2026-06-10 17:01
Boris Cherny, head of Claude Code at Anthropic, has shifted from prompting AI directly to "writing loops." This approach, also known as loop engineering, focuses on designing systems for coding agents to operate autonomously. Peter Steinberger and Google engineer Addy Osmani emphasize that this method transforms coding agents into long-running execution systems, enhancing their efficiency and capabilities. Loop engineering combines scheduled execution, isolated workspaces, and persistent...
Source: The New Stack
Janakiram MSV
2026-06-10 16:46
AWS recently announced the general availability of its Graviton5 processor, which powers two new Amazon EC2 instances: M9g and M9gd. These instances are designed for general-purpose workloads and high-speed local storage, respectively. 🖥️ The Graviton5 features a doubled core count of 192, promising up to 25% better compute performance than its predecessor. Notably, these instances utilize the sixth-generation Nitro System, which includes the new Nitro Isolation Engine. 🔒 The Nitro Isolation...
Source: The New Stack
Frederic Lardinois
2026-06-10 16:40
Last week, Microsoft disabled 73 GitHub repositories after a malware attack aimed at stealing developers’ credentials. The incident originated from a malicious commit uploaded to the durabletask repository. Despite the shutdown, Microsoft has not disclosed how many developers were impacted or specific details about the breach. GitHub stated the repos violated its terms of service but did not provide further clarification. This event highlights ongoing security concerns in the industry, as...
Source: The New Stack
Meredith Shubel
2026-06-10 14:00
🚀 Databricks has launched OpenSharing, the successor to its Delta Sharing protocol. This new update supports Apache Iceberg REST Catalog clients and includes on-premises storage vendors like Everpure, MinIO, and Qumulo. OpenSharing enhances secure data sharing by extending capabilities to AI models and unstructured data, helping organizations share assets effectively. Since its debut in 2021, Delta Sharing has become widely adopted, with thousands of partners benefiting from its open...
Source: The New Stack
Frederic Lardinois
2026-06-10 13:00
🚀 Ramp is introducing Applied AI Solutions, a service that deploys forward deployed engineers (FDEs) into finance teams to enhance automation in accounts payable, procurement, and more. Unlike traditional AI tools, this high-touch service focuses on understanding the unique context of each enterprise. Ramp aims to bridge the gap between AI ambition and results, ensuring that finance decisions are informed and governed effectively. #AI #Finance #TechInnovation #Automation #BusinessSolutions
Source: The New Stack
Matthew Burns
2026-06-09 19:58
🌐 In early June, three companies launched AI coding agents aimed at enhancing team collaboration. Cognition introduced Devin Desktop, providing a unified console for teams. Microsoft unveiled Rayfin, which manages agent-built app deployments, while Augment Code rolled out Cosmos to coordinate multiple agents. These innovations signify a shift from individual tools to team infrastructures, similar to the evolution of version control systems. Teams now have a shared harness that remembers...
Source: The New Stack
Janakiram MSV
2026-06-09 19:30
🚀 Anthropic has launched Fable 5, its first general Mythos-class model, with essential guardrails for safety. 🔍 Fable 5 is available now via API and select platforms, priced at $10 per million input tokens. Users with certain subscriptions can access it for free until June 22. 📈 The model excels in benchmarks, scoring 80% on SWE-Bench Pro, outperforming other models in various tasks. 🗓️ After June 22, usage will require credits, pending capacity improvements. #AI #Anthropic #Fable5...
Source: The New Stack
Frederic Lardinois
2026-06-09 18:48
AI is transforming software security, and the Java ecosystem is adapting. 🌐 Broadcom recently announced significant security updates for the Spring Framework, marking a crucial moment in its 23-year history. This response comes as reported vulnerabilities surged over 1,700% within a month. AI is now identifying vulnerabilities at an unprecedented pace, shifting the focus to rapid fixes. Notably, Spring is utilized by over half of Fortune 500 companies, making its security vital as Java...
Source: The New Stack
Darryl K. Taft
2026-06-09 16:17
🚀 Lindy, a no-code AI agent platform, has transitioned from Anthropic to DeepSeek, claiming significant cost savings and enhanced performance. 💰 This shift comes amid rising inference costs that challenge sustainable AI deployment. Companies like GitHub and Uber have experienced substantial billing increases, prompting a reevaluation of AI spending. 💡 The Linux Foundation is responding by launching the Tokenomics Foundation, aimed at creating standards for AI token costs. #AI #TechNews...
Source: The New Stack
Paul Sawers
2026-06-09 14:40
Medium’s recommendation system aims to keep readers engaged by processing user activity signals and correlating them with new articles. 📚 The feature store plays a crucial role in this system, enabling real-time data storage and retrieval to support fast user interactions. Andréas Saudemont explains how the team improved their data model to handle over 1 million operations per second effectively. Learn more about their challenges and solutions! 💡 #Medium #DataScience #MachineLearning...
Source: The New Stack
Cynthia Dunlop
2026-06-09 13:00
🚀 Code reviews are facing a critical shift. Instead of focusing solely on code, experts suggest prioritizing reviews of intent, specs, and acceptance criteria. This change addresses the bottleneck caused by AI-generated code, which outpaces human review capabilities. Research shows that as code changes grow, the usefulness of comments declines. In fact, teams using AI now merge 98% more PRs, leading to longer review times. AI tools can assist in reviews, but they bring challenges like non-...
Source: The New Stack
Ankit Jain
2026-06-09 10:00
The article discusses the end of the "tokenmaxxing" trend in AI, where companies competed to use the most AI tokens without clear business outcomes. As costs rise, firms are shifting from reckless token usage to stricter budget control. Revenium, an observability tool, has introduced AI Insights to help businesses recover wasted AI spending by providing actionable recommendations. This marks a significant pivot in AI management as organizations seek efficiency and optimization. 🌐📊💡 #AI...
Source: The New Stack
Chris J. Preimesberger
2026-06-08 19:25
Memory constraints are now a major challenge for tech teams, driven by hardware limitations and rising costs. 📉 Over the past year, prices for high-bandwidth memory and DRAM have surged by 170%. This has prompted a shift from a buy-all-you-can approach to data-driven optimization strategies. AI is emerging as a solution, helping enterprises optimize workloads and reduce costs by up to 50%. 🌐💡 Learn how to transition from inefficiency to smarter IT practices. #TechTrends #AI...
Source: The New Stack
Jennifer Riggins
2026-06-08 19:05
🚀 Microsoft is encouraging enterprises to transition from Azure Repos to GitHub, despite recent reliability issues at GitHub. The platform has faced outages impacting critical services like CI/CD pipelines, prompting public apologies. 📊 GitHub's usage has surged, processing 1.4 billion commits monthly, with AI contributing significantly. To handle this load, GitHub aims for 30 times its current capacity. 🔄 Microsoft’s new tool, Enterprise Live Migrations (ELM), offers a smoother transition,...
Source: The New Stack
Paul Sawers
2026-06-08 19:01
🚀 Anthropic has released Claude Opus 4.8, introducing dynamic workflows in Claude Code. This feature allows Claude to function like a team of developers, enabling it to create its own orchestration scripts and run multiple subagents simultaneously. 💻 According to Anthropic, this upgrade significantly reduces project timelines, with tasks that typically take months now completed in days. 🔍 Ole Lehmann highlights that dynamic workflows elevate Claude’s capabilities, likening the AI’s role to a...
Source: The New Stack
Jessica Wachtel
2026-06-08 17:35
🚀 Anthropic has announced a limited-time promotion that doubles the five-hour usage limit in Claude Cowork at no extra cost! This offer is available for users on Pro, Max, and Team plans, including legacy Enterprise users. It runs from June 5 to July 5, 2026, with no action needed to participate. Note that this promotion applies only to Claude Cowork, while limits for other Claude products remain unchanged. #Anthropic #ClaudeCowork #AI #TechNews #Promotions
Source: The New Stack
Adrian Bridgwater
2026-06-08 14:00
🚀 Apache Cassandra 6.0 is now in alpha, shifting its focus from new features to taking back tasks historically handled by application teams. This version introduces the Accord consensus protocol, enhancing multi-partition workflows by moving some coordination work into the database itself. Another key change is the implementation of Transactional Cluster Metadata (TCM), which improves cluster metadata coordination and addresses some long-standing challenges. Explore the implications of these...
Source: The New Stack
Anil Inamdar
2026-06-08 13:00
AI agents are transforming workplace applications, but this shift poses risks for identity and access management (IAM). Traditional IAM was designed for human users, while AI agents operate in dynamic, unpredictable ways, leading to potential security concerns. If AI agents are given poorly managed static credentials, it can result in unauthorized access and operational issues. To mitigate these risks, organizations need robust monitoring of AI agent activities to ensure compliance with...
Source: The New Stack
Adrian Bridgwater
2026-06-08 12:00
🚀 Microsoft is shifting its focus in the AI landscape. At Build 2026, they introduced significant updates to Microsoft Foundry, emphasizing reliability over just capability. Key updates include a hosted agent infrastructure and new tools for developers, ensuring production-ready AI applications. The Foundry's managed runtime allows seamless deployment without the need for rewrites. This change highlights Microsoft's commitment to building dependable AI solutions that meet real-world demands....
Source: The New Stack
Darryl K. Taft
2026-06-08 11:00
🚀 Microsoft recently announced exciting updates to Visual Studio at the Build 2026 conference. The key focus is on integrating AI agents into the development process, allowing developers to use their own AI models and keys. This change aims to support teams with specific compliance and environment needs. 🔑 Additionally, Microsoft is embedding AI directly into existing tools like the debugger and profiler, facilitating quicker issue identification and resolution. #VisualStudio #AI...
Source: The New Stack
Darryl K. Taft
2026-06-07 16:30
AI coding tools are showing significant results, with adoption rates rising from 76% in 2024 to 90% in 2025. 📈 Deployment rates have surged from 357 to nearly 1,000 per month, representing a 175x improvement for high-performing teams. However, increased speed alone is not enough; direction is crucial. 🎯 Effective product velocity combines speed with strategic focus to enhance outcomes. Organizations must prioritize a balanced approach to thrive in this rapidly changing landscape. ⚖️...
Source: The New Stack
Paul Stovell
2026-06-07 15:00
🚀 Microsoft has announced that its agent runtime, OpenClaw, is now free, opting to focus on the layers surrounding it. At Build 2026, Microsoft launched Scout, its first always-on work agent, powered by OpenClaw. This move mirrors the Android model, where the base is free, but value lies in the services around it. Microsoft is enhancing enterprise policy controls and integrating with various partners, positioning itself strategically in the agent landscape. #MicrosoftBuild #OpenSource...
Source: The New Stack
Janakiram MSV
2026-06-06 15:00
At Snowflake Summit 26 in San Francisco, the focus shifted from large language models to the concept of the agentic enterprise. AI's role is evolving from writing code to orchestrating complex workflows with minimal human intervention. Christian Kleinerman, Snowflake’s EVP of Product, detailed this approach, highlighting how it simplifies the data lifecycle. Snowflake's new tools aim to enhance developer efficiency by reducing the need for multiple apps, allowing developers to focus on...
Source: The New Stack
Chris J. Preimesberger
2026-06-06 14:00
🚀 At AI Native DevCon, Netlify CTO Dana Lawson shared insights on the evolving role of software engineers. He emphasized that writing code is no longer the core of the job. Instead, understanding agent experience (AX) is key. As AI enables more people to create, engineers must focus on overseeing production and ensuring seamless collaboration between humans and AI. This shift brings both opportunities and uncertainties for developers as the industry rapidly changes. #TechInnovation...
Source: The New Stack
Jennifer Riggins
2026-06-06 11:00
Transitioning AI from Jupyter Notebook to production requires significant changes in mindset and engineering practices. 🛠️ In notebooks, models thrive in controlled environments, but production demands scalable, observable systems that handle uncertainty. Key factors include high accuracy, reproducible training, and robust monitoring. 📈 Effective experimentation is crucial; chaotic testing can lead to failures in production. A disciplined approach ensures models perform reliably in real-world...
Source: The New Stack
Zziwa Raymond Ian
2026-06-06 10:31
🚨 Important AI Development Alert! 🚨 OpenClaw's use of Gavriel Cohen’s code highlights a significant issue of accountability in AI agents. Cohen discovered his work in OpenClaw without consent and stepped away from the project. This raises questions about code ownership and transparency in AI technologies. With AI systems gaining autonomy, the need for clear accountability is critical. Developers are now concerned about whether their code is being used without acknowledgment. Stay informed...
Source: The New Stack
Matthew Burns
2026-06-05 19:03
🚀 Making apps is now more accessible, but monetizing them can be challenging. Replit aims to bridge this gap by creating a financial stack for vibe-coded apps. 🛒 Their latest feature includes a Shopify integration, allowing users to easily set up custom storefronts without prior e-commerce experience. 💻 Users can create a fully functional store in about ten minutes, connecting their Shopify account to finalize their setup. #VibeCoding #Replit #Ecommerce #Shopify #AppDevelopment
Source: The New Stack
Paul Sawers
2026-06-05 19:01
Cloudflare has acquired VoidZero, known for its JavaScript development tools like Vite. Both companies emphasize that VoidZero's open-source and community-driven ethos will remain unchanged. The developer community's reaction is mixed, with some praising the acquisition for its potential benefits, while others express concerns about future changes and vendor lock-in. Cloudflare aims to support the continued development of VoidZero's projects. #Cloudflare #VoidZero #OpenSource #WebDevelopment...
Source: The New Stack
Adrian Bridgwater
2026-06-04 19:36
Cursor has adjusted its pricing model in response to industry shifts, cutting costs by 20% for its Teams plan and introducing a new Premium tier for power users. This change comes as flat-rate pricing models in AI coding are being replaced by consumption-based billing. In addition, Cursor has launched a usage pool for its Composer model and enhanced spend alert features for better budget management. This aims to give organizations clearer control over AI expenses. #AICoding #TechNews...
Source: The New Stack
Paul Sawers
2026-06-04 19:30
🚀 Google has launched Gemma 4 12B, a powerful new model designed for standard laptops. It boasts high-performance, multi-modal intelligence while requiring just 16GB of VRAM. 📊 This model performs nearly as well as the 26B version but with a significantly smaller memory footprint. Developers are excited about its accessibility for advanced reasoning tasks, even offline. 🎤 Notably, Gemma 4 12B supports native audio inputs, streamlining the processing of both audio and visual data. #GoogleGemma...
Source: The New Stack
Meredith Shubel
2026-06-04 19:22
Snowflake is addressing challenges developers face with tool fragmentation and context switching. The company introduced Snowflake CoCo, an AI-based coding agent designed to streamline enterprise development. CoCo helps automate workflows and develop apps with ease. Additionally, Snowflake launched Datastream, a Kafka-compatible service for real-time data streaming, enhancing AI app creation. Learn how these innovations aim to boost productivity without sacrificing governance. #Snowflake #AI...
Source: The New Stack
Chris J. Preimesberger
2026-06-04 18:53
Autonomous agents are evolving beyond chatbots, now capable of reasoning and planning. However, they face a significant challenge: databases. 🔍 Andy Pavlo from Carnegie Mellon University highlighted the complexities at the Percona Live 2026 conference. Databases require high correctness and performance, making them difficult for agents to manage effectively. ⚙️ Pavlo discussed two types of AI agents—tuning and coding. Tuning agents aim to optimize databases, traditionally a job for human...
Source: The New Stack
Chris J. Preimesberger
2026-06-04 18:35
The article highlights that the main challenge for enterprise AI is not technology, but rather operations. As hybrid environments grow complex, managing various applications and infrastructures becomes difficult. Latha Vishnubhotla from HPE notes the significant time spent on maintaining systems after initial setup. The piece emphasizes the importance of the operations platform in navigating this complexity, especially with the rise of AI workloads. Read more about how HPE’s GreenLake is...
Source: The New Stack
TNS Staff
2026-06-04 18:12
🚨 Cybersecurity alert fatigue is a growing concern for SOC teams, facing over 3,000 alerts daily, with 73% being false positives. This challenge costs U.S. enterprises approximately $3.3 billion annually. 🔍 Effective alert tuning is essential, but it comes with risks. Analysts must balance reducing noise without missing genuine threats. 💡 Continuous monitoring and measurable tuning are critical to maintain security without creating blind spots. #Cybersecurity #AlertFatigue #SOC #ITSecurity...
Source: The New Stack
Todd R. Weiss
2026-06-04 17:18
The rise of autonomous databases is reshaping the data landscape, but human expertise remains crucial. 🤖 Vadim Tkachenko from Percona emphasizes that while automation is increasing, the role of database administrators (DBAs) is evolving rather than disappearing. They will shift from maintenance to focusing on data architecture and governance. Machine learning will take over routine tasks, allowing humans to concentrate on strategic issues like data integrity and access patterns. 🌐 This...
Source: The New Stack
Chris J. Preimesberger
2026-06-04 16:45
Securing Kubernetes is increasingly complex, especially with the rise of AI workloads. 🌐 AI introduces new risks, expanding the attack surface due to dynamic processes and unpredictable traffic. Traditional security measures may not suffice. ⚠️ Notably, advancements in zero trust networking are essential. Managed services like Azure Kubernetes Service are enhancing security by providing network-isolated clusters, reducing the risk of data exfiltration. 🔒 Understanding these challenges is...
Source: The New Stack
Mary Branscombe
2026-06-04 16:21
🚀 Asana has launched Dash, an AI assistant designed to streamline work from Slack and email into trackable tasks. This personal AI "chief of staff" helps users manage follow-ups and provides daily briefs on pending items. 🗂️ Asana also enhanced its AI “teammates,” offering industry-specific solutions and better integration with tools like Gmail and Slack. Learn more about how Asana aims to improve workplace productivity through AI. 🤖✨ #Asana #AI #WorkplaceInnovation #Productivity #TechNews
Source: The New Stack
Frederic Lardinois
2026-06-04 16:12
🚀 Nvidia has launched its new model, Nemotron 3 Ultra, featuring 550 billion parameters. It is now available on platforms like Hugging Face and ModelScope. This model leverages a mixture-of-experts technique and supports up to 1 million tokens in context. Nvidia highlights its speed, claiming it is significantly faster than earlier models and can save users up to 30% on token costs. However, while it outperforms competitors like Kimi-K2.6 and Qwen-3.5, it still trails behind some benchmarks...
Source: The New Stack
Frederic Lardinois
2026-06-04 15:23
🚀 Asaf Wiener, CEO of Mate Security, emphasizes the importance of managing AI model costs in startup operations. He identified high expenses early on, prompting a structural shift in how his team handles AI models. By breaking down costs, they can make informed decisions on feature pricing and performance. This approach highlights the growing significance of inference costs for AI startups, shaping their financial futures. 📊💡 #AIStrategy #StartupLeadership #CostManagement #TechInnovation
Source: The New Stack
Matthew Burns
2026-06-03 20:10
The AI cost crisis is gaining attention as the Linux Foundation announces the launch of the Tokenomics Foundation. This body will focus on creating standards and best practices for AI token consumption economics. The foundation aims to clarify the costs associated with AI, which have rapidly outpaced traditional finance models. Major companies like Google and Microsoft are backing this initiative. As AI token usage grows, so do the expenses, with recent data showing a significant increase in...
Source: The New Stack
Paul Sawers