The Potential of Using Artificial Intelligence (AI) as an Accelerator for Social and Economic Development

In recent years, there has been remarkable progress in developing and deploying AI across all sectors of society and the economy, and the application of AI is increasingly permeating all areas of social, economic, and defense activities in countries around the world. Although the term "artificial intelligence" has caused considerable debate internationally, it is clear that the solutions created under this name have become a powerful tool to support human thinking and cognitive activities.

It has already been proven that AI creates the following advantages for any purposeful activity:

• Reducing the time required to achieve the goal
• Avoiding errors arising from human thinking biases
• Improving transparency
• Enhancing the rationale for conclusions and decisions
• Ensuring coherence between multiple decisions being made
• Enabling rapid monitoring of decision implementation without lateral effects
• Automating and making safer processes that humans repeatedly perform or that are hazardous to health
• Creating and effectively utilizing a knowledge base from the results of previously created human knowledge and experience.

‍

AI can never completely replace human cognitive activities, but it has the ability to substitute or support many complex cognitive operations that are essential for creating knowledge and solutions. In particular, AI provides the ability to quickly, inexpensively, and without errors perform cognitive operations that are extremely time-consuming and highly dependent on individual thinking patterns and perspectives, such as searching, generating, collecting, classifying, comparing, processing information, and drawing conclusions.

‍

Numerous international studies have shown that the use of AI can reduce the cost of goal-driven activities by 30-50%, increase productivity by up to 40%, and increase customer satisfaction by 15%. These advantages can be created across all sectors of society. These advantages explain why public and private sector organizations in many countries are interested in using AI as widely as possible.

‍

While it is undoubted that AI creates these advantages, it is essential to establish the necessary environment and resources to create these advantages and develop appropriate infrastructures. Failure to prepare these conditions will not only prevent AI from creating its advantages but will also result in a tremendous waste of resources. Well, proven international practice suggests that it is appropriate to plan and coordinate the following activities to create the prerequisites:

• Clarify the concept of AI and the consequences of its use
• Establish the necessary workflow of the activities of the user organizations to effectively use AI for any goal-driven activity
• Develop official definitions for commonly used core concepts (work, goal, plan, law, regulation, duty, etc.) to be understood in the same meaning
• Develop measurement units for planning and monitoring activities
• Develop commonly shared quality standards for data/information used for drawing conclusions and making decisions
• Develop a unified model of processes and activities that stakeholders perform to achieve a commonly shared goal
• Digitize and automate the planning and monitoring of goal-driven activities
• Develop intellectual solutions to generate, store, integrate, and process high-quality data to support AI solutions
• Train personnel with the necessary skills and expertise.

‍

Clarifying the Concept of Artificial Intelligence

The prevailing view that AI can perform thinking, reasoning, and other operations without human involvement based on pre-programmed algorithms, generating all the knowledge needed by humans, limits the possibility of effective and accessible usage of AI. In other words, the view is that humans do not need to think because AI is capable of thinking and making the necessary decisions on their behalf. Unfortunately, we still do not fully understand the cognitive operations that make up human thinking and reasoning, nor do we have sufficient understanding of the roles these operations play in implementing complex and broad-ranging goal-driven activities. Therefore, rejecting this broad view and instead considering AI as a tool that creates advantages such as saving time, increasing accessibility, and assisting with cognitive and physical operations that are time-consuming, costly, and hazardous to health for humans to perform themselves would be more appropriate. An approach based on this understanding would create the conditions for developing and using AI solutions that are accessible and effective across all sectors of human society.

‍

Improving the Goal-Driven Activities

AI solutions create advantages in supporting human activities by performing cognitive operations that can recognize certain characteristics of a goal-driven activity. Although, if that activity is disorderly, the usefulness of the AI-based solution is greatly reduced. Therefore, it is fundamentally important to bring these activities into an organized state. Especially, the cognitive operations performed by humans are often disorganized and iterative in nature. However, these operations are not directly visible or perceptible to the human eye but rather occur in an extremely complex process between living cells and neurons. For this reason, it is difficult to fully monitor, diagnose, and control human activities using AI, and the conclusions and recommendations produced by AI tend to have only a probabilistic character, similar to a doctor saying, "Your lungs are 80% likely to be inflamed."

‍

Another factor contributing to this weakness is poor management and lack of internal coherence in goal-driven activities. The more disorganized the goal-driven activity is, the more limited the ability to recognize its properties in detail. Therefore, in order to enable diagnosis, digitization, and the more accessible and effective use of AI solutions, it is necessary to bring goal-driven activity into an orderly state. Although this is not easy, humanity is investing tremendous resources into creating a generally accepted model.

This approach is evident by adopting practices to standardize the dominant properties of organizational governance or operational activities in many countries. For example, the international governance standard ISO 37000 is being used to organize the activities of all types of organizations. By consistently implementing standards like ISO 37000 for governance in the activities of multiple interrelated organizations, including government agencies, it is entirely possible to digitize and make their activities coherent with each other, thereby increasing the scope and variety of AI solutions that can be used.

‍

Defining and Aligning Core Concepts

As AI solutions perform certain cognitive operations using specific signals, symbols, numbers, letters, words, concepts, and phrases created by humans, the performance, quality, and results of these operations directly depend on the properties of the elements used to perform them. For example, since the meaning of commonly used words or concepts can be general and subjective, AI solutions cannot understand, translate, or compare them in a single way, and thus, the output of the AI solution will be based on a probability estimate with considerable uncertainties. One way to overcome this weakness is to align the meanings of core concepts commonly used within an organization or industry. This alignment would enable AI to be used more effectively in cross-sectoral or large-scale coherent activities.

‍

Developing and Improving Activity Measurement Units

A key advantage of using AI-based solutions is the ability to quickly and accurately measure, evaluate, and draw conclusions about specific properties of something or a phenomenon without human involvement over a short period of time. Unfortunately, in most cases, people neglect to develop a solution for translating the properties they want to evaluate and draw conclusions about into numerical values. To perform these evaluation and conclusion operations, it is essential to develop measurement units for measuring the state of the properties. Without appropriate measurement units, it is impossible to generate quality data, and without quality data, it is impossible to draw quality conclusions and make quality decisions, furthermore forcing decisions to be made based on assumptions and biases. Therefore, it is essential to precisely identify the properties of social relationships and development that need to be measured and develop interrelated measurement units essential for translating these properties into numerical values. While developing such units is a highly complex task, it can be solved using systems engineering methods.

‍

Developing Common Quality Standards for Data/Information Used for Drawing Conclusions and Making Decisions

Since AI solutions generate new data and process existing data, defining and adhering to common requirements for data quality will create a fundamental advantage in deploying and effectively using AI solutions more widely. Working with incomplete or non-harmonized data causes AI to make overly general probability-based conclusions and explanations. While several international organizations have defined quality criteria for data, unfortunately, there are no universally accepted criteria. One factor hindering the development of such criteria is the chaotic notions and assumptions in our minds about the concept of data itself. Such chaotic notions also adversely affect the ability to bring cognitive processes into an orderly state. This situation can be resolved by modeling cognitive processes down to the level of operations.

Developing a unified model of processes and activities to achieve a commonly shared goal of

organizations that wish to use advanced AI solutions in their operations must necessarily model their activities into a specific pattern. Such modeling creates many advantages, such as directing participants towards a common goal, establishing alignment, creating a quality data repository, enabling diagnosis and feedback, and enabling sustainable improvement. On the other hand, the elements that make up a goal-driven activity constantly change, limiting the ability to model activities. However, the rapidly developing field of systems engineering in recent years provides an opportunity to overcome this difficulty, with the axiom-based system model playing a crucial role in achieving this goal. Researchers at TUSSolution have developed an axiom-based system model of goal-driven activities, and the results of using this model in numerous studies, analyses, and development efforts suggest that it is possible to model goal-driven activities in detail.

‍

To create the conditions for more effective use of advanced AI solutions in social and economic development activities, the following measures are recommended:  

‍

• Develop and enforce official shared definitions of common pillar concepts
• Develop and enforce standard criteria and metrics for measuring data quality commonly used in planning and monitoring activities
• Develop a unified governance model for goal-driven activities based on the ISO 37000 standard using social systems engineering methods
• Execute the governance model 
• Develop or select a platform to digitize the developed model.

Implementing these steps enables AI to offer more tailored ideas for concluding, developing solution concepts, and modeling the entire goal-driven activity.

‍