Alibaba one-ups Google with convincing AI voice assistant

Alibaba demoed its already operational AI voice assistant at a recent event, with a service that is more robust than Google's Duplex which wowed the internet earlier this year

7Dec

Alibaba impressed delegates with its AI customer service voice assistant which handle calls for its logistics company Cainiao, during a presentation at the 2018 Neural Information Processing Systems Conference.

The unveiling was surprising as, earlier this year, tech giant Google unveiled its own prototype AI voice assistant at the I/O developer conference. The Google Duplex shocked audiences with its incredibly life-like tone of voice with caused many on social media to voice their fears for a future where people no longer know if they are talking to a person or a machine.

However, while Google insinuated that they still had a way to go with the Duplex, Alibaba's AI voice assistant is not only already taking calls from customers, but seemed more advanced than the Duplex, at least from the prerecorded demo they played. MIT Technology Review published the translated from Mandarin conversation the AI had with a customer regarding the delivery of a package:

Agent: Hello, I am Cainiao's voice assistant. I am—

Customer: Hello.

A: Yes, hi, you have a package scheduled for morning delivery to 588 Culture West Road. Is it convenient for you to receive?

C: Who are you?

A: I am Cainiao's voice assistant. I'd like to confirm your morning delivery to 588 Culture West Road. Does that work for you?

C: I’m not home in the morning.

A: Then do you have another address that you'd like to use?


Visit Innovation Enterprise’s Machine Learning Innovation Summit, part of the DATAx New York festival, on December 12–13, 2018


Experts have pointed out that in the space of a 30-second conversation, the voice assistant effortlessly handled three common problems for conversational AIs. The first was interruption, the ability for an AI to react appropriately when interrupted and carry on relaying the necessary information despite it. Next was its ability to respond to nonlinear conversational answers, when a response is not related to the preceding question. Finally, it demonstrated implicit intent – the ability for an AI to read between the lines and infer meaning.

These skills are difficult for most modern AI voice assistants to handle. However, due to Alibaba's size, it has access to a huge amount of data which AliMe, its machine intelligence and technology lab, can use to train its speech recognition and natural language processing (NLP) engines through a variety of call scenarios.

Customer service is a big part of Alibaba's workload as the company handles up to 50,000 customer service calls a day, so it is very incentivized to perfect this kind of technology. It also has plans to expand its digital voice assistant services into its food ordering services and a price haggling chatbot.

How it departments should evolve to deal with cyber attackssmall

Read next:

How IT departments should evolve to deal with cyber-attacks

i