Google Assistant is reportedly undergoing a significant transformation, shifting towards generative AI, as revealed in an internal email disclosed by Axios. The Assistant team leaders are recognizing the potential of a supercharged Assistant fueled by cutting-edge LLM (large language model) technology, prompting them to make organizational adjustments to pursue this vision.
The decision to explore generative AI doesn’t stem from merely wanting to experiment with new technologies. Rather, it appears that other companies have already demonstrated the power of generative AI in public, putting Google in a position where catching up is a top priority. The gradual transformation will unfold over the upcoming months as they fine-tune their approach.
While many chatbots and assistants have already integrated LLM technology, its practicality for this specific sector of technology remains unproven. In the past, services like Assistant, Alexa, and Siri acted more like Mad Libs, where users provided specific subjects and verbs to execute tasks, serving as simple but effective digital interfaces.
However, the question arises whether it’s truly an improvement when asking a simple question about the duration of a drive to the beach results in an answer influenced by the entirety of the Western canon. Although it may initially be amusing to request the weather forecast in the form of a sonnet, such novelty tends to wear off over time, akin to asking Alexa for a joke.
While LLMs possess the ability to engage in conversational threads, this capability may not be desirable when interacting with navigation systems or seeking recommendations for sushi joints. Perhaps, a more suitable approach would be to develop an interface that can seamlessly handle both types of interactions and access the advanced capabilities as needed. Google is positioning itself strategically, ensuring that it is prepared for whichever direction the technology evolves.
