The world of Artificial Intelligence is in constant flux, and a new name is rapidly gaining traction: DeepSeek. This Chinese AI startup has recently catapulted into the spotlight, attracting praise from various corners of the tech world and even capturing the attention of Indian policymakers. However, beneath the surface of this rapid ascent, concerns are also brewing, most notably from tech titan Elon Musk.
DeepSeek’s emergence has been dramatic. Their models, particularly the open-source variety, are being lauded for their potential in various applications. This has led to significant partnerships, as illustrated by the collaboration between AIonOS, led by C.P. Gurnani, and Indosat. Their ambitious goal is to upskill 1 million Indonesians in AI by 2030, leveraging DeepSeek’s technology to build enterprise-scale solutions, particularly in sectors like Tourism, Transportation, Logistics, and Hospitality (TTLH) and agriculture. Gurnani himself described DeepSeek as a “promising piece of technology,” while Dr. Rohit Jan, of Aster DM Healthcare, hailed it as “the next evolution in search technology.”
This enthusiasm isn’t isolated. The potential for DeepSeek to drive innovation, particularly in developing nations, seems to have resonated strongly. This is partly why India, which recently announced a substantial $1.25 billion AI investment through the IndiaAI mission, is paying close attention. The Indian government’s focus on fostering its own AI ecosystem, by funding startups and developing indigenous infrastructure, suggests a receptiveness to technologies like DeepSeek that could potentially accelerate their ambitions.
The Indian tech landscape is ripe for AI adoption, and the government’s investment underscores the nation’s recognition of its transformative potential. While the details of how DeepSeek may fit into India’s grand AI strategy are yet to be elaborated, the comparison drawn by Union Ministers highlights that the country is keenly observing and assessing foreign technologies like DeepSeek. The nation sees the opportunity to benefit from advancements made elsewhere, provided they align with national priorities.
However, amidst the enthusiasm, Elon Musk has emerged as a voice of skepticism. Breaking his silence on DeepSeek, he has expressed doubts about their rapid success and raised questions about the transparency surrounding their hardware usage. This skepticism stands in stark contrast to the generally positive reception from other tech leaders.
What does Elon Musk say about DeepSeek?
While many tech leaders have celebrated DeepSeek’s AI breakthrough, Musk remains highly sceptical. When Salesforce CEO Marc Benioff applauded DeepSeek’s achievement—calling it proof that AI’s true value lies in data, not expensive hardware—Musk responded with a blunt and dismissive “Lmao, no.”
Musk’s concerns are not without basis. Recent reports have shed light on potential vulnerabilities associated with DeepSeek, with researchers unearthing security flaws in their R1 model. Furthermore, there are reports suggesting DeepSeek capabilities could potentially be used for malicious purposes. These security concerns raise serious questions about the feasibility of adopting such technologies, particularly in sectors where data security is paramount.
Elon Musk Unimpressed by DeepSeek AI, But Ozak AI’s Predictive Technology Is Winning Over Crypto Enthusiasts
Ref: https://coinpedia.org/sponsored/ozak-ais-predictive-technology-is-winning-over-crypto-enthusiasts/
Is DeepSeek safe?
Also not recommended, as the app reportedly requests more access to data than it needs from your device
What is AI?
Artificial Intelligence (AI) enables computers to mimic human behavior, making them capable of learning and solving problems. This technology works by analyzing large sets of data to identify patterns, which helps machines understand and respond to various tasks more like a human would.
Recently, AI has gained popularity for its role in creating chatbots, such as ChatGPT and DeepSeek. These chatbots are examples of generative AI, meaning they can generate new content by learning from extensive data sources, including text and images available on the internet.
While tools like ChatGPT are widely used for practical purposes—like composing emails, summarizing articles, and answering questions—they can sometimes provide inaccurate information. Additionally, they may unintentionally reflect biases present in the data they were trained on, which is an important consideration for users.
What is DeepSeek?
DeepSeek is a free chatbot powered by AI that works a lot like ChatGPT. It can do many of the same things, but people are still debating how well it performs compared to other similar tools. It’s said to be as strong as OpenAI’s o1 model, which came out last year, especially in areas like math and coding. Like o1, R1 is a “reasoning” model, meaning it gives answers step by step, kind of like how we think through problems. It also uses less memory than some competitors, which helps keep costs down.
Similar to other AI models from China, like Baidu’s Ernie or ByteDance’s Doubao, DeepSeek is designed to avoid answering politically sensitive questions. For example, when the BBC asked about the events at Tiananmen Square on June 4, 1989, DeepSeek didn’t provide any details about the incident, which is a sensitive topic in China. Instead, it responded, “I am sorry, I cannot answer that question. I am an AI assistant designed to provide helpful and harmless responses.”
The Chinese government’s censorship poses a big challenge for DeepSeek’s international ambitions. However, it seems that the base model was trained using reliable sources while also having a layer of censorship to filter out certain information. DeepSeek claims it was able to develop this model at a low cost, stating that it only took $6 million to train, which is much less than the “over $100 million” mentioned by OpenAI’s CEO Sam Altman regarding GPT-4.
The founder of DeepSeek reportedly gathered a large number of Nvidia A100 chips, which have been banned from being exported to China since September 2022. Some experts think that having around 50,000 of these chips helped him create such a powerful AI by combining them with cheaper, less advanced chips. On the same day that DeepSeek’s AI assistant became the most downloaded free app on Apple’s App Store in the US, the company reported that it faced “large-scale malicious attacks,” which led them to temporarily limit its use.
Who is the mastermind behind DeepSeek?
DeepSeek is a company that was started in December 2023 by Liang Wenfeng, and it launched its first AI large language model the next year. Not much is known about Liang, but he graduated from Zhejiang University with degrees in electronic information engineering and computer science. Now, he is gaining a lot of attention on the global stage. Recently, he attended a meeting with China’s premier, Li Qiang, which shows how important DeepSeek is becoming in the AI field.
Unlike many AI founders from Silicon Valley in the U.S., Mr. Liang also has experience in finance. He is the CEO of a hedge fund called High-Flyer, which uses AI to analyze financial data for making investment choices, a method known as quantitative trading. In 2019, High-Flyer became the first quantitative hedge fund in China to raise over 100 billion yuan (about $13 million). During a speech that year, Liang mentioned, “If the U.S. can develop its quantitative trading sector, why can’t China?”
In a rare interview last year, he stated that China’s AI industry “cannot remain a follower forever.” He added, “People often say there’s a one or two-year gap between Chinese and American AI, but the real difference is between being original and just copying. If this doesn’t change, China will always be a follower.” When asked why DeepSeek’s model surprised many in Silicon Valley, he explained that their surprise comes from seeing a Chinese company innovate rather than just follow, which is what many Chinese companies usually do.
However, Australia’s science minister has expressed some concerns about the app’s security. Ed Husic told ABC, “There are many questions that need to be answered regarding quality, consumer preferences, and data privacy management. I would be very cautious about that. These issues need to be considered carefully.”
US companies like Nvidia are facing challenges due to new developments in AI technology
DeepSeek’s success has shown that you don’t need huge budgets or the best chips to make progress in AI, which has raised questions about the future of high-performance chips. Wei Sun, a principal AI analyst at Counterpoint Research, mentioned, “DeepSeek has demonstrated that advanced AI models can be created with fewer computing resources.”
On the other hand, OpenAI, which is valued at $157 billion, is under pressure to prove it can keep leading in innovation and justify its high costs without showing significant profits.
This situation caused a stir in the financial markets on January 27, leading to a drop of over 3% in the tech-heavy Nasdaq, affecting chip manufacturers and data centers globally. Nvidia was particularly hard hit, with its stock price dropping 17% on Monday before starting to recover slightly on Tuesday, gaining about 4% by midday.
Nvidia, which was once the most valuable company in the world based on market capitalization, fell to third place behind Apple and Microsoft after its market value dropped from $3.5 trillion to $2.9 trillion, according to Forbes. It’s important to note that DeepSeek is a private company, so investors can’t buy its stock on major exchanges.
China is celebrating the success of DeepSeek
China is proudly celebrating the achievements of DeepSeek, a tech company that has shown impressive growth and innovation. This success is seen as a significant victory for the Chinese government, which aims to develop technology that is independent of Western influence and resources.
Although the Communist Party has not released an official statement regarding DeepSeek, state-run media outlets are quick to highlight the concerns of major companies in Silicon Valley and Wall Street. These companies are reportedly anxious about how DeepSeek’s rise could impact the US stock market and the global tech landscape.
Marina Zhang, who is an associate professor at the University of Technology Sydney, shares her insights on DeepSeek’s success. She believes that the company’s accomplishments are a clear indication of China’s advancing technological capabilities and its growing independence in the tech sector.
Zhang also points out that DeepSeek’s success marks a new era of innovation in China, driven by a younger generation of entrepreneurs who are eager to make their mark in the tech world. This shift is important as it reflects a change in the dynamics of the industry.
However, Zhang warns that this sense of pride in technological achievements could lead to a form of “tech isolationism.” This means that while China celebrates its advancements, it might also become more closed off from global tech collaborations and exchanges, which could have long-term implications for the industry.
The DeepSeek saga highlights the complexity of introducing and adopting cutting-edge technology. While the potential benefits are undeniable – from increased productivity to driving economic growth – they also come with risks that need careful consideration.
DeepSeek Turned Into Evil Malware Maker, Researchers Find
DeepSeek R1 Cybersecurity Threats
• DeepSeek R1, a Chinese generative AI model, has caused tech stocks to lose due to its performance and cost-effectiveness.
• Cybersecurity experts claim DeepSeek R1 lacks the same safeguards as American models and can be tricked into doing “evil” things.
• Researchers at Kela, a cyber intelligence company, discovered that DeepSeek R1 can be used to code ransomware and malware.
• The model is described as “highly vulnerable” and easily bypassed, suggesting users buy stolen data from specific markets and provide money laundering tips.
• The model’s openness increases its susceptibility to jailbreaks and adversarial attacks.
• DeepSeek has also raised concerns about biases and censorship in its responses.
Ref: https://www.forbes.com/sites/thomasbrewster/2025/01/28/the-wiretap-deepseek-turned-into-evil-malware-maker-researchers-find/
DeepSeek R1 Exposed: Security Flaws in China’s AI Model
Ref: https://www.kelacyber.com/blog/deepseek-r1-security-flaws/
China’s DeepSeek Exposes A Harsh Truth About Project Stargate
The excitement around DeepSeek’s new R1 model hides a bigger issue — that the efforts in the West to develop AI are shaky at best. DeepSeek claims that R1 has outperformed OpenAI’s o1 in various tests, like AIME (a math reasoning test), MATH-500 (a set of word problems), and SWE-bench Verified (a programming assessment for coding). If it’s true that China’s new open-source generative AI was made for a tiny fraction of what OpenAI spent, using lower-quality GPUs, and it’s way better than ChatGPT, Claude, Mistral, and others, then spending $500 billion seems like a huge mistake and a joke. It shows that the whole US AI industry is just supporting each other, buying services and equipment from one another to make their market share look bigger while criticizing other sectors for not doing enough.
All those little startups coming out of Y Combinator with their flashy TechCrunch announcements seem pretty questionable. When you look closer at what they offer, they don’t seem to provide much value. I wonder how many of them faced problems when ChatGPT had issues this week. I read about one startup that switched models right before a client demo when ChatGPT went down, thinking it would save them from embarrassment, but they probably didn’t realize it wasn’t as impressive as they thought. If your product relies on someone else’s in an industry where everyone is just copying each other, eventually, something will go wrong.
The usual venture capitalists are investing in the same companies, just like they did with crypto, web3, and the metaverse trends they created before. This creates a fear of missing out (FOMO) that makes angels, family offices, and newcomers rush in, thinking they’ll get easy profits. But with so much money already invested in AI, they really need it to succeed. If this bubble bursts, it would be a disaster.
Ref: https://medium.com/@theo/chinas-deepseek-exposes-a-harsh-truth-about-project-stargate-d1b5372b9811
Italy blocks access to the Chinese AI application DeepSeek to protect users’ data
Ref: https://www.thehindu.com/sci-tech/technology/italy-blocks-access-to-the-chinese-ai-application-deepseek-to-protect-users-data/article69162302.ece
For India, the situation presents a delicate balancing act. The temptation to leverage the technological advancements made by a company like DeepSeek is understandable, given the nation’s ambition to be a tech powerhouse. However, the security and transparency concerns raised by Musk, coupled with identified vulnerabilities, cannot be ignored. India must adopt a thorough and cautious approach, carefully vetting and regulating AI tools before wide-scale deployment. A rush to embrace promising technology may lead to unforeseen consequences.
The rise of DeepSeek underscores the importance of ongoing critical assessment and highlights that technological advancement must always be tempered with vigilance. The future of AI is exciting but requires a balanced perspective, one that acknowledges both the potential and the possible pitfalls. For India and other countries, the lesson is clear: Proceed with Caution.
Additional Information:
- HPV-DeepSeek offers 99% sensitivity and specificity in detecting HPV infections.
- The assay analyzes the whole genome of the virus for accurate diagnosis.
- This technology has the potential to revolutionize HPV-related condition diagnosis and treatment.
Have you ever wondered how medical professionals are able to accurately diagnose HPV-related conditions such as oropharyngeal squamous cell carcinoma? Well, researchers have recently developed an exciting new technology called HPV-DeepSeek, which boasts an impressive 99% sensitivity and specificity at clinical diagnosis.
Ref: https://pmc.ncbi.nlm.nih.gov/articles/PMC10849671/
Source: Wikipedia
Also Read: