Nieman Lab.
Predictions for
Journalism, 2025.
The typical inverted-pyramid news article is like airplane food — edible but to no one’s taste. Since its inception in the 1800s, the story format is a functional, practical way to organize information in the absence of any knowledge about what the members of their audience want and need.
But in 2025, we can know our audience and tell stories for each member. Now, we can treat articles as collections of discrete pieces of information. The facts, quotes, and background references are stored individually and assembled (and reassembled) in many “articles” to serve the individual audience members.
Creating understanding, not simply informing, should be the goal of every journalist. In 2025, journalism must build a new generation of knowledge management tools. These tools treat individual information as a customer relationship management system treats sales contacts. They will store new sources, facts, and connections between them. They will also record the provenance of each piece of information.
Reporters and editors should offer their perspectives on essential facts and concepts. Fluid stories can be assembled — instead of the old, imperfect, immutable story we are used to.
These “living” stories can adapt and recreate themselves. When a reader selects a story, the length, depth, format or complexity can be changed as they read it. New information is instantly incorporated. A user’s scroll speed while reading will trigger the system to change the complexity of the language — making it more straightforward if needed. Prior reading history might suggest a reader needs more (or less) background information. Tracking may show a reader is more or less likely to click on audio or video information, suggesting new formats for the story.
It’s not enough to tell readers a fact or to share a quote. The goal of journalism must be to ensure that each reader truly comprehends and can act upon what they learn. Generative AI will help us assemble and reassemble these stories to meet the needs of our audience members. Technology will finally help journalists give their audience members a better version of the news article, one which favors understanding, not just informing. This new approach will help journalism transition from a one-to-many model to a one-to-one-for-many model.
The monolithic, canonical article had value, but it was primarily economic: Spend as little as possible to create the article. Show it to as many people as possible. Hopefully, those people will view advertising, purchase subscriptions, or pay for merch. Audience satisfaction had little economic value, so media organizations needed more incentive to improve storytelling.
Jeremy Gilbert is a professor in digital media strategy at Northwestern University.
The typical inverted-pyramid news article is like airplane food — edible but to no one’s taste. Since its inception in the 1800s, the story format is a functional, practical way to organize information in the absence of any knowledge about what the members of their audience want and need.
But in 2025, we can know our audience and tell stories for each member. Now, we can treat articles as collections of discrete pieces of information. The facts, quotes, and background references are stored individually and assembled (and reassembled) in many “articles” to serve the individual audience members.
Creating understanding, not simply informing, should be the goal of every journalist. In 2025, journalism must build a new generation of knowledge management tools. These tools treat individual information as a customer relationship management system treats sales contacts. They will store new sources, facts, and connections between them. They will also record the provenance of each piece of information.
Reporters and editors should offer their perspectives on essential facts and concepts. Fluid stories can be assembled — instead of the old, imperfect, immutable story we are used to.
These “living” stories can adapt and recreate themselves. When a reader selects a story, the length, depth, format or complexity can be changed as they read it. New information is instantly incorporated. A user’s scroll speed while reading will trigger the system to change the complexity of the language — making it more straightforward if needed. Prior reading history might suggest a reader needs more (or less) background information. Tracking may show a reader is more or less likely to click on audio or video information, suggesting new formats for the story.
It’s not enough to tell readers a fact or to share a quote. The goal of journalism must be to ensure that each reader truly comprehends and can act upon what they learn. Generative AI will help us assemble and reassemble these stories to meet the needs of our audience members. Technology will finally help journalists give their audience members a better version of the news article, one which favors understanding, not just informing. This new approach will help journalism transition from a one-to-many model to a one-to-one-for-many model.
The monolithic, canonical article had value, but it was primarily economic: Spend as little as possible to create the article. Show it to as many people as possible. Hopefully, those people will view advertising, purchase subscriptions, or pay for merch. Audience satisfaction had little economic value, so media organizations needed more incentive to improve storytelling.
Jeremy Gilbert is a professor in digital media strategy at Northwestern University.
