Liang Wenfeng, founder of Deepseekappeared in the center of the technology industry of Chinaalmost immediately. The 39 -year -old man created the Chinese startup of IArtificial NTerligence that captured global attention after being launched to the public.
Liang appeared in a Symposium on January 20after maintaining a low profile for a long time. There he gave a speech by closed doors, under the framework of the event organized by Chinese Prime Minister, Li qiang. Just some interviews in 2023 were the young man’s statements.
Images of the Symposium were published by the local station CCTV, where the reserved Liang can be observed, who was invited to share his opinions on the Chinese government policy. This also led to the talk about Depseek, launched last week and, as they detail from the startup, Use less data at a fraction of the cost of current services, which triggered a global liquidation of technological actions.
Under Liang’s leadership, Depseek deliberately avoided the creation of applications. Insteadconcentrated its talent and research resources in the creation of a model that could match or improve Openai, and wait in the future to continue focusing on avant -garde models that other companies will use to create artificial intelligence products for consumers and companies.
In one of the few interviews he provided in 2023, he said: “China’s artificial intelligence It cannot be in the position of being a follower forever. We often say that there is a gap of one or two years between the artificial intelligence of China and that of the United States, but the real gap is the difference between originality and imitation. “
Who is Liang Wenfeng
Liang grew in the southern province of Guangdong, that during the eighties and ninety years he led the country in the adoption of market capitalism. Under that line, he said that at that time he was surrounded by people who valued the entrepreneurship more than the study, but that he had more academic inclinations.
To the 17 years enrolled in the elitist University of Zhejiangwhere he specialized in electronic engineering and communications before studying a master’s degree in Information and Communications Engineering, which completed in 2010.
Subsequently, In 2015 it co -founded a quantitative coverage fund that uses complex mathematical algorithms to perform transactions, instead of human analysis. It totaled more than 100,000 million yuan (13,790 million dollars) at the end of 2021, but in April 2023 he announced in his Wechat account that he would expand his scope of competition beyond the investment industry and concentrate resources for resources for “Explore the essence of generative artificial intelligence (IAG).”
Deepseek employees are mainly graduates and doctoral students from the best universities in China, who, according to Liang, prefer to work for Deepseek because it addresses the greatest challenges in AI: “What attracts the best talents is obviously the solution of the most difficult problems in the world”he had said.
What is Deepseek? The Chinese alternative to chatgpt that revolutionizes artificial intelligence
The chatbot Deepseekdeveloped in China, has managed to capture global attention by exceeding the leaders of the sector in efficiency and capacities.
This artificial intelligence model, which incorporates advanced reasoning chains, has challenged giants such as Openai and Google.
Deepseek’s success reasons
- Performance: The model Deepseek R1launched on January 20, 2025, specializes in solving complex problems through structured reasoning. In the main benchmarks, he has matched or even surpassed to the OpenA-O1 model, according to a report by The Wall Street Journal. This is remarkable since Depseek uses less advanced hardware than its competitors.
- Accessible cost: Deepseek offers a significantly lower price. While Openai charges 57.80 euros per million generated tokens, Deepseek reduces that cost to only 2.10 euros. This makes it an attractive option for companies and developers.
- Open source: Deepseek’s transparency has also been key. The company has chosen to release the code of its models, allowing the global community to access, improve and adapt technology for other projects. This contrasts with the trend of companies such as OpenAi and Google, who have closed their developments.
Deepseek characteristics
DEPEEEK allows users to activate a called function Deepthinkwhich displays the reasoning chains of the model until reaching the final response, simulating human behavior in their decisions. However, previous versions of the Chatbot have faced criticism for censoring sensitive political content in China.
Future projections
The R1 model, although it requires advanced hardware, is already used to train other smaller ones within the same company. This strategy of using advanced models to create lighter and more specific solutions opens new possibilities in the development of artificial intelligence.
Deepseek has established itself as a formidable competitor, standing out for its power, accessibility and focus on global collaboration. His rise in the sector raises a challenge to the big technological companies, marking a milestone in the development of AI.
Source: Ambito
I am an author and journalist who has worked in the entertainment industry for over a decade. I currently work as a news editor at a major news website, and my focus is on covering the latest trends in entertainment. I also write occasional pieces for other outlets, and have authored two books about the entertainment industry.
