Anthropic researcher Nicholas Carlini used sixteen instances of the Claude Opus 4.6 model, organized as “agent teams,” to develop a Rust‑based C compiler from scratch. Over two weeks and nearly 2,000 Claude Code sessions, the agents produced a 100,000‑line compiler capable of building a bootable Linux 6.9 kernel for x86, ARM and RISC‑V. The open‑source project, released on GitHub, compiles major software such as PostgreSQL, SQLite, Redis, FFmpeg and QEMU, passes 99 percent of the GCC torture test suite, and even runs Doom. The experiment highlights the potential of semi‑autonomous AI coding on well‑defined tasks. Leer más →

A test of four AI coding agents tasked with rebuilding the classic Minesweeper game revealed a blend of successes and shortcomings. While the agents successfully implemented core gameplay mechanics such as chording and flagging, they varied in UI polish, sound options, and development speed. OpenAI's Codex, for instance, took roughly twice as long to produce a functional version compared to Claude Code. The evaluation highlights both the promise of AI-driven development and the current limits that developers must still address. Leer más →

Amazon Web Services introduced three new AI agents—Kiro, Security Agent, and DevOps Agent—designed to operate autonomously for extended periods. Kiro, a coding specialist, learns a team’s workflow, follows spec‑driven development, and can persist context across sessions. Security Agent independently identifies and proposes fixes for code vulnerabilities, while DevOps Agent handles testing and performance validation. AWS CEO Matt Garman highlighted the agents’ ability to take on complex backlog tasks with minimal human oversight, marking a significant step toward AI‑driven software development. Leer más →

After years of stagnant revenue and a 50% staff cut, Replit revived its fortunes by launching the Replit Agent, an AI‑powered coding assistant. The shift from serving professional developers to targeting non‑technical knowledge workers unlocked rapid revenue growth, leading to a $250 million funding round, a $3 billion valuation, and annualized revenue in the $150 million range. Enterprise customers now drive high‑margin earnings, while the company bolsters safety after a high‑profile database mishap. Replit faces competition from AI labs but bets on its unique focus and infrastructure to stay ahead. Leer más →

Emergent, a startup that lets non‑technical users create mobile apps through AI‑driven prompts, announced a $23 million Series A funding round led by Lightspeed. The round also included participation from Y Combinator, Together and several high‑profile angels. Emergent’s platform abstracts the entire software development lifecycle, handling everything from code generation to deployment and bug fixing. Since its launch, more than 1 million people have built over 1.5 million apps on the service. The company aims to broaden its capabilities, including a forthcoming native mobile client and new tools for ideation and discovery. Leer más →