Hope and Fear: the New Industrial Revolution of Generative AI is here

Three years ago I wrote an essay on the future of AI — I warned that huge changes were coming and that nobody was ready. Now, those changes (and a lot more) are here. And despite all of the hype, I believe we are still drastically underestimating the impact of generative AI; Nothing less than a second industrial revolution is at hand, with all of the promise and disruption that entails.

Introduction

In 2020, I wrote an essay titled “Deepfakes Mean the End of Shared Reality, and Nobody is Ready,” highlighting some of the scary and (at that time) largely unanticipated threats posed by advanced AI. The essay didn't get much traction, I think because at the time it was seen as too far-fetched.

Boy, what a difference a few years make. That “far-fetched” future is here with a vengeance. In case you’ve been asleep for the last six months, here are some examples:

Midjourney and stable diffusion enable the creation of promptographs (text-prompted, AI-created images) that are indistinguishable from real photographs. AI music tools are being used to create hits (e.g., AI-generated Drake and The Weekend’s “Heart on my Sleeve”) that copy the vocal style and production of real artists. LLMs (Large Language Models) and Generative Pretrained Transformers (GPTs) are writing code and content that rival skilled human experts.