Connect with us

Published

on

Dado Ruvic | Reuters

On Monday, Chinese artificial intelligence startup DeepSeek took over rival OpenAI’s coveted spot as the most-downloaded free app in the U.S. on Apple‘s App Store, dethroning ChatGPT for DeepSeek’s AI Assistant. Global tech stocks sold off and were on pace to wipe out billions in market cap.

Later on Monday, DeepSeek said it would temporarily limit user registrations “due to large-scale malicious attacks” on its services, though existing users will be able to log in as usual.

Tech leaders, analysts, investors and developers say that the hype — and ensuing fear of falling behind in the ever-changing AI hype cycle — may be warranted. Especially in the era of the generative AI arms race, where tech giants and startups alike are racing to ensure they don’t fall behind in a market predicted to top $1 trillion in revenue within a decade.

What is DeepSeek?

DeepSeek was founded in 2023 by Liang Wenfeng, co-founder of High-Flyer, a quantitative hedge fund focused on AI. The AI startup reportedly grew out of the hedge fund’s AI research unit in April 2023 to focus on large language models and reaching artificial general intelligence, or AGI — a branch of AI that equals or surpasses human intellect on a wide range of tasks, which OpenAI and its rivals say they’re fast pursuing. DeepSeek is still wholly owned by and funded by High-Flyer, according to analysts at Jefferies.

The buzz around DeepSeek began picking up steam earlier this month, when the startup released R1, its reasoning model that rivals OpenAI’s o1. It’s open-source, meaning that any AI developer can use it, and has rocketed to the top of app stores and industry leaderboards, with users praising its performance and reasoning capabilities.

Like other Chinese chatbots, it has its limitations when asked about certain topics: When asked about some of Chinese leader Xi Jinping’s policies, for instance, DeepSeek reportedly steers the user away from similar lines of questioning.

Another key part of the discussion: DeepSeek’s R1 was built despite the U.S. curbing chip exports to China three times in three years. Estimates differ on exactly how much DeepSeek’s R1 costs, or how many GPUs went into it. Jefferies analysts estimated that a recent version had a “training cost of only US$5.6m (assuming US$2/H800 hour rental cost). That is less than 10% of the cost of Meta‘s Llama.” But regardless of the specific numbers, reports agree that the model was developed at a fraction of the cost of rival models by OpenAI, Anthropic, Google and others.

As a result, the AI sector is awash with questions, including whether the industry’s increasing number of astronomical funding rounds and billion-dollar valuations is necessary — and whether a bubble is about to burst.

Read more CNBC reporting on AI

Shares of Nvidia fell nearly 16% on Monday, with chipmaker ASML down nearly 7%. The Nasdaq dropped more than 3%. Four tech giants — Meta, Microsoft, Apple and ASML are all set to report earnings this week.

Analysts at Raymond James detailed some of the questions plaguing the AI industry this month, writing, “What are the investment implications? What does it say about open sourced vs. proprietary models? Is throwing money at GPUs really a panacea? Are U.S. export restrictions working? What are the broader implications of [DeepSeek]? Well, they could be dire, or a non-event, but rest assured, the industry is abuzz with disbelief and speculation.”

Bernstein analysts wrote in a note Monday that “according to the many (occasionally hysterical) hot takes we saw [over the weekend,] the implications range anywhere from ‘That’s really interesting’ to ‘This is the death-knell of the AI infrastructure complex as we know it.'”

How U.S. companies are responding

Some American tech CEOs are clambering to respond before clients switch to potentially cheaper offerings from DeepSeek, with Meta reportedly starting four DeepSeek-related “war rooms” within its generative AI department.

Microsoft CEO Satya Nadella wrote on X that the DeepSeek phenomenon was just an example of the Jevons paradox, writing, “As AI gets more efficient and accessible, we will see its use skyrocket, turning it into a commodity we just can’t get enough of.” OpenAI CEO Sam Altman tweeted a quote he attributed to Napoleon, writing, “A revolution can be neither made nor stopped. The only thing that can be done is for one of several of its children to give it a direction by dint of victories.”

Yann LeCun, Meta’s chief AI scientist, wrote on LinkedIn that DeepSeek’s success is indicative of changing tides in the AI sector to favor open-source technology.

LeCun wrote that DeepSeek has profited from some of Meta’s own technology, i.e., its Llama models, and that the startup “came up with new ideas and built them on top of other people’s work. Because their work is published and open source, everyone can profit from it. That is the power of open research and open source.”

Alexandr Wang, CEO of Scale AI, told CNBC last week that DeepSeek’s last AI model was “earth-shattering” and that its R1 release is even more powerful.

“What we’ve found is that DeepSeek … is the top performing, or roughly on par with the best American models,” Wang said, adding that the AI race between the U.S. and China is an “AI war.” Wang’s company provides training data to key AI players including OpenAI, Google and Meta.

Earlier this week, President Donald Trump announced a joint venture with OpenAIOracle and SoftBank to invest billions of dollars in U.S. AI infrastructure. The project, Stargate, was unveiled at the White House by Trump, SoftBank CEO Masayoshi Son, Oracle co-founder Larry Ellison and OpenAI CEO Sam Altman. Key initial technology partners will include MicrosoftNvidia and Oracle, as well as semiconductor company Arm. They said they would invest $100 billion to start and up to $500 billion over the next four years.

Why China's DeepSeek is putting America's AI lead in jeopardy

AI evolving

News of DeepSeek’s prowess also comes amid the growing hype around AI agents — models that go beyond chatbots to complete multistep complex tasks for a user — which tech giants and startups alike are chasing. Meta, Google, Amazon, Microsoft, OpenAI and Anthropic have all expressed their goal of building agentic AI.

Anthropic, the Amazon-backed AI startup founded by ex-OpenAI research executives, ramped up its technology development throughout the past year, and in October, the startup said that its AI agents were able to use computers like humans to complete complex tasks. Anthropic’s Computer Use capability allows its technology to interpret what’s on a computer screen, select buttons, enter text, navigate websites and execute tasks through any software and real-time internet browsing, the startup said.

The tool can “use computers in basically the same way that we do,” Jared Kaplan, Anthropic’s chief science officer, told CNBC in an interview at the time. He said it can do tasks with “tens or even hundreds of steps.”

OpenAI released a similar tool last week, introducing a feature called Operator that will automate tasks such as planning vacations, filling out forms, making restaurant reservations and ordering groceries.

The Microsoft-backed startup describes it as “an agent that can go to the web to perform tasks for you,” and added that it is trained to interact with “the buttons, menus, and text fields that people use daily” on the web. It can also ask follow-up questions to further personalize the tasks it completes, such as login information for other websites. Users can take control of the screen at any time.

— CNBC’s Michael Bloom contributed reporting.

Don’t miss these insights from CNBC PRO

Continue Reading

Technology

Tesla owners are trading in their EVs at record levels, Edmunds says

Published

on

By

Tesla owners are trading in their EVs at record levels, Edmunds says

A Tesla store in Alhambra, California on March 11, 2025.

Frederic J. Brown | AFP | Getty Images

As Elon Musk wraps up his second month in the White House, Tesla owners are trading in their electric vehicles at record levels, according to an analysis by national car shopping site Edmunds.

The data from Edmunds published on Thursday said that March represented “the highest ever share” it had seen for Tesla trade-ins toward new or used cars from dealerships selling other brands.

Since heading to Washington, D.C. in January as a central figure in the second Trump administration, Musk has been slashing the federal workforce and government spending, and has gained access to sensitive government computer systems and data, though his efforts have been repeatedly challenged in court.

Prior to assuming leadership of the Department of Government Efficiency (DOGE), Musk spent around $290 million last year to help propel President Donald Trump back to the White House.

While investors snapped up Tesla shares after Trump’s victory in November, they’ve been rushing for the exits of late, pushing the stock’s price down by 42% this year. Waves of protests have targeted Tesla facilities in the U.S. and beyond. Other criminal acts of vandalism and arson have targeted Tesla stores, vehicles and charging stations across the U.S.

In addition, Tesla is facing increased competition from EV makers. In January, S&P Global Mobility found Tesla sales declined about 11% year-over-year in the U.S., while Ford, Chevrolet and Volkswagen bolstered their sales of EVs, picking up market share.

“Shifts in Tesla consumer sentiment could create an opportunity for legacy automakers and EV startups to gain ground,” Jessica Caldwell, head of insights at Edmunds, wrote in an email. “As Tesla brand loyalty and interest wavers, those offering competitive pricing, new technology, or simply less controversy could capture defecting Tesla owners and first-time EV buyers.”

The Tesla brand, more than that of any other automaker, is tightly tied to its CEO. In August 2024, Edmunds surveys found that just 2% of car shoppers in the U.S. were unfamiliar with Musk.

Edmunds also said that shopping for new models of Tesla vehicles on its platform dropped to its lowest level last month since October 2022 after peaking as late as November.

Even before Musk began heading up DOGE, Tesla’s brand was suffering. Its brand value fell by 26%, or about $15 billion, in 2024, a second straight annual decline, according to research and consulting firm Brand Finance.

Many car shoppers trade in their Tesla EVs for a newer model Tesla. Edmunds data didn’t account for those transactions.

Tesla didn’t immediately respond to a request for comment.

WATCH: Tesla’s core issues more detrimental than short-term political headwinds

Tesla's core issues more detrimental than short-term political headwinds: Wells Fargo's Langan

Continue Reading

Technology

Micron shares jump on earnings beat, rosy guidance as data center revenue triples

Published

on

By

Micron shares jump on earnings beat, rosy guidance as data center revenue triples

Signage outside the Micron offices in San Jose, California, on Dec. 17, 2024.

David Paul Morris | Bloomberg | Getty Images

Micron shares popped 6% in extended trading Thursday after the company reported second-quarter results that beat analysts’ estimates and offered better-than-expected guidance.

Here’s how the company did:

  • Earnings per share: $1.56, adjusted vs. $1.42 expected by LSEG
  • Revenue: $8.05 billion vs. $7.89 billion expected by LSEG

Revenue increased 38% from $5.82 billion during the same period in 2024, Micron said in a press release. The memory and storage solutions company reported net income of $1.58 billion, or $1.41 per share, up from $793 million, or 71 cents per share, in the year-ago quarter.

Data center revenue tripled, the company said.

Revenue for the fiscal third quarter will be about $8.8 billion, Micron said, topping the $8.5 billion average analyst estimate, according to LSEG. Adjusted earnings will be roughly $1.57 a share, the company said, beating the $1.47 average estimate.

Prior to Thursday’s close, Micron shares were up 22% for the year, while the Nasdaq is down more than 8%.

Micron will host its quarterly call with investors at 4:30 p.m. ET.

Continue Reading

Technology

BlackRock’s head of digital assets says staking could be a ‘huge step change’ for ether ETFs

Published

on

By

BlackRock’s head of digital assets says staking could be a ‘huge step change’ for ether ETFs

Omar Marques | Lightrocket | Getty Images

Appetite for ether ETFs has been tepid since their launch last July, but that could change if some of the regulatory wrinkles holding them back get “resolved,” according to Robert Mitchnick, head of digital assets at BlackRock.

There’s a widely held view that the success of ether ETFs has been “meh” compared to the explosive growth in funds tracking bitcoin, Mitchnick said at the Digital Asset Summit in New York City Thursday. Though he sees that as a “misconception,” he acknowledged that the inability to earn a staking yield on the funds is likely one thing holding them back.

“There’s obviously a next phase in the potential evolution of [ether ETFs],” he said. “An ETF, it’s turned out, has been a really, really compelling vehicle through which to hold bitcoin for lots of different investor types. There’s no question it’s less perfect for ETH today without staking. A staking yield is a meaningful part of how you can generate investment return in this space, and all the [ether] ETFs at launch did not have staking.”

Staking is a way for investors to earn passive yield on their cryptocurrency holdings by locking tokens up on the network for a period of time. It allows investors to put their crypto to work if they’re not planning to sell it anytime soon.

But Mitchnick doesn’t expect a simple fix.

“It’s not a particularly easy problem,” he explained. “It’s not as simple as … a new administration just green-lighting something and then boom, we’re all good, off to the races. There are a lot of fairly complex challenges that have to be figured out, but if that can get figured out, then it’s going to be sort of a step change upward in terms of what we see the activity around those products is.”

The Securities and Exchange Commission has historically viewed some staking services as potential unregistered securities offerings under the Howey Test – which is used to determine whether an asset is an investment contract and therefore, a security. But a more crypto friendly SEC is moving swiftly to reverse the damage done to the industry under the previous regime. Its newly formed crypto task force is scheduled to kick off a roundtable series Friday focused on defining the security status of digital assets.

Ether has been one of the most beaten up cryptocurrencies in recent months. It’s down more than 40% year to date as it has struggled with conflicting and difficult-to-comprehend narratives, weaker revenue since its last big technical upgrade and increasing competition from Solana. Standard Chartered this week slashed its price target on the coin by more than half.

Mitchnick said the negativity is “overdone.”

“ETH … at the second grade level is easier to define … but at the 10th grade level is a lot harder,” he said. “Second grade level: it’s a technology innovation story. … Beyond that, it does get a little more vast, a little more complicated. It’s about being a bet on blockchain adoption and innovation. That’s part of the thesis as we communicate it to clients.”

“There are three [use cases] that we focus on that have a lot of resonance with our client base: it’s a bet to some extent on tokenization, on stablecoin adoption, and on decentralized financing,” he added. “It does take a fair bit of education, and we’ve been on that journey, but it’s going to take more time.”

BlackRock is the issuer of the iShares Ethereum Trust ETF. It also has a tokenized money market fund, known as BUIDL, which it initially launched a year ago on Ethereum and has since expanded to several other networks including Aptos and Polygon.

Don’t miss these cryptocurrency insights from CNBC Pro:

Continue Reading

Trending