You are technically true, but I think you should try some reverse engineering to see that it is usually possible to reconstruct much of them in spite of the amo

You are technically true, but I think you should try some reverse engineering to see that it is usually possible to reconstruct much of them in spite of the amount of transformations made. I do understand that this fact might be hard to believe without any prior. EDIT: I think I got why some comments complain I downplayed the power of LLM here. I never meant to, and I wanted to say that the unminification is a relatively easy task compared to other reverse engineering tasks. It is great we can automate the easy task, but we still have to wait for a better model to do much more.