your timeline convinced you AI is in a bubble. talk to a boomer above the age 35 for 5 minutes. most people don’t even know what claude is.​​​​​​​​​​​​​​​​ kind of wild when you zoom out.

This is weird for me to write, but I think our sprint methods are the best way to work in a world that didn't exist when they were created. Let me explain... When @jakek created the Design Sprint in 2010, the cost of building software was very high. Part of what made his method valuable — and why we embraced it at GV starting in 2012 — was that it allowed teams to run in Plan Mode before executing (yeah that's a Claude Code reference!) Today, the cost of building software is dramatically lower. It's not as low as people on here say, but it's objectively way faster and cheaper to ship software today than it was in 2010. Today, we find that our sprints (the Design Sprint plus new methods like the Foundation Sprint) have taken on a different and likely more important role for teams building software. ❗️Sprints are no longer about reducing uncertainty in the face of high costs. ‼️ Today they are about deciding what to do and how to stand out when the cost of building moves toward zero. I'm seeing this idea pop up everywhere. Builders now emphasize thoughtful planning and definition before setting agents to work. Design and product are more important than ever... but the jobs have also fundamentally changed. @lennysan says that writing PRDs is now the most important technical skill. Everything has changed since 2010. Yet in other ways, nothing has. It's eerie. It's still important for founders — maybe more important than ever — to make good decisions based on good information, then move quickly to validate their hypotheses. Sure, the reasons have changed: it's no longer because of high costs; it's because of low costs. But this work is essential, and despite all the advice out there about WHAT to do (use Markdown this way, set up your environment that way, use agents for this, etc), there are still few frameworks for HOW to do this planning right. Except, of course, for our sprints :) That's why Lenny called them "the missing manual" last year. Which is kinda wild. And there's one more challenge: When it's cheaper for everyone to build, what happens to competition? Yeah, it explodes. If you want to be successful — capture attention, solve a real problem, be reliable — I think you need to leverage opinionated, incisive human thought. That's always been a part of our sprints. By breaking complex decisions into concrete steps, creating focused work time, and working alone together, we help teams get to crisper, better, differentiated perspectives on what to build and how to talk about it. So... I guess this whole post just turned into "yay us"! LOL But seriously... I'm posting this because I believe more than ever in the value of working this way — and continue to see it be absurdly effective for teams building new products. And also, I'm just kind of amazed. It's not often that a tool built for one era turns out to be even more valuable in another. What do you think? What else has proved surprisingly valuable in this wild new world?

Introducing Exa Instant: the first sub-200ms search engine. Faster than Google, it's custom built to power realtime AI products like chat and voice.

The era of the narrow specialist is over. Marc Andreessen predicts the future belongs to the generalist. In the AI age, founders can’t excel in just one domain. They need competence across 6 to 8 different fields. Deep expertise still has its place. But broad knowledge combined with AI leverage is the new superpower.