Supported by AI experts, the HUN-REN competence support programme is set to launch and become accessible to all HUN-REN researchers as of this autumn to encourage higher value-added research. A pilot project to promote the use of artificial intelligence has already been launched, as revealed at the workshop attended by AI Ambassadors.
During the second Ambassadors' Workshop held as part of the AI Action Plan launched in March, AI Ambassadors from research sites presented some good practices in AI use from their respective institutions. Gergely Szertics, AI Advisor to the CEO and Managing Director of AI Partners, then outlined the details of the AI Action Plan's Competence Support Programme.
The programme consists of two parts: competence development and expert support. Developing researchers' AI competencies begins with inspiration and sparking interest, he stressed, noting that HUN-REN HQ relies on inspirational workshops led by AI Ambassadors for this purpose. He explained that to use AI tools effectively, it is essential to know the basics and start using the technologies. This requires clarifying fundamental concepts (such as GenAI or ML basics). Meanwhile, the aim is to establish a "dual-speed system," meaning that advanced users will also have the opportunity to learn about specialised target applications (such as image analysis, big data analysis, and language processing).
Since artificial intelligence is a general-purpose technology, the need to integrate it into research is emerging across more and more fields. To validate and refine these ideas, and to plan research based on validated ideas, HUN-REN decided to establish a system enabling researchers to share their ideas with an AI solutions expert who assists them in identifying the areas and conditions suitable for AI use within specific research projects.
According to the plan, three AI experts from different fields will start working in September. They will be available to HUN-REN researchers for consultations on research plans and available infrastructures. AI support for researchers also includes establishing a broader network of experts, allowing consultations on specialist issues. However, ensuring successful AI-supported research relies on building strong research collaborations, as Szertics emphasised. Therefore, a partnership programme is set to launch, connecting AI researchers from the network with colleagues planning to use AI in other research areas. The aim is to achieve higher value-added research through joint work.
For fine-tuning and testing purposes, HUN-REN HQ has launched a pilot programme, for which applications are expected starting summer from researchers willing to go through the above-mentioned validation process.This autumn, the competence support programme is set to launch and become accessible to all HUN-REN researchers.
In the second part of the meeting, Attila Farkas, a researcher from SZTAKI, presented the LLM-based (large language model) service built on HUN-REN Cloud resources, which operates similarly to the globally recognised ChatGPT. However, a key distinction is that the SZTAKI program is not connected to any third party cloud services, ensuring that all data remains within the HUN-REN Cloud. It was mentioned that while such a service has not previously been available to the network’s researchers, they can now choose from various large language models, for example, for coding purposes. The application is currently in beta testing mode, and those interested in exploring the system's capabilities are welcome to apply.
To conclude the programme, Júlia Torda, Head of Department of Communications at HUN-REN, briefly presented the plans for the HQ's upcoming new website. The revamped site will offer a wide range of services to HUN-REN researchers, with the communications team placing special emphasis on the topic of artificial intelligence. Available resources will include online AI courses and archived educational materials, as well as information on the ethical implications of AI use and application platforms for researchers seeking assistance from HUN-REN HQ experts in using artificial intelligence or booking available machine time ‒ in general, in exploration and utilisation of any AI opportunities.