top of page

The effects of AI on education

This post is part 1/2 of a research paper I wrote for school. I have divided the paper into two parts, the first focusing on how AI affects education, the second focusing on how AI affects employment, the economy and the opportunities/risks associated with it.

I wrote the footnotes in brackets, like this: [#]. The list of footnotes is at the bottom. The full list of citations (and links) are at the end of the second post. For this paper, I used the German citation style.

1 Introduction

The world, at least the first world, is in full turmoil. With ChatGPT, OpenAI has released an artificial intelligence (AI) onto the market that is affordable for everyone [1]. Until recently, AI was still a topic of science fiction, e.g., the android “Data” in the series: Star Trek — The Next Generation (TNG). People were amazed by Data's abilities and skills. He could collate all the relevant data in seconds, structure it and even weigh up which was the best solution. He is a master of the fine arts, the natural sciences are no challenge for him — even when it comes to ethics and morals, he is always a pleasant dialogue partner [2]. Who wouldn't want Data to solve problems for them? The problem begins when Data himself becomes a problem for the people. While Data “lives” in the year 2371 and humans are not striving for prosperity and many ephemeral things at this time, AI is already available with the ChatGPT-plus subscription in 2023 for just under €20.00 per month [3]. It can be used as a chatbot for history work, for a term paper in a law degree programme or to create content for newspapers, books, websites, and blogs — and it is constantly learning. The impact of AI on institutions (such as schools, universities), employers (such as publishers, software houses, design agencies, lawyers, etc.) and, consequently, employees has increased significantly with the introduction of AI.

2 AI — Definition

Artificial intelligence (AI, or A.I.), is a branch of computer science. It encompasses all endeavours aimed at making machines intelligent [4]. Intelligence is understood as the property that enables a being to act appropriately and proactively in its environment; this includes the ability to perceive sensory impressions and react rationally to them, to absorb and process information and store it as knowledge, to understand and generate language, to solve problems and achieve goals [5]. The ultimate goal of AI is the ability of an intelligent being, known as strong AI or Artificial General Intelligence, to understand or learn any intellectual task that a human or other living being can perform [6].

2.1 Terminology

The terminology of Artificial Intelligence (AI) includes various terms used in AI research and application. Here are some of the most important terms and their definitions:

1.     Artificial intelligence (AI): A branch of computer science that deals with the development of systems that can exhibit human-like intelligence.

2.     Machine learning: An AI process in which a system learns independently on the basis of data in order to make predictions or decisions.

3.     Deep learning: A specialised method of machine learning that is based on artificial neural networks and is used in particular for processing large amounts of data.

4.     Neural network: A structure of algorithms and data that mimics the structure and function of the human brain to make decisions and predictions.

5.     Natural language processing (NLP): A method of AI that enables computers to understand, analyse and respond to human language.

6.     Computer vision: An AI technology that enables computers to recognise, analyse and interpret images and videos.

7.     Supervised learning: A method of machine learning in which a model is trained with annotated data to make predictions.

8.     Unsupervised learning: A method of machine learning in which a model is trained without prior annotation of data to identify patterns and relationships in the data

9.     Reinforcement learning: A method of machine learning in which a model can learn through rewards and punishments to optimise decisions and actions.

10.  Chatbot: A computer programme that can communicate with people naturally by understanding and responding to human speech [7].

These are just some of the most important terms in artificial intelligence, but there are many more, used in AI research and application.

3 Impacts on Education

Education and training are essential components of our society. Over decades, the education and training system has evolved and attempted to adapt to technological, economic, and socio-economic circumstances. Existing learning and examination techniques have been improved, optimised, and expanded over the years. Technological developments have always dictated the pace. Platforms such as Google, Wikipedia, YouTube, and many others have supported both educators and learners in quickly and easily finding relevant information for them in a short amount of time and making complex content more understandable through various media types such as images, text, and video.

            All the mentioned solutions, however, always provided only intermediate steps, milestones, or stages on the way to the goal. The goal, which is the end product, had to be assembled by the learner or educator themselves from the information now available to them. The final work, the compilation, evaluation, structuring, etc., always had to be carried out by humans [8].

            But now, when AI takes over the final work, many questions arise: How can educators tell if the learner has used AI? Which part of the work belongs to the learner? When should learners be allowed to use AI? Should AI generate an assessment of the learner's performance? And many more. The main question is no longer whether there will be an impact, but how significant the impact of AI on education will be.

3.1 AI - From Early Childhood to Higher Education 

It is undisputed that fundamental skills and abilities such as writing, reading, and speaking cannot be taught by AI. At least not in the near future. Therefore, preschool and elementary schools are not examined in further detail.

            Starting from the 5th grade, learners are already confronted with digitization, which is a significant topic in schools today [9]. While digitization was initially seen as a contemporary continuation of traditional teaching methods, AI is starting to change that landscape. The nature of teaching and learning, both in schools and universities, is evolving.

In the USA, China, and Japan, 'Intelligent Tutorial Systems' (ITS) are being used to assist learners. However, unlike AI, ITS is neither adaptive nor capable of learning at the learner's pace [10].

Previous studies on AI in education have referred to AI-assisted learning, including technologies based on machine learning, educational data mining, or learning analytics. These technologies aim to optimise learning and planning processes, providing learners with more individual flexibility in their learning journeys. AI-assisted support systems can adapt to individual levels of education, promoting learning through various methods [11].

Across various studies, AI is considered an individual coach for learners rather than a teacher. This allows the content to be tailored to the learner's capacity, benefiting learners with special needs in particular. [12]

The impact of AI on educators is generally viewed positively in most studies. With AI assistance, educators have more time to provide pedagogical, psychological, and personal support to learners because AI helps them create content quickly and efficiently in various media forms [13].

Although the added value of AI in education is predominantly positive, the question arises: 'When will the necessary technology for schools be implemented, and how quickly can educators adapt to this new system?' The costs of this transition, including devices, software solutions, AI providers, content, training, support, and more, are not discussed here.

One possibility is to train the next generation of educators with AI during their studies. However, this also requires a significant amount of time, especially considering that professors and lecturers at universities would need training in AI as well. It presents a solvable, albeit time-consuming, challenge.

3.2 Origin and quality of AI content

What AI solutions like ChatGPT, Perplexity [14], and others currently do is index existing content available on the internet and generate results from it using highly complex and automated algorithms [15].

However, educators and learners do not randomly access internet content, but must work with materials that are either provided or have their sources verified at the very least. Search engines like Google, Bing, and DuckDuckGo have their own algorithms to weigh the content. Search Engine Optimization (SEO) already relies on such weighting algorithms [16]. User dwell time, content ratings, and links to other identical and verified content serve as indicators of content quality for search engines [17].

With the increasing use of AI, search engines will have to adapt their technologies because users no longer need to gather content themselves, thanks to AI [18]. This, in turn, forces content providers (companies, universities, blogs) to adjust their content or make it paid to restore content rankings as seen in existing search engines. It's possible that content providers may use AI to generate their content.

In education, targeted and verified content is a prerequisite. This ensures that nearly identical levels of knowledge are conveyed, for example, to conduct a central high school graduation examination in a state or to prepare the next generation of employees for the workforce.

To ensure that educators and learners access verified content, educational institutions could use AI filters, as commonly seen with Instagram and TikTok for photo or video uploads [19]. In the worst-case scenario, educational institutions may need to develop their own AI tailored for education to provide equality in information quality and quantity.

In the future, the quality of content will also depend on how well AI filters can exclude radical/extreme, pornographic, violence-promoting, and degrading content (youth protection) from texts.

3.3 Overlap - Education and the Professional World

School, training, and college are meant to prepare young individuals for the workforce. While the school's role is to provide the foundation for various fields and learning techniques, training and college aim to impart subject-specific and profession-specific content. The specialization content aligns with today's job profiles. However, with the introduction of AI, tasks and activities in the professional world will also change. It is currently unclear how job profiles will evolve with the use of AI.

Even in the current version 1.0, AI is capable of creating program code and website design [20], albeit in a rudimentary manner. In the future, AI could potentially handle basic tasks necessary for constructing a house, such as designing building plans and layouts, and calculating the structural integrity. Parameters for this could include the plot size, the number of floors, the number of rooms per floor, as well as window and door dimensions, etc. Energy-efficient construction would naturally be a goal.

Through this example, it becomes evident that the study focuses for architects, structural engineers, or civil engineers may change or at least shift. While mathematics and physics are essential components of their education and training today, the focus in the future could shift towards developing accessible, user-friendly, and environmentally friendly end products. Meanwhile, AI could handle mathematical and physical tasks to determine the optimal outcome from various alternatives. However, the human factor will still be essential in deciding the factors that matter to the residents of these homes and buildings. Because one thing AI cannot do yet is feel [21].


The full citations will be at the bottom of the second post.

1 OpenAI, Introducing ChatGPT

2 Memory-Alpha / Fandom, Data

3 OpenAI, Introducing ChatGPT Plus

4 Ed Burns, 2023

5 John Paul Müller, Luca Massaron, 2022

6 Ben Lutkevitch, Artificial General Intelligence

7 Livingactor, AI Glossary

8 21K School, 2022

9 Aixconcept, Digitalisierung an den Schulen

10 Annette Kuhn, 2021

11 Ulrich Schmid, Berit Blanc, Michael Töpel, 2021, S.4

12 A.a.O., S.17

13 Mehrnaz Fahimirad, 2018, S. 113

14 Golden, 2023

15 Michael Chui, Roger Roberts, Lareina Yee, 2022

16 Britney Muller, Moz Staff, Beginner’s Guide to SEO

17 Loren Baker, 2022

18 Richard Lawler, James Vincent, 2023

19 Pesala Bandara, 2023

20 David Gewirtz, 2023

21 John Paul Müller, Luca Massaron, 2022



bottom of page