Digital computers will never achieve human-level intelligence because they are built on the wrong kind of math.
April 29, 2026
Original Paper
The Biological Prerequisite for Artificial General Intelligence: Why Probabilistic Computation Cannot Produce Cognition
SSRN · 6481200
The Takeaway
The AI industry is currently betting that more data and bigger GPUs will eventually create general intelligence. This research argues that probabilistic computation is fundamentally incapable of producing true cognition. Human-level intelligence requires the specific physical properties of biological neural substrates that digital circuits cannot replicate. Scaling existing models will only result in better calculators, not thinking machines. If this theory holds, the path to AGI will require a total shift toward biological or neuromorphic hardware.
From the abstract
This paper argues that Artificial General Intelligence (AGI) and its theoretical successor, Artificial Superintelligence (ASI), are fundamentally unachievable through probabilistic computation alone, regardless of model scale, architectural innovation, or computational investment. We establish that all current AI systems-including large language models, diffusion models, and reinforcement learning agents-operate through statistical pattern matching over structured data representations. While the