Gábor Proszky is also a member of the academic editorial board that adopted a report on the state of the Hungarian language after a wide-ranging discussion forum. The first conclusion of the study is that Hungarian is not in danger. The Széchenyi Prize-winning mathematician and linguist said on the InfoRádió Aréna program that many people fear that Hungarian will disappear one day, but in his opinion this is unfounded, as it is spoken by more than ten million people and, at the very least, Hungarian, in its written form, is also a language that is strongly supported on the Internet.
As he said, previous surveys have found that the real survivors are those languages that find support in the digital world, and nowadays a large amount of Hungarian texts are available on the Internet. He stressed that writers and authors are helped by numerous language correction tools, and that the Hungarian Wikipedia is also “one of the world’s greats”. All this proves that Hungarian has the aforementioned linguistic support on the Internet, so, according to Gábor Prószky, its survival is not in danger. He believes that people are more afraid of Hungarian and “certain words appear that shouldn’t appear”. He added: Extinction is far from inevitable, and it even threatens smaller languages that are used by fewer people. Usually, a political or social reason leads to a language being replaced or disappearing altogether.
As he said, many people think that Hungarian is also in this situation when it comes to professional languages, but he reassured everyone that Hungarian will not be replaced. According to the linguist, it is worth considering at what level a language might be in danger. In general, linguists talk about levels of language, and according to semantics, we can talk about things, so syntax (the part of grammar that deals with the structures of words and sentences), language, and sentence formation are not in danger, according to Gabor. Prószéky, just as our word structure does not change.
He also drew attention to the fact that language change and transformation is a very long process, which may take centuries. Over time, vocabulary changes mainly, new forms appear, and nowadays these new words and expressions have their roots in English. According to the Director General of the HUN-REN Linguistics Research Center, the main reason for this is that English has become the dominant language all over the world, and many concepts and words have been adopted from English by other languages.
There is nothing new in this, since even in previous centuries the Hungarian language was always influenced, for example, by what people and nations lived around us or who invaded Hungary and stayed here for a long time. As a result, Turkish, Slavic, Latin and even Germanic words and expressions appeared and spread in the Hungarian language. “We are now in such a neighborhood with English, but everyone is at the same distance from each other on the Internet today,” commented Gábor Prószky.
What does the Hungarian-developed PULI do better than ChatGPT?
More and more AI-powered language models and chatbots are available for people, such as ChatGPT, Gemini, and Copilot, as well as the first Hungarian language model, PULI, which has 50 billion Hungarian words and its corpus. It represents the Hungarian language better than its large international counterparts. The PULI chatbot was designed and created by the Linguistics Research Center, and Gábor Prószéky also participated in its development. The linguist pointed out that Microsoft granted $10 billion to the OpenAI AI research lab last year to continue developing the ChatGPT model. Knowing this, many may wonder why a Hungarian chatbot is needed? According to Gábor Prószéky
PULI “A little bit yellow, a little bit cloudier, but it’s ours, and that’s what’s really important, because it’s still a Hungarian development.”
He admitted that the Hungarian language model will not be able to compete with ChatGPT and its large international partners, but they hope that it will still be very useful and valuable, because it is “honed for the Hungarian language” and can provide much more Hungarian cultural information “putting knowledge at everyone’s fingertips” than the world’s big programs. The latter, or even Wikipedia, contain 100-150 million words of Hungarian texts, through which the Hungarian language can be understood well, but these programs cannot generate Hungarian cultural content. “For example, they cannot compare the poems of János Arányi and Sándor Petőfi,” explained the Széchenyi Prize-winning mathematician and linguist.
He believes that PULI can be “stronger” than its competitors by disseminating Hungarian cultural content. He pointed out that the large chatbots spread around the world “know Hungarian very well,” but at the same time, there is a big difference between the 130 million words available and the 50 billion words available, which PULI can offer. He admitted that the large international systems are more advanced and sophisticated, but at the same time he stressed that
PULI can be delivered to a Hungarian partner, who can then run it on their own hardware base.
“There is no need to send content that may contain sensitive data to the Pacific Coast or anywhere else in the world,” the linguist added.
“After the world languages, Hungarian and Swedish experts were the first to develop their country’s linguistic model, ahead of Portuguese, for example,” Gábor Broszky noted. “We emerged with Hungarian at a very early stage, and because of this, computer linguistics and Hungarian language technology are now in a very good position. This is not because it is Hungarian, but because any locally sensitive content can be searched in other languages,” he explained.
He stressed that PULI is a trilingual system with English and Chinese texts available in addition to Hungarian texts. In addition to the 50 billion words of connected Hungarian text, 50-60 billion words in English and around 100 billion Chinese characters can also be read thanks to the developed Hungarian model. According to Gábor Prószky, PULI “will know the Hungarian system, Hungarian culture very well, but also the Euro-Atlantic world and the Far East.”
