AsianScientist (Aug. 16, 2017) – By Rebecca Tan – At this point in time, even the smartest technologies can sometimes be outsmarted by a child. Take for example six-year-old Brooke Neitzel who wound up convincing her family’s Alexa digital assistant to order her a US$160 dollhouse and four pounds of sugar cookies. But as Alexa levels up on her conversational skills, shopping sprees like Brooke’s will soon become a thing of the past.
“My 13-year-old interacts a lot with our Alexa at home and she becomes frustrated when it doesn’t recognise what she’s asking it to do,” shared Mr. Steve Leonard, founding CEO of SGInnovate.
“But the point is, just a couple of years ago, Alexa would have understood zero.” What I say to her is to wait a few years and the technology will continue to move along at an exponential rate,” said Mr. Leonard, addressing the crowd gathered at ‘Silicon Valley and its Ripples’, a panel discussion organised by SGInnovate on 20 July 2017.
The panel included entrepreneur and venture investor Mr. Christopher Schroeder; Dr. Noah Raford, the Chief Operating Officer of the Dubai Future Foundation; and Ms Ayesha Khanna, CEO of artificial intelligence (AI) advisory and incubator ADDO AI, who served as the moderator.
“Not to be overly dramatic, but we are at the precipice of a total transformation,” agreed Dr. Raford. “There is no sector of the economy—information or otherwise—that isn’t going to be affected by AI and automation in the next decade.”
How machines learned to learn
AI has truly come a long way, said Mr. Schroeder, who reminded the audience that AI is not actually new but has already experienced one round of boom and bust in the 1950s and 1960s. Citing the example of an automatic Russian-to-English translator developed by the Americans during the Cold War, he shared how AI at the time turned the phrase ‘the spirit is willing but the flesh is weak’ into ‘the whisky is strong but the meat is rotten’.
“The breakthrough came when machines learnt how to learn,” he said. “This happened when computing power grew and we had access to massive amounts of data. But the real essence of it was very sophisticated algorithms that replicate the way the brain works at incredible speed.”
“You can actually see how that happens with a literal translation,” Mr. Schroeder added. “Although language translation is still shockingly hard, AI can now look at a phrase like ‘Paris Hilton was in Paris at the Paris Hilton listening to Paris’ and recognise which ‘Paris’ is the person, which one is the hotel and which one is the music.”
The speed at which AI is improving will only continue to increase, added Dr. Raford, who traced the rapid development of AI innovations coming out of Silicon Valley, from technologies that could only recognise things, to being able to sense patterns and finally being able to understand complex tasks and perform them.
“Last year, Tesla alone clocked 140 million miles driven by self-driving vehicles,” he said. “All the Tesla cars are essentially one giant computer mind. Every time you drive your Tesla on autopilot, it is learning and sending that information back to help other Teslas learn. So, you can see how they can learn exponentially quickly.”
Where does this leave humans?
“At this point I usually get asked ‘What does this mean for our jobs?’,” Dr. Raford continued, adding that labour economists are in consensus that about 50 percent of all jobs will be automatable within the next ten years.
“Anything that is repeatable—any task that you can write down as a series of instructions—will be done by a piece of software or a robot,” he said. “We derive such a sense of personal value and identity from our jobs that the prospect of a future without them leaves us facing some huge social and cultural challenges.”
But the impact of AI and automation is not all doom and gloom, the panellists said. Drawing again from history, Mr. Schroeder recalled how people resisted automation during the Great Depression, when almost 25 percent of all Americans found themselves out of work. While acknowledging that AI and automation are indeed causing a great deal of disruption, he nonetheless noted that the unemployment rate is now only at five or six percent.
“This should help us think about whether there really will be no work at all, or whether there will be different kinds of jobs that could be augmented by AI,” Mr. Schroeder said.
Seeing and seeking opportunities
To illustrate his point, Mr. Schroeder asked the audience if they would hire someone who is unbelievably good at math, incredibly honest, very attentive to data and has incredible customer service skills to boot.
“The answer is: of course. Well, I’ve just described to you a bank teller!” Mr. Schroeder quipped. “This shows that we need to think not just about jobs being taken away, but also broaden our horizons and think about the skills and talents that people already have, and how they can be applied to this world.”
“Our basic hypothesis at SGInnovate is that Singapore does not lack anything that it needs in order to build globally-relevant, technology-intensive companies,” said Mr. Leonard. “The opportunities are there. We don’t lack investment dollars or corporate support, we don’t lack great universities doing great research and we don’t lack an excited government that wants to help.”
Agreeing that Singapore has the unique opportunity to leapfrog Silicon Valley, Ms Khanna shared how her small team at ADDO AI has managed to win deals against much larger and more well-established competitors.
“I think there is a huge opportunity for those who have deep tech skills and for those who are driven and optimistic,” she concluded.
Asian Scientist Magazine is a content partner of the SGInnovate.
———
Copyright: SGInnovate. Read the original article here. Photo: Rebecca Tan.
Disclaimer: This article does not necessarily reflect the views of AsianScientist or its staff.