Artificial intelligence and the science of decisions

You may not be aware or do so unconsciously, but if you work with him, you also work in the science space.

Imagine this: you have made a model that can get support tickets and classify them into different subjects and feelings. With this you can prioritize the support tickets from the critical they are and are directed to the right support team. Sounds very good? But is it really so simple? No. With the model he in the country we are really only half the way to the finish line. If you have decided to make a one like the one I just described, you must have had the intention of optimizing the flow of support tickets. Either for the happiest customers or to reduce costs or perhaps any other business target. However, the way we choose to act on the data we get as a result of it is just as important for the actual one, if not more. When we take an attitude of how to act on the data we get we actually make a pattern. The science that enters these models is not as simple as it may sound. Look at this example:

The support ticket he suggests that at 60% like a new ticket is about completion, 30% about a new feature and average critical marks in feeling. Now it doesn’t look that easy? How do we handle this information? Who should take this ticket? And isn’t it a critical interruption no matter the outcome of the feeling?

I’m not in any way a decisions scientist and I can’t teach anyone here. But what I can tell you for sure is that the decision models on top of it are often left to be a secondary advantage without conscience or strategic approach. And worse still – the decision model is just discussed subsequently We have finished with the models of him. I would argue that it is in making the model of the decision that we actually understand which data we really need, so making it first rarely makes sense as we do not know what we really need.

There are also many traps to be aware of decision -making such as the prejudice of survival (thinking that you have made the right decision because you have made the right result) and many of us think we are better decision makers than we are actually.

If you want to learn more about the decision science, my best advice is to follow the lead scientist on Google Cassie Kozyrkov. It really manages to make the science of decisions at an intelligible level.

So to summarize. If we want to have better results with our solutions, we need to pay more attention to making decisions and in many cases start with it before we go to modeling.