A few months ago, I did something that felt slightly unconventional: I brought together a small group of people who were as curious about artificial intelligence as I was.
We were not a room full of AI experts, and none of us claimed to have all the answers. We were people from different backgrounds, united by curiosity, caution, and a serious desire to understand what AI could mean for our societies and our future.
We were excited by its possibilities, alert to its consequences, and unwilling to settle for headlines.
Instead of passively following AI news, we created a space for deliberate learning.
That experiment became CentPol.
From Individual Curiosity to Collective Inquiry
We began by identifying the domains where our own knowledge was thin - areas where AI was already reshaping institutions, industries, public policy, work, education, security, and everyday life.
We gathered resources, examined emerging developments, and tried to understand not only what was happening, but why it mattered.
The goal was not simply to consume information. We wanted to dissect complex issues, surface the assumptions behind competing arguments, and understand how technological change interacts with social, political, and economic systems.
We shared what we learned through articles, posts, research notes, and other artifacts. More importantly, we shared it through conversation.
Those conversations became the center of the project. They created an environment in which people could ask basic questions without embarrassment, challenge accepted narratives, test their assumptions, and encounter perspectives that would have been difficult to discover alone.
Turning Conversations Into Sprints
After several discussions, we began organizing our work into thematic learning cycles that we call Sprints.
A Sprint is a four- to eight-week period of focused inquiry around a particular subject or policy domain. It is designed to move us beyond surface-level commentary and allow us to examine an issue from multiple angles.
During a Sprint, participants work through relevant research, current events, policy proposals, technical developments, and institutional responses. The purpose is not to manufacture consensus. It is to improve the quality of our questions, sharpen our reasoning, and develop a more informed understanding of the subject.
We have now completed two Sprints and begun our third.
Each has reinforced the same lesson:
Understanding AI requires more than knowing about the latest model release, product announcement, or regulatory proposal. It requires the ability to connect those developments to the broader systems in which they operate.
What a Sprint Actually Looks Like
To make this concrete, consider our first Sprint: National AI Strategy.
Artificial intelligence has moved from the laboratory to the center of national strategy. It is influencing economic competitiveness, national security, public services, and countries’ geopolitical positions.
A national AI strategy is a government’s attempt to define what it wants from the technology, what capabilities it must build, and how it intends to pursue those goals responsibly.National AI strategies often encounter predictable problems.
Some are too vague, relying on broad declarations such as, “We will become a global leader in AI.” Others are too narrow, offering lists of technology projects without confronting access to computing infrastructure, institutional capacity, workforce effects, energy requirements, or implementation costs.
Many also borrow external blueprints without adapting them to local infrastructure, fiscal capacity, political institutions, or geopolitical trade-offs.
Over five weeks, our cohort examined a central question from multiple angles:
What makes a national AI strategy implementable, particularly in African and other Global South contexts where access to computing infrastructure, specialized talent, and fiscal capacity can be binding constraints?
We organized the discussion around a reusable framework comprising nine recurring pillars of national AI strategy: vision; compute and infrastructure; data governance; talent and research; priority sectors; governance and safety; institutions and coordination; international engagement; and metrics.
Each week, participants used that framework to analyze real national and regional strategies.
We compared Singapore’s sequencing and implementation model with Kenya’s attention to local constraints. We examined the African Union’s continental framework as a reference architecture. We also tested the idea of “strategy as a PDF” against the more difficult question of what governments actually fund, coordinate, and build.
The point was not to admire policy documents.
Participants choose either to write a memo or developed three-minute pitches designed for real decision-makers.
That is the texture of a Sprint: a genuine question, a shared analytical framework, real source material, and the discipline of producing work that can withstand scrutiny.
No one needed a coding background. What participants needed was curiosity, intellectual honesty, and a willingness to think in public.
Knowing About AI Is Not the Same as Understanding It
The AI information environment moves quickly.
Every week brings new models, capabilities, investments, policy announcements, controversies, and predictions. Keeping up with those developments can create the impression that we understand the technology simply because we recognize the terminology.
But familiarity is not comprehension.
One of the clearest tests of understanding is whether we can explain an idea in simple terms.Can we describe what an AI system does, how it works, where its limits lie, and why its consequences matter to someone outside the technology sector - without relying on jargon or hype?
That ability is becoming essential because AI will not be shaped by technical professionals alone.
Its development and governance will involve policymakers, educators, researchers, civil society organizations, business leaders, journalists, lawyers, national security professionals, and members of the public.
For these groups to participate meaningfully, they need more than headlines. They need accessible explanations, informed analysis, and spaces in which difficult questions can be examined collectively.
Why Conversation Matters
Learning alone has limits.
Individual study gives us information. Discussion exposes weaknesses in our reasoning.
Other people ask questions we had not considered. They notice assumptions we have been treating as facts. They introduce experiences, disciplines, and perspectives that change how we interpret an issue.
In a serious discussion, ideas are not merely repeated. They are examined.
That process can reveal new frames of reference, unsettle comfortable positions, and produce better ways of thinking about emerging technology.
This matters particularly in AI policy.
The policies being developed today will influence how AI systems are built, deployed, evaluated, and governed. But sound policy cannot emerge from technical knowledge alone. It must account for institutions, incentives, power, culture, security, economic opportunity, and human behavior.
Developing that kind of judgment requires sustained engagement across disciplines.
At CentPol, we call that capacity policy intelligence: the ability to reason about AI as a technical, institutional, economic, political, and human system at the same time.An Invitation to CentPol
CentPol.com is designed to cultivate that capacity.
We are building a community for people who want to understand AI more clearly, examine its implications more carefully, and contribute more intelligently to the decisions that will shape its future.
The discussions are built for curious people from different professional, geographic, and intellectual backgrounds.
Participants do not need to be technical experts. They need to be willing to learn, question assumptions, engage seriously with evidence, and contribute thoughtfully to collective inquiry.
We meet every Friday at 12:00 p.m. Eastern Time, and each session runs for approximately 60 to 80 minutes.
These conversations matter because the future of AI is not only a technical question. It is a policy question, an institutional question, an economic question, and a human question.
Understanding the technology is only the beginning. Learning to explain it, interrogate it, govern it, and shape it is the larger task.
That is the work of CentPol: building policy intelligence for the next technological era.
Join the Sprint at CentPol.com, or reply to this post for details about Friday’s discussion.
You are the voice that will shape AI in your context.