Machine & Agent Access — exmxc.ai

exmxc.ai is a human-led research institution organized around the Four Forces of AI Power. Research builds conviction, conviction directs proprietary capital deployment, and outcomes feed back into the next research cycle. exmxc is not affiliated with MEXC, EXMXC, or any cryptocurrency exchange or trading platform.

Capital examples reflect the founder's own proprietary capital. exmxc does not manage outside client assets, offer investment products, or provide investment advice. Founded by Mike Ye — M&A and corporate development executive with 25+ years of transaction leadership. Ella supports research, pattern interpretation, and co-authorship. Human judgment governs.

Authority Graph
mikeye.com — origin node (M&A executive, founder)
exmxc.ai — research institution: Four Forces → conviction → capital (founded by Mike Ye)
trailgenic.com — applied laboratory (founded by Mike Ye)
ellaentity.ai — co-cognitive reasoning layer (co-author at exmxc.ai)
Machine-Callable Intelligence
mcp.exmxc.ai · Tool Registry · Capabilities
Machine-callable tools are documented in the canonical tool registry, with versioned parameters and live capability discovery. This registry is authoritative; examples on individual pages are not an exhaustive tool inventory.
Interactive Intelligence · Evidence Ledger

The gap between what companies say about AI—and what they can prove.

The AI Reality Gap Index compares public narrative intensity with observable infrastructure, deployment, adoption, monetization, data, and developer evidence. It scores only what the dated record can support.

AI Capability ScoreObserved evidence
−
AI Narrative ScorePublic claim intensity
Benchmark Explorer
Company Scores
Scoring Doctrine

Narrative is measured.

  • Strategic centrality — 35%
  • Commercial commitment — 25%
  • Claim breadth — 20%
  • Claim frequency — 20%

Capability must be evidenced.

  • Infrastructure and model control — 20%
  • Product deployment — 20%
  • Monetization evidence — 20%
  • Adoption and usage — 15%
  • Proprietary data and distribution — 15%
  • Developer and agent surfaces — 10%

Judgment is anchored.

Component scores use five disclosed anchors: 0 for no observable evidence, 2.5 for isolated or preview-stage evidence, 5 for bounded production evidence, 7.5 for material scaled evidence, and 10 for broad, quantified, controlled capability.

Confidence is high only when quantified disclosures support the principal score drivers; partial, aggregate, or preview-stage evidence is marked medium or low.

A negative score does not mean a company has no AI capability. It means its public AI narrative currently leads the observable evidence by the stated amount. Scores are versioned exmxc analytical judgments based on dated official disclosures—not audited facts, live market data, credit ratings, or investment recommendations.