Ideas
The future of public-interest information is being worked out in a lot of places at once. This page collects the thinking we find most useful, from voices across journalism and beyond it, and from our own writing.
How people will find and use information
Four shifts that are dismantling traditional journalism, and why they’re an opportunity, not an extinction.
What AI-mediated news could feel like for readers: more informed, less overwhelmed.
How AI assistants could replace the attention economy with a “curiosity graph,” and who would hold it.
What comes after the algorithmic social feed, and how AI will change the way people connect to information and to each other.
Personalized news is coming. The question is who holds the data about what readers care about, and who controls the tools.
What news is made of
A metadata framework so attribution, evidence and provenance survive when AI uses journalism.
Generative AI as the end point of forty years of media turning into math.
Value is migrating away from content, and creating surprising new opportunities.
Who gets paid
Why journalism’s best chance is building “archipelagos of trust.”
How newsrooms could build their own community spaces on open protocols, where readers keep their identity and contributions.
Why publishers need machine-readable rights and micropayments, not better SEO.
AI can generate truth. The question is who will get paid for it.
The Tow-Knight Center’s first big-picture report draws on a year of our convenings to show where AI is taking public-interest information. News will be intermediated by AI, whether we like it or not. The report maps the information ecosystem in three layers — consumption, synthesis and creation — and shows where newsrooms, technologists and funders can act while the new infrastructure is still being built.
Read the report
Selected posts from our Substack, (Re)Structured News, that dig into the report’s ideas layer by layer.
The big picture
What journalism is for, and the harder tradeoffs AI will force as newsrooms try to stay afloat.
Why people will turn to AI for news whether we like it or not, and why the realistic response is to shape those systems, not resist them.
The one theme running through our Three-Layer Cake report: we can build a better information commons, but not alone.
Consumption
What natural-sounding AI voice could mean for news: new demand, new challenges, and who controls the last mile to users.
How personal AI agents change news consumption, when anyone can ask the questions they’d never ask in public.
Synthesis
Language models have no sense of what’s true. Sannuta Raghu’s News Atoms build sourcing and provenance into news one sentence at a time.
How could an AI agent judge whether a source is credible? Some ideas for teaching it.
Creation
AI licensing deals won’t cover the cost of making news. The case for getting paid for what newsrooms do, not what they make.
Beyond the economics: how journalists see their own role will shape how they respond to AI.