All Articles | OM Consulting

Sketch-style illustration: A woman on the phone gives a thumbs-up and says 'Aye!' as a sign of agreement. On the right, a technically elaborate voicebot with a headset thinks in a thought bubble: 'I? What about you? ...' The bot mishears the Scottish 'Aye' as the personal pronoun 'I' and gets stuck in confusion.

The bot that helps no one: Why chatbots and voicebots fail

You type your question into the small chat window on a website. Perhaps on your bank’s site, perhaps on a mail-order retailer’s site, perhaps on your insurance company’s site. You phrase the question clearly and politely, because you’ve got used to ChatGPT. Shortly afterwards, the reply arrives. It has nothing to do with your question. You try again, keeping it simpler and using different words. The next reply is a boilerplate text from the FAQs. You give up and dial the customer service number. This experience has been documented in studies. A 2024 UK survey by Cavell Group shows that around half of UK consumers now prefer human interaction as the quickest way to resolve customer service issues, while 35 per cent say that automated systems and chatbots fail to deliver satisfactory service. And 45 per cent have tolerated a product problem rather than deal with customer service.

Since ChatGPT, customers have come to expect a contact person who understands their question, keeps the context in mind and, when in doubt, asks for clarification rather than making assumptions. What they get is a FAQ machine in new packaging. Disappointment is inevitable.

Minimalist line drawing: a conductor with his eyes closed raises his baton; in front of him, two musicians are playing the violin at their music stands; a sheet of music bearing the name Ludwig van Beethoven lies on one of the stands.

The myth of the ‘man-month’: more people, later completed

I recently came across a problem that’s often used as a textbook question. An orchestra of 120 musicians takes 40 minutes to perform Beethoven’s Ninth. How long do 60 musicians take to perform the same symphony? Below it, helpfully, it says: Let P be the number of musicians and T the time.

A two-part line drawing. On the left, a figure is chasing a sparrow with a small bell. On the right, the same figure is rolling a huge cannon bearing the inscription ‘AI’ towards the same sparrow. A metaphor for ‘using a sledgehammer to crack a nut’: an oversized tool for a tiny task

Sustainable AI: The right tool saves resources

It is now well established that AI does not run for free. Dr Dina Barbian shows that AI requires energy and water throughout its entire life cycle: from the manufacture of chips and servers, through the training of models, to every single small task.

SA schematic illustration of the explainability of AI decisions. Left: An input document flows along a clearly defined path through a funnel and a cogwheel to an output document. Every step is traceable. Right: An input document flows into a tangled mass of arrows and footprints, from which the same output document emerges. The processing path is not discernible. In the centre stands a person with a magnifying glass, above whom is a thought bubble asking ‘Why???’.

AI Without the Hype: explainability. What decision-makers need to know when using AI

A customer learns that their tariff is being switched to PrePaid. An AI is behind the decision. They ask: Why?

This question has two aspects. The business aspect: Do I myself understand what the system has done? The legal aspect: Do I need to be able to explain the process if the customer objects or a court asks?

The second question is not the subject of this article. It is a matter for the legal department. What matters here is this: if the answer is ‘yes, I must’, this has consequences for the AI involved in the process. This is precisely where the choice of technology becomes crucial.

A stick figure hammers a nail into the wall using a microscope instead of a hammer. A metaphor for using the wrong tool for the job.

First the process, then the tool

Almost every executive presentation I’ve seen recently contains the same sentence: ‘We need to do something with AI.’ What’s rarely mentioned alongside it is the problem it’s supposed to solve. The order is backwards. First comes the desire for AI, then the search for a task that fits it. Many even openly admit what triggered this: at the last conference, everyone was saying they were already using AI.

Stick figures in a car by the lakeside; the sat-nav shows a straight route into the water – an illustration of the ‘Death by GPS’ phenomenon

AI Without the Hype: Death by GPS and automation bias – why we blindly trust AI

Imagine you’re driving through an unfamiliar area. The sat-nav tells you to turn left, the road looks strange, and a sign warns of a flooded ford. You turn left anyway. Sounds absurd? It happens all the time. In English, a specific term has even become established for this: ‘Death by GPS’. A systematic study identified 158 documented cases between 2010 and 2016 alone in which people blindly followed their sat-nav into danger. Fifty-two of these ended fatally.

A stick figure robot with a clinical thermometer in its mouth and a dice in its hand, a thought bubble with possible continuations for ‘The cat is sitting on the... windowsill? Mat? Veranda? Ceiling light?’, an illustration of the probability distribution and temperature in AI language models

AI Without the Hype: A dice or a clockwork mechanism: why AI never says the same thing twice

A language model works with words. For each individual word, it calculates which following word is statistically the most likely. Imagine the beginning of a sentence: “The cat was sitting on the…” The model calculates that “the windowsill” is the most common next word in 60% of cases, “the mat” in 20%, “the stairs” in 5%, and “the veranda” in 1%. “The dishwasher”, on the other hand, almost never occurs.

The model makes its selection from this probability distribution. If it always chose the most probable word, the same answer would come out every time, provided the training data has not changed. It does not. The model “rolls the dice” and sometimes opts for “the mat”, sometimes for “the chandelier”.

PHuman vs robot approving documents: the employee reviews carefully, the robot rolls a dice and produces three different results

AI Without the Hype: People make mistakes too; that is why this sentence is misleading

A few days ago, I carried out a little experiment. I asked a language model how many Rs there are in the word ‘strawberries’. It’s a question nobody would ask in everyday life; you can see the right answer at a glance. The machine replied very quickly and confidently: ‘Two’. I asked again. Two again. It wasn’t until the third attempt that it gave the correct number, with complete conviction.

A relaxed red cat is sitting on a robot vacuum cleaner and is being driven comfortably around the living room

AI Without the Hype: Why AI seems like magic, and how understanding it demystifies it

The science fiction writer Arthur C. Clarke once said: "Any sufficiently advanced technology is indistinguishable from magic." Typically, Clarke’s law is used to describe an encounter with unimaginable technology. To someone from the Middle Ages, my robot vacuum cleaner would seem eerie. We, on the other hand, know exactly what it can and cannot do.

Recreated dialogue with Joseph Weizenbaum's 1966 ELIZA chatbot: simple text responses creating the illusion of a real conversation

AI Without the Hype: The ELIZA effect and why we anthropomorphise AI

During my university studies, I programmed a simple dialogue agent. The principle was straightforward: recognise the input, match it against a pattern, and return the appropriate response. No intelligence. No empathy. No understanding. A programme that merely pretends.

Two armed soldiers guarding an empty park bench: a metaphor for processes that nobody questions anymore

Spotting unnecessary processes: Why municipal utilities guard garden benches

I heard a joke the other day. A new major takes command of a unit and finds two soldiers standing guard next to a garden bench. No one knows why. He asks his predecessor – who says it was already like that when he arrived. Eventually, the major tracks down the two men’s retired predecessor. He simply asks: “What? Has the paint still not dried?”

Illustration: Artificial intelligence in utilities, use cases from customer service to consumption forecasting

AI Without the Hype: AI in municipal utilities. What works today and what doesn’t

Hardly any other topic is currently dominating industry discussions quite as much as artificial intelligence. At every conference, in every strategy paper, in every conversation with software providers: AI is the answer. Exactly which question it answers, however, often remains unclear.

Child learning to ride a bicycle: AI learns the same way, through examples and mistakes

AI Without the Hype: How does AI learn? Explained simply for decision-makers

No child learns to ride a bike by being taught the physics behind it. They get on, fall off, and try again. After a few hours, and a few scrapes, they get the hang of it. The child knows intuitively how to keep their balance, without ever having seen a single formula. AI learns in a similar way. Not through rules that someone programmes into it, but through examples. Lots of examples. And through mistakes.

Figure skater being scored by four judges with different marks: subjective evaluation in practice

Making better decisions: when gut instinct is enough and when data helps

Does that sound like a contradiction to you? Not to me. If you read on, you’ll find out why. Following your gut instinct is a good thing, isn’t it? Or is it? It depends. In both my personal and professional life, I like to use a simple rule of thumb: the more costly the decision or its consequences, the more rational it needs to be.

Coffee cup in the sunlight on a terrace, the weather app got it wrong

AI Without the Hype: Why AI predictions are unreliable and that’s okay

Last summer, I invited some friends over for coffee on the terrace. The app said it would be sunny. We ended up having coffee indoors.