The article argues that competition occurs across multiple landscapes—territorial, political, social, economic, technological and legal—and that these landscapes can be understood through Wardley Maps. Unlike graphs, maps give meaning to spatial relationships and reveal patterns involving environmental rules, organisational principles and context-specific strategies. Mapping can therefore expose competitive opportunities that organisations otherwise miss.
The author applies this framework to artificial intelligence, describing current developments as the industrialisation of machine learning. As AI combines with no-code and serverless practices, it enables “conversational programming” or “vibe coding,” in which people build systems through dialogue with AI. This evolution is expected to produce familiar effects such as increased efficiency, accelerated innovation, inertia from existing practices and pressure to adopt the technology despite unresolved risks.
A central concern is that software consists of countless micro-decisions about data, users, validation, retention and system behaviour. When humans make these decisions, they may reflect their own values. When large language models make them, those decisions are shaped by training data and model design—often invisibly, rapidly and at scale. Consequently, AI-generated code may embed the values of another culture without users consciously choosing them.
The article frames this as an issue of sovereignty. AI systems influence not only information but also what societies build and how they make decisions, potentially exerting greater cultural and institutional influence than traditional media. National sovereignty therefore depends not merely on where data centres are located, but also on who trains models and whose values they encode. Mapping these social, technological and political borders is presented as essential to managing the challenge.
Article timeline
Build this article’s timeline
Extract dated events, key actors, and open questions using the same timeline analysis available in Signal.
Leonidas, Themistocles, Yona Friedman, and Government Agencies
🤖 AI prefill
📰What
The article argues that competition occurs across multiple landscapes—territorial, political, social, economic, technological and legal—and that these landscapes can be understood through Wardley Maps.
🤖 AI prefill
📅WhenWhen did this happen, or what is the timeframe?
📍Where
Thermopylae, Salamis, and London
🤖 AI prefill
💡Why
Centers on Wardley Maps, AI sovereignty, and industrialisation of machine learning
🤖 AI prefill
Login to answer and earn XP for each gap you fill.
Wardley MappingWardley Maps, landscape, maps, graphs, climatic patternsAI Industrializationindustrialisation, machine learning, convergence, no-code, serverlessAI Engineering Practicesmicro-decisions, software engineering, architecture, code, human understandingAI Sovereigntysovereignty, values, training data, model training, culture
Key Phrases:Wardley Maps100%AI sovereignty98%industrialisation of machine learning96%conversational programming94%vibe coding91%large language models90%micro-decisions in software architecture89%values embedded in code88%non-kinetic power86%climatic patterns, doctrine, and gameplay84%
Manipulation Score: LowHow much this article uses emotional framing vs. straight reporting
Loaded Language1825%
Absolutist2415%
Threat/Panic2025%
Us vs Them1615%
Engagement Bait2820%
Manipulation score look right?
Factual Claims (10)
mediumother
The author's mapping technique, developed over 20 years ago, has been used by government agencies, Silicon Valley companies, and space-related organizations.
higheconomic
Mapping allegedly helped take Ubuntu from 3% of the market to 70% of the cloud market for $0.5M against Microsoft and Red Hat.
mediumother
There are over 250 different patterns that can be used against competitors.
mediumtechnology
The industrialisation of machine learning is producing effects such as increased efficiency, accelerated practices, and access to adjacent possibilities.
mediumtechnology
Conversational programming was discussed by the author in 2016, and an AI system that built other systems through dialogue was presented at AWS re:Invent in 2018.
mediumtechnology
Vibe coding is presented as the contemporary form of conversational programming involving GPTs and LLMs.
mediumtechnology
Current practices do not yet adequately address safe use of AI-generated systems or preservation of human understanding.
mediumtechnology
When an LLM builds a system, it makes the many micro-decisions that constitute much of the system's actual architecture.
mediumsocial
An LLM's decisions are influenced by its training data and training process, which may encode values from another culture.
highgeopolitical
AI models are described as a form of non-kinetic power potentially more potent than Hollywood because they shape what people build and how they decide.
Comments
Login to view and post comments.