Latest news with #Llama
CNBC
9 hours ago
- Business
- CNBC
Meta's AI spending spree is Wall Street's focus in second-quarter earnings
Over the years, Meta has built a reputation for using rivals' innovations to bolster its technology. But its decision to copy a Chinese artificial intelligence lab in 2025 in an effort to compete with OpenAI backfired, forcing the company to overhaul its AI strategy. In a rush to mimic the techniques developed by Chinese startup DeepSeek, Meta released a new version of its Llama family of AI models that disappointed third-party developers, according to people familiar with the matter. The reaction was so bad that CEO Mark Zuckerberg decided to spend billions of dollars to revamp the company's AI unit, and he's still considering more shake-ups to Meta's AI strategy, said the people, who asked not to be named due to confidentiality. When Meta reports second-quarter earnings on Wednesday, Zuckerberg will make the case to investors for his AI hiring spree and the company's related strategy shift. Meta's AI blitzkrieg kicked off in June, when it invested $14.3 billion into Scale AI, resulting in the data annotating startup's CEO, Alexandr Wang, joining Meta along with a handful of employees to oversee a cornerstone AI unit. This new Meta Superintelligence Labs will be led by Wang, now Meta's chief AI officer, and former GitHub CEO Nat Friedman, who also joined the company in June along with business partner Daniel Gross. Gross was previously the CEO of AI startup Safe Superintelligence, which Meta tried to buy before being rebuffed by co-founder Ilya Sutskever. While Meta couldn't land the AI pioneer and former OpenAI co-founder, it did hire multiple top researchers from competitors like the ChatGPT maker, Apple and Google to help it regain its footing in the fiercely competitive artificial intelligence market. One such hire was ChatGPT co-creator Shengjia Zhao, who Zuckerberg last week named as his new AI lab's chief scientist. Although Meta's AI talent grab may not result in the company raising its projection for 2025 total expenses, estimated to come in between $113 billion to $118 billion, Cantor analysts said in a note published earlier this month that the investment potentially "moves the target above the low end." Translation: all that hiring comes with a cost, albeit slight. Meanwhile, revenue growth in the second quarter likely slowed to 15%, down from 22% a year earlier, according to LSEG. It would be the slowest rate of expansion for the company since early 2023, and analysts are expecting lower levels of growth in the coming quarters. Zuckerberg believes that new AI talent as part of the Superintelligence unit is worth it if Meta can regain its momentum and potentially create more powerful AI technology that steamrolls the competition, CNBC reported in June. For Meta, Llama 4 represented the company's answer to competing models from rivals like OpenAI, and executives have viewed it as helping the company dominate a potential computing platform of the future. Similar to other Meta-incubated technologies like the PyTorch AI developer tools, the company released Llama to the open-source community, which can then access and use the software for free, subject to certain licensing terms. While the predecessor, Llama 3, was a hit with developers, they haven't taken to Llama 4 because it's seen by some as more difficult to customize and integrate into their apps. That's resulted in many coders preferring Llama 3 over its successor, people familiar with the matter said. Additionally, Zuckerberg lost confidence in his generative AI team and its leadership in part due to a controversy over whether Meta may have gamed certain industry AI benchmark tests, the people said. Llama 4's struggles can be traced back to January, when the sudden rise and ensuing popularity of the open-source R1 AI model by DeepSeek caught Meta off guard, leading to a reevaluation of Llama's underlying architecture, the people said. DeepSeek's R1 is a so-called mixture-of-experts AI model, or MoE. R1 is similar to OpenAI's o1 family of models that can be trained to excel at multistep tasks like solving math equations or writing code. By contrast, Llama's models — before their latest release - were dense AI models, which are generally simpler for most AI developers to fine-tune and incorporate into their own apps, the people said. AI labs like OpenAI and Anthropic, researchers say, have been pushing MoE models to power AI agents that can perform a variety of step-by-step tasks. Those companies keep their designs closely guarded from competitors. OpenAI has been developing an AI model for the open-source community, but CEO Sam Altman said earlier this month that its debut is delayed indefinitely pending safety tests and other reviews. Although Meta has previously published research on MoE models, DeepSeek's release to the open-source community wowed researchers because R1 appeared to be less expensive to train and run compared with other AI models, experts said. Suddenly, Meta executives thought they had a clearer picture into how to create their own efficient and possibly cheaper MoE models, potentially leapfrogging rivals like OpenAI, people familiar with the matter said. Still, some staff members in Meta's GenAI unit pushed for Llama 4 to remain a dense AI model, which though generally less efficient, is still powerful, and Meta originally planned on that architecture acting as the backbone supporting improved voice recognition capabilities, the people said. Ultimately, Meta went with the MoE approach, due in part to DeepSeek's innovations and the promise of pulling ahead of OpenAI, the people said. Meta released two small versions in April and said a "Behemoth" version would come at a later date. But the new MoE architecture disappointed some developers, who were simply hoping Llama 4 would be a souped-up version of Llama 3, people familiar with the matter said. Llama 4 also failed to deliver a significant leap over competing open-source models from China, the people said. Executives at Meta as well as the Superintelligence Labs' high-profile hires are now questioning the company's current open-source AI strategy, and have considered skipping the release of Behemoth in favor of developing a more powerful proprietary AI model, the people said. A Meta spokesperson said in a statement that the company's "position on open source AI is unchanged." "We plan to continue releasing leading open source models," the spokesperson said. "We haven't released everything we've developed historically and we expect to continue training a mix of open and closed models going forward." The New York Times first reported that Meta was considering upending its open-source AI strategy. Despite Meta's AI struggles, the company's core online ad business remains strong, and investors are hopeful that the recent AI investments and hiring will eventually pay off. Zuckerberg said in July that the company would invest "hundreds of billions of dollars" into building out the computing infrastructure needed to power cutting-edge AI projects. "Meta Superintelligence Labs will have industry-leading levels of compute and by far the greatest compute per researcher," Zuckerberg said in a Facebook post. Analysts at Bank of America said in a note this month that Zuckerberg's comments indicate "a sign of confidence in Meta's revenue trajectory." The analysts said that his statement also "implies higher future Capex and Opex," which potentially equates to even more AI spending. "We also see the post as reaching out to AI talent, signaling Meta as a place for AI innovation," the analysts wrote. "We expect AI investment to be a top focus area on the upcoming earnings call, and Meta likely needs to make a case for strong AI returns to drive multiple expansion." Meta and its rivals' pursuit of AI researchers echoes the self-driving car frenzy of 2017, when companies like Google and Uber competed fiercely for talent, doling out "similarly crazy kind of pay packages across the board," said Megh Gautam, chief product officer at deal-tracking firm Crunchbase. "The dynamics still feel very much like a winner-take-all-market, so you're trying your best to give yourself the best shot possible to go make that happen," Gautam said. Investors appear more receptive to Meta's AI spending and strategy shifts, a contrast with a few years ago when the company was heavily pushing the metaverse, said Uday Cheruvu, an analyst and portfolio manager at Harding Loevner, which owns Meta shares. OpenAI, Google and Anthropic are also trying to hire and maintain talent all while continuing to spend billions of dollars on developing their respective AI models, Cheruvu said. "Now with AI, it's not just Meta – everyone else is doing it, so now the euphoria is much higher," Cheruvu said.
Globe and Mail
9 hours ago
- Business
- Globe and Mail
Alibaba's AI Ambitions: 3 Strategic Bets Investors Should Watch Closely
Key Points Qwen is Alibaba Group's AI crown jewel. Alibaba Cloud is evolving into an AI powerhouse. AI is reprogramming Alibaba's core commerce engine. 10 stocks we like better than Alibaba Group › Alibaba Group (NYSE: BABA) has spent the past few years navigating regulatory crackdowns, macro uncertainty in China, and intensifying competition. However, behind the scenes, a massive transformation is underway as the tech giant repositions itself as an artificial intelligence (AI)-native company. For investors, this pivot could mark a turning point. While the headlines still focus on sluggish consumer sentiment and slow e-commerce growth, its AI strategy may be the most important long-term story. Let's break down the three strategic bets Alibaba is making in AI and why they matter. Building a foundational, open-source AI engine with Qwen Alibaba's most important AI initiative is Qwen -- a family of large language models that rivals the capabilities of OpenAI's GPT-4, Meta Platforms ' Llama, and Alphabet 's Google Gemini. The tech giant's seriousness in this area is evident in the ongoing releases of new and better versions of Qwen. The latest version, Qwen3, boasts up to 235 billion parameters and has deep multilingual capabilities (119 languages). This latest version compares favorably to the best models from competitors like OpenAI, Deepseek, and Gemini, achieving competitive results in benchmark evaluations of coding, math, and general capabilities. In other words, Alibaba's LLM is as competitive (if not better) than some of the best globally. Unlike closed models from U.S. tech giants, Alibaba is betting big on open AI -- enabling researchers, start-ups, and governments to use and fine-tune Qwen models freely. This approach is not just technically bold but also strategically smart. It enables Alibaba to drive early adoption globally, particularly in Asia and emerging markets where U.S.-based models are less prevalent. As adoption grows, Qwen and Alibaba's tech ecosystem will become the de facto platform that future AI companies rely on. In other words, Qwen could become the engine powering a new ecosystem of Alibaba-backed AI apps and infrastructure. Alibaba Cloud pivots from infrastructure to AI platform For years, Alibaba Cloud was viewed as a regional infrastructure provider, dominant in China but lagging behind global peers in profitability and product sophistication. But that's changing. The company is now rebuilding its cloud around AI, offering a tightly integrated platform that combines compute, application programming interfaces (APIs), developer tools, and its proprietary Qwen models. For instance, with Model Studio, developers can leverage its LLM to build generative AI applications. This vertical integration is essential, as it moves Alibaba Cloud up the value chain -- from basic hosting to full-stack AI enablement. This move unlocks new customers and expands wallet share among existing ones. Strategically, this also leads to deeper customer lock-in due to higher switching costs. Besides, since Qwen is an open-source platform globally, Alibaba has positioned itself well to grow its presence beyond China, particularly in regions where Amazon Web Services (AWS) and Microsoft Azure are not dominant players. In other words, the shift to AI-native infrastructure could be the key to Alibaba Cloud's long-term growth and expansion. Reinventing commerce with AI At its core, Alibaba remains a commerce company, but even here, AI plays a central role. Across platforms like Taobao and Tmall, the company is embedding generative AI to perform various tasks for merchants and consumers. For merchants, AI can help improve productivity and sales by automating tasks such as product listings, generating marketing content, and providing customer service. For consumers, AI facilitates personalized recommendations and intelligent search. Internally, AI also helps improve productivity and efficiency, especially in areas such as warehousing, fulfillment, and logistics. By fully integrating AI into commerce, Alibaba aims to reinvent itself as it fends off competitors such as Pinduoduo and Douyin. If done correctly, AI can help Alibaba enhance customer delight and increase operational efficiencies, laying the foundation for its next phase of expansion. What does this mean for investors? While the market remains fixated on China risk and e-commerce fatigue, Alibaba is quietly rebuilding its future around AI. From Qwen to cloud to commerce, it's laying the groundwork for a more intelligent, scalable, and global business. The movement toward AI is not a short-term catalyst -- it's a long-term transformation. But if Alibaba executes, investors may look back and see this as the moment it rewrote its growth story. It's a company worth watching. Should you invest $1,000 in Alibaba Group right now? Before you buy stock in Alibaba Group, consider this: The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Alibaba Group wasn't one of them. The 10 stocks that made the cut could produce monster returns in the coming years. Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you'd have $636,628!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $1,063,471!* Now, it's worth noting Stock Advisor's total average return is 1,041% — a market-crushing outperformance compared to 183% for the S&P 500. Don't miss out on the latest top 10 list, available when you join Stock Advisor. See the 10 stocks » *Stock Advisor returns as of July 28, 2025 Lawrence Nga has positions in Alibaba Group and PDD Holdings. The Motley Fool has positions in and recommends Alphabet, Amazon, Meta Platforms, and Microsoft. The Motley Fool recommends Alibaba Group and recommends the following options: long January 2026 $395 calls on Microsoft and short January 2026 $405 calls on Microsoft. The Motley Fool has a disclosure policy.
NDTV
a day ago
- Business
- NDTV
From OpenAI To Meta: All You Need To Know About Chief Scientist Shengjia Zhao
Shengjia Zhao, a former OpenAI researcher, has been appointed chief scientist of Meta 's newly launched Superintelligence Lab, CEO Mark Zuckerberg announced last Friday. Mr Zhao, co-creator of ChatGPT and GPT-4, will work directly with Mr Zuckerberg and Meta's Chief AI Officer, Alexandr Wang, to steer the lab's scientific direction. He is also a co-founder of the lab, which aims to develop artificial general intelligence (AGI) and improve Meta's Llama models. Mark Zuckerberg said Meta was committed to building "full general intelligence" and releasing its research as open source. Who Is Shengjia Zhao? Shengjia Zhao earned his Bachelor's degree from Tsinghua University in 2016, as per his LinkedIn profile. He completed a semester exchange in Computer Science at Rice University and then pursued a PhD in Computer Science at Stanford University. Following his PhD, Mr Zhao joined OpenAI in June 2022 as a Member of Technical Staff, where he worked for over three years until July this year. During this time, he co-created ChatGPT, GPT-4, and several OpenAI mini models, including GPT-4.1 and o3. He has now been appointed as the Chief Scientist of Meta's Superintelligence Lab. Mr Zhao has received several prestigious awards. In 2015, he was awarded the Google Excellence Scholarship. Four years later, he received the Qualcomm Innovation Fellowship (QinF) for his contributions to research in AI while at Stanford. The same year, Mr Zhao was named a JP Morgan PhD Fellow. In October 2022, he was honoured with the ICLR 2022 Outstanding Paper Award, one of the top recognitions in the machine learning research community. Mr Zhao currently lives in San Francisco, California. He is fluent in English and Mandarin. Meta's Superintelligence Lab The Meta Superintelligence Lab (MSL) was launched by Meta to drive progress on its Llama models and pursue its broader goal of developing artificial general intelligence (AGI). According to Mark Zuckerberg, Shengjia Zhao is a co-founder of the lab. While Meta's original AI research is led by Yann LeCun under the FAIR division, MSL operates independently. Meta has poured billions into hiring top talent from rivals, including Google, OpenAI, Apple, and Anthropic. The company also acquired Scale AI for $14 billion, bringing its CEO, Alexandr Wang, on board as Meta's Chief AI Officer. Mr Zuckerberg has committed to investing hundreds of billions more in building vast AI data centres across the US.
India Today
2 days ago
- Business
- India Today
Google DeepMind CEO says Meta's AI talent war is rational because they are behind
The battle for the brightest minds in artificial intelligence has reached fever pitch, and Meta appears determined to buy its way to the top. The company has been aggressively poaching talent from rivals, including DeepMind, and dangling packages worth hundreds of DeepMind's cofounder and chief executive, Demis Hassabis, isn't particularly impressed. In an interview this week, he described Mark Zuckerberg's tactics as 'rational' but also suggested that those motivated purely by money are unlikely to be the ones shaping the future of this year, Meta quietly launched its Superintelligence Labs, a new project led by two high-profile Silicon Valley figures: former Scale AI boss Alexander Wang and Nat Friedman, who previously ran GitHub. The initiative was created to inject fresh life into Meta's AI ambitions after its Llama models, released in April, failed to set the world alight. Zuckerberg himself is said to have been personally involved in the talent hunt, reportedly luring some of the biggest names in AI research with stratospheric offers, packages as high as $200 million a year have been whispered about in industry results are already showing: researchers from OpenAI, Google and even Apple have resigned to join Meta, raising eyebrows across the tech however, believes the deeper motivation for top scientists goes beyond pay cheques. 'There's a strategy that Meta is taking right now,' he told Bloomberg. 'I think the people that are real believers in the mission of AGI and what it can do, and understand the consequences, both good and bad, are mostly doing it to be at the frontier, so they can help influence how that plays out and steward the technology safely into the world.'Meta, he noted, 'right now are not at the frontier. Maybe they'll manage to get back there. And it's probably rational, what they're doing from their perspective, because they're behind and they need to do something. But I think there are more important things than just money. Of course, one has to pay people market rates, and those continue to go up.'The DeepMind CEO also reflected on how dramatically the AI field has changed in just over a decade. 'We couldn't raise any money. I didn't pay myself for a couple of years,' he said of DeepMind's early years.'Now,' he added with a touch of incredulity, 'interns are being paid what we raised as our entire first seed round.' The comment underlines just how much the AI gold rush has distorted valuations and salaries as tech firms race to be first to artificial general intelligence (AGI).For Hassabis, the current frenzy of hiring and spending misses a fundamental point: the real prize lies in being at the cutting edge of research, not in cashing in. DeepMind, acquired by Google in 2014, has built a reputation for scientific breakthroughs, from mastering Go to solving protein structures, and Hassabis clearly sees that mission as more compelling than stock options meanwhile, is betting that an infusion of top-tier researchers will allow it to catch up in the generative AI race, and perhaps leapfrog rivals like OpenAI, Anthropic and, indeed, Zuckerberg's megabucks approach works remains to be seen. But for now, at least one leading voice in AI believes there's more to the future of the field than a blank cheque.- EndsMust Watch
&w=3840&q=100)
Business Standard
3 days ago
- Business
- Business Standard
Why Google DeepMind CEO Demis Hassabis calls Meta's AI poaching 'rational'
Google DeepMind CEO Demis Hassabis has called Meta's poaching of AI talent from rival labs a 'rational' strategy, stating that the Mark Zuckerberg-led company is trying to regain lost ground in the race to build frontier artificial intelligence. Speaking on the Lex Fridman podcast, Hassabis said, 'Meta right now are not at the frontier. Maybe they'll manage to get back there. And it's probably rational, what they're doing from their perspective—because they're behind and they need to do something.' Meta has been making waves in recent weeks with aggressive recruitment efforts, offering compensation packages reportedly as high as $200 million a year to top AI researchers. Zuckerberg's talent push after Llama setbacks Meta's Superintelligence Labs, launched earlier this year, is being led by Scale AI's former CEO Alexandr Wang and ex-GitHub CEO Nat Friedman. The initiative follows lukewarm reception to the company's Llama model releases in April. Zuckerberg has reportedly taken a personal interest in the team-building process. Former OpenAI researchers Shengjia Zhao, Shuchao Bi, Jiahui Yu, and Hongyu Ren are among those who have joined Meta in recent months. 'There's more important things than just money' While Hassabis acknowledged the logic behind Meta's strategy, he stressed that not all AI professionals are driven by financial rewards. 'There's a strategy that Meta is taking right now… I think the people that are real believers in the mission of AGI and what it can do—and understand the consequences, both good and bad—are mostly doing it to be at the frontier, so they can help influence how that plays out and steward the technology safely into the world,' he said. 'There are more important things than just money. Of course, one has to pay people market rates—and those continue to go up,' he added. How AI compensation has soared According to recent federal visa filings reviewed by Business Insider, OpenAI's technical staff earn an average of $292,115, with the top position drawing $530,000. Anthropic's average for technical hires stands at $387,500, while Mira Murati's new venture Thinking Machines Lab is reportedly offering salaries of up to $500,000. Hassabis contrasted this with DeepMind's early days: 'I remember when we were starting out back in 2010, I didn't even pay myself for a couple of years because there wasn't enough money. We couldn't raise any money. These days, interns are being paid the amount that we raised as our first entire seed round.' Other AI leaders have echoed Hassabis's views. 'They get these offers and then they say, 'Well, of course I'm not going to leave because my best-case scenario at Meta is that we make money, and my best case at Anthropic is we affect the future of humanity',' said Anthropic Co-Founder Benjamin Mann.
