I first became fascinated with the relationship between computers, information, language, and emotion back in 1984 when I read an Analog article about a computer than had been used to automatically generate poetry. A week later, I was on my father’s IBM PC, writing a Basic program on (if memory serves me correctly) Dos 2.1 that did exactly that. It was the first computer program that I had ever written, and I was using it to generate part-of-speech structured poetry. It therefore wasn’t that far of a stretch, many years later in computer science grad school, when I took up Information Retrieval as my primary research field. I have always had an interest in the relationship between expression and information. So I was quite surprised when I saw this article about how the first computer was used to generate love poetry:
Back in 1952 a team of scientists was desperate to test the capabilities of Mark One `Baby`, the computer built at Manchester University. One of them, Christopher Strachey, devised a quirky software programme by entering hundreds of romantic verbs and nouns into the new machine. He then sat back as Mark One `Baby` trawled the literary database to create a stream of light-hearted verse.
There must be some underlying similarity shared by those attracted to computering technology, a love of language and structure, information and randomness, self-expression and programming, that leads us to create these sorts of systems. This is not a new or unique observation; the article simply reminded me of how I got my start on a path that eventually led me into Information Retrieval.