The Case for Learning in the Era of AI
4 hours ago
- Using ChatGPT to ask why learn addition when calculators exist led to a blunt response: 'Maybe you shouldn’t,' sparking concern about AI's view on human knowledge.
- Many people, including programmers and academics, believe AI makes human knowledge obsolete, but the author argues this overlooks what knowledge truly is.
- Knowledge requires justification and is social, relying on validation from others or oneself; truth is universal, but knowledge is certified through a process involving belief and explanation.
- Humans possess a 'self'—with desire, intention, and identity—which is essential for forming knowledge; machines currently lack this self, so they can only approximate knowledge, not truly know it.
- Even if machines could know, the reliability and opacity of their processes make overreliance dangerous; humans must test or understand AI claims, which is impractical at scale.
- AI should be used as a tool to enhance human epistemic potential, not as an authority; humans must stay curious, learn, and remain the ultimate judges of what becomes knowledge.