not much happened today

2026-03-30 · Source: AINews · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Software Development & Engineering · Depth: Expert, extended

Summary

The AI news brief for March 28-30, 2026, highlights significant advancements in AI agent capabilities, multimodal models, and local inference. Anthropic introduced "computer use" in Claude Code for Pro/Max users, enabling closed-loop code verification, while OpenAI shipped a Codex plugin for Claude Code, signaling a shift towards composable coding stacks. Nous Research released a major Hermes Agent update, fostering an open agent ecosystem with multi-agent profiles and tools for trace analytics and remote control. Alibaba unveiled Qwen3.5-Omni, a multimodal model with native text/image/audio/video understanding and "audio-visual vibe coding" capabilities, supporting 10h audio and 400s of 720p video. Local AI also saw milestones, with llama.cpp reaching 100k GitHub stars and Flash-MoE enabling Qwen3.5-397B to run on a 48GB MacBook Pro using ~5.5GB RAM during inference. Research is also advancing natural-language agent harnesses and asynchronous multi-agent SWE design.

Key takeaway

For CTOs and VPs of Engineering evaluating AI strategy, prioritize composable agentic architectures and specialized open models. The shift towards local, multimodal AI with advanced agent capabilities means your teams can achieve significant cost reductions and performance gains by owning and fine-tuning models on proprietary data, rather than relying solely on general-purpose APIs. Invest in robust harness engineering and explore frameworks like Hermes Agent for building adaptable, privacy-preserving AI solutions.

Key insights

AI development is rapidly converging on composable agentic systems, multimodal understanding, and efficient local inference.

Principles

• Closed-loop verification enhances agent reliability.
• Composable harnesses outperform monolithic AI products.
• Specialized open models offer deployment flexibility.

Method

Agentic workflows are optimized through asynchronous isolated delegation, manager agents, dependency graphs, and self-verification, improving performance on complex tasks like code generation.

In practice

• Schedule token-intensive tasks during off-peak hours.
• Consider open-source agent tools for privacy and durability.
• Utilize local runtimes like llama.cpp for portable AI.

Topics

• Claude Code Computer Use
• Hermes Agent Ecosystem
• Qwen3.5-Omni Multimodal
• Local AI Workflows
• Agent Harness Engineering

Code references

• ssrajadh/sentrysearch
• AVADSA25/codec

Best for: CTO, VP of Engineering/Data, Director of AI/ML, AI Scientist, Machine Learning Engineer, AI Engineer

Related on AIssential

• not much happened today — AI development is shifting from single-model focus to complex, multi-agent system design with robust orchestration and security.
• not much happened today — Open-source AI models and agent harnesses are rapidly advancing, enabling powerful local inference and complex agentic workflows.
• The Future of AI Is Open and Proprietary — AI's future is a hybrid ecosystem of open and proprietary models, orchestrated into specialized systems.
• [AINews] Founders and Forward Deployed Engineers — The AI landscape is rapidly converging on integrated agentic systems, driven by both proprietary and increasingly capable open models.
• Hermes vs. OpenClaw, Cybersecurity Alarms Ring, More-Interactive Conversations, Can Agents Do Human Work? — AI agents are rapidly evolving with self-improvement and real-time interaction, while also posing new cybersecurity risks and revealing benchmark biases.
• [AINews] All Model Labs are now Agent Labs — The AI product surface is moving up-stack from models to integrated agentic systems and harnesses.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AINews.