The artificial intelligence (AI) realm witnesses rapid developments around the world, reshaping the world of digital transformation and accelerating the pace of AI adoption. In this weekly newsletter, FollowICT is highlighting the most important news and advancements in AI around the globe.
Musk Says xAI Is Building New Data Center
Elon Musk said in an interview with CNBC on Tuesday that xAI is building another data center for artificial intelligence, expected to house one million AI chips from Nvidia.
Musk added that the data center will be located “near Memphis,” where xAI already has a facility with over 200,000 Nvidia chips used to train the company’s Grok model. The Information previously reported that xAI was looking for a new data center so it could purchase more GPUs from Nvidia.
Earlier this year, Musk’s company purchased a one-million-square-foot warehouse in Memphis, Tennessee. An xAI executive told The Daily Memphian in March that the new building could hold up to 350,000 AI chips. It is still unclear whether this is the same site Musk referenced for the million Nvidia chips.
Musk also said he plans to keep buying chips from Nvidia, AMD, and possibly other companies to train and run Grok. He added, “As long as Nvidia makes the best, we’ll keep buying from them.”
Anthropic Releases New Generation of AI Models
Anthropic announced on Thursday new versions of its largest models, Claude Opus 4 and Claude Sonnet 4, according to a blog post.
The company said the updated models perform better on complex, long-duration programming tasks and have improved memory for prior context. The models can also switch between “thinking”, exploring different ways to solve a problem, and “tool use,” meaning they can use external tools, applications, and databases to find answers. Crucially, if a model gets stuck using a tool, it can revert to logical reasoning mode to analyze what went wrong and self-correct.
The new models will be available to Claude users on the Pro, Max, Team, and Enterprise plans. Pricing will remain the same as previous versions of Opus and Sonnet.
Nvidia CEO Says China’s AI Researchers Are “World-Class”
Nvidia CEO Jensen Huang believes that China’s AI researchers are among the best in the world, and it’s no surprise that U.S. companies are hiring many of them.
In an interview with Ben Thompson, author of Stratechery, Huang said, “AI researchers in China are world-class. They’re not just Chinese AI researchers; they are world-class AI researchers.” He added, “You walk through the halls of Anthropic, OpenAI, or DeepMind, and there are a ton of AI researchers from China. Of course that makes sense, they’re exceptional, so it doesn’t surprise me they’re doing exceptional work.”
Huang also said that China is making “amazing achievements” in AI, with companies like DeepSeek and Manus emerging as strong competitors to US-based models.
OpenAI Acquires AI Hardware Startup for $6.5 BLN
OpenAI, led by CEO Sam Altman, announced Wednesday that it has acquired a hardware startup called IO, founded by former Apple design chief Jony Ive. The deal is valued at approximately $6.5 billion, according to a company spokesperson speaking to Business Insider.
Ive and his design firm LoveFrom began collaborating with Altman and OpenAI two years ago, and IO was formed a year later with Scott Cann, Tang Tan, and Evans Hankey. Ive left Apple in 2019 after nearly 30 years.
IO will now merge with OpenAI “to work more closely with research, engineering, and product teams in San Francisco,” according to a press release.
A video posted on OpenAI’s social media shows Altman and Ive discussing their plan to develop “a family of AI devices.”
Altman also revealed that OpenAI is working on a screenless AI device that is aware of its user’s surroundings but is not wearable.
The new devices are expected to help expand OpenAI’s reach to more users, boosting subscriptions and user engagement.
Google Unveils Stunning Array of New AI Products
At its Google I/O 2025 developer conference on Tuesday, Google announced a host of new AI products, including Flow, a new AI-powered video tool aimed at filmmakers.
The tool is powered by three AI models: Veo for video generation, Imagen for image generation, and Gemini for text and prompts.
Google also revealed that its Android development IDE is getting an AI upgrade.
In addition to launching Gemini 2.5 Pro, Android Studio will soon get a new “AI Agent Mode” called Journeys for tackling complex development tasks.
Using Gemini, Journeys lets developers test apps by describing actions and expected outcomes in natural language. Gemini will then run the tests for them.
Google also announced major updates to its Gemini chatbot, including expanded multimodal capabilities, updated AI models, and deeper integration with Google’s product suite.
Starting Tuesday, Google is rolling out camera and screen sharing in the Gemini Live app on both iOS and Android. Powered by Project Astra, this feature allows near-real-time voice conversations with Gemini while streaming video from the phone’s camera or screen.
Google also revealed that its experimental AI project Project Mariner, which browses and interacts with websites, will be available to more users and developers. The tool now supports nearly a dozen simultaneous tasks.
U.S. subscribers to Google’s new AI Ultra plan, priced at $249.99/month, will get access to Mariner. Google says more countries will be added soon. Mariner will also be integrated into Gemini’s API and Vertex AI platform, enabling developers to build AI-agent-powered apps.
Nvidia Plans to Build 1st AI Supercomputer in Taiwan
Nvidia CEO Jensen Huang announced Monday that the company plans to build the world’s first AI supercomputer in Taiwan.
The announcement came during his keynote at Computex, the world’s largest semiconductor industry conference.
Huang said Nvidia will partner with Taiwanese tech giants like Foxconn, TSMC, and the Taiwanese government to create the supercomputer, aiming to support advanced infrastructure and an AI ecosystem.
Microsoft Aims to Use AI to Accelerate Scientific Discovery
At its Build 2025 conference on Monday, Microsoft introduced Microsoft Discovery, a platform that uses AI agents to “transform the scientific discovery process.” The company says the platform is scalable and can handle scientific workloads end-to-end.
Microsoft Discovery is designed to accelerate research through every stage of discovery using powerful AI, from scientific insight generation to hypothesis formulation, result simulation, and analysis.
According to Microsoft, the platform allows scientists and researchers to collaborate with a team of specialized AI agents to achieve results quickly, accurately, and comprehensively by leveraging the latest AI and supercomputing technologies.
Microsoft Launches AI Agent for Programming in GitHub
Microsoft announced the launch of a new AI agent for programming integrated into GitHub Copilot, its development tool suite.
The AI agent can fix bugs, add features, and improve documentation on behalf of developers, the company said at its annual Build developer conference on Monday.
Meta Launches Startup Program to Promote Its Llama AI Models
Meta has launched a new program to encourage startups to adopt its Llama AI models.
The Llama for Startups program offers direct support from Meta’s Llama team, and in some cases, funding. Any US-based startup with less than $10 million in funding, at least one developer, and focused on building generative AI applications is eligible to apply before the May 30 deadline.
In a blog post, Meta said selected participants may receive up to $6,000 per month for six months to help cover costs of building and refining their generative AI solutions. “Our experts will work closely with them to get started and explore advanced Llama use cases that could benefit their startups,” the company added.
Grok 3 from xAI Now Available on Microsoft Azure
Microsoft has become one of the first companies to offer managed access to Grok, the AI model developed by Elon Musk’s AI startup xAI.
Grok 3 and Grok 3 Mini are now available through Microsoft’s Azure AI Foundry, and they come with the same SLAs (Service Level Agreements) that customers expect from any Microsoft product.
Microsoft will bill users directly for Grok, just like it does for other models hosted on Azure AI Foundry.
Amazon Introduces Short AI-Generated Audio Summaries for Products
Amazon announced on Wednesday that it is testing short, AI-generated audio summaries on select product pages.
These summaries are provided by what Amazon calls “AI shopping experts,” who discuss key product features, customer reviews, and relevant information from across the web.
Customers can access the AI-generated product summaries by tapping the “Listen to the Highlights” button in the Amazon Shopping app. Amazon says the feature is being tested on products that typically require more careful research before purchasing.
