100 Days of ML — Day 29 — What happened to that convolutional neural network you promised us…like 10 days ago?

Yeah…about that.

Sorry.

I’m so sorry.

Basically, I wanted to auto-gen photos I was using for a personal Instagram challenge. My wife and I took around 400 photos of me smiling in various styles and settings with various energy. I thought it’d be neat if I could build a CNN to auto-generate that stuff for me.

Problem 1: A GAN is what generates stuff. That’s…actually what it stands for: “Generative Adversarial Network”. It’s backboned by CNN tech, but I said I’d use a CNN to do things and the initial premise was faulty. I’m so sorry.

Problem 2: I don’t have the local tech to handle the amount of data that I have. CNNs and GANs on my practice problems were 28x28 and usually black and white. These photos are up to 1920x1080 pixels and with three colors. They are not the same size, so the data wrangling would’ve been a huge, time consuming mess, and I would’ve needed to purchase GPU or TPU time and I just don’t have the money for that. AI journalism isn’t paying the bills.

So, for me, it’s back to the drawing board, mostly getting new clients so that I can have an income to study this stuff in great detail.

Push this back to the end of the challenge, I guess.

Jimmy Murray is a Florida based comedian who studied Marketing and Film before finding himself homeless. Resourceful, he taught himself coding, which led to a ton of opportunities in many fields, the most recent of which is coding away his podcast editing. His entrepreneurial skills and love of automation have led to a sheer love of all things related to AI.