In an interview on Tuesday, Philomin stated, “This allows agents to provide more personalised and more seamless experiences, especially for complicated tasks.”
For example, he mentioned that the upgraded AI agents could now recall a user’s preference for aisle or window seats on a flight for every subsequent request, something that was not possible in the past.
In addition, Amazon revealed that it has enhanced the Q chatbot it unveiled in November to offer better advice on creating software code, addressing one of the more well-liked applications of generative AI.
Additionally, Amazon announced that it has improved its Bedrock service, which enables companies to develop applications using a variety of AI models, to better assist users in identifying and removing “hallucinations,” which occur when AI generates responses to queries or requests that can be inaccurate or deceptive.
Because they cause users to become distrustful of AI systems, hallucinations have been a persistent issue. For example, earlier this year, Google faced backlash for an AI-powered search function that suggested, among other things, that users add glue to pizza sauce to make sure the cheese sticks to it.
According to Matt Wood, vice president of AI products at Amazon Web Services, the new controls will help minimise the incidence of hallucinations by roughly 75% for specific uses.
Amazon stated in April that AWS, which is in charge of most of the company’s AI work, is on track to generate $100 billion in revenue annually.