5 AI stories explainx.ai reported on May 7, 2026, ranked by reader interest and grouped by topic. Each links to the full write-up with sources.
Built by Yu Shi and team, Kronos treats financial candlesticks as a language—using a hierarchical tokenizer to quantize OHLCV data into discrete tokens, then training autoregressive Transformers for forecasting. Accepted at AAAI 2026 with live demo, finetuning pipeline, and batch prediction support.
Reference agents, skills, and data connectors for FSI workflows—available as Claude Cowork plugins or via Claude Managed Agents API. Includes 10 named agents, vertical skill bundles, partner integrations with LSEG and S&P Global, and 11 MCP data connectors.
At a San Francisco developer event, Anthropic rolled out Dreaming (research preview) to let Claude Managed Agents learn from session transcripts, plus public betas for multiagent orchestration (up to 20 parallel agents), outcomes loops (rubric-based self-checks), and webhooks for external notifications.
On May 6, 2026, Anthropic and SpaceX announced a deal for exclusive use of the Colossus 1 supercomputer, providing 300+ megawatts and 220,000 NVIDIA GPUs. The partnership immediately doubles Claude Code rate limits on Pro, Max, Team, and Enterprise plans, eliminates peak-hour restrictions, and raises API limits.
ByteDance's DeerFlow evolved from a deep research framework to a full super agent harness. Version 2.0 ships with skills, sub-agents, sandboxed execution, persistent memory, Claude Code integration, and support for Telegram, Slack, Feishu, WeChat, WeCom, and DingTalk channels.
Both tools launched in early 2026 as powerful coding agents—Claude Code via terminal with deep reasoning and 1M context, Codex via desktop app for parallel agents and quick tasks. After Anthropic's rate limit boost, developers share strong preferences: Claude excels in code quality, Codex leads in speed.
llms.txt brings structured, curated markdown documentation to your website root so LLMs can quickly access the context they need. Here is why it matters for documentation sites, software projects, and any content you want AI to understand correctly.
Frontier LLM pretraining runs synchronously across tens or hundreds of thousands of GPUs; tail latency and link failures amplify into full-cluster stalls. OpenAI’s MRC reframes datacenter GPU fabrics—multi-plane topologies, packet spraying, trimming, and SRv6 source routing—with an OCP specification and research paper.
The RAG industry is being challenged by a simpler idea: let agents search with grep, glob, and LSP servers instead of pre-indexing everything into vector databases. Here is why agentic RAG is winning for code and structured data.
Instead of scaling reasoning only with bigger parameter counts, HRM and TRM show a second axis: recursive computation at inference time. This guide breaks down what these models do, why they worked on ARC-style tasks, and how to think about recursion versus chain-of-thought and tool use.
BGRemover is an AI-powered tool that allows you to remove video backgrounds effortlessly. With just one click, you can make any video background transparent, making it perfect for content creators and marketers.
PageIndex provides precise, verifiable answers and insights from complex documents. It features a chat interface that allows users to understand documents with explainable answers grounded in the source.
Discovers and invokes agent skills for effective task management in engineering workflows.
Drives development with tests. Use when implementing any logic, fixing any bug, or changing any behavior.
Creates specs before coding to clarify requirements and ensure successful project execution.
Grounds every implementation decision in official documentation for authoritative, source-cited code.
Get each day's AI news in your feed reader: daily RSS · every post