Google’s annual developer conference, I/O 2024, has come to a close, and one theme dominated the event: artificial intelligence. CEO Sundar Pichai set the tone from the start, describing the keynote as Google’s version of “The Eras Tour,” specifically focusing on the “Gemini Era.” Over the course of the two-hour presentation, Google mentioned AI a staggering 121 times, underlining the company’s commitment to this transformative technology.
From the unveiling of the futuristic AI assistant “Project Astra” to the infusion of Gemini into nearly every Google service and product, it’s clear that AI is no longer a mere buzzword but a driving force shaping our digital lives. Let’s take a closer look at the seven most significant AI announcements from Google I/O 2024.
-
Project Astra: Google’s Answer to GPT-4o and CoPilot
Perhaps the most impressive reveal of the event was Project Astra, an “AI agent” designed for everyday life. Essentially Google Lens on steroids, Astra can understand, reason, and respond to live video and audio in real-time. During a demo on a Pixel phone, a user walked around an office, providing a live feed from the rear camera while asking Astra various questions. Gemini, the AI backbone of Astra, quickly processed the visuals and delivered accurate responses, showcasing its multi-modal and long-context capabilities.
Astra’s potential applications are vast, from identifying specific parts of a speaker to recognizing neighborhoods in London. Its generative abilities were also on display, as it swiftly created a band name for a cute pup next to a stuffed animal. While Astra won’t be rolling out immediately, developers and press will have the opportunity to test it at I/O 2024. Interestingly, Google teased the possibility of Astra being integrated into glasses, hinting at a potential comeback for Google Glass.
- Google Photos Gets an AI Boost from Gemini
Google Photos is set to become even more powerful thanks to a major update that fuses it with Gemini. This integration allows users to quickly find specific photos captured in the distant past, such as a note from a loved one, an early photo of a dog as a puppy, or even a license plate. In a demo, Sundar Pichai showed how asking for a license plate number would prompt Photos to deliver an image displaying the plate and its characters.
The new “Ask Photos” feature will roll out to all users in the coming weeks, making even the most disorganized photo libraries more accessible and searchable. This development is sure to make Google Photos an even more compelling choice for users looking to manage their ever-growing image collections.
-
NotebookLM: Simplifying Homework with AI
Parents everywhere will rejoice at the news of NotebookLM’s upgrade, which now includes access to Gemini 1.5 Pro. The demo at I/O 2024 showcased how NotebookLM can create detailed learning guides, quizzes, and FAQs from a notebook filled with notes on a specific topic. This AI-powered feature has the potential to be a better teacher than many parents, making the task of helping kids with homework much less daunting.
Perhaps the most impressive aspect of the upgraded NotebookLM is its ability to output all of the content as audio, creating a podcast-style discussion that’s both engaging and informative. The AI can even respond to interruptions and questions, adapting its explanations to use more accessible metaphors and context. While still a prototype, this feature has parents on the TechRadar team eager to try it out.
-
Searching Google with Video
In an unusual on-stage demo involving a record player, Google showcased an impressive new search trick: the ability to search using a video. By recording a video and asking a question, users can now obtain search results and, hopefully, an answer. This feature is essentially Google Lens for video, differing from Project Astra in that it requires recording and searching rather than working in real-time.
This video search capability is part of a broader infusion of Gemini and generative AI into Google Search, aiming to keep users on the page and make it easier to find answers. Before the video search demo, Google showcased a new generative experience for recipes and dining, allowing users to search using natural language and receive recipes or eatery recommendations directly on the results page.
-
Veo: Google’s Answer to Sora
Google is entering the generative video arena with Veo, its answer to OpenAI’s text-to-video tool Sora. Like Sora, Veo can generate minute-long videos in 1080p quality from a simple prompt, which can include cinematic effects like time-lapses or aerial shots. The early samples are impressive, and users can also upload an input video with a command, allowing Veo to edit the clip to match the request. Masks and tweaks to specific parts of a video are also possible.
While Veo isn’t widely available yet, Google has stated that it will be accessible to select creators through VideoFX, one of its experimental Labs features, in the coming weeks. A broader rollout may take some time, but Google has committed to bringing the feature to YouTube Shorts and other apps, which may have Adobe feeling a bit uneasy.
-
Android’s Gemini Infusion
Gemini is being integrated into the core of Android, allowing it to understand and interact with what’s on your phone’s screen. As demonstrated, Gemini can view, read, and comprehend the content of your device, enabling it to anticipate questions about whatever you’re viewing. This means it can provide context for a video you’re watching, summarize a lengthy PDF, or be ready to answer various questions about the app you’re using.
Additionally, Gemini Nano with Multimodality will launch later this year on Pixel devices, promising faster performance and more on-device processing. One standout feature is Gemini’s ability to listen to calls and alert you in real-time if it’s spam, building upon the existing call screening functionality of Pixel phones.
-
Smarter Google Workspace with Gemini
Google Workspace users are in for a treat, as Gemini integrations and features are set to make a significant impact on their daily workflows. In Gmail, a new side panel on the left allows users to ask Gemini to summarize recent conversations with a colleague, providing bullet points highlighting the most important aspects. Gemini in Google Meet can offer meeting highlights or summarize questions from other participants, eliminating the need for manual note-taking. In Google Sheets, Gemini can help make sense of data and process requests like pulling specific sums or data sets.
The introduction of the virtual teammate “Chip” is perhaps the most futuristic example of Gemini’s integration into Workspace. Chip can reside in a G-chat and be summoned for various tasks or queries. While these tools will initially be available through Workspace and Labs, it’s likely only a matter of time before they make their way to regular Gmail and Drive customers, given Google’s commitment to making AI accessible to all.
The Google I/O 2024 keynote made it abundantly clear that the AI era is upon us, with Google leading the charge in integrating artificial intelligence into our everyday lives. From the groundbreaking Project Astra to the numerous Gemini-powered enhancements across Google’s services and products, the future of technology is undeniably intertwined with AI. As these innovations continue to evolve and become more widely available, it’s evident that the way we interact with our devices and consume information is on the cusp of a profound transformation.
