The Next AI Report Will Not Be a Report
The next AI report—and with it, the future of white papers, policy memos, and strategy documents—will not be a report.
Because the institutional reality is this: no one is reading on the other side.
Not in time. Not in full. Not with the cognitive or organisational bandwidth required for action.
This isn’t just a failure of attention—it is a failure of form.
A misalignment between the static, declarative logic of the report and the complex, recursive, and relational nature of the challenges it tries to address.
The Medium Is the Architecture
In this new reality, the medium is the message—not as metaphor, but as infrastructure.
We are shifting from asymmetric communication—where reports are delivered as inert conclusions—to architectures of dialogue, iteration, and reflexive sense-making.
In this transition, the medium becomes:
Conversational: Built for interrogation, contradiction, and evolution.
Context-aware: Responsive to plural epistemologies and shifting ground truths.
Systemic: Operating within interlinked knowledge graphs and machine-readable taxonomies.
Performative: Not just delivering knowledge, but enabling the conditions for shared understanding and informed agency.
The report dissolves. What remains is epistemic infrastructure—conversational systems rooted in structured data and dynamic reference.
From Content to Cognition Infrastructure
This future is being constructed on libraries of structured data, surfaced through RAG (Retrieval-Augmented Generation) architectures, and expressed through conversational agents.
This is not a UX upgrade. It is a foundational shift in the design of knowing.
Reports are no longer written—they are assembled as conversational scaffolds.
Citations become live queries.
Reasoning becomes traceable and testable.
Meaning is no longer delivered—it is co-constructed in situ.
The production of high-integrity, domain-specific RAGs will become a primary function of institutional knowledge work—defining not only what is known, but how it can be navigated, contested, and updated.
Design Requirements
To operate in this paradigm, we must build:
Referenceable Data Lineages Structured, transparent provenance across temporal, geographic, and ideological dimensions.
Disclosed Sense-Making Logics Epistemic scaffolds made visible—models, thresholds, taxonomies, weightings.
Interrogability-by-Design Confidence, ambiguity, and blind spots must be surfaced—held in conversation, not hidden.
Plural Epistemic Frames Capability to operate across divergent ontologies and disciplines—without collapsing them into false coherence.
Dialogic Density as Asset The cumulative thickness of these dialogic spaces—their memory, richness, and feedback mechanisms—will become one of the most valuable assets of future institutions. Not just knowledge produced, but knowledge experienced and relationally processed.
Strategic Implication
The future of institutional reports lies not in more persuasive narratives, but in designing relational cognition environments.
The question is no longer “What is the insight?” but:
What is the system that allows meaning to be constructed, challenged, and evolved together?
The report is no longer a document—it is a field of engagement.
A living, recursive infrastructure that holds space for doubt, for difference, and for durable shared sense-making.
This is the end of the report.
And the beginning of conversational epistemics as public infrastructure.
Different context but…for quite some time at Kaospilot, I’ve been feeling the tension around asking students to write reports at the end of their semester. It often feels like zombified knowledge, performative reflection that no one reads, including themselves. A ritual of closure that actually closes nothing, aside from a very few.
I’ve been exploring ways to shift this, to compost the report into something more alive, dialogic, and metabolically relevant. But there's comfort in the familiar, even when it’s lifeless (and often almost useless) and the resistance has been quite strong. Even tho the evidence is everywhere.
This post gave me more clarity and inspiration than I expected. I feel even more excited to trust what I’ve been sensing all along: that the report isn’t broken, it’s the form itself that’s out of sync with the complexity we’re trying to learn with.
Thank you, Indy, for naming the shift with such precision and possibility!
This is great. Getting to the realities of what a public infrastructure could be, needs to be.