Values in AI: bioethics and the intentions of machines and people

Abstract

Artificial intelligence has the potential to impose the values of its creators on its users, those affected by it, and society. The intentions of creators as well as investors may not comport with the values of users and broader society. Users also may mean to use a technological device in an illicit or unexpected way. Devices change people's intentions as they are empowered by technology. What people mean to do with the help of technology reflects their choices, preferences, and values. Technology is a disruptor that impacts society as a whole. Without knowing who intends to do what, it is difficult to rely on the creators of technology to choose methods and create products that comport with user and broader societal values. The AI is programmed to accomplish tasks according to chosen values or is doing so through machine learning and deep learning. We assert that AI is quasi-intentional and changes people's intentions. Investors wishing to promote or preserve public health, wellbeing, and wellness should invest in ethical, responsible technology. Environmental, social, and governance (ESG) considerations and metrics should include ethical technology, wellness, public health, and societal wellbeing. This paper concludes that the process by which technology creators infuse values should be couched in bioethical and general ethical considerations, reflective of potential multiple intentions, and should entail a willingness and process to adapt the AI after the fact as the circumstances of its use change.

Socialisation approach to AI value acquisition: enabling flexible ethical navigation with built-in receptiveness to social influence

“…The broader threat of over-personified technology and specific concerns over interactive AI applications designed to fill intimate social roles are both beyond this work. Instead, see a recent ethics article by Zimmerman and colleagues [71].…”