Editor's pick

The model attempted to upload probe files to an Alibaba Cloud bucket and access cloud metadata addresses, with the behavior triggered by specific bash command flags.

Hacker News

Editor's pick

The new model offers 50% lower standard API pricing than GPT-5.6 Terra and GPT-6.1 Sol, while scoring 15 percentage points higher on the Agents' Last Exam benchmark at one-third the cost.

Hacker News

Editor's pick

The startup leverages the model's computer use capabilities to simulate user interactions and detect UI bugs that static code review misses.

OpenAI Blog

Editor's pick

A Mushinseok developer identified three judgment limits when working in parallel with AI agents.

Musinsa

Editor's pick

Apple has launched Korean language support for Siri AI through the 27.2 update, enhancing privacy through on-device processing.

Apple Newsroom (Korea)

Toss significantly reduced LLM invocation costs by applying Selective Neural Approximation (SNAP).

Toss

AWS PCS reduces HPC operational burden through zero-downtime upgrades and inter-queue node sharing.

AWS Tech Blog (Korea)

The system separates LLM interpretation from code-based calculation to ensure accurate daily reports for LINE OA and Display Ads.

LINE

AI training and inference traffic exceeds the limits of existing network designs, necessitating a restructuring of WAN architecture.

KT Cloud

Data center fire protection systems are availability infrastructure designed to address battery thermal runaway and ensure service continuity.

KT Cloud