Why play is producing better software than strategy right now + grab the 4 prompts I use before I build anything

Source: Nate’s Newsletter Date: 2026-02-07 URL: https://natesnewsletter.substack.com/p/why-play-is-producing-better-software

Summary

Early 2025 marked when vibe coding crossed a friction threshold — better context windows, mature agentic patterns, and reliable platforms made building feel like play rather than labor. “Wouldn’t it be funny if…?” projects are producing unexpected commercial success (AI pet portraits at six figures monthly) while market-driven strategic development struggles. The skill that matters most isn’t coding ability but “software vision” — knowing what to build.

Implications

Agent-product positioning thread. The “smartphone camera moment” framing — skill barrier collapsed enough that non-programmers can build — is the right frame for the vibe coding wave’s implications. But Nate’s observation that experienced developers extract disproportionate value means the leveling is real but incomplete: vision and judgment compound, even if execution is democratized.

AI economics thread. Playful, accidental product-market fit (Fable’s pet portraits) suggests that the discovery mechanism for AI-native products is more like exploration than validation. That changes how early-stage AI product investment should work: bet on prolific builders who play rather than strategic builders who plan.

Watch: Whether “software vision” (knowing what to build) becomes a codifiable, teachable skill distinct from both coding and traditional product management — and whether it produces a new practitioner category that sits between builder and product owner.