Communicating in plain English, or “natural language”, is one of the tasks that computers struggle with most traditionally. Even though people use language effortlessly, computers have trouble grappling with the ambiguity and use of context and background knowledge that permeate natural language. But recent advances in machine learning have led to incredible progress in conversational AI and natural language processing and are setting the stage for a world where computers will be able to communicate with us in plain English (or French, or Chinese).
Tools like Google Translate are breaking barriers and making resources that used to be limited to just English-speakers or just Spanish-speakers accessible to people across the world. These tools use neural machine translation, which trains a neural network by feeding them lots of examples of texts that have been human-translated, and they can deduce from the structure of these examples how to handle sentences that they’ve never seen before. This method beats older hand-coded methods where translations would be inferred from simpler statistical models and heuristics.
These tools can even be used in real-time, helping to connect people who would otherwise be unable to communicate by using their smartphone or smart watch to perform translation as they speak with each other.
When you search online, you get helpful autocompletion of your searches as you type. Natural language processing has taken this to the next level, with AI capable of generating entire articles or even programs with just a simple prompt. The GPT-3 system from Open AI is a neural network trained with billions of parameters that has basically learned from all of the data on the Internet, and is able to compose convincing works, to the point of fooling people into thinking it’s human-written.
Amazon made Alexa a household name, with competitors like Siri and Google Assistant not far behind. Voice assistants make it easy to do everything hands-free, from playing music, ordering pizza, to controlling the lights and appliances in your home. These systems are built on natural language understanding technologies that take the audio of your spoken phrases, and interprets them into commands that the system can execute.
These systems are becoming ever more important for search and commerce, and their functionality is growing to include complex multi-stage tasks like planning a trip. The voice assistant vendors are always competing to improve accuracy, reduce latency, and expand functionality to allow you to do more with your voice assistant.
Interacting with your devices using natural language isn’t limited to voice assistants. Natural language technologies are also replacing traditional IVR systems on the phone to make it easier to find what you want when you call a store or a support line.
Text channels like SMS, Facebook Messenger, or Whatsapp are also ideal places to deploy conversational AI, where you can engage with a chatbot in plain English to place an order, learn more information, or ask about the status of a support ticket. These channels are already where we spend a lot of our time on mobile, and making it possible to engage with businesses here broadens the market for all kinds of businesses.
How to use conversational AI in your business
More and more businesses will be able to broaden their reach to include international customers. For example, you can create versions of your website in other languages like Spanish or French, helping to attract traffic from neighbors like Mexico or Canada.
You can also use machine translation to help you in your store. You can improve communication with employees who don’t speak the same language by using translator apps, or you can put up signs in foreign languages to help customers who don’t speak English find what they’re looking for. This is especially important in areas that have a large amount of tourists, who will appreciate being able to find help in their native language.
Conversational AI can also be used to streamline operations and improve efficiency in your store or restaurant. One of the big bottlenecks in operations is answering the phone, which can take employees away from serving customers in the store. Using AI virtual receptionists, you can service callers without distracting your employees, while also providing a level of service that’s much better than traditional IVRs that force callers to navigate complicated menus. Systems that use cutting-edge natural language processing technology like Greet can answer customer questions and help them place orders or make appointments without forcing them to wait through lengthy prompts.
This is the 3rd article in our series on how AI will change your business. Check out the other articles in the series to find out more about autonomous vehicles and more.