When I was young, I moved a lot. Everyone knows it’s a little tricky to make friends, but I hacked it – and found you can make friends in the “wrong way” by sharing your answers for tests. This worked really well, until one teacher noticed all of the students around me were getting 100% on their tests. Instead of assuming I was a great mentor, or that I was going to be a teacher someday, she decided to give the class a pop quiz. That pop quiz, required us to show our work. Turns out it’s really easy to cheat when you’re copying an answer, or memorizing an answer, but as we all know, to really understand something means you can show your work.
Showing your work is the gold standard for physics tests globally, but it’s not the gold standard for Artificial Intelligence (AI). In fact, the industry standard for Artificial Intelligence today is based on neural nets which model the human brain and are inherently a black box. I had the opportunity to talk about the dangers of not showing your work in AI in my recent TEDxRaleigh talk. I invite you to take a look.