By Gaia Gallotti
The crucial role of ethics within the field of artificial intelligence (AI) is one of the most relevant challenges, as increasing concerns are emerging regarding the prevailing preference-based utilitarianism approach in AI ethics. Among others, the “humanistic” ethics for AI represents an alternative ethical perspective known as calling for a more comprehensive understanding of AI and its ethics.
This humanistic approach is based on three elements. Firstly, pluralism, as this new approach acknowledges the different values that influence ethical considerations, such as the various aspects of humanity and other moral components, and eliminates the notion that a single general concept can address all ethical aspects of AI. Thus, this perspective does not reduce AI to a singular master concept.
Secondly, this approach emphasizes the significance of the procedures through which AI reaches decisions, rather than focusing only on the outcomes. In doing so, it highlights that procedural values like transparency, fairness, and accountability are fundamental for a more ethical AI, meaning that the process through which AI systems make decisions should be as ethical as their outcomes.
The last element is participation: this is considered crucial in relation to AI decision-making processes and it includes participation in the workplace, and democratic participation to ensure that AI technologies serve human wellbeing and keep democratic values. Participation enhances autonomy and self-fulfillment while preventing the erosion of democratic principles and the concentration of power.
On the other hand, it is crucial to analyze the potential challenges arising from artificial general intelligence (AGI) and the task of integrating intelligent machines into existing ethical frameworks. As a matter of fact, questions and issues arise about the source of human dignity and its relationship with the value of non-human identities. While we can recognize that AI’s development is uncertain and speculative, it is fundamental to effectively integrate AI technologies into our lives in a way that respects and enhances the well-being and dignity of humans and other beings that share our world.
OCCAM continues its effort to support and promote a more comprehensive and human-centered approach to AI in order to overcome the limitations of simple utilitarianism. A more inclusive consideration of values is needed, in order to emphasize the deep impact of AI on society, work, and democracy.
John Tasioulas; Artificial Intelligence, Humanistic Ethics. Daedalus 2022; 151 (2): 232–243. doi: https://doi.org/10.1162/daed_a_01912