Microsoft tech chief: AI won't cause 'dystopian takeover' by robots

·Finance and policy reporter
The Microsoft sign is shown on top of the Microsoft Theatre in Los Angeles, California, U.S. October 19,2018.  REUTERS/Mike Blake
Richard Potter, chief technology officer at Microsoft Services, spoke at the World of Work conference at the University of Reading. Photo: Mike Blake/Reuters

A senior figure at Microsoft (MSFT) has made a rallying cry for the power of artificial intelligence to fix social problems, despite admitting it could be “unsettling” in daily life.

Richard Potter, chief technology officer at Microsoft Services, said the IT giant was developing AI “for society” as well as a business opportunity.

He said it did not see the future as “some dystopian takeover by Skynet,” referring to the menacing computer system that gains self-awareness and launches a nuclear weapon in the Terminator films.

Potter argued AI could bring significant benefits for humanity, but emphasised that how it was used was a choice for society rather than big tech companies.

In a speech at the World of Work conference in the University of Reading’s Henley Business School on Thursday, he said AI was now the defining technology in digital disruption.

He said the increasing power of silicon chips, the generation of data and “great mathematics” were making it a “scorching summer of AI,” with significant progress in its development.

AI ethics and responsibility

Potter acknowledged new technologies could unsettle people, but said it was right to “expose the darker places, and think through the implications of what we’re doing.”

“The unsettling nature remind us of the responsibility that exists with this technology. There’s so much power in this technology that our focus on responsibility is relentless.”

He said Microsoft itself had its own ethics processes specifically for AI, with all such work in Britain signed off at a high level before it begins.

But he added: “We see the future as less about some dystopian takeover by Skynet, more about the augmentation of humans with technology to develop great outcomes.”

READ MORE: Top HR chief attacks Amazon warehouse jobs in warning to UK firms

Face detection software to verify age and identity

One example of AI’s use he admitted was “unsettling” is Microsoft’s own age verification technology.

Potter said it is used at stores at Microsoft’s own offices, saying he had picked up a small bottle of sparkling wine without attracting attention.

The technology, involving cameras at shelf level within the store and facial detection tools that use AI, seeks to identify any customers under 35.

But he said a younger colleague had been surprised to hear an automated announcement that an attendant was on their way to check his ID immediately after he tried to scan it into his basket.

AI was also being trialled at the same store that could monitor stock levels by analysing camera footage in real time, identifying “the fresh goods like muffins that need to be replenished,” according to Potter.

Such technology enabled one individual to “run the entire store off a Surface Pro [device] in the back,” he said.

READ MORE: A LinkedIn chief perfectly sums up why diversity is critical in artificial intelligence

Uber is now using Microsoft’s wider ‘Face API’ technology to verify the identity of drivers signing in to use its app.

Drivers are now occasionally asked by Uber to take selfies, which Microsoft’s AI system scans and compares to previously provided photos to confirm who they are.

Amazon rolled out its own AI-based verification software, Rekognition, in 2017, with the promise of face detection, analysis and comparison for “user verification, people counting, and public safety use.”

Monitoring vulnerable people at risk of accidents

He said Microsoft was working with “cash-strapped local authorities” to help them use voice recognition software and a voice-based chatbot to talk to, monitor and assist lonely older people “struggling to cope on their own.”

AI devices could analyse their conversations “to understand whether an older person is more lonely this week than before,” and help them place meals-on-wheels orders.

READ MORE: Women twice as vulnerable to losing jobs to robots in the UK

Microsoft joined the Agile Ageing Alliance in 2017, a group looking at technology in social care. It suggested AI could help monitor for and even predict events like falls, helping older people live at home for longer rather than moving into care homes.

“Wearable devices can alert relatives and healthcare professionals if a patient has a fall, for example, or sensors could spot changes in the amount of water being used in the house, suggesting a person is incapacitated,” the company said at the time in a statement.

Potter said he would consider using such services for his own mother, but admitted leaders had to “navigate very carefully” through such new technologies.