OpenAI has rolled out a research preview of its latest GPT-4.5 model for Pro users, with plans to extend access to a wider audience over the next few weeks. The upgraded model boasts better pattern recognition, enhanced emotional intelligence, and an improved ability to generate creative insights.
The new version of GPT-4.5 also addresses a common issue with AI models, ‘hallucinations’ or the generation of false information, reducing its rate to 37.1%, compared to previous models’ higher rates. OpenAI’s CEO Sam Altman acknowledged some logistical challenges in the rollout, including GPU shortages, but assured users that more resources would be added soon.
While GPT-4.5 can handle tasks such as coding and content creation, it currently lacks some advanced features, such as voice and video capabilities. As OpenAI expands its offering, the company continues to refine the model to meet the growing demand for more powerful AI tools.
For more information on these topics, visit diplomacy.edu.
Chinese technology startups are scrambling for fresh funding to capitalise on the growing excitement around AI, fueled by DeepSeek‘s AI breakthrough and President Xi Jinping’s recent support for private enterprises. Companies such as AI-powered optics startup Rid Vision, brain-computer interface firm AI CARE Medical, and robotics maker Shanghai Qingbao Engine Robotics are among those seeking new onshore investments, according to venture capitalist Andrew Qian.
The surge in investment interest comes after a long period of stagnation in China‘s venture capital sector. New Access Capital, which has invested in several AI startups, noted that many businesses are now attracting both cooperation discussions and investment offers. The DeepSeek case has highlighted the rise of innovative Chinese tech firms, moving away from previous trends of copycat startups, and offering hope for a revitalised venture capital environment.
Despite the growing investor enthusiasm, challenges remain. The strict regulatory scrutiny of initial public offerings (IPOs) in China, coupled with heightened tensions between China and the US, complicate the prospects for exit strategies. Venture capital firms are particularly focused on AI-related businesses, with several companies in sectors like AI image generation and medical tech securing significant funding in recent weeks.
However, the overall fundraising environment remains difficult. Data shows a sharp decline in venture capital investments since 2021, and while investor sentiment has improved post-DeepSeek, concerns over the regulatory landscape and geopolitical tensions persist. Some analysts remain cautious about the near-term outlook for IPOs, especially for offshore listings.
For more information on these topics, visit diplomacy.edu.
Three men have been charged with fraud in Singapore as part of an investigation that may be linked to the illegal transfer of advanced Nvidia chips to the Chinese AI company, DeepSeek. The United States is also probing whether DeepSeek used US-made chips, which are prohibited from being shipped to China. This case is part of a wider investigation into potential AI chip smuggling operations involving several individuals and companies across countries, including Singapore.
The charges focus on two Singaporeans, Aaron Woon Guo Jie and Alan Wei Zhaolun, who are accused of falsely claiming that Nvidia chips would not be transferred to unauthorised recipients. A third man, Chinese national Li Ming, is charged with committing similar fraud regarding the shipment of servers. These actions are believed to be connected to DeepSeek’s use of Nvidia chips, although the authorities have not confirmed the company’s involvement.
If convicted, the men could face up to 20 years in prison, a fine, or both. Singapore’s authorities have been actively investigating the smuggling network, with police recently arresting nine individuals in a raid across 22 locations, seizing records and electronic evidence.
Nvidia’s business in Singapore is substantial, making it the company’s second-largest market after the US. However, only a small percentage of its revenue comes from direct shipments to the country. Singapore’s government has emphasised its commitment to strict enforcement of export controls, underscoring that the nation will not tolerate any attempts to evade international regulations.
For more information on these topics, visit diplomacy.edu.
Meta Platforms is seeking to raise $35 billion to help fund the development of its data centres in the US, with Apollo Global Management reportedly leading the financing. The need for increased computational power to support AI systems has driven Meta to focus heavily on expanding its AI infrastructure. The company has already committed to spending up to $65 billion this year on AI development.
The financing deal comes as Meta continues to build new data centres, including a planned $10 billion AI hub in Louisiana, US. The Louisiana centre will run on renewable energy, in partnership with the local utility provider Entergy. Despite the growing demand for data centre space, industry scepticism remains, especially regarding the financial returns from such large-scale investments.
Meta’s push for more AI infrastructure highlights the competitive race among tech companies to meet the rising demand for AI capabilities, with the company planning further expansions in the coming years.
For more information on these topics, visit diplomacy.edu.
Apple has rolled out a new feature called ‘age assurance’ to help protect children’s privacy while using apps. The technology allows parents to input their child’s age when setting up an account without disclosing sensitive information like birthdays or government IDs. Instead, parents can share a general ‘age range’ with app developers, putting them in control of what data is shared.
This move comes amid growing pressure from US lawmakers, including those in Utah and South Carolina, who are considering age-verification laws for social media apps. Apple has expressed concerns about collecting sensitive personal data for such verifications, arguing it would require users to hand over unnecessary details for apps that don’t require it.
The age assurance tool allows parents to maintain control over their children’s data while limiting what third parties can access. Meta, which has supported legislation for app stores to verify children’s ages, welcomed the new tech as a step in the right direction, though it raised concerns about its practical implementation.
For more information on these topics, visit diplomacy.edu.
Meta Platforms is preparing to test a paid subscription model for its AI chatbot, Meta AI, as it looks to compete with industry giants like OpenAI and Microsoft. The test will begin in the second quarter of this year, although significant revenue from the service is not expected until next year. Meta AI, launched in September 2023, offers virtual assistance powered by advanced language models.
CEO Mark Zuckerberg’s plans to boost the company’s AI capabilities include a $65 billion investment this year to expand AI infrastructure. Meta is also working on humanoid robots under its Reality Labs division. As tech companies race to dominate the AI space, Meta’s move to monetise its chatbot aligns with broader industry trends, with Microsoft and Amazon also making substantial investments in AI.
Despite the rising demand for AI services, Meta has not yet revealed the specific pricing for the new subscription service. However, the development highlights the company’s effort to enhance its AI offerings while positioning itself as a strong competitor in the rapidly expanding AI market.
For more information on these topics, visit diplomacy.edu.
TikTok, the popular video-sharing app owned by ByteDance, has unveiled plans to invest $8.8 billion in building data centres in Thailand over the next five years. The announcement was made by Helena Lersch, TikTok’s Vice President of Public Policy, during an event held in Bangkok on Friday. This investment marks a significant move as the company continues to expand its operations in the region.
The specific details of the investment remain unclear, particularly whether it includes a $3.8 billion agreement that was announced by Thailand’s investment board last month. The government’s investment board had previously detailed a deal aimed at boosting digital infrastructure in the country, but TikTok did not provide further clarification on the connection between the two.
This move highlights TikTok’s growing commitment to the Thai market and its broader strategy of increasing local data storage capabilities. As part of its ongoing efforts to expand its global presence, the company is investing in infrastructure to better serve its user base and meet regulatory requirements in key markets.
For more information on these topics, visit diplomacy.edu.
At a conference in Hyderabad this week, pharmaceutical companies highlighted the growing role of artificial intelligence in evolving the drug development process. Companies like Amgen and Parexel showcased how AI can reduce trial times, with Parexel demonstrating a model that speeds up drug safety reports by 30-45 minutes. Industry experts believe that AI can potentially halve the costs and time needed to develop drugs, cutting down the process from over a decade to just a few years.
Beyond speeding up trials, AI is also aiding in drug discovery and medical diagnostics. For instance, AI is being used to repurpose existing drugs for new uses and to analyse medical images, potentially identifying conditions such as cancer that might be missed by human doctors. Medtronic’s CTO stressed that integrating AI across the industry is essential to its future success.
Despite these technological strides, the conference was overshadowed by ongoing uncertainty regarding US President Trump’s proposed tariffs on pharmaceutical imports. Companies, particularly those based in India, expressed concern over the potential impact of these tariffs, with some, like Dr Reddy’s, opting to take a ‘wait and watch’ approach until further clarity emerges. This tariff uncertainty remains a pressing issue for India’s pharmaceutical industry, which is a major exporter of affordable drugs to the US.
For more information on these topics, visit diplomacy.edu.
Deutsche Telekom CEO Tim Hoettges has urged Germany and Europe to invest more in artificial intelligence and data centres to stay competitive with the US and Asia. Speaking after the release of the company’s annual results, Hoettges stressed the need for Europe to increase its computing power, citing a growing demand for data centres that he expects to rise by at least 30%. Deutsche Telekom is already expanding its infrastructure with plans to build four new data centres in Europe, aiming to create one gigawatt of capacity.
Hoettges also emphasised the importance of AI for Europe’s economic growth and sovereignty in the digital age. His comments come as Europe strives to catch up with major AI investments made by the US, with the European Commission pledging to mobilise 200 billion euros for AI development. This contrasts with the US, where private companies have committed up to $500 billion to AI infrastructure.
The call for greater AI investment follows a wave of AI advancements, such as China’s DeepSeek model, which has begun to challenge Western competitors. Hoettges warned that Germany must act quickly or risk falling behind in the global AI race.
For more information on these topics, visit diplomacy.edu.
Lonestar and Phison have successfully launched the first-ever lunar data centre infrastructure aboard a SpaceX Falcon 9 rocket. The mission, set to land on the moon on March 4, is a joint venture to send Phison’s Pascari solid-state drives (SSDs), packed with data from various clients, to the lunar surface. The companies plan to expand this infrastructure, with ambitions to scale up to a petabyte of storage in the coming years.
The idea of building a data centre on the moon originated in 2018, driven by the need for secure, off-Earth storage to safeguard against climate disasters and cyber-attacks. Lonestar’s CEO, Chris Stott, noted that data has become as valuable as oil, if not more precious. The SSDs were rigorously tested for space readiness, ensuring their durability for long-term use without the possibility of on-site repairs.
This mission is just the beginning, as Lonestar and Phison aim to revolutionise data storage with nearly limitless capacity and energy efficiency in space. As AI-driven demand for data storage continues to rise, this lunar project could set the stage for future space-based infrastructure, promising a new era in data management.
For more information on these topics, visit diplomacy.edu.
