| By Distinguished Professor Chen Tsuhan |
We are living in an era of extraordinary technology advances. Dubbed the fourth industrial revolution (Industry 4.0), it represents a fundamental change in the way we live, work and relate to one another. Like the gruelling employment pattern changes witnessed in the previous industrial revolutions, our generation will go through a similar transition. Even as new technologies become the norm, people remain anxious about the impact they will have on their lives, and livelihood. In some cases, these anxieties are well founded. However, a more important question we should be asking is how do we reap the benefits and rewards of AI without compromising our comfort, privacy, security and rights?
Lessons must be learnt from past industrial revolutions to ensure the power of these smart technologies is neither misused nor exploited to the detriment of society. We have to ‘tame’ AI so that it remains a tool that works for us.
AI as a strategy for growth
In Singapore, advancing artificial intelligence (AI) is particularly relevant. With the city state’s economic future dependent on the continued success of its manufacturing, export, and wealth management sectors, the government has sensibly advanced the Smart Nation agenda — to streamline its manufacturing sector and pursue data-driven goals where smart technologies improve every facet of life for Singaporeans, including health care, education and commercial enterprise.
At face value, such a pursuit could be successfully achieved by providing computer scientists and engineers with access to the infrastructure and data resources they need to develop and test the necessary algorithms. At some level, this approach was employed in the development of the global social networks that exist today. However, the flaws in this approach are already becoming evident as social networks struggle to control the spread of false information and malicious content.
Instead, we need solutions that are found in the interdisciplinary fusion of ideas from different fields. Universities are ideally placed to facilitate such ‘purpose-driven’ research, not only due to the depth of expertise across social sciences, computing sciences and engineering, but also because the research need not target short-term payoffs.
Such an approach will enable policymakers to consider the ultimate goal, as well as the risks at play, and design AI technologies that both benefit society, and that society can embrace. At the same time, if individuals are to benefit from the advances Industry 4.0 will bring, they must understand and embrace the changes that are taking place.
The digital imperative — a double-edged sword?
A common challenge faced by many companies wishing to implement AI-based technologies is a reluctance from their employees, who ultimately fear the loss of their jobs. However, according to the World Economic Forum, while 75 million jobs will vanish by 2022, 133 million new positions will replace these.
While most of the tasks currently performed by AI are those that are routine or rudimentary, the capabilities of AI are no doubt rapidly expanding. Across industries, automated systems will ultimately carry out many of the ‘day-to-day’ tasks in the workplace, freeing up the time and effort of employees so that they can focus on more important tasks, such as decision-making, negotiation and offering personalised customer service.
Hence, people ought to take a proactive approach towards their upskilling, and employers must prepare for these changes by encouraging employees to be trained in the necessary areas, allowing them to work with, rather than against, the changes.
People will remain essential in the successful implementation of any AI system. Just recently, the video streaming service YouTube rolled out an algorithm that generates ‘information panels’ alongside what it considers controversial content. However, within a few months of release, these panels began incorrectly linking an accidental fire to past terror incidents. YouTube was forced to admit their systems had made the wrong call, and humans had to intervene.
Evidently, at this stage, humans must continue to work with AI-technologies to vet or guide AI-based decisions, and ensure legal and social standards are met.
Retrain, not replace
For the same reasons that universities play a crucial role in keeping large corporations and governments that deploy AI-based technologies accountable, universities also play a crucial role in educating society, and ensuring both companies and individuals are ‘AI-ready’.
At NUS, we have placed a large emphasis on conducting our smart nation research with a holistic approach. The opening of innovation 4.0 (i4.0) — home of the NUS Smart Nation Research Cluster — in September 2018, demonstrates our commitment towards Singapore’s Smart Nation Initiative. i4.0 brings together researchers, faculty members, students and industry who are involved in data science, analytics and optimisation, AI, as well as cybersecurity, under one roof.
Meanwhile, AI Singapore, a national programme that fosters AI research, conducts courses to improve people’s knowledge and skills in AI. The three initiatives developed by AI Singapore are AI Apprenticeship Programme (AIAP), AI for Industry (AI4I) and AI for Everyone (AI4E), which will equip 12,000 Singaporeans over the next three years with various levels of AI knowledge.
Where do we go from here?
The overarching takeaway is that AI is a natural evolution of information technologies, and we are just witnessing another chapter in the technological revolution. AI is just another tool, even if a highly impressive one. As with the previous three industrial revolutions, while fear and uncertainties were abounding at the beginning, people eventually came to terms with working alongside machines to create the most positive outcome — and Industry 4.0 should be no different. It will be more a matter of cooperation than replacement, with the possibilities of many new doors opening.
About the author
Distinguished Professor Chen Tsuhan is a renowned expert in pattern recognition, computer vision, and machine learning. He is the NUS Deputy President (Research and Technology). He is also the Chief Scientist of AI Singapore which is a national programme in artificial intelligence hosted at NUS.