AI should also be human-centred
Shin Min Daily News, AI，也要以人为本
(Translation)
Some time ago, an uncle from my neighbourhood excitedly took out his phone at a coffee shop to show me a new health app. His children had helped him download it. It was said to be very advanced, using artificial intelligence (AI) to track physical activity, monitor sleep, and even provide health advice.
“This is very good, very smart,” he said. But just as he finished, he smiled and added, “It’s just that I don’t know how to use many of the functions.”
There were all kinds of buttons, charts, and numbers inside the app, and some of the text was very small. He shook his head gently and put his phone away. “Too complicated,” he said. “Maybe only young people can understand it.”
This kind of scene is quite familiar to many people. Today, AI is developing rapidly. It can help doctors analyse medical images, recommend products online, translate languages, and even generate text and images. However, no matter how advanced technology becomes, it ultimately must enter everyday life and be used by people. Often, the problem lies not in the technology itself, but in the failure to truly understand the people who use it.
This is why human-centred design is especially important. Scholar Don Norman has pointed out that many technological products are difficult to use because designers focus too much on the system itself while neglecting the user. In other words, technology should adapt to people, not the other way around.
Beyond design, the social sciences and humanities are equally important in technological development. Fields such as psychology, philosophy, ethics, and behavioural science study how people think, how they make decisions, and how technology should be used. Research shows that when people make decisions, they are often influenced by habits, emotions, and convenience, rather than purely by rationality.
As AI increasingly shapes decisions in real life, we must also pay closer attention to the problems these systems may bring. When algorithms learn from past data, they can sometimes replicate existing biases. AI may also produce answers that are not entirely accurate, a phenomenon known as “AI hallucination.”
Technical knowledge remains important, but the ability to understand people, communicate, create, and solve problems is becoming ever more critical. That is why universities like the Singapore University of Technology and Design (SUTD) have begun integrating AI, design, and the social sciences into its curricula. These programmes not only train students to develop AI technologies but also help them understand how people use technology in real life, while encouraging them to think about human behaviour, ethics, and user experience.
Thinking back to that uncle in the coffee shop, if more attention had been paid to user experience during the design process, perhaps he would not have abandoned the app after just a few minutes but instead used it daily and benefited from it.
As AI becomes increasingly integrated into our lives, the most successful technologies of the future may not be those with the most features or the greatest complexity. Truly successful technology is that which understands people best, and is created for people.
When technology truly understands human needs, it not only makes life more convenient, but also better.
