Claude Science gathers research tools into a workbench
Anthropic emphasizes inspectable outputs and access to computation.
Not just
about technology.
About people, too.
Independent
AI magazine
Since 2026
Anthropic emphasizes inspectable outputs and access to computation.
Anthropic updates planning and tool use for broadly accessible workflows.
IBM evaluates modernization across a project rather than an isolated function.
Yandex updates the language, search, and multimodal models behind its chat.
Google integrates interface interaction into its main Flash model.
Anthropic starts with Slack channels and work assigned through mentions.
Coordinates, block types, and confidence scores make extraction more inspectable.
Google DeepMind describes layered oversight for increasingly capable internal agents.
Z.AI combines a million-token context with adjustable reasoning effort.
OpenAI introduces expert-developed and checked life-sciences evaluations.
AWS describes an agent workflow spanning data, training, and hardware.
The open release targets agents and user-controlled deployment.
Google expands speech-to-speech translation across its products and developer tools.
The program invites work on employment, productivity, and wider economic effects.
Anthropic adds a Services Track and Partner Hub.
OpenAI updates its specialized life-sciences model.
JetBrains targets code and supporting operations with an open-weight release.