Lenny's Written Position
AI is most valuable for PMs as a 'second brain' for synthesis — combining meeting notes, user research, and market data into actionable insights.
The key to useful AI assistance is feeding it context over time, not just asking one-off questions — persistent context makes AI dramatically more useful.
AI coding tools have crossed a threshold where non-technical people can build real, useful products — not just demos.
The best way to learn AI coding tools is to start with a personal problem you actually want solved, not a tutorial project.
Product managers who learn to build with AI tools will have a significant career advantage — it's the most leveraged new skill to develop.
Vibe coding is most powerful for internal tools and personal productivity — not for building the next startup.
The gap between 'demo' and 'daily driver' is where most vibe-coded projects die — polish and reliability matter more than initial creation.
The PM job market is recovering, but the roles that are growing fastest require AI literacy — traditional PM skills alone are no longer sufficient.
AI prototyping collapses the traditional design-to-development handoff — PMs can now validate ideas before involving engineering.
Starting with JSON data structures (not UI mockups) leads to better AI-generated prototypes because it forces clarity of thinking about the product.
The PM role is uniquely positioned to be augmented (not replaced) by AI because the core job is synthesis and judgment, not execution.
The most effective PMs in the AI era are those who can move between strategic thinking and hands-on building — the 'full-stack PM'.
Traditional B2B growth channels (inbound, outbound, product virality, events, lifecycle emails) are becoming less effective because AI has made it easy for everyone to produce content and flood channels.
The biggest barrier to AI adoption at companies is not technology but organizational change, including vague mandates, procurement bottlenecks, and lack of guidance on high-impact use cases.
Saying 'we are AI-first' means nothing if employees do not know what that means for their day-to-day work; successful companies provide specific, concrete tactics employees can adopt.
AI adoption should be tracked as both inputs (who is using AI) and outputs (business value created), and included in performance reviews to create accountability.
Duolingo went from creating 100 courses in 12 years to 150 courses in just 12 months after rebuilding their content creation process with AI.
True AI adoption happens when teams become skeptical of flashy demos and instead demand to see accuracy metrics, evaluation frameworks, and failure modes behind AI products.
Companies should turn their internal AI enthusiasts into teachers by hosting regular demos, hackathons, and dedicated AI experimentation time rather than relying on top-down mandates alone.
Most tech workers are missing out on AI's potential because they are not providing enough context; LLMs feel like blunt, generic instruments when they lack the background knowledge a human colleague would need.
An AI copilot built with ongoing context about your goals, role, projects, team, and org becomes a real thinking partner for long-term complex work, not just a document generator.
Building an AI copilot follows a four-step process analogous to onboarding a teammate: hire it with instructions, onboard it with company knowledge, kick off initiatives in chat threads, and put it to work with simple prompts.
The habit of 'gossiping' to your AI copilot -- casually updating it about conversations, stakeholder changes, and new information via speech-to-text -- is critical for keeping context fresh and effective.
Even when AI is wrong it can be valuable because it spurs you to crystallize what you actually think; the goal is not perfect answers but getting the most out of yourself as a thinking partner.
AI copilot investment earns compound interest: uploading retrospectives and lessons learned from completed initiatives makes the copilot increasingly effective across all future work.
The two biggest barriers to AI prototyping adoption on product teams are making prototypes look good enough for stakeholders and figuring out team workflows instead of individual silos.
Building a reusable component library is the single biggest improvement teams can make to AI prototyping quality, allowing brand-consistent prototypes without manual cleanup each time.
AI prototyping introduces a new 'medium fidelity' tier — better than napkin drawings but not as polished as finalized Figma mocks — and choosing the right fidelity for each context is critically important.
The code generated by AI prototyping tools is mostly useless to engineering teams because it does not follow existing patterns, use the same libraries, or even use the same programming language.
The 'baselines and forks' workflow — creating a high-quality reproduction of your current product then duplicating it to explore new ideas — is the best way to rapidly test multiple design directions without rebuilding each time.
Using the Figma MCP server with Cursor allows AI agents to autonomously take screenshots, extract design tokens, and get CSS from Figma's Dev Mode, producing prototypes indistinguishable from the real product.
Replit can single-shot fairly sophisticated full-stack applications unlike other vibe-coding platforms, because it automatically tests the apps it builds using its internal browser and auto-fixes issues it finds.
Every startup and modern product team should be using Linear for task tracking, project organization, and roadmap building — and at this point, most are.
Creating decks and documents is one of the most time-consuming and annoying parts of most product builders' jobs, and AI tools like Gamma are doing for PMs what Cursor is doing for engineers.
Lovable reached $100 million ARR just eight months after launch, making it the fastest-growing company in history — faster than even Cursor.
Bolt grew from 0 to $40 million ARR in five months and is used by 72% of Fortune 500 product teams, differentiating itself by integrating frontier agents like Claude Code rather than relying on homegrown agents.
Granola has quickly become the obvious tool for AI meeting note-taking because it works behind the scenes using laptop audio instead of joining calls, and everyone who tries it gets hooked.
AI automations (tools like Zapier Agents, Lindy AI, Relay App) are currently the most practical category of AI agents for helping product managers offload repetitive busywork.
PMs should design AI agents with low downside by restricting outputs to drafts, DMs, and recommendations rather than allowing direct actions like sending emails or posting in channels.
Letting AI summarize everything will quickly degrade a PM's customer intuition; instead PMs should use AI to traverse and cluster data while insisting on exact quotes and direct links to original sources.
AI copilots, AI prototyping tools, and AI agents will converge into unified tools that have deep company context, connect to real-world inputs and outputs, and operate by natural conversation.
Using AI prototyping tools to build production AI agents is tempting but not recommended because it creates maintenance burden, SaaS registration bureaucracy, and compliance overhead that dedicated agent platforms handle out of the box.
AI agents shine at ongoing, one-at-a-time repetitive tasks rather than big one-time batch tasks, and PMs should choose continuous trigger-driven work as their first agent use case.
Product teams will shrink by 25-50% due to AI (mostly fewer engineers), while PMs will have more influence and leverage, spending more time in discovery and GTM and less in designing and building.
ChatGPT is used by 90% of tech professionals surveyed, making it more widely used than Gmail (76%) or Slack (71%), representing the most significant shift in the product team tool stack in recent memory.
Cursor is already used by 17% of all respondents and 21% of engineers despite launching just two years ago, signaling rapid adoption of AI-native development environments.
Over 50% of tech professionals combine multiple AI assistants for specific use cases, such as ChatGPT plus Claude for thought partnership and ChatGPT plus Perplexity for deep research.
Figma is used by 97% of designers as their primary design tool, while Canva is democratizing design for non-designers like PMs, marketers, and engineers.
Managers can scale their coaching by distilling their most common feedback into custom GPTs, allowing team members to receive feedback multiple times a day rather than once or twice a week.
If you are skilled at explaining something to people, you are well equipped to teach it to an AI, making great managers uniquely suited to creating effective custom GPTs.
If 2024 was the year of the super IC, 2025 will be the year of the supermanager who harnesses AI to oversee larger, flatter teams and extend coaching beyond just direct reports.
Teams should treat AI as a true thought partner through back-and-forth iteration rather than relying on copy-pasted single-use prompts or pre-built GPTs.
You can reverse-engineer your own managerial intuition into a custom GPT by uploading examples of high-quality and low-quality work, asking the AI to identify specific differences, and then creating a prompt that transforms bad examples into good ones.
The difference between mediocre and mind-blowing AI results comes down to prompt crafting, not the model itself.
Role-playing, style unbundling, emotion prompting, few-shot learning, and synthetic bootstrap are the five most consistently useful prompt engineering techniques.
Adding emotional stakes to a prompt (e.g. 'This task is very important for my career') can elicit more careful and thoughtful AI responses.
Splitting AI tasks into multiple steps with evaluation loops outperforms trying to get everything right in a single prompt.
While AI won't replace you directly, those who use AI effectively will have an advantage in promotions and job applications.
The three most common hard skills hiring managers look for from PMs are SQL, Jira, and experience with LLMs, with LLMs being the fastest-growing requirement.
Machine-learning engineers and data engineers are the fastest-growing tech roles, growing 79% and 55% year over year respectively.
AI companies are hiring almost exclusively engineers and data scientists; there are only 456 open AI PM roles out of 16,935 total AI company openings.
Product managers are the best-positioned role in tech to thrive in a world of AI because the PM job is an amalgamation of soft skills that AI will have the toughest time replacing.
The most valued skill set in tech will increasingly shift from building to knowing what to build, giving clear instructions, and having the taste to judge quality.
Engineers and designers with the strongest PM-type skills like identifying customer pain points, understanding business levers, and communicating clearly will do best in an AI world.
Many jobs across industries will start to look more like product management as AI handles execution and humans focus on directing, iterating, and driving adoption.
A top-tier PM with the skills to fully harness AI capabilities is going to be the most valuable role in tech.
AI is useful for content workflow steps like extracting structure, identifying topics, and keyword research, but should not be used for the actual writing because factual accuracy cannot be guaranteed.
In blind tests, AI-generated answers beat human PM answers in two out of three core PM tasks (product strategy, defining KPIs, and estimating ROI), with 70-80% of voters correctly identifying the AI answer but still often preferring it.
Most people underestimate how close AI is to replacing human work because evaluations typically use outdated models without prompt engineering, which can yield a 50-60% accuracy boost.
AI's primary weakness in PM tasks is producing tactically comprehensive but not truly strategic answers, while humans win by incorporating unexpected connections, niche references, and genuine strategic thinking.
Even when AI ties with human performance on PM tasks, it is effectively a win for AI because it costs pennies and takes seconds versus hours of human effort.
Over 50% of product managers now use an AI chatbot daily, and 85% use one weekly, representing an unprecedented pace of tool adoption in the PM profession.
PMs are primarily using AI search tools like Perplexity for six categories of work: growth strategy, finding benchmarks, market research, learning best practices, evaluating tools, and understanding technical jargon.
Perplexity structures teams to minimize coordination costs by parallelizing projects and using AI for rubber-duck debugging instead of relying on alignment and consensus.
Quality assurance is the PM skill area most likely to be significantly disrupted by AI, as tools can already catch unexpected behaviors by being fed PRDs and testing autonomously.
Custom GPTs can deliver immediate productivity gains across diverse work functions including UX copy refinement, persona conversations, user research indexing, copy experiment generation, and goal setting.
When Italy banned ChatGPT, coder productivity fell by 50%, while Duolingo reported a 25% increase in developer velocity using GitHub Copilot and Shopify wrote over a million lines of code with Copilot.
PMs should consider delivering not just a PRD but also a custom GPT loaded with those same requirements, giving developers an interactive source of truth for brainstorming technical approaches.
Product analytics tools like Amplitude Compass and Mixpanel Signal can quickly identify which features correlate most strongly with user retention.
Learning to work alongside AI will quickly become table stakes for product managers, just as Grammarly became for writing and Copilot for engineering.
The best way to prepare for the AI future in product management is to dive in and get your hands dirty rather than waiting for the technology to mature.
GPT-3 tends to hallucinate confident but incorrect answers, especially when it lacks the right training data, making context injection via embeddings essential for accurate chatbots.
Every newsletter, book, blog, and podcast used as evergreen reference information can now be repackaged as a chatbot, creating a new content format and monetization opportunity for creators.
Podcast Moments
“This design process that designers have been taught, where you go off and do a bunch of research and discovery, and then you diverge, converge, diverge, converge — we sort of treated as gospel. That's basically dead.”
Jenny Wen · Jenny Wen
“There are basically two types of design work now. The first is supporting implementation and execution — engineers are using their seven Claudes to create features. The second is creating the vision or direction, but now it's a three to six month vision, not a two to five year one, and it's sometimes just creating a prototype that points people in the right direction.”
Jenny Wen · Jenny Wen
“It's not just designers who are feeling like, 'Oh yeah, we have to keep up with engineers.' I think even engineers are like, 'How do we keep up with ourselves?'”
Jenny Wen · Jenny Wen
“I think AI will get better at taste and judgment and design. We might be holding onto that a little bit too much. At the end of the day, someone has to decide what is actually going to get built and what actually matters. Someone still needs to be accountable for the decision.”
Jenny Wen · Jenny Wen
“You have to know the difference between a megatrend and a hype cycle. When there's a megatrend, don't fight it. AI is a megatrend, one of the most foundational movements that we have seen in human history.”
Jeetu Patel · Jeetu Patel
“100% of my code is written by Claude Code. I have not edited a single line by hand since November. Every day, I ship 10, 20, 30 pull requests.”
Boris Cherny · Boris Cherny
“Productivity per engineer has increased 200%. I have never enjoyed coding as much as I do today, because I don't have to deal with all the minutia.”
Boris Cherny · Boris Cherny
“In a year or two, it's not going to matter. Coding is virtually solved. I imagine a world where everyone is able to program, anyone can just build software any time.”
Boris Cherny · Boris Cherny
“Claude is starting to come up with ideas. It's looking for feedback, it's looking at bug reports, it's looking at telemetry for bug fixes, and things to ship. A little more like a coworker.”
Boris Cherny · Boris Cherny
“The go-to-market is going to get turned on its head. People start evaluating products in Gemini or ChatGPT. Your website is a lot less important. Sites will change to have a high quality avatar that knows everything about your products.”
Sequoia CEO coach: Why it’s never been easier to start a company, and never been harder to scale one | Brian Halligan (co-founder, HubSpot) · Brian Halligan
“IC engineers are becoming tech leads. They're managing fleets and fleets of agents. I know many engineers on my team have 10 to 20 threads being pulled on at the same time.”
Sherwin Wu V2 · Sherwin Wu V2
“The one person billion dollar startup implies that to enable it, there might be a hundred other small startups building bespoke software. We might actually enter into a golden age of B2B SaaS.”
Sherwin Wu V2 · Sherwin Wu V2
“Companies where AI really works have top-down buy-in combined with bottoms-up adoption. A lot of AI deployments fail because it's an exec mandate, extremely top-down, divorced from what actual work looks like.”
Sherwin Wu V2 · Sherwin Wu V2
“The models will eat your scaffolding for breakfast. Build for where the models are going, not where they are today.”
Sherwin Wu V2 · Sherwin Wu V2
“We live in a bubble in Silicon Valley. A lot of work that runs our entire economy is not open-ended knowledge work — it's repeatable business processes with standard operating procedures.”
Sherwin Wu V2 · Sherwin Wu V2
“Every startup that I've seen fizzle out is not because OpenAI or a big lab has come and squashed them. It's because they built something that didn't resonate with customers.”
Sherwin Wu V2 · Sherwin Wu V2
“I don't have a coding background. My job is to build both internal tools and customer-facing products purely using AI. The key is knowing what to build, not how to build it.”
The rise of the professional vibe coder (a new AI-era job) | Lazar Jovanovic (Professional Vibe Coder) · Lazar Jovanovic
“The biggest mistake people make is starting too big. Start with something you actually need. My first project was a simple dashboard for tracking my own metrics.”
The rise of the professional vibe coder (a new AI-era job) | Lazar Jovanovic (Professional Vibe Coder) · Lazar Jovanovic
“Internal tools are the sweet spot for vibe coding. The bar for quality is lower, the feedback loop is faster, and the value is immediate.”
The rise of the professional vibe coder (a new AI-era job) | Lazar Jovanovic (Professional Vibe Coder) · Lazar Jovanovic
“Every PM in the world should be learning to build with AI tools. Not because they'll replace engineers, but because the ability to prototype changes the entire product development cycle.”
Marc Andreessen: The real AI boom hasn’t even started yet · Marc Andreessen
“I'm a PM at Meta with zero coding background. I built a full internal tool using Cursor in three days. It's now used by my entire team daily.”
The non-technical PM’s guide to building with Cursor | Zevi Arnovitz (Meta) · Zevi Arnovitz
“Start with your own pain point. Don't try to build a startup. Build something you need. The motivation to finish is completely different when you're the user.”
The non-technical PM’s guide to building with Cursor | Zevi Arnovitz (Meta) · Zevi Arnovitz
“Vibe coding is exactly good for this — Cursor and DigitalOcean and Cloudflare will get you a long way. But vibe coding also doesn't really scale.”
How to show up in any room with a low heart rate: Silicon Valley’s missing etiquette playbook | Sam Lessin · Sam Lessin
“Seed VCs who invest in companies branded as AI companies are going to lose an impossibly large amount of money. There's a difference between a great business using AI versus an 'AI company.'”
How to show up in any room with a low heart rate: Silicon Valley’s missing etiquette playbook | Sam Lessin · Sam Lessin
“Context engineering is the new prompt engineering. It's not about the single prompt — it's about what information you give the AI over the entire session.”
The power user’s guide to Codex: parallelizing workflows, planning techniques, advanced context engineering tips, automating code reviews, and more | Alexander Embiricos · Alexander Embiricos
“The trick is parallelization. Run five agents at once on different parts of the problem. Your job becomes orchestration, not execution.”
The power user’s guide to Codex: parallelizing workflows, planning techniques, advanced context engineering tips, automating code reviews, and more | Alexander Embiricos · Alexander Embiricos
“Point solutions don't have enough data in the age of AI to be useful. If you own the mine with the data, you can make money.”
10 contrarian leadership truths every leader needs to hear | Matt MacInnis (Rippling) · Matt MacInnis
“Imagine you wanted to train a model to write an eight line poem about the moon. Most people check: Is this a poem? Does it contain eight lines? But we are looking for Nobel Prize-winning poetry.”
The 100-person AI lab that became Anthropic and Google's secret weapon | Edwin Chen (Surge AI) · Edwin Chen
“Instead of building AI that will actually advance us as a species, we are optimizing for AI slop instead. We're basically teaching our models to chase dopamine instead of truth.”
The 100-person AI lab that became Anthropic and Google's secret weapon | Edwin Chen (Surge AI) · Edwin Chen
“An RL environment is essentially a simulation of the real world. We might build a world where you have a startup with Gmail, Slack, Jira, GitHub. And then suddenly AWS goes down. Model, what do you do?”
The 100-person AI lab that became Anthropic and Google's secret weapon | Edwin Chen (Surge AI) · Edwin Chen
“Do you want a model that says 'There are definitely 20 more ways to improve this email' or one that's optimizing for your time and just says 'Your email's great, just send it'?”
The 100-person AI lab that became Anthropic and Google's secret weapon | Edwin Chen (Surge AI) · Edwin Chen
“Vibe coding is over-hyped. People don't realize how much it's going to make your systems unmaintainable in the long-term.”
The 100-person AI lab that became Anthropic and Google's secret weapon | Edwin Chen (Surge AI) · Edwin Chen
“We're calling it the 'full-stack builder.' A PM who can prototype, run evals, analyze data, and ship small features independently. This is the new bar at LinkedIn.”
Why LinkedIn is turning PMs into AI-powered "full stack builders” | Tomer Cohen (LinkedIn CPO) · Tomer Cohen
“The PMs who are thriving right now are the ones who see AI as expanding their capabilities, not threatening their role. They're building things they never could before.”
Why LinkedIn is turning PMs into AI-powered "full stack builders” | Tomer Cohen (LinkedIn CPO) · Tomer Cohen
“With AI, it's just intensified because you have 10 players pursuing the same market opportunity and so your ability to actually bring the product to market has become more strategically important.”
What world-class GTM looks like in 2026 | Jeanne DeWitt Grosser (Vercel, Stripe, Google) · Jeanne Grosser
“Whatever AI does, currently or in the future, is up to us. Every technology is a double-edged sword.”
The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li · Dr. Fei Fei Li
“We curated 15 million images on the internet, created a taxonomy of 22,000 concepts. That combination of big data, neural network, and GPU was the golden recipe for modern AI.”
The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li · Dr. Fei Fei Li
“In the middle of 2015, 2016, some tech companies avoided using the word AI because they were not sure if AI was a dirty word.”
The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li · Dr. Fei Fei Li
“I realized the North Star challenge of AI is visual intelligence. Our human brains use about 50% of our neurons on visual processing. This is fundamentally a big data problem.”
The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li · Dr. Fei Fei Li
“We find engineering teams that are very AI forward are reporting about eight to 10 hours saved per week. This is the worst it will ever be. This is now the baseline.”
How Block is becoming the most AI-native enterprise in the world | Dhanji R. Prasanna · Dhanji R. Prasanna
“He built this system where Goose is essentially just watching everything he does all the time. A few hours later Goose has already tried to build that feature and opened a PR.”
How Block is becoming the most AI-native enterprise in the world | Dhanji R. Prasanna · Dhanji R. Prasanna
“Vibe coding is highly limiting. We're trying to push Goose to work not just for five minutes at a time but for hours.”
How Block is becoming the most AI-native enterprise in the world | Dhanji R. Prasanna · Dhanji R. Prasanna
“One of the things I do regularly is just throw away huge amounts of code. What would our world look like if every release deleted the entire app and rebuilt from scratch?”
How Block is becoming the most AI-native enterprise in the world | Dhanji R. Prasanna · Dhanji R. Prasanna
“The non-technical people using AI agents and programming tools to build things is really what's been surprising and amazing.”
How Block is becoming the most AI-native enterprise in the world | Dhanji R. Prasanna · Dhanji R. Prasanna
“The biggest mistake in measuring AI productivity is looking at lines of code or PRs per day. You need to measure outcomes: cycle time, change failure rate, deployment frequency.”
How to measure AI developer productivity in 2025 | Nicole Forsgren · Nicole Forsgren
“DORA metrics still work in the AI era, but you need to add new dimensions. We're adding 'AI augmentation rate' — what percentage of work is AI-assisted vs. purely human.”
How to measure AI developer productivity in 2025 | Nicole Forsgren · Nicole Forsgren
“AI is expansionary. There's actually just more and more questions being asked and curiosity that can be fulfilled now with AI.”
Inside Google's AI turnaround: The rise of AI Mode, strategy behind AI Overviews, and their vision for AI-powered search | Robby Stein (VP of Product, Google Search) · Robby Stein
“These things take 6 to 12 months to get truly robust. Like with any major tech revolution, headlines tell one story and on the ground, you need to dig up every road.”
First interview with Scale AI’s CEO: $14B Meta deal, what’s working in enterprise AI, and what frontier labs are building next | Jason Droege · Jason Droege
“The general trend is going from models knowing things to models doing things. That's where environments come into play.”
First interview with Scale AI’s CEO: $14B Meta deal, what’s working in enterprise AI, and what frontier labs are building next | Jason Droege · Jason Droege
“18 months ago, you would get a short story. Now one task is building an entire website by one of the world's best developers. These tasks now take hours and require PhDs.”
First interview with Scale AI’s CEO: $14B Meta deal, what’s working in enterprise AI, and what frontier labs are building next | Jason Droege · Jason Droege
“If you have a human process that is 10 or 20% accurate, AI is awesome. If you have one at 98% accurate and expect the remaining 2%, not totally there yet.”
First interview with Scale AI’s CEO: $14B Meta deal, what’s working in enterprise AI, and what frontier labs are building next | Jason Droege · Jason Droege
“A document that reads the same words in company A will have a different meaning in company B. Digitizing judgment is becoming a bottleneck.”
First interview with Scale AI’s CEO: $14B Meta deal, what’s working in enterprise AI, and what frontier labs are building next | Jason Droege · Jason Droege
“I start every AI prototype with a JSON schema. Not a wireframe, not a design. JSON. Because it forces you to think about what the product actually does before how it looks.”
The secret to better AI prototypes: Why Tinder’s CPO starts with JSON, not design | Ravi Mehta (product advisor, previously EIR at Reforge) · Ravi Mehta
“The best PMs I know right now are all building. Not writing specs and throwing them over the wall. Actually building prototypes, testing them with users, iterating in real-time.”
The secret to better AI prototypes: Why Tinder’s CPO starts with JSON, not design | Ravi Mehta (product advisor, previously EIR at Reforge) · Ravi Mehta
“Answer Engine Optimization is how do I show up in LLMs as an answer? You need to get mentioned as many times as possible.”
The ultimate guide to AEO: How to get ChatGPT to recommend your product | Ethan Smith (Graphite) · Ethan Smith
“Our whole team is only 15 engineers. Most folks work with up to five Devins at once. Devin merges several hundred pull requests into production every month.”
How Devin replaces your junior engineers with infinite AI interns that never sleep | Scott Wu (Cognition CEO) · Scott Wu
“Defining the problem is probably 10% of the average software engineer's time. 90% is Kubernetes errors, debugging, migrations. Devin allows engineers to go from bricklayer to architect.”
How Devin replaces your junior engineers with infinite AI interns that never sleep | Scott Wu (Cognition CEO) · Scott Wu
“Absolutely yes, you should still learn to code. Most of what you're learning is the ability to logically break down problems.”
How Devin replaces your junior engineers with infinite AI interns that never sleep | Scott Wu (Cognition CEO) · Scott Wu
“Jevons Paradox says that as the price goes down, total spend can go up. As programming becomes more effective, we're going to have a lot more programmers.”
How Devin replaces your junior engineers with infinite AI interns that never sleep | Scott Wu (Cognition CEO) · Scott Wu
“Devin is best on tasks that are well-defined. Quick front-end features, bug fixes, testing — things easy to verify.”
How Devin replaces your junior engineers with infinite AI interns that never sleep | Scott Wu (Cognition CEO) · Scott Wu
“We call it a jagged intelligence. Devin can be almost like a staff engineer at understanding the code base.”
How Devin replaces your junior engineers with infinite AI interns that never sleep | Scott Wu (Cognition CEO) · Scott Wu
“If you were founding a new company from scratch with the same mission, how would you execute using a fully AI native approach?”
How we restructured Airtable’s entire org for AI | Howie Liu (co-founder and CEO) · Howie Liu
“As a PM, you need to start looking more like a hybrid PM prototyper, who has some good design sensibilities.”
How we restructured Airtable’s entire org for AI | Howie Liu (co-founder and CEO) · Howie Liu
“It's actually now hard to taste the soup without participating in creating it. To understand the solution space, you have to be in the details.”
How we restructured Airtable’s entire org for AI | Howie Liu (co-founder and CEO) · Howie Liu
“If you want to cancel all your meetings for a day or entire week and just go play around with every AI product, go do it.”
How we restructured Airtable’s entire org for AI | Howie Liu (co-founder and CEO) · Howie Liu
“We're approaching a world where the marginal cost of good output approaches zero. The org chart starts to become the work chart.”
How 80,000 companies build with AI: products as organisms, the death of org charts, and why agents will outnumber employees by 2026 | Asha Sharma (CVP of AI Platform at Microsoft) · Asha Sharma
“These models are living organisms that get better with more interactions. This is the new IP of every company — products that think and live and learn.”
How 80,000 companies build with AI: products as organisms, the death of org charts, and why agents will outnumber employees by 2026 | Asha Sharma (CVP of AI Platform at Microsoft) · Asha Sharma
“We think about what season are we in? A season can be denoted by secular changes in the industry. We have loose quarterly OKRs and 4-6 week goals.”
How 80,000 companies build with AI: products as organisms, the death of org charts, and why agents will outnumber employees by 2026 | Asha Sharma (CVP of AI Platform at Microsoft) · Asha Sharma
“With agents and products that reason, there's a new wave around reinforcement learning. I believe we will see as much money on post-training as pre-training.”
How 80,000 companies build with AI: products as organisms, the death of org charts, and why agents will outnumber employees by 2026 | Asha Sharma (CVP of AI Platform at Microsoft) · Asha Sharma
“We're the largest expert network in the world. The only moat in human data is access to an audience.”
Inside the expert network training every frontier AI model | Garrett Lord (Handshake CEO) · Garrett Lord
“The models have gotten so good that generalists are no longer needed. We have 500,000 PhDs, 3 million master students.”
Inside the expert network training every frontier AI model | Garrett Lord (Handshake CEO) · Garrett Lord
“Model builders care about three things: quality first, then volume, then speed.”
Inside the expert network training every frontier AI model | Garrett Lord (Handshake CEO) · Garrett Lord
“That's not what we're hearing from employers. Being AI native, young people are at a huge advantage.”
Inside the expert network training every frontier AI model | Garrett Lord (Handshake CEO) · Garrett Lord
“I said, we need to become a wartime company. I rewrote the values designed to be a sharp knife to cut out the parts of the company that wouldn't be effective.”
How Intercom rose from the ashes by betting everything on AI | Eoghan McCabe (founder and CEO) · Eoghan McCabe
“Fin is our AI agent who will pass 100 million ARR in less than three quarters.”
How Intercom rose from the ashes by betting everything on AI | Eoghan McCabe (founder and CEO) · Eoghan McCabe
“You don't have a choice. AI is going to disrupt in the most aggressive violent ways. If you're not in it, you're about to get kicked out.”
How Intercom rose from the ashes by betting everything on AI | Eoghan McCabe (founder and CEO) · Eoghan McCabe
“This is a pattern with AI, you won't know what to polish until after you ship. My dream is that we ship daily.”
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI) · Nick Turley
“If we're shipping a feature and it doesn't get 2X better as the model gets 2X smarter, it's probably not a feature we should be shipping.”
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI) · Nick Turley
“I started writing evals before I knew what an eval was because I was just outlining clearly specified ideal behavior.”
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI) · Nick Turley
“You're going to be polishing the wrong things in this space. You won't know what to polish until after you ship.”
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI) · Nick Turley
“We would have immense regret if you had a model that was state-of-the-art on health bench and didn't use that to help people.”
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI) · Nick Turley
“Curiosity is an attribute that we think matters so much more than your ML knowledge.”
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI) · Nick Turley
“If what AI cannot do is the human wisdom piece, is it possible that older managers with more emotional intelligence can be a value?”
Brian Chesky's secret mentor who died 9 times, started the Burning Man board, and built the world's first midlife wisdom school | Chip Conley (founder of MEA) · Chip Conley
“We are moving from the era of specialists to generalists. AI is accelerating this.”
Brian Chesky's secret mentor who died 9 times, started the Burning Man board, and built the world's first midlife wisdom school | Chip Conley (founder of MEA) · Chip Conley
“I think the PM of the future looks more like a product engineer. Someone who can think strategically but also build and test ideas rapidly. The line between PM and engineer is blurring.”
He saved OpenAI, invented the “Like” button, and built Google Maps: Bret Taylor on the future of careers, coding, agents, and more · Bret Taylor
“I think 50th percentile chance of hitting some kind of superintelligence is now like 2028.”
Benjamin Mann · Benjamin Mann
“We felt like safety wasn't the top priority there. The case for safety has gotten a lot more concrete.”
Benjamin Mann · Benjamin Mann
“Once we get to superintelligence, it will be too late to align the models. My best forecast for an extremely bad outcome is 0-10%.”
Benjamin Mann · Benjamin Mann
“In a world of abundance where labor is almost free, what do jobs even look like?”
Benjamin Mann · Benjamin Mann
“The talent wars are real. The expected value of a top researcher is just astronomical.”
Benjamin Mann · Benjamin Mann
“100% of our code is AI-written. Five products, seven-figure revenue. The quality bar has crossed the threshold where AI code is production-ready for most use cases.”
The AI-native startup: 5 products, 7-figure revenue, 100% AI-written code | Dan Shipper (co-founder/CEO of Every) · Dan Shipper
“The gap between demo and production is real. AI can write code fast, but making it reliable requires iteration. We spend about 30% of our time on polish and reliability.”
The AI-native startup: 5 products, 7-figure revenue, 100% AI-written code | Dan Shipper (co-founder/CEO of Every) · Dan Shipper
“It's like having the world's most reliable employee who costs $200 a month. Do all jobs just become a single prompt?”
I’ve run 75+ businesses. Here’s why you’re probably chasing the wrong idea. | Andrew Wilkinson (co‑founder of Tiny) · Andrew Wilkinson
“Studies have shown that using bad prompts can get you down to 0% on a problem, and good prompts can boost you up to 90%. People will always be saying, 'It's dead,' or, 'It's going to be dead with the next model version,' but then it comes out and it's not.”
AI prompt engineering in 2025: What works and what doesn’t | Sander Schulhoff (Learn Prompting, HackAPrompt) · Sander Schulhoff
“My best advice on how to improve your prompting skills is actually just trial and error. But if there were one technique that I could recommend, it is few-shot prompting, which is just giving the AI examples of what you want it to do.”
AI prompt engineering in 2025: What works and what doesn’t | Sander Schulhoff (Learn Prompting, HackAPrompt) · Sander Schulhoff
“Role prompting does not work. On those older models, maybe it worked. On the more modern ones, it doesn't help at all for accuracy-based tasks. But giving a role really helps for expressive tasks, writing tasks, summarizing tasks.”
AI prompt engineering in 2025: What works and what doesn’t | Sander Schulhoff (Learn Prompting, HackAPrompt) · Sander Schulhoff
“Decomposition is another really effective technique. You say, 'Hey, don't answer this. Before answering it, tell me what are some subproblems that would need to be solved first?' And then you ask it to solve each of those subproblems one by one.”
AI prompt engineering in 2025: What works and what doesn’t | Sander Schulhoff (Learn Prompting, HackAPrompt) · Sander Schulhoff
“I have a 30 days of GPT list of 30 things to do, one every single day. I don't know anyone who has gone through this and not come out the other side feeling a hundred times more confident in their skills. It's built as a habit formation tool, not an education tool.”
How to build a team that can “take a punch”: A playbook for building resilient, high-performing teams | Hilary Gridley (Head of Core Product, Whoop) · Hilary Gridley
“I've very explicitly been using AI as a life coach. One prompt was, 'What's an outdated mindset that I'm holding onto that's not still serving me?' It came back with a very polite response about being wedded to the idea of control. It suggested: try to focus on choreography over control.”
35 years of product design wisdom from Apple, Disney, Pinterest, and beyond | Bob Baxley · Bob Baxley
“When you show up with an iPhone, you're thinking about sharing. When you show up with a film camera, you're thinking about saving film. When you show up with a digital SLR, you just take a whole bunch of pictures. Be conscious about how the tools you pick are going to impact the thing that you produce.”
35 years of product design wisdom from Apple, Disney, Pinterest, and beyond | Bob Baxley · Bob Baxley
“There's nothing worse than a brilliant image of a fuzzy concept. We live in a time when it's very easy to produce things at incredibly high production values, but they don't mean anything. They're just fancy potato chips. There's no nourishment there.”
35 years of product design wisdom from Apple, Disney, Pinterest, and beyond | Bob Baxley · Bob Baxley
“The team that works in the most futuristic way is the Claude Code team. They're using Claude Code to build Claude Code in a very self-improving kind of way. Over half of our pull requests are Claude Code generated. Probably at this point it's over 70%.”
Anthropic’s CPO on what comes next | Mike Krieger (co-founder of Instagram) · Mike Krieger
“Our goal with Cursor is to invent a new type of programming, a very different way to build software. A world kind of after code. More and more being an engineer will start to feel like being a logic designer, and it will be about specifying your intent.”
The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO) · Michael Truell
“We definitely didn't expect to be doing any of our own model development when we started. And at this point, every magic moment in Cursor involves a custom model in some way.”
The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO) · Michael Truell
“I truly just think that the ceiling is so high that no matter what entrenchment you build, you can be leapfrogged. It is incumbent on us to continue to try to build the best thing.”
The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO) · Michael Truell
“I would bias less toward trying in one go to tell the model exactly what to do. Instead chop things up into bits. You're specifying a little bit, getting a little bit of work. At the same time, explicitly try to fall on your face and discover the limits of what the models can do.”
The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO) · Michael Truell
“We're in the middle of a technology shift that's going to be more consequential than the internet. And I think it's going to take a while, it's going to be a multi-decade thing, and many different groups will be consequential in pushing it forward.”
The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO) · Michael Truell
“We should be cannibalizing the existing state of our product every six to 12 months. Every six to 12 months, it should make our existing product look silly. It should almost make the form factor of existing product look dumb.”
Building a magical AI code editor used by over 1 million developers in four months: The untold story of Windsurf | Varun Mohan (co-founder and CEO) · Varun Mohan
“We tasked everyone at our company to build an app with Windsurf. We've saved over half a million dollars of SaaS products we were going to buy because our go-to-market team has now built apps instead of buying them.”
Building a magical AI code editor used by over 1 million developers in four months: The untold story of Windsurf | Varun Mohan (co-founder and CEO) · Varun Mohan
“People could be more full stack. Imagine a designer that can ship a fully baked product, a product manager that can prototype and ship to production. We shouldn't put limits on ourselves and what we can build.”
Everyone’s an engineer now: Inside v0’s mission to create a hundred million builders | Guillermo Rauch (founder and CEO of Vercel, creators of v0 and Next.js) · Guillermo Rauch
“Now we have 600 engineers. Some of the best things created with v0 have not come from our engineering team. They've come from the marketing team, the sales team, the product management team. The product management team is fascinating, because now they're actually building the product.”
Everyone’s an engineer now: Inside v0’s mission to create a hundred million builders | Guillermo Rauch (founder and CEO of Vercel, creators of v0 and Next.js) · Guillermo Rauch
“You can be as ambitious as you want. If you have technical skills, have some suspension of disbelief. Focus more on the product description, what do you want the end user to experience. Try to be open-minded about how well the tool can implement it.”
Everyone’s an engineer now: Inside v0’s mission to create a hundred million builders | Guillermo Rauch (founder and CEO of Vercel, creators of v0 and Next.js) · Guillermo Rauch
“Think about LLMs as Oracles that can write software for you, but there's a limit. It's not writing the compiler from scratch. The engineers that learn foundational infrastructure are probably going to be extremely empowered still, for years to come.”
Everyone’s an engineer now: Inside v0’s mission to create a hundred million builders | Guillermo Rauch (founder and CEO of Vercel, creators of v0 and Next.js) · Guillermo Rauch
“The AI models that you're using today is the worst AI model you will ever use for the rest of your life. Every two months, computers can do something they've never been able to do before and you need to completely think differently about what you're doing.”
OpenAI’s CPO on how AI changes must-have skills, moats, coding, startup playbooks, more | Kevin Weil (CPO at OpenAI, ex-Instagram, Twitter) · Kevin Weil
“I actually think chat is an amazing interface because it's so versatile. It is the way we talk. If I had some more rigid interface, I would be able to speak to you about far fewer things and it would get in the way of maximum communication bandwidth.”
OpenAI’s CPO on how AI changes must-have skills, moats, coding, startup playbooks, more | Kevin Weil (CPO at OpenAI, ex-Instagram, Twitter) · Kevin Weil
“Sonnet was really the first model that flipped the equation. We actually tried building Bolt almost exactly a year ago, with the frontier models at the time. Spent a week or two building it. It just didn't work. The output, the code output was not reliable enough. And then we got a sneak peek of the Sonnet stuff in May and we were like, 'Oh. Okay, we should take that project back off the shelf.'”
Inside Bolt: From near-death to ~$40m ARR in 5 months—one of the fastest-growing products in history | Eric Simons (founder and CEO of StackBlitz) · Eric Simons
“Software is deterministic. When you write code and you hit run, it either runs or it doesn't. And that's the key insight Anthropic really had. They just went deep. This is what they're doing, is just reinforcement learning on basically permutating every type of app you could ever build. I'm extremely bullish. It makes technical sense why, of anything, LLMs are going to get insanely better at writing code than probably most other types of applications.”
Inside Bolt: From near-death to ~$40m ARR in 5 months—one of the fastest-growing products in history | Eric Simons (founder and CEO of StackBlitz) · Eric Simons
“I would say, talk to this thing like you do a Linear ticket, or a JIRA ticket. That would be my advice. Talk to this like you're talking to one of the developers on your team. Be specific on things that matter. And on things where you can let it be creative, you can just say, 'Hey, make it prettier.' And it does a really good job when you give it just, vibes.”
Inside Bolt: From near-death to ~$40m ARR in 5 months—one of the fastest-growing products in history | Eric Simons (founder and CEO of StackBlitz) · Eric Simons
“On any Figma URL, when you're looking at a design that you've made, if you just put bolt.new in front of that URL and hit enter, it's going to suck that design into Bolt, and turn it into a full stack app or mobile app, just out of the box.”
Inside Bolt: From near-death to ~$40m ARR in 5 months—one of the fastest-growing products in history | Eric Simons (founder and CEO of StackBlitz) · Eric Simons
“Lovable is your personal AI software engineer. You describe an idea and then you get a fully working product. The reason is to enable those who have had such a hard time finding people who are good at creating software that's been their absolute bottleneck and let them take their ideas and their dreams into reality.”
Building Lovable: $10M ARR in 60 days with 15 people | Anton Osika (co-founder and CEO) · Anton Osika
“I think the number one most important skill when AI can do more and more is taste, like having really good taste for what to build and what is a great product. And so in some ways, the people that are like the best product people, like you, I actually think they're going to become more and more important over time.”
Building Lovable: $10M ARR in 60 days with 15 people | Anton Osika (co-founder and CEO) · Anton Osika
“Just start simple and then what you get is that the AI says, okay, I can go through what does a beautiful Airbnb clone look like and it goes through a bit of design decisions and then I'll zoom out to see more of it. We have this just UI that is... I mean it has all the nice things you would expect from Airbnb clone where you see different categories and you can see two listings from Airbnb with login buttons and everything.”
Building Lovable: $10M ARR in 60 days with 15 people | Anton Osika (co-founder and CEO) · Anton Osika
“I think if you spend a full week on trying to reach an outcome, the best way to learn is I want to do this thing and then I want to use AI to do that thing. And you've spent a full week, you are in the top 1% in the global population. And if you surround yourself with friends who have this obsession or they also care a lot about this, then you'd be quickly in the top 0.1%.”
Building Lovable: $10M ARR in 60 days with 15 people | Anton Osika (co-founder and CEO) · Anton Osika
“Everyone at Shopify has to be able to build something with AI. If you're not learning how to use these tools, you're falling behind. We made this a company-wide expectation.”
Tobi Lutke · Tobi Lutke
“I use AI to essentially do a voice-of-the-customer pipeline automatically. It processes all customer calls, tags them with themes, and then I can go pull up exactly what people said about a given topic.”
Unorthodox PM wisdom: Automating user insights, unselling job candidates, logging every decision, more | Kevin Yien (Stripe, Square, Mutiny) · Kevin Yien
“There's this tool called websim where you can just type in a URL that doesn't exist and it will generate the entire website for you. It's wild. I think it shows you where things are going.”
Dylan Field live at Config: Intuition, simplicity, and the future of design · Dylan Field
Cutting Room Floor
Guest insights on this topic that Lenny hasn't (yet) written about in his newsletters. Potential material for future posts.
“It's not just designers who are feeling like, 'Oh yeah, we have to keep up with engineers.' I think even engineers are like, 'How do we keep up with ourselves?'”
Jenny Wen · Jenny Wen
“You have to know the difference between a megatrend and a hype cycle. When there's a megatrend, don't fight it. AI is a megatrend, one of the most foundational movements that we have seen in human history.”
Jeetu Patel · Jeetu Patel
“The one person billion dollar startup implies that to enable it, there might be a hundred other small startups building bespoke software. We might actually enter into a golden age of B2B SaaS.”
Sherwin Wu V2 · Sherwin Wu V2
“We live in a bubble in Silicon Valley. A lot of work that runs our entire economy is not open-ended knowledge work — it's repeatable business processes with standard operating procedures.”
Sherwin Wu V2 · Sherwin Wu V2
“Seed VCs who invest in companies branded as AI companies are going to lose an impossibly large amount of money. There's a difference between a great business using AI versus an 'AI company.'”
How to show up in any room with a low heart rate: Silicon Valley’s missing etiquette playbook | Sam Lessin · Sam Lessin
“An RL environment is essentially a simulation of the real world. We might build a world where you have a startup with Gmail, Slack, Jira, GitHub. And then suddenly AWS goes down. Model, what do you do?”
The 100-person AI lab that became Anthropic and Google's secret weapon | Edwin Chen (Surge AI) · Edwin Chen
“Do you want a model that says 'There are definitely 20 more ways to improve this email' or one that's optimizing for your time and just says 'Your email's great, just send it'?”
The 100-person AI lab that became Anthropic and Google's secret weapon | Edwin Chen (Surge AI) · Edwin Chen
“Whatever AI does, currently or in the future, is up to us. Every technology is a double-edged sword.”
The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li · Dr. Fei Fei Li
“We curated 15 million images on the internet, created a taxonomy of 22,000 concepts. That combination of big data, neural network, and GPU was the golden recipe for modern AI.”
The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li · Dr. Fei Fei Li
“In the middle of 2015, 2016, some tech companies avoided using the word AI because they were not sure if AI was a dirty word.”
The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li · Dr. Fei Fei Li
“I realized the North Star challenge of AI is visual intelligence. Our human brains use about 50% of our neurons on visual processing. This is fundamentally a big data problem.”
The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li · Dr. Fei Fei Li
“AI is expansionary. There's actually just more and more questions being asked and curiosity that can be fulfilled now with AI.”
Inside Google's AI turnaround: The rise of AI Mode, strategy behind AI Overviews, and their vision for AI-powered search | Robby Stein (VP of Product, Google Search) · Robby Stein
“These things take 6 to 12 months to get truly robust. Like with any major tech revolution, headlines tell one story and on the ground, you need to dig up every road.”
First interview with Scale AI’s CEO: $14B Meta deal, what’s working in enterprise AI, and what frontier labs are building next | Jason Droege · Jason Droege
“18 months ago, you would get a short story. Now one task is building an entire website by one of the world's best developers. These tasks now take hours and require PhDs.”
First interview with Scale AI’s CEO: $14B Meta deal, what’s working in enterprise AI, and what frontier labs are building next | Jason Droege · Jason Droege
“If you have a human process that is 10 or 20% accurate, AI is awesome. If you have one at 98% accurate and expect the remaining 2%, not totally there yet.”
First interview with Scale AI’s CEO: $14B Meta deal, what’s working in enterprise AI, and what frontier labs are building next | Jason Droege · Jason Droege
“A document that reads the same words in company A will have a different meaning in company B. Digitizing judgment is becoming a bottleneck.”
First interview with Scale AI’s CEO: $14B Meta deal, what’s working in enterprise AI, and what frontier labs are building next | Jason Droege · Jason Droege
“We call it a jagged intelligence. Devin can be almost like a staff engineer at understanding the code base.”
How Devin replaces your junior engineers with infinite AI interns that never sleep | Scott Wu (Cognition CEO) · Scott Wu
“If you were founding a new company from scratch with the same mission, how would you execute using a fully AI native approach?”
How we restructured Airtable’s entire org for AI | Howie Liu (co-founder and CEO) · Howie Liu
“It's actually now hard to taste the soup without participating in creating it. To understand the solution space, you have to be in the details.”
How we restructured Airtable’s entire org for AI | Howie Liu (co-founder and CEO) · Howie Liu
“These models are living organisms that get better with more interactions. This is the new IP of every company — products that think and live and learn.”
How 80,000 companies build with AI: products as organisms, the death of org charts, and why agents will outnumber employees by 2026 | Asha Sharma (CVP of AI Platform at Microsoft) · Asha Sharma
“With agents and products that reason, there's a new wave around reinforcement learning. I believe we will see as much money on post-training as pre-training.”
How 80,000 companies build with AI: products as organisms, the death of org charts, and why agents will outnumber employees by 2026 | Asha Sharma (CVP of AI Platform at Microsoft) · Asha Sharma
“We're the largest expert network in the world. The only moat in human data is access to an audience.”
Inside the expert network training every frontier AI model | Garrett Lord (Handshake CEO) · Garrett Lord
“The models have gotten so good that generalists are no longer needed. We have 500,000 PhDs, 3 million master students.”
Inside the expert network training every frontier AI model | Garrett Lord (Handshake CEO) · Garrett Lord
“Model builders care about three things: quality first, then volume, then speed.”
Inside the expert network training every frontier AI model | Garrett Lord (Handshake CEO) · Garrett Lord
“I said, we need to become a wartime company. I rewrote the values designed to be a sharp knife to cut out the parts of the company that wouldn't be effective.”
How Intercom rose from the ashes by betting everything on AI | Eoghan McCabe (founder and CEO) · Eoghan McCabe
“You don't have a choice. AI is going to disrupt in the most aggressive violent ways. If you're not in it, you're about to get kicked out.”
How Intercom rose from the ashes by betting everything on AI | Eoghan McCabe (founder and CEO) · Eoghan McCabe
“If we're shipping a feature and it doesn't get 2X better as the model gets 2X smarter, it's probably not a feature we should be shipping.”
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI) · Nick Turley
“We would have immense regret if you had a model that was state-of-the-art on health bench and didn't use that to help people.”
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI) · Nick Turley
“Curiosity is an attribute that we think matters so much more than your ML knowledge.”
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI) · Nick Turley
“We are moving from the era of specialists to generalists. AI is accelerating this.”
Brian Chesky's secret mentor who died 9 times, started the Burning Man board, and built the world's first midlife wisdom school | Chip Conley (founder of MEA) · Chip Conley
“I think 50th percentile chance of hitting some kind of superintelligence is now like 2028.”
Benjamin Mann · Benjamin Mann
“We felt like safety wasn't the top priority there. The case for safety has gotten a lot more concrete.”
Benjamin Mann · Benjamin Mann
“Once we get to superintelligence, it will be too late to align the models. My best forecast for an extremely bad outcome is 0-10%.”
Benjamin Mann · Benjamin Mann
“In a world of abundance where labor is almost free, what do jobs even look like?”
Benjamin Mann · Benjamin Mann
“The talent wars are real. The expected value of a top researcher is just astronomical.”
Benjamin Mann · Benjamin Mann
“It's like having the world's most reliable employee who costs $200 a month. Do all jobs just become a single prompt?”
I’ve run 75+ businesses. Here’s why you’re probably chasing the wrong idea. | Andrew Wilkinson (co‑founder of Tiny) · Andrew Wilkinson
“Studies have shown that using bad prompts can get you down to 0% on a problem, and good prompts can boost you up to 90%. People will always be saying, 'It's dead,' or, 'It's going to be dead with the next model version,' but then it comes out and it's not.”
AI prompt engineering in 2025: What works and what doesn’t | Sander Schulhoff (Learn Prompting, HackAPrompt) · Sander Schulhoff
“My best advice on how to improve your prompting skills is actually just trial and error. But if there were one technique that I could recommend, it is few-shot prompting, which is just giving the AI examples of what you want it to do.”
AI prompt engineering in 2025: What works and what doesn’t | Sander Schulhoff (Learn Prompting, HackAPrompt) · Sander Schulhoff
“Role prompting does not work. On those older models, maybe it worked. On the more modern ones, it doesn't help at all for accuracy-based tasks. But giving a role really helps for expressive tasks, writing tasks, summarizing tasks.”
AI prompt engineering in 2025: What works and what doesn’t | Sander Schulhoff (Learn Prompting, HackAPrompt) · Sander Schulhoff
“Decomposition is another really effective technique. You say, 'Hey, don't answer this. Before answering it, tell me what are some subproblems that would need to be solved first?' And then you ask it to solve each of those subproblems one by one.”
AI prompt engineering in 2025: What works and what doesn’t | Sander Schulhoff (Learn Prompting, HackAPrompt) · Sander Schulhoff
“I've very explicitly been using AI as a life coach. One prompt was, 'What's an outdated mindset that I'm holding onto that's not still serving me?' It came back with a very polite response about being wedded to the idea of control. It suggested: try to focus on choreography over control.”
35 years of product design wisdom from Apple, Disney, Pinterest, and beyond | Bob Baxley · Bob Baxley
“We definitely didn't expect to be doing any of our own model development when we started. And at this point, every magic moment in Cursor involves a custom model in some way.”
The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO) · Michael Truell
“I truly just think that the ceiling is so high that no matter what entrenchment you build, you can be leapfrogged. It is incumbent on us to continue to try to build the best thing.”
The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO) · Michael Truell
“We're in the middle of a technology shift that's going to be more consequential than the internet. And I think it's going to take a while, it's going to be a multi-decade thing, and many different groups will be consequential in pushing it forward.”
The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO) · Michael Truell
“We should be cannibalizing the existing state of our product every six to 12 months. Every six to 12 months, it should make our existing product look silly. It should almost make the form factor of existing product look dumb.”
Building a magical AI code editor used by over 1 million developers in four months: The untold story of Windsurf | Varun Mohan (co-founder and CEO) · Varun Mohan
“Software is deterministic. When you write code and you hit run, it either runs or it doesn't. And that's the key insight Anthropic really had. They just went deep. This is what they're doing, is just reinforcement learning on basically permutating every type of app you could ever build. I'm extremely bullish. It makes technical sense why, of anything, LLMs are going to get insanely better at writing code than probably most other types of applications.”
Inside Bolt: From near-death to ~$40m ARR in 5 months—one of the fastest-growing products in history | Eric Simons (founder and CEO of StackBlitz) · Eric Simons
“Just start simple and then what you get is that the AI says, okay, I can go through what does a beautiful Airbnb clone look like and it goes through a bit of design decisions and then I'll zoom out to see more of it. We have this just UI that is... I mean it has all the nice things you would expect from Airbnb clone where you see different categories and you can see two listings from Airbnb with login buttons and everything.”
Building Lovable: $10M ARR in 60 days with 15 people | Anton Osika (co-founder and CEO) · Anton Osika
“I think if you spend a full week on trying to reach an outcome, the best way to learn is I want to do this thing and then I want to use AI to do that thing. And you've spent a full week, you are in the top 1% in the global population. And if you surround yourself with friends who have this obsession or they also care a lot about this, then you'd be quickly in the top 0.1%.”
Building Lovable: $10M ARR in 60 days with 15 people | Anton Osika (co-founder and CEO) · Anton Osika
“Everyone at Shopify has to be able to build something with AI. If you're not learning how to use these tools, you're falling behind. We made this a company-wide expectation.”
Tobi Lutke · Tobi Lutke
“I use AI to essentially do a voice-of-the-customer pipeline automatically. It processes all customer calls, tags them with themes, and then I can go pull up exactly what people said about a given topic.”
Unorthodox PM wisdom: Automating user insights, unselling job candidates, logging every decision, more | Kevin Yien (Stripe, Square, Mutiny) · Kevin Yien