Models & Labs
OpenAI Launches GPT-6 Astra, Claims AGI Milestone
WIRED AIhigh confidence
Why it matters
- →GPT-6 Astra could represent a step towards artificial general intelligence, a long-sought goal in AI development.
- →The model's ability to navigate computers and web browsers more effectively than humans could transform automation tasks.
- →Ensuring safety and alignment in increasingly complex AI models remains a critical challenge for the industry.
OpenAI has unveiled GPT-6 Astra, its latest AI model, which it claims could signal the beginning of the artificial general intelligence era. The model excels in navigating computers and web browsers, performing tasks such as booking appointments and searching for jobs more efficiently than humans. OpenAI plans to roll out GPT-6 Astra to paid customers, starting with select organizations. The company emphasizes the model's safety and alignment, though it acknowledges the increasing difficulty in monitoring these aspects as AI capabilities grow. This launch is part of OpenAI's strategy to advance its technology ahead of a potential IPO.
Read originalMore from WIRED AI
© WIRED AIMarket & Regulationbusiness
OpenAI Ends Partnership with Cursor Amid Musk Concerns
OpenAI has made the significant decision to end its partnership with Cursor, a prominent AI coding tool, following its acquisition by Elon Musk's SpaceX. This move is driven by OpenAI's apprehensions about the potential misuse of its technology by Musk's companies, despite the substantial financial impact of losing a billion-dollar customer. As OpenAI gears up for a public offering, it is focusing on maintaining control and independence, even if it means severing ties with a major revenue source. While this decision might affect its standing with developers, OpenAI is relying on its diverse revenue streams to absorb the financial impact.
WIRED AI
© WIRED AIModels & Labsmodels
Nvidia Unveils AI-Powered RTX Spark PCs
Nvidia's new RTX Spark 'superchip' marks a significant step in AI computing, enabling laptops and mini PCs to run AI models locally with impressive power. The Lenovo Yoga 9n 2-in-1, featuring this chip, aims to compete with high-end devices like the MacBook Pro, offering a sleek design and robust performance. This development allows users to execute AI workflows directly on their devices, reducing reliance on cloud services and enhancing privacy. With other manufacturers like Acer and Asus joining the fray, the landscape for AI-capable PCs is rapidly evolving, promising more autonomy and efficiency for AI enthusiasts.
WIRED AI
© WIRED AIMarket & Regulationother
Flock's AI Search Tool Faces Scrutiny
Flock Safety's AI search tool for police departments is under intense examination due to concerns about privacy and potential misuse. The tool enables officers to search for individuals based on descriptions, which raises questions about its accuracy and potential biases. Despite Flock's introduction of safeguards like mandatory case codes and automated audits, experts argue that the technology's reliability and transparency are still problematic. This situation underscores the ongoing conflict between the advancement of law enforcement technology and the protection of civil liberties. As Flock continues to adjust its system, the ethical implications of its use remain a contentious issue.
WIRED AI
More in Models & Labs
Models & Labsmodels
llama.cpp b10794 Release Expands Platform Support
The latest b10794 release of llama.cpp continues its trend of broadening platform compatibility, now supporting a wide array of systems including macOS, Linux, Windows, and openEuler. Notably, this update includes support for Vulkan and ROCm 10.0 on Ubuntu, as well as CUDA 13 on Windows, which enhances performance options for developers using these platforms. While the release doesn't introduce new model architectures, it solidifies llama.cpp's position as a versatile inference runtime across diverse hardware configurations. This update is a testament to llama.cpp's commitment to accessibility and performance optimization for developers working with AI models.
llama.cpp Releases
Models & Labsmodels
llama.cpp b10795 release enhances SYCL fusion
The b10795 release of llama.cpp brings notable improvements in SYCL fusion, specifically by combining operations like RMS_NORM+MUL+ADD and ADD+ADD. This enhancement, under GGML_SYCL_ENABLE_FUSION, boosts performance for supported data types, while unsupported combinations revert to standard methods. The update continues to support a wide array of platforms, including macOS with Apple Silicon, Ubuntu with Vulkan, and Windows with CUDA 12 and 13. This makes llama.cpp a robust choice for developers working across different hardware environments, ensuring efficient AI processing and broad compatibility.
llama.cpp Releases
Models & Labsmodels
llama.cpp b10796 Release Adds n_expert_used_max Function
The latest release of llama.cpp, b10796, introduces the n_expert_used_max function, enhancing the model's ability to handle expert layers. This update addresses previous issues where models with expert layers failed to load due to missing checks. By implementing this function, the software can now better manage the number of experts per layer, ensuring smoother model loading and operation. This release doesn't introduce new models but focuses on refining the existing infrastructure to support more complex configurations.
llama.cpp Releases