Definitions for AI are various and contested.

Instead of delving into these debates, we’ll provide a few proposed definitions and dive into some of the key underlying concepts to give you a framework to understand the rest of the module.

  • One way to understand artificial intelligence is as the imitation of human intelligence by machines. As Sheikh et al note, this is relatively restrictive as it rules out most contemporary applications which are focused on more simplistic tasks, which don’t cover the full breadth of human intelligence (Sheikh et al., 2023).

  • Another way to define AI is to consider what people working on AI are doing when they build AI models. IBM offer the following definition: “artificial intelligence is a field, which combines computer science and robust datasets, to enable problem-solving.” (IBM, n.d.)

  • The UK Government White Paper on AI defines it by reference to the two capabilities which require a regulatory response:

    Adaptivity: Once trained to infer patterns in datasets, some AI models can perform new forms of inference, which they were not explicitly designed for.

    Autonomy: Some models can make decisions without the explicit interference of a human (UKGOV, 2023).

Central to understanding an AI model as something capable of solving problems, or imitating human problem-solving ability, is a basic grasp of algorithms and machine learning.

In simple terms, an algorithm is a set of rules for converting an input to an output – akin to a recipe. It provides a finite procedure to solve a particular problem given input X, the algorithm, based on its coded rules, produces output Y (Prabhu, 2023). In computer programming, the input might be a person clicking a specific tab, the system will follow a set of procedures coded in the algorithm, and then the correct output (opening a certain webpage) will occur. Algorithms are central to how computer programs function. But what makes AI different? How can it adapt to infer things that it isn’t explicitly programmed to?

A lot of what people are talking about when they discuss AI is machine learning. Machine learning is a subset of AI where data and algorithms are used to create systems which imitate the human learning process. We’re able to identify patterns in information presented to us and make predictions based on that information. Similarly, machine learning systems identify patterns in data, and use those patterns to make predictions. Before we dive into some use cases for AI in international development, it will be useful to understand some of the different forms this process can take.

Continue to the next part of the module to discover more about different types of machine learning.