Amazons Alexa+ Launch: A Promising Leap into Generative AI, but Will it Deliver?

The long-anticipated enhancement of Amazon’s assistant is about to be released. Approximately 18 months after the company first introduced the “next-gen Alexa” featuring generative AI, it debuted Alexa+, with preliminary access set to launch in March. Alexa+ will coexist with the traditional version of Alexa and will require a monthly payment of $20, unless you possess a Prime membership, which will allow free usage. This updated assistant will be equipped with the latest enhancements akin to those offered by contemporaries like the revamped Siri or Gemini, such as improved conversational capabilities, enhanced contextual awareness, and the capacity to “summarize intricate subjects” and “propose suggestions based on your preferences.” However, it does introduce a distinct approach, particularly in how Amazon plans to interface with external applications and the broader internet, potentially distinguishing it from competitors.

During the event, Alexa and Fire TV Vice President Daniel Rausch explained three methods through which the new assistant can connect with other services that users may utilize. Firstly, it already has compatibility with “tens of thousands” of integrations, showcasing partnerships with well-known brands like Uber, Sonos, Samsung, and Xbox when Rausch made this claim. This suggests that Amazon collaborated with these partners to ensure their applications function smoothly with Alexa+ via APIs.

Secondly, for the vast majority of online platforms which may not have dedicated apps or the resources to create an API specifically for Alexa+, the assistant is expected to explore the internet for their websites and navigate them on your behalf. Rausch illustrated this point during the presentation by demonstrating how Alexa+ could access the Thumbtack site to hire a professional to repair his oven. Rather than needing to pull out a phone or laptop to navigate menus manually, Rausch could simply communicate his needs and availability verbally, while the AI took care of the rest.

This specific example reminded me of Google’s restaurant-reservation service Duplex, which has been capable of calling businesses on your behalf to book a table since 2018. You only needed to inform Google about your planned date, time, and the size of your party, and it would handle the call for you, even using a voice that sounded human. The way Alexa+ interacts with websites on your behalf appears to be an advanced version of Duplex.

Finally, Alexa+ is capable of collaborating with “AI agents” from various websites, allowing users to interact with just one assistant instead of multiple chatbots. In his presentation, Rausch demonstrated how Alexa+ could connect with the AI music creator Suno to compose a country-style song about bodega cats.

Post-presentation, I engaged with Rausch to gain a deeper understanding of how Alexa+ will interact with various online services. I wanted to clarify if the third approach would include customer service chatbots from entities like Capital One or United Airlines among others. While Rausch mentioned that he could not provide specifics on particular names or services, he did confirm that “the SDK focuses on integrations like this,” indicating it’s designed for users to utilize Alexa to connect with agents and perform tasks.

Rausch recognizes that any friction during the adoption or setup process can deter users. To ensure a smooth transition, upgrading to Alexa+ should not necessitate any new logins or authentication procedures. Users can expect their connected devices, security cameras, and home routines to transfer seamlessly without extra effort.

“The reason why customers appreciate Alexa is due to its ability to simplify complexities,” Rausch noted. Users shouldn’t have to recall the brand of the WeMo plug they purchased just to instruct Alexa to turn off a lamp. “We would never take that away,” he stated. When integrating new services after the upgrade to Alexa+, Rausch assured users that they would either follow the “same methods as before” or find the process even more streamlined, given that “Alexa can assist you in setting these up in numerous scenarios.”

According to Rausch, rather than having to “search through the Alexa app, you simply say ‘Alexa, I want to set up a streaming account with Hulu’ or something similar.” A QR code will be displayed on an Echo device with a screen to facilitate this process, and the assistant is designed to guide users throughout.

“We like to say Alexa is now an expert, and also an expert on herself.”

Amazon is not the only organization that has engineered its assistant to serve as tech support for its users. This method is notably similar to how Siri can assist you in, for instance, recording a video using Cinematic mode or designing a Genmoji. Where better to seek guidance for a product than from the product itself? Why direct users elsewhere for assistance?

“From a customer relationship standpoint, consumers simply desire the task to be accomplished,” Rausch remarked. “They shouldn’t need to be responsible or concerned about any of these details, correct?” He believes that users want plumbing services or restaurant reservations completed without the hassle of navigating websites and making phone calls.

People certainly expect their digital assistant to be accessible and useful, but shouldn’t the accuracy and reliability of these AI services hold equal importance?

When I inquired how Amazon addresses the issues of generative AI’s propensity to hallucinate and occasionally propagate misinformation, he responded, “I believe there has been a misconception in the industry that a model serves as a product.” He also stated, “LLMs form the foundation of the architecture, but they’re not the sole component providing answers.”

In summary, Alexa+ combines Amazon’s knowledge graphs, trustworthy online sources, and collaborations with reputable outlets. “Other products may simply deliver an answer sourced from the LLM. If you’re seeking authoritative information, that is not always the best approach,” he explained.

Rausch emphasized that Amazon has “exercised considerable caution” with Alexa. “Will it make errors? Every software has its flaws,” he acknowledged. “However, we are diligently ensuring it is based on solid knowledge.”

I have yet to experience Alexa+ firsthand, and everything we’ve observed has taken place in carefully curated demonstrations, making it challenging to predict real-world performance. However, if the new Alexa can genuinely collaborate with third-party services in an effortless and efficient manner, it could not only reposition Amazon within the consumer AI landscape but also provide an edge over competitors like Google, Apple, and OpenAI. Given that Amazon pioneered the concept of virtual assistants in homes globally, this may also influence users beyond today’s tech-savvy clientele.

Amazon has introduced Alexa+. Stay tuned with the live updates from the Alexa+ launch event for continuous insights and commentary.