The Brutalist Report

Venturebeat

• How Sakana trained a 7B model to orchestrate GPT-5, Claude Sonnet 4 and Gemini 2.5 Pro [18d]
• Meet ZAYA1-8B, a super efficient open reasoning model trained on AMD Instinct MI300 GPUs [18d]
• Why AI breaks without context — and how to fix it [19d]
• Anthropic Skill scanners passed every check. The malicious code rode in on a test file. [19d]