Rules of Engagement

by Tony Veale of University College Dublin

When you see a child throwing a tantrum on a train, who do you blame: the child, who – though annoying – may not know any better, or the parent, inured to the noise and unwilling to do anything about it? Thinking is something we must all learn to do if we are to do it well, and we must all learn to think socially as well as intellectually to successfully engage with the world. The same is true for our machines: to engage successfully with the world, they must engage successfully with us and with each other. They are like children: when they produce hilariously stupid results, the fault lies as much with us as with them. If we don’t train them to engage with the task at hand, to anticipate the unexpected, and to know when their rules are about to break down, then we all share in the stupidity that ensues.

Bake me a cake as fast as you can

Even bakers now use technology. Email them the picture or text you want on your cake, and they use a special printer (with food dyes for ink) to print your design onto the icing.

This real cake design produced by a New York bakery is a funny example of what can go wrong when people and machines don't engage properly with what they are doing.

The customer used Microsoft Outlook to email the cake message to the bakers. It inserts special HTML tags (code used to format web pages) into its emails to make them look prettier. Unfortunately, the bakery doesn't use Outlook, so the HTML was cut-and-pasted directly into the cake-printing program. We can laugh at the software but the human baker is mostly at fault. Did they really think this was the cake design? Poor Aunt Elsa. If a person can't detect this kind of problem, what hope for our machines?

Breaking Rules

It is commonly believed that creativity is about breaking rules. If you are going to break a rule, start with that one! Nothing is further from the truth. Creativity comes from a hyper-understanding of the rules rather than from a willful ignorance of them. We must know the limits of our rules, and how to tell the difference between a convention and a hard rule. People are creative all the time in chess, poetry and soccer without ever breaking the rules: instead, they break with convention! Creativity requires a deep engagement with the rules, to know where individual initiative can take over. Can computers ever show this kind of engagement with rules? The answer is a qualified "yes": they must be programmed in the right way, not just with hard-coded rules, but with knowledge of their own workings, able to reason about their own rules. Building systems like that is what computational creativity is all about. Ironically, such programs may have more self-knowledge than a human doing the same job. By using introspection to design them we humans can obtain greater self-knowledge of how we ourselves work.

To find out more about computational creativity research check out RobotComix.com which is full of computational creativity comics and cartoons, including a new book Hand-Made By Machines. Enjoy!