How to Use Artificial Intelligence to Enhance Human Moral Intelligence
July 25, 2018 - August 31, 2020
Core Funding Area:
Director: Jana Schaich BorgInstitution: Duke University
Ignorance, confusion, and partiality often distort human moral judgments. To reduce such mistakes, humans can use feedback from additional sources of information. In other words, we need a solution that helps us make accurate moral judgments informed by morally relevant facts from a variety channels.
Jana Schaich Borg and Walter Sinnott-Armstrong propose using algorithms to collect and digest the judgments of large numbers of people to serve as a resource to enhance individual ethical capacities. By pairing the power of AI systems and human wisdom, he hopes to offer a tool for strengthening our moral capacities.
Computers can avoid ignorance, confusion, and partiality, reaching moral judgments that capture fundamental human values. Such moral algorithms cannot replace human moral judgment. Humans must remain in the loop. Still, moral algorithms can provide “tools to empower human moral intelligence.”
The aim of this project is to develop such “new systems for human–machine interactions,” as the Diverse Intelligences Morality in the Machine Age Challenge says. Since 2015, the team has been exploring ways to program human morality into computers, using kidney exchanges as a test case. Now they are extending their previous findings in new and exciting directions.
There are three goals:
1. First, determine how human moral judgments suffer from certain defects like ignorance, bias, and fatigue .
2. Second, compare human moral judgments within and between different groups of people to measure the inevitable variations among groups.
3. Third, combine individual and group moral judgments to guide social choices.
4. Fourth, determine the best way to present information so that decision-makers become more receptive to guidance.
This work could reshape the field of computer science as well as academic and popular understandings of morality and artificial intelligence. In the long term, it might enable practical uses of algorithms to improve human moral judgments and a multitude of public policies.
Opinions expressed on this page, or any media linked to it, do not necessarily reflect the views of Templeton World Charity Foundation, Inc. Templeton World Charity Foundation, Inc. does not control the content of external links.