Google's Voice Assistant to Get Generative AI Boost

Google just might have saved its voice assistant. According to a report by Axios, the tech conglomerate is planning to bring generative AI technologies similar to those that power ChatGPT and its own Bard chatbot to its voice assistant and make it 'supercharged'. 

However, Google hasn’t confirmed the development at the time of writing this article. 

As part of this move, the company has decided to trim "a small number of roles" from the teams working on its Assistant application. According to the report, ​​the move will involve eliminating dozens of jobs from the NLP team and transferring the responsibility to the generative AI team.

Google hasn’t specified the features it intends to introduce to Assistant, leaving room for some exciting possibilities. “We remain deeply committed to Assistant and we are optimistic about its bright future ahead,” Peeyush Ranjan, the vice president of Google Assistant, and Duke Dukellis, the company’s product director, wrote in an email to the team. 

Change of heart

Last year, a report said Google has shifted focus away from its voice assistant to Pixel phones and plans to invest less in developing its Google Assistant voice-assisted search. 

According to reports, Amazon also shifted its focus from Alexa. Earlier this year, Amazon employees faced layoffs in the Alexa division as a result of a substantial operating loss in its hardware unit during the previous year. Similarly, Microsoft's efforts to popularize its Cortana virtual assistant have not yielded the desired success.

However, with the sudden rise of ChatGPT, the desire for a  personalised AI assistant is increasing day by day which can perform petty tasks like ordering grocery and food, taking notes, writing a mail and several other tasks that involve human thinking. 

A voice assistant with its own intelligence can be a game changer. For example, you can ask for food recipes based on the ingredients you have at your home. As most of us regularly order food, groceries, and essential items online, we tend to have our favourite restaurants and shops. If large language models (LLMs) can remember our preferences, it would save a lot of time since we won't have to scroll through the menu and manually add items to the cart.

We would just be able to give voice commands and our order will be placed. Similarly, if we want to send a mail on the go, we can just ask the voice assistant to draft a mail. Instead of going through apps and websites manually, these tasks will become automated and much more convenient. Voice assistants will make various tasks seamless and efficient, allowing us to accomplish things with just a few simple spoken instructions.

LLMs are the future of voice assistants

After experiencing ChatGPT and Bard, it is very difficult to go back to voice assistants like Alexa, Siri and Google Assistant. Voice assistants appear insignificant once we get to know the capabilities of LLM chatbots. That's what happened with Alexa. During its launch in 2015, Alexa experienced a surge of users asking curious and quirky questions, covering everything from the meaning of life to playful wishes. However, as time went by, users became disinterested in Alexa. The devices lacked the capability to provide personalised experiences or generate substantial advertising revenue for the company, causing a gradual decrease in user engagement with Alexa over time. 

Earlier this year, Insider reported that Amazon is looking to add ChatGPT-like generative AI features to Alexa. Large language models, using lots of web data, can sound more human in conversations with people than voice assistants like Alexa and Google Assistant, which were built mainly for natural-language understanding.