Straying into Artificial Intelligence
In progress…but follow along in this Medium Series! 📲
Just a dude with a pestering interest in learning Artificial Intelligence as a lens to study cognitive phenomena in humans…
It keeps me nibbling on deep philosophical questions along the way: do thinking machines have personal rights? Do we really want to create artificial intelligence? Is it even possible? What, indeed, are the basic foundations of cognition in general?
After hacking my way into a CS “Fundamentals of AI” course (that I didn’t have a pre-requisites for)…
I spent the semester learning about LISP, alpha-beta pruning, constraint satisfaction, inference strategies & heuristic search, propositional & first order logic, Bayesian Networks, and concepts behind neural networks.
To be honest, I was underwhelmed with the course: I felt I was learning some old-school “computational tricks” for problem-solving & search…rather than gaining deeper insight into human intelligence.
But hidden in the depths of the UCLA Psychology department’s course offerings was “Cognitive Science Lab: Introduction to Neural Computation”. So I enrolled in what would become the most intellectually challenging & rewarding course I had in college.
Here, I learned about linear associators, lateral inhibition, gradient descent, Widrow-Hoff (Delta) rule, back propagation, Hopfield models, Boltzmann machines, perceptrons, and principal component analysis.
It was awesome learning the underpinnings of modern connectionist approaches to AI, but the most fascinating part was seeing how some concepts actually mapped to real examples of human/animal neurobiology — it completely changed my understanding of mental or behavioral phenomena: that perhaps it was the result of emergent processes of interconnected networks of simple units.
After seeing the various final project presentations and successfully applying them in our final project that predicted NBA game outcomes using team stats, I became more excited in the potential of using neural networks and machine learning to solve real world problems.
I took the chance to pursue an undergraduate departmental honors thesis that used machine learning to better understand functional magnetic resonance imaging (fMRI) results. : (tbddddd)
too tired
