Google has just announced and released some of the biggest updates in their AI lineup, and at the very top of the list is a brand new way to use Google.com to search, called AI mode. This is a significant departure from the traditional Google search experience that users have been familiar with for the past 25 years.

In the AI mode, the search experience is more akin to using Gemini or ChatGPT, where users can engage in conversational searches and receive answers in a more comprehensive and contextual manner. The AI mode is being rolled out in the US, and it represents a significant risk for Google, as it fundamentally changes the way users interact with the search engine.

Unlike the traditional Google search, where users are presented with a list of blue links, the AI mode provides a more curated and AI-powered response, including relevant information, maps, and suggestions for further exploration. This new approach aims to provide users with a more intuitive and efficient search experience, but it also represents a significant shift in Google's core business model.

Overall, the introduction of AI mode in Google.com is a bold and ambitious move by the tech giant, and it will be interesting to see how users respond to this new way of searching and how it impacts Google's market dominance in the long run.

Google has just released two new language models that are now ranked as the top models on the LMA arena.ai chart - Gemini 2.5 Pro and Gemini 2.5 Flash. These models are currently in preview, but you can access them for free through the Google AI Studio.

Gemini 2.5 Pro and Gemini 2.5 Flash are the best-performing large language models available right now, surpassing even the popular ChatGPT models. You can try out these new models by going to the Google AI Studio and accessing the "chat" section, where you'll find both options available.

In addition to the new Gemini models, Google has also improved their Gemini Live feature, which allows you to interact with the AI in a more interactive, conversational manner. You can give Gemini Live access to your microphone, webcam, and screen, and it will provide feedback and assistance as you work.

To further explore the new Gemini models, you can also visit the Gemini website at gemini.google.com, where you'll find the Gemini Pro model currently in preview. This gives you another way to access and test the capabilities of these cutting-edge language models.

Gemini Diffusion is a new and innovative text-generation model introduced by Google. Unlike traditional large language models that predict the next word based on the input text, Gemini Diffusion uses a diffusion process to generate text.

The diffusion process starts with a noisy text input and gradually refines it, step by step, to produce a coherent and meaningful output. This approach allows Gemini Diffusion to generate text that is more diverse and creative, going beyond the typical patterns found in language models.

The demonstration showcased the Gemini Diffusion model's ability to solve a math problem, starting with a noisy input and progressively refining it until the correct answer of "39" was reached. This unique text-generation technique sets Gemini Diffusion apart from other language models and opens up new possibilities for AI-powered text generation.

The release of the Notebook LM mobile app is a significant development, allowing users to access the audio overviews created between hosts on the go. This feature enables you to listen to the summaries and insights while commuting or during other activities, providing a convenient way to stay informed without being tied to a desktop or laptop.

The mobile app's integration with Notebook LM means you can seamlessly access the audio content generated from your own sources, ensuring a personalized and tailored listening experience. Whether you're in the car, on a walk, or simply looking to multitask, the Notebook LM mobile app empowers you to consume the valuable information you need, whenever and wherever it's most convenient for you.

Veo 3, Google's latest state-of-the-art video generation model, is a massive leap in AI video creation. This model not only generates high-quality video but also introduces the ability for the generated content to talk.

The demo showcased the impressive capabilities of Veo 3, where the generated characters were able to engage in natural conversations, complete with appropriate facial expressions and lip movements. This level of realism and interactivity is a significant advancement in the field of AI video generation.

One of the most mind-blowing aspects of Veo 3 is its ability to generate talking video content. The model seamlessly synchronizes the audio and visual elements, creating a truly immersive experience. This feature opens up new possibilities for applications such as virtual assistants, interactive storytelling, and even personalized video messages.

The availability of Veo 3 is a game-changer, and the new pricing tiers introduced by Google, including the Google AI Ultra subscription, provide users with access to this cutting-edge technology. The ability to leverage Veo 3 for video generation, combined with the other tools and features offered in the Google AI Ultra plan, makes it an attractive option for content creators, filmmakers, and businesses looking to push the boundaries of what's possible with AI-powered video.

Gemini now offers three different subscription tiers:

1. Free Plan: This plan is available to anyone with a Google account and provides basic access to Gemini features.

2. Google AI Pro ($20/month): This is a revamped version of the previous Gemini subscription. It includes:

   • Higher limits for the Gemini 2.5 Pro Deep Think model
   • Access to the new Flow filmmaking tool
   • The Whisk image-to-video generation tool (powered by Veo 2)
   • YouTube Premium (ad-free YouTube)
   • 2TB of storage

3. Google AI Ultra ($250/month): This is the top-tier subscription and includes:

   • Unlimited access to the Gemini 2.5 Pro Deep Think model
   • Exclusive access to the new Veo 3 video generation model
   • Flow filmmaking tool (using Veo 3)
   • Whisk image-to-video generation (with Veo 2)
   • YouTube Premium
   • 30TB of storage

The Google AI Ultra tier is required to access the latest Veo 3 video generation capabilities. This subscription also includes a range of other advanced features and tools for filmmaking, image generation, and more.

Whisk is a powerful new tool introduced by Google that allows users to seamlessly generate videos from text prompts. This feature is part of the Google AI Ultra subscription tier, providing users with advanced capabilities for creating dynamic visual content.

With Whisk, users can input a text prompt describing a scene or subject, and the tool will generate a corresponding video. The process is intuitive and straightforward, allowing users to experiment with different prompts and styles to achieve their desired results.

One of the standout features of Whisk is its ability to integrate with Google's advanced text-to-image model, Gemini Diffusion. This integration enables users to generate high-quality images from their text prompts, which can then be used as the foundation for the video generation process.

Furthermore, Whisk offers users the ability to refine and animate their generated videos. Users can apply additional text prompts to refine the video, adjusting elements such as camera movement, lighting, and overall composition. The animation feature allows users to bring their creations to life, adding a dynamic and engaging element to the final output.

Overall, Whisk represents a significant advancement in the field of AI-powered video generation, providing users with a seamless and powerful tool to bring their creative visions to life.

The presenter showcases the interactive capabilities of Gemini Live, a feature within the Google AI Studio. Gemini Live allows the user to engage in a live conversation with the AI assistant, which can provide hands-on support and guidance.

In the demonstration, the user asks the AI to help them fix their bike. The AI assistant demonstrates its ability to:

1. Locate and provide information from the bike's user manual.
2. Search for and recommend a video tutorial on how to fix a stripped screw.
3. Retrieve and highlight the specific email with the required hex nut size.
4. Call the local bike shop to check the availability of a replacement tension screw.
5. Refer back to the user manual to identify the section on replacing brake pads.
6. Suggest options for adding a dog basket to the bike.

Throughout the interaction, the AI assistant seamlessly navigates various online resources and provides step-by-step guidance, showcasing its capability to assist the user in completing the bike repair task. This interactive demonstration highlights the potential of Gemini Live to provide personalized, hands-on support for a wide range of tasks.

Google has introduced a new feature called "Agent Mode" within the Gemini app. This mode allows AI agents to interact with the web and perform tasks on your behalf, under your control.

The key idea behind agents is to combine the intelligence of advanced AI models with access to various tools and services. These agents can take actions on your behalf, freeing you up to focus on other priorities.

For example, if you need to find an apartment for you and two roommates in Austin, with a budget of $1,200 per month and a requirement for a washer/dryer or nearby laundromat, the Agent Mode in the Gemini app can handle this task. It will scour listings from sites like Zillow, apply the specific filters, and even schedule a tour for you, all while you focus on planning the housewarming party.

The agent mode leverages Google's research prototype, Project Mariner, to interact with the web and perform these tasks efficiently. This integration of AI intelligence and web access allows the agent to streamline the apartment search process, saving you time and effort.

Overall, the Agent Mode in the Gemini app represents a significant step forward in the integration of AI assistants into our daily lives. By delegating certain tasks to these intelligent agents, we can free up our time and attention for more meaningful activities.

Google has introduced a groundbreaking feature in Google Meet - real-time speech translation. This revolutionary capability allows users to communicate seamlessly, even when they don't share a common language.

With this feature, participants can speak in their native language, and Google Meet will instantly translate the conversation, enabling all attendees to understand and respond effectively. This technology breaks down language barriers, facilitating more inclusive and productive meetings, regardless of the linguistic diversity of the participants.

The real-time translation feature is powered by advanced AI models, ensuring accurate and natural-sounding translations. Users can simply select the desired target language, and Google Meet will handle the rest, providing a seamless and effortless communication experience.

This innovation from Google is a significant step forward in making global collaboration and communication more accessible and efficient. By bridging language gaps, it empowers teams and organizations to work together more effectively, fostering greater understanding and cooperation across cultural and linguistic boundaries.

Flow is a new filmmaking tool introduced by Google that combines the power of Veo, their state-of-the-art video generation model, with a suite of filmmaking tools. Unlike Veo, which generates single video clips from text prompts, Flow offers a more holistic approach to video creation.

With Flow, users can bring their own assets or generate them using Veo, and then easily manage and reference them as they start creating their projects. This makes Flow more akin to a comprehensive filmmaking tool rather than just a video generation tool.

One of the key features of Flow is its ability to provide cinematic control over the generated footage. Users can describe camera movements, such as "jib up," and Flow will incorporate those into the final video. This level of control allows for the creation of highly polished and professional-looking videos.

Flow will be available as part of the Google AI Pro subscription tier, which utilizes Veo 2 for video generation. However, to access the more advanced Veo 3 model, users will need to upgrade to the Google AI Ultra subscription tier, which is priced at $250 per month.

Overall, Flow represents a significant step forward in the integration of AI-powered video generation with traditional filmmaking tools, providing users with a more comprehensive and versatile platform for creating high-quality video content.

Google has announced a major update to its search capabilities, introducing a brand-new "AI mode" within Google.com. This revolutionary feature allows users to interact with Google's search engine in a more conversational and AI-driven manner, similar to using AI assistants like Chatbot or Gemini.

The AI mode on Google.com provides users with a direct answer to their queries, rather than just a list of search results. This AI-powered search experience includes features like:

• Personalized context: The AI mode will leverage users' Google account data, such as Gmail and Google Drive, to provide personalized suggestions and insights.
• Integrated shopping: AI-powered shopping capabilities will be integrated directly into the search experience, making it easier to find and purchase products.
• Genetic capabilities: The AI mode will incorporate advanced genetic algorithms to provide more accurate and relevant search results.

In addition to the AI mode, Google has also made significant improvements to its Gemini AI models, including the release of Gemini 2.5 Pro and Gemini 2.5 Flash. These models have been ranked as the top-performing large language models, surpassing even the popular ChatGPT.

Users can access the new Gemini models through the Google AI Studio, where they can engage in interactive conversations and utilize features like Gemini Live, which allows for screen sharing and real-time feedback.

Furthermore, Google has introduced a new "agent mode" within the Gemini app, which enables users to delegate tasks to the AI assistant, such as finding and booking apartment listings based on specific criteria.

The future of Google search is undoubtedly AI-powered, with these latest updates showcasing the company's commitment to revolutionizing the way users interact with its search engine and AI capabilities.

The latest updates from Google in their AI lineup are truly groundbreaking. The introduction of AI mode within Google.com is a significant shift in how users interact with the search engine, moving towards a more conversational and AI-powered experience.

The improvements to Gemini 2.5 Pro and Gemini 2.5 Flash, positioning them as the top large language models, offer users powerful tools for various tasks. The addition of Gemini Diffusion, which applies diffusion models to text, is an innovative approach to text generation.

The release of the Notebook LM mobile app and the impressive Veo 3 video generation model, with the ability to generate talking videos, showcase the advancements in AI-powered multimedia creation. The new subscription tiers, including the Google AI Pro and Google AI Ultra plans, provide users with access to these cutting-edge features and capabilities.

Overall, these updates from Google demonstrate the company's commitment to pushing the boundaries of AI technology and redefining how users interact with search, language models, and multimedia generation. The future of AI-powered experiences is here, and Google is leading the charge.