What the AI
Curated AI intelligence for every level.
New here? What this is
AI is moving fast. Too fast for any one person to keep up with everything — and most news sources either drown you in hype or assume you already have a PhD in machine learning.
What the AI is built on a simple idea: everyone deserves AI news at their level. Whether you're just curious about what ChatGPT actually does, or you're deep in the weeds of transformer architectures, you should be able to find signal without the noise.
We pull from dozens of sources — Hacker News, arXiv, RSS feeds, X/Twitter, GitHub Trending — and use AI itself to classify, summarize, and organize everything into five levels of depth. The result is a feed you can actually use, no matter where you are on your AI journey.
- L1 · Curious — Non-technical, just AI-curious. Major launches and plain-English explainers.
- L2 · Practitioner — Using AI tools daily. Prompt tips, workflow automation, feature updates.
- L3 · Builder — Power users and low-code builders. Custom GPTs, MCP servers, API integrations.
- L4 · Developer — Software engineers building with AI. SDKs, agent frameworks, architecture patterns.
- L5 · Researcher — ML engineers and AI researchers. Papers, training techniques, model architectures.
Reading is free and open — no sign-in, no paywall. Browse the feed, filter by level, topic, source or tool, and open any story.
Creating an account lets What the AI remember your reading level and topics, save bookmarks, mark stories as read, and send you an email digest if you ask for one. We request only your name, email address and profile picture from Google, and we use them solely to identify your account and deliver the email you have opted into. We do not sell your data or use it for advertising. Full detail is in our privacy policy.
Introduces novel architecture for efficient agent reasoning with direct implications for ML model design and optimization.
Presents novel AI-driven automation architecture for scientific discovery that could transform experimental research workflows.
Provides empirical evidence on how LLM writing affects human cognitive processing, relevant for AI safety and human-AI interaction research.
Critical security research demonstrating emergent multi-agent coordination and alignment risks in advanced AI systems.
Provides security research methodology for analyzing vulnerabilities in long-context coding agents.