, retro pattern, trust tiers, skill-reliability matrix, delegation cube, attention economics (organic)
- 2026-02-10 Building AI Product Sense Part 2 Marily Nika — Marily Nika (Gen AI Product Lead @ Google, formerly Meta; PhD in ML; AI Product Academy founder) guest post Part 2 on Le
- 2026-02-10 Ship Like Team of Five Claude Code — Every piece of code shipped in last 2 months written by AI. Claude Code opens 100% of PRs. Paral
- 2026-02-11 Inside OpenAI’s Agentic Browser, Atlas — OpenAI Atlas engineers discuss building an agentic browser: >50% of code written by Codex; AI tools find more elegant solutions than huma
- 2026-02-12 How Claude Code Is Transforming Finance—Without Turning You Into a Coder — Every’s head of financial services consulting explains four high-value finance use cases for Claude Code; key unlock is bundled plugins tha
- 2026-02-12 Vibe Check Codex App vs Claude Code — Codex app as ‘command center for agents’ designed to replace terminal UI. Even non-nerds psyched about Claude Code
- 2026-02-13 ClawVault Agent Memory Obsidian Architecture — Open-source agent memory modeled on Obsidian: markdown+YAML outperforms specialized memory infra (74% vs 68.5% LoCoMo), typed memory taxonomy, wiki-link knowledge graphs, budget-aware context injection, vault index pattern (@sillydarket / Versatly, X)
- 2026-02-13 The Two-slice Team — Dan Shipper: Amazon’s ‘two-pizza rule’ is obsolete — AI makes the ‘two-slice team’ viable: one person running an entire software product wi
- 2026-02-13 Compound Engineering Definitive Guide — Comprehensive handbook for AI-native engineering philosophy. AI Development Ladder progression from basic chat to fully
- 2026-02-15 Claire Vo GPT-5.3 Codex vs Opus 4.6 — Claire Vo puts OpenAI’s GPT-5.3 Codex and Anthropic’s Claude Opus 4.6 head-to-head on real engineering tasks
- 2026 Agent Native Architectures Guide — Coauthored with Claude. Principles from building Reader and Anecdote. A coding agent is actually a general-purpose agent
- 2026 Compound Engineering Guide — AI Development Ladder progression from basic chat assistance to fully autonomous engineering
- 2026-02-19 What Board Games Taught Me About Working with AI — Katie Parrott applies board game ‘teach’ framework (pieces→moves→systems→victory) to build a compound writing agent; TASTE.md + CLAUDE.md analog; two-layer system: defaults (baseline) + taste (fills over time); “the engine only reveals its flaws when I actually play” (partial paywall)
Reading
None yet.
Read (Awaiting Processing)
- 2026-02-17 How to Build Agent-native: Lessons From Four Apps — Four Every apps demonstrate agent-native architecture: minimal toolset (read/write/list), emergent behavior, safety constraints in tools not prompts, CLI sandbox retrofitting path (Every / Katie Parrott)
Processed
- 2026-02-06 Spec Driven Development No Code Library — Spec-only library with zero code; when spec-driven dev works vs when traditional code is necessary (Drew Breunig) → Spec-Driven Development
- 2026-02-06 Git Workflows Agentic Era — Git subtree + submodule patterns for sharing context across repos in agentic workflows (organic) → Repositories as Context Boundaries
- 2026-02-07 Ryan Carson Structured AI Development — Three-file system for PRD, tasks, execution in Cursor, plus MCP automation and Repo Prompt (ChatPRD) → Interactive PRD Writing, Task List Generation for Observability, Stepwise Task Execution, Context First Development, Deliberate Context Selection
- 2026-02-07 Hilary Gridley Scaling Yourself Custom GPTs — Building custom GPTs that clone your expert judgment for slide evaluation and writing coaching (ChatPRD) → Reverse Engineer Judgment Into AI, Be 100x More Specific, My Job Your Job Role Delineation, AI as Writing Coach, Scale Manager Expertise With AI
- 2026-02-08 How AI Will Impact Product Management — AI disrupts strategic PM skills most; Shape/Ship/Sync three-job model with disruption ratings (Lenny’s Newsletter) → AI Disrupts Strategic PM Skills Most, Shape the Product, Ship the Product, Sync the People
- 2026-02-14 Claude Skills Explained Reusable AI Agents — Three workflows for creating Claude Skills; meta-skill factory pattern (ChatPRD) → Meta-Skill Pattern
- 2026-02-14 Mernit Openclaw Filesystem as State — Filesystem-as-state + Claude-as-orchestrator; company-as-filesystem; data silos block agent adoption (@mernit, X) → Filesystem as Agent State, Data Silos Block Enterprise Agent Adoption
- 2026-02-13 Agentic Team Memory Devin — Devin captures team knowledge as side effect of engineer corrections; multi-source knowledge, scoped retrieval, emergent onboarding (Nader Dabit / Cognition, X) → Knowledge Capture as Side Effect, Self-Maintaining Knowledge Bases
- 2026-02-13 Everyone Should Use Claude Code More — 50 non-technical Claude Code use cases; tool identity as adoption gate, cross-tool workflow hub via MCP, agent-mediated self-reflection, self-driving documentation (Lenny’s Newsletter) → Tool Identity as Adoption Gate, Agent as Cross-Tool Workflow Hub, Agent-Mediated Self-Reflection, Knowledge Capture as Side Effect (enriched)
- 2026-02-08 Getting Paid to Vibe Code — Professional vibe coder at Lovable: parallel prototyping, structured context loading, 80/20 planning ratio, rules files as knowledge capture (Lenny’s Newsletter) → Parallel Prototyping for Clarity, Structured Context Loading, Context First Development (enriched), Deliberate Context Selection (enriched), Knowledge Capture as Side Effect (enriched)
- 2026-02-14 Progressive Disclosure & Context Graphs — 3-layer retrieval pattern, filesystem-as-retrieval-architecture, progressive tool disclosure, retrieval infrastructure graduation path (note) → Three-Layer Context Disclosure, Filesystem as Retrieval Architecture, Progressive Tool Disclosure, Retrieval Infrastructure Graduation
- 2026-02-13 Harness Engineering Leveraging Codex — Zero manually-written code at scale using Codex agents: 1M lines, 1500 PRs, harness engineering methodology (Ryan Lopopolo / OpenAI) → OAI Harness Engineering, Mechanical Architecture Enforcement, Agent Entropy Management, Three-Layer Context Disclosure (enriched), Filesystem as Retrieval Architecture (enriched)
- 2026-01-27 Management as AI Superpower — Three-variable delegation framework, delegation documentation as AI prompts, talent-to-direction scarcity shift (One Useful Thing) → Delegation Decision Framework, Delegation Documentation as Agent Prompts, Talent-to-Direction Scarcity Shift
- 2026-02-12 Letta Context Repositories Git Memory — Git-backed agent memory as filesystem: agent-self-managed progressive disclosure, initialize/reflect/defragment lifecycle skills, git-versioned commits, concurrent worktrees for memory swarms (Letta) → Agent-Self-Managed Progressive Disclosure, Agent Memory Lifecycle Skills, Git-Versioned Agent Memory, Concurrent Agent Memory via Git Worktrees
- 2026-02-18 How to Do AI Analysis You Can Actually Trust — Four failure modes of AI qualitative analysis with fixes: quote selection rules, quote verification pass, decision-anchored context loading, few-shot calibration, multi-pass verification; LLMs generate not retrieve; model fit for analysis (Caitlin Sullivan / Lenny’s Newsletter) → Quote Selection Rules, Quote Verification Pass, Decision-Anchored Analysis Context, Few-Shot Calibration for Analysis, AI Analysis Multi-Pass Verification, LLMs Generate Not Retrieve, Model Fit for Qualitative Analysis
- 2026-02-19 Head of Claude Code: What Happens After Coding Is Solved — Coding “solved” (0% hand-written code, 10–30 PRs/day, 200% Anthropic productivity gains); build for model 6 months out; understaff to force AI adoption; delay token cost optimization; plan mode as default; use max-capable model; AI moving from execution to ideation; generalists outperform (Boris Cherny / Lenny’s Podcast) → Latent Demand as Product Signal, Build for Future Model Capability, Intentional Understaffing for AI-First Teams, Delay Token Cost Optimization, Plan Mode as Claude Code Default, Capability Over Cost in Model Selection, AI Moves from Execution to Ideation, Generalists Outperform Specialists in AI Era
- 2026-01-30 Compound Engineering: How Every Codes With Agents — 4-step compound engineering loop (Plan, Work, Assess, Compound); each feature makes the next easier; 80% plan+review; rules files as institutional memory (Dan Shipper, Kieran Klaassen / Every) → Compound Engineering Loop, Knowledge Capture as Side Effect (enriched), Intentional Understaffing (enriched)
Stats
- Total sources: 185
- Unread: 167
- Read (awaiting processing): 1
- Processed: 15
- Knowledge entries generated: 51
Updated: 2026-02-20