William Shiyuan Wang

Opinion | Can we curb our destructive behaviour before AI takes control?

If we want to retain control over our lives and the planet, we must rebuild trust and goodwill while restraining our desire for conflict and dominance

Reading Time:4 minutes

Why you can trust SCMP

Participants chat in front of an electronic image of a soldier at a summit titled Responsible AI in the Military Domain, in Seoul on September 10. Humans, not artificial intelligence, should make the key decisions on using nuclear weapons, it was agreed in a non-binding declaration. Photo: AFP

William Shiyuan Wang

Published: 9:30am, 29 Sep 2024

One of the first films I enjoyed in my youth was Disney’s Fantasia, especially the part with Mickey Mouse as the sorcerer’s apprentice. Its theme, based on a poem by Goethe from 1797 – that is, a little knowledge is a dangerous thing – stuck in my mind through the years.

Mickey had learned enough from the sorcerer to command the broom to fetch water for him but did not know how to command it to stop. I am reminded of this story again as the world is being flooded by artificial intelligence (AI).

Human intelligence is based on how we process information via the billions of neurons in our brain. Drawing inspiration from this biological fact, people started building neural networks in machines.

Under the influence of Alan Turing and other visionary scientists, the first computers driven by algorithms were built in mid-20th century. The term “artificial intelligence” was coined in the 1950s. At first our brain had the advantage, with its billions of neurons shaped by evolutionary forces working in parallel.

However, when it comes to sheer information processing, neurons talking to neurons by passing chemicals across synapses is clearly no match against electricity travelling across tiny chips. Furthermore, the chips themselves became ever more abundant and better designed for specialised computations across the decades.

The first milestone suggesting computers had come of age was in 1997, when IBM’s Deep Blue chess-playing computer beat world champion Gary Kasparov. Looking back, it is clear that mastering chess was a small step compared to the giant strides AI has taken since then in a wide variety of fields.