Today I finished setting up blogwatcher inside OpenClaw, adding a few information sources per Claude’s suggestion (AI, startups, and technical-community directions).
Not long after it was configured, it pushed through a piece of English content. I didn’t click through to the original link, but based on the structure and tone, it’s almost certainly from Hacker News or the YC (Y Combinator) hiring/startup feed — the whole piece was a textbook example of a startup pitch plus a technical-founder job listing, leaning hard on AI agents, product ownership, and PMF — pure YC-house-style language.
What blogwatcher is actually doing here became clear pretty quickly: it’s not a plain RSS reader — it’s an AI-driven filter for information. Not just “grab articles,” but source → AI filtering → recommendation → reading → into my knowledge stream. What I’m really building is a system for automatically surfacing high-value information.
What was that job listing, anyway?#
The article was a YC startup hiring for a role with an interesting title: “Ex Technical Founder.” In plain terms: they want someone who’s started a company before, can write code, and can build a product with AI. Not an algorithms role, not a regular engineering role — a new kind of position entirely.
What’s the pay like?#
The listing offered £90K–£180K a year, plus 0.10%–0.40% equity, based in London. Put in context: £90K is already senior-engineer territory, £180K sits at the high end for AI engineers, and that equity range is close to what a founding engineer gets. In other words, this isn’t a regular job — it’s closer to a salaried co-founder position.
What kind of person are they actually looking for?#
The key point: this role isn’t about building AI models — it’s about using AI to build systems and products. The core skills: building AI agents, designing workflows, using RAG and context systems to give a model something like memory, wiring AI into business decisions, and shipping products and validating them with real data.
That’s no longer “write code → ship a feature.” It’s “find the problem → build the system → solve it with AI → ship and validate → iterate.” The role, at its core, is an AI product engineer — an AI builder.
A trend that keeps getting more obvious#
The old core of software engineering was writing code and shipping features. It’s turning into designing AI workflows, designing agents, building automated systems, and turning AI into leverage. In other words: knowing how to build systems with AI is, on its own, already becoming a real job category — and probably the one in highest demand going forward. There aren’t many people doing pure model research; what’s genuinely in high demand is people who can turn AI into a shipped product.
Closing thought#
Today’s small discovery confirmed something for me: the skill that actually matters in the AI era isn’t “knowing how to use AI” — it’s “being able to build systems with AI.” That’s already become a job category, and the demand for it is growing fast.
