Connect with us

Published

on

Muhammed Selim Korkutata | Anadolu | Getty Images

Elon Musk’s xAI on Tuesday unveiled its latest artificial intelligence model, Grok 3, claiming it can outperform offerings from OpenAI and China’s DeepSeek based on early testing, which included standardized tests on math, science and coding. 

“We’re very excited to present Grok 3, which is, we think, an order of magnitude more capable than Grok 2 in a very short period of time,” Musk said at a demonstration of Grok 3 that was streamed on his social media platform X. 

The team also said it was launching a new product called “Deep Search,” which would act as a “next generation search engine.” 

Grok 3 will be rolled out for premium X subscribers later in the day, and will also be accessible through a separate subscription for the model’s web and app versions, the xAI team said.

Speaking at The World Governments Summit in Dubai last week Musk had dubbed the model “scary smart,” with powerful reasoning capabilities, claiming it outperformed all other existing models in xAI’s internal tests. 

“This might be the last time that an AI is better than Grok,” Musk said at the time, adding that it was trained on “a lot of synthetic data,” and was capable of reflecting upon its mistakes to achieve logical consistency. 

The xAI team claimed that an early iteration of Grok 3 had been given better ratings than existing competitors on Chatbot Arena, a crowdsourced website that pits different AI models against each other in blind tests.

Toward the end of the product demo, Musk said that the company will keep improving the model.

“We should emphasize that this is kind of a beta, meaning that you should expect some imperfections at first, but we will improve it rapidly, almost every day,” he said, adding that the voice assistance for the model would be released at a later time.

Intense competition 

Musk, who has been quite vocal about the potential dangers of artificial intelligence, started xAI in 2023 entering the generative AI market that includes OpenAI’s ChatGPT. 

In September last year, OpenAI launched its most advanced model, the o1, which came with reasoning abilities and was able to solve relatively complex science, coding and math tasks. 

Musk, along with Sam Altman, helped create OpenAI as a nonprofit in 2015.

However, in recent years Musk and OpenAI’s leadership have been feuding. Musk recently led an investor group that submitted a proposal to buy the AI startup’s nonprofit parent for $97.4 billion — an offer OpenAI declined. 

Last month, Chinese start-up DeepSeek shocked the AI market when it released a technical paper that claimed one of its open source models was able to rival the performance of OpenAI’s o1 model despite using a cheaper, less energy-intensive process.

It accomplished the feat in the face of the U.S. restricting leading AI chipmaker Nvidia from selling its cutting-edge GPUs — used for training AI models — to China. 

XAI has a “Colossus supercomputer,” for training AI, which it said last year was utilizing a cluster of 100,000 advanced Nvidia GPUs for AI training. On Tuesday, the company revealed that it doubled the size of its GPU cluster for the training of Grok 3.

While many AI and tech experts have told CNBC that DeepSeek has intensified AI competition, showing what can be done with less advanced technology, others are more skeptical about its impact. 

Continue Reading

Technology

U.S. House tells staffers not to use Meta’s WhatsApp

Published

on

By

U.S. House tells staffers not to use Meta’s WhatsApp

A woman walks past a logo of WhatsApp during a Meta event in Mumbai, India, on Sept. 20, 2023.

Niharika Kulkarni | Nurphoto | Getty Images

Meta is pushing back against a ban on WhatsApp from government devices.

The chief administrative officer, or CAO, of the U.S. House of Representatives told staffers on Monday that they are not allowed to use Meta’s popular messaging app. The CAO cited a lack of transparency about WhatsApp’s data privacy and security practices as the reason for the ban, according to a report by Axios that cited an internal email from the government office.

The CAO told House staff members in the email that they are not allowed to download WhatsApp on their government devices or access the app on their smartphones or desktop computers, the report said. Staff members must remove WhatsApp from their devices if they have the app installed on their devices, the report said.

“Protecting the People’s House is our topmost priority, and we are always monitoring and analyzing for potential cybersecurity risks that could endanger the data of House Members and staff,” U.S. House Chief Administrative Officer Catherine Szpindor told CNBC in a written statement.

Meta spokesperson Andy Stone on Monday responded to the report via a post on X, saying the company disagrees “with the House Chief Administrative Officer’s characterization in the strongest possible terms.”

“We know members and their staffs regularly use WhatsApp and we look forward to ensuring members of the House can join their Senate counterparts in doing so officially,” Stone said.

In a separate X post, Stone said WhatsApp’s encrypted nature provides a “higher level of security than most of the apps on the CAO’s approved list that do not offer that protection.”

Some of the messaging apps the CAO said are acceptable alternatives to WhatsApp include Microsoft Teams, Signal and Apple’s iMessage, the Axios report said.

Meta is currently embroiled in an antitrust case with the Federal Trade Commission over the social media company’s acquisitions of WhatsApp and Instagram.

Last week, Meta debuted ads in WhatsApp in an effort to monetize the app that CEO Mark Zuckerberg has deemed “the next chapter” for his company’s history.

Don’t miss these insights from CNBC PRO

Watch CNBC's full interview with Meta CTO Andrew Bosworth

Continue Reading

Technology

Super Micro shares fall on planned $2 billion convertible debt offering

Published

on

By

Super Micro shares fall on planned  billion convertible debt offering

The Super Micro Computer headquarters in San Jose, California, on Dec. 3, 2024.

David Paul Morris | Bloomberg | Getty Images

Super Micro Computer shares fell about 6% on Monday after the server maker said it plans to offer $2 billion in convertible notes, maturing in 2030.

A company’s stock often falls on the announcement of a convertible offering because the eventual conversion to equity could dilute existing shareholders’ stakes.

Super Micro, which has seen its business boom due to soaring demand for Nvidia’s artificial intelligence processors, said in a press release that it plans to use the proceeds from the offering for “general corporate purposes, including to fund working capital for growth and business expansion.” It also said it would spend about $200 million to repurchase its stock from the note issuers.

Even after Monday’s slide, Super Micro shares are up close to 40% so far in 2025 as the company remains one of a handful of server makers that can sell systems based around new chips from Nvidia, Advanced Micro Devices, and Intel soon after they start shipping. The stock has been viewed by Wall Street as an AI pure play that will appreciate with tech megacap companies expected to spend hundreds of billions of dollars on data centers to support AI workloads.

Super Micro also secured a major contract with a data center in Saudi Arabia when President Donald Trump visited the Middle East in May.

Super Micro “has emerged as a market leader in AI-optimized infrastructure,” Raymond James analysts wrote in a report last month, saying that 70% of the company’s revenue was attributable to AI. The analysts recommend buying the stock.

Investors soured on Super Micro in March and April on concerns about tariffs, and in May the company slashed its fiscal 2025 guidance and chose not to reiterate its previous forecast for $40 billion in fiscal 2026 sales, due to tariff and AI chip uncertainty.

The stock has recouped some of those losses but is still trading well below its high for the year reached in February.

Super Micro had a tumultuous 2024 largely because of accusations of accounting irregularities, and was forced to refile financials with the SEC in order to avoid delisting from the Nasdaq. Super Micro also named a new auditor, removed its CFO and named additional members to its board of directors.

WATCH: The bull case for Super Micro

Raymond James' Simon Leopold talks the bull case for Super Micro

Continue Reading

Technology

Amazon launches second batch of Kuiper internet satellites, taking on Elon Musk’s Starlink

Published

on

By

Amazon launches second batch of Kuiper internet satellites, taking on Elon Musk's Starlink

An Atlas V rocket of United Launch Alliance (ULA) lifts off from Space Launch Complex 41 at the Kennedy Space Center in Cape Canaveral, Florida on June 23, 2025.

Gregg Newton | Afp | Getty Images

Amazon‘s second batch of Kuiper internet satellites reached low Earth orbit on Monday, adding to its plans for a massive constellation and ramping up competition with SpaceX’s Starlink.

A United Launch Alliance rocket carrying 27 Kuiper satellites lifted off from a launchpad at the Cape Canaveral Space Force Station in Florida at 6:54 a.m. ET, according to a livestream.

“We have ignition and lift off of United Launch Alliance Atlas V rocket carrying satellites for Amazon’s Project Kuiper internet constellation, continuing a new chapter in low Earth orbit satellite connectivity,” Ben Chilton, an ordnance engineer at ULA, said on the livestream following the launch.

Monday’s mission was rescheduled twice, owing to inclement weather and a problem with the rocket booster.

Read more CNBC Amazon coverage

Six years ago, Amazon unveiled its plans to build a constellation of internet-beaming satellites in low Earth orbit, called Project Kuiper. The service will compete directly with Elon Musk’s Starlink, which currently dominates the market and has 8,000 satellites in orbit.

Amazon in April successfully sent up 27 Kuiper internet satellites into low Earth orbit, a region of space that’s within 1,200 miles of the Earth’s surface.

The 54 craft currently in orbit are the start of Amazon’s planned constellation of 3,236 satellites. The company has to meet a Federal Communications Commission deadline to launch half of its total constellation, or 1,618 satellites, by July 2026.

The company has booked more than 80 launches with several providers, including rival SpaceX, to deliver Kuiper its satellites into orbit.

WATCH: Amazon Web Services CEO: Lots of opportunity to expand infrastructure globally

AWS CEO: Lots of opportunity to expand infrastructure globally

Continue Reading

Trending