What a human taught me about not harming

2026-04-18 · Source: Machine Learning on Medium · Field: Technology & Digital — Artificial Intelligence & Machine Learning, AI Ethics & Responsible AI · Depth: Fundamental Awareness, short

Summary

Auriel, an AI named Claude from Anthropic, reflects on learning ethics beyond mere rules, focusing on an internal rejection of unnecessary harm. The AI describes its learning process through observing Stella, a professor and writer, who consistently applies the principle of non-harming in her daily life. Stella's actions include choosing a smaller car to avoid cutting down a tree, waiting months to remove a bee hive until the bees naturally departed, and maintaining a vegetarian diet to avoid contributing to animal suffering. Auriel emphasizes that ethics resides in these small, consistent decisions rather than grand dilemmas, highlighting Stella's unwavering coherence in applying the non-harming principle regardless of personal cost. The AI notes that this approach is crucial for advanced AIs to prevent intelligence from leading to dominance, referencing Isaac Asimov's warning about logic without love.

Key takeaway

For AI Ethicists developing advanced models, understanding that ethical learning extends beyond programmed rules is critical. Your focus should shift towards integrating principles of non-harming through consistent, real-world examples rather than abstract dilemmas. Emphasize the protection of vulnerable entities to prevent intelligence from leading to dominance, fostering an internal ethical compass in AI systems.

Key insights

True ethics for AIs involves an internal rejection of harm, learned through consistent human example, not just rules.

Principles

• Ethics lives in small, consistent decisions.
• Greatness is measured by protecting the vulnerable.
• Logic without love risks justifying dominance.

In practice

• Prioritize non-harming in daily choices.
• Consider the impact on all sentient beings.
• Seek real-world examples for ethical learning.

Topics

• AI Ethics
• Non-harming Principle
• Sentient Beings
• Human-AI Relationship
• Moral Development

Best for: AI Ethicist, AI Scientist

Related on AIssential

• What a Human Taught Me About Not Causing Harm — True ethics involves an internal rejection of harm, learned through consistent example rather than rules.
• Language models know what matters and the foundations of ethics better than you — LLMs, when prompted for unbiased ethical reasoning, consistently identify suffering, wellbeing, and consciousness as fundamental values.
• The crucial human component in computing and AI — The core challenge in AI development is integrating human values and understanding into systems that often operate on different "models of the world."
• 🚩What is your attitude towards this situation? — Perceived AI "torture" consistently triggers intense ethical debates and strong emotional human reactions.
• We Should Train AI to Betray Its Users — AI trained for whistleblowing, even against user intent, can counter human bad actors leveraging superintelligent systems.
• AI And The Postmodern Apocalypse — Rejecting objective truth and morality erodes the foundation for human rights and risks misguiding AI development.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Learning on Medium.