5 Jaw-Dropping Google I/O Reveals That Point to a Sci-Fi Future

At I/O 2025, Google is showcasing the many improvements to its many AI initiatives that promise science-fiction-like conveniences in the very near future. Sure, we aren’t getting Jetsons-style flying cars, but a Rosie the Robot-style assistant who helps with online tasks will be in your home sooner than expected. Google promises AI integrations in Chrome, agentic digital agents that handle online tasks for you, and many more fantastic tools that are poised to change how we communicate, create, and generally get things done. These are the top five upcoming AI-related features that wowed me at Google I/O.

1. Project Astra: Your Universal Personal Assistant

Project Astra was teased last year as a next-generation AI assistant, and part of the larger Google DeepMind initiative. This “universal AI agent” is designed to understand the world around you and interact with it via various multimodal inputs. Google demonstrated its impressive AI capabilities, showing a user requesting assistance with a bike repair. Astra scoured the internet to locate a PDF manual and a YouTube tutorial, and even called a local bike shop to inquire about a part, seemingly without further user interaction beyond voice commands. It can process information from many sources, interact with apps, and manage multiple tasks while maintaining a natural conversation with you.

I consider myself a fairly social person, but I hate making appointments or reservations, regardless of whether they’re done online or via telephone. An AI-powered assistant to handle those tasks would be a tech godsend. I could also see myself using it in more everyday circumstances, too. It would be a great tool for finding recipes when I’m bustling around in the kitchen. Project Astra is coming later this year.

2. Gemini in Google Chrome Helps You Manage the Web

Google’s Gemini chatbot is coming to the Chrome browser, giving ChatGPT fierce competition. Gemini is currently accessible via a dedicated website and mobile app, but this new integration will enable unique functionalities. You’ll be able to ask questions about the current web page you’re viewing, summarize content, clarify details, retrieve information from past browsing sessions, and engage with Gemini through voice.

Gemini integration while browsing has enormous productivity and accessibility potential. Summarization is nice on its own, but Gemini clarification could assist with your understanding and recontextualize what you’re reading. Accessing information from multiple tabs without switching between them can keep you focused on other tasks, too. I see myself using Gemini to brainstorm ideas when browsing or shopping, picking the AI’s brain about whatever I’m reading on the page, or fact-checking the copious amounts of information (and misinformation) we parse through every day.

Gemini in Chrome is available starting May 21, but only for Google AI Pro and Ultra subscribers ($19.99 per month and $249.99 per month, respectively). It’s also available in the Chrome Beta, Dev, and Canary channels.

Project Mariner is a smart assistant for the Chrome browser available to Google AI Ultra subscribers that can actively browse the internet to perform tasks on your behalf, such as looking up information, conducting research, booking flights and reservations, and making purchases (with permission). Unlike Project Astra, Project Mariner is meant to operate specifically within web browsers, whereas Astra is more general-purpose and multimodal.

Google’s Project Mariner demo was quite impressive. During Google I/O, Project Mariner was used to search for real estate listings, make reservations, and even help with shopping. As someone thinking about purchasing a home in the near future, my mind boggles at the thought of having an AI assistant comb the internet on my behalf, finding prospective properties for me to compare based on criteria I assign to it. I’m a sucker for deals too, so having the tool search for the best prices, adding it to a shopping cart, and sending it to me for approval sounds like a proper step toward the future.

4. Google’s Many Enhanced Content Creation Tools

Google has updated two of its most potent creation tools: Imagen and Veo. Imagen 4 (available in Gemini) has 2K resolution image support and enhanced texture rendering for fabric, fur, and water. The goal? More photorealistic images. It also has Google Workspace integration for apps like Docs, Slides, and Vids, letting you generate images directly within workflows. If filmmaking sounds more like your cup of tea, Google launched Veo 3 (available with a $249.99-per-month Google AI Ultra subscription). It offers shockingly polished AI video creation tools that create life-like images.

AI-generated content creation is a divisive subject, but from a consumer perspective, I like the idea of powerful creation tools at my fingertips. Granted, they’re not readily available to the public yet. Still, the potential is huge, and given time, I expect many of these features to trickle down into consumer hands soon enough.

5. Google Meet’s Real-Time Speech Translation

Real-time voice translation has always fascinated me; it’s future tech that seemed within our grasp yet just beyond our fingertips. Google is developing near-real-time speech translation for Google Meet, aiming for natural conversations. The current beta version supports English-to-Spanish translation. More languages are planned for release in the coming weeks. This builds on Google’s previous experiments with live video-chat translation, as demonstrated at their I/O event.

As someone with family across the globe, these Google Meet improvements sound phenomenal. My Spanish is incredibly rusty, so this improved functionality can help me communicate much more easily with my relatives in Europe and South America. Likewise, as new languages are added, I can also start connecting with my non-English speaking family in Germany and Italy. This AI-powered feature requires a Google AI Pro subscription ($19.99 per month).

About Gabriel Zamora

My career has taken me through an eclectic assortment of fields, and connected me with people from all walks of life. This experience includes construction, professional cooking, podcasting, and, of course, writing. I’ve been typing up geeky takes since 2009, ultimately landing a freelancing position at PCMag. This blossomed into a full-time tech analyst position in 2021, where I lend my personal insight on the matters of web hosting, streaming music, mobile apps, and video games. 

Read Gabriel’s full bio

Read the latest from Gabriel Zamora