Sociotechnical Envelopment of Artificial Intelligence: An Approach to Organizational Deployment of Inscrutable Artificial Intelligence Systems
Социотехническое окружение искусственного интеллекта: подход к организационному внедрению непрозрачных систем искусственного интеллекта
2021-01-01
SCID: 54.1/k86vsumf
Discuss with AI
explainability-performance trade-offinscrutable artificial intelligenceorganizational AI deploymentresponsible AIsociotechnical envelopment
Figures from the paper
Abstract (AI)
The paper presents an approach for implementing inscrutable (i.e., nonexplainable) artificial intelligence (AI) such as neural networks in an accountable and safe manner in organizational settings. Drawing on an exploratory case study and the recently proposed concept of envelopment, it describes a case of an organization successfully “enveloping” its AI solutions to balance the performance benefits of flexible AI models with the risks that inscrutable models can entail. The authors present several envelopment methods—establishing clear boundaries within which the AI is to interact with its surroundings, choosing and curating the training data well, and appropriately managing input and output sources—alongside their influence on the choice of AI models within the organization. This work makes two key contributions: It introduces the concept of sociotechnical envelopment by demonstrating the ways in which an organization’s successful AI envelopment depends on the interaction of social and technical factors, thus extending the literature’s focus beyond mere technical issues. Secondly, the empirical examples illustrate how operationalizing a sociotechnical envelopment enables an organization to manage the trade-off between low explainability and high performance presented by inscrutable models. These contributions pave the way for more responsible, accountable AI implementations in organizations, whereby humans can gain better control of even inscrutable machine-learning models.
Key Findings
1
Envelopment methods include defining interaction boundaries, carefully curating training data, and managing input and output sources.
2
Operationalized envelopment enables organizations to retain greater human control over inscrutable machine-learning systems and supports more responsible AI implementation.
3
Sociotechnical envelopment combines social and technical factors to manage risks associated with inscrutable AI beyond purely technical controls.
4
The case study shows that sociotechnical envelopment helps balance the high performance of flexible AI models against their low explainability.
5
The paper presents an approach for accountable and safe organizational deployment of inscrutable, nonexplainable AI systems.
Research Object
Inscrutable artificial intelligence systems deployed in organizational settings
Research Subject
Sociotechnical envelopment for accountable and safe deployment, including management of model boundaries, training data, inputs, outputs, and the trade-off between explainability and performance
Publication Details
Publication Date
2021-01-01
Journal
Publisher
ISSN
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest