Amazon is testing the use of artificial intelligence (AI) in its shopping app to generate summaries of some product reviews left by customers.
The feature provides a brief overview of what shoppers liked and disliked about the product, along with a disclaimer that the summary is “AI-generated from the text of customer reviews."
Consumers often use reviews to decide whether or not they will buy a product or service, or eat at a restaurant, but it's not clear how summarizing all the reviews based on purchases could help that decision.
The reviews are summarized into one short paragraph using generative AI. Each summary includes positive and negative feedback, shows highlighted keywords in bold, and ends with mixed opinions from multiple customers. It appears that the tags on its review platform, which enable the consumer to click on words and sort through the related reviews, might come into play.
Rather than aggregate reviews, Amazon needs to develop AI tools that identify genuine reviews, he said.
A mobile listing for a children's “Magic Mixies” cauldron toy says that buyers gave positive feedback around its “fun factor, appearance, value, performance, quality, charging, and leakage.”
“However, the majority of customers have expressed negative opinions on these aspects,” the summary states, according to CNBC. “For example, some customers have paid over $100 for a toy that wasn’t worth it, while others have experienced issues with the product’s quality and charging.”