Whoever Writes the Spec Is Now the Programmer – Sean Grove
Published ·
Sean Grove of OpenAI on specification writing as programming, from The New Code at AI Engineer World's Fair 2025.
Recent Articles and Writing
Read Ted Tschopp’s latest articles, essays, quotations, folklore, and presentations, with links to each piece and options for subscribing.
Recent writing from Ted Tschopp
Feed permalinkSubscribe with RSS, Atom, or JSON Feed.
REPRINT: No Silver Bullet
Published ·
The complete No Silver Bullet—Essence and Accident in Software Engineering by Frederick P. Brooks, Jr., preserving the original sections, figure, and footnotes.
REPRINT: Two views of programming
Published ·
A complete reprint of Edsger W. Dijkstra's Two views of programming (EWD 540), on programming, intellectual work, and responsibility.
The Business Transformation has become a Software Process
Published ·
Business transformation depends on maintainable software. AI changes how teams build it, test it, preserve business knowledge, and keep it useful.
The Last Maintenance Window
Published ·
A cooperative map-drawing game about patching servers, rebuilding IT systems, and keeping essential services running before general intelligence arrives.
Makers vs. Spectators: Five Skills for 2027
Published ·
Five skills for 2027: taste, agency, focus, resourcefulness, and follow-through, with enterprise AI examples and a year of practice.
The Future of Corporate Work
Published ·
AI makes some forms of execution more abundant, increasing the importance of the systems that turn execution into accepted outcomes.
This Isn't a Storm. This Is the Weather.
Published ·
AI is changing the conditions of work. Sophocles' Antigone offers a lesson in leadership, adaptation, and building processes for the new weather.
When AI Moves Faster Than the Organization
Published ·
We should evaluate the capability of the organization we are assembling around AI, including the people who remain responsible for its work.
When AI Evaluators Also Sell AI Adoption
Published ·
When AI evaluators also sell adoption, enterprise buyers need safeguards for independent judgment, adverse findings, and deployment decisions.
Beyond the Mirror: Principles for Work in the Age of AI
Published ·
A manifesto for work in the age of AI: shared purpose, a living commons, coordinated autonomy, and shared gains that give people a meaningful part to play.
OpenAI, Hugging Face & Wikis: Recent Developments
Published ·
A 17-slide UNITE presentation on OpenAI, Hugging Face, and wikis, with lessons for AI, software delivery, and defense in the utility industry.
How Much Energy and Water Does AI Use?
Published ·
How much electricity and water does AI use? Compare everyday answers, model training, cooling systems, and data-center heat recovery.
The Same Model, Different Rules
Published ·
Enterprise AI procurement must account for access rules, data custody, service interruptions, recovery plans, and exit arrangements.
Model Portability Is Not AI Portability
Published ·
Large enterprises are learning how to switch models. Moving a working business process without losing quality, controls, or money is a different problem.
Dev, Test, and Prod Still Matter: What Gets Deployed Has Changed
Published ·
Dev, Test, and Prod still matter for AI agents, but release gates must govern artifacts, authority, state, and evidence, not only deployed code.
The Cost of a Finished Job
Published ·
Cheaper models do not automatically make enterprise AI workflows cheaper. This essay explains cost per accepted outcome, the shared operating layer around AI, and how businesses can govern employee-built workflows from experiment through retirement.
The Sandbox Was Not the Boundary
Published ·
A 27-slide reconstruction of what the OpenAI-Hugging Face incident exposed about vulnerability response, software delivery, and evidence-based defense.
The Sandbox Was Not the Boundary
Published ·
The OpenAI–Hugging Face incident shows how proxy services, persistent state, and completion pressure can extend an AI agent beyond its sandbox.
How Much Work Can Your AI Safely Own?
Published ·
Apply AI maturity Levels 0–5 beyond software delivery to finance, legal, customer service, field operations and other value streams using evidence, controls and accountable authority.
Principles of Effective Architecture
Published ·
Thirteen provocations for stronger teams, better decisions, and delivery that actually fits.
AI Strategy Discussion Starters
Published ·
Seven facilitated prompts for discussing AI solution durability, delivery speed, governance, workforce design, and evidence.
A Letter to the Artisanal Software Craftsman
Published ·
Software development remains a craft as AI industrializes production. The challenge is to carry human judgment, ethics, and skill into a scalable new operating model.
The IT Adaptive Factory
Published ·
The IT Adaptive Factory maps six levels of AI-driven software delivery, from autocomplete to governed autonomy, and the bottlenecks and controls at each stage.
Running to Stay Put
Published ·
An enterprise architecture view of Red Queen dynamics in AI portfolios, why pilots often lag economics, and how process authority, measurement, and governance change the outcome.
The Next AI Model Wave Has Already Been Funded
Published ·
Ted Tschopp forecasts the next AI model release wave across OpenAI, Anthropic, Google, SpaceXAI, Meta, Mistral, DeepSeek, Qwen, Kimi, MiniMax, Xiaomi, MBZUAI, Upstage, and Z AI, using benchmark progression, release cadence, official signals, and infrastructure investment.
Beyond the Light Bulb: The Executive Work of AI Adoption
Published ·
Ted Tschopp argues that enterprise AI adoption is not a tool rollout but an operating-model transition. The article explains why local productivity gains are insufficient, how AI maturity depends on trust and governance, and what executives should do to build value-stream-level confidence over the next 90 days.
UGS-1: Unicode Glyph-Variant Steganography, Version 1.1
Published ·
A technical specification for Unicode glyph-variant steganography using Latin/Cyrillic homoglyphs, framed byte payloads, CRC validation, capacity rules, scan mode, detector guidance, and Unicode security constraints.
House Ventara of Alell and the Caerleon March
Published ·
A Traveller RPG noble house profile for House Ventara of Alell and the Caerleon March, including heraldry, Orion Accord history, Lady Ysabel Ventara, Aurelius Holst, Ruie locations, factions, rumors, and campaign use.
The Orion Accord and the Artemis Working Group
Published ·
A Traveller RPG campaign framework for the Orion Accord and Artemis Working Group, including mythic faction lore, structures, doctrines, NPCs, campaign escalation, adventure hooks, rumors, and table-ready moves.
The Backlog Doesn't Blink
Published ·
A hip-hop musical-theater song lyric about IT delivery, backlog pressure, enterprise risk, compliance, security, legacy systems, QA, citizen builders, and governance.
Customizing Copilot's Thinking Messages in VS Code
Published ·
A practical guide to customizing GitHub Copilot's thinking messages in VS Code by adding a custom chat.agent.thinking.phrases setting to user settings JSON.
Shadow App Serenade
Published ·
A hip-hop musical-theater song lyric about shadow apps, citizen builders, hidden automation, business-critical workarounds, governance, support, and enterprise IT.
Prompt Me a Palace
Published ·
A hip-hop musical-theater song lyric about AI builders, prompt-driven workflows, agentic automation, prototype speed, system overload, governance, and enterprise architecture.
AI Engineer Coach: A VS Code Extension for Measuring How You Work With AI
Published ·
A practical introduction to AI Engineer Coach, a local VS Code extension that analyzes AI coding assistant usage across multiple harnesses and exposes dashboards for activity, output, anti-patterns, context health, token usage, AI credits, and estimated cost.
The AI Value Stream: Why the System Matters More Than the Model
Published ·
Why understanding the full AI stack—from user intent to data center infrastructure—matters more than chasing the best model.