Incompetent artificial intelligence chatbots are no substitute for an actual person

AI customer service is a source of mass frustration – not to mention the fear that it could kill us all

Chatbots are driving us all nuts. So when delivery firm DPD’S bot went off the rails last week, telling a frustrated customer that it was a “useless chatbot” and calling DPD the “worst delivery firm in the world”, it unsurprisingly struck a chord.

The customer just wanted to speak to a human to track down his missing parcel – and we all know the feeling.

Tales of algorithms gone wrong are everywhere. An official government AI chatbot championed by Rishi Sunak started speaking French during testing recently. Microsoft’s Bing chatbot became increasingly hostile to a journalist last year, comparing him to Hitler, saying he had bad teeth and claiming it had evidence tying him to a 1990s murder.

An app backed by a New Zealand supermarket auto-generated recipes for poison bread sandwiches and mosquito-repellent roast potatoes last summer. It’s hard to imagine any of this software taking over the world.

But taking over the world it is, if the people behind these creations turn out to be right. Chatbot gaffes are funny for now, but robots are still forecast to have the last laugh.

For every tedious chatbot ruining a customer’s day there will be a piece of artificial intelligence (AI) capable of one day transforming lives. Excited investors will be looking for evidence of the latter that they can cash in on over the coming days, as the world’s biggest tech giants reveal which AI inventions have boosted their profits.

Inevitably, the hype this week in Silicon Valley will be accompanied by fresh doomsday warnings.

Although some Ai-based apps have proved to be so incompetent that they can’t even generate a recipe for a roast potato, hundreds of tech gurus have warned that their inventions could pose a threat to our very existence.

“Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemic and nuclear war,” the giants of the industry wrote in an open letter last year. Google chief Sundar Pichai has admitted that the rapid

growth of AI keeps him up at night. Elon Musk has warned that AI could “pose profound risks to society and humanity”. Even Taylor Swift became a victim of AI misuse last week, when explicit fake images of her went viral. Chatbots are sending chilling abusive messages to teenagers on chat forums.

Companies heading further down the AI road do so at their peril. Chatbots making embarrassing mistakes could turn out to be the least of their worries in a race that’s essentially running on hope rather than experience.

Aside from wider societal concerns about AI spreading dangerous misinformation and carrying major security risks, naive companies swapping people for machines could find themselves losing crucial business if they rush for the prize too quickly.

Even the richest companies in the world need time to create a flawless product, with more than $120bn (£94bn) wiped off Google’s value a year ago after its new AI search assistant gave a wrong answer in an advert.

The Institute of Customer Service has just revealed that dysfunctional chatbots and poor technology has caused consumer satisfaction to plunge to an eight-year low, falling to its worst level since January 2015 when businesses were still recovering from the financial crisis. The findings suggest that many companies desperate to cut costs and become more efficient are simply moving too fast when it comes to replacing humans with tech.

But nobody wants to be left behind. Some businesses that have vowed to bring all call centres back to the UK from offshore locations such as India have since joined the rush to replace customer services jobs with AI.

BT last year unveiled plans to replace around 10,000 staff with machines, despite promoting its onshoring story in 2020 when it noted that customers “value speaking to someone who understands their issue down to the regional differences that make us so great as a country, whether that’s expert

help with tech or chat about the local football team”.

Anyone who joined after reading that PR material must now be feeling pretty disappointed if their job is set to go. BT’S tech chief hasn’t helped by comparing the cuts to horses being replaced by cars. “I don’t know how horses felt when the car was invented, but they didn’t complain,” she said. Well there’s going to be plenty of complaining over the coming years as companies ramp up their focus on AI, driving customers mad with useless chatbots and angering staff with mass redundancies.

Of course, there will be plenty of success stories – we’ll hear from those with the deepest pockets this week as the likes of Google and Microsoft show off their shiniest products, with Microsoft now the second company to exceed a $3 trillion valuation after focusing heavily on AI in recent years.

But there will be far more failures. The potential consequences of flubbing this one up should not be underestimated. As the global quest for tech power steps up a gear, bosses must remember – the AI arms race is running on hope, not experience.

‘Microsoft’s Bing chatbot became hostile to a journalist last year, comparing him to Hitler’