By: Theodore Tong

AI, or artificial intelligence, has made many important advancements in the past years. AI bots, like OpenAI’s ChatGPT, can write programming code, essays, trivia questions, and poetry.

However, AI’s weakest pillar is math. As Kristian Hammond, computer science professor and AI researcher at Northwestern University says, “The A.I. chatbots have difficulty with math because they were never designed to do it.” AI is definitely better off doing English writing homework than algebra.

According to the American Enterprise Institute, “Paul von Hippel, an associate dean at the University of Texas, pointed out ChatGPT’s inadequacies in teaching Geometry, while The Wall Street Journal highlighted similar struggles of the AI tutor Khanmigo.”

In 1940, the Electronic Numerical Integrator and Computer, or ENIAC was invented, and so were many other computers. At that time, computers were simply described as “math on steroids.” Computers calculate numbers within an instant. Unfortunately, AI wasn’t as good as its predecessors. It had hit a solid brick wall.

That is, until in the mid 2010’s, a new technology, called a neural network, modeled after the human brain, broke that wall. By using a different approach, it began to strike gains.

The neural network works by giving the AI vast amounts of data, like books, websites, news articles, and pdfs. Then, when the AI bot is asked a question, it goes through all the data, and is able to predict which word or phrase is most likely to come up next, just like humans.