Executive Summary
Artificial intelligence is currently evaluated by its outputs—optimising actions, selecting paths, and transforming information into behaviour. In The Web That Knows: Intelligence Beyond Computation, Beyond Brain, Genie Capital argues that this reductionist view mistakes the outputs of intelligence for intelligence itself.
To redefine intelligence, the report establishes a rigorous three-layer framework: computation (supplying the mechanism without stakes), cognition (supplying the embedded, self-maintaining stakes), and consciousness (supplying the felt integration and perspective required for true abstraction). Measured against this criterion, current AI is super-computational but pre-cognitive—a system with immense processing power, but no world to inhabit, no self to maintain, and no gaps to feel. True intelligence, by contrast, is found in living, distributed systems like the mycelial networks of a forest, which demonstrate genuine agency, memory, and nested relationality.
Core Insights
- The Fallacy of Computational Reductionism: Treating large-scale symbol manipulation or parameter scaling as true intelligence collapses categories. Computation processes information without skin in the game; true cognition requires stakes and vulnerability.
- The Three-Layer Architecture: Intelligence requires the complete stack working in concert—computation for mechanism, cognition for survival and stakes, and consciousness for the perspective needed to orient toward the absent.
- The Mycelial Mind as a Living Benchmark: Fungal networks function as distributed minds, maintaining complex trading relationships, signaling unseen environmental threats, and demonstrating genuine abstraction and nested betweenness.
- Damage vs. Wounding: While current AI models can only be "damaged" through corrupted parameters or switches, embedded living systems can be "wounded" by the loss of relational dimensions—experiencing a cognitive phantom limb when severed from their ecological partners.
Value to Stakeholders
- For AI Researchers & Founders: Moves the paradigm beyond raw scale and parameter expansion, providing a philosophical and structural blueprint for what authentic intelligence and abstraction require.
- For Executives & Boards: Offers profound insights into systemic resilience, distributed governance, and recognizing the limits of purely mechanistic decision-making tools.
- For Investors: Highlights the vital distinction between super-computational software tools and true adaptive agency, sharpening strategic foresight in technology assessment.
Download the Full Report
Published by Genie Capital Ltd under Creative Commons (CC-BY-NC-SA 4.0). Re-use and distribution are permitted with proper attribution for non-commercial purposes.