An artificial intelligence (AI) team at Stanford University apologized for plagiarizing a large language model (LLM) from a Chinese AI company, which became a trending topic on the Chinese social media platforms, where it sparked concern among netizens on Tuesday. We apologize to the authors of MiniCPM [the AI model developed by a Chinese company] for any inconvenience that we caused for not doing the full diligence to verify and peer review the novelty of this work, the multimodal AI model Llama3-V's developers wrote in a post on social platform X. The apology came after the team from Stanford University announced Llama3-V on May 29, claiming it had comparable performance to GPT4-V and other models with the capability to train for less than $500. According to media reports, the announcement published by one of the team members quickly received more than 300,000 views. However, some netizens from X found and listed evidence of how the Llama3-V project code was reformatted and similar to MiniCPM-Llama3-V 2.5, an LLM developed by a Chinese technology company, ModelBest, and Tsinghua University. Two team members, Aksh Garg and Siddharth Sharma, reposted a netizen's query and apologized on Monday, while claiming that their role was to promote the model on Medium and X (formerly Twitter), and that they had been unable to contact the member who wrote the code for the project. They looked at recent papers to validate the novelty of the work but had not been informed of or were aware of any of the work by Open Lab for Big Model Base, which was founded by the Natural Language Processing Lab at Tsinghua University and ModelBest, according to their responses. They noted that they have taken all references to Llama3-V down in respect to the original work. In response, Liu Zhiyuan, chief scientist at ModelBest, spoke out on the Chinese social media platform Zhihu, saying that the Llama3-V team failed to comply with open-source protocols for respecting and honoring the achievements of previous researchers, thus seriously undermining the cornerstone of open-source sharing. According to a screenshot leaked online, Li Dahai, CEO of ModelBest, also made a post on his WeChat moment, saying that the two models were verified to have highly similarity in terms of providing answers and even the same errors, and that some relevant data had not yet been released to the public. He said the team hopes that their work will receive more attention and recognition, but not in this way. He also called for an open, cooperative and trusting community environment. Director of the Stanford Artificial Intelligence Laboratory Christopher Manning also responded to Garg's explanation on Sunday, commenting "How not to own your mistakes!" on X. As the incident became a trending topic on Sina Weibo, Chinese netizens commented that academic research should be factual, but the incident also proves that the technology development in China is progressing. Global Times

NEW DELHI, July 9 (Reuters) - India's beleaguered Paytm (PAYT.NS), opens new tab has secured approval from a government panel that oversees investments linked to China to invest 500 million rupees ($6 million) in a key subsidiary, three sources with direct knowledge of the matter said. The approval, which still has to be vetted by the finance ministry, will remove the main stumbling block to the unit, Paytm Payment Services, resuming normal business operations. Paytm Payment Services is one of the biggest remaining parts of the fintech firm's business, accounting for a quarter of consolidated revenue in the financial year ended March 2023. A separate unit, Paytm Payments Bank, was wound down this year by order of the central bank due to persistent compliance issues, triggering a meltdown in Paytm's stock. The government panel had earlier held back approval due to concerns about the 9.88% stake in Paytm held by China's Ant Group. India has intensified scrutiny of Chinese businesses since a 2020 border clash between the two countries. All in all, Paytm has been waiting for the nod from the government panel for about two years and without it, it would have had to also wind down its payment services business, which was forbidden from taking on new customers in March 2023. Once the approval has been formalised, it will be able to seek a so-called "payment aggregator" licence from the Reserve Bank of India. The sources, two of whom are government sources, declined to be identified as the decision has not been formally announced. India's foreign, home, finance and industries ministries, whose representatives sit on the panel, did not reply to emails seeking comment. A Paytm spokesperson said the company does not comment on market speculation. "We will continue to make disclosures in compliance with our obligations under the SEBI Regulations, and will inform the exchanges when there is any new material information to share," the spokesperson said.

WhatsApp beta version 2.24.14.20 has a new feature that allows users to share photos with Meta AI. The AI chatbot will analyze uploaded images and provide information or context about the content. Users may be able to request specific edits to their photos directly through Meta AI, though the extent of this feature is still unknown. As the battle for AI dominance heats up, Meta is adding a new trick to its AI chatbot, Meta AI, which is already part of Facebook, Instagram, and WhatsApp. While Meta AI already has impressive text capabilities, such as replying to questions, suggesting captions, and holding conversations, users cannot currently share or upload photos to the Meta AI chat. WaBetaInfo has uncovered the exciting new feature in the WhatsApp beta for Android version 2.24.14.20. This feature will allow Meta AI to interact with photos shared by users, reply to photos, and even edit them. As shown in the attached screenshot, WhatsApp is testing a new camera button in the Meta AI chat, designed to function similarly to the camera button in regular chats. This addition will allow users to manually share photos with Meta AI, a capability that is currently unavailable. With this new functionality, users will be able to ask questions about their photos, presumably allowing users to ask the AI to identify objects or locations or provide context about the photo’s content. Moreover, the screenshot suggests that Meta AI will also offer the option to edit photos, enabling users to make changes to their images directly within the chat by sharing a prompt. The exact scope of this image editing feature remains unclear, leaving us to wonder if it will be limited to simple tweaks or if it will unleash a powerful AI-driven photo editing suite. The possibilities are both exciting and intriguing, and this feature could definitely be a big hit, especially if it performs as promised. While this new image-sharing feature would mean Meta will analyze and face-scan the photos you upload, the screenshot includes a disclaimer indicating that users will have the option to delete their photos whenever they want. As of now, it seems that the feature is still in development, so it might be some time before we finally get to see it roll out publicly. Recently, we also reported about WhatsApp working on an “Imagine Me” feature that would allow Meta AI to generate AI avatars of you based on a set of your photos. WhatsApp in our newsletters WhatsApp is a leading messaging app, keep up to date on the latest, and learn about more Android apps today!

SEOUL, July 9 (Reuters) - Samsung Electronics (005930.KS), opens new tab said on Tuesday it won an order from Japanese artificial intelligence company Preferred Networks to make chips for AI applications using the South Korean firm's 2-nanometre foundry process and advanced chip packaging service. It is the first order Samsung has revealed for its cutting-edge 2-nanometre chip contract manufacturing process. Samsung did not elaborate on the size of the order. The chips will be made using high-tech chip architecture known as gate all-around (GAA) and multiple chips will be integrated in one package to enhance inter-connection speed and reduce size, Samsung said in a statement. South Korea's Gaonchips Co (399720.KQ), opens new tab designed the chips, Samsung said. The chips will go toward Preferred Networks' high-performance computing hardware for generative AI technologies such as large language models, Junichiro Makino, Preferred Networks vice president and chief technology officer of computing architecture, said in the statement.

The world's deepest diving pool, Deepspot, opened this weekend near the Polish capital Warsaw. The 45.5-meter pool contains artificial underwater caves, Mayan ruins and a small shipwreck for scuba divers and free divers to explore. Deepspot can hold 8,000 cubic meters of water, more than 20 times the capacity of a normal 25-meter swimming pool. Unlike ordinary swimming pools, Deepspot can still open despite Poland's COVID-19 epidemic prevention restrictions because it is a training center that provides courses. The operator also plans to open a hotel where guests can observe divers at a depth of 5 meters from their rooms. "This is the deepest diving pool in the world," Michael Braszczynski, 47, Deepspot's director and a diving enthusiast, told AFP at the opening yesterday. The current Guinness World Record holder is a 42-meter-deep pool in Montegrotto Terme, Italy. The 50-meter-deep Blue Abyss pool in the UK is scheduled to open in 2021. On the first day of Deepspot's opening, about a dozen people visited, including eight experienced divers who wanted to pass the instructor exam. "There are no spectacular fish or coral reefs here, so it can't replace the ocean, but it is certainly a good place to learn and train safe open water diving," said 39-year-old diving instructor Przemyslaw Kacprzak. "And it's fun! It's like a kindergarten for divers."