Introduction to Generative AI

Different forms of AI

Artificial Intelligence is something that has been around for a long time and is used in a variety of ways. However, the area is only in its infancy and will be developed more and more. In fiction and popular culture, AI and its capabilities have been described in several different ways, often as some form of "technological evil". However, this picture does not correspond well with the capacity of today's technology. Roughly speaking, different forms of AI can be divided into three levels.

Narrow AI

Narrow AI, also known as weak AI, is the type of AI that we see in use today. It is designed to perform specific tasks or solve particular problems without human intervention. Examples of narrow AI include chatbots, recommendation systems like the ones you'll find on Netflix or Spotify, and image recognition technology. These systems are "narrow" in the sense that their capabilities are limited to the task for which they were programmed, and they lack the ability to apply their intelligence beyond these specific areas.

General AI

General AI or general AI is a theoretical form of AI that has not yet been achieved. It refers to systems that can perform intellectual tasks that today only a human can, i.e. critical thinking and learning from experience and applying knowledge in new situations.

Superintelligence

Superintelligence goes a step further than general AI and refers to a hypothetical AI that not only matches, but surpasses human intelligence in all aspects, including creativity, general knowledge, and social skills. A superintelligent AI would have the ability to improve itself autonomously, leading to a rapid and unpredictable development of its intelligence and capabilities. The idea of superintelligence evokes both fascination and concern, with discussions surrounding its potential impact on society and how to ensure that its goals are in tune with human values and security.

In short, narrow AI is what we have today, general AI is a target for the future, and superintelligence is an even more advanced and speculative possibility.

Artificial intelligence

Artificial intelligence is based on several scientific disciplines, including computer science, psychology, linguistics, and cognitive science. The central goal is to create systems that can perform tasks that require human intelligence, such as learning, interpreting language, recognizing patterns, and solving problems. These abilities are built on a few basic principles and techniques:

Machine learning

Machine learning is at the heart of many AI systems and enables machines to learn from data, instead of being explicitly programmed to perform a specific task. This means that an AI system can be improved over time, based on new data and experiences. In academic teaching, this can be used, for example, to develop new, more personalized learning platforms or digital assistants that adapt to each student's learning style and progress.

Deep learning

Deep learning is a specific branch of machine learning that uses multi-layered neural networks (deep networks) to perform complex tasks. Deep learning is particularly powerful for processing and interpreting large amounts of unstructured data, such as text, images, and audio. For educators, this offers the opportunity to integrate sophisticated analytical tools into research or to use advanced tools to assess student work by identifying key concepts and themes.

Natural Language Processing (NLP)

Natural language processing (NLP) is an advanced field of artificial intelligence that focuses on the interaction between computers and human language. It aims to make computers capable of understanding, interpreting, and generating human language in a way that is both meaningful and efficient. By combining techniques from linguistics and computer science, NLP systems can analyze language structures, understand contexts, and interpret user intent.

The core tasks of NLP include text analysis, understanding the semantics of language, and natural language generation. This data enables applications such as automatic summaries of texts, language-based interfaces, and language learning tools. One of NLP's biggest challenges is dealing with the inherent complexity of language, including irony, slang, and different contexts. To address this, NLP uses advanced algorithms and large amounts of text data to learn and understand the nuances of language.

Cognitive simulation

This specialization of AI focuses on modeling and simulating human thought process and cognition. In an academic context, this can be applied to create virtual labs where students can experiment with different scenarios and see the results of their hypotheses in real time, without risks or high costs.

What is GenAI?

GenAI is an evolved form of artificial intelligence that differs from traditional AI systems through its ability to create new and unique content. This can include anything from texts and images to music and video material. For you as a teacher, this offers an opportunity to explore new ways of imparting knowledge and engaging students, while raising questions about how knowledge in the future should be measured and valued.

Large Language Models: The Heart of GenAI

A crucial component of generative AI is large language models, such as GPT (Generative Pre-trained Transformer). These models have been trained on rich text data to understand and generate natural language. Their ability to produce coherent and relevant text makes them invaluable tools for creating teaching materials or stimulating discussions in the classroom, for example.

GenAI in Practice

By understanding how GenAI works – its learning process based on deep learning and neural networks – you as a teacher can begin to explore how this technology can be applied in your own field of study. Prompts, or the instructions you give the AI, play a critical role in the quality of the generated content, opening up a variety of educational applications.

Challenges and opportunities

As with any new technology, there are challenges, such as questions around data quality, bias, and ethical considerations. For you as a higher education teacher, it is important to navigate these questions consciously, while also exploring GenAI's potential to enrich teaching and stimulate critical thinking among your students.

GenAI represents an exciting future in education and learning. By increasing your understanding of this technology, its possibilities and limitations, you as an academic teacher can take the lead in integrating GenAI into your teaching in a meaningful and responsible way.