Experts compare notes on AI’s future
THE fast-growing development of artificial intelligence should be used to augment – not replace – human capability and opportunity. This was the view of experts at an interactive session on artificial intelligence (AI) at the World Economic Forum’s Annual Meeting in Davos-Klosters, Switzerland, on Tuesday.
With the advent of the Fourth Industrial Revolution and the accompanying technological innovations and advancement, especially in the field of AI, it was stressed that AI development should be guided by the overarching principle that technology should not replace human capability, but rather support it.
Experts further agreed that technology and access to technology should be democratised, and said it was essential to provide people with the relevant knowledge and skills to lay the groundwork for a more egalitarian and sustainable era of cognitive computing.
Ginni Rometty, the chairman, president and chief executive at IBM Corporation in the US, which has taken the lead in cognitive computing within the information technology industry and has developed the advanced AI platform Watson, said transparency was imperative to develop trust in cognitive computing.
Soon, everyone would be working with AI technologies and people would want to know how they were designed, by which experts and using which data.
“Humans need to remain in control of it,” Rometty said, adding that it was imperative that technology be created for, by and with people.
Panellists agreed that ethical and legal concerns must be factored in at the start of the design process, underlining the importance for customers, lawyers, ethicists, scientists and technology developers to work together.
Highlighting the need to democratise technology design, Joichi Ito, director of Massachusetts Institute of Technology’s media lab, said it was worrying that the demographic in Silicon Valley consisted of mostly white men.
He gave the example of a facerecognition technology that failed to recognise dark faces, reflecting a lack of diversity among the engineers who designed it.
“AI is still a bespoke art; the customer cannot imagine the tool yet,” he said, suggesting that stakeholders, including the customer, the lawyer and the ethicist, have a say in technology creation.
Satya Nadella, the chief executive at Microsoft Corporation in the US, said his organisation was focusing on how to make technology broadly accessible.
He cited the success of Microsoft’s Skype Translator, the speech-to-speech translation application available for free download.
Speaking of the challenges that lay ahead, Nadella said many questions remained to be answered, such as how to fix responsibility for decisions made by algorithms that humans had not written, and whether the AI surplus that would be created would be shared equitably.
“Overall, world gross domestic product growth is not stellar,” Nadella said. “We actually need AI.”
To ensure that AI and the Fourth Industrial Revolution helped solve the pressing problems of today, it was important to help train people for the jobs of the future, he said.
He added that in a world with a surfeit of AI, human values such as common sense and empathy would be scarce and that these were the values that the citizens of tomorrow would need most to make humanity the very best it could be.
Ron Gutman, the founder and chief executive of HealthTap, an online application that brings patients and doctors together, said AI would create new jobs that did not exist today.
For instance, sensors and wearables provided so much data that it would become possible to move from reactive to proactive medicine, creating a new ecosystem of jobs. Rometty highlighted her idea of “new collar” jobs, which pivoted on the belief that the skills needed for tomorrow’s jobs were not just the high-end, high-technology skills that could only be acquired through a traditional university degree.
Many jobs, such as those of cloud computing technicians and service delivery specialists, would need skills often obtained through vocational training or in non-traditional ways.
She emphasised, though, that everybody would need retraining. Ito agreed, noting that everyone would have to acquire an understanding of AI, and education systems would have to be made more dynamic. – ANA
AI IS STILL A BESPOKE ART; THE CUSTOMER CANNOT IMAGINE THE TOOL