Modern Media Consumption is Broken

Modern media consumption is broken. The way we read things in 10 years is not going to be the same. That shouldn't be surprising because every 10 years the way we read things has changed. Radio, to records, to broadcast TV, to cable, to websites, to streaming, Podcasts, and blogs, to TikTok.

AI is the newest media beast. What it allows me to do right now is dictate into my computer, split that into multiple media formats, posted on multiple platforms, and expose myself in multiple ways to many different audiences. AI voice to text, AI algorithmic content feeds, and a lot of automation power this.

I'm creating for self expression, so this enables me to amplify that to an insane degree. However, many creators and companies do not create for self expression. They create for visibility. Being seen is another way to generate income. If all that matters is being seen, then what you say and even who says it is not important.

What this means is that new technology like AI will allow more honest creators like myself to fire hose content to every available platform. And while that amplification is happening, less scrupulous creators and companies will use AI to analyze markets, generate content tailored to them, and distribute it automatically.

This media problem is not new. New technology and automation allows people to automate media creation as well. Radio began with regional coverage limiting broadcasts to county wide distribution. A radio program would get recorded and then physically sent to multiple radio stations to increase its broadcast range. Physical media like records followed a similar plan.

Television as well was initially broadcast, and later upgraded to distribution via coaxial cable. These cables were laid nationwide, allowing for media distribution instantaneously on a household by household basis. This now meant that different homes could subscribe to different cable packages, providing them access to customized programming based on what they subscribed to. This was still limited, but it increased the number of channels the average household had from 10 to 100.

With the Internet this number exploded into the thousands, but beyond that. For any individual television channel, you now had the equivalent of accessing every piece of media they owned at the exact same time instead of waiting for it to air. In addition, individual authors, had the same access to distribution. The only limitation at this point was exposure and the physical limits to how much a single person could create in one day.

Bad filler media is as old as media itself. People have been creating things for the sole purpose of filling time as a captive audience has always been lucrative.

With AI, the limitations of individual human creation have been removed and you're now getting Sam Altman tweeting things like I can have AI make a podcast about the things my kids like and turn that on in the car on the way to school.

Media explosions always have the effect of creating a completely new dynamic of consumption. Radio let people hear music at any time instead of having to learn to play instruments. Recorded media let people specifically select the music they wanted to hear. Many musicians could now learn a song by ear through playing a record over and over again instead of having to read sheet music or have someone teach them directly.

tapes and CDs made this process so much cheaper and smaller that the average person could listen to almost any band they were interested in while walking, driving, or any other situation. Streaming allowed people to listen to any song at any time without having to ever go to a store or own a single album. This meant that music could go viral in ways that it couldn't before because you could listen to an entire song for free the instant someone told you about it.

AI creates a sea change as well in that the average person can now generate thousands of songs a day. To put this in perspective, the average person can now generate and distribute online more songs in a single day than that same person has the time to listen to in that same day. A single person can generate more than 24 hours of AI music in one day and also get a full night sleep.

It's extremely difficult to predict how this will play out, but it's not unreasonable to expect a new type of media consumer who creates within an AI ecosystem, custom content for themselves without ever leaving it or being exposed to media created by other people. It's not surprising to expect the people will create content in the same way that farmers grow crops, and use AI tools to either distribute that media to pick winners based on metrics or use other criteria to select winners. The examples of this type of media creation and consumption are endless. What they have in common is that the act of creating media is now just as easy as consuming it.

This will change consumption as well. There will be and already are many people who consume AI content without knowing it.

Hank Green, a creator, famous for being a human centered, intellectual recently went through a scandal centered around his addiction to AI and using generative content in his YouTube videos. AI can create content, whole scale, writing, videos, audio, and more. But we often overlook creators or creative teams, who use AI to enhance what they do. AI generated effects, having AI turn a writers room conversation into a script, having AI do research which you can then narrate, and more. Because AI is so ubiquitous it is going to be impossible to tell the difference 100% between someone who did everything themselves, and someone who used AI in a specific part of the process.

Because of this, the attitude towards consuming AI media will change. You will as always have a broad-spectrum of response to it, people who hate hate it, people who are indifferent and people who seek it out. They will have a broad range of opinions in that spectrum on AI media consumption. It's not unreasonable that in the same universe and AI Luddite, who refuses to consume anything that hasn't been verified as AI free will exist within the same world as somebody who is entire existence is consumed by self generated AI content universes.

People are already dating AI chat bots. They see them as sentient and in some cases better partners than the real people they are married to. I write real code by hand and solve discrete math problems for fun in my free time. It is no exaggeration when I say that AI is complex auto complete. There is a little more to it than that and to explain it to you I would have to explain algorithms like soft max, multidimensional vector weighting, and more. There are already a plethora of amazing articles on how AI actually works. It's important to know that it boils down to, on a philosophical level, auto complete in the same way that your phone does. It has no ability to verify truth, have emotion, or remember anything that wasn't already passed into its text box during its current run.

The Turing test was invented as a philosophical exercise. It's not a real test that verifies anything in a scientific sense. What it does is it shows that if you make auto complete or a chat, bot, convincing enough, scientists can verify on the back end that it is in no way human or even close to human and yet it will still fool humans into thinking that it is. This is also not new. There is a long history of robots, tricking people into thinking they're human and humans tricking people into thinking they're robots. There isn't a deeper philosophical answer to this other than that it's easy to trick people and there have been industries based around tricking people as old as time itself. A casino is a great example. Tabloids are another. The list goes on and on. It's more realistic that people are falling in love with auto complete than it is that over the course of five years humanity went from not having anything close to artificial intelligence, to then inventing it and spreading it all over the world.

Because of this another one of the sea changes that AI is going to introduce in the media ecosystem. We will have both a flood of media that is so large that the average person cannot know where to begin, and we will have humans confusing the fake for the real at a pace, never before seen in human history.

Most likely you already know this. These are not new conversations around AI or the media ecosystem. And it's not why I wrote this article. This article I'm writing because the media ecosystem is going to change. It's going to need to change and it's also going to get forced to change. Things that are going to be important are going to less be the collection of media you own and more around other things like a stable brand image that can consistently output something that people expect, hijacking attention in new never before expected ways, being able to distribute and sift through ten to 100 times the amount of content per second that the largest providers like TikTok and Facebook sort through each day, being able to target, individuals, more accurately, arbitrary ways of looking through audiences and media that are not perfect by any means, but get the job done in relation to the gigantic amount of media there is.

I have a hunch that trust and attention hijacking are going to be two very large areas in the new media ecosystem. And sitting behind them will be the infrastructure concerns around storing and distributing more information a day than we've historically generated as a species.

In tech, the assumption almost always follows more law, that humanity will find a way to double capacity every year. This is not a guarantee. It's just a trajectory. Experts remind us of that fact every year and every year against all odds more law is doubled. It is hard to say one, but there is always a physical limit to doing something. Science just lets us extend that horizon in ingenious ways. It's not unrealistic to assume that AI content generation is going to not just tax but completely erode, destroy flood or jam up our current global data infrastructure.

In that timeframe, it's reasonable to assume that companies will work on both doubling capabilities of handling that much media as well as going in the opposite direction which is reducing consumption and restricting transmission. You might see Internet service providers, reducing the amount of upload speed that they provide people or charging more for it. You might see consumers being more selective about who they're consuming media from. You might start to see distribution of media become less free than it is currently. That said there is a significant amount of profit in controlling in distributing media so at the same time, which you can see already companies like Facebook are attempting to create entire AI first media ecosystems centered around user created an automatically generated AI content, which is completely free to the users. This is likely to be a push pole problem, both in terms of scientific limitation as well as user/audience reaction to these issues.

I think it's going to be interesting to watch how media companies evolve over the next 10 years and I do think it's going to have a trajectory that is heavily based around handling a enormous amount of content throughput that can never be verified as completely AI free.