Ai, Society and teach innovation

Quali sono le sfide principali che la tutela della privacy vive oggi di fronte alle nuove tecnologie? Il Vicepresidente del Garante per la protezione dei dati personali offrirà una panoramica dei temi più caldi, per poi affrontarne alcuni in particolare.

How judges and lawyers figure out what a word means within a legal text? This has always been a thorny question. Among various patterns of legal interpretation, the corpus linguistics approach is a rising one. It exploits big data to understand the ordinary meaning of a word for legal purposes.

Fighting the pandemic is pitting health and privacy against each other in largely unprecedented ways. A conversation between two prominent legal thinkers will compare the approaches to contact tracing apps and mass surveillance in the United States and the European Union.

We will begin the seminar by asking our guests the following questions:

Within the West, many countries and supranational legal systems are struggling to find ways to reconcile the fight against the COVID-19 with the needs of preserving fundamental rights, such as privacy, mobility, etc. A lot of debates revolve around the deployment of AI-based technologies, which track down people and contagion, as they would make extreme physical constraints, such as lockdowns, unnecessary. What are the main legal issues within the EU and the U.S. that such possibilities need to face?

As we expect that the virus will stay with us for long, we are likely to need digital tools for long as well. Once the Schengen treaty returns in full force and EU citizens resort to moving quite freely within EU territories, the apps we are thinking of will ensure inter-state tracing? Is the US facing with analogous issues of decentralization?

It is rather frequent to hear the opinion that we are overestimating the issues of privacy, given the importance of health. Moreover, it is common parlance that the big tech companies already know so much about us, that is paradoxical to fear the Government and massive utilization of data to fight the virus. Are privacy issues almost fictional, in light of the tech giants’ omnipotence?

A growing body of work shows that many problems in fairness, accountability, transparency, and ethics in machine learning systems are rooted in decisions surrounding the data collection and annotation process. We argue that a new specialization should be formed within machine learning that is focused on methodologies for data collection and annotation: efforts that require institutional frameworks and procedures. Specifically for sociocultural data, parallels can be drawn from archives and libraries.

Archives are the longest standing communal effort to gather human information and archive scholars have already developed the language and procedures to address and discuss many challenges pertaining to data collection such as consent, power, inclusivity, transparency, and ethics privacy. We discuss these five key approaches in document collection practices in archives that can inform data collection in sociocultural machine learning.

Timnit Gebru is the technical co-lead of the Ethical Artificial Intelligence Team at Google. She works on algorithmic bias and data mining. Timnit earned her doctorate under the supervision of Fei-Fei Li at Stanford University in 2017. She is an advocate for diversity in technology and is the cofounder of Black in AI, a community of black researchers working in artificial intelligence.

Tommaso Gecchelin is a physicist and industrial designer from Padua, in Italy. He’s the founder, inventor and CTO of Next Future Transportation inc. developing modular autonomous pods. His mission is to merge science and design to create useful and elegant solutions to solve major world problems in the field of logistics and transportation, politics and decision making, environment and sustainability.