Google I/O is one of the most viewed technology conferences in the world. Like most technology conferences in the past year, AI was the elephant in the room at Google I/O 2024 as well. The embrace of the “Gemini Era” by Google was evident in almost all its products and services.

The Gemini Revolution

Even though Google was a bit late into the world of GenAI, Gemini was a game changer for them. Gemini has become even more powerful by making advancements in multimodality and context.

• Multimodality: Google developed Gemini in such a way that it can understand text, images, video, code, audio, and more. Now the Gemini has become more context-aware, meaning that it can process different kinds of data formats and understand their relationships.
• Long context: Everyone loves an AI that can process immense amounts of data. Google understood this and doubled the context window to 2 million.

Reimagining Google Search

Search was the bread and butter of Google at its inception, and Google has come a long way in it. Search is now undergoing a huge transformation by leveraging Gemini. In earlier days, users were asked to do Google Searches in certain ways so that the search engine could provide better results, but now users can ask queries in their natural language, no matter how big the query is. We can ask questions, chain multiple queries, and much more. Today, the bigger the prompt, the happier the AI is to provide results. Also, the Search will proactively provide additional information based on the context of the query. Some of the advanced features in the new Search are:

• AI Overviews: In our busy lives, AI Overviews will play a significant role in the coming days. It synthesizes the best information from the web and then provides concise answers based on the context, all in the blink of an eye.
• Multi-Step Reasoning: With the help of Gemini, Search will now break down complex user queries into manageable parts and then analyze these against data across the web to provide results.
• AI-Organized Results Pages: Google Search will present results after Gemini organizes the web pages dynamically based on the context of the user query. The pages are organized based on diverse perspectives so that even hidden gems of the World Wide Web will be presented, encouraging users to explore new avenues.

Workspace: More AI, More Productivity

Gemini has revolutionized Google Workspace as well. Gemini has made Workspace smarter and more efficient. Gemini will summarize lengthy email threads and answer users’ queries about the content in their inboxes. Also, the lines between individual Workspace apps are blurring. Gemini will seamlessly integrate information and workflows across Gmail, Drive, Docs, Calendar, and more.

Android with Gemini

Smartphones have made our lives easier and faster, but now we are looking for even more speed and efficiency. Google is bringing the power of Gemini even closer to the hands of its users through Android.

• Circle to Search: This feature was first introduced on Samsung devices, but it is now part of the Android OS. Instead of a simple search, Gemini will now get the context in which the search is done and share the information, whether it’s a translation, a recipe, or step-by-step instructions for solving a math problem.
• Gemini: Android users will be saying goodbye to Google Assistant because the new assistant will be Gemini. It can anticipate user needs, suggest relevant actions, and provide instant assistance within any app.
• Gemini Nano: It is the on-device AI model from the Gemini family. It can work offline on users’ devices so that they can access the power of AI even when they are not connected to the internet, and at the same time, it respects user privacy by synthesizing data on their own devices. Also, Gemini Nano will proactively identify and warn users about potential scams, analyzing real-time audio from phone calls to detect suspicious activity.

Developer’s Delights

There were many moments in Google I/O this year that brought smiles to developers’ faces. A lot of updates in AI and Android have made the lives of developers easier.

• New Gemini Models: Google has introduced two new powerful Gemini 1.5 models:
  • Gemini Pro: It is optimized for handling complex tasks requiring high-quality responses.
  • Gemini Flash: As the name says, Flash is fast, just like the superhero. It produces responses faster but with less accuracy than the Pro model.
    Both models offer native multimodality, a 1 million token context window, and competitive pricing, making them accessible to developers of all sizes.
• Jetpack Compose: Google recommends Jetpack Compose to build applications on Android. It has started its journey on Android TV and Android for cars as well. This year, lots of updates were introduced in Compose. Three new scaffolds, namely, Adaptive Navigation Suite Scaffold, List Detail Pane Scaffold, and Supporting Pane Scaffold, help developers build apps that can adapt to various form factors, including flippables and foldables.
• Kotlin 2.0: The Compose compiler is now part of the Kotlin repository, so there is no need for version juggling any more. The K2 compiler is significantly faster than its predecessor, thereby making compilation faster.

The Dawn of a New Era

At Google I/O 2024, the company emphasized its commitment to democratizing AI. AI development is no longer a niche field for experts, and it is becoming more accessible to everyone. Google offers a wide range of AI models and tools, empowering developers of all levels to experiment with and integrate AI features into their creations. By bringing more AI to Android devices, even in offline mode, they have made AI accessible to millions of Android users.

Google I/O 2024 provided a powerful snapshot of AI’s rapid evolution and its potential to reshape our world. However, the journey of AI is only in its initial phase. It’s an exciting time to be alive, to witness the unfolding of AI’s remarkable potential, and to participate in shaping its future for the benefit of all humankind.