For long now, chatbots have been generating weird, non-relevant answers. Of course there are also good examples of how chatbots can improve your life: Aeromexico has implemented a chatbot (and very recently also integrated full Whatsapp support) that works well and have gained worldwide appraisal for its performance.
However, in general, chatbots don’t seem to generate the traction some people might have expected since roughly two years ago, when Facebook launched their chatbot API, back in April 2015 (see my blog article on how we used that first API with blockchain integration). This post is not to explain why it didn’t take off or what fundamental technical and psychological problems slow down the adoption, instead I wanted to focus on a very useful application of a chatbot, today.
Maybe you’ve read about the recent Equifax hack that occurred on September 7th, whereby more than 143 million credit cards may have been stolen. Joshua Browder created a chatbot to help the victims in filing their ‘lawsuit’, without the need to talk to a to ‘a real-life lawyer’. It basically does nothing more than help you skip a bunch of very complicated forms, that people like you and me wouldn’t understand (let’s call it: law-talk).
Although very basic in its nature, it raised a very interesting point for me: this is a very good use for a chatbot. In a similar fashion RPA (Robot Process Automation) tries to help employees in being faster in their work, chatbots might be a good solution to save time and bring hassle-free experiences to complex services (such as filing a lawsuit). The underlying process is straightforward and is not optimized for ambiguous questions, but in the end, it avoids millions of people having to do a very emotional and hard task of getting a compensation for their loss.
Not rocket science, not AI, not machine learning, not cognitive intelligence; just a smart way of helping people get what they deserve, frankly.