A Time of Global Change

This text was written in October 2022, before generative models became everyday tools. It is published as it was: a forecast made at the time, not a statement of where we stand today.

Neural networks will soon change the familiar idea of what a profession is and what creative work means. Demand for programmers, designers and other specialists will shrink: much of what requires their work today, a model will be able to do from a brief. What remains are the operators who set the task and correct the result. Car assembly was once sought-after and prestigious work, and then people were largely replaced by robots. Intellectual work is heading for the same kind of change.

Next will come directors, screenwriters and writers. Already today, some fan fiction written by ordinary people turns out to be more interesting than the original work. But inventing a story and turning it into an actual film are completely different capabilities. A person may have talent and imagination, and no budget, no crew, and no access to the film industry.

Neural networks will be able to remove that barrier. The script, the visuals, the acting — all of it will be generated once you can formulate the right request. Anyone with an idea and the ability to develop it will be able to make films. Talents that simply never had a way to show themselves will start to surface. Less will depend on where you were born and which opportunities you happened to be given.

Imagine someone wants to shoot their own version of Snow White. They make it with a model in ten minutes and post it. One shot in noir: Snow White watching a dramatic scene to Don't Stop Believin'. An unusual combination of image, action and music — and that scene breaks the internet. The person becomes known, because their idea landed with the audience.

Masterpieces will be able to appear not only in studios, but in the personal libraries of ordinary users. On platforms like Netflix, everyone could have an account where they post their own films, and the algorithms would help surface the work that is genuinely worth watching.

But a problem appears here: the easier it becomes to create, the more content there will be. And the faster it will start to lose its value.

You can already feel this with games. There used to be one game, and you played it to death. Now the choice is enormous, but because of the constant stream, fewer and fewer things really catch you. And when something does, it does not hold for long. What happens when films, books and any other work can be produced almost instantly?

If everyone can make films, how will anything become a cult work? Will we drown in an endless stream of stories? Perhaps traditional cinema will survive as an elite art for a narrow circle, and the director of the future will be the person who knows how to formulate their intent for a model.

And these changes will not stop at creative work. I expect that neural networks will be able to act as judges: to go deeply into the circumstances of a person's life, analyse their personality and deliver a verdict. They may even learn to reconstruct details that memory has lost.

We are beginning to develop very quickly, geometrically. Ahead of us is a time of global change. A time when realising your imagination becomes far easier, and what you create loses its value far faster.