The new world order in the context of Artificial Intelligence technologies

Before 2020, one of the most discussed topics in the context of technology and globalization was how firms that build and operate social networks were taking their algorithmic dominance even further, becoming "supranational organizations" equal to or even superior to states. Various approaches were used to explain this. For example, in the controlled hegemony approach, while the use of global digital networks is free, the state maintains strict legal oversight over them; in the mandatory hegemony approach, the state is unable to implement legal sanctions and administrative controls due to high demand for social networks because it cannot build a national social network infrastructure with domestic technology firms (a kind of helplessness); in the alternative hegemony approach, the use of global digital networks is free but their boundaries are determined by legal controls, with both local and global networks coexisting; and in the most extreme case, strict local hegemonic approaches were discussed, such as in Russia and China, where there are only state-controlled local digital networks.

Big technology companies took full advantage of the benefits of this period up to 2020, which we define as Web2, becoming the world's largest trillion-dollar firms and began investing in the next technology, artificial intelligence (AI). For example, Microsoft invested approximately 13 billion dollars in OpenAI last year. OpenAI announced that they spend 5 billion dollars every year to continue the development of ChatGPT. Google, on the other hand, announced that it has spent approximately 50 billion dollars to develop and deploy its AI application named Gemini.

Today, it can easily be said that we are beginning to experience a shift in axis on this issue. States have now realized that technology firms can use this extraordinary power as a pressure element against other states and have increased their interest in this matter. For example, Trump announced in his second term that the US would spend 500 billion dollars on AI infrastructure investments with the Stargate project. By establishing a Department of Government Efficiency, he appointed Elon Musk—the owner of Tesla, one of the largest technology firms and the world's top-selling car manufacturer, the developer of the AI application named Grok, and the richest person in the world—along with Vivek Ramaswamy to this post. In fact, technology companies have effectively entered state administration in this way.

For six years, the US has been implementing a trade embargo aimed at restricting China's access to high-tech products. The main goal here is to prevent China from surpassing the US in the production of technological innovations. For example, they have blocked the sale of powerful graphics processors used for AI operations, such as those newly released by Nvidia, to China, as well as Chinese mobile phone companies like Huawei and social media companies like TikTok.

On the other hand, with news announced by China last week, it began to be discussed that China was not affected by these restrictions at all; on the contrary, it had produced an AI application called DeepSeek (https://chat.deepseek.com/sign_in), which is open to everyone and more successful than ChatGPT, at a cost of 6 million dollars, which is 100 times cheaper and uses 30 times less processing power. Immediately afterward, the stocks of technology companies investing in AI on US stock exchanges, especially Nvidia, and cryptocurrencies, especially Bitcoin, suffered serious losses. The loss in the stock market alone was 1 trillion dollars. Moreover, since China's DeepSeek announcement coincided with the period when annual reports are prepared on the US stock market, technology companies are prohibited from making statements that would affect the market, and therefore they cannot make a statement to curb this decline.

Again, according to a recent statement made by the US, it is a matter of discussion that the US will share high-tech products only with 18 states it defines as "allies" and will implement this restriction within the next few months. Turkey's name is not among these allies. With this statement, the US has also revealed its intention to be a monopoly in technology. Because everyone now realizes that in the 21st century, mastering AI technologies in particular means directing the new world order in line with one's own interests.

The fact that the DeepSeek incident broke out at exactly this time carries the meaning of a "counter-move" to such a world order idea. In fact, DeepSeek was still using Nvidia's older processors. In other words, in a sense, these restrictions caused China to be more creative in innovation instead of stopping it.

The fact that China is producing cheaper AI applications that perform better with less processing power means that the US hegemony in this field, that is, in the field of innovation, will soon change in favor of China. Because it has been seen that development can be done without spending such large amounts of money. This also means that these AI startups will not be able to find large investments in the future as they have until now, and projects that require large expenditures like Stargate have already begun to be questioned.

One of the biggest problems encountered in AI development studies is that AI systems threaten the ecological balance because they consume too much energy, and the other is that huge amounts of data are needed to train AI systems. For example, while the energy used to train GPT-3, the old version of ChatGPT, is equivalent to the annual electricity consumption of 100 houses, the amount of energy used to train the latest version, GPT-4, is equivalent to the annual electricity consumption of 5,000 houses. On the other hand, it is mentioned that the stock of high-quality textual data on the Internet used to train AI will be exhausted by 2028. To this, it should be added that newspapers like the WSJ are suing OpenAI for using their news texts to train AI. As a solution to such data restrictions, some AI companies are planning to turn to quality data sources such as textbooks and produce "synthetic data".

While all this is happening, it should not be overlooked that it is being discussed that income distribution may deteriorate further with AI and that even totalitarian regimes may increase. There are truly big problems waiting to be solved ahead of us. That is why it is very important to increase our awareness on the subject.