New research has shown that in order to create human-like artificial intelligence, the technology must be able to make emotional assessments. A Russian professor has proposed a way to achieve this.
At a recent U.S. conference on biologically inspired cognitive architectures, Professor Alexei Samsonovich from the National Research Nuclear University in Russia proposed an artificial intelligence test based on a simple computer game. The game involves interaction between a human and the computer program, whereby both manipulate virtual people on a computer display. The game utilises actions with emotional content: A player may strike or push aside his co-player, or say “hello” and step aside to make way for him, or help him move off a stone or get out of a trap. This forces players to engage in different types of social relationships including mutual trust, subordination and leadership.
Normally, the artificial intelligence test is considered passed if a person is unable to guess whether his co-player is a machine or a human. In Samsonovich’s case, the idea is that the machine should have an emotional advantage over the average human player, which will manifest itself in players’ wish to rescue the machine first. Furthermore, multiple behavioural parameters of the player and the machine will be calculated during the game, characterising the inner worlds of both. For the machine, these parameters should in the future, become statistically identical to human behaviour.
Approximately a year and a half from now, the researcher hopes to create a machine capable of drawing up plans, setting goals and establishing lasting social relationships with humans. Named Virtual Actor, it will hopefully possess both emotional and narrative intelligence and understand the context of events and the turns they could take.