If you've ever hated a computer, felt like screaming at an automated answering service or just got too disheartened to learn how to use a mobile phone, you will recognise the problem at the interface between human beings, who have feelings, and information technology, which does not.
But Roddy Cowie, professor of psychology at Queen's University Belfast, is on a crusade to make life easier for those who struggle with machines.
Professor Cowie is co-ordinator of Humaine, a European Union-funded network across a dozen countries.
His aim is to get computers to adjust to a style of communication that suits people instead of pushing people to adapt to computers. The Humaine project brings together techies and psychologists to form teams that know both about machines and people.
"You can tell from how I talk that I'm an enthusiastic, excitable, slightly scatty guy," Professor Cowie said. "You process that automatically, without thinking."
The challenge is to gather information on how we do this and to make it available to technological devices. "If you can't document what people are feeling, you can't get machines to do likewise," Professor Cowie said.
He has written a system for recognising voice emotion that includes some 400 features, from pitch of voice and loudness to hesitation, repetition and length of vowels.
This is being used in France to monitor calls to call centres: when someone becomes distressed, they are rerouted from the automatic system to a real person. Other projects include a responsive tourist guide: if a sightseer is enthused by its museum recommendations, it suggests bookshops, but if she is bored, it moves on to shopping malls.
There are also facial recognition programmes, but Professor Cowie warned there would have to be an element of personalisation for these.
A Serbian undergraduate at Queen's investigated how people in Belfast and Serbia rated images showing displays of emotion. "We (in Belfast) think we are quite emotional, but we're comparing ourselves to the English. What we regarded as strong was regarded as very weak or not emotional at all in Serbia."