Raphael Stevens

At a summit in Paris on February 10th and 11th, tech bosses vied to issue the most grandiose claim about artificial intelligence. “AI will be the most profound shift of our lifetimes,” is how Sundar Pichai, Alphabet’s boss, put it. Dario Amodei, chief executive of Anthropic, said that it would lead to the “largest change to the global labour market in human history”. In a blog post, Sam Altman of OpenAI wrote that “In a decade perhaps everyone on earth will be capable of accomplishing more than the most impactful person can today.” More

In early February Openai, the world’s most famous artificial-intelligence firm, released Deep Research, which is “designed to perform in-depth, multi-step research”. With a few strokes of a keyboard, the tool can produce a paper on any topic in minutes. Many academics love it. “Asking OpenAI’s Deep Research about topics I am writing papers on has been incredibly fruitful,” said Ethan Mollick of the University of Pennsylvania. Some economists go further. “I am *sure* for B-level journals, you can publish papers you ‘wrote’ in a day”, said Kevin Bryan of the University of Toronto. “I think of the quality as comparable to having a good PhD-level research assistant, and sending that person away with a task for a week or two,” said Tyler Cowen of George Mason University, an economist with cult-like status in Silicon Valley. More

Eight automakers including BYD, at least nine financial securities companies, three state-owned telecommunications operators and smartphone brand Honor are among the many that have rushed to integrate with DeepSeek in the last week.
“This is quite unprecedented,” Wei Sun, principal analyst of artificial intelligence at Counterpoint Research, said in an email Monday. She pointed to the rate of adoption, scale of business integration and breadth of specific industries covered.
A big factor in the widespread interest is timing, as well as DeepSeek’s open-source availability in China.

Dado Ruvic | Reuters

BEIJING — Chinese businesses are tapping DeepSeek’s newest artificial intelligence model to see how it can improve productivity.
The Chinese AI model took the world by storm in recent weeks after showcasing its reasoning process and claims to undercut rival OpenAI’s ChatGPT on cost — despite U.S. restrictions on Chinese access to the advanced semiconductors needed to develop the tech.

Eight automakers including BYD, at least nine financial securities companies, three state-owned telecommunications operators and smartphone brand Honor are among the many that have rushed in the last week to integrate with DeepSeek. Cloud computing operators Alibaba, Huawei, Tencent and Baidu have all offered ways for clients to access DeepSeek’s latest model.
“This is quite unprecedented,” Wei Sun, principal analyst of artificial intelligence at Counterpoint Research, said in an email Monday. She pointed to the rate of adoption, scale of business integration and breadth of specific industries covered.
“When we have all of these, we know it’s making a big social and economic impact,” she said.

Optimism over artificial intelligence has spread to Chinese stocks. UBS said Wednesday that AI-related Chinese stocks are up by 15% since the start of the year, outperforming the broader MSCI China Index by 9%.
A big factor in the widespread interest is timing. DeepSeek released its latest R1 model on Jan. 20, and news of its low-cost reasoning capabilities prompted a global tech stock sell-off on Jan. 27 — just as millions of urban workers in China were returning to their hometowns to celebrate the eight-day Lunar New Year holiday.

As a result, less developed parts of China gained greater understanding of AI and its impact, a topic previously limited to conversations in China’s largest cities, said Wenhao Zhang, CEO of the Beijing-based consumer marketing consultancy Doodod.
“It’s a major education of the market. This will push the entire ecosystem’s development,” he said Tuesday in Mandarin, translated by CNBC.
Zhang, who studied AI at Tsinghua University, founded Doodod in 2012 to build customer engagement through social media analysis. He said the company — which counts China Merchant’s Bank and Toyota as clients — started looking at DeepSeek’s offerings late last year, and began using it more after the R1 release in late January.
DeepSeek, founded in 2023 out of a quantitative hedge fund, had released a basic version of R1 in November, and a V3 model in December. It launched a smartphone chatbot app in January.

Open-source local deployment

Another attractive factor for businesses is that DeepSeek’s models are open-source, allowing individuals and companies to download and customize it.
DeepSeek also advertised drastically lower prices for applications to use its tech versus that of OpenAI. ChatGPT is not officially available in mainland China and requires users to provide an overseas phone number and payment method from a supported country such as the U.S.
DeepSeek changed the perception that AI models only belong to big companies and have high implementation costs, said James Tong, CEO of Movitech, an enterprise software company which says its clients include Danone and China’s State Grid.
He said Movitech started integrating an earlier version of DeepSeek in the fourth quarter of last year, helping boost sales by about 25% from the same period in 2023. The company plans to launch a new DeepSeek-integrated application by the end of March to improve clients’ ability to make decisions, he said.
Many recent videos on Chinese social media have showed off how to run a local version of DeepSeek on Apple’s Mac mini.
Apple Mac mini online sales in China climbed significantly from November to January, versus the same period the year prior, according to data from consultancy WPIC. The electronics-focused JD.com site recorded unit sales of around 20,200 in January, up from nearly 19,400 in December and around 12,250 in November, the data showed.
DeepSeek’s affordability is pressuring more expensive AI models to cut prices, enabling more businesses to adopt the tech, said Chim Lee, senior Asia analyst at the Economist Intelligence Unit. He added that open-source models allow finance, banking and healthcare businesses — which are subject to stringent data protection rules in China — to develop AI applications locally.
“It is still very early to point to concrete business applications, but a key takeaway is that DeepSeek will accelerate the commoditization of AI,” Lee said.
Beijing is also increasing support. China’s national supercomputing network announced Tuesday that eligible companies and individuals can obtain three free months of DeepSeek access, along with subsidized computing power.
The network is similar to OpenAI’s Trump-backed Stargate project in the U.S. for building AI infrastructure — with the potential for “even faster scaling,” Winston Ma, adjunct professor at NYU School of Law said Wednesday. He is also the author of “The Digital War: How China’s Tech Power Shapes the Future of AI, Blockchain and Cyberspace.” 

Not centered on DeepSeek

The rush to try out DeepSeek doesn’t mean it will be the only AI provider for Chinese companies. Developers in the U.S. and China are regularly releasing new models.
Movitech also uses Alibaba’s Qwen AI model, Tong said, noting that the market wants the tech that can lower costs and produce results the most, whether it’s OpenAI or DeepSeek.
HangHang AI, which has invested several hundred million yuan to develop AI solutions for companies across 20 industries, uses a range of models, said partner and COO Shu Weibing.
Many people first used “Baidu, then realized it wasn’t as good as Kimi, then it wasn’t as good as [ByteDance’s] Doubao, which also cut prices,” Shu said in Mandarin, translated by CNBC. “Now it’s DeepSeek.”
It remains to be seen how much generative AI can boost productivity and profits.
Shu predicts that small businesses and companies that integrate AI with hardware will benefit more than large, consumer-facing internet platforms, whose AI work so far, he said, has focused more on boosting efficiency rather than creating new consumer services.
Despite AI models’ falling prices, “small and medium-sized businesses may still be in a period of wait-and-see” for adopting the tech due to the relatively high cost for a full deployment, including computing power and customization, Mike Fang, senior director analyst at Gartner, said Wednesday in Mandarin translated by CNBC.
But the consulting firm predicts that by 2027, the average price to access a generative AI model will be less than 1% of what it costs now — and that by 2029, 60% of Chinese businesses will have incorporated AI into their primary products and services, forming the top drivers of revenue growth. More

The Consumer Financial Protection Bureau sent termination notices to several dozen employees, according to people with knowledge of the situation.
The affected staff were mostly those with probationary status, said the people, who asked for anonymity to speak candidly after orders to stop all agency work, including speaking with reporters.
Several of those being laid off had already accepted federal buyout offers, said one of the people.

Acting director of the Office of Management and Budget Russell Vought speaks with reporters during a press briefing at the White House in Washington on March 11, 2019.
Jonathan Ernst | Reuters

The Consumer Financial Protection Bureau sent termination notices to several dozen employees late Tuesday, according to people with knowledge of the situation.
The affected staff were mostly those with probationary status, said the people, who asked for anonymity to speak candidly after orders to stop all agency work, including speaking with reporters.

Being on probation means the employee is in a trial period, often lasting a year or two, after starting a new government position, and does not reflect performance, the people said.
The move comes amid a broader effort under President Donald Trump to trim federal staff. The Office of Personnel Management asked federal agencies for lists of all recently hired workers because they are the easiest to terminate, NBC News has reported. That has stoked fears of layoffs at places as disparate as the Federal Bureau of Investigation and the Environmental Protection Agency.
CFPB staff have been on edge since late last week, when operatives of Elon Musk’s Department of Government Efficiency gained access to the agency. The CFPB headquarters have since been shuttered, while employees were told by acting CFPB director Russell Vought not to do any bureau work. Both Musk and Vought have called for the elimination of the CFPB.

‘First salvo’

“This is an unlawfully-executed mass firing,” said Johanna Hickman, senior CFPB litigation counsel who said she received the agency’s dismissal notice. “It’s almost certainly the first salvo in the dismantling of this agency, and a significant percentage of the federal workforce.”
Hickman, who said she started in her CFPB role in June of 2023, said the agency’s new leadership didn’t follow established federal protocol for dismissing probationary employees. “A lot of us are prepared to fight, and we are examining all our legal avenues,” she said.

The terminations have sowed more confusion at the bureau, as several of those being laid off had already accepted federal buyout offers, said one of the people.
Some being dismissed received form letters that did not include their specific names and titles, but left some fields filled with generic placeholders, said this person.
“Unfortunately, the Agency finds that you are not fit for continued employment because your ability, knowledge and skills do not meet the Agency’s current needs,” the CFPB told some who were dismissed, according to people who received the notices.
The terminations hit the CFPB’s enforcement division in particular because of a push under former director Rohit Chopra to boost hiring of enforcement lawyers, said another person. The agency had about 1,700 employees before the job cuts.
The CFPB declined to comment.

Don’t miss these insights from CNBC PRO More

It all seems to add up to something big. On a daily, sometimes hourly, basis, Elon Musk claims that his team of fiscal commandos has found yet more government fraud, terminated another wasteful contract or even scrapped an entire agency. Mr Musk’s supporters believe that, through tech wizardry and sheer willpower, he is slashing the federal deficit in a way that has eluded politicians for years. But this narrative has a glaring flaw: our review of official data shows that Mr Musk’s efforts have scarcely made a dent in spending. More