
Tech • AI • Robotics • Game
Anthropic is pushing Claude beyond chat with Claude Meetings, a feature that joins calls directly from calendar invites. It appears as an active participant, takes notes live, answers written questions, and surfaces information from connected workplace tools when permissions allow. The product aim is to turn meetings into working sessions rather than transcription events. That positions Claude more directly against meeting assistants and broader office suites.
Whiteboards are framed as shared canvases built during the call, with teams and Claude editing the same workspace in real time. Unstructured discussion can be reorganized into timelines, ownership views, and date-based plans spanning months. Users can also ask Claude to analyze the board itself, including identifying crowded areas or standout projects. The feature extends the model from note-taking into live project structuring.
Anthropic also outlined Sheets and upgraded Data tools as part of a broader workplace productivity push. The package suggests Claude is being positioned as an interface for documents, tabular work, and internal knowledge retrieval inside the same product flow. Combined with in-app connectors, the new features point to a tighter link between conversations and enterprise systems. The strategy is to keep work inside Claude from discussion through execution.
A central pitch is that teams should leave calls with usable artifacts already produced. Instead of waiting for a post-meeting recap, participants can review notes and boards as they are created during the discussion. After the meeting, Claude can convert a whiteboard into a shareable document for wider circulation. That reduces the lag between planning, alignment, and follow-up.
A grower in Slovenia used Claude as a practical troubleshooting tool during sweet potato cultivation after repeated setbacks. The farm treated the crop as a commercial experiment despite skepticism from buyers and supermarkets. With limited formal AI expertise, the operator used the model to interpret field observations and images during the season. The case shows Claude being applied well outside office workflows.
The sweet potato project suffered through two years of costly mistakes before results improved. Early attempts were described as near-total failures, underscoring how poorly adapted methods can sink a nontraditional crop. Those losses led to closer field monitoring and a more systematic trial-and-error approach. The eventual turnaround was built on iterative correction rather than a single breakthrough.
One acute problem came when about one-third of mother tubers failed to sprout, threatening planting plans. The grower reportedly fed photos and observations into Claude to help narrow likely causes and possible responses. That use case highlights image-informed, context-specific diagnosis as a practical value proposition for general models. In agriculture, the appeal is speed and second-opinion support rather than full automation.
The farm’s operator, a biologist raised in Vienna, brought scientific curiosity to a multigenerational family farm in Slovenia. Sweet potatoes were initially dismissed as a bad commercial bet by retail buyers, yet the farm persisted. The result was a record harvest after repeated failures, close observation, and AI-assisted troubleshooting. The story illustrates how Claude is now surfacing in both enterprise software and field-level operational decisions.