Superintelligence or Super-Equipped Intelligence?

The term "superintelligence" has achieved near-consensus among AI research leaders, yet the claims gathered under this label rest on a category error: treating capability amplification as architectural transcendence. This paper argues that what researchers describe when invoking superintelligence, including hardware advantages, processing speed, recursive self-improvement, and computational scale, are instances of super-equipped intelligence: systems with the same cognitive architecture operating with better resources and faster execution. The argument draws on the Coherence Resolution Mode (CRM) framework, which identifies intelligence as the capacity to resolve incoherence under constraint through eight functionally complete modes arising from three constraint types and four resolution mechanisms. If this framework correctly characterises the architecture of intelligence, then optimal calibration represents a ceiling that scaling cannot transcend. Increased computational power changes which outcomes can be reached, not the cognitive operations used to reach them. The terminology carries practical weight. Superintelligence implies preparing for incomprehensible cognitive superiority. Super-equipped intelligence requires preparing for familiar dysfunction operating at civilisational scale and electronic speed. Clarity about what advanced AI systems actually are enables better responses than mystification about what they might become. Version note: Version 2 – corrected attributions to G. Hinton in Sections 2.1 and 2.2; no substantive change to the argument.