❔How did ‘The Natural Selection of Brain Wiring’ inspire neural networks in Machine Learning?

Reeshabh Choudhary
3 min readAug 23, 2023
A Neural Network

👆Background: Synaptogenesis (brain development) starts in spinal cord, during its fifth week of gestation (embryogenesis) and in cortex after seven weeks. A synapse is the communication point between two cells (neurons). During first 2 years of birth, around 15,000 synapses are produced on every cortical neuron, which corresponds to a rate of 1.8 million new synapses per second. In short, brain is a dense forest or mesh and for its cognitive abilities, it is crucial that right set of synapses are wired while processing any signal.

⛓Brain processes any action or experience when particular few hundred neurons come together to form synapses with each other for transfer of signal. This process is called Brain Wiring. Our senses (such as sight, hearing, touch, taste, and smell) receive information from the external environment. This information is transmitted as sensory signals to the brain for further processing. Neurons process and transmit information through a combination of electrical and chemical signals. When a neuron receives a signal, an electrical impulse travels along the neuron’s axon. At the synapses, chemical neurotransmitters are released to transmit the signal from one neuron to another. The transmission of information between neurons occurs through synapses. Neurotransmitters released from the presynaptic neuron bind to receptors on the postsynaptic neuron, allowing the signal to be transferred from one neuron to another. Neurons integrate signals from multiple sources, making complex decisions based on the combined input. This process enables the brain to perform various cognitive functions, and operates as a feedback loop.

🤔Out of billions of neurons and a quadrillion or so synapses, how right set of synapses and neurons get properly hooked together during Cognitive Inference?

🔬Science: Brain’s cognitive ability take time to develop. For Example, a child is not able to develop motor skills required to catch a ball, just right after the birth. So, it learns from experience, and keeps correcting its synapse and neuron connection process. With more experience to nature and emotions, brain choses fittest or most useful synapses from the forest of quadrillion synapses.

🧠Use it or Loose it: More brain functions, more synapses are active, greater amount of neurotransmitters are released, better brain’s cognitive abilities. Adhering to Darwinian selection process, Brain starts pruning inactive synapses right after 2 years of birth and continues this process till adolescence.

So, it over produces synapses and the cuts loose them based upon experience, something similar to Back propagation, where a Neural network adjusts the network’s connection weights based on the discrepancy between predicted outputs and actual target values, iteratively refining the network’s ability to make accurate predictions.

It is no wonder that a young child’s environment while growing up is largely responsible for shaping up his abilities through out the lifespan. Children lose around 20 billion synapses/day between early childhood and adolescence. Pruning helps in streamlining the thought process, once of the reason why we take time to be mature!

However, this also means that with maturity comes a less flexible and less creative brain. To cut short: in the age of Mama’s Boy and Daddy’s princess syndrome with pampering shower, it seems very hard to produce strong leaders of younger generation, as their synapse formation is mostly under utilized. Sounds harsh but hard times for children is actually a good thing. It develops their character and creativity. But, will parents understand science or will keep living with their ‘feelings’?

Time will tell.

#AI #neuralnetworks #brain #ML

--

--

Reeshabh Choudhary

Software Architect and Developer | Author : Objects, Data & AI.