Claude outsmarts ChatGPT 🏆, cinematic AI filmmaking 🎬, Google Drive auto-summarizes 🎯

🛠️ Product Updates

Text

Anthropic’s Claude app adds voice chat with Google perks for subscribers

Anthropic is bringing voice capabilities to Claude mobile apps starting May 28, offering natural spoken conversations powered by Claude Sonnet 4. The feature includes five distinct voices, real-time on-screen summaries, and seamless switching between voice and text without losing context. Free users can send 20-30 voice messages per session, while paid subscribers gain voice access to Google Calendar and Gmail, with Google Docs integration reserved for Enterprise accounts. This launch, which includes conversation saving and safety measures against impersonation, keeps Anthropic competitive with OpenAI and Google in the increasingly voice-first AI assistant market.

Perplexity’s new deep research mode takes on ChatGPT and Gemini

Perplexity is upping its research game with a new "Deep Research" mode that will let users access advanced models like Claude 4 Opus while performing iterative searches across hundreds of sources. The upgrade promises comprehensive reports with multi-step reasoning capabilities, directly challenging similar features from ChatGPT and Gemini. Alongside this, Perplexity is adding personalization options that let users leverage past searches for improved responses—with privacy controls including incognito mode restrictions and the ability to delete saved memories. The enhancements position Perplexity to reclaim competitive ground in AI-powered research assistance after recent advances by rivals.

OpenAI tests sign-in with ChatGPT to challenge Apple and Google

OpenAI is testing a "Sign in with ChatGPT" feature, allowing users to access third-party apps using their ChatGPT credentials. The company launched a developer interest form Tuesday, targeting businesses of all sizes to gauge potential integration partners. With approximately 600 million monthly active users, this strategic move positions OpenAI to compete directly with established identity services from Apple, Google, and Microsoft. Developers can already preview the sign-in experience in Codex CLI, OpenAI's AI coding tool, with incentives including API credits for linking accounts. First teased by CEO Sam Altman in 2023, this feature represents OpenAI's expanding ambitions beyond AI chat into broader digital identity services.

Google Drive’s Gemini AI now smartly summarizes videos and tasks

Google Drive's latest Gemini update brings powerful video intelligence to Workspace users starting May 28. The AI can now quickly summarize videos and extract action items, allowing users to grasp key content without watching entire clips. Users can even ask specific questions about video highlights. Rolling out first to rapid-release domains and then to scheduled-release domains by June 16, this enhancement builds on Gemini's existing ability to summarize documents and folders—further cementing Google's commitment to AI-powered productivity across its ecosystem.

Microsoft’s AI-powered gaming copilot launches beta with smart tips and recs

Microsoft's Copilot for Gaming has entered early beta on iOS and Android via the Xbox mobile app, offering gamers an AI-powered companion that draws on their play history and achievements. The initial release helps users view recent accomplishments, get personalized game recommendations, and access tailored tips—with remote game installation coming in future updates. The context-aware AI assistant features customizable voice options and understands what you're playing, positioning it as an intelligent second-screen experience. Beta testing is currently limited to select regions including the US, Australia, and Singapore, signaling Microsoft's push toward AI-enhanced gaming experiences ahead of competitors.

Coding

New Relic and GitHub Copilot team up to automate code fixes end-to-end

New Relic has integrated with GitHub Copilot's Coding Agent, automating the entire software issue lifecycle from detection to resolution. The partnership enables New Relic to spot performance problems in AI-generated code, create detailed GitHub issues, and validate fixes after Copilot drafts and submits corrections. This end-to-end automation significantly reduces incident response time, freeing developers from tedious maintenance work to focus on building new features. Microsoft's Julia Liuson highlighted how such integrations make development tools more scalable and intelligent across the software lifecycle.

Image

Myntra and Google Cloud use AI to revolutionize home decor shopping

Myntra has teamed up with Google Cloud to unveil 'Dream Room Inspirations,' a generative AI feature transforming home decor shopping. Powered by Google's Imagen 3 on Vertex AI, the tool lets shoppers visualize interior designs across styles like Bohemian Chic and Modern Minimalist before purchasing. Users simply select their preferred room type or design theme to generate customized interior visualizations. With over 20 million Indians shopping for home decor online in 2024, this innovation strategically bridges the gap between inspiration and purchase, positioning Myntra as a tech-forward leader in the competitive e-commerce space.

Google Flow brings cinematic AI filmmaking to your fingertips with easy scene building

Google unveils Flow, an AI filmmaking tool now available to US subscribers through Google AI Pro and Ultra plans. This creative powerhouse integrates Veo, Imagen, and Gemini models to help storytellers produce cinematic scenes with unprecedented ease. Flow's standout features include intuitive camera controls, a robust scene builder, and asset management tools that maintain visual consistency across projects. Users can describe complex scenarios in plain language while creating custom, reusable characters and elements. The companion Flow TV library showcases user-generated content with visible prompts, helping creators learn from and build upon others' techniques.

💡 Insights

Tabnine goes all-in on air-gapped AI as rivals bail out

Tabnine has doubled down on air-gapped AI development, becoming the sole provider as competitors like Codeium abandon the space. Their platform enables public sector, defense, and aerospace organizations to run AI coding assistants entirely within secure networks—no internet connectivity required. The offering addresses critical security concerns by eliminating external data transmission and providing transparent, traceable AI outputs with customizable enterprise guardrails. Tabnine is aggressively courting abandoned customers with free migration incentives, reflecting growing demand for genuinely secure AI tools in classified environments where data sovereignty and compliance aren't negotiable.

Anthropic’s Claude 4 Sonnet takes on ChatGPT-4o and wins big

Anthropic's Claude 4 Sonnet has emerged as a formidable challenger to OpenAI's ChatGPT-4o in head-to-head comparisons, winning in five of seven test categories. The new Claude model demonstrated superior capabilities in productivity planning, storytelling depth, idea generation, emotional support, and nuanced critical thinking. While ChatGPT-4o still excelled at tone matching and tied in practical reasoning, Claude 4 Sonnet's deeper emotional intelligence and stronger long-form reasoning represent significant advances for Anthropic as it positions its latest models to compete directly with OpenAI across both consumer and enterprise AI applications.