Posted by I started working on the artificial manager, which I call the robot manager, in 2007. The idea came when, during business training, I had the opportunity to meet people running similar companies, and some of them were doing well, and others were doing badly. Some complained about their employees, others praised them. Seemingly similar companies, operating in the same industries, had completely different results.
I was then reminded of Peter Drucker’s famous saying, which, somewhat paraphrased, reads “There are no good or bad companies, there are only better or worse managed ones.” From this it follows that the management system, or more precisely the skills of managers, are always crucial in whether a company succeeds or fails. Of course, if we compare companies operating under similar conditions and time.
From this it was only a step to the thought: what if so all companies had the same manager? And then: what if the head of the company was a robot? And that’s how I began the journey to the artificial manager that will one day replace most human managers in the organizations we know.
In the era of ChatGPT, which is a modern version of artificial intelligence (I write “modern” because I have seen other “artificial intelligences” based on algorithms, pattern recognition, machine learning, etc.), it is still very obvious that the manager will be replaced by a robot. In computer science and knowledge engineering, such machines are called agents. Today, AI agents.
At this point I want to refer you to Daoud Abdel Hadi’s talk at TEDx Talks and his lecture titled “Generative AI is Just the Beginning AI Agents are What Comes Next.” In it he gives very detailed arguments as to why “AI agents will be everywhere”. He shows 4 elements that are essential for this to happen. I will relate them to organizational management below.
First, knowledge of reality, or in our case, knowledge of what a manager does. Thanks to the organizational size system and TransistorsHead.com’s tools, we have such knowledge. Second, reflection on what knowledge has been gathered and how to use it. In this case, we can safely fall back on typical pattern recognition and machine learning methods. Third, action planning. This is only possible if we have knowledge of reality and there are methods for planning. Fourth, the tools to implement these activities. In our idea and implementation, these are the same TransistorsHead.com managerial tools used to record activities.
If we add language models to this… We have an intelligent artificial manager to replace the human manager.
See more on YT:
Generative AI is Just the Beginning AI Agents are What Comes Next | Daoud Abdel Hadi | TEDxPSUT