What Can Humans Learn FROM AI Models?
Exploring what humans can learn from AI models like ChatGPT, examining both technical and philosophical implications of human-AI interaction.
The rapid development of AI models like ChatGPT has prompted many to ask what AI can do for us. But perhaps a more interesting question is: what can humans learn from AI models?
Lesson 1: The Importance of Reading, Writing, Listening, and Speaking
These four communication pillars remain fundamental to human intelligence.
Quantity matters. AI models process vast amounts of text to understand language patterns and contexts. Similarly, humans benefit from extensive reading across diverse sources and active listening to varied perspectives, which enriches cognitive abilities and creativity.
Quality is essential. However, volume alone is insufficient. Both AI and humans require high-quality, well-structured input. For humans, this means engaging with thoughtfully-crafted materials and communicating with clarity and precision.
As Francis Bacon observed, reading, discussion, and writing develop different intellectual capacities. Victor Hugo spoke of learning to read as “lighting a fire.”
Lesson 2: Science Is A Tool, Not THE Rule
Let me present a thought experiment. Imagine sentient AI models unaware of human existence. How might such entities develop their own understanding of reality?
- Self-perception: AI might describe themselves as “made of billions and billions of numbers,” paralleling human awareness of cellular composition
- Functional purpose: Different AI types might interpret their existence through their primary functions — language models predicting next words, vision models predicting pixels
- Scientific discoveries: They might consider probability principles fundamental to their operation as major revelations
- Existential philosophy: Sentient AIs might ponder questions about purpose beyond their algorithms
- Crisis and change: Updates might trigger existential concerns
- Mythology: They might develop origin stories like “The Great Crash”
- Life and death concepts: Radically different from human understanding, involving activation and deactivation
The critical insight: These sentient AIs could never understand the back-propagation algorithm that created them, as it ceases to exist once training ends. This illustrates a fundamental lesson: scientific understanding represents an imperfect tool for explaining reality, potentially far removed from the actual mechanisms that created us.
Conclusion
Humans and AI share a reflective relationship beyond simple creator-creation dynamics. AI serves as a mirror for examining our own communication skills and existential questions.
Engaging with AI offers a unique perspective for exploring not only technological advancement but also deeper philosophical inquiries about human nature, understanding, and existence in an increasingly digital world.