Things you probably didn’t know about DeepSeek.
Artificial Intelligence (AI) has received a lot of excitement in the community with the release of a new open-source model called “DeepSeek.R1,” developed by the Chinese startup DeepSeek. Released on January 20, this app quickly rose to the top of the free charts on Apple’s App Store, surpassing OpenAI’s ChatGPT. According to DeepSeek, in tasks such as mathematics, coding, and natural language reasoning, the performance of this model can be compared to that of heavyweight, well-known models like OpenAI’s.
DeepSeek is known as Artificial Intelligence Fundamental Technology Research Company Limited. This firm was established in July 2023. As an innovative technology startup, DeepSeek is dedicated to developing cutting-edge large language models (LLMs) and related technologies.
Since officially launching its first model, “DeepSeek LLM,” in January of last year, the company has made continuous progress. According to US media reports, in December, the startup launched its open-source LLM “V3,” which surpassed all of Meta’s open-source LLMs and competed with OpenAI’s closed-source GPT-4-O.
The recently released model R1 has achieved a significant technical milestone, allowing artificial intelligence to automatically emerge with reasoning abilities using deep learning methods. Unlike traditional approaches such as Chain of Thought (COT) and Supervised Fine-Tuning (SFT), DeepSeek has established its identity in the AI industry by adopting Reinforcement Learning (RL) as its core training method.
When asked about the meaning of “DeepSeek,” its latest R1 chatbot responded, “The name reflects the company’s mission to deeply explore and advance the core technologies of artificial intelligence, aiming to push the boundaries of AI innovation and application.”
According to a technical report, DeepSeek’s manufacturing cost is approximately 5.57 million USD, which is significantly lower compared to other models. Experts believe that the progress achieved by DeepSeek “demonstrates the potential of advanced artificial intelligence at a very low cost.” The development of the AI industry has long depended on increasing computing power, and in this context, the low-cost DeepSeek model has the potential to update the AI landscape.
Experts also note that DeepSeek’s new model is particularly impressive in how effectively they have executed an open-source model. Open-source access allows researchers, developers, and users to access the model’s core code. Similarly, the parameters that determine how the model processes information enable users to customize, modify, or expand the model according to their needs.
It can be said that DeepSeek has greatly benefited from open-source principles, and as a result, has demonstrated a strong commitment to contributing to the sharing of knowledge and the collective development of technology. According to technology experts, DeepSeek’s approach is based on new ideas because their work is published and open-source, allowing everyone to benefit from it. This is undoubtedly the power of open research and open-source.
American experts also acknowledge that DeepSeek’s business and development model is open-source, which is seen as a powerful and successful model for science, technology, and business. While DeepSeek’s American counterpart, OpenAI, initially started as an open-source organization, it later transitioned to a closed-source model.
Emphasizing the importance of fostering collaboration and innovation through open-source principles, DeepSeek’s founder, Liang Wenfeng, has already made it clear that building a strong technical ecosystem is a top priority. Liang clarified the company’s stance by stating, “We will not choose closed sources.”
Washington (Daily Pakistan Online) The US Congress has banned its employees from using the Chinese chatbot “DeepSeek”. Congressional offices have been officially warned not to install this chatbot on their official phones, computers and other devices.
According to NDTV, a notice issued by the House Chief Administrative Officer said that DeepSeek is currently under review and its use in Congress has been declared unauthorized. The notice warned that cyber attackers are already using this chatbot to spread malicious software and damage devices. In view of this concern, Congress has implemented security measures to restrict the use of DeepSec on all government devices. Now no employee will be allowed to install DeepSeek on a government computer, phone or tablet.
This is not the first time that Congress has banned any software related to artificial intelligence. In 2023, Congress banned the free version of ChatGPT, while only allowing the paid version for specific purposes. Similarly, last April, the use of Microsoft Copilot was also banned.
Note: This is the author’s personal viewpoint and does not necessarily reflect the opinion of the organization.
If you would like to write for Daily Pakistan, please send your writings to the email address (atiqbaloch909@gmail.com) to send.