OpenAI Transforms ChatGPT into a Shopping Assistant

OpenAI has rolled out new shopping features within ChatGPT, enabling users to receive personalized product recommendations, compare prices, view visuals, and access direct purchase links—all within the chat interface. Check out the announcement on OpenAI’s X post.

Alibaba Launches Qwen3 Amid Intensifying AI Competition

Alibaba has unveiled Qwen 3, an advanced upgrade to its flagship AI model, featuring hybrid reasoning capabilities aimed at improving adaptability and efficiency for app and software development. This release marks a strategic move in China's escalating AI race, following recent advancements by competitors like DeepSeek and Baidu. Explore the Qwen3 repository on GitHub.

Meta Hosts LlamaCon to Promote Open-Source AI

Meta has launched LlamaCon to spotlight advancements in open-source AI to assist developers in creating applications and products. The event precedes the Meta Connect conference scheduled for September and underscores Meta's commitment to investing in AI and supporting the developer community. You can RSVP your presence on their official website.

Adobe Unveils AI-Powered Creative Tools at Adobe Max London

At Adobe Max London 2025, Adobe showcased a suite of AI-driven enhancements across its Creative Cloud applications. Key features include the Firefly Image Model 4 for hyper-realistic image generation, Firefly Boards for collaborative moodboarding, and natural language editing in Photoshop. If you are a content creator or a designer, you should discover how these tools can boost your creative workflows.

Microsoft’s AI Recall Feature Now Available in Copilot+ PCs

Microsoft has officially released its long-awaited Recall feature for AI-enabled Copilot+ PCs. Recall continuously captures screen snapshots to help users retrieve previously viewed content. After addressing early privacy concerns raised by security experts, Microsoft implemented significant security adjustments before the feature's release.

Microsoft’s BitNet, an Ultra-efficient CPU-based AI Model Gains Traction

Microsoft researchers have introduced a groundbreaking AI model that operates efficiently on standard CPUs, eliminating the need for GPUs. This model utilizes ternary weights (-1, 0, 1), significantly reducing energy consumption by up to 96% compared to traditional models. You can check out the Hugging Face model card and the technical paper here.

Gemini 2.5 Series Introduces a “Thinking Budget”

Google’s Gemini 2.5 Flash now includes a “thinking budget" that allows developers to control the model's reasoning depth, enabling a balance of performance and computational cost. This flexibility caters to diverse application needs, from quick responses to complex problem-solving. Learn more about the thinking budget here.

The Washington Post Warms Up to GenAI, Partners with OpenAI

The Washington Post has entered a strategic partnership with OpenAI to expand its journalism access through integration within ChatGPT. This collaboration allows ChatGPT to display summaries, excerpts, and links to The Post’s original reporting in response to user queries, marking a significant step toward greater public access to reliable and timely information. This could very well be the beginning of a new era where top content creators embrace GenAI for a more user centric future. Read the full announcement on OpenAI’s blog.