Connect with us

Published

on

Qwen3 is Alibaba’s debut into so-called “hybrid reasoning models,” which it says combines traditional LLM capabilities with “advanced, dynamic reasoning.”

Sopa Images | Lightrocket | Getty Images

Alibaba released the next generation of its open-sourced large language models, Qwen3, on Tuesday — and experts are calling it yet another breakthrough in China’s booming open-source artificial intelligence space.

In a blog post, the Chinese tech giant said Qwen3 promises improvements in reasoning, instruction following, tool usage and multilingual tasks, rivaling other top-tier models such as DeepSeek’s R1 in several industry benchmarks. 

The LLM series includes eight variations that span a range of architectures and sizes, offering developers flexibility when using Qwen to build AI applications for edge devices like mobile phones.

Qwen3 is also Alibaba’s debut into so-called “hybrid reasoning models,” which it says combines traditional LLM capabilities with “advanced, dynamic reasoning.”

According to Alibaba, such models can seamlessly transition between a “thinking mode” for complex tasks such as coding and a “non-thinking mode” for faster, general-purpose responses. 

“Notably, the Qwen3-235B-A22B MoE model significantly lowers deployment costs compared to other state-of-the-art models, reinforcing Alibaba’s commitment to accessible, high-performance AI,” Alibaba said. 

The new models are already freely available for individual users on platforms like Hugging Face and GitHub, as well as Alibaba Cloud’s web interface. Qwen3 is also being used to power Alibaba’s AI assistant, Quark.

China’s AI advancement

AI analysts told CNBC that the Qwen3 represents a serious challenge to Alibaba’s counterparts in China, as well as industry leaders in the U.S.  

In a statement to CNBC, Wei Sun, principal analyst of artificial intelligence at Counterpoint Research, said the Qwen3 series is a “significant breakthrough—not just for its best-in-class performance” but also for several features that point to the “application potential of the models.” 

Those features include Qwen3’s hybrid thinking mode, its multilingual support covering 119 languages and dialects and its open-source availability, Sun added.

Open-source software generally refers to software in which the source code is made freely available on the web for possible modification and redistribution. At the start of this year, DeepSeek’s open-sourced R1 model rocked the AI world and quickly became a catalyst for China’s AI space and open-source model adoption.  

“Alibaba’s release of the Qwen 3 series further underscores the strong capabilities of Chinese labs to develop highly competitive, innovative, and open-source models, despite mounting pressure from tightened U.S. export controls,” said Ray Wang, a Washington-based analyst focusing on U.S.-China economic and technology competition.

According to Alibaba, Qwen has already become one of the world’s most widely adopted open-source AI model series, attracting over 300 million downloads worldwide and more than 100,000 derivative models on Hugging Face. 

Wang said that this adoption could continue with Qwen3, adding that its performance claims may make it the best open-source model globally — though still behind the world’s most cutting-edge models like OpenAI’s o3 and o4-mini.  

Chinese competitors like Baidu have also rushed to release new AI models after the emergence of DeepSeek, including making plans to shift toward a more open-source business model. 

Meanwhile, Reuters reported in February that DeepSeek is accelerating the launch of its successor to its R1, citing anonymous sources.

“In the broader context of the U.S.-China AI race, the gap between American and Chinese labs has narrowed—likely to a few months, and some might argue, even to just weeks,” Wang said. 

“With the latest release of Qwen 3 and the upcoming launch of DeepSeek’s R2, this gap is unlikely to widen—and may even continue to shrink.”

Continue Reading

Technology

Night owl bitcoin traders: Soon there’ll be an ETF just for you

Published

on

By

Night owl bitcoin traders: Soon there'll be an ETF just for you

Cheng Xin | Getty Images

A newly proposed exchange-traded fund would offer exposure to bitcoin, much like other popular ETFs tracking the world’s oldest cryptocurrency. But, there’s a twist: The fund would trade bitcoin-linked assets while Wall Street sleeps. 

The Nicholas Bitcoin and Treasuries AfterDark ETF aims to purchase bitcoin-linked financial instruments after the U.S. financial markets close, and exit those positions shortly after the U.S. market re-opens each day, according to a December 9 filing to the Securities and Exchange Commission.

The fund would not hold bitcoin directly. Instead, the AfterDark ETF would use at least 80% of the value of its assets to trade bitcoin futures contracts, bitcoin exchange-traded products and ETFs, and options on those ETFs and ETPs. 

The offering would capitalize on bitcoin’s outsized gains in off-hours trading.

Hypothetically, an investor who had been buying shares of the iShares Bitcoin Trust ETF (IBIT) when U.S. markets formally close, and selling them at the next day’s open, would have scored a 222% gain since January 2024, data from wealth manager Bespoke Investment Group shows. But an investor that had bought IBIT shares at the open and sold them at the close would have lost 40.5% in the same time.

Bitcoin was last trading at $92,320, down nearly 1% on the day. The leading cryptocurrency is down about 12% over the past month and little changed since the beginning of the year. 

The proposed ETF underscores jockeying among sponsors to launch ETFs tracking all kinds of cryptocurrencies, from altcoins like Aptos and Sui to memecoins such as Bonk and Dogecoin. The contest has only accelerated under President Donald Trump, who has pushed the SEC and Commodity Futures Trading Commission to soften their stances on token issuers and digital asset exchanges. 

Since being approved under the prior administration in January 2024, more than 30 bitcoin ETFs have begun trading in the U.S., according to data from ETF.com.

Continue Reading

Technology

Cisco’s stock closes at record for first time since dot-com peak in 2000

Published

on

By

Cisco's stock closes at record for first time since dot-com peak in 2000

Chuck Robbins, chief executive officer of Cisco, participates in a Bloomberg interview at the World Economic Forum in Davos, Switzerland, on Jan. 17, 2024.

Stefan Wermuth | Bloomberg | Getty Images

Few companies were as hot in early 2000 as Cisco, whose networking equipment served as the backbone of the internet boom.

On Wednesday, Cisco’s stock surpassed its dot-com peak for the first time. The shares rose almost 1% to $80.25, topping their prior split-adjusted record or $80.06 reached on March 27, 2000. That’s the same day that Cisco passed Microsoft to become the most valuable publicly traded company in the world.

Back then, investors saw Cisco as a way to bet on the growth of the web, as companies that wanted to get online relied upon the hardware maker’s switches and routers. But following a half-decade boom, the dot-com bubble burst just after Cisco reached its zenith, a collapse that wiped out more than three-quarters of the Nasdaq’s value by October 2002.

While the market swoon eliminated scores of internet highflyers, Cisco survived the upheaval. Eventually it started to grow and expand, diversifying through a series of acquisitions like set-top box maker Scientific- Atlanta in 2006, followed by software companies including Webex, AppDynamics, Duo and Splunk.

With its gains on Wednesday, Cisco’s market cap sits at $317 billion, making it only the 13th most valuable U.S. tech company. In recent years, the stock has badly trailed tech’s megacaps, which have been at the center of the new boom surrounding artificial intelligence.

The AI market has reached a level of euphoria that many analysts have compared to the dot-com era. Instead of Cisco, the modern infrastructure winner is Nvidia, whose AI chips are at the heart of model development and are relied up by the other major tech companies that are all building out AI-focused data centers. Nvidia has a market cap of $4.5 trillion, roughly 14 times Cisco’s current value.

But Cisco is angling to benefit from the AI craze, with CEO Chuck Robbins in November touting $1.3 billion in quarterly AI infrastructure orders from large web companies. Total revenue approached $15 billion, which was up 7.5% year over year, compared with 66% growth in 2000.

Shares of Cisco are up about 36% so far in 2025, outperforming the Nasdaq, which has gained about 22% over the same period.

WATCH: Cisco CEO on latest quarter: AI demand from hyperscalers is accelerating

Cisco CEO on latest quarter: AI demand from hyperscalers is accelerating

Continue Reading

Technology

Oracle set to report quarterly results after the bell

Published

on

By

Oracle set to report quarterly results after the bell

Larry Ellison, Oracle’s co-founder and chief technology officer, appears at the Formula One British Grand Prix in Towcester, U.K., on July 6, 2025.

Jay Hirano | Sopa Images | Lightrocket | Getty Images

Oracle is scheduled to report fiscal second-quarter results after market close on Wednesday.

Here’s what analysts are expecting, according to LSEG:

  • Earnings per share: $1.64 adjusted
  • Revenue: $16.21 billion

Wall Street expects revenue to increase 15% in the quarter that ended Nov. 30, from $14.1 billion a year earlier. Analysts polled by StreetAccount are looking for $7.92 billion in cloud revenue and $6.06 billion from software.

The report lands at a critical moment for Oracle, which has tried to position itself at the center of the artificial intelligence boom by committing to massive build-outs. While the move has been a boon for Oracle’s revenue and its backlog, investors have grown concerned about the amount of debt the company is raising and the risks it faces should the AI market slow.

The stock plummeted 23% in November, its worst monthly performance since 2001 and, as of Tuesday’s close, is 33% below its record reached in September. Still, the shares are up 33% for the year, outperforming the Nasdaq, which has gained 22% over that stretch.

Over the past decade, Oracle has diversified its business beyond databases and enterprise software and into cloud infrastructure, where it competes with Amazon, Microsoft and Google. Those companies are all vying for big AI contracts and are investing heavily in data centers and hardware necessary to meet expected demand.

OpenAI, which sparked the generative AI rush with the launch of ChatGPT three years ago, has committed to spending more than $300 billion on Oracle’s infrastructure services over five years.

“Oracle’s job is not to imagine gigawatt-scale data centers. Oracle’s job is to build them,” Larry Ellison, the company’s co-founder and chairman, told investors in September.

Oracle raised $18 billion during the period, one of the biggest issuances on record for a tech company. Skeptical investors have been buying five-year credit default swaps, driving them to multiyear highs. Credit default swaps are like insurance for investors, with buyers paying for protection in case the borrower can’t repay its debt.

“Customer concentration is a major issue here, but I think the bigger thing is, How are they going to pay for this?” said RBC analyst Rishi Jaluria, who has the equivalent of a hold rating on Oracle’s stock.

During the quarter, Oracle named executives Clay Magouyrk and Mike Sicilia as the company’s new CEOs, succeeding Safra Catz. Oracle also introduced AI agents for automating various facets of finance, human resources and sales.

Executives will discuss the results and issue guidance on a conference call starting at 5 p.m. ET.

WATCH: Oracle’s debt concerns loom large ahead of quarterly earnings

Oracle's debt concerns loom large ahead of quarterly earnings

Continue Reading

Trending