AI Pulse

OpenAI has introduced AgentKit, a tool designed to streamline the process for enterprises to create agents, consolidating various tools into a unified platform. This announcement was made during OpenAI’s DevDay in San Francisco and positions AgentKit as a competitor to existing platforms like Zapier. The tool aims to reduce the complexity previously associated with agent […]

The ongoing cognitive migration driven by advancements in generative AI is a subject of significant interest, particularly from the perspective of those familiar with the technology sector. As someone with over 40 years in the industry, my experiences range from coding and systems analysis to IT project management and consulting. Although based primarily in the

A study conducted by Shanghai Jiao Tong University and the SII Generative AI Research Lab highlights that training large language models (LLMs) for complex tasks may not necessitate extensive datasets. The research introduces a framework named LIMI (Less Is More for Intelligent Agency), asserting that effective machine autonomy stems from the strategic curation of high-quality

OpenAI will host its third annual DevDay conference on Monday in San Francisco, gathering over 1,500 developers. The event is set against a backdrop of growing competition in the artificial intelligence sector, specifically from companies like Google, Anthropic, and Meta. OpenAI recently achieved a milestone when its Sora video generation app topped the Apple App

OpenAI’s annual developer conference, DevDay, commenced with the announcement by co-founder and CEO Sam Altman regarding a new “Apps SDK.” This SDK enables the development of applications within the ChatGPT environment, allowing companies to monetize these apps using OpenAI’s recently launched Agentic Commerce Protocol (ACP). Users will have the capability to access various applications without

Huawei’s Computing Systems Lab in Zurich has announced a new open-source quantization technique for large language models (LLMs) called Sinkhorn-Normalized Quantization (SINQ). This method aims to decrease memory requirements while maintaining output quality. The research team has made the code available on GitHub and Hugging Face under an Apache 2.0 license, facilitating free use and

For over fifty years, computing has predominantly followed the Von Neumann or Harvard architectural models, which have been foundational to CPUs, GPUs, and specialized accelerators. While newer architectures like Very Long Instruction Word (VLIW) and dataflow processors emerged to address performance constraints, none provided a holistic departure from the existing paradigm. Deterministic Execution represents a

AI technology is increasingly being promoted as a solution for automating numerous tasks, including customer service and software development. This has led to expectations that AI could significantly reduce the need for human workers. However, there are concerns that these expectations may not align with the current capabilities of AI. A technologist with experience in

OpenAI’s Sora is reportedly working on collaborations featuring fictional characters. Bill Peebles, head of the Sora team, indicated that multimedia cameos of copyrighted characters are planned for the platform, with further details expected soon. This development aims to address ongoing challenges related to unauthorized content, such as depictions of characters in inappropriate or illegal scenarios

AMD has entered into a five-year partnership with OpenAI to supply six gigawatts of processors for AI data centers. This collaboration aims to enhance OpenAI’s infrastructure to support the increasing computational requirements for AI applications, including ChatGPT. The initial phase of the agreement will involve the deployment of AMD Instinct MI450 GPUs, with one gigawatt