AI & Computing Realm
The study of artificial minds, computational architectures, and the boundaries of what machines can know and do. This realm spans from the mathematical foundations of information (Turing, Shannon, Gödel) through physical hardware (neuromorphic chips, quantum processors) to the emergent phenomena of large-scale AI systems.
Computing is the realm where biology meets physics meets engineering. The brain is the existence proof that cheap, efficient, general intelligence is physically possible — and every architecture question in AI is ultimately a question about what the brain figured out that we haven’t yet.
Active Pages
Hardware & Architecture
| Page | Summary |
|---|---|
| tech-neuromorphic-computing | Brain-inspired chips — Intel Loihi 3 (2026), IBM NorthPole, Hala Point; 1,000× more energy-efficient than GPUs for sensory tasks; spiking neural networks |
Coming soon
- Transformer architecture — how attention actually works; why it changed everything
- Quantum error correction — the key barrier to useful quantum computers; 2025 progress
- Cellular automata — Conway’s Game of Life and emergent universes
- The halting problem — Turing’s undecidability and its implications
- Swarm intelligence — ant colonies, bee hives, bird flocks computing without a leader
Cross-Realm Hotspots
This realm connects most densely with:
- biology — the brain is the reference architecture; octopus distributed cognition is the existence proof that neuromorphic design works; gut-brain axis as biological analog of distributed AI
- physics — brain criticality, turbulent dynamics, quantum computing, information as physical phenomenon (Landauer’s principle)
- history — Antikythera Mechanism as first analog computer; Jacquard loom as first programmable machine; computing history is a history of physical instantiation of abstract logic
- textiles — Jacquard → Hollerith → IBM punch card → digital computing; the lineage is literal
- philosophy — consciousness, the Chinese Room, free will under neural determinism, what “understanding” means
Key Tensions & Open Questions
- Symbolic vs. sub-symbolic AI — transformers learn statistical regularities; do they understand anything? (Chinese Room argument)
- Energy wall — training frontier models requires gigawatt-hours; neuromorphic chips offer 1,000× efficiency gains but only for constrained tasks
- Criticality — brains operate near a phase transition between order and chaos to maximize information processing; should neuromorphic chips be tuned to criticality rather than stability?
- Embodiment — does intelligence require a body? Robotics, embodied cognition, and the gap between LLMs and physical agents
- Scaling limits — is transformer scaling hitting diminishing returns? What comes after scaling?
- Alignment — how do you build a system that wants what we want, when “what we want” is itself contested?