Drones… the cloud… 3D printing… self-driving cars (even ones that strike up conversations with you): so far, so familiar. Astonishing though they are when you think about them, these are now well-recognised chunks of what we now know will be our future. But there was also one other piece of technology mentioned at IBM Connect 2017 that those in the know are even more excited about, but seems far less sexy… Chat Bots.
The person who mentioned them was Gary Theis, the Chief Technology Officer at ZUMATA. The company sits between slick travel websites and the things people are actually booking – hotel rooms and the like. He likens the job to tidying up a horrendous bundle of cords at the back of a massive computer, making sense of the mess and making sure that you find the best way to match the data provided by hotel companies up to the way the customers are using the websites.
The application of clever cognitive systems seems obvious for ZUMATA. It crunches data, sorts it out, makes it more meaningful. It’s able to comb hotel reviews for sentiment, and use machine learning to give personally-relevant results.
But if you push this idea even further, then you end up with chat bots, which are in effect little more than a clever interface that allows you to find the information you want in a more conversational way. For travel websites, that means that instead of drop-down menus and search terms, you stick in a couple of replies to questions, perhaps spoken rather than typed, and – just like a proper travel agent – the computer is able to finesse your query into exactly what you’re looking for.
Think about it as an easier and more personal experience. Think back to your local shop where the owner is able to deal with you personally, and even second guess what you want. ‘Progress’ then meant a supermarket, where you navigate the aisles yourself. The next stage was a website with lists of products that you could buy, clicking on them one after another. A few clever algorithms meant that it sometimes prompted you to buy X because you bought it last time, or because it is related to Y, which is already in your shopping trolley.
Bots take this a stage further, and in fact a stage closer to that early situation of being with a local shop owner who knows you personally. The interface is personal: you might just speak, perhaps to say you’re baking a cake for your nephew. The chat bot is able to understand your preferences (plus those of your nephew, thanks to your connections in social media networks). It might suggest chocolate cake and ask if you want the ingredients, or for it to simply order one for you. You reply. You might even hold up a sketch that your nephew did of his ideal cake. It’s done.
This level of natural conversation is devilishly hard to achieve. It relies on a Watson-like level of computing power that can do everything from language processing and tone analysis to image recognition and social media mapping. Drones, 3D printing and self-driving cars might get all the headlines, but chat bots really do have the power to revolutionise how companies and customers do business.
Originally published on ogilvydo.com