Can I guess that the machine’s makers were intelligent, without guessing their motivations?
We can guess optimization, but I’d avoid unconsciously assuming it wasn’t built by an unintelligent optimizer, such as some weird alien evolutionary process or non-intelligent creature/hive, without more extra-terrestrial data.
Otus, I’d pick your first option. I don’t think the safety of (2%, 2%) is worth the 0.5% expected value loss. If the risks were huge, say (-20%, 30%) and (2%, 2%) I’d probably play safe.