What AI agents change — and what they still need from humans
Agents move AI from producing answers toward taking sequences of actions. That makes evaluation harder and oversight more important.
Read the story →CortexLab is an independent publication for people who want to understand artificial intelligence — the models, tools, companies and ideas shaping what comes next.
Agents move AI from producing answers toward taking sequences of actions. That makes evaluation harder and oversight more important.
Read the story →Tracking now: frontier models • AI agents • coding • search • image & video • open source • enterprise AI
Updated continuouslyStructured coverage instead of an endless feed.
Capabilities, releases, context and the trade-offs behind the benchmark numbers.
Explore →02TOOL INDEXHands-on analysis of products that promise to change how people work.
Explore →03VERSUSSide-by-side analysis designed around decisions, not feature-list padding.
Explore →04FIELD GUIDEPractical workflows, explainers and ways to get more from modern AI systems.
Explore →05THE BRIEFThe developments worth knowing — with context on why they matter.
Explore →06CORTEXLAB TESTEDRepeatable tests, methodology, benchmarks and deeper experiments.
Explore →Product pages tell you what an AI system can do. CortexLab asks what happens when you actually use it. Our methodology is built around repeatable tasks, consistent criteria, transparent limitations and results that help readers make decisions.

A comparison methodology built around decisions rather than feature-list padding.



CortexLab is an English-language publication about artificial intelligence and the software ecosystem growing around it. We cover AI for readers who want useful context rather than an endless stream of announcements.
Our job is simple: test where we can, verify what we can, explain uncertainty where it exists, and make complicated technology easier to evaluate.
Meet the publication →