The use of AI technologies in our institute

The use of various AI (machine learning, “artificial intelligence”) technologies is becoming an indispensable element of scientific research and our everyday lives. Research at our institute is supported by these technologies in two ways: on the one hand, generative AI services (e.g. ChatGPT, Google Gemini, Microsoft BingAI, SciSpace, GitHub CoPilot and HUN-REN’s own generative AI service) can support tasks that occur in scientific life, such as discovering and summarizing scientific results and trends, generating scientific program code, writing reports; on the other hand, machine learning models can be developed to perform a task performed by an expert, which can greatly reduce the time needed to complete tasks that generally require an expert. HUN-REN aims to make the most efficient use of AI solutions within the research network by appointing AI ambassadors in the different research institutes. For this task, researchers from institutes who are interested in computer science, programming and AI technologies, but also have experience in research, have been selected to bridge the gap between researchers and AI technologies

A little more than half a year ago, in mid-September 2024, István Bozsó, a colleague of our institute, started his work as an AI ambassador. István Bozsó was previously introduced to AI technologies as the coordinator of a HUN-REN funded institute project. During the first semester, István Bozsó assessed the areas where different AI technologies could be applied within the institute and presented the most important developments in the field and the AI tools provided by the HUN-REN centre to colleagues during several institute workshops. In a series of institute presentations in October, several institute colleagues presented how they have successfully applied machine learning technologies in their research. István Bozsó is also assisted by Dr. András Horváth, a researcher at the Faculty of Information Technology and Bionics at Pázmány Péter Catholic University and a colleague of our institute. In the next few weeks, we will briefly present in a series of newsletters on machine learning techniques already successfully applied in our institute and ongoing research in this area.