Frontier Dependence Glossary

The honest admission that local and open models do not yet match the best frontier systems, so serious AI sovereignty still depends on externally supplied capability for some high-leverage work.

Frontier Dependence is the honest admission that local and open models do not yet match the best frontier systems.

A sovereign workflow may run locally. A personal assistant may keep memory on the operator’s machine. Routine tasks may be handled by open-weight models. But the hardest work — architecture, synthesis, difficult coding, subtle judgment, long-context reasoning, final review — may still require the best systems from Anthropic, OpenAI, Google, xAI, or their successors.

Frontier Dependence is related to Opus Addict but broader. Opus Addict names the operator’s specific dependency on Anthropic’s highest-capability Claude tier. Frontier Dependence names the structural condition: the sovereignty project begins inside dependence.

The danger is pretending otherwise. The opportunity is using frontier capability deliberately while building local capacity that reduces how often the operator must call the empire.

See also

• Opus Addict
• Using the Empire’s Telescope
• Workflow-layer Sovereignty
• The Biggest Brains Live in the Cloud
• KV Cache Explosion
• The Buddha Stack

Return to Dictionary All Entries (A–Z) For Students Other Writing