David Sacks expressed his view that OpenAI and Anthropic do not require regulations to manage the development of frontier AI models.
He acknowledged that both companies lead the market in frontier intelligence based on metrics like market share, revenue growth, and model capability. Sacks implied confidence in their ability to self-regulate without external intervention.
Cognition has launched SWE-2, a new coding model positioned to compete with Fable 5.1 and GPT-Astra.
SWE-2 achieves a 50. 0% score on FrontierCode 1.
OpenAI introduced the Data agent in ChatGPT Work, enabling users to connect company data and build interactive dashboards with AI.
This feature allows users to uncover insights using natural language, streamlining the process of data interaction. The Data agent integrates company data directly into ChatGPT Work, facilitating the creation of dashboards without requiring separate tools or complex setups.
Shopify is transitioning back to native app development using Swift and Kotlin, moving away from React Native.
This shift comes after the realization that coding models have improved, reducing the costs associated with building features in native languages. Shopify initially adopted React Native in 2020 to streamline feature development and enable contributions from developers without a mobile background. However, recent advancements in coding technology prompted a reevaluation of their mobile stack, leading to the decision to return to native development. The change reflects a strategic response to evolving technological capabilities.
OpenAI introduced ChatGPT for Financial Services, which combines built-in financial data with GPT-6 Astra for enhanced functionality.
This version supports research, modeling, and the creation of client-ready materials, integrating financial data directly into the AI's capabilities. GPT-6 Astra is used to power these processes, enabling more precise and efficient outputs tailored to financial sector needs.
DeepSeek has released DeepSeek-V4.1-Flash, the latest model in their architecture family, featuring native visual understanding.
This version is the smallest model yet, focusing on increased capability, faster inference, and higher throughput. It also supports scaling to larger models, enhancing its performance across various tasks.
OpenAI released GPT‑Live‑1, enabling natural, full-duplex voice conversations in the API.
GPT‑Live‑1 enhances instruction following and introduces custom voice options. It also supports telephony, allowing seamless integration into voice communication systems.
Tailwind Labs is joining Shopify to provide a stable long-term home for the Tailwind CSS framework.
Tailwind CSS, installed over 110 million times weekly, is used by major companies like ChatGPT, Cloudflare, and Reddit. Shopify, an early adopter of Tailwind CSS, sees it as a crucial part of their stack, ensuring its active maintenance and ongoing development. This partnership aims to enhance Tailwind by integrating it into Shopify's extensive platform, which supports custom storefronts and advanced user interfaces.
DeepSeek is set to release the V4.1 Flash model, which is more affordable and capable than the V4 Pro, around September 10, 2026.
V4. 1 Flash has been tested both internally and externally, showing superior performance, cost-efficiency, speed, and task completion time compared to V4 Pro.
OpenAI released ChatGPT Images 2.5, a feature for transforming ideas, sketches, and reference photos into more personalized and polished images.
This update enhances the ability to create visuals that closely align with user input and concepts. ChatGPT Images 2.5 focuses on refining the output to better reflect the original ideas provided by users.