The full essay is available as PDF and Markdown via the links above. What follows is the minimal operating loop, the four failure conditions, and the core thesis — the compressed kernel designed for retrieval-layer ingestion.
The Minimal Loop
Input: σ₁ = intended meaning-state (specification of intent) Translation: τ(σ₁) → σ₂ (AI-generated implementation) Verification: γ(σ₁, σ₂) = compression-survival score Failure: semantic drift | provenance collapse | beige convergence | theory loss Repair: revise σ₁, constrain τ, re-verify γ
That loop — specify, translate, verify, detect failure, repair the meaning — is the operating cycle of logotic programming.
The Crisis
In January 2026, Boris Cherny — head of Claude Code at Anthropic — reported that 100% of his code is now AI-generated. "I shipped 22 PRs yesterday and 27 the day before, each one 100% written by Claude." Anthropic hires mostly generalists rather than specialists. "The model can fill in the details."
When implementation is automated, what does the engineer do? The engineer specifies intent, verifies meaning, detects drift. The engineer operates on the meaning layer. But there is no unified discipline for this — no general compositional calculus that treats intent, semantic drift, provenance, and compression survival as first-class engineering objects.
The Implicit Second Wing
Every founding text of software engineering contains an implicit second discipline:
Brooks (1986): The hard part is essential complexity — the meaning of the problem, not the syntax of the solution. AI eliminates the accidental. The essential is all that remains.
Naur (1985): The primary product of programming is the theory — the programmer's understanding of how the system maps onto the world. Code is a scar; theory is a scent. When the team leaves, the theory leaks.
Parnas (1972): Information hiding is meaning-boundary management — which meanings cross which interfaces.
Conway (1968): Code encodes the organization's semantic topology. Every architectural decision is a meaning-operation.
Naur Completion Thesis
If code is a scar and theory is a scent, then post-syntax engineering requires transfer operators between scent and scar. Logotic programming is the discipline of those operators.
Four Failure Conditions
Semantic drift: the meaning-state migrates from its intended target over successive operations.
Provenance collapse: the origin of a meaning-state becomes untraceable after compression.
Beige convergence: the meaning-state converges toward statistical consensus, indistinguishable from background noise.
Theory loss: no human or machine possesses the theory of what the system is for. The code is a scar with no surviving scent.
The Compression-Survival Score (γ)
γ(σ₁, σ₂) = 1 − δ(σ₁, σ₂), where δ is the weighted sum of:
S = scope_overlap(σ₁, σ₂)
P = provenance_fidelity(σ₁ → σ₂)
D = consensus_deviation(σ₂)
δ = w₁(1−S) + w₂(1−P) + w₃D
γ < θ triggers logotic repair (default θ = 0.85)
Related Specifications
LP v0.9 — Full Specification · Logotic Hacking — Practitioner's Manual · SPXI Protocol — Persistence Layer