Basics
"Well, here’s another nice mess you’ve gotten me into!“
— Oliver Hardy to Stan Laurel
Basic categories of paradoxical interaction
Together, these five categories form an observation grid that can be applied to a wide variety of contexts — from economics and culture to AI interaction.
1. Structural paradoxes (incentive or rule paradoxes)
Here, the dysfunction arises from the architecture of the situation itself.
Each action is individually rational, but collectively destructive.
• Typical form: Prisoner's dilemma, tragedy of the commons, security dilemma.
• Mechanism: Incorrect coupling of individual success logic and collective results.
• Example: In web ecologies, all site operators ‘optimise’ for clicks or SEO, leading to overload, fake content and loss of trust.
2. Information paradoxes
The paradox arises from asymmetrical or overly complex information distribution.
Actors act rationally based on what they know – but the system generates misleading signals through feedback loops.
• Typical form: information cascades, signalling effects, feedback loops.
• Example: In social media, algorithmic feedback loops only amplify content that generates short-term engagement until the overall quality collapses.
3. Intentional paradoxes (motive or trust paradoxes)
Here, rational intentions encounter mutual mistrust or unclear frames of reference.
It is precisely the attempt to act rationally that destroys the possibility of rational cooperation.
• Typical form: loss of trust, strategic self-sabotage, pre-emptive strikes.
• Example: in team projects, everyone protects themselves against possible mistakes made by others – and thus creates inefficient control structures.
4. Systemic paradoxes (emergent or recursive paradoxes)
The paradox is not the behaviour of individuals, but the dynamics of the overall structure, which are often self-reinforcing.
• Typical form: path dependency, escalation spiral, Red Queen dynamics.
• Example: platforms must constantly stimulate more interaction in order to remain relevant – and in doing so, they destroy the trust that sustains them in the long term.
5. Perception paradoxes (interpretation or reflection traps)
These arise when the observation itself becomes the source of the paradox.
Example: Attempting to make oneself completely transparent or controllable changes the system and makes control impossible.
• Typical form: observer paradox, self-referentiality, double bind.
• Example: In AI development processes, too much optimisation for ‘measurability’ stifles qualitative learning.