Learning to love AI: Dr Ayesha Khanna sees a new way forward with artificial intelligence
Originally posted by Adam Courtenay @ intheblack.com
For Ayesha Khanna, artificial intelligence must be an aide to humanity, not a destroyer of it. Ahead of her appearance at the World Congress of Accountants (WCOA), Khanna speaks to CPA Australia about AI, accounting and putting humans first.
Many in the financial world have a problem with technology. We’re looking down the barrel of a digital world, driven by things we can’t touch (like bitcoin) and things we can’t control (like blockchain).
There are fintechs and bank bots of every description on every virtual street corner, putting their hands up to replace the analogue norm of everyday finance. Forget what you’ve seen and done, they seem to be telling us; get disintermediated or die.
The digital world is encroaching and many of us are frightened, but there’s no need to be, says one of the world’s foremost thinkers and educators on artificial intelligence, Dr Ayesha Khanna.
Even if the fintechs and banking alternatives are seeing opportunities after the Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry, Khanna is not. This new world we seem to be forced into entering is not about market share or efficiency, or even eradicating human error, according to Khanna, who runs Singapore-based AI consultancy ADDO AI. It is certainly not about displacing for displacement’s sake. For Khanna, AI must be an aide to humanity, not a destroyer of it.
Data as an aide to humanity
To illustrate her point, Khanna explains how her firm was approached by the Dubai Government, which wanted to be a player in the new, big data-driven economy. It wanted to be savvy and smart, and create a one-stop shop for all government services.
It sounded ideal: pay your tolls, permits and fines on a single platform. The problem was that an integration of over 50 government agencies in one swoop was never tested by the end user – the tax-paying citizen. Nobody asked the people what they wanted, explains Khanna, and few used the new integrated platform. If a technology is not human-centric, she says, it will always fail.
Khanna’s firm redesigned Dubai’s suite of services by personalising the experience. It ended up looking more like Facebook and Netflix and less like a piece of faceless online bureaucracy. The people came in droves.
“Netflix and Google know how to personalise experiences,” she says. “Give people things they find interesting. That should always be the approach – and it’s not just for the digital route, but also the way you build housing, apartments and institutions.”
According to Khanna, technology “has to tango with humans” – it has to give us what we need for a more relevant, productive and higher quality life.
Arts and algorithms
Khanna’s road to becoming one of the world’s AI cognoscenti has been a circuitous one. She wasn’t born a fintech genius. She grew up in Lahore, Pakistan, in a conventional, middle-class and highly intellectual environment. Her education was solidly grounded in the arts, but she wasn’t shy of mathematics either.
“I’m Asian,” she laughs, “and being Asian, we all do a lot of mathematics.” That said, her education smacked more of Tolstoy than trigonometry.
Her father was a highly placed civil servant and her mother an English literature professor, and both took on extra studies mid-career. Her mother was a visiting scholar at Harvard University and her father completed a master’s degree at Fletcher School of Law and Diplomacy at Tufts University in the US.
Her home buzzed with discussions on politics, literature and philosophy. She was a single child but was never pushed into being anything or doing anything.
“I was never alone,” she says. “It was no doubt a pretty sheltered life, but with lots of extended family, plenty of cousins and grandparents.”
There were no real mentors, other than the natural curiosity imparted by her enlightened parents. There was no talk of going into business, and the computer sciences or engineering worlds were not even alluded to.
“I want to give girls the creative confidence to participate in any company using AI. I want them to be bold enough to collaborate and question the method.”
When Khanna took the state school-leaving exam with 53,000 other students in the early 1990s, she came first. It granted her a passport to Harvard, where she would study economics.
“I couldn’t believe it, and neither could my parents,” she says. “They told me to go check back with the examination board to see if there hadn’t been a mistake.”
Khanna is self-deprecating, but her love of the humanities has never waned. When most Aussie kids might have been hitting the surf in summer, Khanna was offering her services at Pakistan’s Human Rights Commission and writing plays in her spare time.
Her economics degree had less relevance, she believes, than the people who surrounded her. She shared dorms with the smartest young minds in Eastern Europe, who taught her that there was mystery and poetry in science and technology.
“They looked at maths and science not in a rote, painful, grade-oriented way, but in an artistic way,” she recounts.
“They were all Olympiads at mathematics, the very best in their respective countries – Romania, Croatia and Russia – and they looked at science and technology in a way I’d never even thought of.”
Khanna never really looked back. She did a master’s degree at Columbia University in operations management, and worked as a software engineer in New York. By the mid-2000s, she was working on algorithmic trading systems and risk management systems. She had met and married her husband Parag Khanna, a renowned geopolitical strategist and academic. Then Wall Street collapsed.
Putting humans first
While she was not involved in constructing collateralised debt obligations or any of the problematic financial products, Khanna says she still felt partially responsible for the losses people experienced.
“Even if I had not built those models, I was a part of the system,” she says. Few saw the corrupt end game and when it was obvious, most people she knew didn’t care.
“So many showed so little remorse,” she says ruefully. If this was a seminal moment in her thinking, it surely worked. Products might be technically highly evolved, but what’s their point without a humane outcome?
She returned to studying, taking a doctorate at the London School of Economics, which focused on information systems and innovation. Her thesis was on smart cities and her contention that smart cities had to put humans first.
Khanna and her husband’s decision to relocate to Singapore has certainly been a smart one. Her career trajectory since their arrival has been stratospheric. She has written books on smart cities and become a renowned speaker and thinker on the role of artificial intelligence throughout the world.
Allied to her interest in AI has been her desire to see women of all faiths, backgrounds and races take up science and technology. It is why she started 21C (21st Century) Girls. The registered charity teaches school girls coding and runs a series of AI workshops for polytechnic students.
There is a story often told about her. At a hackathon a few years ago, Khanna came over to watch a young girl assembling parts for a robot. In mid construction, the girl’s mother stopped her daughter and shooed her away.
“Oh, my daughter, that’s not her interest,” the mother said. Khanna was more than a little bemused when the mother duly thrust her son in front of her. He was allowed to play with robots, the girl wasn’t.
“I want to give girls the creative confidence to participate in any company using AI. I want them to be bold enough to collaborate and question the method. I want them to be able to say they can appreciate what someone tells them but have the confidence not to agree. Imparting that kind of confidence gives me a lot of pleasure.”
Smart city Singapore: the world's leading info-state
Khanna and her husband chose Singapore, not just as the best base for her company, but also because it is the smartest city in the world. It is not just a city-state, but also the world’s leading “info-state”, leveraging its knowledge and technology, not its economic or political might. Here, they could live the dream.
One has the strong impression Khanna could not pursue her profit-oriented AI consultancy without simultaneously engaging in her non-profit yet singular desire to see women gain greater access to technology. The two go hand in hand.
“I don’t have a problem with people not being wealthy, but I do have a problem with people having unequal access to things,” she says.
“I’m interested in anything that improves the life of the urban emerging middle class. It could be transportation, financial inclusion or the personalisation of citizen servicing – all these different techniques using AI point to a lower cost of service, higher quality of life and the democratisation of access to services they didn’t have before.”
AI and accounting
Yes, Khanna says, the world of artificial intelligence is coming to accountants and financial professionals, and the basics of accounting practices will be hugely disrupted. This is because there remains so much routine, standard and repetitive work in the business.
“Then again, one would be hard-pressed to find any profession in which there is not a part of it that is routine and repetitive – that’s the part that can be codified and structured.”
Accountants will likewise need to up their game – less rote work, more creative thinking. A smart accountancy firm, she says, will think of AI as a facilitator and a colleague, not as a master.
“If you treat it like a black box and listen too much to it, you may forget it’s been highly designed and is prone to its own bias. It’s not the be all and end all – you have to have the confidence to question it.
“Work with it, not for it. Its pure logic is not beyond questioning. You have every right to interrogate it and every right to understand it.”
Nor do we all have to be technical geniuses to adopt AI.
“If you can learn enough so that you can collaborate with the highly technical people working in AI, then together with the shared knowledge, it will move your company to being more successful,” Khanna says.
A smart company not only has the algorithms at hand, it can also explain them and know what drives them. Conversely, if a company cannot explain the workings of its technology and how it affects clients and individuals, it should not be using it.
“If a company uses machine intelligence to refuse medical insurance to a person, you better know why – you can’t let a machine do it. Accountability, governance and transparency are far more important than the products of algorithms.”
Living in an AI world
Khanna is well aware that many countries – including Australia – are standing on the edge of a precipice. Fear is a big part of not making the leap towards automated intelligence. There are fears of industrial dislocation and mass unemployment.
Part of the problem is demographic. Countries in Asia and South America are embracing technology with greater alacrity than their counterparts in the West because their populations are younger. That said, she says Australia has excellent technology schools, on par with much of the Western world.
Khanna sees the problem as educational and says her aim now is to promote her firm globally and help companies and cities understand what is on offer as a new industrial revolution begins to unfold.
ADDO AI is about to open a Silicon Valley office, and Khanna wants to do more writing on smart cities and the human benefits of artificial intelligence. She also wants to interest more women in coding, but doesn’t see that as the end game. The end game is a technological “even playing field” for everyone, which does not discriminate by age, gender, station in life or geographic position.
She is now helping to develop the first global online AI finance course and will be doing much of the lecturing. Already the course has been accepted by a number of prominent banks in Singapore as an educational tool.
“The point is never to discount your own depth of experience and understanding,” Khanna says.
“Don’t shrink back when this new wave comes. Just remember: you are the real driver. You have so much to offer the AI world.”