Microsoft’s new chat-bot, Xiaolce is an AI that utilizes a deep learning algorithm to chat with humans and also hear at the same time
The bot has already been tested in China and has already had 30 billion conversations
The AI is available for free across Asia
When Google released their chat-bot, Google Duplex, the world was left wondering how far machine learning and AI had come and how far it was yet to go, but Microsoft soon crashed their party 2 weeks later with their very own chat-bot Xiaolce. The bot was released at Microsoft’s AI Innovate event. It could talk and sound exactly like humans.
The project was initially launched in August 2017 and by April the bot could listen and speak at the same time, similar to how humans can. The bot was tested with millions of users in China and engaged in social conversations. During the event the company revealed some exciting developments and features of the project. Below is a summary of what they released:
- The company is making the chat-bot available of partners and developers to be made use of in their applications.
- New features are to be added than can create customized audio stories in a short period of time. The team claims that with inputs from different kids and parents the bot will be able to come up with a short story in about 20 seconds!
- For personalizing background music and story characters the bot will be utilizing a specifically trained deep learning model. This free of cost audio service is already available across Asia since June 1, 2018.
- Microsoft has acquired a conversational AI company, Semantic Machines, in Berkeley, California to make their NLP products and features even better.
According to a post by Harry Shum, Microsoft’s executive Vice President for AI and research revealed that the chat-bot had over 30 billion conversations with over 100 million friends and has talked with about 600,000 people over the phone. You can read more about the blog here.
During the event the chat-bots capabilities were displayed live as the AI had called a user and a very natural conversation with them. They also have a video about this you can check it out here.
So let’s summarize:
Microsoft also mentioned that they will be working with a non profit organisation to create the worlds biggest audio library with short stories and information for children who are blind or visually impaired. This is an amazing feat undertaken by Microsoft to use AI for helping people around the world.
A question does arise about ethics, where the bot can be used to make fake audios. But Microsoft has promised that when this AI is used, the person on the other end of phone will know that they are talking to an AI.