Missions

This post-doctoral position contributes to a broader research programme on the data and micro-task workers who inconspicuously contribute to the production of Artificial Intelligence (AI). Beyond high-profile engineers and data scientists, these lower-level workers perform tasks of data generation (for example, creating audio recordings or texts), annotation (for example, transcribing audio files or labelling images), algorithm evaluation (for example, verification of chatbot replies or search results), and others. Precarious and underpaid, they operate far from the core sites of AI production, often in low-income countries. To shed light on the identities, lived experiences, and actual contribution to AI of these workers, the study combines economic-sociological analyses with advances in large oral-corpus linguistics and its capacity to recast methods and tools from speech technology and machine learning. Specifically, we aim to harness the full richness of oral data, especially from interview recordings and transcriptions, going beyond the identification of key themes and considering a broader range of signals – from regional accents and dysfluencies to tone of voice and word choice. Computationally enhanced linguistic analysis provides the techniques to capture these features systematically and in detail.

Activities

The successful candidate will help the project team to design, pre-test, and field questionnaires and in-depth interviews with workers from Portuguese and Spanish-speaking countries. This new team member will be part of the sociology group while liaising closely with linguists and computer scientists to ensure quality of data and pre-processing. The role may also involve the testing of tools like automated transcription software as well as analysis of textual data- with NLP tools.
The post-doctoral researcher will have the opportunity to join a consortium of highly motivated scientists, primarily based in France but with partners in Argentina, Brazil, Germany, and Spain. The candidate will participate in collaborations with other teams working at the project within a multidisciplinary framework, in a joint undertaking between academia and a private company. There will be regular meetings with the experts involved, and potentially with external stakeholders. The post-doc will also present and co-publish research results in international conferences and/or journals.

Skills

- Organizational and interpersonal skills are necessary, including the ability to manage own workload, to coordinate team efforts, and to collaborate effectively across diverse groups to ensure effective project completion.
- Strong background in political economy, economics of conventions, or socio-economics is essential.
- Prior research experience on digital labour and/or the platform economy (also known as gig economy) is essential.
- Fieldwork experience with questionnaires and/or interviews is essential, preferably through online tools.
- Good experience with use of natural language processing (NLP) tools is essential.
- Ability to work in an interdisciplinary setting involving sociologists, linguists, and computer scientists, both from academia and industry, is essential.
- Ability to work in an international setting, involving collaboration with partners overseas, is essential.
- Fluency in English is essential.
- Fluency in French and/or Spanish is highly appreciated.
- Fluency in Portuguese is desirable.

Work Context

The post-doc will be based at CREST (Centre of Research in Economics and Statistics), at ENSAE Paris, located within the Institut Polytechnique de Paris (IP Paris) campus in Palaiseau. CREST is a large quantitative socio-economic science lab, which includes a Computational Social Sciences unit focused on data digital technologies and AI in the social sciences.
The post-doc's activities are part of the project VOLI ('Voices from Online Labour: Inequalities in digital earning activities across countries'), funded by ANR in 2024-28 and led by Paola Tubaro. The project aims to shed light on how class, gender, race and other gaps persist in labour platforms. To do so, VOLI innovatively combines hypotheses and methods from sociology, large-scale corpus linguistics, and speech technology/ AI to tackle the emergent economic and societal risks that coalesce around the nexus between online platform labour, automation, and social inequalities. In parallel, the methods developed within this interdisciplinary project aim to advance the linguistic study of speech variation, while also building and testing new tools for automated transcription.
In addition to CREST, the partners of VOLI include the Social and Economic Sciences department of Telecom Paris, also part of IP Paris; LISN, Laboratoire Interdisciplinaire des Sciences du Numérique, at Université Paris-Saclay; and CRISCO, Centre de Recherches Inter-langues sur la Signification en COntexte, at Université de Caen Normandie. Another partner is Vocapia Research, a French R&D company and software publisher in the speech and language technology field. International partners are Weizenbaum Institut Berlin (Germany), Universitat de València and Universitat Autònoma de Barcelona (Spain), Universidade do Estado de Minas Gerais (Brazil), and CONICET (Argentina).

Constraints and risks

The successful candidate is expected to participate actively in the activities of the CREST laboratory and the VOLI project. The position does not involve any specific risks.

To apply
Please submit two files: (1) a detailed CV, including a complete list of publications and names of three potential referees, and (2) a cover letter explaining your motivation and qualifications for this job.