Alibaba Cloud AI aims to predict singing contest winner

Staff writer
Future TV Asia

Alibaba Cloud’s “Ai,” an artificial intelligence program was put on the spot to predict the winner at the grand finale of “I’m a Singer,” the popular Chinese reality television produced by Hunan TV.

The competition is major annual event and attracts significant public participation in China. Ai predicted the winner by using neurological networks, social computing and emotional perception.

Min Wanli, chief scientist for AI at Alibaba Cloud said the result was jointly created by TV audience, public judges, as well as the seven contestants.

“It is very random and almost impossible to predict using human intelligence, and we aim to achieve real-time prediction by Ai,” said Min.

“In a previous round, Ai predicted two of the top three winners on April 1,” said Min. “We believe that it will achieve a better performance after learning and evolving over the past week.”

The program has the potential to understand human emotions, gather insights in real-time, and evolve through strong computing and machine-learning capacity.

With this capability, Ai identified and assessed factors that may affect the result, including popularity of the songs, the singers’ voice pitch and energy, audience response and online discussions, to name a few.

The program created and used a dynamic computer model to predict the result by computing both fact-based logical data and subjective emotion-based data.

Ai’s prediction and the judges’ voting were processed independently and did not affect each other.

In the future, Ai will be applied to areas such as personal assistance, weather analysis, smart cities and social trend predictions.

Commentary

Goodbye BSS/OSS, hello MMB (Money Making Bits)

Billing is forever, but BSS and OSS are outmoded concepts in this new age of creative pricing

Billing is forever, but BSS and OSS are outmoded concepts in this new age of creative pricing

Steven Hartley/Ovum

Ovum predicts that there will be major opportunities and challenges for the TMT sector