Artificial Intelligence, Values and Alignment

  title={Artificial Intelligence, Values and Alignment},
  author={Iason Gabriel},
This paper looks at philosophical questions that arise in the context of AI alignment. It defends three propositions. First, normative and technical aspects of the AI alignment problem are interrelated, creating space for productive engagement between people working in both domains. Second, it is important to be clear about the goal of alignment. There are significant differences between AI that aligns with instructions, intentions, revealed preferences, ideal preferences, interests and values… 

Challenges of Aligning Artificial Intelligence with Human Values

It is shown that although it is not realistic to reach an agreement on what humans really want as people value different things and seek different ends, it may be possible to agree on what the authors do not want to happen, considering the possibility that intelligence, equal to their own, or even exceeding it, can be created.

The Challenge of Value Alignment: from Fairer Algorithms to AI Safety

It is suggested that more attention needs to be paid to the question of 'social value alignment' - that is, how to align AI systems with the plurality of values endorsed by groups of people, especially on the global level.

AI Development for the Public Interest: From Abstraction Traps to Sociotechnical Risks

It is shown that for each subfield, perceptions of PIT stem from the particular dangers faced by past integration of technical systems within a normative social order, and a roadmap for a unified approach to sociotechnical graduate pedogogy in AI is presented.

AI Loyalty: A New Paradigm for Aligning Stakeholder Interests

This work suggests criteria for assessing whether an AI system is acting in a manner that is loyal to the user, and argues that AI loyalty should be deliberately considered during the technological design process alongside other important values in AI ethics, such as fairness, accountability privacy, and equity.

Reinforcement Learning Under Moral Uncertainty

A formalism that translates insights from the fields of moral philosophy and machine ethics to that of machine learning is proposed, which is then trained to act under moral uncertainty, highlighting how such uncertainty can help curb extreme behavior from commitment to single theories.

Decolonial AI: Decolonial Theory as Sociotechnical Foresight in Artificial Intelligence

This paper highlights problematic applications that are instances of coloniality, and using a decolonial lens, submit three tactics that can form a declonial field of artificial intelligence: creating a critical technical practice of AI, seeking reverse tutelage and reverse pedagogies, and the renewal of affective and political communities.

AI should embody our values: Investigating journalistic values to inform AI technology design

It is argued HCI design can achieve the strongest possible value alignment by moving beyond merely supporting important values, to truly embodying them.

A narrowing of AI research?

An analysis of the thematic diversity of AI research in arXiv, a widely used pre-prints site, suggests that diversity in AI research has stagnated in recent years, and that AI research involving private sector organisations tends to be less diverse than research in academia.

Implementations in Machine Ethics

A trimorphic taxonomy is introduced to analyze machine ethics implementations with respect to their object (ethical theories), as well as their nontechnical and technical aspects.

AI Ethics and Value Alignment for Nonhuman Animals

This article focuses on the two subproblems—value extraction and value aggregation—discussions challenges for the integration of values of nonhuman animals and explores approaches to how AI systems could address them.



Expected Utilitarianism

It is shown that if one believes that a beneficial AI is an ethical AI, then one is committed to a framework that posits 'benefit' is tantamount to the greatest good for the greatest number, and if the AI relies on RL, then the way it reasons about itself, the environment, and other agents, will be through an act utilitarian morality.

Social choice ethics in artificial intelligence

The normative basis of AI social choice ethics is weak due to the fact that there is no one single aggregate ethical view of society, and the design of social choice AI faces three sets of decisions.

The global landscape of AI ethics guidelines

A detailed analysis of 84 AI ethics reports around the world finds a convergence around core principles but substantial divergence on practical implementation, highlighting the importance of integrating guideline-development efforts with substantive ethical analysis and adequate implementation strategies.

Artificial Intelligence: the global landscape of ethics guidelines

A global convergence emerging around five ethical principles (transparency, justice and fairness, non-maleficence, responsibility and privacy), with substantive divergence in relation to how these principles are interpreted; why they are deemed important; what issue, domain or actors they pertain to; and how they should be implemented.

The Arc of the Moral Universe and Other Essays

In this collection of essays, Joshua Cohen locates ideas about democracy in three far-ranging contexts. First, he explores the relationship between democratic values and history. He then discusses

Justification and Legitimacy*

In this essay I will discuss the relationship between two of the most basic ideas in political and legal philosophy: the justification of the state and state legitimacy. I plainly cannot aspire here

Normative Uncertainty as a Voting Problem

Some philosophers have recently argued that decision-makers ought to take normative uncertainty into account in their decisionmaking. These philosophers argue that, just as it is plausible that we

Value Alignment or Misalignment - What Will Keep Systems Accountable?

It is proposed that a hybrid approach for computational architectures still offers the most promising avenue for machines acting in ethics and critically considers how such an approach can engage the social, norm-infused nature of ethical action.

Social Choice and the Value Alignment Problem *

Principles alone cannot guarantee ethical AI

Significant differences exist between medical practice and AI development that suggest a principled approach may not work in the case of AI, and Brent Mittelstadt highlights these differences.