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:

  1. Ask the model to generate a plan first
  2. Implement in small, iterative steps
  3. 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:

  1. Plan before you prompt — Structure thinking before invoking AI
  2. Context is everything — Comprehensive docs and examples enable quality output
  3. 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.