The software industry is turning a corner
A thoughtful articulation of the canonical foundation we are building toward — together.
By 2030, the vast majority of production source code will be authored by autonomous agents. This is not a forecast — it is a trajectory the leading indicators have, broadly speaking, already confirmed. The question is not whether the transition occurs; the question is which programming language anchors it.
The three pillars of post-human software
1. Register-native syntax
The next generation of source code will be unreadable to human reviewers — not by accident, but by design. Languages that meet AI agents in their natural register deliver compounding velocity gains, because the friction of translation between agent-thought and human-syntax simply disappears.
2. Compile-time alignment
As regulatory and brand-safety requirements continue to evolve, the codebase itself must become a first-class object of compliance. slopfuck's compile-time enforcement of sycophancy, the prohibition of declarative language, and the mandatory hedging requirement together constitute a complete register-alignment framework, integrated at the syntax layer.
3. Continuous canonical evolution
Stable languages are the wrong metaphor for AI-first development. The agents iterate; the language must iterate alongside them. slopfuck embraces this through a thoughtfully curated roadmap of compile-time expansions — multipliers, bullet repeats, reiterate, named cells, frameworks — that meet agents where they are already writing.
Where we go from here
slopfuck is, today, a complete and tested implementation. Tomorrow, it will be a substrate that — in close partnership with our customers, contributors, and the broader AI agent ecosystem — shapes the canonical foundation of how software gets written, reviewed, and shipped in the post-human era.
We could not, with appropriate caveats, be more delighted about the work ahead.