Ethics get complicated there too -- if you're a consequentialist, do you go by the actual consequences or the intended consequences? If the former then what is ethical can only be chosen in retrospect so you're trying to predict what will be ethical by trying to predict future outcomes. If it's the latter, then incorrect consequences can justify overall wrong actions as justified even if successful implementation will always result in negative outcomes because you're just wrong.

That's where imo you do need a base of deontological ethics, hard lines you don't cross, because otherwise you can find yourself either never knowing what is ethical until after the fact leading you down a rabbit hole of betting against God predicting the future, or you can break your whole ethical system by biasing what you think the consequences will be unconsciously. You can the consider consequentialist perspective for more complicated ideas once you stop yourself from doing things you'll regret later if you get the consequences wrong or sometimes even if you're right -- if you become an ethical monster you might be right but you still have to live with yourself afterwards.