Can someone tell me how ChatGPT and Claude cannot get reverse engineered?

Can someone with more wrinkles in their brain tell me how the closed weight models will not be reverse engineered as AI becomes smarter? I am not a ML engineer so explain plz? By studying an output, it can analyze the unknown pattern of itself and reverse engineer the weight of the model. I know this is some kind of distillation and AI is a black box that you can’t replicate exactly, but if AI gets smarter, wouldn’t a distillation also be similar in capability as well? And with Chinese models fo

CryptoAlerta — análise de criptomoedas e mercado em tempo real