Could a computer scientist build a brain?
stankerstjens.github.io
August 15, 2026
36 min read
🔥🔥🔥🔥🔥
48/100
Summary
A program is proposed to enable a single cell to build itself into a brain using only genomic information. The program must be compact enough to fit within a genome and efficient enough to complete development in a timely manner.
Key Takeaways
- A computer scientist can approach the problem of brain development by writing a program that a single cell executes to build itself into a brain, constrained by genomic size and developmental time.
- The genome must specify a developmental program that is compact enough to fit within about 1 gigabyte and must complete the task within a year.
- The design of the brain may be recoverable by understanding how a genome specifies neural development, rather than through disassembly of the brain itself.
- Current machine learning methods do not effectively generate large functional networks from compact descriptions, highlighting a potential area for improvement inspired by biological development.
Community Sentiment
Positives
- The exploration of cognitive properties in high-dimensional systems opens up exciting possibilities for AI development, pushing the boundaries of what we consider intelligence.
- Commenters appreciate the complexity of modern AI specifications, suggesting that while the foundations may not be complicated, the real challenge lies in innovation and scale.
Concerns
- Skeptics point out that traditional computing models, like the von Neumann architecture, are ill-suited for replicating the nuanced processes of biological cognition, limiting AI's potential.
- There's a sense that current approaches to AI may be missing critical aspects of how brains function, and simply simulating biological processes won't lead to true understanding.
