1. médialab Sciences Po
  2. Actualités
  3. Journée doctorants TransNum

Journée doctorants TransNum

Comme chaque année, TransNum consacre une séance de son séminaire aux travaux de jeunes chercheur.e.s.

Rendez-vous, Séminaire TransNum

Cette année, le 25 juin 2021 (9h30-12h30), la séance sera consacrée aux travaux de deux doctorant.e.s et d'une jeune docteure. 

Chaque présentation de 20 minutes de leurs travaux, sera suivie par 35 minutes de discussion collective. 


9h30 : Aifang Ma - “The conundrum of Chinese Internet and social media firms caught in an authoritarian setting: between compliance and “feigned” compliance during the coronavirus period”

It is generally argued that state-society relations in authoritarian regimes are highly unequal, with the state crowding out society in governing the country without being held accountable. As the largest authoritarian regime in the world, China is characterized among others by its indisputably strong state capacities with which the Party-State occupies a dominating position in comparison with firms and civil society. However, several cautionary notes should be added in order to reconsider, not how the Party-State is powerful, but how firms and the civil society alike are not powerless.

The relations between the Party-State and the Chinese civil society have been discussed in great depth in the current literature dealing with rights defense movements.

In order to complete the discussion of state-society relations in China, this paper pays special attention to Chinese Internet and social media firms (thereafter ISMF). It aims to look into how the ISMF in China creatively deal with the instructions of the Party-State that run against their pursuit of commercial interests.

Operating in an authoritarian regime indeed poses a conundrum to firms. On the one hand, they have to comply with regulators’ requirements in terms of what types of contents are allowed to circulate over the virtual space, such that their licenses are not revoked. On the other hand, they have to prop up attractive and valuable information for their consumers, who are more likely than regulators to pull up the hit rates for firms. This paper reviews content regulation policies in China and studies three strategies mobilized by Chinese ISMF to protect their interests as commercial entities when faced with unfavorable content regulation rules: 1) adaptation of the contents taken from official news outlets, 2) reporting scandals involving public authorities sporadically and 3) echoing with online discussions in order to reduce the possibilities of being sanctioned.

Aifang Ma est doctorante en science politique au Centre d'études européennes de Sciences Po.

10h30 : Jean-Marie John Mathew - "Some Critical and Ethical Perspectives on the Empirical Turn of AI Interpretability"

We consider two fundamental and related issues currently faced by Artificial Intelligence (AI) development: the lack of ethics and interpretability of AI decisions. Can interpretable AI decisions help to address ethics in AI? Using a randomized study, we experimentally show that the empirical and liberal turn of the production of explanations tends to select AI explanations with a low denunciatory power. Under certain conditions, interpretability tools are therefore not means but, paradoxically, obstacles to the production of ethical AI since they can give the illusion of being sensitive to ethical incidents. We also show that the denunciatory power of AI explanations is highly dependent on the context in which the explanation takes place, such as the gender or education level of the person to whom the explication is intended for. AI ethics tools are therefore sometimes too flexible and self-regulation through the liberal production of explanations do not seem to be enough to address ethical issues. We then propose two scenarios for the future development of ethical AI: more external regulation or more liberalization of AI explanations. These two opposite paths will play a major role on the future development of ethical AI.

Jean-Marie John Mathew est doctorant à l'institut Mines Télécom en sciences de gestion sur le sujet de l'éthique de l'intelligence artificielle

11h30 : Apolline Taillandier - “Des machines féminines aux algorithmes sexistes. Identité, justice et exclusion dans les débats sur l’intelligence artificielle depuis les années 1980”

Ce projet postdoctoral vise à retracer l'histoire des critiques féministes de l'intelligence artificielle et le rôle des débats sur la justice et la nature des rapports de sexe et de race dans la redéfinition des projets d’IA depuis les années 1980. D’une part, il vise à clarifier l’inscription des questionnements liés au genre dans le champ de l’informatique à partir d’analyses d’archives et d’entretiens avec des chercheurs et chercheuses aux Etats-Unis, au Royaume Uni et en France, dans le contexte du développement des épistémologies féministes (e.g. Harding, Hintikka, Fox Keller), des études féministes de la science et de la technologie (Adam, Wajcman) et des contestations de l’IA symbolique (Turkle et Papert). Il propose ensuite de cartographier les conceptions de la justice, de l’identité, et de l’intelligence dans les débats actuels sur la justice algorithme et le développement de méthodes d'IA éthique, responsable ou durable, par une analyse du champ transnational de la recherche en IA et une étude de plusieurs cas de programmes de recherche croisant machine learning et philosophie morale et politique.

Apolline Taillandier est docteure en science politique et chercheuse postdoctorale à l'Université de Bonn et à l'Université de Cambridge (CST-LCFI/POLIS). Ses recherches croisent l'histoire des idées politiques et l'étude des sciences et technologies. Sa thèse soutenue le 28 mai 2021 était consacrée aux discours libéraux des projets transhumanistes contemporains.