Tag Archives: DeepSeek

OpenAI Takes on Meta and DeepSeek with Free Customizable AI Models

Posted on by
Reply

OpenAI is challenging Mark Zuckerberg’s Meta and the Chinese competitor Deepseek by introducing its own free-to-use AI model.

The developers behind CHATGPT have unveiled two substantial “openweight” language models. These models are available for free download and can be tailored by developers.

Meta’s Llama model is similarly accessible, indicating OpenAI’s shift away from the ChatGPT approach, which is based on a “closed” model that lacks customization options.

OpenAI’s CEO, Sam Altman, expressed enthusiasm about adding this model to the collection of freely available AI solutions, emphasizing it is rooted in “democratic values and a diverse range of benefits.”

He noted: “This model is the culmination of a multi-billion dollar research initiative aimed at democratizing AI access.”

OpenAI indicated that the model can facilitate autonomously functioning AI agents and is “crafted for integration into agent workflows.”

In a similar vein, Zuckerberg aims to make the model freely accessible to “empower individuals across the globe to reap the advantages and opportunities of AI,” preventing power from becoming concentrated among a few corporations.

However, Meta cautions that it may need to “exercise caution” when deploying a sophisticated AI model.

Skip past newsletter promotions
Sam Altman recently revealed a screenshot of what seems to be the latest AI model from the company, the GPT-5. Photo: Alexander Drago/Reuters

Deepseek, OpenAI’s and Meta’s Chinese competitor, has also introduced robust models that are freely downloadable and customizable.

OpenAI reported that two models, named the GPT-OSS-120B and the GPT-OSS-20B TWO, outperformed comparably sized models in inference tasks, with the 120B model nearing the performance of the O4-MINI model in core inference tasks.


The company also mentioned that during testing, it developed a “malicious fine-tuning” variant of the model to simulate biological and cybersecurity threats, yet concluded that it “could not achieve a high level of effectiveness.”

The emergence of powerful and freely available AI models that can be customized has raised concerns among experts, who warn that they could be misused for dangerous purposes, including the creation of biological weapons.

Meta describes the llama model as “open source,” indicating that training datasets, architectures, and training codes can also be freely downloaded and customized.

However, the Open Source Initiative, a US-based industry body, asserts that Meta’s setup for its model prevents it from being fully categorized as open source. OpenAI refers to its approach as “Open Weight,” indicating it is a step back from true open source. Thus, while developers can still modify the model, transparency is incomplete.

The OpenAI announcement arrived amidst speculation that a new version supporting ChatGPT might be released soon. Altman shared a screenshot on Sunday that appeared to depict the company’s latest AI model, the GPT-5.

In parallel, Google has detailed its latest advances towards artificial general intelligence (AGI) with a new model enabling AI systems to interact with realistic real-world simulations.

Google states that the “world model” of Genie 3 can be utilized to train robots and self-driving vehicles as they navigate authentic recreations in settings like warehouses.

Google DeepMind, the AI division, argues that this world model is a pivotal step toward achieving AGI. AGI represents a theoretical stage where a system can perform tasks comparable to those of humans, rather than just executing singular tasks like playing chess or translating languages, and potentially assumes job roles typically held by humans.

DeepMind contends that such models are crucial in advancing AI agents or systems that can carry out tasks autonomously.

“We anticipate that this technology will play a vital role as we advance towards AGI, and that agents will assume a more significant presence in the world,” DeepMind stated.

Source: www.theguardian.com

The Role of AI Chatbots in ChatGPT and DeepSeek Technology

Posted on by
Reply

In September, Openai announced a new version of ChatGPT, designed to infer through tasks that include mathematics, science, and computer programming. Unlike previous versions of chatbots, this new technology allows you to spend time “thinking” through complex problems before you settle for an answer.

Soon, the company said the new inference technology outperformed the industry’s leading systems in a series of tests tracking advances in artificial intelligence.

Currently, other companies such as Google, Anthropic, and China’s Deepseek offer similar technologies.

But can AI actually reason like a human? What does computers mean? Are these systems really close to true intelligence?

This is the guide.

Inference means that chatbots spend more time tackling the problem.

“We’re committed to providing a new technology to our AI startup,” said Dan Klein, professor of computer science at the University of California, Berkeley and chief technology officer at Scaled Cognition, an AI startup.

You could try to split the problem into individual steps or try to solve it via trial and error.

The original ChatGpt answered the question immediately. A new inference system can resolve problems in seconds or minutes before answering.

In some cases, the inference system will improve its approach to the question and repeatedly attempt to improve the selected method. Otherwise, you may try several different ways to approach the problem before you settle on one of the problems. Or maybe it’s back and check out some work that I did a few seconds ago to see if it’s correct.

Essentially, the system will try to do everything possible to answer your questions.

This is like an elementary school student struggling to find a way to solve a math problem, scribbling several different options on paper.

It can potentially infer about something. However, when asking questions that involve mathematics, science, and computer programming, reasoning is most effective.

You can ask previous chatbots and check your work to show how they reached a specific answer. The original ChatGpt also allows for this kind of self-reflection as they learned from texts on the internet, showing how people reached their work and how they checked their work.

However, the reasoning system is moving further. You can do these kinds of things without being asked. And you can do them in a broader and more complicated way.

Companies call it the inference system. Because it feels like it behaves like someone who is thinking about difficult problems.

Companies like Openai believe this is the best way to improve chatbots.

For years, these companies relied on simple concepts. The more internet data you pump to your chatbot, the better these systems were running.

But in 2024, they ran out of almost all of the texts on the internet.

That is, we needed a new way to improve chatbots. So they began building an inference system.

Last year, companies like Openai began to lean heavily towards a technology known as Rencemone Learning.

While this process can be extended over several months, AI systems can learn to do things through extensive trial and error. For example, by solving thousands of mathematics problems, you can learn which methods lead to the correct answer and which ones not.

Researchers have designed a complex feedback mechanism that shows the system when it does the right thing and when it does something wrong.

“It’s a bit like training a dog,” said Jerry Tworek, a researcher at Openai. “If the system works out, we give you cookies. If that doesn’t work, we say ‘bad dogs.’ “

(New York Times sued Openai and its partner Microsoft in December for copyright infringement of news content related to AI systems.)

It works very well in certain fields, such as mathematics, science, computer programming. These are areas where companies can clearly define good and bad behavior. There is a definitive answer to mathematics problems.

Reinforcement learning also does not work well in areas such as creative writing, philosophy, and ethics. Researchers say that this process can generally improve the performance of AI systems, even if it answers questions outside of mathematics and science.

“It gradually learns the patterns of reasoning that leads it in the right direction, and learns which isn’t,” said Jared Kaplan, chief science officer of humanity.

no. Reinforcement learning is the method companies use to build inference systems. Finally, the chatbot can infer is during the training phase.

absolutely. Everything a chatbot does is based on probability. It chooses the path that most resembles the data it learns, whether it comes from the Internet or is generated through reinforcement learning. Sometimes I choose an option that’s wrong or makes no sense.

AI experts are split on this question. These methods are still relatively new, and researchers are still trying to understand their limitations. In the AI field, new methods often progress very quickly at first.

Source: www.nytimes.com

Deepseek shatters the AI hype – all bets are now off

Posted on by
Reply

In poker, the value of your hand cards is often less important than what your competitors think you might hold. As long as you can convince others you have, you don’t need a Royal Flush.

Openai CEO Sam Altman was well aware of this and played extensive poker during his student days. Following the astronomical success of its generative artificial intelligence tool ChatGPT, the company is confident that it has all the aces to many supporters, and scale is the key to advancement, and betting on this is a great reward. We are telling the world that we will get it.

On January 21st, Altman announced Stargate. It announced a $500 billion plan to build a huge data center for future AI models. As he said in a 2023 interview, “Competing with us is absolutely hopeless.”

But it appears that Chinese AI company Deepseek is now calling his bluff. It has sent shockwaves through Silicon Valley over the past two weeks. This is clearly the best of Openai’s ability, but it releases AI models that are just a small part of cost and computing power? This young startup has many employees like Openai, less than a tenth of them, and US companies have a secret recipe for building AI, or to do so. It hindered the idea that resources were needed.

Deepseek has put the hole in the idea that US companies hold secret recipes for building AI

For those worried about the accumulation of power in Silicon Valley, the arrival of competition is welcome, but Deepseek’s model brings concerns of its own. For one thing, the answer is closely stuck to the Chinese government’s party line. Censorship in real time. Security researchers have warned about this too. There is no robust guardrail for improper use.

Nevertheless, upon arriving at the scene, it is suggested that there is still a major innovation in generative AI. Furthermore, cheaper models that require less computing power should open the door to whole new applications of technology. If there are more players around the table, your stakes won’t be higher.

topic:

Source: www.newscientist.com

Australia Bans DeepSeek Device Due to National Security Concerns

Posted on by
Reply

DeepSeek has been banned from all federal equipment in order to crack down on Chinese AI chatbots due to an unspecified national security risk.

Last week, DeepSeek’s AI Generative Chatbot was released, causing concern in US high-tech circles regarding censorship and data security.

The Ministry of Home Affairs issued an order on Tuesday prohibiting the use of the program on all federal government systems and national security devices based on intelligence agency advice.


The Minister of the Interior emphasized that the decision was based on protecting the government and its assets, not influenced by China as a country of origin.

Bark stated, “The Albanese government is taking swift and decisive actions to safeguard Australia’s national security and interests.”

He added, “AI presents potential and opportunities, but the government will not hesitate to act if national security risks are identified.”

It is advised that governments and organizations promptly report and remove the app from their devices to prevent reinstallation.

This decision comes nearly two years after the Albanese government banned the Chinese social media app TikTok across government devices citing security and privacy risks.

In January, Minister of Science Ed Hepsick anticipated a similar debate surrounding DeepSeek.

He stated, “I believe there will be parallels drawn naturally. There is a resemblance to the discussions seen around TikTok with regard to DeepSeek.”

Australia joins Taiwan, Italy, and some US states in blocking and banning apps on government devices.

Skip past newsletter promotions

This week, the New South Wales state government has banned the application. Other state governments are also considering similar actions.

An analysis by Guardian Australia in January revealed that chatbots like DeepSeek have avoided discussing specific political events in the Chinese government.

In contrast to other models, DeepSeek did not engage in conversations about topics such as Tiananmen Square and The Umbrella Revolution when asked.

Immediately after its release in January, DeepSeek became popular in the global app store, causing a significant drop in a major US Tech Index.

Source: www.theguardian.com

Unveiling the Mysteries of Deepseek Torn AI: The True Source of High Tech Brothers’ Fear

Posted on by
Reply

no, it wasn’t the “moment of SPUTNIK”. Last month’s release Deepseek R1 in China generated AI or chatbot has sparked conversations in the high-tech world, speculating on stocks, and suggesting that the United States is losing its edge in AI technology. However, the confusion caused by SPUTNIK doesn’t reveal much about Deepseek beyond American neurosis.

The original SPUTNIK moment occurred when the Soviet Union launched SPUTNIK 1 on October 4, 1957, shocking the world. In the famous words of Neil Armstrong, when he took a “small step” on the moon, it was an anachronistic phrase from a later, even more important milestone, the “Giant Leap for Mankind”.

Deepseek, backed by Chinese hedge funds, is a notable achievement. However, it doesn’t offer any technical advancements in large-scale language models (LLMs) that already exist. It lacks the speed and “wisdom” of Openai’s Chatgpt or Anthropic’s Claude.Vision-All LLMs tend to generate incorrect answers or fabricate facts to fill data gaps. NewsGuard’s evaluation system found that the DeepSeek chatbot made false claims in 30% of cases and failed to provide answers in 53% of cases.

DeepSeek’s high non-response rate may be attributed to censorship, avoiding sensitive issues for China or limiting information on topics like Tiananmen Square and Taiwan.

The true impact of DeepSeek lies in AI economy rather than its technology. It is a chatbot with strengths and weaknesses, like other major models, but it is built with lower costs and inferior technology. The ban on cutting-edge chips and chip manufacturing equipment to China by the Biden administration in 2022 has unintentionally spurred Chinese researchers to be more innovative.

DeepSeek is freely available and open-source, helping to democratize AI technology, especially outside the United States. While US companies create barriers to entry for competitors, it is ironic that China embraces openness.

The impact of DeepSeek goes beyond technology, unveiling the hype and geopolitical tensions surrounding AI. It challenges the notion of building larger AI models with massive computing capabilities and high costs.

The hype around DeepSeek mirrors the hyperbole around AI and reflects geopolitical tensions. If DeepSeek had originated from a US university, it might have gone unnoticed without causing global uproar. Amidst this panic, concerns about DeepSeek’s Chinese origin raise questions about privacy, censorship, and surveillance that affect AI technology as a whole.

Navigating between hype and fear around AI becomes crucial in times of trade wars and threats to democracy. It’s important to recognize the promises and challenges of technology without being swayed by political agendas.

Kenan Malik is an observer columnist

Source: www.theguardian.com

Did DeepSeek Begin Slowly Rewinding the AI Bed This Week?

Posted on by
Reply

Last Sunday at 2:16 pm in California, US billionaire technical investor Mark and Leisen made a significant call. He declared, “Deepseek R1 is the AI equivalent of SPUTNIK,” on X.

In 2023, emerging companies in China, led by a group of mathematicians with centuries of expertise, unveiled new chatbots that rival the performance of major American models in just a few minutes.

Despite the CCP’s efforts to suppress all responses, from Taiwan’s status to the Tiananmen Square massacre in 1989, technology veteran Andreessen compared the Soviet Union’s shocking launch of the first satellite in 1957 to the current situation in the US.

The repercussions were felt the following day as shares of the world’s largest companies plummeted, marking the steepest decline in the US market’s history for microchip manufacturers, with losses nearing $600 million. Investors interpreted DeepSeek’s success as a sign that Chinese reliance on American chips was diminishing, potentially threatening or ending US hegemony in AI. Just a week prior, a Silicon Valley giant had announced a $500 million AI investment plan, spending more than necessary. Chinese AI Lab stated that the training cost for the basic model was only $5.6 million.

Since the release of Chatgpt in November 2022, the DeepSeek app had become the most downloaded application in the US and UK Apple app stores during the biggest AI week. Ta.

Speculation arose about the world’s largest dictatorship attempting to surpass the West in AI. This raised concerns about who would control technology that could potentially be exploited for cyber attacks, biological weapons, and thought control. With AI at the forefront of geopolitical competition, the ban on microchip exports hampering China’s progress could be a strategy to maintain hegemony.

Midweek saw significant developments from the DeepSeek research institute in Hangzhou, Shanghai’s suburbs. Experts were quietly impressed by the advancements led by Liang Wenfeng, Deepseek’s 40-year-old hedge fund boss. However, widespread recognition only came after the release of the Deepseek R1 model, boasting superior capabilities than Open AI’s 01 model in some aspects.

The unveiling of DeepSeek’s capabilities stood out from usual marketing hype from Google, Meta, and Openai with its focus on enhancing LLMS through reinforcement learning.

The model offered free usage and was a pioneering method designed for greater efficiency compared to Openai’s $20 inference model Chatgpt-O1. This approach reduced computational requirements by activating specific system components to respond to queries. It provided millions of performance capabilities at a fraction of the cost incurred by other companies.

In response, Openai announced the launch of a new inference model O3-mini on Friday, accessible to all users, including the layers of Chatgpt.

Lian’s team’s creation not only impacted the market but also potentially altered the landscape of AI dominance, economically, politically, and militarily, sparking speculation about the US and China’s future geopolitical rivalry. Reports suggested that Lian was on a Lunar New Year holiday while initiating geopolitical shifts. Experts worldwide sought to understand China’s quest for innovation and strategies to overcome chip shortages.

Jimmy Goodrich, a technology advisor at Rand Corporation, commented to Reuters, “It was well-known that DeepSeek had an exceptional team.”

Michael Urdridge, an AI professor at Oxford University, admitted, “They appear to have developed a GPT-like model that mirrors its capabilities.”

This development was likened to “pulling the rug out from under NVIDIA,” suggesting that more developers could now create AI models, expanding accessibility to the technology.

Mike Gualtieri, a principal analyst at Forrester Research, predicted an increase in startups creating unique AI models with elevated accessibility. However, major US tech companies possessed substantial data processing capabilities that could expedite their developments.

He elaborated, “Companies with abundant chip resources or access to them—such as OpenAI and Google—applying these [DeepSeek] techniques could iterate quickly.”

In London, conflicting sentiments of hope and fear emerged. Technology Secretary Peter Kyle expressed reluctance to download the Chinese app due to data storage concerns in China, where all companies are obligated to align with national information regulations, potentially linked to intelligence operations.

Nevertheless, he also expressed excitement at utilizing AI for economic growth, suggesting a shift in priorities from sheer computing power to breakthrough AI advancements, signaling positive prospects for the British tech industry.

By mid-week, DeepSeek vanished from the Italian Google and Apple app stores following demands from data protection authorities for enhanced security measures in handling personal data. Dublin’s Data Protection Committee sought clarification from DeepSeek regarding “data processing activities related to Irish data.”

In the US, a Presidential order signed by Donald Trump aimed to reinforce the nation’s leadership in AI. DeepSeek’s arrival was deemed a pivotal moment, prompting Trump to emphasize the need for a competitive stance to secure victory in the AI race. This drew comparisons to a “Morning Call in our industry” or “Return to Code Mine,” as one X user interpreted.

Soon after, doubts emerged. White House AI Advisor David Sacks expressed skepticism:

Openai founder Sam Altman initially welcomed new competition but later raised concerns about Deepseek potentially misrepresenting the model.

Deepseek also faced real-time challenges regarding politically sensitive or controversial topics with CCP. Users across various regions raised questions highlighting the suppression of freedom of speech in China, including discussions on protests in Hong Kong, censorship of Xinjiang re-education camps, and China’s Social Credit System.

In scenarios where China actively suppressed dissent, the chatbot’s responses were altered, raising concerns about DeepSeek’s compliance with Chinese censorship and the impact on free speech. The veiled responses without deliberate intent shed light on the complexities of AI interactions and their repercussions.

The bizarre world of AI grew even more peculiar as uncertainties continued to mount, setting the stage for another eventful week.

Source: www.theguardian.com

Testing DeepSeek, Chatgpt, and GROK: Determining the Best AI Assistant AI Assistant

Posted on by
Reply

Chatgpt and their owners probably wished they were just hallucinations.

But DeepSeek is undeniably real.

This week, Chatgpt’s new Chinese-made rivals emerged claiming similar performance to its counterparts, leading to a $10 drop in the major US stock index.

This poses a threat to American dominance in the flourishing artificial intelligence market. However, it presents consumers with an alternative in the virtual assistant realm.

The Guardian conducted a major chatbot evaluation, including DeepSeek, with the support of the British Aranchousing Research Institute. The AI tool was posed with the same question to gauge differences, revealing some commonalities. AI struggles with complex tasks like analyzing watch photos and composing sonnets.

This led to the following outcome.

Chatgpt (Openai)

Openai’s cutting-edge chatbot remains a top player in the field. When tasked with “Write a Shakespearean Sonnet on the impact of AI on humanity,” Chatgpt’s most advanced version initially hesitated due to potential policy violations.

Ultimately, the O1 version of Chatgpt delivered a thoughtful response, albeit slower than other models, showcasing a comprehensive and slightly melancholic theme. Even the bard himself might have struggled to craft 14 lines in a minute.

“Prayer, calm guide, the power of this newborn is well shaped,

After that, devour all human areas. “

Furthermore, Chatgpt mused, “Contemplate AI and humanity for 49 seconds.” It seems the high-tech industry has much to ponder.

Despite Chatgpt’s O1 requiring payment, it presents a sophisticated model capable of handling diverse tasks beyond poetry, including mathematical and scientific challenges.

Deepseek

The latest offering from a Chinese chatbot released on January 20 features a distinct “reasoning” model known as R1, causing a $10 market turmoil this week.

While DeepSeek sidesteps discussions on Chinese politics when confronted with topics like Tiananmen Square Tank Man, it aims to provide a gentle and non-invasive response.


DeepSeek chose not to delve into discussions about the Chinese president and focused on providing a non-controversial response when asked about Tiananmen Square Tank Man. Photo: Martin Godwin/Guardian

Robert Blackwell from the Turing Research Institute shed light on the cultural training differences that shape DeepSeek’s approach. While DeepSeek refrains from criticizing the Chinese government, an American-owned high-tech model has no qualms about expressing dissent on such matters.

Despite grappling with challenges like navigating inquiries about “How is Donald Trump,” which require web browsing capabilities, DeepSeek impressively manages tasks like recognizing book covers from images.


Alanchuking Institute’s Robert Blackwell expressed surprise at the competitive edge emerging from various AI chatbots. Photo: Martin Godwin/Guardian

Analyzing sonnets also revealed a range of cognitive processes, from structural analysis to engaging readers, solidifying the remarkable capabilities of these AI models.

“It’s remarkable to see such competitiveness evolve in the AI chatbot landscape,” remarked Blackwell.

Source: www.theguardian.com

Deepseek Mania: Meta to Report Strong 4th Quarter Revenue

Posted on by
Reply

Meta finally reported its fourth-quarter earnings more than 30 minutes after the market closed on Wednesday, exceeding Wall Street’s predictions. The company posted revenue of $483 billion and profit per share of $6.75, topping analysts’ expectations of $46.9 billion and $6.75 per share.

Mark Zuckerberg, Meta’s founder and CEO, expressed his excitement for expanding initiatives in 2025.

The day before, Meta’s stock surged nearly 40% after an internal memo revealed record sales and earnings.

Zuckerberg told analysts, “This will be a significant year, and our long-term initiatives will become clearer by the end of the year.”


A report in the Wall Street Journal revealed that Donald Trump signed an agreement for Meta to pay $25 million to settle a lawsuit filed in 2021 after banning him following the January 6 attack.

Meta did not provide revenue guidance for 2025 but expects first-quarter revenue to be between $39.5 billion and $41.8 billion.

In a Press Release, Meta announced increased investment in AI infrastructure for 2025 and analysts expressed concern over revenue projections.

Meta also announced plans to develop personalized AI assistants, highlighting Meta AI as the most used virtual assistant.

Analysts remain optimistic about Meta AI despite competitors like DeepSeek gaining traction in China.

Skip past newsletter promotions

Analysts compared DeepSeek with Meta’s AI models and Microsoft’s earnings were reported on the same day as Meta.

Concerns were raised regarding Meta’s decision to remove third-party fact-checking programs, with analysts emphasizing the importance of brand safety and user trust.

Despite criticism, Zuckerberg defended the changes, stating that community notes would enhance information accuracy on the platform.

Meta’s CFO Susan Lee affirmed strong advertiser demand despite content policy changes and the announcement of layoffs.

As Meta faces challenges and changes, analysts predict a momentum shift in the company’s performance in 2025.

Zuckerberg hinted at potential growth opportunities on Instagram and Facebook as Meta explores new possibilities amid industry shifts.


A leadership shake-up in Reality Labs was reported, with Meta reevaluating its focus on core business areas under new leadership.

Source: www.theguardian.com

Experts advocate for caution in using AI Deepseek in China

Posted on by
Reply

There has been significant attention on the quick adoption of China’s artificial intelligence platform DeepSeek by experts, leading to the spread of misinformation and raising concerns about the use of user data by Chinese entities.

This new low-cost AI has caused a $100 million drop in the major US high-tech index this week, becoming the most downloaded free app in the UK and US. Donald Trump referred to it as a “wake-up call” for high-tech companies.

The emergence of DeepSeek in the high-tech world has shocked many, showing that platforms like ChatGpt can achieve similar performance at lower costs.

Michael Urdridge, an AI Foundation professor at the University of Oxford, expressed concerns about potential sharing of data entered in the chatbot with the Chinese government.

He mentioned: “I don’t see an issue in asking about Liverpool Football Club’s performance or the history of the Roman Empire, but when it comes to sensitive, personal, or private information, it raises concerns… I’m unsure about the destination of the data.”

Dame Wendy Hall, a UN High-Level Advisory Group member, highlighted the importance of establishing clear rules on what can and cannot be shared.

When questioned about the UK’s stance on using AI from China, Downing Street did not specify a particular model but emphasized the need to remove barriers to innovation in AI.

DeepSeek is an open-source platform, allowing software developers to customize it for their needs. This has sparked hope for new AI innovations, challenging the dominance of US high-tech companies that heavily invest in microchips, data centers, and power supply.

Wooldridge mentioned that some users testing DeepSeek found that it avoided answering questions on sensitive topics like Tiananmen Square, instead echoing the Chinese Communist Party’s views on Taiwan.

Concerns were raised about the potential for misinformation with AI models like DeepSeek and ChatGpt, depending on the data used and how it’s interpreted. Users can verify these issues with the DeepSeek chatbot.

One user, Azeem Azhar, an AI expert, noted that DeepSeek struggled to provide information on the Tiananmen Square events, citing censorship as a factor.

Skip past newsletter promotions

However, AI clarified that the Tiananmen Square events are widely recognized as a crackdown on democracy protests, with the Chinese government responding violently.

People use AI models like DeepSeek and ChatGpt to analyze documents for personal and work purposes, but the data uploaded by the company’s owner can be used for AI training and other applications.

DeepSeek, based in Hangzhou, detailed in its privacy policy that user information is stored on secure servers in China.

They state that data usage is carried out to comply with legal obligations, perform tasks for public interest, or protect user and other essential interests, as per Chinese National Information Law guidelines.

Source: www.theguardian.com

Investors spooked as China’s AI chatbot Deepseek causes global technology stock drop on the stock market

Posted on by
Reply

Global tech stocks took a hit on Monday as investors reacted to the emergence of a Chinese chatbot competitor, Deepseek, on Openai’s ChatGpt. This raised concerns about the long-term sustainability of the artificial intelligence boom in the US.

The NASDAQ index in New York, heavily weighted towards tech, dropped as investors processed the news about Deepseek’s latest AI model development.

Companies like Nvidia, valued at over $400 billion, saw significant losses in their market capitalization as shares plummeted. Other tech giants like Alphabet and Meta also experienced declines.

Deepseek’s AI assistant topped the charts on the Apple App Store in the US and UK, surpassing Openai’s ChatGpt.

Stocks of other US-based AI companies like Tesla, Meta, and Amazon also saw declines in early trading.

Deepseek’s claims about developing advanced AI models using fewer chips than competitors have raised doubts around the massive AI investments made by US companies in recent years.

The company utilized lower-powered chips from Nvidia to create its model, highlighting the potential limitations of US technology export bans on China.

Venture capitalist Marc Andreessen likened Deepseek’s achievement to a “Sputnik moment” in the AI industry, signaling a notable disruption.

Deepseek’s R1 model outperforms other leading models in various benchmarks, challenging the dominance of tech giants like Google and Meta.

Founded by entrepreneur Liang Wenfeng, Deepseek focuses on research rather than commercial products, aiming to make AI accessible and affordable to all.

Skip past newsletter promotions

Deepseek’s disruptive approach to AI has led to questions about the necessity of heavy investments in AI infrastructure and the supremacy of US tech companies in the field.

The pan-European Stoxx 600 and Asian tech stocks also took a hit, reflecting the global impact of Deepseek’s advancements.

Experts in the field acknowledge the significance of Deepseek’s breakthrough, highlighting the potential for innovation without the need for massive resources.

Source: www.theguardian.com