Huawei's AI lab denies that one of its Pangu models copied Alibaba's Qwen
The division, called Noah Ark Lab, issued the statement on Saturday, a day after an entity called HonestAGI posted an English-language paper on code-sharing platform Github, saying Huawei's Pangu Pro Moe (Mixture of Experts) model showed "extraordinary correlation" with Alibaba's Qwen 2.5 14B.
This suggests that Huawei's model was derived through "upcycling" and was not trained from scratch, the paper said, prompting widespread discussion in AI circles online and in Chinese tech-focused media.
The paper added that its findings indicated potential copyright violation, the fabrication of information in technical reports and false claims about Huawei's investment in training the model.
Noah Ark Lab said in its statement that the model was "not based on incremental training of other manufacturers' models" and that it had "made key innovations in architecture design and technical features." It is the first large-scale model built entirely on Huawei's Ascend chips, it added.
It also said that its development team had strictly adhered to open-source license requirements for any third-party code used, without elaborating which open-source models it took reference from.
Alibaba did not immediately respond to a Reuters request for comment. Reuters was unable to contact HonestAGI or learn who is behind the entity.
The release of Chinese startup DeepSeek's open-source model R1 in January this year shocked Silicon Valley with its low cost and sparked intense competition between China's tech giants to offer competitive products.
Qwen 2.5-14B was released in May 2024 and is one of Alibaba's small-sized Qwen 2.5 model family which can be deployed on PC and smartphones.
While Huawei entered the large language model arena early with its original Pangu release in 2021, it has since been perceived as lagging behind rivals. It open-sourced its Pangu Pro Moe models on Chinese developer platform GitCode in late June, seeking to boost the adoption of its AI tech by providing free access to developers.
While Qwen is more consumer-facing and has chatbot services like ChatGPT, Huawei's Pangu models tend to be more used in government as well as the finance and manufacturing sectors.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Daily Mail
30 minutes ago
- Daily Mail
Microsoft boss offers callous advice to workers who'd just been fired because of AI advances
A Microsoft executive is facing backlash after suggesting that recently laid-off employees use artificial intelligence to cope with unemployment. The company cut approximately 9,000 jobs last week, many in its gaming division, as it continues to shift focus and invest tens of billions of dollars into artificial intelligence. Matt Turnbull, an executive producer at Xbox Game Studios Publishing, posted AI-generated prompts to LinkedIn offering laid-off workers help with career planning, résumé writing, and even emotional support. The post, which has since been deleted, sparked swift criticism across social media where users called him 'out of touch' and 'tone-deaf.' 'These are really challenging times,' Turnbull wrote, encouraging displaced employees to use chatbots to help manage feelings of impostor syndrome and reframe their layoff experiences in a more positive light. His message included suggestions for using AI to develop 30-day job search plans and tailor résumés for different industries. The remarks came just days after Microsoft shut down several major studios, including Arkane Austin and Tango Gameworks, following layoffs at Xbox, Activision Blizzard, and Bethesda. While Microsoft has committed roughly $80 billion to expanding its AI infrastructure, critics say the company's rapid downsizing of human staff highlights a growing divide between its investment in automation and its treatment of employees. 'I know these types of tools engender strong feelings in people, but I'd be remiss in not trying to offer the best advice I can under the circumstances,' Turnbull wrote. Daily Mail reached out to Turnbull for comment, but has yet to receive a response. The debate surrounding his now-deleted LinkedIn post highlighted a broader conversation about AI's role in the workplace. Turnbull's advice emphasized that AI could not replace the 'lived experience' of workers but suggested it could help them get 'unstuck faster, calmer, and with more clarity' in a time of limited mental energy. 'I know these types of tools engender strong feelings in people, but I'd be remiss in not trying to offer the best advice I can under the circumstances,' he added. However, experts have said that his comments lacked empathy and for the irony of recommending AI as a coping tool when Microsoft is at the forefront of AI development, potentially displacing human workers. The backlash was immediate, with many former Microsoft employees and industry professionals calling Turnbull's comments insensitive. Eric Smith, a Zenimax Online producer who was laid off this week, said: 'Jesus Christ, read the room dude.' In a now-deleted LinkedIn post, Matt Turnbull offered AI prompts for career planning and emotional support, which received criticism for its tone-deafness and sparked swift backlash Paul Murphy, a Game designer, wrote: 'I'm sure you're trying to help, but WTF. You hired these folks once.' 'At least give them the respect they've earned and assume they have the ability to write their own resumes,' he added. The irony of the situation became even more pronounced when you consider Microsoft's role as both a leader in AI and a company overseeing large-scale layoffs. The company has made substantial investments in AI tools like Copilot and ChatGPT, which are designed to help workers with tasks like writing, coding, and problem-solving. However, the growing use of these tools in the workplace raises concerns about the displacement of workers and the ethical implications of replacing human roles with machines. As millions of people begin to turn to AI for personal and professional support, the relationship between these tools and job security becomes even more complex. AI platforms like ChatGPT and Google's Gemini have experienced tremendous growth, especially in the United States, where millions of people are now actively using these tools. ChatGPT alone has over 100 million active monthly users globally, with a significant portion based in the US. The number of users in the US has skyrocketed, with the tool being used for a range of applications, including career coaching, resume assistance, and even mental health support. In 2024, it was estimated that approximately 40 percent of Americans had used some form of AI, with tools like ChatGPT seeing particular growth among professionals looking to improve productivity or navigate job transitions. The growing adoption of AI technologies is happening at a time when major tech companies, including Microsoft, Meta, and Google, are laying off thousands of employees. Microsoft alone cut over 25,000 jobs across multiple sectors from 2023 to 2025, with many positions in roles now being automated or supplemented by AI tools.
Daily Mail
30 minutes ago
- Daily Mail
Joe Rogan flips the God debate on its head with shocking theory that 'we created him'
Joe Rogan has come to a mind-bending conclusion about life, fearing that humanity has misinterpreted what reality is and we're actually in the process of creating God. While interviewing computer scientist Roman Yampolskiy on the Joe Rogan Experience podcast, the men debated the possibility that reality is a giant simulation and humans are building a God-like supercomputer using artificial intelligence (AI). According to Rogan's theory, humanity has misinterpreted ancient prophecies regarding the second coming of Jesus Christ and Judgement Day, saying the creation of this AI super intelligence is the final chapter before our reality resets. 'This whole idea of Jesus coming back, well maybe it's real. Maybe we just completely misinterpreted these ancient scrolls and texts and what it really means is that we are going to give birth to this,' Rogan explained. Yampolskiy, an author and researcher in AI safety, added to Rogan's theory, suggesting that reality is just an ongoing cycle of Big Bangs - the explosion that kickstarted the universe - starting and restarting life over and over again. During the episode released on July 3, Rogan added that he feared God 'is actually created by human beings creating this infinitely intelligent thing that can essentially harness all of the available energy and power of the universe and create anything it wants.' Yampolskiy even contended that many of the religions around the world have one thing in common: 'What they all agree on is that there is [a] super intelligence which created a fake world.' In previous podcasts episodes this year, Rogan has repeatedly warned that AI was just years away from reaching a God-like state as more tech companies pour resources into its development. Rogan suggested that people have misinterpreted events like the second coming of Jesus Christ, claiming that humans would create the intelligence they believed was returning from heaven Rogan asked Yampolskiy if the physical world we see was created by God, then what existed before this all-powerful being created the human race. 'Ideas, just information,' the computer scientist replied during the podcast. 'Just God. God was bored And it was like, let's make some animals that can think and solve problems. And for what reason?' Rogan countered. That's when the podcast host theorized life was an ongoing simulation and people had the idea of God's return to Earth backwards, claiming that people would summon this powerful intelligence themselves through the creation of AI. The theory that all life exists within a computer simulation has been debated for decades, and it's even been entertained in blockbuster movies like 'The Matrix.' While Rogan suggested that deepfakes of Bigfoot, UFOs, and other impossible images could be proof we're living in a virtual reality, scientists have argued that there is more basic evidence that this world is not real. Melvin Vopson, an associate professor in physics at the University of Portsmouth, has argued that gravity may be a sign that we're all living in a virtual simulation. According to Vopson's new paper in AIP Advances, our universe is the 'ultimate computer' and gravity's pull on Earth and in outer space is the computer trying to keep its vast amount of data organized. As for who's running the simulation, Yampolskiy explained that one possibility is life as we know it could simply be a replay of past events being examined by the future human race. Rogan then asked Yampolskiy why an all-powerful AI would even bother creating a reality for our consciousnesses to exist in. Yampolskiy explained that it's impossible to know who is controlling this program, whether it's an alien intelligence, humans from the future, or an AI super intelligence has already been built and is running these simulations by itself. Therefore, it's also impossible to know why people would be needed in the virtual world, adding that he and Rogan could have had the same interview thousands of times and never knew it was happening. 'Is it entertainment? Is it scientific experimentation? Is it marketing? Maybe somebody managed to control them and [are] trying to figure out what Starbucks coffee sells best. And they need to run [an] earth-sized simulation to see what sells best,' Yampolskiy explained. 'Think about this decade... We're about to invent intelligence and virtual worlds, God-like inventions. We're here. There's a good chance that's not just random,' the AI researcher added, suggesting we're already stuck in a computer simulation. In any scenario, Rogan theorized that the end of the simulation was nearing as tech giants get closer to creating the all-knowing intelligence which has been foretold in many prophecies. Whether humanity is already locked in a simulation or we're on the road to creating a God-like intelligence in a self-fulfilling prophecy, Rogan claimed that within 50 to 60 years 'there will be one day a virtual simulation of this reality that's indistinguishable from reality.'
The Independent
2 hours ago
- The Independent
F1 could return to V8 engines as soon as 2029, FIA president says
Ben Sulayem also addressed the ongoing discussion about expanding the grid, specifically the possibility of a Chinese team occupying the final 12th slot. He reiterated his belief that the sport requires more cars rather than an increased number of races. "The time will come when we feel it is right to open an expression of interest," he said regarding the 12th slot. He stressed that any new entrant must genuinely contribute to the sport's sustainability: "We are not here to upset other teams. It won't be just go and do it for the sake of doing it. It has to be worth it for us. The team has to add value to sustaining the business of Formula One." General Motors' Cadillac brand is already slated to join as the 11th team next season, a move that faced initial resistance from existing teams and Formula One's owners, Liberty Media.
