link1s.site

Stanford AI project team apologizes for plagiarizing Chinese model

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

ChatGPT： Explained to Kids（How ChatGPT works）

Chat means chat, and GPT is the acronym for Gene Rate Pre trained Transformer. Genrative means generation, and its function is to create or produce something new; Pre trained refers to a model of artificial intelligence that is learned from a large amount of textual materials, while Transformer refers to a model of artificial intelligence. Don't worry about T, just focus on the words G and P. We mainly use its Generative function to generate various types of content; But we need to know why it can produce various types of content, and the reason lies in P. Only by learning a large amount of content can we proceed with reproduction. And this kind of learning actually has limitations, which is very natural. For example, if you have learned a lot of knowledge since childhood, can you guarantee that your answer to a question is completely correct? Almost impossible, firstly due to the limitations of knowledge, ChatGPT is no exception, as it is impossible to master all knowledge; The second is the accuracy of knowledge, how to ensure that all knowledge is accurate and error free; The third aspect is the complexity of knowledge, where the same concept is manifested differently in different contexts, making it difficult for even humans to grasp it perfectly, let alone AI. So when we use ChatGPT, we also need to monitor the accuracy of the output content of ChatGPT. It is likely not a problem, but if you want to use it on critical issues, you will need to manually review it again. And now ChatGPT has actually been upgraded twice, one is GPT4 with more accurate answering ability, and the other is the recent GPT Turbo. The current ChatGPT is a large model called multimodality, which differs from the first generation in that it can not only receive and output text, but also other types of input, such as images, documents, videos, etc. The output is also more diverse. In addition to text, it can also output images or files, and so on.

The Apple Watch is reportedly getting a birthday makeover

Apple is planning to revamp its smartwatch as its 10th birthday nears. The improvements include larger displays and thinner builds, Bloomberg reported. The revamped watches may also get a new chip, which could enable some AI enhancements. The Apple Watch is about to turn 10, so Apple is planning a birthday revamp, including larger displays and thinner builds, Bloomberg reported. Both versions of the new Series 10 watches will have screens similar to the large displays found on the Apple Watch Ultra, the report said. The revamped watches are also expected to contain a new chip that may permit some AI enhancements later on. Last month, Apple pulled back the curtain on its generative-AI plans with Apple Intelligence. Advertisement It hopes the artificial-intelligence features will prove alluring enough to persuade consumers to buy new Apple products. The announcement has been generally well received by Wall Street. Dan Ives of Wedbush Securities wrote in a Monday note that the "iPhone 16 AI-driven upgrade could represent a golden upgrade cycle for Cupertino." "We believe AI technology being introduced into the Apple ecosystem will bring monetization opportunities on both the services as well as iPhone/hardware front and adds $30 to $40 per share," he added. Apple's stock closed on Friday at just over $226 a share, up 22% this year and valuing the company at $3.47 trillion. That puts it just behind Microsoft, which was worth $3.48 trillion at Friday's close. The tech giants have been vying for the title of the world's most valuable company in recent months — with the chipmaker Nvidia briefing claiming the crown last month. Apple also announced some software updates for the watch at its Worldwide Developers Conference last month. The latest version of the device's software, watchOS 11, emphasizes fitness and health, introducing tools that allow users to rate workouts and adjust effort ratings. WatchOS 11 will also use machine learning to curate the best photos for users' displays. Apple has previously used product birthdays to release new versions of devices. The iPhone X's release marked the 10th anniversary of the smartphone. However, it's not clear exactly when Apple plans to release the revamped watches, Bloomberg said. The company announced the Apple Watch in September 2014, with CEO Tim Cook calling it "the most personal product we've ever made." Apple did not immediately respond to a request for comment made outside normal working hours.

Rotten eggs chemical detected on Jupiter-like alien planet

WASHINGTON, July 8 (Reuters) - The planet known as HD 189733b, discovered in 2005, already had a reputation as a rather extreme place, a scorching hot gas giant a bit larger than Jupiter that is a striking cobalt blue color and has molten glass rain that blows sideways in its fierce atmospheric winds. So how can you top that? Add hydrogen sulfide, the chemical compound behind the stench of rotten eggs. Researchers said on Monday new data from the James Webb Space Telescope is giving a fuller picture of HD 189733b, already among the most thoroughly studied exoplanets, as planets beyond our solar system are called. A trace amount of hydrogen sulfide was detected in its atmosphere, a first for any exoplanet. "Yes, the stinky smell would certainly add to its already infamous reputation. This is not a planet we humans want to visit, but a valuable target for furthering our understanding of planetary science," said astrophysicist Guangwei Fu of Johns Hopkins University in Baltimore, lead author of the study published in the journal Nature, opens new tab. It is a type called a "hot Jupiter" - gas giants similar to the largest planet in our solar system, only much hotter owing to their close proximity to their host stars. This planet orbits 170 times closer to its host star than Jupiter does to the sun. It completes one orbit every two days as opposed to the 12 years Jupiter takes for one orbit of the sun. In fact, its orbit is 13 times nearer to its host star than our innermost planet Mercury is to the sun, leaving the temperature on the side of the planet facing the star at about 1,700 degrees Fahrenheit (930 degrees Celsius). "They are quite rare," Fu said of hot Jupiters. "About less than one in 100 star systems have them." This planet is located 64 light-years from Earth, considered in our neighborhood within the Milky Way galaxy, in the constellation Vulpecula. A light-year is the distance light travels in a year, 5.9 trillion miles (9.5 trillion km). "The close distance makes it bright and easy for detailed studies. For example, the hydrogen sulfide detection reported here would be much more challenging to make on other faraway planets," Fu said. The star it orbits is smaller and cooler than the sun, and only about a third as luminous. That star is part of a binary system, meaning it is gravitationally bound to another star. Webb, which became operational in 2022, observes a wider wavelength range than earlier space telescopes, allowing for more thorough examinations of exoplanet atmospheres.

Enhance Your Photos With NASA's Sharpening Technique

Incredible space photos like those from NASA don't look as stunning straight out of the telescope. They need significant processing, and a crucial part of that is sharpening. Coming to you from Unmesh Dinda with PiXimperfect, this fascinating video explores the APF-R plugin, developed by award-winning astrophotographer Christoph Kaltseis. APF-R stands for Absolute Point of Focus, and it's designed to enhance photo details without creating halos or artifacts. This technology has been used by space agencies with telescopes like the James Webb, and now, you can use it in Photoshop. The plugin allows for non-destructive editing, meaning you can adjust the radius and detail level without permanently altering your image. This feature is crucial for astrophotography, where preserving original details is vital. The video shows how APF-R compares to Photoshop's built-in sharpening tools. The plugin offers multiple rendering methods, each suited for different types of images. For instance, the "Center Weighted" method provides a balanced sharpening effect without making the image look overprocessed. Dinda explains how to fine-tune these settings to achieve the best results, making it clear why this plugin is a game-changer for photographers looking to enhance their images with precision. One notable feature of APF-R is its ability to work with different image types. The video demonstrates how the plugin enhances not just space photos but also landscapes and portraits. For portraits, APF-R can bring out skin textures and eye details without creating the unwanted halos that traditional sharpening methods often produce. Dinda also shows how to combine APF-R with other Photoshop tools, like Smart Sharpen, for even better results. This versatility makes APF-R a valuable addition to any photographer's toolkit. The plugin's cost is $50, which Dinda considers a bargain given its advanced capabilities. There's also a Creative Bundle subscription that includes APF-R and 20 other tools, offering great value for those looking to expand their editing options. Dinda provides discount codes in the video description, making this sophisticated tool more accessible. Check out the video above for the full rundown from Dinda.

Samsung Electronics wins cutting-edge AI chip order from Japan's Preferred Networks

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.