The online interactive magazine of the Association for the Advancement of Artificial Intelligence

Cover of AI Magazine issue 267

Vol 42 No 1: Spring 2021 | Published: 2021-04-12

 

Zeitenwende — Turn of Eras

A new wave of innovation disrupts our social and economic life as we have lived it in the 20th century. On the one hand, artificial intelligence (AI) has the potential to increase our efficiency and ability to cope with the changes that are coming. On the other hand, AI itself also causes disruptions that may be managed best by contributions from other fields. We aim to start an interdisciplinary discourse with this special issue and bring together a set of articles from AI researchers, sociologists, economists, and psychologists who discuss five basic questions: How can AI help? How will AI affect economy and work? How does AI affect democracy? How does AI affect our togetherness? And how does AI affect the human self?

Contributors

Meinolf Sellmann
GE Global Research

Bart Selman
Cornell University

PDF

Rethinking the Maturity of Artificial Intelligence in Safety-Critical Settings

Artificial intelligence, in the form of machine learning, has the potential to transform many safety-critical applications such as those in transportation and healthcare. However, despite significant investment and impressive demonstrations, such technologies have struggled to live up to their promises. To this end, this article illustrates that machine learning fundamentally lacks the ability to leverage top-down reasoning, a critical element in safety-critical systems. This is especially important in situations where uncertainty can grow very quickly, requiring adaption to unknowns. This fundamental lack of contextual reasoning, combined with a lack of understanding of what constitutes maturity in artificial intelligence-embedded systems, has significantly contributed to the failures of these systems. Demonstrations where safety-critical artificial intelligence-enabled systems function as if they were almost operational should not be a substitute for testing. Instead, companies and regulatory agencies need to work together to develop clear criteria and certification protocols before such technologies are made publicly available.

Contributors

Mary L. Cummings
Duke University

PDF

Artificial Intelligence and the Future of Work: A Proactive Strategy

One of the greatest challenges and opportunities of our time lies in harnessing the innovative potential of emerging technologies to help achieve a more prosperous and just society. Discussions of how to do so are at the center of debates over how artificial intelligence, machine learning, and associated tools might affect the future of work. In this article, I will outline a proactive strategy for addressing these issues.

Contributors

Thomas A. Kochan
MIT Sloan School of Management, Institute for Work and Employment Research ISIP member, AIAA Associate Fellow, SPIE Fellow, and IEEE Fellow.

PDF

Artificial Intelligence Community and Self

The means by which artificial intelligence systems will both interact and influence individuals and community are considered, with a description of current capabilities such as video analytics, agency, natural language processing, and online data analysis. Open research questions such as the roles of analogies, associative memory, and the grounding problem are discussed, with speculation presented regarding the possible ramification of this research agenda. Such topics will include how artificial intelligence and humanity might co-evolve; artificial creativity; and life without work.

Contributors

Peter Tu
General Electric

PDF

Procedural Democracy and Electronic Agents

In recent years, firms offering political campaigning and fundraising services have claimed that artificial intelligence is used to maximize their clients’ interests. Numerous news reports suggest that artificial intelligence capacities have influenced high-profile voting events; publications decrying the threat posed to democratic procedures are proliferating. This article takes a skeptical view of such claims, with an argument in six sections. First, the received principles of procedural democracy in relation to voting events are outlined. Second, the kind of services in question are considered. Third, to what extent these services actually involve artificial intelligence is pondered. Fourth, how such technological capacities may bear upon the tenets of procedural democracy are raised. Fifth, it is argued that it is unlikely that those tenets are indeed significantly undermined in practice. And sixth, this article concludes by considering the implications of it being proven that such technological capacities do play a decisive role in voting events.

Contributors

Suman Gupta
The Open University

PDF

Artificial and Human Intelligence in Mental Health

While technology has dramatically changed medical practice, various aspects of mental health practice and diagnosis remain almost unchanged across decades. Here we argue that artificial intelligence — with its capacity to learn and infer from data the workings of the human mind — may rapidly change this scenario. However, this process will not happen without friction and will promote an explicit reflection of the overarching goals and foundational aspects of mental health. We suggest that the converse relation is also very likely to happen. The application of artificial intelligence to a field that relates to the foundations of what makes us human — our volition, our thoughts, our pains and pleasures — may shift artificial intelligence back to its earliest days, when it was mostly conceived of as a laboratory to explore the limits and possibilities of human intelligence.

Contributors

Mariano Sigman
Universidad Torcuato Di Tella

Diego Fernandez Slezak
Universidad de Buenos Aires

Lucas Drucaroff
Universidad de Buenos Aires

Sidarta Ribeiro
Federal University of Rio Grande do Norte

Facundo Carrillo
Universidad de Buenos Aires

PDF

Machine Learning Techniques for Accountability

Artificial intelligence systems have provided us with many everyday conveniences. We can easily search for information across millions of webpages via text and voice. Paperwork processing is increasingly automated. Artificial intelligence systems flag potentially fraudulent credit-card transactions and filter our e-mail. Yet these artificial intelligence systems have also experienced significant failings. Across a range of applications, including loan approvals, disease severity scores, hiring algorithms, and face recognition, artificial-intelligence–based scoring systems have exhibited gender and racial bias. Self-driving cars have had serious accidents. As these systems become more prevalent, it is increasingly important that we identify the best ways to keep them accountable.

Contributors

Been Kim
Google Brain

Finale Doshi-Velez
Harvard University

PDF

Using a Machine Learning Tool to Support High-Stakes Decisions in Child Protection

Machine learning decision support tools have become popular in a range of social domains including healthcare, criminal justice, and child welfare. But the design of these tools often fails to consider the potentially complex interactions that happen between the tools and humans. This lack of human-centered design is one reason that so few tools are actually deployed, and even if they are, struggle to achieve impact. In this article we present the example of the Allegheny Family Screening Tool, a machine learning model used since 2016 to support hotline screening of child maltreatment referrals. We describe aspects of human-centered design that contributed to the successful deployment of this tool, including agency leadership and ownership, transparency by design, ethical oversight, community engagement, and social license. Finally, we identify potential next-steps to encourage greater integration of human-centered design into the development and implementation of machine learning decision support tools.

Contributors

Rhema Vaithianathan
Center for Social Data Analytics

Diana Benavides-Prado
Centre for Social Data Analytics

Erin Dalton
Allegheny County Department of Human Services

Alex Chouldechova
Carnegie Mellon University

Emily Putnam-Hornstein
University of North Carolina

PDF

The Case Against Registered Reports

Registered reports have been proposed as a way to move from eye-catching and surprising results and toward methodologically sound practices and interesting research questions. However, none of the top-twenty artificial intelligence journals support registered reports, and no traces of registered reports can be found in the field of artificial intelligence. Is this because they do not provide value for the type of research that is conducted in the field of artificial intelligence?

Contributors

Odd Erik Gundersen
Norwegian University of Science and Technology

PDF

The Role of Open-Source Software in Artificial Intelligence

With this publication, we launch a new column for AI Magazine on the role of open-source software in artificial intelligence. As the column editor, I would like to extend my welcome and invite AI Magazine readers to send short articles for future columns, which may appear in the traditional print version of AI Magazine, or on the AI Magazine interactive site currently under development. This introductory column serves to highlight my interests in open-source software and to propose a few topics for future columns.

Contributors

Jim Spohrer
IBM

PDF

AAAI's New Award for the Societal Benefits of AI — An Interview with Richard Tong

In this interview, conducted in early 2020 by Ashok Goel, Richard Tong, the chief architect and general manager of Squirrel AI Learning’s US operations, discusses adaptive learning, challenges facing AI in education, and the Squirrel AI Award for Artificial Intelligence to Benefit Humanity.

Contributors

Richard Tong
Squirrel AI

Ashok Goel
Georgia Institute of Technology

PDF

The Role of Open-Source Software in Artificial Intelligence

With this publication, we launch a new column for AI Magazine on the role of open-source software in artificial intelligence. As the column editor, I would like to extend my welcome and invite AI Magazine readers to send short articles for future columns, which may appear in the traditional print version of AI Magazine, or on the AI Magazine interactive site currently under development. This introductory column serves to highlight my interests in open-source software and to propose a few topics for future columns.

Contributors

Jim Spohrer
IBM

PDF

Artificial Intelligence’s Grand Challenges: Past, Present, and Future

Innovative, bold initiatives that capture the imagination of researchers and system builders are often required to spur a field of science or technology forward. A vision for the future of artificial intelligence was laid out by Turing Award winner Raj Reddy in his 1988 Presidential address to the Association for the Advancement of Artificial Intelligence. It is time to provide an accounting of the progress that has been made in the field, over the last three decades, toward the challenge goals. While some tasks such as the world-champion chess machine were accomplished in short order, many others, such as self-replicating systems, require more focus and breakthroughs for completion. A new set of challenges for the current decade is also proposed, spanning the health, wealth, and wisdom spheres.

Contributors

Ganesh Mani
Frank Chen
Steve Cross
Thomas Kalil
Vanathi Gopalakrishnan
Francesca Rossi
Ken Stanley

PDF

Organizing a Successful Artificial Intelligence Online Conference: Lessons from the 13th Symposium on Combinatorial Search

The 13th Symposium on Combinatorial Search (SoCS) was held May 26 to 28, 2020. Originally scheduled to take place in Vienna, Austria, the symposium pivoted toward a fully online technical program in early March. As an in-person event, SoCS offers participants a diverse array of scholarly activities including technical talks (long and short), poster sessions, plenary sessions, a community meeting and, new for 2020, a Master Class tutorial program. This article describes challenges, approaches, and opportunities associated with adapting these many different activities to the online setting. We consider issues such as scheduling, dissemination, attendee interaction, and community engagement before, during, and after the event. In each case, we report on the approaches taken by SoCS, then give a post hoc analysis of their effectiveness and discuss how these decisions continue to impact the SoCS community in the days after SoCS 2020. This work will be of interest to organizers of similar conferences who may be considering the switch to an online format.

Contributors

Daniel Harabor

Mauro Vallati
University of Huddersfield

PDF

Advances in Theory and Applications of Artificial Intelligence

The 33rd International Conference on Industrial, Engineering, and Other Applications of Applied Intelligent Systems was held in Kitakyushu, Japan on 22 to 25 September 2020. This report provides an overview and summary of the conference.

Contributors

Hamido Fujita
University of Granada

Philippe Fournier-Viger
Harbin Institute of Technology

Jun Sasaki
Iwate Prefectural University

Moonis Ali
Texas State University

PDF

Report on the First and Second Workshops on Hierarchical Planning Held at the International Conference on Automated Planning and Scheduling

Hierarchical planning has attracted renewed interest in the last few years. Consequently, the time was right to establish a workshop devoted entirely to hierarchical planning — an insight shared by many supporters. In this article, we report on the first International Conference on Automated Planning and Scheduling workshop on hierarchical planning held in Delft, The Netherlands, in 2018 as well as on the second workshop held in Berkeley, CA, USA, in 2019.

Contributors

Pascal Bercher
Australian National University

Daniel Höller
Saarland University

Gregor Behnke
University of Freiburg

Susanne Biundo
Ulm University

Vikas Shivashankar
Amazon Robotics

Ron Alford
MITRE Corporation

PDF