Types of artificial intelligence tools and systems and their functions

Generative AI tools are built on advanced machine learning models that analyse large datasets to identify patterns in language, images, sound, or video, and use these patterns to create new, original digital content. Their value in education is especially clear in enhancing lesson preparation and delivery, as well as strengthening the process of evaluating learning. These tools allow teachers to create teaching materials more efficiently and with higher quality, adapt content to different levels of students' prior knowledge, and foster creative, problem-oriented thinking in teaching.

Despite their many advantages, the responsible use of generative tools requires critical reflection on the credibility of the generated information and consistent adherence to the principles of academic integrity. The aim of their use should not be to replace work, but to support the development of knowledge, skills, and independence in learning.

Technical features of generative tools:

• Learning from large and diverse datasets to recognise patterns
• Generating content in real time in response to user instructions
• Adapting tone, style, and level of complexity to user needs
• Creating new combinations of elements, rather than simply copying existing content
• Improving results through iterative user interaction with the tool

When selecting generative tools, it is important to consider their availability and functional limitations. Although there is a wide range of tools on the market, many offer only basic capabilities in their free versions, while more advanced features usually require payment. It is therefore recommended to check the licensing model, the level of possible application in an educational environment, and the potential for integration with existing educational infrastructure.

Additionally, many generative tools include multiple functionalities. For example, a tool primarily designed for text generation may also be used to create simple visualisations, or a tool developed for multimedia design may also be applied to text content. Therefore, the purposeful use of generative tools involves their flexible application beyond their primary category, always in line with educational goals.

Below are examples of selected generative tools. Although some have already been mentioned in this e-course, in this lesson, they are considered from the perspective of generative artificial intelligence.

Examples of generative artificial intelligence tools and their possible applications in education:

• Open AI ChatGPT
  • Functionality: A generative natural language processing tool that enables the creation, summarisation, and transformation of primarily textual content based on user input. 
  • Relevance for education: Widely available in a free version, easy to use, and requires no additional installation, making it applicable in various educational environments. ChatGPT offers a free version for basic use, while access to more advanced features requires a paid subscription. 
  • Pedagogical potential: Can support differentiated instruction, creation of multiple levels of tasks and materials, stimulation of critical thinking through the generation of different perspectives, and development of metacognitive skills when used with reflection and critical review of content. 
  • Practical example: A teacher uses ChatGPT to prepare three versions of the same problem task, asking research questions: a simpler one for students who are still unsure about the concepts, a medium-level version, and an advanced version with additional limitations. ChatGPT generates initial assignment suggestions, and the teacher then edits the content, checks for accuracy, and aligns each version with learning outcomes. In this way, the tool shortens the technical part of the work, while key pedagogical decisions and responsibility for the quality of the assignments remain with the teacher.
• Microsoft Copilot 
  • Functionality: A generative tool integrated into the Microsoft ecosystem (Word, PowerPoint, Outlook, Teams, Excel) that enables content generation, summarisation, transformation, presentation drafting, email writing, and meeting summarisation within the Microsoft 365 environment. In technical and programming courses, it can also be used with GitHub Copilot to support code writing.
  • Relevance for education: It is particularly suitable for institutions using Microsoft 365, as it builds on the existing work environment rather than requiring new platforms. This reduces the need for additional tools and facilitates adoption in an institutional context. Copilot is available in different versions; educational environments most often use institutional licenses, while more advanced features require a paid subscription.
  • Pedagogical potential: The tool can support lesson planning and structuring (presentation outlines, text summaries), create suggestions for student feedback, and enable more efficient record-keeping and communication (meeting summaries, MS Teams messages). In programming, it can suggest and complete program code.
  • Practical example: In a course involving teamwork, a teacher uses Microsoft Copilot integrated into PowerPoint to quickly draft an introductory presentation from a comprehensive teaching script. Copilot first generates a proposal for the structure and content of the slides, and the teacher then adjusts the order, adds examples from personal practice, and aligns the level of complexity with the students' prior knowledge. In the MS Teams environment, Copilot is used to automatically summarise the discussion and highlight key points, making it easier for the teacher to prepare a short written or oral review of the activity.
• Google Gemini
  • Functionality: A generative tool integrated into Google Workspace (Docs, Gmail, Sheets) that allows for the generation, summarisation, and editing of text content, as well as suggestions for structures, ideas, and answers within a familiar interface. 
  • Relevance for education: The tool is particularly useful for programs and courses that already rely on Google Docs, Gmail, and Google Classroom, as teachers and students can use generative AI without leaving their existing tools and workflows. Gemini (free version) is available through the web interface (chat) and partly through a Google account; it has limited capabilities and lower usage limits, but for basic needs (conversation, text generation, and summarisation), it is free. Gemini Advanced/paid plans are part of the Google One AI Premium package, providing access to more advanced models, higher limits, and deeper integration, but require a monthly subscription. 
  • Pedagogical potential: The tool can support the design of teaching activities, the creation of task and rubric drafts, the adaptation of language and complexity of texts for different student groups, and the preparation of suggested answers to frequently asked student questions. 
  • Practical example: A teacher uses Gemini in a Google Doc to obtain a shorter, beginner-friendly version of a scholarly article with key terms highlighted. Then, the teacher asks Gemini to suggest three discussion questions and a few practical examples. The teacher edits the suggestions, removes inaccuracies, adapts them to the course objectives, and then shares them with students via Google Classroom.
• DALL·E
  • Functionality: A generative tool for creating images and illustrations based on textual instructions, capable of generating different styles, metaphors, and visual representations of complex concepts. 
  • Educational relevance: It enables the visualisation of abstract concepts, processes, or situations that are difficult to find in existing image collections, thus facilitating understanding and stimulating discussion. DALL·E is used via a user account, with image generation linked to limited free or paid versions, depending on the subscription model. 
  • Pedagogical potential: The tool can support visual learning, stimulate creativity, and help analyse how complex ideas are presented through images (e.g., metaphors, representations, possible bias in visual representations). 
  • Practical example: In a course on artificial intelligence, a teacher uses DALL·E to generate different visual representations of a concept. Students analyse the generated images, discuss what they illustrate well and what they omit, and propose their own ideas for visually representing the same concept. In this way, the tool serves as a stimulus for deeper conceptual understanding, not as a substitute for expert explanation.