ResearchMajorOct 15
OpenGCUA: Toward a Fully Open Stack for Generalist Computer Use
Upvotes329Authors15
Loading…
Glossary term
A capability that lets an AI operate a computer by viewing the screen and using the mouse and keyboard.
Computer use lets a model operate a computer as a person does: it looks at screenshots, moves the pointer, clicks and types. That lets it use software with no special integration.
It is slower and less reliable than direct integrations, but works with any app.
It extends AI to all the software people already use, but acting on a real screen raises safety and security questions.
ResearchMajorOct 15
Upvotes329Authors15