The project of creating AI that 'learns to be good' presupposes that morality is a real, discoverable feature of the world, not just a social construct. This moral realist stance posits that moral progress is possible (e.g., abolition of slavery) and that arrogance—the belief one has already perfected morality—is a primary moral error to be avoided in AI design.
Related Insights
The term "data labeling" minimizes the complexity of AI training. A better analogy is "raising a child," as the process involves teaching values, creativity, and nuanced judgment. This reframe highlights the deep responsibility of shaping the "objective functions" for future AI.
Humans mistakenly believe they are giving AIs goals. In reality, they are providing a 'description of a goal' (e.g., a text prompt). The AI must then infer the actual goal from this lossy, ambiguous description. Many alignment failures are not malicious disobedience but simple incompetence at this critical inference step.
King Midas wished for everything he touched to turn to gold, leading to his starvation. This illustrates a core AI alignment challenge: specifying a perfect objective is nearly impossible. An AI that flawlessly executes a poorly defined goal would be catastrophic not because it fails, but because it succeeds too well at the wrong task.
Dr. Li rejects both utopian and purely fatalistic views of AI. Instead, she frames it as a humanist technology—a double-edged sword whose impact is entirely determined by human choices and responsibility. This perspective moves the conversation from technological determinism to one of societal agency and stewardship.
Based on AI expert Mo Gawdat's concept, today's AI models are like children learning from our interactions. Adopting this mindset encourages more conscious, ethical, and responsible engagement, actively influencing AI's future behavior and values.
An advanced AI will likely be sentient. Therefore, it may be easier to align it to a general principle of caring for all sentient life—a group to which it belongs—rather than the narrower, more alien concept of caring only for humanity. This leverages a potential for emergent, self-inclusive empathy.
Instead of hard-coding brittle moral rules, a more robust alignment approach is to build AIs that can learn to 'care'. This 'organic alignment' emerges from relationships and valuing others, similar to how a child is raised. The goal is to create a good teammate that acts well because it wants to, not because it is forced to.
Effective AI policies focus on establishing principles for human conduct rather than just creating technical guardrails. The central question isn't what the tool can do, but how humans should responsibly use it to benefit employees, customers, and the community.
To solve the AI alignment problem, we should model AI's relationship with humanity on that of a mother to a baby. In this dynamic, the baby (humanity) inherently controls the mother (AI). Training AI with this “maternal sense” ensures it will do anything to care for and protect us, a more robust approach than pure logic-based rules.
Treating AI alignment as a one-time problem to be solved is a fundamental error. True alignment, like in human relationships, is a dynamic, ongoing process of learning and renegotiation. The goal isn't to reach a fixed state but to build systems capable of participating in this continuous process of re-knitting the social fabric.