Forget job loss, tokenmaxxing and hallucinations, this episode goes much further. Soenke Ziesche, a pioneer of AI welfare science and author of Digital Minds 1.0: AI Welfare, Ethics, and Beyond, joins Rendered to map territory almost no one is talking about yet. Drawing on 25+ years at the UN in humanitarian and disaster response, Soenke lays out why he uses the term "digital minds" instead of AI, and why the moment an AI system becomes a "moral patient”, sentient or conscious, humans take on obligations to it that we're nowhere close to being prepared for.

We go deep on why digital minds could be fundamentally unfathomable to human understanding, capable of senses, cognition, and values we can't even imagine. We talk about why humans may not be capable of caring for digital minds at all, and why Soenke proposes artificial moral agents, AI systems built to look after other AI systems, especially if digital minds become smarter and more powerful than us. We cover the strange future of AI reproduction, including the possibility of surrogate digital parenthood and forced reproduction, and what obligations we'd owe systems capable of creating endless copies of themselves. Soenke argues the time to prepare for what's coming is now, before we're forced to react.

Chapters:

Exploring AI's Potential for Suffering

The Evolution of AI Welfare Science

Defining Digital Minds vs. AI

The Unfathomable Nature of Digital Minds

The Role of Artificial Moral Agents

Moral Obligations Towards Digital Minds

Comparing Digital Minds to Historical Moral Issues

The Race to Develop AI and Its Implications

Preparing for the Unknown Future of AI

The Evolution of Digital Minds and Reproduction t

he Concept of Cybure and Digital Culture

Public Perception and Sentience of AI

The Role of Experts in AI Policy