All posts tagged: Coding

Apple Pulls Vibe Coding App ‘Anything’ From App Store, Escalating Enforcement

Apple Pulls Vibe Coding App ‘Anything’ From App Store, Escalating Enforcement

Published by skeptic

Apple has removed a “vibe coding” app from its App Store, reports The Information. AI app building app “Anything” was pulled from the ‌App Store‌, and Anything co-founder Dhruv Amin was told that his app violated Guideline 2.5.2. “Vibe coding” is a term used for code generated using AI based on natural language with no coding experience necessary. Anything and other apps like it let users create apps, websites, and tools with text-based prompts. Apple started removing vibe coding apps from the ‌App Store‌ earlier in March, and the company said that certain features in the apps that were pulled violate code execution rules. In a statement to MacRumors, Apple said that there are no specific rules against vibe coding, but the apps have to adhere to longstanding guidelines. Apple specifically mentioned Guideline 2.5.2, which is the rule Anything apparently violated. Apps should be self-contained in their bundles, and may not read or write data outside the designated container area, nor may they download, install, or execute code which introduces or changes features or functionality …

Nvidia’s Nemotron-Cascade 2 wins math and coding gold medals with 3B active parameters — and its post-training recipe is now open-source

Nvidia’s Nemotron-Cascade 2 wins math and coding gold medals with 3B active parameters — and its post-training recipe is now open-source

Published by skeptic

The prevailing assumption in AI development has been straightforward: larger models trained on more data produce better results. Nvidia’s latest release directly challenges that size assumption — and the training recipe behind it may matter more to enterprise AI teams than the model itself. The open-weight model’s Cascade RL post-training pipeline, detailed in Nvidia’s technical report, offers a reproducible blueprint for enterprise teams building domain-specific reasoning systems without training from scratch. Nemotron-Cascade 2 is an open-weight 30B Mixture-of-Experts (MoE) model that activates only 3B parameters at inference time. Despite this compact footprint, it achieved gold medal-level performance on three of the world’s most demanding competitions: the 2025 International Mathematical Olympiad (IMO), the International Olympiad in Informatics (IOI), and the ICPC World Finals. It is the second open model to reach this tier, after DeepSeek-V3.2-Speciale — a model with 20 times more parameters. Why post-training is becoming the real competitive advantage Pre-training a large language model from scratch is enormously expensive — on the order of tens to possibly hundreds of millions of dollars for frontier …

Cursor admits its new coding model was built on top of Moonshot AI’s Kimi

Cursor admits its new coding model was built on top of Moonshot AI’s Kimi

Published by skeptic

AI coding company Cursor launched a new model this week called Composer 2, which it promoted as offering “frontier-level coding intelligence.”  However, an X user posting under the name Fynn soon claimed that Composer 2 was “just Kimi 2.5” with additional reinforcement learning — Kimi 2.5 being an open source model recently released by Moonshot AI, a Chinese company backed by Alibaba and HongShan (formerly Sequoia China).  As evidence, Fynn pointed to code that seemed to identify Kimi as the model. “[A]t least rename the model ID,” they scoffed. It was a surprising revelation, since Cursor is a well-funded U.S. startup that raised a $2.3 billion round last fall at a $29.3 billion valuation, and is reportedly exceeding $2 billion in annualized revenue. Also, the company didn’t mention anything about Moonshot AI or Kimi in its announcement. However, Cursor’s vice president of developer education Lee Robinson soon acknowledged, “Yep, Composer 2 started from an open-source base!” But he said, “Only ~1/4 of the compute spent on the final model came from the base, the rest …

Google AI Studio 2.0 Adds Antigravity Full-Stack Coding

Google AI Studio 2.0 Adds Antigravity Full-Stack Coding

Published by skeptic

Google AI Studio 2.0 offers a structured approach to full-stack development, combining features like context-aware coding and backend automation to support the creation of production-ready applications. A key component is the Antigravity Coding Agent, which minimizes repetitive tasks by intelligently adapting to the coding context. For example, when building user-facing features such as leaderboards, the agent automates much of the setup, allowing developers to allocate more time to problem-solving. These capabilities are outlined by Universe of AI, providing a clear look at how the platform supports developers at various skill levels. Dive into how Firebase integration simplifies backend processes, including database configuration and user authentication, so you can focus on core functionality. Discover the advantages of real-time multiplayer support, which handles synchronization for interactive applications and learn how session persistence ensures consistent development across devices. This practical overview equips you with actionable insights to effectively use Google AI Studio 2.0 in diverse development scenarios. Intelligent & Context-Aware Development TL;DR Key Takeaways : Google AI Studio 2.0 introduces advanced features like context-aware coding, real-time collaboration and …

Mistral’s Small 4 consolidates reasoning, vision and coding into one model — at a fraction of the inference cost

Mistral’s Small 4 consolidates reasoning, vision and coding into one model — at a fraction of the inference cost

Published by skeptic

Enterprises that have been juggling separate models for reasoning, multimodal tasks, and agentic coding may be able to simplify their stack: Mistral’s new Small 4 brings all three into a single open-source model, with adjustable reasoning levels under the hood. Small 4 enters a crowded field of small models — including Qwen and Claude Haiku — that are competing on inference cost and benchmark performance. Mistral’s pitch: shorter outputs that translate to lower latency and cheaper tokens. Mistral Small 4 updates Mistral Small 3.2, which came out in June 2025, and is available under an Apache 2.0 license. “With Small 4, users no longer need to choose between a fast instruct model, a powerful reasoning engine, or a multimodal assistant: one model now delivers all three, with configurable reasoning effort and best-in-class efficiency,” Mistral said in a blog post. The company said that despite its smaller size — Mistral Small 4 has 119 billion total parameters with only 6 billion active parameters per token — the model combines the capabilities of all Mistral’s models. It …

Cursor’s new coding model Composer 2 is here: It beats Claude Opus 4.6 but still trails GPT-5.4

Cursor’s new coding model Composer 2 is here: It beats Claude Opus 4.6 but still trails GPT-5.4

Published by skeptic

Cursor, a San Francisco AI coding platform from startup Anysphere valued at $29.3 billion, has launched Composer 2, a new in-house coding model now available inside its agentic AI coding environment, and it offers drastically improved benchmarks from its prior in-house model. It’s also launching and making Composer 2 Fast, a higher-priced but faster variant, the default experience for users. Here’s the cost breakdown: That’s a big drop from Cursor’s predecessor in-house model, Composer 1.5, from February, which cost $3.50 per million input tokens and $17.50 per million output tokens; Composer 2 is about 86% cheaper on both counts. Composer 2 Fast is also roughly 57% cheaper than Composer 1.5. There’s also discounts for “cache-read pricing,” that is, sending some of the same tokens in a prompt to the model again, of $0.20 per million tokens for Composer 2 and $0.35 per million for Composer 2 Fast, versus $0.35 per million for Composer 1.5. It also matters that this appears to be a Cursor-native release, not a broadly distributed standalone model. In the company’s announcement …

Apple Pulls Vibe Coding App ‘Anything’ From App Store, Escalating Enforcement

Apple Quietly Blocks Updates for Popular ‘Vibe Coding’ Apps

Published by skeptic

Apple has quietly blocked AI “vibe coding” apps, such as Replit and Vibecode, from releasing App Store updates unless they make changes, The Information reports. “Vibe coding” tools allow users with little to no programming experience to build apps or websites using natural language prompts. Their accessibility has driven rapid adoption among both developers and non-technical users. Apple told The Information that certain vibe coding features breach long-standing ‌App Store‌ rules prohibiting apps from executing code that alters their own functionality or that of other apps. Some of these apps also support building software for Apple devices, which may have contributed to a recent surge in new ‌App Store‌ submissions and, in some cases, slower approval times, according to developers. An Apple spokesperson said the policy is not targeted specifically at vibe coding apps. However, some people familiar with the matter said Apple was close to approving updates for Replit and Vibecode after the developers agreed to modify how their apps preview generated content or remove certain capabilities altogether, such as creating apps for Apple …

COBOL Is the Asbestos of Programming Languages

COBOL Is the Asbestos of Programming Languages

Published by skeptic

Early in the Covid-19 pandemic, the governor of New Jersey made an unusual admission: He’d run out of COBOL developers. The state’s unemployment insurance systems were written in the 60-year-old programming language and needed to be updated to handle the hundreds of thousands of claims. Trouble was, few of the state’s employees knew how to do that. And the crisis went beyond New Jersey, just one of many states that depended on these unwieldy systems. By one rough calculation, COBOL’s inefficiencies cost the US GDP $105 billion in 2020. You might think New Jersey would have replaced its system after this—and that Covid was COBOL’s last gasp. Not quite. The state’s new unemployment system came with a number of quality-of-life improvements, but on the backend, it was still made possible by a mainframe running the ancient language. COBOL, short for Common Business-Oriented Language, is the most widely adopted computer language in history. Of the 300 billion lines of code that had been written by the year 2000, 80 percent of them were in COBOL. It’s …

Y Combinator-backed Random Labs launches Slate V1, claiming the first ‘swarm-native’ coding agent

Y Combinator-backed Random Labs launches Slate V1, claiming the first ‘swarm-native’ coding agent

Published by skeptic

The software engineering world is currently wrestling with a fundamental paradox of the AI era: as models become more capable, the “systems problem” of managing them has become the primary bottleneck to real-world productivity. While a developer might have access to the raw intelligence of a frontier model, that intelligence often degrades the moment a task requires a long horizon or a deep context window. But help appears to be on the way: San Francisco-based, Y Combinator-backed startup Random Labs has officially launched Slate V1, described as the industry’s first “swarm native” autonomous coding agent designed to execute massively parallel, complex engineering tasks. Emerging from an open beta, the tool utilizes a “dynamic pruning algorithm” to maintain context in large codebases while scaling output to enterprise complexity. Co-founded by Kiran and Mihir Chintawar in 2024, the company aims to bridge the global engineering shortage by positioning Slate as a collaborative tool for the “next 20 million engineers” rather than a replacement for human developers. With the release of Slate V1, the team at Random Labs …

I Tried Vibe Coding the Same Project Using Different Gemini Models. The Results Were Dramatic

I Tried Vibe Coding the Same Project Using Different Gemini Models. The Results Were Dramatic

Published by skeptic

Vibe coding is a lot of fun when you know the general gist of the process. It’s as easy as talking to an AI chatbot and having it code up an app for you, but it requires time and patience to iron out issues. I’ve created several vibe coding projects, but there are always new ways to test how good these outputs can be, especially when you consider the model you’re using. With so many AI models to tinker around with, they can produce significantly different results, especially if you don’t have a solid plan in mind. I wanted to see how the lighter models compare to the “thinking” models, as Google and OpenAI refer to them. These lighter models vary in name: Google’s Gemini interface calls it Fast (although the model is actually called, for example, Gemini 2.5 Flash), while OpenAI calls it Instant. I decided to perform an experiment using two models to create the same project. First, I created a project from beginning to end using Google’s Gemini 3 Pro, and I …