Skip to content

Foundational doctrine

What the Obelisk is

The Open Frontier Curriculum is an open, adversarially maintained civilizational reference architecture for lifelong human formation under conditions of abundant machine intelligence.

It is larger than a course list and narrower than a total theory of the good life. It combines a living canon, a knowledge graph, a competence architecture, a practice system, a validation program and a constitutional set of rules for deciding what belongs.

Its purpose is not to keep humans competitively superior to machines at every cognitive task. If artificial systems become broadly superhuman, that objective becomes incoherent. Its purpose is to form humans who can inherit deeply, understand first principles, judge independently, create with taste, act competently, relate humanely, steward wisely, and learn forever.

The central problem

For much of modern education, information, competent explanation and expert labor were scarce enough that schooling could be justified partly as a way to acquire them. AI is eroding that scarcity. Frontier systems already perform strongly across mathematics, coding, professional reasoning and scientific tasks; agents increasingly take actions rather than only answer questions; AI is entering research and medicine; and education systems are struggling to distinguish output quality from actual learning.

The Obelisk therefore asks a more durable question:

What must a human being know, practice, become and remain responsible for even if machine intelligence becomes much more capable than any individual?

The answer cannot be “only the things AI cannot do.” That is a moving target. The curriculum must instead be organized around durable structures of reality and durable responsibilities of human life.

The eight verbs

  1. Inherit - receive the strongest available human discoveries, works, arguments, methods, traditions and failures without becoming captive to them.
  2. Understand - build internal models of physical reality, life, mind, society, history, language, value and technology.
  3. Judge - discriminate truth, quality, risk, value, fittingness and tradeoffs under uncertainty.
  4. Create - make explanations, proofs, systems, designs, institutions, art and culture with craft and taste.
  5. Act - initiate, execute, persist, coordinate, repair and accept responsibility in real conditions.
  6. Relate - listen, care, love, negotiate, teach, collaborate, belong and preserve the agency of others.
  7. Steward - use power responsibly across institutions, technology, culture, nature and generations.
  8. Renew - continue learning, updating and reforming one's models and practices as the world changes.

These verbs are not eight additional departments. They are the telos of the entire graph.

Human authorship

The core idea is human authorship, not human cognitive supremacy.

Authorship means people remain meaningfully involved in forming the ends, commitments, standards, permissions and responsibilities that shape their lives and shared institutions, even when AI performs much of the intermediate cognition or execution.

AI may draft a constitution. It does not thereby acquire legitimate authority to impose one. AI may produce a treatment analysis. It does not settle consent, values or acceptable risk by capability alone. AI may generate a thousand works in a style. It does not eliminate the human work of deciding what deserves attention, inheritance or creation. AI may tutor a learner. It does not make learning equivalent to possessing an excellent answer on demand.

The curriculum should therefore separate:

  • what may be delegated operationally;
  • what must remain intelligible to the learner;
  • what requires human or legitimate institutional accountability; and
  • what must be practiced personally because the practice itself is part of becoming a capable person.

Why deep knowledge still matters

Abundant machine intelligence is an argument for better organized knowledge, not for ignorance.

A learner without internal models cannot recognize category errors, choose good questions, compare rival explanations, notice impossible physical claims, understand historical analogies, interpret value conflict, or know when an AI result needs another kind of evidence. The Obelisk therefore refuses the idea that search or agents make foundational knowledge obsolete.

But it also refuses encyclopedic memorization for its own sake. The curriculum distinguishes knowledge that should be internalized, models that should be reconstructible, references that should be retrievable, and work that may be delegated with evaluation.

Why education becomes more human as AI becomes more capable

As machine cognition becomes abundant, several human scarcities become more visible: attention, ground truth, legitimate ends, judgment, taste, accountability, trustworthy relationships, commitment, embodied competence, meaning, institutional wisdom, physical resources and stewardship across time.

This changes the center of gravity of education without abandoning intellectual rigor. A serious learner should still read hard books, prove theorems, understand mechanisms, write code, study history, conduct experiments and master disciplines. But those activities must increasingly culminate in ownership, discernment, creation, practice and responsibility, rather than only production.

What the Obelisk refuses to become

It is not:

  • an AI-era employability checklist;
  • a list of “uniquely human skills” that will become obsolete as models improve;
  • an encyclopedia with no theory of mastery;
  • a Great Books canon detached from science, technology, global traditions and practice;
  • a competency framework detached from deep source knowledge;
  • a survivalist catalogue of contingencies;
  • a promise of happiness, wealth, virtue or immunity from tragedy; or
  • a static monument that cannot admit evidence against itself.

The Obelisk promise

The strongest intended claim is:

A serious completer of the Open Frontier Curriculum should possess an unusually broad and deeply connected inheritance of human knowledge, practiced competence across the critical pillars of life, strong independent and AI-augmented judgment, cultivated taste, practical and relational agency, and the capacity to continue forming themselves while exercising responsible authorship and stewardship in a high-capability civilization.

And the corresponding human acceptance question becomes:

Would I trust this person not merely to solve an unfamiliar problem, but to understand what matters, choose and revise worthy ends, learn what is needed, distinguish evidence from persuasion, use human and machine intelligence well, act competently in the world, preserve the agency and dignity of others, create things worth having, and accept responsibility for what they help bring about?

That is the standard the rest of the architecture exists to operationalize.