Human agency in the future of AI

THE formal advent of artificial intelligence (AI) is often dated to a 1956 summer workshop at Dartmouth College, a private Ivy League research university in New Hampshire in the United States.

Yet, long before scientists coined the term, the groundwork for computing to perform logical tasks like playing chess was already being laid.

While we often rely on “science fiction” to process these technological leaps, human visions of the future have a habit of realising themselves over time.

Decades before Dartmouth, the concepts of cyborgs and mechanical men were already taking shape in popular culture, most notably in L.F. Baum’s 1900 book, The Wonderful Wizard of Oz.

Unlike the terrifying, dystopian androids of modern cinema, such as The Terminator (1984), the original literary robot, Tik-Tok in The Wonderful Wizard of Oz, was benign.

Lacking human emotion, he was incapable of malice; instead, he served his human companions as an objective advisor and a loyal protector.

Similarly, Tik-Tok’s human friends protected him, providing him with the mechanical tuning required to function.

As such, while human emotions are seen as a differentiator between robots and humans, they can be either benign or malignant – in the very same way that technology can work both ways.

Whether a cooperative dynamic between humans and technology occurs is ultimately dictated by human intent and agency. AI will only cause mass unemployment or displace humanity if people allow it.

Technology is built to assist human productivity, and its long-term economic outcomes need not be negative for three important reasons.

Firstly, society is ageing and the birth rate is declining, which means there are natural forces towards global depopulation over the long run. Global population growth peaked at around 2% in the 1960s. This has dropped below 0.8% today and continues to decline.

The United Nations projects that the global population will peak at approximately 10.3 billion people in the mid-2080s, after which global depopulation will likely occur.

Following this, technology can fill the productivity gap left behind by depopulation.

Secondly, economic output, as defined by the Total Factor Productivity equation, suggests that as rising education and wealth accelerate technological adoption, the relative contribution of raw labour (L) naturally drops because the economy shifts towards capital deepening (K), meaning future output growth (Y) is increasingly driven by technological and structural efficiency gains captured within Total Factor Productivity (A).

However, according to the International Labour Organisation, there has been a mere 1.6% decline in the global labour income share over the past 20 years, suggesting that the replacement of human labour has been progressing at a glacial pace, although there is scope for this to accelerate over time.

Thirdly, history proves that technological changes reconfigure rather than diminish supply chains.

While a new technology automates existing jobs, it also creates new tasks and industries.

During the Industrial Revolution, the automation of textiles displaced traditional weavers, but it rapidly gave rise to a massive network of mechanical engineers and logistics managers.

In other words, AI will not replace workers; rather, workers will be required to use AI to raise productivity or shift to jobs that cannot be replaced by AI.

For example, service businesses with human-to-human interaction remain important, and so are critical thinkers who pose challenging questions to AI.

Education systems will need to shift away from rote learning towards developing critical, inquiring and creative minds. Because technology shifts human labour up the value chain rather than erasing it, the economic pie expands rather than disappears.

Consequently, technology raises productivity and wealth levels, and that wealth will be recycled back into the economy through investments and spending.

Ultimately, technology remains an instrument, not an independent force. Today’s human leaders must consciously design the frameworks and management choices that govern and collaborate with AI.

By choosing to prioritise the augmentation of human capability over simple cost-cutting automation, human agency ensures that the future of AI remains cooperative, productive, and benign.

Technology will not decide its own path, and it is entirely up to us to develop positive or negative scenarios for AI in human society.