Author: Advanced AI Editor

If you are interested in building your very own local deep research AI assistant, you might be interested in Google’s Gemma 3 AI models. They represent a significant advancement in artificial intelligence, offering a compact yet robust solution tailored for local deployment. Derived from the larger Gemini series, these models combine high performance with a strong emphasis on privacy and accessibility. Featuring multimodal capabilities, support for 140 languages, and the ability to generate structured outputs, Gemma 3 is engineered to meet diverse AI research assistant and productivity demands. These models are open source and optimized for local use, allowing you…

On Wednesday, OpenAI announced the release of two new models—o3 and o4-mini—that combine simulated reasoning capabilities with access to functions like web browsing and coding. These models mark the first time OpenAI’s reasoning-focused models can use every ChatGPT tool simultaneously, including visual analysis and image generation. OpenAI announced o3 in December, and until now, only less capable derivative models named “o3-mini” and “03-mini-high” have been available. However, the new models replace their predecessors—o1 and o3-mini. OpenAI is rolling out access today for ChatGPT Plus, Pro, and Team users, with Enterprise and Edu customers gaining access next week. Free users can…

NVIDIA announced that GPU cloud platform CoreWeave is among the first cloud providers to bring NVIDIA GB200 NVL72 systems online at scale, with Cohere, IBM and Mistral AI using them for model training and deployment. The CoreWeave-NVIDIA relationship is well-known – NVIDIA has invested heavily in CoreWeave when CoreWeave was a private company and now, as a publicly held one, and CoreWeave, which is an NVIDIA preferred cloud services provider, has adopted NVIDIA GPUs in its AI cloud infrastructure. Last year, the company was among the first to offer NVIDIA H100 and NVIDIA H200 GPUs, and was one of the first to demo NVIDIA…

Deepseek logo is seen in this illustration taken, January 27, 2025. — ReutersWASHINGTON: The Trump administration is weighing penalties that would block China’s DeepSeek from buying US technology and is debating barring Americans’ access to its services, The New York Times reported on Wednesday.The launch of China’s low-cost AI model, DeepSeek, has rattled the AI ecosystem. The US government has since taken steps to crack down on the Chinese start-up and its support from chip maker Nvidia.Nvidia’s AI chips have been a key focus of US export controls, as US officials aim to prevent the most advanced chips from being…

Alibaba Group Holding (NYSE BABA) co-founder Jack Ma appeared at Alibaba Cloud’s campus in Hangzhou on Thursday to celebrate its 15th anniversary, SCMP reports. On Wednesday, Alibaba’s cloud unit showcased its Blossom project, which aims to accelerate AI adoption by providing its clients with infrastructure, AI models, data, and other services. The company upgraded global AI tools with Qwen-Max, a new reasoning model, and $1/year developer kits. Also Read: China’s DeepSeek Teams Up With Tsinghua University To Raise AI Bar, Boost Reasoning Capabilities Alibaba committed at least 380 billion yuan ($52 billion) in cloud computing and AI infrastructure over the…

Copilot Vision, Microsoft’s AI assistant feature that can interpret what’s on your screen and help you use apps, is now available for free use within the Edge browser, Mustafa Suleyman, CEO of Microsoft AI, announced on Bluesky today. Vision is a “talk-based experience,” as Microsoft calls it, meaning you use it by speaking into the air, then waiting for Copilot to respond.Suleyman says if you opt into the feature, Copilot Vision can “literally see what you see on screen.” Suleyman suggests having Copilot Vision guide you through a recipe while you cook or having it “decode” job descriptions “and jump…

Gemini is getting even more powerful as Google has rolled out the AI assistant’s all-new video generation feature. The new video generation feature will be powered by the company’s all-new Veo 2, the tech giant’s latest text-to-video model. Paid subscribers can generate up to eight-second videos at 720p resolution using simple prompts. Currently, Google states that the more details the prompt includes, the better the video will be. According to Android Headlines, the output will be an MP4 file in a 16:9 ratio. Users can access the feature by clicking on the dropdown menu within the model. After each video…

A group of ten law professors have submitted a so called amicus brief as part of the AI copyright legal battle between several authors and Meta setting out the arguments for why training AI models with existing content does not qualify as ‘fair use’ under American copyright law. It follows the recent submission of another amicus brief by four different professors who argued the opposite. The new brief insists that Meta’s use of the authors’ work when training its Llama AI model is “a commercial use that takes the expression in those works”, is “not transformative” and does not qualify as…

April 16, 2025 — CoreWeave this week became one of the first cloud providers to bring NVIDIA GB200 NVL72 systems online for customers at scale, and AI frontier companies Cohere, IBM and Mistral AI are already using them to train and deploy next-generation AI models and applications. Thousands of NVIDIA Blackwell GPUs are now turning raw data into intelligence at unprecedented speed, with many more coming online soon. Credit: NVIDIA. CoreWeave, the first cloud provider to make NVIDIA Grace Blackwell generally available, has already shown incredible results in MLPerf benchmarks with NVIDIA GB200 NVL72 — a powerful rack-scale accelerated computing platform…

Image: Anthropic Anthropic has introduced a new Research function in its Claude artificial intelligence model, enabling the system to autonomously perform multi-step investigations. Claude can deliver a well-reasoned response with verifiable citations in just minutes, aiming to strike a “balance of speed and quality.” The AI startup says Claude responds to prompts “agentically,” independently working out what information it needs to grab next to build the correct response. It “explores different angles of your question automatically and works through open questions systematically,” Anthropic said in its press release. The Research feature is currently in early beta and can be accessed…