Amazon job postings indicate ChatGPT-like conversational AI for online store

Amazon, the online retail giant, has long been quiet about its conversational artificial intelligence plans, even as its rivals Google and Microsoft make strides in developing and deploying chatbots that can interact with users and answer their questions.

But a new pair of job listings may just offer a glimpse of Amazon’s ambitions. The vacancies, which were first discovered and reports Bloombergdescribed a new search functionality for Amazon’s webstore with a chat interface powered by a technology similar to ChatGPT, one of the world’s leading natural language AI systems.

The job postings suggest Amazon is looking for engineers who can help create “an interactive conversational experience” that allows customers to ask product questions, compare products, receive personalized suggestions, and more.

The posts also liken the project to a “once-in-a-generation search transformation,” similar to how the Mosaic browser revolutionized the Internet in the 1990s.

Here is the full summary of the now deleted job offer:

“We are reimagining Amazon Search with an interactive conversational experience that helps you find answers to product questions, conduct product comparisons, receive personalized product suggestions and much more, making it easy to find the perfect product for your needs. We’re looking for Amazon’s best and brightest to help us realize and deliver this vision to our customers right away. This will be a one-time transformation for Search, just as the Mosaic browser made the web easier three decades ago. If you missed the 1990s – WW, Mosaic and the creation of Amazon and Google – you don’t want to miss this opportunity.”

A new Amazon shopping experience

If realized, the project would represent a significant departure from Amazon’s current search interface, which relies on keywords and filters to help customers find what they’re looking for. It would also put Amazon in more direct competition with other tech giants that have invested heavily in conversational AI, such as Google, Meta and Microsoft.

When asked for comment, an Amazon spokesperson told VentureBeat, “We are investing significantly in generative AI across all of our businesses.” The company did not provide additional details about its job openings or its conversational AI strategy.

However, Amazon has shown signs of interest in recent months. In October, it announced Bedrock, a cloud-based platform that allows developers to build and scale chatbots and other applications using large language models. Amazon also unveiled two new language models of its own called Titan, which it said were trained on billions of words from books, news articles and other sources. The announcements show that the company is making an effort to keep pace with other technology conglomerates in the field.

Amazon was once considered a pioneer of conversational AI when it released one of the first mainstream AI assistants, Alexa, which launched in 2014 and quickly became a popular feature in millions of homes and devices. While Amazon Alexa has never managed to capture the success that text-based solutions like ChatGPT have, the company reported by January 2019, it had sold more than 100 million Alexa devices.

While the hype around Alexa devices has died down, many similar capabilities are now used in text-based chats. For example, Microsoft’s Bing Chat is a conversational AI tool that allows users to purchase products or services through a text-based conversation. Bing Chat can show options, compare prices, and even generate ads based on user queries. Google, Meta and others have also experimented with conversational AI in different domains.

However, Amazon’s advantage may lie in its vast collection of user data and e-commerce expertise. Amazon has published several research papers on conversational search, including one on ConvSearch, a dataset of over 10,000 dialogues collected from real users across various topics and domains. The paper, presented on the SIGIR 2021 conference, shows how ConvSearch can be used to analyze and model user behavior and preferences in conversation search scenarios.

Amazon’s CEO Andy Jassy also hinted at the potential of conversation search in the company’s latest version call income. He said large language models and generative AI capabilities have improved significantly over the past year, opening up new opportunities for transforming customer experiences. “They’ve been around for a while, but honestly the models weren’t that appealing about 6, 9 months ago,” he said. “And they’ve gotten so much bigger and so much better, much faster that it really represents a remarkable opportunity to transform virtually every customer experience that exists, and many that don’t exist and that weren’t as easily possible before.”