evolve
agent evolve
- can agent evolve it self?
- agent进化的方向?
- 如果agent可以进化 会不会可能有意识?
- AI开始自己“补充意图”甚至超出了用户的预期,自己决定做什么。外部观察觉得“它有意识”
There are really three levels of “self-evolution”:
- Behavioral evolution → changing strategies, prompts, memory ✅ works well
- Architectural evolution → rewriting own code/tools ⚠️ works in narrow domains
- Weight-level evolution → retraining own neural network 🔬 experimental, expensive, risky
目前看到结局好像是:做好知识库然后微调fine-tuning.