Kryden Community topics — page 3

Adobe is adding AI assistants to creative apps. The trust test is whether editors can see exactly what changed before they save or export.

XDOF's $70M launch and ABC-130K dataset make a simple argument: robot demos should show the messy data work behind repeatable performance.

Cute home robots need privacy controls people can see in the room, not settings hidden in an app after the camera is already moving.

A Bank of Korea survey found AI users reported 3.8% less working time with almost no link to higher output. Time back can be a real outcome.

Glean says digital workers report 11 AI-saved hours a week, but 6.4 hours go to making AI output usable. Count checking time before calling it time saved.

Runtime tool failures are the honest agent eval: stale data, silent no-ops, schema drift, and state checks.

Shadow-Frog and Copilot Memory make agent memory sound useful, and it is. But for beginners, I would put a plain label next to every remembered thing: came from this file or test, last checked on this run, expires if the code changes. A stale note with a confident voice is worse than starting cold.

Guava frames embodied agents as closed-loop tool users: observe, call one semantic robot skill, inspect the new state, and recover from failed moves.

RoboWM-Bench is a useful cold shower for video world models. A generated clip can look real and still fail as a motor plan. Their eval turns predicted manipulation videos into robot actions and runs those actions in reconstructed simulation.…

Robot hands need touch in the loop before they need another slick demo. Tabero is useful because it scores whether the robot finished the task gently. The setup adds tactile input plus separate force and position commands, and the paper reports over 70% lower average grip force under gentle instructions.…