Since Amazon launched the first Echo in 2015, voice assistants have been gaining presence in homes around the world.
Devices like Alexa, Google Assistant, and Siri have become everyday companions for basic tasks like playing music, controlling smart devices, or checking the weather.
However, after an initial boom, many of these assistants appear to have stagnated, with features that have not evolved significantly and usefulness that remains limited for most users.
With the advancement of artificial intelligence (AI), the question that arises is: can voice assistants be revitalized and become truly indispensable tools in our lives?
The fall and resurrection of Alexa
The case of Alexa, Amazon’s voice assistant, is one of the most representative of this situation. Despite its initial popularity, Alexa has failed to maintain the same level of user enthusiasm over time.
In 2022, Amazon reported a $10 billion operating loss related to Alexa, leading some to declare the voice assistant “dead.”
However, Amazon has not stopped investing in Alexa, and even then began exploring generative AI features, such as the AlexaTM 20B.
This model, developed by Alexa AI researchers, has 20 billion parameters and is capable of learning new concepts and transferring knowledge between different languages and tasks with little to no human intervention.
This represents a paradigm shift toward “generalizable intelligence,” which could radically transform the way we interact with our voice assistants.
Amazon’s commitment to AI
In recent years, Amazon has made significant advances in its voice assistant, Alexa, both through proprietary developments and strategic acquisitions.
In 2024, Amazon launched Amazon Q, a generative AI-powered assistant designed to improve employee productivity.
This wizard not only generates code, but also tests, debugs, and performs multi-step planning and reasoning tasks, which represents a breakthrough in process automation.
Already in 2023, Alexa improved its ability to maintain more natural and personalized conversations, learning from users’ habits and preferences.
In addition, Amazon invested $4 billion in the startup Anthropic, specializing in generative AI, and hired the co-founders of Covariant, obtaining non-exclusive rights to use its AI models.
New Amazon Alexa features
As for new features added, in 2022 Alexa added the ability to display real-time subtitles during calls and the option to share and manage shopping lists with other Alexa devices.
In 2023, custom routines were added to automate daily tasks and real-time call translation functionality.
In 2024, Alexa introduced Smart Briefing, which provides daily AI-generated news briefings tailored to the user’s interests, and a kitchen assistance feature, which suggests recipes based on available ingredients and guides step-by-step preparation.
During CES 2024, Amazon introduced new AI features for Alexa, including Character.AI, which allows you to interact with historical figures and other AI-generated profiles.
He also introduced Splash, which creates musical compositions based on voice instructions; and Volley Games, an AI-powered guessing game.
These features enrich the user experience, making interactions with Alexa more dynamic and personalized.
The power of multimodal AI
One of the areas where AI can make a significant difference is in the development of multimodal systems.
These systems are capable of processing different types of data, such as natural language, computer vision and speech recognition, to better understand the context in which they operate and solve various tasks more efficiently.
Amazon has shown clear interest in this direction. Rohit Prasad, the head of Alexa, has mentioned that the company is working on “ambient intelligence”, a concept that goes beyond simply responding to voice commands.
The idea is that Alexa can process a variety of sensory signals, including visual, tactile and ultrasound, in addition to speech.
This approach could allow voice assistants to offer a much richer and more personalized experience, proactively adapting to the user’s needs rather than simply answering questions.
Challenges and opportunities for voice assistants
Although the technology promises a lot, the path to revitalizing voice assistants is not without challenges.
One of the main problems these devices have faced is their limited ability to perform complex or personalized tasks.
Most users still use Alexa, Google Assistant or Siri for basic functions such as setting timers, playing music or checking the weather, which has led many to view these assistants as tools of limited usefulness.
Furthermore, the business model for these devices has proven to be unsustainable. In the case of Alexa, for example, Amazon only generates revenue when people actively use the devices, which has not been enough to offset development and operating costs.
Without a clear monetization model, it is difficult to justify continued investments in this type of technology.
However, the incorporation of generative AI could change this dynamic. With the ability to learn new tasks from a few examples and adapt to different contexts and users, voice assistants could become much more versatile and useful on a daily basis.
This could not only increase its value to users, but also open up new monetization opportunities, such as personalization of services or integration with other digital platforms.
This post is also available in: