The Economics of the Cloud Market Workshop

Background and objectives

Public cloud computing is changing the face of the digital industries. Whereas firms, governments and non-profit organizations used to run most of their computing in-house, first through computers and then through servers, in recent years more and more of the world’s computations and data storage are carried out “in the cloud”. This raises important analytical and policy questions.

This workshop will provide a platform to share recent research results with the aim of helping policy makers better understand the economics of the sector and helping academics develop policy relevant work.

The event will feature academic presentations and two round-table discussions on the following themes:

1st Roundtable - Issues in the regulation and analysis of the cloud economy

The following questions will be discussed, focusing both on the policy issues and on the understanding of the underlying economics:

• Why has the concentration of the cloud industry increased in recent years?
• What are the limits of interoperability in the cloud? Can third party service providers help with interoperability?
• Switching costs are very high in the cloud. To what extent is that intrinsic to the industry and to what extent is that a consequence of the choices made by users and by cloud service providers?
• What role do egress fees play in the strategies of the cloud service providers?

2nd Roundtable - The cloud and AI

There is a profound link between AI and the cloud. In particular, the massive computing power needed for the development of the cloud creates both deep analytical and important policy issues, which have not been prevalent enough in the current policy debates. The round table will focus on the ways in which we understand the challenges of the infrastructure of AI for regulators and academics. Two examples:

• All the Cloud Service Providers have important investments in some AI models. To foster a healthy competition in the AI market, it is important that all developers of AI models have access to quality computing power. The technical and contractual difficulties associated with this objective will be discussed alongside what competition and regulatory authorities can do to foster this aim.
• There is limited access to specialized microprocessors, and Nvdia, for instance, distributes chips according to its strategic needs. This can be interpreted as a strategy to foster a diversified and innovative ecosystem of users. It can also be interpreted as an anticompetitive strategy. The tools that economics has to analyze these strategies and to distinguish between them will be examined.

Organizing committee

Jacques CREMER (TSE)
Gary BIGLAISER (UNC, TSE)
Alexandre DE CORNIERE (TSE)
Andrea MANTOVANI (TBS Education, TSE)

Conference secretariat

Emilie LOPEZ - emilie.lopez@tse-fr.eu

Partners

We thank the partners of the TSE Digital Center for their support.