Voice assistants hold immense promise. Since Apple’s Siri and Amazon’s Alexa first entered our lives over a decade ago, their most compelling use remains setting timers. Despite competition from Google’s Assistant and Samsung’s Bixby, innovation in this space has stalled.
These assistants frequently misinterpret commands, exhibit poor listening skills, and fail to fulfill their potential as proactive, intelligent digital helpers.
Enter Generative AI
Generative AI technology promises to transform voice assistants from novelty items to necessary tools. Apple recently announced at its Worldwide Developers Conference plans to infuse Siri with generative AI, giving it crucial skills in context and conversation. This development could finally deliver on the original promise of smart digital assistants.
Apple’s new initiative, Apple Intelligence, aims to provide Siri with “all-new superpowers,” including improved language understanding, personal context awareness, and the ability to take action across apps.
For instance, you could ask, “Siri, what time does Mom’s flight land?” The assistant would then search your mail and messages for the information. A follow-up question like, “How long will it take me to get there?” would prompt Siri to pull up a route and ETA via Maps.
Simplifying Commands
The new Siri aims to address fundamental issues with voice assistants, such as requiring precise commands. Instead of needing to say, “Siri, set a timer for 10 minutes,” you could use a more casual phrase like, “Siri, set an alarm for—ooh, wait, no, set a timer for 10 minutes.
Actually, make that 5,” and the assistant will understand. We designed these improvements to make interactions more natural and less frustrating.
The Power of Generative AI
Current voice assistants, including Siri and Alexa, mimic human intelligence through command-and-response programming and machine learning. However, generative AI and large language models (LLMs) could enable voice assistants to generate responses based on what they’ve learned, rather than just reacting with existing knowledge. This could lead to more conversational and smarter voice assistants, significantly enhancing their usefulness.
Challenges and Caution
Despite the potential, creating a superintelligent voice assistant is challenging and risky. Voice assistants can control devices and services in our homes and phones, making it crucial to ensure they don’t make errors. Apple has carefully sandboxed its ChatGPT integration with Siri to prevent issues.
Amazon is also developing a new, improved Alexa. The company has integrated generative AI into parts of Alexa, but the new version is still in development. Last fall, Amazon announced a “smarter and more conversational Alexa” powered by a new Alexa LLM, promising a better understanding of conversational phrases and context. However, this superpowered Alexa remains in limited preview as the company works to merge old and new capabilities.
Slow and Steady Approach
Apple’s new Siri will launch in the fall, initially as a beta version. It won’t be available on Apple’s voice-forward home devices like the HomePod and Apple TV, nor on the Apple Watch, likely due to processing power limitations. However, the smart home is a key area for more intelligent voice assistants, offering opportunities to bridge personal and home spaces and simplify smart home management.
Dave Limp, the former head of devices and services at Amazon, mentioned that the new Alexa LLM has received training on hundreds of smart home APIs. This could enable Alexa to proactively manage smart home devices like lights and thermostats. In contrast, Apple’s new Siri will operate locally for privacy purposes, potentially necessitating updated hardware for smart home use.
Looking Ahead
Although a new era of voice assistants is on the horizon, significant changes will take time. Companies like Google are also exploring new avenues. Google Assistant has yet to undergo a major AI overhaul, with the company focusing on its new AI-powered Gemini assistant. We remain to see if Google will integrate this with its current assistant or launch a new one.
The future of smart voice assistants is promising, especially for companies that successfully merge personal and home assistant capabilities. Imagine a HomePod welcoming you home with personalized updates, alerting you to traffic delays, and ensuring your EV is fully charged and ready. This vision of smart assistants is far more advanced than merely setting timers, marking an exciting evolution in voice technology.
The information is taken from The verge and The New York Times