2020 Symposium Publication Abstracts
Some of the panelists from our 2020 Symposium: Eyes on Me will be contributing related articles to the Law Review for publication after the event. Below are the abstracts.
Beyond Privacy and Transparency: Three Additional Features Innovative Businesses Should Build into Algorithmic Designs
By Peter K. Yu (余家明)
Commentators have widely noted the urgency of building privacy and transparency into algorithmic designs. While those features are important and badly needed, this essay calls for the inclusion of three additional features into those designs: inclusiveness, intervenability and interoperability. The inclusion of these features will be especially important in the age of artificial intelligence and in view of the greater deployment of machine-learning algorithms.
Gaming Algorithms and Other Time Paradoxes
By Roger Ford
A recurring debate over decision-making algorithms concerns their susceptibility to gaming: If someone knows an algorithm will make a decision, she might be able to game it, potentially eliminating many of the benefits of using the algorithm in the first place. So, for example, an algorithm designed to grade student essays might easily be fooled by gibberish, rewarding complex sentences and vocabulary instead of logical reasoning. Scholars and engineers have proposed and debated responses like making algorithms less transparent, which could make them harder to game but have adverse consequences for fairness and accuracy.
Gaming, however, is not the only problem of its kind with decision-making algorithms; instead, it is one example of a broader class of time paradoxes that arise when algorithms make decisions. The fundamental assumption of a decision-making algorithm is that the world in which it will be deployed is the same as the one in which it was developed, but this assumption breaks when people respond differently in the two worlds. A student would not intentionally submit gibberish to a human grader but might happily do so with an algorithm. But people can respond differently for reasons beyond gaming. Patients might respond differently, for example, to medical decisions made by humans and algorithms. They might trust one more or understand it better, and so be more likely to follow instructions. Critically, this means that the correct decision might itself be different in the two worlds. And it suggests remedies beyond simple opacity may needed.
The US Copyright Act: 2020 and Beyond
By Jaia Thomas
The US Copyright Act was enacted on October 4, 1976, well before the advent of Google, Facebook, Netflix, Instagram, Spotify, iPhones and Emojis. While society has continued to evolve in the areas of technology and entertainment, copyright law has been slower to adapt. As 2020 ushers in a new decade, the time is ripe for the US Copyright Office to usher in a new framework for modern day copyright registration and protection. This article will examine those provisions of The US Copyright Act that are outdated and those which will need to be updated in this new decade.
This article examines the US Copyright Act and provides ways in which three specific areas must be updated to better align with today’s current technological and entertainment landscape. Specifically, the article examines: (i). copyright registration subject matter, (ii). music copyright infringement and (iii). Fair Use. First, the article examines the current list of subject matter available for copyright registration. Considering such recent cases as Oracle v. Google, the article examines what types of new subject matter should now be eligible copyright registration. Second, district and federal courts saw their fair share of music infringement cases in the last decade. The article examines some of the most seminal music copyright infringement cases and offers solutions for a more streamlined and efficient system for music copyright protection and licensing. Lastly, while performing a comparative analysis of recently enacted international statutes, the article offers an updated approach to the doctrine of fair use.