Nexus

We often speak of the “information age” as though more information automatically led to more knowledge and better decisions. Yuval Noah Harari challenges this equation. Information networks need not primarily be truthful to be effective. They can coordinate people by stabilising shared stories, categories, money, laws or bureaucratic procedures.

The guiding question of Nexus is therefore: how do information networks connect truth, order and power—and what changes when AI itself generates and selects information and helps shape decisions?

1. Information connects—even without truth

A message can describe facts. But it can also create relationships: a promise, membership, a rule or a shared narrative. Large societies function because many people trust such symbolic orders and coordinate their actions accordingly.

This explains why sheer information volume offers no protection against manipulation. False or heavily simplified narratives can be particularly effective at creating bonds. Information literacy therefore begins by asking what function a message fulfils: does it inform, coordinate, legitimise power or capture attention?

2. A network’s quality shows in its capacity to correct itself

Harari highlights self-correcting mechanisms: independent courts, science, free media, opposition and internal controls. They do not guarantee truth. But they create procedures through which errors can be challenged, examined and revised.

This also applies to organisations. When bad news does not travel upwards, metrics replace criticism or dissenting voices fear repercussions, the information network becomes blind. Psychological safety, transparent decisions and clear escalation routes are therefore both practical Future Skills and organisational responsibilities.

3. AI is more than another communication channel

Earlier media predominantly conveyed human-generated content. AI systems can generate content, conduct conversations, model preferences and personalise suggestions for action at enormous scale. Harari sees a qualitative shift: non-human actors become active nodes in the information network.

This does not produce a certain prediction of the future. But concrete design questions arise: who sets the goals? How are sources and uncertainties made visible? Who can challenge an automated decision? Where does a responsible human remain accessible?

Your 3-minute exercise: examine the information network

Choose an important message, metric or AI output to which you are currently expected to respond:

1. Minute 1 – Origin. Who or which system produced it? What source, selection or translation lies in between?
2. Minute 2 – Function and power. What action is the information likely intended to trigger, and who gains influence as a result?
3. Minute 3 – Self-correction. What alternative source, follow-up question or route of appeal could expose an error?

What we take from this

Nexus connects information and AI literacy with democratic responsibility. In WEITER7 , this can become a learning experiment: for one week, document the source, uncertainty and correction opportunity for a recurring decision—then examine whether the decision became more transparent.

What feedback would my information network need to allow in order to discover an error of its own?

Context: This insight summarises selected arguments from Harari’s broad historical interpretation in our own words. The book connects numerous eras and examples; individual analogies and predictions are open to debate and do not replace specialist research. The exercise and Future Skills connection are the Institut’s own applications.

Sources and further reading: Yuval Noah Harari’s official book page ↗ · Publisher’s announcement for Nexus ↗