ThursdAI Special: Google's Big week - Anti Gravity, Gemini 3 and Nano 🍌 with Kevin, Ammaar and Kat

Quick Overview

The episode features a discussion with Kevin Ho, Ammaar Reshi, and Kat Kampf about the recent releases of Google's Gemini 3, Anti-Gravity IDE, and Nano Banana Pro, focusing on the iterative development and user experience improvements, particularly how these tools enable more creative and less code-heavy application building.

Key Points: Google released Gemini 3 and Anti-Gravity IDE, followed shortly by Nano Banana Pro. The hosts recorded the conversations at the AI Engineer Code Summit in New York, shortly after the releases. Kevin Ho, AI Evangelist at Weights & Biases and host of ThursdAI.news podcast, interviewed the guests. The team focuses on making AI-powered development easier, moving away from low-level coding towards higher-level abstraction. Anti-Gravity IDE allows developers to generate code/UI elements by describing them, with the model understanding context like the current directory structure. The ability to send screenshots or videos of desired output directly to the model (like a 'live coding demo') is a key feature. The team is actively iterating based on user feedback, particularly noting the success of the visual/multimodal capabilities of Gemini 3.

Context: This segment is a special edition of the ThursdAI.news podcast, recorded at the AI Engineer Code Summit in New York, featuring Alex Volkov (Host) interviewing Kevin Ho (Lead Product and Engineering at Anti-Gravity), Ammaar Reshi (Lead, Product and Design on AI Studio), and Kat Kampf (PM for Vibe Coding & AI Studio). The discussion centers around the exciting developments from Google that week, specifically the release of Gemini 3, the Anti-Gravity IDE, and Nano Banana Pro, and how these tools are changing the development workflow.

Detailed Analysis

Alex Volkov kicks off the podcast by welcoming Kevin Ho, Ammaar Reshi, and Kat Kampf, noting the incredibly busy week for Google following the releases of Gemini 3, Anti-Gravity IDE, and Nano Banana Pro. Volkov mentions he recorded conversations with the guests right before the AI Engineer Code Summit in New York. The discussion immediately dives into the capabilities of Anti-Gravity IDE, described as a tool that simplifies coding by understanding context like directory structure and allowing developers to create complex, multi-modal applications with simple prompts. Ammaar Reshi highlights the 'lightbulb moment' users experience when they can simply ask the model to perform tasks (like generating an infographic layout or sending a screenshot/video of desired output) and receive functional code/UI elements back, which is a significant step beyond previous models. Kat Kampf emphasizes that AI Studio is designed to be the home for builders, making it easy to start with the Gemini API, even for non-coders, and to iterate quickly by editing generated code or assets directly. She notes that the ability to tailor the experience based on user intent (e.g., coding vs. non-coding tasks) is a key focus. The conversation shifts to the future, with Alex Volkov expressing excitement about the potential for AI agents to manage complex tasks like running unit tests or even managing production environments (like Google Search or infrastructure), noting that while we are not fully there yet, the foundational capabilities are being built now. Ammaar agrees, pointing out that the seamless integration of these features within AI Studio, where the model understands the context of the task, is crucial for developer productivity and creativity. The panel concludes by emphasizing the rapid pace of innovation in the space, noting that the advancements seen in the last year are astounding.