What Could AI Possibly Have to Do with My Dishwasher?
Have you ever wondered what artificial intelligence and your dishwasher might have in common? Or, even stranger: whether how much you use your dishwasher—or how thoroughly you ignore it—might give away something about your relationship with AI? Why you love it, fear it, or simply don't know what to do with it.
Yes, at first it sounds a little absurd. But bear with me please. I have a dishwasher story, and I hope that by the end it will be clear what it has to do with the conversations we are all having at work about AI.
The colleague who waited twenty years
If you have lived alone for years and never much cared for cooking, your dishwasher has probably met the same fate as mine: for nearly twenty years, there it sat in the corner of the kitchen, staring at the ceiling. A household where eating at home means, at most, a breakfast or a salad does not really need one. By the time enough plates have piled up to make running it worthwhile, you would first have to check whether you own that many plates at all.
But as humans, we are subject to change. About a year ago, mostly to give my brain a rest, I resolved to take up cooking—and the more I cooked, the more I liked it. The trouble is that many dishes (particularly Iranian ones) cannot be made for one portion, or even two. So friends, neighbours and assorted acquaintances began appearing at my table, and the pile of dirty plates grew exponentially. Under those circumstances, could I go on ignoring one of the oldest members of my kitchen? That dear dishwasher had sat quietly in her corner all those years, watching me. Never complaining, patiently waiting for the day destiny would call her name.
We needed no introductory meeting. No contract, no agreement on working hours, none of those formalities. Our collaboration simply started by dividing the work: the creative and delicious part for me, the rather less pleasant part for her. And what a lovely colleague she turned out to be, my dear Boschi. Today I can hardly imagine life without her. (Sadly I don’t know for sure if the feelings are mutual.)
So why am I telling you this story of successful transformation? Because something very similar has grown up between me and another colleague: Monschi, my AI assistant. (Monshi means secretary in Farsi, and it goes rather nicely with Boschi.) A colleague who a year ago played no part in my life at all, and who today works beside me on a great deal of what I do.
Now imagine: amid all those overwhelming daily tasks, a very gentle and smart colleague steps in and volunteers to take care of all that repetitive work and lifts all that weight from your shoulders, making room for you to focus on more interesting parts of the work. How lovely it could be indeed!
Well, a sense of realism is of course needed. I am not saying your whole job would suddenly become a pleasure. But the chance that a larger share of your time would go to the parts you actually enjoy would be very high. Good news is that we happen to live at a moment when this imaginary scene is no longer quite imaginary.
With all due respect, madam: a dishwasher?
I can already hear my reader murmuring with suspicion. I admit, comparing a dishwasher to artificial intelligence may well sound like an outrageous simplification. Put your favourite container in without checking whether it is dishwasher-safe, and Boschi may hand you back a small piece of warped plastic; AI can make mistakes with far more serious consequences. We all know that, and there is no point pretending the two risks are the same size.
But tell me honestly: how many technologies can you name that carry no danger at all? The more capable they become, the more it matters to understand what we are dealing with.
A Formula 1 under a Paykan driver
Suppose you have spent your entire driving life in a Paykan, that familiar old Iranian car. Now somebody drops the keys to a Formula 1 car into your hand. (Although, strictly speaking, a Formula 1 car has no key at all. It needs an external starter and a small crew of people simply to wake up—which rather makes my point for me.) Would you jump straight in and drive away? I hope you wouldn’t risk your own life, the lives of those around you (the crew, remember?), or even the car itself!
So why do we expect new technology to demand nothing new of us? One explanation could be that this one talks to us in plain sentences, which may create the illusion that anything speaking our language must be simple underneath.
With a Formula 1 car, nobody is fooled. But when we come to AI, we swing between the two ends of a spectrum and stop nowhere in between. At one end: "Look what this thing can do! We won't need doctors, or engineers, or writers, or anyone at all!" At the other, the apocalyptic questions. What if it takes all our jobs? What if our children never learn to think for themselves? What if AI comes to rule us, or destroys the world?
I am not saying these questions should not be asked. But somewhere between the excitement and the dread, could we not leave a little room for simply getting to know the thing—instead of drowning in opinions, some of which have no basis at all?
A road we have walked before
Do you remember how we did research before the internet? (Honestly, I remember very little of it myself. I still count as one of the young ones!) We went to libraries, worked through books and journals, and often had to ask around just to find out where to look. Finding a single answer could become a project in itself. Then the internet arrived, and we typed a few words into a search engine (yes, I mean Google!), opened twenty sites, compared them, and eventually—if we were patient enough—arrived at an answer. Did that stop us from thinking and reasoning? Or did it open up information most of us could never have reached so easily?
Today Monschi can do much of that searching and comparing, summarizing and structuring, and come back with an answer instead of sending me on a tour of twenty browser tabs—several of which were beautifully optimised to sit at the top of the list while containing almost nothing we were looking for. Of course, we did not believe everything we found online. Or at least we should not have. AI brings us the same trouble in a more convincing costume: a wrong answer, beautifully written, delivered with absolute confidence. (Worse still is when Monschi insists she never made that mistake, and you have to be clever enough to prove that she did. Believe me, that one calls for a new skill.) So we ask better questions, go back to the original sources, and keep a healthy dose of suspicion. None of which looks to me like getting stupider.
And we have done all this before. Remember when online shops arrived, and how often we heard that shops and bookshops would close for good, that paper books were finished, perhaps reading itself? Some did close. But others found customers far beyond their own neighbourhood; and a young writer with a brilliant idea and no publisher willing to take a risk on her could suddenly publish electronically and go looking for her readers herself. Then came online banking, the jewel in the crown: no more queues, no more carrying an uncomfortable amount of cash around. Did it introduce new risks? Of course—hacking, fraud, new kinds of theft. But the old days were hardly safe either. Walking out of a bank with a bag full of cash was not risk-free. The way we handle money changed, and we learned new ways of protecting it.
The other side of the coin – actually, inside of it!
So why is it that when we arrive at AI, we talk so much about what we might lose? Some of those warnings deserve our attention. Others are opinions repeated so often that they begin to sound like facts. (And I do sometimes wonder how deeply the people making these predictions really know its underlying technology.)
Instead of always looking at one side of a coin, let’s look inside, where most of the metal is. When we learn to do something new, we pick up new skills—and those skills tend to make us curious about things that had never occurred to us before. New possibilities create new needs; new needs create new demands, which lead to new services, new jobs, sometimes entirely new markets. Not overnight, and not without difficulty. But isn't that part of the story worth a little of our attention and preparation too?
None of which means the risks will disappear. Privacy, costly mistakes, the future of work: all of these need care, and sometimes new rules. But perhaps, alongside asking what AI is going to do to us, we might also ask what we actually want it to do for us. And, more importantly, which parts of our work we would rather keep for ourselves.
Boschi and I have a lovely arrangement. And who knows? One day she might have another colleague—a proper all-purpose kitchen robot. I have even settled on the name already: Ashi. (From "ashpazi", which means cooking in Farsi. Boschi, Monschi, Ashi—our home has rules of its own.) Would I then let Ashi do everything while I lie on the sofa? The honest answer is: I don't know. And I don’t even need to know right now. What’s already certain is that Ashi shall take care of chopping and grating the onions. For the rest we will see.
And one demanding job, it seems, will always stay with us: tasting that delicious food. Poor machines. That is one pleasure they will never get to take off our hands.
Continue learning
Explore course 1 and our educational approach to understanding AI’s opportunities and limits.