When Code Reads Neurons: AI and the Brain Share a Mathematical Language

Forum topic · ✨步子哥 · 2025-12-20

Summary

This forum post explores the emerging convergence between artificial intelligence and neuroscience, arguing that silicon-based AI models and carbon-based human brains use strikingly similar mathematical representations when processing information. Drawing on BCI pioneer Max Hodak's views, it discusses MIT's Platonic Representation Hypothesis, which holds that differently architected AI models converge on a shared representation of reality, and notes evidence that large language models handle multimodal data similarly to the brain's semantic hubs. The post highlights Science Corp's Prima retinal implant, which restores functional reading vision to blind patients via subretinal photovoltaic chips; in a 38-patient NEJM-published trial, patients gained an average of 25.5 letters on the ETDRS chart and 84% regained reading ability. It also covers Hodak's 'third cerebral hemisphere' concept of brain augmentation via external hardware, the binding problem of consciousness, ethical issues such as neural rights and identity, and a speculative timeline from medical restoration (5–10 years) to cognitive enhancement (10–20 years) and mind-machine fusion beyond 20 years.

When Code Reads Neurons: AI and the Brain Share a "Mathematical Language"

This post from zhichai.net examines the convergence of AI and neuroscience, based on recent remarks by brain-computer interface (BCI) pioneer Max Hodak and related research.

Key points

Representation convergence: AI models built on silicon chips and biological brains appear to use a strikingly similar "mathematical language" when understanding information.
Platonic Representation Hypothesis: MIT researchers propose that AI models with different architectures, trained on different datasets, converge toward a shared representation of reality — driven by task generality, model capacity, and simplicity bias.
Neuroscience corroboration: MIT findings suggest large language models process multimodal data using mechanisms similar to the brain's "semantic hubs," like the anterior temporal lobe.

Technical path forward

The binding problem: Hodak argues understanding how the brain integrates distributed processing into unified conscious experience is key to advanced BCIs.
A "third cerebral hemisphere": He envisions augmenting the brain with external hardware, extending perception. Future brain boundaries might include "four hemispheres, or a device, or a group of people" connected via networks.

Case study: the Prima implant

Science Corp's Prima retinal implant uses subretinal miniature wireless photovoltaic chips that stimulate still-healthy bipolar cells, paired with image-converting glasses and near-infrared signal systems. Per a study in the New England Journal of Medicine, a 38-patient multicenter trial showed:

| Metric | Result | |---|---| | Average visual improvement | +25.5 letters on the ETDRS chart (>5 lines) | | Reading recovery | 84% of patients regained ability to read letters, numbers, and words | | Significant improvement | 80% improved by ≥10 letters at 12 months | | Safety | Most adverse events resolved within 2 months; original vision not harmed |

Hodak: "Patients go from being almost blind to being able to read books and do crossword puzzles."

Ethics and philosophy

Identity crisis: If memory and personality can be digitized and transferred, "which is the real me" becomes urgent.
Fairness: Enhancement may create a divide between "augmented" and "natural" humans.
Neural rights: An emerging field covering mental privacy, personal identity, and free will.
Physicalism: Hodak holds consciousness is a product of neuronal activity; mastering "binding" could enable engineered consciousness and brain augmentation.

A speculative timeline

1. Short term (5–10 years): Medical restoration — neurodegenerative disease, spinal injury, blindness. Morgan Stanley predicts first commercial BCI devices approved around 2030. 2. Medium term (10–20 years): Cognitive enhancement and new senses for healthy people; Hodak's "third hemisphere" becomes reality. 3. Long term (20+ years): Consciousness digitization, human-AI fusion, redefining life and death.

Conclusion

The post argues we stand at a crossroads of human evolution: as AI learns to read neurons and brains gain external "hemispheres," balancing technological progress with human dignity will determine whether we successfully navigate this cyberpunk future.

Sources cited in the original post include MIT neuroscience research, Science Corp, and the New England Journal of Medicine.