January 25, 2026

I often place "Artificial Intelligence" in quotations because the cultural meaning ascribed to it is not the traditional meaning. AI could have a number of pages in the dictionary because of the many meanings it has been assigned. Automatons and robots with intelligence surpassing humans has been a tale as old as humanity itself. Talos the Bronze giant of Crete could be an example, judging those who would attack his island. It could also have started more literally in 1956 at Dartmouth college where the college first declared their program for artificial intelligence studies. I will note that they believed they could make a "general intelligence surpassing human intelligence within the decade"... Sound familiar?

I like to trace the origins a bit further, to the father of the modern field of Computer Science, Alan Turing. He invented one of the first computers, the Turing Machine, which was capable of making independent decisions based on a set of rules and a tape of input. You could see why I see this as the beginnings of AI. After Dartmouth's new college in 1956, the next big milestone I would point to is the creation of the first neural network in 1958 by Frank Rosenblatt, the Perceptron.

Rosenblatt also represents a diverging path of AI. He believed that the best way to model intelligence was to model the human brain (they didn't know much about the brain in the 60s). And that was how the neural network became the default model moving forward, but imagine if we had taken a different path. What if we had modeled AI after something we actually understood well. Not that the 50s/60s didn't understand the brain, but.... they were RADICALLY wrong. So how can an algorithm based on a model of intelligence that isn't even accurate be intelligent?