Amazon refines AI offerings to keep pace with industry rivals

Getting your Trinity Audio player ready...

Tech giant Amazon (NASDAQ: AMZN) has announced a series of improvements to its artificial intelligence (AI) offerings to keep pace with industry heavyweights amid a frantic grab for market share.

The improvements, designed to improve user experiences, were confirmed by the company’s VP for generative AI, Vasi Philomin, at a high-profile conference in New York. Speaking to attendees, the Amazon executive confirmed that the improvement applied to all existing AI offerings, including those from its cloud subsidiary.

Amazon’s generative AI offerings received additional memory to improve their applicability for enterprise scenarios, allowing users to perform complex tasks relatively easily.

“This allows agents to provide more personalized and more seamless experiences, especially for complicated tasks,” said Philomin.

The biggest winner of the upgrades appears to be Amazon’s Q, a generative AI chatbot developed for enterprises to provide “context-aware” answers to user queries. However, the chatbot trails OpenAI’s ChatGPT and Google’s (NASDAQ: GOOGL) Gemini in popularity and performance, a trend that Amazon is keen to reverse.

With the new improvements, Q’s abilities are expected to be at par with those of industry leaders, particularly in terms of coding, logic and long-term reasoning. The chatbot will boast improved personalization and customization features, an expanding use case for business intelligence while being impervious to hallucinations.

Philomin disclosed that the new upgrades will reduce instances of hallucination in its AI models by nearly 80%, a feat that could increase the accuracy of responses for users. Large language models (LLMs) have been grappling with hallucinations since going mainstream, triggering major gaffes with ChatGPT and Gemini coming under fire for glaring errors.

Outside of the software side of things, Amazon Web Services (AWS) has expressed confidence in meeting the computing demands of enterprises joining the AI bandwagon. The cloud service provider has disclosed plans to build new data centers in different regions, banking on its grit and innovative prowess to put it ahead of its peers.

On course for a highly profitable year

Despite falling down the pecking order for generative AI systems, Amazon is on course to hit a new all-time high for revenues. The technology company is still reeling from the highs of scoring its largest quarterly operating profit in Q1, spurred by AI’s rising adoption and demand from AWS.

“We’re seeing strong demand signals from our customers on the AWS side,” said Amazon CFO Brian Olsavsky. “They’re signing longer deals with larger commitments, many with generative AI components.”

The company has not thrown its weight around yet with blockchain technology in the same intensity as it has done with AI, prompting experts to hail the timing of the pivot as “genius.”

Meanwhile, the successful testing of Teranode in a private network environment on AWS stands as another significant achievement for Amazon. The pilot was conducted across distributed regions, including Ireland, Central Canada, Seoul, North Virginia, Oregon and Mumbai, yielding an average throughput of over one million transactions per second (TPS), but Teranode Director for the BSV Association, Siggi Oskarsson, said with the software’s current architecture, it is possible to push this to up to above a million per second.

“To go beyond that, we have to rearchitect,” said Oskarsson, noting that it is even “theoretically possible” to hit six million TPS. However, he said that the Teranode team is currently keeping it at a minimum of one million TPS due to the costs of operation.

The successful testing not only validates blockchain’s scalability but also holds significant relevance for enterprises, which frequently handle massive amounts of data and are often prime targets for cyberattacks.

In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system that ensures data input quality and ownership—allowing it to keep data safe while also guaranteeing the immutability of data. Check out CoinGeek’s coverage on this emerging tech to learn more why Enterprise blockchain will be the backbone of AI.

Watch: Blockchain & AI unlock possibilities