A three-year effort by hundreds of engineers worldwide resulted in the publication in March of 2019 of Ethically Aligned Design (EAD) for Business, a guide for policymakers, engineers, designers, developers and corporations. The effort was headed by the IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems (A/IS), with John C. Havens as Executive Director, who spoke to AI Trends for an Executive Interview. We recently connected to ask how the effort has been going. Here is an update.
EAD First Edition, a 290-page document which Havens refers to as “applied ethics,” has seen some uptake, for example by IBM, which referred to the IEEE effort within their own resource called Everyday Ethics for AI The IBM document is 26 pages, easy to digest, structured into five areas of focus, each with recommended action steps and an example. The example for Accountability involved an AI team developing applications for a hotel. Among the recommendations was: enable guests to turn the AI off, conduct face-to-face interviews to help develop requirements; and, institute a feedback learning loop.
The OECD (Organization for Economic Cooperation and Development) issued a paper after the release of an earlier version of EAD attesting to the close affinity between the IEEE’s work and the OECD Principles on AI. The OECD cited as shared values “the need for such systems to primarily serve human well-being through inclusive and sustainable growth; to respect human-centered values and fairness; and to be robust, safe and dependable, including through transparency, explainability and accountability.”
John C. Havens, Executive Director, IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems
Based on the success of working to produce a more focused document based on the full EAD, Havens said he is now working with other groups to produce focused guides on EAD for parents, for artists, for the financial industry, as well as for health and for sustainability.
It can be difficult for the IEEE effort not to get buried in the myriad of principles documents surrounding ethical AI development, “maybe 80 of them,” Havens said. Usually the IEEE does get mentioned prominently. Since its release, the EAD has spawned 13 standards working groups within IEEE, and two more will be launched in coming months. “A more pragmatic instantiation of the ideas will come to light,” Havens said.
Another effort is around an effort to enable programmers, engineers, technologists and business managers to better consider how the products and services they create can increase human well-being. (See IEEE P7010 Well-Being Metrics Standard for Autonomous and Intelligent Systems.) Well-being is defined in the draft document by a spectrum of measures, including environmental and educational issues, feelings and mental health.
“The logic is how to have an applied ethics foundation methodology that syncs with how engineers work,” said Havens. “We’re trying to make it easier for them to get the cross-pollinated support they need from non-technical colleagues.” He noted that many of the contributors to the EAD for the IEEE are in academia, so that its principles are likely to be incorporated into college curriculums, where students will be absorbing it.
Asked if he could cite an example of how the IEEE’s ethical AI effort is playing out, Havens mentioned Dr. AJ Moon, an experimental roboticist now on the faculty of McGill University in Montreal, who is on the executive committee for The IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems that produced the EAD. She is also the founder and director of the Open Roboethics Institute. That site is publishing research on topics such as whether AI in healthcare has been “ethics-proofed.”
In the area of personal data privacy, the EAD talks about “data agency” and new ways individuals can grant permission for how their data is used. Asked for an example of how this is playing out, Havens mentioned an app called Private Kit: Safe Paths, developed by a team at the MIT Media Lab led by Ramesh Raskar and included developers from Harvard, Facebook and Uber.
This app is now being used to help track the spread of the coronavirus. It works by sharing encrypted location data between phones in a way that does not go through a central authority. In this way, a user can see if they might have come in contact with someone carrying the coronavirus, without knowing who it might be. A person using the app who tests positive, can choose to share location data with health officials, who could then make it public.
Getting the word out is now the challenge for that effort. “The user adoption strategy will leverage network effects,” Raskar stated in a recent interview in Digital Trends. “We built a web tool for health authorities to disseminate privatized trails. We will use this two-sided network effect to first push the health players in focused, localized site, then let users nudge their acquaintances so collectively they have a ‘peace of mind.’”
Before joining IEEE, Havens worked as a consultant for Gliimpse, a company acquired by Apple in 2016, that practiced “data sovereignty” in the way it allowed personal health data to be shared with healthcare providers directly. The company was founded in 2013 by Anil Sethi and Karthik Hariharan. The software essentially allowed an individual to compile a personal Electronic Health Record, to be shared directly to healthcare providers.
“Gliimpse provided a portable version of your health data, which empowers you as a patient,” Havens said. It can also be used to send healthcare information on children to summer camps, for example.
Sethi later left Apple to form Citizen, a consumer health technology company with a mission to give patients control of their health records. He started the company after the loss of his sister Tamira to breast cancer; she made him promise to do something about health care before she died.
The IEEE Ethics effort is also concerned with defining “well-being metrics.” Overall, well-being is defined as the experience of health, happiness and prosperity, according to Psychology Today. Fortunately, lots of related measurable data is provided in the latest version of the World Happiness Report, recently issued by the Sustainable Development Solutions Network, which is supported by a wide range of impressive organizations and individuals, including the Gallup World Poll. “Gallup has the most data on individually oriented well-being,” Havens said.
The UN report conveys the notion of “self-reporting” on life satisfaction on a zero to 10 scale, with zero being pure misery. This notion of providing a Life Satisfaction score is based on where a person lives in the world and their subjective reporting of their individual experience. The UN report explores, for example, why citizens of Nordic countries have a high life satisfaction. “The point for our work is that well-being needs to be taken into account by engineers or anyone when designing systems,” said Havens. “The UN report provides specific quantitative and qualitative empirical data to help anyone creating AI Systems to do just that.”