The Great Information Repricing

A paper on the future of content says that many publishers are sitting on an intelligence goldmine - but they just don't know it.

by Rob Corbidge

Published: 15:20, 23 April 2026

Proper food for thought is actually quite rare in the media industry. The current themes that give us hope or fear, or sometimes both at the same time, are fairly well-trodden ground. As someone who writes on such themes, I'm aware that our collective media industry posture can be somewhat suspicious and defensive. Outside of some notable legal actions, we don't take the fight to Big Tech or Big AI much at all, rather we are reactive to them.

So some genuinely different thinking is welcome, even if you don't agree with it all. In that vein, Francesco Marconi has just published a paper about the future of the information market, which is both well-argued and bold. To explain his thesis, it's worth quoting him right out of the gate:

"The companies that will monetize truth are the ones that stop selling it as something to read and start selling it as something to use."

It's in the "use" that his ideas sit. If information is usable, then that means it's worth paying for, and as AI systems make producing quantity terribly cheap, this means, as Marconi puts it, that "The media industry is splitting into three different species. The Intelligence Business, the Attention Aggregator, and the Public Good. Only the first has pricing power in an era of abundant content. The second faces structural collapse not because awareness lacks value but because AI is making it free. The third will not survive as a business and it needs a different funding model entirely."

The whole thing is well worth taking the time to read.

Of course, Marconi (formerly of the AP and WSJ), starts out with the effect AI is having on the publishing industry, meaning an abundance of low quality, derivative content. He brings to this fact another line of thinking, which is that almost every publisher has the high quality information people are willing to pay for in-house, but don't realise it. That is, they have both a content operation and an intelligence operation and, as he puts it, the trick is whether a publisher grasps that they have "intelligence assets trapped inside a content wrapper". 

It's important to note Marconi's background. He's been head of R&D for the WSJ and worked on AI and News Automation for the AP. There can be little doubt as to the quality of his insight. Importantly I feel, he's witnessed the value that financial data terminals, such as Bloomberg, bring to their user base, and why that user base is so willing to part with a lot of money to have access to them. It's a useful place to start from when you're working out what value is.

He uses the examples of the FT and WSJ. Both make far more money from their intelligence businesses than they do from their main products. Indeed, the difference in profit margins is quite staggering. But, as Marconi says, the journalism didn't change, the customer did.

He argues that for the sites he calls "Attention Aggregators", the game is already up. Featured snippets and social media were already chewing up the territory of such outfits before the arrival of AI. AI has just accelerated the process. "The Attention Aggregator's product was the click. When comprehension replaces curiosity, the click has no reason to exist."

It's even worse for his Public Good category, which includes sites including local media. There is, according to his thesis, almost no commercial slot for them to occupy in this age, but there is clearly a societal need for them. Other sources of funding must be sought.

Marconi clearly values the human reporter. He frequently refers to them, and recognises the unique information-gathering abilities a flesh-and-blood journalist has. Indeed, he mourns what happens when an experienced specialist reporter is made redundant, noting that it's not just the stories they wrote that are gone, but everything they understood about the area in which they reported. That, he says, is where the intelligence layer sits, in that understanding. This is marketable information.

"What's happening isn't that newsrooms are becoming factories. It's that the factory part of the newsroom, the writing, formatting, translating, summarizing, is being automated away, exposing what was always underneath: a very small number of people doing the work that actually matters."

It's a fascinating read, and a perspective altering one. A good example of that would be his take on Google, and AI Overviews. Marconi points out that Google is damaging its own advertising inventory by building its "no click" content world.  

"Every query answered without a click is a query that generates less advertising revenue. The company is simultaneously the infrastructure provider that media companies depend on and the entity most actively destroying their economics. It's also destroying its own, just more slowly."

Google killing itself? I'll buy that intelligence. Journalism will survive them, but we have to understand what we have and value it accordingly.