Issue #14

The Context Window — July 21, 2026

What's new in AI, explained simply, with something to try.

Tuesday, July 21, 20264 min read

This Week's Stories

MODELSStory 1

Google Launches Gemini 3.6 Flash Model

Google has released a new AI model called Gemini 3.6 Flash. This model is now available on their Model Garden platform. It aims to enhance AI capabilities for various applications.

Why This Matters

Gemini 3.6 Flash is a new AI model (a type of software that learns from data to perform tasks like language understanding and generation). This model is designed to improve performance in tasks such as language understanding and generation. For everyday users, this means AI tools could become more accurate and responsive, making interactions smoother and more efficient. The Model Garden is a platform where developers can access and experiment with Google's AI models, so this release provides new opportunities for innovation in AI applications.

What You Can Try

Visit Google's Model Garden at the provided URL and explore the Gemini 3.6 Flash model. If you're a developer, try integrating it into a small project to see how it enhances language tasks.

Source: console.cloud.google.com

TOOLSStory 2

Claude Code Team Shares Insights in Fireside Chat

Cat Wu and Thariq Shihipar from Anthropic's Claude Code team discussed their tools at the AI Engineer World's Fair. They covered topics like Claude Tag, coding agent security, and tool design. The full session is now available on YouTube.

Why This Matters

This chat offers a peek into how Anthropic develops and uses AI tools like Claude Code and Claude Tag. Claude Tag, a Slack integration, is now handling a significant portion of product engineering tasks, showing how AI can streamline workflows. The discussion also highlights evolving best practices, such as simplifying system prompts for better model performance. Understanding these insights can help users and developers optimize their use of AI tools, making their work more efficient and effective.

What You Can Try

Watch the full fireside chat on YouTube to learn directly from the Claude Code team. Pay special attention to their insights on system prompts and how they use AI tools internally. Apply these insights to streamline your own AI workflows.

Source: simonwillison.net

TOOLSStory 3

Adobe Adds Generative AI to Indigo Camera App

Adobe has updated its Indigo camera app with new generative AI tools. Originally launched to mimic a natural SLR look for iPhone photos, the app now includes AI features to enhance photography. These updates do not use Adobe's Firefly AI models.

Why This Matters

Generative AI in photography can automatically enhance or transform images, making it easier to achieve professional-looking results without extensive editing skills. In the Indigo app, this means users can apply creative effects or improve photo quality with minimal effort. Generative AI (AI that creates new content from existing data) can help users experiment with different styles and effects, much like having a digital assistant that suggests artistic improvements. This makes high-quality photography more accessible to everyone, not just professionals with expensive equipment.

What You Can Try

Download the Indigo camera app on your iPhone. Try using the new AI tools by selecting a photo and applying the 'Artistic Enhancement' feature to see how it transforms your image. Notice the difference in quality and style.

Source: theverge.com

Weekly Insights

Concept of the Week

Function Calling

Function calling is like asking a specific tool in your toolbox to do a job for you. In programming, it means using a defined set of instructions (or a "function") to perform a task, which helps keep code organized and efficient. This matters because it allows developers to reuse code, making software easier to maintain and update. For example, when you use a calculator app to add numbers, you're calling a function that performs the addition for you, simplifying your task.

Tool of the Week

Otter.ai

Otter.ai is a tool that transcribes spoken words into written text, making it easier to capture conversations, meetings, or lectures. It's useful for anyone who wants to keep accurate notes without having to write everything down manually. You could try using it today by recording a short meeting or a class lecture, and then review the transcript to ensure you didn’t miss any important points.

Share this issue

The Context Window — July 21, 2026

This Week's Stories

Google Launches Gemini 3.6 Flash Model

Why This Matters

What You Can Try

Claude Code Team Shares Insights in Fireside Chat

Why This Matters

What You Can Try

Adobe Adds Generative AI to Indigo Camera App

Why This Matters

What You Can Try

Weekly Insights

Function Calling

Otter.ai

Prompt of the Week

Build in 2 Minutes

Trend to Watch