X

At Google I/O, Gemini Really Wants to Talk With You

Gemini is getting an upgrade to being more conversational and having greater app integration.

Imad Khan Senior Reporter
Imad is a senior reporter covering Google and internet culture. Hailing from Texas, Imad started his journalism career in 2013 and has amassed bylines with The New York Times, The Washington Post, ESPN, Tom's Guide and Wired, among others.
Expertise Google | Internet Culture
Imad Khan
3 min read
Gemini AI

Google Gemini logo on phone.

James Martin/CNET

Google Gemini, the search giant's competitor to ChatGPT, is getting a major upgrade allowing it to be more conversational, along with increased cross-app compatibility and greater intelligence overall, the company said at its annual Google I/O developer's conference on Tuesday. 

Early access to Gemini 1.5 Pro is coming to Gemini Advanced subscribers. Like Anthropic's Claude or OpenAI's ChatGPT, Google's Gemini too has a subscription tier for its AI chatbot with more advanced features, costing $20 per month. Google says Gemini 1.5 Pro has the ability to analyze data, can better understand images and has a context window of 1 million tokens. 

Watch this: Google Introduces Gemini AI Upgrades to Gmail and Chat

When an AI chatbot analyzes a sentence, it breaks down and parses each word and punctuation. The more tokens an AI chatbot has, the more text it can analyze. With 1 million tokens, Gemini can parse through a 1,500-page document with 30,000 lines of text. Google says it'll double the context window to 2 million tokens later this year.

Gemini Live is a new conversational experience coming to Gemini Advanced subscribers. Gemini Live will allow Advanced subscribers to talk to Gemini directly via the mobile app on iOS and Android. It uses advanced speech models allowing you to converse with Gemini more naturally. For example, you can interrupt Gemini mid-conversation and it'll adapt, changing its speech patterns. Later, Google plans to release an update allowing you to use your camera to have Gemini respond to your surroundings in real time.

Google's large feature rollout for Gemini comes as OpenAI continues to dazzle with more advanced and sci-fi-esque features. Yesterday, OpenAI revealed a new version of its AI engine called GPT-4o. Like Gemini 1.5 Pro, GPT-4o can also be interrupted mid-sentence to create a more natural-sounding experience. Unlike Gemini 1.5 Pro, GPT-4o is available to both free and paid subscribers, although the paid version has five times greater capacity. 

Today's Gemini announcements gives Google the opportunity to reset the conversation around its AI chatbot, which has been plagued with snafus since launch. Early on, Gemini, then named Bard, got a key piece of information wrong in its initial presentation, leading to a plummeting stock price. Earlier this year, Google had to disable the image generation in Gemini after it began depicting historical figures, like the pope or German Nazis, as women or people of color. 

AI Atlas art badge tag

As Big Tech and Wall Street rally around AI, with players like Meta, Nvidia and Microsoft seeing record stock prices, those who bring the best AI products to market will see the greatest valuations.

Gemini will also see deeper integration into Google's suite of apps like Gmail, Maps and YouTube Music. Google says Gemini Advanced users will be able to create more complex trip itineraries in the coming months, taking into account the travel time between sights along with restaurant and museum recommendations. Gemini will later see integrations with Calendar, its Tasks management app and its Keep notes app. YouTube Music integration rolls out today. 

Like ChatGPT's custom GPTs, Google too is allowing you to create customized versions of Gemini called Gems. You can tell Gemini to create a personalized training coach to or a patient math tutor and it'll oblige.