On the same day that Donald Trump was sworn in for the second term as the 47th president of America, the world's most powerful country, on the same day, that is, on 20 January 2025, an emerging world power nation in China, an AI company called 'Depsic' launched its new 'R-1' model in the market.

Media and analysts around the world focused on studying the impact of Trump's second term and the hundreds of executive orders he has issued suddenly began to understand and assess the impact of Dipsic and his R-1 model. Interest in Dipsic peaked when the US stock market fell sharply on January 27. 

Nvidia, a company that makes GPUs (graphics processing units) used for generative AI training, lost about 600 billion and the entire US stock market lost about a trillion US dollars that day.

Generative AI, including ChatGPT, Jemenai and Sengai, challenging the American dominance in the AI ​​field, while many questions remain unanswered about Dipsic, its low cost and excellent products like R-1 are estimated to further democratize the AI ​​field and be useful for small countries and organizations like our country. In addition, this new product of Dipsic is likely to intensify the 'AI race' between the US and China, and it is also likely to affect small countries like Nepal. 

Dipsic, its role in democratizing the AI ​​sector and the possibility of a possible 'AI race' in Nepal may be relevant to discuss here. Although AI has been around for a long time, we are in the era of generative AI. Although AI is being used in many software and apps such as Tiktok, e-commerce, face recognition, robotics, AI is available at the level that the common people can see, know and use, it is through ChatGPT provided by an American organization called Open AI at the end of 2022.

A service provided by Google's Gemini, Meta's Lama, Perplexity, etc., including ChatGPT, is Generative AI, where text, images, audio, video, music, etc. are produced according to the user's instructions. Deepsik, which is currently in discussion, will also work in the same field of generative AI.  chatGPT is the most popular in the field of

Generative AI and even now many people understand that 'AI' is ChatGPT. With the success of ChatGPT, intense competition was also seen in this sector. Google came up with Gemini and Meta came up with Lama. Looking at generative AI development so far, apart from France's Mistral, the best and most trusted companies in this field are American companies.

Until recently, US dominance in the field was almost universally accepted, and second-ranked China was estimated to be at least six months behind the US, according to the Stanford Institute AI 2024 Index report. The allure of generative AI, its potentially bright future, and America's singular dominance in the field have led to huge sums of money from around the world pouring into US AI. As such, shares of Nvidia, which dominates the manufacturing of GPUs essential for AI model training, are seen to have increased by nearly 500 percent in the last 2 years 2023-2025. Thus in the field of generative AI 

America has taken several steps to protect its hegemony. The US has banned the sale of powerful AI chips to China since 2022, including Nvidia. Despite such bans, the new development of Dipsic has drawn the attention of many to this organization and its work. 

Even though Dipsic Generative is the first organization in China to earn this level of buzz in the AI ​​field, many companies including Alibaba, 01.AI have been working in China before. But Deepsik's R-1 model has surprised many unexpectedly. At the beginning of 2023, around the same time ChatGPT was being talked about around the world, 40-year-old Liang Wenfeng founded this company.

Among some aspects of AI opened by Dipsik, the model based on open source is important, which is not only free to use, but can also be modified according to the needs of the user. This is the big revolution brought by Dipsik in the field of generative AI. R-1 Model Like other companies around, it was not possible to use ChatGPT's 01 model without paying expensive fees. There was also concern that the gap between those who can pay and those who can't can be widened when high-quality models such as ChatGPT 01 have to be paid expensively and only low-quality models are available for free.

Developing a generative AI model like this requires high-level technology, human resources and huge sums of money. Generative AI model development, use, and commercial and strategic purposes in that area are currently the domain of powerful countries and groups, and relatively weak countries like Nepal are bound to be limited to the status of mere users. Concerns about alienation from potential economic and strategic benefits in the field of generative AI are gaining momentum.

In this case, Dipsic can be considered to have played a role in reducing potential concerns and the AI-gap to some extent by providing a high-class generative AI model like R-1 in the open source category. In this sense, Dipsic has opened up an easy and cheap way to develop open source models by gathering some tools and capabilities even in countries or groups or individuals that do not have the high level technology, human resources and large amount of money required for generative AI model development like Nepal.

has made easy access to advanced models even for universities or AI communities who do not have access to the advanced models of ChatGPT or Google in the past and who are independently studying and researching in the field of AI. Many have praised Dipsic for making the field of Generative AI more inclusive and democratic by making it simple and easy to access even the weakest users.  This move by Dipsik has added excitement among

users. In the context of the widespread use of generative AI at the public level in Nepal as well, the opportunity to use a high-quality AI model for free for relatively low-income Nepalis has added excitement.  Dipsik has shown an alternative way even in terms of the costs involved in preparing a

generative AI model. According to various studies and estimates, the R-1 model released by Dipsic is believed to have been prepared at a cost of only about three percent of ChatGPT's 01 model. In this way, Dipsic has questioned the justification of excessive spending by ChatGPT, Google, Nvidia, etc. in the field of generative AI, while on the other hand, it has shown that by working in a slightly different way for countries and groups that are relatively weak in cutting-edge technology and human resources, even in a sophisticated field like generative AI, it is possible to work or get results in a cheap but effective way.  With the discussion of

dipsik, many questions have been raised about it. The world wants to know from Dipsic transparently how such an advanced model was made in such a relatively short and cheap way. ChatGPT, Microsoft, and even the US government are thinking of investigating whether their information and technology has been used irregularly in the R-1 model of Dipsic. Not only

, there have also been complaints about the content produced by Deepsik's R-1 model. Complaints have surfaced that Dipsic refused to answer on China's power politics and strategically sensitive issues such as Tibet, Taiwan or the Tiananmen incident of 1989, and showed China's favor in other countries' disputes with China. 

Dipsic's leap is also likely to intensify the expected AI war between the US and China. Dipsic's R-1 model has not only challenged the US dominance in the field of generative AI, it is estimated that China has emerged as an AI power close to the US. Hinting about this, US President Trump recently said that there is a "wake-up call" in the AI ​​competition between the US and China. The possibility that the US may take strong measures to prevent China's possible AI hegemony can also be understood from what Trump indicated.