Will AI replace <insert XYZ>?
Quick, 5-minute thought dump on the future of work
Today I animated a video of a girl applying makeup, but suddenly, her reflection started doing terrifying things that she wasn’t. Is this a metaphor for AI taking over humans?
With each passing day, advances in AI has made people question what really is the significance of their work in the long run. People wonder if their job will remain once AI becomes better or even if the industry itself will remain. A month or two ago, it was all about coding, software engineers sharing tweets about the day of doom coming or threads that started with “Coding is so over”. This week, with the advent of easier image generation, there are similar statements about graphic designers and animators. Why will I pay for someone to make an animated movie if I can make it all by myself at home?
My Experiments with Art
As someone who’s long been interested in animation and motion graphics, I saw this week as an opportunity to dabble and create some easy animated projects. The goal was (still is) to create a music video with a storyline I have in mind. I started with the bare minimum effort by generating a few images and feeding it into an AI tool (in fact, a bunch of different tools) that could convert images into videos. The output was commendable but no where near what I wanted. There were a lot of hallucinations, random characters walking in from thin walls, characters suddenly disappearing. It was both amazing and horrible at the same time - amazing because, I can’t imagine how long it would take to manually make this and horrible because it is no where good enough to post.
But this made me wonder if the model itself was not good enough or if it was a me problem. With code, I am often able to use generative AI to great effect. It has really fastened my workflow and I am able to output much more than I was normally used to. Then why not with animation?
The weight of experience
The advantage of generating code with AI is that I code everyday. I know what I want, I know how it should work and I know what “feels” right. We are often quick to dismiss it but experience is so valuable in AI that we don’t realize whatever makes a tool so efficient for us is that we already know what we want and when it doesn’t work a certain way we are able to steer it to get the output we want.
The word “Copilot” is actually very apt here because that’s what these tools really are - they aren’t replacing you. At the moment AI is more like cruise control than like a self-driving car. It helps immensely if you already know how to drive. But if you have never driven in your life before, this isn’t getting you anywhere real except maybe inside a closed ground.
Is AI reducing the value of work?
I don’t think so, at least, not at the moment. I do feel more empowered than ever to build things faster and better than I could ever do before. It’s enabling me to focus on interesting things rather than doing mundane stuff. And I think that’s how it’s going to be for a while.
Everybody will be prompt engineers, it’s just that what they are prompting will be different. I will be prompting code, because I know what bounds to set, what to make it do and when it does generate, I know how to make it “feel” right. Animators are going to be around and they are going to be using AI to speed up their workflows. They are going to use AI far more efficiently than I, an animation noob, can. They will be able to prompt the AI to generate the “right” feeling videos better than I can.
It’s easy to discount the value of experience and the value of studying something and really understanding it but AI is really going to amplify the individuals who actually know what they want to do and have spent the time and effort to study it.
It’s exciting.
What happens when AI really transcends human potential?
Wouldn’t that be a good thing? The truth is, a lot of the jobs we all do today are all made up. It’s not like all of us are saving lives or making the world a better place. Maybe once AI takes over the mundane stuff like automating emails and collecting data from dashboards, we all can actually focus on what matters? Explore life beyond the Earth? Build time machines? Figure out the purpose of life and do some soul-searching? Maybe it’s a good thing in the long run!
If you liked this short essay, you might also enjoy all the other ways I dabble with AI. For example, I recently translated an app into 100+ languages with AI! Read about it here -
