Amazon Entered AI Race with New Foundation Models

 

Amazon (NASDAQ: AMZN) unveiled its latest foundation model of artificial intelligence (AI) platform, called "Nova," at the annual AWS conference on Tuesday, Dec. 3. It marked Amazon's entry into the competition of automated services with tech giants like Adobe (NASDAQ: ADBE) and Meta (NASDAQ: META). Amazon CEO Andy Jassy announced that the new model addressed developers' demands for better latency, lower costs, and enhanced fine-tuning capabilities.

 

Nova also means that Amazon is erasing its reputation for lagging behind its competitors in AI applications. Rohit Prasad, head of artificial general intelligence at Amazon, pointed out that the Nova model can outperform its competitors with its faster speed. He is confident that Amazon will grow to be the industrial leader despite its AI model being pre-mature.

 

The major feature of Nova is the Nova Reel tool which enables users to create six-second videos from one text or image. This tool is potentially useful for product displays.

 

Concurrently, Amazon introduced Canvas for generating images from text and announced future AI models capable of processing text, images, speech, and video interchangeably. Amazon also previewed an AI-enhanced version of its Alexa voice assistant, codenamed Banyan. Banyan has been delayed due to accuracy and speed issues. Nevertheless, Prasad, Banyan's supervisor, is positive about Banyan playing a voice assistant role in the AI field.

 

To reduce the entertainment industry's concerns about AI content, Amazon equipped its Nova with the watermarking ability to avoid spreading harmful content. However, some industrial professionals still worry about copyright violations.

 

#ArtificialIntelligence #amazon