Name the decision
What must become clearer, safer, faster or possible?
Solutions for real institutional work
Solutions for Super Intelligence—built with AI, governed by people.
AFA helps organizations extend what their people can understand and accomplish by connecting strategy, information, AI models, agents, infrastructure, institutional knowledge, robotics, learning and measurable outcomes.
AFA terminology
Super Intelligence is the frame. AI remains the technical vocabulary.
Super Intelligence is AFA’s human-centred description of intelligence extended beyond the practical limits of unaided cognition through computation, accumulated knowledge, models, machines and other instruments of intelligence.
We continue to use artificial intelligence (AI) for models, APIs, standards, regulations, procurement categories, vendor products and established search language. The transition is additive, not a denial of the technical field.
Two words matter: AFA’s Super Intelligence is not the same claim as the established one-word term superintelligence, commonly used for an intelligence exceeding human intelligence.
Read AFA’s definition →CHOOSE YOUR STARTING POINT
Three connected ways to build intelligence.
Start with the decision you need to make. Explore focused capabilities within each chapter, or discuss your situation with us.
01 / Strategy, governance & adoption
Decide what intelligence should do.
Begin with organizational readiness, accountability, information protection and workforce capability.
02 / Intelligent applications & systems
Connect intelligence to the work.
Bring approved knowledge, tools and workflows into useful, governed applications.
03 / Infrastructure & physical intelligence
Put intelligence where it belongs.
Evaluate control, compute and physical deployment requirements before selecting technology.
AFA is not a product catalogue
Use the problem, information and responsibility to choose the architecture.
AFA begins by understanding the work and only then decides whether the answer is a policy, training program, Copilot configuration, agent, private model, accelerated compute environment, Marbles use case, Physical AI demonstration or creative experience.
Identify the information
What knowledge, research, data or IP is involved?
Set authority
What may AI do and what remains human-controlled?
Choose the environment
Which models, systems and infrastructure fit the workload?
Build evidence
Pilot, measure, learn and scale only where justified.
Labs connect architecture to evidence
Demonstrate before scale.
The proposed Reference Lab and developing AI Discovery Lab give AFA a practical path from advisory work into hands-on evaluation, customer demonstration and applied R&D.
Sovereign AI & Physical AI Reference Lab
Proposed Toronto environment for accelerated computing, private AI, model evaluation, agents, simulation and robotics.
Reference Lab →Atkinson AI Discovery Lab
Developing mobile experience designed to bring practical AI demonstrations and learning to institutions and communities.
Discovery Lab →Continue the journey
Related pathways
Bring us the problem, not a shopping list.
AFA will start with the work, information, people, risks and outcomes already in front of you.
Start with a Workshop