·
Finding signal on Twitter is more difficult than it used to be. We curate the best tweets on topics like AI, startups, and product development every weekday so you can focus on what matters.
PLUS: DeepSeek's model theft exposed
Naval says AI is turning the Internet into the Library of Babel, infinite content with no signal.
Anthropic caught DeepSeek, Moonshot, and MiniMax running 24,000 fake accounts to steal Claude's capabilities through 16 million exchanges.
Naval calls it: vibe coding is the new product management. English is now a programming language for building apps.
Simon Willison published a guide to agentic engineering patterns for getting better results from Claude Code and similar tools.
Greg Isenberg's blueprint: 3 person teams building $100M companies by automating workflows and pricing outcomes, not seats.
Cursor just made all OpenAI models 30% faster by upgrading everyone to WebSockets with Responses API.
OpenClaw's security researcher says stop treating it like a multi-user bus. It's designed for one user, many agents.
Theo claims you should delete your CLAUDE.md file and he's got a study to back it up.
Anthropic built an AI Fluency Index tracking 11 collaboration behaviors across thousands of Claude conversations to measure effectiveness.
Guillermo Rauch clocked LinkedIn at 11 seconds to load a skeleton-free profile. Someone build FastIn already.
Proliferate is the open source platform for company specific coding agents that can actually do end-to-end work.
DHH on open source: don't wish for lighter responsibility, become stronger. The burden is the reward.
Mailsign got acquired after being vibe coded in days by someone who couldn't code three months ago. First internet money.
Theo's paying $250 monthly and Gemini 3.1 Pro broke his code then stopped working due to capacity issues.
Jack Friks never hits his Claude Code limit because he uses two agents max and reviews every diff.
Ethan Mollick warns: collect your hard problems and good ideas now. Agency without direction isn't valuable.
Confluence Labs hit 97.9% on ARC-AGI-2, focusing on learning efficiency where data is sparse and experiments cost money.
Return Signals tackles the $300B returns problem in online fashion with proactive high touch support for brands.
Dan Shipper on investment bankers: they just discovered Claude Code and think SaaS is dead. Everyone in tech yawns.
Sebastian Raschka dug in and found SWE-Bench Verified has flawed tests that reject 59% of correct solutions.
Turn hours of scrolling into a five minute read.