AI News Feed

Open-source models, agents, and infrastructure — curated daily

FrameworksCrewAI18 May 2026

CrewAI 1.14.5a7: 1.14.5a7

What's Changed Documentation - Update changelog and version for v1.14.5a6 Breaking Changes - Deprecate function_calling_llm field Contributors @greysonlalonde, @heitorado

GitHub Releases
ModelsOpen Source18 May 2026

Reverse engineering Android malware from popular Chinese projectors

Hacker News discussion: "Reverse engineering Android malware from popular Chinese projectors" — 71 points. Trending in the AI community.

Hacker News
FrameworksLangChain18 May 2026

Tuning Deep Agents to Work Well with Different Models

Deep Agents was previously designed in a generic way to work well across model families. Today we’re adding model-specific profiles to adjust prompts, tools, and middleware. We ship profiles for OpenAI, Anthropic, and Google models out of the box, which we see leads to a 10–20 point jump on a subset of tau2-bench over the default harness.

LangChain Blog
ModelsOpenAI15 May 2026

How sales teams use Codex

See how sales teams can use Codex to create pipeline briefs, meeting prep packets, forecast reviews, account plans, and stalled-deal diagnoses from real work inputs.

OpenAI Blog
ModelsMistral17 May 2026

Mistral's CEO: Europe has 2 years to stop becoming America's AI 'vassal state'

Hacker News discussion: "Mistral's CEO: Europe has 2 years to stop becoming America's AI 'vassal state'" — 99 points. Trending in the AI community.

Hacker News
ModelsOpen Source17 May 2026

Show HN: Semble – Code search for agents that uses 98% fewer tokens than grep

Hey HN! We (Stephan and Thomas) recently open-sourced Semble. We kept running into the same problem while using Claude Code on large codebases: when the agent can't find something directly, it falls back to grep, reading full files or launching subagents. This uses a lot of tokens, and often still misses the relevant code. There are existing tools for this, but they were either too slow to index on demand, needed API keys, or had poor retrieval quality.Semble is our solution for this. It combines static Model2Vec embeddings (using our latest static model: potion-code-16M) with BM25, fused via RRF and reranked with code-aware signals. Everything runs on CPU since there's no transformers involved. On our benchmark of ~1250 query/document pairs across 63 repos and 19 languages, it uses 98% fewer tokens than grep+read and reaches 99% of the retrieval quality of a 137M-parameter code-trained transformer, while being ~200x faster.Main features:-

Hacker News
ModelsOpen Source16 May 2026

PART Telescopes – Bringing radio astronomy within reach of rural schools

Hacker News discussion: "PART Telescopes – Bringing radio astronomy within reach of rural schools" — 125 points. Trending in the AI community.

Hacker News
InfrastructureWeights & Biases4 May 2026

Weights & Biases at #NYTech Week

Tech Week is a16z’s annual conference — not one venue but hundreds of events across the country, drawing over 100,000 engineers, founders, and investors. This year we’re hosting five events across Boston and New York. Whether you’re a researcher, a builder, or somewhere in between, there’s something here for you. And if you use Weave […] The post Weights & Biases at #NYTech Week appeared first on Weights & Biases.

Weights & Biases Blog
InfrastructureWeights & Biases4 May 2026

Weights & Biases at #BOSTech Week

Tech Week is a16z’s annual conference — not one venue but hundreds of events across the country, drawing over 100,000 engineers, founders, and investors. This year we’re hosting five events across Boston and New York. Whether you’re a researcher, a builder, or somewhere in between, there’s something here for you. And if you use Weave […] The post Weights & Biases at #BOSTech Week appeared first on Weights & Biases.

Weights & Biases Blog
InfrastructureWeights & Biases4 May 2026

Agentic AI self-correction: How to build systems that fix their own mistakes

On this page Why AI must learn to self-correct The principles of agentic reasoning Architecting a self-correcting system Autonomous reflection loops Overcoming common failures MCP and advanced governance Summing things up The dream of AI has always been autonomy. But true autonomy isn’t just about finishing a task; it’s about recognizing when you’ve taken a […] The post Agentic AI self-correction: How to build systems that fix their own mistakes appeared first on Weights & Biases.

Weights & Biases Blog