When one is to define what Artificial Intelligence (AI), they may be quick to say that it is the technology being used by Ironman in the Marvel movies. While that is AI, we are far from achieving that. So, what is AI and how are we using AI in the world today?
OpenGov had the honour of interviewing Prof. Chen Tsuhan, Deputy President (Research and Technology) and Distinguished Professor at National University of Singapore (NUS). He also serves as the Chief Scientist of AI Singapore, a national programme in artificial intelligence.
Prof. Chen is a renowned expert in pattern recognition, computer vision and machine learning. He joined NUS from Cornell University, Ithaca, New York, United States, where he had been the David E. Burr Professor of Engineering since 2009.
He gave OpenGov insights into the place of AI in the world today and of the innovations and uses surrounding it.
AI in Singapore
Prof. Chen said that Singapore is already using AI but “it is not enough”. While AI technology is being used, it is still scratching the surface.
He said that the medical, legal and financial institutions are the big users of AI and that they will greatly benefit from using it.
He gave examples of how AI can be used to look at financial statements and provide financial advice. In the legal sector, lawyers can use AI to compare various cases and provide advice. The medical field could potentially explore ways in which radiologists can be assisted by AI so that their diagnosis will be more efficient and more precise.
Prof. Chen said that speech recognition is used by doctors for recording down details into medical records. But AI can do much more than transcribing speech into medical records. He said that the future of AI in the medical field should be such that the machine will assist the doctor to provide diagnosis or advice to patients.
Advancements done to machine-learning will allow machines to work more efficiently than humans, mostly in low-level tasks.
Adoption of AI
Prof. Chen shared that the term AI and the field of AI research were created more than 50 years ago. AI was initially a new technology which created a wave of excitement amongst people. It then went through “AI winter”, where its usage was low as it was a hard task. Today, it has grown to perform better than humans and new AI techniques continue to grow.
He said that one of the reasons as to why there still is some hesitance towards the adoption of AI is because humans are now getting worried that they will be replaced by machines. Prof. Chen said that humans need to understand better how AI can better help them. AI technology needs to be improved and be embraced and used even more.
He stressed that AI can and must be used in the right way such as for performing low-level tasks. This allows businesses/doctors to spend more time with their customers/patients this way. AI helps tasks be more personal and make us more human.
On the shift of mindset that people should adopt, Prof. Chen gave this example, “People were once worried about cars at first but now they are easily embraced.” He said that AI can create jobs instead of taking them away as people are needed to understand how AI can be applied better, the impact of it, and of how it is to be utilised.
This will also create jobs in the legal sector where a new study of law/new lawyers are needed to govern AI.
AI and data analytics/literacy
Prof. Chen said that companies such as Google, Facebook and Amazon have been leveraging on AI to study their data and boost their business profits.
AI is used to study their data and identify customer trends and behaviours to predict and promote products to their customers.
AI can be used in capacity building and awareness creation of employees in an organisation. Prof. Chen said that AI Singapore has programmes to train people to be more data literate. “AI for everyone” is one such programme which is a 3-hour course which gives insights into how AI technologies and applications can be used to create the best opportunities.
Prof. Chen stressed that data literacy is important as it will make people more aware of how AI can make us more human and mitigate the challenges in low-level tasks.
AI- related innovations by AI Singapore
Prof. Chen shared about the various ways in which AI technology is being employed in Singapore.
AI Singapore Grand Challenge- it is the promotion of AI technology for identifying and monitoring high blood pressure, high blood sugar, and high blood cholesterol. Medical advice can be provided by machines based on the results.
Grab-NUS AI lab is one such collaboration where AI Singapore implements AI technology on data provided by Grab and conduct data analytics to predict how many customers are going to call for the next ride. They study customer routines of booking timings and locations travelled to. With this, traffic time can be cut down by 30% just by using data analytics.
AI, when used properly, can help systems to be more secure. Prof Chen said that a lot of data breaches are by human actors. With that in mind, AI can be used to analyse people’s behaviours for anticipating potential threats. It can also be used for the analysis of traffic patterns on servers.
Studying of employee’s behaviours such as the sites they have visited can be done. The machine can also be fed with information on previous breaches for it to learn and be able to better identify the next time.
AI can identify novel or “fishy” entrances into systems and detect the weaknesses in them.
He said that white hackers can be used to test how strong a system.
Government agencies and AI
Prof. Chen said that AI Singapore is the linkage for government agencies to employ AI technology. AI Singapore is funded by the Info-communications Media Development Authority of Singapore (IMDA) and National Research Foundation Singapore (NRF).
He shared that AI Singapore is working on a project for public housing which uses AI algorithms for studying lift operations and predicting if they will be functioning properly. This is called predictive maintenance. This will allow for the better allocation of which lifts need constant checks and maintenance based on the AI algorithms which track the usage of the lifts.
Prof. Chen also shared on current research work which is focused on studying how speech recognition can be used to identify and detect emotions. He said that Singapore is the best location for countries to test if their speech and facial recognition systems are effective as Singapore is very diverse in its language and pronunciations.
“AI Singapore carries the responsibility for AI to work better and we encourage organisations to use their AI technology in Singapore to test and determine if their system works well,” he said.
Prof. Chen said the diversity of data is more important than the amount of data for AI to work best.
Training a machine with the same data will not make it better but diverse data will. He acknowledged that getting data for testing of AI technology is challenging due to security measures such as PDPA for protecting data.
He shared about differential privacy, a method where AI can study your data without knowing who you are. He said data can be encrypted and provided to the AI machine for analysis and only be decrypted by the data source. This is also known as homomorphic computing.
“In the AI world, we don’t ask how much data do you have? We ask how diverse is your data?” said Prof. Chen. He concluded with the reiteration that AI should be embraced and adopted into practices for producing the best outcomes.