When designing and deploying emerging technologies like artificial intelligence and autonomous systems, engineers and policymakers face a fundamental paradox: it is nearly impossible to accurately predict the long-term social, ethical, and institutional impacts of a system before it is built, yet waiting until deployment to observe these impacts is often dangerous and irreversible.
Traditional empirical evaluation methods fail in this regime. If you ask stakeholders abstract questions about algorithmic ethics or data privacy, they often default to high-level platitudes. If you attempt to build functional prototypes, you run into steep technical barriers, high costs, and an inability to simulate long-term societal feedback loops.
To resolve this, a growing body of research utilizes speculative narrative prototyping. By grounding abstract technical capabilities in fictional but realistic scenarios, future-dated media, and hypothetical end-user products, researchers create low-stakes simulation environments. These narratives act as cognitive triggers and boundary objects, allowing both experts and laypeople to anticipate harms, uncover latent human needs, and stress-test ethical frameworks before a single line of production code is written.
Translating Abstract Capabilities into Concrete Artifacts
The primary reason speculative narratives work is that they drastically lower the cognitive load required to understand complex, unbuilt systems. Non-technical stakeholders cannot easily evaluate source code, system architectures, or abstract policy descriptions. Speculative prototyping translates these abstractions into concrete, everyday experiences.
This insight is not entirely new. Even in the early days of computing, pioneers recognized the need to bridge the gap between abstract research and societal impact. Firschein et al. demonstrated that translating raw technical capabilities into postulated end-user products provided experts with a standardized unit of analysis to systematically assess commercial viability and ethical desirability.
In modern participatory design, this translation allows non-experts to engage with highly technical concepts. For example, when evaluating complex data practices like cross-dataset linkage and de-identification, Groenewald et al. used concrete, scenario-based elicitation to help older adults provide nuanced ethical feedback. Rather than forcing participants to understand the underlying data engineering, the narrative contextualized the consequences of the technology.
[Abstract Technical Capability]
│
▼ (Speculative Translation)
[Postulated Artifact / Fictional Scenario]
│
▼ (Cognitive Trigger)
[Nuanced Stakeholder Ethical Evaluation]
By using fiction as a cognitive trigger, designers bypass the need for traditional prototyping skills. As De Vos et al. argue, science fiction stimuli empower individuals without formal technical training to articulate complex cause-and-effect relationships and moral dilemmas. Presenting real-life application scenarios helps participants immediately contextualize a technology's use and anticipate its downstream impacts (Weinberger et al.).
Creating Neutral, Low-Stakes Simulation Environments
Beyond lowering cognitive barriers, speculative narratives serve a vital psychological function: they remove immediate political, economic, and institutional biases.
When researchers interview practitioners or organizational leaders about the failures of current systems, they frequently encounter defensive posturing or corporate public relations alignment. Coston et al. found that by shifting the focus to the future through speculative scenario probing, researchers could bypass these immediate defensive reactions. Discussing hypothetical projects allowed institutional stakeholders to openly surface systemic and structural barriers they might otherwise conceal.
On a philosophical level, narrative fiction provides a neutral environment to stress-test ethical frameworks against extreme edge cases that do not yet exist in empirical datasets. Burton et al. and Goldsmith et al. demonstrate that using science fiction scenarios allows AI practitioners to evaluate competing ethical frameworks—such as deontology versus utilitarianism—without the clouding effect of immediate personal or commercial interests.
This methodology treats speculative fiction as a "social laboratory." As Balazs notes in an analysis of historical utopian literature (Dictionnaire critique de l’utopie aux temps des Lumières), speculative narratives have long functioned as rigorous testing grounds for emerging social sciences, allowing authors to model complex variables like public health, demography, and governance within a controlled, theoretical framework.
Mapping Technology to Everyday Routines and Affective Needs
When engineering teams evaluate a new technology, they typically focus on functional constraints and optimization metrics. However, the friction between automated systems and human society usually occurs at the level of daily habits and emotional expectations. Speculative narratives force designers to look at these messy, non-linear human elements.
One prominent framework for this is the concept of "social practice imaginaries." Developed by Viaene et al., this design-led forecasting method shifts the focus away from hardware specifications and onto the "doings and sayings" of daily life. By constructing speculative scenarios around domestic routines, the authors demonstrated how automated learning systems frequently break down during "crises of routine"—revealing value tensions that standard technical evaluations miss.
| Evaluation Method | Primary Focus | Key Outcome |
|---|---|---|
| Traditional Technical Prototyping | Functional constraints, optimization, hardware/software specs | Verifies technical feasibility under ideal conditions |
| Speculative Narrative Prototyping | "Doings and sayings" of daily life, emotional expectations, institutional barriers | Reveals value tensions, systemic friction, and "crises of routine" |
Similarly, narratives can expose latent psychological needs before a technology is physically realized. In the domain of autonomous vehicles, Lee et al. utilized guided scenario imagination—structured prompts that strip away current operational constraints—to uncover the emotional and affective expectations passengers have for future vehicle interiors.
When these narratives are brought directly to practitioners, they act as powerful diagnostic tools. Muller et al. illustrated this by constructing a design fiction tracking a software engineer's relationship with a generative AI assistant. By evolving the narrative from a helpful coding partner into an invasive tool for organizational surveillance, the authors demonstrated how speculative stories can be deployed as "technology probes" to capture real-world developer reactions to long-term sociotechnical threats.
Operationalizing Participatory Speculative Design
To move these narrative methods from artistic exercises to reproducible scientific frameworks, researchers have developed structured, multi-stage methodologies.
For example, Mun et al. introduced the PARTICIP-AI framework, a four-stage speculative design surveying method that democratizes technology governance. Instead of asking the public to react to pre-determined industry roadmaps, the framework guides laypeople through a structured pipeline:
┌────────────────────────┐ ┌────────────────────────┐
│ 1. Brainstorm Future │ ──> │ 2. Extrapolate Concrete│
│ Use Cases │ │ Scenarios │
└────────────────────────┘ └────────────────────────┘
│
▼
┌────────────────────────┐ ┌────────────────────────┐
│ 4. Evaluate Opportunity│ ──< │ 3. Analyze Harms and │
│ Costs of Banning │ │ Benefits │
└────────────────────────┘ └────────────────────────┘
This structured progression ensures that public input moves past superficial enthusiasm to grapple with the real trade-offs of development versus non-development.
To ground these speculations in local realities, researchers often anchor narratives within specific cultural and geographic contexts. Hohendanner et al. deployed a Participatory Speculative Design framework using future-dated media reports to explore the metaverse in Japan. By anchoring the fiction in local challenges—such as disaster preparedness and an aging population—the authors showed how speculative design can move past industry hype to surface localized ethical risks.
Finally, these narratives are frequently paired with physical or structured prompts to facilitate cross-cultural collaboration. Väljataga et al. combined fictional learning scenarios with physical "value cards" across seven European countries. The realistic narratives served as "boundary objects"—shared conceptual spaces that allowed stakeholders with vastly different backgrounds to align on how specific technical features might disrupt privacy, accessibility, and social dynamics.
By structuring speculative user narratives, designers gain a proactive mechanism to uncover latent ethical and privacy concerns long before a system is deployed (Brandão et al.), ensuring that the technologies of tomorrow are built with a rigorous, anticipatory understanding of human values.
Key Foundational Work
Forecasting and Assessing the Impact of Artificial Intelligence on Society (1973) — contributed 1 concept elements to this direction.
Scenarios: Models, Transformations and Tools (Dagstuhl Seminar 03371) (2003) — contributed 1 concept elements to this direction.