Futures friction: a new approach to how media organisations remember, learn and reimagine

Stop for a moment. Step away from your quarterly goals. And ask yourself: Will the formats, products or services that I am investing my time and resources in still be relevant in a few years?

The assumptions behind our choices often seem so obvious that we rarely question whether they still hold. If media organisations only make room for ideas that fit today’s strategy, they risk rejecting the very ideas that could help them remain relevant as the surrounding system changes. How can we make room for that kind of reimagining if the way we develop, assess and share ideas remains shaped by the very assumptions we need to question?

Working with media organisations, it’s becoming clear how difficult it is to connect innovation across teams and build on what others have learned. Ideas struggle to survive daily operations or find support beyond the team that developed them. Those that challenge the current business are often measured by metrics that overlook their potential future value. So, ideas get stuck or are squeezed into a different logic than originally intended, while others may never learn what the team tried to achieve or which assumptions shaped its decisions. An unsuccessful experiment can still produce valuable learning if we preserve the context needed to understand what happened. A good idea might not fit the existing project but be gold in another context.

Could connected second brains make learning more accessible across teams, and might futures friction help question our assumptions in light of possible futures?

The weight of history in news organisations

News organisations carry a particular legacy that needs to be considered in how they preserve and share organisational knowledge. Protecting sources, keeping a scoop from competitors and guarding rights are important parts of the work and culture. Those instincts also shape how knowledge moves within the organisation, including outside the newsroom. Competition between desks can make it harder to share what they have learned, and the reasoning behind a project may stay with the people who made it.

At the same time, there is a kind of exhaustion from the ongoing waves of change in media, from print to web, mobile and platforms. Any practice that asks media organisations to share and question what they know must be built with this history in mind: within the boundaries that source protection and rights require, and with as little extra effort as possible.

Where do you start when people are exhausted by change and used to protecting knowledge? By making it easier to learn from what others have tried and why. Technology could bring that learning into existing workflows, making the organisation’s collective knowledge a starting point for futures-driven innovation.

Second brains: shared organisational knowledge layers

One concept I’m interested in exploring in this relation is the idea of a second brain: a way of organising knowledge outside our heads so we can return to it, connect it and use it in new situations. The term was popularised by Tiago Forte as an approach to personal knowledge management, but the ambition to extend our memory through technology has an even longer history. In his 1945 essay ‘As We May Think’, Vannevar Bush imagined the memex, a desk-like device for storing books, records and communications. People would connect these materials through trails of association that they could revisit and share with others. AI could make it easier to build, maintain and draw on such connections, opening new possibilities for how organisations use their accumulated knowledge.

AI could make this knowledge easier to retrieve and connect across an organisation, helping people examine the evidence and reasoning behind earlier decisions.

It all starts with collecting contextual data: the purpose of an experiment, the audience insights it drew on, what was tried, why a decision was made and where people disagreed. A report saying “user demand was low” tells us little about why but knowing the context can give another team something to work with when that context is accessible and reusable.

Second brains operate at several connected levels: individual second brains that hold employees’ own knowledge, team second brains that capture shared experiments and learning, and an organisational second brain that preserves strategic reasoning across projects and helps connect innovation to the organisation’s purpose and values.

AI researcher Andrej Karpathy describes a related approach in his LLM Wiki: a collection of linked Markdown files that an AI model writes, organises and updates from source material. People select the sources and guide the work, while the original material remains separate from the AI-generated knowledge.

Making context machine-readable allows people and AI agents to work with it across projects. Sources, dates and the conditions behind a conclusion need to travel with it, as does the distinction between evidence and interpretation. Otherwise, a confident interpretation could become accepted as evidence simply through repeated retrieval. Formats and protocols are emerging to support different parts of this.

Futures friction: using futures to inform the present

Futures friction brings futures thinking and foresight into the innovation process. It introduces deliberate friction to test ideas against possible futures, considering changes that might happen in both the short and long term.

A convincing AI-generated answer can make an idea feel settled before we have explored other possibilities. This is the risk of premature convergence. Adding intentional futures friction helps us challenge the assumptions and biases behind that sense of certainty, making room to stop, think and iterate our ideas.

Futures friction builds on a futurist’s way of thinking. Thinking in futures means recognising that the conditions we work within could develop in several plausible directions and exploring what those differences would mean for the decisions we make today. It requires us to look beyond current short-term strategies and pay attention to signals that could challenge what we take for granted. Some changes may be incremental; others could fundamentally alter how people access information or whom they trust, which are existential questions for media to pay attention to. At the same time, many needs and behaviours may endure.

This also means staying willing to change our minds as we learn. Short cycles of testing and reflection help us explore these questions in ongoing innovation work. Each experiment should add to the context others can draw on.

Shared context and futures friction need to work together. A second brain helps turn shared context into infrastructure for remembering and learning. Futures friction brings questioning into that infrastructure, helping organisations explore possible futures and recognise what they may need to reimagine.

A reimagined media organisation in 2030

What might this look like in practice? Imagine a media organisation in 2030, where teams draw on connected second brains and use futures friction to question their ideas.

A journalist returns from a walk with a colleague. They have been discussing an idea for a new local information service that could help residents understand how political decisions connect over time and what they mean for their lives. It would bring together the organisation’s reporting, public records and knowledge of the community. Developing it would require editorial, product and technology teams to work together.

The colleagues have chosen to record their conversation and process it on a local device, where it stays private. Afterwards, an AI agent suggests a short outline for the journalist’s individual second brain, capturing the idea, their reasoning and the questions they have yet to resolve.

She reviews the outline and quickly adjusts which parts to share with her colleagues in the team. Her private notes and the original recording stay in her individual second brain. The approved outline becomes part of a team second brain, where colleagues can develop the idea together. She controls what moves out of her personal space; contributing an idea does not give the organisation access to everything behind it.

The team likes the idea and draws on the organisational second brain. Within minutes, an agent finds an experiment from four years earlier. Its summary says demand was low, but the original evaluation reveals something more specific: people found the proposed service useful but disliked having to install yet another proprietary app.

This is the value of shared context. The team can see what was tried, why it was judged unsuccessful and whether that conclusion is relevant to the new idea. The organisational second brain also preserves the strategic reasoning behind the earlier project, including its purpose of helping residents understand local decisions.

Now futures friction enters the process. Would the idea make sense if residents increasingly expected information to reach them through their own AI assistants? An agent helps the team explore that possibility alongside an alternative: what if people resisted AI mediation and wanted direct contact with local journalists? Observations that residents have agreed to share give the team reasons to take both possibilities seriously.

The team decides to stress-test several ways of accessing the service. Its purpose remains the same, but the assumptions about how to fulfil it remain open to reimagining. The decision, its reasoning and the unanswered questions enter the team second brain. Learning that could help other projects is also shared through the organisational second brain.

Residents can follow how the service is being developed through the reasoning and evidence the team has chosen to share. They can question its choices and contribute their own experience. Their feedback adds to the shared context behind the project.

A few weeks later, a competitor launches a similar service. The team uses futures friction to revisit its assumptions: if this kind of information becomes widely available, what would make its own service useful? It adjusts the experiment and records what changed and why in the team second brain. That learning also reaches the organisational second brain, where other teams can draw on it.

In this scenario, AI helps collect, connect and retrieve knowledge, leaving more room for thinking, meeting colleagues and spending time with residents. Leadership has protected that time and recognised contributions to collective learning. The same efficiency could have gone to cuts; here, the organisation has chosen to invest it in relationships, on-the-ground journalism and futures-driven innovation.

Connected second brains help the organisation remember and learn. Futures friction keeps that learning open to challenge as conditions change. Together, they give teams a stronger basis for deciding which ideas to fund, test and reimagine.

Many conditions would need to hold for this future to take shape, and it is unlikely to unfold exactly as imagined here. The value of scenarios is to help us explore what could change and what we might need to prepare for, including the choices about ownership, time and resources that make this way of working possible.

Where to start?

Back to 2026. I met a software developer who described working with his second brain as adding ingredients to a huge pot, stirring them around and then delivering. His accumulated knowledge provides the context for each new task, changing how he works even as he is still trying to understand the implications. Media organisations could learn from that approach by working more systematically with context.

Start with one current innovation project. Bring together relevant context: earlier experiments, audience insights, signals of change and the reasoning behind the proposal. Make it machine-readable and preserve its sources. Agree who controls the knowledge, what can be shared and reused, which AI systems can access it and how you retain control if you change providers. Source protection, privacy, copyright and cybersecurity need to shape these choices from the start. What stays local, and what is shared through cloud services? What knowledge can be exported to other systems to avoid knowledge lock-in? Who owns what knowledge? Begin in a controlled environment with limited access and carefully selected material, then assess what can be connected more widely.

This takes practice, from organising shared context to questioning what AI retrieves and thinking in futures. Leadership needs to protect time for it, recognise contributions across teams and allow learning to change priorities. A second brain will be of little use if people are rewarded for holding knowledge back or the organisation refuses to reconsider its direction.

This also connects to a call for existential innovation: rethinking how media organisations sustain themselves and what they exist to do. For me, shared context and futures friction could help create the conditions for that work, making it easier to learn across teams, question inherited assumptions and give unfamiliar ideas room to develop in a time where existential questions need to be asked.

So, before returning to your quarterly goals, consider which ideas never make it onto the agenda because they challenge the current logic. Give one of them time, resources and the benefit of what others have already learned. Then use futures friction to explore where it could become relevant.

A futures-driven media industry begins with decisions like these, including the decision to take an idea seriously before it fits today’s strategic environment.

This article was developed with support from my own second brain, using it to question assumptions and explore the reasoning. The intention, arguments and editorial responsibility remain mine.

We are developing a free, downloadable Futures Friction Assistant as part of AI MediaLab Denmark to support futures-driven innovation in media, a collaboration between the H3 AI Lab at the Copenhagen Institute for Futures Studies and Center for AI DMJX, supported by the Google News Initiative.

The Futures Friction Assistant will become freely available worldwide in December 2026 for anyone to use with their own second brain.

Follow the project here (in Danish): https://ai-center.dk/projekter/ai-medialab