Let's start with a riddle. What do Florence Nightingale, a city lawyer named 'RAVN ACE' and a Japanese cucumber farm have in common?
If you thought "the unlikeliest set up to a joke", you're not entirely wrong. However, the answer I was looking for, is artificial intelligence (AI).
Florence Nightingale, born in 1820, was more than just the 'lady with a lamp'. She pioneered a statistically driven approach to healthcare that has formed the basis of much modern AI approach and thinking.
RAVN ACE isn't just any old lawyer; it's an AI system. It is able to do the work of a barrister 2,000 times faster. Recently, RAVN ACE assisted the UK's Serious Frauds Office in a bribery and corruption investigation that resulted in a GBP 671 million settlement.
Looking East, the Japanese cucumber farm involves the use of AI to automatically sort cucumbers by size, shape, colour and other attributes. It was created by Makoto Koike, an embedded systems designer from Japan who wanted to help reduce his parents' burden by automating their cucumber farming.
All around the world and everywhere around us, artificial intelligence is making the world a better, more efficient place. History may show us that AI-related concepts aren't entirely new. Yet, it is undeniable that AI is growing at an exponential rate in modern times, thanks to faster computer chips (Moore's law), a data explosion (100 million pictures are posted to Instagram everyday) and cloud computing.
AI and UN Sustainable Development Goals
I shared these points recently at a talk series hosted by CIMB. The talk was part of the bank's commendable efforts to develop a better understanding of emerging technological trends and their impact on sustainability. When asked to present on "AI and Sustainability", I thought of the just-under 50% of South East Asians who don't own bank accounts—what role could AI, sustainability or even financial institutions fill in their lives?
Recognising that sustainability is such a wide term, my research focused on the United Nation's Sustainable Development Goals (UN SDG's). Otherwise known as the Global Goals, they are a universal call to action to end poverty, protect the planet and ensure that all people enjoy peace and prosperity. In total, there are 17 Goals and 169 targets—hardly a constraint!—but enough to begin with.
For instance, SDG 1 and 2 concern eradicating poverty and ending hunger. Large parts of the world are hungry and impoverished. UN statistics indicate that 783 million people globally live on less than $1.90 a day, while 815 million people are undernourished.
With AI, researchers at Stanford University are able to predict poverty by using satellite imagery. They are able to identify the most poverty-stricken regions in any given African nation, facilitating humanitarian work. The AI algorithm is said to predict poverty with an accuracy of between 88-99%.
From an agricultural standpoint, we know that in 20 years, Earth will need to feed 9.6 billion people. At present, mother nature's plant yields are already not increasing fast enough. Experts at Carnegie Mellon University are using AI, machine learning, robots and drones to solve the emerging global food crisis through what's called "precision farming".
In India, dairy cows are being fitted with RFID tags that transmit data to the cloud, which is then AI-analysed and sent as SMS alerts to farmers on what attention the cows need – dietary, health, etc. Some call it "Cows to Cloud", adding a new meaning to the cow who jumped over the moon. Whatever it's termed, it protects and enhances the livelihood of farmers.
Here, AI is being used for all sorts of things, from detecting early-warning medical conditions to smart medical devices and imagining. For example, healthcare company Enlitic created an AI that is able to detect and prevent cancer better than humans. Apple's latest iWatch can monitor heart-rates, detect abnormalities and contact emergency services.
Closer to home, Malaysia's own Dr Dhesi Raja, developed AIME (Artificial Intelligence in Medical Epidemiology), an award winning system that's capable of detecting dengue outbreaks up to three months early. AIME was used at the 2016 Rio Olympics and is being tested in Penang.
AI Has Its Challenges Too
Beyond the positives, AI does have its challenges. An infamous cautionary tale involves Google Flu Trends. Once hailed as a solution to flu forecasting, it got it oh-so-wrong in 2013 by being 140% off forecast.
AI has also been accused of sexism. Not only are AI assistants predominantly female (Siri, Cortana, etc.), there have been reports of image searches for "CEO" returning only pictures of white men and a tendency to display fewer high-paying executive jobs to women.
In the legal system, an AI programme used to profile convicts and assist judges in determining sentencing based on the likeliness of recidivism has also been found to discriminate against minorities.
In 2017, a new sinister phenomenon known as 'Deepfake' emerged which involved using AI to superimpose the faces of celebrities onto porn videos to near perfection. Most sites have banned deepfakes, but iterations of it are out there (Google "Obama Deepfake").
And of course, Elon Musk has warned that AI could one day lead us to World War III while the late Stephen Hawking has said AI could be the worst event in history of our civilisation.
Alas, let us not let the fear of the unknown dampen the positives of AI which I believe outweigh the negatives. With driverless cars, personalised medicine, and robots allowing us to travel more conveniently, live longer, and free our time to focus on more meaningful pursuits (spiritual, family etc), it is up to us to harness AI for the good of humanity.
Technology is on a forward march, whether we like it or not. AI's advantages and biases, are in fact, a reflection of us as a society. If we strive to become better as human beings, let this also be the same of the things we create.