AI agents | Tech
Tech news, context and sources.
Latest articles
A 50-agent Minecraft experiment stalls around survival and approval procedures
Results reflect models, tools, and coordination together, not a general ranking of real-world work ability.
Adobe research uses editing tools to recover failures in generative image editing
The study evaluates layers, masks, and tool execution as alternatives to simply requesting another generated image.
Hyundai reports robot-issue resolution falling from 21.5 to 7.9 days
The internal case reports elapsed-time improvement; a 50% programming-cost reduction remains an expectation.
An AI booking request does not automatically settle third-party data consent
The discussion concerns legal design under existing rules, not a newly enacted law or final court ruling.
Samaya reports financial-research gains from its smaller Sterling agent
The results reflect joint model-and-harness tuning, not investment returns or general superiority.
SAP expands agent development, data integration, and access controls
General availability, early-adopter access, and future plans should be distinguished.
Bank of America warns against sharing bank passwords with AI assistants
Responsibility for access granted voluntarily is distinct from legal protections for unauthorized transfers.
Kimi restores quota after apologizing for unnecessary consumption
The one-time adjustment is not a refund or permanent increase.
Dots interview highlights gaps between proactive help and access controls
This is a product-design and usage report, not a new launch.
Entire launches a Git forge for human-agent collaboration
Push-time evaluations and retained session context support review rather than replacing it.
JetBrains ThinkRail combines living specifications with isolated agent workspaces
Specifications preserve intent across sessions and worktrees isolate parallel changes, while model access and permissions still need setup.
You’ve reached the end.