Can Artificial Intelligence Unite Nations?

The most popular word of the 90s was globalization. The bipolar world order, which began immediately after the Second World War, ended after the fall of communism. A new wave of globalization began under the leadership of the United States. Developments in transport technology and widespread free trade had already brought industrial goods production to a global dimension. The spread of the internet has paved the way for the free movement of information around the world. It was suggested that the end of nation-states was near, as the world became a global village. The trade union, which flourished in Europe starting in 1945, was crowned in 1993 with the free movement of products, services, people, and money. In the following years, the European Union expanded in waves to include Eastern Europe and the Balkans, and a united Europe was formed.

Globalization continues to exist as an effective trend throughout the world, but nationalism continues to maintain its place on the agenda. I think it is normal, as not everyone in the world benefits equally from globalization's blessings. Although it was said that the world would become a global village and that nation-states would completely lose their importance, the practice was different. In the globalization process, undeveloped countries lost power. The Middle Eastern and Balkan states, who lost their national integrity, suffered greatly due to internal and external wars. Nationalist and protectionist policies continue to find supporters in developed countries. For reasons such as immigrants' share of the job market, cultural differences, and local businesses' damage from global competition, this school continues to be effective in policy in Europe and the United States.

Which is more effective at being a nation? Common Language? History and culture Union? Religious Union? Legal order? Currency? Army?

I think that language is the most important factor in the formation and survival of nations. The world was a global place to some extent before nation-states. At various times in history, major imperial structures such as the Roman Empire, the Ottoman Empire, and the British Empire established global order on the lands they ruled. The nationalism that arose in France in modern times gradually brought the end of Empires. Nation-states have created a market economy within themselves. Language unity largely sets the limits of these markets. Of course, there are countries where more than one mother tongue is spoken, but citizens of those countries agree on a common language.

Thanks to rapidly developing translation programs, artificial intelligence technologies seem to be candidates for removing the language barrier between peoples of different countries. When will AI-powered translation systems become more useful? It would be appropriate to make a few statements about these systems to decide this.

Although the theoretical basis of translation systems dates back to the 1930s until the 1980s, these systems were tried to produce hand-written program codes. From the 1980s, machine learning methods began to be used for translation. The difference was that instead of hand-written codes on linguists' recipes, algorithms were created on rules learned from bilingual versions of texts that were subject to translation. Deep learning methods, which became possible with the development of computer capabilities in the 2010s, played an important role in achieving the current language-to-language translation level.

The models that form the basis for language-to-language translation are created today by the deep learning method, also called Deep Artificial Neural Networks. In training deep learning models, huge data sets of millions of words are used in two languages subject to translation. A special deep learning method called Long Short Term Memory (LSTM) has been developed for processing the language for free text processing purposes. Deep learning models gained a kind of short-term memory thanks to the LSTM method. Thus, when comparing two data sets based on letters, words, and sentences, they do not have to re-learn the relationships repeatedly. The deep learning method, which can process words by taking into account the words before and after them, is efficient in quality. Developments such as the use of GPUs instead of CPUs, distributed data architectures have made models that learn from a much larger data pool possible. Since development in Information Technologies is exponential, it would not be wrong to expect rapid translation software development.

In the meantime, deep learning is also used to transform handwriting into computer text, voice into text, and text into voice. So it is possible with today's technology that our voice goes to the other side in the other language when we talk to someone on the phone. As processors become more powerful, such systems will become more user-friendly over time. Can computers translate at the human level? Probably the situation will vary. Translators will be out of a job? It seems difficult for any software to translate what a master translator does. Artificial intelligence will probably be an aid in improving their productivity rather than replacing translators. I think the same applies to other professions that seem to be under threat.

Once artificial intelligence removes the language barrier, another important pillar of nation-states will disappear. It will probably be more difficult for the new generation, who uses crypto money instead of their national currency and interacts with people worldwide through the internet, to adopt nationalist values.

Since each effect creates its own reaction, it is difficult to predict what will happen in the future. I hope that the development will be towards a multicultural, egalitarian, cosmopolitan world.

Image Source: https://unsplash.com/