The recent developments in the field of artificial intelligence (AI) are pushing the boundaries of what AI systems can achieve, particularly in terms of their cognitive abilities and ethical considerations. There is a growing emphasis on integrating metacognitive capabilities into AI systems to enhance their robustness, explainability, cooperation, and safety. This shift towards 'wise AI' aims to create systems that not only perform tasks intelligently but also navigate complex, real-world situations thoughtfully, accounting for diverse human values and avoiding harmful actions. Additionally, the ethical implications of AI, particularly in healthcare and disability justice, are being foregrounded, with a focus on inclusive design and the intrinsic motivation for kindness in AI systems. The intersection of AI with other emerging technologies, such as Web3, is also being explored, highlighting potential synergies and gaps that need to be addressed. Overall, the field is moving towards more holistic and human-centered approaches to AI development, balancing technological advancement with ethical responsibility.
Integrating Wisdom and Ethics in AI Development
Sources
Clean for Haskell Programmers
Visualizing the Evaluation of Functional Programs for Debugging
Debugging Functional Programs by Interpretation
Taking AI Welfare Seriously
Introduction to AI Safety, Ethics, and Society
Imagining Better AI-Enabled Healthcare Futures: The Case for Care by Design
AI Should Challenge, Not Obey
Imagining and building wise machines: The centrality of AI metacognition
Using Assurance Cases to Guide Verification and Validation of Research Software
Towards evaluations-based safety cases for AI scheming
AI Horizon Scanning -- White Paper p3395, IEEE-SA. Part III: Technology Watch: a selection of key developments, emerging technologies, and industry trends in Artificial Intelligence
Watson: A Cognitive Observability Framework for the Reasoning of Foundation Model-Powered Agents
How do practitioners gain confidence in assurance cases?
Disability data futures: Achievable imaginaries for AI and disability data justice
We Urgently Need Intrinsically Kind Machines
Combining Theory of Mind and Kindness for Self-Supervised Human-AI Alignment
Intersections of Web3 and AI -- View in 2024
Clinicians' Voice: Fundamental Considerations for XAI in Healthcare