Google I/O 2024 Top Highlights: Major upgrades to Gemini 1.5 Pro, New models, Gen AI for search & More

Ayushi Trivedi 15 May, 2024 • 6 min read


The inference of Google I/O, the tech giant’s annual conference, has left the world buzzing with excitement over a plethora of groundbreaking announcements and technological marvels that promise to redefine the landscape of innovation. From significant strides in artificial intelligence to ingenious enhancements in Google’s product ecosystem we will look into all the launches revealed in this event.

Major Feature Update: A Deeper Dive into Google’s Cutting-Edge Offerings

Gemini 1.5 Pro

Sundar Pichai, Google’s CEO, announced an exciting update to their AI model, Gemini. The latest version, Gemini 1.5 Pro, is now available to all users via Gemini Advanced. This powerful AI tool has a context window of 1 million tokens, allowing users to upload and analyze vast amounts of data, be it text or video. With this update, developers and enthusiasts alike can access Gemini’s advanced capabilities.

Additionally, Pichai shared that an even more powerful version, Gemini 1.5 Pro with a context window of 2 million tokens, will be available for developers in a private preview. This exclusive offering boasts the longest context window in the world, enabling complex and in-depth analysis. Users can now upload 30,000 lines of code or one hour of video content onto Gemini Advanced, unlocking new possibilities for AI-powered insights and creations.

“Ask Photos” Feature

Google Photos enthusiasts have reason to rejoice with the introduction of the “Ask Photos” feature, unveiled by CEO Sundar Pichai during the keynote. This innovative feature harnesses the power of artificial intelligence, powered by Gemini, to revolutionize the search experience within Google Photos. Unlike traditional keyword-based searches, “Ask Photos” enables users to pose natural language queries, such as “What’s my license plate number again?”.

When users input such queries, Google Photos utilizes advanced AI algorithms to analyze contextual clues, location data, and historical information associated with the user’s photo library. This comprehensive analysis allows Google Photos to swiftly retrieve relevant images that match the query, providing users with precise and intuitive search results. For instance, querying about a license plate number prompts Google Photos to search past captures. It identifies images containing vehicles with visible plates.

Scheduled for release in the summer, the “Ask Photos” feature promises to redefine the way users interact with their photo libraries. By enabling natural language queries and leveraging AI-powered search capabilities, this feature enhances convenience and efficiency, allowing users to effortlessly rediscover and organize their memories with ease. “Ask Photos” marks a significant stride in Google’s quest to integrate advanced AI for superior user experiences. It’s a testament to Google’s commitment to innovation.

AlphaFold 3

DeepMind’s AlphaFold 3 has emerged as a revolutionary advancement in the field of molecular biology and drug discovery. AlphaFold 3 is an AI-powered model that specializes in predicting the three-dimensional structures of proteins, which play crucial roles in various biological processes within the human body. Proteins, complex molecules, fold into intricate shapes essential for deciphering their functions and interactions. Understanding their structures is crucial.

What sets AlphaFold 3 apart is its unparalleled accuracy and efficiency in predicting protein structures. By harnessing the power of generative AI algorithms, AlphaFold 3 can accurately predict the 3D structures of proteins with remarkable precision. This capability provides scientists with invaluable insights into the molecular mechanisms underlying diseases and biological processes. Accurate protein structure prediction is crucial in drug discovery, as diseases like cancer and Alzheimer’s are caused by specific protein abnormalities. Understanding these structures allows scientists to design targeted drugs for effective treatment.

AlphaFold 3 accelerates protein structure prediction, accelerating drug discovery efforts. This capability enables researchers to identify potential drug targets more efficiently. It streamlines the development process and opens up new possibilities for discovering novel treatments for diseases. This groundbreaking technology has the potential to revolutionize molecular biology and drug discovery.

Latest Innovations of Google I/O

Here’s an in-depth exploration of the pivotal launches and revelations unveiled during this momentous event.

Project Astra

At Google I/O, the tech giant unveiled Project Astra, a revolutionary leap forward in AI assistant technology. This cutting-edge project promises to redefine the future of virtual assistants by harnessing the power of smartphone-captured video and advanced voice recognition algorithms. Through seamless integration with Google’s ecosystem, Project Astra delivers contextually relevant responses to user queries, providing a truly personalized experience.

During an engaging demo at Google I/O, Project Astra showcased its remarkable capabilities by assisting users in real-time tasks. From aiding in coding conundrums through visual analysis to helping locate misplaced items like glasses, Project Astra demonstrated its versatility and potential to streamline daily tasks. Google’s commitment to innovation in AI is evident through Project Astra. It signifies Google’s dedication to enhancing user experiences.

Google I/O

Elevating AI Capabilities with Gemini 1.5 Pro and Gemini Flash

Google unveiled Gemini 1.5 Pro and Gemini Flash, both aimed at advancing AI processing capabilities. Gemini 1.5 Pro emerges as a powerhouse iteration tailored to meet the complex demands of developers worldwide. Boasting an expanded token capacity, Gemini 1.5 Pro offers developers access to a 2 million token context window. It effectively doubles the model’s capacity compared to previous versions. This enhancement empowers developers with advanced tools and functionalities, allowing them to tackle complex challenges with unparalleled precision and efficiency.

Gemini 1.5 Pro features a multi-model architecture and a long context window. This allows it to process diverse data types and retain contextual information, enhancing performance across domains.

In contrast, Gemini 1.5 Flash prioritizes lightning-fast performance without compromising on accuracy or reliability. While specific details regarding token capacity in Gemini Flash are not disclosed, its primary focus lies in delivering rapid execution of tasks across various domains. This iteration of Gemini is ideal for scenarios where speed and agility are paramount, such as real-time data analysis, instant decision-making processes, or high-frequency trading environments. Gemini Flash, despite its lightweight design, maintains versatility and multi-model capabilities, ensuring seamless integration with existing AI workflows.

Google I/O

Expanding Content Protection with SynthID

The expansion of SynthID, Google’s watermarking tool, represents a pivotal advancement in content protection strategies. SynthID now extends its reach beyond images to encompass text and video, equipping content creators with a comprehensive solution to safeguard their intellectual property across a wide array of media formats. SynthID embeds unique identifiers into content, allowing creators to assert ownership and deter unauthorized use. This reinforces Google’s commitment to combating digital piracy and enhancing online security.

This expansion of SynthID signifies a proactive approach by Google to address the evolving challenges of intellectual property protection in an increasingly digital landscape. By offering a versatile and robust toolset, Google empowers content creators to assert control over their creations and preserve the integrity of their work across diverse mediums. As digital content proliferates, SynthID aids in combating copyright infringement and content piracy. It reflects Google’s commitment to supporting creators and intellectual property rights.

Seamless Integration of Gemini Across Google Services

Gemini’s integration into various Google products signifies its versatility and transformative capabilities. By enhancing search functionalities in Google Photos, Gemini enables users to swiftly locate and organize their images with unprecedented ease and efficiency. Moreover, its integration bolsters security measures on Android devices, providing real-time protection against fraudulent calls and ensuring a safer mobile experience. Gemini integrates AI solutions across the Google ecosystem, enhancing user experiences. It improves productivity and convenience in daily tasks.

Upholding Ethical AI Principles

Google reaffirmed its commitment to ethical AI development by implementing innovative strategies like Red Teaming. Through Red Teaming, Google aims to proactively identify and address potential vulnerabilities and instances of misuse within AI technologies. Google rigorously tests and challenges existing systems to ensure responsible AI deployment. This fosters trust and upholds ethical conduct among users.

Charting the Course for Future Innovation

Looking ahead, Google outlined a visionary roadmap for AI-powered experiences, unveiling upcoming innovations like Gemini Live and VideoFX. Gemini Live introduces a new realm of interaction with AI through voice commands. Meanwhile, VideoFX offers creative expression through high-resolution video generation, setting the stage for a future driven by immersive, intelligent experiences.


Google I/O’s conclusion heralds a new era of innovation. Google prioritizes ethical principles in technological advancement. With these monumental launches and developments, Google cements its position as a vanguard of artificial intelligence. Therefore, charting a course towards a future where technology serves as a catalyst for positive societal transformation and human advancement.

Follow us on Google News to stay updated with the latest innovations in the world of AI, Data Science, & GenAI.

Ayushi Trivedi 15 May 2024

Frequently Asked Questions

Lorem ipsum dolor sit amet, consectetur adipiscing elit,

Responses From Readers