Six dimensions define the candidate case.
Bounded solo provenance
One human materially directs, selects, integrates and governs the eligible formation set. Material hidden human formation contribution changes eligibility.
Company-scale substance
The work must have meaningful breadth and depth. Naming volume alone is not enough.
Structural compression
Multiple decision surfaces that would normally be distributed across roles are compressed into one founder-directed AI-native workflow.
Documented externalization
Architecture, artifacts, chronology and records survive outside the founder’s memory and can be reconstructed by qualified reviewers.
Strategic-value plausibility
The eligible stack must be substantive enough to justify serious technical, IP and commercial diligence of potentially high strategic value.
Falsifiability
The claim must expose defeat conditions. A framework that cannot return a negative result is positioning, not evaluation.
What this category does not mean.
Not an official status
No institution currently certifies MZN as a “one-person unicorn.” The phrase is a category hypothesis under challenge.
Not a current $1B valuation
Unicorn-scale strategic value is a diligence question, not a public self-certified valuation.
Not “one human manually did everything”
AI systems and tools are part of the operating model. The question is human formation control and material human contribution, not tool-free execution.
Not eternal solo status
Phase 3 experts, partners and teams can enter later without retroactively rewriting a bounded Phase 2 provenance record.
Phase 1 was team-executed and is explicitly outside the Phase 2 solo claim. Phase 3 is where independent validation, legal/IP work, pilots and selective team formation are expected.
Old proxies may become incomplete before new standards exist.
Headcount, funding rounds, org charts, PR footprint and physical-event presence remain useful signals in conventional company evaluation. They are not automatically sufficient measures of AI-native formation capacity. The response is not to lower standards; it is to test the underlying work directly.
Headcount
Useful for organizational scale, but not a direct measurement of founder-directed capability compression.
Funding & recognition
Useful external signals, but not substitutes for provenance, technical review, IP diligence or product evidence.
Constraint/output context
Constraints can contextualize the formation process. They do not independently prove quality, novelty or valuation.
Rankings, festivals, invitations and model opinions can be dated reasons to investigate. They are not category certification, technical validation or value proof.
Open the External Signals recordThe public definition became narrower as the review problem became clearer.
The published articles are useful because they show the category being formed in public. They are not all current rules. Earlier formulations are preserved as dated sources; the canonical OPU pages carry the current definition.
Organizational possibility
Early essays explored AI-orchestrated founders and the collapse of functions that once required larger teams.
Category + metrics
The thesis shifted toward definition, structural compression, attention limits and the inadequacy of team size or visible spend as complete proxies.
Standards + disclosure
Public essays introduced profile criteria, reconstructable documentation and the tension between evaluability and protected asymmetric value.
Bounded two-axis review
The later public formulation separates solo provenance from asset substance, applies the claim to a bounded phase, and exposes defeat conditions.
Historical publication is evidence of conceptual evolution, not an exemption from correction. Where an older article conflicts with the canonical Phase 2 boundary, current maturity labels or current disclosure terminology, the canonical site controls.
Read the published thesis trail