We are clever enough to create artificial intelligence. Now we are trying to work out how much we can trust it, what we can safely leave to it, and whether we are even ready for what we have created.
We are increasingly handing tasks over to it because it saves us time. That is perfectly understandable. But it is easy to cross the line between using it to help us understand something and accepting its answer because we do not have time to check it.
Sometimes we do not even notice when this happens. We ask it to summarise a document we have not read. Then we make a decision based on that summary. When someone asks us why, we can repeat the explanation we were given, but we do not know what was left out.
I see no reason to pretend we do not need tools like these. We use them because they can help us work better. But an answer we receive in a few seconds sometimes deserves more time than it took us to ask the question.
Because when a decision affects someone's work, health, or life, it is not enough to say that this is what the screen said.
Comments