Friends with benefits?
Forth etech no-optimists, artificial intelligence may well be as close as we get to a super power. But, for the technopessimists, the rise of artificial intelligence could be hastening our own demise. So is this burgeoning‘ super power’ a blessing or a cu
We’ve all dreamed of having a super power at some point in our lives. As a child you may have longed to fly out of your bedroom window into the night sky like Peter Pan. Perhaps you wanted to read other people's minds, live forever, or turn the clock back to reverse a regret or to save a life.
That’s not going to happen. But with the rise of artificial intelligence, some believe we finally have an opportunity to augment our human experience and create a true super power.
As a report from Chapman Tripp and the Institute of Directors called 'Determining our future: Artificial Intelligence' says: “The goal of much AI research is to push forward the boundary of machine intelligence with the eventual goal of creating artificial general intelligence – a machine that could successfully perform any intellectual task in any domain that a human can.”
For many, the idea of a machine performing a task as well as or, worse still, better than a human is a chilling proposition. But even if you’re in this concerned camp, the spread of artificial intelligence as it seeps deeper into all of our lives is, as Kevin Kelly’s book puts it, inevitable. There is too much economic incentive, but, as history has shown, technological advances are not without their dangers. So can we get the balance between man and machine right?
So what exactly is artificial intelligence? You’ve probably heard terms like AI, machine learning and deep learning spouted every way you turn these days. And while they are all intertwined, they are not the same.
In short, deep learning is part of machine learning, which is part of AI. Intel’s Nidhi Chappell, head of machine learning, puts it succinctly when he says: “AI is basically the intelligence – how we make machines intelligent – while machine learning is the implementation of the computing methods that support it. The way I think of it is: AI is the science and machine learning is the algorithms that make the machines smarter.”
AI and machine learning already influence many aspects of our lives – from facial recognition to automated trading to voice activated assistants to recommendation engines – and it’s set to impact many more in the coming years. New Zealand aims to be a keen surfer on this technological wave, and Science and Innovation Minister Paul Goldsmith launched The AI Forum of New Zealand (AIFNZ) in Wellington in June.
The chair of the new AI Forum, an initiative by NZTech, is Stu Christie, who is also an investment manager at NZ Venture Investment Fund with close to 30 years of industry experience behind him. So why launch the organisation now? He puts it down to a few things, like the collection of a mass amount of data, being able to process data of such scale, advances in machine and deep learning, and advances in sensory tech.
“So there’s been a whole bunch of different technologies and the capacity to be able to process that technology which is now bringing that to the fore,” Christie says. “So all those components are coming together to be able to make [AI] happen.”
SCIENCE NON- FI CTION
Ian Watson, an Associate Professor in Computer Science at the University of Auckland, has over 20 years expertise in AI and he says he initially got into the field through an interest in science fiction when he was a kid.
“When I went into computer science the only real area of computer science that interested me was AI,” he says.
He predicts New Zealand will see a lot of applications for AI in agriculture.
“We are now at the point where we can see that there will be robots for example, that could run a whole milking shed and you wouldn’t need the milker there. We can see robots now that would be capable of picking fruit, which of course would have a lot of impact on seasonal work.” Before too long, he says it’ll be drones inspecting the fence lines and monitoring stock rather than farmers.
Unlike Watson, Chris Auld, the director of Developer Experience at Microsoft NZ, says he’s a data guy – but he’s also a technologist, business strategist and a Microsoft Most Valuable Professional (MVP) who has happened to train as a lawyer. And for those of you who have ever been caught in Auckland traffic, he’s got some good news as it is in the early stages of a project with Auckland Transport to try and alleviate the gridlock.
“We’re talking with them about these sorts of technologies and their potential to help with congestion monitoring, congestion modelling, congestion alleviation – the ability to look and see through this image or video analysis where congestion might be and then to make intelligent decisions about how we change traffic light timings and work to reroute the network to ease that congestion.
“So there are huge opportunities in that sort of simulation and modelling. We have an initiative that we’re running around the world focused on traffic management and also traffic safety, driven by artificial intelligence and machine learning.”
SI ZE MATTERS
Machines l ack the capacity to be racist. Machines l ack the capacity to be misogynistic or sexist. Machines j ust l ack the ability to be an arsehole. So we should celebrate that fact. Chris Auld / Microsoft NZ
Christie says the world is waking up to AI, but, because New Zealand is small and agile, we’re less
encumbered by structural issues in terms of our economy and more able to embrace the changes.
“We have an open labour force; we are easy to do business with; we’re a heavily connected first world country but small enough also to be able to collaborate very closely together.”
He points out New Zealand is not leading edge in AI as the deep research and development is largely being done by the tech giants offshore.
“So we’ve got to recognise our position in the market and actually leverage sustainable, competitive advantages that we may have,” he says, explaining the opportunities are biggest in agriculture,g, manufacturing,g, infrastructure and transportation.
However, he does give special mention to Soul Machines (see profile page 97), which is developing remarkable lifelike avatars that display emotional intelligence.
“They are a standout for New Zealand right now. It’s just incredible what they are doing, revolutionising that particular touch point, that customer interface. It’s also enlightening people in terms of what a digital employee may be.”
Mark Rees, general manager of product – small business for accounting software giant, Xero, says what is so exciting about AI is “often you don’t have the ability to look at everything apart from the averages, but with some of these tools you can really see what is the underlying structure in the data, which is really fascinating. It’s like discovery; it’s revealing, like archaeology.”
While there is plenty of chatter about the potential for automation to take jobs, he says AI is set to change the accounting process for the better and, in around five to ten years, low-value, commoditised data entry for accountants will be low-friction, perhaps even completely automated, and will allow them to do more productive things.
“We provide really smart, alerting recommendations that helps business advisers optimise the performance of their business customers. That’s what they focus on, not the mechanical side of data entry or tax preparation but the machines are really helping the business advisers give really smart advice to their customers and the businesses are run better because of that.”
Although building AI into the business offering will help Xero’s advisers, it’s a disruption to them too.
“Our strategy is that we want to help the accountants change their business into more high value services – it is a disruption and with any disruption, people have to make choices about how they respond to that, but I think it does provide a real opportunity for them to adapt their businesses and focus on business advice … I think the misconception is that it’s something radically new when it’s progressively been baked into our experiences.”
BE CAREFUL WHAT YOU WISH FOR
Sarah Hindle is general manager of Tech Futures Lab, and the founder, Frances Valintine, also sits on the AIFNZ Board. Hindle has advised CEOs throughout her career on how to remain ahead of the competitive curve when rapid change is imminent. She also studied philosophy at university, so she takes a slightly different, more holistic view of this shift and the impact it may have on our human existence.
“I think what is becoming really clear now is actually we are computers and AI is showing us that the space between our ears is actually not that much different from something that we can create with a machine.”
Because of the rapid developments in this area, she says it is vital that we start having a conversation “of a nature that we have never had at any other point in history, which is how do we really want to live our lives? Do we need to be working 9-5? What does the purpose of life look like? How might we survive without getting an income five days a week? What other options does that open up for us as a civilisation? I think that’s the most exciting thing – just as a trigger for reconsidering our whole existence.”
Tech Futures Lab launched in July 2016 and, with many of its partners also very involved in AI development, Hindle says it has worked with 3,000 people and 250 companies across every sector to ‘agitate’ that conversation.
“Of course you want to give people the security that it’ll all be fine, but I really think it’s in our hands as to whether we really make this the greatest thing that humans have ever done by really having a chance to recast that social contract and what it looks like for us as humans and eliminate poverty and solve diseases and have a life where we do what we want. Or, we could really muck it up and have a very split society.”
Personally, she doesn’t believe that everyone will slide into a new job once they have been booted out of their old one by a cheaper, more efficient machine.
“I think actually what we are going to need to do is figure out a way whereby we don’t all have to be employed 40 hours a week to survive as dignified human beings. We need to have a very different conversation about what it means to be a valuable member of society and to be a human, so I think that my greatest reservation is our ability and knowledge and willingness to have those conversations and to have them quickly enough so that people can live a good life.”
The report by Chapman Tripp and Institute of Directors also indicated that lower socioeconomic communities would be the ones most likely to feel the effects from AI development, with low-skilled and repetitive jobs at the highest risk of being taken over by technology.
Another recent report by The Royal Society stated 35 percent of jobs in the United Kingdom could have more than a 66 percent chance of succumbing to automation in the next few decades. But it also said “common ground on the nature, scale, and timing of potential changes to the world of work as a result of machine learning is hard to find”, so, at present, there are only guesses.
Jeremy Howard, the founder of and deep learning researcher at fast.ai, has 25 years of machine learning study behind him. And, in a TED talk in December 2014 he mixed up “the wonderful and terrifying implications of computers that can learn”.
“The machine learning revolution will be very different. The better computers get at intellectual activities, the more they can build better computers to get better at intellectual activities. So this is going to be a change the world has never seen before, so your previous understanding of what’s possible is different.
“Computers right now can do the things that humans spend most of their time being paid to do, so now’s the time to start thinking about how we’re going to adjust our social structures and our economic structures to be aware of this new reality."
FI ND AND REPLACE
Associate professor Watson says a major threat is the wider societal impact resulting from advances in AI.
“It’s all very well for an individual company to decide to lay off a third of its workforce – but then if every company in that sector decides to
The world i s awakening to AI but New Zealand’s small and agile nature does make it more responsive and also we have l ess i ncumbent structural i ssues i n terms of our economy. Stu Christie / AIFNZ
lay off a third of their workforce then suddenly you’ve got an awful lot of people who don’t have jobs to go to and that is potentially catastrophic.
“Of course, if it’s left to individual companies to make decisions then they have to make decisions based on their bottom line – on their return to shareholders, that’s their responsibility. So really society as a whole needs to think about this and think about the impacts.”
Many believe one of America’s most common jobs – driving trucks – could soon be extinct due to the rise of autonomous vehicles. So what will those millions do? To address this, the Forum’s Christie says we need to make sure to have “a reiterative education system so that people can retrain in their lives and do that in ways which can get them up to speed and adaptable and an accepting society which does accept that essentially people are going through that process – the investment will also have to come from businesses, not just the individuals to carry the burden of that retraining”.
“The real opportunities here aren’t removing people from the loop, they are giving people better tools to make person to person interactions better,” adds Microsoft’s Auld.
Auld agrees with Christie that New Zealand is well-positioned to navigate this shift due to its close relationship between citizens and government. “We had the Minister presenting [at the Forum launch]; you can bump into the Prime Minister at the airport. We don’t have many countries in the world that are like that. We have a country that is really amenable to flexible, adaptable, smart regulation, so I think that’s going to be key.”
In Seattle earlier this year, the annual three-day Microsoft Build conference took place. Curiously, Microsoft CEO Satya Nadella opened it with a frank warning to the deep technologists in attendance of creating a dystopian reality not dissimilar to George Orwell’s 1984. Entrepreneur billionaire and SpaceX/Tesla CEO Elon Musk has gone as far to say AI is “our greatest existential threat” while Professor Stephen Hawking has warned humans will be helpless to compete with AI and will be ultimately ‘superseded’.
During Techweek this year, Watson gave a lecture exploring the questionable impacts and ethical implications of AI and says there is one area that AI should be forbidden to enter.
“I think probably the only area that one would definitely say you don’t want AI is in terms of autonomous weapons systems – definitely not. It’s perfectly feasible now that those systems could acquire their own target and be allowed to fire rockets but there’s a large number of people who think that shouldn’t be permitted, that there should always be a person in the loop, who can be held responsible for making the decision,” he says. “Why would we want to release weapons out there that can make their own decisions as to whether or not they should shoot us?”
Microsoft’s Auld also says autonomous weapons systems are a great example because “there’s something unique about going to war. It requires a human, someone to make moral decisions. I think that we should avoid putting machines into places where they have to make moral decisions, because they can’t make moral decisions.”
To Auld, that’s something to be taken advantage of.
“Machines lack the capacity to be racist. Machines lack the capacity to be misogynistic or sexist. Machines just lack the ability to be an arsehole. So we should celebrate that fact. We need to be careful about how we build these machines so that they don't make biased decisions accidentally. But artificial intelligence is not like humans; it doesn’t have the innate tendency to cast judgement.”
CHECKS AND BALANCES
Auld also attended the Seattle conference but says the bleak future some are worried about is a long way off.
“The thing about dystopian futures is they’re an extremely long way away,” he says, pointing out that there has been technological disruption and tech-driven unemployment for a long time.
“I think the disruption to people’s lives is probably going to occur less quickly than it has in the past. I think we’ll see the positive benefits accrued far more quickly than we find the negative consequences. But that’s not to say there won’t be negative consequences.”
The AI community here and around the world is working on putting controls in place for their creation. Google’s DeepMind, a world leader in AI research and a company Musk himself invested in, has developed an AI ‘off-switch’. New Zealander Dr. Shane Legg is a cofounder of the London startup that was established in 2010 and was snapped up by Google four years on for about £400 million.
The Future of Life Institute launched a programme in 2015 to research AI safety, funded largely by a donation from Musk. Partnership on AI was formed to explore best practices on AI technologies and as an open platform to discuss the impacts of AI. Non-profit OpenAI is an AI research company, furthering a safe path to artificial general intelligence.
Watson mentions Bill Gates’ suggestion that robots should be taxed if they are doing work, just as humans are.
“That tax revenue could obviously be used for social security but it could also be used as a
lever to control how fast automation is rolled out – if the tax is quite high then the AIs are not as economically efficient, they’re not as attractive. And if the tax is super low then they are very attractive so policy makers could play with that tax to control how fast or slowly AIs are deployed and I’ve got no idea how governments would tax something like that but they seem to be perfectly capable of taxing anything they feel like. I’m sure they would be able to think of a way of doing it.”
GREAT POWER, GREAT RESPONSIBILIT Y
But is it likely that AI will ever reach humanlevel intelligence? A report from the Obama administration late last year said we won’t see machines “exhibit broadly-applicable intelligence comparable to or exceeding that of humans” in the next 20 years, but Google’s director of engineering Ray Kurzweil certainly thinks we will.
“By 2029, computers will have human-level intelligence,” Kurzweil said in an interview early this year, during the SXSW Conference in Texas.
And more technologists and visionaries agree with him. IEEE Spectrum asked a number of them, including Rodney Brooks and Nick Bostrom, when we will have computers as capable as the brain and nearly all said it would happen, but the time frame ranged from ‘soon’ to hundreds of years away.
Microsoft’s Auld says it’s a deeply epistemological question, but adds, “I don’t think we’ll ever get there, and that’s probably a good thing”.
Although Hindle shares the concerns of the likes of Musk and Hawking, she says AI will redefine what it means to be human and what our lives will look like.
“There are lots of scary things about AI and I would be lying if I would try to deny that, but I think what is exciting about it is it almost gives humans a super power – it doesn’t just improve what we’re doing but it kind of gives us this extra capability by being able to access information at speed that we’ve never had before.”
Geoff Colvin, the author of Humans are Underrated: What High Achievers Know That Brilliant Machines Never Will, is confident humans and AI will live alongside each other. He says the greatest advantage we have over technology is that which we already possess and are hardwired to want only from each other – things like empathy, creativity and humour – and that we must develop those abilities. Whether you are dreading a Terminator- style future, or dreaming of the way AI will improve our lives, one thing is certain: AI is already here and only gaining momentum. So, as Hindle says, “we’ve got to move with the machines, not against them, because we can’t stop it”.
Why would we want to release weapons out there that can make their own decisions as to whether or not they should shoot us? Ian Watson / University of Auckland