DHAKA, Feb 14, 2025 (BSS) – As the AI race intensifies, DeepSeek and ChatGPT are poised to shape the future of artificial intelligence.
 
DeepSeek, a rising Chinese AI chatbot, is rapidly advancing in open-source AI development, challenging established tech leaders with its high-performing models like ChatGPT and ?Gemini of USA, Copilot of Canada and others.

 
Meanwhile, OpenAI’s ChatGPT continues to evolve, integrating cutting-edge features and expanding its influence across industries.
 

With both companies pushing the boundaries of AI capabilities, the competition could drive groundbreaking innovations, reshaping how humans interact with technology in the years to come.
 
However, the Chinese AI startup, DeepSeek made global headlines after topping app download charts and triggering a drop in US tech stocks.
 
In January, it introduced its latest model, DeepSeek R1, asserting that it could compete with OpenAI’s technology while being far more cost-effective to develop.
 
Its swift rise shook investor confidence, wiping out billions from chip giant Nvidia’s market value and challenging the assumption that American companies would lead the rapidly expanding AI industry.
 

BBC quoted President Donald Trump as he described the development as a "wake-up call" for US businesses.
 
DeepSeek's cost advantage over American AI companies has already brought significant changes to the tech market as Silicon Valley entrepreneur Marc Andreessen mentioned DeepSeek as one of the "most groundbreaking innovations" in the AI world.
 
What is DeepSeek?  
 
DeepSeek is a free AI chatbot similar to ChatGPT, used for various tasks, though its effectiveness compared to rivals is debated.  
 
It is reportedly as powerful as OpenAI's o1 model in areas like mathematics and coding. As a "reasoning" model, DeepSeek generates responses step by step, mimicking human thought processes.  
 
Developed at a fraction of OpenAI’s costs—just $6 million compared to GPT-4’s estimated $100 million—DeepSeek has also minimized the impact of US chip restrictions. 
 
Its founder stockpiled Nvidia A100 chips, banned from China since 2022, and possibly combined them with cheaper alternatives for efficiency.  
 
DeepSeek's lower memory usage further cuts costs, contributing to its success as the most-downloaded free app on Apple’s US App Store at launch. However, the same day, it faced "large-scale malicious attacks," forcing temporary registration limits and website outages.  
 
Like other Chinese AI models, DeepSeek avoids politically sensitive topics. When asked about the 1989 Tiananmen Square massacre, it provided no details, aligning with Chinese government censorship.
 
 
What is Artificial Intelligence?
 
AI enables computers to mimic human behavior by learning from vast amounts of data and recognizing patterns. This allows machines to solve problems, engage in conversations, and predict behaviors.
 
In recent years, AI has powered chatbots like ChatGPT and DeepSeek—forms of generative AI that creates new content by analyzing text and images. However, these tools can also generate inaccuracies and reflect biases from their training data.
 
Millions use AI for everyday tasks, from writing emails and summarizing text to coding and studying.
 
Who are DeepSeek’s users?
 
DeepSeek’s AI app can be downloaded for free from apps store and the company’s website. Due to being free, the app quickly became the top downloaded app on the Apple Store. However, some users have reported difficulties signing up for the app. It has already been recorded as the highest-rated free app on the apple app store in the United States.
 
What does DeepSeek do?
 
DeepSeek has become popular mainly for its powerful AI assistant tool, which operates similarly to ChatGPT. According to the description on the app store, it is designed to "enhance your productivity and answer your questions."
 
When rating the app, users have commented that it "makes writing more thoughtful."
 

Who is Behind DeepSeek?  
 
DeepSeek was founded in December 2023 by Liang Wenfeng, launching its first AI language model the following year.  
 
Little is known about Liang, a Zhejiang University graduate in electronic information engineering and computer science, but he has quickly gained international attention. He was recently seen at a meeting hosted by China’s Premier Li Qiang, highlighting DeepSeek’s growing influence.  
 
Unlike many Silicon Valley AI entrepreneurs, Liang has a finance background. He is the CEO of High-Flyer, a hedge fund that uses AI for quantitative trading. In 2019, High-Flyer became the first Chinese quant hedge fund to raise over 100 billion yuan ($13 million), according to BBC.  
 

Liang has emphasized China's need to move beyond being a follower in AI. In a rare interview, he suggested Silicon Valley was surprised by DeepSeek because it emerged as an innovator rather than just another Chinese firm imitating US technology, mentioned BBC.  
 
 
Differences between DeepSeek and ChatGPT:
 
ChatGPTis developed by OpenAI, a US-based company, while DeepSeekis developed by a Chinese startup, making it a competitor in the global AI race.
 
ChatGPT has a high development cost, with OpenAI's GPT-4 model reportedly costing over $100 million to develop. Besides, DeepSeek claims to have developed its R1 model for just $6 million, significantly cheaper than its rivals, making it a cost-effective option.
 
ChatGPT uses a large-scale transformer-based language model (GPT series), known for its ability to generate text based on large datasets. DeepSeek also utilizes a “chain of thought” model, which builds answers step-by-step, simulating human reasoning, leading to more understandable and reliable responses.

In terms of Performance and Computing Power, ChatGPT typically requires high computing resources, especially for advanced models like GPT-4. On the other hand, DeepSeek asserted to have achieved similar performance with 95 percent less cost and uses less computing power than models like GPT-4, which allows it to operate more efficiently.
 
On Pricing Model, ChatGPT offers free access to a basic version, but its advanced models (like GPT-4) come with a subscription fee of about $20 per month. DeepSeek is completely free to use, both for general users and developers. Developers can even download and integrate DeepSeek's code into their projects at no cost.
 
Regarding Popularity and Accessibility, ChatGPT is widely used across various sectors, including businesses, education, and personal use, but it is restricted by access costs for advanced versions.  
But DeepSeek has gained significant popularity, especially in China, due to its cost-effectiveness and easy accessibility, even surpassing ChatGPT in some download charts.
 
For Training Data and Political Sensitivity, ChatGPT generates responses based on global datasets, which can lead to biases or errors but is generally open to various topics. 

DeepSeek, like many other Chinese AI models, avoids politically sensitive subjects due to government censorship. For instance, it will not provide information on controversial topics like the Tiananmen Square massacre.
 

In summary, while both ChatGPT and DeepSeek provide advanced AI-driven tools, DeepSeek stands out with its affordable, efficient performance, free access, and targeted appeal in the Chinese market, posing competition to more established models like ChatGPT.
 
The reason Silicon Valley is trembling in fear of DeepSeek:
James Vincent, former editor of the tech news website The Verge, in an article published in Guardian mentioned that R1 has proven that it’s possible to create high-quality AI with lower costs and less power usage. It is 95 percent cheaper than OpenAI’s o1 and uses only one-tenth of the computing power compared to Meta’s LLaMA 3. 
 
The ability to create advanced AI on such a small budget is truly remarkable, raising questions about the expensive methods used by major US tech companies.
 
R1 is a "chain of thought" model. This means that when it answers a question, it logically builds and explains its response step by step, much like how a human thinks and forms an answer. 
 
In contrast, typical AI models simply retrieve information from a database to provide an answer. However, R1 can organize and explain its thoughts, making its answers more reliable and understandable.
 

This is why R1’s capabilities are considered comparable to OpenAI’s o1 model. ZeroOne also uses a similar chain of thought approach. The performance of both models is nearly on the same level.
 

Another major reason is that R1 is easy to use and completely free. Using OpenAI’s o1 model costs $20 per month, but R1’s app is entirely free. Not only is it available for general users, but it’s also fully free for developers. They can download R1’s code and use it in their businesses or projects, whereas other advanced AI models are typically available for a fee. This accessibility has significantly boosted R1's popularity.
 
Overall, it’s clear that the AI world is highly volatile, and predicting its future is challenging. Where DeepSeek’s R1 will take the AI landscape is something only time will tell, he added.
 
Why Were US Companies Like Nvidia Affected?  
 
DeepSeek’s success challenges the idea that AI progress depends on massive budgets and top-tier chips, raising uncertainty about the future of high-performance semiconductors.  
 
"DeepSeek has shown that advanced AI models can be built with limited computing power," says Wei Sun, principal AI analyst at Counterpoint Research, quoted him BBC. 
 
Meanwhile, OpenAI, valued at $157 billion, faces scrutiny over whether it can sustain innovation or justify its high valuation without significant returns.  
 
DeepSeek’s lower costs unsettled financial markets on January 27, triggering a broad sell-off. The Nasdaq dropped over 3 percent, impacting chipmakers and data centers globally.  
 
Nvidia’s stock plunged 17 percent on January 27 before recovering next day. Once the world's most valuable company by market cap, it fell to third place behind Apple and Microsoft as its valuation dropped from $3.5 trillion to $2.9 trillion, according to Forbes.  
 
As a private company, DeepSeek is not publicly traded on major stock exchanges.
 
How Has China Reacted to DeepSeek’s Impact?  
 
DeepSeek’s success is a major win for China, which has been striving for tech independence from the West.  
 
While the Communist Party has not commented, Chinese state media highlighted how Silicon Valley and Wall Street were "losing sleep" over DeepSeek, claiming it was "overturning" the US stock market, according to BBC.  
 
"In China, DeepSeek's rise is celebrated as proof of the country's growing technological strength and self-reliance," says Marina Zhang, an associate professor at the University of Technology Sydney. She sees it as validation of "China’s Innovation 2.0," a new wave of homegrown tech leadership led by young entrepreneurs.  
 
However, she cautions that this enthusiasm could also fuel "tech isolationism."
 
Why are many countries imposing bans on DeepSeek?
 
Countries like South Korea, Australia have imposed bans on the use of DeepSeek by government employees this week.  
 
According to a report by Qatar-based media outlet Al Jazeera, countries including the United States are considering similar bans on government devices, citing “security concerns”.
 
 
United States: the Wall Street Journal reported that US lawmakers are planning to introduce a bill to block DeepSeek on all government devices. 

On January 31, NASA blocked DeepSeek on its systems and employees' devices.  A week earlier, the US Navy instructed its members to refrain from using DeepSeek.
 
In South Korea, a spokesperson from the Ministry of Trade, Industry, and Energy announced that the use of DeepSeek on government employees' devices has been temporarily banned. 
 
On January 31, the country's Personal Information Protection Commission stated that it would write to DeepSeek asking for clarification on how user data is managed. The Hydro and Nuclear Power Department blocked all AI applications for its employees last month.
 
Australia's Home Affairs Department announced a ban on DeepSeek on all government devices. Home Affairs Minister Tony Burke stated that the ban is to protect Australia's national security and interests.
 

Italy's data protection authority announced on January 30 that it had instructed DeepSeek to limit the processing of data from Italian users. Two days earlier, the authority had asked DeepSeek to clarify how user data is stored and managed.
 
Taiwan has also imposed a ban on DeepSeek in government offices, citing "security risks."
 

All of these countries have expressed concerns about the potential security risks associated with the Chinese app. Additionally, they have highlighted the lack of transparency regarding how user data is stored and used.