Argyro Karanasiou & Dimitris Pinotsis to Give 11/30 Wed@Hariri Lecture
Wednesday, November 30, 2016
3:00 PM – 4:30 PM, networking and refreshments at 2:45 PM
Hariri Institute for Computing
111 Cummington Mall, Room 180
In collaboration with Boston University’s Reliable Information Systems and Cyber Security Group (RISCS), the Hariri Institute for Computing will be hosting Dr. Argyro Karanasiou (Senior Lecturer of Law, Bournemouth University) and Dr. Dimitris Pinotsis (Visiting Research Scientist, MIT) for a Wednesday@Hariri seminar
Emergent Normative and Legal Aspects of Automated Systems: The Intricacies of Machine Learning Algorithms
Abstract
In the advent of modern neural networks and machine learning, Automated Decision Making (ADM) has relied upon a plethora of algorithmic approaches and has found a wide range of applications in marketing automation, social networks, computational neuroscience, robotics and other fields. Given its range of applications, the purpose of this talk is to shed light on the legal questions posed by the broad use of ADM, which are key in determining all distinct rights and obligations of the programmer and the consumer. As a case study, driverless cars will be considered. This will also allow us to explore the several layers of human and machine interaction, which stretches beyond the scope of “data protection”. The distinct degrees of human interference and complexity in the underlying algorithms challenge axiomatic concepts in law, such as the notion of “personhood”. To understand this better, the discussion will then focus on opening the “black box”: how do deep neural networks find observed patterns in Big Data? The main aim of the talk is to provide a toolbox for measuring algorithmic determination in automated/semi-automated tasks; another black box yet to be opened.
Dr Argyro Karanasiou is an Assistant Professor (Senior Lecturer) specializing in IT and Media Law, affiliated with the Centre for Intellectual Property, Policy & Management (CIPPM) and with the Data Science Institute (DSI) at Bournemouth University (United Kingdom). Since August 2016, Argyro is a Visiting Research Fellow at the Internet Society Project (ISP) Center – Yale Law School. Her research discusses techno-legal conceptual frameworks towards a decentralized internet regulation with a particular focus on media ownership and user empowerment. Currently, Argyro is working on Deep Learning and its implications for the agent’s autonomy in automated systems. In the past, she has been involved in media related projects with the Council of Europe (Regional Expert on online media and reconciliation in South Eastern Europe) and with the OSCE Representative on Freedom of the Media. In 2013 (Indonesia) and 2015 (Brazil), Argyro was awarded an Internet Society IGF Ambassadorship and in 2014 she was named a PbD Ambassador by the Information and Privacy Commissioner in Ontario, Canada. In 2016, Argyro joined the EFF’s group of experts on Free Trade Agreements and Digital Services. Recently, Argyro was invited to submit evidence to the Royal Society on automated decision making and deep learning. Her current projects span a wide range of topics from IoT/wearable tech to big data, bioinformatics and mesh networks. Argyro tweets @ArKaranasiou.
Dr Dimitris Pinotsis is a Visiting Research Scientist at the Picower Institute for Learning and Memory and the Department of Brain and Cognitive Sciences at the Massachusetts Institute of Technology (MIT). He is also an Honorary Senior Research Associate at the Wellcome Trust Centre for Neuroimaging at University College London (UCL). Dimitris holds a PhD and an MSc in Mathematics from the Department of Applied Mathematics and Theoretical Physics (DAMTP) of the University of Cambridge, UK. His research has been funded by the US Air Force Office of Scientific Research, UK Research Councils (EPSRC) and the Wellcome Trust. It spans diverse areas including machine learning, the analysis of big data in neuroimaging, theoretical neurobiology and nonlinear systems in mathematical physics. In recent work, Dimitris exploits deep neural networks and hierarchical Bayesian inference to answer questions in attention, memory and decision-making. Dimitris is an Expert Reviewer for US Air Force Office of Scientific Research, the Austrian Science Fund and the Italian Ministry of Health. He tweets at @dimitrispp and shares his work at researchgate.net.