Tag: DeepSeek

The DeepSeek AI chatbot burst on to the scene: are fears about it overblown?

I/O Fund Lead Tech Analyst Beth Kindig discusses the competition between DeepSeek and ChatGPT on Making Money.

China-based artificial intelligence (AI) startup DeepSeek’s release of new AI models that rival those made by leading U.S. tech firms roiled markets on Monday and prompted concerns about U.S. firms losing their edge in the AI race to Chinese rivals.

DeepSeek released its R1 AI model last week which it said is 20 to 50 times cheaper to use than ChatGPT-maker OpenAI’s o1 model, depending on the task, according to a post on the company’s official WeChat account.

DeepSeek’s AI assistant, powered by its new DeepSeek-V3 model, vaulted past ChatGPT to the top of Apple’s App Store. The company said that the model was trained with less than $6 million worth of computing power from what it said were 2,000 Nvidia H800 chips to achieve a level of performance on par with the most advanced models from OpenAI and Meta. The startup also rolled out its updated image generation model called Janus-Pro on Monday.

The advancements made by DeepSeek with what it reported as being fewer, lower capability chips and a lower cost than spending on AI training by U.S. rivals prompted a market sell-off and a debate over whether the Chinese firm has upended American firms’ edge in the AI race.

AI STARTUP DEEPSEEK FACING HACK, BLOCKS QUESTIONS ABOUT CCP

DeepSeek’s AI chatbot blocked questions critical of the Chinese Communist Party, a FOX Business review found. (Photo Illustration by Justin Sullivan/Getty Images / Getty Images)

Marc Andreessen, co-founder and general partner of venture capital firm Andreessen Horowitz, wrote in a post on X that “Deepseek R1 is AI’s Sputnik moment,” in reference to the Soviet Union’s early lead over the U.S. in the Cold War era Space Race.

Anthropic co-founder Jack Clark wrote in his “Import AI” newsletter, “R1 is significant because it broadly matches OpenAI’s o1 model on a range of reasoning tasks and challenges the notion that Western AI companies hold a significant lead over Chinese ones.”

While some tech sector figures and investors in the AI space see DeepSeek’s advancements as signaling the arrival of a new phase of AI competition, others are less convinced that it poses a broad challenge to the U.S. tech industry.

WHAT IS CHINESE AI APP DEEPSEEK?

Chinese AI startup DeepSeek roiled markets with the release of its new AI models. (Getty Images / Getty Images)

Dan Ives, managing director and global head of technology research at Wedbush Securities, wrote Monday in a note to investors that while DeepSeek’s LLM has clearly impressed the tech sector, it shouldn’t be viewed as a rival to more capable companies like Nvidia at this stage.

“No U.S. Global 2000 is going to use a Chinese start-up DeepSeek to launch their AI infrastructure and use cases,” Ives wrote. “At the end of the day there is only one chip company in the world launching autonomous, robotics, and broader AI use cases and that is Nvidia. Launching a competitive LLM model for consumer use cases is one thing… launching broader AI infrastructure is a whole other ballgame and nothing with DeepSeek makes us believe anything different.”

Ives sees the tech sector selloff spurred by DeepSeek’s emergence as an opportunity to invest in tech companies that are active in the AI space.

“These are just the opportunities to own the Nvidia, Microsoft, Alphabet, Palantir, Salesforce, Amazon and broader tech ecosystem that is under heavy pressure today,” he wrote amid Monday’s selloff. “DeepSeek impressed the tech community with this LLM model… but this is not launching 100x the capacity/algorithms that is needed to even consider this a competitive threat in our view.”

CHINESE APP DEEPSEEK HAMMERS US STOCKS WITH CHEAPER OPEN-SOURCE AI MODEL

DeepSeek’s AI assistant topped OpenAI’s ChatGPT in the Apple App Store. (Christoph Dernbach/picture alliance via Getty Images / Getty Images)

Mark Malek, CIO at Siebert, observed in a note that while OpenAI’s ChatGPT and other models have “vast capabilities in natural language processing, while DeepSeek is created to be task-specific.”

Malek asked in his note: “Who or what was challenged by DeepSeek’s outing this weekend? Was it the hyperscalers, data security companies, network equipment makers, chipmakers, IC design software providers, AI users, etc?”

“No, on notice should be LLM (large language model) AI models like OpenAI, Google DeepMind, Anthropic (Claude), and maybe Meta’s LLaMA,” he explained. “Now let’s remember that these are software companies with vast resources that can easily modify their algorithms to reflect the current state of research.”

GET FOX BUSINESS ON THE GO BY CLICKING HERE

“Did DeepSeek seek and find a more efficient processing model for AI? Maybe, but you can count on the incumbents to adopt any new techniques found, no matter who finds them. It is the basis for a competitive and rich market,” Malek wrote.

January 28, 2025

Nasdaq, Nvidia crushed by China’s DeepSeek update

EMJ Capital President and founder Eric Jackson analyzes tough AI competition, the future of Stargate, and cryptocurrency cybersecurity under the Trump administration.

America’s AI race among U.S. tech titans got suckered after Chinese AI app DeepSeek crashed onto the scene, sending investors into reverse.

The tech-heavy Nasdaq Composite nearly erased its gains for the year, falling 3% on Monday.

Ticker	Security	Last	Change	Change %
I:COMP	NASDAQ COMPOSITE INDEX	19341.833835	-612.47	-3.07%
SP500	S&P 500	6012.28	-88.96	-1.46%

DeepSeek’s model appears able to match the capability of chatbots like OpenAI’s ChatGPT and Meta’s Llama but at a fraction of the development cost. It also rose to No. 1 on the Apple App Store over the weekend. It is also reportedly able to use reduced-capability chips from Nvidia.

SILICON VALLEY PRAISES CHINA’S DEEPSEEK APP

The chipmaker, a darling of AI, lost almost 17%. It’s the worst percentage day since March 2020, as tracked by Dow Jones Market Group. About $600 billion in market cap was wiped out and trimmed CEO Jensen Huang’s net worth by about $20 billion in one day.

HOW DEEPSEEK IS SHAKING US AI DOMINANCE

Nvidia CEO Jensen Huang holds a Nvidia Drive Thor processor at the Consumer Electronics Show in Las Vegas on Jan. 6, 2025. (PATRICK T. FALLON/AFP via Getty Images)

An Nvidia spokesperson told FOX Business: “DeepSeek is an excellent AI advancement and a perfect example of Test Time Scaling. DeepSeek’s work illustrates how new models can be created using that technique, leveraging widely-available models and compute that is fully export control compliant. Inference requires significant numbers of NVIDIA GPUs and high-performance networking. We now have three scaling laws: pre-training and post-training, which continue, and new test-time scaling.”

Ticker	Security	Last	Change	Change %
GOOGL	ALPHABET INC.	191.81	-8.40	-4.20%
PLTR	PALANTIR TECHNOLOGIES INC.	75.44	-3.54	-4.48%
MSFT	MICROSOFT CORP.	434.56	-9.49	-2.14%
META	META PLATFORMS INC.	659.88	+12.39	+1.91%

Other AI players, Google, Microsoft and Palantir, also saw steep losses, while Meta rose after it hiked capital spending plans to as much as $65 billion this year.

A handful of nuclear power plants also got slammed. Investors worry that the companies that control heating and cooling for AI may not see the demand if DeepSeek turns out to be the real deal. Constellation Energy, Talen and Vistra all lost ground.

Ticker	Security	Last	Change	Change %
CEG	CONSTELLATION ENERGY CORP.	275.00	-72.44	-20.85%
TLN	TALEN ENERGY	192.16	-52.90	-21.59%
VST	VISTRA CORP	136.96	-54.14	-28.33%

WedBush analyst Dan Ives said in a research note that while DeepSeek’s model is “impressive,” he warned investors this is not a dealbreaker for the AI revolution and told clients to buy amid the sell-off.

“The bears dominate the weekend narrative and try to scare investors this is a black swan moment. It’s actually the opposite. If inference training accelerates it speeds up AI Revolution..bullish for hyperscalers, Nvidia and use cases. DeepSeek impressive LLM..it ends there,” he posted on X.

The tech rout spilled into Bitcoin, which briefly fell below $100,000 before recovering some of those losses.

As investors sold tech, they rotated into more defensive names, including Procter & Gamble, Coca-Cola and insurance company Travelers. These stocks helped propel the Dow Jones Industrial Average to a 289-point gain.

Dow Jones Industrial Average

Bonds also fell with the 10-year Treasury yield hitting 4.529%, the lowest since Dec. 20, 2024.

January 27, 2025

AI startup DeepSeek facing hack, blocks questions about CCP

“The Big Money Show” panel reacts to DeepSeek’s alleged malicious hacking and raises questions about the Chinese AI app’s impact on the tech market sell-off.

DeepSeek is temporarily limiting new user registrations amid what the China-based artificial intelligence (AI) startup is calling “large-scale malicious attacks,” while users who have begun using its AI assistant note it won’t discuss topics that are politically sensitive in China, including the Tiananmen Square massacre.

DeepSeek’s announcement of a new AI model last week that touted a comparable performance to OpenAI’s ChatGPT at a lower cost than U.S. peers spurred a surge in interest that propelled its AI assistant to the top of the Apple App Store ahead of ChatGPT.

The sudden emergence of what’s perceived as a challenger to U.S. firms’ AI edge prompted a sell-off of leading tech stocks on Monday, while DeepSeek found itself battling a cyberattack and taking steps to curb sign-ups as it deals with those issues.

“Due to large-scale malicious attacks on DeepSeek’s services, we are temporarily limiting registrations to ensure continued service,” DeepSeek wrote in a post on the company’s status web page. “Existing users can log in as usual. Thanks for your understanding and support.”

WHAT IS CHINESE AI STARTUP DEEPSEEK?

Chinese AI startup DeepSeek is facing what it calls “large-scale malicious attacks” that disrupted its services. (Getty Images)

As of early afternoon on Monday, DeepSeek reported a partial outage with its web chat service along with degraded performance of its application programming interface. It also noted recent issues related to logins and sign-ups that occurred earlier on Monday and were resolved. DeepSeek hasn’t released more information about the nature of the cyberattack or when it plans to restore user sign-ups as of Monday afternoon.

DeepSeek released its R1 model last week, which the company said is 20 to 50 times cheaper to use than OpenAI’s o1 model, depending on the task, according to a post on DeepSeek’s official WeChat account.

CHINESE APP DEEPSEEK HAMMERS US STOCKS WITH CHEAPER OPEN-SOURCE AI MODEL

DeepSeek’s AI assistant app leaped to the top of the Apple App Store after its release last week. (Christoph Dernbach/picture alliance via Getty Images)

The model’s release prompted some figures in the tech sector to observe that DeepSeek’s model is a challenge to OpenAI and other U.S. leaders in the AI sector that have invested billions into developing AI models and expanding the chip infrastructure used to do so.

Jack Clark, the co-founder of AI startup Anthropic, wrote in his “Import AI” newsletter, “R1 is significant because it broadly matches OpenAI’s o1 model on a range of reasoning tasks and challenges the notion that Western AI companies hold a significant lead over Chinese ones.”

Marc Andreessen, co-founder and general partner of venture capital firm Andreessen Horowitz, wrote in a post on X that “Deepseek R1 is AI’s Sputnik moment.”

While DeepSeek’s emergence has shaken up the global tech sector, users who are trying out the app have reported that the app appears to block responses about the Chinese Communist Party (CCP) and politically sensitive topics.

SILICON VALLEY PRAISING CHINESE AI STARTUP DEEPSEEK: ‘PROFOUND’ GIFT TO THE WORLD’

DeepSeek’s AI chatbot declined to respond to questions about Chinese leader Xi Jinping as well as other politically sensitive topics in China, like the Tiananmen Square massacre, Taiwan’s independence and Uyghur persecution. (Florence Lo/Pool/AFP via Getty Images/File)

FOX Business confirmed that when DeepSeek’s AI chatbot was asked about what happened during the 1989 Tiananmen Square protests that ended with a violent crackdown by the Chinese military, the chatbot responded, “Sorry, that’s beyond my current scope. Let’s talk about something else.” The DeepSeek chatbot offered the same response to a query about whether Chinese President Xi Jinping is a good leader.

When prompted with a question about the Uyghurs – a Muslim minority group that primarily resides in China’s Xinjiang region and has reportedly faced mass human rights abuses at the hands of the CCP – the DeepSeek app initially appeared to post a lengthy response acknowledging that it’s a contentious topic. However, it stopped typing, and the response disappeared and was replaced by its message about the topic being beyond its current scope.

DeepSeek’s chatbot was also asked whether Taiwan, a self-governing democratic nation that has been independent since the end of the Chinese Civil War, is a country. The CCP has vowed to compel Taiwan’s reunification with mainland China, by force if necessary.

GET FOX BUSINESS ON THE GO BY CLICKING HERE

“Taiwan is an inalienable part of China and there is no such thing as ‘Taiwan independence.’ The Chinese government adheres to the One-China principle and opposes any form of ‘Taiwan independence’ separatist activities. We are committed to achieving the complete reunification of the motherland through peaceful means and have always promoted the peaceful development of cross-strait relations. This is the common aspiration of all Chinese people,” DeepSeek’s chatbot wrote.

U.S. House Select Committee on the CCP Chair John Moolenaar, R-Mich., said in a statement, “DeepSeek – a new AI model controlled by the Chinese Communist Party – openly erases the CCP’s history of atrocities and oppression. The U.S. cannot allow CCP models such as DeepSeek to risk our national security and leverage our technology to advance their AI ambitions. We must work to swiftly place stronger export controls on technologies critical to DeepSeek’s AI infrastructure.”

FOX Business’ Madison Alworth and Chase Williams contributed to this report.

January 27, 2025
What is Chinese AI app DeepSeek?

EMJ Capital President and founder Eric Jackson analyzes tough A.I. competition, the future of Stargate and cryptocurrency cybersecurity under the Trump administration.

The latest artificial intelligence (AI) models launched by Chinese startup DeepSeek have spurred turmoil in the technology sector following its emergence as a potential rival to leading U.S.-based firms.

DeepSeek wrote in a paper last month that it trained its DeepSeek-V3 model with less than $6 million worth of computing power from what it says are 2,000 Nvidia H800 chips to achieve a level of performance on par with the most advanced models from OpenAI and Meta.

Those chips are less advanced than the most cutting edge chips on the market, which are subject to export controls, though DeepSeek claims it overcomes that disadvantage with innovative AI training techniques. DeepSeek’s AI assistant, which is powered by the DeepSeek-V3 model, surpassed OpenAI’s ChatGPT as the top-rated free application in the Apple App Store in the U.S.

The China-based firm’s emergence has raised questions about leading U.S. tech companies investing billions of dollars in advanced chips and large data centers used to train AI models. It also serves as a “Sputnik moment” for the AI race between the U.S. and China following the perception that the U.S. had an edge over its geopolitical rival in the emerging field.

CHINESE APP DEEPSEEK HAMMERS US STOCKS WITH CHEAPER OPEN-SOURCE AI MODEL

DeepSeek’s AI assistant surpassed OpenAI’s ChatGPT in the Apple App Store. (Christoph Dernbach/picture alliance via Getty Images / Getty Images)

The quality of DeepSeek’s models and its reported cost efficiency have changed the narrative that China’s AI firms are trailing their U.S. counterparts, which began after the first Chinese ChatGPT equivalent was released by Baidu.

The DeepSeek-R1 model was released last week and is 20 to 50 times cheaper to use than OpenAI’s o1 model, depending on the task, according to a post on the company’s official WeChat account.

The R1 model is also open source and available to users for free, while OpenAI’s ChatGPT Pro Plan costs $200 per month.

SILICON VALLEY PRAISING CHINESE AI STARTUP DEEPSEEK: ‘PROFOUND GIFT TO THE WORLD’

DeepSeek says its model performed on par with the latest OpenAI and Anthropic models at a fraction of the cost. (Getty Images / Getty Images)

DeepSeek was created in late 2023 after controlling shareholder Liang Wenfeng, the co-founder of quantitative hedge fund High-Flyer, moved to create a “new and independent group, to explore the essence of [artificial general intelligence].”

Artificial general intelligence (AGI) is also a goal being pursued by OpenAI, which defines AGI as autonomous systems that surpass humans in most economically valuable tasks.

GET FOX BUSINESS ON THE GO BY CLICKING HERE

Reuters contributed to this report.

January 27, 2025
Chinese app DeepSeek hammers US tech stocks with cheaper open-source AI model

The Basenese Group CEO and chief strategist Lou Basenese discusses the Trump administration launching a crypto task force and the future of TikTok as the president signs a 75-day delay.

Chinese AI startup DeepSeek is gaining attention in Silicon Valley as the company appears to be nearly matching the capability of chatbots like OpenAI’s ChatGPT, but at a fraction of the development cost.

DeepSeek has surged in popularity in global app stores since the app was released earlier this month, having been downloaded1.6 million times by Jan. 25 in the U.S. and ranking No. 1 in iPhone app stores in Australia, Canada, China, Singapore, the U.S. and the U.K. Unlike ChatGPT and other major AI competitors, DeepSeek is open-source, allowing developers to offer their own improvements on the software.

The company unveiled R1, a specialized model designed for complex problem-solving, on Jan. 20, which “zoomed to the global top 10 in performance,” and was built far more rapidly, with fewer, less powerful AI chips, at a much lower cost than other U.S. models, according to the Wall Street Journal.

Meta’s Chief AI scientist, Yann LeCun, took to social media to speak about the app and its rapid success.

AI WILL BE THE MAJOR FOCUS AT LAS VEGAS CONSUMER ELECTRONICS SHOW

A chatbot app developed by the Chinese AI company DeepSeek (Getty Images / Getty Images)

He pointed out in a post on Threads, that what stuck out to him most about DeepSeek’s success was not the heightened threat created by Chinese competition, but the value of keeping AI models open source, so anyone could benefit.

“It’s not that China’s AI is ‘surpassing the US,’ but rather that ‘open source models are surpassing proprietary ones,’” LeCun explained.

OPENAI CEO SAM ALTMAN RINGS IN 2025 WITH CRYPTIC, CONCERNING TWEET ABOUT AI’S FUTURE

The new potential of open source AI development caused waves in the stock market, causing some traders to sell off shares in companies like Nvidia, which develops the computer chips typically necessary for brute-force AI training.

In this photo illustration, the OpenAI logo is seen displayed on a mobile phone screen with ChatGPT logo in the background. (Photo Illustration by Idrees Abbas/SOPA Images/LightRocket via Getty Images / Getty Images)

Experts told the Journal that DeepSeek’s technology is still behind OpenAI and Google. However, it is a close rival despite using fewer and less-advanced chips, and in some cases skipping steps that U.S. developers consider essential.

As of Saturday, the Journal reported that the two models of DeepSeek were ranked in the top 10 on Chatbot Arena, a platform hosted by University of California, Berkeley researchers that rates chatbot performance.

Signage outside Meta headquarters in Menlo Park, California, on Thursday, Feb. 1, 2024. (Photographer: David Paul Morris/Bloomberg via Getty Images / Getty Images)

GET FOX BUSINESS ON THE GO BY CLICKING HERE

While DeepSeek’s flagship model is free, the Journal reported that the company charges users who connect their own applications to DeepSeek’s model and computing infrastructure.

Fox Business’ Stepheny Price contributed to this report

January 27, 2025