LillyThe rise of another AI model like Dall-E has some people wondering if it will replace graphic designers. GPT-3 didn’t replace writers, GitHub Copilot isn’t going to replace engineers, and Dall-E won’t be replacing visual artists either.
What is machine learning good at?
Machines are good at automating and scaling manual tasks. A machine processes a 12 hour Twitch stream quickly and efficiently. They don’t get bored like humans.
Machine learning excels in a well defined problem scope, when you can tell the machine exactly what you want. For example, machines can follow instructions like:
- “Go through this Twitch stream and return timestamps when someone is happy. Here’s a definition of happy.”
- “Take this timestamp and find the closest silent moment from before and 30 seconds after. Here’s the threshold for silence. Clip just that moment.”
- “Watch this clip and transcribe what the people are saying. Then summarize what’s happening into 1-2 sentences.”
The above process outlines the work required to create social media clips from a long Twitch stream. Without software, it’s work that creators and editors often have to manually do if they want to prevent engagement loss for their long form content.
What can’t machines do?
Humans have to be the ones strategizing, thinking abstractly and supplying creative direction. Their expertise on the “who”, “what”, and “why” can’t be outsourced to a machine.
Content strategy is figuring out what types of content to make and for who. That planning is necessary for content to have the most impact. Writing code tutorials for a print newspaper doesn’t make sense. On the other hand, creating video tutorials for YouTube is a great place to reach many developers. Understanding what motivates the audience (the “why”) is a human to human experience.
By automating the boring and time consuming tasks, humans can get back to doing the work that energizes and motivates us.
Are you doing something boring, manual, and necessary with your content? Let us know - we’d love to help.
