The Pioneers of AI: Marvin Minsky and the SNARC

Zahid Parvez
2 min readJan 25, 2023

In 1951, Marvin Minsky and Dean Edmonds, supported by John von Neumann, built the very first neural network computer, the computer was known as SNARC.

The Stochastic Neural Analog Reinforcement Calculator (SNARC) embodied a type of neural network that is designed to learn from experience and improve its performance through a process of trial and error (similar to reinforcement learning).

This machine was well suited for problems that involved sequential decision-making and made use of prior experience, for example — navigating out of a maze. In fact this is the problem Minsky had set out to solve.

The original SNARC system consisted of a set of 40 artificial neurons (specifically Hebb synapses) that are connected to each other in a network. Each neuron receives input from other neurons and produces an output that is passed on to other neurons in the network. The network is trained by adjusting the strengths of the connections between neurons based on the outcomes of previous trials. The training of this network was manually done by the operator who would provide feedback to the machine when it produced the right outcome.

Image of a Hebb synapses (source)

While this type of network is not used commonly today, at the time, SNARC systems are considered to be one of the first pioneering attempts at building artificial…

--

--

Zahid Parvez
Zahid Parvez

Written by Zahid Parvez

I am an analyst with a passion for data, software, and integration. In my free time, I also like to dabble in design, photography, and philosophy.