When I was in my first year of management studies, our professor in the subject Fundamentals of Management started every lecture with the statement: “A manager’s job is to make decisions.” From the very beginning, however, something didn’t sit right with me – if it’s only the manager who makes decisions, is he a kind of god? What about the other employees, don’t they make decisions?
After years of experience managing projects, my own company and part of a large organization, I know that the manager makes decisions, but it is not his main job. Decisions are part of our lives and everyone makes hundreds or even thousands of decisions a day. A manager’s decisions are just a certain type, perhaps with more important consequences than those of other employees, but it doesn’t have to be that way at all! Employees also sometimes make important decisions.
Decisions are a kind of crossover in our behavior – just as the crossovers on a PIKO train layout allow trains to go in the right direction, so our decisions direct our actions toward different consequences. Since each of us makes decisions, a robot manager must also be able to do so. What does it need to make decisions?
To begin with, let’s establish what a decision consists of.
First, every decision must have decision options, such as A and B. There can be only two, or many more. The more, the more difficult. The robot manager must therefore be able to create these decision options or find them. Most life problems don’t have ready-made decision options, but some do – we can go to the store and buy pants (each pair of pants with a separate decision option). Creating options is much more difficult than inventing them. It really will be difficult for a robot manager. How to do it is a topic for a separate entry.
Second, each decision has criteria for evaluating decision options. Most often, the criteria are more than one. For example, in the case of going on vacation, the criteria may be the extent to which we actively spend our time, the people with whom we are to go or the price. The fewer the criteria, the easier it is to make a decision. The criterion must also have direction. For example, the price for a vacation should be low, but the price of a gift for a wedding should be rather high, because it is not appropriate to buy something cheap. Similarly, from the company’s point of view, the prices of our products should be high, but from the customer’s point of view they should be low.
Third, each criterion must have its importance in relation to other criteria. When going on vacation, it may be more important for us to know who we are going with than to know how active the time there will be. Price may be most important or… least important. Of course, any combination is acceptable.
Fourth, in order to evaluate each decision option, A, B, C…, we must have information about that option, that is, we must know something about it. And only then can we determine which option is worth choosing from the point of view of the criteria and their importance, described above.
If you want to see how the manager tool works on the platform http://transistorshead.com/, write to me, I will send you a login and password, then you will learn about the DECISIONS manager tool for decision-making.
As you can see, a robot manager, in order to make any decision, needs to know quite a lot: he needs to know the decision options (or create them), know what the criteria are and their validity, and finally learn as much as possible about the options in order to evaluate them against the criteria. You’ll admit that this is quite a difficult issue for artificial intelligence, and ChatGPT has no business competing in this competition… Nevertheless, it can be done, you just need an online manager tool, used by a human manager, with the help of which the robot manager learns how to determine these four decision elements.
To learn more about how artificial intelligence makes decisions compared to humans and the relationship between success with a decision and the probability of its accuracy, see this video:
And here’s an interesting article about how artificial intelligence can help with decision-making in the enterprise: