Artificial Intelligence makes art, knows more than many humans and works faster than they do. But will people accept AI-controlled social robots working in the service industry or entertaining those in need of care?

What does a robot need to have to be accepted as a social partner by a human being? Does it need a face? Should the machine understand — or even show — emotions?

The psychologist, neurologist and philosopher Agnieszka Wykowska, currently researching at the Italian Institute of Technology in Genoa, says: “We tend to humanize everything. We even see faces in car hoods. This is further reinforced whenever a robot demonstrates humanlike behavior.

In a care home for the elderly in Rendsburg, the film shows what sort of relationship forms between residents and robots. Hannes Eilers from the Kiel University of Applied Sciences is carrying out tests there with robots for health insurance companies. The robots sing with the elderly people, play games or demonstrate physio exercises. The one thing they’re not allowed to do with them is pray. The systems there function autonomously. This means they can’t access an AI server, so they abide by data protection laws.

But AI servers are already controlling much of our communication. They don’t just suggest what we should read, eat or buy next: ‘chatbots’ also serve as personal contacts. At the Massachusetts Institute of Technology (MIT) in Boston, the scientist Hossein Rahnama is working on perfecting the appearance and communication skills of chatbots like these. His view: “We now have access to such immense computing power and data that we can create a digital version of every person. Before too long, we can even make them sentient.

In future, will we be able to tell the difference between a flesh-and-blood human, and their digital clone?