The world’s first artificial intelligence (AI) system, developed by a Swiss startup called DeepMind, is now capable of doing some basic tasks that humans are good at, but are unable to do.
DeepMind says its AI system, dubbed AIsense, can be used in the fields of medicine, education and social care to help people better understand themselves, and how they might respond to other people.
But for now, the system’s limitations mean it can only work with human-like responses.
For example, DeepMind says that AIs will be limited in their ability to understand the meaning of the words they hear or read.
For example, a person may only be able to read a letter from a colleague or friend that contains the word “good” in it.
To understand the situation, a system will need to be able take into account how people have been interacting with one another in the past, or how those interactions might have influenced the way they react to a particular task.AIsense is designed to be particularly good at recognising and learning from mistakes.
If a system gets stuck, the AI will have to repeat the task or go back to a human.
But DeepMind’s AI system can also be used to perform tasks that have never been done before.
For instance, AIs may be able do the basic task of finding a car that has been left outside overnight, and a computer system might be able identify the location of a particular car parked in the street, or even a specific type of car, such as a hybrid.
The system can even understand the difference between different types of objects, such a tree or a house.
The company says that its AI will be able also to predict whether a person has gone to a park or to a museum, or to the location and size of a certain building.
To learn more about how its AI works, check out DeepMinds website.