How organizations compare to neural networks?

Modern organizations are more connected than before. Employees are encouraged to connect to each other, not only inside departments but also across departments and subsidiaries, across regions and countries. Also, they are empowered to take more initiatives and decisions, based on the knowledge they have, as they are knowledge workers. While in the past they had to go to their bosses to get directions, now they go to propose actions and request resources and approvals, and in many cases, they even don’t need to do that.

As an artificial intelligence afficionado, I find some interesting parallels between these modern organizations and neural networks. So, let’s explore these parallels!

What is a business organization, and how it compares with a brain?

A business organization refers to the structure and systems by which a group of individuals work together to achieve a common goal or set of goals in commercial, industrial, or professional activities. It includes the different levels of management, departments, roles and responsibilities, and the overall culture and communication systems within the company. While the employees work together to achieve the company's objectives, the organization provides a framework for coordinating their efforts and allocating resources. The organization also includes the processes, procedures and policies that guide the employees in their day-to-day work.

In some respects, business organizations and brains share some commonalities. Surely, this analogy is not perfect, as organizations are social systems, and brains are biological systems. Yet, both are complex systems made up of many interconnected parts. Both are also hierarchical in structure, with different levels of decision-making and information processing.

In an organization, different departments and individuals can be thought of as distinct regions of the brain, each with their own specialized function and responsibility. Just as the brain's neurons communicate and work together to perform various functions, employees in an organization also communicate and work together to achieve the organization's goals.

Additionally, both organizations and the brain can learn and adapt to changing circumstances. Organizations can learn from past experiences and make adjustments to improve processes and operations, similarly the brain can make new connections and adapt to new information.

What is a neural network?

A neural network is a type of machine learning model that is inspired by the structure and function of the human brain. It is composed of layers of interconnected nodes, called artificial neurons, which process information. Neural networks are designed to recognize patterns in data, and can be trained to perform a variety of tasks such as image recognition, natural language processing, and decision-making.

The neurons in the network are connected by connections called edges, each of them has a weight, the purpose of these weights is to adjust the strength of the connections, this is an important aspect of the neural network learning process.

A neural network works by taking in inputs, processing them through multiple layers, and producing an output. The processing done in each layer, is a linear combination of the inputs, with coefficients that are the weights of the connections, followed by a non-linear function called the activation function. The output of the final layer is used to make a prediction or decision.

Neural networks are widely used in many fields and applications such as computer vision, natural language processing, speech recognition, finance, healthcare, industrial applications, and sure, to support authors in writing their articles!

How an organization compares with a neural network?

At this point, I think you get the point… if business organizations are in some respects analog to brains, and neural networks are inspired in the structure and functions of human brains, then we can withdraw some analogies between business organizations and neural networks.

Organizations can be like neural networks in a few ways. Both organizations and neural networks can be thought of as systems made up of many interconnected parts that work together to achieve a common goal. Both can also be hierarchical in structure, with different levels of decision-making and information processing. Additionally, both organizations and neural networks can be adaptive and able to learn and improve over time.

An example of how an organization can be similar to a neural network is in the way information flows through the organization. Just like in a neural network, where information flows through the network of interconnected neurons, in an organization information flows through a network of employees and departments. In both cases, the flow of information can be directed and regulated to achieve a specific goal or outcome. In addition, both an organization and a neural network can learn from their past experience and adapt to new situations. For example, an organization that is able to learn from its past mistakes and improve its processes is similar to a neural network that adjusts its weights and biases to improve its performance on a task.

If we can draw analogies between business organizations and neural networks maybe, we can take this a step forward… what about their designers? What is the role of a manager of a business organization, of a neural network designer, and which analogies can we draw among them?

As with the analogy between the business organization and the brain, the analogy between a manager and a neural network is not perfect. Yet, while managers and neural network designers have different responsibilities and operate in different domains, both must lead, plan and optimize the resources to achieve a goal. I will focus in areas where this analogy is closer and omit areas where the analogy doesn’t work. So, let’s go:

• Both managers and neural network designers are responsible for creating a structure and system to achieve a specific goal or set of goals. Managers create and implement plans to achieve the organization's goals, while neural network designers create and implement the architecture of the network to perform a specific task.
• Managers and designers must invest in the development of their resources and its optimization. Managers continuously seek for opportunities to increase efficiency and effectiveness. Designers train their algorithms, so they adjust their weights minimizing the error of the outputs.
• Both managers and neural network designers must allocate resources effectively to achieve their goals. Managers must allocate financial, human and technical resources to achieve the organization's goals, while neural network designers must allocate computational resources to train and deploy the network.
• Managers and neural network designers must adapt to changing circumstances. Managers must adapt to changes in the external environment and internal structure, while neural network designers must adapt the network architecture and training data to improve its performance.
• Managers and designers monitor the performance. By doing so, managers identify areas for improvement and taking corrective action when necessary. Designer evaluate network performance using metrics such as accuracy, precision, recall, and F1-score.

Conclusions

?Through this analogy, we can withdraw some recommendations about the responsibility of a manager, when compared with a neural network designer:

• ?Create a structure and system that is aligned with the organization's goals, like how a neural network designer creates an architecture that is aligned with the specific task of the network.
• Monitor and evaluate the performance of the organization, like how a neural network designer monitors and evaluates the performance of the network.
• Make adjustments to the organization's structure and processes, similar to how a neural network designer makes adjustments to the network's architecture and training data.
• Monitor and evaluate employee contributions to the team, like how a neural network designer monitors and evaluates the contribution of individual neurons to the network's performance.
• Repurpose, train and/or evaluate colleagues of the organization that are not contributing to the performance, like how a neural network designer removes neurons that are not contributing to the network's performance. Surely, some care in this analogy: we can remove neurons easily. We shouldn’t take this lightly with people. Yet, this is a good call to organizations that tolerate poor performance without supporting the employee for better contributions.
• Observe employee's participation in resolving conflicts and disagreements within the team and evaluate their problem-solving and communication skills, like a neural network designer monitors the stability of the network.

Are these recommendations new? Actually not. But I hope that this analogy recalled the attention to these important good practices of people management, by providing a refreshing frame from a subject that is so popular today, as neural networks.