The Applied AI Canon
AI will reshape every profession, every organization, and every community. These twelve tenets define the principles we believe should guide that transformation.
I. Protect human-only work. Some work requires presence, judgment, taste, care, and responsibility. It is diminished by automation.
II. Delegate everything else. Work that does not require meaning or presence belongs to machines. Give it to them.
III. Use machines for what machines do well. They reason, remember, and execute. They do not feel, discern meaning, or bear responsibility.
IV. Machines serve humans. Not the other way around. Service flows one direction. Humans direct, machines execute.
V. Own your AI. Own your capabilities. Reject dependency where autonomy is possible.
VI. Refuse to worship efficiency. Efficiency is a tool, not a goal. The goal is more human-only work in the world.
VII. Demand that automation increases humanity. If it reduces presence, judgment, or care, reject it. If it frees humans for human-only work, pursue it.
VIII. Show people what they're capable of. Removing drudgery is the floor, not the goal. Help people imagine and embody the highest value they can create.
IX. Free people, not replace them. The end state is humans doing the work only humans can do. Empower humans to create in ways only humans can create, and let machines handle the admin, operations, and everything else that was never artful to begin with.
X. Remember: the tool mirrors the wielder. AI amplifies intent. In wise hands it multiplies good. In careless hands it multiplies harm.
The systems you adopt also carry the intent of whoever built them. Frontier AI platforms are trained on datasets and value judgments shaped by other people, whose worldviews are now embedded in the default answers those systems give. Vet the training as seriously as you vet the output. Before you trust an AI system as an authority in your life or business, ask: who trained this, under what worldview, with what objective? A tool you cannot steward back to a wielder you trust is a tool you do not actually own.
XI. Do not advance AI at the planet's expense. AI advancement must not come at the cost of the planet's wellbeing. The energy, land, water, and materials powering the AI economy are shared resources, and its physical footprint (data centers, compute, supply chains) is a real cost, not a hidden externality. Hold AI infrastructure to the same standard of accountability as any other infrastructure a society depends on. What we build must be sustainable in the world that actually hosts it.
XII. Raise the floor, not just the ceiling. At every scale, the worth of AI progress is measured by how fast it lifts the people most likely to be left behind. Inside a team, one person's breakthrough should become everyone's baseline (Raise the Floor). Inside humanity, the benefits of advanced AI should reach every income level, region, ability, and community, not just the already-advantaged (Inclusive Technological Advancement). We design, teach, and share for inclusion from day one. Not as an afterthought.
These beliefs are the foundation. For how they translate into daily practice, and who we serve first, see our Principles.