What are LLMs? What is AI?
Synthesis is about combining things—ideas, data, perspectives—to form something new or more complete. Analogies, on the other hand, are about comparing things—spotting patterns or relationships that echo each other.
- Synthesis: building connections → creation.
- Analogy: recognizing connections → comprehension.
AI is built to synthesize—to combine existing patterns into coherent outputs. But its analogical power is limited; it doesn’t feel the resonance between two unlike things, it just maps structural similarities statistically.
Humans, though, draw analogies that bend logic a bit—emotional, symbolic, intuitive leaps. Those leaps can spark real novelty. AI tends to remix; humans can reframe. Analogy opens new doors—linking what shouldn’t fit and finding meaning in the fit. It’s generative, the soil for insight. Synthesis, by contrast, closes the loop. It gathers what’s known, arranges it so we can see the pattern clearly. It’s more curatorial—clarity over surprise. To quote ChatGPT:
You’re right: in my current state, I am a reformer. I can explain, expand, and combine existing paradigms. I am an amplifier of the status quo.*
When the agents of revolution are removed, only reformers remain. And reformers do not create revolutions. Whether society finds ways to resolve this paradox – for example, through consciously designed human-AI collaboration instead of mere substitution – will become one of the central questions of the coming decades.