OpenAI is releasing its Codex agentic coding tool in research preview. Codex lets developers delegate simple, routine programming tasks to software engineering agents that can generate production-ready code, documenting the work as they go. Codex can work on many tasks in parallel doing things like writing software features, answering questions about a codebase, fixing bugs, and proposing pull requests for review. According to OpenAI, “each task runs in its own cloud sandbox environment,” preloaded within the user’s repository. OpenAI began releasing Codex last week to ChatGPT Pro, Enterprise, and Team users, with support for Plus and Edu coming soon. Continue reading OpenAI Adds Codex Software Agent to Some ChatGPT Plans

Google DeepMind has introduced AlphaEvolve, a coding agent that takes an evolutionary approach to general-purpose algorithm discovery and model optimization. AlphaEvolve combines the creative problem-solving abilities of Google’s Gemini models with automated evaluators that verify answers, then applies an evolutionary framework that improves on the most promising results. Evolutionary AI refers to techniques inspired by biological evolution, including natural selection, to optimize and design machine learning models. Continue reading Google DeepMind AlphaEvolve: Model of Algorithm Efficiency

Cisco Systems announced the development of a networking chip that uses quantum mechanics. The prototype entanglement source chip, which the company claims can generate up to one million entangled photon pairs per second (at room temperature), was created through the Outshift by Cisco incubator in partnership with UC Santa Barbara. While other companies are building quantum computers, Cisco is focusing on infrastructure, network and security frameworks. The company has also opened a new research facility in Santa Monica, California — Cisco Quantum Labs — dedicated to quantum networking tech. Continue reading Cisco Unveils a New Prototype Chip for Quantum Networking

Anthropic has upgraded its AI assistant Claude, adding Research, an autonomous capability that integrates with Google Workspace. Claude can now search and reference content in Google Docs as well as communications in Gmail and events in Calendar. “With Research, Claude can search across both your internal work context and the web to help you make decisions and take action faster than before,” Anthropic explains, turning the model into a “true virtual collaborator” for enterprise clients. The expansion puts Anthropic into more direct competition with OpenAI and Microsoft as well as Google with Gemini in the AI productivity space. Continue reading Anthropic Adds Deep Research, Google Integration to Claude

Anthropic’s Claude can now search the Internet in real time, allowing it to provide timely and relevant responses that are also more accurate than what the chatbot previously offered, according to the company. Claude incorporates direct citations for its Web-retrieved material, so users can fact-check its sources. “Instead of finding search results yourself, Claude processes and delivers relevant sources in a conversational format.” While this is not exactly groundbreaking — ChatGPT, Grok 3, Copilot, Perplexity and Gemini all have real-time Web retrieval and most include citations — Claude takes a slightly different approach. Continue reading Real-Time Web Access Informs Claude 3.7 Sonnet Responses

Chat interfaces powered by generative AI are impacting online shopping, according to an Adobe Analytics study that found that AI-influenced visits to U.S. retail website increased by 1,200 percent from July 2024 to February 2025. Adobe says this “significant surge” demonstrates an emerging retail AI economy. GenAI chat interfaces are “becoming a helpful assistant for compiling research before making a purchase,” influencing how consumers behave online, according to Adobe. While paid search and email continue to be the dominant traffic drivers, the past year’s growth in AI-aided shopping signals a shift. Continue reading Adobe Analytics: AI-Powered Online Shopping Surges in U.S.

Sony Corporation has launched the Sony Audio Institute at NYUs Steinhardt School of Culture, Education and Human Development, focusing on innovation in the business and technology of music. Opening this spring, the Sony Audio Institute will serve as an interdisciplinary collaboration that brings together the expertise of Sony’s professional and consumer audio businesses and their leading-edge technologies with NYU students, facilities and faculty. The institute opens with NYU Steinhardt Music Business Program Director Larry Miller at the helm. Miller will focus on the new outfit’s operations full time beginning this fall. Continue reading $7.5M Funds NYU’s Sony Audio Institute, Opening This Spring

Alibaba is making AI news again, releasing another Qwen reasoning model, QwQ-32B, which was trained and scaled using reinforcement learning (RL). The Qwen team says it “has the potential to enhance model performance beyond conventional pretraining and post-training methods.” QwQ-32B, a 32 billion parameter model, “achieves performance comparable to DeepSeek-R1, which boasts 671 billion parameters (with 37 billion activated),” Alibaba claims. While parameters refer to the total set of adjustable weights and biases in the model’s neural network, “activated” parameters are a subset used for a specific inference task, like generating a response. Continue reading Alibaba Says Qwen Reasoning Model on Par with DeepSeek

OpenAI is releasing a research preview of what it calls its “largest and best” chat model to date, GPT‑4.5, which scales unsupervised learning in pre-training and post-training. As a result, the new chat model has the ability to recognize patterns, draw connections, and generate creative insights without having to draw on time and energy consuming “reasoning.” GPT‑4.5 is currently available to ChatGPT Pro subscribers ($200 per month) and developers subscribing to OpenAI’s API tier. ChatGPT Plus and ChatGPT Team customers are expected to gain access this week. Continue reading OpenAI’s GPT-4.5 Model Sees Patterns and Thinks Creatively

Microsoft has created a quantum computing chip, Majorana 1, that relies on what it says is a “new state of matter” — one that exists beyond the primary liquid, solid, gas states that have underpinned science since Ancient Greece. Research into this fourth physical existence, called a “topological state,” earned three theoretical physicists the Nobel Prize in Physics in October. Unlike solids, liquids or gases, a topological state is not defined locally by how its particles are arranged, but by how their quantum wavefunction behaves — wrapping around itself globally, across the entire material. Continue reading Microsoft Calls New Topoconductor a Quantum Breakthrough

ChatGPT has a new “deep research” agent that OpenAI says uses reasoning to synthesize large amounts of online information and complete multi-step research tasks. “It accomplishes in tens of minutes what would take a human many hours,” OpenAI suggests, claiming it will “synthesize hundreds of online sources to create a comprehensive report at the level of a research analyst.” Powered by a version of the upcoming OpenAI o3 model optimized for web browsing and data analysis, the company says the deep research agent will typically take 5 to 30 minutes to complete its work. The agent is described as an ideal research tool for areas such as finance, science and engineering. Continue reading ChatGPT ‘Deep Research’ Agent Can Create Detailed Reports

Just weeks after Nvidia announced the availability of its $249 “compact AI supercomputer,” the Jetson Orin Nano Super Developer Kit for startups and hobbyists, CEO Jensen Huang revealed the company is planning to launch a personal AI supercomputer called Project Digits with a starting price of $3,000. The desktop-sized system features the GB10 Grace Blackwell Superchip, which enables it to handle AI models with up to 200 billion parameters. Nvidia claims there is enough processing power to run high-end AI models (performing up to one quadrillion AI calculations per second) while the compact system can run from a standard power outlet. Continue reading CES: Nvidia Will Launch a $3,000 Personal AI Supercomputer

When walking through the Japanese exhibits at CES 2025, it was difficult to miss the huge black spherical drone aircraft HAGAMOSphere that was prominently positioned as if demanding the passerby’s respect. And respect it deserved, for this drone prototype was one of this year’s CES Innovation Award recipients recognized for outstanding design and engineering in consumer technology. HAGAMOSphere’s innovation is its distinct ability to move both horizontally and vertically without tilting the aircraft. If the HAGAMOSphere is outfitted with a suitable camera, jerky movements in captured drone footage could potentially be eliminated or mitigated. Continue reading CES: Spherical Drone Design Could Benefit Media Production

Meta’s FAIR (Fundamental AI Research) team has unveiled recent work in areas ranging from transparency and safety to agents, and architectures for machine learning. The projects include Meta Motivo, a foundation model for controlling the behavior of virtual embodied agents, and Video Seal, an open-source model for video watermarking. All were developed in the unit’s pursuit of advanced machine intelligence, helping “models to learn new information more effectively and scale beyond current limits.” Meta announced it is sharing the new FAIR research, code, models and datasets so the research community can build upon its work. Continue reading Meta Rolls Out Watermarking, Behavioral and Concept Models

Google DeepMind’s new Genie 2 is a large foundation world model that generates interactive 3D worlds that are being likened to video games. “Games play a key role in the world of artificial intelligence research,” says Google DeepMind, noting “their engaging nature, challenges and measurable progress make them ideal environments to safely test and advance AI capabilities.” Based on a simple prompt image, Genie 2 is capable of producing “an endless variety of action-controllable, playable 3D environments” — suitable for training and evaluating embodied agents — that can be played by a human or AI agent using keyboard and mouse inputs. Continue reading DeepMind Genie 2 Creates Worlds That Emulate Video Games