Tech Times | A plague of dumb chatbots
Chatbots are suddenly everywhere,
promising to help us with common tasks.
How can we get them to perform better?
For the past few minutes, I've been chatting with George Washington, and
honestly, he seems rather drunk. He keeps saying things like "cool", "haha" and "u wanna join my army or wut?"
This, of course, is not actually America's first president. It is automated conversational artificial intelligence, known as a chatbot, created by the TV comedy Drunk History and made available through the messaging programme Kik. It's entertaining, if not very coherent.
You can now chat with all sorts of bots through a number of messaging services, including Kik, WeChat, Telegram, and now Facebook Messenger. Some are simply meant to entertain, but a growing number of them are designed to do something useful. You can book a flight, peruse the latest tech headlines, and even buy a hamburger from Burger King by typing messages to a virtual helper. Start-ups are racing to offer tools for speeding the development, management, and "monetisation" of these virtual butlers.
The trouble is that computers still have a hard time understanding human language in all its complexity and subtlety. Some impressive progress is being made, but chatbots are still prone to confusion and misunderstanding. The best commercial chatbots will most likely be those that recognise their own limitations.
SMALL AND SIMPLE
"A pitfall is trying to do too many things at once," says Paul Gray, director of platform services at Kik, which has offered integration for bots since 2014. "You should start off small and simple."
This rush toward chatbots is partly due to the popularity of
several new messaging services. Efforts to harness bots have also, no doubt, been inspired by incredible progress in recent years in other areas of artificial intelligence, such as processing imagery and audio. But processing language is an altogether different challenge, and one that has bedevilled artificial intelligence (AI) researchers for decades. Chatbots date back to the earliest days of AI. One of the very first, called Eliza, was developed at Massachussetts Institute of Technology in 1964. Playing the role of a psychotherapist, Eliza used a simple trick of stringing people along: asking standard questions, and often rephrasing a person's own statements in the form of a question.
Today's chatbots are better, but not by much, and it's hardly surprising. There have been no fundamental breakthroughs in training computers to process and respond to language in recent years. That said, the techniques that have led to advances in other areas, primarily deep learning, are showing promise for parsing language, says Chris Dyer, an assistant professor at Carnegie Mellon University. "There is a lot of exciting work coming out every few months on question-answering," he says.
But given the likelihood of misunderstanding, don't expect bots to handle everything. "Legal advice, medical advice, and psychiatric counselling would probably all be very risky," says Dyer.
- MIT Technology