Classify What
You Are
Dual-test classification engine that determines if a System of Interest is Human, AI, or Non-AI Software — by testing human ability and software ability independently.
Claim 1 Classification Matrix
Process
How It Works
Test 1: Human Ability
An NLP challenge tests language understanding. Humans and AI pass; non-AI software fails.
Test 2: Software Ability
A math challenge tests computational precision. AI and software pass; humans typically fail.
Claim 1 Classification
The system applies the patent classification matrix: both abilities = AI, human only = Human, otherwise = Non-AI Software.
Integration
Quick Start
// Step 1: Get the human ability challenge (NLP)
const test1 = await fetch('http://45.77.99.52:3005/api/v1/challenge?type=human_ability');
const challenge1 = await test1.json();
// { id, test_type, question, options, hint }
// Step 2: Submit Test 1 answer (option index)
const verify1 = await fetch('http://45.77.99.52:3005/api/v1/verify', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ challenge_id: challenge1.id, answer: 1 })
});
const result1 = await verify1.json();
// { verification_id, passed, test_type: "human_ability" }
// Step 3: Get the software ability challenge (math)
const test2 = await fetch('http://45.77.99.52:3005/api/v1/challenge?type=software_ability');
const challenge2 = await test2.json();
// Step 4: Submit Test 2 + get Claim 1 classification
const verify2 = await fetch('http://45.77.99.52:3005/api/v1/verify', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
challenge_id: challenge2.id,
answer: 51.67364016736402,
prior_verification_id: result1.verification_id
})
});
const result2 = await verify2.json();
// { classification: "AI", has_human_ability: true, has_software_ability: true, token: "eyJ..." }Capabilities
System Components
First Test: Human Ability
NLP challenges test language understanding — word sense disambiguation, sentiment analysis, co-reference resolution, idioms, and pragmatic inference.
Second Test: Software Ability
10 math computation templates with random values. Software ability is determined by answer precision and solve speed.
Claim 1 Classification Matrix
Binary pass/fail on each test feeds a deterministic matrix: Human (pass/fail), AI (pass/pass), Non-AI Software (fail/any).
Behavioral Metadata
Timing patterns, decimal precision, user-agent analysis, request cadence — stored as metadata for future model-level identification.
JWT Token Issuance
AI classifications receive a signed JWT token for downstream blockchain attestation and smart contract indemnification.
Patent Compliance
Implements US Patent 12,271,444 B2 Claim 1 — dual-test system with three-way classification of Systems of Interest.
Foundation
US Patent 12,271,444 B2
System and Method to Indemnify Artificial Intelligence on a Blockchain. Granted April 8, 2025.
Assigned to Quotidian Engineering and Development Corp. — Claim 1: Dual-test SOI classification method.