You buy an item on Amazon and want to know what other buyers think about it. But this is a popular product that has attracted hundreds, if not thousands of reviews. Instead of reading them all, you will now be able to consult a summary generated by artificial intelligence and designed to summarize the opinions of all these buyers.
On Monday, Amazon announced the launch of its new AI product evaluation option. Now accessible to a significant sample of mobile buyers in the United States, the generative AI technology will display a short paragraph on the product page, highlighting the main features and summarizing the general opinion of customers. The strengths generated by the AI will also allow you to choose a specific attribute of the product, such as “ease of use” or “reliability”, to see excerpts of reviews mentioning this factor.
The company started testing this new feature in June, but it has just generalized it. If the option is available for a product that you are viewing, a review section appears with a brief summary and a mention indicating that it is generated by artificial intelligence from the text of customer reviews.
screenshot by Lance Whitney/ZDNET
The use of AI to generate review summaries raises a number of questions
Amazon is not the only online seller to offer AI-generated product review summaries. Last week, another seller, Newegg, announced using ChatGPT to condense customer reviews. The feature is called “Review Bytes”. For now, the option is only accessible on the desktop version of the website and only for products that have a certain minimum number of reviews.
The use of AI to generate review summaries raises a number of questions. First of all, can we be sure that the AI faithfully reflects the many opinions it summarizes? Secondly, how does the AI avoid or exclude fake reviews?
To answer the first question, Amazon explains that it continues to test and refine its AI models in order to improve the experience. As for the second question, fake reviews are a problem that has plagued Amazon and its customers for years. And this is a problem that the company continues to try to combat.
For AI summaries, Amazon said it only incorporates reliable reviews from verified purchases
Amazon said it continues to invest in resources aimed at stopping fake reviews before they appear. To detect fraudulent reviews and unusual behavior, the company uses machine learning models to analyze thousands of data points, including relationships with other accounts, login activity and review history. For AI summaries, Amazon said it only incorporates reliable reviews from verified purchases.
The company has also filed lawsuits against fake review brokers who contact customers through websites, social media and messaging services and trick them into writing fake reviews in exchange for money or free products. In 2022, Amazon prevented more than 200 million alleged fake reviews from appearing online.
Source: “ZDNet.com “