Generative AI should encourage creativity, but delivers simplification

In the early days of generative AI, less than two years ago, we would visit ChatGPT to marvel at all the things we could create: silly haikus, poetry, and scenes from “Seinfeld.” Generative AI promised unlimited output, and we used it to create more media, without much question. In the blink of an eye, things came into existence.

But the fun soon ended, and a devastating new functionality took over: using generative AI to compress all existing material and present it in simplified form. We don’t live in a new creative age. We live in the age of abstracts.

It’s easy to look at this advancement with skepticism: This is what we get from AI tools that haven’t (yet) reached their creative potential. But perhaps these tools are simply responding to human desire.

The more we create digitally, the faster we want to digest all that content. And this started long before anyone had heard of large language models (LLMs).

It started with bloggers chewing long newspaper articles into a juicy pulp and serving readers only the most palatable or spiciest parts. Soon, the internet in general became faster and easier to read, too.

YouTube has added chapters and summaries to long videos, and Google searches now take you directly to the relevant part of a video.

As podcasts have gained popularity, so have playback controls: they now allow you to listen at 1.2x, 1.5x or even 3.5x speeds. On Spotify, record labels release pop songs alongside their sped-up versions, perfect for sharing on social media where attention spans are short.

On TikTok, some creators have gone even further: you can watch recaps of entire movies in a handful of two-minute clips. The ideal LinkedIn post? A short, concise piece.

By trying to force our brains to compete with the infinite capacity of the machine, we only move faster and faster.

Digital tools have streamlined media production, and the Internet has streamlined distribution. But instead of engaging with articles, videos, and podcasts, we simply devour the “content.” Faced with a buffet of words, images, and sounds, we sample everything but savor nothing.

LLMs have automated this phenomenon and are summarizing virtually every aspect of life faster, more affordably, and in some cases better than humans. Zoom summarizes your meetings, Microsoft summarizes your spreadsheets and documents, and Apple will soon do the same for your emails and notes.

Adobe can summarize any PDF and restructure any data into a chart, while Canva reduces your endless branding document into concise slides. Amazon condenses thousands of product reviews into a few pros and cons.

Google, Bing, Anthropic, and Perplexity summarize information from websites, intercepting potential readers. LinkedIn offers a button to summarize any article on your feed – this way, you don’t even need to read someone else’s post with your summary.

In a way, summarization provides a necessary efficiency: it keeps us from drowning in our ocean of content, a never-ending tide. But when you employ AI as a mass compression algorithm for culture, something inevitably gets lost in the compression.

When CDs first arrived, they were billed as a perfect digital copy of the music. The digital version, however, was an engineer's interpretation of the music: the bits of information that captured the unique timbre of a vocal or a drum beat were left out. Complicated frequencies were boiled down into simplified code.

Decades later, we all recognize that vinyls—sound waves pressed into plastic, scratched by a needle—are much closer to the original recording. Likewise, GPT compression—and other compression technologies—rob us of the voice, humor, reasoning, and nuance that come with real human thought.

The more we create digitally, the faster we want to digest all that content.

There’s nothing wrong with cutting back on time and cutting out the clutter, especially if it takes us away from the screen for a while. But we’ve all read the studies—or summaries of them—about the addictive properties of digital exposure. Teens are currently spending up to nine hours a day staring at pixels.

The goal of all these efficiencies is to give us more time to access even more content. And since the media we’re digesting tends to be the most complex and most nourishing, our extra minutes are spent devouring empty calories.

All I know is that no one will ever listen to a podcast at 1x speed again. By trying to force our brains to compete with the infinite capacity of the machine, we only move faster and faster. Down the road, we will be surprised that we ever had time to live at ease.

---

ABOUT THE AUTHOR

Mark Wilson is a senior writer at Fast Company. He has been writing about design, technology, and culture for nearly 15 years. learn more

---