Unveiling the Future of Interaction: Google’s Gemini Live Sets a New Standard in AI Conversations

To return to this article, go to My Profile and then click on Saved stories.

What is Gemini Live and Its Functionality?

This week, Google unveiled an array of new devices, including the Pixel 9 smartphones and the latest wireless earbuds. All these devices are powered by Google's Gemini, an AI-driven assistant. Introduced at the beginning of the year, Gemini has become the primary assistant for the Pixel 9 lineup and is currently installed on millions of Android devices globally. Now, there's a fresh method to interact with this AI chatbot known as Gemini Live, which is gradually being introduced.

Google has unveiled its answer to OpenAI's GPT-4o, introducing a feature that allows users to interact with its assistant in a more conversational manner, aiming for the fluidity of a regular human dialogue. This feature is now available to Gemini Advanced members, who pay a monthly fee of $20, by clicking the small Live icon situated in the lower right corner of the Gemini application. Plans are in place to expand this offering to the iOS application and include additional languages in the near future.

Sissie Hsiao, who holds the position of vice president of Gemini experiences at Google, shared with WIRED that the company's new chatbot is far from being just another version of the Google Assistant. This new interface has been entirely redeveloped from the ground up with generative AI technology. Hsiao highlighted that in the process of developing the Assistant, they consistently received two main requests from users. “Firstly, users have been seeking an assistant that communicates more smoothly and naturally, allowing them to converse in their usual manner without needing to adjust their speech. Secondly, they're looking for an assistant that's more adept at tackling complex issues in their lives, beyond just handling straightforward tasks,” explained Hsiao.

Broadcasting Directly from Google

Initiate the Gemini app, and you're greeted with a pristine interface, illuminated by a soft light emanating from its lower edge. Engage with the assistant without even unlocking your phone or activating the screen. This feature extends to Google's latest Pixel Buds Pro 2, enabling you to interact without using your hands, even when your phone is stowed away. You have the liberty to select from 10 distinct voices, each offering a unique tone, accent, and delivery. Upon concluding your interaction, the app presents a complete transcript of your dialogue, a record you can revisit anytime within the Gemini application.

Gemini Live revolutionizes the way voice assistants work by allowing users to interject during conversations, making for a smoother interaction. This feature proves particularly beneficial given Gemini's tendency to be rather verbose. The platform plans to enhance user experience further by integrating with various applications through add-ons, although many of these integrations are still in development. For instance, users will soon be able to request Gemini Live to retrieve a party invite from Gmail, inquiring directly about its time and place, avoiding the hassle of manual search. Additionally, it could facilitate recipe searches and prompt the addition of necessary ingredients to a Google Keep shopping list. Google has announced that these integrations with its applications, including Keep, Tasks, Utilities, Calendar, and YouTube Music, are expected to roll out in the near future.

As the year progresses, Google plans to integrate Project Astra, its computer vision technology showcased at the May developer conference, into Gemini Live. This enhancement will enable users to utilize their smartphone's camera application to interact with and inquire about real-world objects in live time. For instance, if you come across a concert flyer, you could request Gemini to save the event dates in your calendar and to remind you to purchase tickets.

Conversing with AI

Up until now, our interactions with voice assistants have mostly been functional and to-the-point. However, engaging with Gemini Live presented a new dynamic, as starting a dialogue with the AI felt somewhat uneasy. This was a significant leap from simply requesting a weather update, curtain adjustments, or dietary advice for pets from services like Google Assistant or Alexa. Those interactions might include an occasional additional query, but they weren't designed for the continuous exchange that Gemini Live aimed to provide.

Hsiao shared with me her fondness for engaging with Gemini Live while commuting back from her job. She initiated a dialogue on the topic of the Paris Olympics and Celine Dion's performance at the inauguration. "Could you provide me with some details regarding the track she performed?" inquired Hsiao. The AI supplied information about the song's background, its composer, and its significance. Through this exchange, Hsiao learned about Celine Dion's ability to perform in Chinese.

"She expresses her astonishment, explaining, "This really showcases the way in which we can discover new things; it's a form of engagement with technology that wasn't possible prior to this era of inquiry and dialog. This instance merely scratches the surface of the potential that lies ahead with the Gemini assistant."

During my demonstration, I queried Gemini about what I should have for dinner. It inquired whether I was in the mood for something light or a more substantial meal. Our conversation continued, with suggestions being made back and forth. When Gemini proposed a shrimp-based meal, I falsely claimed to have a shrimp allergy, prompting it to then suggest salmon. I mentioned that I didn't have any salmon available. "Perhaps you could grill some chicken breasts and mix them into a salad with grilled vegetables and a light vinaigrette," it proposed. I requested a recipe, and it began to detail the process step by step. Although I interrupted it, I realized I could revisit the Gemini application later to retrieve the recipe.

I can envision adopting this method going forward whenever I'm curious about a topic, and keeping the dialogue going even after Gemini provides a response to my first question. However, several worries persist: Why doesn't it directly cite or reference the sources of its information? How can I be sure of the reliability of its statements? Hsaio mentions that upon leaving Gemini Live, you have the option to click on the small "G" icon beneath the transcribed text to verify its accuracy and conduct your own searches on Google.

Increasingly, I'm convinced that this represents the next wave in how we search for information. The idea is simple: you pose a question, receive your answers, and continue the conversation to deepen your understanding. However, the issue with Gemini is its tendency to provide lengthy answers, leading to significant wait times before you can ask additional questions. While it's possible to cut it off and proceed, interrupting a voice assistant feels somewhat impolite. I'm not keen on being discourteous!

What's the Current Location of Google Assistant?

Amidst the buzz around Gemini and Gemini Live, you might be curious about the whereabouts of Google Assistant. Clicking on your profile symbol within the Gemini application reveals a choice to revert to Google Assistant for those longing for the familiar interface, though it's uncertain how much longer this feature will persist. At present, Google Assistant possesses capabilities not yet available in Gemini, facilitating a transition between the two. "Gemini is progressively gaining the ability to perform these tasks independently," notes Hsiao.

Earlier in the month, Google unveiled its latest Nest offerings, alongside news that Google Assistant is set to receive a more lifelike voice, with several of its functionalities enhanced by Gemini's advanced language models. This upgrade means you could inquire whether a FedEx courier has arrived at your door, and it would be capable of interpreting this information from your video doorbell's recordings. Additionally, motion notifications are expected to become much more detailed, moving beyond simple alerts like “person detected.”

This indicates that we currently possess dual assistant services, and it appears Google is entirely comfortable with this situation for now. According to Hsiao, Gemini will serve as your individual assistant, managing tasks such as checking your calendar and email invitations, all based on your personal information. Meanwhile, Google Assistant will act as the household's shared assistant, given its role as a device for the whole family. "Individuals are not keen on the idea of having their private emails accessible via voice command on a communal speaker in the living room, where a visitor could potentially inquire, 'Hey Google, what is in Julian's email?'"

This seems like a formula for a marketing nightmare. Keeping tabs on the numerous versions of Gemini already in circulation is challenging enough (remember, Gemini debuted as “Bard” in its preview phase last year). Additionally, this could result in some features being restricted depending on the device being used, as a measure to stop a visitor from peeking at your emails. Imagine the annoyance if you become accustomed to commanding Gemini on your smartphone to perform a function, only to have the Assistant on your Nest Speaker decline to execute the task when you leave your phone in a different room.

"Hsiao mentions that they are in the initial stages of determining the brand identity. Setting branding concerns aside, the priority is to ensure that the assistant meets user needs effectively, whether it's being used on personal mobile devices or within the home environment, addressing their specific requirements."

Image Credit: Julian Chokkattu

Purchasing through the links in our articles may result in us receiving a commission. This contributes to the funding of our journalistic work. Find out more.

Explore Further…

Dive into the World of Politics: Subscribe to our newsletter and tune into our podcast

Exploring the Outcomes of Providing Individuals with No-Cost

Not everyone experiences weight loss with Ozempic.

The Pentagon is planning to allocate $141 billion towards a catastrophic device.

Gather with us at the Energy Tech Summit happening on October 10th in Berlin.

WIRED OFFERS

Exclusive Dyson Airwrap Promotion: Complimentary $60 Case and $40 Gift Bonus

Enjoy Up To An Additional 45% Discount on Our August Sale

Discount Code for Vista Print: Save 20% on Certain Signs

Discount Code for Newegg: Save 10%

Subscription Plans for Peacock Premium Begin at Only $7.99 Per Month.

Discover DJI's Academic Discounts and Learning Deals for 2024

Additional Content from WIRED

Evaluations and Instructions

© 2024 Condé Nast. All rights reserved. Purchases made through our website may generate a commission for WIRED as part of our affiliate agreements with various retailers. The content found on this website is protected and cannot be copied, shared, broadcasted, stored, or utilized in any form without the explicit written consent of Condé Nast. Advertisement Choices

Choose a global website