“It’s about the people, stupid.”
I’ve lived by these five simple words throughout my 20+ year career helping people thrive during periods of rapid technological change.
Why? Because everything is so big and fast in the emerging technology space that it’s easy to forget people.
The financial numbers are huge. The pace of product development is relentless. Organizations are struggling to figure out how to adapt in real time. People are being asked to operate outside their comfort zone.
When technology is driving rapid change, I’ve learned people benefit most from first principles thinking, consistent strategy, and foundational skills they can rely on over the long-term.
I developed the Doing AI Efficiently Operating System to meet this need.
What You Won’t Find Here
The AI education and implementation space is loud and crowded right now. Many people are delivering:
Information about how to use the latest large language models such as Fable, Astra and their siblings
Articles describing the latest agentic workflows such as loop and graph engineering
Tutorials about Claude Cowork, ChatGPT and other interface systems released by frontier labs
These topics are well-covered elsewhere. While information about models, agentic workflows and LLM user interfaces like Claude Cowork will be referenced, they won’t be the main focus.
What You Will Find Here
I developed the Doing AI Efficiently Operating System to help you (as an individual, part of a team or organization) gain foundational, first-principles-based knowledge and skills that will aid your success. Here’s how each part of the OS works together to achieve this goal.
Grasp: Gain Knowledge
It all starts with having a solid understanding of AI. Not just generative AI, but other technologies such as machine learning, world models and robotics. Learning about how the market is evolving and what to expect in the near future is also important. This information will help you separate hype from reality and be less surprised by rapid developments in the field. The Doing AI Efficiently OS delivers insights across all these areas.
In addition to the educational content published in this newsletter, you’ll also have access to the:
Thinkers on AI Writer’s directory, featuring emerging and established writers publishing content on AI. The directory is refreshed every 36 hours.
Growing AI terms glossary
Coming Soon: The Thinkers on AI Writer’s Directory
Coming Soon: The AI Glossary
Discern: Practice Judgement
One of the most important roles people will have is judging whether AI outputs are correct, valid and appropriate. Judgement is required in diverse areas, such as checking internal documentation and writing using AI.
You’ll find newsletter articles providing insights about judgement-related topics such as the risks of using AI for writing, and how to conduct high-confidence research in the age of AI-dominated research.
Generative AI is Making Writing Deathly Boring
Welcome to the first installment of the AI Research Beat. This is a regular series where I highlight an interesting study that sheds light on how AI is impacting work.
The OS also features practical education on that will develop your judgement (see below).
Ward: Guard Against AI Risks
Generative AI is powerful, but also risky. The AI Security Guard platform, which I’ve been developing since February 2026, features many resources, including a the popular free AI Security Action Pack.
The newsletter will feature key AI security insights along with links to security-related content and software from AI Security Guard.
Execute: Use AI Excellently
The OS features many AI execution-related resources, including newsletter articles and experiential learning experiences.
Coming Soon: Clankyopolis AI Skill-Building Game
In addition to online virtual interactive courses, the flagship educational experience will be Clankyopolis. This educational game will put players in the role of Administrator of Clankyopolis, a fictional robot city.
As Administrator, you’ll practice key AI skills such as:
Evaluating AI outputs
Conducing data and informational AI quality control
Implementing agentic workflows, from selecting models to evaluating their performance over time
The game will enable you to deepen your understanding of AI implementation in a safe, fun environment.
Honor: Taking Responsibility for AI Outputs
An important step toward mastering AI is being able to take responsibility for the quality and robustness of AI outputs. Every resource, training solution and product in the OS supports this objective.
What to Do Next
Subscribe: In addition to receiving research, analysis, tips, tools and other free resources, subscribers receive discounts on OS educational experiences
Spread the word: If you’re already a subscriber, thank you! Please help this community grow by spreading the word
Contact me with questions: If you have questions about the OS, how to use AI effectively, send me a message. I read every message
Thanks for reading.