TAKEAWAYS
Have you seen young salespeople approaching uncles and aunties at neighbourhood MRT stations to promote artificial intelligence (AI) courses? Have you yourself been stopped in your tracks when you were at a shopping mall and asked the question, “Want to take an AI course? Can use SkillsFuture funding.”
Whether it is in the news, on social media or even in the streets, AI seems to have taken the world by storm. Globally, governments and organisations are proactively championing the technology and investing billions of dollars in AI-related projects, with the goal to unlock AI’s massive potential.
AI is a multi-faceted topic, and some dimensions are worth highlighting.
Given the power of AI and what it can accomplish, ambitious organisations are looking to leverage its propositions. This is often driven by the leadership team. Ms Audrey Ong, who drives transformation in her role as Managing Director of Transformation and Innovation at Keppel Ltd, shares a common stumbling block – the disparity between the AI goal set by the leadership team and the operational reality at the working level.
The role of the leadership team mainly focuses on the big picture, including setting the overall direction and long-term strategy of the organisation in response to market trends and opportunities. On the other hand, the utilisation of AI at the ground level is vastly different. It is more concerned with optimising business processes, attaining greater accuracy on financial reporting, achieving faster documentation turnaround, and so on.
“When leaders formulate their ‘North Star’ mandate for AI, it is crucial to take into account the operational reality that exists within the organisation. For instance, whether the North Star is achievable, whether AI will add value to the working team and, most importantly, whether it allows the organisation to become more competitive,” Ms Ong says.
Ultimately, the leadership team will need to demonstrate how AI can substantially benefit the working team from an operational perspective. This will go a long way towards getting the buy-in from the rest of the organisation to drive AI transformation forward as one coherent united team.
Given the lightning pace of AI advancement, organisations are trying to keep up with the developments so as not to fall behind others. However, if they are not careful, they run the risk of AI fatigue creeping into their working teams. AI fatigue is the mental exhaustion and burnout caused by the incessant pressure to learn, use and manage AI tools in daily life and at work. It is inevitable that some people will adapt more quickly or be better at AI, while others just require more time to get familiar with new AI tools.
With AI, the speed of churning out documents for review has become so fast that backlogs are building up. For Ms Yi Suqin, Head of Data and Analytics at SGX Group, the underlying objective should remain the same regardless of whether AI is used or not. “What you are trying to achieve does not change. Even when we use AI, we should not deviate too much from the deadline,” she maintains. To manage AI fatigue, the goalposts and deadlines originally set should not be unduly moved forward or shortened.
In addition, the design of AI solutions can play a part to help working teams to some degree. Since the leaders in the large language model (LLM) space have yet to be determined, organisations should consider making their AI solutions LLM-agnostic instead of strictly hinging on one particular LLM. Working teams can still have a consistent user experience at the front end while organisations get the flexibility of switching LLMs at the back end.
It is a well-known fact that AI can hallucinate to give wrong information. This is a major risk to finance and accountancy where accuracy of numbers is paramount. “The key thing is that users need to be conscious that AI is not always error-free,” states Ms Yi. In essence, the person who uses the AI model is still responsible and accountable for any consequences from using the outputs, even though they are generated by a machine.
When communicating with the teams she works with, Ms Ong likes to put forward an analogy. “Think of AI as a very smart intern fresh out of school. But, an 18-year-old intern still lacks the life experiences to make sound decisions and therefore needs to be trained.”
To guard against over-reliance on AI, Ms Ong offers one useful recommendation: never put a question to AI without having first thought through it yourself. She likens it to sitting for a test, where knowing the model answers in advance will influence how you respond to the questions. Instead, you should arrive at your own answer first, before you input it into different AI models for critique. This, in itself, can be a meaningful learning process. An added advantage is that the brain is engaged in creative thinking and idea generation. The brain behaves similarly to muscles – the more it is used, the stronger it becomes.
One thing is crystal clear – AI is not going anywhere. In the near future, AI could be so ubiquitous that even uncles and aunties will need to have some basic knowledge of AI to carry out their daily activities.
The next time someone approaches you at the mall with the question, “Have you attended any AI course before?”, perhaps you can consider stopping for a chat, to see what is on offer.
Tune in to CA Listen for more information about AI and its impact.
