Microsoft chief executive Satya Nadella has said it will be down to technology giants to define ethics in the coming era of artificial intelligence, and to plan developments so that society is not left bereft by automation-led unemployment.
Speaking at his company's annual Ignite and Envision events in Orlando, Mr Nadella fielded questions about the human impact of Microsoft's ongoing development of AI, and warned governments could eventually be required to pay citizens to perform different, human-centric tasks.
While Microsoft uses its conferences to educate customers on its latest product offerings, which increasingly use AI components to augment performance, Mr Nadella said the company had established an ethics committee to make judgments on how intelligent systems should act in the future.
Governments around the world, including Australia, are coming under pressure to prepare the workforce for a future when AI-driven automation has rendered some jobs redundant, and Mr Nadella said it made sense for the companies developing the technology to take responsibility.
He said Microsoft's committee included a diverse range of employees, including AI experts, design professionals, computer scientists and anthropologists.
"It is very, very clear that we want to make that design choice of augmenting humans," Mr Nadella said.
"Because even as we bring about new levels of automation and efficiency we have to make sure we don't degrade the human quality."
Mr Nadella said there were conflicting views about the likely evolution of AI, with Elon Musk painting a worrying picture of hyper-efficient robots effectively dominating the workforce.
Taking responsibility for algorithms
However, he said he believed that if companies like Microsoft took responsibility for the algorithms they created, then they could help improve rather than replace human skills.
Referring to Microsoft's earlier AI experimentation with a conversational Twitter Bot called Tay, he said the company had learnt it was important to retain control of the systems as they learnt, rather than let them evolve in the wild.
Tay caused embarrassment for the company when it reflected the views expressed to it by Twitter users, and was scrapped after making aggressive and racist remarks.
"The experience around Tay taught us that you can't abdicate [responsibility for the algorithms we create], we need to take more accountability," Mr Nadella said.
"We should have designed Tay to be resilient to even human infallibility so to speak ... [With employment] the first thing with new technology is there is going to be displacement and the best insurance a society has against displacement is education."
Mr Nadella said companies like Microsoft needed to work with schools and universities to develop new methods of educating those who may be displaced by technology.
Basic national income
He suggested governments could ultimately be required to fund a basic national income scheme, to encourage underemployed professionals into people-focused service jobs, to serve society and maintain their "human dignity".
"In a world where there is abundance of AI what is going to be scarce is real intelligence. What I mean by that is the human qualities that we all take for granted, like empathy," Mr Nadella said.
"That will be in scarce supply and so people-on-people jobs like elder care and care for people with special needs, will need some combination of policy changes as well as these jobs so that the wages of those jobs are good.
"I believe that we will have to come together as a society to solve - without falling victim to the fallacy of limited labour or thinking all jobs are going to be gone - but in making sure we are skilling people for the right jobs of the future and in creating these new jobs whether they are people-on-people jobs or new jobs of art design or creativity using digital media."