We gave a talk in which we focused on one part of AI, namely large language models (LLM) and the ChatGPT tool and its use in education. It can be a great help for students when writing papers, doing homework, or creating projects. Teachers, on the other hand, will appreciate the creativity of ChatGPT when setting up tests, assigning homework, or creating presentations for teaching. However, it should be kept in mind that ChatGPT should only be used as an auxiliary tool, for example, to create a syllabus, but not as a default resource. Why?
- isn’t always right,
- he likes to make things up, even sources,
- he may be able to “spin” an issue in a way that an expert would believe, but the old adage of trust but verify applies here.
So we touched on the difference between the public version of GPT-3.5 and the paid version of GPT-4, or the latest version, GPT-4 Turbo. In general, GPT-4 is less prone to contrivances and handles different languages better, including English. We reminded that the quality of the output is directly dependent on the quality of the input, and that it is useful to be aware that any result can be subjective. The talk also included a hands-on workshop where we made available our own instance of the interface on GPT-4.
A key theme of our workshop was safety when working with AI. In a school environment, there is a need to be more vigilant, as using AI can increase the risk of cyberbullying and enable, for example, the creation of ‘deepfake’ photos or videos. The aim was to educate the audience about the potential of AI, highlighting its effective use, while also highlighting the risks associated with the unwitting use of this technology.
We thank CUOK for the invitation and look forward to further opportunities for collaboration not only in the field of AI!
Are you also interested in learning more about the possibilities of AI, ChatGPT and its practical use not only to streamline your work? Feel free to schedule a no-obligation consultation.