The attack of artificial intelligence

May 27 2019 Jutta Rahenbrock

“Hi, I’m Nobby. How can I help you?” “So, how do we get to the restaurant? I want to have breakfast” “The elevators to the rooms are in 100 meters on the right-hand side. We’re currently offering two days for the price of three. Just € 139 if you decide right now …”

Artificial intelligence is on the advance

Artificial intelligence! Ten years ago, let’s say, it still sounded like something from Frankenstein. Today, up and down the country we are dreaming of the infinite expanses that AI will open up for us. To me, this cocktail tastes like a shot of limitlessness and a hefty dose of “Anything is possible”. We’re seeing the first critical voices, who want to see regulations for the use of AI and clearer monitoring.

Still more style than substance

The reality looks a bit different. In two ways – firstly because not everything that’s marketed under the AI label is actually AI. Frequently, what is promoted as AI is more of a structured interview that is conventionally programmed. Many operators of talking robots, for example, admit this. An update involves additional lines of program. But there is rarely any genuine self-learning, analysis and reaction.

Secondly, what AI can do today is a long way from what we expect of it. Nobby is not the only example. Have you ever shopped on Amazon? I once ordered a red dress from them. No sooner had I unpacked, tried on, and paid for the dress, thanks to the online advertising networks I was bombarded with recommendations for red dresses. What happened with me is no exception. Trainers, shoelaces, dishwashers – “artificial intelligence” seems to view us as repeat offenders of the worst kind. Compared to this, the economic picture of homo oeconomicus is innocuous – we are apparently homo repetitens, hamsters with no horizon beyond the hamster wheel. We are red dress fetishists, dishwasher collectors, trainer hoarders – people preparing for a zombie apocalypse in which only shoelaces will survive as the ultimate form of currency – good for the person who stocks up.

How hard can it be to understand that someone who buys a particular kind of item has already got what they want? If anything, offer me a pair of red shoes to go with it. Or underwear. Or a hat. Or at least a green dress – that would make the label “intelligent” more appropriate.

On the way to AI collaboration?

In the retail sector at least, AI is developing into the public face of companies. In the past we had some magic moments, when the hotline operator quickly sent us a spare part or resolved a problem in seconds by remote servicing. But equally, in the past we’ve been wound up about less helpful or downright incompetent service staff. Today we can laugh at clumsy AI – which they tell us is taking away jobs and raising service to an unprecedented level. But there will come a time when it will annoy us to be treated as manic cash cows who need a second, third and fourth children’s swing for the garden. We will start to value real stores in actual shopping areas again. And we’ll want to buy from real people who understand us. I find it hard to believe that this is the purpose of AI.

Of course, it’s not really fair to make jokes about the whole range of AI. It has admittedly already overcome numerous obstacles – image recognition, voice recognition, voice output for example – and these all work brilliantly. And I’m sure that in the coming years we will get new impetus from AI in the field of collaboration. I only hope that we’ll be dealing with something with a more sophisticated nature. Otherwise the dream of making our lives easier will soon turn into a nightmare. And health insurers will have to finance a whole new disease – depressive AI mobbing.

About the author

Jutta Rahenbrock

Manager Product Marketing, T-Systems International

Show profile and articles