Ying Shaowei recalls the first time he tried ChatGPT when it came into public consciousness in November 2022. I’m speaking to Ying at the newly renovated NCS Hub at Ang Mo Kio, where he holds the position of Chief Scientist.
“It was like an iPhone moment — or driving a Tesla for the first time,” he tells me with a slight grin. “The programme blew me away with how human-like its responses were. It’s actually more than human-like.”
For the uninitiated, ChatGPT is an AI-powered chatbot developed by San Francisco-based AI research lab, OpenAI, based on the GPT-3.5 architecture. The chatbot engages in natural language conversations with humans and has been trained on a vast corpus of text to generate coherent and meaningful responses to various questions and topics.
Today, it’s nearly impossible to talk about AI advancements without referencing the AI chatbot and how it has changed (and upended) our lives.
And while the world is divided squarely between fearing and revering ChatGPT, Ying Shaowei, though impressed by its abilities, feels that it may be, “you know, too perfect. Superhuman even.”
Related: Find out how corporations are tapping artificial intelligence for the future
AI can benefit humanity
Ying is primed to speak authoritatively about AI. He is, after all, currently the senior partner and Chief Scientist of NCS.
Singapore’s NCS is a leading technology services firm operating across Asia Pacific, providing services and solutions in, amongst many things, consulting, digital, technology, and cybersecurity.
Before becoming NCS’ chief scientist, the Chinese High School and National Junior College alumnus was the founder and CEO of DataSpark, a big data analytics and insights company.
Ying Shaowei’s long list of appointments pre-DataSpark included being Singtel’s head of strategy and analytics at Singtel Group Digital Life, an associate principal at McKinsey, and deputy director at the Singapore Ministry of Trade and Industry.
One of the many initiatives NCS initiated was a collaboration with Singapore Prison Services to harness digital tech to improve the digital literacy of rehabilitees. This is in hopes of breaking the cycle of re-offending and easing former inmates back into society.
Ying predicts that it is only a matter of time before AI plays similar roles in uplifting the community and benefiting humanity.
“The use of intelligent technology is becoming more ubiquitous in our work and personal lives, driven by wider data availability, the increasing availability of compute power, and more accessible AI algorithms.”
A fear of progress
Perhaps this ubiquity is what inspired the letter written by the Future of Life Institute to Pause Giant AI Experiments, signed by such digital luminaries of thought as Yoshua Bengio, a professor at the University of Montreal and widely considered a pioneer of modern AI, historian Yuval Noah Harari, Apple’s co-founder Steve Wozniak, and Skype co-founder Jaan Tallinn.
“The fear of AI stems from concerns about how it may develop and become independent,” Ying explains when I ask why he thinks humans fear AI so much.
The motivation behind the letter, he posits, could be to give humanity time to catch up to OpenAI or to express genuine concern about the potential risks associated with AI. “AI is like a nuclear weapon, and in the wrong hands, it can lead to unintended outcomes,” Ying warns.
The letter called on ‘all AI labs to immediately pause for at least six months the training of AI systems more powerful than GPT-4’.
For Ying, the concern about how technological advances may affect humanity is nothing nascent. “We can look at every industrial revolution and see that there have always been concerns about how it will affect the human population,” he reminds me. “It’s not that different in this case.”
Ying Shaowei on ethics
Still, Ying admits that the pace at which AI is developing is a serious cause for concern. “It’s happening so much faster than before — the COVID-19 pandemic accelerated it.”
A World Economic Forum piece penned by Kerri Reynolds in February 2021 reported that the pandemic compelled organisations to invest more heavily in automation through AI to expedite remote working, enhance the user and customer experience, and decrease costs.
“Today, in a post-COVID world, generative AI (a technology that can generate realistic images, audio, and text) is attracting a lot of concern because of its rapid pace. It’s moving too fast for most people to feel comfortable with,” he explains.
There’s also the issue of the ethical implications of AI technology. On this topic, Ying Shaowei is unwavering in his certainty. “AI cannot be taught ethics in the same way as humans,” he emphasises. In fact, Ying says that AI can and should be considered an entirely new species, given that it lacks the same knowledge model of the world that humans possess.
“Large language models are simply prediction engines based on input data,” he explains. “In contrast, humans don’t require as much data to learn and understand concepts such as distinguishing a cat from a dog.”
The point here, Ying emphasises, is that we need to recognise that AI has a different learning process than humans and sets explicit boundaries and governance to ensure ethical use.
“Southeast Asia will probably always be playing catch-up to the western world due to the rapid advances in AI. It’s already happening.”
Ying Shaowei on a new digital divide
Related: AI honcho Dr Ayesha Khanna on how tech has changed her mindset
A new unavoidable divide
Beyond the breathless pace of change, it is worth examining how AI further deepens the rift between the mature and homogenous markets of Europe and North America and the developing region of Southeast Asia.
At the heart of the problem: language. “ChatGPT, a large language model, is trained in English — but less than five per cent of the world’s population uses English as their first language.”
And while efforts are underway to indigenise large language models — such as building localised foundation models — Ying reminds me that not all translation engines effectively capture cultural nuances, phrases, or words unique to a particular language. This, he worries, would create a new kind of digital divide.
“Southeast Asia will probably always be playing catch-up to the Western world due to the rapid advances in AI. It’s already happening,” Ying tells me with nary a hesitation.
“It’s no longer a digital divide — it’s an AI divide. Countries with access to advanced chips and processes will continue to forge ahead and build even bigger leaps over the rest of the world.”
A potential equaliser
The future, if Ying is to be believed, looks grim. Still, by his own admission, Ying wrote in a Business Times piece that ‘AI is underutilised and brimming with untapped potential, especially when it’s projected to uplift 10 to 18 per cent of the region’s gross domestic product by 2030’.
“That’s an equivalent of US$1 trillion (S$1.3 trillion) in economic value,” he adds. Surely that’s motivation enough to swiftly narrow the divide. Ying is hopeful, but he tells me it’s not that simple. “There are two opposing forces at play here,” he elaborates.
“On one hand, people with access to advanced AI technologies will continue to advance their capabilities. Conversely, the open-source movement is making smaller versions of language models available to communities with fewer resources.”
And while these pruned-down models may not be as powerful as their larger counterparts, Ying is quick to elaborate that they can level the playing field and provide access to AI for those with less.
Still, despite all the potential pitfalls, inequality, and growing divide that AI may potentially bring, Ying has faith in the human spirit of innovation. It is, in fact, the one thing that gives him hope.
“We have innovated ourselves out of the pandemic, and we are now facing an unprecedented rate of innovation that potentially would disrupt us,” Ying reminds me. He is unshakeable in his belief that we humans can further innovate to save ourselves from the negative consequences of our past innovations.
“Don’t discount the ability of the human mind and the human spirit to adapt — that’s what defines us as a human race.”
