Kevin Weil (VP Science @OpenAI): Notes that AI is solving open problems with increasingly elegant proofs as models improve. https://x.com/kevinweil/status/2039200605672284572
Box CEO Aaron Levie: Argues the ultimate rate limiter on productivity gains from agents will be critical enterprise concerns like security, compliance, governance, and the ability to review agent work. Companies won’t take on every risk—there’s no free lunch with AI productivity. https://x.com/levie/status/2039216522028257549
Amjad Masad (CEO @replit): Shared excitement about a new release and noted a “new psyop just dropped.” https://x.com/amasad/status/2039214420585492863 https://x.com/amasad/status/2039194288777478259
Thariq (Claude Code @anthropicai): Posted about getting a SNAIL and shared a /buddy post that gained significant engagement (1814 likes). https://x.com/trq212/status/2039201699907469344 https://x.com/trq212/status/2039201498996035924
Peter Steinberger (OpenClaw): Commented on takedown requests with engagement (300 likes). https://x.com/steipete/status/2039156882041123035
Nan Yu (Head of Product @linear): Shared “Still true” post and a meme about apologizing to DHH. https://x.com/thenanyu/status/2039169970278400157 https://x.com/thenanyu/status/2039167941225111754
Dan Shipper (CEO @every): Recommends trying Proof. https://x.com/danshipper/status/2039149696166318320
Aditya Agarwal (GP @SouthPkCommons, Ex-Facebook/Dropbox): Reflects on Facebook’s early “fix bugs in production” philosophy and wonders if the modern version is public source code. https://x.com/adityaag/status/2039198010681565665
Swyx: Notes no Codex growth in March and links to data. https://x.com/swyx/status/2039222156815716437
Garry Tan (President & CEO @ycombinator): Posted about “boiling an ocean” and shared some memes. https://x.com/garrytan/status/2039240608435572881 https://x.com/garrytan/status/2039237008338595850 https://x.com/garrytan/status/2039235860143739070
Data Driven NYC — “Can One Person Build a Billion-Dollar Startup?”
The Takeaway: Andrew from General Intelligence Company of New York argues that AI agents have finally matured enough to enable one-person billion-dollar companies, and he’s building the “middle management layer” to orchestrate specialized agents.
Andrew founded General Intelligence to enable the one-person billion-dollar company—organizations that run themselves with AI. After launching Co-founder (their first product in September), which got 8,000 startups using it to automate operations, they asked: an assistant isn’t a company, so what’s next?
His framework for a product-based company: create product → sell it → support it → operate. Their breakthrough was realizing you need a management layer—a manager agent that oversees coding agents, browser agents, and infrastructure agents. “We call this super optimization.”
Their new product, Co-founder CTO, is the first fully autonomous engineering department—all in the cloud. It plans and delegates to coding agents (like Claude Code), uses Versel preview environments, monitors deployments, writes testing plans, runs browser agents to test, merges automatically, and checks infrastructure. They’ve seen 400% speed improvement internally, averaging 25 PRs per engineer per day. Sometimes they spend more on tokens than salaries—in one day, $4,000 on Opus tokens.
“We’re starting to shift our human capital to intelligence.” In the next 6 months, they’re adding customer support and full revenue/sales stack.
https://youtube.com/watch?v=rX3CxlGm-00
Generated through the Follow Builders skill: https://github.com/zarazhangrui/follow-builders