The impact of AI on work
Many debates are taking place on the future of work in the face of the development of Artificial Intelligence: how many professions will be affected in the more or less near horizon, with what consequences (Frey and Osborne, 2017)? But, in fact, many professions are already experiencing great changes due to AI. It is these current perspectives that we wish to question through this workshop bringing together theoretical considerations and above all practical examples.
Monday 21 March 2022, from 2pm to 5pm
Ecole normale supérieure
Salle de séminaire du département de sciences des données
(3e étage, couloir entre les escalier B et C)
45 rue d’Ulm, 75005 Paris
Free entrance, no registration required
Program
2-3pm: Antonio Casilli (Télécom Paris / EHESS): “An end-to-end approach to the ethics of AI: the socio-economic dimensions of the production and deployment of automation”
Abstract: The contemporary debate on ethical AI embraces the interests and values of producers of automatic solutions, mainly located in Northern countries as well as in some emerging countries. Taking into account the contributions of digital social sciences, this presentation proposes an “end-to-end approach” to the ethics of AI, and highlights a socially responsible and ecologically sustainable technology. Its aim is to expand the scope of current research to include the systematic study of the social, economic and environmental contexts of AI production. In particular, we propose to take into account the extraction of natural resources and the work of annotation and data enrichment necessary to train, validate and sometimes even imitate AI systems. An end-to-end ethical AI aims to promote the autonomy and uphold the dignity of invisible AI workers, mainly located in the Global South.
Download Antonio Casilli’s slides
3-3:45pm : Thierry Poibeau (CNRS, ENS/PSL & Institut 3IA Prairie): “How machine translation has changed the translation profession”
Abstract: Machine translation (MT) tools have long remained highly imperfect, and attempts to use them in the workplace have been partial and experimental at best. The situation has changed in recent years, with the advent of neural methods that have made it possible to obtain translations of acceptable quality in different languages. This development has had a significant impact: some of the translations in the professional environment are now carried out automatically, and a new profession has even appeared: that of post-editor, which consists of marginally correcting the translations carried out automatically. We will examine the consequences of this state of affairs: consequences on the quality, costs and situation of professional translators. Is the influence of MT only negative for professional translators? Does it still open up new opportunities? We will try to answer these questions during the presentation, which will address the influence of AI on intellectual professions, which until recently were thought to be protected from this type of evolution.
Download Thierry Poibeau’s slides
3:45-4pm : Break
4-5pm : Discussion: AI and work, how to ensure training, equal opportunities and diversity?, with Tanya Perelmuter (Abeona Foundation) and Laure Bourgois (entrepreneur, director of Codataschool)
Download Tanya Perelmuter’s slides
Download Laure Bourgois’ slides
Biographies of the speakers
Laure Bourgois is a doctor in Artificial Intelligence, an expert in Symbolic AI and digital simulations. After 17 years of experience in R&D (Orange Labs, IFSTTAR, etc.), in 2017 she founded Codataschool (certified training organization specializing in AI and Big Data), with the aim of bringing together AI specialists for a school of computing that favors creativity and ethics.
Antonio A. Casilli is a professor of sociology at Télécom Paris, the telecommunications engineering school of the Institut Polytechnique de Paris. He is also an associate researcher at the Ecole des Hautes Etudes en Sciences Sociales (EHESS). He coordinates the DiPLab (Digital Platform Labor) research team and is one of the founders of the INDL (International Network on Digital Labor).
Tanya Perelmuter is co-founder and director of strategy and partnerships at the Abeona Foundation. She previously led partnerships at Riskdata, a fin-tech platform. Tanya is a data technology expert, with a particular interest in the risks and opportunities created by innovation in artificial intelligence. She is the general rapporteur of a report by the Institut Montaigne on algorithmic biases. Tanya holds an engineering degree from Columbia University, and a data science certification from École Polytechnique.
Thierry Poibeau is research director at the CNRS, assigned to the LATTICE laboratory (CNRS, ESN/PSL & U. Sorbonne nouvelle). He also holds a chair at the 3IA Prairie Institute. He is a specialist in automatic natural language processing and is interested in the consequences of the growing use of AI techniques on society.
References
Frey, C.B. and Osborne, M.A., 2017. The future of employment: how susceptible are jobs to computerisation?. Technological Forecasting and Social Change, 114, pp. 254-280.