Google supercomputer creates its own 'AI child' that can outperform any machine made by humans I BABY STEPS Google artificial intelligence supercomputer creates its own ‘AI child’ that can outperform its human-made rivals I

A GOOGLE supercomputer has created an “AI child” which can outperform its man-made rivals.

The incredible machine named NASNet becomes smarter through “reinforcement learning” which sees it report back to its “parent” computer when completing tasks.

GOOGLE RESEARCH

2

The AI child can recognise people and object while watching a video in real time better than its man-made counterparts

The AI (artificial intelligence), which was created earlier this year, is able to recognise objects such as people and cars while watching real time video.

NASNet is controlled by a neural network called AutoML which was created by humans at Google Brain.

The parent AI teaches its offspring to do specific tasks which are repeated thousands of times.

Researchers have found that NASNet is 82.7 per cent accurate at predicting images correctly which means it has “outperformed all other computer vision systems,” reports Science Alert.

2

Artificial Intelligence is becoming is becoming incredibly sophisticated and there are fears that the technology could destroy the human race

The system has overall made itself four per cent more efficient through reinforcement learning with its parent.

And because the system is “open source”, the Google Brain team said they hope “the larger machine learning community will be able to build on these models to address multitudes of computer vision problems we have not yet imagined.”

But the rapid development of AI has sparked fears the technology could destroy the human race.

Tesla CEO Elon Musk is one of those who has called for artificial intelligence to be regulated because it poses a “fundamental risk to the existence of human civilisation.”

• The machine called NASNet learns through 'reinforcement learning'
• It reports back to its AI 'parent' and then leanst how it can do it better
• AI can recognise objects, such as people, cars, handbags and traffic lights
• The findings show automation could create more AI all by themselves

• Google researchers have created an 'AI child' that can outperform its human-made counterparts.  
  The machine learns through 'reinforcement learning' which means it trains for a task, reports back to its AI 'parent' and then learns how it can do it better.
  This particular AI recognises objects, such as people, cars, handbags and traffic lights, in real time in a video.
  The creation of this AI child is proof some machine-made programmes are now more accurate than ones created by humans.