We still have time to make choices about AI

THAT technology will replace jobs is a truism that has endured over millennia. Yet, it is clear that the speed at which technology, particularly artificial intelligence (AI), will do that is likely to be exponential from now. About two months ago, a Goldman Sachs report estimated the ballpark to be some 300 million jobs, and earlier this month, the World Economic Forum predicted some 83 million job losses over the next five years.

The headlines have been arresting too: Generative AI tools like ChatGPT burst into the mainstream late last year, but it is already capable of passing exams from US law and business schools (even if it apparently failed Singapore’s Primary School Leaving Examination), giving the rest of us a flavour of how rapid the advancement of AI really is.

This early glimpse into the vast potential of AI forms a parallel with the emergence of social media about two decades ago, The New York Times columnist Thomas Friedman pointed out recently. At a time when social media was sold as a platform that gave people a voice and enabled them to connect with one another, “we did not understand how much social networks would be used to undermine the twin pillars of any free society – truth and trust”, he wrote.

Whether AI will save humanity or deliver our doomsday is a matter of debate, but if we have any interest in choosing the former, now is the time to enact the right policies to guide its development down the path that we want it to go. Indeed, where do we want it to go? It is a question that needs to be answered from the get-go. If AI is to be a beneficial force that liberates us from menial work and brings us to new peaks of social and scientific development, then there is no question that its use must be regulated to ensure that it is used for the common good, and in a way that is ethical and respects data privacy.

The only way to steer AI towards the common good is to ensure that such a powerful tool does not end up in the hands of a minority with questionable ethics, for that is a recipe that will perpetuate – or exacerbate, even – inequality and poverty in ways worse than what the neoliberal order has done. Since generative AI is built using data and content supplied by the public on the Internet, AI should not in fact be owned by any private company.

Daron Acemoglu, an economist from the Massachusetts Institute of Technology, argues in his new book that one key issue with the current development of AI and robots is that it is too focused on mimicking human intelligence, which is a reason why it has ended up displacing jobs. Instead, technology should create “marginal productivity” gains, he says, and this means using AI tools that allow workers to expand the range of services they can provide and elevate their contributions, an endeavour that would, hopefully, lead to higher wages. One example is allowing nurses to use AI tools in the diagnosis of diseases.

SEE ALSO

The artificial intelligence stock-market rush

DECODING ASIA

Navigate Asia in
a new global order

Get the insights delivered to your inbox.

Like all other transformative technologies in the centuries past, AI will herald a new era for humanity, which is why that opportunity to make things right should not be squandered, even amid other pressing and concurrent challenges – the climate crisis, geopolitical tensions and the looming economic slowdown. But this is a choice that will determine whether AI is something to be celebrated or feared.