’ As countries seek to build AI capacity and position themselves for a global economy with AI at its centre, it’s increasingly valuable for policymakers to understand national AI levers of power, comparative advantages, and dependencies.
I’ve built AI Power Atlas, an interactive map to help enable this sort of analysis. It shows how countries are positioned across six separate levers of AI power—compute, model capability, talent, data and digital infrastructure, industrial adoption, and governance / strategic autonomy—and lets the user compare any two side by side. Rather than collapsing everything into a single ranking, it keeps the levers separate, so you can see where a particular country is stronger or weaker on AI levers of power.
It also maps dependencies: who relies on whom for things like advanced chips, semiconductor equipment, energy, and frontier models.
A few notes on the approach:
Scores are shown as ranges, each with a confidence level and a short rationale.
Where the public evidence is too thin to score responsibly, it says “unknown” instead of defaulting to zero.
The same data is available as a sortable table and can be exported.
At this stage it focuses on the 20 or so countries with the most substantial AI ecosystems. As more countries develop their AI economies, the map can be updated and built out over time.
It’s an early prototype with a hand-curated dataset, so I’d welcome feedback from anyone working in AI governance, semiconductors, or industrial policy as I continue to build it out.’
New Tool for Researching AI Power Balance Across Countries
I just released a new AI research tool to explore countries’ AI strengths, weaknesses, and comparative advantages. Welcome any feedback!
https://ai-atlas-1wjxkq84e-covifranklins-projects.vercel.app/
’
As countries seek to build AI capacity and position themselves for a global economy with AI at its centre, it’s increasingly valuable for policymakers to understand national AI levers of power, comparative advantages, and dependencies.
I’ve built AI Power Atlas, an interactive map to help enable this sort of analysis.
It shows how countries are positioned across six separate levers of AI power—compute, model capability, talent, data and digital infrastructure, industrial adoption, and governance / strategic autonomy—and lets the user compare any two side by side. Rather than collapsing everything into a single ranking, it keeps the levers separate, so you can see where a particular country is stronger or weaker on AI levers of power.
It also maps dependencies: who relies on whom for things like advanced chips, semiconductor equipment, energy, and frontier models.
A few notes on the approach:
Scores are shown as ranges, each with a confidence level and a short rationale.
Where the public evidence is too thin to score responsibly, it says “unknown” instead of defaulting to zero.
The same data is available as a sortable table and can be exported.
At this stage it focuses on the 20 or so countries with the most substantial AI ecosystems. As more countries develop their AI economies, the map can be updated and built out over time.
It’s an early prototype with a hand-curated dataset, so I’d welcome feedback from anyone working in AI governance, semiconductors, or industrial policy as I continue to build it out.’