Show notes
We are producing more information than at any point in human history. Generative AI is accelerating that process again, dramatically reducing the cost of creating text, images, video, analysis, advertising and software. Yet one critical resource has not increased alongside it: human attention.
There are still only twenty-four hours in a day. Our cognitive capacity remains constrained. And as information becomes increasingly abundant, the economics of scarcity begins to shift towards the limited resource required to consume it.
In this episode of The Intelligent Economy, Dr Michael D’Rosario examines the economics of attention and asks what happens when an effectively unlimited supply of information competes for a fundamentally limited human resource.
The starting point is Herbert Simon’s influential observation that an abundance of information creates a scarcity of attention. More than fifty years later, that idea provides a powerful way of understanding the business models that came to dominate the digital economy. Search engines, social platforms, recommendation systems and digital advertising markets do not simply compete for users. They compete for minutes, engagement and opportunities to influence behaviour.
That distinction matters because the monetary price of many digital services is zero, but their economic price is not. Attention has become an input into a vast commercial system in which platforms, advertisers and algorithms compete to identify, allocate and monetise opportunities to reach us.
Generative AI changes this equation again.
As the marginal cost of producing another piece of content falls, the volume of material competing for our attention can increase dramatically. Firms can generate thousands of advertisements, product descriptions, emails, videos and personalised messages at very low cost. Information becomes cheaper, but the human capacity to process it does not.
Paradoxically, making information abundant may therefore make attention even more valuable.
But AI may also change the attention economy in a more fundamental way.
For much of the digital era, firms have attempted to infer what consumers want from searches, clicks, purchases and behavioural data. AI assistants allow us increasingly to state our intentions directly. Rather than searching through dozens of products, we may simply tell an agent what we want and ask it to find, compare or eventually purchase the best option on our behalf.
That shifts the economic relationship from inferred intent towards declared intent.
And it raises a new set of questions. Who owns that intention? Who gets access to it? Can companies pay to influence the recommendations made by our agents? And if an AI assistant is acting on our behalf, is it optimising for us, for an advertiser, for a commercial partner, or for the company operating the model?
The episode also considers what happens when machines themselves become an important audience for commerce. If AI agents search for products, compare insurance policies, evaluate hotels, negotiate prices and filter information before it reaches us, businesses may increasingly find themselves competing not simply for human attention, but for machine selection.
That could reduce search costs and strengthen competition. It could also create powerful new gatekeepers.
The organisations controlling the interfaces between consumers and an almost unlimited supply of information may occupy some of the most economically valuable positions in the AI economy. Control over those gateways raises questions about market power, information asymmetry, principal-agent problems and the creation of new forms of economic rent.
Yet there is a more optimistic possibility.
AI may increase what we might call our effective supply of attention. If an AI system can compare utility plans, search hundreds of flights, process administrative information or sort routine correspondence without requiring an hour of active human involvement, it effectively releases attention for other purposes.
Perhaps one of the most valuable applications of artificial intelligence will therefore be not producing more things for us to consume, but reducing how much information we need to consume in the first place.
The Economics of Attention ultimately asks a larger question about the intelligent economy.
If information is becoming almost infinitely abundant while human attention remains scarce, where does economic value move next?
And if attention is one of our most limited resources, should intelligent technology really be designed to capture more of it, or should genuinely intelligent technology give some of it back?
