Ai Model Training
- All
- News
- Videos
-
Microsoft’s Mustafa Suleyman Highlights the Need to be Self-Sufficient in AI: Report
Microsoft’s AI chief, Mustafa Suleyman, said the company must become self-sufficient in AI, training frontier models using its own data and compute. In an interview, he highlighted Microsoft’s goal to build end-to-end AI systems under its new Superintelligence team. The move follows a revised deal with OpenAI that now allows Microsoft to indepe...
-
OpenAI Faces Backlash from Studio Ghibli, Bandai Namco Over AI-Generated Anime Videos
OpenAI has been urged by Japan’s Content Overseas Distribution Association (CODA) to stop training its AI models on copyrighted anime and gaming content. CODA, representing major publishers behind Dragon Ball, Naruto, One Piece, Final Fantasy and Studio Ghibli classics, said Sora 2 outputs closely resemble protected works, potentially indicating ...
-
Grab Superapp Says AI Models Struggle to Understand Asian Languages
Grab says it built its own lightweight vision model after finding both proprietary and open-source AI systems struggled with Southeast Asian languages, leading to errors and high latency in document processing. The move underscores a broader concern: frontier models from firms like Google, OpenAI and Anthropic still lack robust multilingual support...
-
OpenAI Turns to Amazon in $38 Billion Cloud Services Deal After Restructuring
OpenAI has entered a seven-year agreement with Amazon worth $38 billion to purchase cloud services, after last week's restructuring that granted it more autonomy. The deal provides OpenAI access to hundreds of thousands of Nvidia GPUs, essential for training and operating its AI models, including ChatGPT, as it accelerates its artificial intelligen...
-
Apple Shares Massive Dataset to Help Researchers Build Nano Banana-Like AI Models
Apple has unveiled Pico-Banana-400K, a new open-source dataset featuring 4,00,000 real and AI-edited image pairs for training text-based image editing models. Available under a research-only licence, it aims to help developers build better image-editing AIs. The release comes as Apple’s own AI progress appears to have slowed, with its long-promis...
-
Adobe Introduces AI Assistant in Photoshop, New AI Audio and Video Tools in Firefly
Adobe has unveiled a major lineup of AI upgrades across Photoshop, Express, and Firefly, including new AI assistants and audio-video generation tools. The update introduces Firefly Image Model 5 with multi-layered editing, custom Firefly training for individuals, and integrations with ElevenLabs and Topaz. These features mark Adobe’s biggest step...
-
Researchers Use AI to Predict Storm Surges Faster and More Accurately
AI-powered models are transforming storm surge forecasting by generating accurate coastal flood predictions in minutes. Using past storm data and neural networks, these systems outperform traditional physics models, helping coastal communities prepare faster as climate change drives stronger hurricanes and higher seas.
-
Nvidia DGX Spark Supercomputer With Grace Blackwell Chipset to Go on Sale Starting October 15
Nvidia has announced the DGX Spark, touted as the world’s smallest AI supercomputer, which will start shipping on October 15. Priced at $3,999 (roughly Rs. 3.5 lakh), it offers one petaflop of AI performance and 128GB of unified memory. Despite its compact size, Nvidia says it can train and fine-tune large AI models efficiently.
-
Apple Sued Over Use Of Copyrighted Books To Train Apple Intelligence
Apple was hit with a lawsuit in California federal court by a pair of neuroscientists who say that the tech company misused thousands of copyrighted books to train its Apple Intelligence artificial intelligence model.
-
Microsoft Azure Unveils Nvidia GB300 NVL72 Cluster Built for OpenAI’s AI Workloads
Microsoft Azure has launched the NDv6 GB300 VM series, the first supercomputing-scale cluster of Nvidia GB300 NVL72 systems. Designed for OpenAI’s demanding AI workloads, it features 4,600 GPUs, high-speed NVLink/NVSwitch connectivity, and 37TB of memory per rack. The system supports agentic and multimodal AI workflows, enabling model training in...
-
Elon Musk Says xAI Game Studio Will Release an AI-Generated Game in 2026
Elon Musk announced that xAI’s gaming division, xAI Game Studio, will release an AI-generated game before the end of next year. The company has also posted a job listing for a video game tutor to train its AI model, Grok, in gaming concepts. The announcement coincides with updates to Grok Imagine’s video generation capabilities.
-
LinkedIn Wants to Train AI Models on User Data, But This One Step Stops It
LinkedIn is updating its terms to begin collecting user data from November 3 to train its artificial intelligence (AI) models. The professional networking platform confirmed that profile details and public posts will be used, while private messages will remain excluded. Some data will also be shared with Microsoft and affiliates for targeted ads, t...
-
China's Deepseek Says Its Hit AI Model Cost Just $294,000 To Train
Chinese AI developer DeepSeek said it spent $294,000 on training its R1 model, much lower than figures reported for US rivals, in a paper that is likely to reignite debate over Beijing's place in the race to develop artificial intelligence.
-
Google's Viral Gemini Banana Trend is Fun, But It Comes With a Privacy Risk
Gemini’s Nano Banana trend has gone viral on social media, surpassing 500 million image generations. The custom AI model lets users upload images and apply detailed edits with prompts. While popular, it has reignited privacy concerns, as Google stores all user conversations and uploaded images, which can be used to train its AI systems and servic...
-
Google VaultGemma: 5 Things to Know About the AI Model That Puts Privacy First
Google has unveiled VaultGemma, a one-billion-parameter privacy-focused AI model trained using differential privacy techniques to prevent memorisation of sensitive data. By adding calibrated noise during training, the model avoids exposing personal identifiers like addresses or credit cards. While this approach reduces accuracy and increases comput...
-
Microsoft’s Mustafa Suleyman Highlights the Need to be Self-Sufficient in AI: Report
Microsoft’s AI chief, Mustafa Suleyman, said the company must become self-sufficient in AI, training frontier models using its own data and compute. In an interview, he highlighted Microsoft’s goal to build end-to-end AI systems under its new Superintelligence team. The move follows a revised deal with OpenAI that now allows Microsoft to indepe...
-
OpenAI Faces Backlash from Studio Ghibli, Bandai Namco Over AI-Generated Anime Videos
OpenAI has been urged by Japan’s Content Overseas Distribution Association (CODA) to stop training its AI models on copyrighted anime and gaming content. CODA, representing major publishers behind Dragon Ball, Naruto, One Piece, Final Fantasy and Studio Ghibli classics, said Sora 2 outputs closely resemble protected works, potentially indicating ...
-
Grab Superapp Says AI Models Struggle to Understand Asian Languages
Grab says it built its own lightweight vision model after finding both proprietary and open-source AI systems struggled with Southeast Asian languages, leading to errors and high latency in document processing. The move underscores a broader concern: frontier models from firms like Google, OpenAI and Anthropic still lack robust multilingual support...
-
OpenAI Turns to Amazon in $38 Billion Cloud Services Deal After Restructuring
OpenAI has entered a seven-year agreement with Amazon worth $38 billion to purchase cloud services, after last week's restructuring that granted it more autonomy. The deal provides OpenAI access to hundreds of thousands of Nvidia GPUs, essential for training and operating its AI models, including ChatGPT, as it accelerates its artificial intelligen...
-
Apple Shares Massive Dataset to Help Researchers Build Nano Banana-Like AI Models
Apple has unveiled Pico-Banana-400K, a new open-source dataset featuring 4,00,000 real and AI-edited image pairs for training text-based image editing models. Available under a research-only licence, it aims to help developers build better image-editing AIs. The release comes as Apple’s own AI progress appears to have slowed, with its long-promis...
-
Adobe Introduces AI Assistant in Photoshop, New AI Audio and Video Tools in Firefly
Adobe has unveiled a major lineup of AI upgrades across Photoshop, Express, and Firefly, including new AI assistants and audio-video generation tools. The update introduces Firefly Image Model 5 with multi-layered editing, custom Firefly training for individuals, and integrations with ElevenLabs and Topaz. These features mark Adobe’s biggest step...
-
Researchers Use AI to Predict Storm Surges Faster and More Accurately
AI-powered models are transforming storm surge forecasting by generating accurate coastal flood predictions in minutes. Using past storm data and neural networks, these systems outperform traditional physics models, helping coastal communities prepare faster as climate change drives stronger hurricanes and higher seas.
-
Nvidia DGX Spark Supercomputer With Grace Blackwell Chipset to Go on Sale Starting October 15
Nvidia has announced the DGX Spark, touted as the world’s smallest AI supercomputer, which will start shipping on October 15. Priced at $3,999 (roughly Rs. 3.5 lakh), it offers one petaflop of AI performance and 128GB of unified memory. Despite its compact size, Nvidia says it can train and fine-tune large AI models efficiently.
-
Apple Sued Over Use Of Copyrighted Books To Train Apple Intelligence
Apple was hit with a lawsuit in California federal court by a pair of neuroscientists who say that the tech company misused thousands of copyrighted books to train its Apple Intelligence artificial intelligence model.
-
Microsoft Azure Unveils Nvidia GB300 NVL72 Cluster Built for OpenAI’s AI Workloads
Microsoft Azure has launched the NDv6 GB300 VM series, the first supercomputing-scale cluster of Nvidia GB300 NVL72 systems. Designed for OpenAI’s demanding AI workloads, it features 4,600 GPUs, high-speed NVLink/NVSwitch connectivity, and 37TB of memory per rack. The system supports agentic and multimodal AI workflows, enabling model training in...
-
Elon Musk Says xAI Game Studio Will Release an AI-Generated Game in 2026
Elon Musk announced that xAI’s gaming division, xAI Game Studio, will release an AI-generated game before the end of next year. The company has also posted a job listing for a video game tutor to train its AI model, Grok, in gaming concepts. The announcement coincides with updates to Grok Imagine’s video generation capabilities.
-
LinkedIn Wants to Train AI Models on User Data, But This One Step Stops It
LinkedIn is updating its terms to begin collecting user data from November 3 to train its artificial intelligence (AI) models. The professional networking platform confirmed that profile details and public posts will be used, while private messages will remain excluded. Some data will also be shared with Microsoft and affiliates for targeted ads, t...
-
China's Deepseek Says Its Hit AI Model Cost Just $294,000 To Train
Chinese AI developer DeepSeek said it spent $294,000 on training its R1 model, much lower than figures reported for US rivals, in a paper that is likely to reignite debate over Beijing's place in the race to develop artificial intelligence.
-
Google's Viral Gemini Banana Trend is Fun, But It Comes With a Privacy Risk
Gemini’s Nano Banana trend has gone viral on social media, surpassing 500 million image generations. The custom AI model lets users upload images and apply detailed edits with prompts. While popular, it has reignited privacy concerns, as Google stores all user conversations and uploaded images, which can be used to train its AI systems and servic...
-
Google VaultGemma: 5 Things to Know About the AI Model That Puts Privacy First
Google has unveiled VaultGemma, a one-billion-parameter privacy-focused AI model trained using differential privacy techniques to prevent memorisation of sensitive data. By adding calibrated noise during training, the model avoids exposing personal identifiers like addresses or credit cards. While this approach reduces accuracy and increases comput...
