The troubling side of the AI revolution

It was 1985, and Rupert Murdoch was about to spend upwards of a billion dollars buying another Australian newspaper group. In the newsroom of The Melbourne Herald, bulky beige boxes had just been installed on every desk, replacing the typewriters and stacks of flimsy rectangular copy-paper that until then had been our means to create and deliver the news to our feared overlords (subeditors).

The older and more senior journalists eyed those grey screens, with their blinking green cursors, with a mix of suspicion and horror – but I was 19, fresh to the industry, and keen as mustard to embrace this new marvel: the word processor.

The idea that a computer might be used for one single purpose – writing a story then sending it electronically to another room in the same building – seems quaint now, but this was nearly 40 years ago, when technology moved fairly slowly – one innovation, one smallish step at a time.

It was easier back then to assimilate a fairly simple leap forward into your daily work routine.

Fast-forward to the early 2000s, and I clearly remember reading the following words in the intro to a TVNZ story about smartphones: ‘‘but will we really want to carry a computer around in our pockets?’’

In recent years my kids have found that ancient clip on YouTube and still rib me mercilessly about it. ‘‘How could you have been so shortsighted, Mum?’’

It’s hard to recognise the sweep of technological advancement in the moment – or at least it has been, until now. Now, a normal citizen must scramble to keep up. It’s been comfortable to think we can keep a desultory eye on the development of artificial intelligence and its implications, but the darn thing is moving so quickly.

Newsfeeds are packed with stories about the latest iteration of ChatGPT and its generative AI cousins – the implications for learning and job security are mindblowing.

Even if you don’t consider yourself at the front line of technological uptake, you’d better be nimble; artificial intelligence is about to become part of your daily life. This can be scary if, like me, you clearly remember a life with paper maps and rotary-dial telephones rather than Google Maps and social media.

In 2020, the World Economic Forum’s Future of Jobs report employed some neat understatement when it noted ‘‘the workforce is automating faster than expected’’. On the chopping board were 87 million jobs worldwide by 2025, it predicted – and on the other side of the ledger, 97 million new jobs created by this robot revolution.

Two months ago, a Goldman Sachs report put the likely number of jobs lost at 300 million. Even the forecasting of wholesale change is failing to keep up.

It’s hard to recognise the sweep of technological advancement in the moment – or at least it has been, until now. Now, a normal citizen must scramble to keep up.

If you’re in sales, customer service, bookkeeping, truck, courier or taxi driving, proofreading, retail – the news is not great. If you’re a lawyer, a CEO, an educator, PR or events manager, or (clever you!) in computer science, you can apparently rest easy, at least for now.

Coming into the workforce, or considering a career change? Data science, data labelling, computer engineering and machine learning is the canny bet.

That World Economic Forum report suggests a full 50% of employees expecting to stay in their jobs for the next five years will have to retrain in their core skills.

Just staying competitive will require organisation-wide change.

Unfortunately, adapting to what we’re given may be the easy part. Ensuring those tools are fair, unbiased and properly regulated – that’s the part we’re failing at already.

A new book by my former colleague Tracy Spicer pokes into the dark corners of the AI revolution, and finds bias (notably, misogyny and racism) is becoming baked in.

Some of the anecdotes in the book are hilarious, like the woman who asked an AI bot what to use when her favourite brand of tampons sold out, and was given ‘white mushrooms’ as its top pick.

Some are just plain awful – Nigerian tech worker Chukwuemeka Afigbo’s 2017 video of an automatic soap dispenser in a Marriott hotel that worked for a white colleague but not for him, for example.

As Spicer points out, AI innovation is being hustled along by a small group of white men in Silicon Valley: ‘‘They’re creating a perfect world in which technology works really well – for them.’’

Some say the book falls short when it comes to solutions; too focused on the individual rather than the wider need to dismantle power-norms and the patriarchy.

But as we stand dithering on the brink, frozen by lack of understanding of this fourth technological revolution, it is, at least, a useful place to start.