By Chris Coniglio, Head of Innovation at North Kingdom
In the past, information of any kind was scarce. Before the printing press, it was valued in a way we wouldn’t understand today. Few people had the skills to create written or illustrated information, and even fewer had the means to distribute that information broadly. Literacy was far from the norm and access to information or distribution was reserved for a select few.
Today, information is more than commonplace. It’s piped into our consciousness through feeds designed to drive engagement by companies monetizing attention. Desktop publishing and Web2.0 have given the world the tools and means to not only create but distribute their work with the world at the push of a button.
However, the other edge of the sword is that with decreased scarcity comes decreased value. Content has global reach within milliseconds of its inception and often sparks a multitude of threads upon its creation on multiple platforms. A single post can generate entire libraries worth of opinion, comments, reaction videos, stitches, and memes. Because information is more accessible, and because there is more of it, it is less valuable. It is more information than even our machines can meaningfully process, let alone our meager biological hardware. For the present however, the authors of most information and content are people. We write, draw, photograph, video, and model a massive amount of content daily but the quantity of information that humans generate is a pale blue dot in the night sky in the set of all possible content which could be generated.
Tomorrow, machines will be capable of generating a quantity of information orders of magnitude greater than the sum of all user generated content today and at a similar or higher level of quality. If we follow the line describing the value of information over time from cuneiform to TikTok into the future, we can predict that the value of information approaches zero. In other words, the death of content.
Seventy two years ago, the father of cybernetics Norbert Wiener predicted that machines would do more and more things which used to be exclusive to human beings. He showed a line between two points. The first point: during the industrial revolution, machines gained the ability to work as with the steam engine and the bulldozer. In the early twentieth century, machines gained the ability to calculate as with Wiener’s anti-aircraft guns, and Turing’s enigma machine. Now finally in the twenty-first century machines have gained the ability to create with generative adversarial networks and latent variable models. Machines are now able to write text, generate images, video, audio, 3D content, and they’re able to do it increasingly well and an unprecedented scale. It is not the third wave of AI but the third wave isn’t necessary to produce a seismic shift in our relationship to information. What does this entail for the future? Let’s examine three possible implications.
First, it means that we will see an increasing shift from the paradigm of information or content as value to one of algorithms as value. When people have their own rich, detailed, and dynamic worlds filled with hyper-personalized libraries of content and information created for them by machines, what is valuable to them? The value increasingly becomes how to search, filter, sort, and prompt content rather than the content or information itself. Even today, how we are able to interact with, contextualize, and navigate information is of greater importance than the meager possession of it.
The second implication relates to the web. We saw a first iteration of the world wide web in which an interconnected web of documents enabled a new type of communication and an unprecedented proliferation of information. This was followed by a second version defined by the scale of content which the collective internet hive-mind was able to churn out, otherwise known as user generated content. We are standing at the edge of a third version of the world wide web defined by the shift from user to machine generated content. The level of utility, personalization, entertainment, and value that an MGC driven web will create casts a shadow which dwarfs today’s buzzwords in their entirety. To view the next iteration of the web as merely decentralized, solely 3d content, or simply connecting Fortnite with Minecraft underestimates the scale and breadth of the eventual but gradual shift in much the same way thinking of the internet as merely Geocities might look to us today.
Finally, user generated content gave us many of the largest companies of today: Google, TikTok, Twitter, Instagram, Snapchat, Youtube, LinkedIn, the list goes on. These companies have developed algorithms which match content created by all users to the interests of a given user. It’s similar to linear regression. Given a set of data points about a user, we can predict which of a set of possible points best matches their unique preferences. However, what happens when we no longer have to reach into a grab bag of content provided by users, and instead can generate content dynamically and individually for each user? We can create content so incredibly hyper-personalized that engagement becomes a given, rather than a goal. Suddenly, we’re starting to approach the video tape from Infinite Jest.
These are only a few considerations around the impact that the continual growth of machine capability has on human affairs, business, and culture but it’s by no means an exhaustive list. It’s incredibly inspiring and terrifying to see how this shift is unfolding today and into the future.
Content is dead and as we hand over our pencils, cameras, paintbrushes, and notebooks to the machines we build, it will be more than just interesting to find out if its successor is the benevolent artist or the tyrant.