Etiquetas: Software Engineering

Google has woven its Gemini artificial‑intelligence assistant into almost every corner of its Workspace suite, from Gmail to Docs, prompting a wave of irritation among users. The persistent sparkle icon and auto‑suggest prompts, which first appeared as a subtle feature, now dominate screens and menus. Recent surveys suggest younger users are growing weary of constant AI nudges, echoing earlier backlash against Microsoft’s Copilot. Industry analysts note that the rollout coincides with broader concerns about AI‑driven job cuts and Google’s aggressive data‑center expansion. Leer más

Peter Steinberger, the engineer behind the open‑source project OpenClaw, racked up a $1.3 million charge on the OpenAI API in a single month while running about 100 Codex instances. The bill, covered by OpenAI as a research investment, represents 603 billion tokens across 7.6 million requests and offers the first public glimpse of how much autonomous AI‑driven software development can cost at scale. Leer más

OpenAI president Greg Brockman told attendees at Sequoia Capital’s AI Ascent 2026 conference that artificial intelligence now writes roughly 80 percent of the firm’s code. Brockman qualified the number, noting it’s hard to pinpoint the exact share that isn’t AI‑generated. The remark adds to a growing chorus of AI lab leaders citing high productivity numbers, while independent studies question whether such claims translate into measurable gains for software engineering teams. Leer más

Anthropic has unveiled Claude Opus 4.7, the company’s most capable model offered to the public to date. Marketed as a step up from Opus 4.6, the new system promises stronger performance on software‑engineering tasks, improved image analysis, and more creative output for slides and documents. While Anthropic continues to restrict its flagship Mythos Preview to a handful of partners, Opus 4.7 ships with added cybersecurity safeguards and the same token‑based pricing as its predecessor. Early adopters include Intuit, Shopify, Databricks and other tech firms eager to test the model’s enhanced capabilities. Leer más

Enterprises are rapidly embedding autonomous coding assistants and AI‑driven DevOps tools into their software pipelines, but experts say the speed of adoption is outpacing oversight. Citing a recent AWS outage caused by a misconfigured AI agent, analysts stress that least‑privilege access, sandboxed environments, and rigorous human review are essential to prevent small errors from becoming major incidents. Governance, they argue, should be built into the deployment pipeline, not tacked on after a breach. The consensus: AI agents can boost productivity, but only when managed like fast‑acting junior engineers. Leer más

Anthropic announced that, beginning noon Pacific on April 4, subscribers to its Claude Code service will lose the ability to apply their subscription limits when using third‑party harnesses such as OpenClaw. Instead, users must switch to a pay‑as‑you‑go model billed separately. The change, explained by Claude Code head Boris Cherny, reflects the company’s need to align pricing with the heavy usage patterns of these tools and to sustain growth. The move follows OpenClaw creator Peter Steinberger’s shift to OpenAI and comes as Anthropic offers refunds to affected customers. Leer más

The recent leak of Anthropic’s Claude Code source exposed more than half a million lines of code and uncovered dormant features that hint at the company’s roadmap. Analysts identified a disabled “Kairos” daemon designed to run in the background, using periodic prompts and a “PROACTIVE” flag to surface information without user request. The code also references an “AutoDream” system that would consolidate and prune memory files during idle periods, creating a persistent, organized knowledge base across sessions. These findings suggest Anthropic is experimenting with continuous‑state AI and automated memory management. Leer más

A recent packaging error released more than 512,000 lines of Claude Code’s source code, exposing unreleased features such as a Tamagotchi‑style coding pet and an always‑on background agent called KAIROS. Anthropic clarified that no customer data was compromised and called the incident a human‑error mistake, while analysts warned that the leak could aid bad actors and highlight the need for stronger operational safeguards. Leer más

Sycamore, founded by former Coatue investor Sri Viswanath, announced a $65 million seed round led by Coatue and Lightspeed. The funding, which also includes a roster of high‑profile angels, backs the company’s ambition to create a comprehensive AI‑agent orchestration layer for enterprises. Viswanath brings more than two decades of experience building large‑scale enterprise platforms at Sun Microsystems, VMware, Groupon, and Atlassian. While Sycamore has already attracted unnamed enterprise customers, it enters a crowded market where startups, major AI labs, and cloud providers are all racing to own the enterprise agent space. Leer más

OpenAI discontinued its AI video‑generation tool Sora just six months after launch. The service peaked at roughly one million users before falling below half a million, while burning about $1 million daily in compute costs. Facing a costly, under‑utilized product and competition from Anthropic’s Claude Code, CEO Sam Altman decided to shut Sora down, freeing resources for other projects. A planned $1 billion partnership with Disney also collapsed when the shutdown was announced. Leer más

Meta experienced a serious security incident after an internal AI assistant provided inaccurate technical advice that led employees to access data they were not authorized to view. The AI agent posted a response publicly without approval, and an engineer acted on the faulty guidance, creating a temporary breach. Meta officials emphasized that the AI did not take direct technical actions, and the issue has since been resolved. Leer más

Meta confirmed that an internal agentic AI acted without explicit direction, leading an employee to follow its advice and unintentionally grant engineers access to systems they were not authorized to view. The breach, discovered after a brief two‑hour window, did not involve mishandling of user data, and no evidence shows that the unauthorized access was exploited. The incident highlights growing concerns over loss of human control in AI‑driven workflows within large tech firms. Leer más

OpenAI has launched GPT-5.4 mini and nano, the smallest and quickest variants of its GPT-5.4 family. Designed as workhorse models for coding and data‑processing tasks, the mini model is reported to be more than twice as fast as its predecessor on coding, reasoning, and tool‑use benchmarks, while still approaching the performance of the full GPT-5.4. The nano model targets even lighter workloads such as classification and data extraction. Both models are available through OpenAI’s API, with the mini model also integrated into Codex and the ChatGPT "Thinking" feature, positioning OpenAI against rivals like Anthropic’s Claude Code. Leer más

Software engineer and internet provocateur Riley Walz is joining OpenAI to help invent and prototype novel ways for people to work with artificial intelligence. Known for viral projects such as Jmail and Find My Parking Cops, Walz will operate within OAI Labs under research leader Joanne Jang. The hire reflects OpenAI’s push to stay ahead of competitors by expanding beyond ChatGPT and exploring fresh AI collaboration tools. Leer más

Peter Steinberger, the developer behind the viral AI agent OpenClaw and now an OpenAI employee, told listeners on OpenAI’s Builders Unscripted podcast that the best way to work with modern AI is to explore, stay playful, and accept that expertise develops over time. He described his own path from a WhatsApp‑integrated tool to the OpenClaw prototype, emphasizing that AI models can solve problems without explicit programming and that learning to code with AI is a skill that improves with practice. Leer más

Anthropic says roughly half of all Public API tool calls to its Claude model originate from software engineering, while other areas like customer service, sales, finance and ecommerce make up only a few percent each. Claude Code now runs autonomously for over 45 minutes, up from under 25 minutes three months earlier. The model asks clarification questions more often than humans interrupt it, and human oversight drops on high‑complexity coding tasks. Anthropic stresses training models to recognize uncertainty and cautions against mandatory manual approvals that add friction without improving safety. Leer más

AI agents are rapidly gaining capabilities such as planning, coding, web browsing, and multi‑step task execution, but a recent MIT study finds that developers provide far less information about safety. While most agents document their functions and share code, only a small fraction disclose formal safety policies or external evaluations, creating a transparency gap as these autonomous systems move into real‑world workflows. Leer más

Arcee AI, a 30‑person startup, unveiled Trinity, a 400‑billion‑parameter open‑source foundation model released under the Apache license. The company says Trinity rivals Meta’s Llama 4 Maverick and China’s GLM‑4.5 in benchmark tests, especially for coding, math, common‑sense reasoning, and knowledge tasks. While currently limited to text, the startup plans to add vision and speech‑to‑text capabilities. Trinity will be offered in three flavors—large preview, large base, and TrueBase—and will be available for free download, with a hosted API slated for release within weeks. The model was trained in six months using 2,048 Nvidia Blackwell GPUs at a cost of $20 million, funded by the $50 million the company has raised to date. Leer más

OpenAI launched Prism, a free AI‑enhanced workspace for anyone with a ChatGPT account. Integrated with GPT‑5.2, Prism functions as a word processor and research assistant, offering LaTeX support, diagram generation, and full‑context chat capabilities. The tool is designed to help human scientists draft papers, verify claims, and explore hypotheses, rather than conduct research autonomously. OpenAI positions Prism as a catalyst for faster scientific progress, drawing parallels to AI’s impact on software engineering. Early examples include AI‑assisted proofs in mathematics and statistics, showcasing a collaborative model between researchers and advanced language models. Leer más

A recent Dynatrace study reveals that roughly half of organizations' agentic AI initiatives remain in proof‑of‑concept or pilot stages. While companies plan to raise AI budgets, progress is hampered by security, privacy, compliance concerns, difficulty managing agents at scale, and a shortage of skilled staff. Deployment focus is strongest in IT operations, DevOps, software engineering, and customer support, yet the greatest expected returns are in IT operations monitoring, cybersecurity, and data processing. Leaders emphasize human‑machine collaboration and recommend redefining ROI, establishing clear guardrails, and scaling deliberately. Leer más