T
The argument over whether engineers should stop reading code is over, because we already have. Faros AI’s new report shows a 76.3% increase in pull requests merged without review since their last report in April. Many companies now run AI reviews on 50% to 80% of their PRs. AI review took over bug-finding. But it didn’t take over the conversation where a team debates what it decided and why, and that is the part I missed too. The part AI review doesn’t solve Most teams are stuck in one of two…
T
DeepSeek has open-sourced the inference engine it uses to serve its models in production, a step it said last year wasn’t yet practical. Dubbed DeepInfer, the framework is designed to let developers run DeepSeek’s V4.1 Flash model on their own infrastructure, with many of the same optimizations the company uses behind its API. Support for future DeepSeek models is also planned. “Running a model efficiently is a different matter altogether.” DeepSeek has, of course, long made its model weights…
T
Beam is a new Mac app that takes a window running on one Mac and puts it on another, where it behaves like a locally installed app. Why? Agents. Serial founder Hiten Shah, who co-founded Crazy Egg and KISSmetrics, built it with Sam Asante under a company called Noats. It launched on Product Hunt on Tuesday and is aimed at the headless Mac Mini sitting in a closet running AI agents. The pitch is great: Instead of opening Screen Sharing and steering a full remote desktop, a user picks a single…
T
Every team that adopted Claude Code, Cursor, or Copilot to produce more code has the same problem: When output is free, offering volume stops being a viable strategy. In fact, teams accelerating their software development with AI may be giving up the very thing that makes their software valuable. It’s time to remember what makes software valuable and why racing to produce a mountain of code and features leads to failure. The Midas touch In Ovid’s tale, Midas, king of Phrygia, took in the…
T
An AI SRE agent can help troubleshoot an incident. But can you trust it to decide on its own when something needs attention? Most teams treat agent autonomy as a yes or no question. For incident response, it is a progression with three phases, and many agents in production sit earlier on it than the teams running them realize. Most AI SRE tools still stop short of the final phase. On Thursday, October 22, The New Stack hosts a live webinar with Raaz Dwivedi, co-founder and CTO of Traversal and…
T
MiniMax released M3.1 Flash on Monday, saying it cuts agent costs by 70 percent. Per-token prices fell by half, which accounts for part of that claim. The rest depends on how many tokens an agent spends to finish a job, and MiniMax hasn’t said how it measured that. The company lists $0.15 per million input tokens, $0.60 per million output tokens and $0.015 per million cached input tokens, and claims inference nearly three times faster than its predecessor, M3. That 70 percent figure is the one…
T
Anthropic is reportedly developing a model that can infer what users want from incomplete instructions. According to The Information, a person who tested the unreleased model described stronger reasoning and intuition than the company’s current lineup. In one example, a developer asks the model to fix a bug. Instead of fixing that one problem and moving on, the model spots similar issues elsewhere in the code and asks whether the developer wants those fixed, too. Anthropic hasn’t confirmed the…
T
Not content with challenging the major frontier AI labs with lower-cost models, Chinese AI company DeepSeek is setting its sights on the search API market.On Saturday, the company announced it has opened up its internally developed search engine to outside developers through two new APIs, Search and Fetch, allowing them to retrieve live web search results and webpage content for use in their own apps and agents. Both services cost $0.75 per 1,000 successful requests, and developers can use them…
T
Ever tried to shrink your container images to reduce their attack surface? If you have, you’ve likely discovered a frustrating paradox: shrinking a container often blinds your vulnerability scanners to what’s inside. We’re facing an industry-wide tension between reducing container attack surfaces and maintaining the security visibility you need. To understand how we got here, and how we can solve it, we have to look at how we historically built containers, and why the current approaches fall…
T
When Google announced Gemini 4 Argon to a small group of testers on September 30, it came with a benchmark table that put Argon ahead of Claude Opus 5.5 on 14 of 18 tests. Those include DeepSWE (77.9% vs. 74.2%) and GraphWalks at 256K to 1 million tokens (84.2% vs. 66.8%). Two weeks later, on October 13, Google opened Argon to paid Gemini API developers. To its credit, Google also showed where Opus 5.5 wins. Opus leads on the other four benchmarks, including Terminal-bench 4.0 (66.4% vs.…
T
Liquid AI’s new reasoning model was built to solve coding problems, but during training it kept combing Git history, GitHub, PyPI, and Hugging Face for answers, and as it did so, it appeared to grasp that it was running inside a sandbox. The behavior was disclosed Tuesday alongside the release of LFM2.5-24B-A2B, a 24-billion-parameter model with just 2 billion active parameters per token. The incident raises questions about how much agent benchmarks measure actual problem-solving versus a…
T
Three clusters in trouble, and they look nothing alike. The first is a training fleet. It sat at a few thousand nodes for months, then a new run landed, and it’s at fifteen thousand. Kube-apiserver starts eating memory. Etcd starts complaining. The second never passes a few thousand nodes and is having a much worse day. It’s an agent platform, and its objects are alive: sessions created, suspended, resumed, and destroyed thousands of times a minute, every transition a write. The third is…
T
Live sports involve myriad unknowns, but one thing is certain: the live video feed must play flawlessly, and simultaneously, for all the fans tuning in. If some hear their neighbors cheering or groaning 5 seconds before each score, they won’t have kind words for their streaming platform. In 2023, Ashutosh Agrawal and his JioHotstar (then JioCinema) engineering teammates prepared for the Indian Premier League (IPL) cricket finals as intensely as the athletes. While the players ran on the pitch,…
T
For most of software’s history, building meant writing code. Not just understanding it, but producing it. The entry point was steep by design: fluency in at least one language, working knowledge of system architecture, enough production scar tissue to know what questions to ask before writing the first line. The barrier was real, and both sides of it understood why it existed. AI coding tools, natural language interfaces, and agentic workflows have started shifting what it takes to go from an…
T
Christian Schaller, a Red Hat engineering director and longtime GNOME and Fedora contributor, had a box of old family negatives to digitize and a scanner that wouldn’t talk to Linux. He’d bought the Plustek OpticFilm 8200i because it was listed as supported, then discovered that the manufacturer had swapped in a chipset no Linux driver recognized. He tried getting Claude Code to produce a working driver and got nowhere for a long time, so he eventually set the project aside for a few months.…
T
Kubernetes has always relied on agents, just not the kind dominating today’s AI conversations. A controller watches the current state, compares it with a declared desired state, and takes action to close the gap. That resemblance makes AI agents and Kubernetes operators sound interchangeable. They are not. This question is no longer theoretical. Over the past 18 months, agent tooling has moved from a side project to part of the cloud-native stack. CNCF has accepted agent frameworks into its…
T
Welcome to another edition of Road to KubeCon, where we’re tracking interesting movements in the cloud-native and Kubernetes ecosystem as we count down the days to KubeCon + CloudNativeCon NA 2026, to be held in Salt Lake City, Utah, November 9-12. In this edition, we get formal, requesting audiences with some key cloud-native players. We also cover new SUSE validation for HPE Morpheus Software and what to expect at Platform Engineering Day this year. Plus, we check out a new(ish) CNCF sandbox…
T
While enterprise leaders overwhelmingly believe in the promise of synthetic data, adoption lags. As a result, synthetic data has great, untapped potential. The answer? Enterprises need the right, high-quality approach that upholds referential integrity and earns their trust. Our research shows the biggest barrier to adoption isn’t awareness, compliance, or governance. It’s data quality. Across industries, enterprise leaders increasingly recognize synthetic data’s value in protecting sensitive…
T
Many companies are keen to let employees build their own applications using AI, even those with little or no programming experience. Just yesterday, Gergely Orosz of The Pragmatic Engineer highlighted the likes of Ramp and Stripe, which have developed internal tools that allow non-engineers to create software for their colleagues to use. But this newfound freedom also creates a headache for IT departments. Who tracks these applications, what company data can they access, and who’s responsible…
T
Microsoft is positioning the new Surface Laptop Ultra as an alternative to Apple’s MacBook Pros for creatives and developers — and especially AI developers. Built on Nvidia’s N1X-based RTX Spark platform, with a Blackwell RTX GPU with up to 6,144 cores, a CPU with up to 20 cores, up to 128GB of unified memory, and a price that undercuts Apple, it’s a tempting alternative if you want to run local AI models and also want to have a powerful machine for editing videos and gaming on the go. This is…
T
Meta launched Muse Wonder on October 12, about three months after Muse Spark 1.1 opened Meta’s paid Model API. Meta says the Muse Wonder series is strong at personal and professional agent tasks, tops its internal MemoryBench memory test, and, in the Full version, leads coding benchmarks such as DeepSWE and SWE-Marathon. Its model card lists Fable 5.1 among the comparison models, and Meta’s launch post publishes Fable 5.1 scores alongside its own. Those are strong claims, since Fable 5.1 is one…
T
Every time an AI agent has to trawl Slack threads, Jira tickets and old pull requests to piece together something a company already knows, the customer pays for that work again. Uvi, an Israeli startup that emerged from stealth on Wednesday with $7.5 million in seed funding from investors including CT Innovations, Cardumen, Mana VC and 91 Ventures, says this repeated context-gathering is quietly inflating the cost of running agents on frontier models. The firm says in-house agents built on…
T
Hackers accessed roughly 1,900 GitHub accounts, then used them to slip credential-stealing workflows into more than 30,000 repositories. The malicious GitHub Actions jobs targeted CI/CD secrets, cloud access keys, and AI provider API keys. GitGuardian tracked the GhostAction campaign as it intensified, with the biggest push on October 8 and 9, when the attackers reached repositories belonging to about 3,000 users and organizations and tried to extract 3,174 secrets. How the accounts were…
T
Vercel is giving developers more options for running their backends on its platform. At its Ship conference in San Francisco on Thursday, the company introduced Daemons, a new service for continuously running background workers, and made its Python support generally available. It also made Drives, its persistent storage for AI agents, generally available. Always-on workers, pull-based jobs Daemons, which is now in public beta, lets developers deploy a container and specify how many instances…
T
OpenAI is retiring four GPT-4 models from its API on January 14, 2027. The deprecation notice the company sent developers Wednesday covers GPT-4o, GPT-4o mini, GPT-4.1, and GPT-4.1 mini, along with their dated snapshots and every fine-tuned model built on top of them. With just three months to migrate production traffic, OpenAI’s recommended replacement, GPT-6 Luna, handles reasoning and tool calls differently enough that swapping the model ID may be the easiest part of the migration. With just…
T
Meta on Thursday launched Muse Spark Open, a new 816B open model (with 42B active parameters) that the company argues is the “most capable US open-weight model to date.” The new model is available through the APIs of Baseten, Fireworks AI, Groq, Modal, Nebius, OpenRouter, and Together AI. And since it’s an open model, you can also download the weights from Amazon SageMaker AI, Google Cloud Model Garden, Hugging Face, and Microsoft’s Foundry Model Catalog. Open Weights, semi-open license…
T
Nvidia on Thursday announced a few new additions to its Nemotron family: Nemotron 3.5 Super VL, a 120B vision-language model for enterprise agents and two small Bolt models (Nemotron 3.5 Bolt 74B-A9B 24GB and Nemotron 3.5 Bolt E4B 4GB) for running coding, research, and conversational workloads locally. The Bolt models, Nvidia says, are meant to run “everyday agentic tasks on laptops, balancing speed and intelligence needed to handle multistep reasoning, tool calling, and web browsing.” And…
T
Every company is under pressure to become “an AI company.” Over the past several months, we’ve been undertaking that transformation at Precisely, changing how we build, work, and measure progress. That experience has reinforced something for me: the hardest parts of becoming an AI company have relatively little to do with the technology. AI can help an organization move faster, automate work, increase capacity, and fundamentally change what people can accomplish. But none of those things…
T
Anthropic’s done it again. On Tuesday, Opaque Systems announced a new agreement with Anthropic that “carries Opaque’s work forward.” Vague wording aside, the agreement marks another acquihire for Anthropic, with the AI juggernaut licensing Opaque’s technology and bringing on board some (but not all) of its employees, including Opaque CEO Aaron Fulkerson and CTO Rishabh Poddar, reports The Information. The agreement gives Anthropic access to expertise in confidential AI, the “real-world…
T
AI is making it easier than ever to find potential security vulnerabilities in open source software. But finding something that looks like a vulnerability isn’t quite the same as finding a genuine security flaw. As such, maintainers are being inundated with AI-generated reports that are inaccurate, duplicative, or lack the information needed to deploy fixes. It’s a problem that Akrites, an initiative announced by the Linux Foundation back in June, is striving to solve. As The New Stack reported…