New Developments in AI Technology with Alexa+ and Siri

Even giants like Amazon and Apple are struggling as they stitch generative AI into their widely-adopted voice assistants. Alexa and Siri are but the tip of the technological iceberg. In an effort to achieve the best possible experience for their users, both companies’ trips to success have seen important progress and mistakes.

In February Amazon introduced Alexa+. This latest incarnation of its voice assistant is supposed to come up with unique responses on the fly. This new generative ability is what sets Alexa+ apart from its forbearer—which mostly just used a set of preloaded animations to guide user interactions. Still, Alexa+’s launch was not without its shortcomings. To start with, it lacked some basic functionality. It was incapable of hosting third-party apps such as GrubHub, or telling you bedtime stories, or helping you come up with ideas for birthday presents.

Amazon is currently working on a web-browsing agent named Nova Act to beef up Alexa+. This new technology is designed to create responses with up to 90% accuracy. Today’s most sophisticated multi-step AI agents are challenged to achieve accuracy rates in the 30%-60% range. Amazon CEO Andy Jassy admitted the challenges faced along the way in development. He acknowledged that doing the same with generative AI into Alexa has been met with multiple bumps in the road and setbacks.

Apple’s Siri is experiencing a revolution. As CEO Tim Cook said, the company just needs more time to make sure Siri has the right features. Siri, similar to Alexa, defaults to canned responses. It can do all the practical stuff, too—setting timers and reading your texts. Unlike Google, Apple has walked away with its share of challenges as the transition to generative AI for Siri has been equally fraught with complexity.

Both companies see generative AI as a valuable tool to transform the way users interact. Our intent is to design a powerful, exciting experience that meets user needs in an interesting way. Our goal is to provide personalized and contextually appropriate responses. Yet the path to realizing these ambitious goals is rocky.

Tags

Leave a Reply

Your email address will not be published. Required fields are marked *