For example, if you tell it to the robot of ç Bring something to drink ”, it should probably estimate that it is related to tea or water, not gasoline or vegetable oil. This is open for a person without words, but not for a robot.
The new test contains an example of 2,000 household demand, which is often not fully understood. It helps to check whether the robot can recognize such situations and indicate that it only needs.
Ambik writers propose to divide uncertainty into three species:
-
Preferences User (tea or fruit juice),
-
Widespread Information (Do not offer vinegar as a drink),
-
Question Safety (Do not give a hot child).
Most of the modern models still cope with such tasks badly – on average, they react correctly in only 20% of cases. This can be fixed with a new data set.
Ambik can be used to teach both language and robot systems. In addition, it is suitable for tests where the actions consist of several steps – for example, pour water, put a cup of cup and serve.
The data set is open to everyone and will help create more “estimated” and suitable Helmetre robots.
Source: Ferra
I am a professional journalist and content creator with extensive experience writing for news websites. I currently work as an author at Gadget Onus, where I specialize in covering hot news topics. My written pieces have been published on some of the biggest media outlets around the world, including The Guardian and BBC News.
