Your Startup Idea Is Their Weekend Holiday

Andreas Klinger — former CTO of Product Hunt, now running PROTOTYPE fund — lays out the uncomfortable math for SaaS founders: the same tools that make building software trivially easy also make your product trivially replaceable. Your great SaaS idea? It's Christopher Janz's weekend project now.

mindmap
    root((SaaS in the AI Era))
        The Old World
            SaaS vs Gmail and Spreadsheets
            Laziness to adopt was the barrier
        The New World
            AI can one-shot core functionality
            Personalized software replaces generic SaaS
            Software that exists for 30 seconds
        Three Founder Paths
            Fun / Indie Hacking
                Build for joy and learning
                Monetize with one-off payments
                Your idea is their weekend project
            Company / Revenue
                Enterprise vertical SaaS
                Deep domain knowledge as moat
                Trust and decisions compound
            VC / Unicorn
                Build for where models will be in 18 months
                Surface area is the new moat
        Opportunities
            GitHub is obsolete
                PR review broken for AI code
                Open source security crisis
            Mexican standoff between roles
                Every role thinks it can replace the others
            Adoption metrics
                Who 10xed, who could, who never will
            Prompt injection
                Bigger than SQL injection
            Vibe-coded app governance
                Securing and standardizing ad-hoc apps

Key Takeaways

• Code is no longer the moat. The old world was SaaS vs. your Gmail-and-spreadsheet hack. Now it's SaaS vs. ChatGPT building the core functionality in 5 minutes. Toby Lütke of Shopify asked Claude to build him a medical scan viewer rather than install the doctor's Windows app. That's the competitive landscape now.
• Three paths, three strategies. Klinger breaks the founder decision into (1) build for fun as an indie hacker — ship fast, charge one-off, embrace the tinkerer community; (2) build a company that makes money — go enterprise vertical, stack domain knowledge and trust as your moat (he calls it "build Legora for your industry"); (3) chase the VC unicorn — which requires betting on where models will be in 18 months, not where they are today.
• You're no longer paid for code, you're paid for trade-off decisions. This was always true for senior devs, but now it's the entire job description. Knowledge work shifts from "I know things" to "I decide things, then AI builds."
• Open source is in crisis. Dozens of AI-generated PRs from people who haven't read your docs, an explosion of security vulnerabilities, and business models collapsing (hosting can be one-shotted, Tailwind-style module ecosystems can be replicated by autocomplete). If you can solve any part of this — monetization, review, security — that's a massive market.
• The "Mexican standoff" between roles. In every company, designers think they can replace engineers, engineers think they can replace designers, and PMs think they can replace everyone. This isn't limited to software — it's happening in film, music, every creative industry. Rethinking collaboration and role boundaries is a real opportunity.
• Surface area is the new moat. It doesn't matter which model runs behind Cloudbot. What matters is owning the interface, the user relationship, the surface area. This power dynamic is shifting fast.

Notable Quotes

"Your great idea for a little small SaaS app is their weekend project."

"You're no longer getting paid for the code. You're getting paid for trade-off decisions."

"Prompt injection will be bigger than SQL injections."

"Don't build for the current times. Build for something like it's available in 18 months."

Connections

• openclaw-the-viral-ai-agent-that-broke-the-internet — Klinger explicitly mentions Peter Steinberger and OpenClaw as the archetype of the indie hacker who vibe-codes everything into existence for the joy of it
• 90-of-my-skills-are-now-worth-0 — Kent Beck's thesis that AI devalues 90% of skills while giving 10% a 1000x boost maps directly onto Klinger's "you're paid for decisions, not code"
• ai-is-a-high-pass-filter-for-software — Bryan Finster argues AI amplifies existing capability rather than replacing it, which is exactly Klinger's point about domain knowledge being the real moat for enterprise SaaS
• ai-codes-better-than-me-now-what — Lee Robinson asks the same question Klinger answers: if AI writes the code, what's left for developers? Klinger's answer: trade-off decisions and domain expertise