curl-impersonate, a modified version of curl, can impersonate the TLS and HTTP handshakes of major browsers, allowing users to bypass browser fingerprinting techniques employed by web services, promoting a more open and private web experience.
Orbital Materials, founded by an ex-DeepMind researcher, pilots an AI-designed purple powder that captures carbon more efficiently at data center temperatures, potentially reducing costs to $200/ton and enabling companies to offset emissions from energy-intensive data centers.
BrowseComp, a novel AI benchmark, evaluates web browsing skills by posing 1,266 complex queries requiring synthesis of information across multiple websites, with a deep research model achieving 51.5% accuracy, surpassing other models' performance on this challenging task.
Canva's Visual Suite 2.0 introduces AI-powered tools like visual spreadsheets, AI design partner, data storytelling, and personalized content creation, revolutionizing how teams collaborate visually in the AI era.
Google's new Gemini 2.5 Pro and Flash models boast enhanced thinking capabilities, enabling advanced AI applications, while Veo 2 offers high-quality video generation from text and image prompts, and the Live API allows developers to build real-time interactive applications.
Google aims to merge its AI models Gemini and Veo, combining Gemini's multimodal capabilities with Veo's understanding of real-world physics learned from YouTube data, to create an enhanced universal digital assistant with improved physical world comprehension.
YouTube throws its weight behind the 'No Fakes Act' to combat unauthorized AI-generated replicas, allowing individuals to request removal while balancing innovation, as it expands its pilot program enabling creators to detect and remove AI copies of themselves.
AI agents' energy usage varies based on model size, modality, and output length, necessitating a balance between performance and sustainability through efficient model selection, open-source frameworks, and task-appropriate approaches.
Google has unveiled Ironwood, a colossal seventh-generation Tensor Processing Unit (TPU) designed for inference, boasting unparalleled computational power over 24 times greater than the world's largest supercomputer, marking a groundbreaking leap in energy-efficient inference capabilities.
IBM has launched the z17 mainframe, engineered specifically for AI and generative AI workloads, featuring high-capacity Telum II processors, compatibility with IBM's Spyre accelerator chips, and new performance management and operating system tools.