Rauchg declares YES-CODE: an entire category of software, 'no-code', was built under the presumption that code is expensive, difficult, and scarce. Coding agents have forever changed the equation. Code is now cheap, easy, and abundant.

He recalls @cramforce being asked by an analyst if Vercel is a no-code platform. Without hesitation: 'No, it's the absolute opposite. It's a yes-code platform.' No-code took shortcuts and set hard ceilings on performance and sophistication.

Vercel's mission: create the easiest cloud for agents that you never graduate from. 462 likes.

Peter Yang argues that narrow-use-case SaaS is getting harder to monetize because: 1) AI skills can solve the same problem more flexibly and personally, 2) AI-native agents like Codex/Claude Code have personal context and memory that a standalone SaaS or chatbot doesn't, 3) People will pay hundreds or thousands for services (human touch is rare) but compare a $20/month SaaS to their Claude/ChatGPT subscription.

His honest caveat: 'I guess I'm in a bubble and most people have not set up their own AI skills yet.'

TRQ declares Workflows the biggest upgrade to Claude Code's capabilities since skills and subagents. He dove deep with Sid Bhide to figure out best practices and examples—particularly excited about the non-technical tasks it enables for Claude Code. 3339 likes.

This is a major signal: Claude Code is expanding beyond coding into general-purpose automation with structured workflow support.

Steipete announces the privilege of working with Microsoft to bring OpenClaw to enterprises. 3413 likes—the second-highest engagement of the day.

This signals that agent platforms are now ready for enterprise distribution through major cloud partnerships.

Thibault Sottiaux (OpenAI) announces Codex updates: business plan users can now host and share websites, vastly improved plugins and skills for broad roles, and visual annotations to give feedback to your agent in docs, slides, sheets and more. 1416 likes.

The annotation feature is notable: it's a new interaction modality for agent feedback beyond text.

Zara Zhang cites OpenAI's latest Codex report: knowledge workers now represent about 20% of Codex users and are adopting it more than 3x as fast as developers. Fastest growing task types for knowledge workers: Data Analysis (110% growth week over week), Research (+37%), and Knowledge Artifacts (+36%).

This is the signal that AI tools are breaking out of the developer niche into the broader knowledge worker economy.

Levie predicts that as token budgets take on a larger part of operating expenses, model routing is the inevitable conclusion. By understanding different work patterns in your domain and having strong evals, you'll be able to cost/performance optimize effectively.

We're still at the point where most use-cases need frontier performance, but soon you'll peel off individual use-cases to lower-cost models. Products that can intelligently route workflows to the right tier of model will aggregate more demand. 299 likes.

Rauchg frames education in the age of AI: start with the language—the linguistic surface is your roadmap. Much like an SDK has function definitions, human language is the new API to the world.

His deeper point: there's never been an expert in any field without mastery over its language. The distinction is that English alone could not produce tangible things before. You had to translate it into machine instruction through learning or delegation. You can now go direct. 765 likes.