“Humanity is a kind of biological boot loader for A.I.” - Elon Musk
Humans have long thought that machines will never be as creative as us, yet as we build more sophisticated systems, we are surprised by what our machines produce.
Our last blog was written by a neural network built to detect fake news content written by other neural networks. In this blog we will explore the ramifications when humans are no longer the primary actors responsible for media content generation.
Dramatic Reduction in Production Cost
Last week’s blog took 1/5th of the time to produce than other blog posts published this year. The model generated content had near perfect grammar and the narrative was on point; it would only require minor effort for a human to clean up and send off to production. It is still early days with this technology and if machine written content improves like image recognition and editing, there is no doubt that human and machine content will be indistinguishable to the untrained eye. This potential for collaborative and fully automated forms of content generation will make it extremely inexpensive to pump out content. The effective price of machine written content will be the compute time to generate it, which for today is pennies. Audio and video are not far behind following the same evolution we see in text. When content is so easy to produce, the core question to ask is; who will have the time to consume it all?
With face filters advancing to a state where they are almost indistinguishable from reality, we are not far from a time when we will see purely fictitious characters rise in popularity. Reanimating popular figures of the past may also see a rise in popularity. The line between what is real or not, will fade. How long before a development team behind a digital actor is given an academy award?
A Future Full of Personalized Content
Recommendation engines are the backbone of some of the most valuable companies in existence. Today content is selected for you, soon, content will be created for you. The dramatic cost reduction in content creation thanks to machine learning systems will allow content to be written and produced on demand. Think Black Mirror’s: Bandersnatch but on steroids. Imagine an entire series based on your preferences and decisions.
How to Stay in the Game
The future will be filled with more and better targeted content chasing the same eyeballs. Mountains of content will be produced and never experienced. Using chess as an example, the best teams are composed of both players and machines, therefore, winning firms will leverage the strengths of A.I. with a human touch.