Teams Compete To Create The Best AI Chatbot And Win $3.5M Alexa Prize


Eight teams are working to build chatbots using Amazon’s resources and win the Alexa Skills $1.5 million prize (with $2 million in other grants and prizes).

In February, Amazon announced the eight finalists for the competition: Alana, Heriot-Watt University, Edinburgh, Scotland; Alquist, Czech Technical University, Prague, Czech Republic; Emerson Conversational Search Agent, Emory University, Atlanta, GA; EVE, Brigham Young University, Provo, UT; Gunrock, UC Davis, Davis, CA; KTH Fantastic, KTH Royal Institute of Technology, Stockholm, Sweden; SlugBot, UC Santa Cruz, Santa Cruz, CA; and Tartan, Carnegie Mellon University, Pittsburgh, PA.

Each team will receive a $250,000 research grant, Alexa-enabled devices, free Amazon Web Services (AWS) to support their development efforts, access to new Alexa Skills Kit (ASK) APIs, and tools, data and support from the Alexa team.

A $500,000 prize will be awarded to the team selected for creating the best socialbot. The second- and third-place teams will receive prizes of $100,000 and $50,000, respectively. Additionally, a $1 million research grant will be awarded to the winning team’s university if their bot achieves the grand challenge of conversing coherently and engagingly with humans for 20 minutes with a 4.0 or higher rating.

The Verge gave an inside look at the competition, where finalists are using Amazon’s resources, including basic speech recognition tools from Alexa, free computing power from Amazon Web Services, and training data from tens of millions of Alexa users. The chatbots went online in the U.S. last month, and feedback from users will help the teams improve their creations before the final judging process in November.

Through the competition, Amazon is able to enlist some of the brightest talent in AI to work on its platform, while some finalists might even be given the chance to join the eCommerce giant in the future.

And all of this work will help to bring Amazon to its ultimate goal: a future where Alexa can hold human-like conversations. A more talkative Alexa would become more personable, which is an entity users could relate to “as a friend, as a companion,” said Rohit Prasad, chief scientist for machine learning at Alexa.