Connect with us

Published

on

In this article

Electric vehicle maker Tesla rolled out a long-awaited software update Friday night that allows customers to request access to its controversial Full Self-Driving Beta (FSD Beta) software.

The move delighted fans of CEO Elon Musk and Tesla, but it risks drawing the ire of federal vehicle safety authorities who are already investigating the automaker for possible safety defects in its driver-assistance systems.

FSD Beta is an unfinished version of Tesla’s premium driver-assistance software, FSD, which the company sells in the U.S. for $10,000 upfront, or $199 a month.

FSD is marketed with the promise of enabling a Tesla to automatically change lanes, navigate on the highway, move into a parking spot, or roll out from a parking spot to drive a small distance at a slow pace without anyone behind the wheel.

FSD Beta gives drivers access to an “autosteer on city streets” feature, which has yet to be perfected and enables drivers to automatically navigate around urban environments alongside other vehicles, pedestrians, bicyclists and pets without moving the steering wheel with their own hands. Drivers are supposed to remain attentive, however, with both hands on the wheel and prepared to take over driving at any time.

None of Tesla’s driver assistance systems — including the company’s standard Autopilot package, premium Full Self-Driving option, or FSD Beta — make Teslas autonomous.

The company previously made FSD Beta available to about 2,000 people, a mix of mostly employees and some customers, who test it out on public roads even though the software hasn’t been debugged.

The new download button could ostensibly lead to a rapid expansion in the number of participants who are not trained regulatory officials.

Government response

Tesla CEO Elon Musk gestures as he visits the construction site of Tesla’s Gigafactory in Gruenheide near Berlin, Germany, August 13, 2021.
Patrick Pleul | Reuters

Last week, when CEO Musk announced new details about the FSD beta button, Jennifer Homendy, the head of the National Transporation Safety Board, voiced concern over the company’s plans in an interview with The Wall Street Journal.

Homendy said, “Basic safety issues have to be addressed,” before Tesla expands FSD Beta to other city streets and regions. The NTSB chief was also displeased that the company was conducting testing of the unfinished product with untrained drivers on public roads in lieu of safety professionals.

Homendy also remarked — and in interviews with Autonocast, an industry podcast, and the Washington Post — that Tesla’s use of the term Full Self-Driving for a “level 2” driver assistance system is misleading and confusing.

Musk himself said last week in a tweet that FSD Beta now seems so good it can give drivers the wrong idea that they don’t need to pay attention to driving while FSD Beta is engaged, even though they are supposed to remain attentive and at the wheel at all times.

On Saturday, after Tesla enabled the “request full self-driving beta” feature in its vehicles — a fan blog named Teslarati shared a post on Twitter asking, “Does Tesla have a fair chance after NTSB Chief comments?”

Musk replied to them on Twitter with a link to the Wikipedia biography of Homendy. While Musk has previously urged his tens of millions of followers on Twitter to alter a description of his own career on Wikipedia, he shared this link to Homendy’s bio there without comment.

CNBC reached out to Tesla and the NTSB — neither was immediately available to comment on Saturday.

Safety score

Musk has been promising Tesla owners an FSD beta download button for months. In March 2021, he wrote in a tweet that the forthcoming button would give users access to the latest FSD Beta build as soon as their car connected to Wi-Fi.

He changed that approach, however. Now, Tesla has a calculator it uses to give drivers a “safety score,” and determine who will be allowed to get and use FSD Beta software.

Screen-shots shared with CNBC by Tesla owners with FSD indicate that the company’s “safety score” is akin to an insurance risk factor score.

Tesla’s systems tabulate a drivers’: “Predicted Collision Frequency, Forward Collision Warning per 1,000 Miles, Hard Braking, Aggressive Turning, Unsafe Following Time, and Forced Autopilot Disengagements,” according to correspondence and screenshots viewed by CNBC.

Tesla’s system does not, at this time, appear to measure and account for how often drivers fail to keep their hands on the wheel, how quickly they take over driving when prompted, or how consistently they keep their eyes on the road.

Only users who have a great driving record for a full week, in Tesla’s view, may gain access to FSD Beta.

Before Tesla released its FSD Beta button (and the 10.1 version of FSD Beta, which is expected this weekend, too) CNBC asked the California DMV Autonomous Vehicles Branch how pervasive and safe FSD Beta-equipped vehicles have been in use in the state so far.

The DMV declined an interview request but said, in an e-mailed statement:

“Based on information the information Tesla has provided the DMV, the feature does not make the vehicle an autonomous vehicle per California regulations. The DMV continues to gather information from Tesla on its beta release – including any expansion of the program and features.  If the capabilities of the feature change such that it meets the definition of an autonomous vehicle per California’s law and regulations, Tesla will need to operate under the appropriate regulatory authorization. Regardless of the level of vehicle autonomy, the DMV has reminded Tesla that clear and effective communication to the driver about the technology’s capabilities, limitations and intended use is necessary. The DMV is reviewing the company’s use of the term ‘Full Self-Driving’ for its technology. Because it is ongoing, the DMV cannot discuss the review until it is complete.”

Continue Reading

Technology

Alibaba shares rise as AI drives 34% cloud sales jump

Published

on

By

Alibaba shares rise as AI drives 34% cloud sales jump

Alibaba showcase its AI technology application achievements from Alibaba Cloud at the World Artificial Intelligence Conference in Shanghai, China on July 26, 2025.

Cfoto | Future Publishing | Getty Images

Alibaba delivered better than expected revenue in its fiscal second quarter as sales in its key cloud computing division accelerated.

Alibaba’s New York-listed shares were around 4.3% higher in premarket trade as investors looked past a plunge in profitability.

Here’s how the company did in its fiscal second quarter ended Sept. 30 versus LSEG estimates:

  • Revenue rose 5% to 247.8 billion Chinese yuan ($34.8 billion) versus 242.65 billion yuan the previous year.

Investors are focused on Alibaba’s cloud computing division which books its revenue related to artificial intelligence. Over the past few quarters, Alibaba’s cloud revenue growth has accelerated.

Alibaba reported a 34% year-on-year rise in cloud computing revenue to 39.8 billion yuan versus expectations of 37.9 billion yuan. That growth rate was faster than the 26% notched in the June quarter.

The Chinese tech giant said its investments in AI were helping its cloud unit.

“Robust AI demand further accelerated our Cloud Intelligence Group business, with revenue up 34% and AI-related product revenue achieving triple-digit year-over-year growth for the ninth consecutive quarter,” CEO Eddie Wu said in an earnings statement on Tuesday.

How Alibaba quietly became a leader in AI

In September, the company said it plans to increase spending on AI models and infrastructure development, on top of the 380 billion yuan ($53 billion) over three years it announced in February. Alibaba said on Tuesday it has spent around 120 billion yuan in capital expenditure toward AI and cloud infrastructure over the past four quarters.

Earnings before interest, taxes, and amortization (EBITA), a measure of profitability, increased by 35% to 3.6 billion yuan for its cloud division.

Alibaba has emerged as one of China’s leading AI players. On Monday, Alibaba said its Qwen app, the Chinese giant’s rival to OpenAI’s ChatGPT, surpassed 10 million downloads within the first week of its public launch. The app is powered by Alibaba’s Qwen artificial intelligence models.

Investors look past profit drop

Meanwhile, the company has been investing heavily in the cut-throat instant commerce market. This a product offering from Alibaba and some of its Chinese e-commerce rivals that promises super-fast delivery on certain items.

Investment in this new segment has weighed on the profitability of Alibaba’s overall business even as cloud computing remains strong.

Overall adjusted EBITA, a profitability measure closely-watched by analysts, fell 78% year-on-year to 9.1 billion yuan, with Alibaba attributing this partly to its investments in quick commerce.

But investors appear to be looking past this because of the growth acceleration at the cloud computing business and Alibaba’s core China e-commerce division which houses revenue from its online shopping platforms Taobao and Tmall as well as the quick commerce initiative. China e-commerce revenue rose 16% year-on-year to 132.6 billion yuan, with growth coming in faster than the previous quarter.

Revenue from quick commerce surged 60% year-on-year in the quarter versus 12% in the quarter before.

“In our consumption business, quick commerce continued to scale with significant improvement in unit economics and drove rapid growth in monthly active consumers on the Taobao app,” Wu said.

Continue Reading

Technology

Nvidia shares fall 3% on report Meta will use Google AI chips

Published

on

By

Nvidia shares fall 3% on report Meta will use Google AI chips

Jensen Huang, NVIDIA founder and CEO, has a Q&A session at a press conference during the APEC CEO summit on October 31, 2025 in Gyeongju, South Korea.

Woohae Cho | Getty Images News | Getty Images

Nvidia shares fell on Tuesday after The Information reported that Meta is considering using chips designed by Google.

Shares of Nvidia were 3.6% lower in premarket trade. Google-parent Alphabet was trading 3% higher after a more than 6% rally on Monday.

On Monday, The Information reported that Meta is considering using Google’s tensor processing units (TPUs) in its data centers in 2027. Meta may also rent TPUs from Google’s cloud unit next year, the publication reported.

Google launched its first-generation TPU in 2018 and it was initially designed for its own internal use for its cloud computing business. Since then, Google has launched more advanced versions of its chip that are designed to handle artificial intelligence workloads.

TPUs are a customized chip and experts say this gives Google an advantage over rivals as it can offer customers a highly efficient product for AI.

If Meta uses the TPUs, it would be big win for Google and potential validation of the technology.

Shares of Broadcom, which helps Google design its TPUs, were up more than 2% in premarket trade on Tuesday after an 11% rise the day before.

Meta reportedly in talks to use Google's AI chips

Nvidia remains the market leader with its graphics processing units (GPUs) that have become the main piece of hardware underpinning the huge AI infrastructure buildout. While Nvidia’s dominance is unlikely to be dislodged in the near term, Google’s TPUs add further competition into the AI semiconductor market.

Companies building AI infrastructure have been searching for a more diversified supply of chips to reduce reliance on Nvidia.

Meta is among the biggest spenders on AI infrastructure, with the company projecting its capital expenditure to stand between $70 billion to $72 billion this year.

The share price moves come amid continued debate around whether there is an “AI bubble” and stretched tech company valuations.

Nvidia has been central to the debate and the company last week reported a stronger-than-expected sales forecast for the current quarter but technology stocks fell after.

Continue Reading

Technology

CNBC Daily Open: Alphabet to omega in AI?

Published

on

By

CNBC Daily Open: Alphabet to omega in AI?

A Google logo is at the announcement of Google’s biggest-ever investment in Germany on November 11, 2025 in Berlin, Germany.

Sean Gallup | Getty Images News | Getty Images

Alphabet on Monday resuscitated the artificial intelligence trade, which had been flagging the previous week. Its stock jumped 6.3%, lifting associated AI names such as Broadcom, Micron Technology and AMD. Major indexes rallied, with the Nasdaq Composite posting its best day in six months.

Investors were particularly enthusiastic about Broadcom because it helps to design and manufacture Google-parent Alphabet’s custom AI chips. In other words, the more market share Alphabet’s AI offerings gain, the greater the benefit to Broadcom — rather like Nvidia and the broader AI sector at the moment. Broadcom shares surged 11.1% on this notion, making it the S&P 500’s top gainer.

But while investors may cheer Alphabet’s leadership on Monday, not everyone wants it to have the last word.

“Some investors are petrified that Alphabet will win the AI war due to huge improvements in its Gemini AI model and ongoing benefits from its custom TPU chip,” Melius Research analyst Ben Reitzes wrote to clients in a Monday note. “GOOGL winning would actually hurt several stocks we cover — so prepare for volatility.”

Approaching the market’s moves from another angle, Melissa Brown, managing director of investment decision research at SimCorp, said it’s a concern when just one stock lifts the market. “That just doesn’t seem to me to be a sustainable force behind driving the market higher over the next however many days,” she added.

Alphabet on Monday may have brought about alpha — in the sense of market outperformance and potentially beginning a new phase of AI enthusiasm — but letting it be the omega as well could pose problems for investors.

What you need to know today

U.S. tech stocks roar back. The Nasdaq Composite popped 2.69%, its best day since May 12, on investors enthusiasm over Alphabet. Other major indexes rose in tandem. Asia-Pacific markets were mostly Tuesday as AI-related stocks ticked up.

Record outflows from BlackRock’s bitcoin ETF. The iShares Bitcoin Trust ETF has seen an exodus of $2.2 billion this month as of Monday stateside, according to FactSet data. That’s almost eight times more in losses than last October, or its second-worst month on record.

Sandisk joins the S&P 500. The flash storage vendor will replace marketing company Interpublic Group in the index before trading begins on Nov. 28 stateside. Shares of Sandisk jumped 7% in extended trading on Monday.

Trump has back-to-back calls with Xi and Takaichi. But the Beijing-Tokyo spat is unlikely to be resolved soon. U.S. President Donald Trump has stayed publicly silent, adding uncertainty for Japan and Taiwan at a tense moment. 

[PRO] The S&P 500’s dividend yield is looking dismal. For investors who are still looking to hold dividend-paying stocks, however, research firm Trivariate Research has a few suggestions on the top performers.

And finally…

MUMBAI, INDIA – OCTOBER 22: Executive chair at the South Korean automaker Hyundai Motor Group Euisun Chung and managing director and CEO at India’s National Stock Exchange (NSE) Ashish Kumar Chauhan and Jaehoon Chang, Chief Executive Officer (CEO) and President of Hyundai Motor Company pose for a photo during the listing ceremony of Hyundai Motor India for its initial public offering (IPO) at the NSE in Mumbai, India on October 22, 2024.

Anadolu | Anadolu | Getty Images

Continue Reading

Trending