The Inequality Of Attention
But is the time consumers spend with media still a useful proxy for its value? A lot less than it used to be. For most of the history of media, the correlation was close enough that equating them made sense. But the relationship between time and value has been weakening for years and AI will stretch it to the breaking point. Increasingly, what matters is not how long people spend with your content, but how much they care.[...]
both the utility consumers derive from a minute of media and the amount of that utility that providers can capture now vary much more widely.
A wrinkle in the inelasticity of attention. More time spent does not necessarily mean more value for the consumer or more revenue for the business.
Shapiro continues:
Confronted with finite time and infinite choice, consumers are increasingly compressing media to extract more utility out of less time.[...]
Sometimes content owners offer the compressed product themselves—NFL RedZone, the skip button during the opening sequence of a Netflix show, or a daily briefing from The New York Times. But consumers are also compressing media on their own: listening to podcasts at 2x speed, skipping or scrubbing, watching user-generated film and TV recaps (my son recently recommended that I watch Dune 2 in its entirety, but watch a five-minute recap of Dune 1 first), watching movies cut up into TikTok clips, and, as we'll discuss below, using AI to summarize or synthesize articles, podcasts, videos, etc. The growing prevalence of compressing media is a big change that upends the time ≈ value logic. The two-minute clip on Reels of the key moments of a podcast almost certainly has much higher utility per minute than the whole podcast and, for some listeners, maybe more utility overall than enduring a three-hour snoozefest.
Consumers are no longer just allocating their time among media experiences; they are modifying those experiences themselves.
Media modification to fit consumer time allocation.
The complication is what happens when the original starts being made for the extract.
The funny thing about facsimiles is that they often don't just copy, they compress. They reproduce by reduction, and output a reliable representation of the most resonant feature. Many of our compressions are lossy.[...]
The compression reaches backwards in time and changes what is created.
A facsimile can eventually change the thing it copies. The original is forced to accommodate the construct.
Books become easier to summarise.
Films become collections of clips.
Songs become engineered for the 6-second snippet you'll recognise them by.
Conversations slowly congeal into messages that can be screenshotted.
Remember that a light is not the sun.
Compression can give people more of what they want in less time. But pacing, context and build-up can be critical complements of the experience. They cannot always be reduced to obstacles to the good bit.
The inequality of attention means that the version that people watch most, the one people care about most and the one that earns the most may be three different experiences for three different modes of consumption and cognition.