Changing the world with AI
Changing the world with AI
For the past two decades, Dr Tan Kar Han has been on the cutting edge of artificial intelligence (AI), developing products and solutions and building world-class R&D teams. An integral part of the Singapore Global Network in the U.S., he is now living in the Bay Area with his family. He shares some thoughts on what Singapore’s AI future may hold.
It has been more than twenty years since Dr Tan Kar Han left Singapore for graduate school in the United States to study artificial intelligence (AI).
“I was an undergrad at NUS working on neural networks for my final year project and I remember a young professor telling me that “AI is dead”.
Undeterred, Kar-Han pressed on, and has since remained at the forefront of AI development. For the past two decades, he’s led teams to invent new applications for AI with the likes of Mitsubishi Electric, Epson, HP and AI chip startup Novumind.
Dr Tan giving a talk at the Capital University of Physical Education and Sports in Beijing as part of a discussion on an AI research program for the new Beijing International Olympic Studies Institute
Between 2017–2019, he returned home to build and lead a new R&D division at Singtel where its sole purpose was to create “world-changing products”.
One of his many projects was to bring the Marina Bay Cruise Centre into the future with a comprehensive suite of products, including using video analytics to monitor queueing times at key checkpoints as well as helping law enforcement identify unauthorised taxi drivers.
Dr Tan with the NCS Robotics Engineering team
Singapore is “hot” for AI
Dr Tan believes Singapore is the furthest it has ever been on its AI journey, and it’s just getting started.
“AI has never been so “hot” in Singapore as it is now. I see a lot of young kids getting excited about machine learning and AI: I was a judge for the junior category of IMDA’s Code XtremeApps and my jaw dropped seeing primary school kids dabbling in machine learning.”
According to Kar-Han, this is in line with Singapore’s push to become a Smart Nation.
“I feel that the Smart Nation initiative of creating a city-scale testbed for experimentation is a great idea, and I was impressed by the desire for many government agencies to adopt new technology solutions and opportunities for small businesses to grow while participating in the grand experiments.”
Dr Tan with SCALE@NTU Corporate Labs (a.k.a Singtel-NTU AI Lab)
What Singapore’s National AI Strategy means for the country
As part of the Smart Nation journey, Singapore recently announced the National AI Strategy, which plans to deepen the use of AI to transform the economy.
“This is a big deal since it brings a fresh focus to AI in Singapore,” shared Dr Tan.
“Singapore has invested in digitalization, cloud, supercomputing infrastructure, and Industry 4.0 technologies. AI is the missing link that enables many new possibilities which motivated the investments in the first place.”
One notable aspect of this is how AI will improve the way Singapore is currently running.
“The National AI Strategy outlines several major infrastructure projects where AI can be incorporated to enhance these city-scale systems. These projects can accelerate the evolution of these systems and serve as leading edge examples of significant AI applications.”
Dr Tan delivering his keynote message at NCS TechConnect 2018.
The challenges of AI
Despite the advantages of AI, Dr Tan does acknowledge that there are still very important considerations, especially in AI’s current stage, where bias in AI is a valid concern.
“We have already seen examples of AI and algorithmic systems exhibiting behaviours that are no longer acceptable for humans, such as discrimination based on race or gender. Without paying special attention, data sets and the resulting AI risk becoming biased,” warned Dr Tan.
An AI of tomorrow
While there is no way of truly knowing what AI is capable of in the years to come, there are three AI trends that Dr Tan is particularly excited about today.
Self-supervised learning – truly adaptive systems that learn and improve over time on their own, breaking the bottleneck of requiring manual data labelling.
Deep generative neural nets – new methods of synthesizing realistic or stylized content with the possibility of shaping the output in ways not seen before.
Low-power hardware – allowing neural networks to be incorporated into a wider range of devices, making them still capable of intelligent behavior while untethered from the cloud.
Dr Tan and NCS team presenting home-brewed concierge robot “Kompass”, at the official opening of Changi General Hospital’s new medical centre
Dr Tan is naturally also optimistic about the future of AI in Singapore, especially for those planning to make it a career.
“The new national focus on AI should create a lot of job opportunities. I’d say that demand for AI, machine learning, and analytics skills has already outstripped supply in Singapore and most places.”
For those planning to join the AI industry, Dr Tan’s advice is to just get started.
“Virtually everything needed to learn about AI is available for free: open access papers, open source code, books, video lectures, and tutorials online that make it easy to get started so find something that interests you and get your feet wet. The field is still advancing at a breathtaking pace, so you have to be experimenting all the time.”
“I feel like a kid in a candy store – it’s a fantastic time to be a scientist and engineer.”
If you are interested in connecting with Dr Tan, please drop us an email today.