AI, dead internet theory, and why analogue is a future-proof business strategy

The internet is abuzz over Google NotebookLM’s eerily realistic-sounding podcasts. Last week, MIDiA wrote about how it will disrupt primarily information-driven podcasting, rather than entertainment podcasts with human flair (for now). But the implications go further. 

The online world is fickle, and AI-generated content easily plays to its rules simply because AI was built using all of its data. Look no further than social media’s spam plague, with AI-generated content being suggested to users, and comments sections filled by bot interactions – reminiscent of the ‘dead internet’ theory. 

In music streaming this is viewed as scamming, with AI-generated songs boosted by bot listens known to syphon millions out of the streaming economy. However, on social media, generative AI is  largely considered fair game – and social is set to become a de-facto home of entertainment moving forward. 

The content conundrum

It is great on paper to be able to produce informational podcasts or background ad tracks or even short videos faster and cheaper with the help of AI. In the era of cultural commodification, it is easy for anyone from studio executives to marketing interns to assume that all content is good content in the saturated attention economy; you just need ears and / or eyes on it to get the point across and the dollars in.

But this is incredibly short-term thinking. Google’s NotebookLM is a great example of how AI can almost instantly build out professional-seeming digital content on just about anything. Yet, despite this technology’s clear and conversational output, it is also remarkable for being entirely unremarkable.

Once you get over the human-sounding chatter, it is striking how banal the outputs can be; they sound like every podcast you have ever listened to, and how you imagine all the ones you haven’t. This holds pretty much true for all AI-generated content. Part of this is intentional as a product design, of course – what would be the use in something that looked or sounded nothing like what it was supposed to? – but part of this is also an inescapable constraint. Because generative AI works using datasets and predictive text-type language generation, its output naturally falls into the dead average of everything ever that has already been produced in that category. And who remembers average? 

The point of it all   

Producing swathes of polished yet mediocre content that no one really wanted to take the time or energy to think through and make, for people who do not really care to spend much time thinking about it, just seems like a lot of money and effort for not very much at all. 

This is, perhaps, a broader flaw in the market for AI companies – most of which still do not properly have long-term business models because of the energy needs and copyright issues at play – but, for entertainment companies especially, there should be a huge push to differentiate from what is rapidly becoming a sea of indistinguishable same-ness in online spaces. 

Movies and music are not simply valuable because of their functionality. You cannot commodify them the way industrial production can mass-produce cutlery or clothing, because they serve no practical use. 

Music and film are valuable because, as art forms, they are nuanced, unique expressions of the human experience that audiences can connect with because they go far deeper than the surface of aesthetics (the memorable ones, anyway). 

This cannot happen in an algorithmic void, generated by AI to reflect the digital average with an eau de Uncanny Valley. Nor can genuine creativity really stand out when served in six second clips side-by-side with the AI slop. 

How much quality, innovative art will be buried by the inevitable AI content dumps?

But real-world touchpoints – be it a big music event, or simply needing to stand up and physically put a record on – stand out. And, helpfully, are more profitable per-head than ad impressions and per-stream rates. 

Digital detoxing

Despite users’ interest in going offline, their digital usage is unlikely to waver without bigger intervention. The emerging analogue shift is a value-driven one, rather than one driven by time spent. While audiences may still engage most often with online propositions, the offline ones are more important to them. Attention overwhelm and constant distractions do not leave much headspace for memorable moments; putting down the phone and doing something in person has more of an impact. 

AI adds greater uncertainty into the online mix, which will only push users to prioritise offline options more strongly. Rather than fighting this, entertainment companies should lean into it – reducing the digital demands that keep audiences trapped in an addictive cycle of online activity, and instead encouraging them to step away and engage more meaningfully offline. This creates longer-lasting, deeply rooted fan sentiment and cultural context – and, crucially, can bring in higher margins for the companies behind them. 

Leave the internet to the bots and the braindead. If you want to build something that lasts, maybe it is time to go outside.