Machine learning has been the focus of Google Inc. for some time now. A new paper published last week offered insight into what all happens inside the laboratories in Mountain View, California.

Google published earlier this month details of psychedelic dreamscapes that envisions what the artificial intelligence technology called Artificial Neural Networks (ANN) dreams about. The internet company revealed last week through a paper details of the exchange between a human researchers and the artificial intelligence.

A series of troubleshooting problems that involved VPN, crashing browsers etc. were the initial challenges that the machine had to take on. And the machine did better than expectations.

However, the problems emerged when the test came to the philosophical exchange. The machine is a very good display of artificial intelligence, but the technology has yet not achieved the level to know what we humans do even have no complete knowledge about.

The technology seemed to be in tatters when researches presented posted questions about an abstraction like morality. The machine replied when asked to discuss morality and ethics that it was in no mood for a philosophical debate. The machine was then asked about to decide the topic it wanted to talk and the machine replied "nothing".

The AI was fed with knowledge of troubleshooting and several movie scripts to power its neurons, explained the paper. This limited the machine's ability to work only with things it had the information for.

An Oxford-based academic has warned that AI will one day surpass human beings.