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
Australia pledges to provide more funds to Pacific island banks to counter China's influence
Australia pledged on Tuesday to increase investment in Pacific island nations, offering A$6.3 million ($4.3 million) to support their financial systems. Some Western banks are cutting ties with the region because of risk factors, while China is trying to increase its influence there. Some Western bankers have terminated long-standing banking relationships with small Pacific nations, while others are considering closing operations and restricting access to dollar-denominated bank accounts in those countries. "We know that the Pacific is the fastest-moving region in the world for correspondent banking services," Australian Treasurer Jim Chalmers said in a speech at the Pacific Banking Forum in Brisbane. "What's at stake here is the Pacific's ability to engage with the world," he said, with much of the region at risk of being cut off from the global financial system. Chalmers said Australia would provide A$6.3 million ($4.3 million) to the Pacific to develop secure digital identity infrastructure and strengthen compliance with anti-money laundering and counter-terrorist financing requirements. Experts say Western banks are de-risking to meet financial regulations, making it harder for them to do business in Pacific island nations, where compliance standards sometimes lag, undermining their financial resilience. Australia's ANZ Bank is in talks with governments about how to make its Pacific island businesses more profitable amid concerns about rising Chinese influence as financial services leave the West, Chief Executive Shayne Elliott said Tuesday. ANZ is the largest bank in the Pacific region, with operations in nine countries, though some of those businesses are not financially sustainable, Elliott said in an interview on the sidelines of the forum. "If we were there purely for commercial purposes, we would have closed it a long time ago," he said. Western countries, which have traditionally dominated the Pacific, are increasingly concerned about China's plans to expand its influence in the region after it signed several major defense, trade and financial agreements with the region. Bank of China signed an agreement with Nauru this year to explore opportunities in the country, following Australia's Bendigo Bank saying it would withdraw from the country. Mr. Chalmers said Australia was working with Nauru to ensure that banking services in the country could continue. ANZ Bank exited its retail business in Papua New Guinea in recent years, while Westpac considered selling its operations in Fiji and Papua New Guinea but decided to keep them. The Pacific lost about 80% of its correspondent banking relationships for dollar-denominated services between 2011 and 2022, Australian Assistant Treasurer Stephen Jones told the forum, which was co-hosted by Australia and the United States. “We would be very concerned if there were countries acting in the region whose primary objective was to advance their own national interests rather than the interests of Pacific island countries,” Mr. Jones said on the first day of the forum in Brisbane. He made the comment when asked about Chinese banks filling a vacuum in the Pacific. Meanwhile, Washington is stepping up efforts to support Pacific island countries in limiting Chinese influence. "We recognize the economic and strategic importance of the Pacific region, and we are committed to deepening engagement and cooperation with our allies and partners to enhance financial connectivity, investment and integration," said Brian Nelson, U.S. Treasury Undersecretary for Counterterrorism and Financial Intelligence. The United States is aware of the problem of Western banks de-risking in the Pacific region and is committed to addressing it, Nelson told the forum's participants. He said data showed that the number of correspondent banking relationships in the Pacific region has declined at twice the global average rate over the past decade, and the World Bank and the Asian Development Bank are developing plans to improve correspondent banking relationships. U.S. Treasury Secretary Janet Yellen said in a video address to the forum on Monday (July 8) that the United States is focused on supporting economic resilience in the Pacific region, including by strengthening access to correspondent banks. She said that when President Biden and Australian Prime Minister Anthony Albanese met at the White House last year, they particularly emphasized the importance of increasing economic connectivity, development and opportunities in the Pacific region, and a key to achieving that goal is to ensure that people and businesses in the region have access to the global financial system.
TikTok to introduce a new feature that can clone your voice with AI in just 10 second
Use of AI is certainly the hottest topic in the tech industry and every major and minor player in this industry is using AI in some way. Tools like ChatGPT can help you do a wide range of task and even help you generate images. The other thing is - Voice Cloning. OpenAI recently introduced a voice engine that can generate clone of your voice with just 15 seconds of your audio. There is no shortage of voice cloning tools on the web which can help you do the same. The newest tech giant which is going to use AI to clone your voice is - TikTok. We all know TikTok, posting short videos with filters, effects and all other kind of things. So TikTok found a way to use the voice cloning AI in its app. TikTok is working on this feature, which does not seem to really have a proper name, it just references it as "Create your voice with AI" and "TikTok Voice Library". In the latest version of TikTok I came across some strings which indicates that TikTok is working on it. I was also able to access the initial UI which introduces the feature and was able to see the terms and condition of "TikTok Voice Library" which user have to accept in order to use the feature. Here are the screenshots from the app- As you can in the screenshot above, this is the initial screen which a user will see for the first time they access this feature. Tiktok claims that it can create an AI verison of your voice in just 10 seconds. The generated AI voice clone can be used with text-to-speech in TikTok videos. It also outline the process of how it will work. You have to record yourself speaking and TikTok will process the voice and use information about your voice to generate your AI voice. When it comes to privacy, your AI voice will stay private and you can delete it anytime. Tapping the "Continue" button brings "TikTok Voice Library Terms" screen which a user should definitely read, you can see here and read as well - How it will work After agreeing to terms and conditions I was introduced with a screen where TikTok will show some text and user have to press the record button while reading the text. Now unfortunately I did not see any text. This is probably because the feature is not fully ready or the backend from which it fetches the text is not live yet. Manually pressing the record button and saying random things also shows an error. So, it's also not possible to provide any sample voice generated with it and see how it compares to other voice cloning competitors. If it starts working someday, it will process your recorded voice and generate AI version of your voice. Here is a screenshot of that screen - My guess is that whenever the feature starts working, users have to clone voice only one time and the saved AI voice can be used through the text-to-speech method to add voice in your videos. You just have to type the words, choice is yours :p
Wto: Members have more trade promotion measures than restrictions
The latest trade monitor released recently by the World Trade Organization shows that between mid-October 2023 and mid-May 2024, WTO members continued to introduce more trade promotion measures than trade restrictive measures. The WTO said it was an important signal of members' commitment to keep trade flowing amid the current geopolitical uncertainty. According to WTO statistics, during the monitoring period, WTO members adopted 169 trade promotion measures on commodities, more than the 99 trade restrictive measures introduced. Most of the measures are aimed at imports. Commenting on the findings, WTO Director-General Ngozi Okonjo-Iweala said that despite the challenging geopolitical environment, this latest trade monitoring report highlights the resilience of world trade. Even against the backdrop of rising protectionist pressures and signs of economic fragmentation, governments around the world are taking meaningful steps to liberalize and boost trade. This demonstrates the benefits of trade on people's purchasing power, business competitiveness and price stability. The WTO monitoring also identified significant new developments in economic support measures. Subsidies as part of industrial policy are increasing rapidly, especially in areas related to climate change and national security.
"Corrupt Politicians GPT" "Fiscal Bill GPT", Kenyan protesters use AI to "protest"
In the past few weeks of anti-government activities in Kenya, AI tools have been creatively used by protesters to serve protests. According to the US "Flag" News Agency on July 5, protests in Kenya triggered by the 2024 fiscal bill are still continuing. In the past few weeks, Kenyan protesters, mainly young people, have creatively developed a series of AI tools to assist anti-government activities. The Kenyan government expressed concern about the risks associated with the use of AI tools in protests. Kelvin Onkundi, a software engineer in Kenya, developed the "Fiscal Bill GPT", which operates similarly to ChatGPT and can receive questions about the fiscal bill and generate responses. Martin Siele, a reporter from the "Flag" News Agency, analyzed: "The 'Fiscal Bill GPT' can convert professional terms in many legislative fields into easy-to-understand information for protesters, helping Kenyans understand the potential impact of the fiscal bill." Another software engineer, Marion Kavengi, developed the "SHIF GPT" to provide Kenyans with information about the upcoming Social Health Insurance Fund (SHIF). In addition to AI tools designed to help people understand controversial policies, protesters have also developed "Corrupt Politicians GPT" to assist protest demonstrations. After entering the name of a politician on the platform, the platform will generate a list of corruption scandals about the politician in chronological order. Developer BenwithSon wrote on the social platform X on June 28: "'Corrupt Politicians GPT' allows people to search for any scandal related to any politician. I have seen some leaders stand at the forefront of the political arena, but they are corrupt behind the scenes." Kenyan Chief Minister and Foreign Minister Mudavadi issued a communiqué to ambassadors of various countries in Nairobi on July 2 local time on protests and relevant government measures, expressing concerns about the use of AI and false information in protests. Mudavadi said: "AI technology is used by people with ulterior motives, which will fill the global information system with false narratives." The Kenya Times reported on June 30 that AI technology enables people to force the government to increase transparency and strengthen accountability, and its role in Kenyan political activities is becoming increasingly prominent. Martin Siller believes that AI is reshaping African political behavior in many ways. AI is a new tool for both governments and opposition parties in Africa, but Kenya is one of the African countries with the most developers, and its young protesters are particularly good at using AI technology to fight the government. The 2024 fiscal bill voted and passed by the Kenyan National Assembly on June 25 clearly stated that additional taxes will be levied to repay the interest on high sovereign debt, triggering large-scale demonstrations. After President Ruto announced the withdrawal of the tax increase bill on the evening of the 26th, demonstrations in many parts of Kenya continued. According to Reuters on July 3, Kenyan anti-government protesters are re-adjusting their activities to prevent the protests from turning into violent incidents.
Morning Bid: Eyes switch to inflation vs elections, Powell up
A look at the day ahead in U.S. and global markets from Mike Dolan After an intense month focused on election risk around the world, markets quickly switched back to the more prosaic matter of the cost of money - and whether disinflation is resuming to the extent it allows borrowing costs to finally fall. Thursday's U.S. consumer price update for June is the key moment of the week for many investors - with the headline rate expected to have fallen two tenths of a percentage point to 3.1% but with 'core' rates still stuck at 3.4%. With Federal Reserve chair Jerome Powell starting his two-pronged semi-annual congressional testimony later on Tuesday, the consensus CPI forecast probably reflects what the central bank thinks of the situation right now - encouraging but not there yet. But as the U.S. unemployment rate is now back above 4.0% for the first time since late 2021, markets may look for a more nuanced approach from the Fed chair that sees it increasingly wary of a sudden weakening of the labor market as real time quarterly GDP estimates ebb again to about 1.5%. There were some other reasons for Fed optimism in the lead up to the testimony. The path U.S. inflation is expected to follow over coming years generally softened in June, amid retreating projections of price increases for a wide array of consumer goods and services, a New York Fed survey showed on Monday. Inflation a year from now was seen at 3% as of June - down from the expected rise of 3.2% in May - and five-year expectations fell to 2.8% from 3%. Crude oil prices are better behaved this week, too, falling more than 3% from the 10-week highs hit late last week and halving the annual oil price gain to 10%. The losses on Tuesday came after a hurricane that hit a key U.S. oil-producing hub in Texas caused less damage than many in markets had expected - easing concerns over supply disruption. Before Powell starts speaking later, there will also be an update on U.S. small business confidence for last month.