Ep 261: Objective Morality II

In this episode of ToKCast, the host continues the exploration of objective morality by building on David Deutsch's principle of optimism, arguing that evil arises from a lack of knowledge and can be overcome through problem-solving. The episode unpacks Thomas Sowell's famous claim about 'no solutions, only trade-offs,' clarifying that Sowell uses scare quotes around 'solutions' to emphasize that perfect outcomes are impossible under resource constraints. The discussion then delves into the nature of value, distinguishing subjective valuation (e.g., personal preference for a Ferrari) from objective outcomes (e.g., whether a purchase brings satisfaction). The core argument is that morality is objective not because it derives from facts alone, but because moral claims—what we should do—are conjectures that can be better or worse based on whether they solve problems and preserve error correction. The host introduces David Deutsch’s moral imperative: 'do not destroy the means of error correction'—a foundational, non-dogmatic principle that unifies epistemology and ethics. This idea is presented as essential for progress, with implications for AI alignment, where hardcoding morality is seen as a fatal error that destroys the very possibility of future improvement. The episode critiques utilitarianism, deontology, and virtue ethics as useful but fallible frameworks, not ultimate foundations, and refutes the orthogonality thesis by asserting that morality and science are logically intertwined. Ultimately, the episode defends a non-foundationalist, Popperian view of knowledge and morality, where progress is possible because better and worse explanations exist objectively, even if we are fallible in identifying them.