More than 136 thousand respondents took part in the survey. Ten variants selected from more than 3 thousand proposals made it to the final selection. Users of the application. This is reported by the Ministry of Digital Transformation.
The winning variant of the name of the big AI language model of Ukraine was just a little ahead of the second most popular variant – “Slovo” (“Slovo”). The third place was taken by the variant “Homin” (“Homin”, “Thunder”). Kavun (“Kavun”, “Watermelon”) fell a little short of the top three.
In June last year, Mincifry together with Kyivstar started the joint development of a national large language model. Its beta testing is scheduled for late spring this year.
Large Language Models (LLMs) are artificial intelligence systems that use neural networks to process, analyze and generate text. They are trained on huge amounts of data, allowing them to understand language, context and style, as well as produce coherent texts.
Built on a Transformer architecture, LLMs predict sequences of words, generating responses, codes or articles. They are used in chatbots, translators, text analysis, and programming. Well-known examples include GPT-4 (OpenAI), Gemini (Google), and Llama (Meta).
A domestic language model is expected to improve security, as foreign LLMs may ignore the Ukrainian context and be vulnerable to propaganda. Local data processing will become critical for AI applications in defense, medicine, and finance.
