Subprojects

This project comprises a series of individual sociotechnical case studies as well as contributions to a forensic toolkit. Connecting all these is the question of what sort of methodology is commensurate to the unfolding AI paradigm—here, the project took as its point of departure a synthesis of approaches from sociohistorical, design, and technical research. A useful structuring principle has been the pipeline underlying modern AI: the three distinct components of dataset, model, and application. The project takes specificity as a tenet to uphold, whether of models, architectures, procedures, epistemologies… conversely, between knowledge and tools there is a continuum to be explored.

Sociotechnical case study
TechnicalDesignSociohistorical
DatasetModelApplication

A Pedagogy of Machines: Technology in Education and Universities in Translation

The project intends to reverse the assumption that “AI is transforming education” to investigate how “education transforms AI”.

Sociotechnical case study
TechnicalDesignSociohistorical
DatasetModelApplication

AI Design Interventions for Social Diversity

How could new epistemologies and methodologies for AI systems design be developed based on critical insights from feminist and postcolonial science and technology studies?.

Sociotechnical case study
TechnicalDesignSociohistorical
DatasetModelApplication

AI Interpretability and Accountability in the Humanitarian Sector

Investigates pitfalls of the use of ML in the humanitarian sector.

Sociotechnical case study
TechnicalDesignSociohistorical
DatasetModelApplication

AI Interpretability and Linear Cause-Effect Models in Medicine: Is Non-Linear Diagnosis Possible?

Exploring the relationship between the traditional causal-linear explanation model and the knowledge generated by AI; investigating the conceivability of non-linear causality in biomedical diagnosis..

Forensics toolkit
TechnicalDesignSociohistorical
DatasetModelApplication

Exploratory machine learning interfaces

This project produces several interactive pedagogical interfaces on specific machine learning concepts..

Toolkit + Case study
TechnicalDesignSociohistorical
DatasetModelApplication

Exposing.ai – the production pipeline of facial recognition systems

Investigation, catalogue, and provenance of the datasets that enable facial recognition systems.

Sociotechnical case study
TechnicalDesignSociohistorical
DatasetModelApplication

Interpretability and Accountability of AI Systems in the Sciences

Investigates potential strategies and methods to expose the epistemic failures of visual AI systems in the natural and social sciences.

Forensics toolkit
TechnicalDesignSociohistorical
DatasetModelApplication

Latent mechanistic interpretability

This project seeks to provide a benchmark for evaluating the extent to which a model is mechanistically interpretable.

Sociotechnical case study
TechnicalDesignSociohistorical
DatasetModelApplication

Scopic regimes of neural wiring

Investigating the ideologies of vision of the neural wiring itself.