Industry Alignment
ASDLC formalizes patterns that industry practitioners are discovering independently. This page tracks convergent thinking from engineers, researchers, and organizations navigating agentic development.
Convergent Frameworks
Boris Cherny: Plan-and-Iterate Discipline
Source: The Peterman Podcast (December 2025)
Role: Principal Engineer, Anthropic; Creator of Claude Code
Framework:
- Ask the model to generate a plan first
- Implement in small, iterative steps
- Write by hand where you have strong technical opinions
ASDLC Mapping:
| Cherny's Term | ASDLC Equivalent |
|---|---|
| "Plan first" | Spec-Driven Development |
| "Small iterative steps" | Context Gates |
| "Human reviewing" | L3 Autonomy |
| "Skilled pilot" | Instructor-in-Cockpit |
Key Quote: "Speed is seductive. Maintainability is survival."
Andrej Karpathy: Vibe Coding Definition
Source: Twitter/X (February 2025)
Role: Former Director of AI, Tesla; Co-founder, OpenAI
Contribution: Coined "vibe coding" to describe natural language → code generation without formal specifications
ASDLC Response: Vibe Coding concept article documenting the pattern and its bounded applicability
Matt Watson: Product Thinking as Core Competency
Source: LinkedIn (January 2026) and Product Driven (2023)
Role: 5x Founder & CTO, CEO of Full Scale
Thesis: Vibe coders outperform average engineers because they think about product outcomes, not just implementation. The AI era makes product thinking mandatory—AI handles "just build this" work; humans must decide what matters.
Key Insight:
"For years, we rewarded engineers for staying in their lane, closing tickets, and not rocking the boat. Then we act surprised when they don't think like owners... Product thinking isn't a bonus skill anymore. In an AI world, it's the job."
ASDLC Mapping:
| Watson's Principle | ASDLC Equivalent |
|---|---|
| "Understand the user and the problem" | Product Thinking |
| "Give context, not just tasks" | Context Engineering |
| "This will bite us later" (risk recognition) | Adversarial Code Review |
| "Clearly explain what and why" | Spec-Driven Development |
ASDLC Response: Product Thinking concept article—why Specs exist to force product thinking before code generation
Rasmus Widing: Product Requirement Prompts (PRPs)
Source: GitHub (2024-2025) and LinkedIn
Role: Engineering Leader; Creator of PRP methodology
Thesis: "What's the minimum viable specification an AI coding agent needs to plausibly ship production-ready code in one pass?" PRPs are "PRD + curated codebase intelligence + agent runbook."
Core Principles:
- Plan before you prompt — Structure thinking before invoking AI
- Context is everything — Comprehensive docs and examples enable quality output
- Scope to what the model can reliably do — Bounded execution units
PRP Structure:
| PRP Component | ASDLC Equivalent |
|---|---|
| Goal (what needs building) | The Spec — "What" |
| Why (business value) | Product Thinking |
| Success Criteria | Context Gates |
| All Needed Context | Context Engineering |
| Implementation Blueprint | The PBI |
| Validation Loop | Quality Gates |
Key Quote: "Simple idea. Hard to do consistently. That's where the skill is."
ASDLC Response: Product Requirement Prompt concept article mapping PRP to ASDLC terminology
Industry Data Points
Google: 30% AI-Generated Code
Finding: As of 2024, approximately 30% of Google's code is AI-generated
Challenge: First year where copy-pasted code exceeded refactored code
ASDLC Concern: Highlights the maintainability crisis from unchecked AI generation without formal specifications
Anthropic: Claude Code Adoption
Finding: 80-90% of Claude Code's own codebase is now written by Claude Code
Impact: 70% productivity increase per engineer
ASDLC Note: Demonstrates the power of agentic development when properly structured. Anthropic's success validates the need for rigorous frameworks like ASDLC to manage this transition.
Forrester: Technical Debt Crisis
Prediction: 75% of tech leaders will face moderate-to-severe technical debt by 2026
Root Cause: Speed prioritized over maintainability in AI-assisted development
ASDLC Solution: Spec-Driven Development + Context Gates provide the structural discipline to prevent this outcome
Convergent Principles
Across these independent sources, a consistent pattern emerges:
| Observed Problem | Industry Response | ASDLC Framework |
|---|---|---|
| AI generates code without understanding architectural constraints | Cherny: "Write by hand for strong opinions" | Spec-Driven Development |
| Code works but violates implicit contracts | Google: Copy-paste exceeds refactoring | Context Gates (deterministic validation) |
| Speed creates unsustainable technical debt | Forrester: 75% will face debt crisis | L3 Autonomy (human oversight) |
| Engineers trust AI too completely | Cherny: "Plan first, then iterate" | The Spec (contracts before code) |
ASDLC is not prescriptive dogma—it's the formalization of hard-won lessons from practitioners at the frontier of agentic development.
Contributing
Know of research, frameworks, or industry voices that align with ASDLC principles? Submit a suggestion.