AI for Social Good—A Faustian Bargain

Abstract

Artificial intelligence (AI) technology sits in our pockets and homes as smart devices in the guise of friendly personal assistants. We seem to have made a Faustian bargain with these smart devices powered by AI technology to help us live our life, to help us master our schedules, keep fit, keep up with our work and family communications, keep us entertained, allow us to shop from home, and among other things to manage our banking and finances from home. While making us immensely efficient or productive, all our personal information is being harvested. These smart devices and their application programs are engineered to make users addicted to look for “notifications” and constantly feed their data as “updates.” This is the Faustian bargain of we seem to have made at an individual level. In the public domain and for humanity at large, is there another one? AI is also seen to be contributing to “Social Good” in many domains such as health care, agriculture, manufacturing, energy management, transportation, and governance. At the root of most modern AI, applications are an adaptive machine learning component that needs properly labeled data. Data-centric algorithms create an insatiable quest for all kinds of data, leading to unprecedented profits for mega international corporations. They have engineered applications and platforms to harvest personal data from millions of people. Governments don’t seem to mind if the corporations are seen to provide services and aid governance. In this paper, we wish to review efforts and approaches where governments and industry take up more responsible positions on the acquisition of data and regulate its use. Is it possible that Faustian bargain of private data for services might be regulated or positioned more equitably? Formulation of standards by organizations such as the IEEE and their stance might be a way forward.