In June, a caged robot at the Volkswagen plant killed a 22-year-old technician by pushing the victim against a metal plate. He was assembling the robot that was supposed to grab and configure car parts. The incident was soon forgotten, especially after the maker of Beetle recently got into a bigger problem with the emission test software scandal.

However, the incident is proof of what Tesla CEO Elon Musk has warned of robots with artificial intelligence (AI) one day killing mankind. Volkswagen attributed the incident to malfunction likely to human error.

But Toby Walsh, an AI professor at the University of New South Wales and researcher at Data61, assures that it would be at least 50 years before robots become so sophisticated as to have the ability to become killer robots, reports the Sydney Morning Herald.

What the world should worry, in the meantime, is the “stupid” technology or autonomous weapons systems such as drones that are used for warfare, according to a leak from classified US military documents. The document, now called “drone papers,” state that over five months of operations in Afghanistan, almost 90 percent of people killed during drone strikes were not the intended targets. The papers were published in The Intercept.

Walsh warns that the 90 percent could even go higher without human intervention. To stop the further manufacture of these technology, which if it falls into enemy hands could have disastrous effect, Walsh was part of the 2,000 AI researchers who signed the open letter in July to ban killer robots.

He will give the issue another push this week in New York at a side event of the UN First Committee on Disarmament & International Security meeting. Many nations support the move, but big powers such as the US, China, Israel and Britain have been quiet, while Australia has no formal statement on its stand.

“The US claims there’s always a human somewhere in loop but it’s only a little matter of computer code to take that human out of the loop,” Walsh points out.

In September, Walsh said that the 2013 Oxford estimate that 47 percent of all employment in the US is at risk of being automated using AI could be on the low side. He says, “It’s hard to think of a job that a computer ultimately won’t be able to do as well if not better than we can do,” quotes TechTimes.

Contact the writer at feedback@ibtimes.com.au or tell us what you think below

RELATED STORIES